Epistemology and Practice: Durkheim's The Elementary Forms of Religious Life

http://www.cambridge.org/9780521651455

This page intentionally left blank

Epistemology and Practice

In this original and controversial book Professor Rawls argues thatDurkheim’s The Elementary Forms of Religious Life is the crowningachievement of his sociological endeavour and that since its publica-tion in English in 1915 it has been consistently misunderstood. Ratherthan a work on primitive religion or the sociology of knowledge, Rawlsasserts that it is an attempt by Durkheim to establish a unique epistemo-logical basis for the study of sociology and moral relations. By privilegingsocial practice over beliefs and ideas, it avoids the dilemmas inherent inphilosophical approaches to knowledge and morality that are based onindividualism and the tendency to treat concepts as the limit of knowl-edge, both tendancies that dominate western thought. Based on detailedtextual analysis of the primary text, this book will be an importantand original contribution to contemporary debates on social theory andphilosophy.

Anne Warfield Rawls is Associate Professor of Sociology at BentleyCollege, Waltham, Massachusetts. She has a background in both sociol-ogy and philosophy and has published extensively on social theory andsocial justice.

Epistemology and PracticeDurkheim’s The Elementary Formsof Religious Life

Anne Warfield Rawls

Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo

Cambridge University PressThe Edinburgh Building, Cambridge , UK

First published in print format

- ----

- ----

© Anne Warfield Rawls 2004

2005

Information on this title: www.cambridg e.org /9780521651455

This book is in copyright. Subject to statutory exception and to the provision ofrelevant collective licensing agreements, no reproduction of any part may take placewithout the written permission of Cambridge University Press.

- ---

- ---

Cambridge University Press has no responsibility for the persistence or accuracy ofs for external or third-party internet websites referred to in this book, and does notguarantee that any content on such websites is, or will remain, accurate or appropriate.

Published in the United States of America by Cambridge University Press, New York

www.cambridge.org

hardback

eBook (EBL)

eBook (EBL)

hardback

http://www.cambridge.org/9780521651455

http://www.cambridge.org

For Ty and Martin

Contents

Acknowledgements page ix

Introduction 1Durkheim’s Epistemology: the Neglected Argument 7Epistemological Crisis 8Religion and Reason 11Perception versus Emotion 15Mis-Communication Between Disciplines 17Order of Argument 22

1 Durkheim’s Outline of the Argument in the IntroductoryChapter 28

Section i: Consideration of Religion 33Section ii: The Introduction of Epistemology 46Concluding Paragraphs 68

2 Durkheim’s Dualism: an Anti-Kantian Anti-RationalistPosition 72

Durkheim’s First Distinction: “Double Man” 78Second Distinction: “Two Layers of Knowledge” 90The Social as Sacred versus the Individual as Profane 100Sociology of Knowledge: Idealism versus Concrete Practices 101Conclusion 105

3 Sacred and Profane: the First Classification 108Durkheim’s Conception of Religion 112Animism versus Naturism 124Totemism 135Conclusion 137

4 Totemism and the Problem of Individualism 139The Totem as the Origin of the Sacred 141Totemic Classification as a Logical System 149The Status of Individual and Sexual Totemism 152Conclusion 161

vii

viii Epistemology and Practice

5 The Origin of Moral Force 162Mana as the Origin of the Idea of Moral Force 164Logic and Collective Representations 177Personhood and Myths versus Rites 188Conclusion 192

6 The Primacy of Rites in the Origin of Causality 194Book III, Chapter One: The Negative Cult 196Book III, Chapter Two: The Positive Cult 202

7 Imitative Rites and the Category of Causality 212Book III, Chapter Three, Section i: Description of Imitative Rites 212Book III, Chapter Three, Section ii: The Principle Behind

Imitative Ritual 213

8 The Category of Causality 230Book III, Chapter Three, Section iii 234Durkheim’s Socio-Empirical Argument for Causality 258

9 Logic, Language and Science 262Section i: Practice/Real versus Belief/Ideal 266Section ii: Religion and Science 283Section iii: The Social Origins of Logic and Language 288

10 Durkheim’s Conclusion Section iv: Logical Argumentfor the Categories 301

The Six Categories of the Understanding 301

Conclusion 316The Development of Two Conflicting Durkheims 321The Fallacy of Misplaced Abstraction 324The Sociological Dilemma 326Scientific Things versus Social Things 328Recognizable by Design 330Intelligibility as a Constraint on Practice 334

Bibliography 339Index 345

Acknowledgments

The research comprising this book has been in the works for almost fifteenyears and during that time has benefited from a great deal of support,advice and encouragement. The debt I owe is large and the number ofpeople who were generous enough to lend their valuable time to thesupport of this project humbling.

That Durkheim was making an argument for an epistemology in theclassical sense, and that he intended this argument to ground his entiresociology, first came to me as I was teaching a graduate seminar at WayneState University in the early 1990s. Good ideas have often come tome in this way during public close readings of texts, and I think of thephenomenon as a dialogue in the best Socratic sense. To the students inthat and subsequent seminars, in particular Gary David, Lynetta Mosby,Bonnie Wright, Derek Coates and Jennifer Dierickx, I owe a great deal.

Through this dialogue I was able to hear myself think in ways that areimpossible for an individual alone. I hope those students understand thecontribution they made to my thinking. I also hope that they learned fromthe exercise that the study of any single great text is at the same time astudy of almost everything. The idea that one must sample a large numberof great texts to get a broader view I think is wrong. Mastery is essential –but sampling tends toward superficiality. Great texts like true lovers donot reveal themselves easily. They require a deep commitment of time,openness, energy and even kindness. The prevalence of the tendencyto skim, sample, classify and judge has had unfortunate consequencesfor the teaching of sociological theory. Any author of major significanceholds that status precisely because they struggled with the most importantdilemmas of their time. Classical texts don’t reflect “dead men’s ideas,”as it has become popular to say, rather they reflect the birth pains andcontinuing trauma of the world we live in.

Many colleagues supported my efforts. In the early stages of writing Isent the first draft of a manuscript to Randy Collins and Norbert Wiley,both of whom took the time to make very extensive and helpful com-ments. They were particularly helpful in orienting me toward aspects

ix

x Acknowledgments

of Durkheim scholarship with which I had been unfamiliar. With theiradvice and encouragement I was able to write the article that was pub-lished in the American Journal of Sociology in 1996 as “Durkheim’s Epis-temology: The Neglected Argument.” Others who were helpful at thispoint were Donald Levine, Charles Lemert and Harold Garfinkel. DavidBritt and Albert Meehan also read and commented on early versions.Garfinkel and Lemert in particular offered essential support over thecourse of the project.

During one of our phone conversations I confided to Harold that I waswriting about the epistemological argument in Durkheim’s ElementaryForms. Expressing great skepticism, as one would expect, he asked tosee a copy of the manuscript. Much to my delight he considered what Ihad written to be in essential respects more empirical than theoretical,and proceeded to encourage me in various ways. Over the next few yearsHarold incorporated his own thinking on Durkheim into manuscriptsthat he was then working on, eventually deciding to subtitle the bookwhich I edited with him “Working Out Durkheim’s Aphorism.”

Harold also sent a draft of the article to various people, a number ofwhom replied directly to me. I thank all of them and in particular PierreBourdieu, for their comments and their time. They must have wonderedat Garfinkel’s interest in the manuscript and I hope that in the writing ofthis book I have in some small measure answered their questions.

I was also the beneficiary of discussions with several students ofParsons, including Garfinkel and Joe Feagin, of their recollections of hav-ing studied The Elementary Forms and other works by Durkheim, with Par-sons. Joe took the time to talk with me at a point when he was particularlybusy. Their recollections of Parsons’ reading of Durkheim, experiencedfirst hand as graduate students in Parsons’ theory courses, and discus-sions of the difference between that reading and my own – in particularwith regard to Durkheim’s treatment of practice – were particularly valu-able. I was struck by the experience reported of reading the book againand finding the actual text somehow transformed by my own reading.

There is no way really to properly thank Charles Lemert for his supportof my work. His involvement with this manuscript has been a major factorin its publication. His comments on the manuscript were extensive andhelpful and his encouragement profoundly heartening.

To Roberto Serrai I owe the loan of his apartment in Florence (Firenze)in the fall of 1998 in which to complete the first full draft of thebook. The inspiration of an eighth floor balcony overlooking the birth-place of modern humanism cannot be overestimated. I need also thankRoberto for his fine translation of my work on Durkheim into Italian.While in Firenze Alessandro Pizzorno and Gianfranco Poggi, both at the

Acknowledgments xi

International Institute in Fiesoli at the time, were gracious with their timeand advice, inviting me out for wonderful dinners and lunches to discussDurkheim’s texts.

Wes Sharrock and several of his graduate students made themselvesavailable for discussions of Durkheim, Garfinkel and practice over beerin a Manchester pub in the summer of 2001. These discussions inspiredrevisions to both my introduction to Ethnomethodology’s Program: WorkingOut Durkheim’s Aphorism and to this book.

I am indebted to the editors and translators of the Mauss review, AlainCaille and Stephane Dufoix in particular, for translations of my work onDurkheim in the review and for their efforts as organizers of the GEODEconference on globalization and social theory in the summer of 2003bringing together a diverse spectrum of social theorists in Paris. Theirefforts with regard to the subsequent translation and publication of paperspresented at that conference are much appreciated. I owe a special thanksto Lorenza Mondada for accompanying me on that occasion.

Thanks are owed also to Michel de Fornel, editor of Enquete, and thevarious commentators on my article for that journal (to appear in thespring of 2005). I thank Albert Ogien also for our discussions of my workin Paris in the summer of 2003.

My thanks to Sandro Segre of the University of Genoa, Gino Muzzettoand his colleagues at the University of Pisa, Carmen Leccardi and NinoSalamone, of the University of Milan-Bicocca, and members of theItalian Sociological Association, who invited me to present a series oflectures and to speak at a conference on ethnomethodology and phe-nomenology at the University of Genoa in January 2004. They were gra-cious hosts and offered an invaluable opportunity to present and discussthe developing connection between Durkheim, Garfinkel, practice andglobalization in my work. I found many Italian colleagues with similarinterests and greatly enjoyed long hours of discussion and debate withinterested and committed scholars over fabulous food. Thank you.

I have also to thank members of the Charles Institute in Prague fortheir invitation to speak at a Thursday seminar in January 2004. A spe-cial thanks to Zdenek Konopasek for forging a unique line of inquiry inPrague – which he has done in the face of the growing popularity of sta-tistical sociology there – and for his intelligence and fortitude in doingso. The insights from our talks helped cement several issues in place.

To Svetlana Bankovskaya and her students and colleagues (particularlyVictor and Dimitri who were sometimes my able escorts) in Moscow whoinvited me to lecture on both Durkheim and Garfinkel at The HigherSchool of Economics in May 2004 as I was putting the finishing toucheson this manuscript. I also very much enjoyed my talks with Andrei Korbut

xii Acknowledgments

(of Minsk University) who is collaborating with Svetlana on a translationof Garfinkel’s work into Russian. The discussion of translation issueswith regard to texts has been very instructive in deeply theoretical waysand am greatly indebted to both Svetlana and Andrei for the opportu-nity. The hospitality of my hosts in Moscow was exceptional and thefood and conversation everything one could hope for. The vibrant socialdynamic unfolding there at such speed echoes Durkheim’s concern withthe increasing importance of practice in modern society and providedmuch food for both thought and conversation, particularly with regardto the problem of marginalization, the focus of Svetlana’s research.

This book was preceded by the publication of several articles and Imust thank the editorial staff at the AJS and also John O’Neal, CraigCalhoun and Norman Denzin for their guidance and support in seeingthese articles through the publication process in journals they edited. Imust thank also David Fassenfest who was instrumental in helping topublish my recent work on Durkheim’s Division of Labor in Society.

Several colleagues have stood by me from the start, and been generousin support of my career even though they may not always have understoodthe way in which it was unfolding. I thank especially Doug Maynard, PeterManning and Harold Garfinkel in this regard. Doug was instrumental inbringing me to the University of Wisconsin as a postdoctoral fellow inthe 1980s and has worked with me on the institutional development ofethnomethodology for many years. I place a high value on his friendship.

To Harold I owe a debt I am only just beginning to realize the enor-mity of and will never be able to repay. To have had the opportunity ofdiscussing ideas with him over two decades is an immeasurable gift.

Peter Manning is one of a very few people with a mastery of a broad lit-erature on social theory and philosophy who also sustain a deep and abid-ing commitment to interaction as a locus for the doing of both good workand good theory. Conversations with him over many years, beginning in1987 when he asked me to replace him at the University of Michiganduring a sabbatical year, have been a constant source of both inspirationand good sense. Most importantly, Peter has always been a touchstoneagainst which I could test my ideas.

I also owe a debt of gratitude to various persons who have disagreedwith me over the years. They have often pushed me to clarify my ideasin useful ways. I would like particularly to thank Jeffrey Alexander andWarren Schmaus in this regard. Warren provided an opportunity tosharpen my discussion of epistemology in my response to his reply inAJS. Jeff has for many years been generous in spite of disagreementsbetween us and magnanimous in the face of my continued criticism.

Acknowledgments xiii

To those who tried to teach me philosophy and social theory, TomMcCarthy, Alisdaire MacIntyre, Bernard Elevitch, Erasim Kohak, KurtWolff, Harold Garfinkel, John Findley, Gila Hayim, Ephriam Isaac, JeffCoulter, George Psathas, Francis Waksler and James Schmidt, I thankyou. Jeff Coulter and George Psathas in particular helped to focus a rebel-lious mind. To Francis Waksler I owe an additional debt for awakeningmy interest in social theory and introducing me to both ethnomethodol-ogy and phenomenology. A special thanks to Burten Dreben who readover my AJS paper and my reply to Schmaus and made helpful commentsand suggestions.

Bentley College has provided an invaluable haven at a critical point inmy career and the friendship and support of my colleagues, in particularTim Anderson and Gary David have quite literally kept me going throughdifficult times. Tim in particular has encouraged my work from the firstand it is a great pleasure to be answerable to him now.

My sons Martin and Ty have shown more appreciation and under-standing of my preoccupations with thought over the years than I couldever have asked for.

Finally, I thank my father for instilling in me a deep respect for classictexts and the belief that clarifying matters of reason and justice can makea real difference.

Introduction

Emile Durkheim’s Les Formes Elementaires de la Vie Religieuse: LeSysteme Totemique en Australie, published in French in 1912, the first ofDurkheim’s major works to be translated into English in 1915, as The Ele-mentary Forms of the Religious Life,1 offers a theory of mutual intelligibilityachieved through orders of practice, a position that his earlier writings onsocial order assumed, but did not explain, and as such is the crowningachievement of Durkheim’s sociology.2 The book, generally treated eitheras a work on primitive religion, or a sociology of knowledge, and elabo-rately and consistently misunderstood since the beginning, constitutes, infact, Durkheim’s attempt to set his earlier works on a firm epistemologicalfooting. This he achieves by elaborating a theory of practice, as the basisfor mutual intelligibility, which would establish a unique epistemologicalbasis for sociology and the study of moral relations.

The Elementary Forms presents a careful and thorough historical andcomparative argument for the empirical origin of six basic ideas, or cat-egories of the understanding, identified by the philosophical debate asessential to epistemological validity (time, space, classification, force,

1 There was an abridged translation of Sociology and The Social Sciences in 1905, but TheDivision of Labor, Rules, and Suicide were not translated until 1933, 1938 and 1951 respec-tively.

2 The 1915 translation, by Ward Swain, was published by the Free Press. A new translationby Karen Fields, with the title The Elementary Forms of Religious Life, was published in1995, also by the Free Press. There is also an abridged translation by Carol Cosman,published by Oxford University Press in 2001, with the title The Elementary Forms ofReligious Life. The latter is a translation of the 1991 French edition of the text and allpage references are to that edition. All references to Durkheim’s text in this book willinclude page references to the original French edition as well as to the 1915 and 1995complete English translations. In each citation an asterisk will precede the date of thetext from which the quotation was taken. In one or two cases, page numbers for the 2001translation will also be given. I began working with the 1915 translation in 1990. Newtranslations and editions are appearing faster than I can keep up with them. A practiceof numbering paragraphs and sections in social theoretical works, as is done for majorphilosophical works, would simplify the citation process.

1

2 Epistemology and Practice

causality and totality).3 This argument, which treats religious practiceas the foundation of social life, speaks to the current context of conflictbetween religion and what is generally referred to as secular culture in anera of globalization. The ascendency of the secular in the west, followingthe bloodbaths of the Protestant Reformation, reduced the scope of reli-gion in public life. This has left many people feeling that the major publicinstitutions lack a moral base. The resulting erosion of trust has fosteredan increasing tendency to turn to traditional religious institutions as away of rebuilding a sense of “community.” Recent studies which showthat religious participation makes people feel more secure and promotes ageneral sense of well-being should footnote Durkheim, who argued thatreligion would always be necessary, not only to foster feelings of well-being, but also to ground essential ideas arising from those feelings.4

3 That Durkheim’s argument focuses on only six categories (time, space, classification,force, cause, and totality) which he makes empirical arguments for the origin of in TheElementary Forms has been missed. Commentators on the epistemology apparently do notgrasp the exclusivity of the list and it is represented in different ways by different commen-tators, often with an “etc.” This may be due in part to references which Durkheim himselfmakes in several places to categories in the work of Aristotle and other philosophers. Thefirst reference in The Elementary Forms, for instance (Durkheim, ([1912:12–13]∗1915:21–2; 1995:8–9)) is to Aristotle’s list of categories and includes “personality” and “num-ber” which do not appear as categories in Durkheim’s argument (although he doesmake an argument for the origin of personality). Number appears again in ([1912:12–13]∗1915:21–2; 1995:8–9) and personality again in ([1912:26–8]∗1915:31–2; 1995:17–18). These instances all appear in the introduction where Durkheim is making referenceto Aristotle’s list of categories and to the general philosophical problem with regard tocategories. These are not the six categories which he argues for the empirical validity ofin the body of the text, however. One result of this misunderstanding is that one of themost important of the six categories, classification, is generally treated as a survey of clas-sification practices and not as a category in its own right. One reviewer commented “Thequestion that nags me most is why Durkheim’s analysis of classification has promoted somuch research while his analysis of the categories (space, totality, time, force, causality,etc.) has promoted so little.” I think the question answers itself. Just as this reviewer didnot recognize classification as one of the categories (a list to which they also added an“etc.”) the general sociological public have also not recognized Durkheim’s studies ofclassification as having anything to do with his epistemology. Therefore, while the epis-temology has been almost totally ignored, the part of it which focused on classification,because it has been misinterpreted as a survey of symbolic systems, has received a greatdeal of attention.

4 For instance in a recent editorial by George Will, in The Boston Globe, he reports that sci-entists have discovered that there is a biological need for people to connect and that thismay manifest in religious participation. He writes that “The scientific fact, if such it is,that religious expression is natural to personhood, does not vindicate any religion’s truthclaims. A naturalistic hypothesis is that the emotions of religious experience have neu-robiological origins: The brain evolved that way to serve individual and group survival.”The only real difference between this argument and Durkheim is that having stressedthe importance of social connectedness throughout his article, Will, and the scientistshe cites, want to locate the origin of everything that has to do with social connected-ness in the individual. This, as Durkheim points out, will not work. To be sure there are

Introduction 3

However, Durkheim’s argument made some very important distinc-tions between forms of religious practice in modern and traditional lifethat the current discussion proceeds in ignorance of. It is the function ofreligious practice in establishing essential shared sentiments and ideas thatDurkheim argues is a necessary foundation for social life, not religiousbeliefs. For Durkheim this means that much of what is currently consid-ered secular has, in fact, taken over and fulfills the functions of religiouspractice. If religious practice is closely tied to belief then it will necessarilybe at odds with a national, not to speak of an international set of socialrelations, which must be based on something like what Norbert Eliascalled civility: a set of civil practices that operate to create social unity inthe absence of shared beliefs. Durkheim argued in The Division of Labor(Book III Chapter Two), that a sense of unity and well-being based onshared belief, while it is comforting to group members, ultimately threat-ens the security and solidarity of an advanced division of labor because itleads inevitably to exclusive groupings within the larger collective. Whatis needed in a modern context is solidarity based on shared practice notshared belief.

In arguing that religion played an essential role in establishing a sharedknowledge base, Durkheim was rejecting existing approaches to the prob-lem of knowledge, replacing explanations that began with the individualwith his own socially based argument that knowledge is created by theshared experience of enacted practices. His argument privileges enactedsocial practice over beliefs and ideas, an innovation that avoids dilem-mas inherent in philosophical approaches to knowledge and morality thatare based on individualism, and the privileging of beliefs and ideas overpractices; both dominant tendencies in western thought (Rawls 2001).

The problem of intelligibility, whether acknowledged or not, lies atthe center of any social theory. Persons cannot cooperate to maintain asocial order unless they can communicate. Therefore, the two problems,of order and intelligibility, are not separable. A Durkheim whose earlierwork was favored, and who was interpreted as not having addressed the

limitations that can be placed at the doorstep of the individual: the inability to transferideas into the heads of others, a natural inclination for survival that would prevent socialconnections from forming. The solution to these problems comes from social relations.Religious practices that work by creating an emotional response in the person solve theseproblems. Of course human biology has to cooperate, but as Will also reports, scientistsare saying that social relationships alter the “hardwiring” of the brain, and the biology ofthe human body. So, it is obvious even within his own article that human beings are beingshaped by social relations and that biology is made that way to serve the needs of society.This was Durkheim’s argument, much criticized at the time. The explanation cannot befound in the individual if the changes come from society. (George Will, The Boston GlobeSeptember 22, 2003, A11).


problem of intelligibility, assuming instead a naively positivist approachto knowledge, has seemed increasingly irrelevant to contemporary soci-ology since at least the 1950s when it began to become apparent thatintelligibility was, if anything, the more important of the two questions.Then, interpreted by Jeffrey Alexander and others as a proponent of thesociology of knowledge in the 1980s, an interpretation that favored thelater work, Durkheim was resuscitated as a precursor of the postmod-ern critique. This interpretation, however, while engendering a renewedinterest in The Elementary Forms, continued a trend that had begun beforethe 1920s of treating the early and later work as fundamentally different.

It is important, then, that enacted practices, which are the keystoneof Durkheim’s epistemology, offer a focus on the mutual achievement ofintelligibility through practice in a way that supports the arguments ofhis earlier work, and thus reveals his overall position as having a unityof vision with unexpected relevance to contemporary debates over thecentrality of interaction and moral issues to social thought.

Durkheim, who is generally thought of as a macro theorist of socialorder, and a champion of the status quo, was in fact focused on the prob-lem of intelligibility and the limits that the need for moral reciprocity atthe level of local enacted practices impose on social forms as a prerequisitefor the achievement of intelligibility. Taking seriously the argument thatenacted practices constitute the foundation of intelligibility and socialorder entails that orders of practice, or interaction orders, come beforeand underlie institutional orders and their corresponding accounts (Rawls1987). Durkheim had argued in The Division of Labour ([1893]1933) thatorders of practice replace shared belief as the foundation of solidarity inan advanced division of labour (Rawls 2003). Maintaining a commitmentto such orders is thereby a moral imperative in a context of globalization.

Durkheim’s position is, in fact, aligned with contemporary interac-tionist arguments that are usually considered to have departed signifi-cantly from classical social theory. Yet, not only was Durkheim engagedin making a distinction in 1912, similar to the one made by Goffman andGarfinkel, between orders of practice and institutions, and articulatingthe moral commitments required (in Durkheim’s case between prospec-tive practices and retrospective accounts), but, he had already outlinedthe argument in Book III of The Division of Labor in Society in 1893 (Rawls2003). The distinction, however, would not begin to be taken seriouslyuntil C. Wright Mills distinguished between following rules and acting inways that are accountable to rules in “Situated Action and Vocabularies ofMotive,” in 1940, and it would not achieve any widespread impact untilHarold Garfinkel introduced “institutional contexts of accountability” asa way of understanding the relationship between prospective orders of

Introduction 5

local practice and the way institutional contexts orient practices towardaccounts in Studies in Ethnomethodology, in 1967.

While the details of Durkheim’s marriage of epistemology and enactedpractice are worked out only in The Elementary Forms, the argument asit appears there is a logical extension of ideas presented in earlier pub-lications in which Durkheim argued for the importance of science andmethods, social facts, the distinction between mechanical and organicsolidarity, and the development of classifications through religious prac-tice. The Elementary Forms was intended once and for all to clarify thesearguments, setting them on a strong and unique epistemological foun-dation which placed social practices rather than the individual and theirbeliefs and ideas at the center.

Allusions by Durkheim to what would later become his epistemologyappear in, and are central to, the arguments of The Division of Labor inSociety ([1893]1933). The Rules of the Sociological Method ([1895]1982)and Suicide (1895). In The Division of Labor Durkheim used many exam-ples drawn from Australian aboriginal religious practices to illustrate hispoint that solidarity, in what he called a mechanical grouping, depends onthe mutual enactment of practices designed to align the emotional lives ofmembers of the group, producing what he referred to there as “collectiveeffervescence.” He also argued in that text (Book III Chapter One) thatbeliefs are only secondary and retrospective phenomena, arising from theattempts of participants to explain the feelings generated in them by theirmutual enactment of shared practices. His point is that these explana-tions are not designed to represent their underlying causes and purposesand, therefore, necessarily distort knowledge of social relations, the sameargument that he would make later in The Elementary Forms.

Because of this distortion shared beliefs appear to be essential to socialsolidarity. However, they are in fact not essential. It is the enactment ofshared practices that is essential and, therefore, attempts to fix social prob-lems by strengthening beliefs (as for instance by strengthening traditionalreligious communities – or through a general philosophy – as Comte hadproposed, a problem that Durkheim takes up in Book III Chapter Oneof The Division of Labor) when beliefs no longer support the necessarypractices, are misguided. This should become clearer, Durkheim argues,as the role of shared practices increases with the advance of the division oflabor, and practices come to overshadow shared beliefs (see Rawls 2003for an extended discussion).

The epistemology of The Elementary Forms also extends Durkheim’searlier arguments regarding the empirical status and scientific validityof what he called “social facts” in the Rules of the Sociological Method([1895]1982) and Suicide ([1897]1951). The emphasis on social facts,


generally interpreted as naively positivist, changes its character when itis understood that for Durkheim the recognizability and validity of socialfacts are produced only in and through participation in social practices.The argument can then be seen to involve processes of mutual socialconstruction to an extent that is very contemporary. Garfinkel’s referencesto “Durkheim’s Aphorism” are intended to underscore the importance ofinspecting “social facts” for their dependence on the situated occasionsof their construction, in and through the mutual enactment of practices,in particular social scenes (Garfinkel 2002). This is a form of analysis thatDurkheim outlines, but did not and, given the limitations of the empiricalmaterials available at the time, could not complete.

Durkheim also touches on the epistemological argument in three otherworks. The essay on Primitive Classification ([1901]1963 co-authoredwith Marcel Mauss) outlined the parameters for the origins of the cat-egory of classification, but did not attempt to distinguish the social logicof the concept (the sociology of knowledge) from its genesis in enactedpractice (the epistemology). The lectures on Pragmatism (1913–14),published in English as Pragmatism and Sociology (1983), worked outthe classical epistemological problem in some detail and critically eval-uated the pragmatist5 solution to the problem which was emerging inDurkheim’s day. But, Durkheim’s own epistemology is not elaborated inthat work.

The essay “The Dualism of Human Nature and its Social Condi-tions,” published in 1914, two years after The Elementary Forms, clarifiedDurkheim’s position that human reason arose as a result of participationin social practice, contrasting reason with what Durkheim consideredto be a preexisting animal or biological nature of the human being.6

The essay bears similarities to Rousseau’s argument, in the Discourse onthe Origin of Inequality ([1757]1999), that reason developed only afterhumans became social, and elaborates a theme concerning the primacy

5 The Pragmatist position is also sometimes referred to as social constructivist and as suchit was also criticized by Durkheim. The difference between Pragmatist constructivismand Durkheim’s position is that it begins from the perspective of the individual and con-structs outward. Even forms of constructivism that began with concepts institutionalizedthrough language would conflict with Durkheim’s practice-based constructivism. Thissame conflict exists today between different forms of interactionist constructionism. Sym-bolic Interactionists sometimes take the Pragmatist position, while others, more closelyfollowing Goffman and Garfinkel take a solidly social view consistent with Durkheim.For Durkheim both the individual and social are ultimately constructed through enactedsocial practices. But, it is the assembled group doing the constructing through sharedpractices, not the individual.

6 The argument of this essay has sometimes been confused with an earlier article, “Indi-vidual and Collective Representations,” written in 1898, in which Durkheim criticizedthe radical empiricism of William James. This confusion has resulted in a long history ofmisinterpretation in the secondary literature (see Rawls 1997 for an elaboration).

Introduction 7

of the social condition that is evident in Durkheim’s earlier lectures onRousseau (published in English along with his Latin thesis in a volumeentitled Montesquieu and Rousseau, in 1960). Durkheim’s emphasis on thesocial ultimately constitutes an important critique of the Enlightenmentfocus on the individual and in this regard his position is similar to Marx.7

0.1.0 Durkheim’s Epistemology: the Neglected Argument

Although Durkheim’s work has been the subject of extensive criticism andcommentary, and hundreds of books on Durkheim, and on The Elemen-tary Forms, in particular, have been written, somehow in the process, theoriginal epistemological argument made by Durkheim in The ElementaryForms has been almost completely neglected. Most often his sociologyof knowledge is treated as if it were intended to be an epistemology. Itwas not. When the epistemology is mentioned, the argument is generallymisunderstood, summarily dismissed, and Durkheim’s position charac-terized as naive or contradictory. Some commentators even claim thatDurkheim ignored epistemology altogether.8

Authors typically dismiss Durkheim as merely another Kantian, Carte-sian rationalist, empiricist, or pragmatist, often combining one or moreof these labels, ignoring the incompatibility between them. In spite of theobvious contradiction in attributing multiple conflicting positions to onethinker in a single work, these attributions all seem to be fairly generallyaccepted, often by the same scholar, and sometimes in the same sentence.The fact that scholars continue to rely on secondary source traditionsregarding Durkheim’s text contributes to the persistence of this problem(see Rawls 1997a for an extended discussion). The neglect of the text inthis regard, the reliance on secondary sources, and the almost universalresort to philosophical positions criticized by Durkheim, in attemptingto explain his argument, is somewhat puzzling, as the epistemologicalargument is not a minor theme of Durkheim’s text.

While it is true that the epistemological argument in its entirety appearsonly in The Elementary Forms and not in Durkheim’s other works, there itis the major preoccupation of the work and is laid out systematically overits entire course. The whole book, each discussion of religion, philoso-phy, anthropology, or aboriginal society, is a careful empirical elaboration

7 Durkheim’s position is also similar to Marx in arguing that social inequality perpetuatedby shared beliefs constitutes a fundamental contradiction of industrial capitalism. SeeBook III of the Division of Labour and Rawls 2003 for an extended discussion.

8 For instance, Nisbet maintains that Durkheim ignored the epistemological question alto-gether. Although Giddens rejects the argument that there are two different Durkheims,he says that Durkheim is a Kantian who argued for the social origins of the elementaryforms of reason, a misunderstanding of both Durkheim and Kant.


of Durkheim’s epistemological claims. The examples are so specificallytied to the epistemological argument that Durkheim even differentiatesbetween the sorts of totemic rites in terms of their correspondence tothe development of different categories of the understanding, and not interms of their similarities and differences in more conventional structuralor conceptual terms. That is, different rites generate different sorts ofsocio-empirical experiences and therefore give rise to different categories.Rites of sacrifice and oblation, for instance, generate the general categoryof force, while imitative rites generate the general category of causal-ity. Thus, certain rights lay the foundation for the development of otherrights/ideas, and the organization of the book by types of right reflectsthe correspondance between certain rights and certain categories. Thisorganization of the discussion of religious ritual in terms that are dictatedby epistemological concerns has most likely been confusing to scholarswho did not recognize the epistemological focus of Durkheim’s argument.

The epistemological argument is essential to an understanding ofDurkheim’s overall position. In fact, it isn’t too much to say that sociologyitself cannot properly be understood without Durkheim’s epistemology.Durkheim intended the epistemology to lay a foundation for valid socio-logical argument. Understanding The Rules of the Sociological Method, forinstance, in the absence of Durkheim’s epistemology, leads to the curiousresult that Durkheim, and the discipline that he is said to have founded,appear to be positivist, when Durkheim provided the proof that it was not.The argument that social facts have an “objective” reality that is witness-able in its details does not mean that Durkheim was a positivist if thosesocial facts are mutually constructed through enacted practices. The dis-tinction between practices and concepts allows Durkheim to argue thatpractices which are publically enacted can be seen and heard, whereasconcepts and ideas cannot. The discipline, including social theory proper,in continuing to privilege concepts over practices has developed in adirection which Durkheim would have repudiated for its epistemologicalcontradictions.

0.2.0 Epistemological Crisis

When Durkheim initially articulated his epistemology, questions of epis-temological validity and scientific knowledge were hotly debated. At theend of the nineteenth century an epistemological crisis of major pro-portions had been reached. It looked as if all claims to knowledge werehopelessly relative; that no knowledge was valid. Philosophers had arrivedat this dilemma after more than a century of debate over the arguments ofDavid Hume and Immanuel Kant. Epistemology had, from its beginnings

Introduction 9

in Greek philosophy, struggled with a separation between thought andreality, occasioned by essential differences between the two: thought con-sisting of concepts, which are general and continuous; reality consistingof flux and change. Proponents of the newly developing sciences in thesixteenth and seventeenth centuries wanted to establish an empirical basisfor key scientific concepts. In the 1690s it seemed as if John Locke hadsucceeded in establishing this possibility through a careful analysis of whathe called simple ideas. But, by 1735 Hume had introduced convincingproofs that key ideas, like causality, on which, he argued, all statementsof fact depend, have no basis in individual perception.

The implication of Hume’s argument was that empirically valid knowl-edge was not possible. When in 1754, Kant found a way of addressingHume’s dilemma, arguing that certain key ideas, referred to by Kant as“the categories of the understanding,” exist a priori in the human mind,philosophers flocked to embrace Kant’s solution to Hume’s dilemma. Butthe solution came at a price. After Kant, epistemology had to deal witha further separation between thought and reality created by the facultyof human understanding: because the categories of the understandingwere considered by Kant to be a priori, natural reality would always beperceived in terms of human categories of thought, and never in itself. Inthe case of both Hume and Kant, then, human ways of perceiving andthinking were thought to add something to reality which was not there inthe original. As a consequence, it seemed impossible for human knowl-edge to stand in the sort of empirically valid relationship with reality thatwas required by science.

Durkheim had, throughout his career, been a proponent of science.He believed that many social problems were exacerbated by unscien-tific “solutions.”9 His task, as he saw it, was to establish valid empiricalgrounds for the study of social relations, and in particular those social rela-tions that were properly moral relations, which determined the possibilityof rational, stable and equitable social life. For this he needed to groundhis studies on an epistemology that would establish social and moral rela-tions as possible subjects of valid empirical study. Durkheim situatedhis argument within the context of the epistemological debate betweenempiricism (including Pragmatism) and what he called apriorism: thatis, between Hume and James on the one hand, and Kant on the other.10

9 In particular Durkheim criticized socialism for advocating broadscale social reforms onthe basis of unfounded assumptions about the relationship between shared beliefs andsocial solidarity in an advanced division of labor context. See The Division of Labour([1893]1933) Book III and Rawls 2003. See also Durkheim ([1895–6]1958).

10 William James had become popular in France around the turn of the century and hadbeen invited to Paris to lecture. Thus, Durkheim was confronted in his own intellectualcircle with James as a compelling proponent of empiricism.


Durkheim also extended his consideration to empiricist and a prioristexplanations of the origin of religious ideas, and in particular, the originof the sacred.

Durkheim’s epistemological argument, articulated in the central chap-ters of The Elementary Forms, locates the origin of the fundamentalcategories of human thought, or reason, not in individual perception,as Hume had argued, nor as a transcendent and innate aspect of themind, as Kant had argued, but rather, in the shared emotional experi-ence of those ritually produced moral forces created by the enactment ofconcrete practices in the midst of an assembled group. This constituted aradical departure from the existing alternatives and promised to addressthe inherent dilemmas in novel ways.

Durkheim felt that he had established an epistemological foundationfor sociology that would allow it to address the great and pressing ques-tions of moral philosophy that were increasingly being abandoned in hisday.11 He felt that modern society was heading toward a moral abyss,because of a failure to achieve justice. In the past, he felt, sufficient moralguidance had always come from society. However, due to the degree ofreligious and cultural pluralism in modern society, religious and culturalinstitutions based on shared belief could no longer provide the moralguidance needed for society as a whole. Therefore, that guidance wouldhave to come from broadly based secular institutions. These he arguedwere failing to deliver sufficient justice to support personhood and intel-ligibility in a modern context. Durkheim felt that a scientific study ofsociety would reveal how society had been able to produce moral feelingsin the past, and also explain the current period of moral mediocrity. Forthis he needed to establish a valid, empirically based, science of society.His aim, in this regard, is no different in The Elementary Forms than it hadbeen in the earlier Rules of the Sociological Method, the opinions of variouscritics notwithstanding.

During the course of the twentieth century, due to a growing consensusthat an argument for empirical validity could not be made, philosophersincreasingly abandoned the classical form of the epistemological ques-tion, which required empirical validity, in favor of a neo-Kantian, andfinally a Pragmatist, or social constructivist, approach to knowledge as

11 The reasons for the abandonment of moral philosophy were the same as those for theabandonment of epistemology. Moral philosophy depends on arguments from “reason”and with the abandonment of epistemology, “reason” had given way to Intuitionism,a school of moral philosophy that was in many ways the counterpart of Pragmatism inepistemology. Hume and Kant had each ventured into epistemology only in order toestablish a basis for their moral philosophies; Hume basing his argument on the passionsbecause he could not establish reason, and Kant basing his argument on reason, becausehe thought he could. When Kant’s epistemology fell, Intuitionism, like Pragmatism,reigned.

Introduction 11

justified belief.12 This process was already well underway by 1912. Forthe neo-Kantians, reality was forever out of the reach of human knowl-edge. The Pragmatists and social constructivists treated social consensusand socially accepted definitions of meaning as the true measures defin-ing the limits of validity. The persistence of the dilemma can be seen inthe influence of postmodern and pragmatist approaches within the socialsciences and humanities today.

According to Durkheim, however, this abandonment of the classicalquestion, in favor of a consensus theory of truth, only appears to be neces-sary because the epistemological question has been cast, by both empiri-cists and a priorists alike, in individualist terms: in the form “how canindividual perceptions of natural reality give rise to valid knowledge of thatreality?” Durkheim argued that because individual ideas are not the origin ofhuman reason, this way of posing the question makes it appear to be unsolvable.According to Durkheim, replacing the individualist approach of tradi-tional philosophy, with an approach solidly embedded in enacted socialpractice, was the only possible solution to the epistemological dilemma.

A sociological approach to epistemology is necessary, for Durkheim,because knowledge begins with relations between persons, not with theindividual. The individual human perceiver does not exist outside of, orbefore, society. Therefore, to begin with the individual is to begin withthe result of a social process reifying the individual by treating it as if ithad an independent existence.13 The epistemological question remainsunsolvable because the process through which human individuals aremade rational and human in the first place remains unexamined.

0.3.0 Religion and Reason

Along with other social thinkers (trained as philosophers) who have cometo be known as the theoretical founders of sociology, Durkheim rejectedthe rational individualism of Enlightenment philosophy.14 Replacing

12 For a more extensive treatment of the demise of classical epistemology within philosophyproper see my reply to Schmaus in AJS (Rawls 1998).

13 In criticizing individualism in this regard Durkheim’s critique of classical epistemologyis similar to Marx’s critique of classical economics. Marx argued that classical economicsrendered relationships, the most fundamental economic phenomena, in his view, invis-ible, by focusing on the individual. The “real” economic forces, according to Marx,consisted of relations between producing persons and classes of persons, and thereforecould not be grasped by any approach that either began with, or was epistemologicallybased on, the individual, or more general economic ideals. The problem with an indi-vidualist approach to epistemology, according to Durkheim, is similar to the problemwith an individualist approach to economics; individualism renders those fundamen-tal relationships and practices, that constitute both society and the rational individual,invisible.

14 Stjepan Mestrovic, in his numerous books and articles on Durkheim treats it as sig-nificant that both Durkheim and Schopenhauer rejected Enlightenment philosophy.


rational individualism with a socially based epistemology was an essen-tial element of the classical socio-theoretical arguments of Marx, Mead,Weber and Durkheim, although Durkheim was the only one to articulatean epistemology as such. Misunderstanding this thrust leaves Enlight-enment elements in contemporary sociological theory that are highlyproblematic. The postmodern attempt to eliminate these Enlightenmentelements from sociology would be unnecessary if the epistemologicalbasis of classical sociological theory had been understood.15

Enlightenment philosophy was born at that moment in western historywhen religions and the unity of cultural values had lost their ability toensure peace and harmony within a people and provide adequate moralguidance for political and social relations. Not only the development ofinternational relations, but the multiplication of religions within countriesand a growing cultural pluralism meant that, for the first time in history,religion and culture created divisions within a single people instead ofpromoting unity and harmony (religion and culture had, of course, alwaysprovoked conflict between peoples). Religions and cultures which conflictwith one another cannot provide an adequate foundation for political, ormoral, order and as a consequence the social stability of western societyfound itself on a very different footing, after the sixteenth century, fromthe unity experienced by homogeneous groups.16

Enlightenment philosophy constituted an attempt to transcend theseemerging divisions and the cultural, religious, and social structural differ-ences which created them, by founding moral imperatives on universalideas, or universal reason. The positing of a “rational being” as a uni-versal phenomenon was a keystone of Enlightenment philosophy. To berational, however, Enlightenment philosophers thought that individuals

For Mestrovic this is grounds for finding similarities between their two positions.It is true that along with most nineteenth-century philosophers both Durkheim andSchopenhauer rejected Enlightenment philosophy. However, they reject it in very dif-ferent ways. Enlightenment philosophy posited the rational individual. Schopenhauerrejected reason in favor of the will, but did not reject the individualist orientation ofEnlightenment philosophy. Therefore, he ends up championing the individual will.Durkheim rejects the individualist orientation without rejecting the emphasis on rea-son. Therefore, he ends up with a collective source for reason, with very little role forthe individual will. That the two both reject Enlightenment thinking does not mean thatthey have anything in common.

15 Because Symbolic Interactionism has origins in Pragmatism which retains an individu-alist perspective some interactionists identifying themselves as Symbolic Interactioniststake an individualist view. This is not true of all Symbolic Interactionists. It is certainlynot true of interactionism in general, nor is it true for Sociology as a whole. It is onlyby misunderstanding the social construction of self and ideas which is the underlyingpremise of most sociology that Post-Structuralists still feel the need to transcend Enlight-enment ideas.

16 This became the distinction between Gemeinschaft and Gesselschaft, which Durkheimreferred to as mechanical and organic forms of solidarity.

Introduction 13

needed to be stripped of all cultural and religious overtones; strippedto the bare reason, so as to rise “above” cultural disagreements. Thisis one of the reasons why western philosophy considers the social to benecessarily contingent.

Thus, the image of the rational individual was necessarily conceivedof as asocial in order to eliminate culture from its makeup.17 Againstthis context, the classical social theorists pointed out that there are manydifferent experiences of self and other, and differing forms of reason.Different relationships between persons and their societies create theseexperiences. In all cases, reason and self are generated by social rela-tionships. In his lectures on Rousseau, Durkheim pointed out that theargument that reason has a social origin first appears in Rousseau’s Dis-course on the Origins of Inequality, even though for Rousseau, once reasondid develop, it took on a universal character with no social content. But,for Rousseau there is no asocial condition of the rational being: there isno rational being, there is not even a recognizable “person” at all, withoutsocial relations.

Classical sociological theorists, building on this line of argument,insisted that the image of the rational individual, promoted by Enlight-enment philosophers as the foundation of a universal ethics, could not infact be the starting point for a truly universal conception of the person,because the idea of the rational individual was itself a result of westernsocio-economic development; an artifact of the division of labor in mod-ern society. In other words, reason itself, and the rational individual, inthe form in which the Enlightenment project generally proposed it, wasitself a cultural phenomenon.

When the classical social theorists argued that the individual had asocial origin, they also rejected the philosophical prejudice against thesocial as necessarily contingent. Durkheim argued that the social ori-gin of the rational individual did not necessarily lead to epistemologicalrelativism. Furthermore, he believed that underlying social forms wouldprove to have stable, even necessary, features over time. Durkheim’s socio-empiricism18 focuses on a dynamic relation between the group mem-ber, as a participant in ritual social processes, and the social processestheir participation enacts. The rejection of the rational individual and itsreplacement by the social is not meant to result in relativism. Durkheimconceives of some social relationships as so basic and necessary that theymust occur in some form in all cultures in order for human reason to exist.There would be some variation in the form of the rituals themselves, but

17 Rousseau and Hume and later John S. Mill rebelled to some extent against this viewemphasizing emotions and Durkheim follows their lead.

18 I use this word to distinguish Durkheim’s empiricism from a naturalistic positivist empiri-cism. For Durkheim social facts have a special empirical presence as social forces.


the experiences, and the categories of reason produced, would not vary.Durkheim replaced what he saw as the relativism of the Enlightenmentindividual, with universal and necessary social practices.

The purpose of religion in human history is, according to Durkheim, toprovide the enacted practices necessary to generate the emotional experi-ences and their corresponding ideas essential to rational beings. Withoutthe enactment of such ritual practices, he argues, human reason wouldnot develop and therefore society could not be sustained. This is whatDurkheim means when he argues that society precedes the individualrational being. It is not an argument for historical precedence.

Certain practices, Durkheim argued, understood as sounds and move-ments, could create the same sentiments in all participants in assembledgroups.19 These sentiments in turn constitute a basic set of shared ideaswithout which, he argued, mutual intelligibility could not be achieved.The attempt by collective groups to explain the unique sentiments felt bythe group when enacting ritual practices, gives rise to elaborate narrativeswhose purpose is to explain the origin of those feelings. These narrativescome to be known as religious beliefs. These narratives do not, however,locate the “real” cause of the feelings, or emotions, in the enactment ofpractice itself. Rather, they explain the emotions created through enactedpractice by invoking great natural and supernatural forces. These culturalrepresentations and beliefs may, in turn, become social forces in their ownright. However, for Durkheim, they bear only an allegorical relationshipto enacted practices and cannot produce the essential emotions on theirown.

19 The idea that all participants would get the “same” sentiment needs to be explored.For instance, would men and women experience the same sentiment? If both men andwomen enact the same ritual together and the ritual creates the unity of the group thenboth would experience that as causality in the same way. On the other hand, if onlymen participate in one ritual and only women in another, then the moral boundariesseparating men and women may be strengthened by the ritual. However, if the ritualsboth enact causality the resulting ideas should still be the same. Participation in enactedritual would make men and women have different ideas only if the rituals themselveswere enacting fundamentally different forms of association. If women, for instance, takepart in more rituals of solidarity while men take part in more rituals of differentiation,then their fundamental ideas may end up being different. But, if they experience someof each form of practice they would all have the same ideas, but in different proportions.It would only be a matter of a different balance between fundamental ideas. They wouldboth have the same ideas, and should be able to communicate, but different expectationswith regard to practice would play the guiding role in shaping interaction orders ofpreference and expectation. Given that practices are more important than ideas, however,the real problem is that the preferences for local interaction orders could be slightlydifferent, producing misunderstanding. See Deborah Tannen (1990) for a discussion ofthis problem in gender relations and Rawls (2000) for a discussion of what I refer to asInteraction Orders of Race.

Introduction 15

While natural forces could only be perceived by individuals as par-ticulars, social forces, Durkheim argued, are inherently dynamic andcontinuous and, when experienced as such by groups of persons assem-bled together to enact practices, provide an empirical source for sharedconcepts. It is the experience of moral force 20 that creates the rationalidea, not the belief system. This solves the problem of social contin-gency, because whereas beliefs and rituals might differ greatly, it is hardto see how the experience of moral force by the assembled group couldbe different from case to case. Experiencing causality as a moral force,for instance, would in essential respects have to be the same every time aritual produced or enacted causality as a moral force. Otherwise, it wouldnot make sense to say that the ritual enacted causality.

This way of addressing the gap between thought and reality replaces theindividualist approach, which characterized both empiricism and aprior-ism, with an epistemology that is socially based. The resulting episte-mology treats concrete social practices as natural processes whose func-tion is to make general categories of thought available to their humanparticipants. In this way Durkheim hoped to overcome the epistemolog-ical dilemma resulting from individualism and address the problem ofmutual intelligibility, the foundational issue from a sociological perspec-tive. Social order is only a secondary issue, important because mutualintelligibility requires it. In rejecting the individual as a starting point,the way is opened for Durkheim to explain the origin of necessary basicconcepts in terms of concrete social processes; while nevertheless avoid-ing cultural relativity, something that had not been done before.

0.4.0 Perception versus Emotion

Attempts by empiricists to explain the origin of general ideas through adetailed logical analysis of individual perception, and the logical relationbetween objects in perception, had concluded that not only logical rela-tions, but all relations, are properties added by the mind, and not part ofthe perceived object in its own right. Thus, the classical empiricist attemptto establish a direct relationship between perception and an underlyingnatural reality ended in skepticism. Durkheim criticized classical empiri-cism for failing to explain the possibility of deriving even the simplest

20 The emotional character of the perception of moral force is an important issue. By claim-ing a distinction between emotional experience and perceptions which come through thefive senses Durkheim reinforces his argument that the experience of social forces is notsubject to the same problems as the perception of natural forces. Not only do socialforces have continuity in their own right, but they are perceived through a different fac-ulty of mind; an emotional faculty. This point Durkheim shares in essential respects withHume. See Rawls 1996b for further discussion of this point.


general ideas directly from experience. He said ([1912:19]∗ 1915:27;1995:13) that: “Classical empiricism results in irrationalism; perhaps itwould even be fitting to designate it by this latter name.” However, Kant’sargument, “apriorism,” as Durkheim referred to it, presented no viablealternative, as it resulted from accepting the empiricist dilemma, andthen treating the impossibility of generating the categories empirically asthe basis for establishing their a priori status. Furthermore, while the apriori argument solves the problem of generalization, it renders generalconcepts purely ideal; a path that Durkheim refused to follow. Ideas, forDurkheim, come only after, and never before, concrete social relation-ships.

Durkheim also argued, in his lectures on Pragmatism, that, becauseof its inherent individualism, even the “radical empiricism” of WilliamJames retained the problems of classical empiricism. Because he replacedthe dualism of thought and reality with individual action as a dynamicconnection between the two in a context of utility, James had to giveup the possibility of truth and logic in any classic sense. The consensustheory of truth that resulted claimed to be a social solution, in so faras it focused on consensus between persons. But, as Durkheim pointedout, the social consensus proposed by James did not treat the social asprimary, but rather as a derivative of individual thought and action: aconsensus based on utility and the necessity for compromise, not basedon empirical validity. As a consequence, Pragmatism is only able to claimsufficient validity to proceed with a minimum of mutually coordinatedaction.

While rejecting both empiricism and apriorism, Durkheim also rejectedPragmatism. Current attempts to revive Durkheim’s position by castingit within a Pragmatist framework ignore the force of his criticism. FromDurkheim’s perspective unless some common and valid conceptual foun-dation can be established that is shared by all persons, the problem ofexplaining both individual knowledge and inter-subjective communica-tion, and hence morality, will remain unsolvable and truth and knowledgewill remain indeterminate. Pragmatism, in giving up this possibility, hefelt, gives up too much. Pragmatism does not take a sufficiently socialpoint of departure. By contrast with James, Durkheim argued that, inestablishing the validity of the categories, individual sense perceptionmust be completely replaced by the shared emotional experience of prac-tices that are inherently social, in order to explain the origin of empiricallyvalid categories of the understanding.

Durkheim saw the philosophical dilemma with regard to epistemology,as a result of privileging the individual and ideas, over collective prac-tice, and argued that the problem could only be solved by substituting a

Introduction 17

collective approach for individualism, and practices for beliefs. Throughthis substitution, Durkheim was able to approach the problem of thevalidity of knowledge in the classic sense, while at the same timeaccommodating contemporary Pragmatist and Hermeneutic (interpre-tive) objections to classical epistemology.21 Consequently, in approach-ing epistemological questions, Durkheim did not have to fall back on ajustified belief argument, as his philosophical contemporaries were forcedto do. Neither is his argument idealist. Many times, over the course of TheElementary Forms, Durkheim reiterates the claim that his epistemologicalposition presents an entirely new alternative that is completely differentfrom, and superior to, the others. He believed that a sociology based onhis arguments would replace philosophy in this regard.

0.5.0 Mis-Communication Between Disciplines

Unfortunately, the literature on Durkheim’s work in general, and partic-ularly with regard to The Elementary Forms, treats his interest in redefin-ing philosophy as strictly secondary to his various sociological interests,which are elaborated as a positivist interest in science, and a separate,and idealist, interest in religion. While it is certainly true that Durkheimwas primarily a sociologist, his philosophical interests cannot be sepa-rated from his sociological theory, as it was through his epistemologicalargument that Durkheim sought to establish the possibility of sociologyand its ascendency over philosophy.

Furthermore, Durkheim is generally characterized as a poor philoso-pher whose attempts to “dabble” in philosophical argument do not meritserious attention. But, Durkheim, like other classical social theorists, pos-sessed a more substantial background in philosophy than most of hiscritics (both his own contemporaries and more recent critics). He was aphilosopher by training, completing a “Latin Thesis” on the philosophyof Montesquieu as part of the requirement for his PhD. He lectured onRousseau and other philosophers, including James and classical empiri-cism. Trained as a philosopher, writing for others with the same training,he could assume that his readers had some familiarity with the argumentsof the empiricists, pragmatists, and apriorists with which he wrestled.Unfortunately, since Durkheim’s time, a system of education has devel-oped within sociology that pays very little attention to either social theoryor philosophy.

21 For a discussion of the relationship between the arguments of Durkheim and James seemy “Durkheim and Pragmatism,” Sociological Theory, 1997. See also Hans Joas for abest case of the argument that Durkheim is a Pragmatist.


An enslavement to a history of ideas approach, based on labels, thatnot only privileges context over argument, but also tends to privilege sec-ondary commentators over the primary sources they write about, seemsto have developed in sociology. Labels are allowed, even encouraged, tostand in for theory itself. This emphasis on labeling, rather than on exam-ining the details of argumentation, leaves the average sociologist with noway of understanding the basis of the discipline in which they work. Stu-dents generally take only one or two courses in social theory, usually fromsociologists whose main interest is in some substantive area, who do notwrite about theory themselves and almost no one expects or wants stu-dents to specialize in theory. In this context, it is generally accepted thattheorizing is a matter of interpretation.22

In keeping with this trend, those commentators who do take time todiscuss Durkheim’s epistemology generally begin by naming the debate hewas addressing, using words like empiricism, apriorism, neo-Kantianism,Cartesian rationalism. Rarely do they explain what the terms mean, orlocate anything specific in Durkheim’s text to which the words are sup-posed to refer.23 Furthermore, for decades, labels have been used whenreferring to debates between classical social theory and philosophy, whichare sociological hybrids that do not carry the same meaning in sociol-ogy as in philosophy; words like “social epistemology,” “positivism,” and“realism” versus “idealism.” Furthermore, these terms have no appli-cation to classical social theory in any case because they assume anindividualist and non-social constructivist position that Durkheim andother major classical social theorists rejected. Their use has obscured thephilosophical debate from sociological view, thus, further obscuring thephilosophical significance of the arguments. The reader has no greaterunderstanding of what the argument was about after reading such com-mentary than they did before they read it. Furthermore, they will haveno idea where to go in the text to find what is being referred to.

Philosophers on the other hand, who presumably have the backgroundto recognize Durkheim’s argument, do not, because of its essentiallyheretical form. Although Durkheim was trained as a philosopher, in

22 This is another case, like philosophical individualism, of a method of work creatingthe problem that eventually defeats it. Approaching texts as examples of types of theo-ries leads inevitably to an infinite regress of theoretical interpretation. But, that is notbecause theory itself consists only of interpretation. Approached differently, stepwise asarguments, theories are as concretely empirical as any other social construction.

23 In fact, one of the major objections to my interpretation of Durkheim has been theprevalent understanding of his position as Kantian. An examination of the works onDurkheim that are considered authorities in this regard, however, will show that theclaim that Durkheim was a Kantian was made by fiat, by a process of labeling, notargumentation.

Introduction 19

advocating an empirical study of the social as a corrective to philosophy,and in rejecting rational individualism, he was essentially rejecting thephilosophical starting point. Philosophers generally consider the social,or empirical, to be arbitrary and accidental, whereas the individual hasseemed to be a necessary and logical starting point. From a philosophi-cal perspective, replacing the individual with the social leads inevitably tocontingency. How can the absolute, purely logical, or “ought,” be derivedfrom the imperfect thing that “is”? This question is a philosophical sta-ple. This was the question when Pragmatism embraced the social. But,according to Durkheim the problem with Pragmatism was not that thesocial was introduced into the equation, but rather that the social thatwas introduced remained essentially subordinate to individual percep-tion, projects and ideas.

In contrast with James, Durkheim put the social squarely at the cen-ter of his epistemology. In addition to violating this philosophical taboo,which may have been enough in itself to convince philosophers that hiswork was not worthy of consideration, Durkheim also failed to adequatelyaddress key points in the argument where philosophers could be predictedto have problems with his approach. In particular, Durkheim offered asocial origin for the categories of the understanding without always deal-ing adequately, in sections where the argument is introduced and sum-marized, with the problem of whether persons need to generalize fromexperience in order to absorb these categories (which would seem to rein-troduce problems inherent in the empiricist position), or whether, as asocial framework they are absorbed as whole cloth, so to speak.

As organizing structures of thought which require no generalizationon the part of individuals because of their social origin, Durkheim’s cat-egories might avoid the empiricist dilemma. An extended and thoroughdiscussion of how this is possible in the Introduction, in addition to muchlater in the text, would have helped philosophers to take his argumentseriously. Durkheim also failed to clearly distinguish between collectiverepresentations as they appear in his sociology of knowledge and collec-tive representations as they apply to his epistemology, sometimes usingdifferent terminology, but other times using the same term for both cases.

As a consequence, some critics have argued that Durkheim’s claimswith regard to his epistemology reveal an ignorance of the epistemo-logical debate in philosophy. But, Durkheim’s argument is at all pointscouched in a sophisticated understanding of what was involved in thatdebate. Durkheim reviews empiricism, apriorism and Pragmatism, indetail, pointing out problems with each. Reason, according to Durkheimhas its origins in inherently social experiences, not in individual percep-tion. Hume had shown that individual perception could not be the basis


of a valid epistemology in the classic sense. Durkheim analyzed Hume’sargument several times in The Elementary Forms, and agreed completelywith Hume’s conclusion. But, Durkheim argued that Hume’s emphasis onindividual perception had itself caused the problem that Hume was trying, butunable, to solve.24

According to Durkheim the shared emotional experience of moralforces was the real origin of the categories of the understanding. Notbeginning with the social would therefore, necessarily cause problemsfor an epistemological argument because it would leave out the origin ofknowledge: the thing allegedly being examined. However, it was equallyproblematic to focus on social beliefs and values as a source of knowledge,as many followers of Durkheim, including Parsons and Levi-Strauss,did. The social facts that needed to be examined were practices, notbeliefs or narratives. These, Durkheim argued were only retrospectiveconstructions that obscured, without preserving, the underlying socialfacts. Durkheim argued that it was only during certain ritual activities,or practices, performed within assembled groups, that social experiencecould eclipse individual perception in the manner required for the devel-opment of rational categories of thought.

It was the process of enacting practices, not the resulting concepts, thatcould serve to ground epistemology, Durkheim argued. The categoriesof rational thought were a direct result of such social experiences, andnot available to the individual outside of, or prior to, participation inparticular sorts of ritual social contexts. Contact with the culture mightgive the individual the expected uses of words, as Wittgenstein argued,but not the shared experiences that they “call up” (which in order tobe mutually intelligible must be shared, or collective). Obviously, not allconcepts will be mutually intelligible in this way. Some would functionas collective representations and have use meaning only. But, certain keyconcepts would have their origin in the emotional experience of sharedenacted practice, and this would be sufficient to ground the rest.

Durkheim believed he had shown that some essential social rituals arenecessary, both for the development of human reason, and for the con-tinuation of society. He referred to these practices as religious, whetherthey had recognizably religious beliefs accompanying them or not. The

24 Durkheim’s argument, in this regard, is similar to Wittgenstein’s point that it is theassumption that words have meaning through referential relations with things, whichhe referred to as an Augustinian picture language theory, that had caused the “prob-lem” of meaning as it had come to be known in philosophy. He argued that situatingmeaning within a social context as a function of “use” practices solved the problem.Philosophers have displayed a resistence to this argument that is similar to their treat-ment of Durkheim, attempting to incorporate both positions into a more conventionalphilosophical framework than either argument can tolerate.

Introduction 21

relationship between religious practices, mutual intelligibility and socialsolidarity explains the connection between Durkheim’s epistemologicalargument and his concern for questions of ethics, morals and social ordergenerally. Neither society, or stable relations between persons, could, heargued, continue to exist without such rituals. Human beings would notbecome rational beings in the first place, nor could they remain rational,without the continuation of society, which in turn depends on continuedenactment of the required rituals.

This meant that in modern pluralistic societies, wherein religion nolonger provides rituals that bind the whole group, secular rituals wouldhave to take the place of religious rituals in providing a foundation for rea-son and morality. It also meant, as Durkheim elaborated at great lengthin The Division of Labor, that those rituals could not be based on sharedbeliefs if they were to work in a modern and changing society.25 Further-more, a science of these social practices was necessary. Sociology wasintended to be such a science of practice. The model Durkheim used forsecular practice was laboratory science, the practices of which are heldin common by working groups of scientists even when they disagree ata theoretical level: practice driving theory, or belief, not the other wayround.

The implications of Durkheim’s theory of practice for moral philosophyare profound.26 As Durkheim argued in The Division of Labor, if peoplemust use practices to achieve social order and intelligibility in an advanceddivision of labor without the justification of beliefs, then they must beable to feel a personal commitment to practice. Under such conditionsequal opportunity to participate in practice and guarantees of personalfulfillment through participation will become a necessary prerequisite forsocial solidarity in such a society (Rawls 2003). If the society does notprovide for what Durkheim refers to as “justice” in this regard, then those

25 David Reisman’s discussion in The Lonely Crowd (1950) of the modern character whoseorientation is toward the others in situated practice rather than to inner conviction orgroup values, is one of the first attempts to come to grips with the effect of this changeon the person.

26 Communitarian moral philosophy bears some relationship to Durkheim’s argument,and some communitarians have made use of Durkheim’s argument in their work. Thedifficulty is that they tend to treat beliefs and ideas as the essential aspects of ritualthat must be preserved. Or, alternatively, to treat ritual as important, but for no reasonother than social bonding and the preservation of cultural identities. Durkheim’s focuson practice as the basis for “rational” personhood, in a more universal sense, seemsto have been lost. The real implication, I believe, is with regard to the viability of apractice conception of moral philosophy. Both Kant and Rawls, 1953, made compellingarguments based on a practice conception, but, the practice conception itself fell victimto problems with rules that emerged both with regard to Wittgenstein and game theoryin the mid 1950s. Durkheim offers the outline of a viable argument for the necessity ofpractices.


individuals and groups who feel excluded will fail to commit themselvesto the society. Fragmentation, alienation and anomie will result. Theseconditions, in turn, will undermine the stability necessary as a foundationfor contractual relations in an advanced division of labor: a fundamentalcontradiction. But, this contradiction is not the necessary consequenceof advancement in the division of labor.

If structural inequalities could be eliminated and justice therebyachieved all members of the society would have equal access to situatedpractices. Then at each next moment something like Goffman’s workingconsensus, or Garfinkel’s notion of trust, could operate as a guaranteebetween all participants in any given social practice. Even though peo-ple experience themselves as increasingly individualized, members of anadvanced division of labor under conditions of justice would neverthe-less be able to participate fully and commit themselves equally to eachnext situated practice in which they find themselves engaged. In doing sothey are required to trust one another to perform practices recognizablyand make sincere presentations of self. In exchange they make a com-mitment to try to the best of their ability to preserve the intelligibility ofthe situation and the self presentation of others. This reduces ambiguity,contingency and interpretation.

Thus, both contractual relations in a division of labor context andintelligibility at the level of enacted practices require justice. The first inthe form of a just social contract and the second in terms of a strongworking consensus that offers equal moral guarantees to all.

0.6.0 Order of Argument

In order to clarify the argument, and put various misreadings to rest, thisbook presents a systematic and close reading of Durkheim’s text. The aimis not to write “about” Durkheim’s text, but rather, to present that textas a series of detailed steps in an argument. This has required followingclosely the order of argument in the text, and the book is divided intochapters that take up issues in the text in the order in which Durkheimpresented them. The task, as I see it, is to find a way of reading thetext that makes sense of each of Durkheim’s arguments and leaves nocontradictions.27

27 There is a certain arrogance involved in reading a classic text and assuming that thereader is able to see contradictions that the writer overlooked. This may happen whenreading student papers, and even all too often with colleagues. But when a book hasbeen a classic for a century, it is a sure bet that the author has made no simple mindedmistakes. If the author appears to be stupid, my advice is that the reader has not yetgrasped the point of the text. Durkheim was not a stupid man. Therefore, it is a goodidea to read it again.

Introduction 23

The central chapters of The Elementary Forms present an extensiveand detailed argument for six categories of the understanding: time,space, classification, force, causality, and totality. Unfortunately, how-ever, Durkheim’s commitment to empirical detail results in long sectionson Totemism which are essential to his epistemological argument, butare so long and so apparently focused on Totemism per se, that theirepistemological significance is generally missed. For example, Durkheimtakes up Animism in order to argue against an a priori explanation oftotems. Similarly, he takes up the arguments of Frazer and Tylor in orderto argue against a classical empiricist explanation. The consideration ofindividual and sexual totems constitutes an argument that totems do nothave an individual origin.

In each case Durkheim argues against either empiricist individualism,or what he refers to as “apriorism.” Because he built his epistemologyon the details of the actual enactment of totemic rites, an argument thattreated those details as having an individual origin, or as existing a priori,would contradict his epistemological position. Therefore, the social originof totems is an issue critical to his epistemological argument, and hislengthy discussions of totems constitute necessary logical steps in hisoverall argument. However, his argument is only tangentially concernedwith Totemism.

Unfortunately, these long sections on Totemism not only have beentreated as a consideration of totems in their own right, but as explo-rations of “conceptual systems,” rather than as elaborations of the empir-ical details of enacted practice as an essential part of Durkheim’s episte-mology. The tendency to treat epistemology as fundamentally a matterof ideas tends to further obscure Durkheim’s point. As a consequence,most scholars have concentrated their attention on the first and last chap-ters of The Elementary Forms, where the epistemological argument is onlysketched (the presentation there taking a more familiar conceptual form),while ignoring the epistemological argument made in the central chap-ters. A curious misreading of the text has resulted.

The Introduction and Conclusion present their own difficulties. Theintroduction is an earlier elucidation of the argument, which appearedin 1909, as “Sociologie religieuse et theorie de la connaissance,”28 andleaves epistemological and sociology of knowledge issues relatively,although not entirely, undistinguished (Lukes, 1973:408). The conclu-sion presents another difficulty. It takes up the sociology of knowledgeafter the epistemological argument has been completed and assumes that

28 An earlier section on sociology and philosophy was omitted from The Elementary Forms,according to Lukes (Lukes, 1973:582).


the reader has an understanding of that argument. The Conclusion doesnot summarize the overall position. The epistemological argument itselfappears only in the central chapters, which have been generally ignoredin this regard, leading scholars to infer the epistemology from the Intro-duction and the Conclusion where it is only vaguely sketched. Thus, thecareful relationship worked out in the central chapters, between specificpractices and the corresponding categories of the understanding they pro-duce, is missed with the result that Durkheim is interpreted instead ashaving focused on symbolic belief systems.

Because the epistemological significance of the central chapters hasbeen missed, the epistemological argument has never been recognizedas such. Even the best treatments of Durkheim’s theory of knowledge,David Bloor (1982) and Jeffrey Alexander (1988) being notable in thisregard, treat it as an argument about the sociology of knowledge and not anargument, in the classical sense, about the nature of mind and the origin ofhuman reason. Yet, it is clear from a close reading of the central chaptersthat Durkheim articulated an epistemology in the classical sense. Thesociology of knowledge is a distinct secondary argument which is meantto rest on this foundation. In his introduction to the 1983 English editionof the lectures on Pragmatism (Pragmatism and Sociology), John Allcock(1983:xl) wrote that “a rounded assessment of Durkheim’s epistemologyhas yet to be undertaken” and there has been none since. Allcock alsopointed out that discussions of the nature of Durkheim’s social facts andhis sociology of knowledge have been the primary focus of Durkheimscholarship and have been confused with his epistemology.29

The aim of this book is to analyze closely the argument of the centralchapters of The Elementary Forms and set them in the general context ofthe book as a whole.

The first chapter considers Durkheim’s Introduction, and the argu-ments which he outlined there; in particular his statements regardingepistemology and religion. It also considers various misunderstandingsof that Introduction. The second chapter takes up Durkheim’s dualism,an argument that plays an important role in the Introduction, but whichis not really worked out there. Durkheim returned to this argument in1914 in “The Dualism of Human Nature and its Social Conditions,” inan attempt to clarify misunderstandings of The Elementary Forms whichled to the mis-identification of Durkheim as a Kantian, and also obscuredthe importance of the distinction between sacred and profane as a first

29 Unfortunately, while Allcock’s essay is one of the best in many regards, he offers thelectures on Pragmatism as the best source for the epistemology.

Introduction 25

classification. The pivotal position of the “sacred” in Durkheim’s argu-ment leaves his position without foundation when his epistemologicalinterest in the sacred is not understood, creating the false impressionthat the reason Durkheim focused on religious rites was because he wasinterested in religion and the sacred per se, and not in epistemology andethics.

Following the Introduction, Durkheim divided The Elementary Formsinto three books. Book I examines principle questions and defines reli-gion. Book II examines various theories about the origin of the principlereligious beliefs, and in the process describes hundreds of totemic ritesin minute detail. Book III focuses on religious rites and practices and themoral forces they create. This book approaches those issues as follows:Chapter Three considers Durkheim’s critique of animism and naturismand the origin of the sacred as he presented it in Book I. Chapter Fourconsiders his discussion of Totemism as it appears in the first five chap-ters of Book II. Chapter Five presents the arguments for moral force, souland personality, key factors in his epistemology, which Durkheim workedout in Chapters Six through Nine of Book II. Over the course of Book II,Durkheim took up, rather exhaustively, the range of accepted argumentsabout the origins of religious beliefs, dismissing them one after the other.By the end of Book II, he had completed his argument that the origin ofbeliefs and concepts cannot be explained without focusing on practicesand rites first. Rites and practices explain beliefs, not the reverse. It is acharacteristic of Durkheim’s style of writing that the central argument onwhich his own position rests is articulated last, in Book III, after havingdisposed of the competition, so to speak.

Chapter Six takes up issues raised in Chapters One and Two ofBook III, where Durkheim considered the primacy of rites over prac-tices. The relationship between specific types of rites and the categoriesthey produce are examined. For instance, rites of sacrifice give rise togeneral feelings of sacred and sacred time on which the experience ofother categories depend. Chapter Seven introduces the imitative ritesthat are the foundation of Durkheim’s argument for causality in the firsttwo sections of Chapter Three, Book III. Chapter Eight examines theargument for the empirical origins of the concept of causality as pre-sented by Durkheim in Chapters Three and Four, of Book III. Causalityis the most important concept for an epistemology, and if Durkheim canestablish it empirically, he can rest on that argument. This concludes thestep by step consideration of arguments presented in the central chapters.Chapter Nine considers the first three sections of Durkheim’s Conclusionwhere he discusses logic and language, and Chapter Ten considers thearguments of Section iv, which is the only part of the conclusion to discuss


the epistemology. These chapters are particularly important and difficult,because the conclusion received disproportionately more attention fromcritics over the years, and consequently, is responsible for more than itsshare of misunderstandings.30 The Conclusion is also important, becausein it Durkheim outlines arguments not made in the body of the text, andattempts to connect everything into one overarching sociological posi-tion. The relationship between his summary outline in the Conclusion,and the argument of the central chapters has not been well understood.

It might be asked why bother with a close reading of the text? What,after all, do highly abstract questions of epistemology have to do withsociology? Aside from pointing out that there is a problem if sociologyis built on an epistemological argument that it has not yet understood,and that there is no social order without mutual intelligibility, there areimportant reasons for revisiting Durkheim’s text. In an age of global-ization, as the division of labor advances around the world, many prob-lems are arising that Durkheim warned against. In particular the socialfunction of religious practices is being confused with religious beliefs.If persons, in an attempt to overcome the increasing contingency andinsecurity of modern life, turn to traditional religious communities thatexclude nonbelievers, increased fragmentation will result. Practices, onthe other hand, have the potential to be inclusive. It was Durkheim’s posi-tion that an international cult would have to be based on shared practicesthat do not discriminate between beliefs.31

30 There is a clear pattern in the prominent secondary sources of citing the introductionand conclusion while ignoring the central chapters. Three early critics writing in English,William Dennes, Charles Elmer Gehlke and Charles Schaub, seem to have shaped theway Durkheim’s text would be read for the rest of the century. A close look at theircitation patterns raises many issues: Schaub’s references to individual being are to theIntroduction except for one to the force section (on whether totems have an individualorigin). References to concepts are to the last Chapter. References to categories are to theIntroduction except the last two citations at the top of p. 321, which come from thesection on force and the Conclusion and lead into the idealist argument. As Schaubmakes his transition from a description of the categories to his characterization of themas idealist, he switches his citations from the Introduction to the Conclusion. Gehlke’sarticle quotes the “Individual and Collective Representations” twenty times in the firsttwenty-five pages and fifteen times in the next twenty-five. He references The ElementaryForms not once in the first twenty-five pages and only sixteen times in the next twenty-five. This out of a total of 165 references to Durkheim’s works appearing in the first fiftypages of his manuscript. To compound this, Dennes proceeds by paraphrasing Gehlkeparaphrasing Durkheim. Schaub, while making fewer references to the “Individual andCollective Representations,” and consequently doing a marginally better job of graspingessential points of Durkheim’s epistemology, nevertheless makes a crucial reference tothe “Individual and Collective Representations” that appears to be a combination of twopassages quoted by Gehlke, again suggesting that Gehlke, not Durkheim, is Schaub’ssource for the quotation. See Rawls 1997 for an elaboration of this citation pattern andits consequences.

31 See The Division of Labor, Book III Chapters one and Two and Rawls 2003.

Introduction 27

There are also, I believe, good reasons to return to a serious readingof classical sociological texts in general. While it is often argued thatcontemporary social theory has gone beyond the classics, I think theclassics may still, in fact, be ahead of us. Certainly sociologists todayknow more, that is, command a greater number of facts. But, because ofserious misinterpretations of classical texts, combined with an increasingspecialization in education that leaves social theory essentially untaught,as sociologists we are losing sight of the big picture.

The serious study of practice requires detailed study of practices asthey are being enacted in natural settings. Durkheim’s argument pre-sumes that religious rituals have purposes and effects that are evident intheir details. The dominant assumption of contemporary sociology, thatorder is not evident in any single case and thus must be aggregated acrossa great number of cases, contradicts Durkheim’s assumption. System-atic studies of local orders of practice require qualitative observationalresearch coupled with analysis that does not rely on coding data. Codingprocedures lose the detail that the study of practice requires. This typeof detailed research is usually seen as departing from classical sociology.But, in fact, it is the only type of contemporary research that is consistentwith the parameters for the study of society that Durkheim laid out.

Classical social theorists were involved in a grand project. NotParsons’ grand project of respecifying social order mathematically, buta rather grander project of reinventing philosophy as a social science.Contemporary scholars seem to have concluded, on the basis of philo-sophical criticisms, and the apparent impossibility of accommodatingmultiple perspectives within a comprehensive moral viewpoint, that thegrand project, referred to increasingly as theory with a capital “T,” isimpossible, and as a consequence, have tended to retreat to theories ofwhat has come to be called, the middle ground. But, classical theorists hadin common a rejection of those same philosophical criticisms from whichwe now retreat. They all engaged in the serious project of arguing thatthe study of social processes and relationships fundamentally recast thetraditional questions and overcame their limitations. Careful reconsider-ation of the theoretical issues that the classical social theorists confrontedat the outset, and in particular the problematic relationship between soci-ological and philosophical ideas, needs to be undertaken. Without sucha reappraisal we will most likely continue to reproduce a discipline that isin almost every way just what the classical theorists argued against. And,we will continue to invoke their names as justification for doing so.

1 Durkheim’s Outline of the Argument inthe Introductory Chapter

Durkheim makes large claims with regard to his epistemological argu-ment in the introductory chapter of The Elementary Forms. He promisesto provide an empirical demonstration of an ever present and universalreality. That reality, according to Durkheim, is not religion, or any par-ticular cultural or social form, but rather “man,” (l’homme) and thatwhich is necessary in order for persons to become recognizably human.Durkheim is referring to nothing less than an empirical proof of the socialcauses of the development of human reason. The claims are large and thescope of the argument outlined impressive. Nevertheless, the introduc-tory chapter of The Elementary Forms is significant, not so much for whatit says about Durkheim’s epistemology, as for what it leaves unsaid.

The epistemology is announced, but not worked out, in the Introduc-tion. The emphasis there is placed on reviewing opposing arguments, andjustifying the choice of Archaic religion as a focus. In the central chapters,by contrast, Durkheim’s epistemology is worked out in detail. As a con-sequence, the Introduction and the central chapters present Durkheim’sargument in ways that appear to conflict on essential points; the “talkabout” the epistemology in the Introduction, and the “demonstrationof” the epistemology in the text, differing in substance. In addition, con-tradictory statements about the empirical validity of various ideas seemto appear one after the other in the Introduction. The argument of thecentral chapters, on the other hand, is careful, methodical and consistent.

Furthermore, while many of the apparent contradictions in the Intro-duction may be the result of trying to talk about the epistemology beforethe descriptions of the moral force of social practice have been presentedin the body of the text, it is not at all clear that the epistemological argu-ment, as it appears in the central chapters, is actually discussed in theIntroduction. What Durkheim believed he could successfully argue formay have undergone a transformation over the course of completing thebook. Whether it is merely a problem of talking about the epistemology inadvance of the demonstrations in the text, or whether he actually meanttwo different things, Durkheim’s claims in the Introduction concerning

28

Durkheim’s Outline of the Argument 29

the importance of his argument, seem out of proportion with the epis-temology as presented in the Introduction. Those claims are, however,substantiated by the central chapters of the text.

In light of these problems, the fact that most commentators seem tohave relied disproportionately on the Introduction and Conclusion, ratherthan the central chapters, for an understanding of Durkheim’s epistemo-logical argument, has been consequential. Not only is the epistemologynot worked out in the Introduction in any detail, but, there are, in fact,several features of the Introduction itself that serve to obscure the epis-temology. First, Durkheim seems to make the claim in the Introductionthat the text will explain the origins of all logical ideas. But the episte-mology, as presented in the body of the text, only explains the originsof six categories of the understanding and the idea of the sacred.1 It isonly in the Conclusion that Durkheim returns to the question of logicin general (aside from cursory mentions), and there only in summaryform; Second, in the Introduction, Durkheim makes claims with regardto both an epistemology, for which he claims universality and necessity,and a sociology of knowledge, in which the validity of ideas is culturallyrelative. However, he does not clearly distinguish between the two sortsof argument. Third, Durkheim opens his argument with a long discus-sion of religion. This creates the impression that his focus is on religion,when actually religion is of interest to Durkheim primarily because, as thefirst arena within which enacted practices created and sustained moralforce, religion is synonymous with the creation of human reason, and theresulting experience of dualism.2

Durkheim treats the question of the origin of the idea of the sacred, thefirst expression of moral force, as the key to a viable empirically based epis-temological argument.3 Consequently, he devotes the entire first section

1 Durkheim may have been thinking of personality as a type of category when he wrotethe Introduction where he lists personality as one of Aristotle’s categories. In Book II,Chapter Eight, Section vi, when Durkheim discusses the idea of the person, he identifiesthe idea of the person, or personality, with the sacred, or soul, as the result of moral force,and therefore possibly as a category. However, he does not refer to it as a category in thatcontext, and does not describe its genesis in the experience of moral force the way he doesfor the other categories. He also does not include it in the list of categories elaborated inthe Conclusion. In so far as what Durkheim means by personality is the first idea of thesacred, or soul, then personality would be the first product of moral force. It might besaid to be an expression of moral force in general and not a particular category. In so faras he refers to the particular person, Durkheim seems to mean something like “rationalbeing” by personality, or personhood. Again, in this instance, personality would not be acategory, but rather, reason itself. For an extended discussion of this issue see 5.3.0.

2 In fact, in the Conclusion it becomes quite clear that Durkheim sees little differencebetween secular celebrations, science and religion.

3 See my Chapter Three for a discussion of the relationship between the idea of the soul,the first dualism between sacred and profane, and the development of human reason.


of the Introduction, and part of the second section, to a discussion ofreligion. His pursuit of the origin of the sacred will also dominate Book I.For Durkheim, the sacred, as the first idea of moral force, is the keyto epistemology. What Durkheim is interested in are practices that aremutually shared and collectively enacted, to a degree sufficient to explainthe creation of collective feelings of sacred and profane and the categoriesof reason they give rise to. He is not interested in religion per se. He doesexplain this in the Introduction. But the long justifications of his approachto religion ironically tend to further obscure the point.4

Further confusing the issue, in the Introduction, Durkheim treats log-ical concepts, like contradiction, as though they were included in his listof categories of the understanding; those ideas for which he will provideempirical evidence of their origin in religious practice. In so far as hemeans to refer to the distinction between sacred and profane, he willindeed show a social origin for this “contradiction.” But, that is not thesame thing as arguing that the idea of contradiction itself has an originin social practice, which he does not do. In fact, he seems to includecontradiction with resemblance and other basic animal abilities to think.Believing that Durkheim has argued for an empirical origin in religionof even the basic ability to discern left from right, commentators havecriticized his argument for being impossibly naive (Dennes 1924). Thebasic ability to discern left from right is necessary for survival and would,therefore, have to precede the development of religion (see the discussionof dualism in Chapter Two). It is certainly required prior to the ability toexperience the moral force of left and right, or sacred and profane space.This criticism has led to the assumption that Durkheim is either assum-ing an innate rationalism, or that his argument is circular and thereforeuntenable. The criticism of the argument as circular has been particularlyinfluential.

However, Durkheim never even attempts to make an argument for thesocial origin of the idea of contradiction in the body of the text. The cat-egories for which he actually provides evidence of an empirical origin inreligious practice are not the basic logical processes implied by the Intro-duction. Durkheim introduces what he refers to as a “dualism” of humannature that allows for basic logical abilities, that are not empirically valid,to precede the development of empirically valid categories of the under-standing. These distinguish the basic animal from the socially developed

4 It is less clear in a contemporary context what criticisms Durkheim’s justifications weredesigned to anticipate. But, if we realize that he was speaking to a very “Christian” orientedwestern world and saying that religion was about society, we realize that he had reason toworry and that probably much of the disgust with which the book was received was dueto this position on religion. Of course, he was also defending the equality of aboriginalsin Europe in 1912, a time when Blacks and Jews were hardly considered human by theso-called “master races.”


human. The ability to discern left from right is one of the abilities thatDurkheim clearly allows is prior to the development of human reason.In the Conclusion he says that the individual animal already has some ofthese abilities, but that they are not useful in a social context, becausethey are in an individual form and therefore strictly incommunicable.5

Thus, Durkheim does not in fact make the mistake that he appears tosay that he will make. This issue will be discussed further in my ChapterTwo, where Durkheim’s dualism is considered in detail.

Because of these and other difficulties with Durkheim’s Introduction,it is essential to clearly delineate the various arguments presented thereand to read them in the context of the methodical step by step empiricalpresentation of the epistemology in the body of the text, without whichthe argument makes no sense. A distinction between categories of theunderstanding and socially transmitted ideas in general emerges inthe body of the text. The empirical demonstration of the categories inthe body of the text also clearly departs from the confusion of categorieswith logic in general that threatens the coherence of the Introduction.

The fact that the introductory chapter was written several years beforethe body of the text may be a contributing factor. First published in1909 as “Sociologie Religieuse et Theorie de la Connaissance,”6 whichis its subtitle in The Elementary Forms, it appears to have been completedbefore Durkheim had worked out the empirical arguments of the centralchapters in their final form. The lack of a distinction between the soci-ology of knowledge and the epistemology in the Introduction, and alsobetween the categories of the understanding and logic in general, sug-gests that Durkheim’s thinking about the mechanics of his proof had notprogressed much beyond the argument of Primitive Classification (1901),when he first wrote the introductory chapter in 1909.

In Primitive Classification, Durkheim and his co-author Marcel Mauss,had pointed out relationships between many systems of conceptual classi-fication and religious beliefs and practices among indigenous Australianpeoples. However, that work only hinted that in some unspecified waythe resulting concepts were transmitted from the group to individualpersons, or between individuals. Because of this, the argument of PrimitiveClassification suggests that what Durkheim meant by classifications weresets of ideas in a sociology of knowledge, and not cases of the social gen-eration of the category of classification. It is only in The Elementary Formsthat the epistemological argument with regard to classification emerges.There are hints of an emerging epistemology in Primitive Classification, in

5 It is clear from Durkheim’s discussion that what the individual has as basic animal abilitiesare empirical generalizations, such as Hume argued for, rather than innate reason as somecommentators (e.g. Lukes and Schmaus) have assumed.

6 The English translation is, “Religious sociology and the theory of knowledge.”


the importance accorded to the experience of the sacred, but Durkheimleft the philosophical issues unspecified. Unfortunately, because Primi-tive Classification is a simpler text, it has often been substituted for TheElementary Forms in discussions of Durkheim’s epistemology.

Durkheim’s epistemology challenges basic philosophical assumptionsregarding the empirical validity of certain essential ideas. Therefore, hiscomments on the philosophical debate over epistemology, in the Intro-duction, are an important indication of what he believes to be the sig-nificance of his own epistemological argument. It is important for soci-ological readers with no background in philosophy to understand howDurkheim’s argument for the validity of socio-empirical experience fitsinto the context of philosophical debate, so that the care with whichDurkheim built his argument, and the relevance of that argument to bothsociology and philosophy can be made clear. In this regard, Durkheim’scomments and criticisms regarding philosophical debates over epistemol-ogy will be inspected in some detail. Only against a clear backdrop of thephilosophical discussion can the steps in Durkheim’s argument be under-stood and his claim that archaic religions provide an empirical founda-tion for the origin of the categories of the understanding be meaningfullyassessed.

Because of the necessary tightness of epistemological argumentation,loose interpretations of such arguments tend, by their very nature, to bemisleading. Is it Durkheim’s argument, or the commentators who areguilty of “looseness”? I believe the truth is that, apart from the Introduc-tion, Durkheim’s argument was so tight, and followed a classical form ofargument so closely that, in the context of the apparent contradictionsin the Introduction itself, it was missed altogether. However, the Intro-duction, read in the context of the body of the text, as an announcementof the main argument, rather than as the argument itself, ceases to be anobstacle to understanding that text.

Durkheim’s Introduction is divided into two sections of two and threeparts respectively. There are five main issues taken up there that will beconsidered in this chapter; two in Section i, involving the relevance ofreligious studies to sociology and philosophy, and three in Section ii,where Durkheim’s own epistemological argument is introduced. InSection i, Durkheim First, explains why archaic religion is relevant tounderstanding the religious essence of human reason; and, Second, arguesthat religious studies have been faulted by an individualistic bias. InSection ii, Durkheim First argues that religion bears a special relation-ship to epistemology, and that, if looked at socially, provides a newway of addressing epistemological questions; Second, characterizes the


epistemological debate of his day between empiricism and apriorism, set-ting out his arguments against philosophical and religious individualism;and, Third, outlines his own position: to establish the “true” relationshipbetween religion and epistemology and show that the categories of theunderstanding have an empirical origin in those moral forces enactedby religious practice. A secondary issue involving two arguments for“dualism,” and their significance for Durkheim scholarship, will be takenup in my Chapter Two.

In an effort to reconstruct Durkheim’s epistemology without takingthe steps of the argument out of context, the organization of this chap-ter follows, as closely as possible without repetition, the organizationof Durkheim’s introductory chapter. Inconsistencies between the Intro-duction and chapters in the body of the text will be addressed, andDurkheim’s supporting arguments, consisting of detailed discussions ofenacted practice, which appear in great detail in the body of the text, willbe reconstructed, point by point, in later chapters.

1.1.0 Section i: Consideration of Religion

In Section i of the Introduction Durkheim lays the foundation for his argu-ment that collective religious practice is the origin, or cause, of humanreason. In order to demonstrate the relationship between religious prac-tice and reason Durkheim proposes to examine an archaic religion thatstill fulfills its underlying social function without unnecessary elaboration.Because he understands ([1912:1]1915:13; *1995:1)7 that this choiceof starting point will most probably be viewed by the critics as strange(etrange), and as he says ([1912:5]1915:15–16; 1995:3–4), the startingpoint for an argument largely determines its outcome, he promises to paygreat attention, in the pages and chapters ahead, to a justification of hischoice.

7 Each time a reference is made to Durkheim’s text page numbers are given for three publi-cation dates to indicate where that passage appears in both complete English translationsand in the original French text. I have once or twice also referenced a new abridged editionin English. An asterisk is placed before the date of the text from which the citation hasactually been taken in each case. This is necessary for several reasons, there are signifi-cant differences between the translations. Both English translations are widely used. It isalways best that readers be able to check a reading against the original French, rather thanrelying on English translation. The page numbers differ significantly making it difficult tofind the corresponding passages if the page numbers are not supplied. I have tried in eachcase to use the English translation that I felt captured the original French text best andthe reader will see that I have switched back and forth between texts with some frequency.I have sometimes used the original French in addition to English and I have also severaltimes offered an alternative translation where I felt there was misunderstanding. The ideais to stay as close to the original text as possible.


In addressing these issues, in Section i, Durkheim makes two mainpoints. First, in justifying his treatment of Australian Totemism as reli-gion, he argues that all religions have certain fundamental properties andcauses in common. Causes are in this case the necessary prerequisites forbecoming human that religions fulfill.8 Therefore, looking at the sim-plest set of religious practices, he argues, is the best way of finding thesecauses without the interference of elaborate belief systems; and, Second,he argues that the study of religion has been obscured by approachesthat are, on the one hand, individualistic, and, on the other, treat beliefsystems themselves as the purpose of religion.

According to Durkheim, the prevalence of studies that speak disparag-ingly of primitive religions, and claim that they are not really religionsbecause they do not have a belief in a deity, is the result of an indi-vidualist approach to religion that focuses on faith or beliefs. Only whenapproached as a collective social phenomenon, with an emphasis on prac-tices, does Durkheim believe that religion can be properly understood.Durkheim defends ([1912:2–3]1915:14; *1995:2) his decision to treat anarchaic religion as an equal with modern religions, arguing that seen insocial terms, “all religions are true. All fulfill given conditions of humanexistence.” He argues that if archaic religious practices fulfill the functionand purpose that religions fulfill, then no matter what form they take, theyare religions.

1.1.1 Relevance of Archaic Religions

The First issue Durkheim takes up in Section i of the Introduction is thequestion of why he has chosen to examine an archaic religion insteadof a modern religion. He explains that the study of “archaic” religiouspractice is better suited to revealing the everpresent causes of “a presentreality” ([1912:1]1915:13; *1995:1) with which he is primarily con-cerned. According to Durkheim, that “reality is man” ([1912:1]1915:13;*1995:1). He does not say that the reality is religion, and it will becomeapparent over the course of the Introduction that his interest in religionis secondary to his interest in human reason. It is only because of therole that Durkheim believes religion, and more specifically, enacted reli-gious practice, plays in the development of human reason, that he hasexplored it.

In laying out his epistemology, Durkheim explains, he focused onarchaic religious beliefs and practices, for methodological reasons. He

8 Durkheim often uses the word cause, or causes (also causes in French) to mean anunderlying need. It often means something more like survival of the fittest: an underlyingfunction or need that explains why a social form persists over time, than direct causation.


felt that religious rituals in, what he called, “simple” societies wereshared to a degree sufficient to support the formation of general cat-egories of thought, without the distraction of elaborate accompanyingbelief systems. This focus on practices over beliefs will emerge over thecourse of the text as essential to Durkheim’s position. The argument thatDurkheim makes at this point treats the causes, or functions, of religionas the essence of what religion is really all about, treating religious beliefsas merely secondary phenomena. Archaic religions more easily reveal thisessential distinction because they are simpler in two ways. They are rela-tively free of narrative elaboration and therefore, do little more than fulfilltheir underlying functions. Furthermore, they are shared to a degree thatmakes it possible to study the effects of religion on a whole society atonce, without having to take individual differences into account.

Although the same relationship between practice and reason is ulti-mately to be found in both modern and archaic societies, modern society,Durkheim says ([1912:7]1915:17; *1995:5), does not exhibit the samedegree of homogeneity: “Neither religious thinking nor religious practiceis shared equally among the mass of the faithful. The beliefs as well asthe rites are taken different ways, depending on men, milieux, and cir-cumstances . . . Under such conditions, it is difficult to perceive whatmight be common to all.” In, what Durkheim called, “simple” societies,on the other hand, almost everything people did was part of religion. Thisall pervasiveness of religious belief and practice in traditional communi-ties, according to Durkheim, makes archaic religions better suited thanmodern religion for the study of the relationship between ritual practiceand the development of human reason. In modern culture, only formalreligions, divorced from everyday life, remain.

The word religion, as Durkheim uses it in The Elementary Forms, shouldnot be understood in this modern sense. By religion Durkheim meant notreligious belief, or formal institutions of religion, but the all pervasivenessof religious practice, and its penetration into all aspects of daily life, intraditional societies. He argues that some aspects of traditional religiouspractice must be preserved by modern religions.9 But, many aspects ofmodern religion, generally thought of as essential to the idea of religion,are not included in Durkheim’s definition of religion. Although it does not

9 Given Durkheim’s argument, there are three ways of explaining existence of the categoriesof the understanding in modern life: First, that formal religions and their practices are stillsufficient to produce the categories; Second, that the categories are no longer produced,or are insufficiently produced, and the meaning of the fundamental concepts is onlyconventional; or, Third, that other routines of daily life have taken over the function ofreproducing the moral forces in modern society that religious rituals used to enact inarchaic societies. Durkheim seems to opt for a combination of the three but with the thirdbecoming predominant in modern life.


appear until the next chapter, Durkheim’s ([1912:65]*1915:62; 1995:44)definition of religion is relevant here: “A religion is a unified system ofbeliefs and practices relative to sacred things, that is to say, things setapart and forbidden-beliefs and practices which unite into one singlemoral community called a church, all those who adhere to them.”

Durkheim referred to the resulting “single moral community” as achurch regardless of whether or not there was an identifiable meetingplace, or corresponding institutional organization. His definition of a“church,” and his particular emphasis on practices, accommodates thesocial relationships inherent in a traditional community, not the insti-tutional organization and dogma of a modern church. According toDurkheim ([1912:60]*1915:59; 1995:41) “A society whose members areunited by the fact that they think in the same way in regard to the sacredworld and its relations with the profane world, and by the fact that theytranslate these common ideas into common practices, is what is called achurch.”10

In justifying his focus on archaic religious practices, Durkheim([1912:7]1915:18; *1995:5) says that in archaic societies “The group reg-ularly produces an intellectual and moral uniformity of which we knowonly rare examples” in modern society. In other words, archaic religionsconform more closely with Durkheim’s definition of a “church” thanmodern religions that are located in actual church buildings. Persons inmodern society must have experiences of moral force that generate thesame categories of the understanding, according to Durkheim. But thoseexperiences do not generally occur in contexts of belief and/or practicethat are the same for everyone.

In archaic society, on the other hand, the moral experience of the cat-egories by individuals is more likely to be identical for all members of thegroup. According to Durkheim ([1912:8]1915:18; *1995:5), the practiceof religious rites is more equally distributed: “Everything is common toeveryone. The movements are stereotyped; everyone executes the sameones in the same circumstances; and this conformity of conduct merelytranslates that of thought.” In modern society, by contrast, there is a divi-sion of ritual labor, resulting in a differentiation of experience. This, plusthe development of elaborate belief systems, makes it difficult to studythe relationship between enacted social practice and the development ofcategories of the understanding in a modern context. In Archaic religionsthe relationship between practices and underlying social function is mucheasier to see. It is Durkheim’s focus on practices and their effects, over

10 While the language here suggests that beliefs precede practices, Durkheim will make itclear that in the first instance practices must come first.


beliefs, and social function over individualism, that largely determinesthis selection.

Durkheim also emphasizes the relationship between identical soundsand movements and the development of reason. His argument in thisregard is essential to an understanding of his position, because it placesthe emphasis, in explaining the origin of the categories, on the experi-ence of enacting sounds and movements in common, not on learningwords, or mastering systems of belief. This focus on shared sounds andmovements is a huge departure from traditional epistemological argu-ments that focused on individual experience in attempting to explainthe validity of general ideas. More importantly, it renders the argumentempirical, as Durkheim claimed, and not idealist, as it is generally inter-preted. Durkheim’s “theory of practice,” as I call it, will be explored ingreater detail in my Chapter Six.

Finally, there is an emphasis on underlying causes, needs, and in a spe-cial sense, functions, that begins to emerge in the first part of the Intro-duction as part of the justification for taking archaic religious practicesseriously as religion. This functional argument will also play a significantrole in Durkheim’s critique of individualism, in Section ii of the Intro-duction. Archaic religions he says ([1912:9]1915:19; *1995:6, emphasisadded), aid in the explanation of religion because they more easily revealtheir underlying purpose: “Because the facts are simpler, the relationsbetween them are more apparent. The reasons men invoke to explaintheir actions to themselves have not yet been refined and revamped bysophisticated thought: They are closer and more akin to the motives thatcaused those actions.”

Durkheim believed that there are real causes of, and needs for, reli-gious phenomena that can be discovered. These causes address neces-sary social functions. It is these underlying causes that Durkheim eluci-dates in his epistemology. Because the collective representations that arepassed down over the course of history distort the view of the underlying factsand necessities that caused the development of the practices and obscurethis underlying relationship, a primitive religion is most likely to revealits underlying causes. Durkheim argues ([1912:10]1915:19–20; *1995:7,emphasis added) that “As [religious thought] progresses historically, thecauses that called it into existence, though still at work, are seen no moreexcept through a vast system of distorting interpretations.”

It is important to note the emphasis on causes, as opposed to interpreta-tions. Durkheim’s epistemology focuses on empirical relations of practice,not on relations between ideas or beliefs (the sociology of knowledge).While The Elementary Forms has consistently been interpreted as idealist(elaborating a sociology of knowledge of relations between ideas), there


is strong talk throughout the book of underlying needs and causes, andof socially occasioned empirical happenings that fulfill those causes andneeds. Durkheim speaks of empirically observable religious practices as“truths” related to satisfying these underlying social needs.

Durkheim knew that some people would think it was an error to treatarchaic and modern religions as if they bear the same measure of truth.In his own defense, he argued ([1912:3]1915:14; *1995:2) that “it is afundamental postulate of sociology that a human institution cannot restupon error and falsehood.” If there are practices of longstanding in anysociety, then there must be an underlying need for those social practices,according to Durkheim.11 Religious practices are “true” in so far as theirpurpose is to fulfill an underlying need and they do in fact fulfill thatneed.

This grounding in social needs, provides, according to Durkheim([1912:3]1915:14; *1995:2, emphasis added), a guarantee that a socialinstitution that persists over a great length of time is “true” in his spe-cial sense that it fulfills some real social necessity: “Therefore, when Iapproach the study of primitive religions, it is with the certainty that theyare grounded in and express the real . . . What I criticize in the schools Ipart company with is precisely that they have failed to recognize it.” Onthe basis of this argument, Durkheim will defend aboriginal intelligence,and the legitimacy of aboriginal religious practice, throughout the book.

Durkheim argues that the existence of any society depends on thedevelopment of a body of ideas that are sufficiently shared to facilitatecommunication and group unity. This is a fundamental social need thatlies beneath religious practice and dictates that the categories must, andtherefore will, be generated. Furthermore, Durkheim assumes that thisis an everpresent human need. If Durkheim can make the argument thatthe purpose of religion is to fulfill this need, then he will also have shownthat the purpose of archaic religions is the same as the purpose of modernreligions.12 The purpose of both types of religion, their function, so tospeak, is to provide the experiences of moral forces that persons need inorder to develop the categories of the understanding.

At the end of the Introduction, and later in the body of the text,Durkheim explains this underlying necessity that acts as the “cause” ofreligious practice. He argues that all peoples must have some occasions for

11 In fact, he will say that other longstanding practices and narratives, like the philosophicalposition on dualism, have this same sort of truth. See my Chapter Two.

12 Of course, in so far as what produces unity of thought and feeling in a society may havechanged over time, it may in fact turn out that modern religions are less true than archaicreligions, because they no longer serve, as well, the purpose of creating unity of thoughtand feeling. In modern society “Interaction Orders” in every-day-life, and professionaland civic associations, may have taken over this function to a considerable degree.


assembling so as to produce moral force, or collective effervescence. ForDurkheim, the possibility of shared knowledge and mutual intelligibilitydepends entirely upon the collective enactment of those shared practiceswhich produce moral force and through the experience of moral force thecategories of the understanding. Without participation in ritual assem-blies, and the performance of ritual practices, the categories would notbe presented to experience, and all contact between minds would be lost.According to Durkheim ([1912:23]*1915:30; 1995:16), “If men did notagree upon these essential ideas at every moment, if they did not have thesame conception of time, space, cause, number, etc., all contact betweentheir minds would be impossible, and with that all life together.”13

If individual consciousness were left to itself, words would only be signsof internal states and communication would be impossible. Durkheimreiterates this point later in the text during his discussion of classificationin Book II, Chapters Six and Seven. At that point, Durkheim introduces([1912:278–332]*1915:223–264; 1995:197–233, emphasis added) hisown view of the significance of symbols in the performance, and memory,of ritual practice:

In fact, if left to themselves, individual consciousnesses are closed to each other;they can communicate only by means of signs which express their internal states.If the communication established between them is to become real communica-tion, that is to say, a fusion of all particular sentiments into one common sen-timent, the signs expressing them must themselves be fused into one single andunique resultant. It is the appearance of this that informs individuals that they are inharmony and makes them conscious of their moral unity. It is by uttering the same cry,pronouncing the same word, or performing the same gesture in regard to some object thatthey become and feel themselves to be in unison.

Durkheim’s argument is that “real” communication requires, what G. H.Mead ([1924]1934) called a significant symbol; one which calls up thesame response in both the speaker and the hearer. The sign that can callup a common sentiment is necessary for a genuine meeting of minds.In Archaic religions, according to Durkheim, the totem is the unitary“cry” which gives the experience of moral unity and creates the commonsentiment. It is able to do this because on prior occasions it has beenthe symbol in the name of which moral force was enacted and sharedemotion experienced. Practices enacting the moral character of the socialworld, and the symbols which solidify and maintain those practices, likethe totem, when enacted in social assembly, create the categories of the

13 This raises a whole other question as to whether arbitrary naming works outside of acontext of shared practices which could support a high degree of nominal expectations.It also throws into question poststructural assumptions about meaning. It would seemthat a high degree of social order would be required to support nominal categories, whichwould explain why poststructuralist studies often focus on coercion.


understanding. These categories are not conventional symbols havingmeaning by consensus, or agreement. If they were, they could never callup the same response in all parties and they would not have empiricalvalidity.

Durkheim treats symbols as a surface phenomena beneath which lieconcrete social relations. It is not a referential relationship, but a causalone. The concrete social relations, that is, the practices, cause the feelingsthat are “called up” by the totemic symbol. Other symbols lie at a muchgreater distance from their causes. In speaking of symbols, Durkheimsays ([1912:3]1915:14; *1995:2, emphasis added): “We must know howto reach beneath the symbol to grasp the reality [and here he means socialreality] it represents and that gives the symbol its true meaning.” Accordingto Durkheim, all symbols are ultimately caused by social relations. But,only the categories, and the totemic emblem, retain the common sen-timent. Other symbols, and collective representations have, over time,become merely interpretations, and as such grow further from their causesover time. Even the categories themselves, would become distanced fromtheir causes over time, if they were not constantly reenacted. There-fore, because of the potential distance between symbols and practices,Durkheim’s interest in religious symbols and practice, with regard to theepistemology, is not in the symbols per se, but in their relationship to anunderlying reality enacted by the rites.

Durkheim’s focus on rites, in his justification of his choice of archaicreligion, is really the key to his epistemological argument. When jus-tifying his treatment of archaic religious practice as legitimate religion,Durkheim makes an important distinction between the social need for rites,and their justification in beliefs. Persons need to collectively enact rites inorder to become rational beings. They need to collectively enact rites inorder for there to be anything to create beliefs about. They need beliefs,on the other hand, in order to give narrative meaning to the rites whichthey must perform, so that they will continue to perform them.

According to Durkheim ([1912:3]1915:14; *1995:2, emphasis added)there are “true reasons” for the performance of rites, which are entirelyseparate from religious beliefs:

The most bizarre or barbarous rites and the strangest myths translate some humanneed and some aspect of life, whether social or individual. The reasons the faithfulsettle for in justifying those rites and myths may be mistaken, and most often are,but the true reasons exist nonetheless, and it is the business of science to uncoverthem.

Myths and practices are both hearable and observable, told anddone, in the assembled group. As enacted practices, they have functions


and “true reasons” separate from what individuals believe about them.These social functions constitute the “truth” about religious practice,for Durkheim, and can be studied empirically by focusing on the enact-ment of religious rites. Belief, or “reasons,” when detached from enactedpractices, are less tangible and often more personal or conventional (i.e.“true” only by consensus). This distinction between practice and beliefis an important key to Durkheim’s argument that all religions address thesame human needs, and will be taken up in detail in my Chapter Six. Itis a distinction that, in fact, organizes the text, as Durkheim examinesbeliefs in Book II and rites in Book III.

1.1.2 Religion: Individual or Collective in Origin

The Second issue that Durkheim examines in Section i of the Intro-duction is the question of whether religion is an individual or a socialphenomenon. This question is critical to his argument that the social,or shared, aspects of religion cause the development of reason and isaddressed extensively in the central chapters. Religion understood as amatter of individual belief, or revealed faith will not support the empiri-cal argument for the development of the categories in collective enactedpractice that Durkheim makes in the body of the text. Individual rev-elation is particular to a given individual, and not communicable, andas such, religious practice could not provide a social origin for the cate-gories. It would remain within the limitations of empiricist individualism(discussed in Section ii).

Durkheim begins this argument by criticizing the individualism inher-ent in the philosophical and anthropological approaches to religion. Hesays ([1912:6]1915:16; *1995:4) that: “All they do is analyze the idea theyhave of religion.” This dialectical, or introspective, method is necessar-ily individualistic. According to Durkheim, an individualistic approachobscures the essentially social nature of religious phenomena, makingreligion look like an aggregate of individual ideas, when it is, in fact, notjust social in form, but, Durkheim argues, exists only to fulfill underlyingsocial needs.

Durkheim argues that the tendency to take an individualist approach,because it obscures the underlying social needs that religion addresses,makes archaic religions, which address those social needs in the sim-plest and most direct manner, appear to be illogical. In other words, ifone treats archaic religions in terms of their belief systems, instead of interms of the underlying needs which their practices address, they seemabsurd. Archaic religions are so pure in fulfilling only social functions,


that beliefs are essentially irrelevant to understanding them.14 Durkheimargues that only an approach to explaining the origin of human reasonthat is based in social experience can reveal the meaning and coherenceof archaic religions. In the body of The Elementary Forms, he demon-strates at length, that popular arguments concerning the lack of validity,and the superstitious origins, of primitive religious ideas; arguments thattend to treat these religious belief systems as naive and even silly, are theresult of the individualist assumptions made by empiricist and aprioristphilosophers and anthropologists.

Durkheim consistently defends archaic religious practice from variousdetractors throughout the text. His argument that all social practices thatsurvive the test of time are equally true, because they fulfill the samepurpose, is brought into play in this regard. This is interesting, becauseDurkheim’s so-called “functional” approach is generally referred to as“conservative.” Yet, here we find it as the basis for both his defense ofthe legitimacy of “primitive” religious systems, and his insistence that anindividualist approach, while it may be the accepted form of thought inwestern society, completely obscures the philosophical issues and socialphenomena that the approach claims to clarify. Social functionalism, inthis case, is best seen as opposed to individualism, rather than to otherversions of social explanation. Both Durkheim’s assertion of the legit-imacy and cogency of primitive religions, and his assertion that philo-sophical individualism results in obscuring the essential nature of societyand human knowledge, pose serious challenges to accepted religious,political, and philosophical views. This is not a conservative position.

It is, according to Durkheim, not possible to develop an accurate under-standing of archaic religion, or of any other social phenomena for thatmatter, from a perspective of individualism. This is because the beliefs,the only part of religion available from an individual standpoint, areessentially irrelevant. Archaic religion is essentially the enactment of basicsocial functions that are necessary for the development of shared ideas.Religious beliefs, or accounts arise as retrospective accounts for the prac-tices, which, in turn, develop in response to underlying social needs.These accounts do not provide access to the underlying reasons for thepractices.15 Durkheim argues ([1912:27]1915:31–33; *1995:18) that “toknow what the conceptions that we ourselves have not made are made of,

14 The idea that religion is about beliefs is for Durkheim a western, division of labor, postreformation conceit.

15 Over the course of the text, Durkheim will develop this idea of beliefs as accounts thatdistort actual practices. It is interesting to speculate on the relationship between thisdeveloping argument and the theory of accounts as it developed later in the work ofKenneth Burke, C. Wright Mills, and Harold Garfinkel. In Garfinkel’s case in particularthe distorting relationship between accounts and practices is clear.


it cannot be enough to consult our own consciousness.” Analyzing eitherreligion, or the conditions of knowledge in general, from an individualperspective, makes a nonsense of both religion and epistemology.

In criticizing the individualistic, introspective, approach to religion,Durkheim ([1912:8–9]1915:19; *1995:6) draws an analogy between reli-gion and the family, arguing that it used to be thought that “the fatherwas the essential element of the family,” but then anthropologists dis-covered matriliny. His point is that it makes a big difference whether onetakes an individualistic view, or, whether the broader social elements ofa phenomena are taken into account. By introspection, western thinkerswere led to believe that the father is the essential element of the family,because those thinkers were western European men, and men headedthe western European family. But, a broader social view revealed that thesocial functions of the family can be fulfilled by many different forms. Thesame is true, according to Durkheim, if a broader social view of religionis taken.

Durkheim argues that an exclusive focus on modern religions has ledto the erroneous conclusion that certain ideas, like the idea of a deity, arean essential aspect of religion. However, a broad social analysis of archaicreligion shows that this is not the case. Archaic religions do not havedeities, properly speaking, but they are clearly forms of religious prac-tice and, Durkheim argues, serve the same social needs. Here the differ-ence between Durkheim’s idea that religions serve social functions, andWeber’s treatment of religion as sets of ideas and beliefs, is important. ForWeber ([1921]1968), archaic religions are magic, and not religion, pre-cisely because their practices attempt to cause effects in things. Treatingthe ideas and beliefs themselves as the purpose of the rites, led Weber tothis conclusion. For Durkheim, however, the beliefs are essentially irrel-evant. The practices are religion precisely because they do indeed attemptto cause effects through their rites, and in the process fulfill the need to pro-duce moral force and human reason.16 Furthermore, archaic peoples arecorrect, according to Durkheim, in thinking that through their practicesthey are creating effects that have moral significance. They are creatingmoral forces. This need for the creation of moral force is, he argues, thesame need that motivates modern religions.

Moreover, archaic religions, more clearly than modern religions, revealtheir origins in social needs. Whereas in modern religions one could pos-sibly (mistakenly) take the beliefs themselves literally as the purpose ofreligious practices, one cannot do so with archaic religions. Durkheimsays ([1912:10–11]1915:20; *1995:7) that “in the primitive religions,

16 In this regard Durkheim and Weber have perspectives on religion that are quite theopposite of one another.


the religious phenomenon still carries the visible imprint of its origins.”This imprint, according to Durkheim, is the social function of religiouspractice in creating the categories, not an individual, or group, expressionof faith.

Also, early in the Introduction, Durkheim makes a careful note of whathe means by the “origin” of religion. He does not want to be misun-derstood as speaking of an “historical” origin of religious thought, andsays ([1912:10–11]1915:20; *1995:7) that “there is nothing scientificabout the question” of historical origins. Rather, Durkheim ([1912:10–11]1915:20; *1995:7) is interested in locating the underlying social causesof religious thought: “I would like to find a means of discerning the ever-present causes on which the most basic forms of religious thought andpractice depend.” According to Durkheim ([1912:6]1915:17; *1995:4,emphasis added) the external resemblances between religions “presup-pose deeper ones. At the foundation of all systems of belief and all cults,there must necessarily be a certain number of fundamental representations andmodes of ritual conduct that . . . have the same objective meaning everywhereand everywhere fulfill the same functions.” Durkheim does not mean by thisthat the religious beliefs must be the same. Although he will argue thatthere must always be a distinction between the sacred and the profane.What he argues is that all religious practice has as its purpose to producecertain “representations” that are fundamental to the human understand-ing. Fulfilling this need is a social function because society cannot exist,reason cannot develop, and, coordinated action cannot occur, unless thisneed is fulfilled.

For Durkheim, religion has one underlying purpose: the creation ofthe categories of the understanding and the moral force that is theirequivalent. This is why there is religion in all societies. According toDurkheim ([1912:6]1915:17; *1995:4, emphasis on religion in original,other emphasis added): “It is these enduring elements that constitutewhat is eternal and human in religion. They are the whole objective con-tent of the idea that is expressed when religion in general is spoken of.”This claim, ([1912:6]1915:17; *1995:4, emphasis added) that a “certainnumber of fundamental representations and modes of ritual conduct” are the“whole objective content” of the idea of religion, runs contrary to receivedDurkheim scholarship. Durkheim is not referring to conventional beliefsor conventional representations whose meaning is given by a social con-text, but rather, to representations that have the “same objective meaningeverywhere.”17

17 The meaning of “the same” here will be discussed in greater detail in later chapters onthe development of particular categories of the understanding. Durkheim clearly states


When Durkheim uses the word “objective” with regard to the purposeof religion in generating the categories, he makes it clear that for himreligion does not consist essentially of systems of belief or narrative. Noris his analysis of religion about cosmologies of conventional concepts.Religion does not serve different purposes in each society. Durkheim’sanalysis of the fundamental characteristics of religion does not concernreligion in any conventional sense. In fact, most scholars who study reli-gion would certainly object to Durkheim’s subordination of religion toboth epistemology and underlying social functions.

Far from placing his emphasis on analyzing cosmologies of concepts,narratives and beliefs, Durkheim ([1912:11]1915:21; *1995:8) is inter-ested in archaic religions because of their lack of conceptual elaboration:“Their very lack of elaboration makes them instructive, for in this way theybecome useful experiments in which the facts and the relations amongfacts are easier to detect.”18 If what Durkheim meant by “facts” wasreligious narrative belief systems, this would make no sense. Westernreligions would clearly provide a better source for elaborate systems ofbelief. Archaic religions better serve the sociological purpose becausethere, what Durkheim calls the “facts” (and I would here say that “facts”refers to the actual performance of the rites, and the relationship betweenthat performance and the underlying needs) are easier to detect. In form,there is not much difference between this argument and the argumentof the Rules of the Sociological Method, generally considered to be as pos-itivist as The Elementary Forms is alleged to be idealist. Both deal withunderlying social needs and social relations as the “social facts” that arethe key to explaining social phenomena and the human individual. And

that the categories will differ from society to society. He also states that they are alwaysand everywhere the same. This has led to some confusion. Some scholars argue thatDurkheim only intended social causes to add content to innate categories, while othersargue that the argument as a whole makes no sense. I will try to show that Durkheimargued for sufficient variation to make it clear that the categories cannot be innate, whilenevertheless maintaining that in essential respects time is time, space is space, and causeis cause.

18 In the 1915 translation their lack of conceptual elaboration was translated as “On the con-trary, they are rudimentary and gross; we cannot make of them a sort of model whichlater religions only have to reproduce. But even their grossness makes them instruc-tive . . .” In the original French the passage reads, “Elles sont, au contraire, rudimen-taires at grossieries; il ne saurait doc etre question d’en faires des sortes de modeles queles religions ulterieures n’auraient eu qu’a reproduire. Mais leur grossierete meme lesrend instructives . . .” While the original French leaves the sort of “grossness” involvedambiguous, the passages immediately surrounding the quote make it clear that Durkheimmeans that the relationship between social facts is more easily seen and secondary char-acteristics do not get in the way. As he will say throughout, narrative elaboration is thesecondary characteristic that he wishes to avoid. I believe that the 1995 translation bestcaptures the spirit of the passage.


both focus on methods of sociological research which use empirical evi-dence with regard to those social facts.19

1.2.0 Section ii: The Introduction of Epistemology

It is in Section ii of the Introduction that Durkheim first raises the issue ofepistemology explicitly. This discussion has three parts; First, Durkheimargues that the study of religion, as he has outlined it in Section i, providesa new way of addressing epistemological issues that have appeared to beunresolvable. To make his point, Durkheim outlines the origin in religiouspractice of time, space and logic; Second, he provides an elucidation of theepistemological dilemma in philosophy, as he sees it, characterizing theopposing positions as apriorism and empiricism; and Third, he explainshow the argument of The Elementary Forms will provide a new solution tothat dilemma. What he does not do in this Section is to provide a detailedaccount of his own argument. That account is given only in the bodyof the text. Here he focuses instead on the general characteristics andimplications of his position.

While this discussion is important, it is confused by several issues.The discussions of time and space do not make the relationship betweenreligious practices and categories as clear as his later presentations in thebody of the text. Furthermore, there are discussions with regard to logicthat suggest a problematic equivalence between the categories and logic.This is doubly misleading because Durkheim does not follow through onwhat he says with regard to logic in the body of the text. While he doesfinally turn to a discussion of logic in general in the Conclusion, thatdiscussion seems to have been confused with the epistemology. All of thiscontributes to the impression that Durkheim is announcing a sociologyof knowledge, even though he is using the word epistemology explicitlyand referring to classical epistemological dilemmas, which he claims notonly to address, but to provide a solution to.

1.2.1 Relation Between Religion and Epistemology

In the First part of Section ii, Durkheim discusses the role of religion inthe development of human reason. Whereas in Section i, in his justifica-tion of Archaic religion, Durkheim argued that an individual view renders

19 The misunderstanding of Durkheim’s position in this regard has led to the belief that hisearly work was positivist and his later work idealist (see 0.3.0). In reality Durkheim wasnever a positivist in the modern sense and his early work was as innovative sociologicallyas his later work. The failure to see the grounding in a sociological epistemology hascontributed greatly to the confusion.


religious function, and hence Archaic religion, absurd, his argument hereis that an individual view obscures the epistemological relevance of reli-gious practice. He argues that communication and social cooperationrequire persons to have certain basic logical ideas in common and thatthe development of shared logical ideas cannot be explained by eitherempirical individualism or apriorism. The basic ideas that people havein common cannot be a priori, according to Durkheim, because theyvary too much from place to place. But, individualist empiricism can’texplain them either, because they are much more constant than the prob-lem of generalization would allow for (see 1.2.2 for a discussion of thisproblem).

Therefore, Durkheim says, the only viable explanation is that societiesthemselves have to provide for this sameness. Because persons in all soci-eties develop the same categories of reason, something common to allsocieties would have to exist that could produce such shared categories.Because religious practice is common to all societies, Durkheim argues,the sameness could be provided for by religious practices. In fact, heargues that the reason why societies develop religious practices in thefirst place is in order to develop a sameness and unity of thought amongthe members of the social group. He does not make this as a teleologicalargument, however. Rather, he states the argument in a conditional form:Societies cannot exist where sameness of thought does not exist to a sufficientdegree. Therefore, only groups that manage to develop religious rites thatare able to cause the ideas essential to this sameness and unity of thoughtwill become societies.

In this second section, Durkheim works slowly toward his thesis thatreligion can provide this necessary sameness of thought, and therefore,that religious practice is the key to epistemology. At first, his interest inthe epistemological significance of religion is only hinted at ([1912:12–13]1915:20–21; *1995:8); “There is an aspect of every religion that tran-scends the realm of specifically religious ideas. Through it, the study ofreligious phenomena provides a means of revisiting problems that untilnow have been debated only among philosophers.” It is not clear fromthis, or from the statements immediately following this passage, whetherhe is outlining a sociology of knowledge or an epistemology.

The argument continues to be ambiguous, in this regard. Durkheimargues ([1912:12–13]1915:20–21; *1995:8) that: “If philosophy and thesciences were born in religion, it is because religion itself began by servingas science and philosophy. Further, and less often noted, religion has notmerely enriched a human intellect already formed but in fact has helped toform it.” This statement, however, could still refer either to a sociology ofknowledge, wherein religious explanations served the purpose of scientific


explanations, or to a modified form of classical rationalism, in which theoutline of the categories remains apriori.

However, the discussion immediately following these remarks, makes itclear that Durkheim intends to indicate that the form of knowledge, as wellas the content has a social origin ([1912:12–13]1915:20–21; *1995:8):“Men owe to religion not only the content of their knowledge, in sig-nificant part, but also the form in which that knowledge is elaborated.”Arguing that religion shapes the form of knowledge, however, still doesnot indicate whether Durkheim is articulating a sociology of knowledge,or a classical epistemology. It does however, eliminate both classical ratio-nalism and apriorism as possible interpretations of his position.

But, as Durkheim begins to discuss the categories themselves he makesclaims that are not consistent with a sociology of knowledge argumenteither. His first articulation of the categories ([1912:12–13]1915:22;*1995:8) is a case in point: “At the root of all of our judgements, thereare certain fundamental notions that dominate our entire intellectual life.It is these ideas that philosophers, beginning with Aristotle, have calledthe categories of understanding: notions of time, space, number, cause,substance, personality.”20 This list of categories makes it clear that inspeaking of the categories Durkheim is not referring to arbitrary rep-resentations of religious beliefs, or saying that each society develops adifferent set of fundamental ideas. He refers rather to something likeAristotle’s categories of the understanding: one set of ideas common tohuman reason in all societies, in all times.

Furthermore, both here, and earlier in the Introduction ([1912:2–7]1915:14–17; *1995:2–4), where he introduces the idea of a fundamen-tal set of universal ideas that constitute the truth of religious practice,Durkheim discusses the categories in terms that suggest that they haveuniversal characteristics that are incompatible with a sociology of knowl-edge position. He says that the categories ([1912:12–13]1915:21–22;*1995:8–9): “correspond to the most universal properties of things. Theyare like solid frames that confine thought. Thought does not seem tobe able to break out of them without destroying itself, since it seemswe cannot think of objects that are not in time or space, that cannotbe counted, and so forth.” It is this sort of statement, clearly incom-patible with a sociology of knowledge position, that has been inter-preted as indicating that Durkheim is a classical rationalist; that is,that he is positing universal categories that then get filled in by socialexperience.

20 This list of categories is different from the list given in the Conclusion. It is the list givenin the Conclusion that corresponds with the proofs given in the body of the text.


But, the interpretation of Durkheim as a classical, or Cartesian ratio-nalist also conflicts with the text. Durkheim is very clear that the sharedcategories of thought do not exist unless society makes them. A clas-sical rationalist posits reason as an enduring and underlying feature ofthe universe, or of man. The idea of “creating” reason is not compati-ble with a classical rationalist position. The need to make categories ofthe understanding is, for Durkheim, however, the everpresent underlyingneed driving societies toward religious practice. There would be no needfor religion without this need for categories, which Durkheim calls the“whole objective content” of religion. While Durkheim certainly believesin the possibility of human reason, and therefore refers to himself some-times as a rationalist, his claim that the categories of the understandinghave a social origin, eliminates the possibility that his argument is rational-ist in any conventional philosophical sense. He is certainly not a classical,or Cartesian rationalist as has often been claimed.

In the fourth paragraph of Section ii, Durkheim says ([1912:13]1915:22; *1995:9) that the point, as he has made it so far, has a cer-tain interest, but that what he will say next “gives it its true significance.”The next point turns out to be a slightly different version of his argumentin Section i, that not only the categories, but religion itself is an essen-tially social phenomenon. What is different about the argument here isthat Durkheim argues not only that the categories come from religion,but that religion itself is born to fulfill this social need. While he hasalready argued at length in Section i that religion serves social functions,and has a social origin, this time the point has an epistemological rele-vance. In saying that the categories have an origin in a necessary socialphenomenon, in the context of his epistemological argument, Durkheimis claiming for them an origin in social necessity and not in the contin-gency of individual religious experience ([1912:12–13]1915:22; *1995:9,emphasis added):21

Religious representations are collective representations that express collectiverealities; rites are ways of acting that are born only in the midst of assembled groupsand whose purpose is to evoke, maintain, or recreate certain mental states of thosegroups. But if the categories are of religious origin, then they must participatein what is common to all religion: They too must be social things, products ofcollective thought.

Religion is, for Durkheim, a set of public practices that produce cer-tain shared mental states. It is not the result of individual mental states,

21 This equation of individual experience with contingency and social experience withnecessity reverses the usual philosophical view. Durkheim will take it up again in theConclusion. See my discussion in Chapter Nine of this book.


but rather their social cause. This treatment of both religion and thecategories, as fundamentally social phenomena, is the central pivot ofDurkheim’s argument. It allows him to turn the insurmountable dilem-mas of the empiricists into supports for his own position. The impossibilityof explaining the origin of empirically valid general ideas in individual expe-rience, becomes the necessity for the group experience of enacted moral forces.It is this idea that the categories are born in the midst of rites, ways ofacting, whose purpose is to evoke those categories, that Durkheim worksout in detail in the central chapters.

Durkheim argues that the purpose of religious rites is to evoke mentalstates in assembled groups. Religious worship is not an expression of eitheran individual, or a collective state of mind. Rather it evokes such a state.Durkheim says that this collective state is a “reality” that in the momentof production can be felt by participants. The collective representations,making up his sociology of knowledge, that develop afterwards to describethis process are, by contrast, expressions of collective realities, and thereforehave an only indirect basis in reality. This is why beliefs inevitably distortsocial reality.

Following this general discussion of the social origin of the categories,Durkheim introduces issues with regard to the social origin of the specificcategories of time and space22 and more general issues with regard to thesocial origin of logic. This discussion has been highly problematic becausemany critics treat it as his main empirical demonstration of the origin ofthe categories. In fact, Durkheim’s discussion of time, space, and logic,in the Introduction is, like his discussion of classification in PrimitiveClassification, very preliminary. Therefore, taking this to be Durkheim’sshowcase argument has caused problems.

With regard to time and space, Durkheim, tries to show that it is onlythe social divisions of the category of time that allow the category tohave any meaning. He argues ([1912:14]1915:22–23; *1995:9) that timewould not make sense if it did not have any divisions: “We can conceiveof time only if we differentiate between moments.” This differentiation

22 Durkheim discusses differences between himself and Kant over the status of time andspace as categories. One of the differences that Durkheim notes is that while he includestime and space in his list of categories, for Kant they are built into the faculty of percep-tion. Durkheim ([1912:10–11]1915:20–21; *1995:8ff ) considers them to be categories“because there is no difference between the role these notions play in intellectual life andthat which falls to notions of kind and cause.” Kant had made time and space intuitionsbecause he couldn’t see how a perception could be made without time and space. Theother categories are applied to perceptions by the faculty of judgement after they enterthe mind. If time and space were located in the faculty of judgement perceptions not inspace and time would have to be put into space and time by the faculty of judgement. Toavoid this problem, Kant made them intuitions, not categories. Because, for Durkheim,the categories reflect emotional experience the problem is not the same.


he says ([19112:14]1915:22–23; *1995:10), has a social and not an indi-vidual origin: “The category of time is not simply a partial or completecommemoration of our lived life. It is an abstract and impersonal frame-work that contains not only our individual existence but also that ofhumanity.”23

In distinguishing between individual and social existence in his dis-cussion of time, Durkheim refers to a distinction between the social andthe individual nature of persons that he called the “dualism of humannature.” This is a very complicated notion, involving the idea that socialrelations completely transform the person. It has been mistakenly identi-fied by scholars with either a Cartesian or a Kantian mind/body dualism.However, Durkheim’s dualism involves a very different sort of distinctionbetween the social being and the individual. In his lectures on Rousseau([1892]1960) Durkheim credits the philosophical precedent for this ideato Rousseau’s essay On the Origins of Inequality, not to Kant or Descartes.While the discussion of dualism is essential to the argument of the Intro-duction, understanding what Durkheim meant by dualism requires anextensive discussion, much of the evidence for which must be drawn fromother publications. Therefore, a chapter devoted entirely to Durkheim’sdualism (Chapter Two) will follow this chapter.

Durkheim argues that persons in a given society not only share thesame idea of time, but that the idea of time always corresponds to moralaspects of the society ([1912:15]1915:23; *1995:10): “It is not my timethat is organized in this way; it is time that is conceived of objectivelyby all men of the same civilization.” Social time is not the same thingas the experiences of the individual organism. According to Durkheim([1912:15ff]1915:23ff; *1995:10ff), the category of time is not comprisedof sense impressions: “the complexus of sensations and images that servesto orient us in duration . . . The first are the summary of individualexperiences, which hold only for the individual who has had them. Bycontrast, the category of time expresses a time common to the group –social time, so to speak. This category itself is a true social institution.”

The same argument, Durkheim says ([1912:15]1915:23; *1995:10),applies to the category of space. Space cannot be conceived of as inde-terminate any more than time. Durkheim ([1912:15]1915:23; *1995:10)characterizes his argument, that “If purely and absolutely homogeneous,[space] would be of no use and offer nothing for thought to hold onto,” as standing in contrast to Kant. An undifferentiated space could not

23 Durkheim’s position on time is superior to James (and Schutz who followed James) innot needing an “objective” measure of time to contrast with the individual experienceof time. For Durkheim, the social category of time is neither objective or individual, butrather constructed by the group.


be experienced, he argues ([1912:15]1915:23; *1995:10): “Spacial rep-resentation essentially consists in a primary coordination of given senseexperience. But this coordination would be impossible if the parts of spacewere qualitatively equivalent, if they really were mutually interchange-able.” It is the moral qualities of space and time that give them form.Then Durkheim ([1912:16]1915:24; *1995:10–11) asks “But where dothese divisions that are essential to space come from?” He gives severalillustrations of his argument that the ideas of time and space vary fromplace to place and that in such cases they correspond to the social/moralstructure of the groups that hold them in some obvious way.

The ambiguity in the discussion of time and space is that in someplaces it is clearly a sociology of knowledge argument in which a socialframework of space and time ideas are learned and passed down throughthe society. In the Introduction, Durkheim gives no indication that thecategories of space and time are experienced emotionally by participants.The discussion at these points is much closer to the argument of PrimitiveClassification than to the epistemology that emerges in the central chap-ters of The Elementary Forms. However, in the body of the text, Durkheimconnects the ideas of space and time more explicitly with the direct expe-rience of moral divisions and forces experienced during religious ritual.Although, even there, time and space do not get as convincing a treatmentas force and causality.

One problem is that none of the arguments for the categories can besuccessfully made until it has been established that the ideas of the sacredand of moral force are the result of enacted social practice. As this doesnot occur until Book II, Chapters Six and Seven, arguments that comeprior to this point in the text remain somewhat ambiguous. The origin ofthe categories in the concrete experience of moral force by participantsenacting religious ritual, that emerges fully only in the cases of force andcause, is the hallmark of Durkheim’s epistemology.

The discussion of time and space also becomes confused with variousclaims about logic. Durkheim ([1912:17]1915:24–25; *1995:11–12) inmaking an explicit statement, denying that the categories of space andtime are innate, includes basic concepts of logic in his argument: “Farfrom being built into human nature, no idea exists, up to and includingthe distinction between right and left, that is not, in all probability, theproduct of religious, hence collective, representations.” This is a strongstatement of the social constitution of even the basic distinction betweenright and left.

While this might appear to offer support for Durkheim’s epistemolog-ical argument, in fact, this inclusion of logic in general along with hisargument for the social constitution of the categories, confuses the issue.


If Durkheim meant to include all logical ideas, then his argument wouldhave to be a sociology of knowledge. It is not possible that the most basicthought processes have a social origin, although they may be shaped bysocial processes. This would leave animals without basic survival skills,and conflicts with Durkheim’s own argument with regard to dualism.However, nowhere in the central chapters does Durkheim attempt toprovide an origin for these logical concepts in enacted practice, insteadrelying on statements about religious representations to explain logic,which would make it a sociology of knowledge argument. He only dis-cusses logic in summary fashion in Section iii of the Conclusion. Hiscomments on logic in the Introduction, therefore, may unfortunatelyhave overshadowed his argument for the categories, even though he doesnot follow through on this suggestion in the body of the text.

For instance, in the Introduction, Durkheim ([1912:17]1915:25;*1995:12) makes what appears to be a sociology of knowledge argu-ment with regard to the idea of contradiction. He suggests that the verybasic logical notion of contradiction may itself have a social origin. He([1912:17]1915:25; *1995:12) suspects this may be the case because:“the hold that the notion of contradiction has had over thought has var-ied with times and societies.” This variation in the role played by con-tradiction leads to significant differences between aboriginal and modernsocieties. According to Durkheim ([1912:17]1915:25; *1995:12) “todaythe principle of identity governs scientific thought; but there are vast sys-tems of representation that have played a major role in the history of ideas,in which it is commonly ignored.” These vast systems of representation,according to Durkheim ([1912:17]1915:25; *1995:12), are the mytholo-gies and they “deal with beings that have the most contradictory attributesat the same time.” For Durkheim ([19112:17]1915:25; *1995:12), thesevariations suggest at least a partial social origin for the modern “logical”idea of contradiction: “These historical variations of the rule that seems togovern our present logic show that, far from being encoded from eternityin the mental constitution of man, the rule depends at least in part uponhistorical, hence social factors.” However, in his arguments with regardto dualism, the ability to tell right from left is a basic animal ability.

To further confuse things, not only does Durkheim’s argument in thebody of the text not provide logic with an empirical origin in social prac-tice, but he only argues that its social origin is partial, which seems tosupport the popular interpretation of his position as both rationalist andidealist. The argument can appear, as Schmaus (1998) has interpretedit, to be that some, almost empty, universal categories of logic need to befilled with social contents; indicating a sociology of knowledge built on anunderlying foundation of rationalism, because logic is derived partially


from social representations and partially from an underlying reason; notdirectly from experience as required by a “foundational” epistemologicalargument from origins.

It is, therefore, essential to distinguish Durkheim’s claims with regardto logic in general, from his epistemology (discussions of logic can befound in Durkheim’s text in Book II, Chapter Seven, Section iv, and inhis Conclusion). Strong as his claims with regard to logic appear to be,they do not receive the same sort of demonstration of empirical originsin the central chapters that Durkheim offers for the categories of theunderstanding. Contradiction, and the distinction between right and left,are not treated by Durkheim as categories of the understanding. In hisdiscussion of the categories themselves Durkheim will emphasize that itis the direct experience of moral forces like space and time, enacted bythe assembled group, that invests those categories with their divisions,and hence their intelligibility, not the learning of social representations.

Durkheim also makes the misleading claim, immediately after thesestatements about the social origins of logic, that “analogous demonstra-tions concerning the notions of genus, force, personality, and efficacywill be found below” ([1912:17]1915:25; *1995:12). He does not makeanalogous demonstrations of the categories. Those demonstrations arequite different. Furthermore, it is Aristotle’s categories that are listed byDurkheim in the Introduction. But, he does not use these terms in hisdemonstrations of the social origins of the categories in the body of thetext. Personality, if I am right in equating it with soul, is not treated asa category at all, in Durkheim’s argument, but as the first idea of moralforce, or the sacred, on which all the other categories depend. Categoriesreferred to by Durkheim as force, or moral force, and causality, are thosewhich receive recognizable demonstrations in the text. Ideas like soul, orpersonality, and logical concepts in general, are not treated in the sameway, as the Introduction might lead one to believe.

This discrepancy is further evidence that Durkheim’s argument wasevolving as he wrote the text. It is the category of causality (not efficacy),and to a lesser extent totality, in addition to time, space, class, and force,that receive empirical proofs in the central chapters. Causality, which isthe most important category for the empiricists, and received the mostextensive treatment in the central chapters, is not even mentioned on thislist (efficacy is on the list. But it is a weaker idea than causality whichDurkheim defines as necessary force).

It seems as if Durkheim may have been toying with a very differentidea when he wrote the Introduction; an idea for a social explanation ofcategories of logic, and, in particular, Aristotle’s logic. In some places itseems as if this initial idea might have had inherently rationalist elements,


although it is clear by the Conclusion that it does not. However, thecategories that Durkheim does provide empirical discussions of in thecentral chapters, are not inherently rationalist, and speak to the empiricistdebate, rather than to Aristotelian logic, or Kantian categories. Hume, inparticular, devotes the bulk of his argument to the origins of the conceptof causality, as does Durkheim. It may be that the categories Durkheimeventually chose were those that he could match up with various instancesof moral force. Or, it may be that as he worked more closely on his debatewith empiricism, he realized that his proof of the concept of causality, thekey concept in the empiricist argument, was also the key to solidifying asocio-empiricist epistemology.

It is even possible that having already worked out the origin of the cate-gories of the understanding in the enactment of moral forces, Durkheimhad begun thinking his way through how this would help to explain thesocial origin of logic in ways that were also socio-empirically valid. He mayhave meant to complete his proofs for the social origin of logic at a laterdate. We will never know for sure, although the earlier publication date of1909 for the Introduction, suggests that the idea for a socially determinedlogic was written before Durkheim had worked out the mature epistemo-logical argument of The Elementary Forms, and was less well developed.The Conclusion does suggest that the argument for logic would havecome next. One thing is certain, however, whatever Durkheim meant toindicate about left, right, and contradiction in the introductory chapter,it is not the same as his argument for the categories of the understandingin the central chapters.

1.2.2 Durkheim’s Assessment of the Epistemological Debate

In the Second part of Section ii, of the Introduction, Durkheim([1912:18]1915:26; *1995:12) begins an assessment of the epistemologi-cal debate as it has been “up to the present.” In speaking of epistemology,Durkheim ([1912:18]1915:26; *1995:12) wrote that “only two doctrineshave opposed one another.” These “two doctrines” Durkheim refers toas the empiricists (Locke, Berkeley, Hume, and followers) and aprior-ists (Kant and followers).24 In his lectures on Pragmatism, given a year

24 Hume and Kant will be allowed to stand for contemporary representatives of thesepositions which Durkheim may have been more explicitly referencing, with the exceptionof William James who is explicitly mentioned. I refer primarily to Hume, instead of morecontemporary empiricists to whom Durkheim might have been referring because of thestriking similarity between his approach to the problem of general ideas and causality inparticular and Hume’s original argument. The argument belongs to Hume regardless ofwho may have repeated it. It also follows the form of Hume’s original argument.


later (1913–14), he also explicitly criticized the empiricism of WilliamJames. Neither empiricism or apriorism had been able to provide an ade-quate solution to the problem of knowledge and the debate continued inDurkheim’s day.

Durkheim first characterizes, and then criticizes and rejects, both ofthese schools of thought. This is important, as most scholars considerDurkheim’s position in The Elementary Forms to be either Kantian, orPragmatist, and almost all of them consider it to contain inherently clas-sical rationalist elements. A careful reading of this section makes it clearthat Durkheim does not consider himself to be any of these. He comesclosest to adopting the empiricist position, although he rejects both theclassical and pragmatist forms of empiricism, and apriorism, with equalcertainty. The empiricism that Durkheim adopts is, by contrast, distinctlysociological.

The task of an epistemology is to explain the possibility of humanknowledge and to ascertain what relation human experience and ideasbear to the particular objects, or moments, of experience. When knowl-edge is thought to begin with individual perception, certain problemsarise: the things which persons perceive change from day to day, andfrom moment to moment. Nothing is ever exactly the same twice, andthe stream of perception (because persons have not already acquired gen-eral categories of thought, but rather are attempting to derive these ideasfrom that stream) must be constantly changing and yet undifferentiated.

The question from this perspective, is how general concepts can bederived from this undifferentiated stream of particular experiences, whichare literally not the same from one moment to the next, let alone fromperson to person. How can an object be given the same name, with thesame meaning, by two persons who necessarily perceive it differently?Or even by one person perceiving it at two different moments? Or asan undifferentiated perceptual field? How can perceptions which are notexactly the same give rise to the same ideas in a manner that is empiricallyvalid and not in any sense arbitrary? To further complicate the problemof how categories based on sameness can be derived from an undifferen-tiated stream of particulars, perceiving sameness presupposes the abilityto perceive difference.

If one looks at how persons actually use language, objects that are saidto belong to a general class or category, even when referred to by thesame word (e.g., two “tables”, or, two “dogs”), are never exactly like oneanother. On what basis do persons assign some things to one general cat-egory, while others, which may be very similar, are assigned to another?The empiricists argued that unless general ideas were part of a particularsense experience, and therefore directly perceivable in their own right,there must be a process by which individual perceptions are analyzed to


form general ideas. But Hume was able to show that general ideas couldnot be part of a single sense experience. Therefore, they must be formedby a process in the mind. A mental process cannot be empirically valid, asit would of necessity add something to experience. Therefore, the result-ing general ideas could not be empirically valid either. Given the infinitedifferences between things, and the impossibility of explaining the originof valid general ideas given an empiricist starting point, the consistentapplication of both particular terms and general categories is impossibleto give an explanation for, unless one accepts their conventional char-acter, or assumes that persons have apriori knowledge of how to applythem. This latter was the Kantian solution to the dilemma.

However, resorting to a priori explanations, while it preserves a sortof empirical validity, creates insurmountable barriers between empiricaland conceptual reality. Concepts, while perfectly shared between ratio-nal beings, bear no empirically valid relationship to the world in itself.The Kantian solution reproduces the essential problem with the classi-cal rationalist argument that empiricism had sought to overcome: thatif mind is universal and obeys different laws than physical bodies, thenhuman reason is incompatible with the demands of modern science forempirical validity. In the age of Newton, heroic attempts had been madeby Locke to put science on a secure philosophical footing, through aclarification of the empirical foundations of knowledge. Attempts by theempiricists, Locke and Hume, to solve the problem through a detailedlogical analysis of individual perception, and the logical relation betweenobjects in perception, had concluded in Hume’s argument that logicalrelations, indeed all relations, are properties added by the mind, and notpart of the original empirical perception in its own right.

Hume argued that even if particular individual experiences could beconsidered empirically valid, as Locke had argued, all attempts to gener-alize on the basis of them would still be invalid because a general categoryrepresents something that is different from, and added to, the experienceof particular objects and events. Objects do not display their relatedness toone another directly to perception. If general categories represent some-thing not present in the separate particular experiences that have beenadded by the mind to the collection of particulars, then, Hume argued,general ideas have no empirical validity. Thus, the attempt by empiri-cists to establish a direct relation between perception and an underlyingnatural reality that would support science, had ended in skepticism.25

25 It can be argued that Hume was not a skeptic because he was satisfied with his own solu-tion, feeling that the concept of causality as a “habit” sufficed for all practical purposes.However, it would still be true to say that he gave up on the idea of establishing a validrelation between general ideas and experience and that is what is meant by skepticismhere.


On the other hand, Kant’s attempt to solve the problem by positinga limited number of a priori categories in place of the large number ofrational forms posited by earlier rationalists, resulted in a situation whereperfect shared knowledge between individuals can be explained, but thisknowledge is only an imperfect reflection of reality, as all perception is, onthis view, mediated by a priori categories. Thus, reality in itself, is neverperceived. In this case the explanation of the origin of the categories issolved at the expense of the empirical validity of the individual experienceof reality; generating a split between the world as an object of experienceand the world in itself; Kantian dualism.26

In his discussion of this epistemological dilemma, Durkheim ([1912:18]1915:26; *1995:12) first characterizes Kantian, or Neo-Kantianepistemology; “For some the categories cannot be derived from expe-rience. They are logically prior to experience and condition it. They arethought of as so many simple data that are irreducible and imminent inthe human intellect by virtue of its natural makeup. They are thus calleda priori.”

Immediately following this characterization of Kant’s position, Durk-heim ([1912:18]1915:26; *1995:12) characterizes Hume’s empiricism:“For others, by contrast, the categories are constructed, made out ofbits and pieces, and it is the individual who is the artisan of that con-struction.” Following the contrasting characterizations of the two posi-tions, Durkheim ([1912:19]1915:26; *1995:13) criticizes them both,concluding that “both solutions give rise to grave difficulties.” Adoptingthe empiricist position, according to Durkheim ([1912:19]1915:26;*1995:13), creates problems because “the categories must be strippedof their characteristic properties.” They become, as he will say later, anaverage of many particulars that leaves out what was essential to each asa particular. Because empiricists can’t explain how the categories comeinto being they conclude that they must be the result of opinion andhabit; a conclusion that strips the categories of both validity and neces-sity, properties that Durkheim required of the categories.27

Later, at several points in the body of the text, especially with regardto his discussion of classification (Book II, Chapter Three, Section ii)Durkheim returns to the problem of general ideas. In articulating theempiricist dilemma Durkheim argues that because experience is made

26 The rejection of positivism in the twentieth century generally embraced such a splitalthough it tended to be based on culture, not apriori categories. Phenomenology andexistentialism are examples.

27 There is a footnote on habit in the Conclusion ([1912:620ff1]1915:428ff10; 1995:436ff10) in which Durkheim notes that you can only have a habit where somethingalready exists. Therefore, habit cannot explain the genesis of anything.


up of bits and pieces, of particulars, general ideas must be added to expe-rience, since they cannot be found in it. Therefore, general ideas have noempirical validity. This obviously means that they cannot be considerednecessary, universal or empirically valid.28 However, it is Durkheim’sown view that the categories are necessary, universal and empiricallyvalid. Therefore, the categories of the understanding cannot be gen-eral ideas. Durkheim ([1912:19]1915:26; *1995:13) rejects the empiricistposition, saying that its conclusion is unacceptable because the categoriesare necessary: “In fact, they are distinguished from all other knowledgeby their universality and their necessity.”

Durkheim ([1912:19]1915:26; *1995:13) is harsh in his criticismof empiricism: “Classical empiricism leads to irrationalism; perhaps itshould be called by that name.” While he has little better to say aboutthe apriorists, calling them lazy and saying that they have offered nosolution, he notes ([1912:19]1915:26; *1995:13) that, ironically, it is the“apriorists who are more attentive to the facts,” by which he means thatthey at least recognize the universal and necessary character of the cat-egories. According to Durkheim ([1912:19]1915:26; *1995:13) “Sincethey do not take it as self evident truth that the categories are made ofthe same elements as our sense representations, they are not committedto impoverishing the categories systematically, emptying them of all realcontent and reducing them to mere verbal artifices.” This reference tosense impressions is important. For Durkheim, the categories come fromsocial experience, or emotion, not from sense impressions. The advan-tage of apriorism, for Durkheim, was that at least Kant did not begin withsense impressions.

Beginning with sense impressions, Durkheim believes, it is impossibleto arrive at the origin of the form of thought. In speaking of the differencesbetween form and content, Durkheim argues for the impossibility of anempiricist solution to the dilemma of where the framework for thoughtcomes from. Again, this has led to confusion. Some have thought that insaying this Durkheim was advocating apriorism for form and empiricismfor content. He was not. Durkheim argues ([1912:208–9]*1915:171;1995:147, emphasis in original) that the contents of perception: “aremade up of vague and fluctuating images, due to the superimposition andpartial fusion of a determined number of individual images.” According to

28 For other places in the text where Durkheim discusses general ideas see: ([1912:208]*1915:171–2; 1995:147); ([1912:339]*1915:270; 1995:239); ([1912:223]*1915:183;1995:158); ([1912:519–20]*1915:407; 1995:368–9); ([1912:526]*1915:411; 1995:372); ([1912:616–618]*1915:480–81; 1995:433–35); ([1912:630]*1915:490; 1995:442–3).


Durkheim ([1912:208–9]*1915:171; 1995:147) no scholar has identifiedvalid categories with generic images:

this is why every school of thinkers has refused to identify the idea of class with thatof a generic image. The generic image is only the indistinctly bounded residualrepresentation left in us by similar representations, when they are present inconsciousness simultaneously . . . an animal is able to form generic images thoughthey are ignorant of the art of thinking in classes and species.

While the stream of sense perception cannot provide the material for gen-eral ideas, Durkheim argues that social force as an “object” of emotionalexperience may better serve the purpose.

There is a difference between collections of things in perception,and social units which have internal cohesion that can be experienced.According to Durkheim ([1912:208–9]*1915:171; 1995:148) “materialthings may be able to form collections of units, or heaps, or mechanicalassemblages with no internal unity, but not groups in the sense we havegiven the word. A heap of sand or a pile of rock is in no way comparable tothat variety of definite and organized society which forms a class.” Thekey to Durkheim’s position here is the phrase “internal unity.” Socialcollections can be directly experienced through the experience of theirinternal connections, or moral forces. Durkheim is not simply saying herethat social conventions, or conventional social arrangements, supply thematerial for general ideas.

It is interesting to consider Durkheim’s criticism ([1912:19]1915:26;*1995:13) of empiricism as having reduced the categories to “mere ver-bal artifices,” in light of the interpretation of Durkheim’s own argumentas a sociology of knowledge that reduces the categories to conceptualconventions. Essentially, Durkheim’s criticism of empiricism is that theargument that causality has only a conventional meaning, while suffi-cient for all practical purposes, ignores the universality and necessity thatDurkheim believes are an indispensable characteristic of the categories(See my Chapter Eight on causality and Durkheim’s Book III, ChapterThree). If Durkheim had really meant for his own epistemology to beinterpreted as a sociology of knowledge of conventional meanings, thenhe should be in agreement with empiricism on this point. Instead, he crit-icizes Hume ([1912:19]1915:26; *1995:13) for reducing the categoriesto “mere verbal artifices.”

On the other hand, while his criticisms of empiricism might appearto align Durkheim with Kantian rationalism, his criticism of the apri-orists makes it clear that he did not take a classical rationalist, or apri-orist position either. In addition to calling apriorists “lazy,” he arguesthat ([1912:21]1915:28; *1995:14) “the categories of human thought


are never fixed in a definite form; they are ceaselessly made, unmade,and remade; they vary according to time and place.” For a classical ratio-nalist the categories are of necessity universal and fixed in an eternal form.For a Kantian they are innate. For neither are they made and remade.Nor would they require participation in social practice to make them.Durkheim argues that the apriorists have not made their case any morethan the empiricists have. He criticizes them for arguing that becausecategories of the understanding are necessary for knowledge, but cannotbe accounted for using the methods of classical empiricism, they must beapriori. Durkheim insists that the possibility of apriori categories mustitself be explained, not just arrived at by a process of elimination.

Durkheim ([1912:21]1915:28; *1995:14) complains that the aprior-ists “ascribe to the intellect a certain power to transcend experience andadd to what is immediately given. But for this singular power they offerneither explanation nor warrant.” Durkheim considers that the aprior-ists have cheated.29 They have shifted the problem from the origin ofconcepts, to the justification of innatism. They believe that categoriesare necessary, but they cannot explain them, so they conclude that theymust be innate. But this still leaves the problem of where the innate con-cepts originally come from and how they got to be innate in the firstplace. Durkheim ([1912:208–9]*1915:172; 1995:147–8) will say later inthe text that this approach to the problem is a “lazy man’s solution . . . thedeath of analysis.” Since the innate, or apriori, thesis is the key to theKantian argument, to reject it is to reject the argument as a whole.Durkheim is not a Kantian.

The point, Durkheim says ([1912:21]1915:28; *1995:14), “is to knowhow it happens that experience is not enough, but presupposes condi-tions that are external and prior to experience, and how it happens thatthese conditions are met at the time and in the manner needed.” Howdo just those categories that we need, in order to develop reason, comeinto being just because we need them? The argument seems somewhatmagical when posed in this way. The apriorists accept Hume’s argumentthat individual sense experience alone is not enough to explain the ori-gin of the categories. But, their strategy of building categories into theindividual mind does not challenge the individualist premise of Hume’sposition. Durkheim challenges the apriorists to explain the developmentof conditions prior to individual experience that will explain the origin ofthe categories.

29 Husserl similarly accuses Kant of cheating over this issue. He considered Hume to havebeen more honest, even if mistaken in his approach. See my “Reply to Gallant andKleinman” (1985) for a discussion of this issue.


Durkheim believed, as Kant did, that the categories are necessary. Hewas not willing to accept the empiricist, or pragmatist, position that cat-egories, in the classic sense, do not exist. But, he thought that the realchallenge to epistemology lay in explaining the development of condi-tions prior to individual experience that could explain the origin of thecategories of the understanding with sufficient necessity and universality.Durkheim argues ([1912:19]1915:26; *1995:13) that the categories arenecessary because “they are the common ground where all minds meet.”Without shared categories, Durkheim argues, both mutual communica-tion and society would be impossible. Therefore, Durkheim agrees withKant, that wherever society and mutual communication exist, the cate-gories must also exist as a necessary foundation. But, the possibility oftheir existence must be explained.

According to Durkheim, the empiricist and apriorist positions amountto the same thing. Both have in common the assumption that an episte-mological argument must begin with individual perception and explainhow individuals come to share knowledge in common.30 Either the indi-vidual generalizes the categories from individual experience (empiricism),individual minds come ready equipped with innate ideas (Kantian apri-orism), or the world is organized into categories, or natural forms, whichindividual minds have only imperfect perceptions of (classical rational-ism). As a consequence, according to Durkheim ([1912:21]*1915:28;1995:14), in spite of their differences, both rationalism and empiricismresult in perpetuating the same epistemological problem:

If this debate seems to be eternal, it is because the arguments given are really aboutequivalent. If reason is only a form of individual experience, it no longer exists.On the other hand, if the powers which it has are recognized but not accountedfor, it seems to be set outside the confines of nature and science.

Philosophers have not questioned this focus on the individual inher-ent in both the empiricist and apriorist positions. Questioning it wasDurkheim’s central thrust, and the foundation of the discipline he calledsociology. Referred to variously as a Neo-Kantian, classical empiricist,or positivist, Durkheim clearly rejected the individualist premise inher-ent in all of those positions, thereby avoiding their outcomes. He alsorejected absolutely, as an explanation of the origin of the categories ofthe understanding, the Kantian and Neo-Kantian assumption of a set ofideas common to all persons that exists apriori, either in their minds, orin society (as a group mind).

30 Even though apriorism appears to begin with universals, and therefore not to be anindividualist position, it would not need to posit universals, in Durkheim’s view, unlessit were accepting the individualist dilemma in the first place.


For Durkheim the categories must have an origin in the experienceof each and every individual person. But, those experiences must becaused by conditions that occur prior to (or at least simultaneously with)the development of reason in the individual social being, and the sameexperiences must be available to all beings who would become rational.For Durkheim, such conditions occur only during the collective enact-ment of practice, an inherently social, not an individual, or natural phe-nomenon.31 Furthermore, such conditions are available to persons onlyin their capacity as participants in the assembly, an inherently social, notan individual state of being.

1.2.3 Sketch of Durkheim’s Sociological Position

Durkheim rejects individualism and takes an inherently sociologicalapproach to the classic epistemological dilemma. This does not mean,however, that he accepts social relativism, as is usually argued. But, rather,that he gives a social explanation for the valid empirical origin of the cate-gories of the understanding. In Section ii of the Introduction, Durkheimargues that all societies need persons in those societies to develop sharedcategories of the understanding. Because society requires logical consen-sus, it must ensure that logical consensus is created. In order to createlogical consensus, persons must participate in certain types of enactedpractice that create moral force. In making this argument, Durkheimdescribes processes of coercion that are essential to creating the neces-sary logical consensus ([1912:24]1915:30; *1995:16): “Thus, in order toprevent dissidence, society weighs on its members with all its authority.Does a mind seek to free itself from these norms of all thought? Societyno longer considers this a human mind in the full sense, and treats itaccordingly.”

This argument is often interpreted to mean that individuals are forcedto believe in, or think, in terms of the categories of the understanding.There is a big difference between forcing persons to believe, however,and forcing them to participate in social practices that create shared expe-riences. While it may indeed be the case that many, even most, collectiverepresentations are forced on the minds of individuals through variousforms of coercion, the argument that the categories themselves are theproduct of coercion in this way is not consistent with Durkheim’s argu-ment that the six categories of the understanding have a direct experi-ential origin in enacted practice. Coercion itself could not produce the

31 That social is in a sense natural, but still to be distinguished from nature. See the dis-cussion at the end of 1.2.3.


experience of the categories. However, experiencing the moral force thatgives the coercion its power can, and does.

Coercion, however, does accomplish two essential things. It ensuresthat the rites necessary to generate the experience of the categories areperformed, and it prevents rites from being performed that would gener-ate ideas that conflict with the basic premises of the society in question.32

The interpretation of coercion that would be consistent with Durkheim’sepistemological position is that in order to make sure that the categories areproduced, society coerces its members to perform those actions that are necessaryfor, and to abstain from actions that would interfere with, the production of thosemoral experiences that cause the categories. Durkheim ([1912:24]1915:30;*1995:16) writes:

Outside us, it is opinion that judges us; more than that, because society is repre-sented inside us as well, it resists these revolutionary impulses from within. Wefeel that we cannot abandon ourselves to them without our thought’s ceasing tobe truly human. Such appears to be the origin of the very special authority thatis inherent in reason and that makes us trustingly accept its promptings. This isnone other than the authority of society.

The authority of society forces persons to perform rites that createideas without which their thought would cease to be recognizably human.“Thus,” according to Durkheim ([1912:24]1915:30; *1995:17), “thenecessity with which the categories press themselves upon us is not merelythe effect of habits whose yoke we could slip with little effort; nor is thatnecessity a habit or a physical or metaphysical need, since the categorieschange with place and time; it is a special sort of moral necessity that isto intellectual life what obligation is to the will.”33 The external coercionto reproduce the rites creates the categories through the enactment of therites, which in turn, produces internal coercion through moral force.

This internal coercion of the moral force of the categories is differentfrom the experience of sense perception. In arguing that the categorieshave a social origin, Durkheim ([1912:24]1915:30; *1995:16–17) pointsout that they impose themselves on persons from the outside with anauthority that sense perceptions do not have: “This is none other than

32 In modern society, this coercion may have shifted its location away from “formal” insti-tutions and occur more often in everyday interaction in the form of sanctions of thosewho violate expectations at the level of “Interaction Order.” See Rawls 1987, 1991.

33 For Kant, it is a logical necessity based on consistency. For Durkheim it is a feeling ofmoral necessity created in persons by social practices, so that they will not contradict,or undermine, the ways of life, i.e., practices, necessary for the creation of reason andlogic. His note that the categories change by time and place seems to refer to the socialcontext in which the category is experienced, and not to variations in the correspondingmoral forces.


the authority of society passing into certain ways of thinking that are theindispensable conditions of all common action.” Moral force is what par-ticipants feel when the authority of the group is enacted by the practicesand experienced directly. The point of the contrast between sense per-ception and moral force is that sensations do not impose themselves onus in this way. Sense perceptions can be doubted, moral forces cannot.The contrast enables Durkheim to argue that the most impressive forcesare moral forces, not natural forces, and therefore, that the first greatforces that the first religions sought to explain were moral forces and notnatural forces. Durkheim will continue to elaborate this point throughoutBook I.

The idea that the categories are experienced internally as emotionssolves a problem for Durkheim. He argues that the empiricists had madenot one, but two mistakes: first, they focused on the individual; second,they focused on sense perception. The task was therefore to find a sourceof knowledge that did not depend on sense impressions and that wasan inherently collective experience. The difference between the experi-ence of the categories, through enacted ritual practices, and the sensibleapprehension of objects, for Durkheim, is that the latter are experiencedas particulars that could have been otherwise than what they are, whilethe former are experienced as necessary forces, or feelings, that could notbe different. In speaking of this difference, Durkheim refers to the expe-rience of the categories as the experience of moral forces. This Durkheim([1912:19]1915:26; *1995:13) contrasts to the ordinary sense perceptionof natural objects: “Of course, when sensations are present to us, theyimpose themselves on us in fact. By right, however, we remain free toconceive them otherwise than they are and to picture them as occurringin an order different from the one in which they occurred.” Moral forces,on the other hand impose themselves on us “by right,” and we are notfree to consider them otherwise.

According to Durkheim ([1912:19]1915:26; *1995:13), sensation isalways particular: “Sensation or an image is always linked to a definiteobject or collection of definite objects, and it expresses the momentarystate of a particular consciousness. It is fundamentally individual and sub-jective.” Moral forces, on the other hand, always evoke a feeling aboutrelationships in the social group. These relationships are real and consti-tutive of both the group and its members. The truths of these relation-ships are not momentary and are never the property of any particularconsciousness, but rather, are only true because they are shared by theparticipants in the group.

Whereas sensation is an inherently individual phenomenon, Durkheim([1912:19]1915:26; *1995:13) notes that the experience of the categories


has an inherently shared moral quality: “Reason, which is none other thanthe fundamental categories taken together, is vested with an authority thatwe cannot escape at will.” If persons try to think without the categoriesthey “meet sharp resistance” ([1912:19]1915:26; *1995:13). Reason, forDurkheim, is created by moral truths, whereas sensation consists of acci-dents of individual sensation. They make up two completely differentsorts of knowledge ([1912:19]1915:26; *1995:13):

Here, then are two sorts of knowledge that are like opposite poles of the intel-lect. Under these conditions, to reduce reason to experience is to make reasondisappear – because it is to reduce the universality and necessity that characterizereason to mere appearances, illusions that might be practically convenient butthat correspond to nothing in things.

When Durkheim says that reducing reason to experience makes reasondisappear, he means that reducing reason to individual sense perceptionwill have this result.34 For Durkheim, a focus on the moral experienceof persons engaged in enacted practices within assembled groups is theonly possible source for valid categories. These moral experiences are notavailable via sensation. But, only through the emotional experienceof moral force. Later in the text (see my Chapters Seven and Eight)Durkheim consistently refers to this experience as emotion.

It was trying to explain moral truths, truths that “translate” the expe-rience of the moral relationships within the social group, on the basisof individual sensation, in Durkheim’s view, that led the empiricists tothe erroneous conclusion that those moral truths are only opinion.35

Durkheim alludes here to Hume’s argument that while causality cannotbe valid, it nevertheless exists as an opinion and a habit, and in that guiseis important, since without it we could make no statements of mattersof fact. Durkheim argued that by assuming the individual as a startingpoint the two traditional positions create the same dilemmas in spite oftheir different approaches. Then encountering those dilemmas over and overagain, they assume that they are inherent in the problem of knowledge itself.What they do not realize, according to Durkheim, is that their own indi-vidualist approach has created these problems and made them falselyappear to be insurmountable.

34 I believe that experience here refers to individual experience. The loose use of words isone of the problems. The beginning of the sentence makes it clear that he is setting upa contrast by naming experience in the sentence (in French l’experience). However, heseems to eliminate all experience as a source of the categories when he uses the wordexperience, and to indicate a sociology of knowledge argument, when that is clearly nothis intention.

35 By moral here Durkheim means only to convey a sense or emotional feeling of the socialnecessity created by the group. He does not mean ethical, or just, by moral in this context.


Durkheim’s alternative is to drop the individualist approach. Beginningwith social forms instead of individuals, it becomes possible to arguethat the categories of the understanding develop within social forms asthe emotional experiences shared by participants, of those qualities ofmoral forces, enacted in social forms (constituted in and through ritualparticipation in social forms). This is not an argument that social forms,or words, somehow mystically carry meaning and convey it to persons.It is not a theory of symbols, or a cosmology of ideas. The idea is rather,that it is in the process of creating, enacting, and recreating social forms,that participants feel the social, or moral, force of what they are doing.Individual social beings must be participants in social forms in order formoral forces to be experienced by them as qualities in the social forms.Outsiders and observers will not experience moral forces unless theyparticipate fully.

The fact that the categories develop to fulfill a social need, according toDurkheim, should not mislead us into thinking that they are not empir-ically valid. Social forces are also natural forces and the categories theygive rise to, he says, are valid within social contexts. While acknowledg-ing that the social origin of the categories makes them empirically validonly with regard to social phenomena, Durkheim also wants to claim ameasure of validity with regard to natural phenomena and events. Socialforces, Durkheim ([1912:36]1915:31–2; *1995:17–18) says, are as realas natural forces:

The fact that the ideas of time, space, genus, cause, and personality are con-structed from social elements should not lead us to conclude that they are strippedof all objective value. Quite the contrary, their social origin leads one indeed tosuppose that they are not without foundation in the nature of things.

Durkheim says that there is no reason to think that these particular ideasare stripped of all objective value. He does not say that they have perfectobjective validity with regard to nature, however. He takes a middle posi-tion in that regard. He thinks that they have perfect validity with regardto knowledge of social relations, but that they have only some objectivevalidity with regard to natural objects.36

Because society is part of nature and a natural thing, Durkheim believesthat it must be similar in some sense in its basic parts to the rest of nature.This seems in places to be a sort of rationalist argument that all the parts ofnature, including society, must in some fundamental sense be the same.It could also be a pragmatist argument that social relations could only“work” if they were in tune with nature. In other places, such as the

36 The question of validity will be taken up at length in my Chapter Two.


Conclusion to The Elementary Forms and the 19th lecture on Pragmatism([1913–14]1960), Durkheim argues that the relationship between natureand the categories grows closer and more valid over time.

But Durkheim’s argument for the validity of the categories does notrest on this argument. The claim that social functions are natural onlyapplies to his argument that the validity of the categories can be extendedto natural objects (albeit in a qualified way), because nature and soci-ety are part of one greater reality. The way Durkheim finally goes aboutdemonstrating the empirical validity of the categories, making the cate-gories “translate” the collective experience of enacted practices, does notrest on rationalist, or pragmatist, assumptions. Whether the categorieshave any valid application to nature, and what sort of application theymight have is something that Durkheim himself revisits repeatedly. In the19th lecture on Pragmatism his argument for the validity of the categories,when applied to nature, takes the form of a quasi-pragmatist sociologyof knowledge, rather than a rationalist argument. It is possible that heintends something like that here, but that without sufficient articulationit appears to be based on rationalist assumptions when in fact it is not.Durkheim continues to pursue this point, and it figures heavily in hisdualism.

1.3.0 Concluding Paragraphs

Durkheim does two things in the concluding paragraphs of the Introduc-tion. He reiterates his criticisms of empiricism and apriorism and thentakes up the implications for Ethics of his argument that certain ritualconditions constitute the necessary basis for the development of reason.In the last paragraph the argument turns from a debate with apriorismand empiricism, to an argument with utilitarian moral philosophy andclassical economics. That Durkheim’s interest in establishing a secureepistemological footing for sociology ultimately serves his interests inestablishing Ethics on a secure empirical footing becomes clear. Ethicsin a division of labor society, in which religion is giving way to scienceappears as Durkheim’s overarching concern.

In reiterating his debate with apriorism and empiricism Durkheimwrites ([1912:27–28]1915:32–33; *1995:18) that after his epistemologi-cal argument has completely changed things “the categories cease to beregarded as primary and unanalyzable facts; and yet they remain of suchcomplexity that analyses as simplistic as those with which empiricismcontented itself cannot possibly be right.” The categories are empiricaland analyzable, but also highly complex. Much too complex to have anorigin in sense perception. Durkheim argues ([1912:27–28]1915:32–33;


*1995:18) that “they appear as ingenious instruments of thought, whichhuman groups have painstakingly forged over centuries, and in whichthey have amassed the best of their intellectual capital.”

Durkheim then ([1912:27–28]1915:32–33; *1995:18, emphasis add-ed) reiterates the purpose of his study: “Without making these questionsthe direct subject of my study, I will take advantage of all the opportunitiesthat present themselves to capture at birth at least some of those ideasthat, while religious in origin, were bound nevertheless to remain at thebasis of human mentality.” It is not clear why Durkheim should havebeen so tentative in stating epistemology as his objective. While he saysthat epistemology is not the direct subject of the study, Durkheim also([1912:27–28]1915:32–33; *1995:18) strongly implies that the indirectgoal of the study is to take “all the opportunities that present themselves,”to address epistemological questions through this study of religion and tomake, as he says, “certain fragmentary contributions to that science.”37

Finally, having argued that certain ritual conditions constitute the nec-essary basis for reason and coordinated social life, Durkheim takes up theethical implications of his position. The argument turns from a debatewith Kant and Hume over the origin of the categories of the under-standing, to an argument with utilitarian moral philosophy and classicaleconomics. Utilitarian moral philosophy and classical economics havegenerally treated the rational individual (or some combination of ratio-nal individuals) as the highest good and then tried to figure out howbest to maximize the good for all. Durkheim, like Marx before him,argued that this kind of individualist thinking misses the point.38 With-out the ritual cohesion of the social group according to Durkheim, therewould be no rational individual and no communicable thought in the firstplace.

One implication of Durkheim’s argument for moral philosophy is thatit is a mistake to begin with the individual, or to treat the individual as the

37 Maybe we should read between the lines, as excessive humility is sometimes the guisein which great claims were made. Possibly in his academic circle it was an even moreheretical thing to do than what he actually did in focusing on archaic religious practice.However, it may also be that these disclaimers are another artifact of Durkheim’s havingwritten the Introduction first, at a time when he did not yet realize how extensive hisepistemological argument would turn out to be. Certainly the argument that he did makemet with derision and abuse. Maybe it would have been rejected even more completelyhad he approached the problem more directly. One cannot, however, help wonderingwhether things might have been a little better if people had at least understood moreclearly what he was up to.

38 Marx argued that it was a result of capitalist thinking. Durkheim says that the division oflabor promotes individualism and presumably that leads to placing the individual alwaysat the center. Thus, both point at the development of western industrial capitalism asthe cause of the philosophical reliance on individualism.


highest good. To put it another way, if you take the individual to be thehighest good, then you must recognize that the individual is not a given,but has social preconditions that must be fulfilled. These preconditionsthen become absolute moral goods. Because these preconditions involvepractices, Durkheim’s argument also has implications for a practice con-ception of moral philosophy. In making this argument Durkheim buildson Rousseau. The enactment of certain forms of social practice is nec-essary for the development of human reason. Therefore, for Durkheim,society, in a form best able to sustain stable rational beings, somethinglike Kant’s kingdom of ends, is a necessary good, because without it therewould be no human beings.

For Durkheim, human beings are inherently social beings. Withoutthe necessary social overlay, and by social he necessarily means moral,the biological being remains an animal, and is not recognizably human.The individual is only recognizably human to the extent that it possessesshared categories of thought. In the modern period when money, sci-ence, and the division of labor have, to a large extent, replaced religion,the development of common, or shared, categories has, for Durkheim,become somewhat problematic. Persons may not participate in sufficientshared practice to experience moral force. Or, the practices that createmoral forces for them may be threatened by the spread of global cap-italism. They may develop very different collective representations andvery different practices and can lose their commonality; consequently,discussions result in interpretive dilemmas; a cause of anomie (Rawls1999).

Durkheim worried about modern society, believing that the loss ofmoral unity could lead to anarchy in all areas of human existence thatdepend on the categories. This concern led him to write extensively onmorality and education. He thought it was imperative to create new socialforms that would continue to produce sufficient moral force in mod-ern society, because, from his perspective, without moral force persons,quite literally, lose the foundation for shared understanding and fall intothe situation envisaged by Hume, where the perceptual basis for every-one’s concepts is different, but they believe that they are referring to thesame ideas: the postmodern dilemma. His solution was to rest on sharedcommunities of practice – scientific practice being his primary example.Durkheim’s ethics was to be an ethics founded on a clear understandingof the necessary prerequisites for the human social condition: his episte-mology of practice.

In his concluding paragraph Durkheim ([1912:27–28]1915:32–33;*1995:18) makes sweeping claims with regard to the scope of hisargument:


In this fresh formulation, the theory of knowledge seems destined to join the oppo-site advantages of the two rival theories, without their disadvantages. It preservesall the essential principles of apriorism but at the same time takes inspiration fromthe positive turn of mind that empiricism sought to satisfy. It leaves reason withits specific power, but accounts for that power, and does so without leaving theobservable world. It affirms as real the duality of our intellectual life, but explainsthat duality, and does so with natural causes.

Durkheim claims to have offered a solution to the major philosophicalproblems of his day; both epistemological and ethical. Note the empha-sis on the “observable world ” and “natural causes” in this passage. Whileusually interpreted as having posited a cosmology of ideas, or to haveassumed an inherent rationalism, in this passage Durkheim points explic-itly toward the observable social phenomena that he believes explain theorigin of these ideas and he refers to social causes as natural causes.

The duality of intellectual life referred to in this passage is the distinc-tion between pre-rational animal nature and a rational socially causedmind, combined in the same individual person. This idea of dualism isboth critical to Durkheim’s position and highly misunderstood. It will bethe subject of the next chapter.

2 Durkheim’s Dualism: an Anti-Kantian,Anti-Rationalist Position

One of the misinterpretations standing in the way of an understanding ofDurkheim’s epistemology is the belief that he took a Kantian position ondualism. The misinterpretation is consequential in that it supports theclaim that he also took a Kantian position on epistemology. Durkheimdoes take a position on human reason that he refers to as “dualism.” Thisposition is presented, in conjunction with his discussion of epistemology,in the Introduction to The Elementary Forms, and plays a key role in theargument. In fact, Durkheim felt so strongly about the centrality of hisposition on dualism to the argument of The Elementary Forms as a whole,and was so disappointed that the argument was misunderstood, that, inresponse to criticism of that book, he wrote an article devoted entirelyto an explanation of his position on dualism. The article, “The Dualismof Human Nature and its Social Conditions” was published in 1914,in the journal Scientia two years after the publication of The ElementaryForms.

In the Scientia article Durkheim argued that there are two aspects ofeach human being: a pre-rational animal being and a rational social, orhuman, being.1 These two aspects of the person conflict with one another,producing the internal tension that philosophers across the ages havereferred to as dualism. According to Durkheim ([1914]1960:329):

It is this disagreement, this perpetual division against ourselves, that producesboth our grandeur and our misery; our misery because we are thus condemnedto live in suffering; and our grandeur because it is this division that distinguishesus from all other beings. The animal proceeds to his pleasure in a single and

1 A note on the meaning of the terms human and animals would seem to be required here.All significantly social animals would have many of the qualities that Durkheim ascribesto human beings because of the complexity of their social relationships. They would nothave human reason until they were able to perform rituals that created moral forces. But,the progress toward reason probably occurs on more of a continuum than Durkheimsupposed. The first realization of the sacred would only require the sense of unity andgroup membership that many animal groups clearly demonstrate. They might then besaid to have the first category of unity, or classification, and a basic idea of the sacred,without yet having developed the ritual complexity to produce the others.

72

Durkheim’s Dualism 73

exclusive movement; man alone is normally obliged to make a place for sufferingin his life.

This tension between two aspects of the person is, for Durkheim, whatdistinguishes the human from the animal. It is a necessary condition ofhuman existence. But, unlike Kantian dualism, it has social causes.

In the Scientia article Durkheim asserted that the idea of human dual-ism was a foundational principle of The Elementary Forms and expressedsurprise that the critics had not realized this. This lack of understanding,he says ([1914]1960:325–6), is what prompted him to write the article:“Since the critics who have discussed the book up to the present have not –to our great surprise – perceived the principle upon which this explana-tion rests, it seemed to us that a brief outline of it would be of someinterest to the readers of Scientia.”

The principle is a simple one. Durkheim’s position on dualism allowshim to do two things. First, it allows him to maintain that before thedevelopment of the categories of the understanding, as individual ani-mals, persons had the ability to think in basic ways. What society addsare categories of thought with a universal and necessary dimension. It isthis universal and necessary dimension that makes social thought human,according to Durkheim. Second, it explains the origin of the first moralclassification between the sacred and the profane, which Durkheim willattempt to establish in Book I. Unfortunately, given the persistent mis-understandings of Durkheim’s argument with regard to dualism, and itsidentification with both Kantian and Cartesian rationalism, his insistenceon the importance of the argument may have further confused, ratherthan clarified, the interpretation of his work.

Even Durkheim’s use of the word “dualism” has caused much confu-sion; leading many commentators to believe, simply on the face of it, thathe espoused a Kantian, or Cartesian, position.2 The fact that in present-ing his own position he referred favorably to Kant’s dualism no doubthas added to that impression. But, Durkheim’s position is almost a com-plete rejection of Kant’s. Consequently, the identification of Durkheim’sdualism with Kant’s dualism has created serious problems for the under-standing of Durkheim’s epistemology. The most obvious of these is thatif Durkheim believed, as Kant did, that reason existed in the individualprior to experience, he could not possibly have argued, as he did, that the

2 LaCapra, 1964, seems to be a primary source for this belief. But, there is nowhere inLaCapra’s text where the argument is actually made that Durkheim’s position has anythingin common with Kant’s. LaCapra seems to assume that because Durkheim uses the term“dualism” he is a Kantian. Unfortunately, LaCapra’s views on the matter have been andcontinue to be influential.


categories of the understanding have a social origin. This confusion hasalso obscured Durkheim’s focus on practices as concrete and witnessable,a focus which makes no sense from a Kantian perspective, in which theworld in itself is necessarily unknowable.

Contrary to the general view, Durkheim’s position on dualism is reallyquite opposed to the idea that reason is intrinsic to the human mind;an essential element of both Kantian and Cartesian dualism. While heapplauds their recognition of the fact of dualism, he complains that mostphilosophers, including Kant, have focused only on the tension thatresults from dualism, but have not examined its cause. For Durkheim,the cause of dualism is social and not inherent in the human being. Inan argument reminiscent of Rousseau’s thesis in The Origins of Inequal-ity, that human beings did not develop reason until they became social,Durkheim maintains that participation in the enactment of ritual socialpractices transforms the individual into a rational human being. How-ever, the animal pre-rational nature of the individual remains intact. Itis the tension between these two, the rational social being and the pre-rational individual,3 that constitutes dualism, according to Durkheim, notan inherent tension between the body and the mind, or between spiritualand material reality, as Kant and Descartes had argued.

This distinction between the fact of dualism and its explanation is essen-tial to understanding Durkheim’s position. The distinction runs rightthrough The Elementary Forms, as the distinction between underlyingneeds and retrospective accounts (practices versus beliefs).4 Durkheimaccepts dualism as a social phenomenon, or social fact, and, as such, heexpects to find its causes in underlying social forces. While Durkheimappears to treat the Cartesian and Kantian positions with great respect,and several times compares his own position favorably to Kant’s, thisshould not be mistaken for agreement with their explanations of dualism.In fact, Durkheim treats the arguments of the philosophers as significantonly in so far as they supply narrative historical evidence of the existence ofunderlying social forces. He is not interested in philosophical explanations ofthe phenomenon, per se. Furthermore, because none of the philosophers

3 On Durkheim’s view the pre-rational individual never becomes rational. It remains a pre-rational element within the social being. That is one reason why it is a mistake to beginwith individual knowledge in building an epistemology. The individual biological beingis pre-rational. It is only by becoming social that individuals are transformed into rationalpersons. As individuals they have no experiences of reason.

4 This very important distinction has become important in contemporary social theory,with many studies focusing on the importance of accounts and contexts of accountabilityin maintaining institutional forms of life. But, for some reason the distinction betweenpractices and accounts has never been tied back to Durkheim, who seems to have origi-nated it.


considered underlying social forces in their arguments, Durkheim severaltimes faults them for failing to offer any “explanation” of dualism at all.

Durkheim crafted The Elementary Forms as a sociological explanationfor the phenomenon of dualism, and the relationship of that phenomenonto epistemology. Dualism, in the context of the argument as a whole, rep-resents not only the tension between the individual animal and the humanbeing, caused by the development of the categories, but also the distinc-tion between sacred and profane. In fact, in so far as human dualism is adirect result of the experience of the moral force of the division betweensacred and profane, it is an internalization of that moral division whichwould make dualism the first moral division.

In spite of Durkheim’s expression of surprise that the critics had notrecognized the importance of his argument, however, it would be fairto say that, although it appears throughout The Elementary Forms, inthe guise of various distinctions, particularly the distinction between thesacred and the profane, his own argument is not discussed as such in thatwork. It seems that its substance, as well as its importance, remainedobscure even after his elaboration of it in the Scientia article. Curiously,while Durkheim’s position with regard to dualism offers a serious chal-lenge to both Kantian and Cartesian dualism, he did relatively little todistinguish it from these better known versions of dualism. Even in theScientia article, which was addressed to clarifying his position on dualism,Durkheim presented only a short sketch of the argument. He proceededas if his own argument, with regard to dualism, was clear and the problemwas only that the critics had not understood its centrality to his overallposition. The facts of the matter are quite otherwise, however. The criticshave not been reluctant to call Durkheim a dualist, but they have entirelymisunderstood his argument.

In the Scientia article Durkheim does distinguish his position from theearlier Cartesian and Kantian dualisms to some degree. However, thereis less emphasis on differences, than on what the arguments have in com-mon. The belief in dualism, according to Durkheim, is a very old onethat has appeared in all known societies. The belief is always present, heargues, because the phenomenon is always experienced. Durkheim con-sidered dualism to be a “social fact.” Dualism is an experience that hasnecessarily accompanied human intelligence in all times and in all cul-tures, because the social creation of reason, on his view, not only inevitablycreates a dualism between the individual biological being and the socialbeing, but reason itself, as the result of enacted ritual practice, is basedon a dualism between the sacred and the profane.

All dualist positions recognize this phenomena, and therefore, in avery superficial sense, Durkheim aligns himself with all of them. The


fact that he makes no distinction between various dualisms in this regardshould have tipped off his critics that he meant something entirely differ-ent by dualism. In Durkheim’s special sense, of treating a philosophicalargument as narrative evidence of an underlying social fact, of attendingalways to the relationship between accounts and underlying practices, theCartesians and Kantians, and even Plato, cannot ultimately be wrong inarguing for dualism, because one of the most characteristic peculiaritiesof human nature, according to Durkheim ([1914]1960:325–6), is this“constitutional duality of human nature.”

In fact, what Durkheim criticizes most are theoretical positions thatargue that the experience of dualism does not have some basis in fact. Heargues ([1914]1960:326) that “a belief that is as universal and perma-nent as this cannot be purely illusory.” These are almost the same wordsDurkheim used ([1912:3]1915:14; 1995:2) to justify his serious treat-ment of primitive religious beliefs in The Elementary Forms; “a humaninstitution cannot rest upon error and falsehood.” It was Durkheim’spractice not to dismiss beliefs that people have held over long periods oftime and in many different cultures. But, neither did he embrace the beliefsthemselves as true. He treated such beliefs as evidence of an underlyingsocial reality.

In the case of archaic religions, his affirmation was not of the beliefsthemselves. Rather, he looked for the underlying social causes of thosebeliefs. Durkheim’s interest in philosophical dualism, as he describes it inthe Scientia article, is similarly in the underlying social phenomena thatexplain the persistence of the belief in dualism. He did not accept thephilosophical account of dualism, any more than he accepted the reli-gious account of the purpose of ritual practices, or of the origin of thesacred. However, just as Durkheim did not entirely dismiss the valid-ity of religious beliefs simply because it was impossible for the religiousaccounts of them to be literally correct, so also he does not entirely dismissKantian or Cartesian dualism just because he believes that the accountsthey give are not literally correct.

For Durkheim, the error of these earlier dualisms was in not realizingthat dualism had a concrete social explanation. The true philosophicalissue for Durkheim, is to explain where dualism comes from historically,and why it characterizes only the human condition.5 That dualism has asocial origin is really a huge and controversial claim, however. By positinga social explanation for dualism, Durkheim challenged the entire fabric of

5 It is a philosophical issue and not a sociological issue, even though it requires a sociologicalexplanation, because it has to do with the origins and validity of human knowledge.Durkheim’s overall intention seems to be to address philosophical issues sociologically,not to abandon philosophical issues for sociological issues.


both Cartesian and Kantian dualism. The question remains why Durkheimdid not state this challenge more directly.6 He may have assumed that thecontradiction between his social explanation of dualism and the ratio-nalist explanations of Cartesian and Kantian dualism was sufficientlyobvious without further elaboration. However, most critics have missedthe contrast entirely. The dualism that earlier philosophers took to bea real apriori condition of the human intellect, or of the universe itself,Durkheim treated as an experienced artifact of the socially constructedcharacter of the categories of the understanding. The positions could notbe more different.

Great care must, therefore, be taken in reading Durkheim’s text. Itis highly unusual for a critic to treat great works of philosophy as folk beliefs;accounts, that provide evidence of underlying social relations; social facts.His expressed respect for these texts is no different from the respect thathe accords to archaic religious beliefs; both are treated as indicationsof an underlying social reality, and “true” only in so far as they reflectthat reality. Durkheim certainly did not “believe” in archaic religiousbeliefs. Neither did he “believe” in the philosophical arguments of Kantor Descartes. Only a careful reading will reveal his fundamental disagree-ments with the classic philosophical arguments and the important trans-formation that his dualism effects.

Taking the Introduction to The Elementary Forms and the Scientia articletogether, Durkheim’s arguments with regard to dualism can be charac-terized as follows: First, Durkheim argues that the biological individ-ual is not possessed of reason, whereas the social individual is. ThisDurkheim refers to as “double man” or “homo duplex.” He compareshis argument, in this regard, to the classic debate between empiricismand apriorism, and also to the religious belief in a distinction betweenbody and soul, arguing that both distinctions have the same origin;Second, Durkheim must address the issue of whether or not socially gen-erated categories of the understanding create an insurmountable dualismwith regard to knowledge of the natural world. He addresses the ques-tion whether socially generated categories could have empirical valid-ity with regard to natural objects and events? This question involveshim in an extended consideration of the classical philosophical argu-ment in this regard; Third, the first two discussions lay the foundationfor Durkheim’s argument that the idea of the Sacred and Profane hasits origin in the moral force of dualism itself. That is, as the person

6 This is similar to the question raised with regard to the Introduction, of why he did notstate his epistemological objectives more directly. In neither case is there a satisfactoryanswer.


becomes a social being, they feel that one part of themselves has becomesacred while the other part, the individual part, remains separate andprofane.7

Finally, Durkheim also needs to explain what it is about social relationsthat enable them to transform the person into a rational being. While TheElementary Forms was in its entirety an explanation of this, in the Scientiaarticle, Durkheim attempts to summarize his position and in the courseof doing so, reproduces some of the same problems that appear in theIntroduction and Conclusion of The Elementary Forms. As these summaryarguments have been enormously problematic, this argument also will bereviewed.

Although there is some overlap, Durkheim works these positions outdifferently in the Introduction to The Elementary Forms and in the Scientiaarticle. In particular, the Scientia article takes up the philosophical cri-tique in much greater detail, and therefore, is of interest as a record ofthe relationship between Durkheim’s position and other philosophicalpositions, as he saw it.

2.1.0 Durkheim’s First Distinction: “Double Man”

Durkheim ([1912:21–23]1915:28–29; *1995:15) first introduces the ideaof dualism toward the end of his Introduction to The Elementary Forms.Here he introduces the problem of the validity of socially generatedknowledge when applied to natural objects and relations. A discussionof what Durkheim refers to as the “double man” then follows. Accord-ing to Durkheim ([1912:23]1915:28–29; *1995:15), “Man is double. Inhim are two beings: an individual being that has its basis in the body andwhose sphere of action is strictly limited by this fact, and a social beingthat represents within us the highest reality in the intellectual and moralrealm that is knowable through observation: I mean society.” Earlier onthe same page Durkheim tied the idea of dualism to his overall claimthat “Society is a reality sui generis,” maintaining that the two differentlayers of knowledge are not reducible to one another. The individualstates “are wholly explained by the psychic nature of the individual.”While on the other hand, he argues, the categories of the understand-ing can be completely and entirely explained on the basis of socialphenomena. “Between these two kinds of representations,” he writes([1912:22]1915:28–29; *1995:15), “is all the distance that separates the

7 As noted earlier, this emergent feeling of the “sacred” might explain animal behaviorwherein it is clear that the individual has sacrificed themselves, or put themselves at risk,in order to save the group, or even another individual member of the group.


individual from the social: one can no more derive the second from thefirst than one can deduce society from the individual, the whole fromthe part, or the complex from the simple. Society is a reality sui generis.”

In the Introduction, as in the Scientia article, Durkheim points out thathis position on dualism has parallels in the Kantian tradition, in variousclassical rationalist arguments, and even to some extent in empiricism.He ([1912:23]1915:29; *1995:15) makes several favorable comparisonsbetween his own position and that of Kant, introducing his thesis of thedouble man, for instance, with the phrase “as the well known formulahas it,” followed by a reference to apriorism. Durkheim says ([1912:21]*1915:28; 1995:14) about the dualism between two sorts of knowledgethat “our hypothesis keeps this [apriorist] principle intact.”

However, Durkheim’s dualism involves the idea that the categories ofthe understanding, which have a social origin, exist in a perpetual tensionwith the knowledge originating in the five senses that characterizes theindividual pre-rational being. For Kant, on the other hand, the dualismwithin the individual is caused by the fact that the innate rationalism ofthe mind is at war with the inclinations of the body. The fact that Kantposits the categories of the understanding as innate also results in a splitbetween the world of experience and the world in itself.

Durkheim’s ([1914]1960:325) argument, as presented in Scientia,begins with the statement that “society can exist only if it penetrates theconsciousness of individuals and fashions it [individual consciousness]in ‘its image and resemblance.’” Society must, necessarily, take it as apriority to create collective mental states in its members. Groups thatdo not orient toward this priority will not survive. Therefore, Durkheimargues ([1914]1960:325), we can state as a matter of necessity that “agreat number of our mental states, including some of the most importantones, are of social origin.”

In an argument reminiscent of Rousseau in The Origins of Inequality,Durkheim says ([1914]1960:325), that “it is civilization that has mademan what he is; it is what distinguishes him from the animal.” Rousseauhad argued that man in a state of nature had no need for the inclinationsand desires displayed by civilized man. Reason had no survival valuefor the animal. Therefore, Rousseau argued, the “rational” condition ofcivilized man was a social product. The “noble savage” was free andindependent. Civilized man, on the other hand, traded his independencefor social needs, and then developed reason in order to pursue thoseneeds. Morality was also said by Rousseau to be a social creation and didnot apply to man as an animal.

In a similar vein, Durkheim argues ([1914]1960:5) that because themost fundamental ideas that constitute human reason are of social origin,


“the causes and conditions on which civilization depends” are the samething as “the causes and conditions of what is most specifically human inman.” If we change this to say that the causes and conditions on whichsociety depends, which Durkheim argues are found in religious ritual,are the same thing as the causes and conditions of human reason, wehave the argument of The Elementary Forms. In fact, the entire claim thatsociety is a reality sui generis, along with the epistemological argument ofThe Elementary Forms, rests on this argument.

What is most specifically human in man, in Durkheim’s analysis, is theresult of social practices: the categories of the understanding, the collec-tive representations they give rise to, and the dualism of human naturethat possessing categories and collective representations creates. Eventhough this dualism is in some sense experienced as “psychological,” itscauses are the same as those of civilization, and therefore, the explana-tion of “what makes up man,” according to Durkheim ([1914]1960:325),requires an historical, and not a psychological analysis. The essentialcharacteristic of human nature is this dualism: the socially derived cate-gories and collective ideas of the social being versus the particular sen-sations and generic representational abilities belonging to the biologicalindividual.

Society, for Durkheim does not consist merely of contingencies thatare unacceptable as a foundation for epistemological argument. Rather,what are for philosophers contingent social forms, are for Durkheim nec-essary social relations that are identical to human reason. This is anextremely important point and speaks to Durkheim’s claim that societyobeys natural laws. There are prerequisites for the human condition that,he argues, must be met, or rational beings will cease to exist. Eliminatingthe contingencies of individualism, while at the same time embracing theapparent contingencies of social practices, is the only way to establishsocial relations as a foundation for ethics, which is Durkheim’s ultimateaim.

In the Scientia article, Durkheim describes The Elementary Forms ashaving provided an elucidation of this issue. He writes ([1914]1960:325)that, “in attempting to study religious phenomena from the sociologicalpoint of view, we came to envisage a way of explaining scientifically oneof the most characteristic peculiarities of our nature.” This characteristicpeculiarity is what Durkheim refers to as “the dualism of human nature.”The epistemology presented in The Elementary Forms was intended toprovide a scientific explanation of dualism, by providing an empiricalexplanation for the origin of the categories of the understanding in socialforces. Durkheim expresses surprise that the critics have not picked upon this fact.


2.1.1 The Dualism of Empiricism and Apriorism

Durkheim presents his dualism, in the Introduction, as something ofa combination of both empiricism and apriorism. Roughly speaking, itcould be said that Durkheim credits empiricism with providing a descrip-tion of the individual biological being, while he credits apriorism withproviding a description of the social being. Throughout the Introduction,Durkheim moves back and forth between descriptions of empiricism andapriorism in trying to explain his position. Empiricism is made to repre-sent the individual level of knowledge, and apriorism is made to representthe social. This is, in an important sense, what Durkheim means by say-ing that his position combines the strengths of the two schools of thought,while avoiding their weaknesses. He does not mean that he accepts anypart of either argument, per se.

For Durkheim, the biological being is not recognizably human. To behuman, and rational, is to first be social. But, before the social beingdevelops there is a prior animal being. This individual animal being doesnot go away when persons become social. Durkheim makes a distinctionbetween two persons; one that has its basis in the body, and a social beingthat represents society within us. According to Durkheim ([1912:24]*1915:30; 1995:16), the biological being alone, without the developmentof the categories of the understanding, is not recognizably human: “Doesa mind ostensibly free itself from these forms of thought. It is no longerconsidered a human mind in the full sense of the word, and is treatedaccordingly.” In order to be possessed of human reason it is necessaryfirst to become social. The pre-social individual is just like all other ani-mals. Durkheim ([1912:626]*1915:487; 1995:440) returns to this ideain the conclusion of The Elementary Forms: “A man who did not thinkwith concepts would not be a man, for he would not be a social being.If reduced to having only individual perceptions, he would be indistin-guishable from the beasts.” Given the empiricist focus on individual per-ceptions, the implied criticism of empiricism is clear.

Durkheim does not deny that pre-social individuals exist, or that insome sense they think. The pre-social individual has the capacity to sortsense impressions and form generic representations. What Durkheimargues is, that in their capacity as individuals, their experiences andthoughts are quite different from those of persons actively engaged ina social unit. Durkheim ([1912:21]1915:28;*1995:14–15) allows for abasic empiricism within individual perception that is not valid knowledge,but nevertheless in touch with reality: “In fact, that knowledge which iscalled empirical, the only knowledge of which the theorists of empiricismhave made use in constructing the reason, is that which is brought into our


minds by the direct action of objects. It is composed of individual stateswhich are completely explained by the psychic nature of the individual.”8

Several critics have argued that Durkheim’s epistemology is circu-lar because persons would have to be able to perceive differences inorder for the social construction of moral differences to have any effecton them. But, these critics have misunderstood Durkheim’s argument(Rawls 1997). In Book Two, Chapter Three, of The Elementary Formswhere Durkheim discusses the social origin of the category of classifi-cation, he ([1912:206]*1915:170; 1995:146) elaborates further on theidea that the individual biological being has the ability to make genericrepresentations:

It is not our intention to deny that the individual intellect has of itself the powerof perceiving resemblances between the different objects of which it is conscious.Quite on the contrary, it is clear that even the most primitive and simple classifi-cations presuppose this faculty.

Durkheim allows that individual perception includes a natural abil-ity in the pre-social individual to perceive opposition between things.He argues ([1912:206]*1915:170; 1995:146) for instance, that the divi-sion of things into “opposites” appears in so many societies because ahigh degree of contrast leads to a high degree of natural visibility and“intuitions” and “feelings of affinity or of repulsion” for the contrasts.He says ([1912:206]*1915:170; 1995:146) that there “is a certain intu-ition of the resemblances and differences presented by things,” and thatthis has played an important part “in the genesis of these classifications [ofopposites].” In fact, the ability to perceive opposition would have to existbefore the first idea of moral force, the sacred and profane (the mother ofthe other categories), could create the moral feelings about oppositionsand develop those feelings into a sacred organization of classifications.

One of the most influential early critics of Durkheim’s epistemology,William Dennes (1915), maintained that his epistemological argument iscircular because the categories are supposed to come from social experi-ence, but social experience would not be possible unless persons alreadyhad the categories. He argued that they would not be able to feel themoral force of social distinctions unless they could perceive distinctions,per se. He is right. But, Durkheim’s argument is not vulnerable to this

8 The 1915 translation reads: “The knowledge that people speak of as empirical – all thattheorists of empiricism have ever used to construct reason – is the knowledge that thedirect action of objects calls forth in our minds.” The French edition of 1912 reads:“En effet, les connaissances que l’on appelle empiriques, les seules dont les theoriciensde l’empirisme se soient jamais servi pour construire la raison, sont celles que l’actiondirecte des objects suscite dans nos esprits.”


criticism. It is Durkheim’s argument that prior to the development of thecategories, individuals are capable of forming generic ideas based on feel-ings of affinity and repulsion, and contiguity and resemblance. Thereforethey would be able to perceive the required distinctions. These genericrepresentations have no empirical validity. But, nevertheless, these basicabilities which precede the development of valid categories of thoughtare useful, in much the same way that Hume thought they were. BecauseDurkheim’s dualism includes this dimension of individual perception,the argument is not circular.9

The biological individual is limited to the impressions they can getfrom sensation, combined with a basic ability to sort those sense impres-sions. This allows them the ability to form “generic representations.”But, these generic representations are like Hume’s general ideas and onlyappear to have objective validity through custom and habit. Accordingto Durkheim ([1912:19]*1915:26; 1995:13): “A sensation or an imagealways relies upon a collection of objects of the same sort, and expressesthe momentary condition of a particular consciousness; it is essentiallyindividual and subjective.”

Durkheim’s argument is that only the social part of the humanbeing, the part that only exists after the experience of moral force,has empirically valid categories of the understanding. For Durkheim([1912:208]*1915:171–2; 1995:147) the biological individual has onlyrudimentary abilities which could never supply the framework forknowledge:

The feeling of resemblances is one thing and the idea of class another. The class isthe external framework of which objects perceived to be similar form, in part, thecontents. Now the contents cannot furnish the frame into which they fit. Theyare made up of vague and fluctuating images, due to the superimposition andpartial fusion of a determined number of individual images . . . the framework,on the contrary, is a definite form, with fixed outlines, but which may be appliedto an undetermined number of things, perceived or not, actual or possible.

The category of classification is not the same thing as the “feelingof resemblances.” The category itself provides a framework whichcan then be applied to sensation. The sensations themselves remainparticular and disconnected. Durkheim ([1912:21]*1915:28; 1995:14)reiterates Hume’s empiricist argument that feelings of affinity and repul-sion by themselves cannot lead to the development of the category ofclassification: “So between the two sorts of representations there is all

9 In so far as the first moral distinction is dualism, Durkheim also avoids the problem ofcircularity. That distinction comes from a moral feeling, and therefore, does not requireeither a general idea, or a category to precede it.


the difference which exists between the individual and the social, andone can no more derive the second from the first than he can deducesociety from the individual, the whole from the part, the complex fromthe simple.”

The empiricists treated these feelings of affinity and repulsion as theonly possible origin of empirically valid knowledge. For Durkheim, how-ever, these natural abilities of the individual have no empirical validity.While possession of such feelings and ideas may be a necessary prereq-uisite for the development of rudimentary social relations, such feelings,on Durkheim’s view ([1912:19]*1915:27; 1995:13), cannot give rise toempirically valid categories of the understanding, but only to an aware-ness of feelings of resemblance and contiguity, as they do for Hume.

Under these conditions forcing reason back upon experience causes it to disap-pear, for it is equivalent to reducing the universality and necessity which charac-terize it to pure appearance, to an illusion which may be a useful practicality, butwhich corresponds to nothing in reality; consequently it is denying all objectivereality to the logical life, whose regulation and organization is the function of thecategories. Classical empiricism results in irrationalism; perhaps it would even befitting to designate it by this latter name.

Durkheim accepts the limits that Hume’s empiricism places on knowledgeavailable to the individual. But, he argues that not all knowledge has itssource in the individual. Participation in enacted practices with an assem-bled group, allows the individual to transcend those limitations. On theother hand, while Durkheim posits shared categories of the understand-ing as the solution to the problem, his argument is not Kantian.

A Kantian position assumes that the categories are innate and precedeall understanding. Durkheim’s position differs from Kant on two essen-tial points. First, he allows that crude abilities to perceive contiguity andresemblance, as well as crude social behaviors, precede the developmentof the categories and; Second, he argues that the categories are social inorigin. Thus, for Durkheim, the categories are empirical historical neces-sities, but not logical necessities. Social development requires them. Butthe survival of the individual being could do well enough without them.Reason itself, and the imperatives of reason, only come after the devel-opment of the categories.

While Durkheim’s dualism is generally considered evidence that hetakes a Cartesian or Kantian position, because it sets up a dualismbetween rational man and physical man, in fact, Durkheim argues thatthe biological individual is not capable of the exercise of reason: not aCartesian or Kantian position at all. Only in its capacity as a social beingdoes the human possess reason, according to Durkheim; and then the


reason that they possess is emotional and bodily, not conceptual in ori-gin. Furthermore, the Cartesian and Kantian arguments both treat rea-son as a given that is completely incompatible with the organization (orlack thereof) of the “corruptible” world of nature. They would neverposit reason as dependent on what they saw as the “contingencies” ofsocial relations. For Durkheim, however, some aspects of the social weresufficiently necessary to provide a universal and noncontingent basis forcategories of the understanding.

2.1.2 The Dualism of Body and Soul

Persons in all ages have, Durkheim says ([1914]1960:326), “beenintensely aware of this duality.” Persons have, he says ([1914]1960:326),“everywhere conceived of [themselves] as being formed of two radicallyheterogenous beings: the body and soul.” He then elaborates on the dif-ferences between the ideas of body and soul that give this relationshipa dualistic quality. Durkheim does not mean that he himself believesthat dualism really consists of a distinction between body and soul. Heargues that persons in all ages have represented dualism in this way. Justas Durkheim argues that religious beliefs do not literally represent theunderlying reality that explains the origin and purpose of religion, butnevertheless, should be treated as evidence of that underlying reality, soalso he treats the beliefs about dualism, not as literal representations of anunderlying dualism, but, rather as evidence of an underlying social phe-nomenon. Therefore, the fact that Durkheim discusses dualism in termsof body and soul, at this point in his argument, does not mean that hehimself believes that this is what the dualism “really” consists of as anunderlying social cause.

Rather, the persistent beliefs about dualism are treated by Durkheimas evidence of something in the essential nature of the human being thatrequires an explanation. Durkheim does not assume that the causes ofdualism are identical with the beliefs about it. In fact, he enters into anextensive criticism of the various beliefs about dualism. In a manner thatparallels his consideration of religious anthropology in The ElementaryForms, he faults the philosophers who hold these beliefs for not havingbothered to ask what the underlying reality is that explains their beliefsin dualism. The beliefs and narratives about dualism (reason/matter,mind/body, body/soul), like religious beliefs, are, for Durkheim, only anindication of underlying causes. They are socially constructed accountsthat serve particular social purposes. They do not reveal the underlyingsocial facts. Nor do they explain what the dualism really consists of. Infact, they tend to distort and obscure that understanding.


Following a discussion of dualism as having been historically repre-sented in terms of the division between body and soul, Durkheim argues([1914]1960:326) that “a belief that is as universal and permanent as thiscannot be purely illusory. There must be something in man that gives riseto this feeling that his nature is dual.” He does NOT say that he believesthat the underlying cause of this idea is an actual distinction betweenbody and soul. He says only that the belief cannot be “purely illusory.”

2.1.3 Dialectical Relation Between Empirical and ConceptualAspects of Being

While he does not accept that the experience of dualism represents anactual distinction between body and soul, Durkheim argues that therereally are two aspects of each human being that correspond to the ideasof body and soul. The body he relates to sensory input and individ-ual experience. The soul he relates to conceptual thought and social, ormoral, aspects of experience. He says ([1914]1960:327) that these twoparts of the human self each “represents a separate pole of our being,”and that, “these two poles are not only distinct . . . but are opposed toone another.”

As he did in the Introduction to The Elementary Forms, Durkheimcontrasts the individualistic, or empiricist mode of perception with theapriorist or rationalist mode of thought. The individual pole, he says([1914]1960:327), includes “sensations and sensory tendencies.” The“sensory appetites,” he says ([1914]1960:327), “are necessarily egoistic.”He argues that ([1914]1960:327) even after we become social beings,“When we satisfy our hunger, our thirst, and so on, without bringing anyother tendency into play, it is ourselves, and ourselves alone, that we sat-isfy.” It may not be often that we follow purely egoistic urges without as hesays “bringing any other tendency into play,” but we retain the capacityto do so.

Sensations have an entirely different character from the experience ofthe categories. Sensations are particular, and as such, not transferrable(i.e. communicable). In their particularity, according to Durkheim, the“sensation” cannot be detached from the “organism.” He says ([1914]1960:327) that “A sensation of color or sound is closely dependent on myindividual organism, and I cannot detach the sensation from my organ-ism. In addition, it is impossible for me to make my awareness pass overinto someone else.” Other persons can see the same things that we see,but they do not have access to our sensations. This level of individualexperience is, therefore, essentially incommunicable.


Durkheim contrasts perception with the role played by concepts inhis sociology of knowledge. Concepts on the other hand, Durkheimargues ([1914]1960:327), “are always common to a plurality of men.”10

Durkheim argues ([1914]1960:327), that concepts “are constituted bymeans of words” and that words and language are not the words of a singleperson, but rather “the result of a collective elaboration.” Because con-cepts are held in common, they are impersonal. They are shared amongmembers of a social group. According to Durkheim ([1914]1960:327),“By means of them minds communicate.” Persons may individualize andpersonalize the collective representations of their language to some extent,but the individual is not the author of these concepts.11

Durkheim ([1914]1960:327) treats the contrast between sensationsand concepts as parallel to the personal and impersonal, concluding, atthis point in his argument, that these two aspects of “our psychic life are,therefore, opposed to each other as are the personal and the impersonal.”Durkheim ([1912:23]1915:29*1995:16) had argued in the Introductionto The Elementary Forms that: “In the realm of practice, the consequenceof this duality is the irreducibility of the moral ideal to the utilitarianmotive.” The consequence of dualism “in the realm of thought” on theother hand “is the irreducibility of reason to individual experience. As partof society, according to Durkheim ([1912:23]1915:29; *1995:16), theindividual naturally transcends himself, both when he thinks and when heacts.” For Durkheim the utilitarian motive is individual, whereas moralityis inherently social.12 The personal and the impersonal have become thetwo new poles of the argument for Durkheim; replacing empiricism andapriorism; the body and the soul.

Durkheim describes ([1914]1960:327) two beings in each person, oneof whom “represents everything in relation to itself and from its own point

10 Durkheim argues here as though he were a nominalist, and it is important that in theensuing discussion where he says that words have only conventional meanings, he isNOT talking about the categories.

11 Unfortunately, in referring to the distinction between concepts and sensations, at thispoint in the article, Durkheim does not distinguish between collective representations,which are constituted by means of words, and categories themselves, which have theirorigin in common experience. This is a distinction that Durkheim only makes at theend of the article. He also does not address the question of where the first collectiverepresentations, or categories, come from.

12 There is some question as to whether John Stuart Mill ([1861]1998) thought of utilityas an individual motive, however. He may very well have thought of it more as a socialnecessity, something without which the community as a whole would not survive. How-ever, he did think of it as an idea that could be grasped by the individual mind, and notsomething that required the transforming effects of society first. Although there again,he often referred to the society of other persons as the greatest good.


of view.”13 Then there is a second being “which knows things sub specieaeternitis, as if” he says ([1914]1960:327), “it were participating in somethought other than its own.”14 This being, he says ([1914]1960:327),“tends to accomplish ends that surpass its own.” Where the first beingtakes itself always as the end or goal of action, the second being holds endsthat are shared and outside of itself. The social being seeks common endsas opposed to individual ends. Durkheim concludes ([1914]1960:327)that “the old formula of homo duplex is therefore verified by thefacts.”

Durkheim describes ([1914]1960:328) the dualist character of thehuman being as “a double center of gravity.” According to Durkheim([1914]1960:328), persons cannot pursue external “moral ends withoutcausing a split within ourselves, without offending the instincts and thepenchants that are the most deeply rooted in our bodies.” This argumentappears to be Kantian, paralleling Kant’s argument that moral actioncontradicts bodily impulses. Durkheim also refers explicitly to Kant atthis point. He says ([1914]1960:328) “as Kant has shown, the law ofduty cannot be obeyed without humiliating our individual, or, as he callsit, our ‘empirical’ sensitivity.”

However, for Durkheim, this tension has a different source than it doesfor Kant. For Durkheim, it is the contradiction between the individualand the social being that creates the tension between moral and individualinclinations, not the contradiction between body and soul. Durkheimasks ([1914]1960:328), “How can we belong entirely to ourselves, andentirely to others at one and the same time?” This contrast between theindividual and the social is not at all the same thing as Kant’s distinctionbetween the pure rational being and the physical being. For Kant, boththe rational and the physical being are individual. Individual reason onlyobeys moral imperatives because it is purely rational and understandsthat not to be moral is a logical contradiction.

For Durkheim, by contrast, it is the social being and not the individualthat is moral and rational and thereby owes a duty to others. The indi-vidual per se, is not rational and has no moral duties. Furthermore, themoral duty of the social being is not a pure rational duty, as it was forKant. Rather, because its essence as a social and moral being dependson its enacted relationship to, and with, others in a society, the social

13 In fact, in order to be really consistent Durkheim would need to be arguing that thisbeing doesn’t represent anything to itself conceptually. It may make up some generaliza-tions with which to represent its sensations to itself, but these wouldn’t be concepts asDurkheim means that term.

14 There may be some parallel to Marx’s idea of species being here. Both involve the ideaof identification with the species rather than with individual ends.


being has a duty to those social forms, and to those assembled others,that make its existence possible. This duty is both moral and rational.Because society is the necessary condition of its existence, the rationalbeing must preserve society.

Durkheim argues that these two poles of the self live in a sort of dialec-tical relationship with one another. The individual must “be” in orderto experience, but it must reach outside of itself, to society, in order tothink. The individual and social are contradictory states of being. But,both are necessary. Durkheim asks ([1914]1960:328) “to what wouldconsciousness be reduced if it expressed nothing but the body and itsstates?”

For Durkheim ([1914]1960:329), thinking requires concepts. There isa problem according to Durkheim ([1914]1960:329), however, because“sensory reality is not made to enter the framework of our concepts spon-taneously and by itself.” Sensory experience has to be altered to make itfit into conceptual frameworks. Hume showed that this is true no mat-ter what the concepts are. Only by removing the content from expe-rience can it be rendered in conceptual form. According to Durkheim([1914]1960:329), “our concepts never succeed in mastering our sensa-tions and in translating them completely into intelligible terms.” In fact,he argues ([1914]1960:329), “they take on a conceptual form only bylosing that which is most concrete in them.”

What Durkheim is describing here is the problem of empiricism; thatconcepts cannot be made out of empirical sense experience. It wasin order to solve this problem that Kant had argued that reason wasinnate and thereby created the mind body dualism that he is known for.Durkheim, however, deals with the same issue in two very different ways.First, through his sociology of knowledge he is able to argue that soci-ety makes up collective representations (these are not categories). Whilethese do some violence to individual experience, they provide conceptswith which persons can communicate with others.15 However, withoutthe categories these collective representations would be purely nominal,completely out of touch with natural reality, and therefore, incommunica-ble. Therefore, Durkheim’s epistemology offers a Second approach to theproblem. Through his epistemology, via categories that are not concepts,but direct expressions of social experience, Durkheim not only securesempirical validity of a sort for the categories, explaining the possibility

15 The categories are quite different in this regard, because there are experiences thatcorrespond directly to them, so it is important to note that in this paragraph he is nottalking about the categories of the understanding. In fact, in this particular passage hewrites as though there were no empirically valid categories. But, at the end of the articlehe will distinguish the categories from collective representations in general.


of intersubjective communication, but also puts them into some form ofcommunication with the world of nature.

Because of the contradictions between individual perceptions and con-cepts, and also between the individual and society, Durkheim argues([1914]1960:329), that “we cannot simultaneously satisfy the two beingsthat are within us.” According to Durkheim ([1914]1960:329), “it is thisdivision that distinguishes us from all other beings.” The animal livesin a present moment, while human beings must live a dual and contra-dictory existence. Their individual sensible aspect lives in the present,while their social self lives in a moral universe of goals and duties. Onceagain, treating the philosophers’ views as a folk narrative, Durkheim says([1914]1960:329–30), “thus the traditional antithesis of the body andthe soul is not a vain mythological concept that is without foundationin reality. It is true that we are double, that we are the realization of anantinomy.”

The myth, like religious myths, is not vain, but represents an underlyingreality. While the traditional Kantian narrative is not wrong and repre-sents something, however, Durkheim argues ([1914]1960:330), that thequestion of where this duality comes from, what it really represents, andwhy it arises remains unanswered by the philosophers and the psycholo-gists. The explanation, he claims ([1914]1960:330), belongs to sociology:“If this odd condition is one of the distinctive traits of humanity, the sci-ence of man must try to account for it.” This is another very bold claimabout the relationship between sociology and philosophy. Not only doesDurkheim lay claim to epistemology as properly a sociological and not aphilosophical issue, but, in the context of his consideration of the prob-lem of dualism, Durkheim also lays claim to the contradiction betweenconceptual and sensible experience, as a properly sociological, and not aphilosophical problem.

2.2.0 Second Distinction: “Two Layers of Knowledge”

Durkheim’s assertion of a dualism within the individual human being,between two ways of knowing, leads inevitably to the question of whetherthe two forms of knowledge; natural/sensation and social/emotion, canbe validly applied to all phenomena, or only to those which gave riseto them. The essential question is whether socially derived categories ofthe understanding have any valid application to natural phenomena? Orwhether they validly apply even to social phenomena other than thoserituals which gave rise to them? Durkheim takes up this question, at thesame place in his Introduction where he first introduces the thesis of the


double person. It is a problem that Durkheim will also take up moreextensively in his concluding chapter.16

2.2.1 The Question of Validity

Durkheim introduces the problem of validity in terms that directly refer-ence Kant’s position, calling it ([1912:21]1915:28; *1995:14) “The fun-damental thesis of apriorism.” Durkheim says that his own hypothesiskeeps this fundamental principle of apriorism intact (because of differ-ences I give both English translations):

The fundamental proposition of the apriorist theory is that knowledge is made upof two sorts of elements, which cannot be reduced into one another, and whichare like two distinct layers superimposed one upon the other. Our hypothesiskeeps this principle intact ([1912:21]1915:28).

The fundamental thesis of apriorism is that knowledge is formed from two sorts ofelements that are irreducible one to the other – two distinct superimposed layers,so to speak. My hypothesis keeps this principle intact ([1912:21]1995:14).17

According to Durkheim the argument that knowledge is made up oftwo sorts of elements is a fundamental principle of apriorism. In fact,he writes in a footnote to this passage, that dualism is more fundamen-tal to apriorism than innatism. When Durkheim goes on to say that his“hypothesis” keeps that principle intact, he refers only to the fact ofdualism, however, and not to explanations such as Kantian innatism,rationalism, or individualism. Durkheim never says that he accepts theKantian position, only that he keeps the fundamental principle of dualismintact.

As Durkheim outlines the different characteristics of his two sorts ofknowledge, however, they turn out to be very different from Kant’s “twolayers.” Whereas, for Kant, the difference is between two worlds, theworld in itself and the world as presented by human reason; the differencebetween reason and matter, or body; for Durkheim, the two layers consist

16 This issue is also discussed in the 19th Lecture on Pragmatism, in Durkheim’s lectureson Pragmatism published as Pragmatism and Sociology ([1913–14]1983).

17 I have provided quotations from both translations because they differ somewhat. Inthe original French the passage reads: “La proposition fondamentale de l’apriorisme,c’est que la connaissance est formee de deux sortes d’elements irreductibles l’un al’autre et comme de deux couches distinctes et superposees. Notre hypothese main-tient integralement ce principe” ([1912:21]). In a footnote to this passage Durkheimanticipates the criticism that he has identified apriorism with dualism and not withinnatism. He dismisses this objection, arguing that innatism plays only a secondary rolein apriorism. He calls innatism ([1912:21]1995:14) “a simplistic way of portraying theirreducibility of rational cognition to empirical data.”


of individual versus collective forms of knowledge. One layer, accordingto Durkheim, consists of individual knowledge derived from perception,while the other layer of knowledge is processed through the categories ofthe understanding, which are social and collective in origin.

Kant distinguished the world in itself from the world as experiencedthrough the faculty of reason. For Kant there is no individual expe-rience that is not processed through this faculty or through the intu-itions. Durkheim, by contrast, distinguished between individual percep-tion of the empirical world, and the social world as experienced by socialbeings, through socially generated categories of the understanding. ForDurkheim, the two types of knowledge are somewhat independent of oneanother. But both can occur simultaneously.

Although the two types of knowledge, for Durkheim, coexist in a sin-gle human being, he argues ([1912:22]1915:29; *1995:15) that there is agreat distance between the two layers of knowledge: “between these twokinds of representations, then, is all the distance that separates the individ-ual from the social.” Social phenomena are not an aggregate of individualphenomena. Individual experiences are only meaningful to the individ-ual who experiences them. Social phenomena are immediately meaning-ful to the entire group. With regard to individual experiences, thereforeDurkheim argues ([1912:15ff]*1915:23ff; 1995:10ff), man is like otheranimals; by contrast, social experiences, like social time, are common toall members of the group:

Thus we see all the difference which exists between the group of sensations andimages which serve to locate us in time, and the category of time. The first are thesummary of individual experiences, which are of value only for the person whoexperienced them. But what the category of time expresses is a time common tothe group, a social time, so to speak. In itself it is a veritable social institution.Also, it is peculiar to man; animals have no representations of this sort.

The individual animal has memory, can experience before and after andeven anticipate events. But, they have no mutually intelligible social rep-resentations of time.

While Durkheim considers social forms to be distinct from individualforms, he also considers social forms to be natural forms that have natu-ral functions. Social representations express the natural forms of society.“Society is,” as he says ([1912:22]1915:29; *1995:15) “a reality suigeneris; it has its own characteristics that are either not found in the restof the universe or are not found there in the same form. The represen-tations that express society therefore have an altogether different contentfrom the purely individual representations.” Individual and social timeare not the same thing.


If society is a separate form of reality, however, and the categorieshave their origin in this reality, then even if we accept that they arenatural, the question of whether they have an empirically valid appli-cation outside of that reality must be addressed. As Durkheim puts it([1912:25]1915:31; *1995:17): “But if the categories at first do no morethan translate social states, does it not follow that they can be appliedto the rest of nature only as metaphors?” The question of whetherthe categories are merely metaphorical and artificial, purely a matterof social convention, must be considered. Durkheim poses this as arhetorical question, however. He does not think that the categories onlyhave a metaphorical application to natural phenomena. He writes that([1912:25]1915:31; *1995:17) if we think they are limited in this waythen “we would thus return to nominalism and empiricism by anotherroute.”

It is important to understand what Durkheim is arguing in this para-graph, because many scholars attribute a nominalist, or conventionalist,position to Durkheim. But, his point here is that a sociology of knowledgethat treats the meaning of the categories as purely conventional is not aviable argument. Such an argument would certainly be incompatible withDurkheim’s own claims that the categories do have some validity whenapplied to nature, or even the claim that they have any empirical validityat all. While Durkheim argues that the categories have limited validitywhen applied to natural phenomena, he argues that they have perfectvalidity when applied to social phenomena, which would not be true ofconventional or nominal categories.

However, Durkheim does not think that the categories are nom-inal or conventional in any sense: they are part of nature. He says([1912:25]1915:31; *1995:17) that “To interpret a sociological theoryof knowledge in that way is to forget that even if society is a specific real-ity, it is not an empire within an empire: it is part of nature and nature’shighest expression.” While they are very different from individual rep-resentations, social representations also express some reality. I use theword “express” here because the categories are not representations orsymbols, as we tend to think of them. But, they “express” social forces,and as Durkheim says, “call up” the feelings experienced by participantswhen enacting social forces.

Because society is part of nature and a natural thing, Durkheim argues([1912:25]1915:31; *1995:17), it must be similar in some sense in itsbasic parts to the rest of nature: “It is impossible that the fundamentalrelations that exist between things – precisely those relations that the cate-gories serve to express – should be fundamentally dissimilar in one realmand another.” Society is, according to Durkheim, a natural organism that


obeys the same natural laws that govern the rest of nature. It is, therefore,impossible, he thinks, that the categories that “translate” social relationsshould have no application to the rest of nature.

According to Durkheim ([1912:25]1915:31; *1995:17), the funda-mental relations expressed by the categories “stand out more clearly inthe social world,” but, he argues, “it is impossible that they should not befound elsewhere, though in more shrouded forms. Society makes themmore manifest, but has no monopoly on them.”

It is important to point out here that while Durkheim claims to preservethe Kantian principle of dualism, at this point what he does is to describea difference between social experience and the individual experience ofthe natural world that would explain Kant’s feeling, or conviction, thatthe two realms are completely different. For Kant there are two realms ofreality, the rational and the world in itself. But, there is only one form ofknowledge: sensation processed through the manifold of judgement: i.e.,reason. For Kant, the world in itself cannot be known. What Durkheimargues is that there are two different knowable worlds and two differentsorts of knowledge; one corresponding to each. Furthermore, he argues,the two are not entirely inaccessible to one another, as Kant argued.The social world is part of the natural world. Furthermore, each person,because they are dual, has knowledge of both sorts, and therefore, therecan be successful cross-fertilization between the two forms of knowledge.He pursues this theme of cross-fertilization in his comments on the devel-opment of science in the Conclusion.

While the six categories are themselves perfect mirrors of social rela-tions, the narrative elaborations of the experiences that generate them, aswell as the proliferation of collective representations into languages thatfollow the initial development of the categories, pose a different sort ofvalidity issue. In his sociology of knowledge Durkheim concerns himselfwith the question of how closely language, custom and belief can cometo “accurately” portraying either natural or social relations.

The categories express, or translate, the fundamental relations betweenthings. But, collective representations, and even categories, when notdirectly experienced, are “constructed concepts” and “an artifice.”Durkheim argues ([1912:25]1915:31; *1995:17) that when the categoriesare separated by time and interpretation from their “initial meaning” theymay come to “deviate from their initial meaning” and stand only as sym-bols and be artificial: “If when they deviate from their initial meaning,those notions play in a sense the role of symbols, it is the role of well-founded symbols. If artifice enters in, through the very fact that theseare constructed concepts, it is an artifice that closely follows nature.”

As symbols they pose a different sort of epistemological problem.Durkheim argues ([1912:25]1915:31; *1995:17), however, that even


when collective representations deviate from their initial meaning andbecome symbols, they nevertheless “closely follow nature and strive tocome ever closer to nature.”

Durkheim can argue that even constructed concepts that have deviatedfrom their initial meaning are not completely devoid of empirical validity,for two reasons: first, even though they have deviated from their origin,they still have an origin in shared social experiences, and are thus neverpurely symbolic, and; second, they can be corrected by individual senseexperience which is separate from the faculty of reason. The two differentsorts of knowledge can communicate with one another, and over timeknowledge based on individual sense experience can correct “deviations”in the collective representations.

Durkheim maintains ([1912:22ff]*1915:28–29ff; 1995:15ff)that socialrelations are natural relations of a special sort and therefore, there is nodeep incompatibility between reason and the empirical world of nature, orbetween individual and society: “If experience were completely separatedfrom all that is rational, reason could not operate upon it; in the sameway, if the psychic nature of the individual were absolutely opposed tothe social life, society would be impossible.”

While there is no complete separation between the two layers of knowl-edge, however, in Durkheim’s view, the emotional experience of moralforce, and perceptions of natural forces, are entirely different sources ofknowledge and he argued that categories could not arise from perceptionsof natural events. Individual perceptions of natural events can result onlyin generic representations that belong entirely to the individual. But, whilethey arise from separate sources, they each apply, in at least a limited way,to both the empirical and the social domains.

According to Durkheim, the division by resemblance of natural things,by the individual perceiver, is not valid knowledge. Durkheim takesHume’s position that resemblance and contiguity can only give rise toan opinion or habit of thinking in a particular way. The categories arenecessary for empirically valid thought. The categories themselves canonly develop when the social being joins in assembly with its fellows towitnessably enact the moral divisions of their group. However, if cat-egories of the understanding which develop from social processes areadded to the natural ability to perceive resemblance and form genericrepresentations, Durkheim feels that there is then no reason that sortingby resemblance couldn’t have some limited empirical validity with regardto natural phenomena. The validity would not be equal to that for socialphenomena, but not completely out of touch with natural reality either.

It is unclear from a reading of the Introduction alone why Durkheimrefers to his position as keeping the apriorist principle intact, when thereare so many obvious differences between the two positions. There are


other dualisms besides Kant’s and the association has become so prob-lematic that a clearer distinction on Durkheim’s part seems desirable.However, Durkheim had a unique way of treating philosophical argu-ments as if they were folk narratives, or religious beliefs. This led him totreat their essential claims as data, while at the same time providing verydifferent explanations for the same phenomena: a point that will becomeclearer in light of the Scientia article.

2.2.2 Philosophical Critique

In the context of his argument that the philosophers have not provided anexplanation for dualism, in the Scientica article Durkheim considers vari-ous philosophical approaches to the issue. First, he considers philosophiesthat deny dualism. These positions Durkheim ([1914]1960:330) refers toas “empirical monism and idealistic monism.” Durkheim dismisses thembecause he considers that there is sufficient historical narrative evidenceto assume that dualism exists and has a real underlying cause.18 Durkheimwrites ([1914]1960:331), that “it is no different with the absolute ide-alist.” In the case of the absolute idealist, reality is made up entirely ofconcepts instead of sensations. Because it overcomes the contradictionbetween the conceptual and the empirical, the idealist position equallydenies the reality of dualism. Durkheim ([1914]1960:331) argues that“according to the idealist, an absolute intelligence seeing things as theyare would find that the world is a system of definite ideas connected witheach other in relationships that are equally definite.” Sensations, on thisview are only concepts, according to Durkheim ([1914]1960:331): “Theyassume the particular aspect in which they are revealed to us in experienceonly because we do not know how to distinguish their elements.”19

For the idealist, dualism is an illusion created by our inability to dis-tinguish between the elements of concepts. Thus, the world as presentedthrough sensation appears to be particular, but in reality, for the ide-alist, it is not. Durkheim argues that if this view were correct thenthe experience of dualism would decrease as scientific knowledge and

18 I am not sure who Durkheim intends to refer to here, but it sounds much like Locke,or a follower of Locke who did not accept Hume’s critique. Durkheim cannot mean toreference Hume at this point, because he represents this view as assuming that there isno problem with matching concepts and sensations. While there are similarities betweenLocke and Hume, Hume clearly recognized this problem, and Durkheim referencesHume’s recognition of the problem, whereas Locke did not realize that it was ultimatelya problem.

19 This description corresponds to the version of empiricism championed by BishopBerkeley in his Three Dialogues Between Hylas and Philonous, in Opposition to Scepticsand Atheists [1913]. The sceptics and atheists correspond to David Hume and followers.


conceptual clarity increased with scientific progress. That is, as webecome less confused about the nature of sensation, we would realizethat sensation is really conceptual. Durkheim points out that this is notonly not the case, but that the experience of dualism, and the gap betweenconcepts and sensation, seems to increase, not to decrease with scientificprogress.

As evidence of this, Durkheim ([1914]1960:331–332) turning to reli-gion again as the social context within which reason develops, points outthat it is modern religions that stress the dualistic nature of the person,whereas primitive religions present a picture of man in greater harmonywith the universe: “The great religions of modern man are those whichinsist the most on the existence of the contradictions in the midst ofwhich we struggle. These continue to depict us as tormented and suf-fering, while only the crude cults of inferior societies breathe forth andinspire joyful confidence.”

Religion, as the repository of collective narrative, is a rich source ofexamples of underlying empirical realities for Durkheim. Idealism, hesays, fails to take the experience of dualism represented in the collectivenarrative into account. It fails to explain how, if dualism is only an illusion,the experience of it continues to grow, and the narratives to multiplyand become more complex. Thus, idealism fails to explain where theexperience comes from. In light of the fact that Durkheim has often beeninterpreted as an idealist himself, this criticism is important.

Following this discussion, Durkheim turns to theories that embracedualism. These he says ([1914]1960:332), “are those which limit them-selves to affirming the fact that must be explained, but which do notaccount for it.” Durkheim will applaud all of these positions for not deny-ing the evidence of both personal experience and the historical narrativerecord. However, he criticizes them all for not providing an explanationof the dualism they embrace. He argues that none of them have evenattempted to explain why dualism exists.

Durkheim ([1914]1960:332) first considers what he refers to asPlato’s “ontological explanation.” According to this explanation ([1914]1960:332): “Man is double because two worlds meet in him: that of non-intelligent and amoral matter, on the one hand, and that of ideas, thespirit, and the good, on the other.” This position is closest to Durkheim’sown in putting moral considerations only on one side. It also agrees withRousseau’s position in this regard.

What Durkheim says ([1914]1960:332–33) about Plato’s position,however, is what he says about all the positions that affirm dualism: “Butif this answer – completely metaphysical as it is – has the merit of affirm-ing the fact that must be interpreted without trying to weaken it, it does


confine itself, nevertheless, to distinguishing the two aspects of humannature and does not account for them.” Plato’s position has the merit ofaffirming what persons actually experience. But, Plato does not providean account of why this should be the case. This approach Durkheim says([1914]1960:333) “repeat[s] the problem in different terms; it does notresolve it.”

The fact that Durkheim and Plato agree that dualism has a basis inreality does not mean that Durkheim takes Plato’s position, any morethan it means that he takes Kant’s position. It also should be pointedout that the position that Durkheim attributes to Plato here, and thencriticizes, is very close to the Cartesian position that is often attributedto Durkheim.

Durkheim argues ([1914]1960:333) that it is not the “plenitude ofbeing” that is the center of the dualistic struggle, as Plato and the Carte-sians would have it. Rather, human beings stand at the center of thisstruggle. Therefore, the explanation for dualism must be found in the essenceof what it is to be human, and not in conflicting aspects of the universe as awhole. Of course, in placing the human being at the center of the struggle,Durkheim is in essence placing society at the center, because it is societythat makes that being human and rational. The essence of what it is tobe human is social.

Durkheim credits Kant with being the first to place the rational humanbeing at the center of the analysis of dualism. Thus, Kant moved a stepbeyond the Cartesian understanding of dualism which focused on therational and nonrational in the universe as a whole. However, Durkheim([1914]1960:333) says that Kant’s theory, although it focuses on thehuman being as the location of dualism, and is the “most widely acceptedat present offers an even less satisfactory explanation of human dualism: itdoes not base it on two metaphysical principles that are the basis of all real-ity, but on the existence of two antithetical faculties within us.” WhereasPlato has two contradictory characteristics of the universe meeting in ourminds, Kant locates the contradictions entirely in the individual person.Thus, Kant’s focus on the person, while important, is also one-sided: itleaves out the social. Durkheim also criticizes Kant for not explainingwhy this dualism exists in the individual.

The distinction that Durkheim attributes to Kant is between sensitiv-ity and reason; between sensing as individuals and thinking in universaland impersonal terms. He says ([1914]1960:333) that “Kant more thananyone else has insisted on this contrast between reason and sensitivity.”What Durkheim says is, that even if Kant had been correct in charac-terizing dualism in these terms, he would still have offered no solutionto the problem that interests Durkheim, namely figuring out where this


duality comes from and what underlying realities it represents. He says([1914]1960:334):

But even if this classification is perfectly legitimate, it offers no solution to theproblem that occupies us here; for the important thing to determine from ourconsideration of the fact that we have aptitudes for living both a personal andan impersonal life, is not what name it is proper to give to these contradictoryaptitudes, but how it is that in spite of their opposition, they exist in a single andidentical being.

For Durkheim ([1914]1960:334), “merely to give a name to each being,”even if it is the right name, does not address the fundamental question.Durkheim also posits a contrast between reason and sensibility. But, hisfocus is on where the contrast comes from.

In criticizing the philosophers for naming, but not explaining dual-ism, the positions that affirm dualism, he says ([1914]1960:332), “limitthemselves to affirming the fact that must be explained, but do notaccount for it.” Dualism cannot become an explanation for other things,in Durkheim’s view, until it has itself been adequately explained.

Durkheim ([1914]1960:334) finds Kant’s position, with which his ownposition is generally equated, even less satisfactory than others in thisregard: “The theory that is most widely accepted at present offers an evenless satisfactory explanation of human dualism.” It is not the descriptionof what dualism consists of, however, that Durkheim finds less than sat-isfactory, although he will criticize that also, but the failure to accountfor why the dualism exists and what its underlying causes are. While,according to Durkheim ([1914]1960:334), “Kant more than anyone elsehas insisted on this contrast between reason and sensitivity, between ratio-nal activity and sensory activity,” Kant has given no explanation for howthis dualism is possible or where it comes from.

Durkheim argues that ([1914]1960:334) “if we have all too often beensatisfied with this purely verbal answer, it is because we have generallythought of man’s mental nature as a sort of ultimate given which need notbe accounted for.” Durkheim argues that it is in giving an account of theorigin of the categories of the understanding in social processes that thereal origin of the phenomena of human dualism can be explained. Thus,Durkheim’s claim that dualism is the principle upon which the argumentof The Elementary Forms rests.

For Durkheim ([1914]1960:330) the task is to explain what it is aboutthe origin of human reason that results in two contradictory beings com-ing to exist in the same person; in the fundamental question “where dothis duality and this antinomy come from?” Durkheim ([1914]1960:334)argues that a large part of the problem is that there is a tendency to


believe that a question has been answered “when we attach such andsuch a fact, whose causes we are seeking, to a human faculty.” This placesthe person too completely at the center of the phenomenon. Durkheim([1914]1960:334) argues that the social causes of the phenomenon areoverlooked because it is assumed that reason is “a sort of ultimate given.”“But why,” he ([1914]1960:334) asks “should the human spirit, whichis – to put it briefly – only a system of phenomena that are compara-ble in all ways to other observable phenomena, be outside and aboveexplanation?” Why should the faculty of reason not be subject to socialforces?

2.3.0 The Social as Sacred versus the Individual as Profane

The contrast between sacred and profane, experienced as a socially gen-erated moral force, is, for Durkheim, the key to the explanation of wherereason (the categories) and hence dualism, comes from. Without the con-trast between sacred and profane and its enactment by the social group,moral forces could not be felt. The sacred Durkheim identifies with thegroup, the social, and with moral force. The profane he identifies with theindividual (among other things). Because collective concepts are sharedand transcend the individual, they are assigned by Durkheim to the realmof the sacred and not the profane. They are distinct from the individualand the profane and, according to Durkheim ([1914]1960:336), that iswhat “constitutes the essence of their sacred character.”

If society were simply an extension or aggregation of individual desiresmediated by some sort of an agreement, as the utilitarians argued,then these elements of the person would harmonize. The first (sensa-tion) would be a component of the second (reason), and the distinctionbetween the sacred/group and the profane/individual would collapse.The experience of dualism would not exist. This is not the case, how-ever, according to Durkheim ([1914]1960:338), because “society has itsown nature, and consequently its requirements are quite different fromthose of our nature as individuals.” Because society has needs that arecompletely different from individual needs, society must make demandson individuals that conflict with individual needs and which cannot bereduced to them.20 He argues ([1914]1960:338) that “It is evident thatpassions and egoistic tendencies derive from our individual constitutions,while our rational activity – whether theoretical or practical – is dependenton social causes.”

20 Durkheim’s argument with regard to the irreducibility of ritual social orders to individualmotives parallels the idea that Interaction Orders make demands on their own behalfthat are not reducible to either individual motives or institutional rules (Rawls 1987).


Because the rational activity of persons depends on the social, andrational activity becomes more important over the course of history, itis also the case that society becomes more important. Durkheim argues([1914]1960:339) that “since the role of the social being in our singleselves will grow ever more important as history moves ahead . . . allevidence compels us to expect our effort in the struggle between the twobeings within us to increase with the growth of civilization.”

Durkheim speculates that, by contrast, for Kant’s mind body dualism,the advance of civilization and education would bring about a greaterability to use the mind to control the body. One would expect that as civ-ilization advanced, dualism would become less dramatic a conflict withinthe individual. But, on Durkheim’s view exactly the opposite is the case.Because society is becoming more important, the rational social side ofus will make greater demands on the pre-social being and the tensionbetween the two will increase.

Kant’s dualism fares no better in Durkheim’s analysis than religiousbeliefs. Both assume enduring universal qualities in place of social rela-tions in trying to explain dualism. Just as he argues that the reason Kantposits as innate has a social origin, Durkheim argues that the religiousidea of the soul as sacred has a social origin. The reason that is created inand through participation in religious practice has its parallel in the ideaof the soul or intelligence. Religion operates by distinguishing betweenthe sacred and the profane. The soul corresponds to the sacred and thebody to the profane. This does not mean that the soul and body are real,any more than Kant’s mind and body were real, but that the ideas of souland body are created in and through the religious distinction betweensacred and profane, and the categories of reason that are generated fromthat initial moral distinction.

According to Durkheim ([1914]1960:335), “The duality of our natureis thus only a particular case of the division of things into the sacred andthe profane that is the foundation of all religions, and it must be explainedon the basis of the same principles.” This origin of the experience ofdualism in the social experience of moral force, Durkheim says, is whathe was trying to demonstrate in The Elementary Forms.

2.4.0 Sociology of Knowledge: Idealism versusConcrete Practices

While Durkheim wrote the Scientia article in order to clarify the argu-ment for dualism that was an important part of the argument of TheElementary Forms, he was obviously not successful. Instead of realizingthat Durkheim was making an entirely new argument, in which social


prerequisites explain human dualism, the critics seem to have takenDurkheim’s affirmation of the “fact” of dualism, and his positive ref-erences to Kant, as evidence that he took a Kantian position on dualism.

While Durkheim’s treatment of Kant’s argument as a folk narrative,and his consequent lack of interest in producing a comprehensive crit-icism of it, must be at least partially responsible for the mistaken iden-tification of Durkheim’s position with Kant’s, there is another problemwith Durkheim’s presentation of the argument. In The Elementary FormsDurkheim’s epistemology was obscured by the failure to clearly distin-guish between a sociology of knowledge argument and an epistemologicalargument. Unfortunately, in the Scientia article Durkheim continues tovacillate between these two ways of talking about the relationship betweenlanguage and moral force, without clearly distinguishing between the two.

When Durkheim writes that people represent ideas and sentiments inthe form of moral forces, he seems to be treating moral forces as if theyexisted only as ideas and had their force only through belief; a sociology ofknowledge argument. He writes ([1914]1960:335, emphasis added) that:“The ideas and sentiments that are elaborated by a collectivity . . . areinvested by reason of their origin with . . . an authority that cause[s] theparticular individuals who think them and believe them to represent themin the form of moral forces that dominate and sustain them.” It appearshere that the moral forces themselves are only social representations,when in fact, moral forces are what Durkheim argues in The ElementaryForms are ultimately real.

Yet, in other parts of the article, and finally at the end of the arti-cle, Durkheim ([1914]1960:338) makes a very important distinctionbetween the collective representations he has been discussing in some-what ambiguous terms and the categories of the understanding:

In the book that is the occasion of the present study but which we can onlymention here, we have tried to demonstrate that concepts, the material of thelogical thought, were originally collective representations . . . We have even founda basis for conjecturing that the fundamental and lofty concepts that we callcategories are formed on the model of social phenomena.

The categories are treated here as a special type of collective representa-tion.21 “Even” with respect to the categories, he says, the social originsargument can be made. The social origins argument is quite different inthe two cases, however. While all collective representations must evokea collective mental state of some sort, in order to operate as significantsymbols, the categories of the understanding are caused by very specific

21 The same thing happens in the Conclusion where the categories are sometimes calledcollective representations, but are nevertheless dealt with in an entirely separate section.


types of shared experiences, that only occur when certain types of prac-tices are enacted in particular ways, and, unlike collective representationsin general, the categories are identical, not only between individuals, butin some sense between cultures and historical periods as well. Further-more, the categories are identical with the mental states that they expressand invoke, whereas other symbols only more or less correspond withsome collective sentiment.

Another difficulty with the Scientia article is Durkheim’s argument thatthe vitality that these ideas communicate to us is the effect of a special psy-chic operation by which many consciousnesses are fused into one com-mon consciousness. According to Durkheim ([1914]1960:335) “Theyare simply the effects of that singularly creative and fertile psychic opera-tion – which is scientifically analyzable – by which a plurality of individ-ual consciousnesses enter into communion and are fused into a commonconsciousness.” This passage surely contributed to the interpretation ofDurkheim’s argument as a Wundtian group mind theory. It seems topoint away from the concrete enacted practices, that Durkheim workedso hard to convey, and back toward psychology. But, I believe that if readcarefully, it becomes clear that Durkheim is referring to ideas that havebecome attached to, that is, have become symbols of, the effects of moralforces previously felt. This attachment occurs when individuals enter intocommon communion, in enacting practices together. As Durkheim says([1914]1960:335) several lines before this passage: “sacred things aresimply collective ideals that have fixed themselves on material objects.”They don’t work as ideas to bring these feelings to mind unless they havebeen previously experienced as moral forces in such a way as to “attach”the feelings to the ideas.

In the passage immediately following this, Durkheim returns to thequestion of how the feelings originally become attached to the sym-bols.22 He argues ([1914]1960:336) that collective representations mustbe embodied in material objects and actions before they can be sharedbetween minds:

From another point of view, however, collective representations originate onlywhen they are embodied in material objects, things, or beings of every sort –figures, movements, sounds, words, and so on – that symbolize and delineatethem in some outward appearance. For it is only by expressing their feelings, by

22 The paragraph opens with the phrase (in English translation) “From another point ofview.” Since Durkheim is clearly expressing his own point of view in both this and theproceeding paragraph, I can only think that he means to state the contrast between theeffects of collective symbols once they exist, and the idea that they have an origin whichmust be explained. Explaining that origin, instead of just beginning with the ideas, is to“take another point of view,” that is, to take Durkheim’s own point of view.


translating them into signs, by symbolizing them externally, that the individualconsciousnesses, which are, by nature, closed to each other, can feel that they arecommunicating and are in unison.

In this account it is the material and embodied character of symbols, asmovements, sounds and words, used on social occasions, that makes thefeeling of communicating in unison possible.

On this view, collectivity becomes possible only when the same feel-ings are evoked in response to a symbol in many people at the sametime. According to Durkheim ([1914]1960:336), material symbols andembodied movements are necessary because “the things that embodythe collective representations arouse the same feelings as do the mentalstates that they represent and, in a manner of speaking, materialize.” Itis very important that these are not just symbols, as we have come tothink of them as interpreted signs. They do not call out different ideasin different minds; they symbolize collective emotions in such a way thatlooking at them “materializes” those emotions, and they are then feltagain (although less strongly).

Because the mental states that the symbols evoke are collective andhave only collective meanings, the feelings that they arouse must alsohave a collective origin, and cannot just be communicated from indi-vidual to individual. This is the case whether the symbols are collectiverepresentations and therefore not empirically valid, or empirically validcategories. Both are inherently collective forms of thought. This is notso with individual sensations, or with generalizations that are based onindividual sensation. These remain essentially incommunicable, and theproblem of solipsism persists.

The moral forces that these symbols represent are, he says([1914]1960:336), in all cases real: “This system of conceptions is notpurely imaginary and hallucinatory, for the moral forces that these thingsawaken in us are quite real – as real as the ideas that words recall to usafter they have served to form the ideas.”23

The difference is that in the case of collective representations, theconcepts represent the moral forces, whereas categories are caused byand materialize them. In both cases the moral forces are real. Butthe relationship between moral forces and ideas is different. The col-lective representations originate as retrospective attempts to explaincollective experiences. Categories of the understanding are themselves

23 It is not entirely clear what Durkheim means by “real” in this passage. In The ElementaryForms when Durkheim refers to moral forces as real he makes it clear that he meansthings like “if the headman says you are exiled you are.” Sort of like a speech act typeof moral force. But, here he is talking about moral forces having the same sort of realitythat words have.


collective experiences. The vitality of both types of collective representa-tions, however, ultimately originates in the “action of the group,” not justin other ideas. When persons go off on their own, according to Durkheim([1914]1960:336), the vitality of these conceptions is reduced:

It is not extinguished, however; for the action of the group does not ceasealtogether: it perpetually gives back to the great ideals a little of the strengththat the egoistic passions and daily personal preoccupations tend to take awayfrom them. This replenishment is the function of public festivals, ceremoniesand rites of all kinds.

In keeping with his oscillation between arguments that sound idealistand arguments that sound materialist, Durkheim refers to moral forcesagain in the next paragraph as “ideas” again. While this inconsistencyin Durkheim’s language, not distinguishing clearly between collectiverepresentations and categories, has obscured his argument, he certainlybelieved that he had resolved the contradiction between these two posi-tions with his epistemology. He apparently did not realize that his failureto clearly distinguish them would render the epistemological argumentitself invisible.

2.5.0 Conclusion

In spite of the similarities between Durkheim and Kant with regard to thefact of dualism, their understandings of what dualism is, and their expla-nations of its origin, are not at all the same. For Kant, dualism followsfrom the character of innate reason. If reason is innate there is no pos-sible contact between rational beings and the world in itself. Therefore,there is a contradiction between thought and materiality. For Durkheim,persons are both rational/social moral beings and a-moral beings of sensa-tion at the same time. The needs and experiences of the two beings withinus are inherently in conflict, but not unavailable to one another. Socialbeings do exist in an indirect contact with natural phenomena throughthe individual being within them. Although the general ideas that arederived from that sensible contact with the world of nature do not havethe same empirical validity as the categories, or the same communicabil-ity as collective representations, they are, nevertheless, derived directlyfrom contact with natural phenomena, and are not processed through,what Kant called, “the manifold of judgement” or the “intuitions.”24

As a consequence, the sensations available to the individual can cor-rect invalid empirical generalizations made by the individual through a

24 Immanuel Kant, The Critique of Pure Reason 1781.


process of trial and error, as they could for Hume and other empiricists,especially James. Individual generalizations, based on sensation, can also,over time, correct conceptual generalizations about the natural world thatare held by the collective and the social being. Durkheim is describinghere something like the process of normal versus revolutionary sciencein the work of Thomas Kuhn (1958). Thought can only occur withinexisting conceptual parameters, and normal science takes place withinan agreed upon, or socially constructed framework of concepts and prac-tices. However, empirical experience does not respect the same concep-tual boundaries and cannot always be made to fit the existing consensus.Periods of revolutionary science in which thought and experience can-not be made to match would be rare, and would need to overcome theenormous resistance of existing conceptual frameworks. The differenceis that for Kuhn, these periods constitute leaps of “faith.” For Durkheim,they are points where collective knowledge falls back on empiricism. But,these corrections would not be impossible. This sort of correction is notpossible from Kant’s position.

Ultimately, Durkheim’s distinction is completely different from eitherthe Cartesian or Kantian dualisms, because he makes a distinctionbetween two naturally occurring empirically experienced realms ofhuman life that he says exist in the same person: individual sense experi-ence and social experience. Both are directly experienced, neither is ideal.The Kantian and the Cartesian positions, on the other hand, counter-poise a pure realm of reason with the corruptible world, and pure mindwith pure body. Even though Durkheim limits the knowledge of the bio-logical individual to what can be generalized on the basis of sensation,he treats the body and the empirical world as a natural part of the socialrealm, and the social as a part of the natural world.

Whereas Kant’s rational man has no ability to perceive the physicalworld in itself, and is in a constant war to preserve reason from bodilyinfluences, Durkheim’s social individual retains the ability to perceive thenatural world directly. While there may be an internal struggle betweenindividual and social impulses, and conceptual forms shape the way inwhich individual sensations are experienced, for Durkheim, the two sortsof impulses do not belong to two completely different realms, one ofreason and one of matter, as they do for Kant and the Cartesians. ForDurkheim, both are natural, both are necessary, and human progressthrough history is made possible because the two individuals can, soto speak, share information with one another. The biological individualis able, albeit only from time to time and with great difficulty, to addnew information about the natural world to what the social individualknows.


It is Durkheim’s epistemology that allows him to reverse this relation-ship; allowing the natural to enrich the social world of ideas. Because thecategories are empirically valid and socially caused, there is a separation,but no divorce, between natural reality and social reality. It is true thatthe categories are not drawn from individual perception of natural reality.But, they do come from direct experience of social reality, which forDurkheim, is also a part of nature.

If Durkheim had offered only a sociology of knowledge in The Elemen-tary Forms, that argument would not have allowed him to maintain theconnection he claims between the conceptual and the empirical realms.The social end of things would, in a sociology of knowledge with noepistemological underpinning, have no empirical validity, would be com-pletely ideal, and, therefore, completely out of touch with the sensibleworld experienced by the individual biological being. This would be muchmore consistent with the Kantian and Cartesian dualisms. It is the wayin which The Elementary Forms has generally been interpreted. But, it isa position that Durkheim clearly rejects.

Durkheim’s position is a very original and interesting joining and tran-scending of the Kantian and empiricist positions in this regard. He says inseveral places that what he is attempting to do is create a position that fallshalfway between the two. I believe that he meant this in a very profoundsense. Empiricism corresponded with the individual, and apriorism withthe social aspects of Durkheim’s theory. However, by combining them inthe way he did; putting social practice in place of the individual at thecenter of the theory, he completely transformed both. Durkheim arguedthat his position would solve the problems of both empiricism and aprior-ism, preserve the benefits of each, and stand halfway between the two onseveral key issues. He did not mean that he was in any simpleminded waytaking part of each position and trying to cobble up a hybrid halfway inbetween. He meant that there was a completely different way of conceptu-alizing dualism, and that when one takes this sociological position, manyof the older beliefs are supported, while at the same time, the problemsassociated with them are solved.

3 Sacred and Profane: the First Classification

The sacred is the key to Durkheim’s argument. Consequently, Book I ofThe Elementary Forms, consists of an extensive consideration and criticismof various theories of the origin of the distinction between the sacred andthe profane. Durkheim defines religion, criticizes Animism and Natur-ism, and then offers Totemism as an answer to the question of where theidea of the sacred comes from. As Totemism will provide the focal pointof Durkheim’s argument, his arguments with regard to Totemism are ofparticular importance.

Durkheim’s emphasis on the sacred, combined with a lengthy review ofvarious theories regarding the origin of the idea of the sacred, has tendedto create the impression that his focus is on religion, and the relationshipbetween the variety of religious beliefs and classifications in particularsocieties. In fact, however, the distinction between sacred and profane,and the critical review of the anthropology of religion that occupy Book Iare essential to Durkheim’s epistemological argument. The first dualism,sacred versus profane, turns out also to be the first classification. As thefirst type of moral force it not only constitutes classification as the firstcategory of the understanding, but is also an essential component of theenactment of all the other categories. Without the enacted “feeling” ofthe sacred, moral force cannot be created, categories of reason cannot bedeveloped, and society, as a consequence, cannot exist.

Classification is the first category that Durkheim considers in any detailin The Elementary Forms. The variety of social and religious classificationsin societies are, Durkheim argues, all a result of this one first classificationbetween sacred and profane. It is the moral force of that first classificationthat makes it empirically valid. It is the emotional result of a real socialcreation. The multitude of other classifications are artifacts of social rela-tions, but not empirically valid, or universal, like the distinction betweensacred and profane, which is, in effect, the category of classification itself.Durkheim’s epistemological argument with regard to classification is notan argument about the origins of a variety of particular classifications, ashis earlier work Primitive Classification (co-authored with Mauss 1901)

108

Sacred and Profane: the First Classification 109

led many scholars to believe. Rather, it is an epistemological argumentfor the empirically valid social origin of one universal idea, the sacredand profane; the first classification, and the original of the category ofclassification.

The presentation of the argument for sacred and profane, as the firstenacted classification, involves long sections on the anthropology of reli-gion, as well as extensive descriptions of various clan, tribal, and totemicclassifications. These lengthy discussions dominate both Books I and IIof The Elementary Forms.

Because scholars of religion have offered explanations of the origin ofthe sacred which, if accepted, would negate his epistemology, Durkheimneeds to refute their positions and establish an origin for the sacred inthe enactment of social practice. Durkheim’s task is first to show that theidea of the sacred could not have origins in sense experience of naturalphenomena. Then he argues that it could not be an a priori idea. Finally,he argues for the social origins of the sacred, laying claim to the sacredas a sociological phenomenon enacted through social practices.

The idea of moral force is an essential part of this argument. It is themoral force that results from the enactment of the distinction betweensacred and profane that makes a valid sociologically based epistemol-ogy possible, for Durkheim.1 Durkheim argued that the experience ofdualism depends entirely on the distinction between natural and moralforce and the faculty of reason that results from moral force. Accordingto Durkheim, moral force is not a natural or individual development andcannot be perceived through the five senses. Moral force is a social effect,and the response to moral force is emotional, not physical.2 He arguesthat moral forces are created by the collective performance of rites asso-ciated with the distinction between the sacred and the profane. As such,moral force has different kinds of effects than natural forces, and theseeffects are perceived differently from the effects of natural forces.

Scholars of Animism and Naturism, on the other hand, explain the ori-gin of the sacred exclusively in terms of the perception, or mis-perception,of natural forces through the five senses. Totemism, on the other hand,

1 Neither comes first, moral force and sacred are the same thing and are enacted together.Once the sacred exists as an idea it can precede the enactment of moral force, but not thefirst time.

2 In some respects Durkheim’s argument that moral force is a social and emotional phe-nomenon is consistent with Hume, who also argued that moral ideas were emotionalphenomena. However, for Hume the experiences remain individual. Durkheim pointsout that, even though they are treated as emotions, they retain some of the problemsof individualism. Durkheim’s treatment of certain emotions as collective achievements,artifacts of collective practices answers the question of how a group of persons all havethe same emotion, or idea. Hume’s position does not.


is explained in terms of an innate tendency to believe in the sacred.Durkheim’s lengthy consideration of these positions is necessary in orderto establish the social origins of the sacred in moral force, rather than innatural forces, or innate ideas. Durkheim’s analysis of religion, the cri-tique of Animism and Naturism, and the lengthy sections on Totemismthat follow, addresses the question of whether or not these theories pro-vide an explanation of where the distinction between sacred and profanecame from. In each case, Durkheim either argues that these theories havenot given a valid account of the origin of the sacred, or that they assumean innate tendency to believe in the sacred, and therefore, don’t even tryto provide an account of its origin.

Durkheim’s critique of religious anthropology parallels his critique ofepistemology. The animists and naturists generally attempt to explainthe sacred on the basis of sense perception and fail. They fall prey toHume’s dilemma; that general ideas cannot be derived from sense per-ception. The Totemists, in an argument reminiscent of Kant, offer innatehuman tendencies as the origin of the sacred. In both cases, Durkheimargues, in a manner parallel to his criticisms of Kant and Hume, theyhave failed to explain the origin of the idea of sacredness. The animists,in particular Taylor, and the naturists, in particular Muller, must fail,Durkheim says, because sense perception cannot explain the origin ofan idea, like the sacred, that has no counterpart in nature. Durkheim’spoint here is similar to Hume’s argument that the idea of causality couldnot be empirically valid because no counterpart could be found for it insense experience.3 Sacredness, not being a natural phenomena, cannotpresent itself to perception. The Totemists, and in this regard Durkheim,cite Frazier, because they are innatists, assume the existence of the phe-nomenon and therefore, like Kant, fail to provide any explanation at allfor its origin.

Durkheim’s critique of various religious theories in Book I, concludesthat the distinction between sacred and profane could have no othercause except a ritual social cause. Totemism is the form of religion thatDurkheim will argue generates the original distinction between sacredand profane through its rituals. However, as Totemism is explained byFrazier, in innatist terms, it lacks the explanatory power that Durkheimneeded it to have.

Whereas, in his critique of classical epistemology, Durkheim is con-cerned with the origins of six categories of the understanding, in his cri-tique of theories of religion in Book I, he is concerned with the origin of

3 According to Durkheim, causality is also an idea which, like the sacred, is a characteristicof social force, and not of relations between things in nature.


only one idea, the sacred, which he will argue holds the key to explainingthe origin of the six categories of the understanding. Sacredness consti-tutes, for Durkheim, an essential element in the constitution of moralforce. Therefore, if his epistemology is to be valid, sacredness, like moralforce, cannot have an origin in sense perception. Neither can it be innate.It must have an origin in moral forces.

It is Durkheim’s position that the distinction between sacred and pro-fane opens the way for the initial development of human reason. It is thekey element in Durkheim’s dualism of human nature. Durkheim uses thedistinction to explain the difference between modes of perception thatare empiricist and individual and modes of perception that are social andconceptual (see 2.1.1). Before the experience of the distinction betweensacred and profane, persons experience the world in terms that are con-sistent with Hume’s empiricism. After the experience of the distinction,however, persons are able to communicate in terms of empirically validshared categories of the understanding.

When Durkheim complained in the Scientia article, in 1914, that hiscritics had missed the fact that dualism was the underlying theme ofThe Elementary Forms, he meant it quite literally. The point is not justthat the idea of human dualism runs through the entire argument, butthat the epistemology itself is based on the first dualism: the distinctionbetween sacred and profane. The arguments about the sacred are theepistemological argument in essential respects. As Durkheim wrote inthe Scientia article (1914: 335) “The duality of our nature is thus only aparticular case of that division of things into the sacred and the profanethat is the foundation of all religions.” Human dualism is only a secondaryeffect of the fact that the division between sacred and profane, created byritual practices, has created human reason.

While the theories of religion that Durkheim considers in Book I areno longer as influential as they were in Durkheim’s day, the questionswhich he raised with regard to their explanation of the origin of the ideaof the sacred are as pressing today as they were when Durkheim wrotethem. If the sacred is the essential idea upon which reason, moralityand social solidarity are going to rest, then the way in which its originis explained will matter greatly. Durkheim’s criticisms can be adapted toaddress naturalist or innatist approaches to the question.

Durkheim’s argument with regard to the idea of the sacred can bedivided into three parts. First, Durkheim defines religion so that itincludes the totemic practices that he needs as a foundation for hisargument. In the process he criticizes conventional conceptions of reli-gion and distinguishes beliefs from rites. The distinction between sacredand profane must be established as a universal if it is to be the basis of


universal categories. Establishing the universal character of the distinc-tion involves Durkheim in a consideration of the origin of the idea. Also,because Totemism has often been considered magic and not religion,Durkheim considers the distinction between religion and magic. Second,in an argument that parallels his earlier consideration of empiricism andapriorism, Durkheim takes up the arguments of Animism and Natur-ism. Third, Durkheim ends Book I, in Chapter Four, with a preliminaryconsideration of Totemism. In particular, Durkheim must establish thatTotemism does not have individual origins. Book II will be taken up withan elaboration of this argument.

3.1.0 Durkheim’s Conception of Religion

The first chapter of Book I opens with a long and involved discussion ofthe definition of religion. Durkheim’s first task, he says, will be to definewhat he means by religion. Defining religion carefully is important, hemaintains, because otherwise phenomena that are properly religious willbe left out of the analysis, as they have been with other authors. This mat-ters particularly, because the religious phenomena that Durkheim willcredit with creating the categories of the understanding, totemic prac-tices, are not considered, by most scholars, to be religious phenomena atall. Leaving them out leaves scholars with no explanation for the originof the sacred, according to Durkheim, since totemic practices are theirorigin.

Durkheim cannot allow the essential categories to have their origin inbeliefs, because then they would not have an empirical origin. Therefore,definitions of religions as systems of belief, rather than practices, suchas Max Weber’s, are problematic from Durkheim’s perspective. Conven-tional views of religion as primarily systems of belief must, therefore, bedispensed with before he presents his own argument. Although, for pur-poses of argument, Durkheim considers beliefs first, this is only so thathe can criticize a beliefs approach and put it aside. In fact, for Durkheimpractices come first. The beliefs that seem to motivate the practices actu-ally come only after practices as secondary phenomena that develop toexplain the moral force that is felt by participants when certain practicesare enacted. Because the sacred is usually considered to be a belief, andnot a result of enacting practices, Durkheim’s consideration of the majortheories with which he differs must deal with beliefs first. However, hewill argue that the sacred is really an artifact of certain practices. The“motivation” for such practices is to create and maintain the boundarybetween sacred and profane, not a belief in the supernatural accordingto Durkheim.


The first idea that stands at the beginning of the epistemological pro-cess, even before the idea of classification, which Durkheim says is prob-ably the first category, is the sacred. The sacred, like the categories, isdifferent from other ideas, in that it must be enacted through practices inorder to exist. It does not exist first as an idea. Therefore, having religionbegin with the enacted moral force of the sacred is not the same thing ashaving it begin with an idea of the sacred. Furthermore, if religion mustexist first in order to create the idea of the sacred, then the “real” purposeor function of religion must be something other than a celebration of theidea of the sacred, an idea that could not exist until religion had alreadycome into being.

3.1.1 Critique of Conventional Conceptions of Religion

In his attempt to define religion, Durkheim ([1912:33]*1915:38–9;1995:22) first dispenses with those “conventional” definitions of religionwhich, he says, “might prevent [the mind] from seeing things as theyreally are.” His concern is to dispense with definitions of religion thatwould not include Totemism as a religion. Any definition that required abelief in a deity or the supernatural would have this effect, as Totemismdoes not involve such ideas. Furthermore, the conventional belief thatreligion has to do with the supernatural, or with the idea of a deity,is a definition that focuses on ideas rather than on the practices thatDurkheim requires for his epistemology. A definition of religion as ideasand beliefs would only allow Durkheim to argue that the categories comefrom other beliefs and ideas. A practice oriented definition of religion isessential to save Durkheim from this circularity. Totemism has the prac-tices that Durkheim needs to make his argument that moral force is asocial creation. But, it does not have a belief in a deity or the supernatu-ral. Therefore, Durkheim needs a new definition of religion that focuseson practices and includes Totemism.

The first definition that Durkheim challenges requires a belief in thesupernatural as one of the defining characteristics of religion. He saysthat according to this definition no primitive practices could be consid-ered religions because they do not include the idea of the supernatural.He also argues that the idea of “mysteries” is essential to the belief in thesupernatural and that both ideas are very recent developments in religiousthinking. According to Durkheim ([1912:35]*1915:40; 1995:24) prim-itive peoples do not treat their religious rites, or the power which theyattribute to them, as mysteries. On the contrary, they treat these powersas the most obvious and logical things in the world. Durkheim argues([1912:35]*1915:40; 1995:24) that it is a mistake to equate primitive


practices with the mysteries of modern religions: “For those who believein them, they are no more unintelligible than are gravitation and elec-tricity for the physicist of today.” Neither gravitation or electricity canbe seen. Yet, we have no problem believing in them and do not con-sider them to be mysterious. For the primitive nature is not mysterious.Durkheim argues ([1912:37]*1915:42; 1995:25) “that it is science andnot religion which has taught men that things are complex and difficultto understand.”

Furthermore, according to Durkheim, the idea of physical forces,which are also often cited as an origin of the sacred, is a more recentidea coming only after the idea of social or moral forces. He says([1912:35]*1915:40; 1995:24), that since “the idea of physical forcesis very probably derived from that of religious forces; then there cannotexist between the two the abyss which separates the rational from the irra-tional.” The idea of physical forces is derived from the social, just as theidea of religious forces is derived from the social. It is a mistake to treateither physical forces, or supernatural forces, as the mysterious origin ofthe sacred, since both have the same origin in social forces.

Finally, in order to have an idea of a supernatural order of things,Durkheim argues that a society must first have an idea of a naturalorder of things. The idea of a supernatural order requires the expla-nation of a disjuncture between two orders of things. In order for thisdisjuncture to depend on an expectation of regularity in nature, sci-ence would already have to be well developed. Therefore, Durkheimargues ([1912:36]*1915:41; 1995:24), the idea of the supernatural can-not come before the development of science: “In order that certain thingsare supernatural, it is necessary to have the sentiment that a natural orderof things exists, that is to say, that the phenomena of the universe arebound together by necessary relations, called laws.” This, according toDurkheim is a late development.

Consequently, he argues ([1912:36]*1915:42; 1995:25) the origin ofthe belief in the supernatural cannot be reduced to being awestruck inthe face of the unforeseen forces of nature, because events cannot beunexpected, unless there are background expectations against which theytake place. Of course, the logical idea of a natural order would comeeven later than the development of a contingent idea based on repeatedexperience.

Nothing can be unforeseen without science, according to Durkheim([1912:38]*1915:42; 1995:25), because before the development of sci-ence, “The new is a part of nature just as well as its contrary.” In order toexperience the unforeseen one must have clear expectations regarding theforeseen. Consequently, for the untutored, nature seems more random


than patterned. Durkheim says ([1912:38]*1915:42; 1995:25–6) that “Ifwe state that in general, phenomena succeed one another in a determinedorder, we observe equally well that this order is only approximative, thatit is not always precisely the same, and that it has all kinds of excep-tions. If we have ever so little experience we are accustomed to seeing ourexpectations fail, and these deceptions return too often to appear extraor-dinary to us.” According to Durkheim ([1912:38]*1915:42; 1995:26) “Acertain contingency is taught by experience just as well as a certain uni-formity.” Because of this there is no reason for persons to experience alogical disjuncture between the expected and unexpected. There wouldbe no reason to create the idea of a supernatural to explain the one andnot the other.

Furthermore, the disjuncture between the expected and the unex-pected cannot be the origin of the distinction between the sacred andthe profane, because it is not of the same character as the disjuncturebetween the sacred and the profane. The latter is an absolute disjunc-ture. The former is only contingent.

In any case, according to Durkheim ([1912:39]*1915:43; 1995:26),religions generally do more to explain the regular and expected than theunexpected.

Durkheim also dispenses with the idea of divinity as essential to reli-gion. This idea can’t be essential to religion, he says ([1912:40]*1915:44;1995:27), because it would eliminate Buddhism from being considered areligion. According to Durkheim, treating the idea of divinity as essentialto religion not only treats the idea of divinity as being much more impor-tant than it is, it also overlooks the importance of practices to religion.Not only do some religions not have a deity, but according to Durkheim([1912:47]*1915:49; 1995:32) “even within deistic religions there aremany rites which are completely independent of all idea of gods or spiri-tual beings.” Here Durkheim cites the many rules that regulate the mun-dane activities of everyday life. These he will say later come directly fromdistinctions related to the contrast between the sacred and the profane.They do not come from the idea of a deity.

The distinction between beliefs and rites, or practices, is essential toDurkheim’s understanding of both religion and epistemology. Many rites,according to Durkheim ([1912:47]*1915:49; 1995:32), “work by them-selves, and their efficacy depends upon no divine power; they mechan-ically produce the effects which are the reason for their existence.” Itis the effects of such rites to which Durkheim will turn in Book III ofThe Elementary Forms for the source of moral force and the idea of thesacred. In some religions there are sacrifices that are omnipotent, notdivinities.


This is what makes the actual performance of the rite so impor-tant. Durkheim’s emphasis on rites is consistent throughout. Durkheim([1912:31–2]*1915:37–8; 1995:21–2) opened Book I with the idea thathe will examine the various religions in “their concrete reality” in orderto find out what religion really is. Now he says ([1912:48]*1915:50;1995:33): “This is the explanation of the fundamental importance laidby nearly all cults upon the material portion of the ceremonies . . . Sincethe formula to be pronounced and the movements to be made containwithin themselves the course of their efficacy, they would lose it if theydid not conform absolutely.”

3.1.2 Beliefs versus Rites

Having dealt with some preliminary issues of definition, Durkheimbegins, in Book I, Chapter One, Section iii, to distinguish beliefs fromrites. According to Durkheim, religions are made up of many types ofactivity and many sorts of beliefs. He argues that in seeking a defini-tion of the whole, we must first understand the parts. In approach-ing the “parts” of religion, Durkheim ([1912:50]*1915:51; 1995:34)makes the distinction between beliefs and rites that is essential to hisposition: “Religious phenomena are naturally arranged in two fundamen-tal categories: beliefs and rites. The first are states of opinion, and consistin representations; the second are determined modes of action. Betweenthese two classes of facts there is all the difference which separatesthought from action.” It is on rites that Durkheim will ultimately rest hisepistemology.

However, Durkheim’s analysis will begin by focusing on beliefs. WhileDurkheim argues that rites can be causally efficacious entirely indepen-dently from beliefs, rites are considered to be religious only when theirobjective is considered to be sacred. Therefore, a belief in the sacred, asthe objective of religious rites, must be taken into account first when con-sidering religious rites, and according to Durkheim ([1912:50]*1915:51;1995:34) in order to define rites one must first define beliefs.

Beliefs are generally considered to be essential to the definition of reli-gion. However, for Durkheim, it is not beliefs that are causally efficaciousin the development of the categories of the understanding. Even the beliefin the sacred is only important as a result of enacted practice, not as a beliefper se. In fact, Durkheim argues that with the exception of the sacred itdoesn’t much matter what the beliefs are. What is essential, according toDurkheim, is that certain forms of rites be observed. This, which pro-duces the moral force of the sacred, is what allows practices to create thevarious categories of the understanding.


According to Durkheim’s definition of religion, any set of rites, or socialrituals, that contributes to a distinction between the sacred and the profaneconstitutes a religion, regardless of what its beliefs are. The key factor,for Durkheim, is whether or not the rite creates a belief in the sacred, notwhether it assumes such a belief in the first place. The criterion he uses todefine religion is whether or not it is composed of practices, or rites, thatestablish that first classification, or dualism, between sacred and profane.Thus, his discussion of religion is really a discussion of the developmentof the first classification. Durkheim’s focus on religious beliefs, however,does not treat all beliefs equally. He is only concerned with the one idea:the sacred, which he considers to be the defining experience in all religiouspractice.

3.1.3 Sacred versus Profane: A Universal Social Distinction

According to Durkheim ([1912:50]*1915:52; 1995:34) all known reli-gions have one characteristic in common, they all make a distinctionbetween the sacred and profane:

All known religious beliefs, whether simple or complex, present one commoncharacteristic: they presuppose a classification of all things, real and ideal, ofwhich men think, into two classes or opposed groups, generally designated by twodistinct terms which are translated well enough by the words profane and sacred(profane, sacre). This division of the world into two domains, the one containingall that is sacred, the other all that is profane, is the distinctive trait of religiousthought.

Durkheim opens Book I Chapter Three of The Elementary Forms withthe claim that the distinction between the sacred and the profane is auniversal distinction, one that is found in all societies. According toDurkheim, every society develops the same distinction between sacredand profane. This distinction, he says ([1912:53]*1915:54; 1995:36,emphasis added), always takes the form of a division of the world intotwo classifications of things: “The sacred and the profane have always andeverywhere been conceived by the human mind as two distinct classes,as two worlds between which there is nothing in common . . . the factof this contrast is universal.” Even the most basic religious ideas, hesays, are impossible without the notion of sacredness. Durkheim argues([1912:56]*1915:56; 1995:38) that religion always takes the form of adivision between sacred and profane: “But the real characteristic of reli-gious phenomena is that they always suppose a bipartite division of thewhole universe, known and knowable, into two classes which embrace allthat exists, but which radically exclude each other.”


His explanation for this universality, which is his explanation for theorigin of the idea of sacredness, and therefore of religion itself, is alsothe key to Durkheim’s entire epistemological argument. He argues thatthere is nothing in sensation that could account for this idea, but that theperception of moral force in certain ritual settings does produce the feelingof such a distinction. Unfortunately, the explanation for how the ritualproduction of moral force creates the feeling of sacredness, does not beginuntil the end of Book II, in the discussion of moral force, and is mostclearly articulated in Book III, Chapter Three, when Durkheim beginsto focus on rites instead of on beliefs.

Nevertheless, in spite of the distance separating these arguments,Durkheim’s claims with regard to the universality of the concept of sacred-ness are made quite clearly at the beginning of Book I. This is important,as it is impossible to reconcile his claims regarding the universality ofsacredness with the idea that Durkheim is articulating a sociology ofknowledge only and not also an epistemology. The universality claimedfor the concept of sacredness also cannot be reconciled with the idea thatDurkheim took an innatist position with regard to the categories, becausehe insists that the distinction between sacred and profane, as a logical cat-egory of thought, is an artifact of social practice and could have had noother origin.

The argument with regard to the dualism of the sacred and profane isnot merely related to Durkheim’s argument in the Introduction regardingthe dualism of human nature. It is the dualism between sacred and pro-fane that creates the dualism of human nature. For Durkheim (1914:335)human dualism is only an instance of the dualism between sacred andprofane.

Durkheim does not deny that the form in which the division betweensacred and profane manifests itself varies widely between societies. Nev-ertheless, he maintains ([1912:53]1915:54; *1995:36) that the divisionitself is universal: “While the forms of the contrast are variable, the factof it is universal.” Durkheim argues that all societies divide the world intothe same two parts, sacred and profane, and that the moral, or religious,significance of the two parts is created by social forces, and not derivedfrom individual perception of either natural or supernatural forces. Noris the idea of the sacred innate.

3.1.4 The Origin of the Idea of Sacredness

After arguing that the distinction between sacred and profane is the uni-versal defining characteristic of religion, Durkheim considers the ques-tion of what distinguishes the idea of the sacred from the profane? Here


again, he seeks to demolish the existing arguments that derive the dif-ference from natural experience, so that he can argue that what distin-guishes the two is their relationship to the social practices that create moralforce.

The first position that Durkheim challenges is the claim that the dis-tinction between the sacred and profane is modeled on certain hierarchiesfound in nature. Durkheim argues that the distinction cannot be basedon the idea of hierarchy. He points out ([1912:52]*1915:52–3; 1995:35)that not everything that is superior to something else in a hierarchy isconsidered more sacred: “It is not enough that one thing be subordi-nated to another for the second to be considered sacred in relation to thefirst.” According to Durkheim ([1912:52]*1915:53; 1995:35) “There aresacred things of every degree.”

Second, Durkheim points out that hierarchy is not a feature of nature.He argues that hierarchy only occurs in society. Therefore, even if hier-archy were the origin of the distinction, it would still have a social origin.

While the distinction between the sacred and profane is social, accord-ing to Durkheim, unlike most social distinctions, it has a special characteras a moral distinction. Hierarchy is a simple social distinction. There isnothing about hierarchy that implies a moral distinction. Therefore, heargues, the distinction between sacred and profane is not a simple mat-ter of hierarchy, it is also a moral distinction. According to Durkheim([1912:52]1915:53; *1995:35) “One might be tempted to define sacredthings by the rank that is ordinarily assigned to them in the hierarchy ofbeing.” But this, Durkheim says ([1912:52]1915:53; *1995:35), is incor-rect. Subordination versus superordination is not the same distinction assacred and profane.

After taking up the argument with regard to hierarchy, Third, Durkheimconsiders the argument that the natural and perceivable heterogeneitybetween sacred and profane things is the origin of the distinction. Thisdistinction would appear to offer a solution based on the perception ofnatural contrasts. For instance, he points out, the distinction betweensacred and profane often involves contrasts like black and white, nightand day, male and female. These are distinctions that are easy to perceive.He says ([1912:53]1915:54; *1995:36) “However, if the criterion of apurely hierarchal distinction is at once too general and too imprecise,”taking the natural contrast approach, “nothing but their heterogeneity isleft to define the relation between the sacred and the profane.” Just asthe distinction between sacred and profane can not be a simple matterof hierarchy, it also can not be a simple contrast of differences. Unlikenaturally contrasting phenomena, the sacred and profane are in absoluteopposition to one another.


According to Durkheim ([1912:53]*1915:53–4; 1995:36), “In the his-tory of human thought, there is no other example of two categoriesof things as profoundly differentiated or as radically opposed to oneanother.” The distinction cannot come from the perception of naturallyheterogeneous things, because there is nothing like it in nature. Durkheimmaintains ([1912:53]*1915:53–4; 1995:36) that “the traditional opposi-tion between good and bad is nothing beside this; for the good and thebad are only two opposed species of the same class, namely morals, justas sickness and health are two different aspects of the same order offacts, life, while the sacred and the profane have always and everywherebeen conceived by the human mind as two distinct classes, as two worldsbetween which there is nothing in common.”

The sacred and profane are so categorically opposed that a being canonly pass from one realm to the other by undergoing a “metamorphosis.”This, he says ([1912:53]*1915:54; 1995:36) “Puts into relief the essentialduality of the two kingdoms.” In this regard Durkheim discusses initiationrites and other rituals in which the profane is reborn as the sacred. Therebirth motif, according to Durkheim ([1912:53]*1915:54; 1995:36),proves “that between the profane, which he was and the sacred which hebecomes, there is a break of continuity.”

Unlike ordinary opposites that inhabit the same world, the sacred worldand the profane world cannot be allowed to touch one another. Mysticsand saints withdraw into monasteries in an attempt to withdraw com-pletely from the profane world into the sacred world because of,what Durkheim calls ([1912:53]*1915:55; 1995:36) the “antagonism”between the two.

Durkheim says ([1912:56]*1915:55; 1995:38) that this separation isone of the hallmarks of the sacred. We can know the sacred thing, becauseit is always and everywhere what the profane cannot be allowed to touch.According to Durkheim ([1912:56]*1915:55; 1995:38) “The two classescannot even approach each other and keep their own nature at the sametime.” This is clearly not a characteristic shared by natural contrasts.

Following various statements of the deep logical opposition betweenthe sacred and profane, Durkheim arrives once again at a statement of theuniversality of the distinction. “Thus,” says Durkheim ([1912:56]*1915:55–56; 1995:38):

We arrive at the first criterion of religious beliefs. Undoubtedly there are sec-ondary species within these two fundamental classes which, in their turn, aremore or less incompatible with each other. But the real characteristic of religiousphenomena is that they always suppose a bipartite division of the whole uni-verse, known and knowable, into two classes which embrace all that exists, butwhich radically exclude each other. Sacred things are those which the interdictions


protect and isolate; profane things, those to which these interdictions are appliedand which must remain at a distance from the first. Religious beliefs are the repre-sentations which express the nature of sacred things and the relations which theysustain, either with each other or with profane things. Finally, rites are the rulesof conduct which prescribe how a man should comport himself in the presenceof those sacred objects.

Based on the analysis thus far, Durkheim offers ([1912:56]*1915:56;1995:38) his own definition of religion: “When a certain number of sacredthings sustain relations of co-ordination or subordination with each otherin such a way as to form a system having a certain unity, but which is notcomprised within any other system of the same sort, the totality of thesebeliefs and their corresponding rites constitutes a religion.”

But, this definition is only preliminary and he returns immediatelyto the question of the origin of the idea of the sacred, a question towhich he will return again and again throughout the text. Durkheim asks([1912:58]*1915:57; 1995:39), “What has been able to lead men to see inthe world two heterogeneous and incompatible worlds, though nothing insensible experience seems able to suggest the idea of so radical a duality tothem.” His eventual answer will be that the moral forces created throughparticipation in certain rituals are responsible for the absolute boundarybetween the sacred and the profane.

3.1.5 Religion versus Magic

Because of his emphasis on practices, Durkheim needs to make sure thatTotemism is not confused with magic, a social form that is also basedmore heavily on practice than belief. It will be important for Durkheimto establish that Totemism creates a true “sacred,” not just a magicaloutcome. Therefore, the remainder of Book I, Chapter One, is taken upwith a differentiation of religion from magic.

This attempt to distinguish magic from religion leads to Durkheim’sdefinition of a “church.” While magic shares many aspects of religiouspractice and belief, and many of the rites and beliefs appear to be similar,magic does not have what Durkheim refers to as a “church.” Becausemagic works by “fooling” people and therefore, is not truly communalhowever much it is believed in, it doesn’t constitute a church. The feel-ing of sacred can only be empirically valid, for Durkheim, if it is gen-uinely enacted and shared. This is very important to Durkheim’s argu-ment because he considers the church to be the unity of the group, andmoral force cannot be enacted in the absence of group unity. Therefore,magic cannot enact moral force, because it has no church, no group unity,and thus cannot generate the categories, whereas religion can.


According to Durkheim ([1912:60]*1915:59; 1995:41) “The reallyreligious beliefs are always common to a determined group, which makesprofession of adhering to them and of practicing the rites connected withthem. They are not merely received individually by all the members ofthis group; they are something belonging to the group, and they makeits unity.” What allows religions to generate moral force is not that indi-viduals achieve faith as individuals, or that they practice according totheir beliefs as individuals. For Durkheim the key factor is the sharingof both beliefs and practices by a community that is unified by the col-lective practice of those same beliefs and practices. This does not meanonly that the community is unified because all members hold the samebeliefs, but rather, that through the collective performance of certain ritu-als they have created the same ideas in themselves as a collective achieve-ment. This collective creation, he says, is a characteristic of religion, notmagic.

Durkheim maintains that a magician does not have a communityof believers, but rather provides a service. According to Durkheim([1912:61]*1915:60; 1995:42) “The magician has a clientele and not achurch, and it is very possible that his clients have no other relationsbetween each other, or even do not know each other; even the rela-tions which they have with him are generally accidental and transient;they are just like those of a sick man with his physician.” On the otherhand, according to Durkheim, a true religion always binds together acommunity.

Sometimes magicians do gather together to practice and share their art.But, this does not constitute a church because it excludes participationby a participating audience. According to Durkheim, ([1912:63]*1915:60–61; 1995:42) “A church is not a fraternity of clients; it is a moralcommunity formed by all the believers in a single faith, laymen as well aspriests. But magic lacks such a community.”4

Durkheim also deals in this section with the question of whether indi-vidual cults are religions. This is similar to the magic issue because suchcults are celebrated by individuals and not churches. There is an essen-tial difference, however. Durkheim argues ([1912:64]*1915:61; 1995:43)that most individual cults, of patron saints and the like, are just part ofa larger religion: “The patron saint of the Christian is chosen from the

4 Of course, to the extent that religious leaders do not believe in the religious rites theypreside over they might also be said to be “fooling” people and thus performing magic inthis sense. The question is whether or not the rites they perform under such conditionswould still produce moral force. Possibly this is why the morals of religious leaders canbe so damaging to religious communities when they are exposed as lax. For instance, thepopes of the fifteenth century and the demise of the period of Catholic Empire.


official list of saints recognized by the Catholic Church.” According toDurkheim ([1912:64]*1915:61; 1995:43), this is true in all religions thathave personal cults: “In a word, it is the Church of which he is a memberwhich teaches the individual what these personal gods are, what theirfunction is, how he should enter into relations with them and how heshould honour them.” While magic exists outside of religion, individualcults are themselves part of, and defined by, religion. The essential pointis that individual cults cannot be the origin of the idea of the sacred ifthey depend on an existing religious context for their own existence.

Durkheim then offers ([1912:65]*1915:62–3; 1995:44) his final defi-nition of religion: “A religion is a unified system of beliefs and practices relativeto sacred things, that is to say, things set apart and forbidden-beliefs and prac-tices which unite into one single moral community called a Church, all thosewho adhere to them.” This means that religion is by definition collective.It also, by definition, relates to the sacred.

At the end of this section, Durkheim appends a footnote ([1912:ff65–6]*1915:ff63; 1995:ff44) that takes up the issue of an apparent dis-crepancy between this definition of religion and one he had given in anearlier article. His note in this regard is interesting, given the contro-versy over the role played by constraint in Durkheim’s approach to socialorder. Durkheim says that he has been interpreted as arguing that exter-nal constraint is the primary agent producing social solidarity. In severaldifferent footnotes in The Elementary Forms he challenges this interpreta-tion. In this particular footnote he explains that the obligatory characterof beliefs, which has been interpreted as external constraint, is derivedfrom something that is already by its very nature collective.

In other words, it is not the obligation that creates the collectivity,but the collectivity that creates the obligation. He argues ([1912:ff65–6]*1915:ff63; 1995:ff44) that “In this other work, we defined religiousbeliefs exclusively by their obligatory character.” But, this obligatory char-acter is derivative from the collective nature of religious practices. There-fore, he says “The two definitions are thus in a large part the same.”The first definition, according to Durkheim ([1912:ff65–6]*1915:ff63;1995:ff44), “neglected the contents of the religious representations toomuch.” In other words, the significance of the practices and beliefs, theirnecessity, the fact that persons could not be either human or rationalwithout them, was neglected. This created the impression that the obli-gation was an artificial one and thus entirely external and coerced. InDurkheim’s footnote, he says that he never meant that. He always meantthat the obligation was created by the internal necessity of the represen-tations, that they are necessary for reason. But, he believes that his earlierdefinition was misleading and that this one avoids confusion.


3.2.0 Animism versus Naturism

Having made sure his definitions in order, Durkheim says again at thebeginning of Book I, Chapter Two, that his task is to identify a setof religious practices that will explain the origins of the sacred andlay the foundation for his epistemological argument. Durkheim says([1912:67]*1915:64; 1995:45) that he will now “set out in search ofthis elementary religion which we propose to study.” In setting out onthis search however, Durkheim finds that the two most prominent formsin which archaic religious thought has been cast, Animism and Natur-ism, are highly problematic in several respects. According to Durkheim([1912:67–8]*1915:64–5; 1995:45), wherever one looks at religion onefinds two distinct forms of religious thought, side by side, but neverthe-less distinct. “The one” he says “addresses itself to the phenomena ofnature . . . for this reason it has been given the name of Naturism. Theother has spiritual beings as its object . . . This religion of spirit is calledAnimism.”

According to Durkheim ([1912:68]*1915:65; 1995:45–6), these twoforms of religious practice co-exist everywhere: “Now, to explain the uni-versal co-existence of these two sorts of cults, two contradictory theorieshave been proposed.” These theories are problematic from Durkheim’sperspective, however, because they each reduce the one sort of phenom-ena to the other. Either everything is cast as Naturism, or everything iscast as Animism. Thus, the two theories present conflicting versions ofreligious phenomena, even though both deal with the same anthropolog-ical facts.

Just as he earlier agreed with the “facts” on which dualism is based whiledisagreeing with the existing explanations, Durkheim agrees with the factson which Animism and Naturism are based, while disagreeing with boththeories. He argues that both perspectives have some validity, but thata simple reduction of the one to the other will not resolve the difficul-ties. As with his position on dualism, accepting the facts, for Durkheim,is not at all the same thing as accepting the theories. Before going onto elaborate his own position Durkheim intends to resolve the appar-ent contradiction between Naturism and Animism. Just as he soughtto resolve the contradictions within epistemology and classical dualismby showing that both sides corresponded to aspects of a real dualism,with Naturism and Animism, he will argue that as explanations they arewrong, but that the facts which each seeks to explain are correct. Theirapparent contradiction, according to Durkheim, comes from a failure torecognize that each focuses on only one side of the problem of humandualism.


In attempting to articulate the relationship between dualism and thetheoretical positions of Animism and Naturism, Durkheim makes fourbasic criticisms: First, he points out that both theories assume the ideaof sacredness and use it to explain what religion is, instead of explain-ing where the idea of sacredness comes from. Durkheim considers botharguments in considerable detail in order to make this point; Second, bothpositions argue that religious ideas come from perceptions of nature thatare distorted, either through limited intelligence, or through problemswith language. Durkheim’s position requires that these same ideas havea social origin. Therefore, he needs to show that the limited intelligenceand confusion of language arguments of the animists and naturists do notwork; Third, both positions assume that certain beliefs are the definingcharacteristics of religion. Durkheim needs to establish that in all casesthese ideas are not essential and really have their origin in the more funda-mental idea of the sacred and the profane. He will consider the relation-ship between beliefs and practices here and begin the argument that cer-tain practices are efficacious in their own right; Fourth, Durkheim arguesthat Totemism is a religion that demonstrates that the basic idea of sacredand profane comes before the sorts of beliefs that are usually consideredto be essential to religion. He will argue that it is the practice of Totemismthat generates the first distinction between sacred and profane. There-fore, all other religions are, of necessity, derived from Totemism; Fifth, allother approaches assume that Totemism is an individualist phenomenon.Durkheim’s epistemological argument depends on the idea that reli-gion is a fundamentally social phenomenon. According to Durkheim,Totemic religious practice generates the distinction between sacred andprofane. If Totemism itself had an individual origin, then the argumentwould be circular. Therefore, he must demonstrate the social origins ofTotemism.

3.2.1 Animism

According to Durkheim, Animism is an argument of an empiricist sortthat attempts to derive the idea of sacredness from individual experience.The primary theorist of Animism considered by Durkheim is Taylor.Spencer, who Durkheim says took up the argument from Taylor, is alsoconsidered. In order to claim that Animism is the most elementary form ofreligion, Durkheim argues ([1912:69]*1915:66; 1995:46–7), they mustshow that the idea of the soul is “formed without taking any of its ele-ments from an anterior religion; secondly, it must be made clear how soulbecome the object of a cult.” Finally, since the cult of nature and the cultof the soul are, according to Durkheim ([1912:70]*1915:66; 1995:46–7)


always found together, Animism, in claiming exclusively to come first,must also explain “how the cult of nature is derived from it.”

Beginning with individual experience, Durkheim argues ([1912:70]*1915:66; 1995:47), the animist explains that the ideas of soul and thesupernatural are derived from a confusion between dreaming and wakingstates. The animists point out that in many societies persons in a dream-ing state are said to believe that they actually leave their bodies whenthey dream. They argue that the perception that one can leave the bodywhen dreaming is mistaken for a soul that can leave the body. Hence thedevelopment of the idea of spirits.

There are several problems that Durkheim finds with this position.First, he points out that the idea of a soul inhabiting the body is notthe same thing as the idea of a spirit. The usual belief is that the soul isonly transformed into spirit when it leaves the body in death. The souls ofliving persons are not confused with spirits. Thus, according to Durkheim([1912:75]*1915:68–9; 1995:50), the first cults were ancestor cults, notcults addressed to the souls of living persons.

Second, the animists tend to compare primitives to children inmaking this argument; implying that limited intelligence explains theconfusion of dreaming and waking states. According to Durkheim([1912:95]*1915:85; 1995:64), Taylor says that, like the child, the prim-itive thinks of nature in human terms and populates nature with humanspirits. The spirits of inanimate things explain the phenomena of nature.Anthropomorphism thus provides the explanation for how Naturismdevelops from Animism.

Durkheim criticizes ([1912:95]*1915:85; 1995:64) this view, arguingthat anthropomorphism is a late development. Early man, according toDurkheim would not have been able to survive if he could not tell hisdreaming from his waking states. Furthermore, the idea that primitivepeoples cannot tell the difference between inanimate and animate objectsis, he says, absurd: even children can distinguish people from chairs. Onthis point, “but upon this point only,” according to Durkheim ([1912:76–77]*1915:70; 1995:50–51) even Herbert Spencer differs with Taylor:Spencer says that the anthropomorphism argument attributes less intel-ligence to primitives than to a cat who knows that if they touch a mouseit will run.

In his Principles of Sociology (p. 346 and cf 384), according to Durkheim([1912:76]*1915:70; 1995:51), Spencer attributes the development ofAnimism to “numerous errors due to language.” According to Durkheim([1912:77]*1915:71; 1995:51) these errors of language include “the lit-eral interpretation of metaphorical names.” He says ([1912:76]*1915:70;1995:51) that because it is the custom in many societies to name people


after things, the animists argue that there arose an ambiguity where itbecame “very difficult for a primitive to distinguish a metaphor from thereality.”

Durkheim responds that if language is going to be credited with causingthis problem it should be the case that the same problem with languagecauses similar problems in other areas of life. According to Durkheim([1912:77]*1915:71; 1995:51), “In order to explain a fact as general asthe religion of nature by an illusion, it would be necessary that the illusioninvoked should have causes of an equal generality.” There is, however,no evidence that the misunderstanding of metaphors has such univer-sality. As far as Durkheim is concerned this dispenses with Spencer’sargument.

Taylor, however, still remains, and is influential. He has also done animportant service to thinking about religion, according to Durkheim.Prior to Taylor, philosophers just assumed the presence of the soul. Theanimists have at least introduced the idea that the soul requires an empir-ical explanation. They have also introduced the idea that the concept ofthe soul, or self, may vary somewhat from culture to culture. Even if theexplanation the animists give is wrong, offering an explanation at all is,according to Durkheim, a step forward.

Durkheim points out several problems with Taylor’s argument. First,according to Durkheim, the primitive idea of the soul is much morecomplex than Taylor supposes and cannot be explained on his evidence.Second, even if it were not more complex than he assumes, his evidencestill does not explain it. Durkheim argues ([1912:79]*1915:73; 1995:53)that if the idea of the soul is supposed to come from dreams, and thisexplanation is supposed to hold for all historical cultures, then in orderfor it to be universal it should be the only possible explanation of dreams:“But if this hypothesis of a double is to be able to impose itself upon menwith a sort of necessity, it should be the only one possible, or at least,the most economical one.” Durkheim points out that there are otherand also more “economical” explanations; one might see a friend in adream and then discover on talking with that friend that they denied beingthere.

Third, according to Durkheim ([1912:80–1]*1915:74; 1995:54) theargument with regard to the confusion of waking and dreaming statesdoes not credit the primitive with sufficient intelligence to stay alive.Furthermore, experience would show that the senses do not always giveaccurate information. Why then, he asks ([1912:80]*1915:74; 1995:54),would the primitive blindly believe everything that happens when they areasleep, when they do not even always believe their senses when they areawake: “This blind credulity which is attributed to the primitive is really


too simple . . . He cannot live long without perceiving that even whenawake his senses sometimes deceive him. Then why should he believethem more infallible at night than during the day?”

Fourth, Durkheim argues that the animists have assumed an importancefor dreams in primitive life that requires an explanation. Even if everydream could only be explained by a theory of double, or soul, that doesnot explain why men felt the need to explain dreams at all. Durkheim asks([1912:82]*1915:75; 1995:55) why people should base all their beliefs onwhat happens when they are asleep instead of what happens when theyare awake?

Durkheim will suggest that the idea of the sacred precedes the beliefthat dreams provide evidence about sacred things. In order for the prim-itive to believe that dreams are important in this way, he argues, theywould already have to have the idea of the sacred. The fact that theprimitive attributes certain actions to his double, according to Durkheim([1912:82]*1915:75; 1995:55) does not mean that the “dream actuallyfurnished the materials out of which the idea of the double or the soul wasfirst constructed.” Durkheim refers to ([1912:82]*1915:75; 1995:55) thedream explanation as a “retrospective justification.” He argues that theanimists have mistaken a late development and its retrospective justifica-tion for the origin of the process.

Durkheim points out that the evidence with regard to dreams is quitecomplex. People distinguish between their dreams and do not con-sider them all to be of the same sort. Primitives only consider somedreams to be revelations, not all dreams. The dreams that are consid-ered to be revelations are considered as such because they have to dowith sacred things and sacred beings. Therefore, according to Durkheim([1912:84]*1915:76; 1995:56) “these dreams were possible only wherethe ideas of spirits, souls and a land of the dead were already existent . . .Thus, far from having been able to furnish to religion the fundamen-tal notion upon which it rests, they suppose a previous religious system,upon which they depended.” The identification of dreams as revelationsalready assumes, according to Durkheim, the development of the idea ofthe sacred. Therefore, dreaming cannot explain the development of theidea of the sacred.

Durkheim then comes to what he refers to as “the heart of the doc-trine.” The animists argue that it is death that makes the soul sacred. ButDurkheim asks ([1912:85]*1915:77; 1995:57) “why does this soul, bythe mere fact that it is now detached from the organism, so completelychange its nature?” What is involved, according to Durkheim, is a changeof the double from profane to sacred. He asks why death would effect thischange and answers that it would not. Durkheim argues that the idea of


sacred, or mana, must already exist, and then somehow it gets attachedto the idea of soul after death.

Mana, according to Durkheim (1912:88]*1915:79; 1995:59), is theidea of the sacred: “the distinctive character of every sacred being.”He offers numerous examples of how the idea of the sacred attaches tothe souls of those who already had mana while alive and not to everyoneequally.

Furthermore, Durkheim points out that if the animists were right, thecult of the dead should be the earliest cult and it isn’t. Durkheim offers([1912:90–1]*1915:80–2; 1995:60–1) an interesting definition of a cultas “periodically” recurring in support of this claim. Most observanceshaving to do with ancestors among primitive peoples do not have this“periodic” character. Only ancestors who had “mana” before death arehonored with such periodic cults in Australia, according to Durkheim.

Durkheim ([1912:97]*1915:86; 1995:65) makes one final criticismarguing that if the animist theory were true then all religious beliefs wouldbe false since they all have their historical origins in illusions: “If it weretrue, it would be necessary to admit that religious beliefs are so manyhallucinatory representations, without any objective foundation whatso-ever.” According to Durkheim ([1912:97]*1915:86; 1995:65), not onlysacred beings, but the idea of the sacred itself is a mistake from the ani-mist perspective: “From this point of view, then, sacred beings are onlythe imaginary conceptions which men have produced during a sort ofdelirium which regularly overtakes him every day.”

Durkheim rebels against this conclusion. For Durkheim religion cannotbe merely an illusion. If allowed, the conclusion would, he feels, also nul-lify the animist position. Durkheim asks ([1912:99]*1915:88; 1995:67),“What sort of a science is it whose principle discovery is that the subjectof which it treats does not exist?”

3.2.2 Naturism

Like Animism, Naturism explains the origin of the sacred through confu-sions, illusions and ambiguities, and Durkheim’s argument against Natur-ism is in many respects similar to his argument against Animism. In bothcases he argues that the causes which they cite could not have been theorigin of the idea of the sacred and that, in fact, their explanations assumethe prior existence of this idea.

However, the naturists not only approach the argument quite dif-ferently, but the naturists are scholars with very different backgroundsfrom the anthropologists and ethnologists who wrote about Animism.The scholars of Naturism, according to Durkheim ([1912:100]*1915:89;


1995:68) have confined themselves to the study of the great civilizations ofEurope and Asia. Durkheim says ([1912:100] *1915:89; 1995:68) that thenaturists have made a study of the apparent universality of mythic symbol-ism in these civilizations. Durkheim ([1912:101–2]*1915:90; 1995:69)cites Max Muller and Adalbert Kuhn in this regard. By comparing thegreat universal myths the naturists believed they could discover the sim-pler beliefs that underlay their development.

According to Durkheim ([1912:101–2]*1915:90–2; 1995:69–70),Muller, the principle naturist, on whose work he will focus his critique,commences from a principle that is apparently contrary to that of Ani-mism. Durkheim says ([1912:103]*1915:91; 1995:70) that for Mullerthat which Naturism represents is real: “‘Religion,’ he says, ‘if it is tohold its place as a legitimate element of our consciousness, must, like allother knowledge, begin with sensuous experience.’” For Muller, as forthe empiricists, all knowledge begins with perception. Here Durkheimconfronts empiricism again in the guise of Muller who, according toDurkheim ([1912:103]*1915:91; 1995:70), maintained that “there canbe nothing in the mind (or in beliefs) which was not first perceived.”5

But, Durkheim asks ([1912:103]*1915:91; 1995:69–70), “which arethese sensations which give birth to religious thought?” The naturistspoint toward the great forces of nature, fire, the sky. Many deities havenames that originally referred to natural phenomena, such as fire. Becausemany of these words had the same root, Muller concluded, according toDurkheim ([1912:104–5]*1915:92; 1995:70–1) that these words mustalready have designated deities before the various groups split off into sep-arate languages and cultures. From this, Durkheim says ([1912:104–5]*1915:92; 1995:70–1), Max Muller thought he could conclude that the“religious evolution of humanity in general had the same point of depar-ture.” This point of departure being the perception of natural forces.

While Muller considered natural forces to be the origin of a fear andawe that would lead to religious ideas, he argues, according to Durkheim([1912:106]*1915:93; 1995:72), that “religion really commences onlyat the moment when these natural forces are no longer represented inthe mind in an abstract form.” It is always to beings, and not to forces,that religious cults are addressed. Muller argues that the limitations oflanguage are responsible for this transformation.

Here, according to Durkheim, although they begin at the oppositeend, the naturists confront the same problem that was raised by the ani-mists. How does one explain the transformation of inanimate objects into

5 Durkheim presented the phrase in Latin “Nihil est in intellectu quod non ante fuerit insensu.”


animate spirits? Whereas the animists explained this transformation as aconfusion of dreams with reality and an incapacity to distinguish the ani-mate from the inanimate, according to Durkheim ([1912:106]*1915:93;1995:72) Muller argues that language alone “has brought about thismetamorphosis, by the action which it exercises upon thought.”

In applying names to sensations there is necessarily a human construc-tion, as concepts are not equivalent to things. Language also obeys thelaws of thought, not the laws of things. Therefore, applying concepts toperception cannot fail to produce deceptions. There is a lot of Hume’sproblem with concepts and general ideas in this argument.

Durkheim accepts ([1912:106]*1915:93; 1995:72) this limitation:“Thinking consists in arranging our ideas, and consequently in classi-fying them.” According to Durkheim ([1912:107]*1915:94; 1995:73)classification plays an important role in this process: “classifying is alsonaming, for a general idea has no existence and reality except in andby the word which expresses it and which alone makes its individuality.Thus the language of a people always has an influence upon the manner inwhich new things, recently learned, are classified in the mind and subse-quently thought of; these new things are thus forced to adapt themselvesto pre-existing forms.”6

The problem is compounded because the great root words arenot only general, but they express actions, not objects. For Muller([1912:108]*1915:95; 1995:74) this shows that “men generalized andnamed their principal ways of acting before generalizing and naming thephenomena of nature.” According to Durkheim ([1912:108]*1915:95;1995:74) because of their “extreme” generality, “these words could beextended to all sorts of objects which they did not originally include.”

Because language was developed as a way of speaking about humanagency, according to Durkheim ([1912:108]*1915:95; 1995:74) speak-ing of natural forces requires the metaphor of human action. The naturistexplanation is that this metaphor was then taken literally, and the idea thatspirits inhabit things was born. In Muller’s view, according to Durkheim([1912:109]*1915:96; 1995:74–5), “Language thus superimposes uponthe material world, such as it is revealed to our senses, a new world,composed wholly of spiritual beings which it has created out of noth-ing and which have been considered as the causes determining physical

6 Great care should be taken in interpreting passages such as this one. Durkheim is notsaying that it is only as names that words have signification. He is saying that “generalideas”, that is, ideas shared among people about phenomena that are also shared andnot particular, require names in order to be communicated about. He is not making thenominalist argument that the name itself is the reality of the thing. But, rather, that names“call up” shared ideas.


phenomena ever since.” This created a confusion of human agency withnatural forces.

Myths resulted from the attempt to make sense of the resulting confu-sion. If every word has to refer to a spirit or agency, then spirits had tobe invented to correspond to the words. Where there was more than oneword for the same thing, a family of spirits was invented. In this way theproperties of language, and in particular general ideas, created the ideathat spirits animate natural objects and forces.

Durkheim points out several problems with the theory; First, the rootwords are not as similar as Muller assumes, and may be the result ofborrowing through social exchange, rather than evidence of direct com-mon ancestry; Second, that ([1912:111]*1915:97; 1995:76) it is no longeraccepted that the root words allow us to “reconstruct, even hypothet-ically, the original language of the Indo-Europeans”; Third, the Vedicdivinities did not all have an exclusively naturistic character as Mullersupposed.

However, the big problem with Naturism as an explanation of religion,according to Durkheim, is that it assumes, on the one hand, that religionarose as a way of explaining nature, but, then on the other hand, theexplanation bears so little relation to nature as to be practically useless.Durkheim argues ([1912:113]*1915:98; 1995:77) that:

If, then, religion’s reason for existence was to give us a conception of the worldwhich would guide us in our relations with it, it was in no condition to fulfillits function, and people would not have been slow to perceive it: failures, beinginfinitely more frequent than successes, would have quickly shown them that theywere following a false route, and religion, shaken at each instant by these repeatedcontradictions, would not have been able to survive.

Durkheim admits that errors have been passed down to us throughhistory. But, he argues ([1912:113]*1915:98; 1995:77) that “they cannever perpetuate themselves thus unless they were true practically, that isto say, unless, without giving us a theoretically exact idea of the thingswith which they deal, they express well enough the manner in which theyaffect us.” Durkheim will come back to this point again and again withregard to the “truth” of religion. He will maintain that unless there is anunderlying truth that makes it practically efficacious to believe in religion,people would not have developed religion. It is that underlying truth thatDurkheim attempts to explain.

“Thus,” according to Durkheim ([1912:114]*1915:99; 1995:78), “it isonly in appearance that Naturism escapes the objection which we recentlyraised against Animism. It also makes religion a system of hallucinations,since it reduces it to an immense metaphor with no objective value.”


According to Durkheim ([1912:114]*1915:100; 1995:78) “the believer,like the delirious man, lives in a world peopled with beings and thingswhich have only a verbal existence.”

Durkheim faults Muller for one further thing. Instead of facing theconsequence of his theory of religion for contemporary beliefs in a deity,Muller made an arbitrary distinction between religions based on myths,which he called a disease of both language and the mind, and Christian-ity which Muller claimed was based on the truth. Durkheim, of course,does not allow this sort of “false” distinction. He will insist that all reli-gions have the same origin and that none are exempt from the need forexplanation.

Durkheim points out that it is hard to understand why people havebelieved in their myths if their purpose was to provide a practical guide tothe universe, because they rarely have useful effects. However, Durkheimsays ([1912:117]*1915:102; 1995:80) that if religion has a differentpurpose the problem might be avoided: “Let us suppose that religionresponds to quite another need than that of adapting ourselves to sen-sible objects: then it will not risk being weakened by the fact that itdoes not satisfy, or only badly satisfies, this need.” When rites fail peoplemake up reasons that are consistent with the beliefs behind the rite. But,for this to happen so frequently, Durkheim says ([1912:117]*1915:102;1995:80), “it is necessary that these religious ideas have their source inanother sentiment than that betrayed by these deceptions of experience,or else whence could come their force of resistance?” The real purpose ofreligion, Durkheim will argue, to create group unity and shared humanreason, must be fulfilled in order that other shortcomings should be soroutinely excused.

3.2.3 Final Critique and Transition to Totemism

At the end of Book I, Chapter Three, Durkheim returns to the generalcritique of Naturism with which he started. He argues that the attemptto explain the origin of the idea of the sacred in terms of the percep-tion of natural events involves a serious misunderstanding of the natureof the sacred. According to Durkheim ([1912:119]*1915:103; 1995:81),Naturism is based on a prejudice regarding the splendors of nature that isof relatively recent origin: “It is stated as an axiom that in the natural playof physical forces there is all that is needed to arouse within us the idea ofthe sacred; but when we closely examine the proofs of this proposition,which, by the way, are sufficiently brief, we find that they reduce to a prej-udice.” It is supposed by the naturists that nature is such a marvel thatpersons would naturally have worshipped it. But, according to Durkheim


([1912:119]*1915:103; 1995:81), “really, that which characterizes thelife of nature is a regularity which approaches monotony.” Monotony isnot awe inspiring, and according to Durkheim ([1912:119]*1915:103;1995:81), “uniformity could never produce strong emotions.” Thereare sometimes variations, but as Durkheim says ([1912:119]*1915:103;1995:81), “these momentary variations could only give birth to equallymomentary impressions.”

Even if nature were not monotonous, the entire argument is at faultDurkheim argues ([1912:119]*1915:103; 1995:81), because “admiringan object is not enough to make it appear sacred to us.” According toDurkheim ([1912:120]*1915:103; 1995:82), it is a misunderstanding ofthe nature of the sacred to equate it with ideas like admiration or awe:“We misunderstand what the religious sentiment really is, if we confoundit with every impression of admiration and surprise.” Even the argumentthat it is the “infinite” character of nature that impresses humans is insuf-ficient in this regard.

For Durkheim ([1912:120]*1915:104; 1995:82) the real question iswhere the idea “that there are in reality two categories of things, radicallyheterogeneous and incomparable to each other” comes from. Accord-ing to Durkheim ([1912:120]*1915:104; 1995:82) none of the naturalphenomena that the naturists point toward would lead to the idea of adualism: “Nature is always and everywhere of the same sort.” There isnothing about nature and natural forces that suggests such a disjuncture.Pointing out that natural phenomena vary in intensity and importancedoes not solve the problem. Durkheim argues ([1912:120]*1915:104;1995:82) that “sacred forces are not to be distinguished from profaneones simply by their greater intensity.” The sacred is not simply a moreintense version of something that would otherwise be profane.

The sacred is not only different in character from natural phe-nomena, it also has a different cause and purpose. Durkheim argues([1912:121]*1915:104–5; 1995:82–3) that “if religion really was bornbecause of the need of assigning causes to physical phenomena, the forcesthus imagined would have been no more sacred than those conceived bythe scientist to-day to account for the same facts. This is as much as tosay that there would have been no sacred beings and therefore no reli-gion.” Therefore, the purpose and origin of the sacred must be lookedfor outside of the domain of natural phenomena.

Durkheim’s final argument against Naturism is that even if he is wrongand the forces of nature could have inspired feelings of sacredness withregard to the forces of nature, this does not explain the origin of thesacred in primitive religions, because the primitive does not view nature


in such terms. According to Durkheim ([1912:121]*1915:105; 1995:83)the primitive “is in no way conscious that cosmic forces are so supe-rior to his own.” On the contrary, primitive peoples believe that throughtheir rites they exercise dominion over nature. According to Durkheim([1912:122]*1915:105; 1995:83), “His rites are, in part, means destinedto aid him in imposing his will upon the world. Thus, far from beingdue to the sentiment which men should have of their littleness before theuniverse [according to the naturists], religions are rather inspired by thecontrary sentiment.” Furthermore, Durkheim says, if the primitive didworship nature because they were awestruck by its overwhelming powers,then the first things they worshipped would have been the great naturalforces. In fact, the primitive worships the most trivial things.

3.3.0 Totemism

Durkheim opens Chapter Four, the last chapter in Book I, with an intro-duction to Totemism, a full discussion of which will take up the bulk ofBooks II and III. He opens ([1912:124]*1915:107; 1995:85) the chapterwith the observation that the positive conclusion of the preceding analysis(of Naturism and Animism) is that both perceptions of natural events,and human beings, have been ruled out as possible origins of the idea ofthe sacred. The only task remaining, before beginning his own argument,is to dispense with the accepted approaches to Totemism.

Unlike Animism and Naturism, which both assume an empiriciststarting point, the primary scholars of Totemism have, according toDurkheim, taken an innatist position, assuming an inherent tendencyto believe in the sacred. This is extremely unfortunate from Durkheim’sperspective, because he believes that Totemism is the only religion thathas the potential to explain the origins of the sacred. The principal ritualsofTotemism, according to Durkheim, are those that originally created thedistinction between the sacred and the profane. But, in taking an innatistposition the scholarly proponents of Totemism have, as far as Durkheimis concerned, taken a position that essentially negates the question of ori-gins, and hence the potential contribution of Totemism to epistemology.

The Totemists, and here he singles out Frazer, Durkheim identi-fies with a school of anthropology that he says ([1912:132]*1915:113;1995:91), assumes the innate religious nature of man, which, like Kant’sinnatism, completely does away with the question of origins. The con-sequence of this assumption of an innate nature of man was thatthe Totemists were not interested in documenting the development ofTotemism, nor in the details of local forms of religious practice. Totemists


were concerned first and foremost with documenting the universality ofthe phenomenon, in order to support their innatist hypothesis.

Following a review of the development of Totemism as a subject ofserious scholarly interest, Durkheim entertains an interesting method-ological critique. The approach has been too general and not attentiveenough to detail. Durkheim argues ([1912:132]*1915:113; 1995:91) that“since from this point of view, facts have an interest only in proportionto their generality, they consider themselves obliged to collect as largea number as possible of them . . . Our method will not be such a one.”Durkheim’s criticism is that in making an argument that Totemism is uni-versal scholars have focused only on generalities and essentially missedthe thing that is so important with regard to the epistemology.

Durkheim’s methodological criticisms are of particular interest inthat he seems to take a qualitative as opposed to a quantitative posi-tion. Whereas previously, he has criticized his opponents, on theoreticalgrounds, here the critique is primarily methodological. Durkheim is say-ing, in effect, that if what you want to get at has something to do withthe details of social action, in this case religious ritual, then by “heaping”up masses of data and considering only the common elements, the beastfor which you search will be lost. This is the part of the book in whichDurkheim advocates ([1912:134–5]*1915:114–5; 1995:92) what he laterrefers to as a single case method,7 arguing that if the detail to support theargument cannot be found in a close examination of a single case, thenit does not matter how many cases are piled up, it will not be found.Durkheim’s single case will be Australian Totemism.

Durkheim also believes that religion is universal. But, as he says in afootnote with reference to his methodological criticisms ([1912:ff132–3]1915:ff113; *1995:91ff), “Of course, I, too, consider that the principleobject of the science of religions is to arrive at an understanding of thereligious nature of man. But since I see it not as an innate given but aproduct of social causes, there can be no question of determining it whollyapart from the social milieu.” In other words, the “facts” about religion,taken out of the context of actual congregations of people engaging inreligions practices, are meaningless. The reasons why religion is universalare, for Durkheim, to be found in the details of religious practices, in theway in which particular rituals fulfill and create the rational social natureof human beings, not in general properties, in the form of myths, beliefs,and types of rite, that are shared by all religions.

7 There are several places in the text where Durkheim extols the virtues of a single case andargues that if a proposition cannot be demonstrated on the basis of a single case, then nolarger number of cases will prove it, a clear allusion to the empiricist position that eithersomething is there in a single case or it does not exist as such. See 9.1.1.


Durkheim’s discussion of Totemism, and criticisms of its major pro-ponents, will continue throughout Books II and III.

3.4.0 Conclusion

Both Animism and Naturism, according to Durkheim ([1912:123]*1915:106; 1995:84), “undertake to construct the idea of the divine outof the sensations aroused in us by certain natural phenomena, eitherphysical or biological.” But, he argues ([1912:123]*1915:106; 1995:84),this is impossible because “a fact of common experience cannot give usthe idea of something whose characteristic is to be outside the world ofcommon experience.” In order to explain how the ordinary could havebecome sacred, theorists have to bring hallucinations into the argument,making aboriginal religions appear to be absurd.

Durkheim argues that only a social source for the idea escapes from theobjections he has made with regard to the other theories. The particularsocial source that Durkheim will embrace as the explanation of the originof the sacred is Totemism. Durkheim builds his epistemological argu-ment around the religious rites that comprise Totemism. He has taken agreat deal of care to define religion in such a way that the primary moraldistinction, between sacred and profane is the defining characteristic ofreligion. All religions have made this particular distinction and he arguesthat it is the only distinction that holds across religions. The distinctionis not merely a belief. Durkheim argues that it has a material basis. It isan enacted and experienced distinction.

This is important for two reasons. First, for Durkheim, the creation ofthis distinction makes moral force and therefore, morality proper pos-sible. Therefore, the distinction is, in the broadest philosophical sense,the origin of moral reasoning, of human reason itself, and hence of reli-gion as Durkheim defines it. Therefore, its material basis is necessary forthe empirical validity of the categories of reason; Second, whereas mostdefinitions of religion would exclude Totemism, the definition offeredby Durkheim allows him to treat the rites that comprise Totemism asreligious rites. This is important because it is Totemic rites that makepossible his argument that the distinction between sacred and profane iscreated through religious ritual.

Durkheim does not need to work within the framework of religion. Hisargument would still be about the origins of moral reasoning even if he didnot consider Totemism to be a religion. But, the idea of the sacred is gen-erally considered to belong to the domain of religious studies and he needsthe idea of the sacred. Furthermore, the theories which Durkheim needsto challenge, in order to clear the way for his own analysis of Totemism,


are all considered to be theories of religion. Durkheim’s identification ofthe earliest forms of moral distinction and moral reasoning with religionis not an attempt to constitute social phenomena as religious phenomena.On the contrary, it allows him to secularize religion, essentially turningit into a laboratory for the development and study of moral reasoning insociety.

4 Totemism and the Problem of Individualism

Having argued, in Book I, that the distinction between the sacred andprofane, and through that distinction, ultimately, the concept of clas-sification, had its origin in the division of society into totems, it wasthen important for Durkheim to establish in Book II, two things aboutTotemism: First, that totems as emblem and practice had a collective ori-gin, and not an individual origin as many anthropologists had argued;and Second, that it is not the thing represented by the totem that is theorigin of the sacred, but rather the totemic emblem itself. Establishingthese two points also involves a Third argument, about moral force, theexperience that makes the emblem sacred.

If the totemic divisions, which give rise to the concept of classification,had an individual origin, Durkheim’s argument, that the category of clas-sification is available in ritual experience, would be a circular one. That is,he would have avoided the problem of individualism by basing the valid-ity of human reason on social practices and ideas which, ultimately, hadan individual origin. In order to avoid this problem, Durkheim engagesin extensive refutations of those arguments that treat Totemism as anindividual phenomenon.

An equally serious challenge to Durkheim’s position is presented bythe belief, held by most scholars of Totemism, that it is the things rep-resented by totems that initially give rise to the idea of the sacred. If,however, things are the origin of the sacred, then Durkheim’s argumentagainst Naturism, that there is no way of explaining how natural forcesand objects could generate the idea of the sacred, and against empiricismthat such ideas can not come from perception, would also hold againsthis own argument for Totemism as the origin of the sacred. The idea thatthings, like grubs, are the sacred objects of worship, would also presenta problem for Durkheim’s argument that Totemism is a religion; it beingdifficult to sustain the idea that grubs could be sacred in their own right.Durkheim responds to this challenge by arguing that it is only as an emblemof the emotionally experienced unity of the group it represents, as a moral

139


force, and not as the original totemic object, or thing, that the totem issacred.

Unfortunately, Durkheim draws these arguments out over nine chap-ters in Book II, and in the process the argument becomes so thin, andtakes so many side roads, that it tends to get lost. In the 1995 translation,Book II covers pages 99–299, two hundred pages of text (more in the 1915translation where it covers pages 121–333). In the original French it runsfrom page 141 to page 424, a total of 283 pages. In the process of argu-ing against the individual origin of totems, Durkheim provides detaileddescriptions of literally hundreds of individual and collective totems. Healso provides an extended discussion of totems as emblems or represen-tations, involving extensive descriptions of various food prohibitions, inorder to show that the totem, as emblem, or symbol, is more sacred thanwhat it represents.

While the argument always remains on course, Durkheim’s style ofwriting involves so many examples that it is easy for the reader to losethe thread of the argument. One can easily get the feeling while readingBook II that The Elementary Forms is a comparative study of totems,or a treatise on the evolution of totemic classifications. Indeed, this ishow the book has been interpreted for the most part. Yet, the variousarguments made are essential to Durkheim’s epistemology. Because of this, aclear understanding of the epistemological argument, as it makes its waythrough the various chapters of Book II, is essential.

The arguments regarding representation, individualism and moralforce, are not the only arguments in Book II; they are interwoven withother arguments throughout. It might be said, however, that there aretwo major lines of argument presented by Durkheim in Book II. TheFirst issue involves arguments related to the collective versus the individ-ual origin of Totemism, and related ideas regarding Durkheim’s theory ofrepresentation, and occurs primarily in the first five chapters of Book II.Then in Chapters Six through Nine, Durkheim introduces the Secondmajor line of argument, involving moral force as an origin of the sacred,and the logical ideas that result from the experience of moral force.

This chapter and Chapter Five, which follows, will be divided betweenthose two issues. This chapter will focus, for the most part, on the issuesraised in the first five chapters of Book II, while Chapter Five focuses onChapters Six through Nine.

In Chapter One, of Book II, Durkheim introduces the idea that thetotem is sacred only as a symbol, emblem, or representation, and thennot because it represents a sacred thing, but rather, because it creates a feelingof group unity. Chapter Two focuses on food prohibitions to make thepoint that the totem is more sacred than the thing it represents. Durkheim

Totemism and the Problem of Individualism 141

also takes the position that myths that suggest that the idea of the sacredoriginated with things are only retrospective accounts and, as such, cannotspeak to the issue of whether the thing is the original sacred object, orwhether the original sacred object is the totem itself as emblem. In ChapterThree, Durkheim argues that Totemism, like all religions, constitutes acomplete cosmology of the universe, which, he claims, proves its statusas a true religion. This cosmology, or moral organization of the clan, hesays, is also the origin of logical thought. Chapters Four and Five takeup arguments with regard to individual Totemism and sexual Totemism.References to Chapter Seven, which acts as something of a summary ofthe first five chapters, will also appear (although the full discussion ofChapter Seven will not occur until my Chapter Five).

Overall, Durkheim continues to argue that the underlying phenomenonrequiring explanation is the origin of the idea of the sacred, and the result-ing experience of dualism, which taken together, he believes, comprisethe experiences that many myths are addressed to explaining. He arguesthat the only way to explain how the totem came to be considered sacredis to treat the totem as a collective symbol of group solidarity and moralforce.

4.1.0 The Totem as The Origin of The Sacred

Durkheim opens Book II, Chapter One, with a discussion of the rela-tionship between beliefs and rites that takes up again his rather uniqueapproach, sketched in the Introduction, to beliefs, myths, and symbolsas retrospective constructs, or accounts, whose purpose is to “call up,”or explain, prior experiences (and of course, to get members to repeatthem). This discussion of beliefs as accounts appears at this point in thetext as an explanation of the division of the text between Books II and III;Book II dealing with the origin of beliefs and Book III with rites.

Durkheim’s discussion of the relationship between beliefs and rites atthis particular point in the text, however, displays an ambivalence. Herefers to rites as “in principle” coming from beliefs. But, whatever thoseoriginal beliefs might have been, they are forgotten, and existing beliefsand myths are only retrospective accounts. However, in other parts ofthe text he clearly states that rites come first and cause the feelings thatthe beliefs are retrospective accounts of. What I believe he means is that,in general, persons will only perform rites if they believe in them. Thatis the way that “in principle” beliefs come before rites. In spite of theambivalence at this point in the text, however, Durkheim continues toargue that rites precede and create beliefs, and he never, in fact, managesto discuss beliefs without contexting them in discussions of rites. This


means that his text is not actually divided in the way it appears, with a firstbook on beliefs and a second book on rites. Rites seem always, on closereading, to have pride of place. This privileging of practices over beliefshas, I think, been confusing to scholars, which is unfortunate because ithas very important contemporary implications for the understanding ofthe relationship between narrative accounts and practices in the generalpractice of sociology as a discipline, and will be discussed at length in myChapters Six, Seven, Eight and in the Conclusion.1

4.1.1 The Totem as Emblem

Durkheim also argues in Chapter One that the totem is sacred becauseit is a representation, and not because of the qualities of some thing thatit represents. In other words, it is not a symbol of something sacred, asother scholars have assumed. Rather, as a symbol it serves the functionof creating moral force; a sacred function. Through his rejection of natu-ralist explanations, Durkheim had hoped to cast the question of how theoriginal belief in the sacred was to be explained in an entirely new form.If Totemism is to be the answer to the question, as Durkheim posed it,it must not be understood as a belief system in which things are consid-ered to be the origin of the sacred. Unfortunately, in Durkheim’s day,most contemporary views of Totemism treated it as just such a belief sys-tem. Therefore, in order to base his argument on Totemism, Durkheimneeded to replace the interpretations of Totemism popular in his day withan altogether different interpretation of his own. For Durkheim it is onlyas a symbol essential to the creation of the feeling of moral force that thetotem could have come to be considered sacred.2

In arguing that the totem is more sacred than what it represents,Durkheim becomes involved in long discussions, continuing throughChapters One and Two, of forms of Totemism found in different groups.In each case the point of the discussion is that it is the totem that is wor-shiped, not the thing that it represents. Durkheim denies that the ideaof the sacred has its origin in things, as the naturists believed. Rather,he argues, it could only have been an emblem representing the unity ofthe group that first came to have a sacred character. If it is the emblem

1 This is an issue that, I believe, divides contemporary sociology: those who focus onnarrative accounts versus those who focus on enacted practices. There are relatively fewof the latter, as the statistical study of society generally shares a reliance on narrativeaccounts with more “cultural” forms of sociology.

2 This discussion of the totemic symbol is not just about Totemism, but also about meaningas a function of social processes versus as a function of representation. As such it outlinesa philosophy of language that bears a distinct resemblance to Wittgenstein’s emphasis onmeaning through use rather than reference.


that is sacred, the ground of the argument can be completely changed,from individual perception of nature and a corresponding theory of ref-erential meaning, to a consideration of the social function and significa-tion of symbols. Durkheim argues that this will allow the origin of thesacred to then be demonstrated empirically. Durkheim sums up his argu-ment in Chapter Seven, where he concludes ([1912:293–4]1915:235–6;*1995:208): “And so it is in totemic emblems and symbols that the reli-gious source is to be found while the real objects represented by thoseemblems receive only a reflection.”

Durkheim argues that the sacredness of the totem as an emblem isthe characteristic mark of Totemism as a religion. The totem, he argues,is sacred just because of its collective nature. It represents the unity ofthe group. According to Durkheim ([1912:315]1915:251; *1995:221),“The symbol thus takes the place of the thing, and the emotions arousedare transferred to the symbol. It is the symbol that is loved, feared, andrespected.” It is the collective “emotions aroused” that give the symbolits status as a sacred emblem. If the totem had an individual origin itcould not have generated collective emotions, and therefore, could nothave been the origin of the idea of the sacred.

Durkheim makes a comparison between the totem, and emblems andflags that represent the unity of modern groups and families. He argues([1912:158]1915:134; *1995:111) that “the totem is not simply a name;it is an emblem, a true coat of arms, and its resemblance to the heraldiccoat of arms has often been commented upon.” According to Durkheim([1912:158]1915:134; *1995:111), “the totem is in fact a design thatcorresponds to the heraldic emblems of the civilized nations, and eachperson is authorized to wear it as proof of the identity of the family towhich he belongs.” The idea is that even though people in modern societydo not worship things, they feel a reverence for emblems. The function of cre-ating group unity is in both cases the same. Durkheim ([1912:315–16]1915:252; *1995:222) goes on to argue that the “flag” is to modern coun-tries a kind of totem: “The soldier who dies for his flag dies for his country,but the idea of the flag is actually in the foreground of his conscious-ness. Indeed, the flag sometimes causes action directly.” Here he refersto suicidal missions to recover flags, or the flag as a battle cry. Americans,who have no monarchial symbol, have a national anthem about the flag.3

Durkheim says ([1912:315–16]1915:252; *1995:222) that “the totem isthe flag of the clan.” It is an object of worship because of the group

3 In the aftermath of the World Trade Center bombings the potency of the flag as a totem inthe context of modern division of labor society has been demonstrated beyond question.The flag has appeared everywhere and commentators refer to a “feeling” of unity and“pride” in attempting to explain the phenomenon.


unity that it symbolizes and helps to generate as a felt experience for thegroup.

Emotions become attached to the totem because it is necessary forthem to be displayed in a recognizable public and symbolic form in orderfor communication about the emotional experience that is shared to bepossible. The totem is present before the whole group when the entireclan gets together for initiations, at which time the totem is drawn onthe bodies of the initiates and the unity of the group and its membersis literally created. According to Durkheim ([1912:162–3]1915:136–7;*1995:114), “to that image the felt emotions attach themselves, for it isthe only concrete object to which they can attach themselves.” Wearingthe totemic symbol reinforces the feeling that the individual belongs to thegroup. It heightens the experience of group unity. According to Durkheim([1912:192–3]1915:136–7; *1995:114) the totem is so much a symbol ofthe unity of the group that it becomes a part of the members: “they imprintit in their flesh, and it becomes part of them.” Members of the groupbecome so identified with the totem through this process, using it to visu-alize the unity of the group that, according to Durkheim ([1912:162–3]1915:136–7; *1995:114), “generally the members of each clan seek togive themselves the outward appearance of their totem.”

In the same way that collective emotions, when attached to an objector mark, can render it sacred, collective emotions can transform per-sons as well. While persons, in their own right, are part of the pro-fane world, and themselves profane, as members of totems, they aresacred beings. They have been transformed from individual organismsinto human beings and feel the force of that transformation. Accordingto Durkheim ([1912:190]1915:156; *1995:133): “Each member of theclan is invested with a sacredness that is not significantly less than thesacredness we just recognized in the [totemic] animal.” It is through par-ticipation in the totem, and observing its rites, that the person becomessacred. Because of this, according to Durkheim ([1912:190]1915:157;*1995:134), the person regards themself as identical with their totem.The totemic organization gives them a way of thinking conceptuallyabout the unity they feel. In this manner the individual is turned intoa general category. If each individual is identical with some general cat-egory, then all the individuals are the same. This is not only the originof the general idea of classification, but Durkheim will argue in ChaptersEight and Nine, that it is also the origin of the idea of the soul and ofpersonality.

The totemic symbol not only represents the felt unity of the group,but it is necessary for the creation and recreation of that unity. It is the


emblem without which they are not a group. According to Durkheim([1912:170]1915:142; *1995:120), “churingas are not merely useful toindividuals; the collective fate of the entire clan is bound up with theirs.Losing them is a disaster, the greatest misfortune that can befall thegroup.” The various rites that contribute to group unity cannot be per-formed without the Churingas.

It is not the object itself that is sacred. Durkheim ([1912:172–3]1915:143–4; *1995:122) argues that, contrary to the accepted view, itis the totemic “image [that] sanctifies the object on which it is engraved.”The totem as emblem is so sacred that all that is necessary is to drawthe totem on a thing to render the thing sacred. Churingas, for exam-ple, sacred objects among the Aboriginals, are according to Durkheim([1912:172–3]1915:143–4; *1995:121) “merely objects of wood andstone like so many others; they are distinguished from profane thingsof the same kind by only one particularity: the totemic mark is drawnor engraved upon them. That mark, and only that mark, confers sacred-ness on them.” It is, in fact, the totem that renders things sacred andnot the other way round. According to Durkheim ([1912:176]1915:147;*1995:124), as sacred as the nurtunja pole is, it is torn apart after theceremony: “Thus it is no more than an image of the totem-indeed a tem-porary image-and therefore plays its religious role in this rite and in thisrite only.”

Having reviewed the evidence, Durkheim concludes that it supportshis argument that the emblem is more sacred than what it represents,and that it is the emblem itself about which the belief in the sacred hasformed, not things. He argues that only after the emblem became sacreddid the things that it represents come also to have a sacred characterthrough their association with the totem.

Furthermore, Durkheim argues that totems do not stand in a referen-tial relationship to things, such as the totemic object or animal. He pointsout that totems are quite abstract and do not “look” like what they rep-resent. They do not get their meaning from things. But, neither are theypurely conventional representations. They are representations of sacredfeelings, not of things, which explains their abstract character. Accordingto Durkheim ([1912:179]1915:149; *1995:126, emphasis added) thislevel of abstraction signifies that the Australian does not represent thetotem “in order to . . . renew the sensation of it” but rather, “becausehe feels the need to represent the idea he has [of it] by an outward andphysical sign.” Totems are, according to Durkheim, social and abstract,and not individual and sense driven in origin. They are not intended tocall out the sensation of an object they represent. Rather, they are true


significant symbols4 whose function is to call out in the mind of each par-ticipant the same collective idea that all have experienced together before:the idea of group unity, or moral force.

4.1.2 The Case of Dietary Laws

In Chapter Two Durkheim uses the case of dietary restrictions, to sup-port his argument that it is only as an emblem that the totem is sacred.He points out that if the totem was only sacred because of the thing itrepresented then: First, its role with regard to dietary restrictions shouldbe its defining characteristic, because its primary function would be toprotect the thing it represents; and, Second, the thing the totem prohibitsthe eating of should be considered more sacred than the totem. Yet, thethings represented by the totem can be eaten and are not sacred under allcircumstances, whereas the totem itself is always sacred and can never betreated as a profane object without incurring sanctions.

Totems are often characterized by dietary restrictions “The dietaryrestriction, however,” according to Durkheim ([1912:154]1915:131;*1995:108) “is not the characteristic mark of Totemism. The totem is,first and foremost, a name and, as we will see, an emblem.”5 Durkheimargues that the sacredness of the totem comes first from its character asa symbol of the unity of the group, and not from the diverse set of “reli-gious” functions, such as dietary restrictions, that it takes on to serve thetotemic object.

Furthermore, he argues, the dietary restrictions and other practicesthat often accompany Totemism cannot be the reason why Totemismdeveloped, or the reason why it is considered sacred, because the totemicobjects which such practices are designed to protect could not have gen-erated the idea of the sacred. Treating the thing as the sacred wouldrequire treating specific animals and plants as sacred, a position thatDurkheim believes he has already dispensed with in his critique of Ani-mism and Naturism. In spite of having argued this already, in Book I,however, Durkheim provides in this section, literally hundreds of exam-ples to show that the dietary laws demonstrate that it is not the animalsand plants themselves that are considered sacred, but rather the totemicemblem.

4 I believe that there is a significant relationship between this argument with regard tototems and G. H. Mead’s significant symbol. There is a very brief discussion of this at5.1.2.

5 The earlier translation reads (1915:131) “The totem is a collective label.” But the originalFrench (1912:154) clearly accords with the 1995 translation: “Le totem est, d’abord etavant tout, un nom et, comme nous le verrons, un embleme.”


4.1.3 Myths as Retrospective Accounts

There are those who argue that the fact that myths provide an oral historythat attributes the origin of the sacred to things, and mythical personages,and not to symbols, contradicts Durkheim’s argument. Durkheim doesnot deny that believers feel that the things represented by totems are them-selves sacred, and that the belief is elaborated in their mythology. But heargues that this is a later development based on mythical elaborations ofthe experience of the moral force of the totem. This involves Durkheimin a theory of myths and narratives as retrospective accounts that wouldcast doubt on all narrative accounts and always privilege direct observa-tion of practices over the narrative record (although it would also alwaystreat the narrative record as an indicator of underlying social relationsor functions). The totem communicates its moral force, through conta-gion, to the objects and things that it represents.6 This result is recordedin myths. The purpose of myths, in this regard, is to explain the feelingspeople have when they perform a rite, not to explain the origin of therite.

Although myths often offer explanations for the origin of the sacred,Durkheim argues that these explanations are invented after the fact andserve the purpose of reinforcing an already existing belief, rather thandocumenting its history. Because the function of myths is to reinforcepractices and not to be true, they often create a distorting picture of empir-ical reality. But, this does not impede their function as myths. Myths thatattribute a sacred character to things are secondary phenomena createdto explain the feeling of sacredness that becomes attached to those thingsthrough their relationship to the totem. Myths work if they reinforcebeliefs that are necessary to motivate the group members to perform therites. But, they are not the origin of those beliefs. According to Durkheim([1912:172–3]1915:143–4; *1995:121–2), “respect was not caused bythe myth; far from it. If men conceived this myth it was to accountfor the religious respect that those things elicited.” It is not beliefsand myths that cause respect, but rather, emotional feelings of unityand respect that cause the beliefs. For Durkheim ([1912:183]1915:152;*1995:129) it is a mistake to treat myths as evidence of origins: “as arule, the object of myths is to interpret the existing rites rather than to

6 According to Durkheim ([1912:339] 1915:269; 1995*:238–9ff) “Another cause accountsfor a large part of this fusion: the extreme contagiousness of religious forces. They invadeevery object in their reach, whatever it may be. Hence the same religious force can animatethe most dissimilar things, which by that very fact find themselves closely connected andclassified in the same genus. I will return to this contagion below [Book III Chapter One],while showing that it is related to the social origins of the idea of the sacred.”


commemorate past events; they are more an explanation of the presentthan they are a history.”

According to Durkheim ([1912:192]1915:158; *1995:135), there arecertain social needs that myths serve. They make sense of the rites so thatpeople can keep performing them, because they need to be performedin order to continue producing the feeling of sacredness. Therefore, thevariety of rites that require being performed will dictate the sorts of thingsthat will need to be explained by myths. That is why the same sorts ofmyths will be found everywhere. Not because of underlying structuresof either thought, or society, as Levi-Strauss argued, but rather, becausethe underlying need for social practices that create moral force is a humanuniversal.

Durkheim argues ([1912:288–9]1915:231–2;*1995:204) that his anal-ysis of Totemism makes it increasingly clear that myths are a secondarylevel of phenomena:

The impression increasingly is that the mythological constructions, even the mostelementary ones, are secondary products overgrowing a substratum of beliefs –simpler and more obscure, vaguer and more fundamental – that constitute thefirm foundations on which the religious systems were built. This is the primitivestratum that the analysis of totemism has enabled us to reach.

Myths are necessary because persons do not understand the real reasonswhy they must perform religious rites. But, they need reasons in order tokeep on performing them. Myths furnish these reasons.

Durkheim says ([1912:298–9]1915:239–40;*1995:211) that, “Mytho-logical interpretations would doubtless not have been born if man couldeasily see that those influences upon him come from society. But the ordi-nary observer cannot see where the influence of society comes from.”Even today, the commonsense view of social life seriously misrepre-sents the importance of social processes. Scientific advances, such asDurkheim’s own theory of religion, may make it possible in the future, toprovide more accurate explanations for necessary social functions. But,he says ([1912:298–9]1915:239–40; *1995:211), in the meanwhile, “solong as scientific analysis has not yet taught him, man is well aware thathe is acted upon but not by whom. Thus he had to build out of nothingthe idea of those powers with which he feels connected.”

For Durkheim, it is sociology that will ultimately deliver this under-standing of where these moral forces come from. Maybe then a new setof myths can be constructed that are less distorting. But they must stillserve the function of motivating persons to recreate sacred feelings andmoral forces.

Durkheim has sometimes been interpreted as privileging words, orideas, over things, with the result that Durkheim’s argument has been


considered idealist. But, that is quite the reverse of what Durkheim arguesin this section. Totemic symbols are not considered by Durkheim to beideas. On the contrary, he treats them as physical objects, or marks. Theirsignificance is not that they represent ideas either. What they signify arefeelings that were once shared between members of a group, or moreexactly, feelings that, once shared, made persons feel like members of agroup. They call up collective moments. Such shared moments can becalled up by a picture, things, or a word, and in so doing, the feelings thatwere once felt together are renewed. Certainly, the collective momentof the Kennedy assassination generated as much collective myth, and asmany “totems,” as Durkheim’s theory would have predicted. The deathof Lady Diana, Princess of Wales, provided a similar moment of collectiveemotion. The bombing of the World Trade Center is sure to live on inmyth and legend, generating totems for the ages.

Such moments of collective feeling are highly valued, and representingthem in a generally recognizable form enables entire populations to callup such moments for collective reflection. In the case of totems, there arerites that members can use to recreate such moments. In modern society,such occasions are increasingly left to chance, and the emblems of thembecome even more important, since the occasions which produced theoriginal feelings can often no longer be reproduced intentionally.7

4.2.0 Totemic Classification as a Logical System

In Chapter Three Durkheim introduces the relationship betweenTotemism, cosmology and logic, arguing that Totemism divides the uni-verse into kinds. There are two points to this argument. First, Durkheimis trying to establish that Totemism is a religion. He makes this argu-ment because he needs Totemism to be the origin of the sacred, and theidea of the sacred is closely associated with religion. Religions generallyoffer a cosmology. Totemism offers a cosmology. Therefore, because ithas a cosmology, he says, it has one of the essential characteristics of reli-gion. Second, he argues that a cosmology is a necessary consequence of

7 Nor in many cases would one want to recreate them, as, unfortunately, in modern times,such moments, perhaps through lack of constructing others, have become centered ontragedy. Heidegger interpreted the modern preoccupation with what he called “care,” thetrivial tragedies of everyday life or gossip, as evidence of the mediocrity of modern life.While Durkheim also refers to the mediocrity of modern life, he might argue that tragedyhas come to fill the void left by an absence of shared practice. This is not a failure of theindividual to achieve authenticity, as Heidegger thought, but rather, a more general failureon the part of society to reproduce the collective rites necessary to maintain fully rationalbeings capable of moral action. One might also argue, however, in Durkheim’s case, thata preoccupation with daily practice is not the trivial matter that Heidegger thought, butin fact substitutes for ritual practices as a shared context of situated practice in modernlife.


Totemism as a set of moral classifications. This is a different point. Notonly does Totemism have a cosmology, which is the defining character-istic of religion, but cosmology, and the logical thought that follows, inparticular the moral force of the category of classification, is a necessaryconsequence of Totemism, which is how, and why, cosmology came, forDurkheim, to be the defining characteristic of religion.

Durkheim says that Totemism, and all other religions, has this charac-teristic of creating cosmologies, because of the way the elements of thecosmology function as emblems of collective unity, and thereby producecategories of time, space and classification, not because they are createdto serve the purpose of either providing mythological explanations, orindividual explanations of religious rites.

Clan cosmology, Durkheim argues, follows directly from the waytotems create group unity by separating the sacred from the profane.He argues that there are always two original phratries in clan societies,from which derive the idea of genus or class. This is the case becausethere cannot be any social unity, no idea of the sacred, no moral force,without an original division into two. Without the first division there canbe no experience of moral force, nor experience of group unity againstwhich distinctions could develop. All subdivisions follow from the originaldivision into two. The unity of the logical systems created by Totemismreproduces the moral divisions in the society. But, these “logical” sys-tems are always generated by a felt moral unity, and not by a logical andperceived unity. As Durkheim argues ([1912:208]1915:171; *1995:147),a “feeling of similarity” is one thing and the “notion of kind another.”The first is a generic image based on individual perception, the second isthe result of experiencing the moral divisions in the universe created byTotemism in a company of assembled others.

According to Durkheim ([1912:213]1915:173–4; *1995:150) totemicclassifications are moral and not merely logical classifications: “By theirjoining, then, the people of the clan and the things classified in it forma unified system, with all its parts allied and vibrating sympathetically.This organization, which might at first have seemed to us purely logical,is moral at the same time.” Moral classifications are those animated and“vibrated” by moral force, not by the moral beliefs of a particular society, orby the sensations of individual members. He says ([1912:213]1915:173–4;*1995:150) that “The same principle both animates it and makes itcohere: That principle is the totem.”

Durkheim ([1912:205–6]1915:169–70; *1995:145) argues that inPrimitive Classification he “showed how these facts illuminate the man-ner in which the idea of genus or class too form among humans. Theseclassifications are indeed the first that we meet with in history.” If thedistinction between the sacred and profane is ignored, Durkheim might


be interpreted as saying that the social division into totems leads throughthe learning of these conventional divisions to the idea of classification. Theargument of Primitive Classification is generally interpreted in that way.But, the entire argument is based on the felt moral force of the sacred andprofane during enacted practices. According to Durkheim, none of thesedivisions could occur at all without the first division into sacred and pro-fane, and none of them could be experienced without moral force. Thecategory of classification is not learned in any conventional sense, it isdirectly felt in the moral force of the ritual enactment of the totem. Thedivision between sacred and profane creates the first possibility of moralforce, and that, in turn, creates all the social divisions that are to follow.Therefore, the totemic cosmology is truly sacred, because it representsdivisions between things on the basis of moral forces.

The division between sacred and profane, and the resulting moralforces, are the key to the difference between Durkheim’s epistemologyand his sociology of knowledge. Therefore, when the discussion of clas-sification takes place, as in these chapters, with little mention of the ideaof sacred and profane, it tends to create the impression that Durkheim ismaking a sociology of knowledge argument for the origin of the categoryof classification. But, the argument occurs in stages. First, in Book I,Durkheim argues that without the first distinction between sacred andprofane there could have been no emotional experience of classificationat all, and thus no moral force. Then, in Book II, he argues that classifica-tions have a social and not an individual origin. This section, although itspeaks of the category of classification as coming directly from the socialdivisions, that is, the cosmology in a given society, assumes the priordevelopment of the distinction between sacred and profane, which is thebasis of that cosmology. Book II quite naturally assumes the argumentsof Book I. There must always be two phratries to start with, because thefirst division between sacred and profane must always be the beginning.With the aid of felt moral force the totemic classifications create, and thenmirror, the organization of the society.

According to Durkheim, one can get a feeling of similarity from rela-tions that are not possessed of moral force. But, these various perceptionsof similarity and difference are only content. The category of classificationis, according to Durkheim ([1912:209]1915:172; *1995:147–8 emphasisin original), a framework for thought:

Kind is the external framework whose content is formed, in part, by objects per-ceived to be like one another. The content cannot itself provide the frameworkin which it is placed. The content is made up of vague and fluctuating imagescaused by the superimposition and partial fusion of a definite number of individualimages that are found to have elements in common. By contrast, the framework isa definite form having fixed contours, but can be applied to an indefinite number


of things, whether perceived or not and whether existing or possible. . . . Thisis why a whole school of thinkers refuse to identify the idea of kind with thatof generic image, and not without reason. A generic image is only the residualrepresentation that similar representations leave in us when they present them-selves in consciousness at the same time, and its boundaries are indeterminate;but a genus is a logical symbol by means of which we think clearly about thesesimilarities and others like them. Besides our best evidence of the gulf betweenthose notions is that the animal is capable of forming generic images, whereas itdoes not know the art of thinking in terms of genera and species.

This discussion of a “fixed” framework for thought, has led someDurkheim scholars to assume that Durkheim is positing an a priori con-cept of classification as a framework for thought. But, he has alreadyrejected the validity of an a priori approach. Durkheim argues that thefeelings of moral force generated by totemic symbols and practices giverise to those ideas that do not have an origin in generic images or indi-vidual perception. These ideas, this framework for thought, which inthis case he is calling “logical,” but in other cases he identifies withreason, is not a priori, it is caused, he says, by the emotional experi-ence of moral force.8 It is the function of religious rites to supply theexperience of moral forces necessary to create this “fixed framework forthought.”

4.3.0 The Status of Individual and Sexual Totemism

In addition to the Totemism that belongs to the whole clan, Totemismalso appears in various individual forms. According to Durkheim, anothermistake scholars of Totemism have made is in assuming that the individ-ual form is the original source of the clan form. Just as philosophers madetheir mistake by attempting to build a whole logical system on individ-ual perception, scholars of religion, according to Durkheim, have failedto explain the origin of the sacred, because they begin with individualreligious experience, and then attempt to explain the development ofcollective forms of religious practice on the basis of that individual expe-rience. Durkheim argues that individual experience could never have pro-duced collective forms. Collectivity, whether of logical thought, or reli-gious thought, presupposes a generality, or consensus of thought, that, heargues, cannot be explained as an aggregation of individual experience.For Durkheim, individual religious experience, like logic, must have its

8 Logic and reason are not at all the same thing the way Durkheim writes about them.Therefore, his use of the term logic in this section is somewhat confusing with regard tohis overall argument. Nevertheless, I have avoided the temptation to attempt clarification.In this case lack of clarity may do much to explain the consequent misunderstandings.


origin in the collective life in order to be even minimally communicable,let alone empirically valid.

Durkheim’s critique of an individualist interpretation of totems appearsprimarily in Chapters Four and Five of Book II. In Chapter FourDurkheim describes various forms of individual Totemism, particularlythose found among Native Americans, arguing that individual Totemismis a later form that does not appear in all of the earliest societies (which iswhy he can’t use Australian examples). In Chapter Five Durkheim arguesthat not only is Totemism not an individual phenomenon, but that thevery structure of clan societies could not have come into being withoutthe totem. Therefore, the totem and the society must have developedtogether. The totem is thus sacred, not just because it represents the unityof the group, but also because it creates that unity. In the second part ofChapter Five Durkheim reconsiders various theories of Totemism (ini-tially reviewed at the end of Book I), and argues that in order for them tobe true, a “thoroughgoing idiocy” on the part of Aboriginals would haveto be assumed.

4.3.1 The Problem of Individualism

While Durkheim is, in one sense, dealing with the problem of individual-ism in all aspects of his argument, individualism poses a particular prob-lem with regard to the argument he wants to make for Totemism. If totemsfirst arise as ideas in the individual mind, as a result of individual attemptsto explain and understand the world, one popular argument at the timeDurkheim wrote, then, the concept of classification which they give riseto would still only be a generalization from individual experience, albeitindirectly (i.e., individual experience –> totems –> general ideas). Thisis the argument above all else which would threaten Durkheim’s position([1912:251–9]*1915:204 1995:178): “What we must know before every-thing else is whether or not the individual totem is really a primitive fact,from which the collective totem was derived; for according to the replygiven to this question, we must seek the home of the religious life in oneor the other of two opposite directions.”

In seeking the “home” of religious life, the question of individual ver-sus collective origins is the determining one. The positions of Durkheim’sopponents, empiricist individualism (Naturism and Animism), or an indi-vidualism which posited an a priori religious nature of man, would all besupported if the totem had its origin in the individual, as either basedon individual impulse or innate religious tendencies. Durkheim’s other“direction,” the sociological position, could only be established on a col-lective foundation.


Therefore, once Durkheim feels that the status of the totem as emblemis secure, and that the status of Totemism as a religion has been estab-lished, the demonstration of the collective origins of the totem becomeshis main objective.

4.3.2 Empirical Evidence Does Not Support Individualism

In Chapters Four and Five, Durkheim argues that the evidence withregard to individual and sexual Totemism does not support the argu-ment that Totemism had an individual origin. Individual Totemism is veryimportant in some groups and does make use of sacred emblems, just asclan Totemism does. But, Durkheim argues that individual Totemism isjust a special case of clan Totemism, and not the origin of clan Totemism.He points out ([1912:233]*1915:191; 1995:165) that there are manytribes in Australia “where the custom seems to be absolutely unknown.”Sexual Totemism, on the other hand, in which the two sexes are dividedinto two sexual “corporations,” with all the accompanying emblems andproscriptions of clan Totemism, Durkheim considers to stand halfwaybetween individual and collective Totemism. Durkheim suggests thatthe marked distinction, or moral force, separating men and women insuch cultures, who often even live apart, is the origin of this sort oftotemic division, not the sexual attributes of persons, per se. He pointsout ([1912:234]*1915:192; 1995:166) that, by contrast with individualTotemism, sexual Totemism “is found only in Australia and only in asmall number of tribes.” It is thus found in earlier forms of Totemism,but has still not developed in all of them; a demonstration, according toDurkheim, that it is of lesser antiquity than clan Totemism.9

Durkheim argues that if individual Totemism were the primary form itwould be found in all very simple societies, while clan Totemism wouldnot be as universal. The opposite, he says, is the case. Furthermore,Durkheim ([1912:253]*1915:205; 1995:179) argues, the division betweentotems, within clans, indicates a collective origin: “The carefully regulatedway in which the totems and sub-totems are divided up, first between thetwo phratries and then among the various clans of the phratry, obviously

9 There has been a great deal of criticism of Durkheim’s assumptions about which forms ofTotemism are simplest and which are earliest. Happily, for purposes of this analysis, noneof that matters. What we are interested in here is in trying to understand the argumentthat Durkheim was making and why he felt the need to discuss the various forms ofTotemism in so much detail. If the facts are otherwise, then the argument may turn outnot to work. But the first task is to understand what the argument, in fact, was. It is alsoimportant to point out that Durkheim’s data could be wrong in many respects withoutthat effecting his argument. All Durkheim really needs is for individual totems to showevidence of being a later development than clan totems.


presupposes a social agreement and a collective organization.” If collec-tive totems were merely the result of an aggregation of individual totemsthey would not be expected to have such a complex and orderly organiza-tion. Nor would they be expected to display the same basic organizationeverywhere.

Totemism exhibits an overall sameness in the organization of phratriesand clans, according to Durkheim, which could not have its basis in theindividual. Durkheim asks ([1912:253]1915:205; *1995:179), if individ-ual Totemism really were the origin of collective Totemism, then howdoes one explain that “except where totemism is in decline, two clansof the same tribe always have different totems. ” He says that the origi-nal division into two, that creates the original distinction between sacredand profane, always leaves this mark on clan structure in the form of twoopposing phratries. It would not be there if Totemism had an individualorigin.

According to Durkheim, individual totems, where they do exist, servea completely different function from collective totems. Collective totemsorganize the group and signify the individual’s place within the group.Durkheim argues ([1912:238]*1915:194; 1995:169) that clan totems arethe primary mode of social organization in simple societies: Members ofa clan are not all related to one another by blood “their unity comessolely from . . . their participating in the same totemic cult.” Individualtotems, on the other hand, are chosen by, or for, individuals, usuallyafter initiation, and signify special accomplishments, characteristics, or apatron of the individual.

Unlike collective totems, according to Durkheim ([1912:229]*1915:188; 1995:163), individual totems do not always presuppose a commonprinciple uniting all individuals who have the same totem. The individualtotem has a utility, or special relationship, with the individual. It is chosenby, or for, the individual, at a specified point in their life. Taboos, andparticularly dietary restrictions, with regard to individual totems, are alsovery different than those associated with collective totems. Durkheimargues ([1912:229]1915*:205; 1995:179) that:

Collective totemism cannot be deduced from individual totemism except by amisunderstanding of the differences separating the two. The one is acquired bythe child at birth; it is a part of his civil status. The other is acquired during thecourse of his life; it presupposes the accomplishment of a determined rite and achange of condition.

Collective Totemism cannot be derived from individual totems, becauseindividual Totemism serves different needs, and therefore, takes differ-ent forms. According to Durkheim ([1912:257]*1915:208; 1995:182–3),


“the cult which the individual organizes for himself in his own inner con-science, far from being the germ of the collective cult, is only this latteradapted to the personal needs of the individual.” Individuals choose theirindividual totems according to personal characteristics, not because oftheir position in the clan organization.10

If individual Totemism had developed first, he argues, it should be themost general form of Totemism. However, Durkheim points out that notall societies which have collective totems even have individual totems.On the other hand, all totemic societies have collective totems. He alsopoints out that those societies which do not have individual totems, orin which they are least well developed, are the simplest societies (thoseof Australia), not the most advanced. Durkheim ([1912:254]*1915:206;1995:180) argues that if individual totems were first they should be themost highly developed in simple societies and occur everywhere when,in fact, the reverse is the case: “In the great majority of the tribes, it[collective Totemism] alone is found, while we do not know a single onewhere individual totemism alone is practiced.”

Some scholars have pointed out that Australian totems are often par-ticular to a single group and have offered this as evidence of individualorigins. According to Durkheim, however, the particularism of totemsamong some aboriginal groups is not due to their having an individualorigin, nor to the failure of the primitive mind to develop analytic think-ing, as is sometimes claimed. Rather, Durkheim ([1912:280]*1915:225;1995:198) argues “it is the nature of the social environment which hasimposed this particularism.” Clans develop totems that reflect the rela-tionship between the clan, as a group, and the environment in which itfinds itself.

Durkheim also argues that the totems of different clans enjoy a largedegree of autonomy in aboriginal religious organization because of thedistance that often separates clans. This autonomy is even sufficient toprevent the idea that separate totems share the same principle of mana,or moral force, and mana appears in very different forms in different clangroups. But, according to Durkheim, this particularism does not have itsorigin in individualism, but rather in the autonomy of the clan totem,which is a collective, not an individual, totem.

For Durkheim the belief in individualism itself, that he confronts in hiscritics, is a product of the socio-economic relations of modern society,as much as totemic particularism was a product of the socio-religious

10 Sometimes, according to Durkheim, powerful parents arrange to have their sons choosecertain more powerful totems in order to maintain social status. But, this is an option,not an automatic consequence of social status.


organization of aboriginal society. Neither have their origin in the indi-vidual. Like Marx, Durkheim argues that the fundamental philosophicalproblem of the age is mistaking the individualism that is a result of livingin a division of labor society, for the inherent nature of human beings,and then beginning all arguments from there. Both individual and collec-tive totems correspond to the needs of social actors and social practiceswithin an existing social context. They have nothing to do with the indi-vidual per se. According to Durkheim, individualism is always a productof social relations, not the reverse. Not understanding this has, in hisview ([1912:246]*1915:200; 1995:174–5), misled scholars: “The desirefor an undue simplicity, with which ethnologists and sociologists are toofrequently inspired, has naturally led many scholars to explain, here aselsewhere, the complex by the simple, the totem of the group by that ofthe individual.”

4.3.3 The Totem Creates the Unity of the Group

In Chapter Five Durkheim argues that clans could not have come intobeing in their known forms without the totem. The totemic emblem,he argues, makes Totemism and, in fact, all organized social life, possi-ble. According to Durkheim Totemism exists as soon as the first divisioninto two primary clans, or phratries (which required totemic emblems)exists. The first division into two, the first moral dualism, according toDurkheim ([1912:238]1915:194; *1995:169), creates the clan: “Organi-zation based on clans is the simplest we know, for it exists in all its essen-tials the moment a society has two primary clans.” The first moral divi-sion creates the first experience of sacred and profane. When the moralforce of this experience is given conventional symbolic representation,Totemism exists. In referring to clans Durkheim ([1912:238]1915:194;*1995:169) says that “their unity arises solely from having the samename and the same emblem, from believing they have the same rela-tions with the same categories of things, and from practicing the samerites.”

As a result, the totemic organization into clans is as simple an orga-nization as can possibly be found. Therefore, if the origin of the beliefin the sacred nature of the totem can be identified, Durkheim believes([1912:238]1915:194; *1995:169) that he will have discovered “by thesame stroke what kindled religious feeling in humanity.” Because it is thesimplest form of social organization, Durkheim believes that the prin-ciples underlying Totemism hold for all social organizations. The firstprinciple which he is seeking, the principle that is not only the originof religious feeling, but also the origin of both human reason and social


organization, is the moral force of the distinction between sacred andprofane.

According to Durkheim ([1912:261]1915:211; *1995:185), “There isreligion as soon as the sacred is distinguished from the profane, and wehave seen that totemism is a vast system of sacred things. So to explain itis to show how those things come to acquire that trait.” He will add thefinal ingredient to the equation in Chapter Six, with the introduction ofmoral force.

4.3.4 Individualism Implies That Aboriginals Lack Intelligence

In the second section of Chapter Five, Durkheim argues that presumingtotems to begin with the individual creates an additional problem. Itrequires believing that aboriginals are “thoroughgoing idiots,” as he says([1912:250]1915:203; *1995:177), because the totems they have chosencould have no possible individual value. Durkheim consistently defendsthe intelligence of Aboriginals in this and other sections. He argues that itis the attribution of individual choice to Totemism that makes Aboriginalsappear unintelligent. According to Durkheim ([1912:244–5]*1915:199;1995:173–4), if totems were chosen for utilitarian purposes, to assure¨the aid of supernatural beings for instance, as some claim, “then theyshould preferably have addressed themselves to the most powerful ofthese.” Durkheim points out ([1912:244–5]*1915:199; 1995:173–4) thatfar from representing the most powerful entities, totems often representthe most humble: “The beings with whom they have formed this mystickinship are often among the most humble which exist.”

Some theories of Totemism argue that the belief system results fromvarious confusions in the mind of Aboriginals which, Durkheim argues,are inconsistent with the basic ability to survive that Aboriginals com-mand better than persons in industrialized countries. This portrayal ofthe primitive as inept occurs, he says, because, if one begins from anindividual perspective, the choice of particular totems must be explainedon the basis of individual choice. Collective totems, not having been chosenby individuals for individual purposes, and not being derived originally fromthe individual totem, cannot be explained in terms of individual choice withoutdoing an injustice to the individual intelligences that such a choice is attributedto. Durkheim argues ([1912:250]*1915:203; 1995:177) that the sorts ofarguments made to support an individual origin for totems such as bugsand rocks, are not only fantastic and not supported by the facts, butrequire us to think of primitive persons as lacking in intelligence: “Itpresupposes a thoroughgoing idiocy on the part of the primitive whichknown facts do not allow us to attribute to him.”


This belief that primitive people are lacking in intelligence is caused bysupposing that totems are chosen by individuals because of the utility ofthe things they represent. Durkheim argues that a close look at totemicsymbols reveals that they often do not consist of animals or things whichcould be of any possible use to an individual. Therefore, he concludes(again) that they must have some other origin. To argue, as some promi-nent theorists did, that primitive peoples believe that they can hide theirsouls in some species of bug or vegetable when they are in danger, andthen come to believe that they are part of the thing they are hiding theirsouls in, and forget how to differentiate themselves from it, and con-ceptually remove themselves from the bug or rock, is to treat them likeidiots. Durkheim ([1912:249–50]*1915:202; 1995:177) calls such theo-ries “ingenious fabrications.”

But, Durkheim says ([1912:260]1915:210; *1995:184), this theoryalso begs the question of where Totemism comes from: “If it is to beimaginable that human souls are the souls of animals or plants, it mustalready be believed that man takes what is most fundamental to him fromeither the animal or plant world.” In other words, man would not try tohide himself in a bug in the first place unless he already had some ideathat it was possible to do so. According to Durkheim, believing in such apossibility requires the prior development of Totemism.

However, and here he enters into another level of argument, even ifthe idea of hiding in a bug were possible without Totemism, a personaldesire to hide in a bug would still not be sufficient to explain how theidea of hiding in a bug (or the bug itself) came to be thought of as sacred.According to Durkheim ([1912:260]1915:210; *1995:184), “The vaguebelief in an obscure kinship of man and animals is not enough to founda cult. This merging of distinct realms cannot lead to dividing the worldbetween sacred and profane.” What defies explanation on these groundsis how the sacred nature of the totem can be derived from its allegedpractical utility to the individual.

Durkheim also criticizes ([1912:262]1915:211–12; *1995:186)Andrew Lang’s argument that confusions of language explain the devel-opment of totems. Durkheim attributes to Lang the argument that “fromthe moment organized human groups come into existence, each feels theneed to distinguish itself from the neighboring groups with which it isin contact, and, to this end, gives them different names.” Names of ani-mals and plants are used because they are easy to represent by drawings orgestures. But, according to Durkheim ([1912:263]1915:212; *1995:187)Lang believes that primitive peoples cannot differentiate a name from theobject it signifies. As a consequence, when they have named themselvesafter a thing, they come, through linguistic confusion, to believe that they


are identical with that thing. Myths then spring up to explain this strangeplant and animal nature of man. But it all starts out with a confusionabout language. Durkheim will say that this also assumes that primitivepeoples are idiots.

Durkheim goes on to argue ([1912:263]1915:212; *1995:187), as hedid with the “hiding in a bug” theory, that even if this explanation, ofconfusing names with things, were accepted, it would not address thequestion: “From whence does the religious character of totemic beliefsand practices arise?” If Totemism began with a linguistic confusion, thenhow did totems come to be thought of as sacred? The theory not onlystrains credulity and does not accord with the facts, it does not explain theone fundamental thing that, according to Durkheim, must be explained:What makes the totem an object of reverence? Man’s belief that he isan animal or plant, whatever its origin, does not explain why he imputesamazing virtues to that species when he does not impute them to himself.

Lang, like Frazer, avoids this problem by denying that Totemism is areligion, and therefore denying that Totemism contains an idea of thesacred. He says that Totemism becomes sacred later when it comes incontact with more advanced ideas of deities, and he cites various mythsas his source. But this explanation, according to Durkheim ([1912:264–5]1915:213–14; *1995:188), is also problematic because “those very mythsare in conflict with Lang’s idea of Totemism. If the Australians had seenthe totem as nothing more than a human and profane thing, they wouldnot have imagined making a divine institution out of it. If, on the otherhand, they felt the need to relate the totem to a deity, they did so becausethey acknowledged its sacredness. These mythological interpretationsthus display, but do not explain, the religious nature of Totemism.”

Furthermore, according to Durkheim, denying that Totemism is a reli-gion, and arguing that it only became a religion when it came in contactwith more advanced societies that had already developed religion alsocontinues to “beg the question” in that it does not explain where thefirst idea of the sacred came from in the societies they came into con-tact with. Durkheim argues that these theorists are wrong, in any case,because Totemism is more primitive than the religions that it is supposedto have been derived from. But, he also argues that if the sacred aspectsof Totemism were derived from other religions, then the origin of thereligion that it was derived from would still need to be explained. Thetask for Durkheim is to explain the first origin of the sacred, wherever itmay be found, not just to assume it, or to attribute it to “elsewhere.”

According to Durkheim ([1912:266]1915:214–15; *1995:189), deny-ing the religious character of totems is to deny the facts. Totemism exhibitsall the characteristics of the sacred. To argue that it only became a sacred


system later when it came in contact with more “advanced” religions, isto deny this very obvious fact.

4.4.0 Conclusion

Having dispensed with an individualistic approach to totems, and alsowith assumptions regarding their a priori nature, Durkheim returns, inChapter Six, to a discussion of the totem as emblem, and begins to explainthe role of moral forces in the origin of the sacred character of totems.

5 The Origin of Moral Force

Durkheim’s idea of moral force is the key to his epistemological argument.Without it the argument necessarily falls back on either the problem ofgeneral ideas based on sensation, or, if one abandons empiricism andpragmatism, on idealism. Durkheim’s argument is intended to establish athird alternative. That Durkheim has been interpreted as having adoptedboth empiricism and apriorism is a clear indication that his idea of moralforce has not been understood.

The order in which Durkheim presents his argument is at least partly toblame for this misunderstanding. For one thing, the discussion of moralforce only really begins in Chapters Six and Seven of Book II, of TheElementary Forms. But, even then, the argument is not at all satisfactoryuntil the connection between rites and emotions is made definitively inthe discussion of causality, which does not occur until Book III (see myChapter Eight). That later discussion is the only one in The ElementaryForms to clearly and unambiguously locate rites as the cause of a categoryof the understanding. This earlier discussion of moral force still combinesconsiderations of beliefs with rites, which ultimately obscures Durkheim’sargument that beliefs and ideas are caused by rites, although he doesmake the argument repeatedly here, in conjunction with his discussionof beliefs.

It is curious that Durkheim allowed the discussion of moral force toremain ambiguous and that it appears so late in the text. The explanationis probably that Durkheim’s attempt to deal systematically with all priortheories that claim to explain the origin of ideas essential to his argument,leads to his dealing extensively with the beliefs perspectives that he crit-icizes before presenting his own argument. The “systematic argument”that results, however, confuses many issues and leaves his own argumentfor last. Unfortunately, he does not explain this. Once the other positionshave been dispensed with, he seems to feel that his own position, themuch referred to “alternative,” must be obvious. It is, of course, not. Inspite of this curiosity, however, Durkheim is clear at many points in thisdiscussion of moral force that it is rites that cause beliefs and not beliefs

162

The Origin of Moral Force 163

that cause, or explain, rites. Beliefs, he argues, arise only as retrospectiveaccounts of the moral forces created by the collective enactment of rites.

Because Durkheim places his own argument last and embeds it in somuch empirical detail, it is important, as one reads the section on moralforce, to understand that the wealth of detail about the performance ofrites in this section is not extraneous material. Durkheim was attemptingto describe an observable ritual process which he believed caused essentialemotions that are the main ideas that make up human reason. Withoutthe experience of these emotions, he argued, persons lack the capacityfor reason, and thus, for moral reasoning as well.1

In Chapter Six, of Book II, Durkheim introduces the idea that thetotem acts as a force that can be felt by members of the group. This forceDurkheim calls a moral force and often refers to it by the totemic names ofmana or wakan. He hinted earlier, in Chapters One and Three of Book II,at the experience of moral force in his discussion of the feelings membersof the group get from the totemic emblem. But, there he did not presentmoral force as the source of those feelings in any detail. In Chapter Six, hebegins to describe moral forces in empirical detail. Social or moral forcesare, according to Durkheim, as real as physical forces. Mana, moral forceexperienced as emotion, is Durkheim’s answer to the question of why theemblem, as a clan emblem, would be considered sacred.

In Chapter Seven, Durkheim describes Mana as an emotion elicitedby participation in totemic rites, giving participants the feeling of thedivine, psychic energy, and moral authority. Society is experienced as amoral force, but not all the time. He argues that the sacred and profaneoccur in phases. Short periods that are considered sacred are followed bylonger periods which are considered profane. Durkheim has mentionedthe issue of internal versus external constraint in earlier chapters. But, headdresses the issue explicitly in Chapter Seven. He also takes a positionquite opposed to that of most scholars of religion in arguing that becauseof its role in both creating and symbolizing the moral force and the emo-tional excitement of sacred phases, Totemism begins in the experience ofjoy not fear.

The remaining chapters in Book II, Chapters Eight and Nine, dealwith the development of the ideas of soul, personality, and deities. Theidea of soul has been offered by some scholars as the origin of the ideaof the sacred. If this idea is accepted, it would constitute an individual

1 The establishment of empirical grounds for moral reasoning is most likely Durkheim’sultimate goal. That would make this argument a logical outgrowth of The Division ofLabor, originally introduced by Durkheim as a work of moral philosophy. However, moralreasoning is reasoning and the capacity to reason must be established before an argumentfor the empirical validity of moral reasoning can be made.


origin for the idea of the sacred. As with Totems, Durkheim needs theidea of the sacred to have a social origin. Therefore, he needs to establishthat the idea of the soul cannot develop until after the idea of the sacredand offers empirical support for this claim. In Chapter Nine, dualism,and its relationship to myths of the soul and deities, is connected to thefunction of Totemism in creating human reason. Throughout the earliersections, Durkheim has continued to argue that there must be a way ofperceiving social forces that does not involve sensation. In these chaptershe begins to explain what he means by that. But he will still not get to acomplete explanation until the discussion of causality in Book III, wherehe presents moral force in terms of rites instead of beliefs.

This chapter will be divided as follows: Section 5.1.0 deals with theintroduction of the idea of moral force in Chapters Six and Seven. WhileDurkheim continues to argue against the individual origin of totems here,he begins to introduce systematically the idea that the totem is consideredsacred because it generates moral force as the emblem that enables thegroup to create collective unity. Section 5.2.0 deals with the last twosections of Chapter Seven. These consist of summaries of two arguments,the first regarding collective representations and the second regarding theorigins of logical thought. Both of these arguments have been discussed inearlier chapters. Finally, section 5.3.0 will cover the arguments, appearingin Chapters Eight and Nine, with regard to dualism, myths, deities, souls,and personhood.

There is some overlap between the arguments and they don’t alwaysstay neatly within the chapters. Chapter Seven presents a particular prob-lem in this regard because it seems to summarize the argument thus far,and therefore duplicates earlier arguments to some degree. It at first seemscurious that Chapters Eight and Nine should follow a summary of theargument. But, I believe that these chapters act as a transition betweenthe argument of Book II, which focused on the origin of beliefs, and theargument of Book III, which is based on rites. With a few exceptionsthis outline represents the order of the main arguments of Durkheim’schapters.

5.1.0 Mana as the Origin of the Idea of Force

With the objective of establishing social forces as the origin of the idea offorce, Durkheim opens Chapter Six of Book II, with a restatement of hisargument that the idea of force cannot have come from the perceptionof natural forces. Humans have no real control over natural forces, andas Durkheim argues, natural forces rarely have any but predictable andmonotonous qualities. Nevertheless, religions usually contain the idea of


supernatural forces, and mythic explanations of rites often present rites asattempts to control such powers. Since persons have no real control overnatural forces, the belief that they can control these forces cannot, he says,have come from any individual experience of controlling natural forces.2

Therefore, there must be some other origin for this belief. Accordingto Durkheim, there is an energy which emanates from social life andparticularly from the sacred, and the practices surrounding it, which doesgive persons a sense of power and force. Durkheim argues that it must befrom this power, generated by the collective, which he refers to as mana,that persons get the idea that they can exercise power over nature. Notunderstanding where this feeling of power originally comes from, theythen attribute it mistakenly to natural forces.

This mistake is an easy one to make because as a social force, moralforce can be distinctly “felt.” Durkheim says ([1912:270–1]1915:218;*1995:192):

when I speak of these principles as forces, I do not use the word in a metaphoricalsense; they behave like real forces. In a sense, they are even physical forces thatbring about physical effects mechanically. Does an individual come into contactwith them without having taken proper precautions? He receives a shock that hasbeen compared with the effect of an electrical charge.

Mana feels to the participants like an external force and it is capableof moving them both emotionally and physically. Thus, according toDurkheim ([1912:272]1915:219; *1995:192), “the totemic principle isat once a physical force and a moral power.”3

While men have never really “moved mountains,” Durkheim ([1912:299–300]*1915:240; 1995:211–12) argues that “mana,” or moral force,has moved whole communities:

There are occasions when this strengthening and vivifying action of society isespecially apparent. In the midst of an assembly animated by a common passion,we become susceptible of acts and sentiments of which we are incapable whenreduced to our own forces; and when the assembly is dissolved and when, findingourselves alone again, we fall back to our ordinary level, we are then able tomeasure the height to which we have been raised above ourselves.

Moral force creates a special state of being, in which collective actsand sentiments, not available to the individual, are possible. There

2 Believers may object that the deity has the power to make natural forces felt. However,Durkheim’s point is that there is nothing in the way natural forces present themselves tous, whatever their actual causes, that would lead us to the belief that we can control them.Yet, so-called primitive religions have this belief.

3 The 1915 translation says “totemic force” instead of physical force. But the originalFrench is “force materielle.”


is less in coercive, or natural force, to impress us than in moral, orsocial, force according to Durkheim. Even with mana, which has, anddemonstrates, physical force socially, he argues that it is the moralaspect of the force which is most impressive. According to Durkheim([1912:296]*1915:237; 1995:209): “the empire which it holds over con-sciences is due much less to the physical supremacy of which it has theprivilege than to the moral authority with which it is invested.”

Experiencing moral force in common, according to Durkheim([1912:271]1915:218; *1995:192) creates feelings of mutual and recip-rocal obligation: “All the beings that participate in the same totemic prin-ciple consider themselves, by that very fact, to be morally bound to oneanother.” According to Durkheim ([1912:279]1915:224; *1995:197),the common principle that creates such moral obligations is mana or,wakan: “The totem is the means by which the individual is put in touchwith that source of energy. If the totem has powers, it has them becauseit incarnates wakan.”

Moral force is a general principle that can be represented by a variety oftotemic symbols and objects: things. Many dissimilar things, or symbols,are able to evoke the same feelings of moral force in members of the group.This is because the feelings are not generated by any qualities of the thingsthemselves, but by the moral force which they are able to generate assymbols of the group. According to Durkheim ([1912:269]1915:216–17;*1995:190) “the similar feelings that these dissimilar kinds of things evokein the consciousness of the faithful, and that constitute their sacred-ness, can derive only from a principle that is shared by all alike-totemicemblems, people of the clan, and individuals of the totemic species. Thisis the common principle to which the cult is in reality addressed.” Thisprinciple is, of course, moral force.

While mana is in some sense a universal, and has its origins in univer-sal collective needs, it usually takes a particular form. Durkheim insiststhat this particularity should not create the mistaken belief that mana hasits origins in the particular, however, or that the feeling generated by itvaries from place to place. According to Durkheim ([1912:282]1915:227;*1995:200), “since religious forces are localized within definite and dis-tinct social settings, they become differentiated and specialized accordingto the setting in which they happen to be.” All these ideas of mana must, inthe beginning be particular, he argues, because the idea initially is causedby the practice of rites and in its early stages Totemism involved a greatdiversity of rites, none of which encompassed all the others. Accordingto Durkheim ([1912:280–1]1915:225; *1995:199), “the notion of oneuniversal mana could be born only when a religion of the tribe devel-oped above the clan cults and absorbed them more or less completely.It is only with a sense of tribal unity that a sense of the world’s unity


arose.” The experience of mana, however, as moral force, is universaland characterizes all the particular manifestations of mana.

5.1.1 Mana is Not Individual in Origin

As with all important totemic phenomena, Durkheim must argue thatmana is not individual in origin. An individual origin would locateits beginnings either in sensation or in individual demonstrations ofpower, and thereby inherit the problems associated with individual-ism. Durkheim argues ([1912:283–4]1915:228; *1995:201) that “On thegrounds that man is at first ruled mainly by his senses and by sensu-ous representations, it has often been argued that he began by imag-ining the divine in the concrete form of definite and personal beings.”This would have happened either because natural powers were anthro-pomorphized, or because individual beings possessed impressive personalpower; both individualistic explanations. Durkheim, however, says thatneither can explain the origin of the concept of mana because such per-sonal beings, or powers, do not appear within Totemism. According toDurkheim ([1912:284]1915:228; *1995:201) wakan can become “indi-vidualized and fixed upon some definite object or point in space.” But, itis initially a collective phenomenon and only associated with individualswhen they play an important role within the structure of Totemism ingenerating moral force (for instance it often becomes associated with achief or priest).

Furthermore, Durkheim argues, the forces that animate Totemism arenot personal or particular, as one would expect if they had been modeledon the individual, or individual will, but rather indefinite and anonymous.This, according to Durkheim, is because those forces do not come fromparticular beings or locations, but rather from the collective as a whole.

The ability of collective rites to create moral force independently of thebelief in natural or supernatural forces, also belies an individual origin.Here, Durkheim is clear in placing the rite before the belief. Moral force,in such cases, comes to the individual from the group before, and indepen-dently of, anything the individual believes, or has previously experiencedas an individual. Thus, according to Durkheim ([1912:286]1915:229;*1995:202, emphasis added):

It is not surprising that even in religions in which gods indisputably exist, thereare rites that are efficacious by themselves, independent of divine action. Thisis so because that force can attach to words spoken and gestures made, as well as tomaterial substances. Voice and movement can serve as its vehicle, and it can produceits effects through them without help from any god or spirit. Indeed, let that forcebecome primarily concentrated in a rite, and through it that rite will become thecreator of deities.


It is not the belief in a deity that creates the moral force of the rite, butrather the shared enacted performance of the rite that ultimately leads tothe belief in a deity.

In other words, myths about deities are created to explain the felt moralforce of enacted ritual practices. In fact, moral forces, as Durkheim willargue in detail in Book III, are created by rites, that is, by sounds andmovements, not by systems of belief. It is the rites that have createdthe beliefs in the first place. Furthermore, systems of belief, even afterthey have developed a high degree of sophistication, cannot create andsustain moral force in the absence of the periodic performance of rites.4

For Durkheim, it will ultimately be those ideas caused directly by moralforces, experienced during the performance of rites, that have, what hecalls, empirical validity.

5.1.2 Mana as the First Idea of the Category of Force

Mana is not only important because of the key role played by moral forcesin explaining the origin of the idea of the sacred. Mana is also important,according to Durkheim ([1912:290]1915:232; *1995:205), as the firstcommunicable idea of the category of force: “it is the notion of forcein its earliest form.” As the first idea of force that had its source in thesame emotional experience for all members of the group, mana is the firstcommon idea, or significant symbol, of force. A symbol is significant,according to G. H. Mead, when it “calls out” the same idea in the mindsof all participants. The significant symbol is also, for Mead, the mark ofthe transition from animal to human thought. As Durkheim speaks of it,Mana is an idea, a feeling, created when all participants have an identicalexperience of moral force. The totem, as the symbol that “calls out”this feeling in the minds of participants, is the first significant symbol ofmoral force. As such the totem is also the first significant symbol of theidea of force itself. Like Mead, Durkheim argues that it is the collectiveexperience of a symbol, in Durkheim’s case an experience made possibleby the moral force called out by the symbol, that transforms the biologicalindividual into a human being capable of rational thought.

All minds that participate in the rite have the same emotional experience,Durkheim says ([1912:297]1915:238; *1995:209–210): “Because these

4 This is probably why in contemporary culture, when the rites erode there is a felt lack.Generally, persons turn toward beliefs, particularly beliefs about the past, and strive fora strict orthodoxy of belief to restore what has been lost. Of course this sort of religiousorthodoxy, in Durkheim’s view, cannot work and is not necessary. It is a revival of prac-tices, not the enforcement of beliefs that is required. Furthermore, any practices that willsupply the required experiences will do, they do not have to be any particular practicesin order to work.


ways of acting have been worked out in common, the intensity with whichthey are thought in each individual mind finds resonance in all the others,and vice versa. The representations that translate them within each of usthereby gain an intensity that mere private states of consciousness canin no way match.” This idea of force has empirical validity, in spite ofits religious origins, because it is based on a real empirical experienceshared by the participants; created by social reality, in response to socialneeds.

While the idea of force has religious significance, and was said by Comteto be borrowed by science and philosophy from religion, as a religiousidea, it is not based on beliefs, but rather, has its origins, according toDurkheim ([1912:292]1915:234; *1995:206), in real forces and there-fore, has objective meaning:

From religion, philosophy first and later the sciences borrowed it. Such is theintuition Comte already had when he called metaphysics the heir of “theology.”But his conclusion was that, because of its metaphysical origins, the idea of forcewas fated to disappear from science, and he denied it any objective meaning. Iwill show, to the contrary, that religious forces are real, no matter how imperfectthe symbols with whose help they were conceived of. From this it will follow thatthe same is true for the concept of force in general.

Because social forces are real and can be experienced as emotions byparticipants, the idea of force is valid, general, and shared. It is not ideallysuited for use as a scientific concept, according to Durkheim. But, becausereligion is the origin of human reason, ideas originating in religion cannotbe replaced with ideas better suited to science and philosophy, as Comtehad argued. However, because Durkheim defines religion in terms ofpractices, and not in terms of ideas, this is not the problem that Comtesupposed.

5.1.3 Mana as Emotion

The idea that mana can be felt directly by participants as emotion is the keyto Durkheim’s whole argument.5 In Chapter Seven, he begins to arguethat it is the collective experience of emotion that makes force valid as acategory. He argues that people have to give up the idea that all knowledgecomes through the senses. They must learn that the knowledge that comesthrough emotion is the foundation of human reason. Emotion, as collec-tive effervescence, is a very much misunderstood concept in Durkheim’swriting, and it has often been identified as a group mind phenomenon.

5 See my Chapters Six, Seven and Eight and also Rawls 2001 “Durkheim’s Treatment ofPractice” for an elaboration of the importance of emotion to Durkheim’s epistemology.


Nothing could be farther from the argument of the text. Collective effer-vescence is the effect that participants feel when moral forces act uponthem. It does not exist as shared concepts in a group mind. It is directly feltby all as a physical reaction to the enactment of the rite. Moral force pro-duces collective emotions that are common to all participants. According toDurkheim ([1912:300]1915:240; *1995:212), “To strengthen emotionsthat would dissipate if left alone, the one thing needful is to bring allthose who share them into more intimate and more dynamic relation-ship.” It matters little what people think, or believe. What matters is whatthey do.

It is the collective emotions that ultimately constitute the essence ofhuman reason. Returning to his earlier argument regarding emblems,Durkheim says ([1912:293]1915:235; *1995:207 emphasis added) thatthere is nothing in the thing represented by the totem that “resemblesgrand and powerful religious emotions or could stamp upon them a qualityof sacredness.” The idea of the sacred, and the ideas of kind and force, donot come from sense perception, they come through the collective experi-ence of emotion. Durkheim uses the French word “sentiments” to denoteemotions. It is not things that create these emotions, but collective sym-bols, emblems, and rites. Durkheim ([1912:294]1915:236; *1995:207emphasis added) writes that “if the emotion elicited by the thing itselfwere really the determining cause of totemic rites and beliefs, then thisthing would also be the sacred thing par excellence.” He argues that it isthe emotion generated by the emblem that is the origin of the sacred, theemotion is not generated by the thing.

When Durkheim gives the emotion a name he usually calls it respect, orcollective effervescence. This collective emotional experience of respectgives rise to a general consensus about persons and objects that are thefocus of respect. Durkheim ([1912:297–8]1915:238–9; *1995:210)refersto this collective emotion of respect as “opinion.”6 It is important not toconfuse Durkheim’s use of this term with what we generally mean byeither opinion, or public opinion. The word “opinion” is usually used byphilosophers to denote a distinction between belief and knowledge, withopinion occupying a status lower than knowledge. Durkheim does notmean anything like this. For Durkheim the emotions generated by collec-tive action are knowledge. The opinion they give rise to would be collectiveknowledge. At the very least they would be retrospective accounts aboutknowledge. It is because the group has strong feelings about group expe-riences that the group treats certain of its members and certain symbols

6 The French word used is also “l’opinion.”


with respect. This feeling of respect is also what Durkheim means byopinion.

As he did in the argument with regard to emblems, Durkheim com-pares collective effervescence among Australian aboriginals to the pow-erful moral forces felt in modern life. For instance, referring to the moralforce generated by making a political speech before a crowd Durkheimsays ([1912:300]1915:241; *1995:212), “This extraordinary surplus offorces is quite real and comes to [the speaker] from the very group he isaddressing.” According to Durkheim ([1912:300]1915:241; *1995:212)“it is then no longer a mere individual who speaks but a group incarnatedand personified.” Collective effervescence, or moral force, is not some-thing peculiar to Totemism. In order for humans to have reason, moralforces must continue to be generated. Therefore, processes in modern lifemust also generate moral force in similar ways. Totemism has no monopolyon moral force. It only offers the first instance of it.

Durkheim ([1912:302]1915:342; *1995:213) argues that “this stim-ulating action of society is not felt in exceptional circumstances alone.There is virtually no instant of our lives in which a certain rush of energyfails to come to us from outside ourselves. In all kinds of acts that expressthe understanding, esteem, and affection of his neighbor, there is a liftthat the man who does his duty feels, usually without being aware of it.”

The moral being depends on this constant input of moral force. Accord-ing to Durkheim ([1912:302]1915:342; *1995:213):

Thus is produced what amounts to a perpetual uplift of our moral being. Since itvaries according to a multitude of external conditions – whether our relations withthe social groups that surround us are more or less active and what those groupsare – we cannot help but feel that this moral toning up has an external cause,though we do not see where that cause is or what it is. So we readily conceiveof it in the form of a moral power that, while immanent in us, also representssomething in us that is other than ourselves. This is man’s moral consciousnessand his conscience. And it is only with the aid of religious symbols that most haveever managed to conceive of it with any clarity at all.

This moral uplift transforms the biological individual into a humanbeing. In referring to the contemporary speech maker, Durkheim says([1912:300]1915:241; *1995:212), “the passionate energies that hearouses re-echo in turn within him, and they increase his dynamism.It is then no longer a mere individual who speaks but a group incarnatedand personified.”

This need to produce and reproduce collective emotions continues overthe course of history. The emotions do not need to be kept at fever pitchat all times, however. Rather, there are phases of high and low emotionalproduction.


5.1.4 Phases of Moral Force: Time

Durkheim develops the argument that alternating phases of sacred andprofane effect the enactment of moral force, in Chapter Seven, Sections ii,iii, and iv. While both human reason and social solidarity demand thatmoral force be continuously generated, Durkheim argues that there areperiods of greater and lesser moral force. There are phases of social lifecharacterized by a predominance of sacred feelings and periods charac-terized by a predominance of the profane. This is what, he says, createsthe moral force of the category of time: sacred time versus profane time.According to Durkheim ([1912:301]1915:241; *1995:213), periods inwhich “individuals seek one another out and come together more” resultin a heightened “general effervescence that is characteristic of revolution-ary or creative epochs.” It is the act of coming together collectively andenacting social practices in unison that generates moral force.

In some periods people come together more often and create an abun-dance of collective emotion. At other times people gather together lessand feelings of collective emotion become weak. According to Durkheim([1912:301]1915:241; *1995:213), “the result of that heightened activ-ity is a general stimulation of individual energies. People live differentlyand more intensely than in normal times. The changes are not simply ofnuance and degree; man himself becomes something other than what hewas.” The person who focuses on individual needs during profane phases,focuses on collective needs during sacred times; volunteering to fight inwars, or make great personal sacrifices for the group. The individual lives“above” themselves.

According to Durkheim ([1912:212–13]1915:250; *1995:220–1),time is divided between sacred and profane among the Australians toa degree that is very different from modern religions. He says that thismakes the religious periods more intense. In so doing it creates a tangiblemoral distinction between sacred and profane time.

Because, among aboriginals profane time is spent apart from thegroup, Durkheim says ([1912:308]1915:246; *1995:217), nothing moreis required than for persons to congregate in order for them to feel thatthey are in a sacred time: “The very act of congregating is an exceptionallypowerful stimulant. Once the individuals are gathered together, a sort ofelectricity is generated from their closeness and quickly launches them toan extraordinary height of exaltation.” In such collective states, accordingto Durkheim, participants are all feeling the same emotions.7

7 Of course, in order for merely congregating to have the effect of generating moral force,and not just a general excitement, members of the group must be able to anticipate theidea of sacred time; that is, they must already have experienced it. Elsewhere, Durkheim


Durkheim ([1912:308]1915:247; *1995:217–218) argues that gather-ing together to engage in collective activities produces a special state ofbeing: “every emotion expressed resonates without interference in con-sciousnesses that are wide open to external impressions, each one echoingthe others. The initial impulse is thereby amplified each time it is echoed,like an avalanche that grows as it goes along.” The individual state of beingthat takes over during profane phases is completely engulfed by collectiveemotional experience during sacred times.

According to Durkheim ([1912:309]1915:247; *1995:218) collectiveemotions are most often produced by highly routinized rites, not becauseroutinization is necessary to create collective emotion in the first place,but rather, because in order to express what they have already felt togethercollectively, people need to create a collective emblem and perform thesame actions that gave them the feelings of moral force before. He says([1912:306]*1915:246; 1995:217–218): “And since a collective senti-ment cannot express itself collectively except on the condition of observ-ing a certain order permitting cooperation and movements in unison,these gestures and cries naturally tend to become rhythmic and regular;hence come songs and dances.”8 If people could spontaneously performidentical actions routinization would not be necessary. But, this is gener-ally not possible.9

Moral force can be generated by any sounds and movements that areperformed by a group in unison. But, sounds and movements are morelikely to occur in unison and continue to occur in unison, if they areroutinized or take an expected form. It is also the case that the verystrong feelings called up during sacred phases are enabled to continue onduring profane phases, to some degree, because of the availability to theindividual of the totemic image.

argues that much more than just congregating is required in order to render time, or thetotem, sacred in the first place.

8 I have used the 1915 translation in the text at this point because it preserves the emphasisin the original French that Durkheim placed on the performance of rites. The 1995translation reads: “Probably because a collective emotion cannot be expressed collectivelywithout some order that permits harmony and unison of movement, these gestures andcries tend to fall into rhythm and regularity, and from there into songs and dances. Butin taking on a more regular form they loose none of their natural fury.” A more literaltranslation from the French reads ([1912:308]): “Without doubt, because a sentimentcollective cannot express itself collectively without the condition being observed of acertain order that permits the unity (concert) and the movements in assembly, thesegestures and these cries tend likewise to become rhythmic and to become regularized;hence, the songs and the dances.” The 1995 translation, at this point, does not preserveDurkheim’s strong emphasis on the enactment of rites, and on the details of particularforms of rite, appearing to focus on beliefs instead.

9 See Chapters Six and Seven for a discussion of the importance of identical sounds andmovements. See also Rawls 2001 “Durkheim’s Treatment of Practice.”


In this regard Durkheim reiterates his earlier point that the totemicimage is considered sacred because of its ability to call up sacred feelings.But now he ([1912:316]1915:252; *1995:222–3), extends the point toinclude the idea of calling up sacred feelings during profane times:

The image goes on calling forth and recalling those emotions even after the assem-bly is over. Engraved on the cult implements, on the sides of rocks, on shields,and so forth, it lives beyond the gathering. By means of it, the emotions felt arekept perpetually alive and fresh. It is as though the image provoked them directly.Imputing the emotions to the image is all the more natural because, being com-mon to the group, they can only be related to a thing that is equally common toall. Only the totemic emblem meets this condition. By definition, it is common toall. During the ceremony, all eyes are upon it. Although the generations change,the image remains the same. It is the abiding element of social life. So the mys-terious forces with which men feel in touch seem to emanate from it, and thuswe understand how men were led to conceive them in the form of the animate orinanimate being that gives the clan its name.

Without the totemic emblem the social being could not survive in pro-fane times and would give way entirely to the pre-social individual. It isthe emotions that are evoked, in profane times, by the sight or memoryof the totemic image, that preserve the memory of moral force and sustainthe social side of the human through these, often long, periods of time.Persons are only able to preserve their social character through profanetimes by constantly invoking in memory and anticipation, the totemicimage and its quality of moral force. This power that seems to emanatefrom the emblem explains why the members of the group believe that itis sacred.

5.1.5 Internal Sense of Duty versus External Constraint

In Chapter Seven, Durkheim also develops the idea of internal versusexternal constraint. This involves revisiting the idea of dualism. Thereis, according to Durkheim, a difference between the internal sense ofduty that is produced in participants by the experience of moral forcein the assembled group, and the external constraint imposed on per-sons by society in order to induce persons to participate. One of thesecorresponds with immediate feelings that are internally felt; the other,with socially sanctioned beliefs and ideas. The first is necessary, the sec-ond arbitrary (in form). Durkheim ([1912:302–3]1915:342; *1995:214)contrasts the moral forces that come from immediate assembled action,with those merely conventional forms of life that have, as he says, con-gealed in language and custom:


In addition to those free forces that continuously renew our own, there are otherforces congealed in the techniques we use and in traditions of all kinds. We speaka language we did not create; we use instruments we did not invent; we claimrights we did not establish; each generation inherits a treasury of knowledge thatit did not itself amass.

Moral forces have a shared quality that gives them empirical validity.Conventions do not have empirical validity. They must be imposed. Yet,both operate in some sense internally. Conventional ideas are internal-ized, but remain empirically invalid. They do not correspond directlyto felt ideas, or emotions. The ideas resulting from the emotions pro-duced by the performance of rites, on the other hand, are internal-ized as emotions to begin with and do not need to be imposed becausethey are immediately shared. The feelings that these two sorts of socialforces “provoke” in us are, according to Durkheim ([1912:303]1915:243;*1995:214) “qualitatively different from those we have of merely physicalthings.” Physical things are perceived through the senses. Social forces,whether internal and immediate, or external and congealed, are experi-enced differently.

Thus, according to Durkheim ([1912:303]1915:243; *1995:214),“The two sorts of representations form two distinct sorts of reality witha clear line of demarcation between them: the world of profane things onthe one side, the world of sacred things on the other.” Because of thisdistinction, according to Durkheim, we really are in touch with two dis-tinct sorts of reality. While sensations move us from without, the worldof sacred things moves us from within because of the respect that we feelfor the moral forces that it generates.

The emotion of respect that persons feel for moral forces is, accordingto Durkheim ([1912:296]1915:237; *1995:209), a large part of why theyfeel a duty to society:

If society could exact those concessions and sacrifices only by physical constraint,it could arouse in us only the sense of a physical force to which we have nochoice but to yield, and not that of a moral power such as religions venerate.In reality, however, the hold society has over consciousness owes far less to theprerogative its physical superiority gives it than to the moral authority with whichit is invested. We defer to society’s orders not simply because it is equipped toovercome our resistance but, first and foremost, because it is the object of genuinerespect.

Physical forces cannot give rise to the idea of the sacred, because per-sons do not have the emotion of respect for them. This would be trueof any form of external constraint that was not accompanied by moralforce.


Moral forces, by contrast, inspire a special sort of respect thatDurkheim ([1912:296]1915:237; *1995:209) compares with a Kantiankind of duty: “An individual or collective subject is said to inspire respectwhen the representation that expresses it in consciousness has suchpower that it calls forth or inhibits conduct automatically, irrespectiveof any utilitarian calculation of helpful or harmful results.” Durkheimargues ([1912:342]1915:271; *1995:240) that “there is an even morepronounced opposition between sacred and profane things,” than is con-tained in the principle of contradiction. The sacred and profane, he says([1912:342]1915:271; *1995:240), “expel one another from conscious-ness.” Kant’s test for rational thought was that two contradictory ideascannot be thought at the same time without contradiction. Durkheimis arguing that the distinction between the sacred and the profane morethan fulfills this criteria. He ([1912:298]1915:238; *1995:210) says that“The hallmark of moral authority is that its psychic properties alone giveit power.” Here Durkheim favorably compares the rational contradictionsimplied in moral forces to the force of the logical test involved in Kant’sCategorical Imperative.10

Durkheim also takes up the issue of external constraint in some detail ina footnote. He complains ([1912:298ff]1915:239ff; *1995:210ff, empha-sis in original) that he has been interpreted as arguing that externalconstraint is how society works, an interpretation that is still popular.Durkheim rejects this interpretation:

I hope this analysis and those that follow will put an end to an erroneous interpre-tation of my ideas, which has more than once led to misunderstanding. BecauseI have made constraint the external feature by which social facts can be most eas-ily recognized and distinguished from individual psychological ones, some havebelieved that I consider physical constraint to be the entire essence of social life.In reality, I have never regarded constraint as anything more than the visible, tan-gible expression of an underlying, inner fact that is wholly ideal: moral authority.The question for sociology-if there can be said to be one sociological question-isto seek, throughout the various forms of external constraint, the correspondinglyvarious kinds of moral authority and to discover what causes have given rise to thelatter. Specifically, the main object of the question treated in the present work isto discover in what form the particular kind of moral authority that is inherent inall that is religious was born, and what it is made of. Further, it will be seen belowthat in making social pressure one of the distinguishing features of sociologicalphenomena, I do not mean to say that this is the only one. I will exhibit anotheraspect of collective life, virtually the opposite of this one, but no less real.11

10 See Kant’s Prolegomena to Any Future Metaphysics.11 The fact that Durkheim calls moral authority ideal in this passage will probably confuse

the issue. The thing to keep in mind is that he is trying to distinguish it from constraintthat is external and physical, that is, experienced through the senses. What he needs toargue about moral force is that it is felt internally as emotion. He refers to his distinctionbetween internal and external knowledge.


In the Division of Labor in Society, for instance, Durkheim made exter-nal constraint, in the form of punishment, a measure of the type of asociety. That is, whether it was an organic or mechanical form of soli-darity. What he is arguing here, is that external constraint can be used assuch a measure, as a matter of convenience, because it is easily observed.However, external constraint is not the reason why persons do what theydo. External constraint has, for Durkheim, never been other than the“visible” expression of an underlying moral force.

External constraint is like the emblem that calls out the emotion ofmoral force. According to Durkheim ([1912:299]1915:240; *1995:211),“Society’s workings do not stop at demanding sacrifices, privations, andefforts from us. The force of the collectivity is not wholly external; it doesnot move us entirely from outside. Indeed, because society can exist onlyin, and by means of, individual minds, it must enter into us and becomeorganized within us. That force thus becomes an integral part of our beingand, by the same stroke, uplifts it and brings it to maturity.” Society canonly endure when persons come to share ideas and to need a commonlife. This is accomplished when social practices succeed in creating moralforces in its members, joining them from within, and giving them at leastas great a stake in the continued life of the group, as they have in their ownindividual existence. Society cannot maintain itself by means of externalconstraint alone.

5.2.0 Logic and Collective Representations

The last two sections of Chapter Seven, Sections v and vi, take up twoissues that require separate treatment. They summarize aspects of theepistemological argument laid out in the Introduction and that come upagain in the Conclusion (and also in earlier parts of Book II). In Sectionv Durkheim discusses collective representations in relation to philosophy.In Section vi he discusses the development of logical thought, and doesso again in a philosophical context.

5.2.1 Collective Representations

There has been a great deal of confusion about what Durkheim meant bycollective representations. He has consistently been interpreted as thoughby collective representations he meant ideas that had no basis in reality:something like pure conventions of thought. It is this interpretation thatled to the identification of Durkheim as an idealist. He does argue thatcollective representations are social constructions. He also refers, at theend of Section v, to collective representations as making up a natural


system of idealism. But, for Durkheim, that does not mean that collectiverepresentations have no basis in reality. He seems to mean something likea system of ideas that is caused by social reality and that in its evolutionobeys the laws of social reality.

In this section, Durkheim discusses collective representations and con-trasts them to individual representations. He makes it clear through thisdiscussion that all collective representations have their origin in symbolsused by the group to represent moral force. Some collective representa-tions call up the actual experience of moral force, and therefore, have thestatus of valid categories of the understanding, like force, classificationand cause. Others are derived from the feelings generated by moral forcein a more general way and therefore, while not possessing the empiricalvalidity of categories, nevertheless, have an empirical basis. All collectiverepresentations are ideas that have their counterpart in collective emo-tions, either directly or at some degree of remove.

Durkheim opens the section by asking how the clan got theidea of creating emblems to symbolize their collective feelings. He([1912:329]1915:262; *1995:231) says that “by expressing the socialunit tangibly” the emblem “makes the unit itself more tangible to all.”Because emblems are able to express the unity of the social group tan-gibly, he argues ([1912:329]1915:262; *1995:231), the use of emblems“must have arisen spontaneously from the conditions of life in common.”The initial discovery would have been something of an accident. Onceemblems had come to serve this function, however, Durkheim supposes([1912:329]1915:262; *1995:231) that “the use of emblematic symbolsmust have spread quickly.”

According to Durkheim ([1912:329]1915:262; *1995:231) the embl-em “is a constitutive element” of any awareness that society has of itself.“By themselves,” he says ([1912:329]1915:262; *1995:231) “individualconsciousnesses are actually closed to one another, and they can commu-nicate only by means of signs in which their inner states come to expressthemselves.” Individual ideas and feelings, based on sense perception andinner states, are not communicable to other individuals in an unambigu-ous or empirically valid way.

Durkheim is anticipating Wittgenstein here. Purely individual states ofconsciousness cannot be expressed. But, unlike Wittgenstein, Durkheimargues that there are collective states of consciousness, not just conven-tions for the use of words. It is only collective experiences that can be des-ignated by collective signs. These “signs,” that allow for communicationbetween individuals, Durkheim refers to as collective representations.

Collective representations are necessary because of the alterationbetween sacred and profane time. It is not enough for society toperiodically produce collective feelings. Those feelings must carry the


group through profane phases, and must also become a basis for com-munication, if society is to survive. It is not possible for members ofthe group to generate collective emotions all of the time. They mustalso go about the profane business of providing food and shelter. Coop-eration in this regard is much aided by the invention of symbols. But,before persons can come to productively make use of symbols, theremust first be a symbol which calls out the same emotions in everyone.Not all subsequent symbols have to work this way, conventions of usemay suffice for many, but the first, and essential ones, do. Accordingto Durkheim ([1912:329]1915:262; *1995:231–2) “for the communica-tion that is opening up between them to end in a communion-that is,in a fusion of all the individual feelings into a common one-the signsthat express those feelings must come together in one single resultant.”This resultant, or symbol, reinforces feelings of moral unity. Durkheim([1912:229–30]1915:262; *1995:232) argues that “It is by shouting thesame cry, saying the same words, and performing the same action inregard to the same object that they arrive at and experience agree-ment.” This emotional experience is what is embodied in the collectiverepresentation.

There are also individual representations. Durkheim ([1912:330]1915:262; *1995:232) distinguishes collective representations from indi-vidual representations that also “bring about repercussions in the bodythat are not unimportant.” The individual “feels” something, and theirbody reacts physically and emotionally to individual perceptions. “Still”he says ([1912:330]1915:262; *1995:232) “these effects can be treatedas analytically distinct from physical repercussions that come with,or after them, but that are not their basis.” These individual represen-tations are accompanied by “physical repercussions,” but these physicaleffects are not shared with others, and are ultimately not the basis of theindividual representations. The individual representation is a construc-tion of the individual mind. As Hume argued, it has no direct counterpartin perception.

In the case of collective representations, on the other hand, accord-ing to Durkheim, the situation is quite different. He argues that withcollective representations the emotional effects are the basis of theideas and, in some cases, cause them directly. According to Durkheim([1912:330]1915: 262–3; *1995:232, emphasis added):

Collective representations are quite another matter. They presuppose that con-sciousnesses are acting and reacting on each other; they result from actions andreactions that are possible only with the help of tangible intermediaries. Thusthe function of the intermediaries is not merely to reveal the mental states asso-ciated with them; they also contribute to its making. The individual minds canmeet and commune only if they come outside themselves, but they do this only


by means of movement. It is the homogeneity of these movements that makes thegroup aware of itself and that, in consequence makes it be. Once this homo-geneity has been established and these movements have taken a definite formand been stereotyped, they serve to symbolize the corresponding representations. Butthese movements symbolize these representations only because they have helpedto form them.

Collective representations help individual minds to come outside of them-selves. This is made possible when everyone in the group makes the samemovements at the same time. This creates group awareness, and collec-tive emotions, that are not possible for an individual. These emotionscan be called out by the symbols. The symbols correspond directly tothe emotions caused when the group moves in unison. They do not cor-respond to sensations. Therefore, Durkheim argues, they do not needto be constructed by the individual mind. Enacting practices in commoncreates a shared emotional state that is not possible for an individual; thatonly exists as shared, and which, therefore, has a collective existence, anemotional state that can only be called out by a symbol of the practicesthe enactment of which caused that emotion.

While the emotions are powerful, however, if they are not repre-sented by symbols, he argues, they cannot be communicated, but onlyreproduced by enacting the practices again. According to Durkheim([1912:330]1915:263; *1995:232) “Without symbols, moreover, socialfeelings could have only an unstable existence.” The feelings remainstrong only as long as persons are gathered. But, according to Durkheim([1912:330–31]1915:263; *1995:232) “if the movements by which thesefeelings have been expressed eventually become inscribed on things thatare durable, then they too become durable.” Symbolizing the feelingsmakes it possible to call them up at any time.

Collective representations then are not labels that point toward sociallyconstructed meanings. Rather, they are an integral part of what they rep-resent. For this reason, according to Durkheim ([1912:331]1915:263;*1995:233) “we must guard against seeing those symbols as mereartifices-a variety of labels placed on ready-made representations tomake them easier to handle. They are integral to those representations.”Although they are symbols, they do not represent ideas, or things. Nor is thesymbol merely a name. According to Durkheim ([1912:333]1915:265;*1995:234), “the purpose of the image is not to represent or evoke a par-ticular object but to testify that a certain number of individuals share thesame moral life.”

The connection between labels and things is, according to Durkheim([1912:331]1915:263; *1995:233) “not purely conventional.” Neither isit a relationship of representation. Collective representations “call out”


ideas that are produced through the collective performance of sounds andmovements. According to Durkheim ([1912:331]1915:263; *1995:233)they portray “a real feature of social phenomena: their transcendence ofindividual consciousness.” They are in an important sense the means ofthis same transcendence. According to Durkheim ([1912:331]1915:263;*1995:233) “The objectivity of the symbol is but an expression of” theexternality of the moral force generated by the group.

Society, which requires moral force to create its unity and make com-munication possible, requires symbols because without those symbolsmoral force cannot be represented, and the group cannot be made awareof itself except at times of great collective effervescence. According toDurkheim ([1912:331]1915:263; *1995:232), “while emblematizing isnecessary if society is to become conscious of itself, so is it no less indis-pensable in perpetuating that consciousness.” Symbols are not only anecessary part of the process through which rational categories of thoughtare generated in the first place. They are also necessary for the contin-ued presence and evolution of rational, logical, thought. According toDurkheim ([1912:331]1915:264; *1995:233) “social life is only possiblethanks to a vast symbolism.”

Durkheim considers idealism at the end of Section v. Since one of themost serious misinterpretations of his work is that it is idealist, this dis-cussion is important. As with other places in his work where Durkheimappears to agree with a philosophical position, he turns out here to agreeonly with a version of idealism that is so transformed by his own think-ing as to make the comparison entirely misleading. Durkheim ([1912:320–28]1915:255–62; *1995:225–231) says that because the totemicsymbol is so important, the principle of idealism is realized in a naturalform. He argues ([1912:326]1915:260; *1995:229) that “hence, thereis a realm of nature in which the formula of idealism is almost literallyapplicable; that is the social realm.”

But, Durkheim’s ([1912:326]1915:260; *1995:229) espousal of ideal-ism is accompanied by the claim that the ideal is also real: “However, eventhough purely ideal, the powers thereby conferred on that object behaveas if they were real.” Unlike the idealists, Durkheim believes that the ideasthat he calls collective representations, act as real forces, that they are theproducts of forces that act on bodies, and therefore, that they have coun-terparts in the physical world, in so far as the social world is physical.After collective representations have been caused by moral forces all theyneed is to be attached to material things in order to reproduce the moralforces that caused them.

For Durkheim, idealism and realism are not mutually exclusive terms.The ideal and the real are bound together in an eternal dualism. Ideas that


are experienced as social forces are very concrete and real. Similarly, thesocial person cannot escape the dualism of their nature, which couplesthe biological individual, the animal of sense perception, with the humanbeing whose reason is based on concepts. The task, he believes, is tounderstand this, and break the false dichotomy of the ideal and the real.

5.2.2 Logic

In Section vi, of Chapter Seven, Durkheim argues that modern scien-tific logic, and the logic exhibited by primitive societies, are essentiallythe same, and have the same origin: society.12 He also challenges theindividualism of empiricism and pragmatism, arguing that the only wayprimitive man could have developed the ability to reason is by learningnot to rely on perception. In fact, he will argue that the genius of primitivepeople was in learning not to be limited to what can be known throughthe senses. Durkheim says ([1912:340]1915:270; *1995:239) that “whatwas essential was not to let the mind be dominated by what appears to thesenses, but instead to teach the mind to dominate it and to join togetherwhat the senses put asunder.”

Having introduced the problem of logic in the Introduction, andhaving promised there to show that all of logic has a social origin,Durkheim ([1912:336]1915:267; *1995:236) now says that his theoryof Totemism “will be yet another opportunity to observe that logical evo-lution is closely interconnected with religious evolution and, like religiousevolution, depends upon social conditions.” According to Durkheim([1912:339]1915:269; *1995:238) it is social requirements that give birthto the first logical forms “fusing together ideas that at first glance seemdistinct.” It is also social requirements that continue to create logicalforms over the course of history. While primitive logical forms are some-what different from those of modern society, and for that reason some-times appear crude to us, they are in fact, according to Durkheim, the“momentous” beginning of logical human thought, and modern scienceand philosophy owe everything to them.

In arguing for a continuity between primitive and modern logicDurkheim again criticizes those who claim that primitive thinking iscompletely illogical. According to Durkheim ([1912:341]1915:270–1;*1995:240) “It has been said that the participations whose existencemythologies presuppose violate the principle of contradiction and, on

12 Durkheim takes up the question of logic again in Section iii of his Conclusion. Thereit becomes clear that the arguments with regard to logic and the epistemological argu-ment are distinct, though related arguments. Confusing them obscures the epistemologyentirely.


those grounds, are antithetical to the participations that scientific expla-nations involve.”13 But, the primitive never confuses the sacred withthe profane. As we saw earlier, the distinction between sacred andprofane obeys the principle of non-contradiction. Durkheim maintains([1912:341]1915:271; *1995:240) that contemporary western logic isnot as different from primitive logic as is supposed: “Is not postulatingthat a man is a kangaroo and the sun a bird identifying one thing withanother? We do not think any differently when we say of heat that it isa movement, and of light that it is a vibration of the ether, and so on.”Durkheim’s point is a very modern one; that it is not only the primitivewho identifies one thing with another that is not identical to it as a senseperception. A piece of wood feels hard, but looks brown, yet we say itis one thing. All languages, and in fact, all systems of thought, do this.What the primitive identifies as the same does not look the same to us,and vice versa. But, the principle involved, of classifying things accordingto kinds, is essentially the same.

Contemporary logic, according to Durkheim ([1912:336]1915:26;*1995:236) will not allow us to classify “beings that differ not only inoutward appearance but also in their most fundamental properties –such as minerals, plants, animals, and men” in the same logical cate-gories. We demand that the differences between them assign them todifferent categories. “But these distinctions, which seem to us so natural,are” according to Durkheim ([1912:337]1915:267; *1995:237) “not atall primitive.” The merging of what we would consider to be separatecategories is found in all primitive mythologies. According to Durkheim([1912:337]1915:267; *1995:237), “The rocks have a sex; they have theability to procreate; the sun, moon, and stars are men and women, whofeel and express human feelings, while humans are pictured as animalsor plants.”

Durkheim rejects the claim by animists that this is a “confusion” ofthe facts, caused by a tendency to anthropomorphism on the part ofprimitive thinkers. He argues that primitive people classify animals andplants together because that is how they correspond to the moral forces, orsocial divisions, which they experience. This is not anthropomorphism. Itis not that people have made the world in their own image. According toDurkheim ([1912:337]1915:268; *1995:237) “He has no more imaginedthe world in his own image than he has imagined himself in the image ofthe world. He has done both at once.” The person and their world arejoined in a moral unity according to Durkheim ([1912:337]1915:268;*1995:237): “In the way he thought about things, he of course included

13 Here he cites Levy-Bruhl, 1910 Les Fonctions Mentales, page 77ff.


human elements, but in the way he thought about himself, he includedelements that came to him from things.”

This way of dividing the world came from the moral distinctions char-acteristic of the social world; the classification of the world according toclan totems. They could not have come from sense experience. Accord-ing to Durkheim ([1912:337]1915:268; *1995:237), “there was nothingin [sense] experience that could have suggested these mergers and mix-tures.” Sense perception only offers particular and discontinuous images.Nothing in sensation is ever exactly the same as anything else, or eventhe same as itself twice. Therefore, there is nothing in sensation thatwould give the idea of a classification of things. The animal ability tosort by resemblance might lead to invalid classifications, in a Humianway, but the primitive clearly does not sort by this type of resemblanceeither.

On this point Durkheim challenges pragmatism and empiricism. Heargues that what is required in order for human logical thinking to developis that the individual ceases to rely on their senses for information. On thebasis of sensation alone, he argues, the primitive would never have arrivedat the idea of classification, or any of the other categories for that matter.According to Durkheim([1912:337]1915:268; *1995:237–8), “From thestandpoint of observation through the senses, everything is disparate anddiscontinuous. Nowhere in reality do we observe beings that merge theirnatures and change into one another. An exceptionally powerful causewould have had to intervene and so transfigure the real as to make itappear in a form not its own.” Society is that “powerful cause,” necessaryto create the first concepts that are shared among members of the group,and allow for a logical sorting of the world.

Totemism makes this possible because the moral energy symbolized bythe totem creates a conceptual distinction between sacred and profane,between things that belong to the totem and things that do not. Thisway of classifying things is not based on sensation. It creates conceptualrelations that correspond to the emotions felt by all the participants duringreligious rituals. And it is very like Kant’s principle of non-contradiction.

Durkheim argues that sense perception is inherently unable to makeany explanation of the world possible. With regard to Totemism he says([1912:339]1915:269; *1995:239) that, “Of course, the mental habitsit implies prevented man from seeing reality as his senses show it tohim; but as the senses show it to him, reality has the grave disadvan-tage of being resistant to all explanation.” This is true because senseperceptions are particular and disconnected, and according to Durkheim([1912:339]1915:269; *1995:239), “to explain is to connect things toother things.”


Explanations, according to Durkheim ([1912:339]1915:269; *1995:239), make things “appear to us as functions of one another and asvibrating sympathetically in accordance with an internal law that isrooted in their nature.” “Sense perception,” on the other hand, he says([1912:339]1915:269; *1995:239) “which sees from the outside, couldnot possibly cause us to discover such relationships and internal ties.”Logical thought requires making connections between things that arenot apparent to the senses. Therefore, the development of logical thoughteither requires direct access to the connections between things or makesthem up.14

Durkheim ([1912:339]1915:269; *1995:239) agrees with Hume thatwith regard to general ideas, “only the intellect can create the notionof them.” Relations between things cannot be perceived through thesenses. He argues ([1912:340]1915:270; *1995:239) that “when I learnthat A regularly precedes B, my knowledge is enriched with a new pieceof knowledge, but my intelligence is in no way satisfied by an observa-tion that does not carry a reason with it.” Perception presents a regularlychanging series of views of the world, but it does not organize them, orprovide any basis for an explanation of their organization. Durkheim says([1912:340]1915:270; *1995:239 emphasis in original) that “I begin tounderstand only if it is possible for me to conceive of B in some way thatmakes it appear to me as not foreign to A but as united with A in somerelation of kinship.”

The difference between Hume and Durkheim on this point is thatHume argues that because valid general ideas cannot come from percep-tion, they must be arbitrary inventions of the mind. Durkheim will argue,that on the contrary, society would not be possible if they were arbitrary,and therefore, it is a social requirement that there be such mental con-structions and that they be empirically valid. It follows that society canonly exist where social practices have developed in such a way as to createthe emotional experiences necessary to generate the “mental” functionsthat constitute logical thought.

According to Durkheim ([1912:338]1915:268; *1995:238) “It is reli-gion that carried out this transformation; it is religious beliefs thatreplaced the world as the senses perceive it with a different one.” Itis essential to understand that Durkheim is not talking about religiousbeliefs here, but rather, about the emotional effects of enacting practicesin common. The collective enactment of practice transforms the world ofexperience from a world of particular and discontinuous perceptions, to

14 Here Durkheim refers to the problem of general ideas discussed in my Chapters Oneand Two.


a world organized along complex logical principles, in which everythinghas a place. Totemism, according to Durkheim ([1912:338]1915:268;*1995:238), easily explains this development: “Once that belief wasaccepted, the disparate realms were bridged. Man was conceived of as akind of animal or plant, and the plants and animals as man’s kin.”

These religious beliefs in turn, are not arbitrary, but are themselves,according to Durkheim ([1912:338]1915:268; *1995:238) the result of“definite social causes.” Because the existence of the clan depends onbeing able to generate the idea of clan unity, the existence of the clanrequires an emblem that symbolizes the unity of the group. It also requiresbeliefs that will tend toward reproducing the unity of the clan. Thus, thebelief in the clan and the clan totem ultimately depend on, and are causedby, the enacted practices.

The collective, or moral, force, which all participants feel, andwhich is the origin of common sentiments and ideas, is representedin the form of the group emblem, or totem. According to Durkheim([1912:338]1915:269; *1995:238), “Men had no choice but to conceivethe collective force, whose workings they felt, in the form of the thingthat served as the flag of the group.” He argues ([1912:339]1915:269;*1995:238) that “there is no society in which [this cause] is not at work.Nowhere can a collective feeling become conscious of itself without fix-ing upon a tangible object.” The feelings are immediate and shared. But,they must be symbolized before the group, as a group, can focus on thosefeelings as a collective object.

In both primitive and modern society prerequisites for shared mean-ing and order place demands on the forms that both belief and practicewill take. Durkheim ([1912:339]1915:269; *1995:238) argues that “It issocial requirements that have fused together ideas that at first glance seemdistinct.” These social requirements are not met magically just becausethere is a functional need for them, however. They are met through themutual experience of moral forces that create significant symbols amongmembers of the group. Durkheim says ([1912:339]1915:269; *1995:238)that “through the great mental effervescence that it brings about, sociallife has promoted that fusion.”

According to Durkheim ([1912:339]1915:269; *1995:239), while thelogic created by Totemism may seem crude, “we must be careful notto depreciate it . . . it was a momentous contribution to the intel-lectual development of humanity.” This logic made a first explanationof the world possible. The social forces which created it also madepossible the first significant symbols without which communication“about” the world would be impossible. He quite literally means here,that unless Totemism had given humans the first logical abilities those


logical abilities would not exist. There is no innate rational faculty in thisargument.

Durkheim would prefer to applaud primitive peoples for figuring outhow to create the possibility of logical human thought, rather than judgingthe adequacy of the logical systems that were actually developed, by mod-ern western standards. He says ([1912:339]1915:269; *1995:239) that“It was less important to succeed than to dare.” What primitive peoplefirst had to “dare” was to organize their thinking around their emotionalexperience of the moral order of the social world, instead of relying onthe evidence of their senses.

From this point the evolution of logic toward modern science becamepossible. According to Durkheim ([1912:339]1915:270; *1995:239),“As soon as man became aware that internal connections exist betweenthings, science and philosophy became possible.” But, this possibilitywas not inherent in the human mind, and it did not come from sensa-tion. According to Durkheim ([1912:340]1915:270; *1995:239), “reli-gion made a way for them [science and philosophy].” Ironically, for sci-ence to be able to focus on empirical observations, it was first necessaryfor the human mind to stop relying on observation and rely instead oncollective emotion, in order to create the capacity for reason.

Durkheim says ([1912:340]1915:270; *1995:239) that “to make mentake control of sense impressions and replace them with a new way ofimagining the real, a new kind of thought had to be created: collectivethought.” No individual person could develop logical thought. Significantsymbols and shared concepts must belong to more than one individualsimultaneously. For this to be possible, they must not only have the sameexperience at the same time, but, they must have an experience which onlyhas an emotional impact because it is shared, and therefore, transcendssense perception. Durkheim argues ([1912:340]1915:270; *1995:239)that “it is because religion is a social thing that it could play this role.”

If it seems too much to claim that collective thought could easily solvethe problem, whereas individual thought never could, Durkheim had anexplanation. He argued ([1912:340]1915:270; *1995:239) that “If col-lective thought alone had the power to achieve this, here is the reason:“Creating a whole world of ideals, through which the world of sensed real-ities seemed transfigured, would require a hyperexcitation of intellectualforces that is possible only in and through society.”

Durkheim’s point about “hyperexcitation” here is very important.“Creating a world of ideals” has been interpreted by critics as a sociologyof knowledge argument in which concepts and ideas are determined bysociety and then somehow “impressed” on the individual mind. But, hereDurkheim says that collective ideas are only possible because of the role


played by the collective experience of emotion in creating shared ideas inthe minds of participants. The world of ideas is not an arbitrary creation.The possibility of creating collective understanding depends on the jointenactment of the significant symbols, or categories, that result from thecollective experience of emotion. Those emotions will always take a sim-ilar form and it is their form that leads to the collective ideas. Therefore,they always have characteristics in common.

Durkheim has shown how a collective approach explains the origin ofboth the categories of classification and force. If one posits an individualorigin for the development of totems, it is necessary to suppose that theindividual had already developed the concept of classification in order todevise the totems. On this view, logic would have to be either completelyarbitrary, or innate. Durkheim argues that his solution is the only way outof this dilemma. Totems arise “naturally” to meet social needs. Then,the emotional experience of their moral force by members of the groupcreates the category of classification through the division of the world intosacred and profane.

5.3.0 Personhood and Myths versus Rites

In the remaining two chapters of Book II, Chapters Eight and Nine,Durkheim deals with the issues of dualism, deities and personality. Ear-lier, in the Introduction, Durkheim listed personality as a category. Itis in this section that he comes closest to delivering on that promise.Dualism, deities and personality, are all important issues with regard toDurkheim’s theory of Totemism. Usually, scholars of religion assume thatan idea of the soul is the cause, not the consequence of religious thinking.But, Durkheim argues, beginning with the soul once again introduces thespecter of individualism. In Chapter Eight, Durkheim asks what the ori-gin of the belief in the dualism of the soul could be. His answer will bethat the soul is the totemic principle incarnate in the individual. Here hebuilds on his own dualist position as presented in my Chapter Two. Chap-ter Nine explores the question of the origin of the idea of deities. Deitiesarise, Durkheim argues, in order to ensure that rites are performed. Thatis, deities, like myths, arise as retrospective accounts. Thus, everythingpoints ahead toward the discussion of rites in Book III.

5.3.1 Dualism and the Soul

Most scholars of religion, according to Durkheim, argue that the idea of adeity, or spiritual being, is fundamental to the idea of religion. Durkheim


acknowledges that an idea of a soul or spirit appears in all religions fromthe beginning. But, his theory of Totemism requires that the idea of soulbe a result of Totemism, rather than an underlying cause. Otherwise,the soul, as an origin, would reintroduce individualism. He takes up thisproblem in Chapter Eight.

Durkheim argues that the organization of totemic religion, and theexperience of moral force, explain the belief in a soul. According toDurkheim ([1912:355–56]1915:282; *1995:251), “the soul is none otherthan the totemic principle incarnate in each individual.” He gives hun-dreds of examples that demonstrate the correspondence between variousideas of the soul and the organization of the totemic religions in which theyare found. According to Durkheim ([1912:375]1915:297; *1995:265)the fact that the soul is often portrayed in animal form supports hisclaim that the idea of the soul emanates from Totemism. Durkheimargues ([1912:356]1915:283; *1995:252) that “just as society exists onlythrough individuals, the totemic principle lives only in and through theindividual consciousnesses whose coming together forms the clan. If theydid not feel the totemic principle within them, it would not be; it is theywho put it into things. And so it must subdivide and fragment amongindividuals. Each of these fragments is a soul.”

Durkheim also argues that the dualism that he has argued existsbetween the individual and collective aspects of every social being, whichresults from the experience of moral force, contributes naturally to theidea that there are two aspects to a single being, one physical and theother participating in the sacred life of the group. But, once again, thisdualism has a social and not an individual origin. According to Durkheim([1912:375]1915:297; *1995:265) “the soul has always been considereda sacred thing: in this respect it is opposed to the body, which in itselfis profane.” The distinction, he says ([1912:356]1915:383; *1995:251),corresponds to a real distinction within each individual and the soul/bodydistinction comes to represent the experience of that dichotomy: “the soulparticipates in the nature of the subject . . . in this way it comes to havetwo contradictory features.” Beliefs about the soul assume that the soulis immortal and outlives the individual. While this is a belief, Durkheimargues ([1912:377]1915:298; *1995:266–7) that the belief correspondsto the experience of dualism. The soul represents the group in the indi-vidual. It is true, he says, that the group outlives the individual, which isessentially the same thing as the soul outliving the individual.

According to Durkheim ([1912:386]1915:305; *1995:272) the idea ofthe soul is the origin of the idea of personality. By personality he meansindividual personhood, which is a combination of the individual and


social being. He does not mean individualism in the modern westernsense wherein the public self is an individual self, but in the Greek senseof the private individual. Individual personhood thus has a social origin.For Kant, he says ([1912:387]1915:306; *1995:273), “the cornerstone ofthe personality is the will.” The will is the ability to act in accordance withreason, which is an impersonal principle like the sacred totem. If reasonis understood in Durkheimian terms, as a collective phenomenon, thenDurkheim’s idea of personality is also something that only a rational beinghas. The soul is the symbolic expression of the personality. The person-ality is what we have in common with other persons that allows us to beautonomous from the body. It is not the same thing as our individuation,or our individual will, however. People are individuals, in the sense thatdualism supposes, anyway. But, they are only persons when they acquirepersonality, which is socially constructed.

The irony of this position is that, according to Durkheim ([1912:388]1915:306; *1995:273), “what makes a man a person is that by which heis indistinguishable from other men.” Personhood, as Durkheim lays itout, is not conventional individualism. While we get personhood fromwhat we hold in common with others, however, it is also from per-sonhood that we get the experience of liberty. According to Durkheim([1912:389]1915:308; *1995:274–275) “Passion individualizes and yetenslaves. Our sensations are in their essence individual. But the moreemancipated we are from the senses, and the more capable we are ofthinking and acting conceptually, the more we are persons.” To be ableto think collectively and conceptually is to be liberated from our sensualbeing. Liberty is the freedom to think in collective terms. Primitive peo-ples achieve this when they “dare” to think in terms not drawn from sen-sation. Personhood is not the same thing as individualism. It requires thesocial element represented by our sacred nature as social beings. However,according to Durkheim ([1912:390ff]1915:308ff; *1995:275ff), “even ifthe essential element of personality is that which is social in us, fromanother standpoint, there can be no social life unless distinct individualsare associated within it; and the more numerous and different from oneanother they are, the richer it is. Thus the individual factor is a conditionof the personal factor.”

In the Introduction Durkheim had promised ([1912:386]1915:305;*1995:272) to explain the origin of the category of personality. He doesn’tcall it a category here, and he does not provide for it an origin in emo-tions, as he does for force, cause, time, space, and class. However, hedoes say that “the idea of soul long was and in part still is the most widelyheld form of the idea of personality.” Therefore, as a collective represen-tation, the origin of personality is explained in this section and it does,


indirectly, come from the experience of moral force, through the experi-ence of dualism.

5.3.2 Beliefs Develop to Protect and Perpetuate Rites

In Chapter Nine Durkheim provides an explanation for why the idea ofspirits or deities would have developed out of Totemism. This argumentmakes use of the idea that myths are retrospective accounts that explainrites and point toward the rite as the real purpose of the myth. Thus,this chapter makes a transition from a focus on beliefs to a focus on rites.Durkheim argues that the beliefs come into existence in order to protectthe rites. That is their function.

Mythic persons and legendary ancestors, he argues ([1912:395]1915:312–13; *1995:279), were modeled on the idea of souls. Both havetheir origin in the experience of dualism. The mythic persons developto fulfill specific functions within Totemism, not as explanations of anyreal mythic persons. Mythic persons also provide another explanation ofthe experience of dualism by locating spirit powers outside the individualthat correspond with the experience of moral forces.

Durkheim believes that this explanation for the development of mythicpersons also explains why individual Totemism developed out of collec-tive Totemism. He argues ([1912:401]1915:316; *1995:283) that “theindividual totem has all the essential characteristics of the protectingancestor and plays the same role.” The protecting individual is anotherretrospective development that protects the practice of Totemism.

The idea that there are special ancestors developed, he argues, asguardians of special rites. These ancestors only exist to explain the prac-tices and to ensure that the rites are properly performed. This is importantbecause a homogeneity of sounds and movements is required for the expe-rience of collective unity. These deities provide a mythical explanation ofthe importance of that homogeneity; a homogeneity whose importanceDurkheim explains sociologically. The ancestors who protect those ritesthat unify the whole clan, become for that reason, more important thanother ancestors.

Initiation is the principle tribal cult and the most important spirit, ordeity, in most groups is the guardian of initiation. Thus, the developmentof a primary spirit, or deity, is a natural development responding to theneed to protect the primary totemic rites. The deity myths explain whythe rites must be done, and the rites also influence these myths. But,in spite of their importance, according to Durkheim, primitive peoplesdo not “worship” such mythic persons. He says ([1912:408]1915:321;*1995:288) there was no reason to hold special rites in their honor


because “they are themselves no more than the rite personified.” Thesemythic persons, he says ([1912:408]1915:321; *1995:288), exist only“to explain why the practices exist.” All deities, according to Durkheim([1912:408]1915:321; *1995:288), would have the same origin in pro-tecting the totemic practices: they exist to “oversee the manner in whichit is celebrated.” At this early stage, Totemism has not begun to worshipdeities, as modern religions do. But, the same social needs have createdthe mythic persons of Totemism. All religious myths and legends, heargues, play a similar role.

Durkheim now makes the transition from beliefs to rites, which hetakes up in Book III. He says ([1912:424]1915:333; *1995:299) that“Up to now, we have considered these religious representations as if theywere sufficient unto themselves and could be explained only in terms ofthemselves. In fact, they are inseparable from the rites.” The “as if” inthis quote is important. One of the problems with Durkheim’s argumentis that he has seemed to treat beliefs as if they could be separated fromrites. Book III, finally, focuses on a consideration of the rites.

5.4.0 Conclusion

By the end of Book II, Durkheim has gone as far as he can with an expla-nation of the origin of the sacred based on beliefs. He has dispensed withhis major competitors and cleared the ground for his own argument. Hehas also offered a novel approach to the understanding of myths, deities,and symbols in general, as retrospective representations of collectivelyenacted emotions.

The problem with the argument thus far, is that Durkheim cannot reallydescribe how it is that mana works to create the idea of moral force withoutdescribing the rites through which moral force is generated. As he saysat the end of Chapter Seven, beliefs and rites cannot be separated. Andyet the analysis of Book II has tried to do just that. He has worked ritesendlessly into his examples. But, because he has organized the book insuch a way as to separate the discussion of beliefs and rites, and placed thediscussion of rites last, after the discussion of beliefs has been completed,he cannot, at this point, use the detailed examples of rites, the way he willin Book III, to show how rites cause beliefs in any coherent way, or whenhe does try his argument becomes obscure. This leaves time and space,class, and force with less convincing discussions than causality, which isdealt with in Book III, in a full context of rites. This also means thatthe ordering of his argument for the various categories has quite a bit todo with what Durkheim believes he has to argue in a belief framework,


as opposed to what he can argue in a rites framework. Causality clearlyrequires a rites argument, and is therefore the clearest case of Durkheim’sepistemological argument. Because of this, and because of the prioritythat he gives to arguing with others, Durkheim’s empirical argumentswith regard to various categories are oddly split up between Books IIand III.

6 The Primacy of Rites in theOrigin of Causality

While the title of Book III is “The Rites,” the primary task of Book IIIis to establish the social origin of the category of causality. There is nocontradiction here, as the argument for causality constitutes the centraldemonstration that rites precede and cause beliefs, or more fundamen-tally that rites cause the categories of reason. It is the experience of themoral force of that creation; of the necessity involved, that constitutes thecategory of causality.

Before proceeding to a discussion of the origin of causality in enactedpractices, however, Durkheim first begins to establish the process throughwhich religious rites can cause collective ideas in their participants. Hewants to establish that in principle treating rites as the cause of empiricallyvalid beliefs is not subject to certain obvious problems before unveiling hisown argument. It is a slow and cautious approach, typical of Durkheim,which leaves the important argument for the categories deeply embeddedin the text. Because of the length of the argument, the variety of counterarguments Durkheim considers and empirical presentations he makes,the argument for causality will be analyzed in three distinct parts over thenext three chapters.

The first two chapters of Book III, which look like a survey of types ofreligious rites, in actuality constitute a demonstration of the importanceof various rites in relationship to the generation of the sacred. Negativerites protect the sacred while positive rites, according to Durkheim, actu-ally create sacred things. Following these two chapters, the first two sec-tions of Chapter Three describe mimetic rites, those rites that Durkheimwill say cause the category of causality. Durkheim will not deal with theconcept of causality itself until Section iii of Chapter Three, Book III.The argument with regard to the category of causality itself will be dealtwith in the next chapter. This chapter will consider the discussion of ritesand their relationship to beliefs as they occur in the first two chapters ofBook III.

The form of Durkheim’s argument is once again misleading. He pro-ceeds methodically from question to question, attempting to deal with all

194

The Primacy of Rites 195

the preliminary questions before taking on the more important ones. But,he does not say where the argument is going. As a result Book III openswith apparently trivial discussions of various forms of rites. It works itsway only slowly to the more important issue of the relationship betweenthose rites and the category of causality. The discussion of how rites causebeliefs is more central to Durkheim’s argument and more original thanthe earlier discussion of beliefs, which consisted primarily of refutationsof existing arguments. However, because Durkheim embeds it in a longrefutation of all the conventional arguments about rites, arguments thattreat beliefs as primary, and that he feels might be used against him, thepoint is easy to lose.

While these chapters were obviously intended to clarify the importancethat Durkheim placed on rites, because of the way the book is organized,the effect may actually have been to make the section on rites appear tobe less important than the section on beliefs. But, it is a critical mistaketo read Durkheim’s text in this way. It seems to be Durkheim’s style ofargument, at least in The Elementary Forms, to arrive at his most importantpoint last after an exhaustive consideration of all possible conventionaland opposing points of view. This leaves his argument for the social originof causality, the most important argument for an epistemology, waitinguntil Book III, Chapter Three. Even in Chapter Three the elaboration ofrites begins again and continues for two more sections. The argument forcausality begins in almost complete obscurity in the middle of a chapter.

Durkheim produces all of this detail about aboriginal rituals because hisargument for causality requires a focus on rites. He must first establishthe primacy of rites over beliefs. For this reason, the argument couldnot be made in Book II where beliefs were being considered. Even theargument for force was escaping the boundaries of that section. Theproblem is that Durkheim does not say any of this. He does not explainwhat the long sections on rites are leading up to. In characteristic fashion,Durkheim begins Book III with a consideration of how other scholars haveexplained the importance and development of rites in different religions.He assumes that the reader has been able to follow his argument andknows why he is doing this. A review of these sections will attempt topick out the issues considered in them that were critical to Durkheim’sargument.

Durkheim develops his discussion of rites related to the category ofcausality in several stages. First, in Chapter One he considers what herefers to as “the negative cult,” arguing: 1) that only certain forms of thenegative cult, are essential to the idea of the sacred; 2) that there mustbe causes of the negative cult; 3), that the cause is contagion, but askswhere the ability to be contagious comes from; 4) the answer, he says, is


that emotions cause contagion. The identification of emotions with theeffects of rites is a critical move in Durkheim’s argument. He will go onto argue that certain forms of collective emotion constitute a special formof social perception.

Second, in Chapter Two, Durkheim considers the positive cult. 1) Hedescribes the phases of the cult; 2) then he introduces sacrifice as the firstform of sacrificial communion. Sacrifice is important to Durkheim as atype of creation; 3) Durkheim also takes time to address various objec-tions to his position, in particular those of Robertson Smith, that primi-tive beliefs that the deity requires human sacrifice is illogical. Durkheimargues that the belief is not illogical because the idea of the deity wouldcease to exist if the rites were not performed.

This brings Durkheim to the consideration of mimetic rites in the firsttwo sections of Chapter Three, a discussion which will be consideredin the next chapter. A discussion of causality as a category follows inChapter Eight.

6.1.0 Book III, Chapter One: The Negative Cult

Durkheim opens Chapter One by distinguishing, what he calls, nega-tive and positive rites. The distinction is important because he will arguethat participation in a negative cult is a precondition for participationin a positive cult. One must be made ready. But, the negative cult can-not produce the idea of the sacred. Negative rites can only keep apartthings that are already sacred, thereby keeping them sacred. All religionshave prohibitions of this sort. Durkheim also distinguishes negative ritesfrom magical rites. He argues that there are two different kinds of pro-hibitions. Some belong to religion and some to magic. This is a dis-tinction that is always important to Durkheim. He does not want therites on which he bases his argument to have anything to do with magic.Magic is an individual manipulation of means to ends. Religion is a col-lective enactment of sacred things. According to Durkheim ([1912:430]1915:339;*1995:305): “Magic prohibitions propose an entirely secularidea of properties [characteristics that should not mix] nothing more.”Therefore, magic prohibitions do not really have to do with sacred andprofane, just with keeping things apart.

6.1.1 Religious Prohibitions

Rites that have to do with sacred things have an element of necessitythat merely magical practices do not have. Durkheim argues ([1912:430]1915:339;*1995:305) that “religious prohibitions are categorical imper-atives and magic ones are utilitarian maxims.” He will deal in Book III


only with religious prohibitions, not magical ones. He then makes a fur-ther distinction within that set of prohibitions between those that regulaterelationships and those that prevent contact between the sacred and theprofane. The first, he ([1912:431] 1915:340;*1995:306) says, “cannotmake up a cult, proper, for a cult is above all made up of regular relationsbetween the profane and the sacred as such.” The first sort of prohibitionregulates relations within the group. But, he argues that they are not thebasis for the important rites that generate the essential emotions and cat-egories, because within the group the sacred and profane make constantcontact. Durkheim argues that moral forces can only be produced whenthe sacred and profane are sharply defined.

It is the system of prohibitions that directly separates the sacredfrom the profane, according to Durkheim. He argues ([1912:431] 1915:340;*1995:306) that the system of prohibitions “derives directly fromthe notion of sacredness, which it expresses and implements.” It is thissystem that is the basis of the important cults that Durkheim consid-ers. This system of prohibitions “expresses” the sacred, the fundamentalmoral force without which human reason cannot develop, and the cultwhich develops around the sacred will generate the other emotions andcategories.

For the remainder of Chapter One, Section i, Durkheim reviews theprinciple forms of the cult of prohibitions found in Australia. The dif-ferent forms of prohibition are, 1) prohibitions of contact; prohibitionsof eating, touching and speech. Sacred and profane things cannot evenbe mingled. Everyday clothing must be removed before participating insacred rites. Secular occupations must be suspended because, accordingto Durkheim ([1912:439] 1915:346; *1995:311), “work is the preeminentform of profane activity”; 2) Prohibitions of space; religious and profanelife cannot exist in the same space; 3) Prohibitions of time; religious andprofane life cannot exist in the same time. This is of course the real sourceof the idea of time.

6.1.2 Making Ready To Approach the Sacred

In Chapter One, Section ii, Durkheim begins the argument that the neg-ative cult is not just negative. By keeping the person from the profanethe negative cult gets them ready to make contact with the sacred. Thus,Durkheim ([1912:441] 1915:348;*1995:313) says, the negative cult “isin a sense a means to an end; it is the precondition of access to the positivecult.” The negative rites are also different in that they seem to be morepersonal than the positive rites. Durkheim points out that they are oftendone alone. This supports Durkheim’s argument that only the positiverites are true acts of social creation. Negative rites can increase religious


zeal because the person knows what is coming. But, only the positive cultcan actually produce emotions in common.

Then the text ([1912:445] 1915:350;*1995:316) takes up religiousasceticism, which follows directly from religious prohibitions. Believerscan come to devote themselves to the prohibitions as a way of daily life.This is what asceticism is. The discussion in this section seems to be anindirect comment on the development of Protestant asceticism in Europe.Durkheim says that the negative cult can become bloated and invade thewhole of life. While asceticism is not a rare or abnormal form of religiouslife, it normally stays in the background.

What makes the Protestant Ethic, as it developed in Europe, standapart, is that it made everyday work into a sacred object. Durkheim hasbeen describing aboriginal cults that separated work from the sacred. But,in Protestant asceticism the negative cult grows and everything, includingwork, becomes part of it. This is highly unusual. All that remains profanefor Protestant Ethic believers are outsiders, the enjoyment of life, andemotions other than religious emotions. Because religious emotions aresacred, persons should experience no other emotions.1

Pain is one of the necessary conditions of asceticism, according toDurkheim ([1912:446] 1915:351;*1995:317), because our very flesh isprofane and we must pull away from it to become sacred. In this sec-tion, Durkheim gives hundreds of examples of painful rites. Durkheimpoints out that all modern religions also believe in the sanctifying powerof pain. Therefore, he argues, modern rites are based on the same beliefsas Totemic rites. The implication of these comparisons is that Durkheimfeels that he is still trying to establish the status of Totemism as religion.

Pain also functions as a sign or symbol. Durkheim says ([1912:451]1915:355;*1995:320), “Pain is the sign that certain of the ties that bindhim to the profane world are broken.” The profane world is natural andthe human as a biological being belongs within it. In order to preservea distinction between sacred and profane, however, what society does,Durkheim says ([1912:452] 1915:356;*1995:321), is to force persons to“go up the down staircase of nature.” Thus, the unnatural becomes asign of the sacred.

1 For Weber the development of capitalism required a particular form of economic behaviorthat people would only engage in if they had a reason because it involved privation. Weberoffered Protestant Ethic beliefs as that reason. So, Durkheim and Weber are saying thesame thing on this point. The religious reason is only important because it producedthe behavior in question. It was the behavior, and not the belief that caused capitalism,and in Durkheim’s analysis the behavior that also caused the belief. Asceticism is naturalonly as a part of religious practice. And there it is very natural. What seems to happen isthat work becomes an activity oriented toward money, which becomes a sacred totem orsymbol, and consequently should not be spent to support the flesh which is profane.


6.1.3 The Causes of the Prohibitions

In Book III, Chapter One, Section iii, having determined what the systemof prohibitions consists of, Durkheim now looks for the causes of thoseprohibitions. Again, he moves through the text question by question.First, he identifies two causes, a feeling of respect and the phenomenonof contagion. He argues that a feeling of respect causes prohibitions,because the representations that cause these feelings of respect, need to beseparated from the things that contradict them. According to Durkheim([1912:453] 1915:356;*1995:321), “Because of the emotion it inspires, arespected being is always expressed in consciousness by a representationthat is highly charged with mental energy. Hence it is armed in such a wayas to throw any representation that wholly or partly contradicts it far awayfrom itself.” The argument here that representations of the sacred havethe power to repel their opposites builds on a discussion earlier in the text,in which Durkheim described the sacred in terms reminiscent of Kant’sCategorical Imperative: saying that it cannot be held in consciousness atthe same time as the profane.

The sacred also demands respect and must be kept far apart fromthe profane because of the contagiousness of the sacred. Everything thesacred touches becomes sacred through contact.2 Durkheim providespages of examples to support this claim.

Durkheim ends this section with the argument that the action of thesacred in infecting or damaging the profane that approaches too closely,does not depend on a belief in a deity. Rites, he says, are often efficaciousin their own right. One of the effects caused by rites, that does not dependupon beliefs, is contagion. Durkheim ([1912:459] 1915:361;*1995:325)says that “This observation is one that I will have occasion to repeat ineach of the chapters to come.” Indeed he does keep repeating it, becausein each next case he is describing rites that he argues are efficacious intheir own right. There are rites that cause beliefs, whose effects do notdepend on beliefs.

6.1.4 Why is the Sacred Contagious?

Chapter One, Section iv begins with the question of how the contagious-ness of the sacred is to be explained. This is one of those questions thatseems to be about religion, but turns out to be another step in the epis-temology. What Durkheim needs to do first is to argue that it makes

2 Implied in this claim is the idea that the profane cannot be allowed to become sacredthrough contact because the profane is necessary both for sustaining normal life, and fordefining the boundaries of the sacred.


sense to treat contagion as something that really happens. Usually westernthinkers treat the belief in contagion as a matter of illogical “primitive”superstition, or magic. Durkheim needs to argue that contagion is real,because the effects of contagion play a role in his epistemology.

Here Durkheim confronts empiricism, and what he ([1912:459]1915:361;*1995:325–6) calls the “well-known laws governing the asso-ciation of ideas.” He has, of course, argued that these laws of empiri-cism are not the laws governing reason, but only conventions govern-ing what will count as clear thinking in the context of western scienceand philosophy. Even though it is said that the “well educated” manwould never make this mistake, and that “the primitive objectifies theseimpressions naively, without critiquing them,” Durkheim ([1912:459–60] 1915:361;*1995:326) disagrees. He has two responses. The Firstresponse is that sacred effects in modern religions also have the power ofcontagion. Second, he argues ([1912:460] 1915:361–2;*1995:326) thatin everyday life the primitive “does not attribute to one thing the proper-ties of its neighbor, or vice versa.” In other words, a belief in contagionis not evidence that primitive peoples cannot think logically. It is onlywhen speaking of the effects of religious rites, Durkheim says ([1912:460]1915:362;*1995:326), that the thinking of “primitive” peoples appearsillogical to western thinkers: “It is religious thought alone that has amarked inclination toward fusions of this sort.” This, he says, is becausewe have elevated a mythology based on individualism over the socialfacts as they are presented to us in our own religious rituals. We have,in other words, allowed our beliefs to obscure the facts. It is western sci-entific thinking that is obscured by superstition and belief, not primitivethinking.

Durkheim concludes that variations in human intelligence, in the appli-cation of the law of association, for instance, cannot explain the belief incontagion. Religious forces, he says, have the special property of beingexternal to and not intrinsic to bodies. Therefore, they can have relation-ships with things that defy the normal laws of the association of bodies andcan be contagious just like primitive peoples think. This special featureis due to the special nature of religious force, not to primitive thinking.

According to Durkheim ([1912:461] 1915:362;*1995:327) “Religiousforces are in fact only transfigured collective forces, that is, moral forces;they are made of ideas and feelings that the spectacle of society awak-ens in us, not of sensations that come to us from the physical world.Thus, they are qualitatively different from the tangible things in whichwe locate them.” Because the social facts of religious phenomena donot come through the five senses, thinking based on individualism andgeneral ideas is hopelessly distorting when applied to those phenomena.


The contagion of religious forces is explained by Durkheim ([1912:461]1915:363;*1995:327) in the following way: “Since nothing binds themto the things in which we localize them . . . Their intensity pushes themtoward diffusion, which everything facilitates.” The religious forces thatare created by the collective enactment of practices live in the relation-ship between the persons present. Those forces are not located in anyone person or thing. They are felt by all, and as a consequence arefelt by all to be transmittable to anything that comes in contact withthem.

6.1.5 Emotions Cause Contagion

Right at the end of Chapter One, Durkheim ([1912:463] 1915:364;*1995:328) reintroduces the theory of the relationship between emotionsand the creation of ideas that he began to articulate in the discussion offorce, at the end of Book II. This theory of emotion supplies an explana-tion of contagion:

The cause of those [religious] feelings are entirely foreign to the nature of theobject on which they eventually settle. What constitutes those feelings are theimpressions of reassurance and dependence that are created in consciousnessthrough the workings of society. By themselves, these emotions are not bound tothe idea of any definite object. But since they are emotions, and especially intenseones, they are eminently contagious as well. Hence they are like an oil slick; theyspread to all the other mental states that occupy the mind. They pervade andcontaminate especially those representations in which are expressed the variousobjects that the man at that very moment has in his hands or before his eyes:Totemic designs that cover his body, bull roarers that he causes to resonate, rocksthat surround him, ground that he tramps underfoot, and so on. So it is that theseobjects themselves take on religious significance that is not intrinsic to them butis conferred on them from outside. Hence contagion is not a kind of secondaryprocess by which sacredness propagates, once acquired, but is instead the veryprocess by which sacredness is acquired. It settles by contagion; we should notbe surprised that it is transmitted contagiously. A special emotion gives it thereality it has; if sacredness becomes attached to an object, that happens becausethe emotion has encountered the object on its path. It naturally spreads from theobject to all the others it finds nearby.

The emotions generated in persons by enacted practices become asso-ciated with anything they come into contact with. The contagious qualityof sacredness not only finds its explanation in the theory of religion thatDurkheim has offered and provides a confirmation of the theory, but italso explains a feature of primitive mentality; why primitive logic is sodifferent from modern western logic. “Primitive” logic is more closely tied


to the social facts of enacted practices than western logic, which has becomedistorted by a focus on beliefs. Here Durkheim makes a transition to a focuson logic and the theory of knowledge.

Religious thinking, Durkheim says, is the source of logical thought.Because primitive religious forces have the principle of contagion,primitive logic also has this feature. When primitive thinking is not lookedat this way it “seem[s] alien to logical life.” But Durkheim argues thatthe rites have created these ideas: “these fusions and participations haveplayed a logical role, and one of great utility: They have served to con-nect things that sensation leaves separate from one another.” Durkheim isarguing that the primitive mind has succeeded through the logic of sacredcontagion in connecting things that would have remained apart if thoughtwere based on individual reason alone (see 4.2.0). Durkheim argues thatthe primitive belief in contagion is not irrational: “Thus, the sort of fun-damental irrationality that we are at first led to impute to contagion, thesource of that bringing together and mixing, is far from being its distinc-tive mark. Contagion prepared the way for the scientific explanations ofthe future.”

The point is that sensation could never allow persons to connecttogether things that sensation presents as separate. But the emotion gen-erated by religious rites does create a shared emotional experience of suchconnections. Durkheim is not saying here that the things that are con-nected are really intrinsically connected in the ways in which the religiousbeliefs propose. What he is saying is that they really do have in common thatthey are all animated by the same relationship to a religious emotion that is gen-erated by the rite. In connecting these things together the primitive is notwrong about that. In fact, in contrast with Hume’s idea of causality as ahabit of belief, Durkheim is saying here that the primitive is absolutelyand unconditionally right in his thinking with regard to religions emotioninhabiting all of these things. A shared emotional experience generatedby rites does connect them in their own experience. It is a form of internalknowledge, and, according to Durkheim, the primitive person has directaccess to it.

6.2.0 Book III, Chapter Two: The Positive Cult

The negative cult is not an end in itself, it prepares the participant for par-ticipation in the positive cult. The negative cult gives “access to religiouslife,” but, according to Durkheim ([1912:465] 1915:366;*1995:330),“presupposes, rather than constitutes, that life.” In fact, just abouteverything so far presupposes rather than constitutes that life. It is


the positive cult that creates the sacred in the first place. Without thepositive cult, Durkheim argues, the emotions that give rise to the ideasof sacred and profane would never develop.

This chapter takes up the discussion of positive rites in five sections:First, Durkheim discusses the variety of rites that make up the positivecult. They include initiation rites and the Intichiuma. The first not onlycreates new members of the group, but the whole group usually getstogether to participate. The Intichiuma, by contrast, is performed toinsure the reproduction of the species. Durkheim will argue that it isthrough the performance of these rites that the feelings of unity andrespect that give rise to the feeling of the sacred are produced andreproduced; Second, Durkheim describes the phases of rites; Third, heanalyzes rites of sacrifice and their function; Fourth, he outlines Smith’sobjection, that it is a contradiction in the idea of a deity to believe thatdeities require sacrifices; Fifth, Durkheim addresses Smith’s objections,and Sixth, he describes the phases of sacred and profane time.

6.2.1 Phases of Ritual

In the first section of Chapter Two, Durkheim describes the twophases of the Intichiuma rite. According to Durkheim ([1912:467]1915:367;*1995:331), the Intichiuma is celebrated when the good sea-son comes. Each totemic group has its own set of rites for this event. AndDurkheim describes the many ritual types that compose it. The questionthat he says he is trying to answer here is whether they all have a commonorigin. Because, he argues, the positive cult is the origin of the sacred,then, the rites that constitute it must occur everywhere. For Durkheimthis is an empirical question, and they must all be described and explainedseparately before that question can be answered. The discussion focuseson the universality of the Intichiuma because the universality of the ini-tiation rites is already well established.

There is a first phase of the Intichiuma to ensure the well-being ofthe animal or plant. The important thing with regard to Durkheim’sargument are the “means” used in doing this, which he describes indetail. Durkheim also describes how the Witchety Grub clan conductsits Intichiuma. The point is that they believe they insure the reproduc-tion of the totemic species through participation in this rite. Accordingto Durkheim ([1912:475] 1915:373–4;*1995:337) “For the native, theefficacy of these rites is beyond doubt: He is convinced that they mustproduce the results he expects of them, and with a sort of necessity.” Itis this feeling of necessity with regard to the rite that Durkheim connects


with the idea of moral force. He will need to demonstrate that this feelingis generated by the rite and not just associated with it through belief andconvention.

Following what Durkheim refers to as “act one of the feast” there aremany ritual prohibitions. This goes on until another ceremony takes placethat ends the sacred phase and brings everything back to the profane.Durkheim says ([1912:477] 1915:375;*1995:338) that “There is a finalceremony to bring these extraordinary prohibitions to an end and adjournthis long series of rites.” Then he describes the many forms taken by thisending ceremony.

6.2.2 The First Form of Sacrificial Communion

In spite of the endless detailed descriptions of rites at this point in thetext, Durkheim never describes rites or beliefs unless the descriptionsare required by his epistemological argument. Sacrifice is essential toDurkheim’s argument because it supports the idea that primitive peoplesthink of their essential nature as having an origin in their ritual prac-tices. The interest in the rites of initiation and the Intichiuma is that theyconstitute early forms of sacrifice. According to Durkheim ([1912:480]1915:377;*1995:340) “What gives the system of rites just described itsinterest is that it contains all the principal elements, and in the mostelementary form now known, of a great religious institution that was des-tined to become a foundation of the positive cult in the higher religions:the institutions of sacrifice.”

In describing these rites Durkheim is attempting to establish that theyare genuine rites of sacrifice. He refers to them as the first rites of com-munion and explains how this can be the case. Because the members ofthe group believe that they are group members only by virtue of hav-ing something of the sacred in them, they must replenish the sacred inthem in order to remain members of the group. First fruits are the mostpotent, which is why the first fruits in many religions have been essential tosacrifice. According to Durkheim ([1912:480–1] 1915:378;*1995:341),they become social beings through consuming this substance. He says([1912:482] 1915:379;*1995:342) that “The people of the clan cannotremain themselves unless they periodically renew the totemic principlethat is in them.” Sacrificing first fruits binds persons in the group to oneanother.

Durkheim argues that this form of communion is as old as the oldestreligion. This is significant because he will argue that it is the origin ofthe sacred.


6.2.3 Smith’s Objection: Why Would Deities Need Sacrifices?

Durkheim has begun to lay out an argument that treats beliefs aboutsacrifice as making practical sense. The member of the group is reallycreated as a member of the group through the sacrifice. But, there arescholars who will object to Durkheim’s position. They will argue that thefacts do not support his conclusion. In particular, Durkheim singles outRobertson Smith as one who will say that the facts contradict Durkheim’sclaim. Smith believed that sacrifice didn’t make any logical sense becauseit presupposed that deities needed help from humans. If a deity was sup-posed to be all powerful they would not need any help. Thus, the ideaof sacrifice logically contradicts the idea of a deity. According to Smith,tribute was the original form of the rite, not sacrifice, and sacrifice laterbecame confused with it.

Smith arrives at his conclusion by a process of assessing the logical con-sistency between the ideas of deity and sacrifice. If a deity is all powerful,then it is a contradiction to believe that they need anything from humans.But, this is just the sort of analysis of ideas that Durkheim has arguedobscures the underlying social facts from view. The truth of the rite,according to Durkheim, has nothing to do with beliefs and ideas. There-fore, a contradiction between the ideas is irrelevant. Durkheim arguesthat these rites of communion are the earliest form of rite, and that theypresuppose just the sort of reasoning that Smith says is irrational. Theyhave endured because the ideas that are used to justify the rites havenothing to do with their real purpose, which is to recreate the unity ofthe group.

Durkheim argues that there is no idea of a deity and no groupwithout the rite of sacrifice. According to Durkheim ([1912:485]1915:385;*1995:347) “The act of offering naturally awakens in peoplethe idea of a moral subject that the offering is meant to satisfy.” He says([1912:487] 1915:383;*1995:345), “Thus he regularly closes the circlethat, according to Smith, is entailed by the very notion of sacrificial trib-ute. He gives to sacred beings a little of what he receives from them andhe receives from them, all that he gives them.” In other words, the act ofcollective sacrifice recreates the experience of the sacred for members ofthe group. The rite creates the idea of the deity as it creates the group.

It is Durkheim’s argument that these sacrifices gave rise to the ideaof a deity, rather than being constrained by the logic of an alreadyexisting idea of a deity as Smith supposes. Durkheim says ([1912:491]1915:385;*1995:348), “Thus we can believe that the practice of the cultencouraged the personification of religious forces – in a secondary way, nodoubt, but one that deserves notice.” Durkheim is arguing that the rites


come before and create the idea of a deity. Therefore, the logic impliedin the idea of a deity cannot be used to analyze the rites. The rites havea function as social facts that exists before the idea of a deity exists. Theparticipants recreate themselves as moral subjects through the act of sac-rifice. They also create the deity as a moral object. It is not the deity, buttheir own moral and social nature that requires the performance of therite in order to be reborn. This is another of Durkheim’s arguments thatthe rites come first, the beliefs second, and the ideas of personal forcesor deities only after.

6.2.4 Meeting Smith’s Objection: Deities are Created by Sacrifices

Having first argued that Smith’s objection is based on a misconceptionabout the function of rites, Durkheim nevertheless, goes on to arguethat there is another answer to Smith’s question. Durkheim asks whetherthere are any good reasons why deities might need man’s help that donot contradict the idea of a deity? His answer is that deities are sacredonly as long as they are thought of as sacred, and that depends on ritescontinuing to be performed. In other words, since it was the rites thatcreated the idea of sacred deities in the first place, if the rites cease to beperformed, the deities will cease to exist. Therefore, the deities need therites of sacrifice to be performed.3

The problem with this discussion from the standpoint of Durkheim’sepistemology is that although he is arguing that it is through rites ofsacrifice that deities are created and recreated, he refers to this processfor several pages with words like “think” and “believe” that suggest thatdeities are caused by thinking about them. This is not what he means.But, for three pages it is what he says. For instance, he says ([1912:492]1915:386–7;*1995:349), “The sacred beings are sacred only because theyare imagined as sacred. Let us stop believing in them, and they will beas if they were not.” He goes on to say that ([1912:492] 1915:386–7;*1995:349) the beliefs cannot be allowed to weaken “without the sacredbeings losing their reality, because the sacred beings exist only in andthrough their representations.”

Durkheim does not mean that the feeling, or moral force, of the sacredoriginates in beliefs and ideas. What he does mean is that the idea ofsacred beings only exists through representation. There is and can be

3 One problem with this argument is that if we go strictly by an analysis of the concepts,there might appear to be a contradiction between the idea of a deity and Durkheim’sargument that this idea is created by man in the first place. But, Durkheim might arguethat an idea of a deity shared by members of a group would have to have a social origin.Any direct experience of a deity would be incommunicable.


no sensible idea of such beings. The feeling, on the other hand, onlyexists when it is enacted through rituals. Durkheim is not arguing thatthe world of “things” can be known only through representation, thingspose an entirely different problem. Rather, he is arguing that ideas that arenot based on sensation, can only be known, referred to or believed in eitherthrough representation or emotion. On these pages he is emphasizing rep-resentation. But, several pages later he will again begin to emphasize therole played by emotion in creating the idea of the deity in the first place.Earlier he was also quite clear that what gives the representations mean-ing in the first place is the shared emotion generated by the enactment ofrites. So, while Durkheim might sound at this point as if he has reducedknowledge to representations, he is not really even limiting knowledge ofsacred beings to representations. That knowledge originates in the sharedemotional experience of the gathering. But, it can only be communicatedand thought about via representations.

Durkheim really makes two arguments at this point. The First is thatpersons need to think about their supreme beings because if they don’tthink about them they don’t exist. The Second is that if the collective doesnot perform the rites of sacrifice they cannot generate the emotion of thesacred and the deity (and the society) will cease to exist.

While the arguments are different, in both cases the original idea of thedeity is created through the performance of rituals. The contrast betweenknowledge based on individual sensation and the shared emotion gener-ated by the categories is maintained throughout. His earlier argumentthat society only exists if the sacred is produced and reproduced throughthe enactment of ritual practices is repeated in the argument that deitiesonly exist if the sacred is produced and reproduced through ritual prac-tices. The representation argument is a secondary reason for performingthe rites. There are no representations in the first place to believe in untilthe rites create the idea of a deity in the first place. The rites argument isprimary. Durkheim ([1912:495] 1915:388;*1995:351) argues:

If, as I have tried to establish, the sacred principle is nothing other than soci-ety hypostatized and transfigured, it should be possible to interpret ritual life insecular and social terms. Like ritual life, social life in fact moves in a circle. Onthe one hand, the individual gets the best part of himself from society – all thatgives him a distinctive character and place among other beings, his intellectualand moral culture. Let language, sciences, arts, and moral beliefs be taken fromman, and he falls to the rank of animality; therefore the distinctive attributes ofhuman nature come to us from society.

Representation is necessary because sacred emotions are only producedduring special social times. Something is required to keep them alive


in between and to allow them to be communicated between minds.Durkheim argues ([1912:495] 1915:388;*1995:351) that “society existsand lives only in and through individuals. Let the idea of society be extin-guished in individual minds, let the beliefs, traditions, and aspirations ofthe collectivity by felt and shared by individuals no longer, and the societywill die.”

6.2.5 Phases of Sacred and Profane Time

Both society and religion have natural phases of sacred and profane times.This is the case because the sacred deities and society itself exist in aheightened way when the group comes together to perform the rites. Thenover time the representations and feelings thus created begin to lose theirintensity. This process continues until the group once again assembles.The sacred and profane times are experienced entirely differently, and allmembers of the group necessarily experience these differences in the sameway. It is not only time that is distinguished by sacred and profane phases;the places where the group gathers to perform the rites become associatedwith the sacred feelings produced there, while the places that individualsgo to take care of business during profane times, become associated withthe profane. This process of spatial association is the same for everyone.

Because of this, Durkheim argues ([1912:492] 1915:386–7;*1995:349), the representations “achieve their greatest intensity when the indi-viduals are assembled and in direct relations with one another, at themoment when everyone communes in the same idea or emotion. Oncethe assembly is dissolved and each person has returned to his own exis-tence, those representations lose more and more of their original energy.Overlaid little by little by the rising flood of day-to-day sensations, theywould eventually disappear into the consciousness, unless we found somemeans of calling them back into consciousness and revitalizing them.”Representations are like an echo of the actual shared emotions producedby the rite. Like echoes they fade away.

Members of the group come together to perform rituals. Those ritualscreate a world of sacred feelings and ideas that are attached to thosefeelings. The sacred feelings, or beings, are represented through emblems.But, as the members of the group disperse to pursue their profane dailylives it becomes harder and harder to remember the shared feelings thatthe representations are intended to call up. When the representationsbecome too weak, then the members of the group need to gather togetheragain. If they do not, the representations will lose their ability to recallthe sacred feelings and the deities and the society will cease to exist.

Durkheim says ([1912:494] 1915:387;*1995:350):


The emotions aroused by the periodic crises through which external things passinduce the men witnessing them to come together, so that they can see what isbest to do. But by the very fact of being assembled, they comfort one another;they find the remedy because they seek it together. The shared faith comes to lifeagain quite naturally in the midst of reconstituted collectivity. It is reborn becauseit finds itself once again in the same conditions in which it was first born. Onceit is restored, it easily overcomes all the private doubts that had managed to arisein individual minds.

Coming together solves the problem. According to Durkheim([1912:494] 1915:387;*1995:350) “The mental image of the sacredthings regains strength sufficient to withstand the inward or externalcauses that tended to weaken it. Despite the obvious failures, one canno longer believe that the gods will die, because they are felt to live againin the depths of one’s own self.”

It does not matter what the actual rites consist of, as long as they arerites of sacrifice and the whole community participates. Durkheim argues([1912:494] 1915:387;*1995:350) that “No matter how crude the tech-niques used to help the gods, they cannot seem unavailing, because every-thing happens as if they were really working. People are more confidentbecause they feel stronger, and they are stronger in reality because thestrength that was flagging has been reawakened in their consciousnesses.”

This whole section has been an argument to prove Smith wrong andDurkheim now says that the way these rites work proves that the godsreally do need men. He argues ([1912:494] 1915:387;*1995:350) that“Thus the purpose of the cult is not only to bring the profane into com-munion with sacred beings but also to keep the sacred beings alive, toremake and regenerate them perpetually.”

Of course, the actual gestures and words do not create the sacred.Durkheim argues ([1912:494] 1915:387;*1995:350):

To be sure, the material offerings do not produce this remaking through their ownvirtues but through mental states that reawaken and accompany these doings,which are empty in themselves. The true raison d’etre of even those cults that aremost materialistic in appearance is not to be sought in the actions they prescribebut in the inward and moral renewal that the actions help to bring about.

The point of the rites is not to effect a mechanical or material end.Even the most material cults, according to Durkheim ([1912:494]1915:387;*1995:350) seek a moral renewal. What the words and ges-tures do is to put the group into simultaneous action. What is required isthat they all say the same words and make the same movements in unison.

The utilitarian exchange of goods that the rite describes according toDurkheim, is not the invention of “utilitarian” theorists, but, is as old


as religion itself. He argues ([1912:495] 1915:388;*1995:351) that “Itarises from the fact that although sacred beings are superior to men, theycan live only in human consciousness.” Sacrifice is necessary, exchangeis necessary, because the creation of sacred moral subjects dependsupon it.

Durkheim has referred to the relationship between individual, soci-ety and deity as a circle. Within this circle individuals come together toperform a rite together. Through doing so they transform themselves.They attribute this transformation to a deity which they then feel theyneed to feel acting within them. Because they feel this need they con-tinue to enact the feeling of the deity. Thus, they continue to transformthemselves. Durkheim says ([1912:495] 1915:388;*1995:351) that “ifpressing the analysis further and substituting for the religious symbolsthe realities they express, we enquire into the way those realities behavewithin the rite, this circle will seem to us even more natural, and we willbetter understand its sense and purpose.” This is a very important sen-tence because it not only says that the religious symbols express socialrealities, but that the key social realities in question are ones which have a“behavior within the rite.” This, as he says, closes the circle because thereare enacted social realities that create shared emotions which is what thereligious ideas express.

He has answered Robinson Smith’s question by showing that the beliefthat a deity requires sacrifices in order to survive is true. Discover-ing the mental mechanism behind the belief is a first step toward theanswer. Durkheim ([1912:495] 1915:388;*1995:351) says that “When,from beneath these outward and seemingly irrational doings, we haveuncovered a mental mechanism that gives them sense and moral import,we have made a step toward solving this problem.” But, it is also necessaryto discover whether the mental mechanism has any real basis in fact.

It is at this point in the discussion that Durkheim switches from talkingin terms of beliefs and representations to talking about concrete practices.

First, he has attempted to establish a logical mental mechanism. Thatis, that the belief itself is logical because of its results. But then Second,he attempts to establish a real empirical basis. This is a bit confusingsince it is hard to see how the two can really be separated, since all ofthe arguments of the first part seem to be to entail the idea that sacredfeelings are really produced by the rite. But, in this second part he may begoing farther and arguing that it is not just sacred feelings that are beingexperienced, but that the rite is a genuine act of creation. This wouldmove him a step closer to the argument for the category of causality,because participants experience genuine creation, which is a principlepart of the idea of deity and of causality.


Having completed his proofs Durkheim begins summing up. In refer-ring to beliefs, Durkheim says ([1912:496] 1915:389;*1995:352) “Fromthe fact that it can be explained psychologically does not follow that thisbelief has objective value.” In order to establish that it has objective value,he says “it must be possible to establish that the effect of the cult is peri-odically to recreate a moral being on which we depend, as it dependsupon us. Now this being exists: It is society.”

Durkheim then continues with what has now become a socialfacts argument. The importance of religious ceremonies, according toDurkheim ([1912:497] 1915:389;*1995:352) “is that they set collectiv-ity in motion.” It is their social effects that are most important. “Thus,their first result is to bring individuals together, multiply the contactsbetween them, and make those contacts more intimate. That in itselfmodifies the content of consciousness.”

Then Durkheim ([1912:497] 1915:389–90;*1995:352) summarizesthe contrast between ordinary and sacred periods. During ordinary peri-ods the representations “live on their past, and, in consequence, theywould in time be depleted if nothing came to give back a little of thestrength they lose through this incessant conflict and friction.” The con-flict is between the profane needs of everyday life and the sacred rep-resentations of society. Consequently, the positive cult takes a periodicform. During such periods, Durkheim says ([1912:497] 1915:390–1;*1995:352), “The individual soul itself is also regenerated, by immersingitself once more in the very wellspring of its life.”

Durkheim refers to an “impulse toward periodicity” ([1912:498–9]1915:391;*1995:353). Because normal daily life demands instrumentaland utilitarian action it cannot co-exist with the intense periods of pro-ducing the sacred that occur during sacred periods. He has said thisseveral times before and this section probably overlaps with several oth-ers. According to Durkheim it is not the “external periodicity” of theseasons, or of real time that creates this same consciousness. It is the nec-essary periodicity and tension between sacred and profane times. Seasonsare critical periods for nature. But, sacred and profane times are criticalperiods for society and the sacred.

One of the things that Durkheim says happens as societies becomemore complex is that they develop less of a tolerance for this periodicityand find ways of making sacred times occur with more regularity, or atleast stopping the fluctuation.

7 Imitative Rites and the Category of Causality

Following the first two chapters of Book III, Durkheim introduces imi-tative rites in the first two sections of Chapter Three. This discussion isparticularly important because these are the rites that Durkheim arguescreate the category of causality. Therefore, particular attention will begiven to the argument with regard to mimetic rites in this chapter. Inthe first two chapters of Book III Durkheim prepared the reader for thisargument, by showing that rites in general create collective emotions andrepresentations, even society. If he can meet objections to that argument,then he may be able to meet objections to the argument that a particularform of rites creates a particular category. Furthermore, the category inquestion is in a sense the category of creation, and what he has arguedfor two chapters is that rites create something entirely new. Durkheimhas been establishing creation as a result of rites. Now he will establish that“necessary causes” are created through the performance of a particularform of rites.

First, in Section i, Durkheim describes mimetic rites, and Second, inSection ii, he explains the principle behind mimetic rites. This involves1) a distinction between mimetic rites and magic, 2) a further criticismof empiricism, 3) an explicit examination of the idea that rites have pri-macy over practices, 4) a discussion of the purpose of imitative rites, 5)a discussion of their social utility, and 6) the argument that specific ges-tures have no inherent magical efficacy. Durkheim needs to be sure boththat Totemic rites are not confused with magic and that his argumentthat rites have efficacy will not be confused with magic. A discussion ofCausality itself, as a category, follows in my Chapter Eight.

7.1.0 Book III, Chapter Three, Section i: Description ofImitative Rites

For the opening of a very important chapter the beginning of this sectionis remarkable for saying nothing about its significance. In the previouschapter, Durkheim had discussed the positive cult and the negative cult

212

Imitative Rites 213

and their relationship to the sacred and moral force. In Chapter Threehe introduces a particular form of rite, the “mimetic” rite, the enactmentof which, he argues, is the social production of the idea of causality.Durkheim ([1912:501] 1915:393;*1995:355) opens the chapter with thewords: “The techniques just discussed [a reference to rites of sacrificeand oblation discussed in Chapters One and Two] are not the only onesused to bring about the fertility of the totemic species. Others with thesame purpose either accompany them or take their place.” He then goeson to describe mimetic rites in great detail. But, in the opening sectionhe gives no hint as to the importance of the mimetic rites and does notmention causality. If one does not understand already where the argumentis going, one cannot make the connection. His presentation at this pointappears to be purely descriptive.

There are a series of pages describing various mimetic rites. These rites,he ([1912:501] 1915:393;*1995:355) says “are composed of movementsand cries intended to mimic the behavior, or traits, of the animal whosereproduction is hoped for.” The point of these descriptions seems to be tosupply sufficient detail for the reader to understand that the participantsin the ritual really are trying to imitate an animal or plant. Durkheim([1912:504] 1915:395;*1995:357) also gives examples in which partici-pants imitate inanimate objects.1

Durkheim’s reference to the point of the rite being to bring “about thefertility of the totemic species” could easily create the mistaken impres-sion that he is interested in whatever produces the fertility of the totemicspecies, either a naturalistic or a religious notion, and not in the principleof causality. But, in fact, what creates the fertility of the totemic species is,according to Durkheim, that which creates in all members of the groupthe feeling of moral force. That feeling of moral force is in turn created bythe group through the enactment of the ritual. Therefore, in speaking ofbringing about the fertility of the totemic species, Durkheim is referringto his epistemological argument, although he makes no reference to it atthis point.

7.2.0 Book III, Chapter Three, Section ii: The PrincipleBehind Imitative Rituals

It is only in Section ii, of Chapter Three ([1912:508] 1915:398;*1995:360), that Durkheim begins to explain that mimetic rites have an underly-ing social purpose in creating group unity. However, while Section ii is one

1 Durkheim’s reason for making this argument is discussed in 4.1.1. The thing chosen asa totem need not have any inherent virtues. It becomes sacred by virtue of becoming anemblem.


of the more important sections in terms of Durkheim’s theory of causal-ity, and he discusses the idea of causality in conjunction with mimeticrites in this section, he does not mention the “category” of causality atthis point, nor does he discuss its relevance to either philosophy or epis-temology. Rather, he focuses on establishing the claim that mimetic ritesenact a basic principle of causality and that there is an element of creationinvolved in their performance. Durkheim does not make the connectionto epistemology until Section iii, after he has dealt with his opponents andestablished that the idea of causality has social origins, social functions,and immediate and necessary social effects upon participants in mimeticrites.

Durkheim ([1912:508] 1915:398;*1995:360) opens Section ii with theclaim that “all of these rites” he has described “belong to the same cat-egory.” That is, they are all imitative rites. He then proceeds to differ-entiate his own approach to mimetic rites from the standard approach.The language in the paragraph that follows is tricky. Durkheim beginsby outlining the way mimetic rites are “usually” thought of. If the para-graph is not read carefully, the important distinctions made here betweenwhat is “usually” thought, and what Durkheim himself argues, are easilyconfused with the result that Durkheim can appear to be espousing theposition that he is actually criticizing.

According to Durkheim, imitative rites are usually, and he says improp-erly, thought of as “sympathetic magic.” These rites, he says, are alsousually further subdivided, also improperly, into two principles. Thefirst of these two principles he refers to as the principle of contigu-ity ([1912:508] 1915:398;*1995:360, emphasis in original): “Whatevertouches an object also touches everything that has any relationship of proximityor solidarity with that object.” The second he refers to as the principle ofresemblance, usually summarized, he ([1912:508] 1915:398;*1995:360)says as “like produces like.”

Durkheim says that the second principle is the one that the rites thathe has been describing put “into operation.” His way of putting it wouldseem to suggest that he agrees with this way of understanding the prin-ciple, but he does not. In characteristic fashion, Durkheim accepts thesecond principle, but not the empiricist explanation given of it. Con-tiguity and resemblance are the mental functions with which empiri-cists try to explain the principle. According to Durkheim, while he alsounderstands the principle behind the mimetic rites in the form “like pro-duces like,” contiguity and resemblance cannot explain the principle,because it includes the idea of true creation, which is more than contigu-ity and resemblance. This argument follows from his earlier critique of

Imitative Rites 215

empiricism. Durkheim argues that the principle, as enacted by the rite,represents a true creation, not an association of ideas by either resem-blance or contiguity.

When Durkheim ([1912:508] 1915:398;*1995:360) characterizes hisown position, he distinguishes it from the empiricist position, arguingthat “the rites that concern us are a different case.” In Durkheim’s viewthe principle as it is “usually” considered leaves out its most importantelement: the idea that something “entirely” new is being created. Accord-ing to Durkheim ([1912:509] 1915:398–9;*1995:360–1), the rites: “pre-suppose not merely the passage of a given state or quality from one objectinto another but the creation of something entirely new.” Thus, Durkheimchallenges the claim by other scholars that in performing these rites prim-itive peoples are making use of (and in fact using badly) basic forms oflogical thought; the ability to think in terms of contiguity and resem-blance, and nothing more. Durkheim insists that there is a real creationrepresented in the performance of the rite and that primitive peoples arenot wrong in representing it as a creation. The persons who perform therite have the expectation, according to Durkheim ([1912:509] 1915:398–9;*1995:361), that: “The very act of depicting the animal gives birth tothat animal and creates it.” He argues that contiguity and resemblancecannot explain how the performers come to have this expectation, and hesets it as his task to explain where the idea of a pure creation could havecome from.

7.2.1 Mimetic Rites are Religion not Magic

Many of Durkheim’s contemporaries believed that the ideas behindmimetic rites came from magic. Durkheim rejects the idea that mimeticrites are magic, or that the idea of causality comes from magic. He arguesthat in order to have the idea of magic in the first place one would alreadyhave to understand the principle of creation, or causality. One wouldnot try to create something magically, unless one already had the ideaof creation. While magic often includes the idea of causality or creation,the idea of creation, Durkheim argues, could not have come from magic,because magic presupposes it. Therefore, the origin of magical ideas alsoneeds to be explained, and consequently, cannot provide the explanationfor this principle. Durkheim argues that the idea of magic must have comefrom the principle of causality as expressed in religious rites, and not theother way round. What needs to be explained is where both the idea ofmagic and the category of causality come from. The ideas of contiguityand resemblance explain neither of these ideas.


A major theme of The Elementary Forms, which runs throughout thebook, has been that the totemic rites are, properly considered, religionand not magic. The mimetic rites, in particular, have typically been asso-ciated with magic. Therefore, in this section, in which he argues thatmimetic rites are the most basic of all religious rites, and that all otherreligious thought and practice is derived from them, Durkheim is mak-ing an argument that runs contrary to the received thinking in religiousstudies, and in the sociology of religion, at the time when he wrote. Hisposition is completely opposed to that taken by Max Weber, for instance,who considered mimetic rites to be magic, and religion to have developedfrom magic.

This is an important point for Durkheim because if mimetic rites wereconsidered to be magical rites then they would also be individualistic andinstrumental in ways that are problematic, not just in terms of the par-ticular argument with regard to causality, but also in terms of his overallargument for the social origin of the categories of the understanding.None of the categories could have empirical validity if the religious prac-tices that create them began with individual perception. Magic beginswith individual instrumental action and involves the idea that a specificgoal is to be achieved by either the individual or the collective perfor-mance of a rite. It does not involve the idea of the sacred, and cannotexplain the origin of that idea. Nor are there effects of magic that areeither purely social or collective.

This distinction between religion and magic is essential to Durkheim’sargument and he has carefully distinguished religion from magic in theThe Elementary Forms. Additionally, much of what is so distracting aboutBook II of The Elementary Forms is addressed to proving that Totemismis a religion, with a collective and not an individual origin, counteringthe belief that it is magic. Religion, according to Durkheim, is socialand collective in ways that magic clearly is not. Religion, he says, has achurch of more or less equal believers. Magic, on the other hand, has aperformer and an audience. There is no unity in the experience of magicand therefore it has no church, creates no moral forces, and no categories.If causality is to have a collective origin, it must be in religious rites thatact out unity, and not in magic.

Durkheim returns again, at the end of Section ii, to the argument thatmagic cannot be the origin of religion. He ([1912:516–7] 1915:404–5;*1995:366) argues that magic did not come before religion, religioncame before magic: “The precepts on which the magician’s art rests[causality] were formed under the influence of religious ideas.” Theirapplication to secular concerns is secondary. According to Durkheim([1912:517] 1915:404–5;1995*:366), Hubert and Mauss

Imitative Rites 217

showed magic to be something altogether different from crude industry, basedon crude science. The principle of causality, created through religious practice,becomes detached from religion and available to magic and later to science. Theyhave brought to light a whole background of religious conceptions that lie behindthe apparently secular mechanisms used by the magician, a whole world of forces,the idea of which magic took from religion.

According to Durkheim ([1912:517–18] 1915:404–5;*1995:366): “oncethis principle that like produces like took form to satisfy definite religiousneeds, it became detached from its ritual origins and, through a kindof spontaneous generalization, became a law of nature.” The individualand instrumental application of the principle of creation, correspondingto science and magic, is a later manifestation of a causal principle thathas its origins in religion.

7.2.2 Criticism of Empiricism

After his opening arguments, Durkheim returns to a consideration of theanthropological school of Tylor and Frazer. He ([1912:509] 1915:398–9;*1995:361) says that “they call upon the association of ideas” here withregard to causality “just as they do to account for the contagiousness ofthe sacred.” In other words, his criticism of the empiricists on this point,parallels his earlier criticism of their explanation of the origin of the sacred(see 3.2.0). Durkheim ([1912:509] 1915:398–9;*1995:361) argues thatusing this empiricist approach, which uses basic logical ideas, to explainreligious phenomena, “is to misunderstand the specific character of thepractices under discussion.”

One of the problems that Durkheim must face in this section is that hisown interpretation of imitative rites, which is critical to his argument, dif-fers sharply from the interpretation current in his day that imitative ritesare based on a crude misapplication of the idea of causality to things thatresemble each other; an interpretation that leads naturally to the conclu-sion that mimetic rites are magical beliefs whose purpose is to manipulatenature, and not religious beliefs, which must have some connection to thesacred. Durkheim argues that the popular interpretations make no sense.Furthermore, because they misunderstand the nature of the rites, theyunderestimate the intelligence of what is actually being achieved.

According to Durkheim ([1912:509] 1915:398–9;*1995:361) theirapproach might “from one point of view” be “applied somewhat jus-tifiably to the case of bewitchment.” In bewitchment there really are twodistinct things: the image and the thing it represents. Because there aretwo things they can be subjected to the logical operations of contiguity andresemblance. But, bewitchment is a case of magic. In mimetic rites there is


only the image or emblem. According to Durkheim ([1912:509] 1915:398–9;*1995:361) “only the image is given in the mimetic rites we have juststudied, and as for the model, there is none, since the new generation ofthe totemic species is still no more than a hope, and an uncertain hope atthat.” Religious rites have different properties from magical rites. Thereis no model, nothing to which the image refers, and it is the image that iscausally efficacious, just as with all rites involving the totemic image (seesection 4.1.1 the totem as emblem).

The causality implied in mimetic rites is, according to Durkheim, not arelationship of reference. Mimetic rites do not have meaning by setting upa relationship to, or between, objects presented to sense perception. Theycreate feelings which only occur during the performance of the rite andhave meaning only within the context of the group. By contrast, Durkheimargues, that the logical ideas of contiguity, or association, only work asexpressions of relationships between things. There has to be somethingto compare to something else before these two ideas, or ways of thinkingabout things, can be applied. In the case of mimetic rites because the ideais one of pure creation, and the “object” of the rite is created by the rite,and therefore does not exist beforehand for comparison, the principles ofcontiguity and resemblance cannot be applied.

Because these principles cannot be applied to mimetic rites, Durkheimargues that the approach taken by Tylor and Frazer, that explains the ritesas a crude application of the association of ideas, is completely misleading.Durkheim maintains that the association of ideas could not lead to theidea of creation: to the idea of something that does not yet exist. Takenstrictly in classical empiricist terms, however, these rituals not only looklike a false application of crude principles to natural events, they makeprimitive peoples look unintelligent. Durkheim argues that no one wouldbe stupid enough to believe in the literal causal efficacy of imitative rituals,unless there was more to the ritual than the crude perception of custom,habit, similarity, and contiguity. Persons must get reinforcement for thesebeliefs in terms of real causal efficacy.

According to Durkheim ([1912:511] 1915:400;*1995:362) “the gen-eral properties of human nature cannot explain such odd practices.” Thephrase general properties of human nature refers here to the basic logicalabilities that humans possess as animals. In Durkheim’s dualism, theseare pre-rational and pre-social abilities. They are the mental functionsassumed by empiricists, and Durkheim rejects them as an explanation.Taking an empiricist individualist starting point and trying to explainthe belief in mimetic rites by the association of ideas creates the falseimpression that primitive peoples are stupid. As Durkheim has arguedagain and again in The Elementary Forms, beliefs that assume so much

Imitative Rites 219

stupidity cannot be the reason why primitive peoples do what they doand believe what they believe. While Durkheim agrees that people areable to form crude causal connections with no empirical validity betweenactions, or events and effects, based on generalizations from experience,he ([1912:509] 1915:398–9;*1995:361) argues that trying to explain thetotemic rites in terms of contiguity and resemblance alone, as the empiri-cists had done, results in treating them as no more than faulty general-izations.

7.2.3 The Primacy of Beliefs versus Rites

Having argued that the empiricists have misunderstood the nature ofmimetic rites, and that these rites in fact embody the principle of purecreation, Durkheim has to explain how this is possible. He ([1912:511]1915:400;*1995:362) says that the rite through which the members ofthe group affirm their kinship also creates and recreates the group. Thetotemic group exists only if believed in, and the rites create the belief.Durkheim ([1912:511]*1915:400;1995:362) argues that “it exists onlyin so far as it is believed in. And the effect of all these collective demon-strations is to support the beliefs on which they are founded.” The belief,however, is a secondary phenomenon. The actual creation of the moralunity on which the belief rests comes first. Otherwise, it would only bean idea.

Because beliefs arise for the purpose of justifying rites, Durkheim([1912:511]*1915:400;1995:362) says that an empirical study of riteswould explain the origin of beliefs. This is very important as it goesagainst the usual idea that beliefs explain rites. It changes the ordering ofboth argument and research, placing actions before ideas. As Durkheimhad argued earlier in The Rules of the Sociological Method, if there is tobe an empirical science of sociology, there must be an empirical domainof facts. Weber’s approach, which began with ideal types, threatened tocreate a sociological domain that was primarily conceptual.2 Durkheim,on the other hand, insisted that ideas and concepts were secondary phe-nomena, and that they could be given empirical explanation. This aspectof Durkheim’s argument speaks to contemporary dilemmas in sociology.It resembles Garfinkel’s insistence that social order and meaning are inthe “details” of what occurs, and that witnessable orders of social scenesare not conceptual. It is also a precursor to what I have called the fallacy

2 Weber’s ideal types could be argued also to rest on empirically specifiable characteristics,see for example page 8 of Economy and Society. But, in most discussions of them Weberappears to treat them as ideal typical concepts. This is part of a tradition in sociology,including both Parsons and Schutz, of privileging concepts over action.


of misplaced abstraction (see Conclusion). In rejecting positivism, philoso-phers rejected the empirical in favor of the conceptual. The argument isthat persons do not perceive the real world, they perceive a conceptualreality. However, a conceptual sociology has not been able to answer thequestion of where the perceived coherence of experience comes from. Itis this question of coherence that Durkheim set out to address.

In spite of having made this very important and revolutionary argu-ment Durkheim is usually interpreted as having argued that beliefs arethe origin of the rites: that the ideal is the foundation of the real. Notonly does this turn his argument around completely backwards, and treatDurkheim as having been an idealist who treated ideas as primary, whenin fact, he treated social action as primary, but, because Durkheim wasnot making this sort of argument, the conclusions that he draws on thebasis of his own argument appear to be absurd when paired with thesort of idealist argument which he is interpreted as having made. If onetakes an idealist position, then there are certain limitations that one mustaccept. Durkheim’s position appears to be absurd, when interpreted asidealist, because he accepts none of the limitations. The idealist inter-pretation assumes that the categories ultimately rest on beliefs, which areonly ideas. Thus, Durkheim’s claim that the categories are empiricallyvalid appears to be absurd.

There may be a confusion of Durkheim’s position on religion withWeber’s sociology of religion on this point. Weber did argue that sharedbeliefs come first and are the basis for action. He also interpreted themimetic rites as magic and not religion ([1921]1968). If Weber had thenargued that these beliefs had empirical validity he would, of course, clearlyhave been contradicting himself. But, Weber made no such argument.Durkheim takes a position that is in direct opposition to Weber’s positionon both points. According to Durkheim, it is the fact of association, ofacting in common, that causes the feelings, ideas, and beliefs which cometo be identified as religious beliefs. That is what makes them religionsand it is why religion cannot evolve out of magic. To explain religion onthe basis of magic (i.e. individual instrumental behavior) would be likeexplaining society as an aggregate of individual attempts to achieve valuedends, not a Durkheimian position. That the feelings have their origin inthe experience of association, is also the basis of Durkheim’s claim thatthe categories have empirical validity.

7.2.4 The Purpose of Imitative Rites

Imitative rites have collective and social purposes. Durkheim will dis-cuss two such social purposes. First, recreating the unity of the group;

Imitative Rites 221

and, Second, the need for outward expression of moral unity through therecreation of the totemic species.

The first purpose of the totemic rites is to create and recreate the unityof the group by imitating the totemic emblem. According to Durkheim([1912:511] 1915:400;*1995:362) “When they are assembled, then, theirfirst act must be to affirm to one another this quality that they ascribe tothemselves and by which they define themselves. The totem is their ral-lying sign.” The sign is effective, not because it refers to some object, butrather, because it serves as a vehicle for generating moral force. Accordingto Durkheim ([1912:511] 1915:400;*1995:362): “They witness to oneanother that they are members of the same moral community . . . the ritenot only expresses this kinship but also makes or remakes it, for this kin-ship exists only insofar as it is believed, and the effect of all these collectivedemonstrations is to keep alive the beliefs on which it rests.”

However, this purpose of affirming the existence of the totemic speciesis so completely tied to Totemism that, he argues, it would not havemade its way into other religions, unless there were some other, morefundamental reason for the performance. The need of the Australian toresemble his totem cannot be the only cause of the efficacy of the principleof “like produces like,” according to Durkheim ([1912:511] 1915:400;*1995:362) because that efficacy applies to what is most specific to thetotemic beliefs and yet the principle itself can be found in all religions.Therefore, there must be another reason why rites which embody theprinciple of “like produces like” have survived. It cannot have literal effi-cacy because “like,” literally interpreted, does not produce “like.” There-fore, according to Durkheim there must be a second principle involved.Or another social cause to explain why the rites continue to be performed.

It is this second purpose that involves true creation. According toDurkheim ([1912:511] 1915:400;*1995:362) “In fact, the very generalpurpose of the ceremonies in which we have seen it applied is not onlythe one I have just mentioned, fundamental though it is, for they also havea more immediate and conscious purpose: to bring about the reproduc-tion of the totemic species.”Durkheim([1912:512]1915: 401;*1995: 363)argues that “a single concern cannot haunt an entire group of men to thatextent and not become externalized in tangible form.” Their concern withcreating and recreating group unity becomes focused on the animal orplant that represents the totemic group. Furthermore, their unity takes amaterial form: “This thinking in common is inevitably manifested out-wardly by movements.” These movements, he says, work best if they areimitative of the animal or plant. Imitative movements are most effectivefor stimulating a unity of thought and feeling, not for religious reasons,but because “there are no movements that as closely resemble the idea


that fills consciousness at that moment, since they are its direct and almostautomatic translation.” Consciousness is filled with the unity of the groupand the totem as its emblem. Imitating the totemic object through move-ments and cries gives outward form to what is simultaneously in all theirinward experience. This is a first statement, at least in this section, of thepurpose of mimetic rites in producing thought in common.

Just as the unity of the group requires an outward expression in thetotem, the collective performance and emotional experience of the riterequires outward expression in movements in common. According toDurkheim ([1912:512] 1915:401;*1995:363) “The people do their bestto imitate the animal; they cry out like it; they jump like it; they mimicthe settings in which the plant is daily used.”

This need for outward expression gives imitative rites their universalcharacter. It is a need that Durkheim ([1912:512] 1915:401;*1995:363)says is not confined to any single religion: “Nor is this the need of anyone era or caused by the beliefs of any one religion. It is quintessen-tially human.” It is a universal need of collective life. As soon as peopleare gathered together they are almost compelled to do it. “To be sure,”Durkheim ([1912:512] 1915:401;*1995:363) says, “speech is one meansof expressing it, but movement is no less natural. Springing from the bodyjust as spontaneously, it comes even before speech or, in any case, at thesame time.”

Collective life is not possible without the possibility of communicat-ing common experiences between persons. Durkheim suggests that theseimitative movements may even facilitate speech. The problem is thatsome referential words and gestures can obviously communicate individ-ual intentions without common ideas. Other thoughts, however, can onlybe communicated after the development of common ideas. Referentialsymbols, such as “eating gestures,” are not meaningless. But, Durkheimis referring here to cases where the same idea, or symbol of a feeling iscalled up in the mind, not a referential relationship that is more or lessaccurate. In fact, none of the ideas in question have a referential meaning,only shared experiential meaning.3

Next Durkheim poses a further question. “But even if we can thusunderstand how these movements found their way into the ceremony”he ([1912:512] 1915:401;*1995:363) says, “the power that is ascribed tothem” must still be explained. If we take them literally, as reproducingthe physical species, they still seem absurd.

3 This way of explaining meaning in terms of the shared details of experience is close toWittgenstein’s “use” meaning in essential respects. For instance, Durkheim has specifi-cally denied that reference is involved in his critique of Tylor and Frazer above. Unfortu-nately, Durkheim does not work any of this out.

Imitative Rites 223

Durkheim’s contemporaries have assumed that mimetic rites do notreally accomplish their goal. However, Durkheim argues that whether ornot the rite is actually causally efficacious is an empirical matter of fact,not a matter of logic. If the true purpose of the rite can be discovered,then the effectiveness of the rite can be assessed empirically. Therefore,he argues, the possibility that the principle of causality is generated in andthrough imitative rituals needs to be established through careful empiri-cal examination, not simply dismissed. Durkheim argues that instead ofconsidering the principle in its general and abstract form “let us connectit with the system of ideas and sentiments” which the “rites put intopractice” (Durkheim, [1912:511]*1915:400;1995:362).

In other words, a careful and detailed study of the rites in question,reveal that they really do create something new and therefore, the beliefof the primitive that, in performing the rite, they are creating something,not only has a foundation in “fact,” but also in emotions and feelings thatcorrespond to that fact (or are the same thing as that fact).

7.2.5 The Social Utility of Mimetic Rites

While the empiricists had argued that mimetic rites constituted a faultyapplication of the ideas of contiguity and resemblance, Durkheim arguesthat the rites have a real utility in creating shared experiences. Accord-ing to Durkheim ([1912:513] 1915:402;*1995:364) The moral powerof the rite over minds, “Which is real, made them believe in its powerover things, which is imaginary.”4 It is their power to remake the totemthrough ritual interaction, or their necessary role in the remaking of thetotem, which gives participants a sense that they have power. If per-sons use rituals to achieve mutuality, this is something of which they

4 There is a translation problem here in both English translations. I have avoided the prob-lem in my text by supplying only a partial quote. But, it is worth mentioning as it couldeffect the way the text is read. The original French (1912:513) reads “L’efficacite moraledu rite, qui est reelle, a fait croire a son efficacite physique, qui est imaginaire. . .” The1995 (364) translation reads “The power of the rite over minds, which is real, made thembelieve in its power over things, which is imaginary.” The translator, Karen Fields, hasmade a note on the translation at this point (364ff) to the effect that “Here the term ‘moral’refers to mind as opposed to matter.” Presumably, the assumption is that Durkheim istrying to articulate a mind body dualism. What Durkheim contrasts, however, is an imag-inary physical effect of the rites with a real moral effect, not a mind body dualism. He doesnot mean the power of the rite to be an effect on the mind. The 1915 (402) translation iseven worse in this regard: “Thus comes about that men attribute creative virtues to theirgestures, which in themselves are vain. The moral efficacy, which is imaginary; that ofthe whole, to the belief in that of each part by itself.” Here the sentence is very confusedand moral efficacy is presented as imaginary, when it is Durkheim’s point that only themoral effects are real.


can be immediately aware. The rituals are their own effect. According toDurkheim ([1912:529]*1915:414;1995:374) “it is because they serve toremake individuals and groups morally that they are believed to have apower over things.” Once the moral efficacy of the rite had been experi-enced, the belief in this power was extended to a belief in a power overthings, which the rites do not have.

According to Durkheim ([1912:513] 1915:402;*1995:364) “The gen-uinely useful effects brought about by the ceremony, as a whole, are tan-tamount to an experimental justification of the elementary practices thatcomprise it.” Other practices could be substituted. But the combinationof shared practice and performance by the assembled group does have areal result that justifies a belief in the efficacy of the rite.

Because it is the totemic idea which binds them together, and whichthey all have in their minds, it is quite natural and necessary that there berepresentations of the totem accompanying the ritual. But, that should notmislead us into thinking that the idea that these symbols can cause repro-duction comes from their simple resemblance to totemic animals. Thatidea results from the fact that what the symbol represents is the actualityof the members of the totem constituting and reproducing themselves asa moral community.

Because the community only exists in so far as the totems it representsand enacts are believed in, the rituals which make and reinforce beliefin the causal efficacy of the totem also quite literally make and remakethe moral community Durkheim says ([1912:511]*1915:400;1995:362):“The rite does not limit itself to expressing this kinship; it makes it orremakes it.” Members of the totem who participate in the ritual have beenreproduced as members of a moral community. According to Durkheim([1912:513]*1915:402;1995:364), it is the resulting feelings of well beingand moral unity that make participants believe the rite has succeeded.

Without this reproduction, the totem would cease to exist, because itis the enactment of the totemic ritual which causes the totem to exist as atotem, and its continued enactment is necessary to maintain the existenceof the totemic beliefs. By enacting their belief in the causal efficacy ofthe totemic symbol; displaying the totemic symbol to one another ontheir bodies; and, by acting like the totemic animal during the ritual,members of the totem not only reaffirm, but actually “remake” the kinshipgroup in and through the ritual. According to Durkheim ([1912:513]*1915:402;1995:364):

it also exercises a profound influence over the souls of the worshipers who takepart in it. They take away with them a feeling of well-being whose causes theycannot clearly see, but which is well founded. They feel that the ceremony is good

Imitative Rites 225

for them; and, as a matter of fact, they reforge their moral nature in it. How couldthis sort of well-being fail to give them a feeling that the rite has succeeded, thatit has been what it set out to be and that it has attained the ends at which it wasaimed? As the only end which was consciously sought was the reproduction ofthe totemic species, this seems to be assured by the means employed, the efficacyof which is thus proven.

The feeling of participants that the rite has been successful is not merelya personal feeling. Durkheim argues that it is a general feeling, shared withothers, which has a general source.5 Through the ritual the participantsand members of the totem become stronger in their totemic feelings forone another. Their belief in the totem is strengthened. Therefore, thetotemic species is reproduced. The beliefs and feelings of the individualmembers are an essential ingredient in the communal life. If they donot feel a part of the totem then it ceases to exist. It has no existenceindependently of being believed in.

Durkheim argues that once this social utility is taken into account theactions of primitive peoples in performing mimetic rites can be seen asrational, pursuing an immediate end, and not as occurring merely becausethere is a traditional belief.6 The traditional belief exists only to justifyand ensure the performance of a necessary and efficacious rite. Accordingto Durkheim, this is no different from confirming a scientific experiment([1912:516] 1915:404;*1995:365): “To begin with, the moral efficacy ofthe ceremony is real and directly felt by all who take part; therein is aconstantly repeated experience whose import no contradictory experi-ence can weaken.”

Furthermore, contrary to the claims of the empiricists, Durkheim([1912:516] 1915:404;*1995:365) argues that the physical results thatare sought after usually do occur: “It is in fact normal for the totemicspecies to reproduce itself regularly.” So, the empirical facts in generalappear to support the beliefs in this regard. While the empirical facts ofnature cannot explain the origin of the belief (only social facts explainthat) they do serve to confirm the beliefs. The physical species wouldreproduce perfectly well if the rite were not performed (although thetotemic species would not). But, that fact of their reproduction confirms,rather than contradicts the continuation of the rite. “Not surprisingly,”

5 See 8.1.2 for a discussion of the arguments of Hume and Durkheim regarding the statusof “feelings.”

6 This view of traditional action has implications for the distinction made by Weber,Parsons, and Habermas between traditional and rational action. Durkheim’s distinctionbetween mechanical and organic solidarity is not based on a distinction between twosorts of reason. If traditional action is all functionally rational then individual instrumen-tal action is less functionally rational. The idea that traditional action is meaningful butnot rational fails.


Durkheim ([1912:516] 1915:404;*1995:365) says, “since the rites, espe-cially the periodic ones, demand only that nature take its regular course,it seems most often to obey them.”

It becomes clear in this section why Durkheim separated his discussionof beliefs from rites. There is nothing in nature that could lead to the beliefthat if you didn’t do these rites nature would not continue on its course.But, if the feelings created by the rites generate these beliefs, it is alsothe case that there is nothing in the beliefs that contradicts the naturalcourse of things. Therefore, the beliefs will receive constant empiricalaffirmation. As a retrospective justification for action, that is efficaciousin its own right, the beliefs are perfectly suited. If, however, the beliefsare taken to have preceded the rites, and one attempts to explain thebeliefs independently of the rites, Durkheim argues that absurdities areinevitable.

According to Durkheim ([1912:516] 1915:404;*1995:365) this is whatmakes faith “impervious to experience.” When people stop believing,but still continue to perform rites, Durkheim ([1912:513] 1915:402;*1995:364) says that this proves his point about their utility: “The truejustification of religious practices is not in the apparent ends they pursuebut in their invisible influence over consciousnesses and in their mannerof affecting our states of mind.” Preachers who set out to convert peo-ple, he says, do not try to convince them of what they should believe,but rather, try to convert them through participation in rites. This isparticularly evident today in the conversion practices adopted by cultgroups, which focus first on getting the person to participate. Durkheim([1912:514] 1915:402;*1995:364) says that, those who seek to convert,concentrate on “awakening or reawakening the sense of moral supportthat regular celebration of the cult provides.” This is the origin of “faith,”according to Durkheim. It leads one to believe in advance of, and evenin contradiction to, proof.

7.2.6 Specific Ritual Gestures Have No Inherent or Magical Efficacy

The reason that beliefs attach to specific practices, is not because the prac-tices themselves have any specific utility in achieving natural outcomes.The belief that specific gestures and words have a causal efficacy is aresult of the social utility of the rite. According to Durkheim ([1912:513]1915:402;*1995:364), “if value is attached to these various manipula-tions, it is not because of value intrinsic to them but because they arepart of a complex rite whose overall utility is felt.” This again contrastsDurkheim’s position with Weber’s. Weber considers mimetic rites to bevalue-rational or traditional: done purely because they are believed in.Durkheim says that their effects show that they have a utility. The thing

Imitative Rites 227

is that they have collective utility, which is quite different from individualinstrumental utility in Weber’s terms. The rites have no utility for individ-uals, except insofar as the individuals cannot exist without the group. Inthis sense of collective utility, most of Weber’s traditional actions wouldhave utility. But, the creation of collective utility also ultimately has utilityfor the individual social being, because without the performance of therites, individuals would not have reason, and would not be recognizablyhuman.

In spite of the importance of the performance of the ritual, however,the specific physical details of the rituals are, in an important sense, irrel-evant. What is essential is that all members of the group make the samemovements and sounds, and that the rite is performed the same way eachtime. But, almost any sounds and movements in unison will serve thepurpose. Because the physical details of these rituals are irrelevant totheir success, Durkheim ([1912:515]*1915:403;1915:365) argues thatrational criticism of those physical details is also irrelevant. The resultsof totemic rituals are social, and psychical, not physical, and therefore,cannot be meaningfully criticized in terms of a relation between theirphysical details and natural physical results. The real purpose of suchrituals, Durkheim ([1912:515]*1915:403; 1995:365) argues is the repro-duction of the moral community, not the representation of physical, orsupernatural states of affairs.

Durkheim argues that the specific gestures which are enacted in theritual are irrelevant and could be replaced by others. They have causalefficacy, not because of magical or scientific properties, with which theyact on objects or animals, but rather, because, as representations of theshared totemic symbol of the group, they can produce feelings of moralunity in its members, which strengthen the group and hence the totem.According to Durkheim ([1912:515]*1915:403;1995:365): “It is becausethe true justification of religious practices does not lie in the apparentends which they pursue, but rather in the invisible action which theyexercise over the mind and in the way in which they effect our mentalstatus.”

This is a very different kind of causal relation from a physical, or naturalcausal relation, and Durkheim carefully contrasts the two. The physicalgestures have no natural efficacy, their efficacy is purely social (Durkheim,[1912:514]*1915:402;1995:364). But, as a collective social rite the ges-tures do have a real utility. If Durkheim’s argument hinged on relationsbetween natural events, it would be subject to Hume’s skeptical con-clusion, because relations between natural events remain unknowable incausal terms. But, Durkheim sets up a causal relation in which feelingsare also immediately causes. This is a dynamic relation that can be imme-diately felt, not a natural relation between two things, requiring inference.


The empiricist perspective taken by Frazier and Taylor (Durkheim,[1912:510]*1915:399;1995:361), treats the point of the ritual as the lit-eral reproduction of the actual animal species which the totem represents.Given this perspective, Durkheim argues, it will appear that the membersof the totem have made a gross error, because the totemic symbol cannotcause the actual reproduction of biological animals. If the rite is inter-preted as having successfully reproduced the totemic species only if itcauses the biological animal to reproduce, then the belief in the causalefficacy of the ceremony seems so absurd that it is hard to figure out howanyone could make it.

If, however, the rite is, as Durkheim says, not aimed at reproduc-ing the biological species, but rather at reproducing the moral energyof the group, the rite really is causally efficacious. If the rite reproducesthe moral unity of the totemic group, then the rite really does causethe reproduction of the species. It is the effect on the men themselves;their feelings of well being and moral unity, that give them the feelingthe rite has succeeded, not any presumed magical action of the rite onnatural or animal objects.

The causal relation which Durkheim speaks of is a moral one whichhe says operates on our “internal” states. Because the result is in ourconsciousness we can perceive directly the action of the rite on our moralstate.

The contrast with Evans-Pritchard on this point is significant. Evans-Pritchard argued that certain belief systems were impervious to empiricalevidence, or proof, because of the circularity of their logic. He contrastedthese belief systems with science. But, from Durkheim’s perspective, theseare in actuality belief systems that have grown up to justify practices.Neither scientists, nor primitive peoples, use their theories, or beliefs, toaccomplish their practices. Both have well developed practices, or rites,which the beliefs seek to justify and explain. Evans-Pritchard has, fromDurkheim’s perspective made the mistake of treating theories as rulesor recipes. But, they do not function as rules for how to act, only asretrospective justifications or accounts for practices that have their ownprospective order. The difference between scientific belief and religion isthat science is after “truth” whereas religious beliefs have as their objectiveto get people to enact practices that create necessary feelings. As long asthe feelings are achieved, the beliefs are producing the desired empiricalresult. The truth of things, in such cases, can be found only in a study ofthe practices, and not in the beliefs.7

7 But, the beliefs and accounts can be important as indicators of relevant practices. Studiesof accounts, for instance, provide important clues about practices.

Imitative Rites 229

Because the beliefs are not really important in their own right, did notcome first, and really only serve to provide the reasons for the practices(which are causally efficacious in their own right), the beliefs can beanything that will the support the practices. Therefore, it would makesense that the most successful beliefs, with regard to religious practicewhich needs to remain stable over time, would be ones flexible enoughto include most natural occurrences as confirmation of themselves. Thiswould give circular belief systems particular utility. It is not, however, asEvans-Pritchard thought, that these beliefs have no empirical foundation.Or, that they employ circular logic. It is just that the empirical foundationthat they do have is not related to the system of beliefs in the way thatis usually assumed. It is not that they are unrelated. But, that they arerelated only as a secondary system of justifications; what C. Wright Mills(1940) called a “vocabulary of motives.” The beliefs do not try to explain,what Durkheim refers to as, the “real” empirical causes and effects ofthe rites.8 They serve the social purpose of justifying action, and gettingpeople to perform the rites. Scientific beliefs and practices, on the otherhand, should leave room for counter-examples, because their purpose isnot only to maintain the stability of scientific practices, but also to allowfor change, to aid in the pursuit of “truth.”

8 This would explain why racist beliefs, which exhibit circular logic, are so imperviousto empirical demonstration. But, it would also suggest that those beliefs could only bechanged by changing the underlying practices which they serve to justify. As long asthe practices remain intact there will be a purpose for the beliefs. This is the reverse of theway racism and prejudice are usually approached. Typically, attempts are made to changebeliefs and create a “sensitivity” of perspective. Durkheim’s argument would suggest thatthis effort is misplaced and that an emphasis on changing underlying practices would bein order.

8 The Category of Causality

While Durkheim has discussed creation as an inherent property of cer-tain rites in the first few chapters of Book III, his argument for the socialorigin of the category of causality does not begin until Book III, ChapterThree, Sections iii and iv. Because of the centrality of the concept ofcausality to epistemology, Durkheim’s argument for the social originof the concept of causality is the centerpiece of The Elementary Forms.The earlier sections of the book, particularly the extensive sections ontotems and moral force, lay the groundwork for Durkheim’s argumentwith regard to the category of causality. Causality is an essential conceptin philosophy and played a pivotal role in the debate between empiri-cism and apriorism. Philosophers on both sides of the debate agreed thatwithout causality, science would have no foundation. Indeed, knowledgeitself, which Hume argued consists of inferences from effects to causes,would be impossible without the concept of causality.

Following the formulation of the problem by Hume and Kant, philoso-phers in Durkheim’s day adopted several strategies toward explaining theorigin of the idea of causality. Some agreed with Kant that the idea ofcausality was innate. Others argued that it was only an instinct, or asHume had argued, an opinion. These arguments left the idea of causalitywithout any empirical validity. Pragmatism stepped into the breach, inthe decade before Durkheim wrote The Elementary Forms, and offeredthe possibility that while the concept of causality might not have empir-ical validity, it might, nevertheless, have sufficient experiential basis toground a sort of pragmatic validity, or truth by consensus.

This is the dilemma that confronted Durkheim in 1912, and, at leastin part, because of the neglect of Durkheim’s argument, remains essen-tially the position in which epistemology finds itself at present. In fact,the current consensus is that epistemological questions in their classi-cal form make no sense, and the field of inquiry is currently known as“justified belief” and not epistemology.1 But, the question, as addressed

1 See the 1998 reply by Warren Schmaus to my 1996 AJS article, and my response also inAJS 1998.

230

The Category of Causality 231

by Durkheim, retained its classical epistemological form. It remained asearch for the empirical origin of the categories. Durkheim argued thatthe classical dilemma was the result of taking an individualist position.The dilemma would disappear, he argued, given a sociological approach.If Durkheim were able to secure a basis for the concept of causalityin direct experience of social processes he would achieve something ofgreat importance, with serious implications for classical and contem-porary methodological and theoretical debates, in both sociology andphilosophy.

In spite of the importance of the argument for causality to the book asa whole, however, with all other discussions leading up to and buildingfrom it, most discussions of The Elementary Forms completely overlook it.Certainly, the fact that most scholars do not realize that the book consti-tutes an epistemological argument is part of the reason why Durkheim’sargument for causality, only important really as part of an epistemolog-ical argument, is so consistently overlooked. However, the form of theargument is also responsible. The way the book is written, the argumentfor causality can easily appear to be an afterthought. The argument doesnot even appear in the body of the text until just prior to the conclu-sion ([1912]508, 1915:410, 1995:360), and then only covers two shortsections, between 13 and 20 pages in length, depending on the edition([1912]528, 1915:413, 1995:373). It is easy to overlook the centralityof these pages to the overall argument of the book. Following these twosections, the argument for causality quickly gives way to a discussion of“feelings,” or emotions, related to particular forms of rites discussed inChapters Four and Five. Because of the role played by emotions in thecreation of the categories, this discussion is intimately connected to theargument for causality. But, unfortunately, Durkheim uses the terms effi-cacy, goal, and purpose, and not causality in the latter two discussions,and the connection to the epistemological argument is consequently notas clear as it might be. Furthermore, the fact that the discussion of causal-ity follows more than 250 pages on totems, and is intertwined with furtherdiscussions of totemic rites of sacrifice, oblation, and imitative rites, alsocontributes to its being overlooked.

It is not until Chapter Three of Book III ([1912:501–528] 1915:393–413;*1995:355–373) that Durkheim really gets to the argument for theempirical origins of the category of causality. None of the earlier chap-ters indicate that they are building toward this argument. Even in BookIII, Chapter Three, epistemology is not mentioned until the third sec-tion. The first two sections are comprised of descriptions of “movementsand cries” that are designed through imitation to cause the reproduc-tion of the totemic species. These Durkheim refers to as “mimetic” rites.


There is an essential relationship between these rites and the category ofcausality. The relationship between these rites and Durkheim’s epistemo-logical argument, however, is not mentioned until seventeen pages intothe chapter in the 1995 translation ([1912:518]1915:405;*1995:367).

When Durkheim does get to Section iii of Chapter Three, however, hequickly makes an explicit connection between the rites being describedin the first two sections of the chapter (and in fact all the rites that havebeen described in the book) and the epistemological argument outlinedin the Introduction. Durkheim ([1912:518] 1915:405;*1995:367) says:“The principle just explained does not have a merely ritual function: itis of direct interest to the theory of knowledge. In effect, it is a concretestatement of the law of causality and, in all likelihood, one of the earlieststatements of it ever to have existed.”

Durkheim then proceeds to work systematically through the argumentthat the category of causality comes from the experience of moral forcethat is created in and through the performance of mimetic rites. In thisregard he clearly distinguishes the idea of causality, which he says may bean instinct, from the category of causality, which must have an empiricalorigin in the emotional experience of the performance. For Durkheim, allrites described in the book have at least part of their origin and explanationin the fact that the category of causality can only come into being throughthe performance of certain social processes. Durkheim ([1912:554]1915:432;*1995:391) says later, in Book III, Chapter Four, that: “Inall likelihood, the other rites we have studied are no more than variationson this fundamental rite.”

In other words, everything that Durkheim has discussed in The Elemen-tary Forms up until this point, begins with these rites, and the principle ofcausality. Durkheim seems to be arguing, in Chapters Three and Four,that these rites also caused the idea of the sacred, which is the primaryidea behind religion that he set out to explain. This is an illustration ofhow Durkheim works systematically, but in a reverse fashion, toward hisgoal. What he takes first is the thing that must be explained: the sacred.The quest of the whole book is a quest for the origin of the sacred. Whathe gets to last is the explanation: causality.

The discussion of causality as an instinct is of critical importance inthis regard, and I believe has confused the issue enormously. Many criticsassume that Durkheim means that the category of causality is an instinctand therefore, that his position is rationalist or innatist, or even Kantian.They also accuse him of contradicting himself by saying at one point thatcause is innate and at another that it has empirical origins. But, this isnot at all Durkheim’s argument. Durkheim introduces causality as aninstinct in order to address a practical question. His theory depends onthe assertion that rites came before and created beliefs.


Specifically, Durkheim argues that mimetic rites caused the idea of thesacred. This means that there would be no idea of the sacred, no moralforce, and hence no human reason unless, and until, mimetic rites wereperformed. But, there is a problem. It is impossible to explain why theywould have been performed in the first place if there were no idea of causalefficacy. How could people try to employ the idea of “like produces like”if they didn’t have the idea in the first place. In this discussion of instinct,and in his discussion of dualism, Durkheim provides a solution to thisdilemma. There is a basic animal idea of, or instinct about, causality.This is not the category of causality, and is pre-rational. If a person hadthe basic instinct that “like produces like,” they would have a reasonto perform these rituals before any idea of the sacred, or the categoryof causality itself, had developed. This is not true of the other rituals,which are based on the idea of the sacred and make sense only afterthe first experience of moral force creates a basic belief in the sacred,or moral force. But, mimetic rites create the idea of the sacred, so theycannot depend upon it. Durkheim is offering these rites toward the endof the book as an answer to his opening question with regard to the originof the sacred. The mimetic rites alone require a pre-social underlyingexplanation and cause, and Durkheim’s argument that there is a sense ofcausality that exists as an underlying instinct provides it.

The category of causality is something else entirely. By performing therite, the participants create a moral union or association, and thus createthe first feelings of moral force. Thus, working systematically from prob-lem formulation, step by step toward a solution, Durkheim has: First,located the idea of the sacred as the essential idea to explain, and; Second,located the feeling of moral force as essential to the creation of that idea(along the way dealing with a number of embedded issues such as individ-ualism, dualism, emblem, magic etc.); and, Finally, located the enactmentof causality through imitative rites as the origin of the feelings of moralforce and therefore, as the origin of the idea of the sacred. According toDurkheim, this means ironically that the rite that is believed to be a causalrite, the initial belief for which is based only on instinct, actually enactsa real creation of a sacred thing, and in so doing the idea of the sacred,the idea of the efficacy of the rite, and therefore, the feeling of a moralobligation to perform it again, which add up to the category of causality,are all created by the enactment of the rite. While the initial performanceof the rite would not have taken place without some instinctive belief inits efficacy, the category of causality that results from the rite is a realcreation, whether the participants believe in it or not. In other words, thepre-rational individual belief is not the origin, or cause, of the category.The instinct only explains why they would perform the rite. The result,which happens to be the category of causality in this case, is completely


independent from whatever beliefs are behind the rite. It is the rite thatis causally efficacious, not the beliefs.2

Why Durkheim doesn’t start out Book III, or at the very least open thechapter that explains the origins of the category of causality, with somedirections as to where the argument is going, is something of a mystery.But, it is consistent with his general style of argumentation, in whichquestions lead step by step to surprising and undisclosed conclusions.He does start The Elementary Forms with a sketch of his epistemologicalargument, in the Introduction. But, even there he does not say anythingabout the relationship between the various chapters and the overall argu-ment. Nor does he say that his argument will essentially come to rest onemotions created by the performance of totemic rites. The Introductiondoes say that religion will be examined first in order to get to a point aboutepistemology. But, it isn’t clear that Durkheim is describing how the bookitself will be organized, working its way slowly and carefully toward itsmain point, which comes only in the later chapters of Book III.

8.1.0 Book III, Chapter Three, Section iii

In the first two sections of the chapter Durkheim has dealt with causalityas an idea presupposed by mimetic rites, but not as an essential categoryin an epistemology. However, Durkheim opens Chapter Three, Section iiiwith an explicit statement regarding the relationship between the princi-ple of causality, evident in mimetic rites, and the “theory of knowledge.”Durkheim ([1912:518]1915:405;*1995:367) argues that “the principlejust explained,” a reference to “like produces like,” as discussed in Sec-tion ii, “does not have a merely ritual function; it is of direct interest tothe theory of knowledge.” Durkheim ([1912:518] 1915:405;*1995:367)argues that “In effect, it is a concrete statement of the law of causality and,in all likelihood, one of the earliest statements of it ever to have existed.A full-fledged notion of the causal relation is implied in the power thusattributed to ‘like produces like.’” Here Durkheim refers, finally, back to

2 Actually, even though their positions on the sociology of religion conflict, this argumentis compatible with Weber. Because it is not the idea of the Protestant ethic, but ratherthe ensuing behavior that is causally efficacious in creating capitalism. If the middle classdeveloped this religion in order to justify behaviors in which they were already engaged,which seems obvious, then the Protestant ethic is also a secondary belief system, developedin order to justify a preexisting lifestyle. It is the lifestyle itself that is causally efficaciousand creates the beliefs. The beliefs then play a significant role in enhancing certain aspectsof the lifestyle (rites). So, while Weber appears to place his emphasis on beliefs, it couldbe argued that it is the practices that produce those beliefs and then the beliefs accentuateparticular aspects of the practices and in turn transform them. So, they are not really veryfar apart.


the argument that he made in the Introduction, that religion will furnishthe origins of the essential categories of the understanding.

Because this is the only section in the book where Durkheim connectsan idea generated by religion directly to his epistemological argument,while at the same time calling it a “category,” it is of critical importance forthe understanding of his epistemological argument. In this short sectionDurkheim reviews his earlier arguments against empiricism and aprior-ism, his criticisms of Animism and Naturism, and reiterates his claim thathe has discovered a social origin for “the categories of the understanding.”Durkheim builds his argument in this section, with regard to causality,on his prior analysis of the concept of force. He refers to force and moralforce constantly throughout the argument. This is essential because in hisanalysis of force, Durkheim argued that the idea of force has its origin inthe emotions, or feelings, generated by the collective enactment of ritualpractices. Here, in his discussion of causality, he does not place the sameemphasis on causality having an origin in feelings, although he does sayit more than once. Instead, he concentrates on demonstrating that causeis part of the idea of force, and therefore, has the same social origin. Theimportance of feelings is not as evident in this discussion of causalityas it was in the earlier discussion of force. From Durkheim’s perspective,the argument for the origin of force in collective feelings has already beenmade. He argues that causality is only a special kind of force. That causealso has its origin in collective feelings, follows from the fact that the ideaof cause is a type of force.

Unfortunately, there are two things that obscure Durkheim’s argumentat this point. First, the lack of a more direct tie between causality andfeelings leaves this section still reading like a sociology of knowledge atpoints; and, Second, there are at least four passages where the translationcreates the mistaken impression that Durkheim assumes an underlyingreason that precedes collective practices. If the connection from causalitythrough force to emotions is kept in view, the impression that Durkheimis articulating a sociology of knowledge can be avoided. The translationissues will be discussed as they come up in the text.

Because mimetic rites embody one of the earliest statements of thecausal principle, and because Durkheim believes that the causal princi-ple has a social origin in the performance of these rites, a study of mimeticrites can, according to Durkheim ([1912:518]1915:405;*1995:367),reveal the origins of the principle of causality: “Thus, the origins ofthe precept on which mimetic rites rest can explain how the princi-ple of causality originated.” Durkheim distinguishes here between the“precept” upon which mimetic rites rest and the “principle” of causal-ity. Durkheim argues that the idea of causality generated by religious


ritual has the same origin as the category of causality in logic and science.But, he will continue throughout this section to maintain a distinctionbetween the idea of causality based on belief, instinct, or habit, and thecategory of causality that is generated by the experience of moral force.“The first” idea, the religious idea, he ([1912:518]1915:405;*1995:367,italics in original) says “arises from social causes. It has been fashioned bygroups with collective ends in view, and collective feelings express it. Thuswe may presume that the same is true of the second.”

There is a translation issue at this point. The 1995 (367) transla-tion reads: “collective feelings express it.” This suggests that it (the ideaof causality) exists already to be expressed. The 1915 (406) transla-tion reads: “It was elaborated by groups having collective ends in view,and it translates collective sentiments.” This is better in having the ideatranslate sentiments, but it still suggests, although in a more subtle way,that the idea exists already and only “translates” the rites. The originalFrench ([1912]:518) puts the sequence in a different order: “et ce sont dessentiments collectifs qu’il traduit,” is better translated as “It was the collectivesentiments that were translated by [the precept of causality]” reinforcingthe point that force, or causality, comes from the sentiments and not thereverse. It is not “that collective feelings express” an underlying precept,as the 1995 translation suggests, but rather that force is a collective feel-ing that the precept “like produces like” expresses. In the original Frenchthe direction of the action is reversed. The sentiments are given as theorigin of the concept, instead of just expressing or translating it. Giventhe prevalence of the misinterpretation of Durkheim as having assumedapriori categories, or having said that ideas have their origin in otherideas, the fact that in the French the sentiments come first, not the idea,is extremely important. Durkheim says that the idea of cause comes fromcollective sentiments, not that collective sentiments come from preexist-ing ideas. The translations conflict with Durkheim’s central argument inthe text: that the idea is created by the emotional experience of the moralforce of the rites.

8.1.1 Analyzing the Principle

Durkheim’s first step in the argument that the category of causality hassocial origins is to analyze the principle of causality itself. Durkheim([1912:518]1915:405;*1995:367) argues that “to verify whether this isindeed the origin of the elements from which the principle of causality ismade, it is enough to analyze the principle itself.” Analysis of the conceptis sufficient at this point because the empirical demonstration has alreadybeen given in the arguments with regard to the feeling of force. The origin


of the idea of force, in emotions generated by the collective performanceof the rites, furnishes the empirical origin of causality. Durkheim hasonly to demonstrate that the idea of cause is part of the idea of force.The analysis of causality will then confirm what was argued already withregard to force.

Durkheim ([1912:518]1915:405;*1995:367) proceeds to establish thispoint: “First and foremost, the idea of causal relation implies efficacy,effective power, or active force.”3 Durkheim does an analysis of the ideasimplied by, or contained within, the notion of the causal relation. Heanalyzes these implied ideas, which contain the ideas of force and efficacy,because his argument for causality rests on his argument with regard toforce.

Cause is a special aspect of force. According to Durkheim ([1912:518]1915:405;*1995:367), “Cause is force before it has manifested the powerthat is in it.” Durkheim ([1912:518]1915:405;*1995:367) goes on to say:“Effect is the same power, but actualized.” This is an important sec-tion because the claim is made that the idea of force, which Durkheimhas already demonstrated as having an origin in feelings generated bytotemic rites, contains the ideas of both cause and effect. If “cause isforce before it has manifested the power that is in it,” and effect is “thesame power but actualized,” then force is both cause and effect, con-nected in a dynamic relation, after cause has manifested the power that isin it. Force is the power that causality actualizes. Force is what connectscause and effect. By connecting feelings with the idea of force, Durkheimhas explained the origin of causality, because causality is already con-tained in the idea of force. In an interesting way he is arguing thatmoral force equals causality, although force is also more than causal-ity, which is only a special kind of force. All Durkheim has left to do isanalyze the concept and show that cause does have its origins in the ideaof force. Durkheim’s ([1912:518]1915:405;*1995:367) identification ofcause with force explains how it is possible that “humanity has alwaysimagined causality in dynamic terms.”

3 There is only a small point of translation difference here, but it is worth mentioning.The 1915 translation (1915:406) reads: “The first thing which is implied in the notion ofthe causal relation is the idea of efficacy, of productive power, of active force.” There isnot much difference between these translations. But, the two published ones give a senseof “fact” or definition, that is missing from the French. ([1912]:519) “Ce qui est toutd’abord implique dans la notion de relation causale, c’est l’idee d’efficacite, de pouvoirproducteur, de force active.” My own translation ([1912]:519) “That which is before allelse implied in the notion of the causal relation, is the idea of efficacy, of productive power,of active force.” Since the argument that cause is implied in the idea of force is so criticalto his argument the feel of the sentence is important. The 1915 translation suggests thatforce is only the first thing among others. The 1995 translation is better in this regard,but still uses the word “first.”


Durkheim’s ([1912:518]1915:405;*1995:367) next move is to alludeto the philosophical argument: “To be sure, some philosophers deny thisconception any objective basis; they see it only as an arbitrary constructof imagination that relates to nothing in things.” But, while Durkheimwill take up a debate with empiricism on the next page, he does not doso in order to establish that the concept of causality has a basis in real-ity. At this stage in the argument “reality” is not the issue Durkheim ispursuing. He is more interested in the question of origin. The analysistherefore, should be empirical, not logical. It concerns empirical origins,not logic. Durkheim ([1912:518]1915:405;*1995:367) says: “For themoment, however, we do not have to ask ourselves whether it [causality]has a basis in reality; noticing that it exists and that it constitutes, andhas always constituted, an element of ordinary thought (as acknowledgedeven by those who criticize it) is enough. Our immediate purpose is tofind out not what causality amounts to logically but what accounts forit.”

There is a translation issue with regard to this passage that needs to bediscussed in some detail. In both English translations the words give theappearance that Durkheim is saying that the concept of causality “always”constituted an element of “ordinary mentality” prior to social action ofany sort. This reading of the text is inconsistent with Durkheim’s claim, inthe very same lines, that causality has a social origin, and that his interestis in the empirical social origin of the concept, not its logic.4

There is an ambiguity in the French at this point that may havecaused some translation difficulties. Durkheim uses the French word“commune” where the translators have used the word “ordinary” in theEnglish. The word commun in French can mean “ordinary.” The fem-inine case of the word is commune. In this sentence the feminine caseis called for. However, the word “commune” in French, spelled withan “e” can also be a different word, from commun, carrying a sense ofcommunal, or shared living, or thinking. The question is whether or notDurkheim intended the feminine of commun, or the word for collec-tive, commune. It is my sense that Durkheim did not mean to say thatcausality had always been an element of ordinary thought. He meantthat it had always been an element of common, or communal thought.(The newest abridged translation Cosman (2001) does say “communalmentality” p. 271).

The substitution of the word “ordinary,” for commune, with regard tohuman mental equipment in this sentence, supports the interpretation of

4 “Il nais suffit de constater qu’elle existe, qu’elle constitue et qu’elle a toujours constitueun element de la mentalite commune; et c’est ce que reconnaissent ceux-la memes qui lacritiquent.” ([1912]518).


Durkheim as a rationalist, who believed in underlying innate categoriesof the understanding. It sets up a claim that contradicts Durkheim’s ownclearly stated argument that causality has a social origin. If, instead oftranslating the word as “ordinary,” its collective sense is preserved, thenDurkheim can be seen essentially to be arguing that the idea of causalityhas always constituted an element of “shared thought.” That is, from themoment that thought became shared, it had this quality. Durkheim’s dual-ism allows him to say that before ideas were shared thought was different.He follows Rousseau in this regard. So, this statement that causality hasalways been an element of shared thought, after the transforming effectsof sharing, or moral force, have been felt, does not in any way suggestthat this quality of thought is inherent in the human mind.

This point is even more important because Durkheim has used theword “always” in making this claim. Durkheim argued that the categoriesof the understanding were universal, and must appear wherever soci-ety survives. Such categorical claims have frequently been interpreted asindicating an inherent rationalism underlying his social origins argument.When coupled with the translation problem in this section the potentialproblems with the interpretation of the word “always” are multiplied.But, what Durkheim argues is that it is not possible to have shared ideaswithout the experience of force, and cause is a part of force. Therefore,from the first moment ideas were shared, which would follow the firstexperience of force, people would necessarily and always, have the ideaof cause, which is part of the idea of force.

In fact, Durkheim may be “playing” with the double meaning of com-mun, commune and communement in French in several places. What iscommon, is also communal, but never ordinary. He would argue, how-ever, that it is common because it is communal, not communal becauseit is common. There are other French words for ordinary that Durkheimcould have used that would have avoided this ambiguity. These otherwords are “better” for conveying the sense of ordinary, if ordinary wereindeed what he meant. He has already used the word communementearlier in the paragraph to denote a general or shared knowledge withregard to what is implied by causality (that was also translated “usually”and “ordinarily”). It is much more consistent with his own argument forthe word to have communal overtones in both cases. It makes no sensefor Durkheim to have argued both that causality has social origins, andthat it has always existed in the ordinary mentality, in the same sentence.

Durkheim’s next move is to reiterate his argument that society, in theguise of moral force, is the first power that men represented to them-selves as a power. If cause comes from force, and force has a social ori-gin, then cause has a social origin. Durkheim reviews the various forms


in which moral force has appeared in the preceding pages. According toDurkheim ([1912:518]1915:405;*1995:367): “The analysis of the evi-dence has already permitted us to show that, in prototype, the idea offorce was mana, wakan, the totemic principle-various names given tocollective force, objectified and projected into things.” (The translationomitted a comma which has been added. See footnote for discussion).The prototype of the idea of force was collective force, called by variousnames, objectified and projected into things through the collective expe-rience of moral force. If collective force is the prototype of the idea offorce then it is also the prototype of the idea of causality.5

The omission of the comma from the translation, while minor, leadsto a further confusion with regard to the transposition of “imagine” and“think” for “represent” in the sentence immediately following. ThereDurkheim ([1912:518]1915:405;*1995:367) makes the claim that thefirst powers which persons represented as such were social forces. Thetranslation reads: “So the first power that men imagined as such doesindeed appear to have been that which society exerts upon its members.”The point of this statement is to remind the reader that these social man-ifestations of force have been established, in earlier chapters, as the firstpower that persons experienced and created representations of.

The problem with the passage in English translation is that the words“thought of as such” and “imagined as such” are used, where in theFrench Durkheim wrote “represente comme tel” (represent as such).6

5 There is a small translation problem that contributes some ambiguity at this point. Inthe French after Durkheim gives a list of the prototypes of the idea of force, “nomsdivers donnes a la force collective” (names diverse given to the force collective) there is acomma. After the comma he writes: “objectivee et projetee dans les choses” (objectifiedand projected into things). The clause that follows the comma is meant to refer back toeverything on the list, not just to “collective force.” The clause after the comma modifies“force” and Durkheim has given a list of forces. In the French there is a comma indicatingthis. But, in both English translations the comma has been omitted. The impression iscreated that the objectified forces have been given diverse names.

But, Durkheim’s point is that forces are objectified and projected into things. It is theexperience of this that creates the categories. In the English translations the impressionis created that force, in the guise of mana and wakan, stands in a different relation toreality than the various names given to collective forces that are “objectified and projectedinto things.” But what Durkheim actually says is that mana, wakan, l’orenda, totemicprinciples, which are diverse names for collective forces, are [all] objectified and projectedinto things. Mana and wakan are themselves only real as felt, enacted, projected, andrepresented. Treating mana and wakan as real, and “diverse names” as ideal, misses thepoint that mana and wakan are themselves names, and are actualized through symbolicand mimetic ritual behavior. But, they are names that have real effects.

6 The entire sentence in French ([1912]:519) reads: “Le premier pouvoir que les hommesse sont represente comme tel semble donc bien avoir ete celui que la societe exerce sur sesmembres” (The first power that men did represent as such seems to have been that whichthe society exercises over its members).


Durkheim’s argument depends heavily on his distinction between repre-sentation, or collective feeling, and thought, or idea. The English translationscompletely reverse this point, substituting “imagined” and “thought of ”for represent. By, “represent as such” Durkheim could mean that themembers of the collective symbolically represent the collective feeling tothemselves, or collectively represent the collective feeling to themselvesthrough an external symbol, or that they collectively represent the feelingthrough the internal experience of an external form, i.e., mimetic rites.Imagine and think, on the other hand, are terms that refer to the individualmind. Thinking and imagining are done by, and belong to, individuals,not collectivities. Durkheim uses “representation” to refer to somethingsocial, as opposed to something individual.

Durkheim then turns to a consideration of the philosophical argumentwith regard to causality. His debate with philosophy, in this section, willbe over whether the idea of causality could have come from externalexperience. But, he is not abandoning the idea that causality has originsin reality and he will return to the claim that a social origin is a “real”origin in the Conclusion.

8.1.2 Causality Could Not Come From External Experience

Durkheim’s second approach to a proof that causality has social origins,is to establish that the idea of causality could not have other than a socialorigin. He begins a review of the empiricist argument against the originof causality in sense experience. According to Durkheim ([1912:519–20]1915:406–7;*1995:368) “It is obvious at first glance, and recognizedby all, that external experience cannot possibly give us this idea. Thesenses show us only phenomena that coexist with or follow one another,but nothing they perceive can give us the idea of that constraining anddeterminative influence that is characteristic of what we call a poweror force.” The philosophers of empiricism, he says, have shown thatthe idea of causality could not have come from external experience, orsense perception. Sense perception consists only of disconnected ideas([1912:519–20]1915:406–7;*1995:368): “The senses take in only statesthat are realized, achieved, and external to one another, while the inter-nal process that binds these states together eludes the senses.” This is theheart of the empiricist argument: that disconnected perceptions must beconnected by the mind, because the connections between them are notavailable to perception. Causality is, for the empiricists, an idea that con-nects two or more perceptions. But, the senses cannot perceive the causalconnection in its own right, only the result in the form of an achievedexternal state.


Durkheim’s argument is that moral force can be experienced whole,so to speak, as a dynamic force, rather than as disconnected parts. This,however, would not be possible if force were perceived through the senses,but only if it constitutes a special form of social, or internal, experience.By accepting the empiricist argument that sense perception cannot be theorigin of a valid idea of causality, Durkheim strengthens his argument thatthe idea must have a social origin. He assumes that his reader is familiarwith classical empiricism, and therefore understands the empiricist argu-ment for why the idea of causality cannot have an origin in sense percep-tion. However, assuming that the sociological reader is no longer familiarwith this argument, and given its importance with regard to Durkheim’sepistemology, a short review of the argument is in order.

Hume, the author of the classic empiricist dilemma with regard tocausality, argued that all knowledge could be divided into two sorts: math-ematical knowledge and moral/empirical knowledge (Hume, 1777:25).While mathematical knowledge was for Hume (1777:26) a matter of purereason, because numbers were not real, but only logical constructions inthe mind, empirical or moral knowledge depended on the concepts ofcause and effect, and contiguity and resemblance. Hume (1777:26–7)argued that “By means of this relation [cause and effect] alone we can gobeyond the evidence of our memory and senses. . . . All our reasoningsconcerning fact are of the same nature. And here is it constantly sup-posed that there is a connection between the present fact and that whichis inferred from it.”

Hume placed causality at the center of human reasoning. Unless causalrelations could be established between perceptions, according to Hume,sense impressions would remain particular and unconnected. Every infer-ence, every connection between a present fact and a prior condition,requires the idea of causality. Therefore, it seems obvious that personsmake use of such an idea. The question for Hume, as for Durkheim,was not whether the concept exists, because even children and animalsmake use of it, but where it comes from and whether or not it has empir-ical validity. Hume argued that knowledge of cause and effect cannot beattained from sense perception of a single instance. Objects, according toHume (1777:27), do not contain and display their causes. He (1777:30)argued that “. . Every effect is a distinct event from its cause. It could not,therefore, be discovered in the cause.” Yet, as Hume (1777:39) noted,even children make use of the ideas of cause and effect, so they mustcome from somewhere.

Because causality cannot be perceived in the single instance it can onlybe based on what being collected in a series adds to the single instancescollected. According to Hume (1737:78): “as this idea [causality] arises


from a number of similar instances, and not from any single instance,it must arise from that circumstance, in which the number of instancesdiffer from every individual instance. But this customary connexion ortransition of the imagination is the only circumstance in which they differ.In every other particular they are alike.” Generalization is not a solution tothe problem, because generalization over a series of particular experiencesonly adds to those particulars the perceiver’s feelings about the series (asa series), according to Hume, which is not an empirical experience ofcause and effect. Hume argues that all that is added to single instancesby their collection in a series is our feeling about the series.7 There isnothing empirically different about the series that is more than the sumof its parts. Therefore, the feeling persons get about the series is the onlyaddition and must be the basis of the idea.

But, while generalization, or repetition over a series of impressions,cannot produce the impression of causality if it is not present in any ofthe single instances, repetition can produce the belief that similar thingswill keep happening. This belief becomes customary and habitual, andaccording to Hume (1737:78) custom and habit are the origin of thenotion of causality: “when many uniform instances appear, and the sameobject is always followed by the same event; we then begin to entertain thenotion of cause and connexion.” For Hume (1737:78 emphasis added)this notion is not empirically valid, because it originates with a feeling“We then feel a new sentiment or impression . . . and this sentiment is theorigin of that idea which we seek for.” Thus, causality originates in anindividual feeling, not in an empirical experience.8 It, therefore, has noempirical validity and persons are mistaken when, on the basis of customand habit, they believe that the transfer of this feeling to external relationshas any validity.

7 Once again this is very close to Durkheim. However, Durkheim makes it the basis of apositive argument whereas, for Hume the implications are negative. One of the importantaspects of social life that is greater than the sum of its parts is the feeling that participantsget from it. However, since all persons get the same feeling, and since this feeling incommon is the purpose of the collective actions in which persons engage, and necessaryin order for concerted social action to exist, this feeling has great importance, whereasfor Hume it is dismissed as mere feeling.

8 The argument concerning the relation between feelings and the general concept of causal-ity is an important one. Durkheim argues that the concept of causality originates in afeeling. But, for Durkheim this feeling has empirical validity because it is a general feel-ing which is shared by all participants. The feeling is not an individual matter, but isthe social product which is the purpose of the ritual. It is the essence of the ritual. Thefeeling is the ritual creating the unity of the group. Durkheim would agree with Humethat feelings have no validity when they are the product of individual perceptions whichdo not contain general social forces (feelings). The ideas generated by these feelings alsohave no empirical validity when applied to natural objects and relations.


In the Treatise Hume argued that the concept of causality is merely theresult of custom and habit. He maintained that there are no empiricallyvalid experiences corresponding to causality. Therefore, while we makeuse of the concept of causality, we have no “idea” of it, in Hume’s specialsense of the word idea. In such cases a “custom” or “habit” of mind hasintervened to produce a concept that gives the appearance of an idea.

Later in the Enquiries, Hume (1777:47) concluded that causality mustbe an instinct because it cannot be explained any other way.9 However,while he concedes that the idea of causality has an origin in instinct, thisis not an apriorist argument. For Hume the resulting concept has novalidity because it is not empirically based. He (1777:47) rejects the ideathat it could be a faculty of reason: “All these operations are a speciesof natural instincts, which no reasoning or process of the thought andunderstanding is able either to produce or to prevent.” For Hume it is onlythe disposition to believe in causality that is an instinct and causal beliefsare not knowledge. For Hume, knowledge of empirical matters cannotbe arrived at simply by an operation of the mind. When combined withexperience, the instinct to believe in causality leads only to customary andhabitual causal reasonings which the mind adds to experience. Theseadditions themselves have no basis in experience and hence no empiricalvalidity.

From a classical empiricist perspective, the origin of the concept ofcausality would have to be explained in terms of direct individual expe-rience of empirical objects or events in order to have empirical validity.10

What Hume demonstrated was that causality cannot be explained on the

9 Apparently custom and habit required a greater validity in the perception of causalrelations than Hume was willing to allow. Hume’s position in the later Enquiries [1777]appears superficially to be closer to that of Kant than his position in the earlier Treatise[1739] because he refers to the idea of causality as an instinct. In the Enquiries Humeconcluded that the disposition to believe in causality must be an instinct, because, whilewe have no adequate empirical basis for believing in causal relations, even children thinkin terms of them (Hume [1777]).

However, while for Hume the disposition to think in causal terms about the relationbetween the objects around us must be apriori, in the sense that it is an instinct, he isnot taking what is known as an apriorist or Kantian position when he argues this. Kantargues that causality is an apriori faculty of the judgement which is the same in all ofus and therefore that judgments of causality have validity for all humans even thoughthey can have no adequate empirical basis. This is Kant’s “transcendental” argument.Kant is saying that while the concepts have no empirical validity, because they are sharedidentically by all humans they are a valid basis for “human” knowledge.

Hume does not accept this. He does not make the disposition to believe in causalityas a faculty of judgement. He merely argues that thinking in causal terms, or believingin causal relations is an instinct or disposition.

10 This perspective rules out various strategies like Plato’s pre-birth observation of theforms, and other variants on innatism. It also rules out shared mind theories like thoseof Berkeley, Hegel, and later Kant.


basis of individual perception alone, because causality is not a sensiblequality of natural objects and their relations that can be perceived byindividuals in a single case. If the disposition to believe in causality is aninstinct, or a conclusion drawn after perceiving a series of natural objectsor events, it is something added by the mind to perception, not part ofthe original perception, and therefore not empirically valid.

The significance of Hume’s project is that it fails, or rather, that Humeshows that a positivist empiricism (as he conceived it) could not explainthe empirical origin of the essential general categories on the basis ofindividual perception. Durkheim argues that ([1912:519–20]1915:406–7;*1995:368) “For just this reason, the philosophers of empiricism haveseen these different ideas as so many mythological aberrations.”

Durkheim’s own version of the empiricist dilemma is entirely con-sistent with Hume’s position. Durkheim ([1912:519–20]1915:406–7;*1995:368) argues that: “The senses take in only states that are real-ized, achieved, and external to one another, while the internal processthat binds these states together eludes the senses. Nothing they teach uscan possibly suggest to us the idea of something that is an influence or anefficacy.” It is only because moral force is made available to “internal”experience “whole,” so to speak, and that it is a collective and not anindividual feeling, that it avoids this empiricist dilemma.

8.1.3 Internal versus External Experience

The next step in the proof moves from external to internal experience.Durkheim ([1912:519–20]1915:406–7;*1995:368, emphasis added)argues that: “if external experience has no part in the origin of theseideas and if, on the other hand, it is inadmissable that they should havebeen given us ready-made, we must assume that they come to us frominternal experience.” Hume had argued that cause was available internallyas a feeling. But, for Hume, because these feelings were individual, andcreations of the mind, they had no empirical validity. Durkheim is makinga very different argument. He argues that the feeling of moral force is aninternal experience of an external, social force, the experience of whichis shared with the collective. Thus, for Durkheim moral force is availableto internal experience in an empirically valid way.

Given the frequent interpretation of Durkheim as a rationalist, or apri-orist, it is also important to note here that Durkheim rules out the pos-sibility that the idea of causality comes ready-made as “inadmissable.”If external experience cannot be its origin, then it must have an originin internal experience, because it is inadmissable that it be ready-made.This also rules out the sociology of knowledge interpretation. Unless the


categories are at some stage made available to internal experience, theycannot be enforced, or passed down as beliefs, by the authority of societyeither.

Before articulating his own argument with regard to internal experi-ence, however, Durkheim considers the empiricist approach to the ques-tion of the validity of internal experience. Empiricists have made twoarguments concerning the part played by internal experience in the gener-ation of general ideas like causality. First, classical empiricists argued thatbecause concepts, like causality, rely on internal mental processes, theyare not empirically valid. They argue this, even though they agree thatthe mind has direct access to internal experience, because the percep-tion of causality could not, from their perspective, be present in internalexperience. Later, a second argument was advanced, that the internalexperience of being moved by the will might be an actual empirical expe-rience of causality, that is perceived internally, and therefore, at leastpartially empirically valid. It is this second sort of argument that, in part,grounds a pragmatist approach to the problem.

The first issue Durkheim confronts is the problem of the “will.” He([1912:519–20]1915:406–7;*1995:368) argues that “It has often beenthought that the act by which our will comes to a decision, holds ourdesires in check, and rules our bodies could have served as the model forthis construction. In an act of will, it is said, we directly perceive ourselvesas a power in action. Seemingly, therefore, once man came upon that idea,extending it to things was all it took for the concept of force to come intobeing.” According to Durkheim, however, this internal perception of theaction of the will, cannot provide an origin for the idea of force in internalexperience, because internal experience does not have the characteristicsof force that need to be explained.

Durkheim argues ([1912:519–20]1915:406–7;*1995:368) that “aslong as the animist theory passed for demonstrated truth” the argumentthat the will served as the first model of force “could seem confirmed byhistory.” That is, as long as the argument was accepted that the religiousidea of force had its origin in personal forces, it would appear that thegreat historical religions supported the argument that the will could bethe model of such a force. But, as soon as Durkheim’s argument, thatsocial forces are the original model for religious forces, is accepted, thenthe will is no longer a possible model for such historical ideas.

The will argument also casts the dilemma on an individualist foot-ing. The forces perceived are individual and the perception of thempersonal and not shared by the group. Yet, the religious forces thatDurkheim ([1912:520]1915:407;*1995:368) has focused his argumenton are impersonal: “we know that the first forces men imagined are anony-mous, vague, diffuse forces, the impersonality of which resembles cosmic


forces and which therefore stand in the strongest contrast with the emi-nently personal power that is the human will.” Furthermore, accordingto Durkheim ([1912:521]*1915:407;1995:369) impersonal forces havea fundamental characteristic, “their communicability”, that contradictsthe assertion that they could have come from the perception of the will.Individual sense impressions are not communicable, or shared.

Nothing individual, or originating in sense impressions, can be thesource of the internal experience that Durkheim refers to, if it is to beconsidered collective and empirically valid. The will and the “I” are per-sonal, whereas the idea of force is impersonal. Durkheim ([1912:521]*1915:408;1995:369) argues that the forces represented by the totemicrituals are experienced by all participants, whereas personal forces wouldhave been incommunicable (that is, they would not be shared in commonwith others): “the self has just the opposite characteristic: it is incommu-nicable.”

Empiricists, according to Durkheim, often tried to find an origin forgeneral ideas like causality in personal forces. Durkheim argues that if per-sonal forces were the source of the idea of causality, or if they came first,then personal forces, or spirits, would have been the first ones with whichpeople populated the world. They would have been the first totems, butthey were not. Therefore, he says, the empirical evidence shows that nei-ther the will nor other individual internal states could have been the modelfor the idea of causality. Durkheim ([1912:520]*1915:407;1995:368)argues, furthermore, that the particular form of causal relation embodiedin the totemic rite of “like produces like” could not have come from theexperience of personal forces, because personal forces do not posses thecreative power to transmute themselves into other things; a fundamentalcharacteristic of the belief, and another mode of communicability.

Therefore, while Durkheim will argue that persons have access to thecategory of causality because it is experienced internally, he denies thatcausality comes from an internal perception of individual mental or phys-ical states. An argument that causality has its origin in internal individualstates raises the same objections as the argument that it has its originin the perception of external states, as far as Durkheim is concerned. Itmust have an origin in internal experience of external social forces thatare shared with a collective in order to be empirically valid.

8.1.4 Cause as Internal Experience

Durkheim argues that in order for causality to have a valid empiricalorigin in internal experience, the forces that give rise to the idea mustbe impersonal, that is, collective, as well as internal. He will argue thatthe only sort of force that fulfills this condition is collective, or moral,


force. Durkheim ([1912:521]1915:408;*1995:369) says that the experi-ence of moral force satisfies this requirement: “It can come to us onlyfrom our inward experience; the only forces we can touch directly areof necessity moral forces. At the same time, however, they must also beimpersonal . . . Now, the only forces that satisfy this twofold conditionare those that arise from life in common: collective forces.” Moral forcescan be “touched directly.” This is a critical point. Durkheim is not sayingthat the idea is learned, or socially defined. He argues that it comes fromdirect experience of impersonal social forces. Impersonal social forces are“touched” by the participants. But, that experience cannot be individual,it must be collective as well as impersonal in order to satisfy the require-ments of empiricism. Moral forces are able to satisfy these requirements,according to Durkheim, because they are both internal and impersonal.Durkheim ([1912:521]1915:408;*1995:369) argues that “In actuality,they [these forces] are on the one hand wholly psychic, made exclusivelyof objectified ideas and feelings, and on the other hand, they are by defi-nition impersonal, since they are the product of cooperation.11 Being thework of all, they are the property of no one in particular.”

External forces share some properties with internal forces. But theirexternal character prevents them from being directly perceived whole, soto speak, by consciousness. Durkheim ([1912:521]1915:408;*1995:369)says “granted, physical forces have the same property [of spreading spon-taneously and entering subjects from the outside], but we cannot havedirect consciousness of them. Because they are external to us, we can-not even apprehend them as such.” External forces cannot be perceivedas forces “as such.” Durkheim ([1912:521]1915:408;*1995:369) arguesthat “when I run against an obstacle, I have a sensation of confinementand discomfort; however the force causing that sensation is not in me butin the obstacle and thus beyond the range of my perception. We perceiveits effects but not the force itself. This is not the case with social forces.”

According to Durkheim ([1912:521]1915:408;*1995:369) socialforces are experienced differently: “Since they are part of our interior life,we not only know the results of their action, but see them in action.” Socialor moral forces have their entire effect within the individuals who makeup the collective. Moral forces only exist if they are felt internally. Theyonly have efficacy when they are felt. Therefore, what is being felt is theirforce and efficacy. Durkheim argues ([1912:521]1915:408;*1995:369)

11 What is of particular concern in sections like this one is the use of the words psychic andideal. An objectified feeling is the most important part of this. We have a collective feeling.Then we objectify it, create a representation for it, and can treat it as a collective idea.Which, it could be argued is what an objectified idea is. It is certainly not just an ideaand it contributes to a force. It is the force that generates the experience of causality.


that “because this happens entirely within us, we capture in action theconstraining and necessitating influence that escapes us when it comesfrom an external thing.”

According to Durkheim ([1912:521]1915:408;*1995:369, emphasisadded) if moral force is not experienced internally, it does not exist:“The force that isolates the sacred being and holds the profane ones ata distance is, in reality, not in that being; it lives in the consciousness ofthe faithful. Thus the faithful feel it at the very moment that it acts on theirwills to prohibit certain actions and prescribe others.”

The fact that moral force is available to internal experience does not,however, mean that those who experience it interpret that experience ade-quately. According to Durkheim ([1912:521]1915:408;*1995:369) “Ofcourse, we do not always interpret that influence adequately, but we can-not fail to be conscious of it.” The beliefs that represent the collectiveinterpretation of the feeling of force may not grasp it adequately, and thisis the origin of the many beliefs that seek to explain the experience, butthe force will nevertheless make itself felt in consciousness. That feelingis all that is necessary for the development of the category of causality,according to Durkheim. An adequate interpretation will have to await thehistorical development of the sociology of knowledge, which he arguescomes ever closer to an adequate understanding over time.

8.1.5 Hierarchy

The next step in Durkheim’s analysis of the concept of causality is topoint out that moral force has overtones of hierarchy and social authority.This, he argues, shows that the concept bears the mark of its social origin.According to Durkheim ([1912:521]1915:408;*1995:369): “the idea offorce bears the mark of its origin overtly. It in fact entails an idea of powerthat does not go without those of ascendency, mastery, domination-and,correspondingly, of dependence and subordination.” Durkheim arguesthat hierarchy does not exist in nature. Society has created these classifi-cations.

Durkheim’s general ideas about the social origin of hierarchy can betraced to Rousseau. According to Rousseau, before the development ofsociety human animals were free and equal like the other animals. Theorganization of animals in nature was purely individualistic: survival ofthe fittest reigned. The development of society, according to Rousseau,creates a new set of needs. Social animals must develop a new facultyof reason in order to pursue these new needs and goals. Reason, as asocial creation, comes ready made with ideas of power and dominationand over time makes most men unfree although they now have reason.


In Rousseau’s dialectic, the task for civilized man is to use their reasonto break free of the inequalities that are inherent in the social forms thatcreated their reason. This is the deepest form of dialectic. What Durkheimmakes of it here, is only that hierarchy cannot be found in nature, andtherefore, since the idea of force contains the idea of hierarchy, then theidea of force must be a social creation and not derived from the perceptionof natural forces (whether within the individual or in nature).

Following this line of reasoning, Durkheim argues that another rea-son that totemic forces cannot have their origins in personal forces (anidea dealt with earlier in the discussion of the will) is because totemsinvolve hierarchies. According to Durkheim, there are no hierarchies innature. Referring back to the will argument, Durkheim points out thatwhen philosophers, like Kant, write in individualistic terms about some-thing even as personal as the relationship of persons to their own bodiesthey must use terms with a social origin to do so. For instance, Kan-tian rationalism requires “governing” the body and “subjecting” it tothe dictates of reason. These concepts have clear social connotations.According to Durkheim ([1912:523]*1915:409;1995:370), “men havenever succeeded in imagining themselves as forces; mistress over theirbodies, except by introducing concepts taken from social life.” Durkheim([1912:523]*1915:409;1995:370) argues that hierarchies are all sociallyconstructed and it is from society that the general ideas of power haveoriginated: “it is society which confers upon [masters] the singular prop-erty which makes the command efficacious and which makes power.”Therefore, for totemic forces to contain the idea of hierarchy shows thatthey have a social origin.

Durkheim’s argument with regard to human dualism also comes upin this discussion of hierarchy as a social phenomenon. He ([1912:523]1915:409;*1995:370) argues that in order to develop the idea of the will asa personal force the person “had to differentiate himself from his phys-ical double and impute a higher sort of dignity to himself than to hisdouble-in a word he had to think of himself as a soul.” As an individualanimal being, homo sapiens is not able to even imagine the possibil-ity of personal, or individual forces. The animal, Durkheim ([1912:523]1915:409;*1995:370) says, is not even conscious of itself as a personalforce, because it is not social in the same way as humans are and “doesnot impute a soul to itself.” Thus, even the idea of personal forces wouldhave to come after the experience of social forces.

8.1.6 Cause as Habit versus The Principle of Causality

The next step Durkheim takes is to argue that there is a sort of necessityand constraint about the idea of causality that empiricism has not been


able to explain. If causality were only an expectation, or habit of belief,that a certain thing would happen, Durkheim asks, then how could thatexpectation become so strong as to work “apriori,” in advance of theevidence?

Durkheim ([1912:523]1915:409;*1995:370) argues that “the idea offorce is not all there is to the principle of causality.” Causality sets up aprecise relationship in a time series. According to Durkheim ([1912:524]1915:410;*1995:370) “this principle consists in a judgement stating that aforce develops in a definite manner and that its state at each moment of itsevolution predetermines the succeeding state. The first is called cause; thesecond effect; and the causal judgement affirms the existence of a neces-sary conjunction between these two moments of any force.” Empiricismwould predict that persons would come to have a more or less strongexpectation that certain results would follow certain events. However,Durkheim argues that the human experience with the idea of causal-ity is quite different. According to Durkheim ([1912:524]1915:410;*1995:370) “ruled by a sort of constraint from which it cannot free itselfthe mind sets up this relation in advance of any proof.” The person doesnot wait to perform a number of trials to see how strong their expectationof cause and effect will be. According to Durkheim ([1912:524]1915:410;*1995:370–1) the mind “postulates this relationship, as people say,apriori.”

Durkheim is not arguing that the category of causality is apriori, how-ever. He is arguing that when persons have experiences that they wouldsay involve causality, they do not approach those experiences the wayempiricists would predict. They approach those experiences as thoughthey already had an idea of causality that could be applied in the abstractto each next situation. Persons are not born with this principle, but oncethey have experienced the moral force embodied in mimetic rites, theypossess the category of causality. This category acts an external rule thatwill not allow thought to deviate.

Durkheim argues that the principle of causality is not like opinion orhabit either. He says ([1912:524]1915:410;*1995:370): “Now, the prin-ciple of causality has an entirely different character. It is not simply aninherent tendency for our thought to unfold in a certain way; it is a normexternal and superior to the flow of our representations, which it rulesand regulates absolutely. It is endowed with an authority that binds theintellect and goes beyond the intellect; in other words, the intellect isnot its creator.” There is an experience of necessity and constraint aboutthe way the idea is experienced that is not consistent with the empiricistexplanation.

Durkheim ([1912:524]1915:410;*1995:371) argues that “Empiricismhas never succeeded in giving an account of that apriorism and that


necessity. Never have the philosophers of that school been able to explainhow an association of ideas reinforced by habit could produce anythingother than a state of expectancy, a more or less strong predisposition onthe part of ideas to call themselves to mind in a definite order.”

But, while he argues that the idea of causality is more than an expecta-tion or habit of thought, he also argues that making the principle aprioriin Kant’s sense does not solve the problem either. According to Durkheim([1912:524]1915:410;*1995:371) “In this regard it does no good to sub-stitute hereditary for individual habit. The nature of habit does not changebecause it lasts longer than a man’s life; it is only stronger. An instinct isnot a rule.” The idea of causality exhibits a real necessity and externalcharacter, that neither habit nor heredity can explain. A habit, or instinctdoes not have the power to rule representations absolutely that Durkheimattributes to the principle of causality. An expectation, or belief, thatthings will happen in a certain way does not explain this authority of theprinciple of causality over the mind. That necessity, Durkheim says, andthe authority of society to enforce and convey it, can only be explained,if there is a real necessity involved.

8.1.7 Consideration of Empiricism

Durkheim next moves into his argument that moral force, or authorityis behind the necessity which persons feel about the idea of causality.But, he takes this argument in two stages. First, Durkheim ([1912:524]1915:410;*1995:371) hints at his own position, that society is the sourceof the real necessity that alone can explain the principle of causality, beforemoving on to a consideration of empiricism, arguing that: “The rites juststudied enable us to discern a source of that authority that until now hasbeen little suspected. Let us recall how the causal law that the mimeticrites put into practice was born.”

But, before elaborating further on his own position, Durkheim con-siders whether an empiricist explanation of the feelings generated bythe performance of the rites could also explain the feeling of necessity.Durkheim’s method is always to deal with all possible counter-arguments,in a sort of process of elimination, before presenting his own. Durkheimargues that from an empiricist perspective all that is possible is that anassociation of ideas arises. While this argument is somewhat different indealing specifically with feelings generated by the rites, Durkheim arguesthat it has the same weakness as other arguments for causality based on theassociation of ideas that he has already considered. His point here is thatif causality did not have the origin in moral authority which he attributesto it, then it would still fall prey to the classical empiricist dilemma.


One might argue that a certain regularity of expectation is created by theperformance of the rite. However, according to Durkheim ([1912:524]1915:410;*1995:371) this feeling of regularity is not the same thing asthe principle of causality:

The group comes together, dominated by one concern: If the species whosename it bears does not reproduce, the clan is doomed. In this way the commonfeeling that animates all its members is expressed outwardly in the form of definitemovements that always recur in the same way in the same circumstances. And forthe reasons set forth, it turns out that the desired result seems to be obtained whenthe ceremony has been conducted. An association is thereby formed between theidea of this result and that of the actions preceding it. This association doesnot vary from one subject to the other. Because it is the product of a collectiveexperience, it is the same for all who take part in the rite. Nonetheless, if no otherfactor intervened, only a collective state of waiting would result.

The “collective state of waiting” described here is like Hume’s habit ofbelief. Because of a constant conjunction, persons come to expect thatwhenever one thing happens a specific other thing will follow. But, asHume points out, the idea of causality is in such a case only an inference,and has not been perceived in its own right. One cannot claim on the basisof such an argument that causality has actually been produced by the rite.This “collective state of waiting” only produces the belief in causality.But on the basis of such constant conjunction, according to Durkheim([1912:524]1915:410;*1995:371), “an imperative rule of thought wouldnot come into being.”

Durkheim argues that the performance of imitative rights, “like pro-duces like,” involves a belief in causality, not just the association ofideas. This belief in causality is a part of the belief in the totemic rites,because, in order to believe in the rites, one must believe in causality. Thisidea of causality cannot, according to Durkheim ([1912:511]*1915:400;1995:362) come from the mere association of what comes first with whatfollows:

So there could be no question of association, whether correct or not; there is areal creation, and we cannot see how the association of ideas could possibly leadto a belief in this creation. How could the mere act of representing the movementsof an animal bring about the certitude that this animal will be born, and born inabundance?

Because persons need to come together ritually in order to create them-selves and their community, there is both an individual and a group needto believe in the efficacy of the ritual. It is this need for the performanceof the ritual that initially invests society with the authority to demand thatits members believe.


8.1.8 The Authority of Society

The authority of society is made necessary by the dependence of bothsociety and human reason on the continued performance of the ritu-als. Social unity depends upon the performance of the rituals. But, therituals will not develop in the first place, or continue to be performed,unless the participants believe in a principle of causal efficacy. There-fore, beliefs must also be enforced. According to Durkheim ([1912:524]1915:410;*1995:371) “To prescribe that the animal or plant must be imi-tated to make them come to life again is to make ‘like produces like’ intoan axiom that must not be doubted.” Durkheim ([1912:524]1915:410;*1995:371) argues that “Opinion cannot permit individuals to deny thisprinciple in theory, without at the same time permitting them to violate itin their conduct.” In other words, if people are allowed to deny the prin-ciple they will also cease to perform the rituals and then the experienceof the category will not be created. If the principle were doubted, thenthe rite might cease to be performed and the category of causality wouldcease to be experienced.12

Because the categories are necessary for reason, communication, andcooperation, society must impose a belief in the causal efficacy of ritualinteraction as a moral obligation for members of every group. Accordingto Durkheim ([1912:524]1915:410;*1995:371):

Because a social interest of premier importance is at stake, society cannot let thingstake their course, at the mercy of circumstances . . . Society requires this ceremonywhich it cannot do without to be repeated whenever necessary and . . . imposesthem as an obligation. Those actions imply a definite attitude of mind that, inresponse, shares the same quality of obligation.

If causality were only a collective belief, or a representation that couldbe socially learned, even if not directly experienced, then it would not benecessary to force people to take part in the rituals that create the idea ofcausality.

However, causality is not only an opinion created by social obligation.The category of causality is, according to Durkheim, an empirically validemotional experience that cannot be created merely by enforced belief,it must be experienced during the performance of mimetic rites in orderto have empirical validity. The need for people to believe in the rights issecondary to the need to get people to perform the rites. What society

12 It is important to note that this relationship between belief and practice changes inthe advanced division of labor. In modern societies, Durkheim argues, people committhemselves to practices without having corresponding beliefs. Thus, people come to havea direct commitment to the practices, or manners, of life.


must make sure of is that the belief in causality, which is necessary for theperformance of the rite, cannot be doubted, otherwise the rite would notbe performed, and the real feeling of causality would not be produced.

In traditional forms of society religious faith ensures that persons fulfillthis obligation. The needs of society are the reason for religious obliga-tion. Even the self or the human “soul,” as the origin of the personal-ity, is the result of “civilization,” according to Durkheim ([1912:523]*1915:409;1995:370). If the making and remaking of self and intelligi-bility did not depend on the performance of ritual practices, there wouldbe nothing moral about observing or violating those practices. They arenot moral just because social authority says so. Social authority says sobecause they are essential to persons becoming human and thereforemoral beings.

The true purpose of religion, for Durkheim, is, therefore, to establishand maintain common categories of thought.13 Durkheim ([1912:5514–15]*1915:403;1995:365) argues that “The true justification of religiouspractices does not lie in the apparent ends which they pursue, but ratherin the invisible action which they exercise over the mind and in the way inwhich they effect our mental state.” The apparent ends are not important.It is only because shared practices are necessary to sustain intelligibilityand remake individuals in a social framework that they are moral. Accord-ing to Durkheim ([1912:524]1915:410;*1995:371) the ritual belief isreinforced by “a logical principle that is none other than the intellec-tual aspect of the ritual one.” The respect for the experience of forceis transferred to a respect for the authority of society. This respect, inturn, according to Durkheim ([1912:524]1915:410;*1995:371), is givento the ways of thinking and acting which are created by, or required by, theritual: “The respect evoked by society passes into those ways of thinkingand acting to which it attaches value.”

8.1.9 The Category of Causality

Durkheim then returns to a discussion of the “category” of causality.Referring to “a sociological theory of the idea of causality,” Durkheim([1912:526]1915:411;*1995:372) says that “Using this example, we cantest once again how a sociological theory of the idea of causality, and the

13 This argument is similar to Hobbes’ justification of monarchy. But, the philosophicalargument is different. Durkheim is not articulating the need of individuals for a collective,as Hobbes did, but rather speaking directly to the needs of the collective. If there is goingto be any collective thought and action certain prerequisites must obtain. Although theargument that persons could not have common thought without this unity is similar, thelogic of the argument is quite different.


categories more generally, both diverges from the classical doctrines onthis question and accords with them.” As he has maintained throughout,Durkheim insists here that his theory retains important characteristicsof both empiricism and apriorism while diverging from them in areasthat were problems. It is the empirical character of the categories thatDurkheim seeks to preserve, along with their universality, while replac-ing the explanations with a new one of his own. He writes ([1912:526]1915:411;*1995:372) that:

Here, as in apriorism, causality retains the a priori and necessary character of thecausal relation. The sociological theory does not simply affirm it but also accountsfor it and yet does not, as in empiricism, make it disappear while ostensiblyaccounting for it. Besides, there can be no question of denying the part thatbelongs to individual experience. That the individual by himself notes regularity,is not to be doubted. But this sensation is not the category of causality. The first isindividual, subjective, and incommunicable; we make it ourselves from personalobservations. The second is the work of the collectivity, which gives it to us ready-made. It is a framework in which our empirical observations arrange themselvesand which enables us to think about them-that is, to see them from an angle thatenables us to understand one another on the subject of those observations.

Durkheim does not deny individual experience. What he argues is thatthe individual perception of regularity is not the origin of the categoryof causality. But he rejects the a priori explanation of where the ideaof causality comes from. It comes, according to Durkheim, from thecollectivity. It does not come from the collectivity in the form of ideasthat are learned as ideas, however. They must be experienced as emotionsduring the performance of ritual interaction. Providing such experiencesof the categories is the purpose of religion, according to Durkheim.

There is a relationship of sorts between individual sensations andthe emotional experience of the categories, because both are real. But,they are not identical. Nevertheless, because both are real, according toDurkheim ([1912:526]1915:411;*1995:372), the category of causalitycan be applied to individual perceptions, even though it does not comefrom them: “To be sure, if the framework can be applied to the con-tent, that is because it is not without relationship to that content, but theframework does not merge with what it contains.”

Durkheim ([1912:526]1915:411;*1995:372) argues that his position isable to “retain the a priori and necessary character of the causal relation”while also accounting for that necessity empirically. Durkheim’s empha-sis is on empirical explanation and he holds himself throughout to anempiricist standard, rejecting apriorist explanations outright. However,classical empiricism did not come up with that explanation because itfocused on the individual. Given the individual as a starting point the


categories can only be the end result of an individual mental process.This, according to Durkheim ([1912:526]1915:411;*1995:372), was thegreat mistake of empiricism:

In sum, the mistake of empiricism has been to see the causal tie as only a learnedconstruct of speculative thinking and the product of more or less systematic gen-eralization. Pure speculation can give birth only to views that are provisional,hypothetical, and more or less plausible, but views that must always be regardedas suspect. We do not know whether some new observation will invalidate them inthe near future. Therefore an axiom that the mind does and must accept, withouttesting and without qualification, cannot come to us from that source.

Generalization is a faulty tool. It is not adequate, according toDurkheim, to the demands of concerted action and shared meaning.Durkheim takes a very modern position when he argues that such anexplanation does not account for what people are actually able to do andthat therefore, there must be an explanation that does account for howwe are able to do and think what needs to be done and thought. Accord-ing to Durkheim ([1912:526]1915:411;*1995:372), “The demands ofaction, especially of collective action, can and must express themselvesin categorical formulas that are peremptory and sharp and that brook nocontradiction, for collective movements are possible only if they are con-certed, and thus regulated and well defined. They preclude blind groping,which is a source of anarchy.”

Concerted cooperative action would simply not be possible, accord-ing to Durkheim, unless some mechanism developed which would cre-ate empirically valid categories of the understanding. This mechanism,he argues, is religion, and more specifically, the performance of reli-gious ritual. Thus the imperatives of action are, according to Durkheim([1912:526]1915:411;*1995:372) the same as the imperatives of thought:“And since action cannot do without the intellect, the intellect is eventu-ally pulled along in the same way, adopting without argument the theo-retical postulates that practice requires. The imperatives of thought andthose of the will are probably two sides of the same coin.” Consequently,a study of the imperatives of action is inherently epistemological: it con-stitutes a study of the imperatives of thought.

The category of causality, according to Durkheim makes inter-subjective meaning possible. He ([1912:526]1915:411;*1995:372) ar-gues that “it is a framework in which our empirical observations arrangethemselves and which enables us to think about them–that is, to see themfrom an angle that enables us to understand one another on the subjectof those observations.” Without that framework individual sense impres-sions, he has argued, are incommunicable.


8.2.0 Durkheim’s Socio-Empirical Argument for Causality

In the final two chapters of Book III, Chapters Four and Five, Durkheimreviewed the evidence in support of his claim that religion is the origin ofthe category of causality. The main point of the two chapters, in fact, apoint echoed throughout the book, is that while there are many differentforms of religious rite, they all have the same end and object. Havingmade his argument for the origin of the category causality in mimeticrites, Durkheim reviews various other types of ritual, including rites ofinitiation and rites of mourning. His point here is that these rites havethe same object as the others. According to Durkheim ([1912:591–2]1915:460;*1995:417) “the result in all cases is communion among indi-vidual consciousnesses and mutual calming. While the fundamental pro-cess is always the same, different circumstances color it differently. Inthe end, then, it is the unity and diversity of social life that creates at thesame time the unity and the diversity of sacred beings and things.” It issociety that transforms action into something of moral consequence. Themoral consequences, in turn, take reason as their end. The moral forcesof social unity are the origin of the categories.

Durkheim’s argument with regard to the category of causality, isintended as an epistemological analysis in the classical sense. The dif-ference is that Durkheim begins, as did Rousseau, with the idea that thesocial being is transformed by the social relations in which it finds itself.Therefore, an epistemological analysis must begin with the social forcesthat transform the individual, and not with the individual mind. Philoso-phers have always had a hard time with the idea that the social exists asa force that cannot be calculated merely as an aggregate of individualactions. But, from the beginning this has been sociology’s fundamentalpostulate, and it was certainly Durkheim’s. That the objective reality ofsocial facts is sociology’s most fundamental phenomenon was the mainpoint of Durkheim’s Rules of the Sociological Method. Reiterated explicitlyin the Preface to the Second Edition as something “said again” becauseof its importance, that postulate is carried through in his argument forthe empirical origin of the category of causality.

For Durkheim, causality cannot be merely a generalization based ona series of perceptions of the social, this would not allow for collectiveaction and mutual communication. It certainly cannot be a Kantian apri-ori. Rather, he argues, causality must, along with other social forces, beimmediately experienced as a general phenomenon in the single instance.When Durkheim makes this argument he is striking at the heart of theepistemological problem. When he argues that the experienced empiricaleffects of ritual action are more than the sum of the individual parts (i.e.,


that shared ritual has meaning and social, or moral, force and emotionaleffects, which the sum of its parts alone does not) he is making a signif-icant improvement on the empiricist position. The category would thennot be a general idea added by the mind to a series of perceptions ofnatural objects and events, but, rather, created by the experience of ritualsocial process and thereby made immediately available to perception inthe individual instances of the experience of social ritual by participants.

In developing the argument that necessary force, or causality, is a col-lective idea which can be immediately experienced in ritual social action,Durkheim challenged the argument current in his day that the empiricistprinciples of resemblance and association of ideas are sufficient to explainthe development of the idea of causality. This version of empiricism is pos-itivist and Hume would have repudiated it. Durkheim uses Hume’s ownargument against this empiricism, current in his day. He argues that theassociation of ideas cannot create ideas that were not there in the initialseries, it can only produce an expectation reinforced by repetition thatsomething will continue to happen as it has in the past. This, Durkheim([1912:524]*1915:410;1995:371) argues is not the category of causality:

Philosophers of this school [empiricism] have never been able to explain how anassociation of ideas, reinforced by habit, could produce more than an expectationor a stronger or weaker disposition on the part of ideas to appear in a determinedorder. But the principle of causality has quite another character. It is not merelyan immanent tendency of our thought to take certain forms; it is an externalnorm, superior to the flow of our representations, which it dominates and rulesimperatively. It is invested with an authority which binds the mind and surpassesit, which is as much as to say that the mind is not its artisan. In this connection,it is useless to substitute hereditary habit for individual habit, for habit does notchange its nature by lasting longer than one man’s life.

Durkheim argues that the concept of causality has an authority which“binds the mind and surpasses it.” No combination of individual experi-ences (by association, habit, or resemblance) are sufficient to explain thedevelopment of this concept and the force which it exerts. Therefore, itmust have an external empirical origin.

Essentially Durkheim picked up where Hume left off, but with a sig-nificant twist. Where Hume demonstrated the impossibility of an indi-vidualist empiricism, Durkheim accepts that impossibility and argues thepossibility of a socio-empirically based empiricism. By substituting socialforms and the experience of those forms by the individuals enacting them,for individual perceptions of natural phenomena, Durkheim claims tocircumvent Hume’s skeptical conclusion.

The idea of causality is revealed by Durkheim to have been available toexperience all along, on certain sorts of occasions. It is this substitution of


social forms for individual perception which not only provides the key toDurkheim’s solution to Hume’s dilemma, but, consistently through theyears comes to define what is distinctive about sociology as a discipline.The primary unit of analysis for sociology has not been the individual,but various posited levels of social organization.

Durkheim’s treatment of emotion is a case in point. The emotion pro-duced by causality is not just a highly charged emotional state. Thatwould be individual, and could be achieved by the individual alone. Theemotions have the quality of creation and obligation in them. They areemotions that carry the authority of society with them. They are moralforces and immediately experienced as such. As the participants dancethe dance and become emotionally charged, they are feeling the effectsof the creation they are enacting on themselves. There is the thing theyare trying to do: reproduce the species. Then there is the felt effect ofthat creative process as they feel themselves being created and recreatedas members of the clan. They are also feeling the obligation to do this.As they feel themselves being recreated they know that they need to keepdoing this. It is not an obligation artificially and externally imposed, buta felt internal need. If they do not recreate themselves as such they willcease to exist as such and they feel this as they perform the rite: and theyall feel it simultaneously.

The concept of causality still exhibits a degree of relativity, as Durkheim([1912:527]*1915:412;1995:373) acknowledges in the last paragraph ofChapter Three: “The principle of causality has been understood differ-ently in different times and places; in a single society it varies with thesocial environment and the kingdoms of nature to which it is applied.”But, as he says there, this is not a problem. For Durkheim the purposeof the category of causality was not to secure ultimate truth, or transcen-dental validity, that is, the category did not evolve to fulfill philosophicalpurposes. It’s purpose is not philosophical, but social, and its measureis not universal consistency. Categories came into being to fulfill socialneeds. There is no reason the categories cannot exhibit a certain degreeof variation from place to place and still fulfill those social needs.14 Thepurpose of the categories, he argues, is to provide a framework for col-lective thought which variations from place to place and time to time areirrelevant to.

14 A certain inconsistency in referring to the categories as both universal and relative to somedegree may bother the reader. However, if we take universality to mean that causalityis ultimately a very abstract feeling of creation it is hard to see how the idea could bedifferent in different places. What would it mean to say that it was different as an abstractfeeling? On the other hand the practices which give rise to the feeling and the context inwhich they are experienced will differ greatly.


Durkheim ([1912:527]*1915:412;1995:373) also wants to be sure thathis “observations,” with regard to the category of causality, are notseen as “a complete theory of the concept of causality.” I would spec-ulate that Durkheim is here entering into a complex territory. The cat-egories, which are universal and collective, must be used by individ-uals for purposes which are different from their origin. As Durkheim([1912:527]1915:412;*1995:373) notes: “we might well ask ourselveswhether a physicist and a biologist imagine the causal relation in the samefashion.” As he has continually asserted that the categories are univer-sal, he can only mean here that in their applications, which are differentfrom the rituals that gave birth to them, they only have an applicationby analogy. Each purpose will develop its own sociology of knowledgewith regard to the ideas and uses of the terms. That does not mean thatthe category varies from place to place. Even with regard to the appliedconcept, it is clear that Durkheim ([1912:527–8]*1915:413;1995:373)feels that his remarks can “be generalized to a certain degree.”

9 Logic, Language and Science

While the main body of The Elementary Forms is methodical in endlessdetail, the Conclusion is methodical in a very cryptic way. In the Con-clusion, packed into a relatively short number of pages, Durkheim notonly reviews the main arguments of the book; that rites are more essen-tial than beliefs, and that participation in rites creates the categories ofthe understanding; he also makes new arguments about the relationshipbetween religion and scientific argument, logic, and language, and thenrelates these new arguments to the epistemological argument laid outin the central chapters. This involves a closely argued series of pointsthat are carefully organized into four numbered sections. It is only inthe fourth of these sections that Durkheim discusses the categories ofthe understanding. That section will be discussed separately in the nextchapter.

An analysis of various misreadings of The Elementary Forms suggeststhat scholars have tended to read Durkheim’s Conclusion as if it looselysummarized the argument of the book as a whole. It does not. The dis-cussions of logic and language, which are generally taken to be sum-mary arguments, in fact, appear in the Conclusion for the first time,and are quite different from the discussions of epistemology and logicwhich appears in the body of the text.1 Taking the discussions of logicand language to represent a summary of his epistemological argumentmakes that argument appear to be hopelessly contradictory. Citation pat-terns suggest that this misunderstanding has been exacerbated by the factthat scholars have mainly focused on the Introduction and Conclusion,treating the central chapters as merely a discussion of religion that is inci-dental to the overall argument.2 This has been consequential because the

1 Preliminary discussions of logic appear in Book II Chapter Three (discussed in 4.2.0),and Chapter Seven Section vi (discussed in 5.2.0). These discussions do not integratethe idea of logic into either Durkheim’s epistemology or his theory of language, however.They also do not establish the relationship between logic and epistemology, an argumentwhich appears only in the Conclusion.

2 See Rawls 1997 for a discussion of the two most prominent early critics who wrote inEnglish, Charles Elmer Ghelke (1920) and William Dennes (1924) and the relationship

262

Logic, Language and Science 263

central chapters are the only place where the argument for the epistemol-ogy is made. This organization of the text is, I believe, due to the factthat Durkheim is making an empirical argument that can only be madeto appear in the midst of empirical materials. As such it is a very unusualform of argument, and as Durkheim was the first to try it, and providedan Introduction and Conclusion that appeared to be conventional, theinnovation was missed.3

In Section i, Durkheim reiterates his argument that rites are moreimportant than beliefs and points out various problems and fallacies thatresult from treating religion as a matter of beliefs and not practices. Focus-ing on social action, as the essence of religion is not magical or material-istic, Durkheim says, because it is the transforming effect of social actionon persons that creates spirituality in the first place. Since it is the “real”society that creates spirituality and ideas about spirituality the appar-ent conflict between the real and the ideal, which has been considered soimportant by philosophers and continues to be a staple of social theory, isa false one. It also follows that the idea that western religions are superiorto Aboriginal religions, because of the alleged superiority of their beliefs,a point that Durkheim has made repeatedly in the text, is mistaken. Itis practices and not beliefs that are essential to religion, and aboriginalreligions have a more fully articulated practice than western religions.

In Section ii, Durkheim works out the relationship between religion andscience, arguing that there is no real conflict between the two. At the turnof the century this was a big issue. Religion was defending a domain ofbeliefs against science, arguing that scientific practice, particularly whenit dealt with the natural world, was fine, but that it should not try todeal with explanations of human moral and social life, or with religiousbeliefs. The play Doctor Jekyll and Mister Hyde dates from this period, andpresents the moral trials of a psychiatrist who tried to separate the goodand evil in people through a scientific experiment. Of course, the resultis disastrous and the scientist dies for his sins. Like Dr. Jekyll, Durkheimdealt with moral issues scientifically. The Division of Labor had presented

between their misunderstandings of the text and their citation patterns. In spite of clearmisinterpretations of Durkheim’s text and citation patterns that show an over reliance onreadings of Durkheim’s Introduction and Conclusion, both Gehlke and Dennes continueto be cited as reliable sources with regard to Durkheim’s epistemology.

3 The question of how to introduce an argument that needs to remain embedded in empir-ical demonstration is an interesting one. Wittgenstein’s Philosophical Investigations, doesnot have a conventional introduction. Garfinkel has adopted a similar strategy. This mayhelp make it clear that the arguments are unconventional and have the virtue of focusingthe reader on the empirical discussion in the text. However, it may also have obscuredthe theoretical implications of the arguments and led to a misunderstanding of a differentsort. In any case, one suspects that if Durkheim had tried this his work would simply havebeen treated as incomplete.


an argument that there is an increasing need for justice in modern life andinsisted that the science of sociology played a necessary role in establishingthis need (Rawls 2003). The Elementary Forms argues that it is the functionof social practices in generating moral experiences to transform humanbeings into autonomous rational beings. Thus, for Durkheim, society isthe origin of morality and rational moral beings through an active processthat can only be established scientifically.4

In Section iii, Durkheim takes up the question of how logic and lan-guage could have come from society. Here he articulates a view of mean-ing as use, reminiscent of Wittgenstein, and discusses the possibility ofa sociology of knowledge. Unfortunately, Durkheim makes referencesin this section to society as “thinking” and “acting” that have led to theinterpretation of the argument as invoking a “group mind.” It is thereforeessential to point out that, according to Durkheim, society only has thesecharacteristics when it is put into motion by persons who are assembled toenact practices. He is not talking about a group mind, but rather about astate of collective action that only exists, and can only be experienced by,participants who are assembled and actively creating moral force throughthe enactment of practice. The upshot of this argument is that, as soonas there is society, use meanings, and hence logic, have to come into exis-tence. Therefore, the idea common among Europeans, that they havea highly developed logic while aboriginals do not, is wrong. Aboriginallogic is, from Durkheim’s perspective, as highly developed as westernlogic. Aboriginal peoples do not have clearly delimited general ideas. ButDurkheim argues that general ideas have nothing to do with logic. Theyare a manifestation of the western post-capitalist focus on the individual.Ironically, because the same tendency to confuse general ideas with logicthat Durkheim was criticizing was invoked by his critics, the argumentsof this section have done the most to obscure Durkheim’s epistemology.Durkheim’s rejection of general ideas seems, given this confusion, to bea rejection of epistemology in the classical form of the question.

In Section iv, to be taken up in the next chapter, Durkheim deals withwhat he calls a “final question,” of how certain categories of thoughtcame to be social. This is the only part of the Conclusion where he dealsexplicitly with the categories of the understanding. It is essential to seethat he treats the development of categories as a separate question fromthe development of logic discussed in Section iii.

4 It is important to note that Durkheim is not arguing that there is no deity, or that the ideaof a deity is not an important aspect of religion. He is arguing that the basic experiencesthat caused human beings to develop an idea of a deity, are social and emotional, notspiritual. Durkheim’s position is not in conflict with the idea of a deity. If one believes inan all powerful deity, there is no reason in principle why such a being, recognizing thelimitations of human knowledge, could not make itself known in this way.


Durkheim’s argument with regard to the development of logic and lan-guage, and the claim that society as a system of symbols and conventionscontains within it all of the universe and everything that can be known,has been much misunderstood and brought him a great deal of criticism.It has also deflected attention from the epistemology. The argument withregard to logic is similar to Wittgenstein’s position in his lectures of the1930s and 1940s, and was picked up in the 1980s, in particular by DavidBloor (1982),5 as a way of approaching the sociology of science. At thetime however, it was treated as a variant of Neo-Kantianism, with theresult that Durkheim’s references to the social world and its conventionswere interpreted as references to ideal conceptual constructs, instead ofto the concrete witnessable practices that Durkheim intended.6

Yet, throughout the Conclusion, Durkheim continues to insist on theprimacy of practice over ideas, maintaining that only a study of riteswill reveal the social causes of knowledge. The problem, according toDurkheim, is that concepts have been treated as primary. It is a mistakethat he seeks to rectify by locating the origins of meaning in concretepractices. He could not be arguing that knowledge is defined by concep-tual conventions of language and also arguing that empirical observationscan move our understanding beyond these concepts. The irony is thatDurkheim’s position seems to have been interpreted as idealist because of the veryargument in which he was challenging idealism and advocating the possibil-ity of an empirical basis for human knowledge. When Durkheim arguedfor a world of conventional meanings, he meant a socially constructedreal world of social possibilities, more like what Wittgenstein meant byconventions of “use”; meanings specified by concrete contexts that donot live in heads, dictionaries, or referential relations. And even then hewas only referring to the origin of logic, not the origin of knowledge.He was taken to mean, however, that an ideal world of only conceptualpossibilities defined the boundaries of knowledge.

There are two important reasons why the argument that practices aremore important than ideas is a difficult one to make in a western Euro-pean context. First, as Durkheim himself points out, western thought hascome to associate clear logical thinking with having one’s concepts clearlydefined. His own argument that the meaning of concepts has nothing todo with definitions, or, clearly specified general ideas, conflicts directly

5 David Bloor (1982) has combined a Durkheimian with a Wittgensteinian approach tothe study of scientific practice. His work in this regard has been particularly effective.

6 The difference between Wittgenstein and Phenomenology in this regard is that Wittgen-stein argued that meaning was given by concrete witnessable conventions of use, whereasPhenomenologists posit meaning patterns at a conceptual level in the mind. Gurwitschmoved this into gestalts that were somewhere between the concrete and the mind, but stillnot social conventions. Because Durkheim focused on witnessably enacted conventionsand not on beliefs and ideas, his position is closer to that of Wittgenstein in this regard.


with this belief. Second, since the advent of the Protestant Ethic, west-ern religions have focused increasingly on beliefs and have rejected theimportance of rites. In fact, the early Protestants thought of rites as repre-senting the traditional authority of the Medieval Catholic church. In thiscontext, rituals were thought to bind individuals to belief systems suchthat they were prevented from engaging in clear autonomous thinking.Early Protestants were champions of autonomy, and strove in their ownreligions to get rid of, or at least to simplify, all rites so as to promote aclearer focus on well reasoned beliefs. Contrary to the Protestant Ethic,and western religions in general, Durkheim argues that what gives bothreligion and society purpose and efficacy, and creates the moral forcesthat are their essence, or soul, are practices and not beliefs.

A final problem that the Conclusion poses for an understanding ofDurkheim’s epistemology is that there, as in the Introduction and thebook as a whole, Durkheim leaves his explicit discussion of the episte-mology for the end, Section iv. This has the effect, in all three cases,of creating the impression that the book is primarily about other things,particularly if the reader does not carry through to the end of each part.

It is ironic that, in spite of all the care that Durkheim took to develop hisargument step by step through the body of the text, providing exhaustivedetail at every point so that the empirical character of his argument shouldbe clear, it is the discussion in the Conclusion, of logic and language, itsconfusion with the epistemology, and the disproportionate attention itreceived, that seems to have done the most to create an idealist inter-pretation of the work. That he was dubbed an idealist for making anargument in which he challenged the distinction between realism andidealism by taking the position that empirically witnessable practices aremore important than beliefs, or ideas, is the crowning irony.

9.1.0 Section i: Practice/Real versus Belief/Ideal

In Section i of his Conclusion Durkheim reiterates at the outset, theimportance of his argument that practices are more important than beliefsand ideas. A mistaken emphasis on beliefs, Durkheim argues, explainswhy westerners look down on primitive religion as magic. Given thisbeginning, scholars of religion tend to treat rites as merely contingentmaterial manifestations of beliefs. However, according to Durkheim, ritesare not merely signs by which faith is outwardly expressed. Rites are themeans by which faith is created. Rites also give signs their meaning.All rites have a purpose in that they create moral forces not present inthe world before enacted performances created them and consequently,aboriginals are not wrong, or mechanistic, in thinking that enacting riteswill transform them.


The mistaken equation of religion with beliefs has led to many unnec-essary apparent conflicts, according to Durkheim, not only between reli-gions, but between the real and the ideal, between religion and science,between the individual and the collective, and between primitive andwestern peoples. These, he believes, can all be remedied by a propersociological view of religious practices. It is not hard to see how someof these claims would make Durkheim’s argument not only difficult, butunpopular, particularly with persons committed to the idea that the mostimportant thing about religion is beliefs and not practices.

According to Durkheim if a certain sort of practice is the cause of theidea of the sacred, or of causality, in one religion, then it must have a sim-ilar effect in another. He ([1912:594]1915:463;*1995:418) argued thatif so-called primitive religions consist primarily of their practices, thenthis is true also of the so-called more advanced religions: “it is incon-ceivable that the same effect could be sometimes due now to one cause,now to another . . . unless fundamentally the two causes were but one.”Durkheim believed that it was presumptuous to believe that only primitivereligions can be explained sociologically, and that the so-called advancedreligions have no social origin. Yet, this is what Europeans in Durkheim’stime, as well as in our own, generally did and do believe.

Here, as throughout his text, Durkheim asserts the equality of primitivereligion and primitive reason. Modern western logic and religion are, hesays, essentially the same as primitive logic and religion. His insistence onthis point must have infuriated his contemporaries, particularly thoseamong them who believed that western religious beliefs constituted arevealed truth, or those who were racist and believed in the superiority ofEuropeans, which in the Europe of 1912, just prior to World War I, wasan extremely prevalent belief.

9.1.1 Significance of the Single Case

In making his argument that the primary phenomena of religion are prac-tices and not beliefs, Durkheim opens his concluding chapter with theclaim that one thorough study of a single case is enough to establish ageneralizable fact about all religions. Given the contemporary emphasison large sample size, Durkheim’s reasons for focusing on a single case aresignificant. The argument has its origin in Durkheim’s understanding ofHume’s empiricism. If something cannot be found in a single case thenit must be concluded that it was added by the process of generalizationand not present in the original single case. Therefore, it is not only thatthe single case is sufficient to establish the general character and functionof religion, but that if it cannot be established through a single case itcannot be established at all.


In view of this constraint, Durkheim makes the provocative claim thathe can demonstrate what is fundamental about all religions through ananalysis of a single case study of Australian aboriginal religion. He arguesthat primitive religions have the same basic features as western Christianreligions, but that the western focus on beliefs has obscured this fact. Itis Durkheim’s position that all of the great ideas, and all of the princi-ple modes of ritual conduct, have been found to exist within Australianaboriginal religion: Totemism. Therefore, what he has found true ofTotemism is true also of western religions, in spite of the differencesbetween their systems of belief. He ([1912:594] 1915:463;*1995:418)elaborates the principle ideas that exist in both religious systems as “thedistinction between sacred and profane things, the ideas of soul, spirit,mythical personality, national and even international divinity; a negativecult with ascetic practices that are its extreme form; rites of sacrifice andcommunion; mimetic, commemorative, and piacular rites.”

While Durkheim argues that practices are more fundamental thanbeliefs, or ideas, he also argues that there are certain basic ideas, foundin all religions, those six which he demonstrated in the body of the text,that are always and must always be created by religious practices. There-fore, existence of these ideas can be treated as evidence that the neces-sary practices are being enacted. Because Totemism contains all of theseessential ideas it must therefore also contain the practices that createthem. Durkheim believes that what he has found through the study ofTotemism applies to all religions and religious practices everywhere andin all times.

Durkheim has made claims about the efficacy of a single case as thebasis for research several times over the course of the book. Here he reit-erates his argument ([1912:594] 1915:463;*1995:418) that “when a lawhas been proved by a single well-made experiment, this proof is univer-sally valid.” It is an interesting point, given the general interpretationof Durkheim as a proponent of positivist and quantitative methods. Hispoint is that unless social order can be demonstrated in a single instance,or a single case, it cannot be proved to exist. No matter how large thesample size, if the phenomenon can’t be found in each single case, thenit must have been added by the analysis. What Durkheim repeatedlyinsists on is that the ideas and practices he examines are not artificialconstructs, or mere conventions. Rather, it is necessary for some formof practice to produce just the emotional result that they do producewhich is the real basis for rational thought. Therefore, since it is necessarythat the practices for doing so be evident in the single case, and theymust exist, an empirical sociological proof based on a single case must bepossible.


A point on methods. Weber devised ideal types, followed later byParsons with unit acts and the idea of actors orienting toward values,because he did not expect to be able to demonstrate anything on thebasis of a single case. Both Weber and Parsons expected to have to cat-egorize instances of social order under an ideal type that these instancesapproached, but were never recognizable as in any actual case, an assump-tion that Garfinkel has characterized as “Parsons’s Plenum.” Durkheim’spoint is that if the researcher focuses on practices instead of ideas, theproblem changes in two ways: 1) practices, unlike ideas, can be seenand heard and, because they must be recognizable to other participants,they can be given detailed physical descriptions, and; 2) When personsenact practices, it is the physical aspects of practice, not any underlyingmeanings they may have that must be recognizable. It is also these wit-nessable characteristics that have efficacy in creating moral forces. Beliefsmay provide the impetus for gathering, but they cannot by themselvescreate moral forces. The similarities between Durkheim’s position andGarfinkel’s insistence that the orderly properties of members’ methodsmust be exhibitable in any actual case are obvious.

On this view, the methodological dilemma with regard to typificationand categorization that has seemed so intrinsic to sociology turns outto be a byproduct of not treating social orders as witnessable in thefirst instance, but rather, as primarily conceptual and interpreted ordersof experience.7 Obviously, interpretation constitutes a large and impor-tant dimension of social order. But, according to Durkheim, it is not theprimary dimension of social order, arising only after the fact in attempts toexplain social experience. Parameters of recognizability that are embed-ded in practices (as prospective orders) define categories of socialphenomena “naturally.” It is the analyst’s job to discover these, not toconstruct them.

9.1.2 Practices versus Beliefs

The distinction between practices and beliefs shaped the organization ofThe Elementary Forms into two books, one on beliefs and one on prac-tices, and ultimately constitutes both the key to Durkheim’s epistemologyand the foundation of his sociology. The problem, he says ([1912:594]1915:463;*1995:419) is that “most often, the theorists who have set outto express religion in rational terms have regarded it as being, first andforemost, a system of ideas that correspond to a definite object.” Weber’s

7 See Rawls 2002 for a discussion of the problem of typification in Goffman and Giddens,and Garfinkel’s solution to that problem.


sociology of religion, for instance, treats religion as a system of beliefsand ideas. Totemism is considered by Weber to be only magic because itconsists primarily of practices not beliefs.8 The focus of scholars is usu-ally on the correspondence between religious beliefs and their object. Inthis context, the correspondence of beliefs with what the theorist believes isthe truth about religious beings, becomes the standard of religious schol-arship. This means that the object itself, the deity and spirits, must betaken on faith and cannot be established empirically. Therefore, the reli-gious object, against which beliefs are supposed to be evaluated, is itselfonly an idea in the minds of the evaluators. Through circular thinking itpredictably turns out that only western style belief systems are consideredto be religions.

According to Durkheim ([1912:594] 1915:463;*1995:419) “rites haveappeared from this standpoint to be no more than an external, contingent,and physical translation of those inward states that alone were deemedto have intrinsic value.”9 Some western religious groups, the Quakersfor example, tried to eliminate rites altogether. This focus on beliefs,Durkheim argues, is why religion has often appeared to be incompatiblewith science. If religion is seen as a system of beliefs, then those beliefs conflictin essential ways with scientific theories. However, if religion is seen insteadas sets of practices that produce necessary social experiences which havea transforming effect on participants, and beliefs are understood as onlya secondary and contingent result of practices, then religion is no longerincompatible with science. In fact, on this view, religious beliefs andscience turn out to be the same thing, with early religious belief systemsserving as the first scientific cosmologies, an argument that Durkheimconsiders in detail in Section ii.

The prevalent western view of religion as consisting primarily of beliefsis not consistent with the commitment believers themselves have to

8 See Weber, Economy and Society. It is the fact that primitive rites seek to effect an imme-diate end or purpose through religious practices, with no corresponding religious beliefsystem, that led Weber to define primitive rites as magic and not religion. However, asDurkheim points out, all religion has an immediate purpose in creating an altered stateof being in its participants. Durkheim considers magical only those rites, the mysteriesof which are not shared with participants, that are performed only by the magician andhis assistants and not shared by the communicants. Only there, as he says, is there no“church.”

9 It is instructive to point out the similarity between this analysis of the problem andWittgenstein’s charge that it is the “picture theory” of language that is responsible forcreating the appearance of dilemmas in philosophy where there are none. Durkheim issaying that we assume that there are individual feelings that persons attempt to use signsto convey. What he argues is that the shared feelings come before and give meaning tothe signs. Not recognizing this has created the appearance that the problem of meaningis one of connecting signs (beliefs) with what they refer to, when in fact the problem isto understand the social context within which the sign has a recognizable meaning forparticipants: contexts of appropriate use – in detail.


their religious practices either. Durkheim ([1912:594–5] 1915:463–4;*1995:419) points out that believers know that religion is not primar-ily about ideas, but rather that “its true function is to make us act andto help us live.” Religion does not just give believers new ideas, it makesthem stronger and gives them new abilities. As a result of participation inreligious practices, according to Durkheim ([1912:594–5] 1995:463–4;*1995:419), the believer “within himself feels more strength to endurethe trials of existence or to overcome them. He is as though lifted abovethe human miseries, because he is lifted above his human condition.”The person feels stronger because they are stronger. Beliefs are merely asecondary effect of this direct action of practice on the person.10

An idea would not have this power to transform the person. Anidea can, according to Durkheim([1912:594–5]1995:463–4;*1995:419),“only release emotive forces that are already within us, neither creatingnor increasing them.” The emotive forces need to first be created by par-ticipation in enacted practices. Furthermore, in order for energy to enterinto people from the outside there must be some way in which it becomesa part of their internal makeup.

What is essential, Durkheim ([1912:595] 1915:464; * 1995:420) argues,is not that persons think, but that they act: “in short, we must act; andso we must repeat the necessary acts as often as is necessary to renewtheir effects.” “It is the cult that stimulates the feelings,” he ([1912:595]1915:464;*1995:420) argues, that produce the ideas that lead to thebeliefs in the first place. The cult is not just a system of signsor beliefs, according to Durkheim ([1912:595]1915:464;*1995:420,emphasis added), in terms of its practices, it consists of “the sum total ofmeans by which that faith is created and recreated periodically.”

Durkheim confronts the western emphasis on beliefs with the argumentthat religion consists of practices that give meaning to signs. The emphasison ideas, not only in studies of religion, but in sociology and philosophyin general, he argues, is misplaced. The essential thing to know is howpractices produce ideas – in detail.

9.1.3 Social versus Individual Causes of Religion

There has been a movement to “rescue” Durkheim by reinterpreting hisposition in a Pragmatist context. The problem is that while Durkheim and

10 Mike McCallion and David Maines have found in their research on the Catholic churchthat there are difficulties when beliefs/ideas are used as the reason for changing religiousrituals. People do not like their rituals changed. Recent changes with regard to beliefsabout and the placement of the tabernacle have been particularly problematic. When therituals do change the consequences cannot be controlled by beliefs and ideas. There areconsequences in terms of social solidarity and the production of moral forces that mayeventually make themselves felt in terms of changes in the beliefs themselves, but meetinitial resistance.


William James both grounded religion on real experiential antecedents,there are profound differences between their positions. For James, thePragmatist, the experiences in question are primarily individual. ForDurkheim, by contrast, religious experience begins with emotions thatcan only be created through collective enactment of practices and thuscannot be experienced by the individual alone. Durkheim ([1912:595]1915:464;*1995:420) acknowledged that he and James both accepted“that religious belief rests on a definite experience, whose demonstrativevalue is, in a sense, not inferior to that of scientific experiments, thoughit is different.” However, according to Durkheim ([1912:595] 1915:464;*1995:420), it does not follow “that the reality which grounds it shouldconform objectively with the idea believers have of it.” While beliefs mayhave empirical causes, there is no reason to think that those beliefs accu-rately represent their causes. The function of beliefs after all is not to rep-resent their causes, but to bring believers together to enact practices.11

Furthermore, because religious ideas are caused by empirically observ-able rites, and the essential transforming experience that makes per-sons human consists in the enactment of religious practice, the meth-ods of science can be used to study them. Durkheim argues ([1912:595]1915:464;*1995:420) that “to discover what the object [of religion] con-sists of, then, we must apply to those sensations an analysis similar to theone that has replaced the senses’ representation of the world with a scien-tific and conceptual one.” That, Durkheim says ([1912:595] 1915:464;*1995:420) “is precisely what I have tried to do.” Scientists recognized thelimits of individual perception and developed an experimental method.Sociology has to accept the limitations of beliefs and ideas and developan empirical method. In establishing Totemism as the origin of humanreason, Durkheim was trying to make sociology a science, by replacingthe study of ideas, with the study of social practices: not by making it“objective” in a positivist sense.

Durkheim follows Rousseau12 in arguing that without society personsare only animals and animals, as far as we know, do not have moralissues.13 Among animals, Rousseau argues, survival is the rule. It is soci-ety that raises man above himself because it is society that creates the

11 This is not to say that once formed, beliefs do not begin to dictate changes in practices.But, Durkheim would argue that if these changes in practice interfere with the essentialjob of the practices, to create reason and social unity, the changes will result in bringingabout the demise of that society and its religion. It is the practices that are essential, notthe beliefs.

12 See Rousseau ([1757]1999), The Origins of Inequality for the argument that moralitybegins with society. See also 10.1.4 for a discussion of the relationship between Durkheimand Rousseau.

13 It would also follow that to the extent that animals do have what could properly be calleda society they also have moral issues.


ability to reason. When Durkheim writes of “society” in this regard, I takehim to mean enacted social practices, not an amorphous ideal entity calledsociety. This understanding is underscored by Durkheim’s referencesto society as a field of action. According to Durkheim ([1912:597–8]1915:465;*1995:421, emphasis added) “society cannot make its influencefelt unless it is in action, and it is in action only if the individuals whocomprise it are assembled and acting in common. It is through com-mon action that society becomes conscious of and affirms itself; societyis above all an active cooperation.”

Unfortunately, Durkheim’s references to society as comprised of move-ment and action, are sandwiched between other statements that could betaken to stress the conceptual. For instance, Durkheim makes a statementregarding the relationship between “intellectual goods” and “society”that could be given an idealist interpretation. According to Durkheim([1912:597–8] 1915:465;*1995:421), “what makes man is that set ofintellectual goods which is civilization, and civilization is the work ofsociety.” While the phrase, “intellectual goods” in this context could beinterpreted as referring to ideas, Durkheim is emphasizing the role ofpractices, not ideas in this section. By society he means social groups inaction. Immediately following this passage he writes that ([1912:597–8]1915:465;*1995:421, emphasis added) “even collective ideas and feel-ings are possible only through the overt movements that symbolize them.Thus it is action that dominates religious life, for the very reason that societyis its source.”

Even though action dominates religious life and rites are primary,beliefs play an essential role. Because the moral feelings created by collec-tive action are available to self-reflection only when represented in somefashion, moral forces could not function without taking material form:representing themselves materially as beliefs and symbols. However, thesesymbols borrow their essential element, feelings of moral force, from theexperience of enacted practices, and their material representations arewhat is most superficial about them.

9.1.4 Symbols

While collective action creates social or moral forces that are real, they canonly be felt in the moment. Material, or witnessable aspects of collectiveaction are needed to recall those feelings when the group is not produc-ing them. Symbols allow for this. Symbols of felt moral forces becomean important focus of religious belief, Durkheim argues ([1912:597–8]1915:465;*1995:421), “probably because collective feelings becomeconscious of themselves only by settling upon external objects, those veryforces could not organize themselves without taking some of their traits


from things.” Material aspects of rites, even movements, thus can becomesymbols of collective ideas.

However, as symbols they are ultimately only representations of thefeelings generated by the same movements and objects during the per-formance of the rite. The problem is that while they originally comefrom material and emotional aspects of practice, symbols come to beassociated with the narrative religious explanations of where those feel-ings come from, which inevitably follow the successful enactment of acollective practice. Thus, symbols come to be associated with ideas, eventhough in the first instance their efficacy as symbols came only from the factthat as sounds or movements, or as things, they were part of the enactedpractices.

In this relationship between feelings and symbols it is essential to notethat the feelings precede the symbols. The purpose of the practice is toproduce feelings that are called up later by the symbols. The only ideasthat are not secondary to feelings are the categories of the understandingand that is because they are the feelings produced by the practices andonly come into being simultaneously with, as a direct effect of, enactedpractices.

Unless it is understood that religious practices are valued for the trans-forming effect they have on their human participants, religion will appearto be the same as, but not as accurate as, science. According to Durkheim([1912:599–600] 1915:467;*1995:422), religious practices can be madeto look like a chemical reaction, “just as today a body is placed in contactwith a source of heat or electricity in order to heat or electrify it. The pro-cedures used in the two cases are not essentially different.” But, while the“procedures” may not be different, the results of religious practices donot work by transferring mechanical forces from one object to another.Religious practices actually create moral forces that would not otherwiseexist. These forces then can be felt by all the participants.

The enactment of practices by an assembled collective creates a newsocial state of being. The participant feels this transformation and theirfeelings become objectified in religious, or moral forces and the reli-gious symbols and ideas that call them up. Totemism is not, accordingto Durkheim ([1912:599–600] 1915:467;*1995:422) materialistic, eventhough its essence is empirically evident. The idea is ([1912:599–600]1915:467;*1995:422) “not to exert a kind of physical constraint upon . . .imaginary forces.” That would be science, or magic. Rather, heargues ([1912:599–600] 1915:467;*1995:422), the function of religiouspractices is “to reach, fortify, and discipline consciousness.” More thanthis, the point is to create the ability for consciousness to be rational andfor rational consciousnesses to engage in intersubjective communication.


Because the function of religious practice aims at the creation of intellec-tual life, all religions are in this sense spiritualistic, whether or not theirbeliefs include spirits.

9.1.5 The Is versus Ought Dilemma

Durkheim’s argument that human reason and morality have their originsin mundane social practices involves him in a controversy that has cometo be known in philosophy as the “is versus ought” problem. The questionis whether an imperfect source can be the origin of ideas of perfection. Inmoral philosophy the argument goes as follows: something that is imper-fect, that is, human social life, cannot give persons the idea of what oughtto be; that is, perfect morality.

Durkheim ([1912:599–600] 1915:467;*1995:422) poses the dilemmain the form of the question whether it is “the real society, such as it existsand functions before our eyes” that allows society to be the substrate, orcause of religious life? Then he asks whether it would need to be in someideal capacity that society would play this role. Would the idea of per-fection and perfect beings, perfect morality, he asks ([1912:599–600]1915:467;*1995:422) not have to come from “the perfect society, inwhich justice and truth reigned, and from which evil in all its forms wasuprooted?”

According to Durkheim, the problem is that the question itself assumesa false dichotomy between real and ideal. The apparent conflict, hesays, disappears if concrete practices are given priority. It is the realsocial world, according to Durkheim ([1912:601] 1915:468;*1995:423),not an ideal world, that finds itself reflected in religion: “to see onlythe idealistic side of religion is to simplify arbitrarily. Religion is real-istic and every ugliness is to be found in it.” It is not only primitivemythologies that distort the underlying reality so that it is hard to rec-ognize, all religious systems of belief are equally distorting. But, hesays ([1912:601] 1915:468;*1995:423), it is also “true that even if themythologies and theologies allow a clear glimpse of reality, the reality wefind in them has been enlarged, transformed, and idealized.” The sourceof this idealized side of religion is quite literally the source of the idealsthemselves; enacted practices. It is the real that creates the ideal.

Because the real creates the ideal, Durkheim argues, the apparent con-flict between the real and ideal is a false one. Since experiences of nec-essary force and group unity are essential to reason and morality thepractices that create them and the enacted moments of their creationmust occur in all societies. Although the essential moments in these prac-tices exist necessarily in the context of imperfect societies, the moments


themselves can be experiences of perfect unity or reciprocity, both ofwhich are essential to our conception of morality.

Because the real versus ideal distinction is a false dichotomy, when itis taken seriously it becomes impossible to explain how the capacity toidealize could have come from the real. But, if we realize that practicingthe real inevitably leads to creating ideals, the question becomes moot.Creating moral, or social forces, creates ideals. If persons had not createdideals, they would not be social beings. This is not historical materialism,as Durkheim will point out at the end of the section. Moral forces arenot a mere reflection of their material base, but rather a transformation.Something entirely new and ideal is created through religious practice.

Finally, Durkheim argues that it is the ideal that constitutes the under-lying condition of the real. In other words, there are basic human needs,that we usually think of in terms of ideals, but which are in fact functionalprerequisites for cooperative social life and intelligibility. Because theseneeds must be met, they define the underlying conditions of the real.

It is essential to Durkheim’s position that this process of creating idealsis not something that society makes artificially. The ideal society is anexpression of real society, not an arbitrary fantasy. It is a direct result ofenacted practice and the need for such practices to be enacted. Accord-ing to Durkheim ([1912:603–4] 1915:470;*1995:425), “when we set theideal society in opposition to the real society, like two antagonists sup-posedly leading us in opposite directions, we are reifying and opposingabstractions.”

It is interesting that Durkheim argues that the distinction between theideal and the real is, in the case of social phenomena, a reification, becausethe general analysis of Durkheim’s argument, makes use of these conceptsin what Durkheim would have called a reified form. Similarly, theorytextbooks continue to use the terms in just the way that Durkheim arguedagainst. For Durkheim, while the real creates the ideal, the need for creatingbasic common ideas is an underlying condition, or function of the real (enactedpractices).

Furthermore, as by-products of the enactment of social practices, idealscould not be studied separately from the practices that created them. Afocus on the study of symbols, as symbols, necessarily obscures both theproblem of meaning and the problem of social order. What is necessaryis a careful study of the use of symbols in a context of practice.

9.1.6 Origin and Function of Religious Narratives

Durkheim argues that before we accept the capacity to make up ideals asan unexplainable mystery, we should be sure that it cannot be explained


empirically. If it is not necessary to accept a mysterious and innate originfor the capacity to idealize, then we should not accept such an origin.14

Durkheim ([1912:602] 1915:469;*1995:424) then offers “my proposedexplanation of religion” as having “the specific advantage of providing ananswer to this question.” The key to explaining the origin of the capacityto create ideals is the same thing as explaining the origin of the idea ofthe sacred. This is the case, he argues ([1912:602] 1915:469;*1995:424),since what defines the sacred is that the sacred is added to the real.In explaining the origin of the sacred, Durkheim singles out collectiveexperience; the state of effervescence, or moral force, which creates theexperience of the categories. This, he has argued throughout the book,is the origin of the idea of the sacred; the first classification; and there-fore, it follows, the origin of the first ideal. The practices that create thestate of effervescence also, thereby, create both the first empirically validcategories and the capacity to form new arbitrary and conventional ideals.

In a collective state of effervescence, according to Durkheim ([1912:602] 1915:469;*1995:424), “man does not recognize himself; he feelssomehow transformed and in consequence transforms his surround-ings.” Because he feels transformed he makes up stories, Durkheim says([1912:602] 1915:469;*1995:424), “to account for the very particularimpressions he receives,” and in doing so “he imputes to the things withwhich he is most directly in contact properties that they do not have.” Inthis way, the participant in enacted practices, in their attempt to accountfor the transformation that they feel in themselves, creates a world ofideals that did not exist before. “In short,” Durkheim says ([1912:602]1915:469;*1995:424), “upon the real world where profane life is lived,he superimposes another that, in a sense, exists only in his thought, butone to which he ascribes a higher kind of dignity than he ascribes to thereal world of profane life. In two respects, then, this other world is anideal one.”

It is important to note that the distortion thus created by ideals is aneffect of narratives and beliefs made up after the fact by participants toexplain the transforming effects on them of participating in enacted prac-tices. The distortion is not an effect of the categories, which are part ofthe transforming effect of the rites. Because the purpose of myths is toexplain these feelings, the criteria for what is acceptable will not includethe requirement that the myth accurately portray the functions of theritual.

14 Here Durkheim follows a philosophical principle originating with William of Ockhamin the 1400s that the simplest of two arguments is to be preferred, all else being equal.The principle is known as “Ockham’s razor.”


Persons participate in the social world concretely. The experiences thatthey get from this concrete participation, however, seem to them not tocome from the concrete world. These experiences are emotions. As cate-gories they are ideas. They are also ideals. Even though they come fromthe real, they seem as though they must have come from an ideal source.This is so because when persons are by themselves and do these samethings they do not get these feelings. So, in order to explain feelings thatare generated in a population that enacts identical sounds and movementstogether, persons invent an ideal world that they believe moves them inter-nally in this way. They then superimpose this world of sacred ideas uponthe profane world of daily life. In this way the sacred appears to takeprecedence over and define the concrete profane world. But, it was theconcrete world of enacted practices that created the sacred world in thefirst place.

Thus, while persons experience an ideal world, they do so because theylive and participate together in a concrete world that is available for exam-ination. According to Durkheim ([1912:602] 1915:469;*1995:424), “theformation of an ideal is by no means an irreducible datum that eludes sci-ence. It rests on conditions that can be uncovered through observation.”In assembling and concentrating, members of the social group createa dimension of experience that does not otherwise exist. This dimensionis real, but only exists if participants assemble to create it. After the fact, itrequires symbols to recall it. Because the experience of moral force gen-erated by enacted practice is so different from ordinary experience it isexpressed as an ideal, even though it is in fact real and observable. As a conse-quence, Durkheim ([1912:603–4] 1915:470;*1995:425) says, “A societycan neither create nor recreate itself, without creating some kind of idealby the same stroke.”

9.1.7 Society as Idea and the Sociology of Knowledge

While concrete acts create ideas, the evolution of ideals once theycome into being, becomes increasingly detached from the real basisof the original ideas. In this process of detachment, conflicts some-times develop between ideals. According to Durkheim ([1912:603–4] 1915:470;*1995:425) sometimes society “feels pulled in all directions.When such conflicts break out, however, they are not between the idealand the reality, but between different ideals, between the ideal of yester-day and that of today, between the ideal that has the authority of traditionand one that is only coming into being.” The study of these conflicts inthe evolution of concepts over the course of history is a study of relationsbetween ideals. It is this area of study on which Durkheim’s argumenthas had the greatest influence.


While Durkheim is arguing that the evolution of ideas can and shouldbe studied, he also argues that conflicts between concepts are not realconflicts, but only apparent conflicts that result from the increasing distancebetween ideas (accounts of practices) and the practices themselves, overtime. By treating these conceptual conflicts as real, and not studying themconcretely, we reify social differences that do not really exist. Even thoughthe ideals first had their origin in the real, as ideals they are only accountsof the real, they distort their origins, and can only be studied as accounts.The solution to such conflicts is to return to the study of those concretepractices that create the ideas.

While Durkheim argues that beliefs and ideals should be studied intheir own right, they must be studied in a different way from practices.They do not obey the same laws as the real. Beliefs have a life of their ownand must be studied as ideas. This is where the sociology of knowledgeprovides a context for the study of the laws of the evolution of ideas,and the forces, or effects, of beliefs and ideas. This would be a study oflanguage and symbolic exchange similar to that carried on by symbolicinteraction, postmodernism, and cultural sociology today.

Beliefs and ideas as symbols in a language system do not have an originin a direct experience of moral forces as do the six categories of the under-standing. Therefore, they must be learned in a social context. Durkheim([1912:603–4] 1915:470;*1995:425) begins, at this point in the text, tospeak in terms of an “assimilation” of ideas: “It is by assimilating theideals worked out by society that the individual is able to conceive ofthe ideal.” The language suggests that ideals somehow exist in societyand that by joining society the individual “assimilates” them. Durkheim([1912:603–4]1915:470;*1995:425) goes on to say that “it is in the schoolof collective life that the individual has learned to form ideals.”

Durkheim creates a problem for himself with the terms used here.The “school of collective life” could be either enacted practices, or aconceptual world view, or both. It is important to note, therefore, thatthese statements occur in a section in which Durkheim is reiterating hisargument against an innate origin for the categories of the understanding.Durkheim is emphasizing a social, versus an individual, origin of collectiveideas, in an attempt to deny the claim that he accepts Kant’s innatistposition.15 He is not attending to the possibility that he will be interpretedas arguing for an ideal versus a real origin for ideas, he has just dismissedthe distinction between the ideal and the real as a false one.

In the discussion of language that follows, Durkheim ([1912:603–4]1915:470;*1995:425) returns to his emphasis on the concrete: “It is

15 Kant argued that all religions express in a popular form what the innate rational mindwould decide on the basis of pure reason. The argument is based on the idea that allhuman beings are born with an innate capacity to reason.


society that, by drawing him into its sphere of action, has given him theneed to raise himself above the world of experience, while at the sametime furnishing him the means of imagining another.” What is importantabout this passage is that it stresses the “sphere of action,” rather thanthe more passive “assimilation,” as the source of “the need to raise him-self above the world of experience.” Social solidarity and intersubjectivecommunication, require symbols and ideals that make possible the repro-duction of unity, intelligibility, and reciprocal relations with others. As ananimal the individual has no such needs.

The need to think is the cause of social practices; their underlyingcondition of existence. According to Durkheim ([1912:605] 1915:471;*1995:426) “This faculty is not a sort of luxury, which man could dowithout, but a condition of his existence. If he had not acquired it, hewould not be a social being.” Reason is not the condition of the existenceof the biological being, nor an innate faculty of the mind, but rather ofhuman social being and the result of participation in enacted practice.16

Thus, the ability to idealize, which develops as a way of first focus-ing and later expressing moral feelings around the image of the totemicanimal during religious rituals, becomes a necessary condition of man’snew existence as a rational social being. To create that capacity is theunderlying purpose, or function, of religious practices.

9.1.8 Historical Materialism versus Social Transformation

Although Durkheim ([1912:605] 1915:471;*1995:426) argues that ideashave an empirical origin, he explicitly distinguishes his position from his-torical materialism saying that, while he focuses on the real, his positionshould not be understood as “merely a refurbishment of historical mate-rialism.” This is a reference, if not to Marx, to materialist interpretationsof Marx current in Durkheim’s day.17

16 In calling the ability to idealize a faculty, Durkheim may have misled those who werelooking for Kantian, or innatist elements in his position. By “faculty” Durkheim meansan ability to think in concepts which depends on a prior acquisition of the categoriesthrough participation in enacted practice. His argument that this faculty is a necessarycondition of human social existence may have confused the issue by making it seemas though as a necessary condition it must have come before “social existence.” Butwhat Durkheim means by necessary conditions are underlying functions and purposes, notantecedent causes or innate capacities.

17 Durkheim was familiar with the arguments of Marx and those whom he referred to as“the socialists.” He had outlined a project, before writing his dissertation, that wouldhave culminated in a systematic critique of socialism. The Division of Labor was the firstpart of this project. Although some additional lectures were given, the project remainedessentially unfinished. However, it is clear from what he did write (published in English asSocialism and Saint-Simon ([1895–6]1958)) that he was concerned with the tendency to


Durkheim says ([1912:605] 1915:471;*1995:426) “I in no way meanto say that religion simply translates the material forms and immediatevital necessities of society into another language.” On the other hand,Durkheim does insist that social life is real: “I do indeed take it to beobvious that social life depends on and bears the mark of its materialbase.” But, what Durkheim argues is not that materiality directly shapesideals, or categories of thought, which then would be mere effects, orrepresentations, of the real. Rather, he argues that real material practicescreate something completely different, a transformation that does not referback to its origin, and appears to be completely different from the ritethat created it.

Partly, this is because the real, as Durkheim speaks of it, is a social andnot a natural real. While sounds and movements are in a sense concrete,they only have efficacy as social and not merely as material entities. Itis the social element of the practices that is transforming. They createhuman beings out of individual biological beings.18

This synthesis does not just represent those material relations, it createsa whole new world of ideas. The result of the social synthesis to whichDurkheim alludes ([1912:605] 1915:471;*1995:426) is “a whole world offeelings, ideas, and images that follow their own laws once they are born.”The conceptual ability that is created in this way, he says ([1912:605]1915:471;*1995:426), “enjoys such great independence that it sometimesplays about in forms that have no aim or utility of any kind, but only forthe pleasure of affirming itself.” These processes can be studied in theirown right, he argues, and are also properly sociology. Here Durkheimadvocates a sociological study of ideas and the evolution of their relations,and this part of his argument has been embraced by contemporary socialtheory (see for example Alexander and Seidman 1990). Parsons, who waswedded to a more materialistic and individualistic vision of social order,could not, and did not, appreciate what Durkheim was arguing here.

either explain everything in terms of material relations or in terms of ideas, a dichotomythat he rejected. Social relations have material aspects that are essential to them. But,what is most important about these material relations, for Durkheim, is that they createideal ones.

18 This would distinguish Durkheim’s position from some historical materialist positions,but not necessarily from Marx. Like Durkheim, Marx built on Rousseau and the trans-forming effect of social relations on human beings. See especially references to Rousseauin the beginning of the Grundrisse. In fact, most of what Marx considered to be mate-rial, class relations and relations of production, for instance, are in fact social relations.Thus, Durkheim probably would not have seen Marx as a materialist, if he read his workin the original (particularly his early work) but rather, like himself, as focused on thematerial, or empirical, aspects of social relations. There is some evidence in Book III ofThe Division of Labor that, in criticizing socialism, Durkheim largely exempts Marx (seeRawls 2003).


9.1.9 Pragmatism and Religious Individualism

Durkheim concludes Section i with a reconsideration of the question ofPragmatism and radical individualism and asserts not only a social originfor religion, but also the necessity for some degree of religious univer-salism.19 According to Durkheim ([1912:607] 1915:472–3;*1995:427),if radical religious individualism has remained unrealized it is “becauseit is unrealizable in fact.” He does not deny the possibility of individ-ual contemplation. But, he denies that it can take the place of religiouspractice. If religions were essentially a matter of beliefs, then, radicalindividualism would be possible. But, since religious practice is funda-mentally about generating moral forces it must be collective in character.According to Durkheim ([1912:607] 1915:472–3;*1995:427) “The onlyhearth at which we can warm ourselves morally is the hearth made by thecompany of our fellow men.”

Even if the world were really created and managed by the sorts ofbeings that the various religions claim, he says ([1912:607] 1915:472–3;*1995:427), “if they are to have the useful influence over souls that istheir raison d’etre, we must believe in them. The beliefs are only at workwhen they are shared.” It is not enough for spiritual beings to exist andfor individuals to believe in them. It is necessary for the group to assembleto put beliefs to work. The modern conception of an individual contem-plative religion does not do this. According to Durkheim ([1912:603–4]1915:470;*1995:425) “To be sure, collective ideals tend to become indi-vidualized as they become incarnate in individuals. Each person under-stands them in his own way and gives them an individual imprint, someelements being taken out and others being added.” But, the ability towork with ideals, however much it has the capacity to be individualized,does not begin with the individual. It depends on the prior existence of acult that puts beliefs into action.

Durkheim ([1912:609] 1915:474;*1995:428) argues that there is a col-lective cult that is becoming international: “There is no national life thatis not under the sway of an international collective life. The more weadvance in history, the larger and more important these internationalgroupings become.” The limiting condition on the degree of individ-ual difference for collective ideals would be the recognizability of theidea as used by the individual in communication with others.20 Thatis, conceptual forms can change, or individuate only within the limits

19 This is probably a reference to James. But, it would also apply to the post-ProtestantReformation tendency to treat religion as a matter of individual faith.

20 Recognizability is a term used by Garfinkel to refer to the witnessable aspects of ordersof practice, that constitute their identifiability as particular forms of practice.


of recognizability, if they are to remain communicable. The success ofinternational groupings requires thousands of sounds and movementsto become broadly recognizable for communication across groups to besuccessful. This leads to a convergence of collective representations indifferent societies.

In this first section Durkheim can be seen to address the same problemof religious conflict (in Europe) that Enlightenment moral philosophershad been trying to solve by advocating the separation of church and state,and providing a secular blueprint that all could abide by. For his part,Durkheim argued that the religious beliefs that caused all the conflictwere really only derivative phenomena anyway and not what religionwas all about. If the universality of the function of religious practice wererecognized, the differences could be resolved.

9.2.0 Section ii: Religion and Science

In Section ii Durkheim focuses on the relationship between religion andscience. Given the argument of Section i that religion is essentially a mat-ter of rites and not beliefs, Durkheim asks what the relationship betweenreligion and science could be? In his day religion and science were clearlyseen to be in conflict, offering competing explanations of both historyand morality. According to Durkheim the appearance of conflict is basedon the mistaken idea that religion is primarily a system of beliefs. But,religious rites and citizen’s meetings, he argues, serve the same func-tion. Religion is primarily a system of action and, as such, science cannotreplace it. It is only the speculative function that escapes from religioninto science. Ironically, modern religion is different from religions of thepast because it must begin with science. But, it is also the same, because,as a system of action, it goes beyond science and performs a necessaryfunction that science cannot.

9.2.1 The Essence of Religious Practice Must Outlive Beliefs

Durkheim argues that in order for society and human reason to con-tinue, the fundamental elements of religious practice must outlive the sym-bols and beliefs involved in religious thought. Collective feelings and ideasneed periodically to be recreated. This is so, he ([1912:610] 1915:474–5;*1995:429) says, because “There can be no society that does not experi-ence the need at regular intervals to maintain and strengthen the collectivefeelings and ideas that provide its coherence and its distinct individuality.”Social solidarity and human reason both depend on the constant mak-ing and remaking of society and the social participant through enacted


practice. According to Durkheim ([1912:610] 1915:474–5;*1995:429),“This moral remaking can be achieved only through meetings, assem-blies, and congregations in which the individuals, pressing close to oneanother, reaffirm in common their common sentiments.” Whatever formreligious beliefs take, and however much they change, the essential func-tions of religious practice, in putting people into collective movement,must be preserved.

In making the argument that practices, because they are the essenceof religion, must continue, Durkheim compares citizens, in the Europeof his time, commemorating an event in their national life, to princi-ple religious observances of the Christian and Hebrew calendars. Heasks ([1912:610] 1915:474–5;*1995:429) “what basic difference there isbetween Christians’ celebrating the principal dates of Christ’s life, Jewscelebrating the exodus from Egypt . . . and a citizen’ meeting commem-orating the advent of a new moral charter or some other great event ofnational life.” The function of the practices is the same. The differencesbetween the beliefs are irrelevant as long as the practices are enacted. Asfunctions that were previously religious become secularized, the essentialexperiences produced by the enactment of religious practice live on inthem.

According to Durkheim, the current phase stands at a developmentalpoint between the religions of the past and a new type of religion thatwill emerge in the future. “If today we have some difficulty,” he says,([1912:610] 1915:474–5;*1995:429) “imagining what the great feastsand ceremonies of the future will be, it is because we are going througha period of transition and moral mediocrity.” Caught between, whatDurkheim referred to in The Division of Labor [1893]1933 as, mechan-ical and organic forms of solidarity, social forms have lost the abilityto move their participants to former levels of moral unity.21 Past reli-gions espoused ideas that no longer excite us. According to Durkheim([1912:610] 1915:474–5;*1995:429) this is either because “they havepassed so completely into common custom that we lose awareness ofthem or because they no longer suit our aspirations.” In the first case, herefers to the progressive loss of ritual that has characterized western soci-ety since the Protestant Reformation and the industrial revolution. Thesecond refers to early Christian ideas of equality and brotherhood that

21 Many scholars interpret Durkheim to mean that modern society is an organic solidarity.This is clearly wrong given Durkheim’s criteria for organic solidarity as specified inChapter Two on “Forced Labor,” in The Division of Labor. These criteria include thecondition that there can be no inherited wealth or privilege, a condition that clearly hasnot been met in any society. (See Rawls 2003 for an extended discussion).


are no longer egalitarian enough, or democratic enough, to be consistentwith the ideals of a pluralistic modern society.

“Meanwhile,” as Durkheim ([1912:611] 1915:476;*1995:430) says,“no replacement for them has yet been created.” To the degree that soci-ety still functions some practices must continue to be enacted in secu-lar form. However, Durkheim also argues in Suicide ([1897]1951), thatmodern society is characterized by a state of anomie, in which the neces-sary connections, or bonding, between persons and between person andgroup, do not occur often enough and are not strong enough to sustaineither the integrity of individual selves or the solidarity of social forms.This, he argues, produces a general state of anomie in the populationand leads to a much higher rate of suicide than in societies that providean optimum degree of bonding.22 Thus, while practices must provide anecessary minimum of social solidarity and shared ideas in order for soci-ety to continue existing, solidarity may be at a bare minimum in modernsociety, as it currently exists.

Durkheim([1912:610] 1915:474–5;*1995:429) predicts that, this con-dition of uncertainty and moral mediocrity in modern society will not last:“A day will come when our societies once again will know hours of cre-ative effervescence during which new ideals will again spring forth andnew formulas emerge to guide humanity for a time.” This will happenbecause it is necessary that it happen in order for society to continue.These occasions of the future, he says ([1912:611] 1915:476;*1995:430)will be recreated, just like those of the past, “to relive them in thoughtfrom time to time.” Durkheim gives the example of the French Revolu-tion as a secular faith that was once the object of a cult.

While Durkheim ([1912:611] 1915:476;*1995:430) says that it“exceeds human faculties” to predict whether the religions of the futurewill “better suit the reality to be expressed,” than current religions, healso seems to be saying that the religion of the future will be one thatis based on, and reinforces, the idea of human equality, at least bet-ter than current religions. This would be consistent with his argumentin Book III of The Division of Labor that an advanced division of laborrequires justice as a functional prerequisite. He also suggests a progres-sive development toward religious universals, and human equality. Thiswill be caused by increasingly international relations between societies,and increasing individuation of persons within groups, both the result of

22 Durkheim also points out that too much bonding can lead to suicide, as in traditionalsocieties that require ritual suicide for dishonoring the group, or in what Goffman wouldhave referred to as total institutions, the military, gangs, prisons, in which persons willchoose to die for the good of the group.


an increasing division of labor. He also suggests, at the end of Sectionii, that religion will, in the future, be better informed by science (and byscience he probably means sociology) as to its purpose, or function.

9.2.2 Religious Cosmologies

While the essential function of religion in creating unity and sharedideas depends in the first instance on enacted practices, there are otheressential functions of religion that rest on its character as a system ofideas. Practices, according to Durkheim, are turned toward action, whilebeliefs and ideas are turned toward thought. Practices satisfy the needsof collective action and collective communication: universal and timelessneeds. Ideas correspond to the need to explain the relationships betweenthings in the world around us. Since their function is different, ideasmay correspond to needs that are not as universal as practices, needs thatbelong to specific historical periods. According to Durkheim ([1912:611]1915:476;*1995:430), “Since they do not rest on the same conditions,then, there is reason to ask whether ideas correspond to needs as universaland as permanent as the practices do.”

Consequently, the future of religious beliefs must be considered separatelyfrom the future of religious practices. Durkheim ([1912:611] 1915:476;*1995:430) argues that if we take the object of religious ideas to be anexplanation of the supernatural, we will refuse to believe that the spec-ulative function of religion is only secondary and can “be overthrown.”But Durkheim does not think that religion explains supernatural objectsand relationships. If it did, then there would be no way of evaluating howwell any religion expresses its object, because there would be no way ofevaluating the correspondence between idea and object.

If, however, we accept Durkheim’s proposition ([1912:612–13]1915:477;*1995:431), that religious ideas are really about “nature, manand society. The mystery that appears to surround them is entirely super-ficial and fades upon closer scrutiny.” If religion is about social relations,and its purpose is to motivate persons to enact practices, then the reali-ties of religion are similar to the realities of science, and the correspon-dence between idea and reality can be evaluated.23 Both religion andscience seek to clarify relations between persons and the world they live in.“Both,” he says ([1912:612–13] 1915:477;*1995:431), “pursue the samegoal.” Even the “essential notions of scientific logic” he says ([1912:612–13] 1915:477;*1995:431), “are of religious origin.”

23 Durkheim is not arguing here that the words have meaning through a relationship ofcorrespondence, but rather that there is a relationship between the ideas, and the practicesthat created them.


While science aims toward an objective truth, and perfects its methodstoward that end, the aim of religion in explaining relations between thingsin the world remains essentially the same as it was in earlier religiouscosmologies: to keep people performing the rites. “Hence” Durkheimargues ([1912:612–13] 1915:477;*1995:431) “it seems natural that reli-gion should lose ground as science becomes better at performing its task.”What science cannot do is to replace the essential ritual functions of reli-gion that are its essence.

Durkheim is approaching a controversial issue. The idea that religiousbeliefs are a crude form of scientific thinking will be difficult for mostpeople to accept. Similarly, the argument that primitive rites are moreimportant than “refined” modern beliefs will likely not sit well with awestern audience. But, given that rites are always more important thanbeliefs, a religion focused on beliefs is, as he says, something of a problem.

Ironically, according to Durkheim, it was this speculative focus ofmodern religions that made room for the development of science inthe first place. He argues that while primitive religions explained theworld and the cosmos, Christianity allowed natural science to gain instrength, because Christianity considered the material world to be pro-fane. But, while science has been given increased freedom, the worldof souls is still considered sacred. Any science, like psychology, or hisown sociological theory of religion, that seeks to explain soul, or theintellectual life, in empirical or scientific terms, will still run into trou-ble with religion. “Hence,” he says([1912:612–13] 1915:477;*1995:431)“the strong resistance one encounters whenever one attempts totreat religious and moral phenomena scientifically.” Of course thisis exactly what happened to the epistemological argument of TheElementary Forms.

9.2.3 Science is Conceptual, Religion is Action

While religious beliefs have only at best an allegorical truth, through thecomplex relationship in which they stand as accounts to practices, thepractices have immediate truth. Furthermore, unlike science, religionis primarily a system of action, that is necessary for both human beingand social life. According to Durkheim ([1912:613] 1915:477;*1995:432,emphasis added), “insofar as religion is action and insofar as it is a meansof making men live, science cannot possibly take its place.” The essentialfunctions of the practices cannot be replaced by a form of thought.

According to Durkheim([1912:613] 1915:477;*1995:432), “What sci-ence disputes in religion is not its right to exist but its right to dogma-tize about the nature of things.” Because of the advancement of science,


religion will be different in the future. Religion already has to complyto some extent with scientific beliefs. But, the secularization of religiousideas is not the same thing as the secularization of the functions of religionitself. While these may become increasingly detached from the symbolsand ideas that have characterized religion in the past, because these prac-tices create moral human life, they will always have a spiritual, and thus,a religious character. Science in the future can contribute justificationsto the practice of religion, instead of conflicting with it.

9.2.4 Faith Must Rush Before the Truth

The problem is, according to Durkheim, that people will not practicea cult for which they have no beliefs. He argues ([1912:613] 1915:477;*1995:432) that, “To spread or simply to maintain religion, one mustjustify it.” Therefore, finding a way for people to believe something thatwill commit them to the practices is essential. This requires cooperationbetween religion and science. But, while science can refine justificationsof practice over time, religion cannot wait. The purpose of faith is actionnot truth. He argues that ([1912:613] 1915:477;*1995:432) “Theorieswhose calling is to make people live and make them act, must there-fore rush ahead of science and complete it prematurely.” Thus, reli-gious faith will always go before science and because of this, he says([1912:615] 1915:478;*1995:433), “obscure intuitions of sense and sen-sibility often take the place of logical reasons.”

On the other hand, religious faith in a contemporary scientific con-text cannot clash with science. After the development of science, faithmust begin with science ([1912:615] 1915:478;*1995:433) “while exer-cising the right to go beyond science, it must begin by knowing anddrawing inspiration from science.” This will make religions in the futuredifferent from religions of the past. According to Durkheim ([1912:615]1915:478;*1995:433), “There rises a power before religion that, eventhough religion’s offspring, from then on applies its own critique andits own testing to religion.” Here, as in the 19th lecture on pragmatism([1913–14] 1955), Durkheim seems to be suggesting that religious beliefswill become truer over time as they continue to be informed by scientificadvances, but that at the level of practices, religion has been absolutelytrue from the very beginning.

9.3.0 Section iii: The Social Origin of Logic and Language

Having established the relationship between science and religion inSection ii, and having said that science came from religion, Durkheim


then asks how religion could have given rise to science? While Durkheimhas alluded to an argument for the social origins of logic, throughout thetext, and hinted at what his argument will be in various places, the actualdiscussion of the social origins of logic appears only here. Durkheimargues that what science gets from religion is both cosmology and logic,both content and method. But, the relationship between religion andlogic, poses a puzzle. He has argued that religious cosmologies arise toexplain social realities and that as retrospective accounts they distort thatreality. Why and how could such distorting cosmologies have given riseto conceptual logic?

According to Durkheim ([1912:617] 1915:480;*1995:433) “nothingpredisposed society for this role, it would seem, since it is obvious thatmen did not come together for the purpose of satisfying speculativeneeds.” Religion has produced something, that is, logic, that was notits original purpose. However, logical thought so conceived is essential tosociety. As soon as society develops, logic develops. Therefore, logic andconceptual thinking are as old as the human social being.

That Durkheim’s arguments with regard to logic and epistemology aredistinct is underscored by his comments in introducing the argument forthe social origins of logic. These comments show that he considered thesocial origin of logic a more difficult argument than the social origin of thecategories. He ([1912:617] 1915:480;*1995:433) opens the discussion bypredicting that “Some will think it reckless of me to broach a problemof such complexity here. For the treatment it deserves to be possible, thesociological conditions of knowledge would have to be better known thanthey are.” Only a few of the necessary conditions have been specified(this may be a reference to his own epistemological argument, whichmight be said to specify some necessary conditions). But, in spite of hisown objections Durkheim ([1912:617] 1915:480;*1995:433–4) proposesthat “the question is so important and so directly implied by everythingthat has gone before that I must make an effort not to leave it without ananswer.” There must, after all, he says, be a first answer to every question.

9.3.1 Concepts and General Ideas

Durkheim ([1912:617] 1915:480;*1995:434) then begins a discussion ofconcepts, which are he says “The basic matter of logical thought.” Thissection at first seems a little strange, since Durkheim has spent so muchtime in the body of the text answering the question of how society played arole in the formation of the categories. It becomes clear, however, that hetreats the idea of the development of concepts and language as a separateproblem from the development of the categories. He is also asking the


question of how society, as the origin of concepts, also came to be thefoundation of logic, since that could not have been its original purpose.

This section involves Durkheim in a discussion of the relationshipbetween concepts and general ideas. There was an extensive discussion ofthis issue in the body of the text. But, here the issue seems to be a differentone. In the text, the focus was on problems inherent in the process of gen-eralization that prevent general ideas from being valid categories of theunderstanding. Here the focus is on the consequences of treating the for-mation of concepts as a process of generalization. Generalization is a sim-ple process of individual reasoning. According to Durkheim ([1912:617]1915:480;*1995:434) “if we see the concept only as a general idea, asis usually the case, the question of how society could be an origin oflogical thought seems insoluble.” The wider scope of general ideas doesnot, by itself, make them logical. There are concepts that refer only toindividuals and they are also logical. The ideas of individual ancestralheroes, for instance, are true concepts. The individual can make theirown generalizations based on their own perceptions. The result of equat-ing concepts with general ideas, he argues, is that society appears to havenothing to do with the process. Given that beginning point, Durkheimsays ([1912:617] 1915:480;*1995:434) “it is not easy to see why general-ization should be possible only in and through society.”

Durkheim then builds on his earlier critique of general ideas, argu-ing that they are not logical concepts. General ideas only containwhat was in a particular idea. If the particular idea was not in thefirst instance logical, why would the generalization that includes it belogical? Indeed, as Durkheim argued elsewhere, individuals are capa-ble of forming generalizations based on sense perceptions. But, hesays ([1912:617] 1915:480;*1995:434), “If there is nothing logical aboutthe particular ideas, why would the general ones be any different?”Durkheim’s argument with regard to the relationship between generalideas and logic parallels the original problem of general ideas. Since logiccannot be derived from perception, it cannot be equated with generalideas.

9.3.2 Concepts versus Sense Impressions

Having disposed of general ideas as a possible origin of logical thought,Durkheim proceeds to rather directly consider the position of WilliamJames, arguing that concepts could not have come from the “flux” ofindividual experience.24 Durkheim distinguishes concepts from sense

24 The section relating to William James appears in Durkheim’s text at pages ([1912:618–19] 1915:481; *1995:434–5).


impressions in any form. He distinguishes concepts from what he calls([1912:618–19]1915:481;*1995:434–5) “tangible representations,” sen-sations, perceptions, or images. Durkheim ([1912:618–19] 1915:481;*1995:434–5) mentions several properties that he believes distinguishconcepts from tangible representations: “Each [sense representation] islinked with the exact moment in which it occurs.” Furthermore, heargues ([1912:618–19] 1915:481;*1995:434–5), “We are never assuredof retrieving a perception in the same way we felt it the first time; foreven if the thing perceived is unchanged, we ourselves are no longerthe same.” This makes perceptions different from concepts, which donot change over time. “The concept, on the other hand” accordingto Durkheim ([1912:618–19] 1915:481;*1995:434–5), “is somehow out-side time and change.”25 Experience is in flux, concepts are static. Heargues ([1912:618–19] 1915:481;*1995:434–5) that “One might say thatit is in a different region of the mind, a region that is calmer and moreserene.” Concepts resist change. Language consists of concepts, and is,therefore, very different from the world of experience, which it is never-theless used to express.

Because of the differences between concepts and perception, Durkheimargues, we could not have learned concepts through individual expe-rience. When the differences are considered, the social origin of con-cepts becomes clear. Concepts are common to all because they are theresult of cooperative activity. According to Durkheim ([1912:618–19]1915:481;*1995:435) “The system of concepts with which we think ineveryday life is the one the vocabulary of our mother tongue expresses,for each word translates a concept.” They are created in the collectivestate in which all minds and bodies meet. Through social conventions ofuse and participation in social practices we learn to speak a language, andto think in concepts, that could not have come from sense impressions.

At this point the discussion becomes somewhat Wittgensteinian.Socially created concepts form the boundaries of the world that can beknown. However, unlike Wittgenstein, Durkheim argues that there arebasic shared experiences that cause some of these collective represen-tations, and help to shape the form that language takes. Both conceptsand sense impressions represent something external that is experienced.

25 Durkheim sometimes uses the word universal to mean broad or general when speaking ofconcepts. At other times he uses the word general (in French the words used are universel,universalisable, and general). He is ambivalent with regard to the word general becausehe does not want to be misunderstood as equating concepts with general ideas. In thissection in which he is arguing against general ideas it is reasonable to suppose that heuses the word universal with regard to concepts in order to avoid the word general, andthat he does not mean universal here in any strict epistemological sense. He does notmean, for instance, that all concepts have the same universality as the categories of theunderstanding.


But concepts correspond to social forces rather than to sensations. Theconcept is more stable than sensations because society itself, which it is areflection of, is very resistant to change. Therefore, language is not nomi-nal, or arbitrary, for Durkheim, even though it is conventional. In fact, hewill argue that every concept retains something of the social experiencesthat are its cause or motivation.26

In this section Durkheim ([1912:618–19] 1915:481;*1995:435) alsodiscusses special scientific terminology, with regard to which, he says,the scientist must “always do a certain violence” in order to be inno-vative. This is particularly important given the currency of Durkheim’swork in the sociology of science. Durkheim seems to have been the first tocombine an empirically based science with the argument that words arethe boundaries of both thought and knowledge. He also seems to haveseen that scientific innovation would require crossing moral boundariesof both language and scientific faith. Those revolutionary arguments thatdid a “certain violence” to established ways of speaking about things,could be expected, like his own, to meet with moral censure. In thishis arguments are a precursor for Wittgenstein, Thomas Kuhn, contem-porary arguments in the sociology and philosophy of science, and thepostmodern and poststructural critique.

9.3.3 Concepts Can be Passed Between Consciousnesses

One of the characteristics that concepts get by being social, accordingto Durkheim, is universality. Universality is a social requirement becauseconcepts can only be communicated if they are common to all. Accord-ing to Durkheim ([1912:619] 1915:481–2;*1995:435): “It is impossiblefor me to make a sensation pass from my consciousness into somebodyelse’s; it is closely dependent on my body and personality and cannot bedetached from them. All I can do is invite another person to set himselfbefore the same object as I and open himself to its influence. By contrast,conversation and intellectual dealings among men consist in an exchangeof concepts.” In order for concepts to work as a means of communication,it is necessary that they have a common origin.

The fact that concepts can pass between people, while sense impres-sions cannot, Durkheim argues ([1912:619] 1915:481–2;*1995:435),shows their origin, “It is common to all because it is the work of the com-munity.” Concepts would not be able to pass between people if they were

26 I say cause or motivation because some ideas are directly caused by social experiences,and some arise after the fact in an attempt to provide an account of that experience.There is a difference between these two, but both nevertheless bear the stamp of theirsocial origins.


constructions of the individual mind. Therefore, the fact that they can bepassed from one person to another is evidence of their social origin. Theconcept, according to Durkheim ([1912:619] 1915:481–2;*1995:435)“does not bear the imprint of any individual intellect, since it is fash-ioned by a single intellect in which all the others meet, and to which theycome, as it were, for nourishment.” The “single intellect” in which allmeet is society.

This sort of reference to “society” has been interpreted as an indicationthat Durkheim believed in a group mind. But, it is more consistent withDurkheim’s argument to assume that he means an assembled group,acting on an established set of social expectations, through contact withwhich moral forces and/or conventions of use can be experienced.

9.3.4 Language

Durkheim’s discussion of language is short, important and provocative.It has also come in for heavy criticism. But, it is important to note thatsince the epistemology is not based on the discussion of language, andThe Elementary Forms is not about a theory of language, the theory doesnot have to bear any weight.

According to Durkheim, the real world we live in is the social world ofour making, not the natural world. Concepts do not correspond to expe-rience of the natural world and, according to Durkheim ([1912:620–21]1915:483;*1995:436) “Often a term expresses things we have never per-ceived and experiences we have never witnessed.” But, socially madeconcepts can directly correspond to that socially made world. Accord-ing to Durkheim ([1912:620–21] 1915:483;*1995:436) “Even when weknow certain of the objects to which the term refers, we know them onlyas particular examples that serve to illustrate the idea but that would neverhave been enough to form it by themselves.”

Because the idea could not have come from the particular examples thatillustrate it, it must have originated as a broad type. Durkheim argues that([1912:620–21] 1915:483;*1995:436) “Whenever we are in the presenceof a type of thought or action that presses uniformly on individual intel-lects or wills, that pressure on the individual reveals the intervention ofthe collectivity.” The broad, or shared type, which cannot have its originin particulars, is evidence of the collective origin of ideas; the result ofcollective action presses uniformly on individual intellects.

This is a part of Durkheim’s Conclusion that would certainly haveappeared to Parsons to be idealist. Concepts, with a distinctly collectiveorigin, are said to define the boundaries of what can be known. What thecritics will have missed, however, is the emphasis on the social character


of everything that concepts correspond to. Durkheim is not saying thatwords and society are ideal, or that the ideal defines the real. He is sayingthat both concepts and what can be known are social and collective inconcrete and not conceptual ways. As he argued in Section II, it is the realthat creates the ideal, and the ideal ultimately sets functional conditionsfor the real. That is, in order for society to exist it must create the ideal.

The social experiences that concepts correspond to are just as concreteas the individual experiences that sensations correspond to. Durkheim isnot arguing that all experience is just as ideal as concepts. Concepts, ina very real sense are not ideal themselves. He also doesn’t mean thatsociety, as a separate entity, thinks. He means that individuals assembledand in action are society, and that as they act together to enact society, thesame social experiences press upon all of their wills and form the sameconcepts in each of them.

Language is ultimately the mirror of the society and what it knows.According to Durkheim ([1912:620–21] 1915:483;*1995:436) “There isa whole science condensed in words then, a science that is more thanindividual; and it so far surpasses me that I cannot even make all theresults my own.” There is also a footnote about habit in this section.Durkheim notes that habit cannot explain the generation of concepts orlanguage. Prior actions may be crystallized in a habit, but habits cannotexplain the genesis of an idea. It had to already exist to become a habit.27

9.3.5 Concepts as Collective Representations

The argument that language is a mirror of society and all that is known,and that individuals assembled to enact practices together not only acttogether, but also think together, brings Durkheim to his discussionof collective representations. He says ([1912:621] 1915:483;*1995:436)“This point enables me to define the sense in which I say that concepts arecollective representations. If they are common to an entire social group,it is not because they are a simple average of the corresponding individ-ual representations.” Collective representations are not abstractions frommore concrete individual impressions. They are completely different fromgeneral ideas which are, Durkheim says, a simple average of particularexperiences. They are social representations of social things.

27 The idea that habits of speech constituted an explanation of language was an importantaspect of Locke’s empiricism. For Locke, it meant that people often speak in wordsthat they have learned through habit, but which have no meaning for them. The ideaof habit reappeared in James in a slightly different form. There it was offered more asan explanation of how words come over time to have an habitual meaning, as opposedto Locke’s argument that because they are habit, or common acceptation, they have nomeaning. The idea of habitus as an explanation of meaning appeared again in the workof Pierre Bourdieu.


Collective representations are also not abstract. They are concrete andapply directly to collective social experience. According to Durkheim([1912:621] 1915:483–4;*1995:436), “Concepts are not abstract thingsthat have reality only in particular circumstances. They are representa-tions just as concrete as any the individual can make of his own environ-ment, for they correspond to the way in which the special being that issociety thinks about the things of its own experience.” Collective repre-sentations are concrete, because they correspond to the way society (thegroup in action) “thinks about the things of its own experience.”

If this reference to society thinking is interpreted as referring to anabstract entity called society, then what Durkheim says here makes nosense. Abstract entities cannot have concrete thoughts. However, if wetake him to mean the assembled group, each member of which is havingthe same experience of enacted practice as the others, then that expe-rience is concrete and the result would be collective thought. Individualrepresentations are those of sensation, perception, or images. Collectiverepresentations are concepts and the collective emotions that constitutethem.

9.3.6 The Function of Language Leads to Logic

The function of language in the social group leads to a quality of gen-erality, or universality, in language because particulars are not of inter-est to society and the purpose of language is to allow for communica-tion between persons. According to Durkheim ([1912:621] 1915:484;*1995:437) “it is in the nature of society most often to see things in largemasses and in the form they take most generally.” Through his analysisin this section, what Durkheim means by the social contribution to log-ical thought becomes clear. To think logically is to think impersonally.It is to go beyond fleeting individual representations and think in stableand communicable ideas. Language is necessarily impersonal and logi-cal because it expresses the way in which society (in the person of theassembled group) conceives the objects of experience.

If “language” originated with individuals, then it would have the qual-ity of particulars. But, it would also be incommunicable. Because societyrequires universals, concepts develop the quality of universals in order tomeet the needs of intersubjective communication, which also suits con-cepts to the purposes of logic. This quality distinguishes general conceptsthat are the work of society; that is, collective representations, from indi-vidual general ideas, which are averages of particular sense impressionsand, for Durkheim, not communicable in the same way.

As social representations, collective representations add the knowledgeof the group to the knowledge of each individual. According to Durkheim


([1912:621] 1915:484;*1995:437) “To think with concepts is not merelyto see the real in its most general characteristics but to turn upon sensationa beam that lights, penetrates, and transforms it.”

Durkheim presents language as a system of ideas that represents eachsocial system in general logical terms. Ironically, language does this notbecause it serves the function of logic, but because the function of inter-subjective communication requires generality. Generality, however, is alsoa basic requirement of logical ideas.

Social ideas are so inherently general, he says, that some ideas are nec-essarily distorted when individuals think them. Because of the differencebetween collective and individual states of being, and the fact that con-cepts are produced by the collective state, individuals often have troublethinking in conceptual terms without distortion. Misunderstandings hap-pen, he says ([1912:621–2] 1915:484;*1995:437) “because we all use thesame words without giving them the same meaning.” This is almost theopposite of the usual way of putting the problem, where it is assumedthat persons all have different ideas that they try to put into words. Onthis view, the distortion occurs because words do not adequately rep-resent individual thought. What Durkheim is saying is collective repre-sentations, and the conventions of language use, are inherently collective.When individuals try to make language express perceptions that they haveexperienced in their capacity as individuals, distortion occurs becauseindividual thought can never be universal and collective enough to matchthe words entirely.

Not only is language necessarily general, but, because social gener-alities are the blueprint for language, which is then also comprised ofgeneralities, and social relations and functions make their way into thegeneralities of language and language use. He does not mean that con-ventions of language use are the origin of reason. This would be consis-tent with a Wittgensteinian position, and Durkheim’s argument has oftenbeen interpreted in this way. But, this argument is not consistent with thecentral chapters. Nor, is it consistent with many of the claims about logicand reason that he makes in this same section. What Durkheim meansis simply that logic requires general terms and logical relations and thatsocial conventions supply both. According to Durkheim ([1912:622–3]1915:484–5; *1995:437) “Logical thought is possible only when manhas managed to go beyond the fleeting representations he owes to senseexperience and in the end to conceive a whole world of stable ideals, thecommon ground of intelligences.” While the function of language was not toproduce logic, a communicable language has the same requirements as logic. Infact, it is the requirements of communicable language that define logic,not the reverse.


9.3.7 The Problem of The Truth of Concepts

Logic, according to Durkheim, presupposes the experience of a truthdistinct from sense impressions. That truth is the moral force of society.Durkheim asks how people could have arrived at the idea that there isa truth beyond individual sense impressions when there is nothing inindividual experience to suggest it.28 The answer he gives is that theygot the idea through collective experience. Collective experience createsimpersonal thought. This answer refers back to the “is/ought” discussionof Section ii, in which Durkheim argued that experience of an imperfectsocial world creates the idea of a truth beyond that world.

Therefore, simply because society exists, there also exists beyond theimage of sensation a system of collective representations with marvellousproperties and moral force. The history of philosophy, he says, is thehistory of the human individual attempting to bring their own ideas closerto this collective idea of truth and explain where it came from. Accordingto Durkheim, when man thought he understood the causes of concepts,he gave himself the right to make them. “In this way” he says ([1912:623–4] 1915:485;*1995:438), “the faculty of conceptualization individualizeditself.” But, it did not start as an individual capacity.

Having argued this, Durkheim anticipates a criticism of his position.For many scholars the function of a concept is to be true, not to facil-itate agreement among minds. They see the impersonality of conceptsas a consequence of their truth and objectivity, not truth as the conse-quence of generality. Durkheim anticipates the criticism that languageposited in terms of conventional meanings cannot be true. Some he says([1912:623–4] 1915:485;*1995:438), will argue that “a concept wouldseem not to fulfill its raison d’etre unless it was true.” The evolution ofconcepts that Durkheim proposes does seem to move toward truth inthe sense of an “agreement with the nature of things.” But, according toDurkheim, to agree with things is not either the origin or the purpose ofconcepts, and does not happen to all of them.29 In any case, concepts

28 This argument about the social origins of the idea of truth is similar to Durkheim’sargument in the body of the text regarding the social origins of the idea of the sacred.The similarity suggests that there is an essential linkage between the idea of the sacredand moral force and the ideas of truth and logic.

29 There is another interesting parallel to Wittgenstein here with regard to Durkheim’s argu-ment that the assumption by philosophers that the meaning of words must be explainedin terms of a correspondence to things had caused the problem of meaning in the firstplace. According to Wittgenstein words do not have meaning through correspondence.Therefore, this approach to the problem is not only wrong, but renders the problem ofmeaning unsolvable. See the first pages of The Philosophical Investigations (1945) for adiscussion of this problem.


with so-called “scientific objectivity” are only different in degree fromother concepts. Scientific objectivity is also, he says, a matter of faith.

Here and in the 19th lecture on pragmatism Durkheim allows thatover the course of history concepts may come to be more true in thisrepresentational sense of referring to something accurately. But, he willalso argue that being true is not the primary function of concepts. It ismore important for concepts to be in harmony with the system of symbols,beliefs and moral forces in a society, than it is for concepts to be true.

According to Durkheim ([1912:624] 1915:485–6;*1995:438–9) “Theconcept that is at first held to be true because it is collective tends not tobecome collective unless it is held to be true.” Here he is speaking of truein the sense of fulfilling its social function, not corresponding to things.This would seem to hold however, only for certain concepts, becausehe contrasts it with those concepts that come from language which, hesays ([1912:624] 1915:485–6;*1995:439) are not subjected to “any priorcritique.” When we learn a language we accept the words, whether wehave prior experience of the social events or things that they designate.But, words in a language would all have passed this test at some point inthe past. Other concepts come from direct collective experiences. But, hesays ([1912:624] 1915:485–6;*1995:439) there are “only differences ofdegree between those concepts and the ones that draw all their authorityonly from the fact of being collective. A collective representation, becauseit is collective, already presents assurances of objectivity.”

The important question is whether a concept meets its social func-tion. This is not an idealist argument, however, because social functionsare real. Durkheim argues ([1912:625] 1915:486;*1995:439, emphasisadded) that “a collective representation necessarily undergoes a test thatis repeated indefinitely. The men who adhere to a collective representationverify it through their own experience.” The authority of concepts comesfrom their harmony with society, not from a correspondence with things.Durkheim says ([1912:625] 1915:486;*1995:439) “To be believed it isnot enough that they be true.” Concepts could theoretically in fact betrue and still not be believed. The whole question of the truth of a partic-ular concept is, according to Durkheim, a secondary issue. For Durkheim([1912:625] 1915:486;*1995:439, emphasis added) “If they are not inharmony with other beliefs and opinions – in short with the whole set ofcollective representations – they will be denied.” Even our willingness toaccept new scientific concepts depends on our faith in science, he says,not on the objective truth of the concepts.

All collective representations, because they are collective, present assur-ances of objectivity. Because a collective representation comes from col-lective experience it is in agreement with the social nature of things. It is


verified through collective experience. Scientific symbols also have thischaracter. Even when they are scientific, concepts never take their author-ity from their objective value alone. It is not enough to be true. They mustbe in harmony with other beliefs, and with collective experience in gen-eral. They get their objective value first from their moral authority, fromcollective representations, and only secondly from their correspondencewith things.

According to Durkheim, bearing the seal of science carries authorityin the modern world only because we have faith in science. That faith isnot essentially different from religious faith.30 Just like religion, scienceexpresses a state of opinion. This is a perspective on science that wouldhave been rejected in the heyday of positivism, but which sounds muchmore plausible today.

Still speaking of science as faith, Durkheim says ([1912:625]1915:486;*1995:439, emphasis added) that “The value we attribute toscience depends, in the last analysis, upon the idea we collectively haveof its nature and role in life, which is to say that it expresses a state ofopinion. The reason is that everything in social life rests on opinion, includingscience itself.”

It appears that Durkheim is arguing that because logic and languageultimately depend on the state of development of a society, even themost empirically based science is still ultimately a study of opinion. Hespeaks of science and opinion getting closer to the truth. So, it wouldseem that opinion can be more or less true and that he is not arguing thepostmodern position that since we have to think in concepts there is nopoint in empirical methods. Durkheim believes that empirical methodscan lead to innovation in thought. There may be resistance from thoughtand belief. But innovation based on empirical evidence is not impossible.

9.3.8 Conceptual Thought is Contemporaneous With Humanity

It follows from the argument that communicable concepts are logical,that all social beings, and here Durkheim means particularly to indicateaboriginal peoples, in all societies, have been equally possessed of logicalintelligence. Durkheim argues ([1912:626–7] 1915:487;*1995:440) that

30 This might also be interpreted to imply that the laboratory practices of science arelike religious practices in being the real essence of science. If we extend the analogy tolaboratory practices there are some interesting parallels. It is through shared practicesin the laboratory that scientists come to have shared experiences on which to base theirshared scientific language. Furthermore the laboratory practices which have becomethe focus of Harold Garfinkel’s later research (see also Mike Lynch) would be directlyparallel to Durkheim’s research and writing on religion. Both are focused on practicesand according to Durkheim science serves some of the same functions as religion.


“A man who did not think with concepts would not be a man, for he wouldnot be a social being.” It has been possible to argue that logical thoughtand scientific speculation (religious thought) are a modern development,he argues, only because of the identification of the concept with generalideas, instead of with social processes.31 But, he argues ([1912:626–7]1915:487;*1995:440), “conceptualizing is not the same thing as general-izing,” which is an individual mental process. Since logical thought beginswith the concept, and the concept is inherently social, and there canbe no social life without it, then it follows for Durkheim ([1912:626–7]1915:487;*1995:440) that “there has been no historical period when menlived in chronic confusion and contradiction. Certainly, the different fea-tures of logic in different historical periods cannot be overemphasized;logic evolves as societies themselves evolve. But, however real, the dif-ferences should not cause us to miss the similarities, which are no lessfundamental.” Aboriginal peoples were as logical as modern western peo-ples. Their logic was dictated by the general social forms in their society,our logic is dictated by the general social forms in ours. Because our soci-ety has become more international there has been a progression towardmore universal forms. But, this is not because we have escaped from thesocial construction of logic; it is a feature of the social construction oflogic.

31 It is very interesting to note that Durkheim holds a philosophical argument, with regardto the importance of general ideas, responsible for the western centric attitude toward theintelligence of aboriginal peoples. Of course he does not mean that philosophy caused theproblem, that would be inconsistent with the relationship between logic and society thathe outlines. It would be consistent with his position to argue that western philosophydeveloped in a society in which the careful delineation of general ideas had for socialreasons become important. Philosophers then mistakenly confused general ideas andlogic. Because of this confusion many philosophical problems, including epistemology,and the problem of meaning in language could not be solved. Additionally, this focuson general ideas provided an account of western intellectual superiority that was veryconvenient as a justification for colonialism.

10 Durkheim’s Conclusion Section iv: LogicalArgument for Social Origin of the Categories

In Section iv of his Conclusion, Durkheim takes up the origin of the cat-egories of the understanding as a final question of logic, but treats thisas a separate question from the origin of logic in general addressed inSection iii. The discussion of the categories here is important because,given the general tendency to overlook the relevance of the centralchapters, scholars have tended to treat this discussion as Durkheim’sargument that the categories have a social origin. It is essential to under-stand that in this last section, Durkheim does not make an argument thatthe categories do have a social origin. He does not need to, as he hasmade that argument already in the central chapters. What he argues hereis that the categories would not be suited to, their purpose if they did nothave a social origin and therefore must, for logical reasons, have a socialorigin. This is a second logical argument entirely different from, and notintended to be taken in place of, the demonstration of their empiricalorigin in the central chapters. Unfortunately, as he does many times inthe text, for instance with regard to Smith’s objection, Durkheim followsone argument with another that is quite different from the first. The argu-ment that the categories need to have a social origin is not the argumentthat they do have a social origin expressed differently. It is a differentargument.

10.1.0 The Six Categories of the Understanding

One important thing that happens in Section iv is that, over and overagain, Durkheim repeats a list of six categories. Given the ambiguity ofthe introductory discussion, the congruence between this list and the cat-egories that Durkheim actually provides empirical demonstrations for inthe text is important. In Section iii, where he discusses collective rep-resentations, Durkheim does not mention categories at all, But here hekeeps mentioning the same six. This denotes a clear separation betweenthe discussion of the categories and the earlier discussion of collectiverepresentations and underscores the difference between them.

301


In this last section Durkheim also makes it clear that the epistemolog-ical question of the social origin of the categories has been a consistenttheme throughout the text. The question of the social origin of the cat-egories, he says ([1912:627]1915:488;*1995:440) “was set out in theintroduction and has remained implicit throughout the book.” In fact,Durkheim has demonstrated in exhaustive detail in the central chaptersof the text that six categories of the understanding have a social origin.The question he addresses now is not whether or not the categories actu-ally have a social origin, that he has already demonstrated, but rather,why they need to have a social origin.1

Of course, being concepts, it would follow from the discussion of logicthat the categories come from, or are the work of the collectivity. But,Durkheim clearly distinguishes the categories from concepts and logic.The problem with regard to the categories, Durkheim says ([1912:627–8]1915:488;*1995:441), is “more complex.” If all concepts and all of logichave a social origin, then the categories must be suited to work in har-mony with them. But they can’t come from them. On the other hand, ifthe categories had an individual or an innate origin they would not becompatible with the concepts which it is their function to govern. Thepurpose, or function, of the categories is to act as a foundation for reason,for use by social beings, who are required to reason in a social world aboutsocial things, using concepts that take a conventional social form. If thecategories did not have a social origin, Durkheim argues, they could notfulfill this function.

This is a quite straightforward argument. Durkheim has demonstratedrepeatedly in the text that the categories are the direct result of enactedsocial practice. What he argues here is that not only is this empiricallytrue (a contingent matter) but also logically necessary. In order for thecategories to be of any use to social beings, as aids to rational thought thatoccurs within a socially defined context in which all thoughts, ideas, andperceptions, are shaped by social experience, the categories would haveto be compatible with that social world. The only way in which they canbe entirely compatible, he argues, is if they themselves came from that

1 In the text Durkheim says ([1912:627]1915:488;*1995:441) “the question is where theygot this trait.” In French the line reads “Il s’agit de savoir d’ou leur vient ce caractere.”It is important to note just how Durkheim poses the question at this point. He asks wherethe categories get the trait of being social. But, from the ensuing discussion it becomesclear that he means to answer the question why they need to have a social origin. If thequestion is read literally as a question about where, then Durkheim’s argument appearsto be confused because he has already answered that question exhaustively in the text,and what follows is not an answer to the question of where the categories got their socialcharacter, but rather the question why they need to have a social origin.

Durkheim’s Conclusion 303

social world. If they came from individual perceptions, or anywhere else,they would not suit their social purpose. This would be a logical necessityin any possible social world.

In this argument Durkheim anticipates a criticism of his empirical argu-ment for the categories based on the philosophical assumption that theempirical is contingent, and that logic and morality cannot be derivedfrom contingencies. He sets up the argument so that he has two defensesof his position. Arguing that it is a logical necessity that the categorieshave a social origin partially addresses the concerns of the critics. But,he has another argument ready, his earlier argument with regard to theis/ought problem and now he transposes that argument onto his discus-sion of logic. According to Durkheim society, as the origin of logic, isinherently logical. Therefore, a social origin for the categories does notmake them illogical. Attributing a social origin to logical thought and tothe categories of rational thought, Durkheim argues, does not denigratelogic and reason. It is true that notions worked out socially will not bedirectly adequate to their objects. Society is still an individual thing andit particularizes ideas. Therefore, even collective representations containsocially subjective elements. But, as he argued in Section iii, achieving acorrespondence type of truth is not their purpose. In spite of their lack ofcorrespondence to objects, they open the way to ordered thinking; whichis their purpose: they are the foundation of logic.

Durkheim argues that despite the claims of western logicians, mod-ern logic has not evolved and become more systematic by eliminatingits social elements. To the degree that it did eliminate social elements,on Durkheim’s view, it would have been reduced to general ideas andwould no longer constitute logic. Rather, he argues, it has developed asa corollary of a new form of social life: international life. International liferequires logics to expand beyond the narrow social frames they wereborn in. In this fashion, he argues, logical organization begins to becomeautonomous from social organization. The bond that first joined logicalthought to particular collectives becomes increasingly detached, imper-sonal, and universal.

At the very end of Section iv, Durkheim takes up what he refers toas the “vocation of sociology,” which he says is to open a new way tothe science of man. According to Durkheim, up until now there weretwo choices: First, to explain reason on the basis of sensation; mind bymatter; or Second, to connect reason to some reality above experience.What placed us in this difficulty, he says, was that the individual was takento be the natural end. But, a new explanation becomes possible when werealize that above the individual there is society.


Society, he says, is a system of active forces, not a nominal being,not a creation of the mind. Society as a network of mutual emotionalconnections creates reason. Therefore, as a cause of the categories, soci-ety is empirical and doesn’t need to be placed outside of experience.In any case, “before drawing that extreme conclusion” Durkheim says([1912:638]1915:495–6;*1995:448), it is best to find out “whether thatwhich is in the individual but surpasses him,” and could not come fromhim, could have come from the “concretely experienced, reality that issociety.” Society as a concretely experienced reality is not ideal. It is ademonstration of this that Durkheim says he has tried to give.

10.1.1 Why The Categories Need to be Social

While Durkheim has argued that all concepts are social and therefore,as concepts the categories must also be social, he argues ([1912:627–8]1915:488;*1995:441) that “the categories are social in another sense andto a higher degree. Not only do they come from society, but the very thingsthat they express are social. It is not only that they are instituted by societybut also that their content is various aspects of the social being.” In otherwords, it is not just that the categories are “instituted” by society, they arenot just conventional. Their content preserves essential social experiencesthat suit the categories to a social function. The content of the categoriesare those social feelings and collective experiences that make personshuman and social, and hold society together, creating social solidarityand the grounds for mutual communication. That these collective statesare directly experienced, and that these experiences are the categories,makes them quite different from other ideas, whether those other ideasare taken to have conventional, or referential meaning. The categories notonly express social things, but they express the fundamentals of societyand provide a rational basis for thinking about those fundamentals.

The reason that the categories must have a social origin, accordingto Durkheim, is that for ideas to serve the broad function of facilitat-ing communication and social bonding, they must themselves have abroad foundation, and at the same time be compatible with a universe ofpurely conventional social concepts. According to Durkheim ([1912:628]1915:488;*1995:441, emphasis added) “Indeed, the function of the cat-egories is to govern and contain the other concepts. They form the per-manent framework of mental life. But, to encompass such an object, theymust be modeled on a reality of equally wide scope.” For all people to beable to communicate, concepts must have a common origin. However,unless their reasonings with one another are to be entirely conventionalin content, their basic faculty of reason must also have a basis in the


real. According to Durkheim, society is the only possible source of con-crete experiences that are shared broadly enough to provide for universalempirically valid ideas as a basis of understanding. These ideas wouldnot be purely conventional. They would be empirically based.

Because the categories have a social function, to provide for commu-nication and knowledge within a social framework, they have to be com-patible with that framework. They must be universal, therefore they musthave a universal for a model. Kant’s solution, that the categories are uni-versal because they are innate, does not work for Durkheim. He arguesthat an innate individual origin for categories could not explain essentialaspects of social solidarity. Nor could it provide a foundation for thinkingin terms of social conventions. Yet, in spite of not being innate, the cat-egories must have an origin broad enough to allow them to function asuniversals. According to Durkheim, only those social experiences gener-ated by enacted practices, are sufficiently universal to generate universalideas.

10.1.2 Why Individualism Offers an Insufficient Explanation

Having dispensed with an innate origin for the categories, Durkheimargues that individually derived categories could not have the neces-sary social qualities to support the use of reason in a social context.This is not the same argument Durkheim has made earlier that therecould be no individual origin for the categories. That argument involveda critique of empiricism and general ideas. Here Durkheim makes thelogical argument that if it were possible to derive categories from individ-ual perception they would not be functional. The individual lives in timeand space, and has an idea of similarity, regularity etc, as James pointsout. But, it is only in so far as the individual acts as a member of soci-ety that they have a sense of social time, social space and social regu-larity. It is social time in which the individual lives and about whichthey must exercise reason. The same is true of social space, social clas-sifications, and social forces. Abstract ideas of time, space, and forcewould not help the social being to reason with regard to the social worldin which they find themselves. According to Durkheim ([1912:629–30]1915:489;*1995:442)

. . . the regularities that I can perceive in the way my sensations follow one anothermay very well have value for me; they explain why I need to wait for the secondwhen the first of two phenomena whose constant conjunction I have experiencedis given to me. But, that state of personal expectancy cannot be assimilated to theconception of a universal order of succession that imposes itself on all minds andall events.


Durkheim’s point is that even if it were possible to derive universal categoriesfrom the individual mind, or from individual sensation, they would notbe well suited to their purpose, which is to reason about social things andto communicate with others about those things.

Individually derived categories, Durkheim says ([1912:628–9]1915:488–9;*1995:441), would be stuck within the individual’s “narrow hori-zon.” Innate concepts, on the other hand, would have an abstract qualitythat would not suit them to reasoning with the conventionally derivedconcepts and social unities created by social forces. What is needed areconcepts that can express the social world as a totality, a unity that hasno counterpart in nature, or in the individual. Only concepts that havean origin in social forces could have this dual character of social totalityand universality.

10.1.3 The Relationship Between the Categories and a Universe of Concepts

According to Durkheim, the universe (of thought and social experi-ence) takes place within society. It is the total genus outside of whichnothing exists. Empirically valid categories of thought, with an originin social experience, are suited to reasoning within the limits of thisuniverse bounded by socially defined concepts. Durkheim ([1912:630]1915:490;*1995:442–3) outlines his idea of a universe of concepts andthe relationship between society and concepts in general.

Since the world expressed by the whole system of concepts is the world societyconceives of, only society can provide us with the most general notions in termsof which that world must be conceived. Only a subject that encompasses everyindividual subject has the capacity to encompass such an object. Since the uni-verse exists only insofar as it is thought of and since it is thought of in its totalityonly by society, it takes place within society; it becomes an element of society’sinner life, and thus is the total genus without which nothing exists. The conceptof totality is but the concept of society in abstract form.

Categories with an individual origin would not be compatible with such“social systems of concepts.” But, if socially derived categories were definedby this universe of concepts they would themselves be stuck within thatuniverse and be relatively useless for reasoning about it. It is the empiri-cal nature of the social experiences that give rise to the categories thatdistinguish them from social conventions. Without this distinction hisargument would be idealist. This distinction, however, allows him to dis-cuss a progress toward the “truth,” which he does in the very last pagesof the Conclusion (and in the 19th lecture on pragmatism).

It is the world expressed in concepts that members of society perceive.It is thinking about the universe in social terms that makes it available as


a universe in thought. Therefore, the whole of the known and knowableuniverse takes place within society. The universe is, for social beings,exactly as they think about it in social terms. This social ordering ofthe entire universe is what creates the idea that the universe, which isreally composed of an infinite number of things, is really one totality. Inreasoning about such a socially created totality, social beings need sociallyderived, but at the same time universal, and empirically valid categories.

The concepts with which persons speak and think are all part of a“system of concepts” defined by, and modeled on, the social structure ofeach society. In order to underline their social origin, Durkheim points outthat social time and space are ideal rather than coming to us through senseimpressions ([1912:631]1915:490–1;*1995:443), “But what brings outthe extent to which that total space differs from those concrete expansesthat our senses cause us to perceive is the fact that localization is whollyideal and in no way resembles what it might be if it was dictated to usby sense perception.” The categories of time and space that are part ofthe rational faculty will have to be suitable for reasoning about this idealsocial time and space.

What Durkheim seems to be doing on these pages, beginning with thediscussion of totality, is showing how the social origin of each categorysuits it for its job of reasoning with various social concepts, while an originin perception, or innate reason, would not. He is not arguing that thecategories are derived from their relationship with a universe of concepts,however, only showing how they are suited for thinking within such aconceptual universe.

Having discussed time and space, Durkheim turns to a discussion ofcausality ([1912:631]1915:491;*1995:443): “Similarly, the causal rela-tion becomes independent of any individual consciousness from themoment it is collectively established by the group; it hovers above allthe minds and all the individual events. It is a law having impersonalvalidity. I have shown that the law of causality seems to have been bornin just this way.” Words like “collectively established,” are problematic ifthe reader treats this as an argument to establish the empirical origins ofthe categories. It is not. It is an argument that their social origin suits thepurposes that social beings have for them. Durkheim ends the discussionof causality by saying “I have shown that the law of causality seems tohave been born in just this way.” This again, refers back to the argumentof the text. If we take it to refer to an artificial establishment by the group,then we are taking it out of the context in which it was written.

These summaries of the relationship between the categories and socialideas, are not intended to explain where the categories came from, andrefer constantly back to the fact that Durkheim has already made thatargument. The point here is to establish that since human social beings


must reason within a universe of conventional social concepts, their basicrational conceptions must bear some relationship to the reality that con-structed those conventions.

10.1.4 Relationship to Rousseau on the Origin of Reason

After arguing that only categories with a social origin would be use-ful for reasoning with concepts that had a social origin, Durkheimrestates an argument that, in his lectures on Rousseau, he attributes toRousseau.2 The argument is that without society individuals have noneed to develop rational thought. According to Durkheim ([1912:631–2]1915:491;*1995:443–4, emphasis added):

There is another reason why the constituent elements of the categories must havebeen taken from social life: The relationships they express could not becomeconscious relationships except in and through society. Even if, in a sense, they areimmanent in the life of the individual, the individual had neither reason or meansto grasp them, think about them, make them explicit, and build them up intodistinct notions.

Here Durkheim is arguing that even if something like the categories couldbe grasped by an individual, they would have no use for them unless theywere engaged in social cooperation. In other words, it is only in theircapacity as collective beings, that social beings would recognize the cat-egories as being of any use. They have utility only within the contextof social cooperation. Rousseau had argued that reason had no need todevelop as long as persons were existing only as individuals and not yetas social beings. The inherent animal abilities are sufficient for individualexistence. According to Durkheim ([1912:632]1915:492;*1995:444),“Not only does the animal have no others [abilities], but our own indi-vidual practice quite often presupposes nothing more.”

It is important not to be misled by Durkheim’s language in speaking ofabilities that are “immanent in the life of the individual.” What he meansis shown by his going on to detail all the things that the individual animalcan do without the categories. These are only superficially similar to theabilities that the categories put into play. The categories express relations,according to Durkheim, and relations are inherently social. Naturalthings and individual animals do not have, or experience, such relations.

2 Durkheim gave a series of lectures on Rousseau, published in English as Montesquieuand Rousseau (1960), that make it clear that the resemblance between his argument hereand the argument of Rousseau is not a coincidence. Durkheim credits Rousseau withbeing a sociological thinker because he argues, in The Origins of Inequality ([1757]1999),that human reason would never have developed unless persons had gathered togetherinto societies. Therefore, according to Rousseau, human reason had inherently socialqualities.


The sorts of needs and relationships that reason expresses are not partof the individual life of either the animal or the human. Durkheimargues ([1912:632]1915:491;*1995:444) that “To orient his individ-ual self in space and to know at what times to satisfy various physi-cal needs, he [the individual] had no need for a conceptual represen-tation of time or space, once and for all.” The categories are not nec-essary for the animal to go about its daily life. They only become nec-essary when the individual begins to function as a social being. “Thecollaboration of several in pursuit of a common goal,” Durkheim says([1912:633]1915:492;*1995:445, emphasis added), “is possible only ifthere is agreement on the relation between that goal and the means thatmake its achievement possible – that is, if a single causal relation is acceptedby all who are working together in the same enterprise.”

This mutual acceptance of a single causal relation requires the acqui-sition of a conception of causality that is both shared and sufficientlyempirically valid to facilitate achieving a goal. It is a different sort ofproblem from the perception of before and after effects by animals. Thisformulation of the problem is similar to Pragmatism, so it is importantto see that the solution is different.

In writing of the abilities of animals to perceive similarity and dif-ference, and go about their daily tasks, Durkheim says that theseabilities are sufficient if there is no social purpose at stake. He argues([1912:632]1915:491;*1995:444, emphasis in original) that, “Theimpression of deja vu, of something already experienced, implies no clas-sification. In order to differentiate between those things we must seekafter and those we must flee, we have no need to join the effects ofboth to their causes with a logical link, if individual convenience aloneis at stake.” According to Durkheim ([1912:632]1915:491;*1995:444)“These [abilities] would be sufficient for man as well if his movementshad to satisfy individual needs alone.” But, the social being exists in soci-ety and society requires reasoning about social relationships, not merelyabout individual purposes.3 Here Durkheim clearly restates his positionon dualism: reason does not exist before society makes reason necessary.

Thus, reason is both taken from, and suited to, the needs of society.It has no other purpose and no other origin. It has a social purpose, andit must work with social concepts and a socially derived logic. That iswhy it must have a social origin. According to Durkheim ([1912:633]1915:492;*1995:445, emphasis in original), “It is not surprising, then,

3 Note that the individual ability to reason that Durkheim refers to does not require thattwo individuals be in agreement, or that they form their general ideas in the same way.Thus, certain of the problems raised by Hume are avoided. All such action is based onindividual opinion, and without the need for social coordination that would be sufficient.


that social time, social space, social genera [classes], and collective causal-ity should be the basis of the corresponding categories, since it is in theirsocial forms that they were first conceived with any degree of clarity byhuman consciousness.”

This is the breakthrough that creates human reason and makes societypossible. If the ability were not immanent in the animal, the performanceof collective practice would probably not generate shared categories ofthought. But, if the performance of collective ritual did not require coop-eration and generate shared categories of thought, reason would notdevelop. The individual ability could not be turned into a shared andcommunicable faculty of reason without the social process. It also wouldnot be transformed unless there were a social need for it. This is impor-tant, because Durkheim has often been interpreted as having argued thatreason was already present in the animal in an unformed state. He didnot argue this.

Society, Durkheim argues, is only possible with the categories. Thecategories are only necessary in society. Therefore, society needs to gen-erate the categories in order to exist. Durkheim writes ([1912:632–3]1915:492;*1995:444) that “Society is possible only if the individuals andthings that make it up are divided among different groups, which is tosay genera, and if those groups themselves are classified in relation toone another. Thus, society presupposes a conscious organization of itselfthat is nothing other than a classification.” He then goes on to reiteratethe relationship between social functions and the categories of space andtime. But now he presents space and time as social needs that are ful-filled by the practices of time and space that generate the categories oftime and space. The categories, according to Durkheim ([1912:633–4]1915:492;*1995:445) are not nominal or artificial. They are ideas thatsociety finds in its own forces.

10.1.5 Society is Not Illogical

Philosophers have avoided grounding important arguments in socialfacts, because they consider the social to be illogical and contingent.However, since, according to Durkheim, it is society that creates con-cepts and logic, society that creates the need for reason, and society thatis the universal capable of creating the categories, society itself, he argues([1912:633]1915:492;*1995:445), cannot be illogical:

society is by no means the illogical or alogical, inconsistent, and changeable beingthat people too often like to imagine. Quite the contrary, the collective conscious-ness is the highest form of psychic life, for it is a consciousness of consciousnesses.


Being outside and above individual and local contingencies, collective conscious-ness sees things only in their permanent and fundamental aspect, which itcrystalizes in ideas that can be communicated.

The statement of the association between society and logic is clear.However, Durkheim uses language that tends to suggest that society isitself doing the seeing and thinking. Referring to society, Durkheim says([1912:633–4]1915:492–3;*1995:445),

At the same time as it sees from above, it sees far ahead; at every moment, itembraces all known reality; that is why it alone can furnish the intellect withframeworks that are applicable to the totality of beings and that enable us tobuild concepts about them. It does not create these frameworks artificially butfinds them within itself, merely becoming conscious of them. They express waysof being that are met with at all levels of the real but that appear with full clarityonly at the pinnacle, because the extreme complexity of the psychic life thatunfolds there requires a more highly developed consciousness.

These references to society “seeing” and “thinking” have contributedto the false impression that Durkheim has a group mind theory, and thathe argues that the categories are thought into being [by the group mind]rather than being experienced. Both of these tendencies in the text reinforcethe very misinterpretations that they were designed to counteract: that is,that society consists of real forces, the experience of which creates sharedcategories. Durkheim’s references to society being conscious of itself,more properly recall those moments when assembled groups create socialforces and every member of the group feels that force simultaneously.

The result of referring to society as “thinking” has unfortunately beenthat for many critics the categories appear to be the product of collectivethought, instead of collective action. Collective thought requires a collec-tive mind. Collective action does not. It requires a group of social beingsassembled and in ritual motion. This is largely the result of talking aboutthe logical necessity of the categories (which rather naturally results in anemphasis on “thinking” rather than on their empirical origins). Through-out the text, Durkheim has stated clearly that the categories come fromcollective action. In fact, the whole argument of Book II is that the prob-lem cannot be solved by focusing on beliefs and ideas. This should makeit clear that Durkheim rules out collective thought as a solution. It is onlycollective action, he says, that can explain how all members of a groupcome to have the same categories. Therefore, this last section needs tobe read as a set of statements about the logical need for categories to becreated by society in action, not about a collective mind.

However, there are limits to the adequacy of the categories. Durkheimargues ([1912:634]1915:493;*1995:445) that:


Assuredly, this is not to say that notions worked out in that way could be directlyadequate to their objects. If society is something universal as compared to theindividual, it is still an individuality, having its own form and idiosyncracies; itis a particular subject and, consequently, one that particularizes what it thinksof. So even collective representations contain subjective elements, and if theyare to become closer to things, they must be gradually refined. But, crude asthese representations might have been at first, it remains true that with themcame the seed of a new mode of thinking, one to which the individual couldnever have lifted himself on his own. The way was open to stable, impersonal,ordered thought, which had only to develop its own special nature from thenon.

The first categories opened the way for the development of logic. Thecategories themselves, as a faculty of reason do not change. But, over timethe categories and social logic work together to pull social logic ever closerto the truth. This development of logic over time explains the apparentconflicts between western and non-western logic.

10.1.6 The Development of Modern Logic and Consideration of the Future

Truly human thought, that encompasses all groups is, for Durkheim, anideal limit of history toward which we move closer, but probably neverachieve. Thought, he argues, has the ability to transcend the limits ofsocial groups and view humanity in the abstract. Durkheim’s discussionat this point resembles Marx’s argument, in the 1844 Manuscripts, thatonly when persons transcend the categories that divide them into kinds ofpersons can they overcome the alienation of the modern social conditionand realize their true nature as species beings. In both cases it is a restate-ment of the ideal of the Enlightenment, but without the individualism,and in thoroughly sociological terms.

Modern logic does not fundamentally divide the modern person fromthe aboriginal. The development of a modern form of logic, Durkheimsays, has the same origins as the original development. Both are related tothe society in which they are found. According to Durkheim ([1912:634]1915:493;*1995:446), “If logical thought tends more and more to jetti-son the subjective and personal elements that were launched with it, thereason is not that extra social factors have entered in but far more that anew kind of social life gradually developed: international life, whose effecteven then was to universalize religious beliefs.” The effect of internationallife is to universalize religious beliefs because only through doing so arepersons able to communicate with one another. This is particularly truein a context of globalization.


This process of expanding international life led naturally to a process bywhich logic came to be separate from any particular collectivity. Accord-ing to Durkheim ([1912:634–5]1915:493;*1995:446):

As that international life broadens, so does the collective horizon; society nolonger appears as the whole, par excellence, and becomes part of a whole thatis more vast, with frontiers that are indefinite and capable of rolling back indef-initely. As a result, things can no longer fit within the social frames where theywere originally classified; they must be organized with principles of their own; log-ical organization thus differentiates itself from social organization and becomesautonomous.

Thus, the refinement of logic in the modern world is attributed, byDurkheim, to the same social processes that created logic out of socialprocesses in the first place. Not to a detachment from social processes.The “objectivity” of logic is but the reflection of the modern social expe-rience, associated with pluralism within and between societies. “This” hesays ([1912:634–5]1915:493;*1995:446) “it seems, is how the bond thatat first joined thought to defined collective entities becomes more andmore detached and how, consequently, it becomes ever more impersonaland universalized. Thought that is truly and peculiarly human is not aprimitive4 given, therefore, but a product of history; it is an ideal limit towhich we come ever closer but in all probability will never attain.”

While Durkheim argues that logic develops over the course of history,he did not mean to suggest that the categories are themselves subject tochanges in social processes in the same way. The categories always comefrom direct experience of enacted practices. How they fit into a socialsystem of concepts and beliefs, a logic of thought, however, depends onchanges in societies, and in particular, on relations between societies; theinternational dimension.

Durkheim anticipates some of the troubles that his argument will gen-erate, although nothing like the full extent of the misunderstanding. Hewrites ([1912:636–7]1915:495;*1995:447) “Some will be astonished,perhaps, to see me connecting the highest forms of the human mindwith society.” The distance between sensation and logic seems to requirethe world of reason and morality to have been “added to the first by anact of creation,” he says. Here he refers either to Kant or to religiousbeliefs.5 Durkheim ([1912:636–7]1915:495;*1995:447) does not denythe act of creation, however, he only disputes its origin: “to attribute to

4 I take this not to mean that primitive people do not have the categories, but rather, thatthe categories are not innate.

5 Durkheim introduces a discussion of dualism at this point with regard to which I referthe reader in my Chapter Two.


society this dominant role in the origin of our nature is not to deny thatcreation. Society does indeed have at its disposal a creative power that noobservable being can match.”

It is this creative power, the power to produce moral forces, thatexplains the social origin of the categories. According to Durkheim([1912:637]1915:495;*1995:447):

A society is the most powerful collection of physical and moral forces that we canobserve in nature. Such riches of various materials, so highly concentrated, areto be found nowhere else. It is not surprising, then, that a higher life develops outof them, a life that acts on the elements from which it is made, thereby raisingthem to a higher form of life and transforming them.

These forces are real and the categories that they create therefore haveempirical validity.

10.1.7 The Vocation of Sociology

Then in the last paragraph of book, Durkheim discusses the “vocationof sociology” which he says ([1912:637–8]1915:495–6;*1995:448) “isto open a new way to the science of man.” Sociology is to lead reli-gion and moral philosophy into the new era. This is to be done bygrounding sociology on a new epistemology that treats practices as its pri-mary phenomenon. According to Durkheim([1912:637–8]1915:495–6;*1995:448), “Until now, we stood before these alternatives: either toexplain the higher and specific faculties of man by relating them to lowerforms of being – reason to sense, mind to matter – which amountedto denying their specificity; or to connect them with some reality aboveexperience that we postulated but whose existence no observation canestablish.” Before Durkheim made his argument the only choices wereinnatism and empiricism: Kant and Hume (or James).

According to Durkheim “What placed the mind in that difficulty isthat the individual was taken to be the finis naturae. It seemed there wasnothing beyond him, at least nothing that science might discover.” AsDurkheim has argued throughout the text, treating individual experienceas the source of the categories is why their origin could not be explained.Only when a social origin for the categories is admitted, does the epis-temological problem become capable of solution. “But now” he says([1912:637–8]1915:495–6;*1995:448; emphasis added), “a new way ofexplaining man becomes possible as soon as we recognize that above theindividual there is society, and that society is a system of active forces –not a nominal being, and not a creation of the mind.” Focusing on socialorigins constitutes a focus on practices. Just as natural forces can be


perceived through the senses, social forces can be experienced throughthe emotions, through rites and social actions.

The problem of the contingency of empirical evidence has been solved.Thus, in order “to preserve man’s distinctive attributes, it is no longernecessary to place them outside experience.” The experience of the moralforces that constitute society is a real experience. And, society itself isreal, composed of real forces. It is not a nominal being or a creation ofthe mind. Therefore, categories of reason based on social experience dopreserve the essential aspects of reason that are lost with Hume or James.This allows Durkheim to explain fundamental characteristics of both therational and the moral empirically.

Conclusion

There are important implications for both sociology and philosophy ofa failure to appreciate Durkheim’s epistemological argument. Sociology,which was conceived in the attempt to come to terms with the spreadof industrial capitalism in the nineteenth century, is ironically being leftbehind in the quest to come to terms with globalization in the twenty-first.In place of a sociology grounded in Durkheim’s theory of practice, whichwould solidly connect with studies of global practices in business, science,economics and communication, the discipline has turned to Pragmatismand other more conventional positions that place beliefs and motivationsat the center of social life. As a consequence, actual worksite practices,which are essential to the understanding of science and industry, arerendered invisible to the researcher. In some countries a form of sociologyfocused on beliefs and concepts, almost to the exclusion of practices,threatens to render sociology obsolete as researchers with interests in theglobalization of economics, science and technology increasingly come tounderstand the importance of practice in the contemporary context.

At the turn of the nineteenth century Durkheim offered a solution toclassical problems in philosophy intended to replace Pragmatism andmake epistemology and ethics more relevant in the current context.Because of the failure to take up his argument, however, the problemshe addressed still cripple both disciplines. For sociology the implicationsare most obvious. The thinker who most sociologists credit with beingin some important sense a founder of the discipline has been funda-mentally misunderstood. Framing disciplinary debates in the absenceof Durkheim’s epistemology has led to serious misunderstandings notonly of Durkheim’s entire corpus of work but, more importantly, has leftthe foundations of social thought looking deeply conflicted. In buildingon Durkheim’s work, while neglecting his epistemology, the disciplinehas generally placed itself in an untenable epistemological position, asphilosophers have repeatedly pointed out (Winch 1958; Rorty 1979;Turner 1994). More importantly, work which Durkheim would haveseen as central to the discipline, for instance, studies of shared enacted

316

Conclusion 317

practice (interaction at various levels), and the relationship betweenmutual intelligibility, symbolic meaning and social structure, have beenrelegated to the sidelines in the quest to uncover alleged Durkheimian“structures,” “social facts,” or, more recently in the United States, the“logic” of conceptual or “narrative” systems.

In the process, Durkheim’s argument has been split in two (positivist/empirical versus idealist/symbolic) and the two opposing halves have beenpracticed separately in different disciplines and sub-disciplines and ontwo different continents, leaving both without epistemological coherence.It matters a great deal to social science whether Durkheim saw social factsas external constraining entities in their own right, as his position hasgenerally been interpreted, or as enacted practices whose recognizabilityis a feature of their witnessable achievement. In the second case the detailsof the enactment of shared practice are their structure, not the invisible normsand rules posited by the conventional interpretation.

For Durkheim, the analysis of symbol systems cannot be detachedfrom the empirical contexts of use which give them meaning. However,when his argument is misunderstood and symbols are treated as having anindependent meaning the result is that two different types of sociology areindicated, depending on which interpretation of Durkheim is accepted.On the one hand relations between symbols and the normatively orientedactions of individuals can only be revealed via abstract conceptualization,quantitative measurement, and modeling. The details of enacted practice,on the other hand, are open only to a detailed qualitative approach.

The division between these two approaches has been conceptualizedas a distinction between structure and interaction; things that are said tobe large and things that are said to be small. Hence the labels micro andmacro sociology. This is a false rendering of the distinction between thetwo types of sociology. The distinction should be seen rather as between asociology based on concepts, norms, and values, that assumes an approx-imate order, based on the actions of individuals as constrained by normsand values; and, a sociology focused on concrete observable practices, themeaning and function of which is essentially prior to and independentof norms, and the order of which must be observable in each intelligiblecase. The former, although generally considered to be measuring struc-ture, actually measures the orientation of individuals toward normativevalues and relies primarily on data drawn from interviews and surveyswith individuals. The latter, generally characterized as individualist, actu-ally focuses on the practices that comprise the shared recognizability ofsituations in their details.

Durkheim’s argument reverses the assumption that interaction ordersare constrained by so-called larger frameworks of norms and values that


ultimately explain their order. For Durkheim practices, as sounds andmovements, come first and give rise to sentiments, norms, values, rep-resentations and beliefs, not the reverse. As such the needs of practiceconstrain the form that social structure can take. While conceptual andinstitutional orders may take on a life of their own once generated, theyhave their beginning in and are maintained through the enactment ofpractices. Furthermore, when institutional orders fail to meet the chang-ing needs of practice societies fail.

The moral forces that result from enacting practices are considered byDurkheim to be not only the origin of valid categories of the understand-ing but also the social facts at the core of social orders. The implicationsof Durkheim’s treatment of practice as concrete sounds and movements,for the current tendency to favor a conceptual view of practices, are pro-found. At the very least the analysis suggests that sociology, in most ofits conventional faces, has not really been either socially, or empiricallycentered in Durkheim’s sense.

Durkheim’s treatment of practice has implications for both the theoret-ical understanding of “structure” and for preferred methods of research.In order to be consistent with Durkheim’s argument, sociology wouldhave to assume an actual order in the enactment of each individual case,which would be directly available to observational research methods,rather than trying to establish tendencies which must be “modeled.”1

At present, qualitative empirical studies of concrete social practice findthemselves assigned rather unhappily to the interpretive side of the so-called micro macro debate; the current state of understanding classicalsociology leaving their commitment to the details of concrete practicestheoretically unfounded and difficult to defend. An understanding ofDurkheim’s commitment to witnessably enacted practices remedies theperipheral status such studies now have, revealing them as essential notonly to disciplinary thought, but to studies of scientific, technological andbusiness practices worldwide.

As important as these implications are, the implications for philosophyare potentially greater, but less clear. If it is possible to establish a validepistemological argument on the basis of studies of enacted social prac-tice, then current arguments about epistemology, and particularly thoseof Pragmatists and social constructivists, will, as Durkheim recognized in1912, need to be revised. Sociology and a sociological theory of sharedenacted practice, would replace the current vogue for theories of practice

1 See Garfinkel’s discussion of Parsons’ plenum (1988) for an extended consideration of thepractice of modeling a hypothetical order in traditional sociology versus the assumptionmade by Garfinkel that each individual case displays a witnessable order which is availablefor research.

Conclusion 319

based on individual action, systems of belief, or paradigms, and sociologywould find itself at the heart of the epistemological debate, determiningthe criteria of validity, instead of at its periphery.

The implications of a viable theory of practice for moral philosophy arealso profound. The idea that moral reasoning is embedded in constitutiveaspects of practice is at least as old as Kant. His example of lying as apractice with constitutive features that it would be illogical to violate (inthe Prolegomena) being a primary example. In the mid-twentieth centurythe idea took a Wittgensteinian twist in the argument for “Two Conceptsof Rules,” by John Rawls (1953). Goffman and Garfinkel both suggest asimilar moral relevance to constitutive aspects of practice. In their viewboth mutual intelligibility and the social construction of self depend on amutual commitment to shared practice that is deeply moral because bothself and social order fall apart without it. Because of an inherent individ-ualist bias in philosophy a compelling empirical or logical support for apractice view has not emerged there. A comprehensive sociological theorythat demonstrated the constitutive character of the elements of practiceswould bolster a practice view of both morality and social order (see Rawls1983, 1987 and 1990 for preliminary arguments in this regard).

The argument that maintaining the practices necessary to create col-lective moral feelings is fundamentally what religion is and has alwaysbeen about is also of great importance. The identification of religionwith beliefs has led to the idea that religion necessarily creates bound-aries between believers in different religious narratives. Durkheim dis-agreed. He presented religion as sets of practices that not only createemotional solidarity, but also provide the possibility of communicationacross boundaries of belief. Certainly, in its guise as systems of belief,religion does tend to reinforce boundaries. But, if Durkheim is right thatthe beliefs themselves are only incidental to religion, and that it is theenacted practices and the need for group participation in them, that con-stitutes the real function of religion, then situated practice offers a wayof crossing boundaries. On this view the answer to the question how tostrengthen the moral ties between people in modern society would get avery different answer.

Religion as practice, separated from belief, should become more impor-tant as beliefs become less important. Religions, as systems of belief thatjustify participation in practices are weakening their hold on the individualconsciousness, as the division of labor increases because people need todevelop more international, even global, ways of thinking about humanity.But, at the same time that the need for shared beliefs is reduced, peopleretain the need for participation in shared practices. In fact, this needmost likely increases in the modern case, as practices must continue to


maintain intelligibility and social order, now with little justification fromshared beliefs. Consequently, shared public forms of practice that couldprovide both a moral and an epistemological foundation for public lifemust be maintained and the commitment to them must be strong.

Studies of the interface between technology and human work increas-ingly suggest that a failure to take the demands of practice into accountleads to problems in technological application (Lucy Suchman, 1999;John Seeley Brown and Paul Daguid, 2002; Christian Health, 2000). Itis a mistake to focus training on beliefs and attitudes. The design of thetechnology itself must take the human commitment to particular formsof shared practice into account.

This commitment to practice is what Goffman’s involvement obliga-tions and Garfinkel’s “trust” are about. The intelligibility of social actiondepends on its recognizability and the ability to “trust” that the otherspresent are working with the same practices and are who they say theyare. It is not only a social necessity. As Kant argued, it is a matter of purelogic. Rational beings cannot will to violate that which both reason andthe capacity to be social self-reflective beings depends on. They might dothings that violate these principles, but if they are thinking rationally theycould not rationally will themselves to violate the practices that reason andself depend on. The thought itself would be contradictory.

Objections to Durkheim’s epistemology will, and should, remain.While a good case for his position can be made, it is ultimately moreimportant to see his work as initiating an important line of argumentwhich, while not explicitly recognized, has always been critical to thesociological enterprise. It would be a mistake to evaluate his argumenton the basis of a standard which it was incapable of achieving, given thelimitations of research methods at the time. Because Durkheim’s epis-temology depends to a large degree on the empirical details of actualshared enacted practices, details which constitute the witnessable enact-ment of social facts, but, also details which the research practices of thetime shed relatively little light on, it should not be surprising if there is acertain incompleteness in the argument and a degree of inaccuracy withregard to details. Modern technological aids to data collection and thesophistication of contemporary field work, when contrasted with the lackof attention to such details at the turn of the century, guarantee that thiswill be the case.

It is to modern studies of shared enacted practice that we should lookfor an evaluation of the potential of Durkheim’s view: studies which aregenerally seen as having no epistemological or general theoretical impli-cations, but, which turn out to be at the center of the Durkheimian projectproperly understood. Splitting Durkheim’s argument in two halves andturning it against itself has not only obscured the relevance of such studies

Conclusion 321

to the foundation of social thought that Durkheim outlined, but has madesocial theory as a project appear self-contradictory.

11.1.0 The Development of Two Conflicting Durkheims

It is possible to separate the different misinterpretations of Durkheim’stext into several distinct phases, which owe as much to each other, asevolving iterations of misunderstanding, as they do to the economic,social and political contexts in which they were framed.

The first phase of misunderstanding occurred in both Europe andAmerica immediately after the initial publication of The Elementary Forms.In this phase the misunderstandings were occasioned by an implicit indi-vidualism on the part of the critics. Given Durkheim’s insistence thatthe individual social being did not exist as such, assuming such anindividual as a starting position from which to evaluate his theory hadpredictably bad results. As a consequence, influential critics, whose inter-pretations would stand throughout the twentieth century as authoritieson Durkheim’s text, treated Durkheim’s argument as a shoddy piece ofcircular thinking that assumed a preposterous group mind theory. Theyalso argued that the text was fraught with contradictions that revealedDurkheim’s inadequacy as a philosopher.2 These interpretations of thework have, unfortunately, survived the various iterations of misinterpre-tation mostly intact, and, as a consequence, continue to negatively impactDurkheim scholarship today.

The second phase of misinterpretation seems to have been associated inFrance, first with Levy-Bruhl and then with Levi-Strauss, and in Britainwith Malinowski and Radcliff-Brown. In this second phase, Durkheimwas interpreted as an idealist. The Elementary Forms was said to beabout the relationship between social structure and religious beliefs andconcepts: a sociology of knowledge. In the British drive to establish socialscience on an empirical footing, this led to the rejection of Durkheim’swork altogether, while in France, it led to the increasing popularity of hiswork as a study of the logic of conceptual structures in society. As a result,French structural anthropology, particularly in the form made famous byLevi-Strauss, claimed The Elementary Forms as a founding classic, whilein England Durkheim was hardly discussed.

Talcott Parsons reports first encountering Durkheim in the context ofnegative characterizations of his work as idealist, when Parsons was at

2 The most influential initial critics in English, who continue to be cited, were WilliamDennes, Charles Elmer Gehlke, and Charles Schaub. See Rawls 1996b “Durkheim’sEpistemology: The Initial Critique, 1915–1924” for an extended analysis of theirarguments.


the London School of Economics in 1923.3 In spite of this introduc-tion, however, Parsons felt that Durkheim’s early work was suitable toan empirically based sociology and deserved to be saved, even thoughhe agreed with the critics that his later work was hopelessly idealist. ThatDurkheim had himself rejected the distinction between realism and ideal-ism that was increasingly being used to characterize his work, only servedto convince Parsons and others that he had, in his later years, lost hismind.4

In attempting to rescue Durkheim’s early work from the negative eval-uation of The Elementary Forms in Britain and the US, Parsons initi-ated the, by now famous, distinction between an earlier positivist and alater idealist Durkheim. By means of this argument, Parsons was ableto usher in a third phase of misunderstanding; introducing Durkheimto American sociology as a positivist empiricist whose later work wasnot important. Consequently, Parsons incorporated his interpretation ofDurkheim’s early work into The Structure of Social Action ([1937]1968)arguing that Durkheim was a realist and empiricist, not an idealist, ignor-ing The Elementary Forms, and championing the argument of The Divisionof Labor which had only been published in the US in English translationin 1933. Given the political and economic climate of the US in the yearsbetween 1930 and 1960, Parsons may have done well to eliminate fromhis interpretation of Durkheim any elements that smacked of either ide-alism or socialism.

Needless to say, however, the “Two Durkheim Hypothesis,” as it cameto be known, constituted a hopeless butchery of Durkheim’s argument.Not the least of its problems was that it was built on a distinction betweenthe real and the ideal which Durkheim had argued repeatedly was a falsedistinction, but which led in Parsons’ hands to interpreting as positivistDurkheim’s early arguments for an empirical method to ground sociology.These arguments would then come under fire unnecessarily with therejection of positivism in the 1970s.

Not withstanding these problems, this version of Durkheim stood firmin the US until the 1970s. Similarly, the idealist version of Durkheim

3 According to Parsons ([1937]1968:viii, emphasis in original) “In 1924–25 I spent a yearas a research student in sociology at the London School of Economics . . . Durkheimwas of course known in both England and America, but discussions were overwhelminglyderogatory; he was regarded as the apostle of the ‘unsound group mind’ theory.”

4 In The Structure of Social Action ([1937]1968:304–7, 445–7), Parsons dismissedDurkheim’s argument that reason has its origin in emotions with the observation thattoward the end of his life Durkheim had gone “clean over into idealism.” This cavalierattitude towards Durkheim’s position was taken up by many twentieth-century Durkheimscholars and essentially allowed them to make use of what they wanted of Durkheim’swork, without having to take seriously Durkheim’s own views. See also Rawls 1996:468–76 for a discussion of this problem.

Conclusion 323

popularized by Levi-Strauss remained popular in anthropology untilabout the same period. The two versions really represented two sidesof the same coin. Each took one side of a position that Durkheim hadargued only makes sense if the two sides are kept together and both over-looked Durkheim’s account of how emotion and meaning are created inand through practices. Consequently, both sociology and anthropologywere handicapped. In the 1970s, in France, poststructuralism advanceda critique of the position in its guise as structural anthropology. This cri-tique made its way to the US in the late 1970s. In its wake a reappraisalof Durkheim began to occur in the US and Great Britain in sociology.In France, however, the critique tended to come from the humanities,remained essentially individualist and did not involve a reappraisal ofDurkheim to any significant extent.

This critique ushered in a fourth phase of interpretation: a politicallyliberal, sometimes even referred to as “radical,” Durkheim. One of thethings Parsons had done in transforming Durkheim for a politically con-servative American audience, was to make him appear more politicallyconservative than he was. Parsons had also translated Durkheim’s socio-logical method into a mathematical form for a population that believed inthe truth of numbers. While this probably did a great deal to popularizeDurkheim in the earlier period, by the 1970s it had become a prob-lem. What Parsons had played down was Durkheim’s strong egalitariancommitment.5

In the face of increasing criticism of the conservative Durkheim in the1970s Anthony Giddens (1971) and Steven Lukes (1973) began to pointout that Durkheim was not the political conservative that he had beenportrayed as. In The Division of Labor, he had even argued that inequalityconstitutes a contradiction in what he called the functional prerequisitesfor capitalism (Rawls 2003). This created a new interest in Durkheim’swork. The problem is that while this view presented a fairer appraisalof Durkheim’s politics, it did nothing to clarify his epistemology or hisemphasis on practice and, as a consequence, the understanding of hiswork as a whole became increasingly conflicted.

In the 1980s, as sociology began to embrace the postmodern and/orpoststructural dilemma, scholars became attracted to the French versionof Durkheim’s sociology of knowledge, bringing it into sociology. In this

5 Although Durkheim dismissed socialist beliefs as unscientific, as he dismissed all beliefs,because they distort the underlying realities they stand as accounts and justifications of, heargued in Book III of The Division of Labor that inherited wealth and privilege, which arestaples of western capitalism, are socially dysfunctional to the point where if the injusticesof inequality are not overcome he maintained that the division of labor in its modernwestern form would cease to exist.


fashion, Durkheim came to be portrayed as a cultural sociologist, and apragmatist; a fifth phase of interpretation (spearheaded by Jeff Alexanderand Hans Joas respectively). This had the positive result of once againincreasing the interest in Durkheim within contemporary sociology. How-ever, the interpretation of Durkheim as a cultural sociologist and/or apragmatist continued to attribute to his position the same distinctionbetween idealism and realism, beliefs versus practices, the emphasis onindividualism, and the tendency to privilege ideas over practices, that hehad so consistently argued against. His emphasis on culture and ideas wasonly part of a larger argument. Misunderstanding Durkheim’s intentionsin this regard, undervaluing the centrality of practice to his argument andnot appreciating the relevance of his epistemological position to his soci-ology as a whole, was what had divided the interpretation of Durkheimscholarship into two conflicting camps in the first place. It could hardlyrestore the unity of his position.

11.2.0 The Fallacy of Misplaced Abstraction

The contemporary sociological emphasis on concepts is one of themost serious obstacles to Durkheim’s view of practice. The transpos-ing, of Durkheim’s careful rendering of concrete actions, into beliefs, ispart of a general tendency among both sociologists and anthropologiststo translate practices into conceptual terms. This tendency, Durkheimwould argue, results from placing the individual before the social epis-temologically. As a consequence, mental impressions appear to be real,whereas actual physically enacted practices, which are real, appear to bemerely conceptual. I refer to this as the “fallacy of misplaced abstrac-tion.” The abstraction is misplaced because the coherence of sounds andmovements; their mutual intelligibility, must have at least as much todo with the recognizability of the particular sounds and movements thatcomprise an enacted practice, as do beliefs about the practice. A seriesof movements and sounds must be recognizable as a particular sort ofpractice, before beliefs and concepts about that practice can be invoked.

There was, in the earlier part of the twentieth century, a tendencyin philosophy and social science to ignore the level of abstraction nec-essarily involved in descriptions of things. Descriptions were for a longtime treated as either synonymous with the objects they described, or asproblematic only through incompleteness. That words pulled “objects”out of the flux of experience and located them in time and space, inaddition to categorizing them in ways that ignored essential elements ofobjects as particulars, is a problem at least as old as Hume. Yet, scienceand philosophy at the end of the nineteenth century often proceeded as

Conclusion 325

if reality already contained all those characteristics that concepts in factimposed upon it.

Alfred North Whitehead ([1929]1978:7–13) referred to this belief,that conceptual representations of reality as concrete, reflected a realconcreteness in reality itself, as “the fallacy of misplaced concreteness.”Since Whitehead’s time it has become a commonplace that descriptionsof “things” necessarily involve a level of abstraction which makes cer-tain sorts of scientific and philosophical arguments highly problematic.As a result there is a tendency in modern sociology and philosophy totreat experience as equivalent to conceptual systems and correspondingbeliefs, rather than things.

Works which feature the word “practice” in their titles, such asBourdieu’s A Theory of Practice, or Steven Turner’s The Social Theory ofPractices, tend to focus on the conceptual limits of belief at the expense ofthe witnessable details of practice. The fact that knowledge is ultimatelyshaped by the human perceiver is interpreted to mean that concepts,beliefs, and attitudes alone define the limits of human knowledge.

While Whitehead was quite right to point out the fallacy of misplacedconcreteness in the positivism and realism of his own day, it is equallyproblematic to argue that reality exists wholly and entirely as systemsof concepts. I argue that, with regard to the domain of social practice,this is a fallacy in the opposite direction. Hence I refer to it as “thefallacy of misplaced abstraction.” Social practices which, unlike natu-ral events, are essentially concrete witnessable events produced by, for, and inthe presence of others, and produced to physically enact social classifica-tions, have been reduced to conceptual abstractions, as if those abstractionswere not only an inherent characteristic of the social event but definedits publicly recognizable character. This process replaces the meaning-ful empirical enactment of the practice for members with an interpre-tation by the researcher and treats that interpretation as the originalthing.

The “fallacy of misplaced abstraction” needs to be distinguished fromsimple idealism, because it doesn’t treat the world itself as ideal; it isonly the limits of human knowledge that are supposed to be constitutedby ideas. Persons are thought to live in a demonstrably empirical world.But it is believed they can know nothing valid or useful about that worldon an empirical level. The turn is toward an examination of conceptsand their relation to one another within conceptual universes, referred toas “the linguistic turn.” Practices, as recognizable and repetitive soundsand movements, are of no interest on this view. They are thought, likenatural objects, to be seen only through the lens of concepts, and aretherefore considered to be necessarily subject to “the fallacy of misplaced


concreteness.” The obvious contribution of a reading of Durkheim’s ownsociology of knowledge to this fallacy is ironic.

As a consequence of this fallacy, contemporary sociology, philosophy,and the humanities find themselves at something of an impasse. Hav-ing come to see the world as consisting primarily of concepts, beliefs,norms, and values: in a word; texts and narratives, contemporary socialtheory has some problem explaining the relevance and validity of empir-ical research. If society is viewed as primarily conceptual in nature, thenempirical happenings would provide no more than crude clues to what isgoing on in the individual or collective narrative. On this view, one eitherstudies aggregates of individual happenings and hunts for patterns, asclues to underlying beliefs and norms that are said to “constrain” indi-vidual behaviors, or one focuses on interpretation and the possibility of asociological study of interpretive practices. The first leads to quantitativemethods, and the latter to interpretive or narrative sociology.

Having rejected positivist and realist approaches to the problem ofknowledge, and embraced the fallacy of misplaced abstraction, truth hascome to be defined in terms of conceptual structures of belief, a domainwhich is unavailable to empirical inspection. Truth has been reduced to“overlapping consensus” and the idea of an essential human “reason,”which has been the mainstay of western philosophy for several hundredyears has seriously eroded. The disciplines find themselves in the midst ofthe postmodern or poststructuralist dilemma, wherein reality is definedby conceptual structures of belief to which no one has empirical access.

11.3.0 The Sociological Dilemma

The Postmodern and/or Pragmatist dilemma poses particular problemsfor sociology because, unlike the humanities, the domain of the socialis not initially a text, although texts and narratives are common socialphenomena. Not only does the social domain not consist entirely, or evenprimarily, of concepts, it presents itself in the first instance in the form ofsounds and movements which persons make in the presence of, and withan orientation toward, others. In order to treat social reality as essentiallysymbolic or conceptual, concrete social phenomena must first be ren-dered in conceptual form. While this is always a problem, it is particularlycritical in understanding essential worksite practices in a technologicalage.

Treating the coordinated movements of social actors as a text andreducing the meaning of those sounds and movements to concepts isproblematic not only because it introduces relativity and contingencyinto the study of the social, but because it eliminates the empirical elements

Conclusion 327

of that domain in their entirety. Such a reduction of sound and movementto concepts, assumes that the movements and sounds themselves play norole in creating and recreating the meanings, concepts, and social formsthat eventually make their way into texts and narratives. It assumes thatconcrete social practices can be adequately characterized in conceptualterms. Researchers find themselves in the ironic position of trying to studydetailed worksite practices by reducing those details to concepts that losethe worksite practice altogether.

This assumption asks narrative and belief to play too large a role. Nar-rative and belief statements are a specialized form of social practice. They canbe studied as accounts, formulations and narratives. They occur relativelyinfrequently, however, generally only at points when some contingency orambiguity has arisen and, as Durkheim argues, bear a special relation toongoing practices operating retrospectively as accounts or justificationsof what has just occurred, but, not prospectively to order interaction.They do not accurately portray practices either, because that is not theirpurpose. Furthermore, not all recognizably meaningful practices have, orcould have, beliefs, narratives, or accounts, corresponding to their detailsas practices. Some actions are accountable and generate narratives ofaccountability and motive (Mills 1940, Garfinkel 1940, 1967). Othersare enacted with a degree of taken for grantedness that belies a foun-dation in belief and narrative (Rawls 1987, 1989). The latter compriseinteraction orders of conversation and interaction that are utterly basicto intelligibility and social order and as such must be incorporated intoany systematic theory of social order.

The result of trying to avoid “the fallacy of misplaced concreteness” isthat empirical studies of all sorts, including the best classical field stud-ies, have come under fire by postmodernist and poststructuralist thinkers.Fieldworkers have been forced to defend the validity of field experience.6

6 For instance the attacks on Whyte and Becker and the defenses by Emerson and Hallinanof field work. As if the truth of the social organization of men in Boston’s North End con-sists exclusively in their own “beliefs” and “narratives” about that organization rather thanin an analysis of their actual social behavior. If Whyte or Becker or any of dozens of othergreat field researchers had relied for data on reports by informants alone and on the basisof informant narratives alone had claimed an understanding of the social organization ofthe “corner” or “musician culture,” their work would never have received the recognitionit did and they would not even be available as contemporary targets. An interesting irony.We would have wanted to know if the informant’s beliefs about, and accounts of, theirown behavior were “true.” Yet, in the ongoing debates over classic fieldwork, currentrecollections of “formerly held beliefs” have been used to challenge the validity of carefulobservational descriptions of social behavior. This only seems backwards, of course, tothose who recognize that a reduction of social behavior to belief and narrative is a fallacy.

Mills on accounts is relevant here as well. Accounts (a special form of narrative) accord-ing to Mills, are social phenomena and their acceptable form and content is shaped by


The conceptual typifications created by social science are treated as a pri-mary reality in place of the concrete lived experience of actors engagedin enacting social practices. Any appeal to empirical details is treated aspositivist and thereby subject to Whitehead’s fallacy. Nothing but con-cepts, symbols, and the interrelationship between concepts in a systemof symbols is to be treated as real. Culture, gender, language, class, andindeed a whole host of other social phenomena, including social struc-ture itself, have come to be rendered entirely in symbolic or conceptualterms.7

Social experience is treated as an individual confrontation with norms,values, or concepts, within a shared context of group expectations. Thisignores an important aspect of social practices, the recognition of whichis foundational to the discipline of sociology (Rawls 1998). Practices canonly be enacted in assembled groups in the form of a witnessably repro-duced regularity of sounds and movements. Concepts themselves can-not be witnessably produced for others. The actual lived experience ofsocial practices, consists of an elaborate kaleidoscope of concrete empiri-cal detail. Those details are not themselves irrelevant. Durkheim’s theoryof practice points the way toward a solution to this dilemma. It was, afterall, written in the first place as a solution to just these problems as theywere debated at the end of the nineteenth century (Rawls 1997b).

11.4.0 Scientific Things versus Social Things

The original debate focused on the problem of truth in individual andreferential contexts. However, there is a significant difference betweenscience, the humanities, and sociology, which the tendency to rendersocial practices in conceptual terms overlooks. For science the objectsof study are not socially organized in their own right. Therefore, any

the social structures in which they are constructed and toward which they orient. Theaccounts given by Corner Boys in the 1930s were sensitive to the particular social struc-ture in which they were constructing their social lives, the social structure which Whytewas attempting to uncover by reporting their activities, and therefore those accountsmight have shed some indirect light on that social organization. The accounts whichthey or their sons give in 1990 are oriented toward an entirely different framework ofsocial expectations and could hardly be the same. Nor could contemporary accountsprovide clues to a hidden framework of 1930s social expectations toward which they nolonger orient. It seems obvious that the reduction of social reality to concepts wouldalso entail the relativity of the meaning and validity of those concepts to a particulartime and place. But, apparently the postmodernist critique does not always recognize thislimitation.

7 These conceptual terms are often numerical, or statistical and as such masquerade as“facts”. But, statistics are merely numerical renderings of conceptual categories: opera-tionalized concepts to use the formal jargon.

Conclusion 329

conceptual organization distorts the original nature of things. In the phi-losophy of science the recognition of the symbolically mediated characterof experience has led to a focus on “paradigms,” universes of discourse,and systems of symbols as they define scientific practice in scientific lab-oratories. From Kuhn (1958) to modern studies of laboratory practice(Douglas 1966, Bloor 1982), the emphasis has been on studying the pro-cess whereby the apparent concreteness of natural phenomena is replacedby their symbolic counterparts in an attempt to study the influence ofsystems of symbols on scientific practice and discovery.

In the humanities the objects of study generally consist of experiencesrecorded already in conceptual form; in the form of language, narrative,film, visual art, music, or text. Therefore, what any given universe ofsymbols signified at a particular time in history or culture is crucial tounderstanding a particular text. Furthermore, over the course of historycertain voices were allotted very narrowly circumscribed textual forms,while others had essentially no voice.8 There are not one, but many his-tories, each of which may in essential respects be true. There can be noaccess to the original sounds and movements that accompanied the con-struction of those texts and therefore, texts have to be deconstructed toreveal the multiple possibilities from which a particular version has beenselected.9

Sociology stands in a very different relationship to its subject. Whereasthe humanities must try to reconstruct the context in which the text wasoriginally constructed, the sociologist stands in the midst of the ongoingconstruction of texts. Society is produced and reproduced all around usall the time. As persons sociologists are constantly taking part in social

8 Women, for instance, have often had very few textual outlets which were normally availableto them. The publication of religious conversion narratives was one such outlet and it hasbeen argued that the historical recovery of women’s narrative voice requires attention tosuch forms. Similarly, slaves as property show up in tax and inheritance inventories, whilethe lives of the freeborn poor, with the possible exception of birth, death and immigrationrecords, have been largely unrecorded.

9 With historical texts this has recently involved many interesting attempts to recreate theactions or journeys recorded in the text. Engaging the subject of the text in concrete termsoften gives researchers new insights into how things might have been done in a particularhistorical period. Historical reconstruction of the roof over the Roman coliseum, forinstance, shows that one popular model of how the roof was constructed would have beenhighly impractical because the roof could not have been retracted. Other roof models builtto 1/3 scale were not only able to retract, but historians discovered during their recreationthat this type of roof could be cantilevered backwards and sideways for repairs, makingthis a very easy and versatile roof to manage. The enacted aspects of the manipulationof the roof model are what is important to note about this research. Physical movementsinvolved in the process, the physical limitations of bodies, and the mastery of practicesthat weren’t imagined before the reconstruction began were made evident by the physicalenactment of the reconstruction. Thus physical aspects of ancient practices are emergingas essential to an understanding of the historical process.


scenes and they do have access to the original scenes as they unfold. Soci-ology as a discipline is interested in all forms of social life. Therefore, thedata are everywhere. The job is not to interpret already existing texts,but to see how it is that persons recognizably and witnessably producefor one another, out of the infinite detail of lived experience, the mutu-ally understood sounds and movements that are the prerequisite for theeventual construction of texts, narratives, accounts, statistics, and so on.

The task of understanding this social process does not need to giveform to an otherwise formless flux of experience as does natural science.10

Social practices are produced to be understood by others and are imme-diately experienced as orderly and meaningful by competent membersof social groups in most cases. Neither are social practices enacted inconceptual form, as are the typical objects of study in the humanities.Social practices appear to be infinitely varied in their concrete details andyet are recognizable to participants in social groups. The task is not togive multiple interpretations to a fixed text, but rather to explain how sucha variety of sound and movement could come to have a single recognizablyreproducible meaning for any particular assembled group; to understandhow a social world which is composed of meaningful social forms andpractices is recognizably reproduced in and through these social practices.

11.5.0 Recognizable by Design

Social practices are designed to be recognized and understood, in a spe-cific time and place, by a specific assembled population, within a specificsequence of events. Analyzing how the design achieves its purpose is not thesort of distortion that explaining oxygen or gravity from a human per-spective is. The meaning of social practices is not cosmic, or timeless; it isa situated human construction out of witnessable sounds and movements.

That persons also share beliefs and narratives about the practices theyenact at a particular place in a particular time cannot be denied. In fact,institutional accountability requires this. Nevertheless, some sounds andmovements count as relevant within particular social settings and somesounds and movements do not. This is true even though narratives andsymbols may be the same across situations. The ability to produce soundsand movements that can be seen by others to be just the sort of soundsand movements that count in a particular situation is necessary for therecognizability of social practice. The witnessable details of sounds and

10 Not that nature is itself formless. But the human experience of nature is formless untileither human or social meanings are imposed upon it.

Conclusion 331

movements are the key to how and whether they will be recognized andunderstood by real persons in real time as practices of a particular sort; thesort that count in a particular situation. Before shared beliefs and narrativescan come into play this recognizability must be achieved. The details ofthis achievement of recognizability are available to the researcher.

It is when situated practices are not recognized, disrupting the taken forgranted character of the practice, and creating a breach in the ongoinginteraction, that interpretations and narratives come into play. Some-times these narratives take the form of accounts – recognizable reasonsor excuses, for the breach. At other times they stigmatize the individualwho has produced the unrecognizable practice, characterizing them asabnormal or crazy. If categories of race, gender, occupation, or culturalsubgroup etc., are available and can be invoked then the accounts canbecome “outgroup” narratives that create stereotypes of “them.” It is theessential characteristic of outgroup narratives that they are not shared.But, once the interaction has moved to the level of accounts it is becausethe interactional commitments necessary for mutual intelligibility havealready been disrupted to some degree, and accounts, which producemore distance could not re-establish that commitment even if they wereshared (Rawls 1995).

Sociologists who study enacted social practice do not confront a real-ity that is already coded into conceptual typifications. They may codeit as fast as they experience it (although coding is not recommended,as it renders essential details invisible). But, the enactment of practiceis nevertheless there in front of them in all of its infinite detail. Nordo these details need to be coded into sounds and movements out ofa constantly changing flux. The movements that constitute practices areconstructed as recognizable social movements in the first place by the peo-ple who enact them. “Words” are not fluxes of sound on which personsimpose conceptual boundaries, although they are often treated that way.Language is not an arbitrary coding system for sounds. Rather, “word-sounds” are produced to be identifiable as words, and not only as words,but as specific words, in specific sequences of words.

Conversations are constructed so that word-sounds comprise turnswith a recognizable structure, with identifiable turn completion points,pauses, repairs, assessments, and other conversational structures (Sacks1992; Sacks, Schegloff and Jefferson 1974). Participants in the conver-sation watch for these so they will know what to do next. The detail ofsuch constructions is difficult to “see” when, as Garfinkel (1967) pointedout, the process of producing and recognizing such detail is “taken forgranted.” In fact, trying to think about it while enacting practice makesthings seem hopelessly complex. It helps that the research process can be


aided by modern methods of audio and video data recording. However,even traditional field methods, depending on nothing more than obser-vational skills, frequently preserve interaction in enough detail so that thewitnessable sounds and movements that constituted the social order andmeaning of the original scenes can be recognizably reconstructed.

The real problem is seeing the relevance of the details of practice totheir meaning. Treating meaning as a matter of concepts results in tryingto get meaning out of an interpretation of content and/or symbol. Theproblem is that “content” or “symbol” does less to define meaning thanthe sequential form of practice. What is much more effective and happenswhenever there is no breach in practice, is that the sequential juxtapo-sition of words tells participants how to hear them. For instance, wordscan be heard as “invitations” or “brush offs” depending on their place-ment in a sequence of talk. The words used to indicate acceptance of apre-invitation, like “nothing” as a response to the question “what are youdoing tonight?”, may not have a “content” that would indicate the inter-actional involvement work they do in a sequence. What does “nothing”mean after all? But, the meaning of that involvement work will, never-theless, be evident from the placement of the words in a sequence. If aperson says “nothing” then they have invited, they will be expected toaccept the invitation that will follow. If they do not they have not ful-filled their interactional obligations. Saying “nothing” commits one to aparticular course of action.

In spite of these differences between sociology and the humanities andsciences, however, sociologists constantly speak as though the problemsconfronted by science and the humanities are also our problems. Thesocial sciences have for years ignored the concrete witnessability of socialpractices in favor of elaborate conceptual models and corresponding sta-tistical analyses. In fact, sociology has gone much farther than the sciencesin holding itself to a false model of abstraction. In spite of a sensitivity toWhitehead’s fallacy, science still has “things” in front of it and does notin practice substitute descriptions for things. However theorized these“things” necessarily become, and however much the descriptions mayshape the understanding of the thing, science retains relatively unmedi-ated access to its basic things. A science based purely on descriptionwould be no science at all. Scientists remain in the laboratory.

Social science could retain its objects in just this way. Social scientistscould remain in the field and could preserve the details of the practicesthey study. However, social science generally assumes that the under-lying narrative of beliefs and norms, which are in fact only retrospec-tive accounts, and the behavioral tendencies allegedly generated by anorientation toward those beliefs, are more important than the concrete

Conclusion 333

witnessable details of social practice. This is true in the case of both poststructuralism and Parsonian structuralism. It is true for both interpre-tive sociologies and traditional so-called macro sociology. In each case,actions are analyzed in relation to norms and values, or concepts, andnot in their own right. It is assumed that the “reality” of social forcesis to be met with only at a general or conceptual level. Even Goffmanultimately fell victim to this problem, trying to reduce situated action toconceptual typifications (Rawls 2002). It is only in aggregate tendenciestoward norms and values, or in shared beliefs, that patterns of order areexpected to appear. The realm of immediate social experience is treatedas an infinite and chaotic flux.

The only dissenter has been Garfinkel and those in Ethnomethodologyand Conversation Analysis who followed him. One reason why interac-tionists in general have not focused on the details of practice may bethe prevalence among those who have studied interaction for focusingon interaction in breach. In other words, crime, deviance, race, gender,class, etc. If practices only unfold prospectively in a taken for grantedmanner when interaction is unproblematic, then studies of deviance willfind a predominance of narrative and interpretation, and even when fieldresearchers proceed with great care, their research will indeed land themin the midst of accounts and justifications. What has been highlighted bythe prevalence of a focus on deviance is the way in which retrospectiveaccounts are invoked when things break down and life appears to be mas-sively contingent. The prevalence of narrative in situations of breach hasbeen mistaken for the routine order of social life.

For example, when a person has a presentation of self that fits theexpectations of a given situation, they can walk up and participate in thatsituation. They may not know what type of situation they are walkingup to. It may be a game of some sort that they do not know how to play.Nevertheless, with no prior knowledge of the game, they can walk up andthrough a process of watching, trial and error, and being instructed byothers, they can quickly learn enough to participate without problems.When the police walk up to a situation, however, it tends to evaporate infront of them. Because of this they must guess what is going on, fallingback on accounts and narratives, because they do not have access toongoing practices. When field researchers study the police, what theysee is the endless process whereby the police must invoke categories andtypifications in order to make sense of the myriad contingencies withwhich they are faced on a daily basis.

What is not so obvious is that the police only have to do this becausethey are not themselves able to participate in the situated practices itis their job to observe, and therefore must guess what is going on.


Unfortunately, there has been a tendency to generalize insights drawnfrom research on crime and deviance to social theory proper. Thus, theperception is created that understanding what is going on in situationsalways requires categorization, typification, and interpretation. The theo-retical project becomes one of trying to come up with a comprehensive setof categorizations and an adequate theory of the process by which personscome to be able to learn and apply the necessary categories. Thus, whatstarted as a detailed interactional observation of practice, often quicklybecomes cognitive and returns to a focus on concepts. This has hap-pened not only among Postmodernists and Symbolic Interactionists, butalso among some Ethnomethodologists.

11.6.0 Intelligibility as a Constraint on Practice

Durkheim not only rejected the idea that arbitrary social representations;concepts, determine the form of social practice and human reason, hereversed the equation; arguing that universal prerequisites for reasonplace heavy constraints on the forms social practice can take; beliefs andvalues serving retrospectively to support the necessary forms, not to cre-ate them. There can be no society without mutual intelligibility and nointelligibility without logic and reason. Therefore, in order for society tocome into being, practices must develop which create and sustain humanreason. Even after society has been formed, it is necessary that the prac-tices which people engage in be designed to create and sustain mutualintelligibility. Practices which do not, or which, because of social changes,stop being able to do this will stop being enacted.

Durkheim’s arguments, in The Division of Labor, with regard to thenecessity of justice in modern society, along with the argument of Suicidethat an optimum degree of social connectedness is necessary for the sur-vival of the individual, make only slightly different versions of the sameargument. With regard to justice Durkheim argued that because equalityin the conditions of contract are required for equal competition, and thedivision of labor only functions well under conditions of equal compe-tition, justice is a functional prerequisite of the division of labor. Withregard to social connectedness he argued that the integrity of the indi-vidual self (as a social production) requires a certain degree of embed-dedness in social practices. Too much is just as bad as too little. Societieswhose practices produce too high a level of contingency have high rates ofsuicide. Similarly people (often married women according to Durkheim)who find themselves in unfulfilling situations that they see no hope ofchanging have high rates of suicide. In both cases the need for social con-nections and for justice operate as constraints on the form that society can

Conclusion 335

take. In The Elementary Forms the argument is that the need for reason asa functional prerequisite for social order operates as a constraint on theform that social practices can take.

If what Durkheim meant by practices had conformed to the currentgeneral understanding of practices as norms, concepts or beliefs, thenit would have been illogical for him to argue that the need for intelligi-bility acted as a constraint on practice. It would have been illogical toargue that “enacting” a practice would change a person, or cause thedevelopment of rational ideas if practices were ideas. Clearly, in order toenact concepts, norms, or beliefs, one would already have to have them. Thus,interpreting Durkheim’s theory of practice in the conventional way, ren-ders his argument viciously circular, as critics have not been reluctantto point out (Dennes 1924). The conceptual interpretation of practicesalso leaves unanswered the question of where these concepts, beliefs, andnorms come from in the first place, lending credibility to the argumentthat Durkheim must have held a rationalist or Kantian view of the originof the categories of the understanding.

The failure to appreciate the constitutive relationship between enactedpractice and general social forms, has resulted in the current split insociology between those who practice a naive statistical conceptualismand those who reduce everything to a level of conceptual reality. The naiveempiricists of course do not understand the problem. Those who focuson concepts, however, presumably do. There are several consequences ofthis current emphasis on concepts.

First, the details of actual concrete cases of social exchange appear tobe essentially irrelevant, what is currently of interest is the conceptualapparatus, or belief system, which lies behind the behaviors and whichthey may provide some evidence of. But the actual concrete practices areconsidered to be of no interest in their own right, and empirical researchor evidence of any sort is often argued to be essentially irrelevant on theassumption that as concepts define the limits of reality any individual“truth” is as valid as any other.

Second, as Norman Denzin has pointed out, there is an infinite regress,or “abyss,” when concepts are taken to be the primary social reality. Ifall knowledge is shaped by concepts that are somehow socially perpetu-ated then we are all stuck in a circle of unverifiable conceptual fabrica-tion. Also, communication depends upon sharing concepts, but sharingconcepts depends on being socially constructed in ways which reproducethe inequities of the social system. So the possibility of justice, socialchange, or even critical and creative thinking are hard to explain.

Third, there is also the problem of getting concepts into heads in thefirst place. The attempt to deconstruct would be doomed to failure if


what was being deconstructed was really a primary reality with nothingunderneath to be revealed through the deconstruction.11 The idea thatpersons without concepts of their own, or with purely individual con-cepts, acquire general, or social, concepts by coming into direct contactwith concepts themselves, is problematic: at least as problematic as whatDurkheim was alleged by his critics to have argued. What persons comeinto contact with are the sounds and movements that comprise socialpractices. There can be no direct contact with concepts.

Fourth, because Poststructuralism has accepted the classic structuralistposition that either social order is formally institutionally ordered or thereis indeterminacy, essential theories and methods are being discarded. Ithas become popular to denigrate the possibility of “theory with a capi-tal T.”

There is, however, another way of explaining the order and intelligi-bility of action. This alternative proposes that concrete witnessable prac-tices, through the witnessable details of their recognizability, constitutethe meaningful and expected regularities in everyday life. Instead of for-mal institutions providing the order against which meaning is achieved, onthis view, formal institutions themselves are a production of the witness-able sounds and movements that constitute practices (Garfinkel 1967).The formal orders act, as Mills said, as contexts of accountability, butare powerless to prospectively order either actions or meanings (Mills1940).

The solution is to turn to a detailed study of enacted practice asDurkheim proposed. The problem that any argument on a Durkheimianmodel will have to face, however, is the fact that the analysis dependsheavily on the enactment of what Durkheim calls ritual interactions which(at least in formal institutional terms) play a less prominent role in mod-ern industrial society. More mundane practices must have taken theirplace if Durkheim is right.12

11 There is a more complicated form of circularity with statistics. Basically organizationsproduce statistics as accounts, or numerical narratives, of organizational activity. Theseaccounts address various legal and political obligations of organizations with regard tovarious mandates. The statistics are then used to measure the effectiveness of the man-dates, or the participation of populations covered by the mandate in the organization.The organizational account is taken as independent evidence not only to verify the orga-nization’s own activities, but is naively treated as evidence of general social practice. Inboth cases the account has been allowed to stand for the practices. It is something liketrying to verify that a man is making child support payments and taking an affidavitwritten by that same man to show in court as evidence that he had made the payments.Practices that naive wouldn’t last long in the “real” world if their purpose was to pro-duce a “correct” count. However, for purposes of organizational accountability they aretreated as perfectly adequate as Garfinkel noted (1967).

12 Durkheim makes use of a broad definition of religion which fits any generally sharedenacted practices in which members of assembled groups participate: “We have a religion

Conclusion 337

For Durkheim the experience of moral force was the most importantfeature of enacted practice for providing direct experience of the six cat-egories. Mundane enactments of local interaction orders would seem tofulfill this requirement. There are moral obligations at the level of “inter-action order” wherein everything depends on the mutual commitment toenacted practice (Rawls 1987). This gives interaction a moral dimensionand implicates shame, blame, and trustworthiness (Rawls 1990). Theoriginal sacred character of formal ritual practices may have played a rolein the development of the initial category of classification, dividing theworld in two morally, in order that the original division would have moralforce and thereby socio-empirical validity. But, deviance and stigma mayfulfill the same function in modern society.

Durkheim argued that secular orders would necessarily replace formalrituals as the source of order as the division of labor increased and gaveexamples of laboratory science in this regard (1993, Book III, ChapterOne). Garfinkel, Goffman, and Collins have focused on informal localorders in modern society and shown that they do indeed work to estab-lish and maintain mutual intelligibility, self, social order, and morality.Goffman can be interpreted as arguing for an “interaction order” thatis quite separate from institutional social practices (Rawls 1987).13

Garfinkel has focused on the achievement of intelligibility in and throughlocal orders which are identical with the practices which enact them.He has also introduced the idea of contexts of accountability as a wayof studying the phenomenon of institutional constraint. Collins arguesfor the importance of what he calls “ritual interaction chains” in sus-taining social solidarity and selfhood through time. Durkheim’s notionof enacted practice would need to be interpreted along similar lines asincluding mundane enactments of “interaction order” in everyday life inorder to make it current. However, it has the potential to ground moderninteractionism theoretically in a rather grand way.

The essential ingredient of enacted practice was always the perceptionof moral force by participants, which is in principle separable from reli-gious or institutional constraints. The experience of the mutual creationof moral force, in and through the enactment of shared practice, and

as soon as the sacred is distinguished from the profane” (Durkheim [1915]1912:210).We might take an even greater liberty and say that as soon as the way “we do things” isdistinguished from what “is not done” this distinction has been achieved. That is to saythat the recognizability of practices as “our own”, or appropriate to a particular situation,is the key to their being considered sacred.

13 Interaction orders are not random, individually initiated, or “negotiated” orders. Theyare patterned, recurring, and mutually expected orders the regularity of which isdepended on by participants for achieving intelligibility.


the experience of mutual obligation to the enactment are both character-istics of “interaction order” practices (Rawls 1987, 1990). Durkheim’sown suggestion, in the second preface to The Division of Labor, that pro-fessional associations might play a moral role in modern society, corre-sponding to totemic ritual in traditional society, runs along similar lines.

Unless concrete practices continue to be enacted by assembled groups,Durkheim argues, society cannot be sustained, and individual reason can-not be created on the basis of the derivative beliefs and representationsalone. Therefore, a sociology must engage in a detailed empirical studyof the underlying social facts (practices) essential to human reason, socialsolidarity, and morality. This is a position that Durkheim maintained con-sistently from The Rules of the Sociological Method through The ElementaryForms.

Bibliography

Aarsleff, Hans. 1982. From Locke to Saussure. University of Minnesota Press.Adler, Patricia, Peter Adler, and John Johnson. 1992. “StreetCorner Society

Revisited: New Questions About Old Issues.” in The Journal of ContemporaryEthnography. April.

Alexander, Jeffrey C. 1982. The Antinomies of Classical Thought: Marx andDurkheim. Berkeley: University of California Press.

1988. Durkheimian Sociology: Cultural Studies. Cambridge University Press.Alexander, Jeffrey C. and Steven Seidman (eds.). 1990. Culture and Society.

Cambridge University Press.Allcock, John. 1983. Editor’s introduction to Durkheim’s Pragmatism and Sociol-

ogy. Cambridge University Press.Becker, Howard. 1963. Outsiders. Chicago: Free Press.Bloor, David. 1976. Knowledge and Social Imagery. London: Routledge and Kegan

Paul.1982. “Durkheim and Mauss Revisited: Classification and the Sociology of

Knowledge”. Journal of the History and Philosophy of Science. 13(4):267–97.1983. Wittgenstein: A Social Theory of Knowledge. New York: Columbia Univer-

sity Press.Boelen, W. A. Marianne. 1992. “StreetCorner Society: Cornerville Revisited.”

in The Journal of Contemporary Ethnography. April.Bourdieu, Pierre. [1972]1977. Outline of a Theory of Practice. Cambridge Univer-

sity Press.Brown, John Seeley and Paul Duguid. 2002. The Social Life of Information.

Cambridge: Harvard Business School Press.Cladis, Mark S. 1992. A Communitarian Defense of Liberalism: Emile Durkheim

and Contemporary Social Theory. Stanford University Press.Collins, Randall. 1988. “The Durkheimian Tradition in Conflict Sociology”,

in Alexander, Jeffrey C. (ed.), Durkheimian Sociology: Cultural Studies.Cambridge University Press.

1990. “Ritual Interaction Chains”. Symbolic Interaction. 12(1).Coser, Lewis A. 1971. Masters of Sociological Thought: Ideas in Historical and Social

Context. New York: Harcourt.Cuvillier, Armand. 1955. Editor’s introduction to Durkheim’s Pragmatism and

Sociology. Cambridge University Press.Dennes, W. R. 1924. “The Methods and Presuppositions of Group Psychology”,

University of California Publications in Philosophy 6(1):1–182.

339


Denzin, Norman K. 1992. Symbolic Interactionism and Cultural Studies.Cambridge MA: Blackwell.

Doroszewski, W. 1933. “Quelques Remarques sur les Rapports de la Sociologieet de la linguistique: Durkheim et F. de Saussure”. Journal de Psychologie.Vol. XXX.

Douglas, Mary. 1966. Purity and Danger. An Analysis of Concepts of Pollution andTaboo. New York: Praeger.

Durkheim, Emile. [1893]1933. The Division of Labour in Society. Chicago: FreePress. (Trans. 1933).

[1895]1982. The Rules of the Sociological Method. Steven Lukes (ed.). New York:The Free Press. (Trans. 1938).

[1895–6]1958. Socialism and Saint-Simon. Edited by Alvin Gouldner. TheAntioch Press.

([1897]1951). Suicide: a Study of Society. New York: Free Press. (Trans. 1951).[1898]1953. “Individual and Collective Representations”, in Sociology and Phi-

losophy. Glencoe Illinois: Free Press.1907. Letter to the Revue Neo-Scholastique November 8, 1907. In

Durkheim, [1895]1982. The Rules of the Sociological Method. New York: FreePress.

1912. Les Formes elementaires de la Vie Religieuse: Le Systeme Totemique en Aus-tralie. Paris: Librairie Felix Alcan. (Trans. 1915).

[1912]1915. The Elementary Forms of The Religious Life. Chicago: FreePress.

[1912]1995. The Elementary Forms of Religious Life. Karen Fields (trans.)New York: The Free Press.

[1913–14]1955. Pragmatism and Sociology. Cambridge University Press.[1913]1960. “Dualism and Human Nature”, in Emile Durkheim (ed.) Kurt H.

Wolff. Ohio State University Press: Columbus.1960. Montesquieu and Rousseau. Ann Arbor: University of Michigan Press.

Durkheim, Emile and Marcel Mauss. [1901]1963. Primitive Classification.University of Chicago Press.

Emerson, Robert. 1996. Keynote Address Stone Symposium.Fenton, Steve, with Robert Reiner and Ian Hamnett. 1984. Durkheim and Modern

Sociology. Cambridge University Press.Feyerabend, Paul. 1975. Against Method. London: NLB.Fine, Gary Alan. 1993. “Ten Lies of Ethnography: Moral Dilemmas of Field

Research.” In The Journal of Contemporary Ethnography. October.Fuller, Steve. 1989. Philosophy of Science and Its Discontents. San Francisco:

Westview Press.Garfinkel, Harold. 1967. Studies in Ethnomethodology. New Jersey: Englewood-

Cliffs.1988. “Evidence for locally produced, naturally accountable phenomena of

order*, logic, reason, meaning, method, etc. in and as of the essential haec-ceity of immortal, ordinary society,” Sociological Theory ’88, vol. 6, no. 1,pp. 103–09.

2002. Ethnomethodology’s Program: Working Out Durkheim’s Apharism.Boulder, Colorado: Rowman and Littlefield.

Bibliography 341

Gehlke, Charles Elmer. 1915. “Emile Durkheim’s Contribution to Social The-ory”, Columbia University Studies in Economics, History, and Public Law.7–187. 1(VCIII).

Giddens, Anthony. 1971. Capitalism and Modern Social Theory. Cambridge Uni-versity Press.

1978. Emile Durkheim. New York: Penguin.Godel, Robert. 1957. Les Sources Manuscrites du Cours de Linguistique

Generales de F. de Saussure. Addendum. Geneva: Droz.Godlove, Terry. 1989. “Epistemology in Durkheim’s Elementary Forms of Religious

Life,” Journal of the History of Philosophy. 24(3):385–401.Goldenweiser, A. A. 1915. “Review of The Elementary Forms” American Anthro-

pologist. (17):719–35.Heath, Christion. 2000. Technology in Action. Cambridge University Press.Hilbert, Richard A. 1992. The Classical Roots of Ethnomethodology: Durkheim,

Weber, and Garfinkel. Chapel Hill: University of North Carolina Press.Hinkle, R. C., and Hinkle, G. J. 1954. The Development of Modern Sociology.

New York: Random House.Hirst, Paul Q. 1975. Durkheim, Bernard and Epistemology. London: Routledge.Hughes, H. Stuart. [1958]1977. Consciousness and Society. New York: Random

House.Hume, David. [1739]1978. A Treatise of Human Nature. Oxford University Press.

[1777]1975. Enquiries Concerning Human Understanding and Concerning thePrinciples of Morals. Oxford: Clarendon Press.

James, William. [1907]1955. Pragmatism. New York: Meridian Books.Jameson, Frederic. 1972. The Prison-House of Language. New Jersey: Princeton

University Press.Joas, Hans. 1993. Pragmatism and Social Theory. University of Chicago Press.Jones, Robert Alun. 1986. Emile Durkheim: an Introduction to Four Major Works.

Beverly Hills: Sage.Kant, Immanuel. [1781]1965. The Critique of Pure Reason. New York: St. Martin’s

Press.[1754]1998. Ground work of the Metaphysic of Morals. Cambridge University

Press.Kohnke, Klaus Christian. 1991. The Rise of Neo-Kantianism: German Academic

Philosophy Between Idealism and Positivism. Cambridge University Press.Kuhn, Thomas. 1962. The Structure of Scientific Revolutions. University of Chicago

Press.LaCapra, Dominick. 1972. Emile Durkheim: Sociologist and Philosopher. Ithaca

New York: Cornell University Press.Latour, Bruno and Steve Woolgar. 1979. Laboratory Life: The Social Construction

of Scientific Facts. London: Sage.Leaf, Murray J. 1979. Man Mind and Science. New York: Columbia University

Press.Lehmann, Jennifer. 1990. Deconstructing Durkheim. London: Routledge and

Kegan Paul.Lemert, Charles. 1997. Sociology After the Crisis. Boulder Colorado: Rowman and

Littlefield.


Levi-Strauss, Claude. [1958]1963. Structural Anthropology. New York: BasicBooks.

Levy-Bruhl, Lucien. [1910]1966. How Natives Think. New York: WashingtonSquare Press.

[1922]1966. The “Soul” of the Primitive. Chicago: Henry Regnery.Lukes, Steven. 1973. Emile Durkheim: His Life and Work: a Historical and Critical

Study. Stanford University Press.Lynch, Mike. 1997. Scientific Practice and Ordinary Action: Ethnomethodology and

Social Studies of Science. Cambridge Universtiy Press.1982. Editor’s introduction to Durkheim’s The Rules of the Sociological Method.

Marx, Karl. Grundrisse. In Tucker (ed.) The Marx Engles Reader. New York:Norton Press.

Mead, G. H. 1934. Mind, Self, and Society. Chicago: Free Press.Mestrovic, Stjepan. 1988. Emile Durkheim and the Reformation of Sociology. Mary-

land: Rowman and Littlefield.1993. Durkheim and Poststructuralism. New York: Aldine DeGruyter.

Mills, C. Wright. “Situated Action and the Vocabulary of Motives.” AmericanJournal of Sociology. 5:904–13.

Nisbet, Robert. 1974. The Sociology of Emile Durkheim. New York: Oxford Uni-versity Press.

1993. The Sociological Tradition. New Brunswick: Transaction Publishers.Parkin, Frank. 1992. Durkheim. Oxford University Press.Parsons, Talcott. ([1937]1968). The Structure of Social Action. New York: Free

Press.1973. “Durkheim on Religion Revisited: Another Look at The Elementary Forms

of The Religious Life” in (ed.) Beyond The Classics.Poggi, Gianfranco. 1972. Images of Society: Essays on the Sociological Theories of

Toqueville, Marx, and Durkheim. Stanford University Press.Quine, Willard Van Orman. [1966]1977. The Ways of Paradox and Other Essays.

Cambridge: Harvard University Press.Rawls, Anne Warfield. 1985. “Reply to Gallant and Kleinman on Symbolic Inter-

action versus Ethnomethodology.”Symbolic Interaction 8(1):121-40.1987. “The Interaction Order Sui Generis: Goffman’s Contribution to Social

Theory,” Sociological Theory, 5(2):136–49.1988. Reply to Levine and Fuchs. Sociological Theory.1989. “Language, Self, and Social Order: a Re-evaluation of Goffman and

Sacks,” Human Studies. 12(1):¨147–72.1990. “Emergent Sociality: A Dialectic of Commitment and Order.” Symbolic

Interaction. Volume 13(1): 63–82. French translation, 2002, in the MaussReview.

1995. “Narratives as Evidence of Trouble in Interracial Communication.”Midwest Sociological Meetings. Chicago.

1996 “Durkheim’s Epistemology: The Neglected Argument,” American Jour-nal of Sociology. 102(2):430–82. (French trans Enquete 2005)

1997a. “Durkheim’s Epistemology: The Initial Critique, 1915–1924”. TheSociological Quarterly. 37(4).

1997b. “Durkheim and Pragmatism.” Sociological Theory. Volume 15(1):5–29.

Bibliography 343

1998. “Durkheim’s Challenge to Philosophy: Human Reason as a Product ofEnacted Social Practice.” American Journal of Sociology. 104.

2000. “Race as an Interaction Order Phenomenon: W. E. B. Du Bois’s ‘DoubleConsciousness’ Thesis Revisited.” Sociological Theory. 18(2):239–72.

2000. “Durkheim’s Treatment of Practice,” The Classical Journal of Sociology.1(1):33–68.

2002. Editor’s Introduction to Ethnomethodology’s Program: Working OutDurkheim’s Aphorism, by Harold Garfinkel, Rowman and Littlefield: Boulder,Colorado.

2003. Conflict as a Foundation for Consensus: Contradictions of Capitalismin Book III of Durkheim’s Division of Labor. Critical Sociology.

Forthcoming, with Bonnie Wright, “The Dialectics of Belief and Practice,”Critical Sociology.

Rawls, John. [1953]2001. Collected Papers. Cambridge: Harvard University Press.Reisman, David. 1950. The Lonely Crowd. New Haven: Yale Univ. Press.Rorty, Richard. 1979. Philosophy and the Mirror of Nature. Princeton University

Press.Rousseau, Jean Jacques. ([1757]1999) Discourse on the Origins of Inequality.

Oxford University Press.Sacks, Harvey. 1993. Lectures in Conversation 2 vols. Cambridge University Press.Sacks, Harvey, Emmanuel Schesloff and Gail Jefferson. 1974. “A Simplest Sys-

tematics for the Organization of Turn taking in Conversation.” Language50:696–735.

Schaub, Charles. 1920. “A Sociological Theory of Knowledge” The PhilosophicalReview. 29(4):319–39.

Schmaus, Warren. 1994. Durkheim’s Philosophy of Science and the Sociology ofKnowledge: Creating an Intellectual Niche. University of Chicago Press.

Shadish, William R. and Steve Fuller. 1994. The Social Psychology of Science.Guilford University Press.

Sirianni, Carmen. 1981. “Justice and the Division of Labor: a Reconsiderationof Durkheim’s Division of Labor in Society”. Theory and Society. 449–70.

Simmel, Georg. [1918]1978. The Philosophy of Money. Boston: Routledge andKegan.

Simpson, George. 1963. Emile Durkheim: Selections from His Work. New York:Thomas Y. Crowell Company.

Stone, Gregory P. and Harvey Farberman. 1967. “On the Edge of Rapproche-ment: Was Durkheim Moving Toward the Perspective of Symbolic Interac-tion?” The Sociological Quarterly.

Suchman, Lucy. 1999. Plans and Situated Actions: The Problem of Human MachineCommunication. New York: Cambridge University Press.

Tannen, Deborah. 1990. You Just Don’t Understand. Harper Collins: New York.Turner, Stephen. The Social Theory of Practices: Tradition, Tacit Knowledge, and

Presuppositions. University of Chicago Press.Vico, Giambattista. [1744]1948. The New Science of Giambattista Vico. Ithaca

New York: Cornell University Press.Weber, Max. [1921]1968. Economy and Society. Berkeley: University of California

Press.


Whitehead, Alfred North. [1929]1978. Process and Reality. New York:MacMillan.

Whyte, William Foote. ([1937]1993). Street Corner Society: The Social Structureof an Italian Slum. University of Chicago Press.

1992. “In Defense of Street Corner Society.” in The Journal of ContemporaryEthnography. April.

Winch, Peter. 1956. The Idea of a Social Science. London: Routledge & KeganPaul.

Wittgenstein, Ludwig. 1945. Philosophical Investigations. Oxford: Blackwell.Woolgar, Steve and Bruno Latour. 1979. Laboratory Life: The Social Construction

of Scientific Facts. London: Sage.

Index

Abyss, conceptual 335Accounts 141, 327, 331, 333

Retrospective justifications 4, 147–49,188, 226, 229, 327

Accountable to rules 4–5Contexts of accountability 327, 336, 337Religious beliefs as 42, 191

Advanced division of labor 3, 5Aggregates of individual happenings 326,

333Alexander, Jeff 4, 24, 324Ancestral heroes (individual) 290Ancestors (legendary) 191–92Animal abilities 308Animism 125–29, 183–84, 235, 246

(versus Naturism) 23, 109, 110, 124–25Anthropomorphism 183–84Apriorism 9, 15–17, 19, 57, 58, 60–62,

152, 230, 235Causality and 251, 256Dualism of 81–85

Arbitrary(beliefs and representations) 174, 188ideas as 238

Archaic Religion 34–41Defense of 34, 41–42

Aristotle’s categories 48Asceticism 198Association of ideas 200, 218Audio and video recording 332Authority 249, 254–55

Becker, Howard 327Beliefs 14, 20, 24, 34, 44–46, 116–17,

266–71, 283As a specialized form of social practice

327As commitment to practice 288Future of different from future of

practices 286As accounts 42, 141As justifications for practices 320

Distorting 37, 94, 101, 249, 275Represent underlying reality 90Versus practices 3, 41, 76, 113, 115–17,

141, 162, 167, 186, 194–211, 269–71Secondary and retrospective 5, 34–35,

40, 191, 219, 226, 327Arbitrary 174The primacy of rites over 206, 219–20Origin in rites 194, 219Racist 229Scientific 229Totem exists only if believed in

224–25Individual pre-rational 233Must be enforced 254

Berkeley 96Bewitchment 217Biological individual 278, 308Bloor, David 24, 329Bourdieu, Pierre 325Buddhism 115–16

Cartesian Rationalist 7, 49, 73Categories of the understanding 1–2, 3, 8,

9, 12, 16, 17–20, 22–23, 24, 26, 35,39, 48, 53–55, 56, 58, 60–62, 67, 73,83, 84, 93, 98–99, 102, 103–04, 108,110–11, 116, 178, 181, 188, 194–211,216, 230–61, 264, 289, 301, 337

Why they need to be social 304–05Six categories 301–02As a final question of logic 301–15Personality as 188Causality 212–29, 230–57, 261Role of emotion 231, 232Innate 239Necessary for reason and

communication 254Classification as first category

337Categorical imperative 176Categorization 269

345


Causality 15, 54, 212–29, 230–61As creation 194, 233As moral 228As an aspect of force 237Like produces like as concrete statement

of 232Idea of causality 232As instinct 232–33, 243As a habit of thought 244, 250–52As category 232, 233, 244, 255–57Enactment of 233Principle of 236–41Efficacy implied 237, 254And external experience 241–45As internal experience 247–49Durkheim’s argument for 258–61

Cause and effect,connected by force 237in empiricism 251–52

Christian and Hebrew calendars284

Christianity allowed science to gainstrength 287

Church 35–36, 121, 216, 270Circle, individual society deity 210Circular argument 139, 229, 321, 335Circular thinking 270Citation patterns 262Clan totemism 141, 156, 157, 184Classification 83–84, 131, 150, 184

Category of 151As first category 108, 337Moral 73Sacred and profane as first 108As a logical system 149–52, 183

Coat of arms 143Coercion 63–65Coherence 220

Of sounds and movements 324Collective effervescence (emotion) 170,

171, 178, 181, 277And human reason

170Collective experience 247Collective feeling 241, 283Collective forces 200–01Collective life 222

International 282Collective reflection 149Collective representations 19, 37, 50,

95–101, 103–04, 119, 177–82Purpose as foundation of logic 303Emotion embodied in 178–79Collective symbol, totem as 141, 156,

178

Versus individual representations 92, 93,97–98

Performance of ritual 122Collective states 172Collective totems 155–56, 158Collective utility 227Collins, Randall 337Communal

Mentality 238–39Communication 144, 178–79, 180, 247,

254, 283, 292–93, 295, 304, 335Communitarian moral philosophy 21Comte 5, 169Concepts 87, 327–28

as collective representations 294truth of 297–99

Concepts and general ideas 289–90,292–93

Versus sense impressions 290–92Conceptual abyss 335Conceptual reality 335

versus empirical reality 85, 86–90, 294Conceptual systems, sociology of 317, 321,

327–28Conceptual thought contemporaneous

with humanity 299–300Conceptual typifications 331Concrete practices 210, 278

Function of 317Concrete ways of knowing 294Consensus theory of truth 230Constitutive practice, moral relevance of

319Constraint 123, 326Constructed (socially) 180, 238Constructivists, social 318Contagion 147, 199–202, 217

Is real 200Is emotion 201–02Not irrational 202

contexts of accountability 327, 336, 337Garfinkel 4, 337Mills 4–5, 327, 336

Contiguity and resemblance 214, 215,218, 223, 242

Contingencies 333, 334Contradiction (non-contradiction) 183,

205, 320Conventional beliefs and ideas 174–75,

180Conventional divisions 151Conventional interpretation 317

The categories are not conventions 304Conventional meanings 265

Conventions of use 265

Index 347

Conversation Analysis 333Conversational structures 331Conversion, cult groups 226Cosmology, religious 149–50, 286–87Creation, the principle behind imitative

rituals 213–15Rites as 210Causality as 212, 233

Crime 333Critics of Durkheim 105, 265, 266, 293,

317, 321, 326, 335Inherent individualism of 321, 323Initial critics in English 321

Culture 328Custom and habit 244

Death 128–29Deconstruction 329, 336Dennes, William 82, 262, 321, 335Deity 264Deities 115, 160, 163, 167, 168, 188,

191–92, 199, 205–06Only exist if sacred is created through

rites 207Created by sacrifices 206–08

Denzin, Norman 335Details 23, 136, 219, 223, 269, 317–18,

320, 331Relevance of 332

Deviance 333, 337Dietary restrictions 146Discourse on the Origins of Inequality,

Rousseau 6–7Divinity 115, 160, 167Division of Labor, Durkheim 3, 5, 21–22,

177, 263, 280, 281, 284, 322, 323,334

Book III 4, 5, 284, 323Division of labor 319Dreams 127–28Doctor Jekyll and Mister Hyde 263Douglas 329Dualism 6, 14, 24–51, 58, 71, 72–107,

109, 111, 118, 134, 174, 182, 188,309

And Animism and Naturism 125As a social fact 75Of body and soul 85–86, 188, 191Of personal and impersonal 87Kantian 88–89, 93, 98–99, 106–07Cartesian 106–07Double man or homo duplex 77, 78–80,

218Of sacred and profane as first

classification 108, 337

Result of moral force 171And hierarchy 250

“Dualism of Human Nature and its SocialConditions” 6, 14, 24–72, 111

Durkheimargument for causality 258–61argument, organization of 2, 24–26,

28–29, 32–33, 46, 140, 162–63,194–95, 231–32, 307

as cultural sociologist 324as philosopher 16, 17–22, 32, 70–71,

321Philosophy as folk belief 77, 85, 86, 90,

102defense of aboriginal intelligence

158–60, 182–83, 186–87, 199–200,202, 215, 218–19, 225, 267, 300, 303

dualism 72–107, 233, 309Critics of 82–83Critique of Kant 93, 94, 99–100, 101interpreted as Kantian and rationalist

335interpreted as “Radical” 323interpreted as a pragmatist 324

Dynamic relationship, emotions 220,237

Economics, classical 11, 68–69Efficacy 231, 233, 237

Rites not beliefs 234Implied by causality 237, 254Of moral forces only when felt 248

Elias, Norbert 3Emblem 142–44, 145, 146, 154, 163Emerson, Robert 327Emotion 2–3, 10, 13, 15–17, 20, 50,

65–66, 109, 165, 169, 207, 231, 235,256, 278

Versus sensation 85, 86, 109Collective 143, 144, 168–69, 170, 172,

180, 181, 241Individual state 173, 178, 179Real,but only when created collectively

278And totems 144–45, 147And reason 163Mana as 163, 169–71As sentiment 170Respect as 170–71, 175, 199Passion 190Created by rites 212Feelings of Moral force 213, 233Feelings of cause and effect 237And contagion 201–02Corresponding to social facts 223


Mimetic rites and 218Causality and 260And the categories 304

empirical contexts of use 317Empirical is contingent 303empirical validity 167–68, 175, 219–20,

243, 246, 247, 254, 306, 312–14,326, 337

empiricism 9, 111, 125, 130, 182, 184,200, 218, 223, 225, 230, 235, 241,247, 250, 252–53

dualism of 81–85causality 214–15, 238, 244, 251–53,

256Durkheim’s criticism of 217–19Hume 230, 267Locke 294

enacted practice 3, 4, 5, 20, 23English Anthropology 321–22Enlightenment individualism 7, 11–12, 14,

312Enlightenment philosophy 283Enquiries Concerning Human Understanding

244epistemology 5, 46–68, 230–31, 258

crisis 8–11, 230–31empiricists 7, 55, 230, 241classical empiricism 15–17, 235apriorists 55, 235Pragmatist 6, 55categories of the understanding 1–2Durkheim’s 7–8, 10–11, 16–17, 232Locke 9, 57Hume 15–17, 57, 227, 230, 334Kant 15–17, 230James 9, 56Justified belief 230Theory of knowledge 234–35

Ethics 4, 13, 21, 68Ethnomethodology 4–5, 27, 330–34Evans-Pritchard 228Evolution of ideals 278External constraint 123, 174–77, 317External experience 241–47External forces 63–65, 248

Faith must rush before the truth 288Fallacy of misplaced abstraction 219,

324–26Fallacy of misplaced concreteness

(Whitehead) 325Feelings

of cause and effect 237collective 241individual 245

Field Work 327–28, 332, 333Flag 143–44Flesh is profane 198Force 67, 235

Moral 15, 20, 39, 58, 63–65, 83,235

Classification as first moral force 108External 63–65, 248Internal 247–49Idea originates in Mana 164–68, 171Cause as an aspect of 237Impersonal forces 247

Framework of mental life 304Frazer 23, 135–36, 217, 218, 228French revolution 285Function 231, 243Functionalism 34, 35, 37, 38

As a test of truth 38, 42Use of cause 34Of totemism 164, 190, 221Of the categories 304Justice as functional prerequisite for

division of labor 334

Garfinkel 4, 6, 22, 42, 219, 269, 282, 299,318, 319, 320, 327, 331, 333, 336,337

Gehlke, Charles Elmer 262, 321Gemeinschaft versus Gesellschaft 12, 35,

70Gender 328General ideas 37, 56, 57, 58–59, 60, 83,

131, 185, 264, 289–90Generalization 257Generic representations 83Giddens, Anthony 269, 323Globalization 70, 319Goffman 4, 6, 269, 285, 319, 333, 337

Involvement obligations 320Group mind 170, 264, 293, 311, 321Group unity 143–44, 205, 208, 213Grundrisse 281, 319Gurwitsch 265

Habit 58, 243, 293–94Cause as 244, 250–52

Habit of belief 202Habitus, Pierre Bourdieu 294Hallinan 327Heidegger 149Heterogeneity 119–20Hierarchies 119, 249–50Historical materialism 276, 280Hobbes 255humanities 326, 329

Index 349

human nature, general properties of 218human reason 170, 226–27Hume 8, 9–10, 15–17, 20, 57, 95–105,

179, 184, 185, 202, 242–45, 308,309, 315

Problem with concepts 131Durkheim’s critique of 58–60Causality 230, 242–44, 245emotions 13empiricism 267Epistemology and 9–10Skepticism 227Enquiries Concerning Human

Understanding 244Treatise on Human Understanding 244

Hyperexcitation 187–88

Idealism 17–18, 95–105, 149, 178,181–82, 220, 266–83, 306, 325

Durkheim’s critique of 92, 96–97, 275Durkheim interpreted as 105, 265, 266,

293, 317, 321, 326Ideals

create distortion 277explain emotions 278

Ideal types 219Ideas 177, 207, 219, 304Imitative rites and causality 212–29, 233,

251, 253Purpose of 220Mimetic rites 212, 213, 233Causal rite 233

Impressed (ideas on the mind) 187Individual

as a general category 144as a natural end 303, 314

“Individual and CollectiveRepresentations” Durkheim 6, 24–26

Individual (biological) 141, 156, 171, 177,180, 187, 218, 280

Individual purposes 309individual belief 233, 319individual cults 122–23individual experience 125–26, 165, 184,

188individual perception 85, 86, 185, 245,

305individual sensations 256individual representations 178, 179individual totemism 152–61individual will 190individual utility 227individual feelings 245individual forces 250individual generalizations 290

Individualism 3, 11, 16, 19, 42, 62–63,153–54, 156–57, 167, 218, 272,305–06, 321

Durkheim’s argument against 3, 11, 37,41, 56, 58–60, 167–68, 231

Empiricist individualism 23, 125, 182Hume’s 57Kant’s 23, 57, 58Enlightenment 7And classical economics 11Soul and 188, 189–90

Individualist 216, 246, 250Industrial revolution 284Initiations 144, 191Innate 239Instinct 232–33

causality as 232–33, 244institutional accountability 330, 336institutionally ordered 336Instrumental 209, 211, 216, 220, 226Intelligibility 3–5, 10, 16–17, 20, 21–22,

39–40, 144, 178–79, 180, 247, 254,255, 283, 317, 319, 320, 324, 327,331, 334, 337

As a constraint on practice 317Interactional involvement 332Interaction obligations, moral 331, 337Interaction orders 14, 38, 100, 327,

338Of gender and race 14

Interior life 248Internal constraint 174–77Internal knowledge 202Internal experience 247–49

Versus external experience 245–47International 303, 319

collective life 282, 299–300Interpretation 37–38, 249, 269, 317, 318,

326, 332, 333–34Interpretive sociology 333Intersubjective

meaning 257, 292–93communication 295

Intichiuma 203–04Involvement obligations 320Irrational 210Is versus ought 19, 275–76, 297, 303

James, William 9, 16, 17, 56, 272, 305,315

Pragmatism 9Radical empiricism 6, 16

Jefferson, GailJoas, Hans 324Justified belief 11, 16, 17, 230


Justifications, accounts as 229, 327Justice 21–22, 264

Functional prerequisite for division oflabor 334

Kant, Immanuel 7, 8, 15–17, 58, 230, 250,252, 319

Religion and 279Rationalism 73Apriorism 23, 58, 280, 305Categorical Imperative 176, 320Dualism 58, 72–89, 90–96, 98–99,

106–07Epistemology and 9–10Kingdom of ends 70Time and space as intuitions 50Duty 176Contradiction (non-contradiction) 176,

183, 184, 320Will 164, 190And causality 244, 252

Knowledge, social versus individual forms90–96

Kuhn, Thomas, normal versusrevolutionary science 106

Labels 180–81Laboratory science 329, 332LaCapra 73Lang, Andrew 159–60Language 131–32, 262–94, 300, 328

Confusions introduced by 159–60, 183limitations of 130–31social origin of 288–300function of 295

Levi-Strauss, Claude 20, 321, 323Levy-Bruhl, Lucien 321Liberty 190Like produces like 214, 217, 221, 233, 247

Involves belief in causality 253Causal rite 233And theory of knowledge 234

Linguistic turn 325Locke, John 9, 57, 96, 294Logic 30–31, 50, 52–54, 181, 182–85,

188, 200, 201–02, 218, 242, 262–96,300, 301, 334

Logical opposition 120–21Logical system, classification as 149–50,

152Social origin of 288–300, 303Categories of the understanding 301–15Modern 312–14

Logically necessary 302–03Lukes, Steven 23, 323

Macro sociology 333Magic versus religion 121–23, 196, 212,

226–29, 270And mimetic rites 215–17Bewitchment 217Origins in causality 215Comes from religion 216

Magician 122Malinowski 321Mana 129, 163, 240

As a universal 166As emotion 163, 169–71As origin of force 164–71, 240Collective force 240

Marx 7, 11, 12, 69, 88, 157, 280–81, 312Material end 209Material representations 273, 274Mathematical knowledge 242Mauss, Marcel and Hubert 216–17Mead, George Herbert 12, 39–40, 146,

168Mestrovic 11–12Methodological critique 136Methods 269, 318Micro versus Macro 4, 27, 136, 242,

317–18Mill, John Stuart 13, 87Mills, C. Wright 4, 42, 229, 327Mimetic rites 212, 223, 233

Religion not magic 215–17Most basic religious rite 216Image of totem in 217–18Pure creation 218Cause the idea of sacred 233Causal rite 233And causality 251, 253

Modern industrial society 336Modern versus traditional life (organic

versus mechanical) 1–3, 5, 12, 35, 70,171, 186, 198, 285, 312–14, 319,337, 338

Montesquieu and Rousseau 6–7, 13, 17,308

Moral/empirical knowledgeMorality 16–17, 21

Social solidarity 334–38social thought 4, 21Moral imperative 4Moral relations 1–9Justice 10, 21–22Moral classification 73, 150Moral reasoning 138Moral unity 150Moral issues in 134

Moral obligations, interactional 337

Index 351

Moral issues, science of 263Moral force 15, 20, 29–30, 39, 60, 63–65,

83, 118, 121, 154, 235, 314, 337Origin of 162–92Feelings of 213, 233, 277Energy 184And symbolic meaning 103–04Classification as first type 108Totemism and 171Sacred and profane phases 172–74Moral union 233Not perceived through senses 242Can be touched directly 248Separates sacred and profane 249Created by practice 274

Moral Philosophy 10, 21–22, 69–70, 314,319

Enlightenment 283communitarian 21practice conception 21utilitarian 68, 69–70, 87, 166, 196Durkheim and 69–70

Moral unity, outward expression of 221Movements and cries 213, 221–22, 227,

318, 324, 326, 330, 336Movements in unison 222Reduced to concepts 327

Muller, Max 130–33Multiple possibilities, realities 329Mutual intelligibility 1–22Myths 130, 132–33, 147–49, 165, 168

As misleading and retrospective 147–48,168, 188, 191, 275

Function of 148Secondary level 148Mythic persons and legendary ancestors

191–92

Naıve statistical conceptualism 335, 336Narrative and myth 14, 35, 37, 40–41,

44–46, 130, 276–78Represents underlying reality 90, 94, 101Sociology 317Truth of 38, 130, 132And texts 326As a specialized form of social practice

327Outgroup narratives 331

Natural causes 227Natural objects 325Natural versus social forces 14–15, 93,

109, 164, 165–67, 175Naturism 129–33, 235

Versus Animism 109, 110, 124–25Necessity, as cause 250–52

Negative cult 196–202Neo-Kantian 265Nominalism 131, 304Norms and rules 317

Ockham, William of 277Ockham’s razor 277Opinion as collective knowledge 170–71Orders of practice versus institutional

orders 4, 27Original division in two 150, 155, 157Origins of Inequality, Rousseau 272, 308

Pain as a symbol 198Paradigms and labels 17–18, 319, 329Parsonian structuralism 333, 336Parsons’s Plenum 269, 318Parsons, Talcott 20, 269, 281, 321–22Perceived coherence 220Perception, individual 178, 179, 182, 183,

184–85, 241–45, 305Perception versus emotion 15–17, 118,

150, 164, 170Personal forces 250Personality 29, 163, 190

As a category 188As will 190Origin in soul 255

Phenomenologists 265Philosophy,

implications for 316–18, 338individual bias of 319

Philosophical Investigations,Wittgenstein263, 297

Physical forces 114, 175Physical causes 227Picture theory of meaning 270Plato, Durkheim’s critique of 92, 97–98Police 333–34

studies of 333ideas generalized to social theory peoper

334Positive cult 202–11Positivism 4, 8, 17–18, 220, 259, 268, 317,

325Postmodernism 4, 10–11, 327, 334Poststructuralism 12, 323, 327, 333, 336Practice 266–83, 318–19, 325, 335

Theory of 1–3, 14, 21, 336Versus beliefs 3, 41, 112, 115–17, 167,

265, 266–71, 283Orders of 4Study of 27Collective performance of 122, 255Detail 223


Relevance of detail 332Constitutive, moral relevance of 319As moral and epistemological

foundation for public life 320Practices 95–105, 116–17, 269–71

Usually transposed into concepts bytheorists 324, 327

Reduction to belief a fallacy 327Beliefs about 330Details 320Turned toward action 286Future of different from future of belief

286Function of 40–41, 125, 185Concrete 210, 318, 325Laboratory 299As the essence of religion 112–13And social solidarity 2–3Civil 3Enacted shared 4, 5, 20, 23, 75, 185,

264, 316, 320Situated 22Ritual Religious rites 21, 39, 168, 204Prospective 4Routinized rites 173Mimetic rites 215–17Create moral force 274Generate universal ideas 304–05

Pragmatism 6, 7, 9, 10–11, 12, 16, 19, 68,182, 184, 230, 271, 282–83, 309, 318

Consensus theory of truth 11, 16, 17,326

William James 6, 9, 16, 272, 305Religious individualism 282–83

Pragmatism and Sociology 6, 24Pre-invitation (Alene Terasaki) 332Pre-rational 233Presentation of self 333Primitive Classification 6, 50, 52–54, 108,

150–51Primitive religion, defense of 34, 41–42Primitive thinking 200Process 194, 232Profane (and sacred) 144Professional associations 338Prohibitions express the sacred 197, 199Protestant asceticism 198Protestant Ethic 198, 234, 266Protestant reformation 284Pure reason 242

Quakers 270Qualitative (versus quantitative) 27, 136,

317, 318Quantitative methods 268, 317, 326

Radcliff-Brown 321Radical Durkheim 323Rational categories of thought 181Rational individualism 11–12, 19Rationalism

Cartesian 73Kantian 60–62, 73

Rationalist, Durkheim interpreted as 239,245

Real,Forces 315practices as 266–83

Realism versus idealism, false distinctionbetween 275, 322, 325

Real creates ideal 275–76Reason 19–20, 170, 278, 283, 304, 326,

334Result of social practices 6–7, 28, 68–69,

75, 118, 132–33, 164, 190, 249Categories necessary for 254Religion and 11–15, 33–57

Recognizability 22, 282, 283, 317, 324Recognizable by design 330–34Recognizable turn structure 331

ReferentialContexts 328Meaning 304relationship 145, 218, 222, 227, 265

Regularity of expectation 253Religious narratives 276–78

Conversion narratives 329Religion 33–36, 46, 113, 262–300

Cooperation with science 288Is action science is conceptual 287–88Definition of 112, 113–16, 121, 123Broad definition of 336Belief versus practice 112, 115–17, 186,

194–211, 283–86Origin of 44–46As origin of causality 258and Reason 11–15, 21, 187cultural pluralism 12–13, 21as individual belief 41, 49–50and belief 44–46, 319as beliefs about dualism 85, 86–90as distinction between sacred and

profane 95–101versus magic 121–23truth of 38and logic 289real function of 319as practice 319

Religion and logic 202Religion and science the same 270, 283–88Religious cosmology 286–87

Index 353

Religious faith, belief 255Religious forces 200–01Religious individualism, pragmatism and

282–83Religious practice or rites 21

All have same purpose 258Are efficacious 199, 212, 221–22,

223–29, 234Mimetic rites the most basic 215–16,

217Imitative, purpose of 220Primacy of rites over beliefs 206, 219–20Rites, routinized 168, 173Create emotion 212Cause causality 194–211Cause ideas 194Reproduce totemic species 221Details 136, 223Function of 3, 133Ritual 115–16, 184, 204Collective performance of 122Versus belief 167Ritual phases of 203–04

Religious prohibitions as CategoricalImperatives 196–97

Representation 100, 102–03, 139–40, 142,143, 166, 167, 170, 175, 178, 199,206, 207, 208, 241, 254, 273, 274,295

Resemblance and contiguity 214, 215,223, 242

Respect 170–71, 175, 255Respect precedes belief 194

Retrospective accounts 141, 147–49, 188,191

Rites,give signs their meaning 266are primary 273

Ritual Interaction Chains (Collins) 337Rituals,

continued performance of 254interactions, modern and secular 336

Rousseau 6–7, 13, 51, 74, 79, 239,249–50, 258, 272, 281, 308–10

Routinization 173Rules, following versus accountable to 4–5Rules of the Sociological Method, Durkheim

5–6, 8, 10, 219, 258, 338

Sacks, Harvey 320Sacred (and profane) 10, 25, 29–31,

72–101, 107, 117–18, 144, 145, 176,277

Created by Positive cult 202Totem as 142, 160

Totemic emblem as 163, 166, 174Totem as origin of 141–49, 184Cause and purpose 134–35As dualism 75, 77, 95–101, 118As first classification 108, 110–11,

277As social versus individual 93Origin of 118–21, 232Outline of Durkheim’s argument for

110–11Time 163, 172–74, 178, 179Origin in emotion not sensation 170Sacred beings only exist through

representation 206Negative rites protect the sacred 196Separated by negative rites 197Approaching the sacred 197–98Contagion 199–201Separated by moral force 249

Sacrificial communion 204, 205–06Schmauss, Warren 230Schopenhauer 11–12Schuab, Charles 321Science

Of man 303cooperation with religion 288and religion the same 270as faith 299is conceptual religion is action 287–88laboratory 329, 332

Scientificterminology 292objectivity 298symbols 299argument 262–300versus social things 328–30Theories 270

Secondary source tradition 7, 17–18, 26,29, 262, 321

Secular orders 337Secular faith 285Self, social construction of 319Sensation, sense impressions 85, 86, 93,

118, 167, 169, 170, 175, 180, 183,184–85, 187, 190, 200, 201–02, 207,241–45, 247, 303

Versus concepts 290–92Sensible idea 207Sentiments 14, 65–66, 170Sequential form of practice 332Sequence of talk 332Sexual Totemism 152–61Significant symbol 39–40, 146, 168,

188Single case method 136, 267–69


Single intellect 293Situated construction 330Situated practice 22Smith Robertson 205Social causes 227Social constructions 177, 180Social constructivism 6, 10–11Social contract 21–22Social facts 5–6, 37–38, 44–46, 206, 211,

219, 223, 258, 317, 320Dualism as 75

Social forces 248, 249–50are natural forces 67, 93, 175social order 320

Socialism 9, 280Socialist beliefs unscientific 323

Socialism and Saint-Simon, Durkheim 280

Social authority 249Social theory 27, 319Social Theory of Practice Turner 325Social unity 254Society,

Not nominal 304, 314Concretely experienced 304real not ideal 304is not illogical 310–12

SociologyHolding itself to a false model of

abstraction 332as a paradigm science 17–18Epistemological crisis 316

Sociology of knowledge 4, 7, 19, 23–24,53–54, 87, 93, 95–100, 102–03, 105,118, 151, 235, 261, 264, 278–80

Sociology, implications for 316–19, 338Soul 127, 128–29, 163, 188, 189–90, 211,

250Dualism of 188Symbolic expression of personality 190,

255Sounds and movements 14, 37, 168, 173,

180, 209, 221–22, 227, 318, 326,330, 332, 336

Coherence of 324Reduced to concepts 327

Space (and time) 50, 51–54, 307Spencer, Herbert 126Statistical accounts 336Stigma 331, 337Structural Anthropology, French

321–23Structure of Social Action, Parsons 322Suicide, Durkheim 5, 285, 334Sui generis 78–79, 92, 96–97

Supernatural 113–15, 165Symbol, collective 180, 224, 273–75

totem as 39, 141, 142, 156pain as 198generating moral force 221referential 222of feeling 222significant 39–40, 146, 168

Symbolic Interactionism 12, 334Symbolic meaning 40, 317

As effect of moral forces 103–04Symbol systems 329

Taken for granted 327, 331Talk, sequence of 332Taylor 127Texts and narratives 326, 329Theory of knowledge 234–35Theory of Practice Bourdieu 325Theory (social) with a capital “T” 336Thing (s) 139–40, 142, 166, 170, 183–84,

207, 325, 332Thought versus reality 8–9, 15, 16–18Time (and space) 50, 51–54, 163, 172–74,

178, 179, 203–04, 208–11, 305Total institutions 285Totality 306Totem as origin of the sacred 141–49, 174Totem as symbol 39–40, 141, 156Totem creates unity of group 157–58,

224–25Totemic forces 250Totemic rites, causality as essential belief

in 253Totemism 23, 34–41, 109, 110, 121,

135–36, 184, 221, 268Exists only as long as believed in 219,

224–25Logic and 182Defense of 34, 41–42Defense of primitive intelligence

125–26, 127–28, 158–60In defining religion 113Individual Totemism 23, 152–55, 161Clan Totemism 141, 156, 157, 184Sexual totemism 23, 152–61As emblem 142–46, 154, 221As symbol of moral unity 221Emblem as sacred 174And emotion 144–45Collective 155–56, 158Function of 164, 190Symbols 166Versus magic 212Not personal forces 247

Index 355

Totemism as first instance of moral force171

Traditional reason 226Traditional versus Modern (mechanical

versus organic) 1–3, 5, 12, 35, 70,171, 186, 198, 312–14, 319, 338

Translation issues 235, 236, 237, 238–39,240–41

Treatise on Human Understanding 244Trust 2, 22, 337Truth 230, 260, 297, 326, 328

of religious myth 129of concepts 297–99correspondence type 303overlapping consensus 326

Turn completion points 331Turner, Steven 325“Two Concepts of Rules,” John Rawls

319Two Durkheim Hypothesis 4, 7, 320–24Tylor 217, 218, 228Typification, conceptual 269, 331, 333–34

Unity 122, 142–44, 150, 157–58, 181,213, 216, 220

Universal 239, 261, 305Meaning general 291

Universality as a social requirement292

Universe of concepts 306Universe of discourse 329

Use meanings 265, 317Utilitarian 196, 209–10, 211Utility, collective 227, 308Utility, individual 227Utility of mimetic rites 223–26

Validity, empirical 167–68, 169, 175, 194,243, 246, 247, 254, 306, 312–14,326, 337

Value rational 226Vocabulary of motives 229Vocation of sociology 303, 314–15, 325Voices, underrepresented 329, 334

Weber 12, 43, 112, 198, 216, 219, 224–25,226–27, 269, 270

Western industrial capitalism dysfunctional323

Western scientific thinking 200Whitehead, Alfred North 325, 328Whyte, William Foote 327Will, individual 11–12, 190, 246, 250Witnessable 325, 330, 332

achievement 317enactment of social facts 320

Wittgenstein 20, 142, 178, 222, 263, 264,265, 270, 297, 319

Women 329, 334Word sounds 331Working consensus 21–22Wuntian group mind theory 103

Documents

Epistemology and Practice: Durkheim's The Elementary Forms of Religious Life