Problematizing the Luhmannian Constructivist Systems Approach

7/27/2019 Problematizing the Luhmannian Constructivist Systems Approach

1/22

http://csi.sagepub.com/Current Sociology

http://csi.sagepub.com/content/59/6/696

The online version of this article can be found at:

DOI: 10.1177/00113921114197562011 59: 696Current Sociology

Poe Yu-ze WanBungean intervention

(Re-)Problematizing the Luhmannian constructivist systems approach: A

Published by:

http://www.sagepublications.com

On behalf of:

International Sociological Association

can be found at:Current SociologyAdditional services and information for

http://csi.sagepub.com/cgi/alertsEmail Alerts:

http://csi.sagepub.com/subscriptionsSubscriptions:

http://www.sagepub.com/journalsReprints.navReprints:

http://www.sagepub.com/journalsPermissions.navPermissions:

http://csi.sagepub.com/content/59/6/696.refs.htmlCitations:

What is This?

- Oct 20, 2011Version of Record>>

at Univ of Education, Winneba on August 10, 2013csi.sagepub.comDownloaded from
http://csi.sagepub.com/http://csi.sagepub.com/http://csi.sagepub.com/content/59/6/696http://csi.sagepub.com/content/59/6/696http://www.sagepublications.com/http://www.isa-sociology.org/http://csi.sagepub.com/cgi/alertshttp://csi.sagepub.com/cgi/alertshttp://csi.sagepub.com/cgi/alertshttp://csi.sagepub.com/subscriptionshttp://csi.sagepub.com/subscriptionshttp://www.sagepub.com/journalsReprints.navhttp://www.sagepub.com/journalsReprints.navhttp://www.sagepub.com/journalsPermissions.navhttp://csi.sagepub.com/content/59/6/696.refs.htmlhttp://csi.sagepub.com/content/59/6/696.refs.htmlhttp://csi.sagepub.com/content/59/6/696.refs.htmlhttp://online.sagepub.com/site/sphelp/vorhelp.xhtmlhttp://online.sagepub.com/site/sphelp/vorhelp.xhtmlhttp://online.sagepub.com/site/sphelp/vorhelp.xhtmlhttp://csi.sagepub.com/content/59/6/696.full.pdfhttp://csi.sagepub.com/content/59/6/696.full.pdfhttp://csi.sagepub.com/http://csi.sagepub.com/http://csi.sagepub.com/http://online.sagepub.com/site/sphelp/vorhelp.xhtmlhttp://csi.sagepub.com/content/59/6/696.full.pdfhttp://csi.sagepub.com/content/59/6/696.refs.htmlhttp://www.sagepub.com/journalsPermissions.navhttp://www.sagepub.com/journalsReprints.navhttp://csi.sagepub.com/subscriptionshttp://csi.sagepub.com/cgi/alertshttp://www.isa-sociology.org/http://www.sagepublications.com/http://csi.sagepub.com/content/59/6/696http://csi.sagepub.com/


2/22

CS

(Re-)Problematizing theLuhmannian constructivistsystems approach: A Bungeanintervention

Poe Yu-ze WanNational Sun Yat-sen University, Taiwan

Abstract

In the ield o sociological theory, ater the decline o Parsons version o structuralunctionalism, Niklas Luhmanns systems theory came to be seen as the majorrepresentative o systems thinking. While Luhmanns autopoietic systems ontology is

sophisticated and deserves serious consideration, the author argues that the prevailing

identiication o Luhmanns constructivist-oriented approach with the systems approachitsel is problematic mainly or two reasons. First, as the works o the Argentinian-Canadian systems theorist Mario Bunge demonstrate, the systems approach canbe ruitully based on (some version o) scientiic realism. Second, the adherents to

Luhmanns constructivist epistemology in sociology seldom engage in a detailedexamination o the various critiques o anti-realism oered by scientiic and criticalrealists. Drawing on the literature on philosophy o science, especially the writings o

Bunge and a ew leading critical realists, the author suggests that a serious Luhmanniansociologist is obligated to provide more cogent and detailed discussions on philosophy

o (social) science beore taking an opinionated position on epistemology.

Keywords

Mario Bunge, critical realism, epistemology, Niklas Luhmann, operative constructivism,scientiic realism

IntroductionSociologists and social theorists have tended to regard Niklas Luhmanns systems

theory as the major representative of systems thinking. Therefore, scholars

Corresponding author:

Poe Yu-ze Wan, Department o Sociology, National Sun Yat-sen University, No. 70, Lienhai Rd., Kaohsiung

80424, Taiwan.

Email: [email protected]

CSI59610.1177/0011392111419756

Article

Current Sociology

59(6) 696716

The Author(s) 2011Reprints and permission: sagepub.

co.uk/journalsPermissions.nav

DOI: 10.1177/0011392111419756

csi.sagepub.com

http://csi.sagepub.com/http://csi.sagepub.com/http://csi.sagepub.com/http://csi.sagepub.com/


3/22

Wan 697

could entitle their papers Space, boundaries, and the problem of order: A view fromsystems theory (Helmig and Kessler, 2007) or The intersection between systemstheory and grounded theory (Gibson et al., 2005) without specifying which or whoseversion of systems theory they are referring to. This is understandable and to some

extent justified, since Luhmann has indeed developed a sophisticated terminology forgrasping the various dimensions of modern society with all its complexities. Viewedin a positive light, this has been conducive to the reception and teaching of Luhmannstheory outside Germany, facilitating the formation of academic networks centredaround Luhmanns approach.

However, in the very process by which Luhmanns approach was recognized as syn-onymous with systems theory itself, a number of stimulating and fruitful approacheshave been excluded from discussion unintentionally or deliberately. More importantly,Luhmanns constructivist-orientedapproach is often taken for granted as if it were the

only appropriate basis for social systems thinking.The purpose of this article is modest. Without questioning the value of Luhmanns

general theory of self-referential autopoietic systems (cf. Wan, 2011), it merely providesa critical assessment of Luhmanns epistemological project (operative constructivism)mainly on the basis of the philosophical system developed by the accomplishedArgentinian-Canadian systems philosopher Mario Bunge, with an emphasis on its scien-tific realist commitments. Topical questions in scientific research, such as the distinction

between real object and object of knowledge, concept formation, theory evaluationand theory choice, will be discussed to highlight the strengths and weaknesses of

Luhmanns constructivist-oriented approach.

Systems theories: Heterogeneous lines of thought

As is widely known, Luhmanns systems theory is tremendously influenced by phenom-enology,1 Humberto Maturana and Francisco Varelas version of autopoietic theory,Heinz von Foersters second-order cybernetics, and George Spencer-Browns calculus ofform. In other words, the systems theory developed by Luhmann is constructivist-oriented,rooted in specific academic traditions. The problem lies in the fact that Luhmanns pow-

erful rhetoric tends to leave the readers with the impression that his own version of sys-tems theory represents the latest development or paradigm of systems science. Forexample, Luhmann writes:

As we come to modern times the emergence of modern science led more and more to theconclusion that this underlying reality was knowledge itself. (Luhmann, 2002a: 129)

Today the cognitive sciences and the theory of self-referential systems add a new perspective thatcannot be subsumed under idealism or criticized as idealism, that is, insight into the operativeclosure of self-referential systems. A theory of knowledge today that is to be compatible with the

latest developments in science must be able to bear this new perspective. (Luhmann, 2002a: 151;emphasis added)2



4/22

698 Current Sociology 59(6)

This is not the only story available, however. The account of the three waves of systemstheory offered by Keith Sawyer, a leading scholar in the study of social complexity, isespecially worth mentioning here. The first wave is cybernetics in the 1950s and 1960s thathighlights equilibrium and stability under the dynamics of negative feedback loops; with

an emphasis on closed systems, this wave of systems theory significantly influences TalcottParsons structural functionalism. The second wave is the general systems theory (as repre-sented by Ludwig von Bertalanffy and Kenneth E Boulding) from the 1960s to 1980s,which emphasizes the openness, non-linearity and dynamical evolution of systems. WalterBuckley (1998),3 though not as influential as Parsons, is an important figure in this period.The third wave is based on complex dynamical systems theory that developed in the 1990s,laying weight on such concepts as positive feedback loops and path dependence (Sawyer,2005: 1226). Despite the tremendous potential of the third-wave systems theory, themajority of social researchers have abandoned the very concept of system with the decline

of Parsons theory, while systems theory is seen as synonymous with conservatism dis-guised in scholarly arguments. In the field of social theory, Luhmann is the major theoristto have continued the thread of systems thinking.

Luhmann (2006a: 37) claims that his systems theory is indebted to the third and laststage of the development of systems theory, that is, the theory of observing or self-referential systems. But fairly speaking, (1) it is debatable whether the constructivist-oriented theory of self-referential systems is the last stage of the development of systemstheory, since the aforementioned complex systems theory that emerged in the 1990s ismissing from such an account. (2) Even if Luhmann is correct about the historical devel-

opment of systems thinking, the constructivist-oriented approach itself is not beyondcriticism. In this article, the major reference point for mounting a critique of this con-structivist approach is the philosophical system developed by the Argentinian-Canadiansystems theorist Mario Bunge, because while Luhmanns approach is unequivocallyconstructivist-oriented, Bunge is a staunch defender of scientific realism.4 In otherwords, it may be fruitful to compare the contrasting philosophical foundations of thesetwo lines of systems thinking.

In the following sections, by drawing on Bunges and a few leading critical realistsviews on scientific research, I state criticisms of certain aspects of Luhmanns epistemo-

logical project.

De-ontologization of reality: Niklas Luhmanns operative

constructivism

A starting point for understanding Luhmanns epistemological project is his abandon-ment of the distinction between essence and appearance (Ku and Tang, 2008: 7).5 It is inthis context that Luhmann attempts to dispel ontological questions by way of his systemstheory. In his words, The effect of the intervention of systems theory can be described

as a de-ontologization of reality [De-ontologisierung der Realitt] (Luhmann, 2002a:132; 2005: 37). Of central importance to this theme is his epistemological project, oper-ative constructivism. As the following passage indicates:



5/22

Wan 699

There is an external world . . . but we have no direct contact with it. Cognition could not reach theexternal world without cognition. In other words, cognition is a self-referential process. Knowledgecan know only itself. . . . Cognition deals with an external world that remains unknown and, as aresult, has to come to see that it cannot see what it cannot see. (Luhmann, 2002a: 129)

Luhmanns epistemological position is sometimes grouped under the banner of radicalconstructivism, as developed particularly by Ernst von Glasersfeld. Von Glasersfeld(1989: 162) states that radical constructivism boils down to the following principles: (1)the cognizing subject does not receive knowledge passively but actively builds it up; (2)the function of cognition is adaptive, serving the organization of the experiential worldinstead of the discovery of ontological reality. In Rasmussens words, reality or theworld is actually a construction not the things it consists of (an ontological reality). Andsince we all construct our own realities, it is not meaningful to talk of the objective truth

(Rasmussen, 1998: 558). Von Glasersfeld (1991: 28, cited in Rasmussen, 1998: 559)further states that radical constructivism does not want to or is able to be anything else,than a way of thinking over the exceptional world we have access to, and that is the worldof phenomena which we experience. To be sure, a spectre is haunting this passage thespectre of Bishop George Berkeley, or of ontological phenomenalism: the world is asum of appearances, in Kants words (Bunge, 2006: 50).

In upholding knowing without metaphysics and pretentiousness (Holtorf, 1998),Luhmanns epistemological project is consistent with the doctrine of radical construc-tivism, which has been better received in science education due to its emphasis uponactive learning and opposition to spoon-feeding. Even in education, however, it is not

beyond challenge. Bunge (2006: 84) points out that the bulk of modern science is socounterintuitive that it is quite hard to learn, and impossible for an individual to rein-vent by himself. Besides, if the existence of objective truth is denied as recommended

by radical constructivists, criticisms and debates will be irrelevant and teachersexpendable. This judgemental relativism, a logical consequence of radical constructiv-ism, will be detrimental to the ability and willingness of students to learn how to checkfor truth (for more discussions on radical constructivism, see Matthews, 1998;Rasmussen, 1998).

Luhmann does occasionally dub his position as radical constructivism (Luhmann,

1996a: 41, cited in Bjerg, 2006: 59). But in his 1991 article Wie lassen sich latenteStrukturen beobachten?, Luhmann proposes the term operative constructivism, which isalso mentioned sporadically in hisDie Realitt der Massenmedien published in 1995. He

prefers this term to radical constructivism mainly because in his systems theory, refer-ence to the subject is replaced by reference to an empirically observable, operativelyclosed, self-referential system (Luhmann, 1991: 73n, cited in Rasmussen, 1998: 573n).

Therefore, to do justice to Luhmann, it should be noted that his position is not asextreme as some radical constructivists, who tend to doubt the very existence of an exter-nal reality. After all, Luhmann (2002a: 132) states that his theory does not imply that:

. . . reality is denied, for then there would be nothing that operated nothing that observed, andnothing on which would gain a purchase by means of distinctions. It is only the epistemologicalrelevance of an ontological representation of reality that is being called into question.



6/22


7/22

Wan 701

1978, 1986; Sayer, 2000: 1011). Only on the basis of such a realist point of view can weunderstand:

a. Why factual sciences or theories, which are doubly related to reality, namely

semantically and methodologically (Bunge and Mahner, 1997: 95), have alwaysto be tested.

b. Why what some constructivists applaud as the construction of (social) realityis potentially misleading: Although social phenomena cannot exist indepen-dently of actors, most of the time they do exist independently of the particularindividual who studies them (Sayer, 1992: 49). Furthermore, social researchersare not powerful enough to construct the society from scratch. Rather, whatthey often do is to contribute to its reconstruction (Bunge, 1990: 615). Unlessanti-realists provide a suitable semantic theory to establish the superfluity of the

distinction between real object and thought object in scientific research, wemaintain the conclusion that such a distinction has been, and will be, a prerequi-site for any scientific enquiry that cherishes the principle of objectivity (in anon-trivial sense).7

Second, this does not render useless or paltry Luhmanns version of constructivism,which does contain important elements of truth. But here we have to make the essentialdistinction between epistemologicaland ontologicalconstructivism. Luhmanns state-ment that there is an external world . . . but we have no direct contact with it is close to

the mark in the sense that we cannot comprehend the world directly, that is, without themediation of a given conceptual scheme. Therefore, we readily accept that nothing issimply at hand [vorhanden] (Moeller, 2006: 68), but this consideration is far from adeath blow to a full-fledged scientific realism, but part and parcel of it.

Bunges epistemological position is that of a critical (rather than naive) realism.8Naive realists assert that knowledge can reflect, mirror, or correspond to the realitydirectly, but such an uncritical view cannot account forerrors in scientific research andthe attempts to rectify them. Worse, it will be detrimental to advanced research involvingeven more complex theories containing concepts increasingly removed from perception

and intuition (Bunge, 1996a: 355). Consequently, for a critical realist epistemologist,perceptual knowledge, though indispensable, is superficial, incomplete and often wrong,so that it must be enriched with hypothetical or theoretical knowledge (Bunge andMahner, 1997: 134). In other words, indirectknowledge, attained via theories and indi-cators rather then through mere perception or instant intuition, is the deepest (Bunge,2006: 255). In brief, from the perspective of critical realism, scientific research does notguarantee instant, complete and definitive truth (Bunge, 1993: 217).

It follows that this version of realism is not to be confused with reflectionism (e.g.of the early Lenin) or the picture theory of language of the early Wittgenstein. It is aninseparable combination offallibilism and meliorism: all propositions concerning mat-ters of fact are fallible in principle, and they are amenable to corrections. As Bunge(2006: 45) states, the discovery of error is a significant motivation behind scientificresearch, while the faith in the complete truth of some theory only leads to compla-cency and thus stagnation. (It should also be remembered that the recognition of the

http://csi.sagepub.com/http://csi.sagepub.com/http://csi.sagepub.com/


8/22


fallibility of knowledge does not lead to the conclusion that all knowledge is equallyfallible: see Sayer, 1992: 68.) This critical realist epistemology is well in line with epis-temological constructivism, which holds that concepts and their components are human

beings construct. However, epistemological constructivism should not be mistaken for

ontological constructivism, since we construct models of the world, not the world itself(Bunge and Mahner, 1997: 134).9

Nevertheless, a good number of researchers fail to distinguish epistemological fromontological constructivism. For example, while we fully agree that every distinctionmust be understood as contingent, that is, as something that could be different (Luhmann1994: 137), we would not venture the conclusion that:

. . . the second-order observation turns the first-order reality into the observer, and proceedsto observe this constructed observer. It is in this way that things previously construed as given

become contingent, because the second-order observer can claim that other observers are ableto construct different realities by means of different distinctions. (Lu, 2003: 256; emphasisadded)10

To borrow the terms of critical realism, we have no difficulty in accepting epistemicrelativism, which underlines the fact that the world can only be known in terms ofavailable descriptions or discourses (Sayer, 2000: 47) and that all epistemologicalclaims are temporally and spatially located (Wight, 2004: 202). Nevertheless wereject judgemental relativism, i.e., the self-defeating view that one cannot judge

between different discourses and decide that some accounts are better than others(Sayer, 2000: 47). This, of course, brings us to the question of the evaluation andchoice of theories.

Observation, distinction and theory evaluation/choice:Where operative constructivism falls short

Spencer-Brown and Luhmanns works serve well to remind us that all observation pre-supposes a specific distinction that at the moment of its use must be employed blindly(Luhmann, 2002a: 135). Except for this insight, however, there seems not much workingsocial scientists can benefit from. The Danish sociologist Jens Rasmussen (2004: 185;emphasis added) believes that Luhmanns operative constructivism contributes to theradicalization of hermeneutics because it demonstrates that observation alwaysinvolves an observer, and as such it is always biased. An observation (operation) isalready an interpretation; therefore it makes no sense to distinguish between observationand interpretation, since all interpretation involves observation. But to what extent isthis really the case? Since this involves the fundamental disputes between (scientific)realism and (ontological) constructivism, a summarized discussion is helpful.

Bunge and Mahner (2004: 211) argue that the two basic assumptions in any scientific

research are ontological and epistemological realism. Ontological realism is the thesisthat the universe, or reality, exists in se et per se (Bunge, 2006: 251), and the successand especially the failure of scientific theories can only be accounted for satisfactorily onthe basis of this assumption. Sayer (2005: 215) argues convincingly that:



9/22

Wan 703

. . . the fallibility of knowledge and truth claims derives from the very independence of objectsfrom what we think about them. We can only be mistaken if there is something independent ofour thought to be mistaken about. If the objective is collapsed into the subjective, so that thereis nothing outside knowledge or discourse, then these must be infallible.11

Epistemological (critical) realism, as noted earlier, involves the following theses:

1. The world can be known, however partially. It does not claim that our knowledgeof the world is accurate; it suffices that such knowledge be partially or approxi-mately true.

2. Indirect knowledge is the deepest.3. All knowledge concerning facts is incomplete and refutable (moderate

scepticism).

4. Any piece of knowledge can, in principle, be improved upon to make it morecomprehensive or accurate (meliorism).

To these Bunge adds:

5. Any set of facts can, in principle, be accounted for by alternative hypotheses ortheories, but while competing hypotheses and theories are empirically equivalentin some aspects,12 their accuracy or depth may differ in others (moderate

pluralism).6. Objective knowledge is superior to subjective conjecture. Following Nicholas

Rescher, Bunge understands objectivity as follows: A factual proposition is saidto be objective if it refers to real existents in an impersonal manner, and describesthem to the best of its authors knowledge (Bunge, 2006: 31; see also 1983: 80,1556).

Now we turn to the preceding claim made by Ramussen that an observation is alreadyan interpretation and therefore it makes no sense to distinguish between observationand interpretation.

In scientific research, the claim that observation is always theory-laden is

undoubtedly correct, as Bunge (1999: 173) argues that scientific observations, unlikeordinary ones, are designed and conducted in the light of hypotheses. However, thisfact alone does not justify the belief that there are no crucial differences betweentheoretical and observational concepts/terms, even that all observations are alreadydeterminedby theories.

Although observational and theoretical concepts/terms are closely related, they differin varying degrees. For example, colour is the former, while wavelength the latter; inMarxian economic theory, price is the former, value the latter. Besides, scientific dataresult not merely from theory-laden procedures, but also from more or less theory-free

observations such as those of an instrument pointer (Bunge, 2006: 20). Sayer (2000: 41,47, 65n; 1992: 734; see also Kurki, 2008: 2078) also contends that (1) ordinary obser-vations, which usually require a set of concepts instead of theories in the scientificsense, are thus less theory-laden than conceptually mediated; (2) theory-laden is not



10/22


identical with theory-determined: if concepts have already specified everything aboutwhat could be observed, it will be superfluous to make observations; and (3) even if wecan only think within specific conceptual systems, what we can think ofis not necessarilyalready contained in them.

What follows is a brief discussion on an important issue to which, I believe,Luhmanns constructivist approach provides no adequate answer: theory evaluationand theory choice.

Distinctions are necessary but far from sufficient for social science research. Forexample, in the formation of concepts, distinctions can be no more than the point ofdeparture. Most importantly, distinctions are not arbitrary, but have to be guided by the-ory and grounded in evidence. In discussing the role of abstraction in scientific research,Sayer (1992: 138) writes cogently:

Abstractions can be made in various ways. . . . A rational abstraction is one which isolates asignificant element of the world which has some unity and autonomous force, such as astructure. A bad abstraction arbitrarily divides the indivisible and/or lumps together theunrelated and the inessential, thereby carving up the object of study with little or no regard forits structure and form.

In other words, not all observations (including second-order observations) and distinc-tions (including the new distinctions introduced by second-order observations) are ten-able. I submit that Luhmanns constructivist approach is of limited value when it comesto the criteria by which a concept can be assessed, and it thus encounters difficulty inguiding practising social scientists in the formation of fruitful or adequate concepts(for useful discussions on concept formation, see e.g. Bunge, 1983: 16474; Bunge andMahner, 1997: 21524; Gerring, 1999; Goertz, 2006).13 Furthermore, even concept for-mation marks a preliminary stage in social research, with many more steps needed toeventually craft models or theories with explanatory power regarding mechanisms thatunderlie social states of affairs (Schmid, 2011: 139).

Here are a few suggestions and remarks from the perspective of scientific realism.First, the shift of perspectives (of individual and collective observers alike) at the epis-temological level is not tantamount to the changes in natural or social kinds. Therefore,

for all of the attractiveness of the claim that in modern society, by means of the con-struction of change, society is able to continually observe its own conditions, pro-duces new determinations of it, and renders its present conditions contingent andtransient for other conditions to appear (Lu, 2007: 249), it has to be noted that socialchange, as the real object, falls under the purview of ontology, while the self-observationof society (by way of its conception of change), and thereof all the discourses andsemantics about social change, is the thought object in the epistemological domain.Individual or collective observations may and may not exert impacts on the realitythrough certain mechanisms. The point is that observations are not on a par with the

reality observed, and therefore their changes are apt to be out of phase with each other.The conceptual and observational contingency of researchers or actors, as well as thelogical contingency of statements, does not eo ipso render the categories and relationsof social reality contingent. Therefore, the fact that observers treat certain objects as



11/22

Wan 705

contingent may arise from their chaotic abstractions or from what Sayer (1992: 62) callsunwarranted associations. For example, if an observer construes the necessary (inter-nal) relations (i.e. what the object is depends on its relation to the other) between par-ent and offspring or master and slave as contingent, one may conclude that his or

her observation is chaotic and inept.14In addition, an important task for social researchers is to probe into the conditions and

mechanisms through which the (self-)observations of social actors and the social realityinfluence, enable, or constrain (but not determine) each other. One has also to rememberthat some social relations are necessary (e.g. landlord and tenant) rather than contingent(e.g. entrepreneur and best-selling author), with attention paid to the fact that necessaryrelations may be asymmetric in the sense that one object in a relation can exist withoutthe other, but not vice versa (e.g. the relations between money and banking systems)(Sayer, 1992: 90). To appreciate and grasp these dimensions, rational abstractions and

well-conceived concepts are of paramount importance.A few further points can be made. First, the undoubtedly significant differences

between natural and social kinds do not render the basic realist thesis invalid that realobjects and thought objects, though related, have to be distinguished. Based on Bhaskarsdiscussions on the limits of naturalism in social science, Wendt (1999: 6971) spells outfour major differences between social and natural kinds: (1) social kinds are more spacetime specific than natural kinds; (2) the existence of social kinds hinges on the beliefs,intentions and concepts held by social actors; (3) the existence of social kinds also relieson human (discursive) practices; and (4) since most social kinds are inherently relational

(i.e. constituted by social relations), they have both an internal and external structure. Inaddition to Colin Wights illuminating qualifications of these points (Wight, 2006: 534)and Daniel Littles rejection of social kinds as understood in a strong sense (Little,1998),15 I suggest that (1) the fact that most social practices are concept-dependent doesnot entail the identity between society and ideas otherwise wishful thinking will suf-fice. Given that all knowledge is fallible, the knowledge possessed and produced by socialactors is not impervious to correction or criticism (Kurki, 2008: 209; Sayer, 2000: 34;Wight, 2006: 57). (2) Social practices are inclusive of, but not exhausted by, symbolic anddiscursive practices. A good number of human activities are conducted independently of

languages or symbolic systems, particularly those that appear more material and typicallyintervene in or act upon the environment (e.g. drinking water, walking up and down thestairs, or taking up a handful of earth). (3) Human activities take place under specificconditions, which involve not only symbolic or conceptual but material aspects of theworld, and the latter are in a substantive sense independent of peoples perceptions ordescriptions of them (Carter and Sealey, 2009: 72). For example, natural resources, or acertain degree of division of labour and productivity, are required for some types of activi-ties to be possible at all. The unanticipated/unintended consequences of these activitieshave to be taken into consideration as well. (4) Human beings are not separate from nature

but a tiny part of it. Like other entities in the natural world, human beings are entitiesendowed with specific emergent causal powers: they interact with the natural word (suf-fice it to recall Marxs concept of metabolism between humankind and nature), andthereby keep producing a wide range of extra-discursive effects. From this perspective,the semantic structure, on which Luhmannian sociologists tend to focus their attention, is



12/22


far from a self-contained or self-referential system. Its very existence depends on concretereferents, which include not only human (conceptual and material) constructs but naturalentities. A major weakness of semiotics from Saussure onward lies in its exclusive stresson the arbitrary relations between signifiers and signifieds the role of referents is thus

greatly overlooked (Sayer, 2000: 62). Excessive concentration on the signifying practicesand symbolic systems of humankind may divert ones attention from the fact that human

beings are part of nature and lead to anthropocentrism.Second, social reality is partly dependent on the (self-)observations of social mem-

bers, but never reducible to them. One has to distinguish between the (self-)observationof social actors (and the resulting crystallized semantic structure), their discursive prac-tices (the practical relations that social actors construct within specific conceptualframeworks between themselves and concrete objects in the world)16 and the variousconditions and effects (e.g. the unintended consequences) of these practices. In addition

to employing second-order observations of societys self-observation, it is incumbent onsociologists to provide deep explanations of social reality (including the events weexperience and the underlying mechanisms responsible for the events which occur or donot occur due to other interfering mechanisms) by way of rational abstraction, conceptformation and theory construction. These explanatory endeavours are inevitably relatedto ones understanding of the constitution of society (to borrow Giddens phrase),namely ones social ontology. Consequently, epistemology is not all that matters, andavoidance of ontology (or reduction of it to epistemology) leads nowhere fast.

Lastly, a few words regarding the choice of argumentative strategy. Manifest in the

writings of Luhmannian sociologists is the conviction that Luhmanns approach can betaken for granted, or that his reasoning is so conclusive that no one could be gullibleenough to believe otherwise. All one needs to do is to quote Luhmanns arguments to

justify ones constructivist position, or simply to apply his insights to empirical research,as if the dust has long settled on the debate between (whatever versions of) constructiv-ism and realism in philosophy of (social) science. To my knowledge, few followers ofLuhmann have seriously engaged in this debate, other than to have cited those materialsthat Luhmann himself finds compelling (e.g. the arguments of Spencer-Brown, VonFoerster and Gotthard Gnter).

Take William Rasch, an exponent of Luhmann, as an example. He asserts thatReality . . . is not a pattern of objects but an account of such a pattern (Rasch, 2000:1516). The reality is identical to the account of it. Rasch, however, presents no solidarguments for such a bold assertion other than to repeat WVO Quines no less contro-versial statements:

The particulars of physical explanation are eminently revisable, but revision is not the directresult of empirical confirmation or refutation of discrete elements of particular theories.Theoretical descriptions of the world, in other words, are overwhelmingly underdetermined bythe putative object they describe, because narratives to account for empirical data can be spun

in a wide variety of ways. (Rasch, 2000: 15)

The conclusion, naturally, is that description is contingent in the sense that other plausi-ble descriptions are always possible (Rasch, 2000: 18). Nevertheless, anyone with some



13/22

Wan 707

knowledge of philosophy of (social) science would agree that such questions pertainingto the underdetermination thesis are far more complicated than presented above. Quinesarguments serve more as the starting point for discussion than as the final word.

First, as Bunge and Mahner (1997: 132) point out, underdetermination with regard to

the available evidence at a certain time does not entail that further evidence at some latertime will never bear upon the empirical adequacy of the theory or hypothesis. Therefore,the Ptolemaic, Tychonian and CopernicoGalilean models of the planetary system had

been empirically equivalent for some time, but they were no longer so in the secondhalf of the 17th century.

Second, scientists actually seldom face the problem of choosing among severalempirically equivalent theories. As Okasha (2002: 73) notes, there have been relativelyfew cases of underdetermination in the history of science. A more realistic picture is thatit is usually difficult enough (often impossible) to find one theory to fit all the known

data, let alone several, let alone indefinitely many (Worrall, 2000: 353).Third, only empiricists take empirical adequacy as the only criterion for theory choice.

For scientific realists subscribing to what Bunge calls ratio-empiricism,17 empirical ade-quacy is a necessary but not sufficient indicator of truth. Some conceptualindicators arerequired as well. However, neither Luhmann nor his followers seem to have taken seri-ously these questions in advocating their constructivist epistemology.

By contrast, Bunge (1996b: 491) offers 10 conceptual indicators for assessing rivaltheories or hypotheses: (1)Intelligibility: Is it clear or obscure? Can it be elucidated?(2) Logical consistency: Does it contain contradictions or inconsistencies? If it does,

can these inconsistencies be removed without abandoning its most important assump-tions? (3) Systemicity: Is it a conceptual system or part of one, or an isolated piece ofknowledge? (4)Literalness: Is it composed of literal statements or just a metaphor? Ifan analogy, is it deep or shallow? (5) Testability: Can it be checked against conceptu-ally and empirically? (6)Empirical support: If it is tested, are the results favourable?(7) External consistency: Is it in accordance with the bulk of antecedent knowledge?(8) Originality: Is it novel or trite? Does it solve major problems and unify unrelatedresearch fields? (9) Heuristic power: Does it pose interesting questions? (10)

Philosophical soundness: Is it consistent with the philosophy underlying practical sci-

entific research?18

The upshot is that the evaluation of scientific hypotheses and theo-ries is performed on the strength of a whole battery of tests, some empirical, othersconceptual (Bunge and Mahner, 1997: 133).

To construct a normative account of science is not to gainsay the obvious fact thatscientists often accept certain conclusions because of personal interests, the reputationsof other scientists, or because they are inclined to accept assumptions which have becomeconventional in their field (Couvalis, 1997: 148). However, this fact does not refute thescientific realist interpretation of science, nor does it license the convenient conclusionshared by a number of constructivist-relativists that the production of scientific knowl-edge is nothing more than a power game or a self-sustaining convention. On the onehand, I share Okashas view that the claim that social factors represent an importantcausal influence on scientists beliefs needs to be shown on a case-by-case basis, if it isto be taken seriously (Okasha, 2000: 295). On the other, indeed, while social and psy-chological factors do have a role, their effects tend to be outweighed by what Robert



14/22


Merton and CH Waddington call scientific ethos, which inclines scientists to be open-minded and sceptical.

There is no need to agree with Bunges accounts or his version of scientific realism inevery detail, but at least the criteria he puts forward are clear and contestable (rather than

impervious to criticism), and moreover consistent with what critical realists call judge-mental rationalism orjudgemental rationality: one is able to make rational judgementsamong competing views or theories.19

From the above discussion we find it reasonable for Bunge to state that:

Relativists, subjectivists and constructivists have incessantly avoided or ignored the argumentsfavourable to realism. Besides, some of them adopt too narrow a definition of realism. Forexample, they assert that realism maintains that all the scientific truths are totally true, which isevidently not the case at all. (Bunge and Vacher, 1993: 85)

Sophisticated versions of anti-realism are undoubtedly available in philosophy of (social)science (e.g. Larry Laudan), but constructivists in the Luhmannian tradition are all toooften given to piling assertions upon assertions, with scant attention to existing literatureconcerning the matter discussed.

Concluding remarks

The conclusion of this article is as follows.

First, systems theory is a catch-all term covering a multiplicity of research pro-grammes and traditions, of which Luhmanns approach is far from the only representa-tive. In light of the central principles of scientific realism, this article contends thatLuhmannians tend to confound the real object with thought object and thus reduce theontological to the epistemological.

Second, the exponents of Luhmann should not be content with repeating Luhmannsarguments or applying them to social research uncritically. If understood within a frame-work that takes into account the existing discussions in philosophy of (social) science(e.g. the underdetermination thesis, concept formation, theory evaluation and choice),the strengths and weaknesses of Luhmanns constructivist approach will be laid bare anddiscussed in a positive way.

Lastly, consider Kneer and Nassehis (1993: 103) statement that there are a numberof distinctions, a number of different contexts, which cannot be placed together andcompared from an Archimedean point of observation. . . . Thus the absolutely correctview of things is eliminated. . . . Every observation is a contingent construction.Following the discussion of this article, I submit that while the existence of multiple

paradigms may be desirable in social science, the eclectic view that all theories or para-digms have something useful to contribute (Sayer, 1992: 77) is not, because the latterverges on judgemental irrationalism, that is, the idea that you cannot have better or

worse grounds for belief (Bhaskar, 2002: 214), and serves only to protect muddledthinking from criticism. Sayer (1992: 78) argues compellingly that this view is . . . dan-gerous in that it evades critical evaluation my critics are in fact talking about differentrealities. The principle of equality applies only to the moral worth of persons, not to



15/22

Wan 709

empirical beliefs and theories (Sayer, 2000: 48). Therefore, to rid social research of whatSayer provocatively calls Mussolini relativism,20 I believe that only the multiple para-digms that emerge in the course of conversation, comparison and debate are of the utmostvalue and thus worth preserving.

Funding

This article is the result of a research project supported by National Science Council, Taiwan,project number 99-2410-H-110-072.

Acknowledgements

I am deeply grateful to the editor and two anonymous referees for helpful criticisms andsuggestions.

Notes

1. For a general discussion see Paul (2001). Achterbergh and Vriens (2009: 125n) write con-cisely: By introducing the class of meaning processing systems (comprising the classesof social and psychic systems) Luhmann links up with the phenomenological tradition thatconsiders conscious states as both about something . . . and marginally referring to both therest of the world and the past and the future. Luhmann seems to borrow these ideas andapplies them to characterize conscious states and communications, the elements ofrespectively psychic and social systems. Or as Viskovatoff (1999: 500) explains: To makethe autonomy of social systems from psychic systems complete, Luhmann adopted the unu-

sual position that meaning, the medium of both thoughts and communications, is in noway more intrinsic to minds than it is to social systems by developing a phenomenologicaldefinition of meaning. I also agree with Leydesdorff (2006: 109) that Luhmanns elevationof meaning as the central category of his systems theory coincides with the core assumptionin the tradition of symbolic interactionism that inter-human interactions generate meaning insocial situations.

2. See also Luhmanns retirement lecture at Bielefeld University (Luhmann, 1994).3. Buckleys works, especially his views on morphogenesis that is, those processes which

tend to elaborate, create, or change a systems given structure or state. . . . Biologicalevolution, learning, and constitutional amendments are examples of morphogenesis

(Buckley, 1998: 53) have influenced Margaret Archer, a leading critical realist socialtheorist today. Archer, however, has not devoted herself to further developing systemsthinking.

4. Bunge has constructed a new, science-oriented philosophical system (Bunge, 1996a: 12)covering ontology, epistemology, semantics, philosophy of science and technology, moralphilosophy and praxiological theory. His monographs, such as Causality (1959),Philosophyof Physics (1973), The MindBody Problem (1980), Scientific Materialism (1981) and espe-cially his eight-volume Treatise on Basic Philosophy (197489), constitute perhaps the mostcomprehensive and systematic philosophy of the twentieth century (Pickel, 2004: 170n).Bunges influence has been considerable in Latin America, where his works tend to be taken

seriously by systems scientists and social researchers (e.g. Blsamo, 2000). As for Europe,scholars in the German-speaking academic communities have begun to compare Luhmannssystems theory with Bunges systemism and have recognized the latters contribution(Klassen, 2004, 2009; Obrecht, 2005, 2009; Sorg, 2009; Staub-Bernasconi, 2000, 2005,2007), although Luhmanns works remain much more influential. In English-speaking social



16/22


sciences, Bunges contribution is also being explored, utilized and evaluated. For example, in2004, the prominent journalPhilosophy of the Social Sciences devoted two issues to Bungescontributions to the philosophy of social science (e.g. his emphasis on systemism and mecha-nismic explanation), followed by another issue in 2007 that called for rethinking systems

theory (see also Wan, 2011). Incidentally, Luhmann occasionally refers to Bunge in his writ-ings, but he has never sought a systematic engagement with Bunges theory.5. Unless otherwise indicated, all translations from foreign languages are mine.6. The distinction between real object (objet rel) and thought object or object of knowledge

(objet de pense) is suggested by Louis Althusser and adopted by, among others, Andrew Sayer(1992: 479; 2000: 65n). In his 1857 manuscript introducing the Grundrisse, Marx refers tothe former as conceptual totality (Gedankentotalitt), mental fact (Gedankenkonkretum),or conceptual entity (Gedankenganze), and the latter as concrete subject (das realeSubjekt).

7. Sayer (2000: 58ff.) usefully distinguishes three different, often insufficiently elucidated,

meanings of objectivity: (1) value-neutrality or value-freedom; (2) the search for objectiveknowledge; (3) referring to the nature of things regardless of what one thinks of them. What Imean by objectivity here is a synthesis of the second and third. I fully agree with Bunge thatobjectivity is not to be confused with impartiality or value-neutrality: In short, partisanshipmust not be allowed to misrepresent reality, but it may motivate fruitful research (Bunge,2007: 119). Notice that this commitment to objectivity is not to deny that (1) the productionof knowledge is an integral part of (social) reality; (2) human beings are capable of inter-vening in that reality by means of the production of knowledge; (3) their reference to andgrasp of that reality can only take place within a given (though not unchanging) conceptualframework.

8. Bunges usage of the term critical realism should be distinguished from critical realism(CR) as a specific school of thought associated with Rom Harr, Roy Bhaskar, etc.9. In other contexts, Bunge (2001) refers to epistemological constructivism as cognitive

constructivism.10. To quote from Luhmanns text: This multiplicity [of reality], regardless of whether it is a

multiplicity of sides or of perspectives, is itself a product of cognition, resulting from cer-tain types of distinctions (Luhmann, 2002a: 144). Such logical leaps, rooted mainly in thefailure to distinguish different types of constructivism, dominate Luhmanns writings onepistemology.

11. By contrast, it is difficult for intuitionists (e.g. Bergson and Husserl) to account for errors,

since they claim to have direct access to truths (Bunge, 2006: 31). Sayers critique of stand-point theory is also noteworthy: without such concepts as misrepresentation or error (theacceptance of which entails the view of ontological realism that reality exists independentlyof our mind and we may misconstrue this reality), critiques of the prevailing malestreamsocial science will be reduced to complaints about who is making the truth claims, not aboutwhether they are true (Sayer, 2000: 53).

12. This amounts to what is usually referred to as the weak underdetermination thesis: Anytheory has rivals that entail the same actual given observational evidence. The strong under-determination thesis, on the other hand, asserts that any theory has rivals that are equallysupported by all possible observational evidence for it (Devitt, 2005: 7778). My view is

that the former is reasonable and realistic, while the latter is not (e.g. Couvalis, 1997: 1467;Ladyman, 2007: 33740).13. According to Gerring (1999: 3578), a concept involves three aspects: (1) its denotation (the

object to be defined); (2) its connotation or definition (the attributes assigned to the object);and (3) the term covering (1) and (2). A good concept has to bring the three aspects together



17/22

Wan 711

in a well-organized manner. Indulgence in extolling the merits of second-order observationor new distinctions is therefore of little help.

14. Sayer (1992: 160) reminds us to avoid the confusion between logical necessity or possibility,which concern relations between statements, and natural or material necessity or possibility

that concern the relations between things. Certain traditions of social theory (especially criti-cal realism and Marxism) have provided far more sophisticated discussions on the question ofcontingency than those in Luhmanns writings, in which contingency is merely understood assomething . . . neither necessary nor impossible (Luhmann, 1995: 106; see e.g. Clark andYork, 2007; Sayer, 1992, 2000; Williams and Dyer, 2004). Sayer (1992: 8993, 13843) arguesthat it is crucial for social scientists to distinguish between necessary (internal) and contingent(external) relations, because chaotic conceptions rather than rational abstractions may occurwithout such a distinction. Issues like this are beyond Luhmanns purview, and I believe this isclosely associated with his lamentable reduction of the ontological to the epistemological.

15. Illuminating discussions on social kinds can also be found in Groff (2004: 11821),

in which Brian Elliss well-known objections to social kinds are critically examined.Generally I adopt the position of critical naturalism: social, political, and economic struc-tures change over time independent of the activities of social scientist, so social kindsor social facts, and the mechanisms that underlie them, are not immutable subjects ofstudy in the same sense as the physical entities that are the focus of some of the naturalsciences (Bennett, 2008: 210; see also Gorski, 2009: 1656; Little, 2010: 527). Anotherimportant aspect to consider is what Hacking (2007: 293) aptly calls making up people.As he points out, in certain non-trivial circumstances, our sciences create kinds of peoplethat in a certain sense did not exist before. This is because human beings are reactivekinds that constitute moving targets: They are moving targets because our investigations

interact with them, and change them. And since they are changed, they are not quite thesame kind of people as before. The target has moved. I call this the looping effect. Seealso Carter and Sealey (2009: 7681).

16. Such relations usually involve material processes of searching, making contact, separatingout, dividing, combining, activating [and] manipulating (Sayer, 1992: 48).

17. Bunges version of scientific realism is hostile to both abstracted empiricism (to borrowC Wright Mills phrase) and radical rationalism, because neither blind empirical trial norunchecked speculation is of help (Bunge, 1996a: 322).

18. See also Bunge (1998: 388403; 1996a: 1034, 1813, 21617) and Vollmer (1995). Theunderdetermination thesis is highly controversial in philosophy of science, and it is beyond

the scope of this article to explore it in all its aspects. Suffice it to say that Luhmanns follow-ers are obligated to provide more cogent and detailed discussions before they take an opinion-ated position on epistemology.

19. I have trimmed the scope of this article to highlight the philosophical ideas of Mario Bungeand to assess Luhmanns approach mainly in light of Bunges version of scientific realism. Ido not cover all the issues concerning scientific realism and its role in social science.

20. Benito Mussolini once stated that Everything I have said and done in these last years is rela-tivism by intuition. . . . all ideologies are of equal value, [and] all ideologies are mere fictions(cited in Sayer, 2000: 47).

ReferencesAchterbergh J and Vriens D (2009) Organizations: Social Systems Conducting Experiments.

Berlin: Springer.Blsamo IB (2000) Conditions for testing performance. In: Parra-Luna F (ed.) The Performance of

Social Systems. Dordrecht: Kluwer, 201216.



18/22


Bennett A (2008) The mother of all isms. In: Groff R (ed.) Revitalizing Causality. London:Routledge, 205219.

Bhaskar R (1978)A Realist Theory of Science. Hassocks: Harvester Press.Bhaskar R (1986) Scientific Realism and Human Emancipation. London: Verso.

Bhaskar R (2002)From Science to Emancipation. Thousand Oaks, CA: Sage.Bjerg O (2006) Accelerating Luhmann. Theory, Culture and Society 23(5): 4968.Brown GS (1969)Laws of Form. London: Allen and Unwin.Buckley W (1998) Society: A Complex Adaptive System. Sydney: Gordon and Breach.Bunge M (1983) Treatise on Basic Philosophy. Vol. 5. Epistemology and Methodology I.

Dordrecht: D Reidel.Bunge M (1990) Boudon on anti-realism in social studies. In: Weingartner P and Dorn GJW (eds)

Studies on Mario Bunges Treatise. Amsterdam: Rodopi, 613616.Bunge M (1993) Realism and antirealism in social science. Theory and Decision 35(3): 207235.Bunge M (1996a)Finding Philosophy in Social Science. New Haven, CT: Yale University Press.

Bunge M (1996b) Mindbody problem. In: Beaumont JG et al. (eds) The Blackwell Dictionary ofNeuropsychology. Cambridge, MA: Blackwell, 488492.Bunge M (1998)Philosophy of Science, Vol. II, rev. edn. New Brunswick, NJ: Transaction.Bunge M (1999) The SociologyPhilosophy Connection. New Brunswick, NJ: Transaction.Bunge M (2001) Relativism: Cognitive. In: Smelser NJ and Baltes PB (eds) International

Encyclopedia of the Social and Behavioral Sciences, Vol. 19. New York: Elsevier, 13,00913,012.

Bunge M (2006) Chasing Reality. Toronto: University of Toronto Press.Bunge M (2007) Did Weber practise the objectivity he preached? In: McFalls L (ed.)Max Webers

Objectivity Reconsidered. Toronto: University of Toronto Press, 117134.

Bunge M and Mahner M (1997)Foundations of Biophilosophy. New York: Springer.Bunge M and Mahner M (2004) ber die Natur der Dinge. Stuttgart: S Hirzel Verlag.Bunge M and Vacher L-M (1993)Entretiens avec Mario Bunge. Montral: Liber.Carter B and Sealey A (2009) Reflexivity, realism and the process of casing. In: Byrne D and Ragin

CC (eds) The Sage Handbook of Case-Based Methods. Thousand Oaks, CA: Sage, 69100.Clark B and York R (2007) The problem with prediction. The Sociological Quarterly 48(4):

713744.Couvalis G (1997) The Philosophy of Science. London: Sage.Devitt M (2005) Scientific realism. In: Johnson F and Smith M (eds) The Oxford Handbook of

Contemporary Philosophy. New York: Oxford University Press, 767791.

Gibson B, Gregory J and Robinson PG (2005) The intersection between systems theory andgrounded theory. Qualitative Sociology Review 1(2): 321.Gerring J (1999) What makes a concept good?Polity 31(3): 357393.Goertz G (2006) Social Science Concepts: A Users Guide. Princeton, NJ: Princeton University

Press.Gorski P (2009) Social mechanisms and comparative-historical sociology: A critical realist pro-

posal. In: Hedstrm P and Wittrock B (eds)Frontiers of Sociology. Leiden: Brill, 147194.Groff R (2004) Critical Realism, Post-Positivism, and the Possibility of Knowledge. London:

Routledge.Hacking I (2007) Kinds of people: Moving targets. Proceedings of the British Academy 151:

285318.Helmig J and Kessler O (2007) Space, boundaries, and the problem of order.International PoliticalSociology 1(3): 240256.

Holtorf CJ (1998) Knowing without metaphysics and pretension. In: Andersson A-C et al. (eds)The Kaleidoscopic Past. Sweden: Gothenburg University, 9198.



19/22

Wan 713

Klassen M (2004) Was leisten Systemtheorien in der Sozialen Arbeit? Bern: Haupt.Klassen M (2009) Systemtheorie und ihr Transfer in die sozialarbeiterische Praxis. In: Riegler A

et al. (eds) Soziale Arbeit zwischen Profession und Wissenschaft. Vermittlungsmglichkeitenin der Fachhochschulausbildung. Wiesbaden: VS Verlag, 87102.

Kneer G and Nassehi A (1993) Niklas Luhmanns Theorie sozialer Systeme. Eine Einfhrung.Mnchen: W Fink.Ku C-H and Tang C-C (2008) How is sociological enlightenment possible? An assessment of

Niklas Luhmanns theoretical development. In: Hwang RC (ed.) Contemporary EuropeanSocial Theory. Hangzhou: Zhejiang University Press, 320 [in Chinese].

Kurki M (2008) Causation in International Relations. Cambridge: Cambridge University Press.Ladyman J (2007) Ontological, epistemological, and methodological positions. In: Kuipers TAF

(ed.) General Philosophy of Science. London: Elsevier/North Holland, 303376.Leydesdorff L (2006) The Knowledge-Based Economy: Modeled, Measured, Simulated. Boca

Raton, FL: Universal Publishers.

Little D (1998) Microfoundations, Method, and Causation. New Brunswick, NJ: TransactionPublishers.Little D (2010)New Contributions to the Philosophy of History. New York: Springer.Lu K-H (2003) Contingency and time in functionally differentiated societies. In: Hwang RC (ed.)

Modernity, Post-Modernity and Globalization. Taipei: Left Bank, 248276 [in Chinese].Lu K-H (2007) The image of social changes in the sociology. Societas: A Journal for Philosophical

Study of Public Affairs 22: 235264 [in Chinese].Luhmann N (1991) Wie lassen sich latente Strukturen beobachten? In: Krieg P and Watzlawick

P (eds) Das Auge des Betrachters. Beitrge zum Konstruktivismus. Mnchen: Piper,6174.

Luhmann N (1994) What is the case? and what lies behind it? Sociological Theory 12(2):126139.Luhmann N (1995) Social Systems. Stanford, CA: Stanford University Press.Luhmann N (1996a) Protest. Systemtheorie und soziale Bewegungen. Frankfurt am Main:

Suhrkamp.Luhmann N (1996b)Die Realitt der Massenmedien, 2nd edn. Opladen: Westdeutscher Verlag.Luhmann N (2000) The Reality of the Mass Media. Cambridge: Polity.Luhmann N (2002a) Theories of Distinction. Stanford, CA: Stanford University Press.Luhmann N (2002b)Einfhrung in die Systemtheorie. Heidelberg: Carl-Auer-Systeme.Luhmann N (2005) Soziologische Aufklrung 5: Konstruktivistische Perspektiven, 3rd edn.

Wiesbaden: VS Verlag.Luhmann N (2006a) System as difference. Organization 13(1): 3757.Luhmann N (2006b) Cognition as construction; trans. Moeller H-G. In: Moeller H-G Luhmann

Explained. Chicago: Open Court, 241260.Matthews MR (ed.) (1998) Constructivism in Science Education: A Philosophical Examination.

Dordrecht: Kluwer.Moeller H-G (2006)Luhmann Explained. Chicago: Open Court.Obrecht W (2005) Ontologischer, sozialwissenschaftlicher und sozialarbeitswissenschaftlicher

Systemismus. In: Hollstein-Brinkmann H and Staub-Bernasconi S (eds) Systemtheorien imVergleich. Wiesbaden: VS Verlag, 93172.

Obrecht W (2009) Probleme der Sozialen Arbeit als Handlungswissenschaft und Bedingungen ihrerkumulativen Entwicklung. In: Birgmeier B and Mhrel E (eds)Die Sozialarbeitswissenschaftund ihre Theorie(n). Positionen, Kontroversen, Perspektiven. Wiesbaden: VS Verlag,113130.



20/22


Okasha S (2000) The underdetermination of theory by data and the strong programme in thesociology of knowledge.International Studies in the Philosophy of Science 14(3): 283297.

Okasha S (2002)Philosophy of Science. Oxford: Oxford University Press.Paul AT (2001) Organizing Husserl: On the phenomenological foundations of Luhmanns systems

theory.Journal of Classical Sociology 1(3): 371394.Pickel A (2004) Systems and mechanisms.Philosophy of the Social Sciences 34(2): 169181.Rasch W (2000)Niklas Luhmanns Modernity. Stanford, CA: Stanford University Press.Rasmussen J (1998) Constructivism and phenomenology. Cybernetics and Systems 29(6):

553576.Rasmussen J (2004) Textual interpretation and complexity. Nordic Educational Research 24(3):

177194.Sawyer RK (2005) Social Emergence. Cambridge: Cambridge University Press.Sayer A (1992)Method in Social Science, 2nd edn. London: Routledge.Sayer A (2000)Realism and Social Science. London: Sage.

Sayer A (2005) The Moral Significance of Class. Cambridge: Cambridge University Press.Schmid M (2011) The logic of mechanismic explanations in the social sciences. In: DemeulenaereP (ed.)Analytical Sociology and Social Mechanisms. Cambridge: Cambridge University Press,136153.

Sorg R (2009) Welches Wissenschaftsverstndnis braucht die Sozialarbeitswissenschaft? In:Birgmeier B and Mhrel E (eds)Die Sozialarbeitswissenschaft und ihre Theorie(n). Positionen,

Kontroversen, Perspektiven. Wiesbaden: VS Verlag, 2939.Staub-Bernasconi S (2000) Machtblindheit und Machtvollkommenheit Luhmannscher Theorie.

In: Merten R (ed.) Systemtheorien Sozialer Arbeit. Opladen: Leske und Budrich, 225242.Staub-Bernasconi S (2005) Fragen, mgliche Antworten und Entscheidungen im Hinblick auf die

Konzeption oder Konstruktion von (system)theoretischen Anstzen. In: Hollstein-BrinkmannH and Staub-Bernasconi S (eds) Systemtheorien im Vergleich. Wiesbaden: VS Verlag,269298.

Staub-Bernasconi S (2007) Soziale Arbeit als Handlungswissenschaft. Stuttgart: UTB.Viskovatoff A (1999) Foundations of Luhmanns theory of social systems. Philosophy of the

Social Sciences 29(4): 481516.Vollmer G (1995) Der wissenschaftstheoretische Status der Evolutionstheorie. Einwnde und

Gegenargumente. In:Biophilosophie. Stuttgart: Reclam, 92106.Von Glasersfeld E (1989) Constructivism in education. In: Husn T and Postlethwaite TN (eds)

The International Encyclopedia of Education, Supplementary Vol. 1. Oxford: Pergamon Press,

162163.Von Glasersfeld E (1991) Abschied von der Objektivitt. In: Watzlawick P and Krieg P (eds.)DasAuge des Betrachters. Mnchen: Piper, 1730.

Wan PY-Z (2011) Reframing the Social: Emergentist Systemism and Social Theory. Aldershot:Ashgate.

Wendt A (1999) Social Theory of International Politics. Cambridge: Cambridge UniversityPress.

Wight C (2004) Limited incorporation or sleeping with the enemy. In: Joseph J and Roberts JM(eds)Realism, Discourse and Deconstruction. London: Routledge, 201216.

Wight C (2006)Agents, Structures and International Relations. Cambridge: Cambridge University

Press.Williams M and Dyer W (2004) Realism and probability. In: Carter B and New C (eds) MakingRealism Work. London: Routledge, 6785.

Worrall J (2000) Pragmatic factors in theory acceptance. In: Newton-Smith WH (ed.)A Companionto the Philosophy of Science. Malden, MA: Blackwell, 349357.



21/22

Wan 715

Poe Yu-ze Wan is Assistant Professor at the Department of Sociology, National Sun Yat-senUniversity, Taiwan. Current research interests include sociological theory, philosophy of socialscience (especially systems theory, the question of emergence and the project of analytical sociol-ogy) and political sociology (especially the democratic innovations around the world). He is the

author ofReframing the Social: Emergentist Systemism and Social Theory (Ashgate, 2011), whichsystematically examines the implications of Mario Bunges philosophical system for socialresearch. He has also translated a number of major works into traditional Chinese, including,among others, Slavoj ieks The Ticklish Subject: The Absent Centre of Political Ontology, AlexCallinicosMaking History: Agency, Structure, and Change in Social Theory, Richard SennettsThe Fall of Public Man and Andrew Sayers The Moral Significance of Class.

Rsum

(Re-)Problmatiser lapproche constructiviste des systmes de Luhmann:

une intervention de BungeDans le champ de la thorie sociologique, aprs le dclin du onctionnalisme structurel

selon Parsons, on a considr la thorie des systmes de Niklas Luhmann comme la plusreprsentative de cette pense des systmes. Mme si lontologie des systmes

autopotiques de Luhmann est sophistique et mrite dtre prise au srieux, lauteuravance que identifer au pralable lapproche constructiviste de Luhmann avec lapprochesystmique elle-mme est problmatique pour deux raisons. Dabord, comme lont

montr les travaux du thoricien des systmes argentino-canadien Mario Bunge,lapproche systmique peut tre pertinemment onde sur (une version) du ralisme

scientifque. Ensuite les partisans de lpistmologie constructiviste de Luhmann en soci-ologie sengagent rarement dans un examen dtaill des diverses critiques delantiralisme ournies par les ralistes scientifques critiques. En sappuyant sur la

recherche en philosophie des sciences, en particulier les crits de Bunge et de quelquesralistes critiques minents, lauteur suggre quun sociologue srieux partisan deLuhmann doit apporter des arguments plus convaincants et plus dtaills en philosophie

des sciences (sociales) avant dadopter un avis trs arrt en pistmologie.

Mots-cls

Mario Bunge, constructivisme opratoire, pistmologie, Niklas Luhmann, ralisme cri-tique, ralisme scientiique

Resumen

(Re-) Problematizando el abordaje Luhmanniano de los sistemas construc-

tivistas: Una intervencin Bungeana

En el campo de la teora sociolgica, tras el declive de la versin de Parsons del uncion-alismo estructural, la teora de los sistemas de Niklas Luhmann comenz a ser vista

como la principal representante de sistemas de pensamiento. Aunque la ontologa de lossistemas autopoiticos de Luhmann es soisticada y merece seria consideracin, el autor

sostiene que la identiicacin prevaleciente del abordaje de orientacin constructivistade Luhmann con el abordaje de los sistemas, es en s misma problemtica sobretodo por



22/22


dos razones. Primero, como demuestran los trabajos del teorista de sistemas argentino-

canadiense Mario Bunge, el abordaje de los sistemas puede estar ructeramente basadoen (alguna versin de) realismo cientico. Segundo, los adherentes a la epistemologaconstructivista de Luhmann en sociologa raramente se implican en un detallado exa-

men de las varias crticas de antirealismo orecidas por realistas cienticos y crticos.Basndonos en la literatura sobre ilosoa de la ciencia, especialmente en los escritos de

Bunge y unos pocos destacados realistas crticos, el autor sugiere que un socilogoLuhmanniano serio es obligado a proporcionar ms contundentes y detalladas discu-siones sobre la ilosoa de la ciencia (social) antes de adoptar una posicin dogmtica

en epistemologa.

Palabras clave

Mario Bunge, constructivismo operativo, epistemologa, Niklas Luhmann, realismocientfco, realismo crtico

Documents

Problematizing the Luhmannian Constructivist Systems Approach