FROM BLUE SKY RESEARCH TO PROBLEM SOLVING: A … sky research.pdf · scientific knowledge, which we define as new theory that articulates or has the potential to articulate new phenomena

FROM BLUE SKY RESEARCH TO PROBLEMSOLVING: A PHILOSOPHY OF SCIENCE

THEORY OF NEW KNOWLEDGE PRODUCTION

MARTIN KILDUFFUniversity of Cambridge

AJAY MEHRAUniversity of Kentucky

MARY B. DUNNUniversity of Texas at Austin

We examine rationalized logics developed within discourses of the philosophy ofscience for implications for the organization of new knowledge. These logics, derivedfrom a range of philosophies (structural realism, instrumentalism, problem solving,foundationalism, critical realism) offer alternative vocabularies of motive, frame-works for reasoning, and guidelines for practice. We discuss the kinds of knowledgeproduced, the indicators of progress, the characteristic methods, exemplar organiza-tions, and ways in which logics are combined and diffused.

The institution of science is one of the endur-ing contributors to the modern world, providingorganized and established procedures for theaccomplishment of scientific work. Scientificknowledge is organized knowledge in the sensethat its production takes place within andacross formal organizational boundaries. Be-cause of the importance of this type of knowl-edge, there have been many efforts to under-stand its production (see Hessels & van Lente,2008, for a recent review). But one set of dis-courses has been neglected by organizationalscholars—those discourses developed withinthe philosophy of science in answer to the ques-tion “What is science?” (Bortolotti, 2008; Chal-mers, 1999). Within the philosophy of science arange of depictions of scientific activity purportto capture the rational process of scientific dis-

covery. These depictions represent alternativelogics of action that not only describe in ideal-ized terms actual historical examples of famousscientific breakthroughs but also prescribe theway scientific activity should be conducted soas to separate true science from pseudoscience(cf. Lakatos, 1970).

Institutionalized logics of action (defined asorganizing principles that shape ways of view-ing the world; Suddaby & Greenwood, 2005: 38)play a fundamental role in providing social ac-tors with vocabularies of motive, frameworks forreasoning, and guidelines for practice. Theselogics constrain cognition and behavior but alsoprovide sources of agency and change (Fried-land & Alford, 1991; Rao, Monin, & Durand, 2003:795; Thornton & Ocasio, 2008: 101).The purpose ofour article is to examine a range of rationalizedlogics developed within the discourses of thephilosophy of science—logics that profoundlyaffect the research of professional scientists(Laudan, 1977: 59) through their methods anddaily activities (Eddington, 1939: vii), as well asthrough the explanations that scientists “feelcompelled” to offer in justification for their prac-tices (Fuller 2003: 93). The justification of scien-tific activity is increasingly important in themodern world (Hilgartner, 2000), in which theboundaries between science and nonsciencehave become eroded (Ziman, 1996) and in which

For valuable comments and insights on prior versions ofthis paper, we thank discussants and audience members atpresentations at the Academy of Management meeting, theCambridge Realist Workshop, and Oxford University (all in2008); the OTREG meeting at Imperial College, the Univer-sity of Nottingham, and Tilburg University (all in 2009); theDanish Research Unit for Industrial Dynamics (DRUID) sum-mer conference and the Cambridge Judge Business Schoolorganizational behavior reading group (both in 2010.) Thepaper also benefited enormously from the guidance of guesteditor Roy Suddaby and the detailed comments of two anon-ymous reviewers.

� Academy of Management Review2011, Vol. 36, No. 2, 297–317.

297Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyrightholder’s express written permission. Users may print, download, or email articles for individual use only.

there is an insatiable demand for new scientificknowledge (Gibbons et al., 1994; Nowotny, Scott,& Gibbons, 2001: 249).

Logics of action are encoded in the routines oftraining, monitoring, disciplining, and reward-ing professionals (e.g., Friedson, 1970, 2001;Greenwood, Suddaby, & Hinings, 2002). Found-ers bring to their new ventures logics of actionthat continue to influence the structure andpractice of work as the firms grow (Baron, Han-nan, & Burton, 1999). The socialization of newmembers into existing roles (Van Maanen & Bar-ley, 1984; Zucker, 1977) through apprenticeshipsensures the survival of scientific disciplines(Van Maanen & Schein, 1979: 211). Logics of ac-tion are not only routinized in laboratory prac-tice but also provide the basis for rhetorical con-flict in organizations (Suddaby & Greenwood,2005) and can lead to variations in practiceswithin organizations and industries (Lounsbury,2007). Relevant logics of action offer to culturallycompetent actors legitimated discourses for theextraction of organizational resources, particu-larly in fields exhibiting pluralism and change(Dunn & Jones, 2010; Hardy & Maguire, 2008)where basic questions, such as those concern-ing what constitutes a scientific contribution,are often unclear (Overbye, 2002: 7).

It speaks to the legitimated power of dis-courses within the philosophy of science thatthese discourses are now routinely invoked inthe public sphere in debates concerning sciencepolicy (e.g., the debate over global warming;Maxwell, 2010), in business practice (Taleb,2007), and in legal disputes between organiza-tional factions. Thus, in the celebrated Pennsyl-vania case in which a school district tried toassert that intelligent design could be taught asan alternative to evolution, philosophers of sci-ence appeared for both the plaintiffs and thedefendants (Chapman, 2007).

To clarify the discussion, we focus on newscientific knowledge, which we define as newtheory that articulates or has the potential toarticulate new phenomena (Lakatos, 1970). Weinclude within the term theory a variety of forms,including abductive theory (i.e., theoryprompted by surprising observations; Hanson,1958) and theoretical models that posit causalrelationships among terms (cf. Suppe, 2000).From whence does this new theory derive? Wetake the view that new knowledge is stronglyconditioned by logics of action that incorporate

mutual assumptions and orientations. Logics ofaction are expressed, renewed, and changed insocial routines and networks characteristic ofknowledge communities (cf. Giddens, 1984). AsKarl Popper argued, “We approach everythingin the light of a preconceived theory” (Popper,1970: 52). Preexisting assumptions and orienta-tions that are embodied in logics of action arelikely to represent tacit, taken-for-granted back-grounds against which institutional entrepre-neurs provide rational explanations of their ac-tivities (cf. Misangyi, Weaver, & Elms, 2008).

Paraphrasing the Thomas theorem (Thomas &Thomas, 1928: 571–572), we can assert that ifscientific knowledge producers see the worldthrough distinctive ontological and epistemo-logical lenses, this way of seeing will have realconsequences in terms of the organization ofknowledge production. We first derive from thephilosophy of science four characteristic logicsthat represent organizing solutions to the prob-lem of knowledge production (see Figure 1).These representative approaches offer distinc-tive bundles of assumptions and practices and,we suggest, may have different implications forformal and informal organizing. In building newtheory, we formulate empirical predictions con-cerning how philosophies of science as logics ofaction are likely to affect organizing processesand outcomes. These empirical speculationsrepresent opportunities for research rather thanestablished verities.

We suggest, for example, that organizationsthat produce new knowledge may feature notjust one but several or all of the different typesof organizing frameworks discussed. A cluster ofpeople gathered together in a department or alaboratory is likely to share a particular logic ofaction that may be different from logics of actionoperating in other parts of the organization. Ourempirical predictions include comparisons oflogics of action with respect to the problems thatare likely to be pursued, the indicators of prog-ress that are likely to be used, the characteristicmethods each perspective might encourage, andthe kinds of organizations that are likely to exem-plify each perspective (see Table 1).

LOGICS OF SCIENTIFICKNOWLEDGE PRODUCTION

Different positions in the philosophy of sci-ence can be organized according to how they

298 AprilAcademy of Management Review

deal with basic questions of meaning (i.e., on-tology) and knowledge (i.e., epistemology). On-tology concerns the analysis of the types ofthings or relations that can exist. In science, amajor ontological issue concerns whether scien-tific theories represent reality—objects, events,and processes outside the human mind— orwhether scientific theories comprise explana-

tory fictions whose terms (such as electron) areconveniences invented to guide research. Epis-temology concerns how one gains access toknowledge and the relationship between knowl-edge and truth. In science, a major epistemolog-ical issue concerns whether or not scientific the-ories over time move closer and closer to thetruth. The ontological question is “Do scientific

FIGURE 1Matrix of Philosophy of Science Approaches and Associated Logics of Action

Epistemology

Science gets closer and closer to the truth? Yes No

Realist organizing

Pure research logic (e.g., Xerox PARC)

Strong-paradigm organizing

Exploitation logic (e.g., Apple Inc.)

Foundationalist organizing

Induction logic (e.g., Synta Inc.)

Instrumentalist organizing

Problem-solving logic (e.g., Linux)

Yes

No

Ontology

Scientific theories represent reality?

TABLE 1Implications of Philosophies of Science for Organizing

Key Questions Structural Realist Foundationalist Instrumentalist Strong Paradigm Critical Realist

Characteristic goaland logic ofaction?

Discover fundamentalstructure of theuniverse throughpure research

Find hidden patternsin data throughinduction

Truth-independentproblem solving

Create scientificparadigm and exploitits implications

Emancipate peoplefrom prevailingstructures ofpower andoppression

Example of type ofknowledgeproduced?

Scientificbreakthroughs,irrespective ofcommercialimplications

Serendipitousdiscovery ofpatterns in datafrom which newtheory can beformulated

Pragmaticsolutions totheoreticallydefinedproblems

Knowledge and productsconsistent with theoverarching culture ofparadigmaticcommunity

Exposes of powerfulactors’ policiesand actions

Indicators ofprogress?

Causal expression ofrelationshipsamong theoreticalterms; verificationof causal relationsamong terms

Unexpected butreplicablecorrelationsindicative of newdiscoveries;counterintuitivederivations fromfirst principles

Greater number ofimportantproblemssolved

Use of paradigm-definedfacts to solve puzzles;articulation of theparadigm throughempirical work

Challengeprevailing powerstructures, andreimaginepossiblemeaningsattached tocurrent practices

Characteristicmethod?

Mathematical modelbuilding

Data mining Those that areconsideredhistorically andsociallylegitimate

Defined bymethodologicalexemplars within theparadigm

Anthropology ofeveryday life

Illustrativeorganizing?

Self-governingcommunity

Cadre of experts Cross-field,focusedcollaboration

Fortress-likeorganization

Subversive team

Organizationalexamples?

Large HadronCollider; XeroxParc

SyntaPharmaceuticalsCorp.; Google Inc.

Team working tocap Gulf oilspill; Linux

Digital EquipmentCompany; AppleComputer

Greenpeace

2011 299Kilduff, Mehra, and Dunn

theories represent reality?” The epistemologicalquestion is “Does science get closer and closerto the truth?” Ontological and epistemologicaldimensions are represented in Figure 1 in orderto highlight some of the major differences be-tween philosophy of science perspectives.

Realist Perspectives

There are many varieties of realism (Putnam,1987), but, in general, they agree that scientifictheories aim to provide true descriptions of theworld (Okasha, 2002: 59), including the worldthat lies beyond observable appearances (Chal-mers, 1999: 226). Some versions of realism assertthat theoretical terms themselves have “puta-tive factual reference” (Psillos, 1999: 11)—thatterms such as utility function refer to real enti-ties. Realist perspectives agree that scientifictheories replace each other by offering betteraccounts of scientific objects (Putnam, 1975) sothat, over time, science gets closer and closer tothe truth about the world. Because of the affir-mative answers to questions concerningwhether scientific theories represent reality andwhether science gets closer to the truth overtime, realist perspectives are placed in the topleft-hand corner of Figure 1.

Realist perspectives agree, therefore, thatthere is a real world independent of our socialconstructions, that it is possible to assess scien-tific progress toward the truth about this world,and that competing scientific theories can beevaluated rationally in terms of how well theyexplain significant phenomena about thisworld. Further, realist perspectives focus on en-during relations between things, typically in theform of mathematical equations.

Structural realism. Structural realism repre-sents a major breakthrough in terms of a logic ofaction that reconciles two seemingly intractablyopposed arguments that have bedeviled argu-ments for the justification of science (Worrall,1989). On the one hand, the no miracles argu-ment posits that it would be a miracle—“a coin-cidence on a near cosmic scale” (Worrall, 1989:100)—if a theory made many correct empiricalpredictions without being basically correct con-cerning the fundamental structure of the world.This view was put forward originally by Poin-care (1905) but has been advocated in variousforms by many realists (e.g., Popper, 1963; Psil-los, 1999; Putnam, 1975). Opposing this view is

the pessimistic metainduction argument thatthe history of science is a graveyard of onceempirically successful theories (a perspectivealso anticipated by Poincare [1905], as Worrall[1989] points out). If past scientific theories thatwere successful were found to be false, we haveno reason to believe that our currently success-ful theories are approximately true (Laudan,1981).

The reconciliation of these opposing argu-ments involves the claim that as theories in ma-ture sciences change, there is a retention ofstructural content from one theory to the next.For example, the shift from the ether theory oflight to the electromagnetic theory of light in-volved the retention of the mathematical struc-ture expressed in a series of equations such thatat the structural level there is complete continu-ity between the theories (Worrall, 1989). In othercases (such as the transition from Newton’s lawsto those of Einstein), mathematical equationsare retained “as fully determined limiting casesof other equations, in the passing from an oldtheory to a new one” (Psillos, 1995: 18). Structuralrealism, therefore, avoids the claim that theoriescorrectly describe the empirical reality of theworld (defusing the pessimism of the antireal-ists) but accepts that successful theories are ap-proximately true descriptions of the underlyingstructure of the world (accounting for the mirac-ulous success of science).

Structural realist organizing: The logic of pureresearch. In order to gain resources and to intro-duce change into otherwise stable social sys-tems, institutional entrepreneurs “must locatetheir ideas within the set of existing under-standings and actions that constitute the insti-tutional environment” (Hargadon & Douglas,2001: 476). The accepted justification for “bluesky” research has typically been couched interms of a structural realist logic of action. Theso-called linear model or fable justifies blue skyresearch in terms of the necessity that pure re-search scientists delve into the secrets of naturein the absence of tight controls or specific tar-gets so that potential practical applications canbe developed by others in the unspecified futurefor use by consumers (Grandin, Wormbs, & Wid-malm, 2005). Thus, the knowledge workers whoengage in pure research are likely to avoid thetendency to tie their mission to the developmentof specific inventions. Pure researchers arelikely to retain a deep underlying belief in the


coherence of their theoretical frameworks. Newknowledge will tend to be seen as a long-termproject driven by acceptance of causal relationsamong theoretical terms. The structure of theoryfrom a structural realist perspective remains rel-atively unchanging, and it is this very stabilitythat can provide the basis for investing time andresources in innovation (Stein, 1989: 57).

Thus, from a structural realist perspective, itis justifiable to organize massive projects aimedat comprehending the structure of the universe.Projects that seek answers to fundamental ques-tions proceed from the assumption that the pur-pose of science is to map the deep structure ofreality, a reality that is typically assumed to beexpressible in mathematical form (Ladyman,1998) or in terms of theoretical models (Suppe,2000). This unifying assumption facilitates theself-organization of scientists around massivepure research projects so that hierarchical con-trol is often noticeably absent (Knorr-Cetina,1999).

The kinds of questions that are likely to bepursued, therefore, from the perspective of struc-tural realism include, most basically, improve-ments or modifications to fundamental laws(Psillos, 1995) or theoretical models and attemptsto establish evidence to support inferences fromsuch laws. From a structural realist perspective,different ontologies may satisfy the same math-ematical or formal structure, and there is noindependent reason to believe that one of theseontologies is better than another (Psillos, 1995:20). But it is important to remember that thestructural realist believes that theories informus about the structure of the world rather thanabout fictional entities: “realism should involvereference to what ’really’ exists” (French & Lady-man, 2003: 38). Thus, progress from a structuralrealist perspective involves improvements toour knowledge concerning the structure of real-ity and the causal relations among entities, al-though structural realism does not necessarilyentail improvements to knowledge concerningthe objects and properties the world is made of(Ladyman, 1998: 422).

Advances in knowledge, according to thelogic of structural realism, are likely to beachieved by academic researchers who work foruniversities or research institutions (that may befunded by private companies). These advancesare likely to be published in scientific journalsdevoted to pure research and, in some cases,

registered as patents. Pure research advanceswill tend to be taken up by other scholars (asmeasured by citation counts) and by inventorsand others seeking to translate academic re-search into viable products.

Structural realist organizational examples.We suggest, therefore, that exemplars of a struc-tural realist approach to new knowledge pro-duction will tend to be pure-science organiza-tions, such as the Large Hadron Collider (LAC),which employs 2,250 physicists near Geneva,Switzerland, and involves a further 7,750 physi-cists in research collaborations. Research proj-ects can span decades, with the ultimate goal ofunderstanding the fundamental nature of real-ity. The LAC is run on a communal basis, involv-ing laborious negotiations among competinggroups and an arrangement in which all re-search is coauthored by the thousands of phys-icists involved (Merali, 2010).

In the realm of high-tech companies, a famousexample of devotion to relatively pure researchwas Xerox PARC, set up by the Xerox Corpora-tion in a building at the edge of Stanford Uni-versity in 1970 (hence “PARC”—Palo Alto Re-search Center). The research center hired someof the world’s best physicists, mathematicians,materials scientists, computer system archi-tects, and software engineers to pursue funda-mental discoveries in the “architecture of infor-mation” (Chesbrough, 2002: 807). Given millionsof dollars to pursue fundamental research, withthe understanding that material benefits to Xe-rox Corporation would not show up for at least adecade, these PARC researchers produced rev-olutionary discoveries (largely taken up by com-panies other than Xerox), including the personalcomputer, a graphical user interface, a laserprinter, and technology that would later becomeindispensable for the spread of the Internet.

Instrumentalist Perspectives

As its placement in the bottom right-handcorner of Figure 1 indicates, instrumentalism isantirealist in asserting that scientific theoriesare useful instruments in helping predict eventsand solve problems. As one contemporary phi-losopher of science explained this perspective,“Fundamental equations do not govern objectsin reality; they only govern objects in models”(Cartwright, 1983: 129). A variety of different la-bels (instrumentalism, constructive empiricism,


theoretical skepticism, and the philosophy of“as if”; see Horwich, 1991) have been given to theview that one is obliged to believe nothing be-yond the observable consequences of a success-ful scientific theory—“there can be no reason . . .to give the slightest credence to any of its claimsabout the hidden, underlying reality” (Horwich,1991: 1). From this perspective, theories shouldbe judged according to how well they help or-ganize phenomena, facilitate empirical predic-tion, or solve problems in the world (cf. Laudan,1977, 1990), not according to how well they depict“actual” causal processes. Closely related topragmatism (Sleeper, 1986: 3), instrumentalismtreats knowledge as something to be sought notfor its own sake but for the sake of action tosolve problems.

Within the social sciences, this tradition isrepresented by the neoclassical economics viewthat theory serves “as a filing system for orga-nizing empirical material” (Friedman, 1953: 7)and should be judged “by its predictive power”(Friedman, 1953: 8). Important theory tends toprovide “wildly inaccurate descriptive represen-tations of reality, and, in general, the more sig-nificant the theory, the more unrealistic the as-sumptions” (Friedman, 1953: 14). Thus, from thisinstrumentalist perspective, one theory suc-ceeds another not because it moves closer to thetruth but because it represents a more usefulpredictive framework for the phenomena of in-terest. We focus here on the problem-solvingapproach of Larry Laudan, which connects theworld of scientific theory to the solution of prob-lems in the world. Laudan’s philosophy of sci-ence is instrumentalist in the sense defined byJohn Dewey (1903), who established the require-ment that theories be reliable and useful tools inpractical endeavors, such as helping scientistsmanipulate objects and predict outcomes.

Problem solving. According to Laudan (1977),the question of whether a theory is true or falseis irrelevant in determining its scientific accept-ability. What is relevant is whether a theorysuccessfully solves problems (Laudan, 1977: 18).Further, Laudan rejects the view that the historyof science represents a march toward truthabout the world. In his view scientific progressconsists of accepting those research traditionsthat are the most effective in terms of problemsolving (Laudan, 1977: 131). Thus, Laudan’s prob-lem-solving approach is representative of thebottom right-hand corner of Figure 1, being anti-

realist in terms of ontology and epistemicallyinstrumentalist in terms of the progress of sci-ence.

For Laudan (1977), science consists of compet-ing research traditions that differ from the par-adigms discussed by Kuhn (1996/1962) and theresearch programs discussed by Lakatos (1970),in that all the assumptions of a research tradi-tion can change over time (as the research tra-dition tackles new and important problems).Further, a research tradition can spawn rivaland potentially incompatible theories that com-pete with each other and with theories producedby other research traditions in the solution ofproblems. This point of view separates rationalprogress from any question concerning the ve-racity of theories, because progress consists ofincreases in problem solving rather than greaterverisimilitude.

From the perspective of the social organiza-tion of knowledge production, the problem-solving approach of Laudan (1977) recognizesmore clearly than rival approaches the prag-matic nature of scientific progress. Progress in-volves producing more reliable knowledgerather than knowledge that takes us closer tothe truth about the universe (Laudan, 1990: 14).From Laudan’s perspective, a scientist can par-ticipate in two different research traditions orcan synthesize a new research tradition fromcompeting alternatives, and theory born in oneresearch tradition can be separated and movedor taken over by an alternative research tradi-tion that offers more problem-solving capability.Scientists are depicted as pragmatic ratio-nalists who, even as they “accept” theories onthe basis of past success in problem solving, arelikely to “pursue” quite different theories (onesthat may even seem wildly improbable) if thesetheories are seen as offering higher rates ofproblem-solving progress.

Laudan’s approach suggests that “a highlysuccessful research tradition will lead to theabandonment of that worldview which is incom-patible with it, and to the elaboration of a newworldview compatible with the research tradi-tion” (Laudan, 1977: 101). Thus, what is consid-ered the truth is likely to change to accommo-date successful theory. Scientific theories thatare unable to counter the claims of prevailingworld views (even if these world views are putforward in nonscientific domains, such as reli-gion) are unlikely to be effective. Science, in


Laudan’s view, is a fluid, flexible, and changingendeavor, in which the successful scientist isable to juggle alternative theories and enterimaginatively into different research traditions,all in the service of problem-solving activity thatis at the core of scientific work.

Instrumentalist organizing: The logic of prob-lem solving. The problem-solving approach cap-tures the freedom to play around with differenttheories and different traditions of scientificknowledge production in a way that rival phi-losophies of science neglect. The overriding pre-scription of Laudan’s approach is to try and dis-cover the theory that has the highest likelihoodof solving a particular problem, and this mayinvolve working with research traditions thatare mutually inconsistent (Laudan, 1977: 110).The structure of DNA was discovered when sci-entists played with molecular models that re-sembled “the toys of preschool children” (Wat-son, 1968: 38). From this perspective, scientistshave a license to adopt and discard theories andmethods to the extent that they are useful (cf.Feyerabend, 1975) and socially legitimate, with-out any requirement that the scientists committhemselves paradigmatically (cf. Kuhn, 1996/1962) or that they restrict themselves to a set ofunchanging core ideas (cf. Lakatos, 1970).

There is, therefore, an inherent pragmatism inLaudan’s approach (Godfrey-Smith, 2003). Al-though problem-centered organizing does re-quire a certain ideological commitment to what-ever theory happens to be guiding empiricalinquiry, this commitment is minimal in thesense that theory acceptance does not involvethe necessity of believing that the theory is trueor that metaphysical unobservables are real.

We suggest, therefore, that the production ofnew knowledge from this perspective will in-volve scientists getting on with the pragmaticbusiness of investigating the empirical regular-ities in nature without having to believe as truethe grand metaphysical claims embedded intheories. Problem-oriented scientists faced withconflicting theoretical approaches are likely tocompromise in order to “save the phenomena”(Duhem, 1969/1908), in the sense of providing sat-isfactory solutions to important problems (Lau-dan, 1977: 13), irrespective of whether theoreticalpurity is endangered. Because of the problem-solving focus of this logic of action, scientificresearch from this perspective is likely to beamenable to fortuitous spin-offs from attempts

to solve deep intellectual problems (cf. Laudan,1977: 224). There is a greater likelihood that thislogic of action will feature collaborations be-tween university professors and more practi-cally oriented researchers and designers. Forexample, in the problem-oriented design firmIDEO, which produces innovative products forforty industries, the CEO is a professor at Stan-ford, and ten other designers teach product de-sign at the university (Hargadon & Sutton, 1997).

Instrumentalist organizational examples. In-deed, organizations that exemplify a problem-oriented approach to new knowledge productionare sometimes created in response to pressingproblems. Consider, for example, the hybrid or-ganization that was assembled to devise solu-tions to the flow of oil pouring into the Gulf ofMexico following the Deepwater Horizon dril-ling rig explosion on April 20, 2010. This crisisteam included physicists, experts on Mars dril-ling techniques, an expert on hydrogen bombs,and an MIT professor who referenced “goingfaster on my snowboard” among his researchinterests. Literally “anyone . . . who could makea difference was brought in” (a senior BP man-ager, quoted in Tankersley, 2010). Scientific or-ganizations that worked on this cleanup havebegun to contribute new theoretical knowledge(e.g., Camilli et al., 2010).

In the realm of high-tech companies, a prob-lem-solving logic of action is, we suggest, exem-plified in open-source software companies, suchas Apache, Mozilla, and Linux. These companiesoperate on the principle that source code isfreely available to anyone who wishes to ex-tend, modify, or improve it. In the example ofLinux, which has developed an operating sys-tem for computers, the company centers on thefounder Linus Torvalds and 121 “maintainers”who are responsible for Linux modules. Thereare also thousands of user-developers who findbugs and write new pieces of problem-solvingcode. The success of the company has been ex-plained as deriving from the “quantity and het-erogeneity of the programmers and users in-volved in development” (Miettinen, 2006: 177), aprinciple that has been dubbed “Linus’s law”(Raymond, 1999: 41). The variety of different codedevelopers and improvers means that problemsolving is approached in many different ways,with each problem solver using “a slightly dif-ferent perceptual set and analytical toolkit, adifferent angle to the problem” (Raymond, 1999:


43). The open-source software movement haschanged our understanding of the sources ofinnovation in organizations, providing a basisfor new theory development concerning distrib-uted innovation (von Hippel, 2005).

Foundationalist Perspective

Occupying the bottom left-hand corner of Fig-ure 1, foundationalism indicates a combinationof antirealism and the belief that science pro-gresses toward truth. Foundationalist anti-realism was promulgated by logical empiricistswho were influenced by Ernst Mach (1838–1898).Mach “strongly believed that science shoulddeal only in observable phenomena” (Ray,2000a: 104), claimed “that only the objects ofsense experience have any role in science” (Ray,2000b: 245), and conceived of science as re-stricted to the “description of facts” (Wolters,2000: 253). Rudolf Carnap (1891–1970), the mostinfluential logical empiricist in the Mach tradi-tion, attempted to construct all domains of sci-entific knowledge on the basis of individual ex-perience (Carnap, 1928), a perspective that isantirealist in omitting from the realm of exis-tence theoretical unobservables, such as quarks(Creath, 1985: 318). (In contrast, some logical em-piricists, such as Hans Reichenbach [1938],moved toward realism by adopting the beliefthat the physical sciences possess the ontologi-cal authority to tell us which entities, properties,and relations can be said to exist [Crane & Mel-lor, 1990]). Logical-empiricist antirealist founda-tionalism emphasizes empirical data gatheringfrom which scientific knowledge emerges induc-tively (Chalmers, 1999) so that there is a rationalbasis for evaluating new knowledge claims.Theories with greater empirical content aredeemed better than theories with less empiricalcontent.

Foundationalism was one of the most widelydebated conceptions of knowledge productionprior to the revolutionary ideas of Kuhn (1996/1962) and is often called the “received view”(Putnam, 1962; Suppe, 1972). Generally, a foun-dationalist believes there is an ultimate basis ineither empirical data or logical process bywhich knowledge claims can be validated (cf.Ayer, 1952; Kleindorfer, O’Neill, & Ganeshan,1998: 1090). This view blends aspects of logicalpositivism (see Uebel, 1996) and logical empiri-cism (see McKelvey, 2002); indeed, “empiricist

philosophies have often had a foundationaliststructure” (Godfrey-Smith, 2003: 220). More re-cently, foundationalist philosophy of sciencehas resurfaced under the rubric of “reliabilism,”according to which beliefs are justified whenthey are produced by cognitive processes thatare highly reliable (Goldman, 2009). Reliabilismis compatible with antirealism (Beebe, 2007).Traditional received-view foundationalism typ-ically represents a starting point for debate con-cerning the organization of new knowledgerather than the final word (cf. Kleindorfer et al.,1998).

Foundationalist organizing: The logic of in-duction. In terms of the relevance of antirealistfoundationalism for new knowledge productionin the current era, the emphasis on induction,from which scientific knowledge and theoriesemerge, implies collecting lots of data, whichcan be sifted to discover otherwise difficult-to-discern patterns. The prevalence of high-speedcomputers provides a new impetus for this par-ticular orientation. Computer programs can en-able the scanning of databases for correlationsor trends, without any realist presuppositionsconcerning causality or entity existence. Theemphasis within any particular domain is onextracting previously unknown knowledge fromfactual data using quantitative analyses.

In terms of a logic of organizing, foundation-alism would seem to require a small cadre ofexperts who are able to interpret correlationalpatterns in order either to create new theory orto match correlational patterns with existingtheory so that new knowledge can be extracted.There is a danger of authoritarianism in thisemphasis on the interpretation of patterns indata, as has been noted in the debate overevidence-based medicine, where the relevantquestions include who decides what is relevantevidence and who determines the best interpre-tation of this collected evidence (Shahar, 1997).Similarly, in scientific management the searchhas been for the “one best way,” with a relent-less pursuit of improvement through empiricalmeasurement, experiment, and statistical dis-play (e.g., the Gantt chart). New knowledge, fromthis perspective, could be extracted throughclose attention to work processes, rather thanfrom the imaginative promulgation of new the-ory. In the current era the reliance on expertscontinues, but, we suggest, a foundationalistlogic of action these days will tend to direct


supervision toward large data sets rather thanthe labor process. Thus, physicists proliferate onWall Street, bringing their expertise to bear interms of new algorithms to analyze and profitfrom trends in financial data (Bernstein, 2008).The consumers of foundationalist-based knowl-edge are likely to be in the front line of serviceproviders, such as practicing physicians, finan-cial managers, and other professionals.

In terms of the informal structuring of workfrom a foundationalist perspective, therefore,the work process is likely to be highly central-ized around the cadre of experts with specialisttraining who direct the search for empirical reg-ularities that can serve as the foundation for theproduction of new knowledge. From a founda-tionalist perspective, empirical facts remainfacts, even as the world changes and regardlessof whether the facts derive from one disciplinaryarea or another. Therefore, the cadre of special-ists may include people from quite different dis-ciplinary backgrounds and representing quitedifferent historical periods of data representa-tion. There may be a mixing of Ph.D.s in econom-ics and physics, combined together to search forpatterns in financial data. To the extent that thefocus is on finding patterns in data rather thanon pushing forward the boundaries of disci-plines, the common focus on an empirical foun-dation can provide the basis for cohesion.

This empirical approach can take advantageof knowledge collected over periods of time,which is then formalized within a standard setof parameters. The clearest contemporary exam-ple of this approach is data mining. The newknowledge discovered through data mining con-sists of patterns in data that can translate intopossible new products that take advantage ofhitherto unnoticed correlations. Although datamining typically takes advantage of computerautomation and algorithms to generate knowl-edge discovery, it depends on a series of judg-ments, including selecting the knowledge areato be searched, preparing the data set from oftenheterogeneous elements, creating a model thatcan guide the search process, choosing searchalgorithms, interpreting results and testingthem, and using resulting patterns as the basisfor better decision making or new product devel-opment (Goebel & Gruenwald, 1999). Thus, thereneeds to be a team of specialists guiding theautomated process.

Foundationalist organizational examples.This foundationalist approach to organizing forknowledge generation is, we suggest, exempli-fied by Synta Pharmaceuticals, a biopharma-ceutical company focused on the discovery, de-velopment, and commercialization of smallmolecule drugs to treat severe medical condi-tions. Reversing the standard practice in theindustry (which is to start with a theoretical un-derstanding of a disease and then rationallydesign a customized solution), Synta uses massscreening of chemical compounds in the ab-sence of any theory (Gladwell, 2010: 72). Thecompany purchases thousands of chemical com-pounds from around the world, most of whichwere never designed for medical use. It thentests these compounds in batches to see if theyaffect cancer cells. Most compounds have noeffect or prove toxic to all cells. But, once in awhile, a compound proves efficacious againstcancer cells. For example, a compound manu-factured at the National Taras Shevchenko Uni-versity, in Kiev, and purchased by Synta foraround ten dollars, proved effective againstprostate cancer cells. It was an unusual com-pound, “homemade, random, and clearly madefor no particular purpose” (Gladwell, 2010: 73).Had it not been for the atheoretical data miningapproach employed by Synta, this compound’sability to fight cancer cells might never havebeen discovered.

The foundationalist path to the generation ofnew knowledge followed by Synta uses a trial-and-error search for new knowledge that makesno a priori assumptions concerning causalityyet maximizes the possibility of serendipitousdiscovery. This example raises the question asto whether computing power has made possiblegenuinely theory-neutral exploration in a waythat philosophers for decades said could notoccur. This new computer-age foundationalistorganizing does not require guiding theory but,rather, seems to fulfill the empiricist dream ofbuilding science from a foundation of empiricaldata.

In the realm of high-tech companies, GoogleInc. is one example of a company that special-izes in automated data mining as a core princi-ple of its business. The two founders of the com-pany, while Ph.D. students at StanfordUniversity, developed a search engine thatranked websites according to the number of con-nections to other websites. This innovation fa-


cilitated data mining in the enormously com-plex World Wide Web by using informationconcerning which websites had been “voted” tobe the best sources of information by otherpages across the web. The company continuesto focus on data search through mathematicalprogramming in its development of a range ofproducts and services (Girard, 2009).

Strong-Paradigm Perspective

The other off-diagonal perspective in Figure 1derives from the work of Thomas Kuhn (1996/1962) that combines a belief in the actuality ofthe physical world with skepticism about theconvergent-realism claim (endorsed by thestructural realists) that science progresses to anever-closer approach to truth. We take at facevalue the assertion by Kuhn that his philosoph-ical position represents an “unregenerate” real-ism (1979: 415), in the sense that there is a realworld out there—“entirely solid: not in the leastrespectful of an observer’s wishes and desires”(Kuhn, 1990: 10). Kuhn’s position is nuanced bythe assertion that the members of a successfuldisciplinary scientific paradigm define forthemselves what aspects of reality to attend to,change, and adapt.

Paradigmatic community members share ed-ucation, language, experience, and culture andtherefore tend to “see things, process stimuli, inmuch the same ways” (Kuhn, 1996/1962: 193). Thedisciplinary matrix that successful scientistsshare represents reality for them because it se-lects certain objects for investigation and facil-itates the creation of a distinctive social world ofscientific endeavor. Kuhn’s realism does notcommit him to any strong sense that successivescientific theories approach closer to some par-adigm-independent truth (even though muchempirical content may be preserved when onetheoretical paradigm succeeds another; Kuhn,1996/1962: 169). One cannot step outside of his-tory to evaluate truth claims from a paradigm-free, objective perspective (Kuhn, 1996/1962).

Kuhn’s philosophical position is a complexone, but, for our present purposes, we interpretKuhn as affirming that paradigmatic theories doindeed represent reality, although we recognizethat Kuhn did not assume that successive theo-ries represented closer and closer approxima-tions to “what nature is really like” (Kuhn, 1996/1962: 206). As noted in one explanation of Kuhn’s

realism, “Representations arising from attemptsto answer different problems need not meshwell with each other—perhaps the world is toocomplicated for us to get one comprehensivetheory” (Hacking, 1981: 4).

Kuhn modified and clarified his ideas consid-erably over the years in response to critics’ in-terpretations and misunderstandings (seeWeaver & Gioia, 1994, for a careful discussion ofthese issues). Kuhn’s revisions have been de-scribed as putting forward “but a pale reflectionof the old, revolutionary Kuhn” (Musgrave, 1971:296). This revised Kuhn has even been describedas “a closet positivist” (Laudan, 1984: 68). Cer-tainly, it is the original ideas that generatedmuch of the discussion in the philosophy of sci-ence. In our interpretation of Kuhn, we take intoaccount his later emendations while agreeingwith Weaver and Gioia that “early works are notnecessarily invalidated by later ones” (1994:573).

Strong-paradigm organizing: The logic of ex-ploitation. Strong-paradigm organizing, in ourinterpretation, is characterized by a unifiedforce of energetic believers who share funda-mental assumptions about the nature of realityand the practice of research. These believers arelikely to be protective of their ideology, giventhat this ideology has been constructed with theutmost difficulty and constitutes the frameworkwithin which meaningful activity can be con-ducted. Paradigm believers exhibit strong resis-tance to ideological or cultural change.

Knowledge production from this perspectiveis, we suggest, likely to be characterized by arelentless focus on the exploitation of existingknowledge bases. Knowledge production con-sists of such activities as forcing data into exist-ing categories prescribed by the theoretical par-adigm and mopping up remaining corners ofunexploited knowledge—an activity tanta-mount to puzzle solving (cf. Kuhn, 1996/1962).There is likely to be a tendency to exclude com-peting views, given that such outside influencescan disturb the equanimity of paradigmaticpuzzle-solving activity.

Strong-paradigm organizing is likely to ne-glect or ignore persistent anomalies in order tofocus on ingenious technological advances andfixes that are compatible with an overall theo-retical approach. Such organizing will, we sug-gest, tend to feature closed boundaries so as toprotect proprietary knowledge through an em-


phasis on secrecy, patents, copyrights, and con-trols to prevent trade secrets from being stolenby rivals. Cohesive networks are one basis forcompetitive advantage (Coleman, 1990), but bro-kering across groups is likely to promote careers(Burt, 2004).

Strong-paradigm organizational examples.Organizations that exemplify a strong-para-digm approach to new knowledge productionwill, we suggest, tend to have strong, distinctivecultures and ideologies that powerfully shapethe production of knowledge. For example,within some parts of the computer company de-scribed by Kunda (1992), the generation of cut-ting-edge knowledge involved the formulationand dissemination of ideology, the use of grouptestimonials and face-to-face control reminis-cent of brainwashing techniques, and the inva-sion of private life by corporate requirements.We might also think of the now-defunct DigitalEquipment Corporation, where engineerstended to dismiss the possibility of learningfrom rivals or the marketplace and tended tocling to the internal, distinctive culture that con-tinued to shape their lives, even after they hadbeen dismissed from their jobs (Johnson, 1996).In the realm of current high-tech companies in-volved in distinctive innovation, Apple Com-puter exhibits strong-paradigm control overknowledge:

Secrecy is one of Apple’s signature products. . . .Workers on sensitive projects have to passthrough many layers of security. Once at theirdesks or benches, they are monitored by camerasand they must cover up devices with black cloaksand turn on red warning lights when they areuncovered (Appleyard, 2009).

The consumers of strong-paradigm knowledgeproduction are likely themselves to resemblecult members in their enthusiasm for distinc-tively branded products that permit their usersto differentiate themselves in terms of identityand style (Bhattacharya & Sen, 2003).

Strong-paradigm organizing is likely to placethe emphasis on a supportive community of like-minded scientists, engaged within a commonculture, striving to articulate a distinctive visionand solve a set of well-understood problems.Thus, as Table 1 summarizes, the creation of ascientific paradigm itself is an object to be at-tained because it represents a wide-ranging or-ganizing framework for knowledge production,one that picks out certain problems to be solved

while disregarding other problems. A paradigmcommunity shares a disciplinary matrix interms of procedures, exemplars, formulas, andbeliefs. Indicators of progress within such acommunity include success in the solution ofoutstanding puzzles identified by the paradig-matic community. There is likely to be a strongfocus, from this perspective, on new techniquesthat facilitate the articulation of the paradigm,in terms of empirical testing, and that enablesolutions to outstanding puzzles.

A Note on Critical Realism

When choosing characteristic perspectives topopulate the four quadrants of Figure 1, we leftout some important philosophical approachesthat overlap with these four perspectives. Butone contemporary philosophy of science ap-proach—critical realism—stands out as signifi-cantly different from those we have discussedbecause it focuses on the social world of humaninteraction rather than the physical world inves-tigated by physics, chemistry, and the hard sci-ences (Bhaskar, 1998). Because of this basic dif-ference, we treat this approach separately here.The relevance of critical realism for manage-ment scholars has been thoroughly discussedelsewhere (e.g., Reed, 2008; Tsang & Kwan, 1999),so this perspective does not need such an exten-sive introduction here.

Critical realism belongs with structural real-ism in the top left-hand quadrant of Figure 1. Butwhereas structural realism emphasizes the cap-ture of invariant relations in the form of mathe-matical equations, critical realism emphasizesthat the relationships among entities discoveredby social science are likely to be relatively en-during as opposed to completely invariant. Therelatively enduring structure of positions in aparticular culture, for example, would be con-sidered to have causal power over the attitudesand behaviors of those who temporarily occupysuch positions (Archer, 1998).

In focusing on the social world, critical real-ism emphasizes the stratification of reality. Therealm of the real consists of unobservable struc-tures and causal powers, the realm of the actualconsists of events and processes in the world,and the realm of the empirical consists of theexperiences of human beings (Fairclough, 2005).Thus, critical realism emphasizes that there areunderlying structures and forces that are unob-


servable but real in their operations and bestinvestigated through ethnographic and histori-cal research (rather than exclusively throughquantitative analyses of independent and de-pendent variables; Reed, 2008). Critical realists(Archer, Bhaskar, Collier, Lawson, & Norrie, 1998)argue that science, by correcting errors and re-jecting false starts, converges with the truth. Thecritical aspect of this philosophy of science re-lates to the possibility that an explanatory cri-tique of the ways in which structures of poweroperate in society can be emancipatory.

New knowledge proceeding from the perspec-tive of critical realism is likely to challenge ex-isting power structures in industry and govern-ment. In order to break the mold of currentthinking, it would be necessary, from this per-spective, to tackle the past inheritances put inplace by prior thinking that tended to shacklenew discovery. Consumers of such knowledgeare likely to be social activists, interested inradical changes to the status quo.

The logic of action associated with criticalrealism, therefore, is the logic of emancipation.Organizations that exemplify a critical realistapproach to new knowledge production will, wesuggest, tend to be social action organizations,such as Greenpeace, that fight in the publicsphere to seize rhetorical control over the inter-pretation of events (Tsoukas, 1999). Not contentto just produce new knowledge through its ownresearch laboratories at the University of Exeter,Greenpeace is active in direct campaignsagainst deforestation, overfishing, commercialwhaling, global warming, and nuclear power.Thus, this nongovernmental environmental or-ganization attempts to take scientific researchand use it to change the attitudes and behaviorsof people and to produce products that providegreen alternatives to standard technology.

COMBINING AND DIFFUSING THE LOGICS

Given the stark differences between theselogics on basic issues of epistemology and on-tology, the question arises as to which approachwe consider to be the best. In writing this article,for example, which philosophy of science are weimplicitly endorsing? As students of organiza-tions, we are persuaded by a garbage-can ap-proach (Cohen, March, & Olsen, 1972), accordingto which, in any particular knowledge-produc-ing effort, logics of action can be taken to repre-

sent different kinds of solutions available to dif-ferent types of agents endeavoring to tacklestreams of different types of problems. As wehave discussed in the article and summarized inFigure 1 and Table 1, each philosophical alter-native holds different implications for organiz-ing. In terms of our own efforts at knowledgeproduction, we engaged with a pure researcheffort (i.e., the philosophy of science) generallyconsidered to have little relationship to practi-cal endeavors, we applied this philosophicaldiscourse instrumentally to understand the pro-duction of scientific knowledge, and we tried todevelop a fairly tight paradigm organizedaround Figure 1 and Table 1 in order to exploitthe discourses of the philosophy of science. Wehave not tried to engage in data mining throughthe collection of lots of information concerningactual knowledge-producing efforts to see em-pirically what kinds of patterns might be re-vealed, but we recognize that this kind of exten-sive reviewing can yield valuable discoveries interms of underlying patterns (e.g., Van de Ven &Poole, 1995). In short, we anticipate that knowl-edge-producing efforts will generally feature avariety of logics of action in combination.

For example, Figure 2 outlines a hypotheticalknowledge-producing organization that inte-grates a mix of logics of action. Like other orga-nizations, those focused on knowledge produc-tion face technical and environmentaluncertainty. To manage this uncertainty, theknowledge-producing organization is likely tocreate certain parts intended specifically to dealwith uncertainty, thus allowing certain otherparts to carry on the core activities of the orga-nization under conditions of relative certainty(Thompson, 1967). At the technical core of theideal knowledge-producing organization, weenvisage a team of structural realists, scientistswho completely believe in the mission of discov-ering the truth about the universe, including theworld in which we live. For these scientists thereare few doubts about whether theories representreality or whether theories march forward to-ward greater and greater truth. For example, atBell Labs, during its illustrious history, the coreof its knowledge-producing efforts consisted ofthe fundamental physics research group thatwon seven Nobel prizes for discoveries includ-ing the wave nature of matter and the existenceof cosmic microwave background radiation (see


Gehani, 2003, for a description of recent eventsin this organization).

The work of these researchers in discoveringnew phenomena and solutions to theoreticalproblems would need, however, to be bufferedfrom environmental fluctuations by a protectivebelt (cf. Lakatos, 1970) populated by other knowl-edge workers. The job of the protective beltwould be to allow the technical core to operateunder conditions approximating certainty byhelping the organization to adjust to constraintsand contingencies not controlled by the organi-zation. The workers in the protective belt wouldseek to minimize the knowledge-producing or-ganization’s dependence by maintaining alter-natives, actively competing for support and re-sources through a variety of methods, includingalliance building, co-opting, lobbying, and theuse of rhetoric designed to enhance the legiti-macy of the organization and its products(Thompson, 1967). We envisage that this protec-tive belt would include instrumentalists, whowould pragmatically negotiate ways in whichpure, blue sky science could be translated intoproducts and services relevant for problemsarising in society (cf. Hilgartner, 2000). We alsoenvisage a group of foundationalists, activelyscanning data from the environment and match-ing patterns with ideas emerging from a coregroup of structural realist scientists. Further, in

this protective belt around the structural real-ists, we envisage a group of critical realists,able to assess the political structures possiblyobstructing the emergence of revolutionaryideas and able to give advice concerning whatkinds of products and services might be intro-duced most expediently in particular sectorsand how these products and services might becharacterized. The relative power of these differ-ent sets of experts and mediators in the protec-tive belt is likely to change and shift dependingupon the kinds of crises provoked by environ-mental trends (Hickson, Hinings, Lee, Schneck, &Pennings, 1971).

Skilled mediators are likely to be required soas to coordinate the flow of resources andknowledge between these various philosophi-cally disparate groups. Philosophical differ-ences concerning such ontological questions aswhich categories of things are justified can pro-vide arguments against cooperation andchange (Suddaby & Greenwood, 2005). If theknowledge-producing organization must at-tempt to create something like certainty for itstechnical core while also retaining flexibilityand adaptability to satisfy environmental con-straints, we might expect to see a dedicated setof broker-managers, charged with smoothingthe irregularities stemming from environmentalfluctuations while also pushing the technical

FIGURE 2Organizing Logics Combining in Action in Hypothetical Organization

Structural realist:Pure research

Instrumentalist: Problem solving

Foundationalist: Induction

Strong paradigm: Exploitation

Critical realist: Emancipation


core to make necessary modifications as exter-nal conditions change (Thompson, 1967). We seethis whole melange of scientific and knowledge-producing activity immersed in the strong cur-rents of paradigmatic thinking. The ideal orga-nization that we envisage will have been bornin the white heat of a Kuhnian revolution thatbinds the disparate elements together as an in-surrectional force opposed to the prior statusquo. Thus, like Apple Computer, this hypotheti-cal organization will exploit its technical profi-ciency in certain specific areas rather than seekto dominate the marketplace with generic prod-ucts.

At different stages in the development ofknowledge, different organizing logics are likelyto supplement each other. For example, in thefield of synthetic biology, a team of researchersin 2003 successfully created a custom-builtpackage of DNA. The logic of pure science thatdrove their initial work had to be supplemented,over time, with other logics as the team soughtfunding and support to translate basic scienceinto successful pharmaceutical drugs. Led byJay Keasling, a professor of biochemical engi-neering at the University of California at Berke-ley, the team secured $42.6 million to use itspure science discovery to combat a problem ofconcern to the Bill and Melinda Gates founda-tion: the eradication of malaria. With the help ofthese funds, the team started Amyris Biotechnol-ogies, which then collaborated with the Institutefor OneWorld Health, a nonprofit drug maker,and, in 2008, initiated a collaboration withSanofi-Aventis, a Paris-based pharmaceuticalfirm, to bring antimalaria drugs to market in2012 (Specter, 2009). Thus, different organizinglogics may prevail during the different stages ofknowledge production, and firms may creativelycombine different logics over time as they re-spond to the needs and pressures of their evolv-ing institutional environments.

This example illustrates the way in whichpure science gets translated into marketplaceproducts through network alliances, and itraises the question of how network structureaffects the production of knowledge over time.We know that corporate research laboratories,such as Bell Labs, can combine a pure researchgroup with departments devoted to the develop-ment of useful products (such as, at Bell Labs,the transistor, the laser, and the UNIX operatingsystem). But how do different parts of the orga-

nization relate to each other, and how do rela-tions change over time? What kind of “shadowof the past” is likely to haunt knowledge produc-ers in the present? Network research suggeststhat the extent to which knowledge-producinggroups (such as TV production teams) had cohe-sive ingroup ties in the past affects group per-formance in the future, whereas the extent towhich people within the group had ties outsidethe group in the past does not affect future groupperformance (Soda, Usai, & Zaheer, 2004).

Thus, one important question that arises con-cerns whether cohesion continues to influencewhat are considered to be acceptable knowl-edge directions and standards, even as sciencechanges both within the community and outsideit. Creativity research (focused on the produc-tion of Hollywood musicals) has suggested that,within any given community of producers, thesocial structure of the community in terms ofclustering and connectivity can significantly af-fect performance: creative production dependson a fine balance between the clustering of like-minded people and the extent of connectionsacross clusters (Uzzi & Spiro, 2005). The exten-sion and elaboration of these ideas to the pro-duction of new scientific knowledge from a phi-losophy of science approach remains to beachieved.

A network approach could also help investi-gate the ways in which philosophical logics ofaction spread across communities of organiza-tions (DiMaggio & Powell, 1983), given evidencethat social network connections facilitate a gen-eral orientation toward knowledge production,irrespective of organizational affiliation (cf. Sax-enian, 1994). The diffusion of a particular way ofthinking about and doing science can demon-strate a social movement–type fervor (Davis, Mc-Adam, Scott, & Zald, 2005)—a “mob psychology”or bandwagon effect (Abrahamson & Rosenkopf,1993; Rogers & Kincaid, 1981)—that some haveargued detracts from the rationality of scientificprogress (Lakatos, 1970: 140). There would seemto be a considerable difference of opinion con-cerning whether, as Lakatos (1970) claims, suchsocial movement fervor results in only tempo-rary aberrations from the rational progression ofscientific advances or whether rational recon-structions of knowledge progression (of the kindchampioned by Lakatos, 1970) ignore the irra-tionality of so-called scientific progress that ischaracterized by shifts between incommensura-


ble world views (e.g., Feyerabend, 1977). Ifknowledge production is self-correcting in termsof its historical evolution (a view that is compat-ible with Lakatos’s [1970] perspective), then un-derstanding how a logic-of-action bandwagondiverts resources from one research program toanother is still important in helping explain whydiscovery and invention might, in some cases,be delayed. If, however, knowledge productionis path dependent such that if one research pro-gram is supported rather than another the his-tory of knowledge production becomes quite dif-ferent (a view put forward by Noble, 1977), thenunderstanding the spread of logics of actionused to justify resource distribution within andacross knowledge-producing communities be-comes a high-priority task for scholars.

As logics of organizing spread across organi-zational fields (i.e., communities of organiza-tions engaged in related activities; DiMaggio &Powell, 1983) to provide shared schema, prac-tices, and justifications to heterogeneous groupsof organizations engaged in knowledge alli-ances and product development, this is likely tofacilitate collaboration and the formation of al-liances. The field of biotechnology, for example,originated in university labs in the 1970s (Zucker& Darby, 1996), saw the emergence of numeroussmall science-based firms in the 1980s, and hasbeen bringing a number of new medicines tomarket since the 1990s. Because no single orga-nization controls all the competencies requiredto develop and successfully bring a drug to mar-ket, organizations in this field tend to be embed-ded in numerous alliances with other organiza-tions (Powell, Koput, & Smith-Doerr, 1996;Powell, White, Koput, & Owen-Smith, 2005).

When scientific results have to be transferredfrom one institutional context to another, theyroutinely have to be reshaped and recast (Hil-gartner, 1990). Research universities continue topursue blue sky research, whereas industrytends to be the home of more pragmatic, prob-lem-solving work. If academia and industryseek to collaborate, will basic differences inphilosophically based logics of organizinghandicap interactions? Is there a role in suchcollaborations for philosophical brokers trainedin the different philosophical perspectives andable to see where fruitful divisions of labormight be appropriate? Organizational theoryhas emphasized the benefits of brokerage (e.g.,Burt, 1992). However, brokerage, to the extent

that it requires individuals to occupy them-selves in different and disconnected fields, canpose threats to the reputation of individuals inthe eyes of colleagues (cf. Podolny, 2001) andcan, indeed, lead to a perception of the broker asa parasite who feeds on the weakness of others(Serres, 1980). In breaching academic boundar-ies, one “risks the chance of slipping in betweenfields and finding oneself doing work that noone finds relevant” (Pernu, 2008: 32). Gaining abetter understanding of the tactics employed insuccessful brokerage between knowledge-producing entities organized around differentphilosophical logics is a topic for future re-search.

DISCUSSION

If the philosophy of science is underutilized indiscussions of the organization and productionof new knowledge (Grandori & Kogut, 2002: 224),this is, perhaps, evidence of the fragmentationof knowledge into subgenres in the modernAcademy, in which professional philosophy andorganizational studies are separate silos of spe-cialized thinking. In bringing renewed attentionto philosophy of science discourse, our aim herehas been to undertake a creative synthesis inorder to open up new possibilities for theorizingand research concerning the production of sci-entific knowledge.

The philosophy of science, besides providinga distinctive menu of possibilities for manage-ment research (Kleindorfer et al., 1998), also hasthe potential to model the rational production ofknowledge in organizations (Kilduff & Mehra,2008). In this article we have suggested a set ofideal-type logics of action derived from the phi-losophy of science, including the logic of pureresearch (which emphasizes the enduring struc-tural content of scientific theory and justifieslarge groups of specialists’ communal work onmassive projects), the logic of induction (whichemphasizes the investigation and interpretationby a cadre of experts of patterns inherent inempirical data), the logic of problem solving(which emphasizes practical action and an opencommunity of experts with backgrounds thatcross disciplines), strong-paradigm logic (whichemphasizes the relentless articulation of pro-cedures to solve outstanding puzzles withinparadigmatic communities), and the logic ofemancipation (which emphasizes subversive


challenges to prevailing knowledge assump-tions).

Our article has focused on the organizing pro-cess considered as a stream of knowledge (vonKrogh, Roos, & Slocum, 1994), a process of trans-formation by which background assumptionsshared by organizational participants not onlyguide the interpretation of events (cf. March &Simon, 1958) but also facilitate the enactment ofinternal and external environments (Weick,1979) within structures of constraint and controlthat are themselves reproduced by strategic ac-tors (Giddens, 1984). We have argued thatknowledge production is shaped by underlyingassumptions rooted in the philosophy of sciencethat provide different logics for organizing. As-sumptions concerning ontology and epistemol-ogy, often adopted during the formal scientifictraining process, are likely to affect the kind ofresearch scientific knowledge workers pursue,the kind of new knowledge that they produce,and the way they organize to achieve their ob-jectives. This is clear enough in the case of uni-versity researchers (see Crouther-Heyck, 2005,for the influence of logical positivism on HerbertSimon and Holton, 1993, for the influence ofMach’s philosophy on B. F. Skinner), but we the-orize that knowledge workers in other settingsare similarly influenced by the discourses of thephilosophy of science.

We have argued that discourses in the philos-ophy of science shape the logics of organizingadopted by knowledge-producing organiza-tions. But this assumes that causality operatesin only one direction. It is also possible thatresearchers interested in a certain type ofknowledge production may adopt pragmaticallya distinctive discourse associated with a philos-ophy of science in order to justify actions andextract resources from the environment. Further,we suggest that philosophical assumptions areparticularly likely to be invoked during conflictsover funding, status, or credit for new knowl-edge production. Even though we have sug-gested that physicists at the Large Hadron Col-lider are likely to have absorbed a structuralrealist orientation during their training, it is alsolikely that researchers involved in much smallerprojects (e.g., researchers in a small biochemis-try lab) faced with having to justify their workmay resort to a structural realist logic of action.Scientists may use different philosophical viewsin order to legitimize and delegitimize argu-

ments in the eyes of various audiences. For ex-ample, in a dispute over plate tectonics (LeGrand, 1986), “some portrayed themselves asmore concerned with fidelity to data and thusmore empiricist; some portrayed themselves asmaking their claims more precisely falsifiable;and some took the risky strategy of allyingthemselves with a Kuhnian picture of science”(Sismondo, 2010: 127). How knowledge producersembroiled in disputes convince or fail to con-vince audiences of the merits of their views is aquestion that deserves the attention of organi-zational researchers.

Understanding the philosophical underpin-nings of science logics and their implications fororganizing knowledge production may be espe-cially relevant in the current era of changes inscience. In the changing landscape of scientificknowledge production, research groups in uni-versities are considered “quasi firms” that havefrequent knowledge transactions with industry(Kedl, 2009: 229; Oliver, 2008: 195). Lookingaround the intellectual landscape, one sees amarket “of independent epistemic monopoliesproducing vastly different products” (Knorr-Cetina, 1999: 4). This erosion of the demarcationbetween universities and other knowledge-producing organizations and the resultant emer-gence of hybrid organizational forms (Nowotnyet al., 2001) open opportunities for institutionalentrepreneurs (cf. DiMaggio, 1988) to employ arange of logics of action, given the contempo-rary lack of clarity concerning what constitutesa scientific contribution (Ziman, 1996). Some em-inent scientists, for example, have been accusedof launching campaigns employing “disinfor-mation of various sorts coupled with an endur-ing and disgraceful willingness to stick to dis-credited arguments” to influence legislation ona host of issues, from the depletion of the ozonelayer to the death of forests through acid rain(The Economist, 2010: 86; Oreskes & Conway,2010). The entrepreneurial manipulation of insti-tutional logics of action takes place within abroader environment in which government reg-ulators and the public are important stakehold-ers (cf. Misangyi et al., 2008).

We call for greater attention to the use andmisuse of logics of action by organizational rep-resentatives in debates concerning science pol-icy, funding, and legislation. The role of profes-sional philosophers of science as experts inproviding policy advice to organizational actors


could profitably be explored, following the ex-ample of work that has examined the role ofscience experts in policy debates (Hilgartner,2000). Indeed, there are several themes withinthe science and technology studies field thatcould be explored from the perspective put for-ward in this article. These include how thenorms of scientific rationality may be deter-mined to further the class interests of profes-sionals (e.g., Shapin, 1975); how the populariza-tion of science can affect the process ofknowledge production (e.g., Collins & Pinch,1993; Hilgartner, 1990); how scientific phenom-ena themselves can be socially constructed (e.g.,Knorr-Cetina, 1999); and how resources are en-rolled in knowledge networks that combine pa-trons, laboratory equipment, established knowl-edge, and other heterogeneous elements (e.g.,Latour, 1987).

We also call attention to the importance of the“strong programme” in science and technologystudies (Bloor, 1991), particularly the focus on theways in which institutionalized beliefs (such asscientific logics of action) become adopted byrational people even if, to outsiders, these be-liefs are disputed or are seen as less than ra-tional in their operations or consequences (cf.Suddaby & Greenwood, 2005). Researchers, ac-cording to this perspective, need to be self-reflective concerning how they privilege onetype of science over another. One of the princi-ples enunciated by the strong programme is thattrue and false beliefs should be explained bythe same theory (Bloor, 1991: 7). This principlesuggests that the philosophy of science-basedtheory that we have articulated should be devel-oped to be able to explain knowledge produc-tion considered by some to be pseudoscientific(e.g., advances in homeopathic medicine).

We have investigated in this article a set ofrelatively abstract discourses concerning theprogress of science, and we have suggested thatthese discourses are relevant to the productionof new knowledge across a range of scientificorganizations that include but are not restrictedto universities. The discourses of the philosophyof science, we have suggested, can be relevantfor understanding not just how trained scientistsproduce new knowledge but also how the manyother people designated in organizations as“knowledge workers” produce new knowledge.If this article has one overarching conclusion, itis that the philosophy of science can promote

alternative constructions of how knowledge canbe produced, and these alternative construc-tions can facilitate organizational experimentsacross otherwise entrenched knowledge silos.

REFERENCES

Abrahamson, E., & Rosenkopf, L. 1993. Institutional and com-petitive bandwagons: Using mathematical modeling asa tool to explore innovation diffusion. Academy of Man-agement Review, 18: 487–517.

Appleyard, B. 2009. Steve Jobs: The man who polished Apple.Sunday Times, August 16: http://technology.timesonline.co.uk/tol/news/tech_and_web/article6797859.ece?token�null&offset�12&page�2.

Archer, M. 1998. Introduction: Realism in the social sciences.In M. Archer, R. Bhaskar, A. Collier, T. Lawson, & A.Norrie (Eds.), Critical realism: Essential readings: 189–205. New York: Routledge.

Archer, M., Bhaskar, R., Collier, A., Lawson, T., & Norrie, A.(Eds.). 1998. Critical realism: Essential readings. NewYork: Routledge.

Ayer, A. J. 1952. Language, truth and logic. Mineola, NY:Dover.

Baron, J. N., Hannan, M. T., & Burton, M. D. 1999. Building theiron cage: Determinants of managerial intensity in theearly years of organizations. American Sociological Re-view, 64: 527–547.

Beebe, J. 2007. Reliabilism and antirealist theories of truth.Erkenntnis, 66: 375–391.

Bernstein, J. 2008. Physicists on Wall Street and other essayson science and society. New York: Springer.

Bhaskar, R. 1998. General introduction. In M. Archer, R.Bhaskar, A. Collier, T. Lawson, & A. Norrie (Eds.), Criticalrealism: Essential readings: ix–xxiv. New York: Rout-ledge.

Bhattacharya, C. B., & Sen, S. 2003. Consumer-company iden-tification: A framework for understanding consumers’relationships with companies. Journal of Marketing, 67:76–88.

Bloor, D. 1991. Knowledge and social imagery (2nd ed.). Chi-cago: University of Chicago Press.

Bortolotti, L. 2008. An introduction to the philosophy of sci-ence. Malden, MA: Polity Press.

Burt, R. 1992. Structural holes. Cambridge, MA: Harvard Uni-versity Press.

Burt, R. S. 2004. Structural holes and good ideas. AmericanJournal of Sociology, 110: 349–399.

Camilli, R., Reddy, C. M., Yoerger, D. R., Van Mooy, B. A. S.,Jakuba, M. V., Kinsey, J. C., McIntyre, C. P., Sylva, S. P., &Maloney, J. V. 2010. Tracking hydrocarbon plume trans-port and biodegration at Deepwater Horizon. Science,330: 201–204.

Carnap, R. 1928. Der Logische Aufbau der Welt [The logicalstructure of the world]. Berlin: Weltkreis-Verlag.


Cartwright, N. 1983. How the laws of physics lie. Oxford:Clarendon Press.

Chalmers, A. F. 1999. What is this thing called science? (3rded.). Indianapolis: Hackett.

Chapman, M. 2007. 40 days and 40 nights. New York: Collins.

Chesbrough, H. 2002. Graceful exits and missed opportuni-ties: Xerox’s management of its technology spin-off or-ganizations. Business History Review, 76: 803–837.

Cohen, M. D., March, J. G., & Olsen, J. P. 1972. A garbage canmodel of organizational choice. Administrative ScienceQuarterly, 17: 1–25.

Coleman, J. S. 1990. Foundations of social theory. Boston:Belknap Press of Harvard University Press.

Collins, H., & Pinch, T. 1993. The Golem: What everyoneshould know about science. Cambridge: CambridgeUniversity Press.

Crane, T., & Mellor, D. H. 1990. There is no question of phys-icalism. Mind, 99: 185–206.

Creath, R. 1985. Taking theories seriously. Synthese, 62: 317–345.

Crouther-Heyck, H. 2005. Herbert A. Simon: The bounds ofreason in modern America. Baltimore: Johns HopkinsUniversity Press.

Davis, G., McAdam, D., Scott, R. W., & Zald, M. (Eds.). 2005.Social movements and organization theory. New York:Cambridge University Press.

Dewey, J. 1903. Studies in logical theory. Chicago: Universityof Chicago Press.

DiMaggio, P. 1988. Interest and agency in institutional the-ory. In L. G. Zucker (Ed.), Institutional patterns and orga-nizations: 3–21. Cambridge, MA: Ballinger.

DiMaggio, P., & Powell, W. W. 1983. The iron cage revisited:Institutional isomorphism and collective rationality inorganizational fields. American Sociological Review, 48:147–160.

Duhem, P. 1969. (First published in 1908.) To save the phe-nomena: An essay on the idea of physical theory fromPlato to Galileo. Chicago: University of Chicago Press.

Dunn, M. B., & Jones, C. 2010. Institutional logics & institu-tional pluralism: The contestation of care & science log-ics in medical education, 1967–2005. Administrative Sci-ence Quarterly, 55: 114–145.

The Economist. 2010. All guns blazing: A question of dodgyscience. June 17: http://www.economist.com/node/16374460.

Eddington, A. 1939. The philosophy of physical science. Cam-bridge: Cambridge University Press.

Fairclough, N. 2005. Peripheral vision: Discourse analysis inorganization studies: The case for critical realism. Or-ganization Studies, 26: 915–939.

Feyerabend, P. 1975. Against method: Outline of an anarchis-tic theory of knowledge. London: New Left Books.

Feyerabend, P. 1977. Changing patterns of reconstruction.British Journal for the Philosophy of Science, 28: 351–369.

French, S., & Ladyman, J. 2003. Remodelling structural real-

ism: Quantum physics and the metaphysics of structure.Synthese, 136: 31–56.

Friedland, R., & Alford, R. 1991. Bringing society back in:Symbols, practices & institutional contradictions. InW. W. Powell & P. J. DiMaggio (Eds.), The new institu-tionalism in organizational analysis: 232–263. Chicago:University of Chicago Press.

Friedman, M. 1953. Essays in positive economics. Chicago:University of Chicago Press.

Friedson, E. 1970. Professional dominance: The social struc-ture of medical care. New York: Atherton Press.

Friedson, E. 2001. Professionalism: The third logic. Chicago:University of Chicago Press.

Fuller, S. 2003. Kuhn vs. Popper: The struggle for the soul ofscience. Cambridge: Icon.

Gehani, N. 2003. Bell Labs: Life in the crown jewel. Summit,NJ: Silicon Press.

Gibbons, M., Limoges, C., Nowotny, H., Schwartzman, S.,Scott, P., & Trow, M. (Eds.). 1994. The new production ofknowledge: The dynamics of science and research incontemporary societies. London: Sage.

Giddens, A. 1984. The constitution of society. Berkeley: Uni-versity of California Press.

Girard, B. 2009. The Google way: How one company is revo-lutionizing management as we know it. San Francisco:No Starch Press.

Gladwell, M. 2010. The treatment: Why is it so difficult todevelop drugs for cancer? New Yorker, May 17: 69.

Godfrey-Smith, P. 2003. Theory and reality: An introduction tothe philosophy of science. Chicago: University of Chi-cago Press.

Goebel, M., & Gruenwald, L. 1999. A survey of data miningand knowledge discovery tools. ACM SIGKDD Explora-tions Newsletter, 1(1): 20–33.

Goldman, A. 2009. Reliabilism. Stanford Encyclopedia of Phi-losophy. Stanford, CA: http://plato.stanford.edu/entries/reliabilism/.

Grandin, K., Wormbs, N., & Widmalm, S. (Eds.). 2005. Thescience-industry nexus: History, policy, implications.Sagamore Beach, MA: Science History Publications.

Grandori, A., & Kogut, B. 2002. Dialogue on organization andknowledge. Organization Science, 13: 224–231.

Greenwood, R., Suddaby R., & Hinings, C. R. 2002. Theorizingchange: The role of professional associations in thetransformation of institutionalized fields. Academy ofManagement Journal, 45: 58–80.

Hacking, I. 1981. Introduction. In I. Hacking (Ed.), Scientificrevolutions: 1–5. Oxford: Oxford University Press.

Hanson, N. R. 1958. Patterns of discovery: An inquiry into theconceptual foundations of science. Cambridge: Cam-bridge University Press.

Hardy, C., & Maguire, S. 2008. Institutional entrepreneurship.In R. Greenwood, C. Oliver, K. Sahlin, & R. Suddaby(Eds.), The Sage handbook of organizational institution-alism: 198–217. London: Sage.


Hargadon, A. B., & Douglas, Y. 2001. When innovations meetinstitutions: Edison and the design of electric light. Ad-ministrative Science Quarterly, 46: 476–501.

Hargadon, A., & Sutton, R. I. 1997. Technology brokering andinnovation in a product development firm. Administra-tive Science Quarterly, 42: 716–749.

Hessels, L. K., & van Lente, H. 2008. Re-thinking new knowl-edge production: A literature review and a researchagenda. Research Policy, 37: 740–760.

Hickson, D. J., Hinings, C. R., Lee, C. A., Schneck, R. E., &Pennings, J. M. 1971. A strategic contingencies’ theory ofintraorganizational power. Administrative ScienceQuarterly, 16: 216–229.

Hilgartner, S. 1990. The dominant view of popularization:Conceptual problems, political uses. Social Studies ofScience, 20: 519–539.

Hilgartner, S. 2000. Science on stage: Expert advice as publicdrama. Stanford, CA: Stanford University Press.

Holton, G. 1993. Science and anti-science. Cambridge, MA:Harvard University Press.

Horwich, P. 1991. On the nature and norms of theoreticalcommitment. Philosophy of Science, 58: 1–14.

Johnson, K. 1996. Divorced from the job, still wedded to theculture. New York Times, June 16: F11.

Kedl, R. M. 2009. Revolving doors: From academia to industryand back again. Nature Immunology, 10(3): 227–229.

Kilduff, M., & Mehra, A. 2008. Philosophy as core competence.In D. Barry & H. Hansen (Eds.), The Sage handbook ofnew approaches in management and organization: 79–81. London: Sage.

Kleindorfer, G. B., O’Neill, L., & Ganeshan, R. 1998. Validationin simulation: Various positions in the philosophy ofscience. Management Science, 8: 1087–1099.

Knorr-Cetina, K. 1999. Epistemic cultures: How the sciencesmake knowledge. Cambridge, MA: Harvard UniversityPress.

Kuhn, T. S. 1979. Metaphor in science. In A. Ortony (Ed.),Metaphor and thought: 409–419. Cambridge: CambridgeUniversity Press.

Kuhn, T. S. 1990. The road since structure. PSA: Proceedingsof the Biennial Meeting of the Philosophy of ScienceAssociation, 2: 3–13.

Kuhn, T. S. 1996. (First published in 1962.) The structure ofscientific revolutions (3rd ed.). Chicago: University ofChicago Press.

Kunda, G. 1992. Engineering culture: Control and commit-ment in a high-tech corporation. Philadelphia: TempleUniversity Press.

Ladyman, J. 1998. What is structural realism? Studies inHistory and Philosophy of Science, 29: 409–424.

Lakatos, I. 1970. Falsification and the methodology of scien-tific research programs. In I. Lakatos & A. Musgrave(Eds.), Criticism and the growth of knowledge: 91–196.Cambridge: Cambridge University Press.

Latour, B. 1987. Science in action: How to follow scientists

and engineers through society. Cambridge, MA: HarvardUniversity Press.

Laudan, L. 1977. Progress and its problems: Towards a theoryof scientific growth. London: Routledge and Kegan Paul.

Laudan, L. 1981. A confutation of convergent realism. Philos-ophy of Science, 48: 19–49.

Laudan, L. 1984. Science and values: The aims of science andtheir role in scientific debate. Berkeley: University ofCalifornia Press.

Laudan, L. 1990. Science and relativism: Some key controver-sies in the philosophy of science. Chicago: University ofChicago Press.

Le Grand, H. E. 1986. Steady as a rock: Methodology & mov-ing continents. In J. Schuster & R. Yeo (Eds.), The politicsand rhetoric of scientific method: Historical studies: 97–138. Dordrecht, Netherlands: Reidel.

Lounsbury, M. 2007. A tale of two cities: Competing logicsand practice variation in the professionalizing of mutualfunds. Academy of Management Journal, 50: 289–307.

March, J. G., & Simon, H. A. 1958. Organizations. New York:Wiley.

Maxwell, N. 2010. Scientists should stop deceiving us. http://www.guardian.co.uk/commentisfree/2010/mar/12/philosopy-of-science-climate-change.

McKelvey, B. 2002. Model-centered organization scienceepistemology. In J. A. C. Baum (Ed.), The Blackwell com-panion to organizations: 752–780. Oxford: Blackwell.

Merali, Z. 2010. The large human collider. Nature, 464: 482–484.

Miettinen, R. 2006. The sources of novelty: A cultural andsystemic view of distributed creativity. Creativity andInnovation Management, 15: 173–181.

Misangyi, V. F., Weaver, G. R., & Elms, H. 2008. Ending cor-ruption: The interplay among institutional logics, re-sources, and institutional entrepreneurs. Academy ofManagement Review, 33: 750–770.

Musgrave, A. 1971. Kuhn’s second thoughts. British Journal ofthe Philosophy of Science, 22: 287–297.

Noble, D. F. 1977. America by design: Science, technology,and the rise of corporate capitalism. New York: Knopf.

Nowotny, H., Scott, P., & Gibbons, M. 2001. Re-thinking sci-ence: Knowledge and the public in an age of uncer-tainty. Cambridge Polity Press.

Okasha, S. 2002. Philosophy of science: A very short introduc-tion. Oxford: Oxford University Press.

Oliver, A. L. 2008. University-based biotechnology spin-offs.In H. Patzelt & T. Brenner (Eds.), Handbook of bioentre-preneurship: 193–210. New York: Springer.

Oreskes, N., & Conway, E. 2010. Merchants of doubt: How ahandful of scientists obscured the truth on issues fromtobacco smoke to global warming. New York: Blooms-bury Press.

Overbye, D. 2002. Are they a) geniuses or b) jokers? Frenchphysicists’ cosmic theory creates a Big Bang of its own.New York Times, November 9: B7.


Pernu, T. K. 2008. Philosophy and the front line of science.Quarterly Review of Biology, 83: 29–36.

Podolny, J. M. 2001. Networks as the pipes and prisms of themarket. American Journal of Sociology, 107: 33–60.

Poincare, H. 1905. Science and hypothesis. New York: Dover.

Popper, K. 1963. Conjectures and refutations. London: Rout-ledge and Kegan Paul.

Popper, K. 1970. Normal science and its dangers. In I. Lakatos& A. Musgrave (Eds.), Criticism and the growth of knowl-edge: 51–58. Cambridge: Cambridge University Press.

Powell, W. W., Koput, K. W., & Smith-Doerr, L. 1996. Interor-ganizational collaboration and the locus of innovation:Networks of learning in biotechnology. AdministrativeScience Quarterly, 41: 116–145.

Powell, W. W., White, D. R., Koput, K. W., & Owen-Smith, J.2005. Network dynamics & field evolution: The growth ofinterorganizational collaboration in the life sciences.American Journal of Sociology, 110: 1132–1205.

Psillos, S. 1995. Is structural realism the best of both worlds?Dialectica, 49: 15–46.

Psillos, S. 1999. Scientific realism: How science tracks truth.New York: Routledge.

Putnam, H. 1962. What theories are not. In E. Nagel (Ed.),Logic, methodology and philosophy of science: 240–251.Stanford, CA: Stanford University Press.

Putnam, H. 1975. What is “realism”? Proceedings of the Ar-istotelian Society, 76: 177–194.

Putnam, H. 1987. The many faces of realism. LaSalle, IL: OpenCourt.

Rao, H., Monin, P., & Durand, R. 2003. Institutional change inToque Ville: Nouvelle cuisine as an identity movementin French gastronomy. American Journal of Sociology,108: 795–843.

Ray, C. 2000a. Einstein. In W. H. Newton-Smith (Ed.), A com-panion to the philosophy of science: 102–107 Malden,MA: Blackwell.

Ray, C. 2000b. Logical positivism. In W. H. Newton-Smith(Ed.), A companion to the philosophy of science: 243–251.Malden, MA: Blackwell.

Raymond, E. S. 1999. The cathedral and the bazaar. Beijing:O’Reilly.

Reed, M. 2008. Exploring Plato’s cave: Critical realism in thestudy of organization and management. In D. Barry & H.Hansen (Eds.), The Sage handbook of new approaches inmanagement and organization: 68–78. London: Sage.

Reichenbach, H. 1938. Experience and prediction. Chicago:University of Chicago Press.

Rogers, E. M., & Kincaid, D. L. 1981. Communication networks:Toward a new paradigm for research. New York: FreePress.

Saxenian, A. L. 1994. Regional advantage: Culture and com-petition in Silicon Valley and Route 128. Cambridge, MA:Harvard University Press.

Serres, M. 1980. La parasite. Paris: Grasset.

Shahar, E. 1997. A Popperian perspective of the term “evi-

dence-based medicine.” Journal of Evaluation in Clini-cal Practice, 3(2): 109–116.

Shapin, S. 1975. Phrenological knowledge and the socialstructure of early nineteenth-century Edinburgh. Annalsof Science, 32: 219–243.

Sismondo, S. 2010. An introduction to science & technologystudies (2nd ed.). Chichester, UK: Wiley-Blackwell.

Sleeper, L. C. 1986. The necessity of pragmatism: John Dew-ey’s conception of philosophy. New Haven, CT: YaleUniversity Press.

Soda, G., Usai, A., & Zaheer, A. 2004. Network memory: Theinfluence of past and current networks on performance.Academy of Management Journal, 47: 893–906.

Specter, M. 2009. A life of its own: Where will synthetic biologylead us? The New Yorker, September 28: http://www.newyorker.com/reporting/2009/09/28/090928fa_fact_specter.

Stein, H. 1989. Yes, but . . . some skeptical remarks on realismand anti-realism. Dialectica, 43: 47–65.

Suddaby, R., & Greenwood, R. 2005. Rhetorical strategies oflegitimacy. Administrative Science Quarterly, 50: 35–67.

Suppe, F. 1972. What’s wrong with the received view on thestructure of scientific theories? Philosophy of Science,39: 1–19.

Suppe, F. 2000. Understanding scientific theories: An assess-ment of developments, 1969–1998. Philosophy of Science,67(Supplement): S102–S115.

Taleb, N. 2007. The black swan: The impact of the highlyimprobable. New York: Random House.

Tankersley, J. 2010. Engineering a solution to the oil spill. LosAngeles Times, May 21: http://articles.latimes.com/2010/may/21/nation/la-na-oil-spill-houston-20100522.

Thomas, W. I., & Thomas, D. S. 1928. The child in America:Behavior problems and programs. New York: Knopf.

Thompson, J. 1967. Organizations in action. New York:McGraw-Hill.

Thornton, P. H., & Ocasio, W. 2008. Institutional logics. In R.Greenwood, C. Oliver, K. Sahlin, & R. Suddaby (Eds.),The Sage handbook of organizational institutionalism:99–129. London: Sage.

Tsang, E., & Kwan, K. 1999. Replication and theory develop-ment in organizational science: A critical realist per-spective. Academy of Management Review, 24: 759–780.

Tsoukas, H. 1999. David and Goliath in the risk society:Making sense of the conflict between Shell and Green-peace in the North Sea. Organization, 6: 499–528.

Uebel, T. E. 1996. Anti-foundationalism and the Vienna Cir-cle’s revolution in philosophy. British Journal for thePhilosophy of Science, 47: 415–440.

Uzzi, B., & Spiro, J. 2005. Collaboration and creativity: Thesmall world problem. American Journal of Sociology,111: 447–504.

Van de Ven, A. H., & Poole, M. S. 1995. Explaining develop-ment and change in organizations. Academy of Manage-ment Review, 20: 510–540.

Van Maanen, J., & Barley, S. 1984. Occupational communi-


ties: Culture & control in organizations. Research inOrganizational Behavior, 6: 287–365.

Van Maanen, J., & Schein, E. 1979. Toward a theory of orga-nizational socialization. Research in Organizational Be-havior, 1: 209–264.

von Hippel, E. 2005. Democratizing innovation. Cambridge,MA: MIT Press.

von Krogh, G., Roos, J., & Slocum, K. 1994. An essay oncorporate epistemology. Strategic Management Journal,15: 53–71.

Watson, J. D. 1968. The double helix: A personal account ofthe discovery of the structure of DNA. New York: Athe-neum.

Weaver, G. R., & Gioia, D. A. 1994. Paradigms lost: Incom-mensurability vs. structurationist inquiry. OrganizationStudies, 15: 565–589.

Weick, K. E. 1979. The social psychology of organizing. NewYork: Random House.

Wolters, G. 2000. Mach. In W. H. Newton-Smith (Ed.), A com-panion to the philosophy of science: 252–256. Malden,MA: Blackwell.

Worrall, J. 1989. Structural realism: The best of both worlds.Dialectica, 43: 99–124.

Ziman, J. 1996. “Postacademic science”: Constructing knowl-edge with networks and norms. Science Studies, 1: 67–80.

Zucker, L. G. 1977. Role of institutionalization in culturalpersistence. American Sociological Review, 42: 726–743.

Zucker, L., & Darby, M. R. 1996. Star scientists and institutionaltransformation: Patterns of invention & innovation in theformation of the biotechnology industry. Proceedings ofthe National Academy of Sciences, 93: 12709–12716.

Martin Kilduff ([email protected]) is Diageo Professor of Management Studies atthe University of Cambridge. He received his Ph.D. from Cornell University. Hiscurrent research topics (besides philosophy of science theory) include organizationalinnovation, social network cognition, personality effects on network structuring, andthe dark side of emotional intelligence.

Ajay Mehra ([email protected]) is an associate professor of management at theUniversity of Kentucky. He received his Ph.D. from The Pennsylvania State University.His research focuses on the relationship between psychology and the structure anddynamics of social networks.

Mary B. Dunn ([email protected]) is a lecturer in management at theMcCombs School of Business at the University of Texas at Austin. She received herPhD. from Boston College. Her research focuses on professionals’ social networks,knowledge creation, and institutional logics and change.


Copyright of Academy of Management Review is the property of Academy of Management and its content may

not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written

permission. However, users may print, download, or email articles for individual use.

Documents

FROM BLUE SKY RESEARCH TO PROBLEM SOLVING: A … sky research.pdf · scientific knowledge, which we define as new theory that articulates or has the potential to articulate new phenomena