FROM BLUE SKY RESEARCH TO PROBLEM SOLVING: A PHILOSOPHY OF SCIENCE THEORY OF NEW KNOWLEDGE PRODUCTION MARTIN KILDUFF University of Cambridge AJAY MEHRA University of Kentucky MARY B. DUNN University of Texas at Austin We examine rationalized logics developed within discourses of the philosophy of science for implications for the organization of new knowledge. These logics, derived from 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 knowledge produced, 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, providing organized and established procedures for the accomplishment of scientific work. Scientific knowledge is organized knowledge in the sense that its production takes place within and across 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 organizational scholars—those discourses developed within the philosophy of science in answer to the ques- tion “What is science?” (Bortolotti, 2008; Chal- mers, 1999). Within the philosophy of science a range of depictions of scientific activity purport to capture the rational process of scientific dis- covery. These depictions represent alternative logics of action that not only describe in ideal- ized terms actual historical examples of famous scientific breakthroughs but also prescribe the way scientific activity should be conducted so as to separate true science from pseudoscience (cf. Lakatos, 1970). Institutionalized logics of action (defined as organizing 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 for reasoning, and guidelines for practice. These logics constrain cognition and behavior but also provide sources of agency and change (Fried- land & Alford, 1991; Rao, Monin, & Durand, 2003: 795; Thornton & Ocasio, 2008: 101).The purpose of our article is to examine a range of rationalized logics developed within the discourses of the philosophy of science—logics that profoundly affect the research of professional scientists (Laudan, 1977: 59) through their methods and daily activities (Eddington, 1939: vii), as well as through the explanations that scientists “feel compelled” to offer in justification for their prac- tices (Fuller 2003: 93). The justification of scien- tific activity is increasingly important in the modern world (Hilgartner, 2000), in which the boundaries between science and nonscience have become eroded (Ziman, 1996) and in which For valuable comments and insights on prior versions of this paper, we thank discussants and audience members at presentations at the Academy of Management meeting, the Cambridge Realist Workshop, and Oxford University (all in 2008); the OTREG meeting at Imperial College, the Univer- sity of Nottingham, and Tilburg University (all in 2009); the Danish Research Unit for Industrial Dynamics (DRUID) sum- mer conference and the Cambridge Judge Business School organizational behavior reading group (both in 2010.) The paper also benefited enormously from the guidance of guest editor Roy Suddaby and the detailed comments of two anon- ymous reviewers. � Academy of Management Review 2011, Vol. 36, No. 2, 297–317. 297 Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyright holder’s express written permission. Users may print, download, or email articles for individual use only. there is an insatiable demand for new scientific knowledge (Gibbons et al., 1994; Nowotny, Scott, & Gibbons, 2001: 249). Logics of action are encoded in the routines of training, monitoring, disciplining, and reward- ing professionals (e.g., Friedson, 1970, 2001; Greenwood, Suddaby, & Hinings, 2002). Found- ers bring to their new ventures logics of action that continue to influence the structure and practice of work as the firms grow (Baron, Han- nan, & Burton, 1999). The socialization of new members into existing roles (Van Maanen & Bar- ley, 1984; Zucker, 1977) through apprenticeships ensures 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 practices within organizations and industries (Lounsbury, 2007). Relevant logics of action offer to culturally competent actors legitimated discourses for the extraction 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 that these discourses are now routinely invoked in the public sphere in debates concerning science policy (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 to assert that intelligent design could be taught as an alternative to evolution, philosophers of sci- ence appeared for both the plaintiffs and the defendants (Chapman, 2007). To clarify the discussion, we focus on new scientific knowledge, which we define as new theory that articulates or has the potential to articulate new phenomena (Lakatos, 1970). We include within the term theory a variety of forms, including abductive theory (i.e., theory prompted by surprising observations; Hanson, 1958) and theoretical models that posit causal relationships among terms (cf. Suppe, 2000). From whence does this new theory derive? We take the view that new knowledge is strongly conditioned by logics of action that incorporate mutual assumptions and orientations. Logics of action are expressed, renewed, and changed in social routines and networks characteristic of knowledge communities (cf. Giddens, 1984). As Karl Popper argued, “We approach everything in the light of a preconceived theory” (Popper, 1970: 52). Preexisting assumptions and orienta- tions that are embodied in logics of action are likely 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 if scientific knowledge producers see the world through distinctive ontological and epistemo- logical lenses, this way of seeing will have real consequences in terms of the organization of knowledge production. We first derive from the philosophy of science four characteristic logics that 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 for formal and informal organizing. In building new theory, we formulate empirical predictions con- cerning how philosophies of science as logics of action are likely to affect organizing processes and outcomes. These empirical speculations represent opportunities for research rather than established verities. We suggest, for example, that organizations that produce new knowledge may feature not just one but several or all of the different types of organizing frameworks discussed. A cluster of people gathered together in a department or a laboratory is likely to share a particular logic of action that may be different from logics of action operating in other parts of the organization. Our empirical predictions include comparisons of logics of action with respect to the problems that are likely to be pursued, the indicators of prog- ress that are likely to be used, the characteristic methods each perspective might encourage, and the kinds of organizations that are likely to exem- plify each perspective (see Table 1). LOGICS OF SCIENTIFIC KNOWLEDGE 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 of things or relations that can exist. In science, a major ontological issue concerns whether scien- tific theories represent reality— objects, events, and processes outside the human mind— or whether scientific theories comprise explana- tory fictions whose terms (such as electron) are conveniences invented to guide research. Epis- temology concerns how one gains access to knowledge 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 the truth. The ontological question is “Do scientific FIGURE 1 Matrix 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 1 Implications of Philosophies of Science for Organizing Key Questions Structural Realist Foundationalist Instrumentalist Strong Paradigm Critical Realist Characteristic goal and logic of action? Discover fundamental structure of the universe through pure research Find hidden patterns in data through induction Truth-independent problem solving Create scientific paradigm and exploit its implications Emancipate people from prevailing structures of power and oppression Example of type of knowledge produced? Scientific breakthroughs, irrespective of commercial implications Serendipitous discovery of patterns in data from which new theory can be formulated Pragmatic solutions to theoretically defined problems Knowledge and products consistent with the overarching culture of paradigmatic community Exposés of powerful actors’ policies and actions Indicators of progress? Causal expression of relationships among theoretical terms; verification of causal relations among terms Unexpected but replicable correlations indicative of new discoveries; counterintuitive derivations from first principles Greater number of important problems solved Use of paradigm-defined facts to solve puzzles; articulation of the paradigm through empirical work Challenge prevailing power structures, and reimagine possible meanings attached to current practices Characteristic method? Mathematical model building Data mining Those that are considered historically and socially legitimate Defined by methodological exemplars within the paradigm Anthropology of everyday life Illustrative organizing? Self-governing community Cadre of experts Cross-field, focused collaboration Fortress-like organization Subversive team Organizational examples? Large Hadron Collider; Xerox Parc Synta Pharmaceuticals Corp.; Google Inc. Team working to cap Gulf oil spill; Linux Digital Equipment Company; Apple Computer Greenpeace 2011 299Kilduff, Mehra, and Dunn theories represent reality?” The epistemological question is “Does science get closer and closer to the truth?” Ontological and epistemological dimensions are represented in Figure 1 in order to 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 scientific theories aim to provide true descriptions of the world (Okasha, 2002: 59), including the world that lies beyond observable appearances (Chal- mers, 1999: 226). Some versions of realism assert that theoretical terms themselves have “puta- tive factual reference” (Psillos, 1999: 11)—that terms such as utility function refer to real enti- ties. Realist perspectives agree that scientific theories replace each other by offering better accounts of scientific objects (Putnam, 1975) so that, over time, science gets closer and closer to the truth about the world. Because of the affir- mative answers to questions concerning whether scientific theories represent reality and whether science gets closer to the truth over time, realist perspectives are placed in the top left-hand corner of Figure 1. Realist perspectives agree, therefore, that there is a real world independent of our social constructions, that it is possible to assess scien- tific progress toward the truth about this world, and that competing scientific theories can be evaluated rationally in terms of how well they explain significant phenomena about this world. Further, realist perspectives focus on en- during relations between things, typically in the form of mathematical equations. Structural realism. Structural realism repre- sents a major breakthrough in terms of a logic of action that reconciles two seemingly intractably opposed 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 empirical predictions without being basically correct con- cerning the fundamental structure of the world. This view was put forward originally by Poin- caré (1905) but has been advocated in various forms by many realists (e.g., Popper, 1963; Psil- los, 1999; Putnam, 1975). Opposing this view is the pessimistic metainduction argument that the history of science is a graveyard of once empirically successful theories (a perspective also anticipated by Poincaré [1905], as Worrall [1989] points out). If past scientific theories that were successful were found to be false, we have no 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 of structural content from one theory to the next. For example, the shift from the ether theory of light to the electromagnetic theory of light in- volved the retention of the mathematical struc- ture expressed in a series of equations such that at the structural level there is complete continu- ity between the theories (Worrall, 1989). In other cases (such as the transition from Newton’s laws to those of Einstein), mathematical equations are retained “as fully determined limiting cases of other equations, in the passing from an old theory to a new one” (Psillos, 1995: 18). Structural realism, therefore, avoids the claim that theories correctly describe the empirical reality of the world (defusing the pessimism of the antireal- ists) but accepts that successful theories are ap- proximately true descriptions of the underlying structure of the world (accounting for the mirac- ulous success of science). Structural realist organizing: The logic of pure research. In order to gain resources and to intro- duce change into otherwise stable social sys- tems, institutional entrepreneurs “must locate their ideas within the set of existing under- standings and actions that constitute the insti- tutional environment” (Hargadon & Douglas, 2001: 476). The accepted justification for “blue sky” research has typically been couched in terms of a structural realist logic of action. The so-called linear model or fable justifies blue sky research in terms of the necessity that pure re- search scientists delve into the secrets of nature in the absence of tight controls or specific tar- gets so that potential practical applications can be developed by others in the unspecified future for use by consumers (Grandin, Wormbs, & Wid- malm, 2005). Thus, the knowledge workers who engage in pure research are likely to avoid the tendency to tie their mission to the development of specific inventions. Pure researchers are likely to retain a deep underlying belief in the 300 AprilAcademy of Management Review coherence of their theoretical frameworks. New knowledge will tend to be seen as a long-term project driven by acceptance of causal relations among theoretical terms. The structure of theory from a structural realist perspective remains rel- atively unchanging, and it is this very stability that can provide the basis for investing time and resources in innovation (Stein, 1989: 57). Thus, from a structural realist perspective, it is justifiable to organize massive projects aimed at 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 of reality, a reality that is typically assumed to be expressible in mathematical form (Ladyman, 1998) or in terms of theoretical models (Suppe, 2000). This unifying assumption facilitates the self-organization of scientists around massive pure research projects so that hierarchical con- trol is often noticeably absent (Knorr-Cetina, 1999). The kinds of questions that are likely to be pursued, therefore, from the perspective of struc- tural realism include, most basically, improve- ments or modifications to fundamental laws (Psillos, 1995) or theoretical models and attempts to establish evidence to support inferences from such laws. From a structural realist perspective, different ontologies may satisfy the same math- ematical or formal structure, and there is no independent reason to believe that one of these ontologies is better than another (Psillos, 1995: 20). But it is important to remember that the structural realist believes that theories inform us about the structure of the world rather than about fictional entities: “realism should involve reference to what ’really’ exists” (French & Lady- man, 2003: 38). Thus, progress from a structural realist perspective involves improvements to our knowledge concerning the structure of real- ity and the causal relations among entities, al- though structural realism does not necessarily entail improvements to knowledge concerning the objects and properties the world is made of (Ladyman, 1998: 422). Advances in knowledge, according to the logic of structural realism, are likely to be achieved by academic researchers who work for universities or research institutions (that may be funded by private companies). These advances are likely to be published in scientific journals devoted to pure research and, in some cases, registered as patents. Pure research advances will tend to be taken up by other scholars (as measured by citation counts) and by inventors and 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 of understanding the fundamental nature of real- ity. The LAC is run on a communal basis, involv- ing laborious negotiations among competing groups 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 famous example of devotion to relatively pure research was 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 some of 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 millions of dollars to pursue fundamental research, with the understanding that material benefits to Xe- rox Corporation would not show up for at least a decade, these PARC researchers produced rev- olutionary discoveries (largely taken up by com- panies other than Xerox), including the personal computer, a graphical user interface, a laser printer, and technology that would later become indispensable for the spread of the Internet. Instrumentalist Perspectives As its placement in the bottom right-hand corner of Figure 1 indicates, instrumentalism is antirealist in asserting that scientific theories are useful instruments in helping predict events and solve problems. As one contemporary phi- losopher of science explained this perspective, “Fundamental equations do not govern objects in reality; they only govern objects in models” (Cartwright, 1983: 129). A variety of different la- bels (instrumentalism, constructive empiricism, 2011 301Kilduff, Mehra, and Dunn theoretical skepticism, and the philosophy of “as if”; see Horwich, 1991) have been given to the view 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 claims about the hidden, underlying reality” (Horwich, 1991: 1). From this perspective, theories should be 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 to pragmatism (Sleeper, 1986: 3), instrumentalism treats knowledge as something to be sought not for its own sake but for the sake of action to solve problems. Within the social sciences, this tradition is represented by the neoclassical economics view that 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 to provide “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 this instrumentalist perspective, one theory suc- ceeds another not because it moves closer to the truth but because it represents a more useful predictive framework for the phenomena of in- terest. We focus here on the problem-solving approach of Larry Laudan, which connects the world of scientific theory to the solution of prob- lems in the world. Laudan’s philosophy of sci- ence is instrumentalist in the sense defined by John Dewey (1903), who established the require- ment that theories be reliable and useful tools in practical endeavors, such as helping scientists manipulate objects and predict outcomes. Problem solving. According to Laudan (1977), the question of whether a theory is true or false is irrelevant in determining its scientific accept- ability. What is relevant is whether a theory successfully solves problems (Laudan, 1977: 18). Further, Laudan rejects the view that the history of science represents a march toward truth about the world. In his view scientific progress consists of accepting those research traditions that are the most effective in terms of problem solving (Laudan, 1977: 131). Thus, Laudan’s prob- lem-solving approach is representative of the bottom right-hand corner of Figure 1, being anti- realist in terms of ontology and epistemically instrumentalist 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 the research 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 rival and potentially incompatible theories that com- pete with each other and with theories produced by other research traditions in the solution of problems. This point of view separates rational progress from any question concerning the ve- racity of theories, because progress consists of increases in problem solving rather than greater verisimilitude. From the perspective of the social organiza- tion of knowledge production, the problem- solving approach of Laudan (1977) recognizes more clearly than rival approaches the prag- matic nature of scientific progress. Progress in- volves producing more reliable knowledge rather than knowledge that takes us closer to the truth about the universe (Laudan, 1990: 14). From Laudan’s perspective, a scientist can par- ticipate in two different research traditions or can synthesize a new research tradition from competing alternatives, and theory born in one research tradition can be separated and moved or 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 on the basis of past success in problem solving, are likely to “pursue” quite different theories (ones that may even seem wildly improbable) if these theories are seen as offering higher rates of problem-solving progress. Laudan’s approach suggests that “a highly successful research tradition will lead to the abandonment of that worldview which is incom- patible with it, and to the elaboration of a new worldview 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 that are unable to counter the claims of prevailing world views (even if these world views are put forward in nonscientific domains, such as reli- gion) are unlikely to be effective. Science, in 302 AprilAcademy of Management Review Laudan’s view, is a fluid, flexible, and changing endeavor, in which the successful scientist is able to juggle alternative theories and enter imaginatively into different research traditions, all in the service of problem-solving activity that is 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 different theories and different traditions of scientific knowledge 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 likelihood of solving a particular problem, and this may involve working with research traditions that are 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, scientists have a license to adopt and discard theories and methods to the extent that they are useful (cf. Feyerabend, 1975) and socially legitimate, with- out any requirement that the scientists commit themselves paradigmatically (cf. Kuhn, 1996/ 1962) or that they restrict themselves to a set of unchanging core ideas (cf. Lakatos, 1970). There is, therefore, an inherent pragmatism in Laudan’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 empirical inquiry, this commitment is minimal in the sense that theory acceptance does not involve the necessity of believing that the theory is true or that metaphysical unobservables are real. We suggest, therefore, that the production of new knowledge from this perspective will in- volve scientists getting on with the pragmatic business of investigating the empirical regular- ities in nature without having to believe as true the grand metaphysical claims embedded in theories. Problem-oriented scientists faced with conflicting theoretical approaches are likely to compromise 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 theoretical purity is endangered. Because of the problem- solving focus of this logic of action, scientific research from this perspective is likely to be amenable to fortuitous spin-offs from attempts to solve deep intellectual problems (cf. Laudan, 1977: 224). There is a greater likelihood that this logic of action will feature collaborations be- tween university professors and more practi- cally oriented researchers and designers. For example, in the problem-oriented design firm IDEO, which produces innovative products for forty 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 production are sometimes created in response to pressing problems. Consider, for example, the hybrid or- ganization that was assembled to devise solu- tions to the flow of oil pouring into the Gulf of Mexico following the Deepwater Horizon dril- ling rig explosion on April 20, 2010. This crisis team included physicists, experts on Mars dril- ling techniques, an expert on hydrogen bombs, and an MIT professor who referenced “going faster on my snowboard” among his research interests. Literally “anyone . . . who could make a difference was brought in” (a senior BP man- ager, quoted in Tankersley, 2010). Scientific or- ganizations that worked on this cleanup have begun 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, such as Apache, Mozilla, and Linux. These companies operate on the principle that source code is freely available to anyone who wishes to ex- tend, modify, or improve it. In the example of Linux, which has developed an operating sys- tem for computers, the company centers on the founder Linus Torvalds and 121 “maintainers” who are responsible for Linux modules. There are also thousands of user-developers who find bugs and write new pieces of problem-solving code. 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), a principle that has been dubbed “Linus’s law” (Raymond, 1999: 41). The variety of different code developers and improvers means that problem solving is approached in many different ways, with each problem solver using “a slightly dif- ferent perceptual set and analytical toolkit, a different angle to the problem” (Raymond, 1999: 2011 303Kilduff, Mehra, and Dunn 43). The open-source software movement has changed our understanding of the sources of innovation in organizations, providing a basis for 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 combination of antirealism and the belief that science pro- gresses toward truth. Foundationalist anti- realism was promulgated by logical empiricists who were influenced by Ernst Mach (1838 –1898). Mach “strongly believed that science should deal only in observable phenomena” (Ray, 2000a: 104), claimed “that only the objects of sense 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 most influential 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 is antirealist 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 belief that 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 gathering from which scientific knowledge emerges induc- tively (Chalmers, 1999) so that there is a rational basis for evaluating new knowledge claims. Theories with greater empirical content are deemed better than theories with less empirical content. Foundationalism was one of the most widely debated conceptions of knowledge production prior 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 in either empirical data or logical process by which knowledge claims can be validated (cf. Ayer, 1952; Kleindorfer, O’Neill, & Ganeshan, 1998: 1090). This view blends aspects of logical positivism (see Uebel, 1996) and logical empiri- cism (see McKelvey, 2002); indeed, “empiricist philosophies have often had a foundationalist structure” (Godfrey-Smith, 2003: 220). More re- cently, foundationalist philosophy of science has resurfaced under the rubric of “reliabilism,” according to which beliefs are justified when they are produced by cognitive processes that are highly reliable (Goldman, 2009). Reliabilism is compatible with antirealism (Beebe, 2007). Traditional received-view foundationalism typ- ically represents a starting point for debate con- cerning the organization of new knowledge rather than the final word (cf. Kleindorfer et al., 1998). Foundationalist organizing: The logic of in- duction. In terms of the relevance of antirealist foundationalism for new knowledge production in the current era, the emphasis on induction, from which scientific knowledge and theories emerge, implies collecting lots of data, which can be sifted to discover otherwise difficult-to- discern patterns. The prevalence of high-speed computers provides a new impetus for this par- ticular orientation. Computer programs can en- able the scanning of databases for correlations or trends, without any realist presuppositions concerning causality or entity existence. The emphasis within any particular domain is on extracting previously unknown knowledge from factual data using quantitative analyses. In terms of a logic of organizing, foundation- alism would seem to require a small cadre of experts who are able to interpret correlational patterns in order either to create new theory or to match correlational patterns with existing theory so that new knowledge can be extracted. There is a danger of authoritarianism in this emphasis on the interpretation of patterns in data, as has been noted in the debate over evidence-based medicine, where the relevant questions include who decides what is relevant evidence and who determines the best interpre- tation of this collected evidence (Shahar, 1997). Similarly, in scientific management the search has been for the “one best way,” with a relent- less pursuit of improvement through empirical measurement, experiment, and statistical dis- play (e.g., the Gantt chart). New knowledge, from this perspective, could be extracted through close attention to work processes, rather than from the imaginative promulgation of new the- ory. In the current era the reliance on experts continues, but, we suggest, a foundationalist logic of action these days will tend to direct 304 AprilAcademy of Management Review supervision toward large data sets rather than the labor process. Thus, physicists proliferate on Wall Street, bringing their expertise to bear in terms of new algorithms to analyze and profit from trends in financial data (Bernstein, 2008). The consumers of foundationalist-based knowl- edge are likely to be in the front line of service providers, such as practicing physicians, finan- cial managers, and other professionals. In terms of the informal structuring of work from a foundationalist perspective, therefore, the work process is likely to be highly central- ized around the cadre of experts with specialist training who direct the search for empirical reg- ularities that can serve as the foundation for the production of new knowledge. From a founda- tionalist perspective, empirical facts remain facts, even as the world changes and regardless of whether the facts derive from one disciplinary area or another. Therefore, the cadre of special- ists may include people from quite different dis- ciplinary backgrounds and representing quite different historical periods of data representa- tion. There may be a mixing of Ph.D.s in econom- ics and physics, combined together to search for patterns in financial data. To the extent that the focus is on finding patterns in data rather than on 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 advantage of knowledge collected over periods of time, which is then formalized within a standard set of parameters. The clearest contemporary exam- ple of this approach is data mining. The new knowledge discovered through data mining con- sists of patterns in data that can translate into possible new products that take advantage of hitherto unnoticed correlations. Although data mining typically takes advantage of computer automation and algorithms to generate knowl- edge discovery, it depends on a series of judg- ments, including selecting the knowledge area to be searched, preparing the data set from often heterogeneous elements, creating a model that can guide the search process, choosing search algorithms, interpreting results and testing them, and using resulting patterns as the basis for better decision making or new product devel- opment (Goebel & Gruenwald, 1999). Thus, there needs to be a team of specialists guiding the automated process. Foundationalist organizational examples. This foundationalist approach to organizing for knowledge generation is, we suggest, exempli- fied by Synta Pharmaceuticals, a biopharma- ceutical company focused on the discovery, de- velopment, and commercialization of small molecule drugs to treat severe medical condi- tions. Reversing the standard practice in the industry (which is to start with a theoretical un- derstanding of a disease and then rationally design a customized solution), Synta uses mass screening of chemical compounds in the ab- sence of any theory (Gladwell, 2010: 72). The company purchases thousands of chemical com- pounds from around the world, most of which were never designed for medical use. It then tests these compounds in batches to see if they affect cancer cells. Most compounds have no effect or prove toxic to all cells. But, once in a while, a compound proves efficacious against cancer cells. For example, a compound manu- factured at the National Taras Shevchenko Uni- versity, in Kiev, and purchased by Synta for around ten dollars, proved effective against prostate cancer cells. It was an unusual com- pound, “homemade, random, and clearly made for no particular purpose” (Gladwell, 2010: 73). Had it not been for the atheoretical data mining approach employed by Synta, this compound’s ability to fight cancer cells might never have been discovered. The foundationalist path to the generation of new knowledge followed by Synta uses a trial- and-error search for new knowledge that makes no a priori assumptions concerning causality yet maximizes the possibility of serendipitous discovery. This example raises the question as to whether computing power has made possible genuinely theory-neutral exploration in a way that philosophers for decades said could not occur. This new computer-age foundationalist organizing does not require guiding theory but, rather, seems to fulfill the empiricist dream of building science from a foundation of empirical data. In the realm of high-tech companies, Google Inc. 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 Stanford University, developed a search engine that ranked websites according to the number of con- nections to other websites. This innovation fa- 2011 305Kilduff, Mehra, and Dunn cilitated data mining in the enormously com- plex World Wide Web by using information concerning which websites had been “voted” to be the best sources of information by other pages across the web. The company continues to focus on data search through mathematical programming in its development of a range of products and services (Girard, 2009). Strong-Paradigm Perspective The other off-diagonal perspective in Figure 1 derives from the work of Thomas Kuhn (1996/ 1962) that combines a belief in the actuality of the physical world with skepticism about the convergent-realism claim (endorsed by the structural realists) that science progresses to an ever-closer approach to truth. We take at face value the assertion by Kuhn that his philosoph- ical position represents an “unregenerate” real- ism (1979: 415), in the sense that there is a real world out there—“entirely solid: not in the least respectful of an observer’s wishes and desires” (Kuhn, 1990: 10). Kuhn’s position is nuanced by the assertion that the members of a successful disciplinary scientific paradigm define for themselves what aspects of reality to attend to, change, and adapt. Paradigmatic community members share ed- ucation, language, experience, and culture and therefore tend to “see things, process stimuli, in much the same ways” (Kuhn, 1996/1962: 193). The disciplinary matrix that successful scientists share represents reality for them because it se- lects certain objects for investigation and facil- itates the creation of a distinctive social world of scientific endeavor. Kuhn’s realism does not commit him to any strong sense that successive scientific theories approach closer to some par- adigm-independent truth (even though much empirical content may be preserved when one theoretical 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 complex one, but, for our present purposes, we interpret Kuhn as affirming that paradigmatic theories do indeed represent reality, although we recognize that 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 attempts to answer different problems need not mesh well with each other—perhaps the world is too complicated for us to get one comprehensive theory” (Hacking, 1981: 4). Kuhn modified and clarified his ideas consid- erably over the years in response to critics’ in- terpretations and misunderstandings (see Weaver & Gioia, 1994, for a careful discussion of these issues). Kuhn’s revisions have been de- scribed as putting forward “but a pale reflection of the old, revolutionary Kuhn” (Musgrave, 1971: 296). This revised Kuhn has even been described as “a closet positivist” (Laudan, 1984: 68). Cer- tainly, it is the original ideas that generated much of the discussion in the philosophy of sci- ence. In our interpretation of Kuhn, we take into account his later emendations while agreeing with Weaver and Gioia that “early works are not necessarily invalidated by later ones” (1994: 573). Strong-paradigm organizing: The logic of ex- ploitation. Strong-paradigm organizing, in our interpretation, is characterized by a unified force of energetic believers who share funda- mental assumptions about the nature of reality and the practice of research. These believers are likely to be protective of their ideology, given that this ideology has been constructed with the utmost difficulty and constitutes the framework within which meaningful activity can be con- ducted. Paradigm believers exhibit strong resis- tance to ideological or cultural change. Knowledge production from this perspective is, we suggest, likely to be characterized by a relentless focus on the exploitation of existing knowledge 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 of unexploited 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 influences can disturb the equanimity of paradigmatic puzzle-solving activity. Strong-paradigm organizing is likely to ne- glect or ignore persistent anomalies in order to focus on ingenious technological advances and fixes that are compatible with an overall theo- retical approach. Such organizing will, we sug- gest, tend to feature closed boundaries so as to protect proprietary knowledge through an em- 306 AprilAcademy of Management Review phasis on secrecy, patents, copyrights, and con- trols to prevent trade secrets from being stolen by rivals. Cohesive networks are one basis for competitive 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 production will, we suggest, tend to have strong, distinctive cultures and ideologies that powerfully shape the 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 formulation and dissemination of ideology, the use of group testimonials 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 Digital Equipment Corporation, where engineers tended to dismiss the possibility of learning from rivals or the marketplace and tended to cling to the internal, distinctive culture that con- tinued to shape their lives, even after they had been 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 over knowledge: Secrecy is one of Apple’s signature products. . . . Workers on sensitive projects have to pass through many layers of security. Once at their desks or benches, they are monitored by cameras and they must cover up devices with black cloaks and turn on red warning lights when they are uncovered (Appleyard, 2009). The consumers of strong-paradigm knowledge production are likely themselves to resemble cult members in their enthusiasm for distinc- tively branded products that permit their users to differentiate themselves in terms of identity and style (Bhattacharya & Sen, 2003). Strong-paradigm organizing is likely to place the emphasis on a supportive community of like- minded scientists, engaged within a common culture, striving to articulate a distinctive vision and solve a set of well-understood problems. Thus, as Table 1 summarizes, the creation of a scientific 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 paradigm community shares a disciplinary matrix in terms of procedures, exemplars, formulas, and beliefs. Indicators of progress within such a community include success in the solution of outstanding puzzles identified by the paradig- matic community. There is likely to be a strong focus, from this perspective, on new techniques that facilitate the articulation of the paradigm, in terms of empirical testing, and that enable solutions to outstanding puzzles. A Note on Critical Realism When choosing characteristic perspectives to populate the four quadrants of Figure 1, we left out some important philosophical approaches that overlap with these four perspectives. But one contemporary philosophy of science ap- proach— critical realism—stands out as signifi- cantly different from those we have discussed because it focuses on the social world of human interaction 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 discussed elsewhere (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. But whereas structural realism emphasizes the cap- ture of invariant relations in the form of mathe- matical equations, critical realism emphasizes that the relationships among entities discovered by social science are likely to be relatively en- during as opposed to completely invariant. The relatively enduring structure of positions in a particular culture, for example, would be con- sidered to have causal power over the attitudes and behaviors of those who temporarily occupy such positions (Archer, 1998). In focusing on the social world, critical real- ism emphasizes the stratification of reality. The realm of the real consists of unobservable struc- tures and causal powers, the realm of the actual consists of events and processes in the world, and the realm of the empirical consists of the experiences of human beings (Fairclough, 2005). Thus, critical realism emphasizes that there are underlying structures and forces that are unob- 2011 307Kilduff, Mehra, and Dunn servable but real in their operations and best investigated through ethnographic and histori- cal research (rather than exclusively through quantitative 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. The critical aspect of this philosophy of science re- lates to the possibility that an explanatory cri- tique of the ways in which structures of power operate 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 current thinking, it would be necessary, from this per- spective, to tackle the past inheritances put in place by prior thinking that tended to shackle new discovery. Consumers of such knowledge are likely to be social activists, interested in radical changes to the status quo. The logic of action associated with critical realism, therefore, is the logic of emancipation. Organizations that exemplify a critical realist approach to new knowledge production will, we suggest, tend to be social action organizations, such as Greenpeace, that fight in the public sphere to seize rhetorical control over the inter- pretation of events (Tsoukas, 1999). Not content to just produce new knowledge through its own research laboratories at the University of Exeter, Greenpeace is active in direct campaigns against deforestation, overfishing, commercial whaling, global warming, and nuclear power. Thus, this nongovernmental environmental or- ganization attempts to take scientific research and use it to change the attitudes and behaviors of people and to produce products that provide green alternatives to standard technology. COMBINING AND DIFFUSING THE LOGICS Given the stark differences between these logics on basic issues of epistemology and on- tology, the question arises as to which approach we consider to be the best. In writing this article, for example, which philosophy of science are we implicitly endorsing? As students of organiza- tions, we are persuaded by a garbage-can ap- proach (Cohen, March, & Olsen, 1972), according to 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 tackle streams of different types of problems. As we have discussed in the article and summarized in Figure 1 and Table 1, each philosophical alter- native holds different implications for organiz- ing. In terms of our own efforts at knowledge production, we engaged with a pure research effort (i.e., the philosophy of science) generally considered to have little relationship to practi- cal endeavors, we applied this philosophical discourse instrumentally to understand the pro- duction of scientific knowledge, and we tried to develop a fairly tight paradigm organized around Figure 1 and Table 1 in order to exploit the discourses of the philosophy of science. We have not tried to engage in data mining through the collection of lots of information concerning actual 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 in terms of underlying patterns (e.g., Van de Ven & Poole, 1995). In short, we anticipate that knowl- edge-producing efforts will generally feature a variety of logics of action in combination. For example, Figure 2 outlines a hypothetical knowledge-producing organization that inte- grates a mix of logics of action. Like other orga- nizations, those focused on knowledge produc- tion face technical and environmental uncertainty. To manage this uncertainty, the knowledge-producing organization is likely to create certain parts intended specifically to deal with uncertainty, thus allowing certain other parts to carry on the core activities of the orga- nization under conditions of relative certainty (Thompson, 1967). At the technical core of the ideal knowledge-producing organization, we envisage a team of structural realists, scientists who completely believe in the mission of discov- ering the truth about the universe, including the world in which we live. For these scientists there are few doubts about whether theories represent reality or whether theories march forward to- ward greater and greater truth. For example, at Bell Labs, during its illustrious history, the core of its knowledge-producing efforts consisted of the fundamental physics research group that won seven Nobel prizes for discoveries includ- ing the wave nature of matter and the existence of cosmic microwave background radiation (see 308 AprilAcademy of Management Review Gehani, 2003, for a description of recent events in this organization). The work of these researchers in discovering new phenomena and solutions to theoretical problems would need, however, to be buffered from environmental fluctuations by a protective belt (cf. Lakatos, 1970) populated by other knowl- edge workers. The job of the protective belt would be to allow the technical core to operate under conditions approximating certainty by helping the organization to adjust to constraints and contingencies not controlled by the organi- zation. The workers in the protective belt would seek 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, including alliance building, co-opting, lobbying, and the use 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, who would pragmatically negotiate ways in which pure, blue sky science could be translated into products and services relevant for problems arising in society (cf. Hilgartner, 2000). We also envisage a group of foundationalists, actively scanning data from the environment and match- ing patterns with ideas emerging from a core group 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 possibly obstructing the emergence of revolutionary ideas and able to give advice concerning what kinds of products and services might be intro- duced most expediently in particular sectors and how these products and services might be characterized. The relative power of these differ- ent sets of experts and mediators in the protec- tive belt is likely to change and shift depending upon the kinds of crises provoked by environ- mental trends (Hickson, Hinings, Lee, Schneck, & Pennings, 1971). Skilled mediators are likely to be required so as to coordinate the flow of resources and knowledge between these various philosophi- cally disparate groups. Philosophical differ- ences concerning such ontological questions as which categories of things are justified can pro- vide arguments against cooperation and change (Suddaby & Greenwood, 2005). If the knowledge-producing organization must at- tempt to create something like certainty for its technical core while also retaining flexibility and adaptability to satisfy environmental con- straints, we might expect to see a dedicated set of broker-managers, charged with smoothing the irregularities stemming from environmental fluctuations while also pushing the technical FIGURE 2 Organizing Logics Combining in Action in Hypothetical Organization Structural realist: Pure research Instrumentalist: Problem solving Foundationalist: Induction Strong paradigm: Exploitation Critical realist: Emancipation 2011 309Kilduff, Mehra, and Dunn core to make necessary modifications as exter- nal conditions change (Thompson, 1967). We see this whole mélange 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 born in the white heat of a Kuhnian revolution that binds the disparate elements together as an in- surrectional force opposed to the prior status quo. Thus, like Apple Computer, this hypotheti- cal organization will exploit its technical profi- ciency in certain specific areas rather than seek to dominate the marketplace with generic prod- ucts. At different stages in the development of knowledge, different organizing logics are likely to supplement each other. For example, in the field of synthetic biology, a team of researchers in 2003 successfully created a custom-built package of DNA. The logic of pure science that drove their initial work had to be supplemented, over time, with other logics as the team sought funding and support to translate basic science into successful pharmaceutical drugs. Led by Jay Keasling, a professor of biochemical engi- neering at the University of California at Berke- ley, the team secured $42.6 million to use its pure science discovery to combat a problem of concern to the Bill and Melinda Gates founda- tion: the eradication of malaria. With the help of these funds, the team started Amyris Biotechnol- ogies, which then collaborated with the Institute for OneWorld Health, a nonprofit drug maker, and, in 2008, initiated a collaboration with Sanofi-Aventis, a Paris-based pharmaceutical firm, to bring antimalaria drugs to market in 2012 (Specter, 2009). Thus, different organizing logics may prevail during the different stages of knowledge production, and firms may creatively combine 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 which pure science gets translated into marketplace products through network alliances, and it raises the question of how network structure affects the production of knowledge over time. We know that corporate research laboratories, such as Bell Labs, can combine a pure research group with departments devoted to the develop- ment of useful products (such as, at Bell Labs, the transistor, the laser, and the UNIX operating system). But how do different parts of the orga- nization relate to each other, and how do rela- tions change over time? What kind of “shadow of the past” is likely to haunt knowledge produc- ers in the present? Network research suggests that the extent to which knowledge-producing groups (such as TV production teams) had cohe- sive ingroup ties in the past affects group per- formance in the future, whereas the extent to which people within the group had ties outside the group in the past does not affect future group performance (Soda, Usai, & Zaheer, 2004). Thus, one important question that arises con- cerns whether cohesion continues to influence what are considered to be acceptable knowl- edge directions and standards, even as science changes both within the community and outside it. Creativity research (focused on the produc- tion of Hollywood musicals) has suggested that, within any given community of producers, the social structure of the community in terms of clustering and connectivity can significantly af- fect performance: creative production depends on a fine balance between the clustering of like- minded people and the extent of connections across 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 be achieved. A network approach could also help investi- gate the ways in which philosophical logics of action spread across communities of organiza- tions (DiMaggio & Powell, 1983), given evidence that 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 of thinking 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 have argued detracts from the rationality of scientific progress (Lakatos, 1970: 140). There would seem to be a considerable difference of opinion con- cerning whether, as Lakatos (1970) claims, such social movement fervor results in only tempo- rary aberrations from the rational progression of scientific advances or whether rational recon- structions of knowledge progression (of the kind championed by Lakatos, 1970) ignore the irra- tionality of so-called scientific progress that is characterized by shifts between incommensura- 310 AprilAcademy of Management Review ble world views (e.g., Feyerabend, 1977). If knowledge production is self-correcting in terms of its historical evolution (a view that is compat- ible with Lakatos’s [1970] perspective), then un- derstanding how a logic-of-action bandwagon diverts resources from one research program to another is still important in helping explain why discovery and invention might, in some cases, be delayed. If, however, knowledge production is 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), then understanding the spread of logics of action used to justify resource distribution within and across 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 groups of organizations engaged in knowledge alli- ances and product development, this is likely to facilitate 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 numerous small science-based firms in the 1980s, and has been bringing a number of new medicines to market since the 1990s. Because no single orga- nization controls all the competencies required to 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 transferred from one institutional context to another, they routinely have to be reshaped and recast (Hil- gartner, 1990). Research universities continue to pursue blue sky research, whereas industry tends to be the home of more pragmatic, prob- lem-solving work. If academia and industry seek to collaborate, will basic differences in philosophically based logics of organizing handicap interactions? Is there a role in such collaborations for philosophical brokers trained in the different philosophical perspectives and able to see where fruitful divisions of labor might be appropriate? Organizational theory has 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, can pose threats to the reputation of individuals in the eyes of colleagues (cf. Podolny, 2001) and can, indeed, lead to a perception of the broker as a parasite who feeds on the weakness of others (Serres, 1980). In breaching academic boundar- ies, one “risks the chance of slipping in between fields and finding oneself doing work that no one finds relevant” (Pernu, 2008: 32). Gaining a better understanding of the tactics employed in successful brokerage between knowledge- producing entities organized around different philosophical logics is a topic for future re- search. DISCUSSION If the philosophy of science is underutilized in discussions of the organization and production of new knowledge (Grandori & Kogut, 2002: 224), this is, perhaps, evidence of the fragmentation of knowledge into subgenres in the modern Academy, in which professional philosophy and organizational studies are separate silos of spe- cialized thinking. In bringing renewed attention to philosophy of science discourse, our aim here has been to undertake a creative synthesis in order to open up new possibilities for theorizing and research concerning the production of sci- entific knowledge. The philosophy of science, besides providing a distinctive menu of possibilities for manage- ment research (Kleindorfer et al., 1998), also has the potential to model the rational production of knowledge in organizations (Kilduff & Mehra, 2008). In this article we have suggested a set of ideal-type logics of action derived from the phi- losophy of science, including the logic of pure research (which emphasizes the enduring struc- tural content of scientific theory and justifies large groups of specialists’ communal work on massive projects), the logic of induction (which emphasizes the investigation and interpretation by a cadre of experts of patterns inherent in empirical data), the logic of problem solving (which emphasizes practical action and an open community of experts with backgrounds that cross disciplines), strong-paradigm logic (which emphasizes the relentless articulation of pro- cedures to solve outstanding puzzles within paradigmatic communities), and the logic of emancipation (which emphasizes subversive 2011 311Kilduff, Mehra, and Dunn challenges to prevailing knowledge assump- tions). Our article has focused on the organizing pro- cess considered as a stream of knowledge (von Krogh, Roos, & Slocum, 1994), a process of trans- formation by which background assumptions shared by organizational participants not only guide the interpretation of events (cf. March & Simon, 1958) but also facilitate the enactment of internal and external environments (Weick, 1979) within structures of constraint and control that are themselves reproduced by strategic ac- tors (Giddens, 1984). We have argued that knowledge production is shaped by underlying assumptions rooted in the philosophy of science that provide different logics for organizing. As- sumptions concerning ontology and epistemol- ogy, often adopted during the formal scientific training process, are likely to affect the kind of research 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 Herbert Simon and Holton, 1993, for the influence of Mach’s philosophy on B. F. Skinner), but we the- orize that knowledge workers in other settings are similarly influenced by the discourses of the philosophy of science. We have argued that discourses in the philos- ophy of science shape the logics of organizing adopted by knowledge-producing organiza- tions. But this assumes that causality operates in only one direction. It is also possible that researchers interested in a certain type of knowledge production may adopt pragmatically a distinctive discourse associated with a philos- ophy of science in order to justify actions and extract resources from the environment. Further, we suggest that philosophical assumptions are particularly likely to be invoked during conflicts over 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 structural realist orientation during their training, it is also likely that researchers involved in much smaller projects (e.g., researchers in a small biochemis- try lab) faced with having to justify their work may resort to a structural realist logic of action. Scientists may use different philosophical views in order to legitimize and delegitimize argu- ments in the eyes of various audiences. For ex- ample, in a dispute over plate tectonics (Le Grand, 1986), “some portrayed themselves as more concerned with fidelity to data and thus more empiricist; some portrayed themselves as making their claims more precisely falsifiable; and some took the risky strategy of allying themselves with a Kuhnian picture of science” (Sismondo, 2010: 127). How knowledge producers embroiled in disputes convince or fail to con- vince audiences of the merits of their views is a question that deserves the attention of organi- zational researchers. Understanding the philosophical underpin- nings of science logics and their implications for organizing knowledge production may be espe- cially relevant in the current era of changes in science. In the changing landscape of scientific knowledge production, research groups in uni- versities are considered “quasi firms” that have frequent knowledge transactions with industry (Kedl, 2009: 229; Oliver, 2008: 195). Looking around the intellectual landscape, one sees a market “of independent epistemic monopolies producing vastly different products” (Knorr- Cetina, 1999: 4). This erosion of the demarcation between universities and other knowledge- producing organizations and the resultant emer- gence of hybrid organizational forms (Nowotny et al., 2001) open opportunities for institutional entrepreneurs (cf. DiMaggio, 1988) to employ a range of logics of action, given the contempo- rary lack of clarity concerning what constitutes a scientific contribution (Ziman, 1996). Some em- inent scientists, for example, have been accused of 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 on a host of issues, from the depletion of the ozone layer 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 a broader 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 and misuse 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 in providing policy advice to organizational actors 312 AprilAcademy of Management Review could profitably be explored, following the ex- ample of work that has examined the role of science experts in policy debates (Hilgartner, 2000). Indeed, there are several themes within the science and technology studies field that could be explored from the perspective put for- ward in this article. These include how the norms 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 of knowledge 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 technology studies (Bloor, 1991), particularly the focus on the ways in which institutionalized beliefs (such as scientific logics of action) become adopted by rational 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 one type of science over another. One of the princi- ples enunciated by the strong programme is that true and false beliefs should be explained by the same theory (Bloor, 1991: 7). This principle suggests that the philosophy of science-based theory 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 of relatively abstract discourses concerning the progress of science, and we have suggested that these discourses are relevant to the production of new knowledge across a range of scientific organizations that include but are not restricted to universities. The discourses of the philosophy of science, we have suggested, can be relevant for understanding not just how trained scientists produce new knowledge but also how the many other people designated in organizations as “knowledge workers” produce new knowledge. If this article has one overarching conclusion, it is that the philosophy of science can promote alternative constructions of how knowledge can be produced, and these alternative construc- tions can facilitate organizational experiments across otherwise entrenched knowledge silos. REFERENCES Abrahamson, E., & Rosenkopf, L. 1993. Institutional and com- petitive bandwagons: Using mathematical modeling as a 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. c o . 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. New York: Routledge. Ayer, A. J. 1952. Language, truth and logic. Mineola, NY: Dover. Baron, J. N., Hannan, M. T., & Burton, M. D. 1999. Building the iron cage: Determinants of managerial intensity in the early 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 essays on science and society. New York: Springer. Bhaskar, R. 1998. General introduction. In M. Archer, R. Bhaskar, A. Collier, T. Lawson, & A. Norrie (Eds.), Critical realism: 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. American Journal 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 logical structure of the world]. Berlin: Weltkreis-Verlag. 2011 313Kilduff, Mehra, and Dunn Cartwright, N. 1983. How the laws of physics lie. Oxford: Clarendon Press. Chalmers, A. F. 1999. What is this thing called science? (3rd ed.). 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 can model of organizational choice. Administrative Science Quarterly, 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 everyone should know about science. Cambridge: Cambridge University 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 of reason in modern America. Baltimore: Johns Hopkins University 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: University of 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 in organizational 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 from Plato 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 dodgy science. 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 in organization 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. In W. 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 of science. 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 of knowledge: The dynamics of science and research in contemporary 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 to develop drugs for cancer? New Yorker, May 17: 69. Godfrey-Smith, P. 2003. Theory and reality: An introduction to the philosophy of science. Chicago: University of Chi- cago Press. Goebel, M., & Gruenwald, L. 1999. A survey of data mining and 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. The science-industry nexus: History, policy, implications. Sagamore Beach, MA: Science History Publications. Grandori, A., & Kogut, B. 2002. Dialogue on organization and knowledge. Organization Science, 13: 224 –231. Greenwood, R., Suddaby R., & Hinings, C. R. 2002. Theorizing change: The role of professional associations in the transformation of institutionalized fields. Academy of Management Journal, 45: 58 – 80. Hacking, I. 1981. Introduction. In I. Hacking (Ed.), Scientific revolutions: 1–5. Oxford: Oxford University Press. Hanson, N. R. 1958. Patterns of discovery: An inquiry into the conceptual 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. 314 AprilAcademy of Management Review Hargadon, A. B., & Douglas, Y. 2001. When innovations meet institutions: Edison and the design of electric light. Ad- ministrative Science Quarterly, 46: 476 –501. Hargadon, A., & Sutton, R. I. 1997. Technology brokering and innovation 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 research agenda. 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 of intraorganizational power. Administrative Science Quarterly, 16: 216 –229. Hilgartner, S. 1990. The dominant view of popularization: Conceptual problems, political uses. Social Studies of Science, 20: 519 –539. Hilgartner, S. 2000. Science on stage: Expert advice as public drama. 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 theoretical commitment. Philosophy of Science, 58: 1–14. Johnson, K. 1996. Divorced from the job, still wedded to the culture. New York Times, June 16: F11. Kedl, R. M. 2009. Revolving doors: From academia to industry and 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 of new approaches in management and organization: 79 – 81. London: Sage. Kleindorfer, G. B., O’Neill, L., & Ganeshan, R. 1998. Validation in simulation: Various positions in the philosophy of science. Management Science, 8: 1087–1099. Knorr-Cetina, K. 1999. Epistemic cultures: How the sciences make knowledge. Cambridge, MA: Harvard University Press. Kuhn, T. S. 1979. Metaphor in science. In A. Ortony (Ed.), Metaphor and thought: 409 – 419. Cambridge: Cambridge University Press. Kuhn, T. S. 1990. The road since structure. PSA: Proceedings of the Biennial Meeting of the Philosophy of Science Association, 2: 3–13. Kuhn, T. S. 1996. (First published in 1962.) The structure of scientific revolutions (3rd ed.). Chicago: University of Chicago Press. Kunda, G. 1992. Engineering culture: Control and commit- ment in a high-tech corporation. Philadelphia: Temple University Press. Ladyman, J. 1998. What is structural realism? Studies in History 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: Harvard University Press. Laudan, L. 1977. Progress and its problems: Towards a theory of 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 and their role in scientific debate. Berkeley: University of California Press. Laudan, L. 1990. Science and relativism: Some key controver- sies in the philosophy of science. Chicago: University of Chicago Press. Le Grand, H. E. 1986. Steady as a rock: Methodology & mov- ing continents. In J. Schuster & R. Yeo (Eds.), The politics and rhetoric of scientific method: Historical studies: 97– 138. Dordrecht, Netherlands: Reidel. Lounsbury, M. 2007. A tale of two cities: Competing logics and practice variation in the professionalizing of mutual funds. 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 science epistemology. 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 and systemic view of distributed creativity. Creativity and Innovation 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 of Management Review, 33: 750 –770. Musgrave, A. 1971. Kuhn’s second thoughts. British Journal of the 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 a handful of scientists obscured the truth on issues from tobacco smoke to global warming. New York: Blooms- bury Press. Overbye, D. 2002. Are they a) geniuses or b) jokers? French physicists’ cosmic theory creates a Big Bang of its own. New York Times, November 9: B7. 2011 315Kilduff, Mehra, and Dunn 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 the market. American Journal of Sociology, 107: 33– 60. Poincaré, 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. Administrative Science Quarterly, 41: 116 –145. Powell, W. W., White, D. R., Koput, K. W., & Owen-Smith, J. 2005. Network dynamics & field evolution: The growth of interorganizational 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: Open Court. Rao, H., Monin, P., & Durand, R. 2003. Institutional change in Toque Ville: Nouvelle cuisine as an identity movement in 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 the study of organization and management. In D. Barry & H. Hansen (Eds.), The Sage handbook of new approaches in management 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: Free Press. 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 social structure of early nineteenth-century Edinburgh. Annals of Science, 32: 219 –243. Sismondo, S. 2010. An introduction to science & technology studies (2nd ed.). Chichester, UK: Wiley-Blackwell. Sleeper, L. C. 1986. The necessity of pragmatism: John Dew- ey’s conception of philosophy. New Haven, CT: Yale University Press. Soda, G., Usai, A., & Zaheer, A. 2004. Network memory: The influence 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 biology lead 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 realism and anti-realism. Dialectica, 43: 47– 65. Suddaby, R., & Greenwood, R. 2005. Rhetorical strategies of legitimacy. Administrative Science Quarterly, 50: 35– 67. Suppe, F. 1972. What’s wrong with the received view on the structure 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 highly improbable. New York: Random House. Tankersley, J. 2010. Engineering a solution to the oil spill. Los Angeles 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 the Philosophy of Science, 47: 415– 440. Uzzi, B., & Spiro, J. 2005. Collaboration and creativity: The small 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- 316 AprilAcademy of Management Review ties: Culture & control in organizations. Research in Organizational 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 on corporate epistemology. Strategic Management Journal, 15: 53–71. Watson, J. D. 1968. The double helix: A personal account of the discovery of the structure of DNA. New York: Athe- neum. Weaver, G. R., & Gioia, D. A. 1994. Paradigms lost: Incom- mensurability vs. structurationist inquiry. Organization Studies, 15: 565–589. Weick, K. E. 1979. The social psychology of organizing. New York: 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 cultural persistence. American Sociological Review, 42: 726 –743. Zucker, L., & Darby, M. R. 1996. Star scientists and institutional transformation: Patterns of invention & innovation in the formation of the biotechnology industry. Proceedings of the National Academy of Sciences, 93: 12709 –12716. Martin Kilduff (mjkilduff@gmail.com) is Diageo Professor of Management Studies at the University of Cambridge. He received his Ph.D. from Cornell University. His current research topics (besides philosophy of science theory) include organizational innovation, social network cognition, personality effects on network structuring, and the dark side of emotional intelligence. Ajay Mehra (ajay.mehra@uky.edu) is an associate professor of management at the University of Kentucky. He received his Ph.D. from The Pennsylvania State University. His research focuses on the relationship between psychology and the structure and dynamics of social networks. Mary B. Dunn (mary.bowker.dunn@gmail.com) is a lecturer in management at the McCombs School of Business at the University of Texas at Austin. She received her PhD. from Boston College. Her research focuses on professionals’ social networks, knowledge creation, and institutional logics and change. 2011 317Kilduff, Mehra, and Dunn 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.