new knowledge products as bricolage: metaphors … · bricolage: metaphors and scripts in...

NEW KNOWLEDGE PRODUCTS ASBRICOLAGE: METAPHORS AND SCRIPTS IN

ORGANIZATIONAL THEORY

EVA BOXENBAUMCopenhagen Business School and MINES ParisTech

LINDA ROULEAUHEC Montreal

We argue that three epistemic scripts of knowledge production—evolution, differen-tiation, and bricolage—underpin the production—that is, the conception and thepresentation—of new organizational theories. Bricolage of concepts, empirical mate-rial, and metaphors enables the conception of new theories, whereas evolution anddifferentiation, carrying higher academic legitimacy, predominate in theory presen-tation. We develop an integrative model and provide an illustration from organiza-tional institutionalism to delineate how metaphors and scripts influence organiza-tional theory production.

How do new organizational theories come intoexistence? In light of the relatively few organi-zational theories that have been introduced inrecent years (Davis, 2010), we examine how or-ganizational scholars conceive and present newtheories as part of their academic knowledgeproduction. Most notably, we address the epis-temic scripts of knowledge production thatstructure both the conception and presentationof new theories. Epistemic scripts refer to theimplicit, cognitive templates that underpin ourcollective understandings of how new academicknowledge is produced. As such, epistemicscripts represent an institutionalized dimensionof academic knowledge production guiding howscholars develop new theories and how aca-demic audiences recognize a knowledge claimas a new theory. Within organizational theory,we identify three epistemic scripts of knowledgeproduction that organizational scholars use inthe conception and presentation of new organi-zational theories: evolution, differentiation, andbricolage. We describe the characteristics of

each script, analyze their role in theory produc-tion, and examine the possibility of decouplingbetween the scripts that guide the conception ofa new organizational theory and the scripts thatstructure its presentation in academic writing.

Previous inquiries into the emergence of neworganizational theories have developed alongtwo parallel axes, one examining the conceptionof new theories, including their components andprocesses of theory development (e.g., Lewis &Grimes, 1999; McKinley, 2010; Sonpar & Golden-Biddle, 2008; Weick, 1989), and the other lookingat the use of rhetoric in the written presentationof new knowledge (Bonet & Sauquet, 2010; Felin& Foss, 2009; Locke & Golden-Biddle, 1997;Reeves, 2005). Within the former literature,which is significantly more extensive than thelatter, we focus on the stream of research thatexamines the different components that consti-tute theories. Organizational scholars haveidentified empirical material (Alvesson & Karre-man, 2007), theoretical concepts (Whetten, Felin,& King, 2009), and metaphors (Cornelissen, 2006;Cornelissen, Kafouros, & Lock, 2005), to whichwe add epistemic scripts. Since the first two ofthese building blocks are widely recognized el-ements of theory building, we devote our ana-lytical attention to metaphors and scripts.

Metaphors represent a controversial compo-nent of organizational theories (Grant, 2001;Morgan, 1983; Oswick, Keenoy, & Jones, 2003).They were traditionally held to be a source of

Colleagues in the Department of Organization at Copen-hagen Business School, notably Kristian Kreiner, MajkenSchultz, and Morten Vendelø, provided thoughtful input onan earlier version of this manuscript, for which we are grate-ful. We also thank guest editor Roy Suddaby and threeanonymous reviewers for their academic engagement, con-structive approach, and intellectual challenge, all of whichhelped us to sharpen the argument.

� Academy of Management Review2011, Vol. 36, No. 2, 272–296.

272Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyrightholder’s express written permission. Users may print, download, or email articles for individual use only.

pollution in scientific thinking and writing (Pin-der & Bourgeois, 1982), but recent work on thetopic suggests that metaphors constitute a corecomponent in cognitive processing (Cornelis-sen, 2005; Cornelissen & Karoufos, 2008a). Notonly are they inevitable but they also provide avaluable source of imagination, one that in-spires theorists to generate novel propositionsabout their object of study (Bacharach, 1989: 497).They shape the knowledge product as well. Asrecent work on metaphors in organizational the-ory suggests, metaphors are not simply dis-carded after use but remain integrated with the-oretical concepts (McKinley, 2010) and empiricalmaterial (Alvesson, 2003). Metaphors thus per-meate academic knowledge production and theresulting knowledge products. Extending thisline of inquiry, we propose that organizationalscholars form new theories by evoking a widerange of metaphors, which they tacitly assembleand integrate with theoretical concepts and em-pirical material, using epistemic scripts.

Epistemic scripts guide not only the concep-tion of theories but also their written presenta-tion. Academic knowledge about organizationscirculates in the form of texts. When academicswrite papers, they translate organizational real-ity into theoretical terms that are adapted fordifferent audiences (Czarniawska, 1999; Gabriel,2002). In so doing, they employ a variety of tex-tual devices to convey their knowledge claims toreaders and to frame these claims as novel con-tributions to the literature. Empirical findingsshow that notions of “new,” “innovative,”“unique,” and “novel” are frequently used topresent research findings as a substantial con-tribution to academic knowledge (Beyer, Cha-nove, & Fox, 1995; Locke & Golden-Biddle, 1997;Mone & McKinley, 1993). Locke and Golden-Biddle (1997) demonstrated in a detailed analy-sis of organizational journal articles that schol-ars draw on two sets of rhetorical templates toframe new knowledge, one describing the exist-ing body of literature and the other positioningresearch findings as a novel contribution to thisliterature. Scholars combine these templates tocraft a novel contribution to the literature. Al-though this stream of research has added valu-able insight into the textual construction of or-ganizational knowledge, the institutionalfoundation of such knowledge production hasyet to be exposed. We illuminate the epistemicscripts that scholars draw on, more or less im-

plicitly, in order to frame a knowledge productas a legitimate new organizational theory. Usedin academic writing, epistemic scripts serve anormative function that is akin to institutionalpressure. They represent shared understand-ings of knowledge production within aca-demia—that is, the taken-for-granted assump-tions about the nature and emergence of newtheories.

We emphasize this normative dimension ofacademic knowledge production, which coexistswith the technical realm of theory production,including construct definition, hypothesis test-ing, explanatory power, and prediction. The co-existence of a normative realm and a technicalrealm of knowledge production is made possi-ble by decoupling (Meyer & Rowan, 1977). Wepropose that the epistemic scripts used to gen-erate a new theory are often decoupled fromthose employed to present this theory in aca-demic writing. More precisely, we propose thatscholars frequently use the script of bricolage toassemble various building blocks into new or-ganizational theories. This assembly, we argue,is subsequently made invisible in academicwriting when scholars—more or less conscious-ly—adopt other scripts to enhance the academiclegitimacy of their new knowledge product.

We illustrate this decoupling in relation toorganizational institutionalism, an organization-al theory whose foundational texts were writtensome thirty years ago and which has sincegained dominance within the field of organiza-tional theory (Davis, 2010). The authors of thesetexts clearly draw on the scripts of evolution anddifferentiation to present the theory in academicwriting, yet the theory itself contains an assem-bly of metaphors suggesting the extensive useof bricolage during its conception. We attributethis apparent decoupling to different levels oflegitimacy attributed to the three epistemicscripts of knowledge production within the fieldof organizational theory. We visually depict ourargument in the form of an integrative model oforganizational knowledge production, whichemphasizes epistemic scripts as an implicitstructure that governs how new organizationaltheories are produced.

Our argument carries some potential implica-tions for the emergence of new organizationaltheories. Notably, we think that greater insightinto how metaphors and scripts enable scholarsto generate new knowledge may stimulate the

2011 273Boxenbaum and Rouleau

conception of new organizational theories. Sim-ilarly, a better understanding of how epistemicscripts operate in academic writing may enablescholars to reinforce their proposal for a neworganizational theory. Finally, a greater com-prehension of the apparent decoupling betweenthe epistemic scripts used for conceiving neworganizational theories and the epistemicscripts used for presenting them to an academicaudience may renew and stimulate organiza-tional knowledge production. It may open newperspectives for organizational theory builders,just as it may encourage editors and reviewersof prestigious international journals to reflect ontheir role in perpetuating certain institutional-ized scripts of knowledge production whileblocking the diffusion of others. Perhaps implicitpreferences for certain epistemic scripts ofknowledge production contribute to the low ra-tio of new organizational theories that haveemerged since the 1970s (Davis 2010).

In this article we first examine how metaphorsand scripts shape organizational knowledgeproduction, and we subsequently illustrateour argument using foundational texts fromorganizational institutionalism as an exam-ple. The different components of our argumentare then assembled into an integrative modelof organizational knowledge production,which we place in relation to previous litera-ture on the topic. The article concludes with adiscussion of implications for the field of or-ganizational theory.

BUILDING BLOCKS OFORGANIZATIONAL THEORIES

Organizational theory refers to the subset ofsocial science theories addressing organization-al phenomena. Theory has been defined as “anordered set of assertions about a generic behav-ior or structure assumed to hold throughout asignificantly broad range of specific instances”(Sutherland, 1975: 9). To qualify as a theory, as-sertions must form a coherent and parsimoniousframework that is sufficiently general to capturea broad range of empirical situations (Bach-arach, 1989) but precise enough to allow schol-ars to test propositions and hypotheses(Homans, 1958; McKinley, 2010). The purpose oforganizational theory is not only to offer vali-dated knowledge that can explain and predict(McKinley, 2010) but also to suggest “relation-

ships and connections that had previously notbeen suspected, relationships that change ac-tions and perspectives” (Weick, 1989: 524). Orga-nizational theories should provide new ways ofapproaching the inherent complexity and ambi-guity of organizations, supplying managerswith “what Ohmann (1955) called ‘skyhooks’—the psychological anchors that confer order onthe subjective apprehension of experience” (Ast-ley & Zammuto, 1992: 457). They should explain,predict, and delight (Weick, 1979), striking a finebalance between logical coherence and creativeconceptualization (Bacharach, 1989: 513).

Organizational theories are composed of var-ious building blocks, the most important beingempirical material, theoretical concepts, andmetaphors. Empirical material and theoreticalconcepts have received significant attention inorganizational theory, and their respective im-portance for theory building is beyond question(Colquitt & Zapata-Phelan, 2007). In contrast,metaphors represent a less recognized compo-nent of organizational theories. Their contribu-tion to theory building lies in their “generativeimpact” (Cornelissen & Kafouros, 2008a: 367),meaning that they possess a creative potentialto stimulate new perspectives on organizations(Cornelissen, 2005). The imagery conveyed bymetaphors stimulates theorists to generate in-teresting new propositions about organizationallife (Bacharach, 1989: 497) that can later besharpened into theory formulations and testablehypotheses (Cornelissen, 2005; Morgan, 1983;Soyland, 1994; Weick, 1989). Before elaboratingon the creative input of metaphors to theorybuilding, let us briefly review the other twobuilding blocks.

Empirical Material

Empirical material represents an essentialbuilding block of organizational theories (Alves-son & Karreman, 2007), especially regarding in-ductive theory building. For instance, groundedtheory stipulates a systematic recording of em-pirical observations and a rigorous analysis ofthem as the initial steps of theory building(Eisenhardt, 1989; Glaser & Strauss, 1967;Strauss & Corbin, 1994). Organizational re-searchers are also generally encouraged to sim-ulate empirical validation at every step of thetheory-building process (Weick, 1989: 516) in or-der to ensure that the emerging theory fits

274 AprilAcademy of Management Review

closely with the empirical realm (Alvesson &Karreman, 2007: 1265).

Even so, empirical material is not sufficientfor theory building. Using a conceptual lens,theory builders approach empirical materialwith some preconceived notions about organi-zational life (Alvesson, 2003; Astley, 1985; Daft,1983; Feyerabend, 1975; Weick, 1989). It has beenproposed that “we can perceive nothing exceptthrough the knowledge structure in which per-ception is embedded” (Astley, 1985: 497–498) andthat organizational phenomena are so abstractand complex that empirical material does notsuffice to understand them; conceptual catego-ries are required as well (Weick, 1989). As aconsequence, theory builders manipulate em-pirical material conceptually in some way orother to make sense of it (Alvesson, 2003; Alves-son & Karreman, 2007: 1269).

Theoretical Concepts

Theoretical concepts, the core component oforganizational theories, are abstractions of em-pirical phenomena. Concepts, when further ab-stracted and specified, become the constructsthat supply a theory with its conceptual clarityand inherent structure. Construct clarity is ex-pressed through definitions, scope conditions,semantic relationships to other constructs, andcoherence (Suddaby, 2010: 347). For that reason,it is important that scholars carefully craft andspecify theoretical concepts when they engagein theory building (Suddaby, 2010). Scholars maygenerate entirely novel concepts, borrow ortransform concepts from other organizationaltheories, or import them from other disciplines.It is, in fact, quite common for organizationalscholars to borrow concepts and theories fromneighboring disciplines, most notably from psy-chology and sociology (Whetten et al., 2009: 537–538). Theoretical concepts have also been bor-rowed from more distant disciplines, such asbiology (e.g., organizational survival and orga-nizational environment), literary studies (e.g.,narratives and translation), and religious stud-ies (e.g., myths and rituals). The practice of im-porting elements from other disciplines is ap-parently so prevalent in organizational theorythat it has led to a virtual trade deficit with thebasic disciplines (Ilgen & Klein, 1989).

In the process of theory building, scholars linktheoretical concepts to one another to form new

propositions about organizational life (Bach-arach, 1989). Therefore, theoretical conceptsneed to be clearly specified and firmly groundedin empirical phenomena. Furthermore, the pro-posed relationships between them should becarefully aligned with empirical observations toensure that the propositions convey not onlynew but also valid representations of organiza-tional life. Although organizational theorists areprimarily interested in theoretical concepts andempirical material in theory building, organiza-tional theories are also built of metaphors.

Metaphors

Metaphors, a core component of cognitive pro-cessing, consist of the superimposition of asource domain onto a target domain. This trans-fer process may facilitate our understanding ofthe unfamiliar and sharpen our comprehensionof an already known phenomenon (Inns, 2002;Lakoff & Johnson, 1980; Oswick, Keenoy, & Grant,2002). The reason is that perception and knowingare linked “in an interpretive process that ismetaphorically structured, allowing us to under-stand one domain of experience in terms of an-other” (Koch & Deetz, 1981). In fact, metaphorrepresents “a primal, generative process that isfundamental to the creation of human under-standing and meaning in all aspects of life”(Morgan, 1996: 228). Used as a tool for under-standing, metaphors are said to have an “ex-planatory impact” (Cornelissen & Kafouros,2008a; Weick, 1989).

Metaphors serve another function that ex-tends their applicability into the realm of cre-ativity. They possess a “heuristic quality inopening up new and multiple ways of seeing,conceptualizing, and understanding organiza-tional phenomena” (Cornelissen, 2005: 753). Thiseffect is known as a “generative impact” (Cor-nelissen & Kafouros, 2008a; Weick, 1989). Cre-ative processes draw extensively on metaphorsas a heuristic device to stimulate imagination(Lakoff & Johnson, 1980) and to generate insightsthat previously were inconceivable (Cornelissen& Kafouros, 2008a; Inns, 2002; Schon, 1993). Animportant source of creativity lies in the juxta-position of the source domain and the targetdomain (Cornelissen, 2005: 756). Juxtapositionoperates as a catalyst for new insight since“metaphor involves the conjunction of whole se-mantic domains in which a correspondence be-


tween terms or concepts is constructed, ratherthan deciphered, and the resulting image andmeaning is creative, with the features of impor-tance being emergent” (Cornelissen, 2005: 751).In other words, the mapping of one domain ontoanother is an ambiguous act that requires imag-ination and stimulates the creation of newmeaning (Van Maanen, 1995); it is herein that wefind the source of creativity.

Cognitive transfer processes tend to operateimplicitly. We do not need to be aware of ourcognitive processes in order to understand theunfamiliar or to exercise creativity. We are,however, able to deliberately increase ourawareness of cognitive transfer processes bymeans of introspection and research. It is furtherpossible to deliberately stimulate cognitivetransfer processes in order to speed up learningor to enhance creativity. Just as metaphors maybe used proactively for pedagogical purposes,so, too, they may be consciously employed togenerate new insights (Gioia & Poole, 1984;Goffman, 1959; Weick, 1989). Although we canincrease our understanding of how metaphorsoperate and can exploit some of this insight inpractice, it is evident that even extensive effortwill necessarily fall short of any ambition tocontrol cognitive transfer processes. Creativity,in particular, escapes full-fledged formalization.

The cognitive processes associated with met-aphors apply directly to the conception of neworganizational theories. As vehicles of sense-making, metaphors operate as creative cata-lysts in organizational theory building. Theyembody images that stimulate the imaginationand enable theorists to generate novel perspec-tives on organizational life. They can do so in atleast two ways. First, scholars may activate met-aphors subconsciously in the process of analyz-ing complex qualitative data, a process thatnecessarily involves an uncodifiable creativeleap, however small (Langley, 1999). Second,scholars may use metaphors in their cognitiveexperiments with different perspectives on or-ganizational life. Through “disciplined imagina-tion” (Weick, 1989)—that is, systematic applica-tion of different metaphors—scholars mayemploy metaphors as a precursor to new theoryformulations. In either case the metaphors usedfor theory building are integrated into the finalknowledge product. Accordingly, the metaphorscontained within organizational theories canonly be inferred from the way organization is

approached and from the underlying assump-tions about the subject (Smircich, 1983: 341).

The generative impact of metaphors deepensour understanding of how new organizationaltheories are conceived. Although they are per-haps more common in organizational theorythan in many other disciplines (Morgan, 1983),metaphors represent a controversial componentof organizational theories (Cornelissen, 2006;Grant, 2001; Morgan, 1983; Oswick et al., 2002;Oswick et al., 2003; Pinder & Bourgeois, 1982).Some organizational theorists have argued thatmetaphors should be eliminated from organiza-tional theories because scientific knowledge isfundamentally incompatible with figurativelanguage (e.g., Pinder & Bourgeois, 1982). It isincreasingly clear, however, that their elimina-tion is not possible because metaphors are es-sential for generating new insights and formu-lating new organizational theories (e.g.,Cornelissen, 2006; Morgan, 1983; Oswick et al.,2002). Instead, the quest has become one of un-derstanding how metaphors enable the concep-tion of new organizational theories (Bacharach,1989; Cornelissen, 2005; Weick, 1989).

Early research on metaphors in theory build-ing proposed that organizational theories arebuilt around a root metaphor that renders eachtheory unique and coherent. Root metaphorshave been defined as “the dominant or definingway of seeing” (Inns, 2002: 309) or as “a funda-mental image of the world on which one is fo-cusing” (Alvesson, 1993: 116). One of the firstcontributions to this line of inquiry consisted ofa typology that grouped organizational theoriesaccording to their root metaphor (Burrell & Mor-gan, 1979), each root metaphor providing aslightly different understanding of organiza-tions (Morgan, 1983; Smircich, 1983). For exam-ple, a root metaphor of evolution highlights thatorganizations must adapt to their environmentin order to survive, whereas a computationalroot metaphor casts organizations as consistingof interdependent and interconnected parts(Cornelissen & Kafouros, 2008b: 969).

In subsequent research scholars have in-creased the complexity by proposing that orga-nizational theories are composed of primary andsecondary metaphors. Alvesson (1993) suggeststhat secondary metaphors alter the meaning ofthe primary metaphor. For instance, the organi-zation-as-machine metaphor, introduced alongwith scientific management in the early part of


the twentieth century, took on a new meaningwhen it was combined in 1955 with the meta-phors of systemic electrical engineering andcomputer science (Barley & Kunda, 1992; Corne-lissen & Kafouros, 2008b). This increased meta-phorical complexity has added nuance to thisline of inquiry.

The most recent work on metaphors in theorybuilding suggests an even higher level of com-plexity in the metaphorical composition of or-ganizational theories. Empirical research hasshown organizational theories to be composedof multiple metaphors that are combined intofairly complex metaphorical constructs (An-driessen & Gubbins, 2009). Scholars apparentlyrecombine small metaphorical components thatare familiar to organizational theorists (Corne-lissen, 2005) into complex metaphorical con-structs that are temporary and fluid in nature(Cornelissen & Kafouros, 2008b: 957). In this pro-cess theory builders select living metaphorsthat are embedded in a particular sociohistori-cal time and used by a historically developedlanguage community (Tsoukas & Knudsen, 2003:8). The metaphorical construction of a theoreti-cal concept evolves over time, as Andriessenand Gubbins (2009) document in their empiricalanalysis of the notion of social capital. Theyidentified seven conceptual metaphors in thistheoretical concept, some of which were intro-duced in the original publications, while otherswere added later. These recent contributionsadd both complexity and contextual compo-nents to this line of inquiry.

Integration of Building Blocks

Organizational scholars have also begun tostudy the integration of theoretical concepts,empirical material, and metaphors into new or-ganizational theories. Although the integrationof theoretical concepts and empirical material isa widely recognized feature of theory building,many organizational scholars hesitate to inte-grate metaphors in the false belief that meta-phors “contaminate” the scientific endeavor(Pinder & Bourgeois, 1982). As numerous schol-ars have nonetheless documented, theoreticalconcepts are often intertwined with metaphori-cal meaning (Andriessen & Gubbins, 2009; Cor-nelissen & Kafouros, 2008a; Weick, 1989). To ex-emplify, the theoretical term organizationalenvironment draws its metaphorical meaning

from the field of biology, organizational struc-ture from the domain of construction, organiza-tional narratives from literary studies, socialcapital from the study of finance, and organiza-tional culture from the discipline of anthropol-ogy. Such an integration of metaphors and the-oretical concepts is also evident in thetheoretical constructs of “strong and weak ties”(Granovetter, 1985), “organizational identity”(Albert & Whetten, 1985), “organizational learn-ing” (Argyris & Schon, 1978), “loose coupling”(Weick, 1976), “structural holes” (Burt, 1992), and“the invisible hand of the market” (Smith, 1759).These theoretical concepts imply that organiza-tional phenomena are like people (e.g., havingan identity or hands) or like physical entities(e.g., having holes or ties).

A relationship also exists between empiricalmaterial and metaphors. Alvesson (2003) hasshown, for instance, how the metaphors thatguide researchers’ implicit understanding of theinterview affect the empirical data they collect.In light of this increasingly refined literature, itseems somewhat surprising that metaphorscontinue to be a relatively neglected componentof organizational theories (Cornelissen, 2005:753). We propose that this neglect is sustainedby an inattention to epistemic scripts, notably tothe decoupling that exists between the scriptsused to conceive a new organizational theoryand the scripts that scholars subsequentlyevoke in their presentation of the new theory toa scholarly audience. Before turning to an illus-trative example, we examine how epistemicscripts are rhetorically used in knowledge pro-duction.

EPISTEMIC SCRIPTS AND THEIRRHETORICAL USE

Organizational scholars develop new theorieswithin a relatively fixed set of conventions andbehavioral patterns that are known as scripts ofknowledge production (Alpaslan, Babb, Green,& Mitroff, 2006; Donaldson, 2009; Partington &Jenkins, 2007). The notion of script contains met-aphorical elements from the source domain oftheater (Cornelissen, 2005; Goffman, 1959), nota-bly the idea of a stage and of scripted dialogueand behavior. Scripts provide templates for boththe conception and presentation of new aca-demic knowledge, which collectively make upwhat we refer to in this article as knowledge


production. Epistemic scripts are relatively per-sistent cognitive structures that prescribe howtheories should be formulated within the aca-demic tradition (DiMaggio, 1997; Schank & Abel-son, 1977). Anchored in an epistemological tra-dition that embodies assumed representationsof how new knowledge emerges, epistemicscripts capture shared assumptions about theproduction of new knowledge. They implicitlyinform the conception of new theories and, per-haps more explicitly so, the presentation of newtheories to the scholarly community.

When presenting a new theory formulation inwriting, organizational scholars use scripts torelate new knowledge to existing knowledge

(Beyer et al., 1995; Locke & Golden-Biddle, 1997;Mone & McKinley, 1993). Epistemic scripts helporganizational scholars assemble a set of ideasthat either directly or indirectly conveys to aca-demic peers that they are in the presence of anew knowledge product (Donaldson, 2009). Es-sential to knowledge production, epistemicscripts embody the frame of reference that qual-ifies a knowledge product as new. In contempo-rary organizational theory, we find three epis-temic scripts that academics employ to conceiveand represent new organizational knowledge.They are evolution, differentiation, and brico-lage, visually depicted in Figure 1. Each labelcorresponds to a key feature that sets one epis-

FIGURE 1Epistemic Scripts of Organizational Knowledge Production

Evolution Time/complete knowledge Differentiation Paradigm X Paradigm Y

Time/political situations Bricolage

Organizational theory A Organizational theory B

Note: The size of the squares in the differentiation scripts refers to the political influence of that academic community.Different shades in the bricolage script represent different knowledge components.


temic script apart from the others. They repre-sent ideal types that scholars may draw on si-multaneously during the act of knowledgeproduction.

Evolution

Organizational theorists who adhere to theepistemic script of evolution generally positionthemselves as partaking in the collective pur-suit of advancing the frontier of objective knowl-edge about organizations and organizationallife (Donaldson, 1995, 2005; Pfeffer, 1993). Thisscript of evolution refers to the most widely ac-cepted template in science—namely, thatknowledge evolves through trial and error to-ward an increasingly accurate representation ofthe world. According to the script of evolution,the academic pursuit par excellence is to bringcollective knowledge one step closer to objec-tive reality.

This epistemological position tends to alignwith the philosophy of science as developed byKarl Popper (1959). Knowledge becomes morecomplete and accurate as new insights areadded to prior understandings or when falsestatements are rejected on the basis of logicaldeduction or systematic empirical inquiry. Pop-per advanced the notion of falsification as a keyprinciple of scientific inquiry, by which hemeant that new knowledge claims can advancecollective knowledge only if they are formulatedin falsifiable terms and are subjected to empir-ical testing. After rigorous efforts have beenmade to that effect, only propositions and hy-potheses that are not falsified can qualify asobjective new knowledge. Many organizationalscholars endorse theory testing as a particularlyvalid way to ensure the objective advancementof knowledge (Colquitt & Zapata-Phelan, 2007),or at least to reach a consensus on a set ofrelationships describing and explaining the ob-jective world (Bacharach, 1989; McKinley, 2010).The principle of falsification is closely associ-ated with the epistemic script of evolution, al-though the latter is often used in theory devel-opment without an attempt to falsify theproposed theory, a trend that McKinley (2010)deplores as increasingly common in organiza-tional theory.

When used to generate new knowledge, theevolution script guides researchers to explorepossible improvements to an existing theory.

Such improvements could take the form of amore precise or parsimonious theory formula-tion, the identification of conditions under whichthe theory does not apply, or the discovery of anew intervening variable. In the pursuit of suchincremental improvements to an establishedtheory, researchers may inadvertently generateknowledge that could be developed into a newtheory.

The epistemic script of evolution serves notonly to generate new knowledge but also topresent new knowledge claims as being contin-uous with previous knowledge. The rhetoricaluse of this script refers to the ways that authorsstructure the coherence between their proposednew knowledge product and existing knowl-edge. For instance, in a detailed analysis of howtheoretical contributions are textually con-structed, Locke and Golden-Biddle (1997) identi-fied three rhetorical strategies that organization-al scholars use to position their new knowledgeproduct in continuity with existing knowledge—namely, synthesizing coherence, progressive co-herence, and merging conflicting coherence.These rhetorical strategies can be seen as dif-ferent manifestations of the script of evolution,which scholars apply more or less consciouslyto represent their knowledge product as newand unique along an existing line of inquiry.The evolution script also positions the proposednew knowledge product as superior to previousknowledge—that is, as one step of advancementtoward a more accurate or better adapted under-standing of organizations. Assuming that thesubstance of the theory formulation is sound, thescript of evolution, applied appropriately, helpsto convince readers, including editors and re-viewers of scholarly journals, that the proposedknowledge product advances the frontier ofknowledge. In fact, this script seems to be usefulto the conception of new organizational knowl-edge as well as to the scholarly presentation ofsuch knowledge.

Differentiation

The second epistemic script, differentiation,is founded on the premise that there is noshared frame of reference that can encompassall academic knowledge with regard to orga-nizations. Instead, there are differentiated per-spectives that coexist within organizationaltheory and that cannot be ordered hierarchi-


cally (Burrell & Morgan, 1979; Scherer & Stein-mann, 1999). A radical shift in perspective onorganizational life qualifies as a manifesta-tion of the differentiation script. Organiza-tional knowledge is seen as inescapably em-bedded in academic knowledge communitieswhose relative size, complexity, and politicalinfluence change over time (Davies & Fitchett,2005; Hassard & Kelemen, 2002).

The epistemic script of differentiation reflectsthe philosophy of science associated withThomas Kuhn (1962). A core tenet of Kuhn’s workis that the practice of knowledge production un-folds within an academic community thatshares a set of epistemological principles andontological assumptions about the world, whichcollectively form what is known as a paradigm.A paradigm refers to a knowledge base that ismutually incommensurate with that of anotherparadigm, meaning that their core assumptionsabout the world and their modes of knowledgeproduction cannot be integrated with one an-other (Scherer & Steinmann, 1999). Since there isno metaposition from which to compare para-digms and to assess the knowledge generatedwithin them, academic knowledge is inescap-ably tied to the paradigm within which itemerged and, by extension, is relatively im-mune to external contestation (De Cock &Jeanes, 2006; Reed, 1996).

New paradigms supposedly arise in the wakeof a scientific revolution—that is, when the set ofprinciples and assumptions underpinning anexisting paradigm becomes subject to internalscrutiny (Kuhn, 1962). According to Kuhn, thissituation occurs when the results present anabnormality that cannot be resolved withinthe parameters of the existing paradigm. Thisabnormality can then trigger new discoveriesthat provoke the emergence of a new theoryand, eventually, the formation of a new para-digm. As such, novelty occurs inadvertently aspart of a game that is governed by a certainset of rules but whose acceptance requires theelaboration of a new set of rules (Kuhn, 1962:52–53). Time is needed for such divergent in-sights to be recognized and for a correspond-ing new paradigm to take form, a lengthy pro-cess that unfolds outside of the periods knownas “normal science.”

Applied to knowledge conception, the differ-entiation script suggests that researchers ac-tively seek to generate knowledge that is dis-

continuous with existing knowledge. Forinstance, the inductive model of theory buildingknown as grounded theory seems to take itsinspiration from this script. It encourages theo-rists to set aside their assumptions to facilitatethe emergence of new knowledge, a request thatis equivalent to asking researchers to simulatethe conditions that prevail during a scientificrevolution. However, the ability of researchers toactually do so remains somewhat doubtful(Suddaby, 2006). The differentiation script maywell inform the ambition of theory builders, butit does not offer much guidance in terms of howto conceive new theory, aside from encouragingresearchers to approach empirical reality withas little previous knowledge and as few as-sumptions as possible.

When used in knowledge presentation, thedifferentiation script is particularly evocativeas a way of casting a new knowledge productas a distinct new organizational theory. Whenorganizational scholars present their knowl-edge claims as being counterintuitive to trig-ger the interest of readers, they implicitlydraw on the script of differentiation. This isalso the case when they claim to fill a gap inthe literature without extending an existingline of inquiry. According to Locke and Golden-Biddle (1997), organizational scholars draw onthree rhetorical strategies—incompleteness,inadequacy, and incommensurability—to callinto question the existing literature and to al-lude to mistakes in predominant understand-ings within their field of study. In presentingresults that prevailing theories cannot ex-plain, authors set the stage for introducingtheir knowledge claim as the formation of anew theory. In other words, organizationalscholars draw rhetorically on the differentia-tion script to convince readers that their newknowledge product cannot be containedwithin the current stock of organizationalknowledge and that it therefore qualifies as anew theory.

Bricolage

The third epistemic script of knowledge pro-duction, bricolage, refers to an assembly ofreadily available elements. Borrowed from Levi-Strauss’ (1966) seminal work, The Savage Mind,the term bricolage has become increasinglyprevalent in organizational theory (Duymedjian


& Ruling, 2010).1 Levi-Strauss introduced brico-lage as a way of distinguishing mythical knowl-edge from scientific knowledge without neces-sarily ranking them. In his work the notion ofbricolage refers to the way indigenous popula-tions create mythical knowledge from a closeand intimate understanding of the world sur-rounding them. They combine readily availableelements into new representations of the world,a form of knowledge creation that differs fromthe generation of scientific knowledge as prac-ticed in the Western world. In organizationaltheory bricolage therefore refers to the assemblyof different knowledge elements that are readilyavailable to the researcher. They may have theirorigin in various academic disciplines, organi-zational practices, and/or a wider social context.Notably, organizational researchers have com-bined diverse strands of literature, methodolog-ical components, various pieces of theory, andmetaphors to generate new knowledge (Barker,Nancarrow, & Spackman, 2001; Gabriel, 2002;Glynn, Barr, & Dacin, 2000; Marble, 2000; Reed,1996).

Applied to conception, the script of bricolageinvites scholars to produce new knowledgethrough improvisation rather than through ad-herence to a specific theory, method, or para-digm (Bryant & Lasky, 2007). The script of brico-lage casts the researcher as a “bricoleur”—a“flexible and responsive” agent willing “to de-ploy whatever research strategies, methods orempirical materials are at hand, to get the jobdone” (Denzin & Lincoln, 1994: 2). The researcheracts as a handyperson who, rather than invent-ing a new theory or new paradigm, repairs orremodels existing theories by combining vari-ous theoretical concepts, ideas, and observa-tions at his or her immediate disposal. Compo-nents are selected based on contextual factors,such as local constraints on knowledge produc-tion, practical value, and their potential for gen-erating novel insights.2 Using common sense,

the researcher identifies similarities andcomplementarities among the selected ele-ments and integrates them creatively to providecoherent, robust, and actionable insight into anorganizational dilemma. This reinterpretation ofexisting knowledge, mediated by the script ofbricolage, constitutes the source of new knowl-edge.

The epistemic script of bricolage frames or-ganizational theories as fluid constructs that un-dergo transformation in response to contextualfactors. Theories are generated for a specificpurpose, often in response to a particular orga-nizational dilemma, and are later transformedin response to new purposes and contexts. Con-sequently, the frame of reference for assessingknowledge claims is rather more contextualthan fixed (Lowe, Moore, & Carr, 2007). Knowl-edge products are valuable if they, in addition tobeing coherent and robust, respond to the situ-ation in a given place at a particular moment intime.

Used in theory presentation, this script casts anew knowledge product as the assembly of rel-evant elements from the surrounding environ-ment. Such a presentation represents a contro-versial script of knowledge production inorganizational theory. Some organizational the-orists advocate the bricolage script (e.g., Baker &Nelson, 2005; Barker et al., 2001; Duymedjian &Ruling, 2010; Freeman, 2007; Gabriel, 2002), argu-ing, for instance, that it is particularly appropri-ate for organizational realities that have beeninsufficiently studied and for the definition orredefinition of a research object (Barker et al.,2001). Other organizational scholars object to theloose and fluid combination of heterogeneouselements, refusing to acknowledge this post-modern template as an academic mode ofknowledge production (Avital & Germonprez,2004; Pfeffer, 1993; Scherer, 1998; Tadajewski,2009). Since this script is controversial in orga-nizational theory, it is rarely used for the aca-demic presentation of new theories. Yet, as weargue below, it may be rather common in theconception of new organizational theories.

1 Fifty-two articles made reference to bricolage in thescholarly, peer-reviewed journals listed in the database ABI/INFORM between 1992 and 2009, 87 percent of which werepublished after the year 2000.

2 De Certeau (1992), concerned with the invention of dailylife and its tactical features, describes bricolage as a scien-tific activity that resembles a collage of tricks and tacticsfound in laboratories and scientific practice. The aim is tojuxtapose theoretical ambitions rather than to articulate

their foundations. Nevertheless, reflection seeks first andforemost to describe multiple expressions of daily life ratherthan to reflect on the production of knowledge.


Comparison of the Epistemic Scripts

The three epistemic scripts embody differentepistemological positions, but they also differ inother ways. Most important, they are not equallyapplicable to the conception and presentation ofnew organizational theories. For example, thescripts of evolution and bricolage lend them-selves more easily to the conception of newknowledge than does the script of differentia-tion. The former scripts reemploy categories andconcepts that are readily available to the re-searcher, whereas the latter script requires jux-taposition with existing knowledge. As a result,it is difficult for organizational scholars to usethe differentiation script to purposefully gener-ate new organizational theories.

The three scripts partially overlap with oneanother in theory production. While vague in hisdescription of the formation of theories and par-adigms, Kuhn (1962) described the generation ofnew knowledge during scientific revolutions ina way that seems to be identical to the script ofbricolage. In the pursuit of new rules or newtheories, and without a fixed point of reference,researchers resort to what they already knowand can observe. Using familiar elements, theyformulate speculative and imprecise theoriesthat gradually bring them onto the path to newdiscoveries and the generation of new knowl-edge. This account of knowledge production pro-vokes the question of whether it is possible todistinguish between the script of bricolage andthe script of differentiation when it comes toknowledge conception. We maintain that it is.The script of bricolage prescribes the combina-tion of heterogeneous elements into fluid knowl-edge products that acquire value from their us-age. In contrast, the script of differentiationleads to the formation of a new coherent andrelatively stable system of thought whose valueis determined by rules and assumptions. Ratherthan an overlap between the two scripts ofknowledge production, we propose that decou-pling takes place—that is, that different scriptsmay inform the conception and presentation ofnew knowledge. A plausible explanation for thisdecoupling, we argue, is that the three scriptscarry different levels of legitimacy within theacademic community.

An important point we wish to make here isthat the epistemic script of bricolage is not suf-ficiently legitimate to be used rhetorically in the

academic presentation of new organizationaltheories. However, this script appears to be usedmore frequently in the generation of new orga-nizational knowledge than what the academicliterature would lead us to believe. Our conten-tion is that bricolage is a common way to gen-erate new organizational theories, regardless ofwhich script is subsequently evoked to present anew theory to an academic audience. In the il-lustrative example below, we demonstrate theapparent use of bricolage in theory conception,notably of the often neglected element of meta-phors, and examine its decoupling from thescripts that serve to present the theory.

AN ILLUSTRATIVE EXAMPLE:ORGANIZATIONAL INSTITUTIONALISM

Organizational institutionalism refers to abranch of institutional theory that has prolifer-ated within organizational theory for the pastthirty years (Greenwood, Oliver, Sahlin, &Suddaby, 2008). Firmly established as a distinctorganizational theory, organizational institu-tionalism was developed entirely within theworld of academia and is now the most preva-lent theory in the field of organizational theory(Davis, 2010). In fact, institutional theoryemerged as the most frequently used key wordin paper submissions to the Division of Organi-zation and Management Theory (OMT) prior tothe Academy of Management’s annual meetingin 2005: in more than 25 percent of the paperssubmitted, authors included it as one of threekey words (Davis, 2010: 693; Greenwood et al.,2008: 2). As we show in the analysis below, thescrutiny of the metaphorical composition of thistheory suggests that the script of bricolage mayhave been used in its conception, although itspresentation is cast in the epistemic scripts ofevolution and differentiation.

Analytical Procedures

The data consisted of a small number of textsrepresenting the early formulations of organiza-tional institutionalism (also known as neoinsti-tutional theory). We took inspiration from the listof early works cited in the introduction to theThe Sage Handbook of Organizational Institu-tionalism (Greenwood et al., 2008) and used ci-tation scores to identify the most influentialpublications during the first decade of the theo-


ry’s existence (1977–1987). The Social SciencesCitation Index� helped identify journal articlesand books, while Google Scholar pointed to themost cited book chapters in a given book. Fivefoundational texts emerged as the most influen-tial works from the first decade:

1. DiMaggio, P. J., & Powell, W. W. 1983. Theiron cage revisited: Institutional isomor-phism and collective rationality in organi-zational fields. American Sociological Re-view, 48: 147–160. SSCI citations: 3,220.

2. Meyer, J. W., & Rowan, B. 1977. Institutional-ized organizations: Formal structure asmyth and ceremony. American Journal ofSociology, 83: 340–363. SSCI citations: 2,569.

3. Scott, W. R., & Meyer, J. W. 1983. The orga-nization of societal sectors. In J. W. Meyer &W. R. Scott (Eds.), Organizational environ-ments: Ritual and rationality: 129–153. Bev-erly Hills, CA: Sage. SSCI citations for book:562 (and most cited chapter according toGoogle Scholar).

4. Scott, W. R. 1987. The adolescence of insti-tutional theory. Administrative ScienceQuarterly, 32: 493–511. SSCI citations: 442.

5. Tolbert, P. S., & Zucker, L. G. 1983. Institu-tional sources of change in the formal struc-ture of organizations: The diffusion of civilservice reform, 1880–1935. AdministrativeScience Quarterly, 28: 22–39. SSCI citations:430.

We conducted two textual analyses: one of themetaphorical elements and another of the epis-temic scripts used in the presentation of thetheory. Space did not permit for a fine-grainedanalysis of empirical material and theoreticalconstructs, which have already been exten-sively investigated in the production of newknowledge. In the first analysis, which was in-spired by Andriessen and Gubbins (2009), weselected all words whose literal meaning be-longs to a domain other than organizations. Wethen grouped them into source domains (see Ta-ble 1), eliminating words that did not appear atleast ten times in one foundational text. In thesecond analysis, which took its inspiration fromLocke and Golden-Biddle (1997), we identifiedall passages explicitly positioning the newknowledge product in relation to previousknowledge. We grouped the passages accordingto the three epistemic scripts portrayed in Fig-ure 1, and we present selected illustrative exam-ples from the exhaustive list of script-relatedpassages.

Metaphorical Building Blocks

Organizational institutionalism, as reflectedin the foundational texts, contains a wide rangeof metaphors that span the metaphorical do-mains of culture, construction, biology, religion,theater, systems, market, power, and hard sci-ences. Table 1 shows the metaphorical wordsthat appear at least ten times in one founda-tional text, indicating also how many times eachword appears in each text. As the table reveals,some metaphors are widely used in all the texts(e.g., structure and environment), whereas oth-ers appear frequently in one or two texts andrarely or never in others (e.g., myths, ceremony,beliefs, and isomorphism).

Despite the variety of metaphorical domainsevoked in the conception of organizational insti-tutionalism, none of them were new to organi-zational theory. It appears that the reuse of met-aphors from other organizational theoriescreated a bridge that facilitated the adoption oftheoretical concepts from these theories. For in-stance, the metaphorical domain of biology con-nected the new theory to the work of Selznickfrom the 1950s. Decades before organizationalinstitutionalism referred to environment, sur-vival, and variation/selection, Selznick had por-trayed organizations as living, organic formsthat adapt to their environment in order to sur-vive:

Selznick’s institutional approach also empha-sized the importance of history—the “natural his-tory” of the evolution of a living form that isadaptively changing over time, and he stressed aholistic and contextual approach. As Perrow(1986, 157–158) noted: For institutional analysis,the injunction is to analyze the whole organiza-tion. To see it as whole is to do justice to its“organic” character. Specific processes are, ofcourse, analyzed in detail, but it is the nesting ofthese processes into the whole that gives themmeaning (Scott, 1987: 494; emphasis added).

A second example of bridging relates to thenotion of structure from the metaphorical do-main of construction. The concept of organiza-tional structure took its inspiration from thelater work of Max Weber (1949). References tobuildings also existed in organizational theorydecades before the birth of organizational insti-tutionalism. For instance, the metaphor of archi-tecture had been used previously to explain theintegration of multiple organizational elements,such as economic markets, coordination, infor-


mation systems, and bureaucratic function (Me-nard, 1990: 14). The metaphorical domain of cul-ture was not new either, as Smircich (1983)showed in her review of the culture metaphorwithin organizational theory. Culture also ap-peared in other organizational theories prior tothe formulation of organizational institutional-ism. In reusing a metaphor, the authors of thefoundational texts enabled the transfer of theo-retical concepts from other organizational theo-ries.

A similar metaphorical transfer process ap-plied to the borrowing of theoretical conceptsfrom adjacent disciplines, such as anthropology,

sociology, and history. By adopting metaphori-cal domains that were also used in other fieldsof study, the authors of the foundational textsenabled a transfer of theoretical concepts andpropositions from those fields into organization-al theory. This mechanism is illustrated in thefollowing passage from Scott:

With the shift in focus on symbolic aspects ofenvironments and their sources, this version ofinstitutional theory has both contributed to andbenefited from the resurgence of interest in cul-ture. Thus, this institutional theory provides abridge for students of organizations to link to theinsightful work of Berger, Bourdieu, Douglas,

TABLE 1Metaphors Appearing at Least Ten Times in a Foundational Text

Metaphor DomainDiMaggio &Powell (1983)

Meyer &Rowan (1977)

Scott &Meyer (1983)

Scott(1987)

Tolbert &Zucker (1983)

CultureMyths/mythlike 1 52 1 4 0Ceremony/ceremonial(ly)/ceremonialize(d) 6 38 0 1 0Societal/society 6 46 11 22 4Community 2 0 15 1 2

BiologyEnvironment(s)/environmental(ly) 18 66 18 34 4Survival/survive 2 21 3 4 2Field 65 1 5 5 0Selection 13 3 7 1 3Population/populace/populate 11 0 2 2 5

ConstructionBuilding/build/built 2 12 1 2 0Stabilization/stable/stabilize/stability 3 15 0 6 0Structure/structuration/infrastructure 54 112 51 70 35Fragmentation/fragmented 1 0 15 1 0

ReligionBeliefs 0 0 1 26Faith 0 13 1 0 1

TheaterActor(s) 10 1 23 1Perform/performance 7 9 10 10 8

System(De)coupled/(de)coupling 4 18 31 1 0(Sub)system(s) 9 10 39 36 10Network 7 20 6 2 1(De)centralization/(de)central(ize) 11 8 45 5 8

MarketCompetition/compete/competitive 23 6 2 5 2Product(ion)/produce(rs)/producing 7 34 3 15 1

PowerPower(ful) 16 6 1 18 6Dominance/dominant/dominating 10 2 0 1 4Enforce(ment)/force(s) 11 9 0 9 2Coercion/coerce/coercive 10 0 0 6 0

Hard sciencesIsomorphism/isomorphic 52 17 0 2 1Homogeneity/homogeneous/homogenization 27 0 0 0 0


Foucault, Geertz, and Wuthnow, to name onlysome of the leading contributors to the “new”cultural approaches (1987: 499).

As these examples illustrate, metaphors madeup an important building block of organization-al institutionalism. They facilitated the borrow-ing of ideas and theoretical concepts from pre-vious organizational theories, as well as fromother fields of study. The source of novelty laynot in the introduction of a new metaphor but inthe unique combination of metaphorical do-mains into a novel theory formulation.

Our analysis further suggests that the meta-phors employed in the conception of organiza-tional institutionalism were selected and as-sembled using the script of bricolage. Notably,authors of the foundational texts combined met-aphors from very different domains in the samesentence without justifying their pertinence ormutual compatibility. The simplest ones com-bine two metaphorical domains:

• Hard sciences and religion: “The concept ofinstitutional isomorphism is a useful tool forunderstanding the politics and ceremonythat pervade much modern organizationallife” (DiMaggio & Powell, 1983: 150; empha-sis added).

• Hard sciences and biology: “Institutionalisomorphism promotes the success and sur-vival of organizations” (Meyer & Rowan,1977: 349; emphasis added).

• Culture and construction: “Organizationsfail when they deviate from the prescrip-tions of institutionalizing myths: quite apartfrom technical efficiency, organizationswhich innovate in important structural waysbear considerable costs in legitimacy”(Meyer & Rowan, 1977: 352–353; emphasisadded).

More complex forms of bricolage are also evi-dent in the five texts. Numerous passages con-tain metaphors drawn from three or four meta-phorical domains that are combined throughbricolage:

• Construction, biology, culture, and systems:“The building blocks for organizations cometo be littered around the societal landscape;it takes only a little entrepreneurial energyto assemble them into a structure. And be-cause these building blocks are consideredproper, adequate, rational, and necessary,organizations must incorporate them toavoid illegitimacy. Thus, the myths builtinto rationalized institutional elements cre-ate the necessity, the opportunity, and theimpulse to organize rationally, over and

above pressures in this direction created bythe need to manage proximate relationalnetworks” (Meyer & Rowan, 1977: 345; em-phasis added).

• Biology, culture, and construction: “Institu-tional environments are often pluralistic(Udy 1970), and societies promulgate sharplyinconsistent myths. As a result, organiza-tions in search of external support and sta-bility incorporate all sorts of incompatiblestructural elements” (Meyer & Rowan, 1977:356; emphasis added).

• Power, biology, culture, and religion: “A fo-cus on institutional isomorphism can alsoadd a much needed perspective on the po-litical struggle for organizational power andsurvival that is missing from much of popu-lation ecology. The institutionalization ap-proach associated with John Meyer and hisstudents posits the importance of myths andceremony but does not ask how these mod-els arise and whose interests they initiallyserve. Explicit attention to the genesis oflegitimated models and to the definitionand elaboration of organizational fieldsshould answer this question” (DiMaggio &Powell, 1983: 157; emphasis added).

Through this complex and imaginary bricolage,the foundational texts succeed in merging mul-tiple metaphorical domains into a seeminglycoherent knowledge construct. The compositionof the theory appears to be a recombination ofmetaphors that were readily available in orga-nizational theory at the time. None of the meta-phors were novel to organizational theory, yettheir specific combination was unique to thistheory.

Interestingly, the complex assembly of meta-phors into a new knowledge product apparentlydoes not disturb readers. In fact, readers hardlyseem to notice the extensive use of metaphorsand apparent use of bricolage in the conceptionof this theory. As we demonstrate in the follow-ing section, this implicit acceptance can be at-tributed to the use of the other two epistemicscripts— evolution and differentiation—in thepresentation of the new theory.

Scripted Knowledge Production

When organizational scholars introduced orga-nizational institutionalism (or neoinstitutionaltheory) some thirty years ago, they used two epis-temic scripts—differentiation and evolution—toposition it in relation to other organizational the-ories. The differentiation script served to distin-guish the new knowledge product from other or-


ganizational theories, while the script of evolutiondemonstrated its incremental improvement rela-tive to previous organizational theories. Bothscripts appear in all five texts, although in differ-ent proportions. The early texts are dominated bya differentiation script (DiMaggio & Powell, 1983;Meyer & Rowan, 1977; Scott & Meyer, 1983),whereas the later texts favor the script of evolu-tion. All of them present existing theories as inad-equate in some way or another.

The evolution script is employed in the foun-dational texts to show continuity with a previ-ous stream of research within organizationaltheory. The two foundational texts that favor theevolution script presented the new theory as anextension of the institutional approach to thediffusion of innovations within formal structure,thus representing a departure from the contin-gency perspective on this topic. Tolbert andZucker (1983) used the same script to propose aninstitutional explanation of the civil service re-form that municipal governments adopted be-tween 1880 and 1935. They constructed coher-ence by integrating existing theories with casestudy findings to formulate new theory. The evo-lution script is also evident in Scott’s (1987) trac-ing of different versions of institutional theory,which he compared with empirical studies toreinforce the evolutionary development of orga-nizational institutionalism. He simultaneouslyrevealed the existence of a great variety ofconcepts and arguments in organizational insti-tutionalism that indicated a significant internaldifferentiation in need of resolution to ensurethe theory’s further development.

The differentiation script is used primarily inthe foundational texts to present the new knowl-edge product as being divergent from previousorganizational theories. The authors claimedthat this new knowledge product departed fromother organizational theories that were preva-lent at the time, such as contingency and opensystem theories (Meyer & Rowan, 1977), popula-tion ecology (DiMaggio & Powell, 1983; Scott &Meyer, 1983), the organizational network ap-proach (DiMaggio & Powell, 1983), and resourcedependence theory (see Scott, 1987). They main-tained that extant theories were incomplete andneeded specification. Meyer and Rowan af-firmed that prevailing theories had “neglectedan alternative Weberian source of formal struc-ture” (1977: 343), while DiMaggio and Powellfound that prevailing organizational theories

did “not present a fully adequate picture of themodern world of organizations” (1983: 150). Like-wise, Scott and Meyer opened the section enti-tled “Inadequacies of Dominant Models” bywriting that

most existing organizational paradigms deal in-adequately with the pattern of connectednessand disconnectedness among organizations.While there has been great progress during thepast decades in moving outside the formally de-fined boundaries of organizations to take accountof environmental stimulants and constraints andinterorganizational exchanges and ties, presentmodels remain inadequate in important respects(1983: 130; emphasis added).

This identification of inadequacies in existingtheoretical approaches paved the way for theauthors of the foundational texts to introduce analternative perspective, one that renewed thethen predominant view of a formal structure andof the organization-environment relationship.

The authors of these texts also drew on otherstreams of work to justify the need for the newtheory and to make it coherent. For example,Meyer and Rowan (1977: 343) emphasized theimportance of studying rationalized formalstructures by referring to a previous generationof researchers who had demonstrated that infor-mal structure is required for maintaining formalstructure (e.g., sociologists of bureaucracy, suchas Dalton and Downs-Homans) and that nonra-tional elements contribute to the implementa-tion of decisions (e.g., decision-making theoristslike March and Olsen). Scott and Meyer (1983)also anchored their work on societal sectors inprevious literature, writing that “building on theinsights derived from three bodies of work—oncommunity structure, policy implementation,and agency coordination—we argue the utilityof isolating societal sectors for analysis” (1983:129). Scott and Meyer combined these threestreams of research, imported from other fields,to generate the new organizational theory. Toprogressively build coherence into their argu-ment, the authors of the foundational texts alsointegrated previous empirical research on insti-tutional phenomena. DiMaggio and Powell(1983) integrated numerous empirical researchfindings to support their propositions and theiranalytical typology of isomorphic processes.Empirical studies by DiMaggio, Powell, Meyer,Zucker, and Tolbert are abundantly quoted inthis foundational text. Scott and Meyer (1983)


similarly proceeded to elaborate their theoreti-cal concepts and propositions regarding the or-ganization of societal sectors.

It is noteworthy that none of the texts explic-itly cast the new knowledge product in the epis-temic script of bricolage. The absence of thisscript in the presentation of the new theory be-comes all the more remarkable when we scruti-nize the metaphorical content of the new knowl-edge product: bricolage seems to have beenused extensively in the conception of the newtheory. As we discuss in a subsequent section,the preference for using the scripts of evolutionand differentiation in the five foundational textsseems to reflect their higher academic legiti-macy within the community of organizationalscholars.

TOWARD AN INTEGRATIVE MODEL OFORGANIZATIONAL THEORY PRODUCTION

While theories certainly possess a substantialtechnical component, epistemic scripts seem toalso contribute in important ways to the produc-tion of new organizational theories, as we illus-trated in the example of organizational institu-tionalism. They implicitly guide the conceptionof new organizational knowledge and shapehow such new knowledge is communicated toan academic audience. They also help establishcoherence between existing and new knowl-edge, which is required to generate new orga-nizational theories, as well as to understandand assess new knowledge claims. Our generalargument is summarized in an integrativemodel of organizational theory production (Fig-ure 2). This model emphasizes that epistemicscripts operate somewhat independently in thetwo phases of knowledge production. As shownin the example of organizational institutional-ism, the epistemic scripts that scholars use inthe conception of new theories need not be iden-tical to the ones they use for presenting them inwriting. In fact, we propose that they frequentlydiffer—that is, that decoupling is prevalent inacademic knowledge production.

Theory Production Process

According to the integrative model in Figure 2,which should be viewed as a fairly simplifiedtemplate of complex and intertwined processes,theory builders have at their disposal theory

components from the three building blocks ofempirical material, theoretical constructs, andmetaphors. These components, we argue, arerarely selected explicitly, although researchersmay add some elements deliberately for a vari-ety of pragmatic reasons. Our emphasis on met-aphors in this article does not reflect their im-portance relative to empirical material andtheoretical constructs during the conception ofnew organizational theories; all three buildingblocks are needed to build an original theory.

Theory builders creatively integrate selectedtheory components into a new theory using oneor more epistemic scripts of knowledge produc-tion. Within organizational theory, we identifiedthree epistemic scripts—evolution, differentia-tion, and bricolage—which embody different as-sumptions about academic knowledge produc-tion. They represent ideal types and are notmutually exclusively in practice. The process ofintegrating building blocks into theories usingepistemic scripts may unfold subconsciously orwith partial awareness, but it cannot be fullycodified. The output must integrate the selectedcomponents into a relatively cohesive theoryformulation that allows for empirical testing.We propose that the script of differentiation is illadapted for deliberate theory conception, as in-dicated by its shading in Figure 2. As we arguedpreviously, the differentiation script is unlikelyto facilitate theory building because of the in-herent difficulty associated with voluntarily dis-tancing oneself from existing knowledge. Whileboth the scripts of evolution and bricolage areeasier to employ deliberately, we propose thatmany new organizational theories may be con-ceived by means of the bricolage script. Theorybuilders select and combine elements that arereadily available in their surroundings, muchlike the building blocks for organizations thatlitter the societal landscape and that entrepre-neurs easily assemble into a structure (Meyer &Rowan, 1977: 345).

The new theory formulation is subsequentlypresented in writing. This process consists ofeffectively communicating the contents of thenew theory to an academic audience and con-vincing this audience that it is viewing a neworganizational theory. Scholars draw from thepool of available epistemic scripts for this pur-pose, selecting scripts they are familiar withand that they deem capable of conferring aca-demic legitimacy on their new knowledge prod-


uct. Since the scripts of evolution and differen-tiation carry a higher academic legitimacy inorganizational theory than does the script of bri-colage, scholars are likely to spontaneouslychoose one of the two former scripts. This choiceis rarely explicit, we argue, because the scriptsof evolution and differentiation have become soinstitutionalized in organizational theory thatthey are now widely regarded as objective rep-resentations of the relationship between newknowledge and existing knowledge. In contrast,

the script of bricolage is rarely employed as arhetorical device to enhance the legitimacy of anew knowledge product, as indicated by theshading in Figure 2. We contend that this scriptis not (yet) a legitimate mode of academicknowledge production and, hence, is deselectedin writing.

While a new theory formulation may be in-spiring and rhetorically legitimate, it will onlygain recognition as an organizational theory if italso reflects the empirical realm. Empirical val-

FIGURE 2Production of New Organizational Theories

Components

Conception

Presentation

Refinement

Empirical

material

Borrowed

theoretical

concepts

Metaphors

Guidance by

epistemic script of

evolution

Empirical testing

Guidance by

epistemic script of

bricolage

Rhetorical use of

epistemic script of

evolution

Rhetorical use of

epistemic script of

differentiation

Guidance by

epistemic script of

differentation

Rhetorical use of

epistemic script of

bricolage


idation serves to confirm or adjust the originaltheory formulation, a process that is generallypursued by means of subsequent empiricalstudies. Propositions can also be rejected at thisstage. McKinley (2010) maintains that since thisstage represents the very purpose of theory pro-duction, new theories should be formulated sothey enable empirical testing. The results of em-pirical testing rarely lead to the outright rejec-tion of a theory (Davis, 2010) but serve instead torefine the theory formulation. This circularity isindicated in our model with an arrow linkingempirical testing to the three building blocks oforganizational theories. New concepts, new em-pirical material, and/or new metaphors may beadded to the initial theory formulation in re-sponse to empirical testing.

Scope Conditions

The notion of metaphor has been treated insome depth elsewhere (e.g., see Cornelissen2005, 2006; Lakoff & Johnson, 1980; Morgan, 1983;Weick, 1989). Suffice it to say, therefore, that themetaphors built into organizational theories arelikely to vary over time and space. Just as cul-tural variation exists in the metaphors thatshape our implicit understandings of teamwork(Gibson & Zellmer-Bruhn, 2001), so, too, we mayanticipate that organizational scholars from dif-ferent cultures or with different disciplinarybackgrounds will spontaneously select some-what different metaphors when building neworganizational theories.

The construct of epistemic scripts is subject tosome scope conditions as well (see Suddaby,2010). We contend that the bricolage script re-flects our very ability to be creative and that ittherefore represents a universal feature of the-ory conception. It can be inhibited, however, byits academic illegitimacy—for instance, whenresearchers try to discipline their cognitive pro-cesses to reflect the script of evolution or ofdifferentiation. Since disciplined behavior islearned, full socialization into academia maywell be a counterindication for the conception ofnew organizational theories. In contrast, when itcomes to the presentation of new theories, thefrequency and rhetorical effectiveness of differ-ent epistemic scripts probably reflect predomi-nant values in specific communities of practice.

First, we anticipate that time affects whichepistemic scripts are selected for the presenta-

tion of a new organizational theory in academicwriting. The script of evolution can undoubtedlybe traced to the importation of evolutionary the-ory from biology into the social sciences, whiledifferentiation probably gained momentum as alegitimate script of knowledge production in thewake of Thomas Kuhn’s work on paradigms andscientific revolutions. Similarly, the script of bri-colage may currently be gaining legitimacy inresponse to calls for organizational theory tobecome more relevant for organizational prac-tice (Barker et al., 2001; Gibbons et al., 1994; Huff,2000; Schultz & Hatch, 2005).

Second, space impacts the use of epistemicscripts. Just as the predominance of differentorganizational theories varies from one countryto another (Guillen, 1994) so, too, may the fre-quency of epistemic scripts used to present neworganizational theories. For instance, the differ-entiation script may be most prevalent incountries where organizational theorists definetheir mission as one of developing counter-mainstream perspectives, while the evolutionscript may flourish most where organizationaltheory is equated with hard sciences.

Third, reflecting the above considerations, therelative use of the three scripts in the presenta-tion of an organizational theory should fluctuatewith the institutions that prevail in differentscholarly communities. The script of bricolagehas only recently begun to diffuse and is notyet sufficiently legitimate to convince readersthat a new organizational knowledge producthas academic merit. In contrast, organization-al scholars frequently draw on either thescript of evolution or the script of differentia-tion, or on both for that matter, to rhetoricallyjustify the merits of their proposed new knowl-edge product. The latter two scripts have ap-parently become so legitimate that they nowconstitute rational myths of academic knowl-edge production in organizational theory. Wereiterate that this normative component ofknowledge production never stands alone. Justas rational myths coexist with the technicalrealm of organizations (Meyer & Rowan, 1977),so, too, do scripts and metaphors operate inparallel with more substantial components oforganizational theories, such as construct va-lidity, hypothesis formulation, and explana-tory power.


METAPHORS AND SCRIPTS REVISITED

Our work contributes to several streams ofresearch within the field of organizational the-ory, notably the literature on metaphors in the-ory building and the literature on scripts inorganizational knowledge production. Belowwe outline our proposed contribution to thesebodies of literature and offer suggestions forfuture research.

Metaphors in Theory Building

While still controversial in organizational the-ory, metaphors are increasingly recognized fortheir creative potential in organizational theorybuilding. As outlined previously, metaphorsstimulate the imagination by juxtaposing asource domain and a target domain and by in-troducing frameworks that are external to orga-nizational theory yet familiar to organizationalscholars (Cornelissen, 2005; Schon, 1993). Ap-plied to organizational theory building, meta-phors offer new perspectives on organizationsthrough “disciplined imagination” (Weick, 1989)and related systematic processes of imagina-tion (Cornelissen, 2006). In principle, such shiftsin perspective can also be accomplished bychanging methodological or analytical mind-sets (Zyphur, 2009), by engaging in paradigminterplay (Schultz & Hatch, 1996), or by takingadvantage of contradictory theoretical explana-tions of the same phenomenon (Poole & Van deVen, 1989). But metaphors reach wider. Theymake it possible to creatively combine multipleperspectives and to envision entirely new con-ceptual landscapes.

As revealed by the illustration from organiza-tional institutionalism, the range of metaphori-cal domains from which organizational scholarsdraw creatively to build new organizational the-ories can be very extensive indeed. Creativetheory formulations, we argue, draw extensivelyon metaphors from various domains and inte-grate them skillfully with theoretical constructsand empirical material to form imaginative newperspectives on organizational life. These per-spectives are subsequently sharpened and mod-ified by means of systematic empirical inquiry.This proposition resonates with a recent empir-ical study of metaphors in the theoretical con-struct of “social capital” (Andriessen & Gubbins,2009) and with other studies of metaphors in

organizational theory building (e.g., Cornelis-sen, 2005, 2006; Cornelissen & Kafouros, 2008a,b;Weick, 1989, 1995). This previous work suggeststhat the assembly of metaphors is inherent totheory conception.

Our contribution does not lie in metaphoricalassembly per se but, rather, in an explanation ofwhy its existence is not yet widely recognized.Essentially, the assembly of metaphors is hid-den, we argue, behind more legitimate scripts ofknowledge production, which scholars apply toincrease the recognition of their knowledgeproduct as a new organizational theory. In otherwords, institutional forces of academic knowl-edge production, which are reflected in the epis-temic scripts that scholars use to present a newtheory, prevent a wide-scale recognition of themetaphorical assembly that goes into the con-ception of a new theory.

Future research on metaphors in theory build-ing could investigate whether the assembly ofmetaphors that we identified in the formulationof organizational institutionalism is indeed aswidespread in organizational theory as we sug-gest. It would be interesting to conduct a similaranalysis of both earlier and later organizationaltheories, such as Herbert Simon’s work on orga-nizational behavior (e.g., Simon, 1957) and BrunoLatour’s formulation of actor-network theory(e.g., Latour, 1987). Such a diachronic analysismay sharpen the scope conditions of our argu-ment, illuminating the extent to which the selec-tion and the assembly of metaphors are contextdependent. We propose that the particularchoice of metaphors is contextual, while the as-sembly of heterogeneous metaphors in theorybuilding reflects universal mechanisms of cog-nitive processing associated with our very abil-ity to be creative.

Epistemic Scripts

Our work on the epistemic scripts of knowl-edge production extends the work of Locke andGolden-Biddle (1997) on the construction of nov-elty in academic writing. They showed how or-ganizational scholars employ narrative tem-plates as rhetorical devices to craft a novelcontribution to the literature, constructing boththe knowledge product and the gap that it fills.While they examined novel contributions to ex-isting lines of inquiry, we focused on entirelynew organizational theories. We also added the


dimension of knowledge conception, whereasthey exclusively focused on the rhetorical pre-sentation of new knowledge claims. Finally,while they addressed the textual expression ofnew knowledge claims, we exposed the epis-temic scripts of knowledge production that un-derpin these textual expressions. The notion ofepistemic scripts can explain, for instance, whyLocke and Golden-Biddle (1997) found certainrhetorical templates and not others in their em-pirical analysis of academic writing. Our worksuggests that the explanation lies in the aca-demic legitimacy that epistemic scripts conferon new knowledge claims. Operating as institu-tional forces of academic knowledge production,epistemic scripts place new knowledge claimsin a socially constructed relationship with exist-ing knowledge and influence the extent towhich academic readers will recognize suchknowledge claims as a theoretical contributionor an entirely new theory.

We illuminated the epistemic principles thatunderpin each of the three scripts and presentedan integrative model of how they affect organi-zational knowledge production. We further sug-gest that it is not only feasible, as we saw in theillustrative example, but also rhetorically ad-vantageous for scholars to employ several epis-temic scripts simultaneously when presenting anew theory to a potentially heterogeneous audi-ence. This rhetorical multiplicity may increasethe likelihood that a wide community of organi-zational scholars will understand and endorsethe new theory.

Further research is certainly needed on epis-temic scripts. Scholars may examine in moredepth how the different scripts guide theory con-ception. Following previous efforts to study co-herence (Lockwood, 1976; Okhuysen & Bonardi,2011), scholars may also formulate criteria forassessing whether a specific knowledge prod-uct succeeds or fails in achieving coherence ac-cording to the principles that we outlined foreach script. It would be valuable to specify howrigor and coherence of a new theory formulationare achieved depending on the script(s) used.This work is particularly important for the un-derspecified script of bricolage, which other-wise may not diffuse and gain academic legiti-macy. Okhuysen and Bonardi (2011) take someinteresting steps in the right direction, propos-ing conceptual proximity and compatible as-

sumptions as relevant criteria for bricolage. An-other interesting path for future research wouldbe to conduct a synchronic analysis of epistemicscripts in other academic disciplines. Do schol-ars from other academic fields use the sameepistemic scripts to conceive and present newtheories? If not, does the explanation lie in thetype of inquiry within a given field of study, thetiming of an academic field’s formation, or cur-rent institutional forces within the field? Thisline of inquiry may stimulate interesting cross-disciplinary debates.

CONCLUSION

The slow emergence of new organizationaltheories in recent years has sparked an acuteinterest in understanding how new organization-al theories are produced. Organizational schol-ars have made significant advances in terms ofilluminating either the conception or presenta-tion of new theories, but they have left aside therelationship between these two dimensions oftheory production. Although many scholars mayexperience them as being intertwined, these twodimensions represent somewhat distinct ele-ments of organizational knowledge productionthat merit analytical separation. Our worksheds light on the epistemological assumptionsthat underpin each of these two dimensions oforganizational knowledge production. These as-sumptions, we argue, are reflected in epistemicscripts of knowledge production that scholarsemploy implicitly to conceive new organization-al theories and to present them effectively inacademic writing.

During theory conception, epistemic scriptsprovide us with cognitive templates for how toassemble concepts, empirical material, andmetaphors into new theory formulations. Suchformulations must be sufficiently robust to ex-plain and partially predict empirical phenom-ena, but they must also change how we perceiveorganizational life. Metaphors serve as a cata-lyst for bringing about this latter quality of anew theory. Their partial and ambiguous appli-cability to the object of study stimulates theorybuilders to be creative in their interpretationsand to generate new insights. In this article westressed the assembly of multiple metaphors asan expression of theoretical creativity. This cre-ative component must be carefully integratedwith theoretical concepts and empirical mate-


rial to form a valuable new theory. Epistemicscripts facilitate this integration.

When used to present a new theory formula-tion in writing, epistemic scripts relate the newknowledge product to the existing body of aca-demic knowledge in such a way as to reinforcethe perception among readers, including editorsand reviewers, that they are looking at a legiti-mate new theory. New theory formulations thusgain legitimacy not only from their explanatorypower but also from the epistemic scripts thatare stitched into their scholarly presentation.The three epistemic scripts of evolution, differ-entiation, and bricolage confer variable levelsof academic legitimacy on new knowledge prod-ucts, encouraging scholars to adopt the mostlegitimate scripts for presenting their theoryproposal in academic writing. Instrumental insustaining the effects of coherence, novelty, andrelevance associated with new knowledgeclaims, epistemic scripts enhance the perceivedacademic value of new theory formulations.

An important point we want to stress is thatthe epistemic scripts that inform the conceptionof a new theory need not correspond to the epis-temic scripts selected for its scholarly presenta-tion. A new theory may be conceived using thescript of evolution but may be cast in the differ-entiation script when presented to other schol-ars. Notably, we propose that the script of brico-lage is particularly prevalent in the conceptionof new organizational theories yet is almost ab-sent from the presentation of them. We illus-trated this potentially common pattern using or-ganizational institutionalism as an example.The foundational texts of this theory contain acomplex bricolage of metaphors drawn frommultiple metaphorical domains, including cul-ture, biology, construction, religion, theater,market, system, power, and the hard sciences.The theory is nevertheless cast in the epistemicscripts of evolution and differentiation, whichboth enjoy higher academic legitimacy thandoes the script of bricolage. This decoupling canbe attributed in part to the institutional dimen-sion of academic knowledge production. Assum-ing that organizational institutionalism is noexception among organizational theories, brico-lage may represent a rather common mode oftheory building that will remain controversialuntil organizational scholars recognize the insti-tutional forces that govern the academic presen-tation of new organizational knowledge.

Although we emphasize that academic knowl-edge producers make extensive use of epistemicscripts in their work, we do not suggest that theynecessarily engage in a conscious process or anact of manipulation. Academic knowledge pro-ducers are also knowledge consumers who tac-itly reemploy scripts and metaphors that theybelieve to be valuable and relevant for the gen-eration of new knowledge. Such tacitness doesnot exclude the possibility that organizationalscholars could become more conscious of theirown knowledge production and deliberatelyembrace metaphors and scripts when generat-ing and presenting new knowledge. Notably, or-ganizational theory may benefit if scholars dis-tinguish more clearly between the conception ofa new theory and the written presentation of it.A distinction may enable the adoption of themost appropriate script(s) for each dimension ofknowledge production, which could stimulatethe production of new organizational theories. Aclearer distinction may also help readers, in-cluding reviewers and editors, recognize therhetorical use of epistemic scripts and examinemore closely the actual composition of new or-ganizational theories. Our integrative model oforganizational theory production may thus en-courage organizational scholars to become bet-ter able to craft and assess new organizationaltheories.

We also hope that our work encourages schol-ars to become more reflexive about their ownpractices of knowledge production (see Alves-son, Hardy, & Harley, 2008). This need for greaterreflexivity also applies to us, the authors of thisarticle. Our text is written for publication in anacademic journal and, hence, is itself a vehicleof academic traditions and conventions. Wehave applied different scripts and templates tosubstantiate our argument and convince read-ers of our novel contribution. Moreover, our in-tegrative model of theory production and ourarticle as a whole contain a significant numberof metaphors, such as building blocks, knowl-edge products, scripts, rational myths, evolu-tion, and bricolage. These metaphors are notnew to organizational theory, although our com-bination of them is novel. The choice of meta-phors reveals our constructivist position in rela-tion to knowledge production. Our choice ofillustrative example is not neutral either. It re-flects the institutional foundation of our argu-ment about epistemic scripts of knowledge pro-


duction as taken-for-granted elements ofacademic knowledge production. Our generalportrayal of academic knowledge production asan institutionalized activity alludes to our insti-tutionalist inclinations, while our position re-garding bricolage in the production of newknowledge reflects our practice-oriented view ofknowledge production.

Why have so few organizational theories beenintroduced in academic journals over the pastdecade? It is plausible that the review processimplicitly encourages authors to present newknowledge claims as incremental contributionsto an existing line of inquiry rather than asradical breaks with established bodies ofknowledge. For instance, authors who present aknowledge product as discontinuous with exist-ing theories and paradigms—who employ thedifferentiation script alone—may be gently re-oriented by reviewers to position their work rel-ative to a more compatible line of inquiry—thatis, to use the evolution script (see Davis, 2010, fora discussion of theory survival in organizationaltheory). When editors and reviewers in this wayreinforce the evolution script relative to the dif-ferentiation script, they inadvertently discour-age the formulation and publication of new or-ganizational theories. If the aim is to renew thestock of theories in the field of organizationaltheory, we should reinforce the legitimacy of thedifferentiation script and/or the bricolage scriptamong organizational scholars at the center ofthe field, since they are the gatekeepers of themost prestigious academic journals.

A different explanation for the few new theo-ries may lie in the maturity of the field of orga-nizational theory. Most current organizationaltheories were born in the 1970s (Davis, 2010),when the field was still young and undergoingstructuration. Perhaps the field of organization-al theory has become so structured that it is nolonger feasible to introduce new theories into it,provoking instead the formation of new fields ofstudy to host new theory formulations. In thislight the question shifts from the absence of newtheories to the curious persistence of multipletheories within the field. In biology, for instance,Darwin’s theory of evolution has pushed asideLamarck’s evolutionary theory and become al-most synonymous with the field itself. Ratherthan deplore the catch-all features of institu-

tional theory, we should perhaps cherish its in-creasing dominance of organizational theory(Davis, 2010) as a sign of field maturity.

In this article we only examined the produc-tion of new organizational theories, not theirsubsequent diffusion. Future research may in-vestigate the impact of different scripts and met-aphors on the diffusion of new organizationaltheories. Perhaps organizational institutional-ism has diffused widely because its metaphori-cal content and epistemic scripts were particu-larly well chosen for the sociohistorical contextin which it emerged. Or perhaps it is the currentsociohistorical context that “selects” organiza-tional institutionalism as a particularly mean-ingful organizational theory. In any case, thechoice of metaphors and epistemic scripts ofknowledge production may determine not onlythe birth of an organizational theory but also itsvery life cycle.

REFERENCES

Albert, S., & Whetten, D. A. 1985. Organizational identity.Research in Organizational Behavior, 7: 263–295.

Alpaslan, M. C., Babb, M., Green, S. E., & Mitroff, I. I. 2006.Inquiry on inquiry: Scientific inquiry as a reflective pro-cess. Journal of Management Inquiry, 15: 7–16.

Alvesson, M. 1993. The play of metaphors. In J. Hassard & M.Parker (Eds.), Postmodernism and organizations: 114–131. London: Sage.

Alvesson, M. 2003. Beyond neopositivists, romantics, andlocalists: A reflexive approach to interviews in organi-zational research. Academy of Management Review, 28:13–33.

Alvesson, M., Hardy, C., & Harley, B. 2008. Reflecting onreflexivity: Reflexive textual practices in organizationand management theory. Journal of Management Stud-ies, 45: 480–501.

Alvesson, M., & Karreman, D. 2007. Constructing mystery:Empirical matters in theory development. Academy ofManagement Review, 32: 1265–1281.

Andriessen, D. G., & Gubbins, C. 2009. Metaphor analysis asan approach for exploring theoretical concepts: Thecase of social capital. Organization Studies, 30: 845–863.

Argyris, C., & Schon, D. 1978. Organizational learning: Atheory of action perspective. Reading, MA: Addison-Wesley.

Astley, W. G. 1985. Administrative science as socially con-structed truth. Administrative Science Quarterly, 30: 497–513.

Astley, W. G., & Zammuto, R. F. 1992. Organization science,managers, and language games. Organization Science,3: 443–460.

Avital, M., & Germonprez, M. 2004. Reflections on paragram-


matic uses of organizational theory: A response to Yian-nis Gabriel. Organization Studies, 25: 1045–1048.

Bacharach, S. B. 1989. Organizational theories: Some criteriafor evaluation. Academy of Management Review, 14:496–515.

Baker, R., & Nelson, R. E. 2005. Creating something fromnothing: Resource construction through entrepreneurialbricolage. Administrative Science Quarterly, 50: 329–366.

Barker, A., Nancarrow, C., & Spackman, N. 2001. Informedeclecticism: A research paradigm for the twenty-firstcentury. International Journal of Market Research, 43:3–28.

Barley, S. R., & Kunda, G. 1992. Design and devotion: Surgesof rational and normative ideologies of control in man-agerial discourse. Administrative Science Quarterly, 37:363–399.

Beyer, J. M., Chanove, R. G., & Fox, W. B. 1995. The reviewprocess and the fates of manuscripts submitted to AMJ.Academy of Management Journal, 38: 1219–1260.

Bonet, E., & Sauquet, A. 2010. Rhetoric in management and inmanagement research. Journal of OrganizationalChange Management, 23: 120–133.

Bryant, J., & Lasky, B. 2007. A researcher’s tale: Dealing withepistemological divergence. Qualitative Research inOrganizations and Management: An International Jour-nal, 2: 179–193.

Burrell, G., & Morgan, G. 1979. Sociological paradigms andorganizational analysis. London: Heinemann.

Burt, R. S. 1992. Structural holes: The social structure of com-petition. Cambridge, MA: Harvard University Press.

Colquitt, J. A., & Zapata-Phelan, C. P. 2007. Trends in theorybuilding and theory testing: A five-decade study of theAcademy of Management Journal. Academy of Manage-ment Journal, 50: 1281–1303.

Cornelissen, J. P. 2005. Beyond compare: Metaphor in orga-nization theory. Academy of Management Review, 30:751–764.

Cornelissen, J. P. 2006. Metaphor in organization theory:Progress and the past. Academy of Management Re-view, 31: 485–488.

Cornelissen, J. P., & Kafouros, M. 2008a. Metaphors and the-ory building in organization theory: What determinesthe impact of a metaphor on theory? British Journal ofManagement, 19: 365–379.

Cornelissen, J. P., & Kafouros, M. 2008b. The emergent orga-nization: Primary and complex metaphors in theorizingabout organizations. Organization Studies, 29: 957–978.

Cornelissen, J. P., Kafouros, M., & Lock, A. R. 2005. Metaphor-ical images of organization: How organizational re-searchers develop and select organizational metaphors.Human Relations, 58: 1545–1578.

Czarniawska, B. 1999. Writing management: Organizationtheory as a literary genre. Oxford: Oxford UniversityPress.

Daft, R. L. 1983. Learning the craft of organizational research.Academy of Management Review, 8: 539–546.

Davies, A., & Fitchett, J. A. 2005. Beyond incommensurability?Empirical expansion on diversity in research. EuropeanJournal of Marketing, 39: 272–293.

Davis, G. F. 2010. Do theories of organizations progress?Organizational Research Methods, 13: 690–709.

De Certeau, M. 1992. The practice of everyday life (2nd ed.).Berkeley: University of California Press.

De Cock, C., & Jeanes, E. L. 2006. Questioning consensus,cultivating conflict. Journal of Management Inquiry, 15:18–30.

Denzin, N. K., & Lincoln, Y. S. 1994. Handbook of qualitativeresearch. Thousand Oaks, CA, and London: Sage.

DiMaggio, P. J. 1997. Culture and cognition. Annual Reviewof Sociology, 23: 263–288.

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

Donaldson, L. 1995. American anti-management theories oforganization: A critique of paradigm proliferation. Cam-bridge: Cambridge University Press.

Donaldson, L. 2005. Following the scientific method: How Ibecame a committed functionalist and positivist. Orga-nization Studies, 26: 1071–1088.

Donaldson, L. 2009. Where is the jury? The failures in thescientific evaluation processes in organizational sci-ence. Journal of Management Inquiry, 18: 97–100.

Duymedjian, R., & Ruling, C.-C. 2010. Towards a foundationof bricolage in organization and management theory.Organization Studies, 31: 133–151.

Eisenhardt, K. M. 1989. Building theories from case studyresearch. Academy of Management Review, 14: 532–550.

Felin, T., & Foss, N. J. 2009. Performativity of theory, arbitraryconventions, and possible worlds: A reality check. Or-ganization Science, 20: 676–678.

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

Freeman, R. 2007. Epistemological bricolage: How practition-ers make sense of learning. Administration and Society,39: 476–496.

Gabriel, Y. 2002. Essai: On paragrammatic uses of organiza-tional theory—A provocation. Organization Studies, 23:133–151.

Gibbons, M., Limoges, C., Nowotny, H., Schwartzman, S.,Scott, P., & Trow, M. 1994. The new production of knowl-edge: The dynamics of science and research in contem-porary societies. London: Sage.

Gibson, C. B., & Zellmer-Bruhn, M. E. 2001. Metaphors andmeaning: An intercultural analysis of the concept ofteamwork. Administrative Science Quarterly, 46: 274–303.

Gioia, D. A., & Poole, P. P. 1984. Scripts in organizationalbehavior. Academy of Management Review, 9: 449–459.


Glaser, B. G., & Strauss, A. L. 1967. The discovery of groundedtheory: Strategies for qualitative research. New York:Aldine.

Glynn, M. A., Barr, P. S., & Dacin, M. T. 2000. Pluralism andthe problem of variety. Academy of Management Re-view, 25: 726–734.

Goffman. E. 1959. The presentation of self in everyday life.Garden City, NY: Doubleday Anchor.

Granovetter, M. 1985. Economic action and social structure:The problem of embeddedness. American Journal of So-ciology, 91: 481–510.

Grant, D. 2001. Organizations, metaphors and paradigms. InN. J. Smelser & P. B. Baltes (Eds.), International encyclo-pedia of social and behavioral sciences: 10960–10965.Oxford: Elsevier.

Greenwood, R., Oliver, C., Sahlin, K., & Suddaby, R. 2008. TheSage handbook of organizational institutionalism. Lon-don: Sage.

Guillen, M. F. 1994. The age of eclecticism: Current organi-zational trends and the evolution of managerial models.Sloan Management Review, 36(1): 75–86.

Hassard, J., & Kelemen, M. 2002. Production and consumptionin organizational knowledge: The case of the “para-digms debate.” Organization, 9: 331–355.

Homans, G. C. 1958. Social behavior as exchange. AmericanJournal of Sociology, 63: 597–606.

Huff, A. S. 2000. Changes in organizational knowledge pro-duction. Academy of Management Review, 25: 288–293.

Ilgen, D. R., & Klein, H. J. 1989. Organizational behavior.Annual Review of Psychology, 40: 327–351.

Inns, D. 2002. Metaphor in the literature of organizationalanalysis: A preliminary taxonomy and a glimpse at ahumanities-based perspective. Organization, 9: 305–330.

Koch, S., & Deetz, S. 1981. Metaphor analysis of social realityin organizations. Journal of Applied Communication Re-search, 9: 1–15.

Kuhn, T. S. 1962. The structure of scientific revolutions. Chi-cago: University of Chicago Press.

Lakoff, G., & Johnson, M. 1980. Metaphors we live by. Chi-cago: University of Chicago Press.

Langley, A. 1999. Strategies for theorizing from process data.Academy of Management Review, 24: 691–710.

Latour. B. 1987. Science in action: How to follow scientistsand engineers through society. Cambridge, MA: HarvardUniversity Press.

Levi-Strauss, C. 1966. The savage mind. Chicago: Universityof Chicago Press.

Lewis, M. W., & Grimes, A. J. 1999. Metatriangulation: Build-ing theory from multiple paradigms. Academy of Man-agement Review, 24: 672–690.

Locke, K., & Golden-Biddle, K. 1997. Constructing opportuni-ties for contribution: Structuring intertextual coherenceand “problematizing” in organizational studies. Acad-emy of Management Journal, 40: 1023–1062.

Lockwood, D. 1976. Social integration and system integra-

tion. In G. K. Zollschan & W. Hirsch (Eds.), Social change:Explorations, diagnoses, and conjectures: 370–383. NewYork: Wiley.

Lowe, S., Moore, F., & Carr, A. N. 2007. Paradigmappingstudies of culture and organization. International Jour-nal of Cross Cultural Management, 7: 237–251.

Marble, R. P. 2000. Operationalising the implementation puz-zle: An argument for eclecticism in research and in prac-tice. European Journal of Information Systems, 9: 132–147.

McKinley, W. 2010. Organizational theory development: Dis-placement of ends? Organization Studies, 31: 47–68.

Menard, C. 1990. L’economie des organisations. Paris: LaDecouverte.

Meyer, J. W., & Rowan, B. 1977. Institutionalized organiza-tions: Formal structure as myth and ceremony. Ameri-can Journal of Sociology, 83: 340–363.

Mone, M. A., & McKinley, W. 1993. The uniqueness value andits consequences for organization studies. Journal ofManagement Inquiry, 2: 284–296.

Morgan, G. 1983. More on metaphor: Why we cannot controltropes in administrative science. Administrative Sci-ence Quarterly, 28: 601–607.

Morgan, G. 1996. Is there anything more to be said aboutmetaphor? In D. Grant & C. Oswick (Eds.), Metaphor andorganizations: 227–240. London: Sage.

Ohmann, O. A. 1955. ”Skyhooks” with special implicationsfor Monday through Friday. Harvard Business Review,33(3): 33–41.

Okhuysen, G., & Bonardi, J.-P. 2011. Editors’ comments: Thechallenges of building theory by combining lenses.Academy of Management Review, 36: 6–11.

Oswick, C., Keenoy, T., & Grant, D. 2002. Metaphors andanalogical reasoning in organization theory: Beyond or-thodoxy. Academy of Management Review, 27: 294–303.

Oswick, C., Keenoy, T., & Jones, P. 2003. Rethinking organi-zational metaphors: Beyond (M)organization theory. InA. P. Muller & A. Kieser (Eds.), Communication in orga-nizations: Structures and practices: 135–148. Berlin: PeterLang.

Partington, D., & Jenkins, M. 2007. Deconstructing scholar-ship: An analysis of research methods citations in theorganizational sciences. Organizational Research Meth-ods, 10: 399–416.

Perrow, C. 1986. Complex organizations: A critical essay (3rded.) New York: Random House.

Pfeffer, J. 1993. Barriers to the advance of organizationalscience: Paradigm. Academy of Management Review,18: 599–620.

Pinder, C., & Bourgeois, V. W. 1982. Controlling tropes inadministrative science. Administrative Science Quar-terly, 27: 641–652.

Poole, M. S., & Van de Ven, A. H. 1989. Using paradox to buildorganization and management theories. Academy ofManagement Review, 15: 562–578.


Popper, K. R. 1959. The logic of scientific discovery. New York:Harper & Row.

Reed, M. 1996. Organizational theorizing: A historically con-tested terrain. In S. Clegg, C. Hardy, & W. R. Nord (Eds.),Handbook of organization studies: 31–56. London: Sage.

Reeves, C. 2005. I knew there was something wrong with thatpaper: Scientific rhetorical styles and scientific misun-derstandings. Technical Communication Quarterly, 14:267–275.

Schank, R. C., & Abelson, R. P. 1977. Scripts, plans, goals, andunderstanding: An inquiry into human knowledge struc-tures. Hillsdale, NJ: Lawrence Erlbaum Associates.

Scherer, A. G. 1998. Introduction to special issue: Pluralismand incommensurability in strategic management andorganization theory: A problem in search of a solution.Organization, 5: 147–168.

Scherer, A. G., & Steinmann, H. 1999. Some remarks on theproblem of incommensurability in organization studies.Organization Studies, 20: 519–544.

Schultz, M., & Hatch, M. J. 1996. Living with multiple para-digms: The case of paradigm interplay in organizationalculture studies. Academy of Management Review, 21:529–557.

Schultz, M., & Hatch, M. J. 2005. Building theory from practice.Strategic Organization, 3: 337–347.

Schon, D. E. 1993. Generative metaphor: A perspective onproblem setting in social policy. In A. Ortony (Ed.), Met-aphor and thought (2nd ed.): 137–161. Cambridge: Cam-bridge University Press.

Scott, W. R. 1987. The adolescence of institutional theory.Administrative Science Quarterly, 32: 493–511.

Scott, W. R., & Meyer, J. W. 1983. The organization of societalsectors. In J. W. Meyer & W. R. Scott (Eds.), Organiza-tional environments: Ritual and rationality: 129 –153.Beverly Hills, CA: Sage.

Simon, H. A. 1957. Administrative behavior. New York: FreePress.

Smircich, L. 1983. Concepts of culture and organizationalanalysis. Administrative Science Quarterly, 28: 339–358.

Smith, A. 1759. The theory of moral sentiments. Oxford: Clar-endon Press.

Sonpar, K., & Golden-Biddle, K. 2008. Using content analysisto elaborate adolescent theories of organization. Orga-nizational Research Methods, 11: 795–814.

Soyland, A. J. 1994. Psychology as metaphor. London: Sage.

Strauss, A., & Corbin, J. 1994. Grounded theory methodology:An overview. In N. K. Denzin & Y. S. Lincoln (Eds.), Hand-book of qualitative research: 1–18. London: Sage.

Suddaby, R. 2006. From the editors: What grounded theory isnot. Academy of Management Journal, 49: 633–642.

Suddaby, R. 2010. Editor’s comments: Construct clarity intheories of management and organization. Academy ofManagement Review, 35: 346–357.

Sutherland. J. W. 1975. Systems: Analysis, administration,and architecture. New York: Van Nostrand.

Tadajewski, M. 2009. The debate that won’t die? Values in-commensurability, antagonism and theory choice. Or-ganization, 16: 467–485.

Tolbert, P. S., & Zucker, L. G. 1983. Institutional sources ofchange in the formal structure of organizations: Thediffusion of civil service reform, 1880–1935. Administra-tive Science Quarterly, 28: 22–39.

Tsoukas, H., & Knudsen, C. 2003. The Oxford handbook oforganization theory. Oxford: Oxford University Press.

Udy, S. H., Jr. 1970. Work in traditional and modern society.Englewood Cliffs, NJ: Prentice-Hall.

Van Maanen, J. 1995. Style as theory. Organization Science,6: 133–143.

Weber, M. 1949. The meaning of “ethical neutrality” in soci-ology and economics. In E. A. Shils & H. A. Finch (Eds.),The methodology of the social sciences: 1–47. New York:Free Press.

Weick, K. E. 1976. Educational organizations as loosely cou-pled systems. Administrative Science Quarterly, 21:1–19.

Weick, K. E. 1979. The social psychology of organizing. NewYork: McGraw-Hill.

Weick, K. E. 1989. Theory construction as disciplined imagi-nation. Academy of Management Review, 14: 516–531.

Weick, K. E. 1995. Sensemaking in organizations. ThousandOaks, CA: Sage.

Whetten, D. A., Felin, T., & King, B. G. 2009. The practice oftheory borrowing in organizational studies: Current is-sues and future directions. Journal of Management, 35:537–563.

Zyphur, M. J. 2009. When mindsets collide: Switching analyt-ical mindsets to advance organization science. Acad-emy of Management Review, 34: 677–688.

Eva Boxenbaum ([email protected]) is associate professor of organizational studies at theCopenhagen Business School, where she also received her Ph.D. in organizationalanalysis. In addition, she is associated researcher at the Center for ManagementScience (CGS), MINES ParisTech. Her research interests include institutionalizationprocesses, innovation, and cross-cultural translations of new technologies and man-agement practices.

Linda Rouleau ([email protected]) is professor of strategy and organization theo-ries and codirector of the Study Group of Strategy-as-Practice (GePS) at HEC Montreal,where she also obtained her Ph.D. in administrative sciences. Her research focuses onstrategizing and the transformation of strategic sensemaking among middle manag-ers during organizational restructuring.


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.

new knowledge products as bricolage: metaphors … · bricolage: metaphors and scripts in...

Documents