Nahapiet 1997

8/6/2019 Nahapiet 1997

1/26

Social Capital, Intellectual Capital, and the Organizational AdvantageAuthor(s): Janine Nahapiet and Sumantra Ghoshal

Source: The Academy of Management Review, Vol. 23, No. 2 (Apr., 1998), pp. 242-266Published by: Academy of ManagementStable URL: http://www.jstor.org/stable/259373 .

Accessed: 12/01/2011 15:42

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless

you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you

may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .http://www.jstor.org/action/showPublisher?publisherCode=aom. .

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed

page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of

content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms

of scholarship. For more information about JSTOR, please contact [email protected].

Academy of Managementis collaborating with JSTOR to digitize, preserve and extend access to The Academy

of Management Review.
http://www.jstor.org/action/showPublisher?publisherCode=aomhttp://www.jstor.org/stable/259373?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/action/showPublisher?publisherCode=aomhttp://www.jstor.org/action/showPublisher?publisherCode=aomhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/259373?origin=JSTOR-pdfhttp://www.jstor.org/action/showPublisher?publisherCode=aom

8/6/2019 Nahapiet 1997

2/26

' Academy of Management Review1998, Vol. 23, No. 2, 242-266.

SOCIAL CAPITAL, NTELLECTUALAPITAL,AND THE ORGANIZATIONALDVANTAGEJANINENAHAPIETTempleton College, University of Oxford

SUMANTRA GHOSHALLondon Business SchoolScholars of the theory of the firm have begun to emphasize the sources and conditionsof what has been described as "the organizational advantage," rather than focus onthe causes and consequences of market failure. Typically, researchers see suchorganizational advantage as accruing from the particular capabilities organizationshave for creating and sharing knowledge. In this article we seek to contribute to thisbody of work by developing the following arguments: (1) social capital facilitates thecreation of new intellectual capital; (2) organizations, as institutional settings, areconducive to the development of high levels of social capital; and (3) it is because oftheir more dense social capital that firms, within certain limits, have an advantageover markets in creating and sharing intellectual capital. We present a model thatincorporates this overall argument in the form of a series of hypothesized relation-ships between different dimensions of social capital and the main mechanisms andprocesses necessary for the creation of intellectual capital.

Kogut and Zander recently have proposed"that a firm be understood as a social commu-nity specializing in the speed and efficiency inthe creation and transfer of knowledge" (1996:503). This is an important and relatively newperspective on the theory of the firm currentlybeing formalized through the ongoing work ofthese (Kogut & Zander, 1992, 1993, 1995, 1996;Zander & Kogut, 1995) and several other authors(Boisot, 1995; Conner & Prahalad, 1996; Loasby,1991; Nonaka & Takeuchi, 1995; Spender, 1996).Standing in stark contrast to the more estab-lished transaction cost theory that is groundedin the assumption of human opportunism andthe resulting conditions of market failure (e.g.,Williamson, 1975), those with this perspectiveessentially argue that organizations have someparticular capabilities for creating and sharingknowledge that give them their distinctive ad-vantage over other institutional arrangements,such as markets. For strategy theory, the impli-

cations of this emerging perspective lie in ashift of focus from the historically dominanttheme of value appropriation to one of valuecreation (Moran & Ghoshal, 1996).

The particular capabilities of organizationsfor creating and sharing knowledge derive froma range of factors, including the special facilityorganizations have for the creation and transferof tacit knowledge (Kogut & Zander, 1993, 1996;Nonaka & Takeuchi, 1995; Spender, 1996); theorganizing principles by which individual andfunctional expertise are structured, coordinated,and communicated, and through which individ-uals cooperate (Conner & Prahalad, 1996; Kogut& Zander, 1992;Zander and Kogut, 1995);and thenature of organizations as social communities(Kogut & Zander, 1992, 1996). However, notwith-standing the substantial insights we now haveinto the attributes of organizations as knowl-edge systems, we still lack a coherent theory forexplaining them. In this article we seek to ad-dress this gap and to present a theory of howfirms can enjoy what Ghoshal and Moran (1996)have called "the organizational advantage."

Our theory is rooted in the concept of socialcapital. Analysts of social capital are centrallyconcerned with the significance of relation-ships as a resource for social action (Baker, 1990;Bourdieu, 1986; Burt, 1992; Coleman, 1988, 1990;

This research was supported in part by a grant from theSundridge Park Research Fund. We are grateful to JohnStopford, Peter Moran, Morten Hansen, Richard Pascale, MaxBoisot, Wen-Pin Tsai, Nitin Nohria, Paul Willman, AnthonyHopwood, Tim Ambler, Martin Waldenstrom, and threeanonymous referees for their helpful comments on earlierdrafts of this article and in discussions of its subject matter.

242

8/6/2019 Nahapiet 1997

3/26

1998 Nahapiet and Ghoshal 243

Jacobs, 1965; Loury, 1987). However, as Putnam(1995) recently has observed, social capital is nota unidimensional concept, and, while sharing acommon interest in how relational resources aidthe conduct of social affairs, the different au-thors on this topic have tended to focus on dif-ferent facets of social capital. In this article we(1) integrate these different facets to define so-cial capital in terms of three distinct dimen-sions; (2) describe how each of these dimensionsfacilitates the creation and exchange of knowl-edge; and (3) argue that organizations, as insti-tutional settings, are able to develop high levelsof social capital in terms of all three dimensions.Our primary focus, however, is on the interrela-tionships between social and intellectual capi-tal since, as we have already noted, there isalready a clear stream of work that identifiesand elaborates the significance of knowledgeprocesses as the foundation of such organiza-tional advantage. Our aim here is to provide atheoretical explanation of why this is the case.

SOCIALCAPITALThe term "social capital" initially appeared in

community studies, highlighting the central im-portance-for the survival and functioning ofcity neighborhoods-of the networks of strong,crosscutting personal relationships developedover time that provide the basis for trust, coop-eration, and collective action in such communi-ties (Jacobs, 1965). Early usage also indicatedthe significance of social capital for the individ-ual: the set of resources inherent in family rela-tions and in community social organizationsuseful for the development of the young child(Loury, 1977). The concept has been appliedsince its early use to elucidate a wide range ofsocial phenomena, although researchers in-creasingly have focused attention on the role ofsocial capital as an influence not only on thedevelopment of human capital (Coleman, 1988;Loury, 1977, 1987) but on the economic perfor-mance of firms (Baker, 1990), geographic regions(Putnam, 1993, 1995), and nations (Fukuyama,1995).The central proposition of social capital the-ory is that networks of relationships constitute avaluable resource for the conduct of social af-fairs, providing their members with "the collec-tivity-owned capital, a 'credential' which enti-tles them to credit, in the various senses of the

word" (Bourdieu, 1986: 249). Much of this capitalis embedded within networks of mutual ac-quaintance and recognition. Bourdieu (1986), forexample, identifies the durable obligations aris-ing from feelings of gratitude, respect, andfriendship or from the institutionally guaran-teed rights derived from membership in a fam-ily, a class, or a school. Other resources areavailable through the contacts or connectionsnetworks bring. For example, through "weakties" (Granovetter, 1973) and "friends of friends"(Boissevain, 1974), network members can gainprivileged access to information and to opportu-nities. Finally, significant social capital in theform of social status or reputation can be de-rived from membership in specific networks,particularly those in which such membership isrelatively restricted (Bourdieu, 1986; Burt, 1992;D'Aveni & Kesner, 1993).Although these authors agree on the signifi-cance of relationships as a resource for socialaction, they lack consensus on a precise defini-tion of social capital. Some, like Baker (1990),limit the scope of the term to only the structure ofthe relationship networks, whereas others, likeBourdieu (1986, 1993) and Putnam (1995), also in-clude in their conceptualization of social capitalthe actual or potential resources that can beaccessed through such networks. For our pur-poses here, we adopt the latter view and definesocial capital as the sum of the actual and po-tential resources embedded within, availablethrough, and derived from the network of rela-tionships possessed by an individual or socialunit. Social capital thus comprises both thenetwork and the assets that may be mobilizedthrough that network (Bourdieu, 1986; Burt,1992).

As a set of resources rooted in relationships,social capital has many different attributes, andPutnam (1995) has argued that a high researchpriority is to clarify the dimensions of socialcapital. In the context of our exploration of therole of social capital in the creation of intellec-tual capital, we suggest that it is useful to con-sider these facets in terms of three clusters: thestructural, the relational, and the cognitive di-mensions of social capital. Although we sepa-rate these three dimensions analytically, werecognize that many of the features we describeare, in fact, highly interrelated. Moreover, in ouranalysis we set out to indicate important facets

8/6/2019 Nahapiet 1997

4/26

244 Academy of Management Review Aprilof social capital rather than review such facetsexhaustively.

In making the distinction between the struc-tural and the relational dimensions of socialcapital, we draw on Granovetter's (1992) discus-sion of structural and relational embeddedness.Structural embeddedness concerns the proper-ties of the social system and of the network ofrelations as a whole.' The term describes theimpersonal configuration of linkages betweenpeople or units. In this article we use the conceptof the structural dimension of social capital torefer to the overall pattern of connections be-tween actors-that is, who you reach and howyou reach them (Burt, 1992). Among the mostimportant facets of this dimension are the pres-ence or absence of network ties between actors(Scott, 1991; Wasserman & Faust, 1994);networkconfiguration (Krackhardt, 1989) or morphology(Tichy, Tushman, & Fombrun, 1979) describingthe pattern of linkages in terms of such mea-sures as density, connectivity, and hierarchy;and appropriable organization-that is, the ex-istence of networks created for one purpose thatmay be used for another (Coleman, 1988).In contrast, the term "relational embedded-ness" describes the kind of personal relation-ships people have developed with each otherthrough a history of interactions (Granovetter,1992). This concept focuses on the particular re-lations people have, such as respect and friend-ship, that influence their behavior. It is throughthese ongoing personal relationships that peo-ple fulfill such social motives as sociability, ap-proval, and prestige. For example, two actorsmay occupy equivalent positions in similar net-work configurations, but if their personal andemotional attachments to other network mem-bers differ, their actions also are likely to differin important respects. For instance, althoughone actor may choose to stay in a firm because

of an attachment to fellow workers, despite eco-nomic advantages available elsewhere, anotherwithout such personal bonds may discountworking relationships in making career moves.In this article we use the concept of the rela-tional dimension of social capital to refer tothose assets created and leveraged through re-lationships, and parallel to what Lindenberg(1996) describes as behavioral, as opposed tostructural, embeddedness and what Hakanssonand Snehota (1995) refer to as "actor bonds."Among the key facets in this cluster are trustand trustworthiness (Fukuyama, 1995; Putnam,1993), norms and sanctions (Coleman, 1990; Put-nam, 1995), obligations and expectations (Burt,1992; Coleman, 1990; Granovetter, 1985; Mauss,1954), and identity and identification (Hakans-son & Snehota, 1995; Merton, 1968).

The third dimension of social capital, whichwe label the "cognitive dimension," refers tothose resources providing shared representa-tions, interpretations, and systems of meaningamong parties (Cicourel, 1973). We have identi-fied this cluster separately because we believeit represents an important set of assets not yetdiscussed in the mainstream literature on socialcapital but the significance of which is receiv-ing substantial attention in the strategy domain(Conner & Prahalad, 1996; Grant, 1996; Kogut &Zander, 1992, 1996). These resources also repre-sent facets of particular importance in the con-text of our consideration of intellectual capital,including shared language and codes (Arrow,1974; Cicourel, 1973; Monteverde, 1995) andshared narratives (Orr, 1990).Although social capital takes many forms,each of these forms has two characteristics incommon: (1) they constitute some aspect of thesocial structure, and (2) they facilitate the ac-tions of individuals within the structure(Coleman, 1990). First, as a social-structural re-source, social capital inheres in the relationsbetween persons and among persons. Unlikeother forms of capital, social capital is ownedjointly by the parties in a relationship, and noone player has, or is capable of having, exclu-sive ownership rights (Burt, 1992). Moreover, al-though it has value in use, social capital cannotbe traded easily. Friendships and obligations donot readily pass from one person to another. Sec-ond, social capital makes possible the achieve-ment of ends that would be impossible withoutit or that could be achieved only at extra cost.

1 We recognize that this terminology deviates from muchthat is customary in the field of network analysis. In partic-ular, the focus of network analysis is relational data, butincluded under its heading are attributes that we labelstructural here. Scott, for example, describes network anal-ysis as being concerned with "the contacts, ties and connec-tions, the group attachments and meetings which relate oneagent to another .... These relations connect pairs of agentsto larger relational systems" (1991: 3). However, we justifyour usage both through reference to Granovetter and be-cause we believe this terminology captures well the per-sonal aspect of this dimension.

8/6/2019 Nahapiet 1997

5/26

1998 Nahapiet and Ghoshal 245In examining the consequences of social cap-ital for action, we can identify two distinct

themes. First, social capital increases the effi-ciency of action. For example, networks of socialrelations, particularly those characterized byweak ties or structural holes (i.e., disconnectionsor nonequivalencies among players in an are-na), increase the efficiency of information diffu-sion through minimizing redundancy (Burt,1992). Some have also suggested that social cap-ital in the form of high levels of trust diminishesthe probability of opportunism and reduces theneed for costly monitoring processes. It thus re-duces the costs of transactions (Putnam, 1993).Whereas the first theme could be regarded asillustrative of what North (1990) calls "allocativeefficiency," the second theme centers on the roleof social capital as an aid to adaptive efficiencyand to the creativity and learning it implies. Inparticular, researchers have found social capi-tal to encourage cooperative behavior, therebyfacilitating the development of new forms ofassociation and innovative organization(Fukuyama, 1995; Jacobs, 1965; Putnam, 1993).The concept, therefore, is central to the under-standing of institutional dynamics, innovation,and value creation.We should note, however, that social capitalis not a universally beneficial resource. AsColeman observes, "[A]given form of social cap-ital that is useful for facilitating certain actionsmay be useless or harmful for others" (1990: 302).For example, the strong norms and mutual iden-tification that may exert a powerful positive in-fluence on group performance can, at the sametime, limit its openness to information and toalternative ways of doing things, producingforms of collective blindness that sometimeshave disastrous consequences (Janis, 1982; Per-row, 1984; Turner, 1976).The main thesis of the work we have reviewedthus far is that social capital inheres in the re-lations between and among persons and is aproductive asset facilitating some forms of so-cial action while inhibiting others. Social rela-tionships within the family and wider commu-nity have been shown to be an important factorin the development of human capital (Coleman,1988). In a parallel argument we suggest thatsocial relationships-and the social capitaltherein-are an important influence on the de-velopment of intellectual capital. In elaboratingthis argument, we focus on the firm as the pri-

mary context in which to explore the interrela-tionships between social and intellectual capi-tal. Later in the article we consider how ouranalysis may be extended to a wider range ofinstitutional settings.

INTELLECTUALAPITALTraditionally, economists have examined

physical and human capital as key resources forthe firm that facilitate productive and economicactivity. However, knowledge, too, has been rec-ognized as a valuable resource by economists.Marshall, for example, suggests that "capitalconsists in a great part of knowledge and organ-ization.... [K]nowledge is our most powerfulengine of production" (1965: 115). He goes on tonote that "organization aids knowledge," a per-spective also central to the work of Arrow (1974).More recently, Quinn has expressed a similarview, suggesting that "with rare exceptions, theeconomic and producing power of the firm liesmore in its intellectual and service capabilitiesthan its hard assets-land, plant and equip-ment .... [V]irtually all public and private enter-prises-including most successful corpora-tions-are becoming dominantly repositoriesand coordinators of intellect" (1992: 241).In this article we use the term "intellectualcapital" to refer to the knowledge and knowingcapability of a social collectivity, such as anorganization, intellectual community, or profes-sional practice. We have elected to adopt thisterminology because of its clear parallel withthe concept of human capital, which embracesthe acquired knowledge, skills, and capabilitiesthat enable persons to act in new ways(Coleman, 1988). Intellectual capital thus repre-sents a valuable resource and a capability foraction based in knowledge and knowing.This orientation to intellectual capital buildson some central themes and distinctions foundin the substantial and expanding literature onknowledge and knowledge processes. Many ofthese themes have a long history in philosophyand Western thought, dating back to Plato,Aristotle, and Descartes. Two issues are of par-ticular relevance to our consideration of thespecial advantage of organizations as an in-stitutional context for the development ofintellectual capital. These are, first, debatesabout the different types of knowledge that mayexist and, second, the issue of the level of anal-

8/6/2019 Nahapiet 1997

6/26

246 Academy of Management Review Aprilysis in knowledge processes, particularly thequestion of whether social or collective knowl-edge exists and in what form.

Dimensions of Intellectual CapitalTypes of knowledge. Arguably, the most per-sistent theme in writing about the nature ofknowledge centers on the proposition that there

are different types of knowledge. For example, akey distinction scholars frequently make is be-tween practical, experience-based knowledgeand the theoretical knowledge derived from re-flection and abstraction from that experience-adistinction reminiscent of the debate of earlyphilosophers between rationalism and empiri-cism (Giddens & Turner, 1987; James, 1950). Var-iously labeled "know-how" or "proceduralknowledge," the former frequently is distin-guished from know-that, know-what, or declar-ative knowledge (Anderson, 1981; Ryle, 1949). Itconcerns well-practiced skills and routines,whereas the latter concerns the development offacts and propositions.2Perhaps the most-cited and influential dis-tinction of this sort is Polanyi's identification oftwo aspects of knowledge: tacit and explicit.This is a distinction he aligns with the "knowinghow" and "knowing what" of Gilbert Ryle (Polan-yi, 1967). Polanyi distinguishes tacit knowledgein terms of its incommunicability, and Winter(1987) has suggested that it may be useful toconsider tacitness as a variable, with the degreeof tacitness a function of the extent to which theknowledge is or can be codified and abstracted(see also Boisot, 1995).However, close reading ofPolanyi indicates that he holds the view thatsome knowledge will always remain tacit. In sodoing, he stresses the importance of knowing, aswell as knowledge, and, in particular, the activeshaping of experience performed in the pursuitof knowledge.3 Discussing the practice of sci-ence, he observes that "science is operated bythe skill of the scientist and it is through theexercise of this skill that he shapes his scientificknowledge" (Polanyi, 1962: 49). This suggestsboth a view of knowledge as object and of know-

ing as action or enactment in which progress ismade through active engagement with theworld on the basis of a systematic approach toknowing.

Levels of analysis in knowledge and knowing.Another equally fundamental cause for debatewithin philosophical and sociological circlescenters on the existence, or otherwise, of partic-ular phenomena at the collective level. That is,what is the nature of social phenomena that isdifferent from the aggregation of individualphenomena (Durkheim, 1951; Gowler & Legge,1982)? In the context of this article, the questionconcerns the degree to which it is possible toconsider a concept of organizational, collective,or social knowledge that is different from that ofindividual organizational members.

Simon represents one extreme of the argu-ment, stating that "all organizational learningtakes place inside human heads; an organiza-tion learns in only two ways: (a) by the learningof its members, or (b) by ingesting new memberswho have knowledge the organization didn'tpreviously have" (1991a: 176). In contrast, Nelsonand Winter take a very different position, assert-ing that

the possession of technical "knowledge" is anattribute of the firm as a whole, as an organizedentity, and is not reducible to what any singleindividual knows, or even to any simple aggre-gation of the various competencies and capabil-ities of all the various individuals, equipmentsand installations of the firm(1982:63).

A similar view is reflected in Brown and Du-guid's (1991) analysis of communities of practice,in which shared learning is inextricably locatedin complex, collaborative social practices.Weick and Roberts (1993) also report researchdemonstrating collective knowing at the organ-izational level.4 Our definition of intellectualcapital reflects the second of these perspectivesand acknowledges the significance of sociallyand contextually embedded forms of knowledgeand knowing as a source of value differing fromthe simple aggregation of the knowledge of aset of individuals.

These two dimensions of explicit/tacit and in-dividual/social knowledge have been combinedby Spender (1996), who created a matrix of four2 To this recent authors have added the concept of know-

why (Hamel, 1991; Kogut & Zander, 1992).'Indeed, his much-referenced chapter, in which he intro-

duces the tacit dimension, is entitled "Tacit Knowing," not"tacit knowledge."4 See also Walsh's (1995) comprehensive discussion of

organizational cognition.

8/6/2019 Nahapiet 1997

7/26


different elements of an organization's intellec-tual capital. Individual explicit knowledge-what Spender labels "conscious knowledge"-istypically available to the individual in the formof facts, concepts, and frameworks that can bestored and retrieved from memory or personalrecords. The second element, individual tacitknowledge-what Spender labels "automaticknowledge"-may take many different forms oftacit knowing, including theoretical and practi-cal knowledge of people and the performance ofdifferent kinds of artistic, athletic, or technicalskills. Availability of people with such explicitknowledge and tacit skills clearly is an impor-tant part of an organization's intellectual capitaland can be a key factor in the organization'sperformance, particularly in contexts where theperformance of individual employees is crucial,as in specialist craft work (Cooke & Yanow,1993).The other two elements of an organization'sintellectual capital are social explicit knowl-edge (what Spender calls "objectified knowl-edge") and social tacit knowledge ("collectiveknowledge," in Spender's terms). The formerrepresents the shared corpus of knowledge-epitomized, for example, by scientific communi-ties, and often regarded as the most advancedform of knowledge (Boisot, 1995). Across a widerange of organizations, we are currently wit-nessing major investments in the developmentof such objectified knowledge as firms attemptto pool, share, and leverage their distributedknowledge and intellect (Quinn, Anderson, &Finkelstein, 1996).

The latter represents the knowledge that isfundamentally embedded in the forms of socialand institutional practice and that resides in thetacit experiences and enactment of the collec-tive (Brown & Duguid, 1991). Such knowledgeand knowing capacity may remain relativelyhidden from individual actors but be accessibleand sustained through their interaction (Spend-er, 1994). It is the type of knowledge frequentlydistinguishing the performance of highly expe-rienced teams. This shared knowledge has beendefined as "routines" by Nelson and Winter(1982), and it appears that much important or-ganizational knowledge may exist in this form.For example, Weick and Roberts (1993) describethe complex, tacit, but heedful interrelating theyobserved between members of the flight opera-tions team on aircraft carriers, which they sug-

gest may characterize all high-reliability organ-izations.For a given firm, these four elements collec-tively constitute its intellectual capital. Further,the elements are not independent, as Spender(1996) notes. However, in a stylized comparisonof individuals working within an organizationversus the same individuals working at arm'slength across a hypothetical market (in the spiritof Conner and Prahalad's [1996] analysis), weuse the two categories of social knowledge toprovide the crux of our distinction: as Spenderargues, "[C]ollective knowledge is the most se-cure and strategically significant kind of organ-izational knowledge" (1996: 52). Therefore, it ison the social explicit knowledge and the socialtacit knowledge that we focus our analysis oforganizational advantage. This is an importantlimitation of our theory because, by restrictingthe scope of our analysis only to social knowl-edge, we will be unable to capture the influ-ences that explicit and tacit individual knowl-edge may have on the intellectual capital of thefirm.

There is another important way in which welimit our analysis. The potential advantages ofinternal organization over market organizationmay arise from its superior abilities in both cre-ating and exploiting intellectual capital (Kogut& Zander, 1993). We focus here only on the cre-ation of intellectual capital and ignore the ex-ploitation aspects. We have two reasons for im-posing this constraint. First, comprehensiveconsideration of both processes would exceedthe space available. Second, and more impor-tant, the benefits of intraorganizational exploi-tation of knowledge stem largely from missing,incomplete, or imperfect markets for suchknowledge (Arrow, 1974; Teece, 1988; William-son, 1975). Therefore, such advantages histori-cally have been a part of the more traditionalmarket-failure-based theories of the firm. Wherewe go beyond such theories is in our argumentthat internal organization may, within limits, besuperior to market transactions for the creationof new knowledge.

The Creation of Intellectual CapitalHow is new knowledge created? Following

Schumpeter (1934), Moran and Ghoshal (1996)have argued that all new resources, includingknowledge, are created through two generic pro-

8/6/2019 Nahapiet 1997

8/26

248 Academy of Management Review Aprilcesses: namely, combination and exchange.While this argument is yet to be widely scruti-nized, and although it is possible there may bestill other processes for the creation of newknowledge (particularly at the individual level),we believe that these two, indeed, are amongthe key mechanisms for creating social knowl-edge; therefore, we adopt this framework for ourpurposes.

Combination and the creation of intellectualcapital. Combination is the process viewed bySchumpeter as the foundation for economic de-velopment-"to produce means to combine ma-terials and forces within our reach" (1934: 65)-and this perspective has become the startingpoint for much current work on organizations asknowledge systems (Boisot, 1995; Cohen &Levinthal, 1990; Kogut & Zander, 1992). In thisliterature scholars frequently identify two typesof knowledge creation. First, new knowledgecan be created through incremental change anddevelopment from existing knowledge. Schum-peter (1934), for example, talks of continuous ad-justment in small steps, and March and Simon(1958) identify "localized search" and "stableheuristics" as the basis for knowledge growth.Within the philosophy of science, Kuhn (1970)sees development within the paradigm as thedominant mode of progression. Second, manyauthors also discuss more radical change: inno-vation, in Schumpeter's terms; double-looplearning, according to Argyris and Schon (1978);and paradigmatic change and revolution, ac-cording to Kuhn (1970). There appears to be aconsensus that both types of knowledge cre-ation involve making new combinations-incrementally or radically-either by combin-ing elements previously unconnected or bydeveloping novel ways of combining elementspreviously associated. "Development in oursense is then defined by the carrying out of newcombinations" (Schumpeter, 1934: 66),5 a view

endorsed by the recent research of Leonard-Barton (1995).Exchange and the creation of intellectual cap-ital. Where resources are held by different par-

ties, exchange is a prerequisite for resourcecombination. Since intellectual capital gener-ally is created through a process of combiningthe knowledge and experience of different par-ties, it, too, is dependent upon exchange be-tween these parties. Sometimes, this exchangeinvolves the transfer of explicit knowledge, ei-ther individually or collectively held, as in theexchange of information within the scientificcommunity or via the Internet. Often, newknowledge creation occurs through social inter-action and coactivity. Zucker, Darby, Brewer,and Peng (1996) recently have shown the impor-tance of collaboration for the development andacquisition of fine-grained collective knowledgein biotechnology. Their research endorses thesignificance of teamwork in the creation ofknowledge, as identified much earlier by Pen-rose (1959). In developing her theory of thegrowth of the firm, Penrose proposed that a firmbe viewed as "a collection of individuals whohave had experience in working together, foronly in this way can 'teamwork' be developed"(1959: 46).

There are many aspects to the learning em-bedded in such shared experience. They includethe specific meanings and understandings sub-tly and extensively negotiated in the course ofsocial interaction. Importantly, they also includean appreciation of the ways in which action maybe coordinated. For, as Penrose observes, suchexperience

develops an increasing knowledge of the possi-bilities for action and the ways in which actioncan be taken by ... the firm. This increase inknowledge not only causes the productive oppor-tunity of a firm to change ... but also contributesto the "uniqueness" of the opportunity of eachindividual firm (1959:53).An interest in the ways in which such collectivelearning, especially concerning how to coordi-nate diverse production skills and to integrateseveral technology streams, has been at the

'In their theory of the knowledge-creating company,Nonaka and Takeuchi define combination as "a process ofsystematizing concepts into a knowledge system. This modeof knowledge conversion involves combining different bod-ies of explicit knowledge" (1995: 67). They prefer to use dif-ferent terms for those forms of conversion involving tacitknowledge. However, following Polanyi (1967), we believethat all knowledge processes have a tacit dimension andthat, fundamentally, the same generic processes underlie allforms of knowledge conversion. Therefore, our usage of theterm "combination" in this context is more general and is

rooted in our view of intellectual capital as embracing boththe explicit knowledge and the tacit knowing of a collectiveand its members. Our view, thus, resembles more closely theconcept of combinative capabilities discussed by Kogut andZander (1992).

8/6/2019 Nahapiet 1997

9/26


heart of much recent discussion of core compe-tence as the source of competitive advantage(Prahalad & Hamel, 1990) and is suggestive ofthe complex ways in which exchange contrib-utes to the creation of intellectual capital.

The Conditions for Exchange and CombinationIn their analysis of value creation, Moran and

Ghoshal (1996) identify three conditions thatmust be satisfied for exchange and combinationof resources actually to take place. We believethat these conditions apply to the creation ofnew intellectual capital. In addition, however,we identify a fourth factor, which we regard asa prerequisite for the creation of intellectualcapital.The first condition is that the opportunity ex-ists to make the combination or exchange. In ourcontext we see this condition being determinedby accessibility to the objectified and collectiveforms of social knowledge. A fundamental re-quirement for the development of new intellec-tual capital is that it is possible to draw uponand engage in the existing and differing knowl-edge and knowing activities of various partiesor knowing communities (Boland & Tenkasi,1995;Zucker et al., 1996). In the academic worldthe "invisible college" long has been recognizedas an important social network giving valuableearly access to distributed knowledge, facilitat-ing its exchange and development, and therebyaccelerating the advancement of science(Crane, 1972). Clearly, recent developments intechnology, such as Lotus Notes and the Inter-net, have considerably increased the opportuni-ties for knowledge combination and exchange.In addition, however, as the history of sciencedemonstrates, the creation of new intellectualcapital also may occur through accidentalrather than planned combinations and ex-changes, reflecting emergent patterns of acces-sibility to knowledge and knowledge processes.

Second, in order for the parties involved toavail themselves of the opportunities that mayexist to combine or exchange resources, valueexpectancy theorists suggest that those partiesmust expect such deployment to create value. Inother words, they must anticipate that interac-tion, exchange, and combination will proveworthwhile, even if they remain uncertain ofwhat will be produced or how. Writing about theanticipated outcome of a conference of business

practitioners and researchers, Slocum com-ments, "[E]ach of us expects to learn somethingof value as a result of our being here. None of usknows exactly what we are going to learn orwhat path we will take in the pursuit of thisknowledge. We are confident, however, that theprocess works" (1994: ix). This anticipation of orreceptivity to learning and new knowledge cre-ation has been shown to be an important factoraffecting the success or otherwise of strategicalliances (Hamel, 1991). It exemplifies Giddens'(1984) concept of intentionality as an influenceon social action and, in so doing, also acknowl-edges the possibility that outcomes may turn outto be different from those anticipated.The third condition for the creation of newresources highlights the importance of motiva-tion. Even where opportunities for exchange ex-ist and people anticipate that value may be cre-ated through exchange or interaction, thoseinvolved must feel that their engagement in theknowledge exchange and combination will beworth their while. Moran and Ghoshal (1996) seethis as the expectation that the parties engagedin exchange and combination will be able toappropriate or realize some of the new valuecreated by their engagement, even though, asnoted previously, they may be uncertain aboutprecisely what that value may be. For example,while having considerable potential, the avail-ability of electronic knowledge exchange doesnot automatically induce a willingness to shareinformation and build new intellectual capital.Quinn et al. (1996) found, in a study of ArthurAndersen Worldwide, that major changes in in-centives and culture were required to stimulateuse of its new electronic network, and they sug-gest that motivated creativity, which they de-scribe as "care-why," is a fundamental influ-ence in the creation of value through leveragingintellect. In his research on internal stickiness,Szulanski (1996) also found that lack of motiva-tion may inhibit the transfer of best practicewithin the firm. However, Szulanski discoveredthat far more important as a barrier was the lackof capacity to assimilate and apply new knowl-edge.

Accordingly, we propose that there is a fourthprecondition for the creation of new intellectualcapital: combination capability. Even where theopportunities for knowledge exchange and com-bination exist, these opportunities are perceivedas valuable, and parties are motivated to make

8/6/2019 Nahapiet 1997

10/26

250 Academy of Management Review Aprilsuch resource deployments or to engage inknowing activity, the capability to combine in-formation or experience must exist. In their re-search on innovation, Cohen and Levinthal(1990) argue that the ability to recognize thevalue of new knowledge and information, butalso to assimilate and use it, are all vital factorsin organizational learning and innovation. Theirwork demonstrates that all of these abilities,which they label "absorptive capacity," dependupon the existence of related prior knowledge.Moreover, they suggest that an organization'sabsorptive capacity does not reside in any sin-gle individual but depends, crucially, on thelinks across a mosaic of individual capabili-ties-an observation that parallels Spender's(1996) discussion of collective knowledge.

Toward a Theory of the Creation of IntellectualCapital

By way of summary, we have argued the fol-lowing. First, new intellectual capital is createdthrough combination and exchange of existingintellectual resources, which may exist in theform of explicit and tacit knowledge and know-ing capability. Second, there are four conditionsthat affect the deployment of intellectual re-sources and engagement in knowing activityinvolving combination and exchange. Third, inreviewing the burgeoning literature on knowl-edge and knowing, we have encountered muchevidence in support of the view that the combi-nation and exchange of knowledge are complexsocial processes and that much valuable knowl-edge is fundamentally socially embedded-inparticular situations, in coactivity, and in rela-tionships. As yet, we have uncovered no singletheoretical framework that pulls together thevarious strands we can identify in this litera-ture. For example, although a growing body ofwork exists in which scholars adopt an evolu-tionary perspective and identify the special ca-pabilities of firms in the creation and transfer oftacit knowledge, this work has not yet produceda coherent theory explaining these special ca-pabilities. Given the social embeddedness ofintellectual capital, we suggest that such a the-ory is likely to be one that is primarily con-cerned with social relationships. Accordingly,we believe that social capital theory offers apotentially valuable perspective for under-

standing and explaining the creation of intellec-tual capital. It is to this theory we now return.

SOCIALCAPITAL,EXCHANGE,ANDCOMBINATIONSocial capital resides in relationships, andrelationships are created through exchange(Bourdieu, 1986). The pattern of linkages and the

relationships built through them are the founda-tion for social capital. What we observe is acomplex and dialectical process in which socialcapital is created and sustained through ex-change and in which, in turn, social capital fa-cilitates exchange. For example, there is mount-ing evidence demonstrating that where partiestrust each other, they are more willing to engagein cooperative activity through which furthertrust may be generated (Fukuyama, 1995; Put-nam, 1993; Tyler & Kramer, 1996). In social sys-tems, exchange is the precursor to resource com-bination. Thus, social capital influencescombination indirectly through exchange. How-ever, we argue below that several facets of so-cial capital, particularly those pertaining to thecognitive dimension, also have a direct influ-ence on the ability of individuals to combineknowledge in the creation of intellectual capital.Although our primary objective is to explore theways in which social capital influences the de-velopment of intellectual capital, we recognizethat intellectual capital may, itself, facilitate thedevelopment of social capital. Thus, later in thearticle we consider how the coevolution of thesetwo forms of capital may underpin organization-al advantage.The main thesis we develop here is that socialcapital facilitates the development of intellec-tual capital by affecting the conditions neces-sary for exchange and combination to occur. Toexplore this proposition, we now examine someof the ways in which each of the three dimen-sions of social capital influences the four condi-tions for resource exchange and combination wepresented earlier. The specific relationships weidentify are summarized in Figure 1.For the sake of clarity of exposition, we con-sider, in the following analysis, the impact ofeach dimension of social capital independentlyof the other dimensions. We recognize, however,that both the dimensions and the several facetsof social capital are likely to be interrelated inimportant and complex ways. For example, par-

8/6/2019 Nahapiet 1997

11/26

1998 Nahapiet and Ghoshal 251FIGURE 1

Social Capital in the Creation of Intellectual CapitalCombination and Creation of new

Social capital N exchange of - o intellectualintellectual capital capital

(A)Structural dimensionNetwork ties Al Access to parties for

3tri _A2 C3 combining/exchangingNetwork configuration A3 __ - - - intellectual capitalAppropriable organization 4 _6 \

g /tz Anticipationof value\(B)Cognitive dimension r throughcombining/(B) C~nitve dmen~on |Bl G Bf 2 / / nexlctualngaing 1 New

Shared codes and language intellectual capital

//B/3inelletualcapialalShared narratives C aia rae

(C) Relational dimensionTrust _ _ _ _ _ _ _ _ _

NormsObligations d 8 capability

Identification |_ _sL

ticular structural configurations, such as thosedisplaying strong symmetrical ties, have consis-tently been shown to be associated with suchrelational facets as interpersonal affect andtrust (Granovetter, 1985; Krackhardt, 1992). Simi-larly, researchers have highlighted the often-complex interdependencies between socialidentification and shared vocabulary and lan-guage (Ashforth & Mael, 1995).Moreover, not all dimensions of social capitalare mutually reinforcing. For instance, an effi-cient network in structural terms may not be thebest way to develop the strong relational or cog-nitive social capital that may be necessary toensure the effective operation of such networks.Nohria and Eccles (1992), for example, highlightimportant differences between face-to-face andelectronic exchange and propose that usingelectronically mediated exchange to help createa network organization requires more, not less,face-to-face communication. Our primary focuson the independent effects of these dimensions

therefore limits the richness of the present ex-ploration and identifies an important area forfuture work.

Exchange, Combination, and the StructuralDimension of Social Capital

Our main argument in this section is that,within the context of the framework of combina-tion and exchange adopted by us in this article,the structural dimension of social capital influ-ences the development of intellectual capitalprimarily (though not exclusively) through theways in which its various facets affect access toparties for exchanging knowledge and partici-pating in knowing activities. While recognizingthat the structural facets also may be systemat-ically associated with other conditions for theexchange and combination of knowledge, webelieve that these associations are primarily de-rived indirectly, through the ways in whichstructure influences the development of the re-

8/6/2019 Nahapiet 1997

12/26

252 Academy of Management Review Aprilnational and cognitive dimensions of social cap-ital. For example, the strong, symmetrical tiesfrequently associated with the development ofaffective relationships (both positive and nega-tive) may, in turn, influence individuals' motiva-tion to engage in social interaction and, thereby,exchange knowledge (Krackhardt, 1992; Lawler& Yoon, 1996). Similarly, stable networks char-acterized by dense relations and high levels ofinteraction are conducive to the development ofthe different facets of the cognitive social capi-tal we discuss in this article (Boisot, 1995; Orr,1990).

Network ties. The fundamental proposition ofsocial capital theory is that network ties provideaccess to resources. One of the central themes inthe literature is that social capital constitutes avaluable source of information benefits (i.e.,"who you know" affects "what you know").Coleman (1988) notes that information is impor-tant in providing a basis for action but is costlyto gather. However, social relations, often estab-lished for other purposes, constitute informationchannels that reduce the amount of time andinvestment required to gather information.

Burt (1992) suggests that these informationbenefits occur in three forms: access, timing,and referrals. The term "access" refers to receiv-ing a valuable piece of information and know-ing who can use it, and it identifies the role ofnetworks in providing an efficient information-screening and -distribution process for membersof those networks. Thus, network ties influenceboth access to parties for combining and ex-changing knowledge (Al in Figure 1) and antic-ipation of value through such exchange (A2 inFigure 1). The operations of the invisible collegeprovide an example of such networks.

"Timing" of information flows refers to theability of personal contacts to provide informa-tion sooner than it becomes available to peoplewithout such contacts. This may well increasethe anticipated value of such information (A2 inFigure 1), as demonstrated in research on job-seeking behavior (Granovetter, 1973).Such earlyaccess to information may be especially impor-tant in commercially oriented research and de-velopment, where speed to market may be acrucial factor in determining success."Referrals" are those processes providing in-formation on available opportunities to peopleor actors in the network, hence influencing theopportunity to combine and exchange knowl-

edge (Al in Figure 1). They constitute a flow ofinformation not only about possibilities but fre-quently include reputational endorsement forthe actors involved-thereby influencing boththe anticipated value of combination and ex-change and the motivation for such exchange(see Granovetter, 1973, and Putnam, 1993). How-ever, we believe that such reputational endorse-ment derives more from relational than struc-tural factors, which we explore below.Network configuration. Ties provide the chan-nels for information transmission, but the over-all configuration of these ties constitutes an im-portant facet of social capital that may impactthe development of intellectual capital. For ex-ample, three properties of network structure-density, connectivity, and hierarchy-are allfeatures associated with flexibility and ease ofinformation exchange through their impact onthe level of contact or the accessibility they pro-vide to network members (A3 in Figure 1;Ibarra,1992; Krackhardt, 1989).Burt (1992) notes that a player with a networkrich in information benefits has contacts estab-lished in the places where useful bits of infor-mation are likely to air and who will provide areliable flow of information to and from thoseplaces. While acknowledging the importance oftrust and trustworthiness as a factor in thechoice of contacts, Burt (1992) devotes muchmore attention to the efficiency of different rela-tionship structures, arguing, in particular, thatthe sparse network, with few redundant con-tacts, provides more information benefits. Thedense network is inefficient in the sense that itreturns less diverse information for the samecost as that of the sparse network. The benefitsof the latter, thus, derive from both the diversityof information and the lower costs of accessingit.

Jacobs (1965) and Granovetter (1973) havemade similar arguments, identifying the role of"hop-and-skip" links and "loose ties" in informa-tion diffusion through communities. This aspectof diversity is very important, because it is wellestablished that significant progress in the cre-ation of intellectual capital often occurs bybringing together knowledge from disparatesources and disciplines. Networks and networkstructures, thus, represent facets of social capi-tal that influence the range of information thatmay be accessed (A3 in Figure 1) and that be-comes available for combination. As such, these

8/6/2019 Nahapiet 1997

13/26

1998 Nahapiet and Ghoshal 253structures constitute a valuable resource aschannels or conduits for knowledge diffusionand transfer.However, there are important limitations tothe conduit model, in which meaning is viewedas unproblematic and in which the primary con-cern is with issues of information transfer. Forexample, Hansen (1996)has found that weak tiesfacilitate search but impede transfer, especiallywhen knowledge is not codified. Thus, whereasnetworks having little redundancy may be botheffective and efficient for the transfer of infor-mation whose meaning is relatively unproblem-atic, much richer patterns of relationship andinteraction are important where the meaning ofinformation is uncertain and ambiguous orwhere parties to an exchange differ in their priorknowledge. For example, Cohen and Levinthal(1990) have shown that some redundancy is nec-essary for the development of cross-functionalabsorptive capacity. Nonetheless, the generalpoint remains that the configuration of the net-work is an important influence on the accessi-bility of information resources (A3 in Figure 1),although the appropriate level of redundancy iscontingent on the degree to which the parties toknowledge exchange share a common knowl-edge base.Appropriable organization. Social capital de-veloped in one context, such as ties, norms, andtrust, can often (but not always) be transferredfrom one social setting to another, thus influenc-ing patterns of social exchange. Examplesinclude the transfer of trust from family andreligious affiliations into work situations(Fukuyama, 1995), the development of personalrelationships into business exchanges (Coleman,1990), and the aggregation of the social capitalof individuals into that of organizations (Burt,1992). This suggests that organizations createdfor one purpose may provide a source of valu-able resources for other, different purposes(Nohria, 1992; Putnam, 1993, 1995). Such appro-priable social organization can provide a poten-tial network of access to people and their re-sources, including information and knowledge(A4 in Figure 1), and, through its relational andcognitive dimensions, may ensure motivationand capability for exchange and combination(see below). However, such organization alsomay inhibit such processes; indeed, researchdemonstrates how organizational routines mayseparate rather than coordinate groups within

organizations, constraining rather than en-abling learning and the creation of intellectualcapital (Dougherty, 1996; Hedberg, 1981).

Exchange, Combination, and the CognitiveDimension of Social Capital

Earlier in this article, we defined intellectualcapital as the knowledge and knowing capabil-ity of a social collectivity. This reflects our beliefthat, fundamentally, intellectual capital is a so-cial artifact and that knowledge and meaningare always embedded in a social context-bothcreated and sustained through ongoing rela-tionships in such collectivities. Although schol-ars widely recognize that innovation generallyoccurs through combining different knowledgeand experience and that diversity of opinion is away of expanding knowledge, meaningful com-munication-an essential part of social ex-change and combination processes-requires atleast some sharing of context between the par-ties to such exchange (Boisot, 1995; Boland &Tenkasi, 1995; Campbell, 1969). We suggest thatthis sharing may come about in two main ways:(1) through the existence of shared languageand vocabulary and (2) through the sharing ofcollective narratives. Further, we suggest thatthese two elements constitute facets of sharedcognition that facilitate the creation of intellec-tual capital especially through their impact oncombination capability. In each case they do soby acting as both a medium and a product ofsocial interaction.

Shared language and codes. There are sev-eral ways in which a shared language influ-ences the conditions for combination and ex-change. First, language has a direct andimportant function in social relations, for it isthe means by which people discuss and ex-change information, ask questions, and conductbusiness in society. To the extent that peopleshare a common language, this facilitates theirability to gain access to people and their infor-mation. To the extent that their language andcodes are different, this keeps people apart andrestricts their access (Bi in Figure 1).

Second, language influences our perception(Berger & Luckman, 1966; Pondy & Mitroff, 1979).Codes organize sensory data into perceptualcategories and provide a frame of reference forobserving and interpreting our environment.Thus, language filters out of awareness those

8/6/2019 Nahapiet 1997

14/26

254 Academy of Management Review Aprilevents for which terms do not exist in the lan-guage and filters in those activities for whichterms do exist. Shared language, therefore, mayprovide a common conceptual apparatus forevaluating the likely benefits of exchange andcombination (B2 in Figure 1).

Third, a shared language enhances combina-tion capability (B3 in Figure 1). Knowledge ad-vances through developing new concepts andnarrative forms (Nonaka &Takeuchi, 1995). How-ever, as we noted previously, in order to developsuch concepts and to combine the informationgained through social exchange, the differentparties must have some overlap in knowledge.Boland and Tenkasi (1995) identify the impor-tance of both perspective taking and perspec-tive making in knowledge creation, and theydemonstrate how the existence of a shared vo-cabulary enables the combining of information.We suggest it is for all these reasons that re-searchers increasingly recognize group-specificcommunication codes as a valuable assetwithin firms (Arrow, 1974; Kogut & Zander, 1992;Monteverde, 1995; Prescott & Visscher, 1980).Shared narratives. Beyond the existence ofshared language and codes, researchers havesuggested that myths, stories, and metaphorsalso provide powerful means in communities forcreating, exchanging, and preserving rich setsof meanings-a view long held by some socialanthropologists (Clark, 1972; Nisbet, 1969). Re-cently, Bruner (1990) proposed that there are twodifferent modes of cognition: (1) the informationor paradigmatic mode and (2) the narrativemode. The former suggests a process of knowl-edge creation rooted in rational analysis andgood arguments; the latter is represented in syn-thetic narratives, such as fairy tales, myths andlegends, good stories, and metaphors. Accord-ing to Bateson (1972), metaphors cut across dif-ferent contexts, thus enabling the combining ofboth imaginative and literal observations andcognitions. Orr (1990) demonstrates how narra-tive in the form of stories, full of seemingly in-significant details, facilitates the exchanging ofpractice and tacit experience between techni-cians, thereby enabling the discovery and de-velopment of improved practice. The emergenceof shared narratives within a community thusenables the creation and transfer of new inter-pretations of events, doing so in a way thatfacilitates the combination of different forms of

knowledge, including those largely tacit (B4 inFigure 1).

Exchange, Combination, and the RelationalDimension of Social Capital

Much of the evidence for the relationship be-tween social capital and intellectual capitalhighlights the significance of the relational di-mension of social capital. Szulanski (1996) hasfound that one of the important barriers to thetransfer of best practice within organizations isthe existence of arduous relations between thesource and the recipient. Whereas we have ar-gued that the structural dimension has its pri-mary direct impact on the condition of accessi-bility, and the cognitive dimension through itsinfluence on accessibility and combination ca-pability, research suggests that the relationaldimension of social capital influences three ofthe conditions for exchange and combination inmany ways. These are access to parties for ex-change, anticipation of value through exchangeand combination, and the motivation of partiesto engage in knowledge creation through ex-change and combination.

Trust. Misztal defines trust as the belief thatthe "results of somebody's intended action willbe appropriate from our point of view" (1996:9-10). A substantial body of research now exists(Fukuyama, 1995;Gambetta, 1988; Putnam, 1993,1995; Ring & Van de Ven, 1992, 1994; Tyler &Kramer, 1996) that demonstrates where relation-ships are high in trust, people are more willingto engage in social exchange in general, andcooperative interaction in particular (Cl in Fig-ure 1). Mishira (1996) argues that trust is multi-dimensional and indicates a willingness to bevulnerable to another party-a willingness aris-ing from confidence in four aspects: (1) belief inthe good intent and concern of exchange part-ners (Ouchi, 1981; Pascale, 1990; Ring & Van deVen, 1994), (2) belief in their competence andcapability (Sako, 1992; Szulanski, 1996), (3) beliefin their reliability (Giddens, 1990; Ouchi, 1981),and (4) belief in their perceived openness(Ouchi, 1981).Misztal observes that "trust, by keeping ourmind open to all evidence, secures communica-tion and dialogue" (1996: 10), suggesting therebythat trust may both open up access to people forthe exchange of intellectual capital (C3 in Fig-ure 1) and increase anticipation of value

8/6/2019 Nahapiet 1997

15/26


through such exchanges (C2 in Figure 1). Onecan find support for this view in research dem-onstrating that where there are high levels oftrust, people are more willing to take risks insuch exchange (Nahapiet, 1996; Ring & Van deVen, 1992). This may represent an increasedwillingness to experiment with combining dif-ferent sorts of information. For example, Luh-mann (1979) has shown trust to increase the po-tential of a system for coping with complexityand, thus, diversity-factors known to be impor-tant in the development of new intellectual cap-ital. Trust may also indicate greater openness tothe potential for value creation through ex-change and combination (C2 in Figure 1). Boisothighlights the importance of interpersonal trustfor knowledge creation in contexts of high am-biguity and uncertainty: "[W]hen the message isuncodified, trust has to reside in the quality ofthe personal relationships that bind the partiesthrough shared values and expectations ratherthan the intrinsic plausibility of the message"(1995: 153).

As we noted earlier, there is a two-way inter-action between trust and cooperation: trust lu-bricates cooperation, and cooperation itselfbreeds trust. This may lead to the development,over time, of generalized norms of cooperation,which increase yet further the willingness toengage in social exchange (Putnam, 1993). Inthis respect, collective trust may become apotent form of "expectational asset" (Knez &Camerer, 1994) that group members can rely onmore generally to help solve problems of coop-eration and coordination (Kramer, Brewer, &Hanna, 1996).

Norms. According to Coleman (1990), a normexists when the socially defined right to controlan action is held not by the actor but by others.Thus, it represents a degree of consensus in thesocial system. Coleman suggests that "where anorm exists and is effective, it constitutes a pow-erful though sometimes fragile form of socialcapital" (1988: S104). Norms of cooperation canestablish a strong foundation for the creation ofintellectual capital. Becoming, in effect, "expec-tations that bind" (Kramer & Goldman, 1995),such norms may be a significant influence onexchange processes, opening up access to par-ties for the exchange of knowledge (C4 in Figure1) and ensuring the motivation to engage insuch exchange (C5 in Figure 1; Putnam, 1993).

For example, Starbuck (1992) notes the impor-tance of social norms of openness and teamworkas key features of knowledge-intensive firms; hehighlights the significance of the emphasis oncooperation rather than competition, on opendisclosure of information, and on building loy-alty to the firm as significant underpinnings ofthe success of the American law firm Wachtell,Lipton, Rosen and Katz, which specializes in ad-vice on nonroutine, challenging cases. Othernorms of interaction that have been shown to beimportant in the creation of intellectual capitalinclude a willingness to value and respond todiversity, an openness to criticism, and a toler-ance of failure (Leonard-Barton, 1995). Suchnorms may offset the tendency to "groupthink"that may emerge in strong, convergent groupsand that represents the way in which high lev-els of social capital may be a real inhibitor forthe development of intellectual capital (Janis,1982).At the same time, as Leonard-Barton (1995)has shown, norms also may have a dark side;those capabilities and values initially seen asa benefit may become, in time, a pathologicalrigidity.

Obligations and expectations. Obligationsrepresent a commitment or duty to undertakesome activity in the future. Coleman (1990) dis-tinguishes obligations from generalized norms,viewing the former as expectations developedwithin particular personal relationships. Hesuggests that obligations operate as a "creditslip" held by A to be redeemed by some perfor-mance by B-a view reminiscent of Bourdieu's(1986) concept of credential we referred to earlierin this article. In the context of the creation ofintellectual capital, we suggest that such obli-gations and expectations are likely to influenceboth access to parties for exchanging and com-bining knowledge (C6 in Figure 1) and the mo-tivation to combine and exchange such knowl-edge (C7 in Figure 1).The notion that "there is nosuch thing as a free lunch" represents a com-monly held view that exchange brings with itexpectations about future obligations-a viewexplicated in detail by Mauss (1954), Bourdieu(1977), and Cheal (1988). Fairtlough (1994) as-cribes considerable importance to the formal,professional, and personal obligations that de-velop between those involved in cooperative re-search and development projects between dif-ferent organizations:

8/6/2019 Nahapiet 1997

16/26

256 Academy of Management Review AprilPeople in the two companies could rely on eachother.... This was cooperation which certainlywent beyond contractual obligations. It mightalso have gone beyond enlightened self interest,and beyond good professional behaviour, be-cause the scientists liked working together, feltcommitted to the overall project and felt a per-sonal obligation to help the others involved (1994:119).Identification. Identification is the processwhereby individuals see themselves as onewith another person or group of people. Thismay result from their membership in that groupor through the group's operation as a referencegroup, "in which the individual takes the valuesor standards of other individuals or groups as acomparative frame of reference" (Merton, 1968:288; see also Tajfel, 1982). Kramer et al. (1996)have found that identification with a group or

collective enhances concern for collective pro-cesses and outcomes, thus increasing thechances that the opportunity for exchange willbe recognized. Identification, therefore, acts as aresource influencing both the anticipation ofvalue to be achieved through combination andexchange (C8 in Figure 1) and the motivation tocombine and exchange knowledge (C9 in Figure1). We find support for this in the research ofLewicki and Bunker (1996), whose evidence sug-gests that salient group identification may notonly increase the perceived opportunities for ex-change but also may enhance the actual fre-quency of cooperation. In contrast, where groupshave distinct and contradictory identities, thesemay constitute significant barriers to informa-tion sharing, learning, and knowledge creation(Child & Rodrigues, 1996; Pettigrew, 1973; Simon& Davies, 1996).Thus far, we have argued that social capitaltheory provides a powerful basis for under-standing the creation of intellectual capital ingeneral. The various specific links we have pro-posed are summarized in Figure 1. In the nextsection we suggest that the theory also providesa basis for understanding the nature of organi-zational advantage since firms, as institutions,are likely to be relatively well endowed withsocial capital.

SOCIALCAPITAL,NTELLECTUALAPITAL,AND THEORGANIZATIONALDVANTAGEThe last 20 years have witnessed a substan-tial resurgence of interest in the theory of the

firm. During this period, those espousing trans-action cost approaches became increasingly in-fluential, positing, at their simplest, that the ex-istence of firms can be explained in terms ofmarket failure and the greater ability of firms,through hierarchy, to reduce the costs of trans-actions in particular (and relatively restricted)circumstances (Williamson, 1975, 1981, 1985). Thetransaction cost theory of the firm has provedrobust and has been applied across a widerange of issues, but it has also become subject togrowing criticism for a range of definitional,methodological, and substantive reasons (see,for example, Conner & Prahalad, 1996, and Pite-lis, 1993). More fundamentally, as we noted atthe outset of this article, researchers now areseeking to develop a theory of the firm that isexpressed in positive terms (Kogut & Zander,1996; Masten, Meehan, & Snyder, 1991; Simon,1991b)-away from a market-failure frameworkto one grounded in the concept of organizationaladvantage (Moran & Ghoshal, 1996).Increasingly, the special capabilities of organ-izations for creating and transferring knowledgeare being identified as a central element of or-ganizational advantage. We suggest that socialcapital theory provides a sound basis for ex-plaining why this should be the case. First, or-ganizations as institutional settings are charac-terized by many of the factors known to beconducive to the development of high levels ofsocial capital. Second, it is the coevolution ofsocial and intellectual capital that underpinsorganizational advantage.

Organizations as Institutional Settings AreConducive to the Development of SocialCapital

Social capital is owned jointly by the partiesto a relationship, with no exclusive ownershiprights for individuals. Thus, it is fundamentallyconcerned with resources located within struc-tures and processes of social exchange; as such,the development of social capital is signifi-cantly affected by those factors shaping the evo-lution of social relationships. We discuss foursuch conditions here: time, interaction, interde-pendence, and closure. We argue that all fourare more characteristic of internal organizationthan of market organization as represented inneoclassical theory and that, as a result, organ-izations as institutional settings are conducive

8/6/2019 Nahapiet 1997

17/26


to the development of high levels of social cap-ital relative to markets. However, as we subse-quently note, in practice these conditions mayalso occur in some forms of interorganizationalnetworks, thereby enabling such networks tobecome relatively well endowed with socialcapital.Time and the development of social capital.Like other forms of capital, social capital consti-tutes a form of accumulated history-here re-flecting investments in social relations and so-cial organization through time (Bourdieu, 1986;Granovetter, 1992). Time is important for the de-velopment of social capital, since all forms ofsocial capital depend on stability and continuityof the social structure. The concept of embed-ding fundamentally means the binding of socialrelations in contexts of time and space (Gid-dens, 1990). Coleman highlights the importanceof continuity in social relationships:

One way in which the transactions that make upsocial action differ from those of the classicalmodel of a perfect market lie in the role of time. Ina model of a perfect market,transactions are bothcostless and instantaneous. But in the real world,transactions are consummated over a period oftime (1990:91).For example, since it takes time to build trust,relationship stability and durability are key net-work features associated with high levels oftrust and norms of cooperation (Axelrod, 1984;Granovetter, 1985; Putnam, 1993; Ring & Van deVen, 1992). The duration and stability of socialrelations also influence the clarity and visibilityof mutual obligations (Misztal, 1996).

Although, in the main, social capital is cre-ated as a by-product of activities engaged in forother purposes, intentional or constructed organ-ization represents a direct, purposeful invest-ment in social capital (Coleman, 1990, 1993)."These organizations ordinarily take the form ofauthority structures composed of positions con-nected by obligations and expectations andoccupied by persons" (Coleman 1990: 313). Incontrast to the short-term transactions charac-terizing the markets of neoclassical theory, in-tentional or constructed organization representsthe creation and maintenance of an explicit andenduring structure of ties constituting, throughorganizational design, a configuration of rela-tionships and resources usable for a variety ofpurposes-both formal and informal. Moreover,this commitment to continuity facilitates the

other processes known to be influential in thedevelopment of social capital: interdependence,interaction, and closure.

Interdependence and the development of so-cial capital. Coleman (1990) states that socialcapital is eroded by factors that make peopleless dependent upon each other. This appearsespecially so for the relational dimension of so-cial capital. For example, expectations and ob-ligations are less significant where people havealternative sources of support. Indeed, Misztal(1996) has suggested that the recent resurgenceof interest in trust can be explained by the in-creasingly transitional character of our presentcondition and the erosion of social interdepen-dence and solidarity. Yet, most authors agreethat high levels of social capital usually aredeveloped in contexts characterized by high lev-els of mutual interdependence.Whereas markets as institutional arrange-ments are rooted in the concept of autonomy(and institutional economists largely neglect in-terdependence between exchange parties; Zajac& Olsen, 1993), firms fundamentally are institu-tions designed around the concepts and prac-tices of specialization and interdependence anddifferentiation and integration (Lawrence &Lorsch, 1967; Smith, 1986; Thompson, 1967). Inter-dependence-and the coordination it implies-long has been recognized as perhaps the keyattribute of business organization (Barnard,1938). Follet goes so far as to suggest that

the fair test of business administration, of indus-trial organization, is whether you have a busi-ness with all its parts so co-ordinated, so movingtogether in their closely knit and adjusting activ-ities, so linking, interlocking and inter-relating,that they make a working unit, not a congerie ofseparate pieces (1949:61).Such interdependence provides the stimulus fordeveloping many organizationally embeddedforms of social capital. For example, throughproviding the opportunity to create contextscharacterized by the condition of interdepen-dent viability-that is, the requirement that ex-changes are positive in outcome for the systemoverall rather than for each individual memberof the system- organizations considerably ex-tend the circle of exchange that takes placeamong their members (Coleman, 1993; Moran &Ghoshal, 1996), thereby increasing social identi-fication and encouraging norms of cooperationand risk taking.

8/6/2019 Nahapiet 1997

18/26

258 Academy of Management Review AprilInteraction and the development of social

capital. Social relationships generally, thoughnot always, are strengthened through interac-tion but die out if not maintained. Unlike manyother forms of capital, social capital increasesrather than decreases with use. Interaction,thus, is a precondition for the developmentand maintenance of dense social capital(Bourdieu, 1986). In particular, as we noted al-ready, scholars have shown that the cognitiveand relational dimensions of social capital ac-cumulate in network structures where link-ages are strong, multidimensional, and recip-rocal-features that characterize many firmsbut that rarely surface in pure market forms oforganization. Discussing the development oflanguage, Boland and Tenkasi note that it is"through action within communities of know-ing that we make and remake both our lan-guage and our knowledge" (1995: 353). Accord-ing to these authors, such communities musthave space for conversation, action, and inter-action in order for the codes and language todevelop that facilitate the creation of new in-tellectual capital.In a different context Boissevain (1974) showshow multiplex relations are more intimate thansingle-stranded relationships, therefore provid-ing more accessibility and more response topressure than single-stranded relations. Suchrelations typically are imbued with higher lev-els of obligation between network members, aswell as trust-based norms (Coleman, 1990). Fur-ther, Powell (1996) argues that norm-based con-ceptions of trust miss the extent to which coop-eration is buttressed by sustained contact,regular dialogue, and constant monitoring. Headds that, without mechanisms and institutionsto sustain such conversations, trust does not en-sue (see also Coleman, 1990). This echoes Bour-dieu's earlier emphasis on the fundamentalneed for "an unceasing effort of sociability"(1986: 250) for the reproduction of social capitalin its many forms.In neoclassical theory, markets as institu-tional settings are epitomized by impersonal,arm's length, spot transactions. Firms, in con-trast, provide many opportunities for sustainedinteraction, conversations, and sociability-both by design and by accident. Formal organi-zations explicitly are designed to bring mem-bers together in order to undertake their primarytask, to supervise activities, and to coordinate

their activities, particularly in contexts requir-ing mutual adjustment (Mintzberg, 1979; Thomp-son, 1967), change, and innovation (Burns &Stalker, 1961; Galbraith, 1973). Through copres-ence (Giddens, 1984), coloration (Fairtlough,1994), and the creation of such processes as rou-tine choice opportunities (March & Olsen, 1976),organizations also create a myriad of contextsand occasions for the more-or-less planned com-ing together of people and their ideas. Finally,the literature is replete with evidence that or-ganizational life is characterized by a substan-tial amount of conversation: in meetings, confer-ences, and social events that fill the everydaylife of workers and managers (Mintzberg, 1973;Prescott & Visscher, 1980; Roy, 1960). Together,these can be viewed as collective investmentstrategies for the institutional creation andmaintenance of dense networks of social rela-tionships and for the resources embeddedwithin, available through, and derived fromsuch networks of relationships. Alternatively,these meetings and social events provide theunplanned and unstructured opportunities forthe accidental coming together of ideas thatmay lead to the serendipitous development ofnew intellectual capital.

Closure and the development of social cap-ital. Finally, there is much evidence that clo-sure is a feature of social relationships that isconducive to the development of high levels ofrelational and cognitive social capital. Strongcommunities-the epitome of systems of densesocial capital-have "identities that separateand a sense of sociological boundary that dis-tinguishes members from nonmembers" (Etzi-oni, 1996: 9; see also Bourdieu, 1986). The de-velopment of norms, identity, and trust hasbeen shown to be facilitated by network clo-sure (Coleman, 1990; Ibarra, 1992), and the de-velopment of unique codes and language isassisted by the existence of community sepa-ration (Boland & Tenkasi, 1995). Formal organ-izations, by definition, imply a measure of clo-sure through the creation of explicit legal,financial, and social boundaries (Kogut &Zander, 1996). Markets, in contrast, representopen networks that benefit from the freedomoffered to individual agents but that have lessaccess to the relational and cognitive facets ofsocial capital.

8/6/2019 Nahapiet 1997

19/26


The Coevolution of Social and IntellectualCapital Underpins Organizational Advantage

Our main argument thus far has been thatsocial capital is influential in the developmentof new intellectual capital and that organiza-tions are institutional settings conducive to thedevelopment of social capital. We have notedthe significant and growing body of work thatindicates organizations have some particularcapabilities for creating and sharing knowl-edge, giving them their distinctive advantageover other institutional arrangements, such asmarkets. We now pull the strands of our analy-sis together by proposing that it is the interac-tion between social and intellectual capital thatunderpins organizational advantage.

Although our primary aim has been to suggestthat social capital influences the developmentof intellectual capital, we recognize that the pat-tern of influence may be in the other direction.The view that shared knowledge forms the basisfrom which social order and interaction flow is acentral theme in sociology, exemplified in thework of Berger and Luckman (1966) and Schutz(1970). Within organizational analysis, authorslong have suggested that the firm's particularknowledge about how activities are to be coor-dinated underpins its capability to develop andoperate as a social system (Kogut & Zander,1992, 1996; March & Simon, 1958; Penrose, 1959;Thompson, 1967). We represent the influence ofintellectual capital on social capital as a feed-back relationship in Figure 1. More important,however, we believe that it is the coevolution ofsocial and intellectual capital that is of partic-ular significance in explaining the source of or-ganizational advantage.Earlier in the article we noted the dialecticalprocess by which social capital is both createdand sustained through exchange and, in turn,enables such exchange to take place. As Bergerand Luckman observe,

The relationship between man, the producer,andthe social world, his product, is and remains adialectical one. That is, man (not, of course, inisolation but in his collectivities) and his socialworld interact with each other. The product actsback upon the producer (1966:78; see also Bour-dieu, 1977).Giddens, too, examines the self-reproducingquality of social practices, noting that socialactivities are recursive-that is, "continually

recreated by actors via the very means by whichthey express themselves as actors" (1984: 2). ForGiddens this implies a concept of human knowl-edgeability that underpins all social practice.

The discussion of knowledgeability that en-sues suggests the reciprocal quality of the rela-tionship between social and intellectual capitaland is consistent with our emphasis on the so-cial embeddedness of both forms of capital.Since both social and intellectual capital de-velop within and derive their significance fromthe social activities and social relationshipswithin which they are located, their evolution-ary paths are likely to be highly interrelated.Consideration of the reciprocal relationshipbetween knowledge and its social context per-meates the sociology of science (Zuckerman,1988). Mullins (1973), for example, describes thejoint evolution of social interaction, communica-tion networks, and the elaboration of scientificideas and notes that cognitive development isfacilitated by the thickening of communicationnetworks, which then leads to their further elab-oration. Research within organizations offersmany parallel examples (Burns & Stalker, 1961;Leonard-Barton, 1995; Weick, 1995; Zucker et al.,1996). For instance, in a study of change inhealth administration, Nahapiet (1988) de-scribes, in detail, how a new accounting calcu-lus both shaped and was, in turn, shaped by thesocial context in which it was embedded.Discussing Orr's (1990) influential ethnogra-phy of service technicians, Brown and Duguid(1991) provide further insight into this coevolu-tion of knowledge and relationships. Specifi-cally, they describe how technicians achievetwo distinct forms of social construction. First,through their work, and "through cultivatingconnections throughout the corporation" (Brown& Duguid, 1991: 67), technicians engage in theongoing creation and negotiation of shared un-derstanding-an understanding that representstheir view of the world, that is their collectiveknowledge. The second form of social construc-tion, which, according to Brown and Duguid, isalso important but less evident, is the creation ofa shared identity. "In telling these stories anindividual rep contributes to the constructionand development of his or her own identity as arep and reciprocally to the construction and de-velopment of the community of reps in which heor she works" (Brown & Duguid, 1991: 68). In ananalysis reminiscent of Weick and Roberts'

8/6/2019 Nahapiet 1997

20/26

260 Academy of Management Review April

(1993) discussion of collective mind-itself lo-cated in processes of interrelating-these au-thors highlight the mutually dependent and in-teractive ways in which social and intellectualcapital coevolve.

We suggest that this emphasis on the coevo-lution of the two forms of capital provides adynamic perspective on the development of or-ganizational advantage. Spender (1996) arguesthat it is the collective forms of knowledge thatare strategically important, and many authorsclaim that it is these forms of shared tacit knowl-edge that underpin what we have termed the"organizational advantage." It is these collec-tive forms of knowledge, we believe, that areparticularly tightly interconnected with the rela-tional and cognitive forms of social capital withwhich, we have argued, organizations are rela-tively well endowed. Organizations, thus, buildand retain their advantage through the dynamicand complex interrelationships between socialand intellectual capital.

DISCUSSIONAND IMPLICATIONSThe view of organizational advantage we

present here is fundamentally a social one. Wesee the roots of intellectual capital deeply em-bedded in social relations and in the structure ofthese relations. Such a view contrasts stronglywith the relatively individualistic and acontex-tual perspectives that characterize more trans-actional approaches for explaining the exis-tence and contribution of firms. Although wehave identified several ways in which facets ofsocial capital may, indeed, reduce transactioncosts by economizing on information and coor-dination costs, we believe that our theoreticalpropositions go much farther in identifyingthose factors underpinning dynamic efficiencyand growth.In so doing, we note that our arguments areconsistent with resource-based theory in so faras that theory highlights the competitive advan-tage of firms as based in their unique constella-tion of resources: physical, human, and organi-zational (Barney, 1991). Those resources found tobe especially valuable are those that are rare,durable, imperfectly imitable, and nontradable(Barney, 1991; Dierickx & Cool, 1989). Among thefactors making a resource nonimitable are tac-itness (Reed & DeFillippi, 1990), causal ambigu-ity (Lippman & Rumelt, 1992), time compression

diseconomies, and interconnectedness (Dierickx& Cool, 1989), as well as path dependenceand social complexity (Barney, 1991; Reed &DeFillippi, 1990). All of these are features inte-gral to the facets of social capital and to itsinterrelationships with intellectual capital.Thus, we suggest that differences betweenfirms, including differences in performance,may represent differences in their ability to cre-ate and exploit social capital. Moreover, at leastregarding the development of intellectual capi-tal, those firms developing particular configura-tions of social capital are likely to be more suc-cessful. Evidence for this suggestion is found instudies of knowledge-intensive firms that havebeen shown to invest heavily in resources, in-cluding physical facilities, to encourage the de-velopment of strong personal and team relation-ships, high levels of personal trust, norm-basedcontrol, and strong connections across porousboundaries (Alvesson, 1991, 1992; Starbuck, 1992,1994; Van Maanen & Kunda, 1989). The frame-work developed here will provide a useful basisfor further testing these propositions about firmdifferences.

In developing our thesis, we have noted sev-eral limitations in our approach. First, regardingsocial capital, our analysis has concentratedprimarily, although not exclusively, on how so-cial capital assists the creation of new intellec-tual capital. However, we recognize that socialcapital also may have significant negative con-sequences. For example, certain norms may beantagonistic rather than su

Documents

Nahapiet 1997