MANAGING INTRACTABLE IDENTITY CONFLICTS · managing these conflicts. There is recognition that...

MANAGING INTRACTABLE IDENTITYCONFLICTS

C. MARLENE FIOLUniversity of Colorado Denver

MICHAEL G. PRATTBoston College

EDWARD J. O’CONNORUniversity of Colorado Denver

Identity is often at the heart of ongoing intergroup conflicts in organizations. Drawingfrom theories of conflict management, social identity, and organizational identifica-tion, we develop the intractable identity conflict resolution model, which delineates amultiphase process by which the conflicting parties’ identities shift in order to permiteventual intergroup harmony.

At a community hospital in the midwesternUnited States, the medical staff complained thattheir CEO and her administrative team wereignoring, and even thwarting, patient initiativesand recommendations essential to improvingthe quality of patient care. This perceived be-havior on the part of administration threatenedthe very essence of who the physicians believedthemselves to be (those having control over med-ical decisions for their patients). Given thatthreat, the physicians retaliated by attacking thevery core of the administrative team’s profes-sional identity (those in control of the organiza-tion’s future) by refusing to buy into and supportthe vision they had for the hospital. This behav-ior led the administrators to withhold furthersupport for the physicians’ recommendations,and it led to the physicians’ eventual costly andpainful attempts to remove the CEO. Over time,the situation spiraled out of control in that theoriginal dispute multiplied exponentially intonumerous seemingly unrelated battles, such asfights over allocation of funds (O’Connor & An-nison, 2002).

Intergroup conflicts are ubiquitous in organi-zations. These conflicts often involve, but are notlimited to, disputes over interests and resources(Kriesberg, 2003). As in the example above, manyconflicts are also characterized by tensionsstemming from differences in how groups fun-damentally define themselves and from threatsto those self-definitions. Numerous scholarshave noted that when identities are implicatedin a conflict, the conflict tends to escalate, en-compassing an ever-widening number of issues(e.g., Northrup, 1989; Rothman, 1997; Rouhana &Bar-Tal, 1998). Parties then become trapped inan ongoing conflict spiral from which they havedifficulty extricating themselves (e.g., Coleman,2003; Diehl & Goertz, 1993; Zartman, 2005).

The example is illustrative of a widespreadproblem not only in health care but in numeroustypes of organizational settings with interde-pendent groups. Research in a variety of areasis converging on the importance of identity dy-namics in understanding ongoing and persis-tently negative interactions in organizations be-tween and among professional groups andadministrators (O’Connor, Fiol, & Guthrie, 2006;Pratt & Rafaeli, 1997), between artists and busi-ness people (e.g., musicians and symphonyboard members; Glynn, 2000), between laborand management (Fisher, 1983; Haydu, 1989), be-tween external and internal constituencies (Dal-ton, 2003; Lewicki, Gray, & Elliott, 2003), and be-tween and among different demographic groups

We thank Kevin Corley, Jill Purdy, Cliff Young, Peter Bry-ant, Sam Gaertner, Greg Oldham, and the faculty at GeorgeMason’s School of Management, the University of SouthernCalifornia’s Marshall School of Business, and INSEAD, aswell as former associate editor Randall Peterson and threeanonymous reviewers for their helpful comments on earlierversions of this manuscript.

� Academy of Management Review2009, Vol. 34, No. 1, 32–55.

32Copyright 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.

(Friedman & Davidson, 1999; Tomlinson &Lewicki, 2006).

Unfortunately, as Rothman notes, “When peo-ple’s essential identities, as expressed andmaintained by their primary group affiliations,are threatened or frustrated, intransigent con-flict almost inevitably follows. For such con-flicts, conventional methods of conflict manage-ment are usually inadequate and may evenexacerbate the problem” (1997: 5). The difficultyof managing such conflicts creates serious chal-lenges for organizations that serve as arenaswhere these conflicts play out. Identity conflictscan alter members’ attributions of behavior anddistort communication (Friedman & Davidson,1999; Northrup, 1989; Rouhana & Bar-Tal, 1998),which can lead to costly errors, decreased learn-ing in organizations, and increased turnover(Humphreys & Brown, 2002). When groups are inconflict, organizations are also less likely to de-rive benefits such as flexibility and creativity,which are associated with diversity and plural-ity in organizations (Williams & O’Reilly, 1998).While some of these outcomes are true of manyforms of conflict, they are exacerbated whenidentity is involved because such conflicts arelong-lasting and expansive, thereby strainingworking relationships and ultimately threaten-ing the organization’s survival. Moreover, thepotential for conflicts based on group identitiesis likely to increase as organizations continue toglobalize and become more diverse (e.g., Ohlott,Chrobot-Mason, & Dalton, 2004), as professionaland nonprofessional roles become increasinglyinterdependent (e.g., Wallace, 1995), and asunions potentially revitalize in a global econ-omy (e.g., Turner, 2005).

Although there is a lack of systematic treat-ment of such long-standing disputes in the or-ganizational literature (Bunker, Alban, &Lewicki, 2004), the nature of identity-based con-flicts has been well documented in the literatureon intractable social conflicts. However, eventhis literature has yet to converge on a model formanaging these conflicts. There is recognitionthat identity beliefs of the conflicting partiesmust change in order to resolve intractable con-flicts (Kelman, 2006; Northrup, 1989), but weknow very little about the nature of the neces-sary identity changes and even less about howand why they occur.

The main purpose of this paper is to addressthese limitations. We draw on theories of con-

flict management, social identity, and organiza-tional identification to propose a model for mov-ing from intractable identity conflicts (IICs) toenduring intergroup harmony in organizations.We take very seriously the notion forwarded byprior researchers that if identity is part of theproblem, it must also be part of the solution(Kelman, 2006; Northrup, 1989). The proposedmodel identifies a multiphase process by whichthe conflicting parties’ identities shift in order topermit eventual harmonious intergroup rela-tions. Following Kilduff (2006: 252), our modeldevelopment is motivated by problems in theworld, not just gaps in the literature. The papercontributes “actionable knowledge,” which, ac-cording to Argyris, describes, explains, and in-forms users not only about “what is likely tohappen under the specified conditions but howto create the conditions and actions in the firstplace” (1996: 392).

INTRACTABLE CONFLICTS AND IDENTITY

Intractable conflicts are protracted and socialconflicts that resist resolution (Burgess & Bur-gess, 2006; Northrup, 1989; Pruitt & Olczak, 1995;Putnam & Wondolleck, 2003). These labels areoften used to describe ethnonational (Crocker,Hampson, & Aall, 2004, 2005), societal (Sen, 2006),and environmental conflicts (Lewicki et al.,2003). Research on intractable conflicts hastended to take one of two forms: (1) an analysisof their characteristics or (2) a description oftechniques for managing them. Identity hasbeen integral to the first perspective; it is largelyabsent in the second.

Characteristics of Intractable Conflicts

Researchers view intractability as a dynamicproperty of intergroup relations, with conflictsbecoming more or less intractable over time(e.g., Burgess & Burgess, 2006; Crocker et al.,2005; Lewicki et al., 2003; Putnam & Wondolleck,2003). There is considerable consensus on thecharacteristics of these conflicts. One centralcharacteristic of intractable conflicts is that theyare long-standing (e.g., Coleman, 2003; Diehl &Goertz, 1993; Rouhana & Bar-Tal, 1998; Zartman,2005), lasting for years or even generations. Forexample, the union-management conflicts atUnited Airlines have lasted for over twentyyears (Bradsher, 2000). Intractable conflicts are

2009 33Fiol, Pratt, and O’Connor

also believed to be pervasive or chronically sa-lient to those involved (Putnam & Wondolleck,2003; Rouhana & Bar-Tal, 1998). Such pervasive-ness is illustrated in a recent survey of hospitalCEOs in the United States, which showed thatproblematic physician-administrator relationswere among their top concerns, second only tofinancial woes (Evans, 2007). Such conflicts alsoinfiltrate multiple spheres of life (e.g., work andnonwork), weaving together identity and re-source-related issues (Rothman, 1997), some ofwhich are not directly related to the initial con-flict. In the example at the beginning of thispaper, an identity dispute over the legitimacy ofpatient quality initiatives expanded to includenumerous seemingly unrelated resource-basedbattles, ending with attempts by the physiciansto remove the CEO. The tendency to expandbeyond the original dispute increases the com-plexity of these conflicts.

Many researchers view identity as being im-plicated in intractable conflicts. Specifically, in-tractable conflicts are characterized by simplify-ing stereotypes and zero-sum conceptualizationsof identity (Azar, 1986; Coleman, 2003; Kelman,1999, 2006; Putnam & Wondolleck, 2003; Zartman,2005). The identities of parties in intractable con-flicts are negatively interdependent such that akey component of each group’s identity is basedon negation of the other group (Kelman, 1999,2006; Putnam & Wondolleck, 2003). That is, forGroups A and B, a salient part of Group A’sidentity is not being a member of Group B, andvice versa. Furthermore, for one group to main-tain its legitimacy, it must delegitimize theother.

In the identification literature, defining whoone is based on who one is not is called “dis-identification” (Dukerich, Kramer, & Parks, 1998;Elsbach, 1999; Pratt, 2000). Groups who arebound up in IICs are in a state of mutual dis-identification, which is strengthened because ofcognitive simplifications whereby parties ig-nore the potential plurality of outgroup mem-bers’ identities. “The foundations of degradationinclude not only descriptive misrepresentation,but also the illusion of a singular identity thatothers must attribute to the person demeaned”(Sen, 2006: 8). To illustrate, Glynn (2000: 290) ob-served that in a conflict between musicians anda symphony board, both sides engaged inheated, mutual disidentification, with musi-cians perceiving the diverse board members as

“money-grubbing weasels” and board membersviewing all musicians as overly idealistic andunwilling to understand the financial picture.Disidentification has a strong evaluative/affective component (Pratt, 2000); in intractableconflicts, disidentification is expressed in strongemotions, including hatred, pride, and fear(Coleman, 2003; Putnam & Wondolleck, 2003;Sen, 2006; Wedge, 1987). Table 1 provides evi-dence of these characteristics in two conflictivesettings: (1) physician-administrator relations inthe hospital described in the opening vignetteand (2) labor-management relations at UnitedAirlines.

Of all the characteristics of IICs noted in Ta-ble 1, identity is not only the most central butalso helps to explain the presence of the others.For example, as a core construct that links theindividual with larger collectives (Cooley, 1922;Northrup, 1989; Tajfel & Turner, 1979), identityexplains why such conflicts are both salient andemotional. Moreover, since attacks on identitylead to information distortion, it becomesclearer why dialogue and negotiation often failand why conflicts persist over time and becomeinstitutionalized. This perspective is supportedby examining the small body of research on howintractable conflicts develop. This research sug-gests that intractable conflicts often begin whena group’s identity is invalidated by anotherparty (Kriesberg, 1986; Northrup, 1989; Zartman,2005). Since identity is central to how peoplemake sense of the world, this invalidation isfollowed by a distortion of information to fitprior beliefs (Friedman & Davidson, 1999;Northrup, 1989; Rouhana & Bar-Tal, 1998). Ascommunication continues to deteriorate, bound-aries between “us” and “them” become morerigid. Finally, conflicts can become normalized(Stock, 2001) as the conflicting parties begin tocollude to continue the conflict (Crocker et al.,2005; Northrup, 1989; Zartman, 2005).

Because of the centrality of identity in theorganizational conflicts we have discussed, andbecause of its importance in characterizing in-tractable conflicts in the literature (Coleman,2003; Crocker et al., 2004; Gray, 2003), we callsuch conflicts “intractable identity conflicts.”The term highlights that intractable conflictsare difficult to resolve largely because partiesare trapped in ongoing mutual disidentification.We now turn to a discussion of resolution at-tempts.

34 JanuaryAcademy of Management Review

TABLE 1Illustrative Examples of Intractable Identity Conflicts (IICs)

Characteristic

Evidence of IIC at a Community Hospital,1980s–2007 (O’Connor & Annison, 2002;O’Connor & Bujak, 2001)

Evidence of IIC at United Airlines, 1980s–2007 (Bradsher, 2000; Bryant, 1994; Maynard& Walsh, 2004; Wong, 2002)

Long-standing Pressure from physicians resulted in removalof the prior two hospital CEOs. A longhistory of competition and mistrust wasagain coming to a head in the late 1990s,with an attempt to oust a third CEO.

As early as 1985, United Airlines pilots wenton strike to protest the persistentdiversification strategies pursued byCEOs who were also hoteliers (e.g.,Edward Carlson and later Richard Ferris).Conflicts continued over the years andbecame especially fierce after 2002, withUnited’s filing for and subsequentemergence from bankruptcy.

Pervasive/chronicallysalient

Everyone seemed to agree about only onething: there was no basis for physician-administrator trust at the hospital. Seniorphysician leaders were unified in theircommitment to remove the CEO, who wasclearly defined as the enemy. Theseunified physicians were seen as the enemyfrom the point of view of senioradministrators. Projects forwarded byphysician leaders were typically ignored,and the administrative vision wasgenerally disregarded by the physicians.

For much of the period after 1985,employees (especially in the ALPA andIAM unions) were openly critical of themanagement team. Moreover, the crisisremained intense. For example, over aseven-year span, employees made fiveattempts to secure some ownership of thecompany before they succeeded.

Expansion beyondoriginal dispute

One thing was clear: collaboration wasdead, and there was little agreement aboutthe causes of the tragedy. Medical staffleaders and senior administrators couldnot even agree about who should facilitatethe upcoming retreat in the mid 1990s.Medical staff demanded that it be aphysician, but this was unacceptable toadministrators, who did not believe theirviews would be appropriately represented,and vice versa. Unified action amongphysicians and administrators was at anall-time low, and patient and communityinitiatives were hopelessly at a standstill.

Original disputes between pilots and themanagement team were about jobsecurity and wages. Over time, conflictspread to other areas, including mergersand acquisitions, the dropping of Pacificroutes, issues of employee ownership,and issues associated with the company’srecent bout with bankruptcy. Employee-management conflicts expanded overtime to an increasing number of unions(e.g., flight attendants).

Simplifyingstereotypes

Administrators were seen by physicians ashaving little understanding of orappreciation for clinical needs and littlecommitment to the well-being of patients.Physicians were seen by administrators asneither understanding nor appreciating theimpact of their choices on costs and theirimplications for long-term financialviability.

United’s management team viewed pilotsand other employees as having little tono understanding of finance andcorporate strategy. Employees, in contrast,viewed management as too self-interested and uncaring. They feltmanagement did little to protect them.Given management initiatives on suchissues as seniority and subsequentemployee reactions, there is some indirectevidence that management was viewedas not understanding airline culture andas being too driven by the bottom line.

Zero-sumconceptualizations

Physicians believed they must take control ofthe hospital in order to improve quality ofcare. Administrators believed they musttake control of the hospital in order topreserve the financial viability of thiscommunity asset.

Despite the damage to the airline becauseof poor employee-management relations,management teams and labor unionspursued their own interests separately,with one group’s gains seen as the others’loss. For example, pilots and some other

(Continued)


Managing IICsCrocker et al. argue that while “much analyt-

ical work has focused on the causes of theseconflicts and the forces that contribute to theirintractability,” there has been much less re-search on “ending conflict in the so-called in-tractable cases” (2004: 4). Our review of the lit-erature suggests that research has yet tocoalesce around a specific set of techniques ormodels for resolving or managing intractableconflicts. One possible reason for this fragmen-tation may be that much of the work in this areahas focused on specific tactics that are useful inmanaging only certain aspects of the conflict,eschewing prescriptions for managing suchconflicts from their earliest stages up throughthe establishment of enduring intergroup har-mony. For example, research has addressed themanagement of emotion through art, storytell-ing, or joking (e.g., Maiese, 2006; Retzinger &Scheff, 2000) and the management of stereotypesthrough the shifting of conflict frames (e.g.,

Gray, 2003; Putnam & Wondolleck, 2003;Shmueli, Elliott, & Kaufman, 2006). More generaltactics have also been proposed, including thedelineation of the general skills (e.g., listeningand humility) needed by conflicting parties andtheir mediators (e.g., Portilla, 2006), along withspecific suggestions on how to run a workshopfor disputing parties (e.g., Kelman, 2001; Kelman& Cohen, 1986). Some fuller-scale approaches ormodels have been developed—largely to aidthird-party mediators in such conflicts1—in anattempt to embrace larger portions of the con-flict management process (e.g., Linskold, 1986;Rothman, 1997). However, these models have

1 These models include ARIA—Antagonism, Resonance,Intervention, Action (Rothman, 1997); GRIT—Graduated andReciprocated Initiatives in Tension reduction (Linskold, 1986;Tomlinson & Lewicki, 2006); MACABE—Motivation, Affect,Cognition, Behavior, and the surrounding Environment(Pruitt & Olczak, 1995); and transformative mediation (Bush &Folger, 1994; Spangler, 1993).

TABLE 1(Continued)

Characteristic

Evidence of IIC at a Community Hospital,1980s–2007 (O’Connor & Annison, 2002;O’Connor & Bujak, 2001)

Evidence of IIC at United Airlines, 1980s–2007 (Bradsher, 2000; Bryant, 1994; Maynard& Walsh, 2004; Wong, 2002)

employees believed that the managementteam’s success in merger and acquisitionscame at a cost to their seniority systemand their jobs. The management team, incontrast, saw failures in these endeavorsas major blows to its future profits.

Mutualdisidentification

Administrators clearly identified themselvesas not being like those physicians whomthey viewed as complaining a lot, onlyinterested in their own personal well-being, inflexible, and unwilling to examinetheir own failings. For their part,physicians took pride in definingthemselves as distinctively different fromadministrators, whom they viewed asfocused on reducing costs (versus patientwell-being) and padding their resumes tomove on (versus having a long-termcommitment to the community). Given thateach group defined itself in terms of beingdifferent from the other, neither wanted tolisten to or be associated with the other’sperspective.

The management team defined itself as notbeing like pilots and some other employeeswho only thought about their own interestsand not the company’s future profitability.Pilots and other employees disidentifiedthemselves from managers who, theybelieved, did not understand the nature ofthe industry or the work of the employees.Both groups disparaged the other side.For example, one article stated that“management says United is dragged downby the most expensive labor contracts inthe business. The unions maintain thatmanagement has made several horrendousbusiness moves, especially an ill-fatedattempt to buy US Airways that wasblocked by the Justice Department lastyear. Now, they say, United is makinganother by threatening a bankruptcy filing,a move that would probably wipe out thevalue of UAL stock” (Wong, 2002: 2).


typically used classic negotiation and other con-flict management techniques, largely ignoringthe role of identity.

In the rare cases where identity is mentionedas being critical to resolving intractable con-flicts, the specifics for its management are leftvague or undefined. For example, Northrup ar-gued that “if change occurs in the identities of atleast one of the parties, the chances for long-term change are greatly increased, particularlyif the change involves core aspects of identitythat are directly related to the conflict” (1989: 78).However, she did not specify how such identitychange occurs. The concept of identity negotia-tions has also been proposed (Bar-Tal, Kruglan-ski, & Klar, 1989; Kelman, 2006; Rouhana & Bar-Tal, 1998). Here again, however, this work hasfailed to specify how to engage in such a pro-cess in order to remove “the negation of theother party as a central component of each par-ty’s own identity” (Kelman, 2006: 23).

To summarize, our review of the intractableconflict literature suggests that while identity isbelieved to be central to the formation and con-ceptualization of intractable conflicts, there hasnot been (to our knowledge) a systematic at-tempt to understand how the management ofidentity may help to resolve intractability andfacilitate sustained intergroup harmony. Ourpurpose here is to address this gap so as toenrich theory and build actionable knowledge.

A MULTIPHASE MODEL FOR IIC RESOLUTION

The ultimate goal of our model is to delineatethe phases by which groups mired in IICs canmove toward enduring harmonious intergrouprelations. The Cambridge International Dictio-nary defines harmonious as “peaceful,” whichdenotes agreement and accord, in addition to alack of conflict. The model assumes that peo-ple’s readiness to begin a change process doesnot, in and of itself, lead directly to intergroupharmony. Rather, readiness provides the foun-dation for three distinct identity shifts throughwhich the conflicting groups must pass in orderto achieve enduring harmony: decoupled inter-group identities, subgroup identity security, anddual identity strength. Each phase of our modelprovides a necessary but not sufficient condition

for enduring harmony, and successful comple-tion of the prior phase is necessary to carry outeach subsequent phase of the model.

Partly as a function of the different literatureon which we draw, particularly social identitytheory and organizational identification, muchof the process we describe is focused on individ-ual identity dynamics within an intergroup set-ting. However, the overall process is inherentlymultilevel in that individual identity beliefs arelower-level constructs (e.g., identity of an airlinepilot) that aggregate to higher-level constructs(e.g., security of pilots’ identity as a group). Ul-timately, our model proposes relationshipsamong higher-level constructs (e.g., nature andstrength of group-level identities and harmoni-ous intergroup relations).

Figure 1 shows the four phases underlying theIIC resolution model and the process interven-tions and states that mark the passage throughthese phases. The ovals on the left of the figuredepict the interventions thought to lead to thestates shown in the shaded boxes. We reviewevidence but do not provide propositions forthese main effects since (1) each of them hasbeen discussed in existing research and (2) theresults of these main effects are not consistent.To explain these inconsistencies, we forwardpropositions about how previous phases affectsubsequent ones.

We begin our discussion of the model with theultimate end state—enduring intergroup har-mony—and discuss the process interventionsthat have been proposed to achieve such harmo-nious relations. We then note the shortcomingsof these interventions in the context of intracta-ble conflict and propose that the accomplish-ment of the model’s prior state (strong dual iden-tities) makes these interventions more effective.We continue with this pattern, using the limita-tions of current theorizing as the motivation forproposing each of the moderating effects. Begin-ning our discussion of the model with the endstate allows us to demonstrate at each phasethe need for achieving the prior state in order tomove on to the subsequent one.

Beginning at the top of the figure, we knowthat promoting integrative goals and structuresmay lead to intergroup harmony. However, inthe face of intractable conflicts, such goals andstructures are likely to be threatening unlessboth parties can begin to identify with the su-perordinate identity (e.g., a hospital) in addi-


tion to their own subgroup identity (e.g., physi-cian or administrator)—known as holding dualidentities. Research suggests that promoting si-multaneous differentiation and unity of inter-group membership may promote dual identitiesfor the members of each group, but only whenthe superordinate identity does not threaten thedistinctiveness that provides each group itsidentity security. However, the prescribed inter-ventions for developing identity security arenot likely to succeed unless the mutual disiden-tification that is at the heart of IICs is resolved

by decoupling the intergroup identities. Finally,identity decoupling will not result from inter-ventions to promote participants’ ability tomindfully reconsider the nature of intergrouprelations unless they are ready to come to thenegotiating table (known as conflict ripeness).In sum, the move from one state to the next mustbe sequentially ordered to lead to the desiredoutcome of enduring intergroup harmony. Thefollowing expanded descriptions of each phaseare organized around the components of the IICresolution model depicted in Figure 1.

FIGURE 1Intractable Identity Conflict (IIC) Resolution Model


Promoting Integrative Goals and StructuresLeads to Enduring Intergroup Harmony—ButNot Always

The challenge of managing intergroup con-flicts has been a central and consistent theme inthe study of intergroup relations for at least acentury (Sherif, Harvey, White, Hood, & Sherif,1961; Sumner, 1906). The focus of this work hastended to be on ways to get conflicting sub-groups to see themselves as working togetherharmoniously toward common superordinategoals that each of the subgroups can identifywith but that cannot be achieved by any singlesubgroup. Although a number of studies haveprovided evidence consistent with the view thatconflicting groups can learn to work togetherharmoniously if given a superordinate goal (e.g.,Kahn & Ryen, 1972; Worchel, Andreoli, & Folger,1977), the scope of these positive findings ismostly limited to research using ad hoc groupsthat have no history together. In studies of real-life or preexisting subgroups, the findings havebeen inconclusive, indicating that promoting su-perordinate goals does not always reduce con-flicts and sometimes even worsens them (e.g.,Brown, 1978; Skevington, 1980).

A related approach for bringing groups withdifferent orientations together has been struc-tural integration to promote self-categorizationof the subgroups as on the same team. The ra-tionale is that such integration will lead to bet-ter understanding of each other and a greaterawareness of the many similar values and atti-tudes of the “other,” despite the differences(Cook, 1984). Assuming that ignorance maycause prejudice (Stephan & Stephan, 1984), thisapproach seeks to reduce ignorance in order toenhance intergroup harmony. Structural inte-gration mechanisms include promoting directcontact among the parties, establishing liaisonroles, and creating integrated teams and/or de-partments. Here again, support is mixed. Forexample, increasing intergroup contact some-times leads to greater intergroup harmony but togreater hostility at other times (Desivilya, 1998;Worchel, 1986).

Our earlier discussion of the unique nature ofIICs points to why conflicts may actually worsenif a common set of superordinate goals and/orstructural linkages is imposed on group mem-bers. If groups share a long history of beingbound together by mutual disidentification, a

common goal would seem to be heresy. In fact,since each subgroup’s identity may well bethreatened by a superordinate goal that is ac-ceptable to the other subgroup, members arelikely to react with increased levels of disiden-tification. Structural mechanisms, too, will likelyhave little impact at best. As we noted earlier,one of the characteristics of intractable con-flict is that parties in the conflict tend to filternew information to conform with their beliefs(Northrup, 1989). The new information gainedthrough structural integration, therefore, is notlikely to alter the state of ignorance about theother. In fact, it is likely to be filtered in a selec-tive way that provides further support for thenegative stereotypes.

Significant evidence from studies of U.S.health care systems suggests that decades ofattempts to structurally integrate hospitals andphysicians (through joint ventures, mergers, orother structural mechanisms) and decades ofvisioning retreats aimed at unifying them to-ward similar goals have had limited success(O’Connor et al., 2006). For example, one com-mon way to structurally integrate physiciansand administrators in the 1980s and 1990s in theUnited States was through hospitals’ direct own-ership of physician practices. The logic was thatif hospitals owned physician groups, they wouldbe aligned because they would share the samegoals and objectives. For the most part, thesevertically integrated systems failed to producethe intended results, often leading instead to adecrease in physician productivity and an in-crease in ill sentiment toward the system (Holm,2000). Under intractable conflict conditions thereis little space for harmonious relations to takeroot from such integrative activities.

Social identity theorists (Tajfel, 1978; Tajfel &Turner, 1979) have similarly noted that attemptsto bring subgroups together, highlighting theironeness in working toward one set of superor-dinate goals, might produce the unintended re-sult of actually driving them farther apart. Horn-sey and Hogg (2000), for example, found thathumanities and math-science students showedthe strongest intergroup bias when the superor-dinate category (university student) alone wasmade salient. They argued that since socialidentities derive from group memberships andcomparisons with other groups, superordinategoals that focus exclusively on what groupshave in common can threaten a group’s distinc-


tiveness. Such threats can lead to increased at-tempts to differentiate one’s group from theother and to an increased bias against the othergroup (Crisp, Stone, & Hall, 2006). As Huo andMolina note, “An emphasis on a common iden-tity can have a boomerang effect—motivatingthe desire to defend the neglected subgroup andthus highlighting the very group differences thestrategy intended to attenuate” (2006: 360).

These dynamics are likely to be even moreextreme in groups that share a long history ofmutually disidentifying with each other, espe-cially if the superordinate goal is perceived toencompass the interests of the outgroup (Kries-berg, 2003). When identities are perceived aszero sum, any gain for the outgroup will be seenas a loss for the ingroup. For example, at UnitedAirlines, managers and pilots have historicallyframed mergers and acquisitions as win-loseendeavors (e.g., if managers “win,” pilots loseseniority—see Table 1). So while we agree thatthe development of superordinate goals can pro-vide a needed direction for groups and integrat-ing structures can enhance their understandingof one another, these efforts are likely to fail ifgroup members have been engaged in IICs.

The Moderating Role of Dual Identity Strength

Research has provided evidence for a some-what counterintuitive approach for managingidentity perceptions to bring about intergroupharmony (e.g., Gaertner, Bachman, Dovidio, &Banker, 2001; Gaertner & Dovidio, 2000). Whenboth subgroup and superordinate identities aremaintained (referred to as dual identities), onecan anticipate greater acceptance of the oppos-ing subgroup. It follows that the adoption ofdual identities will lead to greater acceptance ofintegrative goals or structures that incorporatethe opposing subgroup’s interests.

Dual identities have, in fact, been associatedwith reduced intergroup bias by group mem-bers. For example, in the Hornsey and Hogg(2000) university student experiment describedabove, the least amount of intergroup bias oc-curred in the dual identity condition, suggestingthe least likelihood of resisting integrativegoals or structures. These results were recentlyreplicated by Crisp et al. (2006), who found thatstudents whose disciplinary training (e.g., hu-manities or science) and university affiliation

were both activated were likewise resistant tointergroup bias. Gaertner and Dovidio (2000) re-ported further empirical evidence of the poten-tial for a dual identity approach to minimizeintergroup bias (and, thus, to pave the way forintergroup harmony) by examining both ethnicand organizational identities: when members ofethnic subgroups in a multiethnic high schoolexperienced a dual identity—that is, identifyingwith both their ethnic subgroup and the schoolas a whole—there was decreased intergroupbias. In a similar vein, Bizman and Yinon (2004)noted that the perception of dual identities inIsraeli secular-religious contexts was the onlypredictor that individuals would use a problem-solving rather than a contentious mode for man-aging conflict. We theorize that holding bothidentities predicted more of a problem-solvingmode because it led to less intergroup bias,opening group members to the possibility ofconceiving of themselves as having enough incommon with their outgroup to look for solutionsacceptable to both groups. Building on thesefindings, we propose that strong dual identitiesstrengthen the impact of efforts to promote inte-grative goals and structures because they re-duce intergroup bias and open the door for har-monious interactions.

Proposition 1: The stronger the dualidentities, the greater the probabilitythat promoting integrative goals andstructures will lead to enduring inter-group harmony.

Promoting Intergroup Differentiation and UnityLeads to Dual Identity Strength—But NotAlways

The literature on dual identities (e.g., Gaert-ner & Dovidio, 2000; Hornsey & Hogg, 2000) sug-gests that priming people to perceive them-selves as both similar to and different frommembers of other subgroups facilitates theadoption of dual identities. These ideas haveyet to be widely incorporated into mainstreamorganizational conflict theories. As Brewer(2001) has pointed out, the implicit assumptionin much of the work on intergroup identity con-flict has been that members’ attachment to theiringroup will inevitably lead to outgroup hostil-ity, making it difficult, if not impossible, to con-ceive of developing dual identities encompass-


ing both a subgroup and a superordinateidentity that includes the outgroup. Yet re-searchers have failed to find any straightfor-ward negative correlation between ingroup pos-itivity and intergroup hostility (Hinkle & Brown,1990). Even at a physiological level, positiveand negative evaluative processes appear tooccur within different, independent neurologicalsystems that may or may not be reciprocallyactivated (Brewer & Brown, 1998). So increasedpositive affect toward the ingroup does not nec-essarily result in increased negative affect to-ward the outgroup, making it possible to holddual identities without significant cognitive dis-sonance.

Roccas and Brewer (2002) have referred to thecapacity to hold dual identities as social iden-tity complexity, which reflects the degree ofoverlap perceived to exist between the sub-groups one is a member of. The less individualsperceive the subgroups they belong to as shar-ing the same members and the same prototypi-cal attributes, the more complex their socialidentities will be. Of course, the complexity re-quired for dual identity development varies. Atone extreme, when there is a sense of continuitybetween the subgroup and superordinate iden-tities (Van Knippenberg & Van Leeuwen, 2001),promoting both is relatively straightforward.They may easily coexist even when different, aslong as they are not inherently in conflict withone another or mutually exclusive. For example,in the aforementioned subgroups in the multi-ethnic high school field study, groups had dualidentities that were unrelated—an ethnic iden-tity and a superordinate high school identity(Gaertner & Dovidio, 2000). Similarly, in lab stud-ies subjects have been successfully primed toadopt noncompeting dual identities (Hornsey &Hogg, 2000). In these settings there was littleconflict between subgroup and superordinateidentities; they could coexist without threaten-ing each other.

At the other extreme, strong dual identitiesare more difficult for people to adopt, andgreater identity complexity is required when theidentities are perceived as conflicting with oneanother. Even if leaders attempt to promote newidentity frames that are not at odds with eachother (i.e., frames not based on mutual disiden-tification), in settings with a long history of in-tractable conflicts, people may distort informa-tion (Northrup, 1989) and ignore the new frames.

Implicit and overlearned conceptions of self inrelation to the “other” often linger long after theactual context shifts (Pelham & Hetts, 1999). Ifthe subgroup and superordinate identities con-tinue to be seen as opposed to one another, it ishighly unlikely that people will adopt strongdual identities, even if they are primed to per-ceive both intergroup unity and differentiation.

This may explain why Gaertner and his col-leagues (Gaertner et al., 2001; Gaertner, Dovidio,Nier, Ward, & Banker, 1999) found that in thecontexts of both banking mergers and blendedstepfamilies, simultaneous perceptions of sub-group differentiation and intergroup unity wereassociated with less favorable intergroup rela-tions than perceptions of a single unified group.The authors speculated that the continued sa-lience of the earlier subgroup identities (alongwith the superordinate identity) may have beenperceived as threatening the primary goal ofcombining the subgroups. An alternative andperhaps more likely interpretation of their re-sults, given the pervasive nature of threats topreexisting subgroup identities (Brewer, 2001;Hornsey & Hogg, 2000), is the reverse but com-plementary interpretation: the superordinateidentity may have been emphasized as the moresalient of the two identities, thus dominatingand threatening each subunit’s preexistingidentity. Since amalgamation was seen as theprimary goal of the mergers/blended families, itis reasonable to assume that intergroup rela-tions suffered in the dual identity condition be-cause members may not have felt secureenough about the value and positive distinctive-ness of their preexisting subgroup for suffi-ciently strong dual identities to develop or bemaintained.

The Moderating Role of Subgroup IdentitySecurity

We suspect that if the subgroup identities hadbeen sufficiently secure in their distinctivenessin the Gaertner et al. (1999, 2001) merged banksand stepfamily studies, the superordinate iden-tity would have been less threatening to theprior subgroup identities, leading to greater ac-ceptance of strong dual identities and an asso-ciated decrease in intergroup bias. Berry’s (1991)work suggests that it is security in one’s ownsubgroup identity, not simply strong positive at-titudes toward one’s group, that predicts toler-


ance of another group. Building on this work, weargue that identity security (a feeling of safetyor protection) may play a more important rolethan identity strength (positive attitudes towardone’s own group) in reducing potential threats.Failure to distinguish between security andstrength may help to explain why no clear link-ages between ingroup positivity and outgroupnegativity have been found (Brewer, 2001).

Researchers have found evidence supportingthe notion that security in one’s subgroup iden-tity is associated with greater acceptance of asuperordinate identity. Identity security, for ex-ample, appears to be integral to building plu-ralistic societies where subgroup and superor-dinate identities are held simultaneously(Hewstone & Brown, 1986; Mummendey & Wen-zel, 1999). Huo and Molina’s (2006) study of anethnically diverse population of Californiansshows that minority groups (e.g., African Amer-icans and Latinos) that experience the securitythat comes from subgroup respect are morelikely to feel positive toward a common Ameri-can identity (cf. Sidanius, Feshbach, Levin, &Pratto, 1997). And although they did not start outwith the condition that subgroup identities arezero sum, Haslam, Eggins, and Reynolds (2003)developed a model (ASPIRe) suggesting thatsubgroups must first be secure in delineatingtheir own needs (i.e., “subcasing”) before sub-group goals can be integrated with superordi-nate goals. Common to these treatments is theassumption that security in one’s subgroupidentity is related to its validated distinctive-ness from the superordinate identity. As we havenoted, attempting to build security by promotinghow subgroups are similar is likely to backfire(Huo & Molina, 2006).

While subgroup identity security is importanteven in group settings where conflicts are rela-tively recent (e.g., in newly merged banks), thepotential for the rejection of superordinate iden-tities, and therefore dual identities, is evengreater when subgroups have been steeped inlong-term intractable conflicts (see Table 1). Wefurther argue that identity security—in the formof validated, distinct subgroup identities—is es-pecially important if subgroup and superordi-nate identities are perceived as competing withone another. Secure subgroup identities createthe psychological safety necessary to acceptdual identities with less defensiveness. Thisleads us to propose that efforts to simulta-

neously promote intergroup differentiation andunity in order to develop dual identities in theface of intractable conflicts are more likely tosucceed if members feel secure and validated intheir own distinctive subgroups.

Proposition 2: The more secure thesubgroup identities, the greater theprobability that the promotion of si-multaneous intergroup differentiationand unity will lead to the develop-ment of strong dual identities.

Promoting Positive Distinctiveness Leads toIdentity Security—But Not Always

We have argued that subgroup members willfeel less threatened by a superordinate identityif they feel secure in their own subgroup’s dis-tinctiveness. A number of researchers have de-scribed distinctiveness-inducing techniques,which often highlight the unique and attractivequalities of the subgroup (Ashforth & Mael, 1989;Dutton, Dukerich, & Harquail, 1994; Fiol, 2002).Linguistic techniques, such as the use of inclu-sive referents (“we”), which strengthen groupbonds and make groups feel special, serve asimilar distinctiveness-defining role (Cheney,1983; Fiol, 1989). Finally, engaging subgroupmembers in common tasks that ignite their pas-sion about fulfilling a meaningful purpose andexplaining the unique purpose of their commontasks are other means of developing secure sub-group identities (Fiol, 2002; Pratt, 2000). Implicitin these techniques is not just that groups aremade to feel distinctive but that such assess-ments are made by building up the ingroup—that is, by promoting positive distinctiveness.

None of the research on promoting subgroupdistinctiveness has addressed the challenges ofachieving security when subgroups are mired inIICs. We have noted that in such cases clearlines of distinctiveness are indeed drawn be-tween the groups, but they are lines that pro-mote mutual “dehumanization” (Northrup, 1989:72)—distinctiveness based on negative compar-isons—which we refer to as negative distinctive-ness. Subgroup members thus may well feel pos-itively secure in relation to the negative other,but this type of security based on negative com-parisons is likely to block, rather than facilitate,the development of dual identities.


Research has described how identity securitymay be promoted by making negative compari-sons, such as promoting identification throughantithesis (Cheney, 1983). In the case of IICs,subgroups mutually disidentify by definingthemselves in opposition to the “negative other.”As Gray argues, when intractable conflicts in-volve negative identity frames (e.g., mutual dis-identification), such “identity challenges callinto question how a group [the ‘other’] has de-fined itself and even its very right to exist” (2003:21). This occurs, for example, when physiciansdefine hospitals primarily in terms of places ofcare giving and view business-minded admin-istrators, at best, as a necessary evil (O’Connoret al., 2006).

Such negative sources of a group’s distinctive-ness are likely to prevent members from seekingpositive distinctiveness. In other words, thesomewhat ironic condition of groups in IICs isthat they have developed a sense of identitysecurity based on their mutual negation of oneanother, therefore sensing no urgency to searchfor other sources of positive distinctiveness (Put-nam & Wondolleck, 2003). Thus, for example,mutual disidentification continued unabated atUnited Airlines, even when the organizationwas in a state of bankruptcy (see Table 1). With-out the removal of negative distinctiveness,there is likely to be limited receptiveness totechniques for promoting positive distinctive-ness in intractable conflict situations. That is,each subgroup already feels good about its su-periority vis-a-vis the other, thus requiring nofurther positive distinctiveness.

The Moderating Role of Decoupled Identities

Relinquishing mutual disidentification as theprimary source of each subgroup’s identity se-curity requires some acknowledgement of thevalidity of the other subgroup’s world view, evenif there remains bitter disagreement (Kelman,2006). For this to occur, the mutually disidenti-fied identities must be decoupled or disentan-gled from one another so that one subgroup’sidentity security is not dependent on the other’sdemise. Intergroup identity decoupling opensthe door for positively based subgroup identitysecurity to take the place of the negativelybased security.

The benefits of disentangling subgroup iden-tities bound by mutual disidentification can be

extrapolated from earlier research. For example,some have argued that under conditions wheregroup identities are in conflict but must bemaintained in an organization, identities shouldbe compartmentalized (Pratt & Foreman, 2000)before attempting to reconcile or bridge them(Pratt & Corley, 2007). Decoupling in our model,however, does not entail physical or temporalseparation or compartmentalization of sub-groups (cf. Breakwell, 1986; McCall & Simmons,1978). Rather, it refers to a psychological state ofseparation. Because groups engaged in IICsidentify themselves, in part, as not being mem-bers of the opposing subgroup, decoupling en-tails cutting out those identity elements that linkthe two parties in the destructive dance of mu-tual disidentification. It requires groups to let goof the assumption that “they” must lose in orderfor “us” to win, opening the possibility for thebattling parties to relate in less destructiveways.

Most of the work on identity decoupling comesout of research on intractable conflicts and in-ternational diplomacy. In particular, Kelman’s(1999, 2006) and Kelman and Cohen’s (1986) ex-tensive efforts to engage Israelis and Palestin-ians in decoupling from their negative interde-pendence stand as exemplars of the need forthis process to occur before either side can gaina strong positive sense of self not linked to thedestruction of the other. Participants and ana-lysts agree that these decoupling efforts contrib-uted significantly to paving the way to the earlynegotiations between the parties by creating anenvironment where the legitimacy of each side’sconcerns was upheld and mutual disidentifica-tion was broken down.

Building on this work, we argue that identitydecoupling must precede attempts to developmore secure positive subgroup identities be-cause it removes the negative distinctivenessthat limits group members’ receptiveness totechniques for promoting identity securitythrough positive distinctiveness. Decoupling theidentities before attempting to make each onemore secure also reduces the risks often associ-ated with “strengthening” former enemy groups(i.e., an enemy group with a strong identity maybe more of a threat if the situation is seen as zerosum). Identity decoupling thus offers the oppor-tunity for members to respond to techniques forpromoting their own (and other) subgroup’s pos-


itive distinctiveness and to become secure intheir own sense of self.

Proposition 3: The more complete theintergroup identity decoupling, thegreater the probability that promotingpositive ingroup distinctiveness willlead to the development of subgroupidentity security.

Promoting Mindfulness Leads to DecoupledIntergroup Identities—But Not Always

For subgroup members to revise their overlysimplified negative views and polarizing stereo-types of the outgroup, they must develop theability to let go of well-established patterns ofperceiving and to begin to see things anew—acapacity that has been referred to as mindful-ness (Fiol & O’Connor, 2003; Langer, 1989; Weick,Sutcliffe, & Obstfeld, 1999). Mindfulness hasthree key characteristics: (1) the creation of newcategories of meaning, (2) openness to new in-formation, and (3) an implicit awareness ofmultiple perspectives. The ability to at leasttemporarily suspend beliefs that have led tosubgroups’ negatively based identity securityderives from the development of these threecharacteristics.

Mindfulness results from a reluctance to sim-plify. Fiol and O’Connor (2003) described mind-less bandwagon behaviors as resulting fromoversimplified interpretations that share someof the same collusive qualities of high levels ofintergroup conflict intractability. Pratt andDoucet (2000: 213) illustrated this black-and-white dynamic by describing how a physicianwas perceived as “anti-patient and anti-profes-sion” when he began to cooperate with a man-aged care provider. The physician’s behaviorillustrates a break from mindless zero-sumbandwagon perceptions and behaviors. His re-luctance to simplify allowed for the possibilitythat managed care providers might not be allbad, thus opening the space to reconsider rela-tionships with them. In the international arena,Kelman (1999, 2006) and Kelman and Cohen(1986) similarly engaged participants in prob-lem-solving workshops, the main purpose ofwhich was to help them form “more differenti-ated images of the enemy” in order for the neg-ative intergroup dynamics to be dismantled(Kelman, 1998: 194).

The mindfulness literature suggests waysthat such differentiation can be developed andnurtured (Fiol & O’Connor, 2003; Langer, 1989;Weick et al., 1999), often through interventionsby third parties. For example, third parties mayredefine, for conflicting parties, the meaning of“paying attention” by asking new kinds of ques-tions, rather than demanding answers to pre-vailing questions (having well-developed re-sponses). They may express conclusions intentative rather than absolute terms and main-tain ambiguity around relevant issues for aslong as possible, resisting the temptation tofind early closure and clarity for the problem-atic issues. All of the strategies for developinggreater mindfulness have to do with shiftingpeople’s assumptions so as to discover unseenpossibilities.

We agree that participants’ abilities to mind-fully rethink the nature of the intergroup rela-tions, to see new categories of meaning besides“us versus them,” and to be open to new infor-mation and perspectives are necessary condi-tions for intergroup identity decoupling, butthese are not always sufficient conditions. Forexample, although physicians may mindfullyunderstand the need to control costs in order toprotect the long-term survival of their local hos-pital, they may nevertheless still be unwilling toconstrain spending in their personal attempts tosave a particular patient’s life. In the case ofintergroup identity decoupling, mindfulness cre-ates the capacity or ability for people to engagein psychologically breaking down the negativecoupling of identities. It does not create the will-ingness to do so.

The Moderating Role of Readiness/Ripeness

Even if mindfully capable of doing so, whywould parties agree to take the first steps ofdecoupling, given that their very identity secu-rity depends on the identities remaining nega-tively coupled? As we noted, intergroup identitydecoupling when groups are engaged in IICsrequires parties to give up that part of theiridentity that is based on the “negative other.”This is extremely difficult, since in these types ofconflicts, negation of the other is not a periph-eral or marginal element of each subgroup’sidentity that can easily be discarded (Kelman,2004). In fact, the intergroup battles themselvesare often a large part of each group’s identity


(Ashmore, Jussim, Wilder, & Heppen, 2001;Northrup, 1989), making it difficult for the con-flicting parties to see a way out or to see oneanother as anything but the enemy. As Zartmansuggests, holding zero-sum identities “cloudsthe parties’ perceptions of a hurting stalemate,since it provides the righteous cause that thriveson pain” (2005: 60). As a result, actions are inter-preted as unilateral, all-or-nothing affairs.

Cutting away the negative sources of identitysecurity (identity decoupling) entails a willing-ness to at least temporarily enter into a state ofidentity insecurity—some level of uncertaintyabout one’s identity. In the literature on interna-tional diplomacy and intractable conflicts,readiness to take on identity insecurity is de-scribed as ripeness. For a conflict to be consid-ered ripe, both parties must be motivated toresolve it (Pruitt & Olczak, 1995). Ripeness in-volves a readiness to commit to a “change in thenature of the relations of the parties from a com-petitive, hopeless, destructive orientation to-wards a more cooperative co-existence with po-tential for mutual gain” (Coleman, 2000: 302).

Conflict ripeness is not likely until people per-ceive either great threat if they fail to give upthe old ways of relating or great benefit as aresult of doing so. An existing body of researchon ripeness has identified two main types ofantecedents: potential for significant sharedpain and/or mutual gain. Specifically, recent,impending, or near-miss crises; a commonthreat; the perception that the current situationis hurting both parties; or added incentives tocooperate (Coleman, 2000; Fiol & O’Connor, 2002;Pruitt & Olczak, 1995; Zartman, 2005; Zartman &Rubin, 2000) have all been posited to serve asmotivators for subgroup members to reconsidertheir relationship with the outgroup. Coleman(2000) further has argued that negative forcesthat induce conflict ripeness may actually addadditional stress to an already burdened rela-tionship. He suggests that removing obstaclesthat prevent ripeness (e.g., decreasing mistrust;Tomlinson & Lewicki, 2006) may be a more effec-tive means of encouraging ripeness. Whateverthe driver, conditions have to be strong enough(e.g., significant pain or gain) to cut throughmembers’ distorted information processing inorder to recognize the necessity to engage inchange. Once motivated to come to the table,individuals in conflicting groups are likely to be

more willing to mindfully disentangle theiridentity from that of their “enemy.”

Proposition 4: The riper the conflict,the greater the probability that effortsto promote participants’ mindfulnesswill lead to decoupled intergroupidentities.

In summary, the IIC resolution model pro-poses that the path from intractable conflict toenduring intergroup harmony involves movingthrough a series of states—from conflict ripe-ness to decoupled intergroup identities, to sub-group identity security, to strong dual identities,and, finally, to enduring harmonious intergrouprelations. After the initial state of ripeness, eachstate is accomplished through the application ofa specific set of interventions, moderated by theprior state (see Figure 1). To begin the process,subgroup members must break free from theirmutual disidentification, an outcome facilitatedby conflict ripeness and the promotion of mind-fulness. The resulting decoupled identities maybe insecure, especially since mutual antipathymay have been a large component of each sub-group’s identity. Thus, we have suggested thatleaders strengthen ingroup identity security bypromoting the groups’ positive distinctiveness.Once subgroup identities are validated and se-cure, they are more likely to provide the solidfoundation from which to adopt a dual identity,an outcome that is facilitated by the simulta-neous promotion of intergroup differentiationand unity. And only when group members holdstrong dual identities will the promotion of inte-grative goals/structures be likely to lead to sus-tained intergroup harmony.

Groups may not need to begin this process atthe “beginning” (developing readiness) if readi-ness already exists; they may not require inter-group identity decoupling if they are not cur-rently experiencing mutual disidentification;promoting positive ingroup distinctiveness maybe unnecessary if the groups are already securein their (positive) subgroup identities; and pro-moting intergroup differentiation and unity maynot be needed if they already hold strong dualidentities. However, our model posits that eachof the prior states must be developed if it doesnot already exist in order to achieve subsequentstates in the model.

Although we have delineated the phases ofthis process as strictly sequential and distinct


from each other, we recognize that it will oftenbe necessary to begin the subsequent phasebefore fully completing the preceding one. Thebeginning of decoupling encourages greaterripeness; positive identity security facilitatesletting go of aspects of disidentification; seeingthe possibility of holding dual identities thatencompass the conflicting subgroups reducesthe potential risk of enhancing the security ofeach one; and beginning to promote compellingsuperordinate goals and/or integrative struc-tures gives subgroup members an overarchingvision with which to identify, in addition to iden-tifying with their own subgroup, thus facilitat-ing the development of dual identities. Our em-phasis on the necessary sequencing of thephases, thus, should not rule out that some de-gree of overlap between each consecutive phaseis likely.

DISCUSSION AND CONCLUSIONS

Theoretical Implications

The main contribution of the IIC resolutionmodel developed in this paper is the identifi-cation of an ordered series of process inter-ventions and necessary prior states to movefrom intractable conflict to intergroup harmony.Subject to empirical verification, we make threecontributions to the literature on intractableconflicts. First, we enrich this research by intro-ducing what is, to our knowledge, the first com-prehensive model of intractability managementthat puts identity center stage in the process. Asnoted earlier, much extant work on resolvingintractable conflicts has failed to systematicallyconsider the role that social identity can play inthe resolution of these conflicts. This oversightmay be one reason why past attempts to man-age these conflicts have met with mixed resultsand why calls continue for new insights intomanaging these conflicts (e.g., Burgess & Bur-gess, 2006; Crocker et al., 2004).

Second, we infuse much of the practice-basedresearch on intractable conflicts with insightsderived from various theoretical traditions. Wedraw most extensively from social identity the-ory, since this theory was designed to examine“impersonal” conflicts that arise between andamong groups (Tajfel & Turner, 1979)—condi-tions that are mirrored, but historically intensi-fied, in intractable conflicts. Specifically, we be-

lieve that research on intractable conflicts,which has only rarely drawn from social identitytheory (see Desivilya, 1998, and Seull, 1999, fornotable exceptions), can be greatly enriched byincorporating the states and intervening pro-cesses that lie between getting conflicting par-ties to the table (ripeness) and creating endur-ing intergroup harmony.

Third, our model highlights areas for furtherresearch on the management of intractable con-flicts. For example, while research has recog-nized the role of emotion (Coleman, 2003; Sen,2006; Putnam & Wondolleck, 2003), the link be-tween identity and emotions in such conflictshas gone largely unexamined. We have arguedthat emotions are central to the mutual disiden-tification process since parties fail to verify thelegitimacy of the other (see also Burke, 1991, andBurke & Harrod, 2005). However, as Sen (2006)has noted, identities are also a source of prideand therefore positive emotion. During intracta-ble conflicts, identities are simultaneously ver-ified by ingroup members and not verified byoutgroup members. This suggests that mixed orambivalent emotions may play a heretofore un-explored role in these conflicts (Wang & Pratt,2008). Given that ambivalent emotions createinstability and may amplify attitudes towardothers (Katz & Glass, 1979), special care must betaken throughout the management process toavoid magnifying minor dislikes among mem-bers into hatred.

Finally, and more broadly, our model has thepotential to contribute to research in areas be-yond intractable conflicts. It highlights themixed results of intervention efforts (e.g., pro-moting mindfulness does not always lead toidentity decoupling) and presents moderatingstates to explain them. As such, the model con-tributes to research on dual identities, identitysecurity, and intergroup identity decoupling.Our model also resonates with research on de-mographic conflict. Diversity researchers havelong recognized the limitations inherent in anassimilation paradigm where the superordinateidentity dominates and in a differentiation par-adigm where subgroup identities dominate(Thomas & Ely, 1996). These paradigms areequally implausible in intractable conflicts. Tothe degree that the parallels hold, diversity re-search may benefit from a model that involvesboth securing subgroup identities and strength-ening dual identities.


Practical Implications

Implementing our model presents a numberof leadership challenges. In the beginning ofthe process, leaders must be aware that re-sources and interests—not just identities—areproblematic. In fact, if researchers who studyconflict ripeness are correct, leaders must rec-ognize that resource- and interest-based issuesmay initially be perceived as more legitimatethan identity issues for bringing people to thetable and beginning the decoupling process(Coleman, 2000; Pruitt & Olczak, 1995; Zartman,2005). So even though intractable conflicts oftenbegin when a group’s identity is invalidated, itmay not be possible to begin the healing pro-cess around issues of identity. This is likely tofeel much too threatening. Resource- and inter-est-based issues may be a safer way to initiallymotivate people to come to the table.

In the first few phases, as people move fromdecoupling to securing their identities, leadersmust do two things in very close order: (1) pro-mote a more differentiated view of the outgroupand (2) promote a more unified view of the in-group. For example, even though physicians area diverse lot, administrative personnel who findthem problematic tend to view them stereotypi-cally as a single entity (“the” physicians), lead-ing to a bipolar dynamic, which must be over-come by promoting differentiated images of “theother.” In the next phase leaders must promotemore unified images of the ingroup. For exam-ple, most physicians do not think of themselvesas belonging to one unified physician group anddo not strongly identify with one another exceptin their common animosity toward administra-tion (O’Connor et al., 2006; see Table 1). By re-moving that negative stimulus, leaders are, infact, stripping individuals of one importantsource of their ingroup identity, making it essen-tial to move quickly to positive ingroup identitydevelopment. In fact, leaders may need to togglebetween the two phases, with the promise ofsubgroup security serving as a motivator to stripaway negative distinctiveness.

In the next few phases, as leaders focus ondeveloping dual identities and integrative goalsand structures, they face the substantial chal-lenge of motivating subgroup members to exertthemselves on behalf of both their ingroup andthe superordinate group that includes the out-group. To start, ingroup members must move

beyond understanding that outgroup membershold multiple identities to realize that they do aswell (i.e., social identity complexity). In addition,the simultaneous adoption of a subordinate andsuperordinate identity requires that the latternot be too closely aligned with either subgroupidentity in order to avoid resistance from theother group. Subgroup identity projection is adefensive mechanism against identity threat orloss and occurs when a subgroup perceives it-self to be more similar to, typical of, and norma-tive for the inclusive superordinate identity cat-egory than the other subgroup (Mummendey &Wenzel, 1999). This projection is then used as abasis for devaluing the outgroup, which is per-ceived to be less similar to the superordinateidentity. In some airlines, for example, this oc-curs when executives claim market power as asuperordinate identity, which may appear to bea projection of the administrative (versus thepilot or aviation) identity of the airline (see Ta-ble 1). To avoid such projection, leaders mustpay attention to the goals of the subgroups andthe superordinate group. In their ASPIRe model,Haslam and colleagues (2003) suggest doing thisby creating superordinate goals that build onthe needs and goals of the subgroups (i.e., “su-percasing”) and creating a feedback systemsuch that members of each subgroup can partic-ipate in the joint goal-setting process (i.e., “or-ganic goal setting“).

Finally, although the literature on conflictripeness discusses the conditions under whichparties will enter into the decoupling phase, weknow very little about when to proceed to eachsubsequent phase in our model. In other words,when are the disidentified subgroup identitiessufficiently decoupled to begin developing se-cure ingroup identities, the subgroup identitiessufficiently secure to begin dual identity devel-opment, and the dual identities sufficientlystrong to promote integrative goals and struc-tures? People’s visible behaviors and language(i.e., what they do and what they say) are indi-cators of shifting identity boundaries. For exam-ple, patterns of who spends time with whom andsymbolic marking of shared intergroup space(Brown, Lawrence, & Robinson, 2005) may reflectchanging group affiliations. Language, too,serves as a marker, indicating people’s repre-sentations of themselves and their ingroup(Cheney, 1983; Fiol, 1989). For example, Fiol de-scribed how changes in the language at Tech-


Co, a technology company undergoing an iden-tity transformation, revealed the changingidentifications of employees with the old “we”and with the new identity the company wasassuming, noting that “language can reflect thestrength of the relationship between individualsand their social groups” (2002: 663). Leaders cantherefore determine the progress in movingthrough each phase of the process we have de-scribed by attending to people’s behavioral andlinguistic cues.

Next Steps

IICs are highly complex. While we believethat managing identity must play a key role inresolving these conflicts, there are additionalissues, beyond the scope of this paper, that mustbe addressed as research in this area movesforward. For example, some of our argumentshave drawn explicitly from research on intrac-table conflicts—a body of research that, to date,has been concerned largely with conflicts be-tween societal groups (e.g., nations, religiousand ethnic groups). We acknowledge that theremay be important differences that can impactintergroup relations across diverse settings, aswell as the success of proposed approaches forconflict resolution. One obvious difference isgroup size, which may impact how informationis disseminated and processed within and be-tween the subgroups, as well as the types ofinteractions that are likely. Issues of size alsoraise issues of scale. That is, how many peoplefrom each group need to be involved in the con-flict management process in order for the mem-bers of the conflicting parties to feel that theirneeds have been adequately represented (Bur-gess & Burgess, 2006)? Groups in organizationsmay also differ from ethnonational groups intheir boundary permeability, with some types ofgroups having memberships that are more per-manent or more permeable than others (Mum-mendey, Klink, Mielke, Wenzel, & Blanz, 1999;Tajfel & Turner, 1979). For example, how muchmore threatening are conflicts based on one’smembership in an ethnic group (e.g., Palestin-ian) versus a professional group (e.g., physi-cian)? Are there differences of membership sa-lience and historical significance, which maymake identity conflicts more or less intense?Future research that tests our propositions

should pay special attention to the potential im-pact of these differences.

We have also very deliberately omitted refer-ence to the impact of relative subgroup statusand power in developing our model. Large inter-group status and power asymmetries wouldlikely lead to domination of one group over theother, effectively suppressing the conflict. Al-though less pronounced status or power differ-entials are potentially relevant in any of ourphases, the specific impact is complex and dif-ficult to predict. Dovidio, Gaertner, and Validzic(1998), for example, conducted a study that ma-nipulated the relative status of groups on thesame and different dimensions of their task ex-perience. Equal status between two groupsworking on the same task produced intergroupbias, but the bias was eliminated when thegroups’ areas of expertise were distinct and val-ued. This suggests that the effects of status arenot direct but involve other intervening vari-ables. Similarly, Rouhana and Fiske (1995) con-cluded that although Israelis were perceived byArab and Jewish students as holding more insti-tutional power (control of tangible elementssuch as the economy) than Arabs, this was ac-companied by a sense of threat rather than se-curity, which was apparently due to the Arabs’greater latent power to determine their ownidentity and ability to endure. This suggests thatdetermining power asymmetry is also not at allstraightforward and cannot easily be related toexperiences of intergroup threat. We leave it tofuture research to flesh out these importantpower and status dynamics.

Naturally, there is a need to test and refine themodel we have proposed. Multiple methods canbe used to examine specific hypotheses or theentire model. For example, we make a strongclaim that decoupling intergroup identities iscritical if attempts to promote positive ingroupdistinctiveness are to lead to subgroup identitysecurity (Proposition 3). A strong test of thisproposition would be to conduct field experi-ments where groups facing intractable conflictsin one organization simply received positivemessages about their respective groups, whilein another organization these positive ingroupmessages were coupled with the techniquesused by Kelman for intergroup identity decou-pling. Subgroup identity security scores in eachorganization could be compared. A weakertest—one that might miss some of the historical


dynamics of IICs—may involve conducting alaboratory experiment where subjects are re-cruited based on the results from a screeningquestionnaire that assesses levels of mutualdisidentification with a relevant outgroup (e.g.,active liberal Democrats and conservative Re-publicans in the United States). As in the quasi-field experiments, control groups receiving pos-itive messages about their ingroup can becompared on their subgroup identity scores(controlling for pretest subgroup identity secu-rity) with those receiving both positive mes-sages and decoupling interventions.

The entire model can be examined via eitherstatic or longitudinal analyses. The former pro-vides information regarding the impact ofpresent levels of each of the variables on pre-dicted model outcomes, whereas the latter pro-vides a test of the implications of alteringlevels of model variables in a manner consistentwith our propositions. Whether we test it stati-cally or longitudinally, we reemphasize that(1) our model proposes sequential phases whereall prior states must be achieved and main-tained for each moderating variable to have theproposed effect, and (2) this may not alwaysrequire that the proposed interventions beginwith establishing conflict ripeness. In many in-tractable identity settings, however, we suspectthat there is little readiness to rethink long-heldidentity beliefs, suggesting the need for inter-ventions to begin by establishing conflict ripe-ness. In such cases interventions guided by ourmodel would need to begin at this earliestphase (see Figure 1) and move through all of thephases presented in our propositions.

A static analysis of our model would entailsurvey measurement of each of the model vari-ables. One would ideally collect multiple indi-cators (e.g., different indicators of dual identitystrength), each measured along a continuum.Within-group agreement of scores from the indi-vidual measures (see Chan’s [1998] “dispersionmodel”) could then be used to assess thestrength of each state. For example, the greaterthe proportion of members who perceive theirsubgroup’s identity to be secure across multiplemeasures, the stronger the group-level constructof subgroup identity security, and, conse-quently, the stronger the moderating effects ofsubgroup identity security will be.

The first phase of static testing would involvea regression analysis of the proposed moderat-

ing effects of different levels of ripeness on therelationship between promoting mindfulnessand intergroup identity decoupling within thedata set. Our model proposes that similar effortsto promote mindfulness will lead to differentdegrees of identity decoupling, owing to differ-ences in conflict ripeness (see Figure 1). Vari-ance in degree of decoupling allows one to testthe moderating effect of intergroup identity de-coupling on the relationship between promotingpositive ingroup distinctiveness and subgroupidentity security. Similarly, differences in thelevel of identity security in the sample allow atest of the moderating effect of identity securityon the relationship between promoting inter-group differentiation/unity and dual identitystrength. Finally, one would test the moderatingeffect of dual identity strength on the relation-ship between promoting integrative goals/structures and intergroup harmony.

The truest test of our model would be a longitu-dinal approach that followed how conflictinggroups evolve over time through the sequentialphases of the model. One way to assess this as-pect of our model would be to conduct multi-method, multiple-site, longitudinal case studies.The test would start with a single case, ideally inan organization where parties in identity conflictare in a state of readiness but not mindfulness (theconditions for readiness having already beenshown in extant research). Researchers wouldthen begin by promoting mindfulness and assess-ing its influence on all of the other states in ourmodel (e.g., decoupled intergroup identities, sub-group identity security, dual identity strength,etc.). Such assessments should involve the use ofmultiple data sources, such as interviews, sur-veys, and documents. At “Time 2” researcherswould continue the first intervention in order tomaintain the state moderating the subsequentstage of the model (e.g., decoupled intergroupidentities), in addition to the next proposed inter-vention (e.g., promoting positive ingroup distinc-tiveness), in order to determine their joint effects.Once again, researchers would examine the ef-fects of these interventions on all of the states inour model. Such assessments would continue tofollow subsequent sets of interventions outlined inour model, ending with an examination of all in-terventions on the impact of promoting integrativegoals/structures (see Figure 1). Throughout thisprocess, researchers would incorporate pattern-matching techniques (Yin, 2003) whereby they


would compare the empirically based patternwith a predicted one—that is, the results of theinterventions with our predicted results.

After conducting this initial case analysis, re-searchers should pursue two types of cases.First, some cases should involve organizationsfacing conditions similar to those in the initialcase study (e.g., in a state of intractable conflict)in order to see whether results can be replicated.Second, some cases should be contrasting (e.g.,in a similar type of organization where there isno intractable conflict) to see if the absence ofintractable conflict can also be predicted by ourmodel (e.g., dual identities are present, identi-ties are secure, etc.). These subsequent casescould be used to confirm or refine our model(Lindesmith, 1947; Yan & Gray, 1994). While theuse of case analyses to test theory is not com-mon in organizational research, it is not withoutprecedent. For example, Lee, Mitchell, Wise, andFireman (1996) used a multiple case study de-sign to conduct the first test of their unfoldingmodel of voluntary turnover.

The measures and intervention techniquesneeded to test our model already exist. For ex-ample, to assess our states (the shaded boxes inFigure 1), several existing measures can beused. Indices of intergroup harmony have beendeveloped (e.g., Banker & Gaertner, 1998).Strength of dual identities can also be ascer-tained using existing identification measures(e.g., Bergami & Bagozzi, 2000; Mael & Ashforth,1992), and Huo and Molina’s (2006) measure ofsubgroup respect can be adapted to approxi-mate subgroup identity security. Kreiner andAshforth’s (2004) disidentification scale can beadapted to examine mutual disidentification—the opposite of decoupled intergroup identi-ties—with the target of disidentification beingthe opposing group (e.g., “I would be embar-rassed to be part of �add competing subgroup’sname�”). Finally, measures of conflict ripenesshave already been developed (Coleman, 2000).

The IIC resolution model’s interventions (theovals in Figure 1) can also be adapted fromexisting practices. For example, techniques forpromoting mindfulness have been developed(Weick & Sutcliffe, 2001). Interventions to buildpositive subgroup distinctiveness as well as topromote simultaneous intergroup differentia-tion and unity can draw on Cheney’s (1983) rhe-torical techniques (e.g., using “we” when talkingabout group members at the relevant level).

And, finally, numerous techniques have beenproposed for promoting integrative goals andstructures (e.g., creating superordinate goals).

To conclude, the repercussions of IIC have longbeen known on the international stage. Accordingto Coleman (2003), nearly one-sixth of the world’spopulation is engaged in or on the verge of engag-ing in identity-based conflicts. And in a recentreview of the literature, Burgess and Burgess re-ferred to intractable conflict as, “arguably, themost destructive force on the planet” (2006: 183).However, researchers in organizational studiesare only recently converging on the role of identityin protracted organizational conflicts and are onlybeginning to assess the damage that can ensuefrom them. As professional organizations continueto take on administrative layers, as nonprofes-sional organizations increasingly depend on pro-fessionals, and as organizations globalize and be-come more diverse, we believe that a focus onmanaging IICs is long overdue.

During off-site retreat sessions, tensions be-tween physicians and administrators at thecommunity hospital were temporarily reducedthrough a series of exercises that created readi-ness to at least consider possibilities beyond all-out warfare. Attempting to capitalize on thisinitial ripeness, the retreat leaders initiatedthree short-term projects jointly sponsored by atleast one physician and one administrator thatappeared to have recognizable, measurablebenefits for both groups. The involved physi-cians and administrators agreed that they wouldreturn to this larger retreat group within a fewweeks to report their outcomes (hopefully dem-onstrating that they could work together effec-tively), thereby providing the foundation for thefollow-up work that was recommended by theconsultants. However, the follow-up work to de-couple the intergroup identities, enhance sub-group security, and develop identity strengthwithin both physician and administrator groupsnever occurred.

While recognition of their mutual interdepen-dence gained during the retreat and follow-upprojects led to somewhat reduced conflict in theshort run, agreement regarding larger initiativeswas never reached. Tensions continued to buildto the point where the CEO contacted one of theconsultants, requesting that he return and againdisarm the conflict that was once more threaten-ing her job. She explained that she had seen no


need to do the intermediate steps he had recom-mended and had, instead, immediately initiatedconversations about creating a common agree-ment among all involved as the basis for reduc-ing the intergroup conflict (according to ourmodel, moving from readiness directly to pro-moting integrative goals). The consultant de-clined to work further with this organization(O’Connor, personal communication).

REFERENCES

Argyris, C. 1996. Actionable knowledge: Design causality inthe service of consequential theory. Journal of AppliedBehavioral Science, 32: 390–406.

Ashforth, B. E., & Mael, F. 1989. Social identity theory and theorganization. Academy of Management Review, 14:20–39.

Ashmore, R. D., Jussim, L., Wilder, D., & Heppen, J. 2001.Conclusion: Toward a social identity framework for in-tergroup conflict. In R. D. Ashmore, L. Jussim, & D. Wilder(Eds.), Social identity, intergroup conflict, and conflictreduction: 213–249. New York: Oxford University Press.

Azar, E. 1986. Protracted international conflicts: Ten proposi-tions. In E. E. Azar & J. W. Burton (Eds.), Internationalconflict resolution: Theory and practice: 28–39. Sussex,UK: Wheatsheaf.

Banker, B. S., & Gaertner, S. L. 1998. Achieving stepfamilyharmony: An intergroup-relations approach. Journal ofFamily Psychology, 12: 310–325.

Bar-Tal, D., Kruglanski, A. W., & Klar, Y. 1989. Conflict termi-nation: An epistemological analysis of internationalcases. Political Psychology, 10: 233–255.

Bergami, M., & Bagozzi, R. P. 2000. Self-categorization, affec-tive commitment and group self-esteem as distinct as-pects of social identity in the organization. British Jour-nal of Social Psychology, 39: 555–577.

Berry, J. 1991. Understanding and managing multicultural-ism: Some possible implications of research in Canada.Psychology and Developing Societies, 3: 17–49.

Bizman, A., & Yinon, Y. 2004. Intergroup conflict managementstrategies as related to perceptions of dual identity andseparate groups. Journal of Social Psychology, 144: 115–126.

Bradsher, K. 2000. Airline deal: The unions. New York Times,http: / /query.nytimes.com/gst/fullpage.html?res�9901EED6143DF936A15756C0A9669C8B63.

Breakwell, G. M. 1986. Coping with threatened identities.New York: Methuen.

Brewer, M. B. 2001. Ingroup identification and intergroupconflict: When does ingroup love become outgrouphate? In R. D. Ashmore, L. Jussim, & D. Wilder (Eds.),Social identity, intergroup conflict, and conflict resolu-tion: 17–41. New York: Oxford University Press.

Brewer, M. B., & Brown, R. J. 1998. Intergroup relations. In D. T.Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook of

social psychology (4th ed.): 554–594. New York: McGraw-Hill.

Brown, G., Lawrence, T., & Robinson, S. 2005. Territoriality inorganizations. Academy of Management Review, 30:577–594.

Brown, R. J. 1978. Divided we fall: An analysis of relationsbetween sections of a factory work-force. In H. Tajfel(Ed.), Differentiation between social groups: Studies inthe social psychology of intergroup relations: 395–429.London: Academic Press.

Bryant, A. 1994. Unions edge closer to gaining control of UnitedAirlines. New York Times, http://query.nytimes.com/gst/fullpage.html?res�9405EEDB1E3CF935A15750C0A962958260.

Bunker, B., Alban, B., & Lewicki, R. 2004. Ideas in currency inOD practice: Has the well gone dry? Journal of AppliedBehavioral Science, 40: 403–422.

Burgess, H., & Burgess, G. 2006. Intractability and the frontierof the field. Conflict Resolution Quarterty, 24: 177–186.

Burke, P. 1991. Identity processes and social stress. AmericanSociological Review, 56: 836–849.

Burke, P., & Harrod, M. 2005. Too much of a good thing? SocialPsychology Quarterly, 68: 359–374.

Bush, R., & Folger, J. P. 1994. The promise of mediation:Responding to conflict through empowerment and rec-ognition. San Francisco: Jossey-Bass.

Chan, D. 1998. Functional relations among constructs in thesame content domain at different levels of analysis.Journal of Applied Psychology, 83: 234–246.

Cheney, G. 1983. The rhetoric of identification and the studyof organizational communication. Quarterly Journal ofSpeech, 69: 143–158.

Coleman, P. 2000. Fostering ripeness in seemingly intracta-ble conflict: An experimental study. International Jour-nal of Conflict Management, 11: 300–317.

Coleman, P. 2003. Characteristics of protracted, intractableconflict: Toward the development of a meta-frame-work-I. Peace and Conflict: Journal of Peace Psychology,9: 1–37.

Cook, S. W. 1984. Cooperative interaction in multiethnic con-texts. In N. Miller & M. B. Brewer (Eds.), Groups in con-tact: The psychology of desegregation: 155–185. Orlando,FL: Academic Press.

Cooley, C. 1922. Human nature and the social order (reviseded.). New York: Scribner.

Crisp, R., Stone, C., & Hall, N. 2006. Recategorization andsubgroup identification: Predicting and preventingthreats from common ingroups. Personality and SocialPsychology Bulletin, 32: 230–243.

Crocker, C., Hampson, F., & Aall, P. (Eds.). 2004. Grasping thenettle. Washington, DC: United States Institute of PeacePress.

Crocker, C., Hampson, F., & Aall, P. (Eds.). 2005. Tamingintractable conflicts. Washington, DC: United States In-stitute of Peace Press.

Dalton, M. 2003. Social identity conflict. Sloan ManagementReview, 45(1): 7–8.


Desivilya, H. S. 1998. Jewish-Arab coexistence in Israel: Therole of joint professional teams. Journal of Peace Re-search, 35: 429–452.

Diehl, P., & Goertz, G. 1993. Enduring rivalries: Theoreticalconstructs and empirical patterns. International StudiesQuarterly, 37: 147–171.

Dovidio, J., Gaertner, S., & Validzic, A. 1998. Intergroup bias:Status differentiation and a common in-group identity.Journal of Personality and Social Psychology, 75: 109–120.

Dukerich, J., Kramer, R., & Parks, J. M. 1998. The dark side oforganizational identification. In D. Whetten & P. Godfrey(Eds.), Identity in organizations: Developing theorythrough conversations: 245–256. Thousand Oaks, CA:Sage.

Dutton, J. E., Dukerich, J. M., & Harquail, C. V. 1994. Organi-zational images and member identification. Administra-tive Science Quarterly, 39: 239–263.

Elsbach, K. 1999. An expanded model of organizational iden-tification. Research in Organizational Behavior, 21: 163–199.

Evans, M. 2007. CEOs focus on finance, docs. Modern Health-care, 37(1): 8–9.

Fiol, C. M. 1989. A semiotic analysis of corporate language:Organizational boundaries and joint venturing. Admin-istrative Science Quarterly, 34: 277–303.

Fiol, C. M. 2002. Capitalizing on paradox: The role of lan-guage in transforming organizational identities. Orga-nization Science, 13: 653–666.

Fiol, C. M., & O’Connor, E. J. 2002. When hot and cold collidein radical change processes: Lessons from communitydevelopment. Organization Science, 13: 532–546.

Fiol, C. M., & O’Connor, E. J. 2003. Waking up! Mindfulness inthe face of bandwagons. Academy of Management Re-view, 28: 54–70.

Fisher, R. 1983. Third party consultation as a method ofintergroup conflict resolution. Journal of Conflict Reso-lution, 27: 301–334.

Friedman, R., & Davidson, M. 1999. The black-white gap inperceptions of discrimination: Its causes and conse-quences. Research in Negotiation in Organizations, 7:203–228.

Gaertner, S. L., Bachman, B. A., Dovidio, J. F., & Banker, B. S.2001. Corporate mergers and stepfamily marriages:Identity, harmony, and commitment. In M. A. Hogg & D. J.Terry (Eds.), Social identity processes in organizationalcontexts: 265–282. Ann Arbor: Sheridan Books.

Gaertner, S. L., & Dovidio, J. F. 2000. Reducing intergroupbias: The common ingroup identity model. Ann Arbor:Sheridan Books.

Gaertner, S. L., Dovidio, J. F., Nier, J. A., Ward, C. M., &Banker, B. S. 1999. Across cultural divides: The value ofa superordinate identity. In D. A. Prentice & D. T. Miller(Eds.), Cultural divides: Understanding and overcominggroup conflict: 173–212. New York: Russell Sage Founda-tion.

Glynn, M. 2000. When cymbals become symbols: Conflictover organizational identity within a symphony orches-tra. Organization Science, 11: 285–298.

Gray, B. 2003. Framing of environmental disputes. InR. Lewicki, B. Gray, & M. Elliott (Eds.), Making sense ofintractable environmental conflicts: 11–34. London:Island Press.

Haslam, S. Eggins, R., & Reynolds, K. 2003. The ASPIRe model:Actualizing social and personal identity resources to en-hance organizational outcomes. Journal of Occupationaland Organizational Psychology, 76: 83–113.

Haydu, J. 1989. Managing “the labor problem” in the UnitedStates ca. 1897–1911. In L. Kriesberg, T. Northrup, &S. Thorson (Eds.), Intractable conflicts and their transfor-mation: 93–106. Syracuse, NY: Syracuse University Press.

Hewstone, M., & Brown, R. 1986. Contact is not enough: Anintergroup perspective. In M. Hewstone & R. Brown(Eds.), Contact and conflict in intergroup encounters:1–44. Oxford: Blackwell.

Hinkle, S., & Brown, R. 1990. Intergroup comparisons andsocial identity: Some links and lacunae. In D. Abrams &M. Hogg (Eds.), Social identity theory: Construction andcritical advances: 48–70. London: Harvester Wheatsheaf.

Holm, C. E. 2000. Next generation physician-health systempartnerships. Chicago: Health Administration Press.

Hornsey, M., & Hogg, M. 2000. Subgroup relations: A compar-ison of mutual intergroup differentiation and commoningroup identity models of prejudice reduction. Person-ality and Social Psychology Bulletin, 26: 242–256.

Humphreys, M., & Brown, A. 2002. Dress and identity: A Turk-ish case study. Journal of Management Studies, 39: 927–952.

Huo, Y. J., & Molina, L. E. 2006. Is pluralism a viable model ofdiversity? The benefits and limits of subgroup respect.Group Processes & Intergroup Relations, 9: 359–376.

Kahn, A., & Ryen, A. H. 1972. Factors influencing the biastowards one’s own group. International Journal of GroupTensions, 2(2): 33–50.

Katz, I., & Glass, D. 1979. An ambivalence-amplification the-ory of behavior toward the stigmatized. In W. G. Austin& S. Worchel (Eds.), The social psychology of intergrouprelations: 55–70. Monterey, CA: Brooks/Cole.

Kelman, H. C. 1998. Interactive problem solving: An ap-proach to conflict resolution and its application in theMiddle East. Political Science and Politics, 31: 190–198.

Kelman, H. C. 1999. The interdependence of Israeli and Pal-estinian national identities: The role of the other inexistential conflicts. Journal of Social Issues, 55: 581–600.

Kelman, H. C. 2001. The role of national identity in conflictresolution. In R. D. Ashmore, L. Jussim, & D. Wilder (Eds.),Social identity, intergroup conflict, and conflict reduc-tion: 187–212. New York: Oxford University Press.

Kelman, H. C. 2004. Reconciliation as identity change: Asocial-psychological perspective. In Y. Bar-Siman-Tov(Ed.), From conflict resolution to reconciliation: 11–124.Oxford: Oxford University Press.


Kelman, H. C. 2006. Interests, relationships, identities: Threecentral issues for individuals and groups in negotiatingtheir social environment. Annual Review of Psychology,57: 1–26.

Kelman, H. C., & Cohen, S. 1986. Resolution of internationalconflict: An interactional approach. In S. Worchel &W. Austin (Eds.), Psychology of intergroup relations (2nded.): 323–342. Chicago: Nelson-Hall.

Kilduff, M. 2006. Publishing theory. Academy of ManagementReview, 31: 252–255.

Kreiner, G., & Ashforth, B. 2004. Evidence toward an ex-panded model of organizational identification. Journalof Organizational Behavior, 25: 1–27.

Kriesberg, L. 1986. Nature, dynamics, and phases of intrac-tability. In C. Crocker, F. Hampson, & P. Aall (Eds.),Grasping the nettle: Analyzing cases of intractable con-flict: 65–97. Washington, DC: United States Institute ofPeace Press.

Kriesberg, L. 2003. Constructive conflicts: From escalation toresolution. New York: Rowman and Littlefield.

Langer, E. J. 1989. Minding matters: The consequences ofmindlessness-mindfulness. Advances in ExperimentalSocial Psychology, 22: 137–173.

Lee, T., Mitchell, T., Wise, L., & Fireman, S. 1996. An unfoldingmodel of voluntary employee turnover. Academy ofManagement Journal, 39: 5–36.

Lewicki, R. J., Gray, B., & Elliott, M. 2003. Making sense ofintractable environmental conflicts. Washington, DC: Is-land Press.

Lindesmith, A. R. 1947. Opiate addiction. Bloomington, IN:Principia Press.

Linskold, S. 1986. GRIT: Reducing distrust through carefullyintroduced conciliation. In S. Worchel & W. Austin (Eds.),Psychology of intergroup relations (2nd ed.): 305–322.Chicago: Nelson-Hall.

Mael, F., & Ashforth, B. 1992. Alumni and their alma mater: Apartial test of the reformulated model of organizationalidentification. Journal of Organizational Behavior, 13:103–123.

Maiese, M. 2006. Engaging emotions in conflict intervention.Conflict Resolution Quarterly, 24: 187–195.

Maynard, M., & Walsh, M. W. 2004. Union asks judge toappoint trustee to run United Airlines. New York Times,http: / /query.nytimes.com/gst/fullpage.html?res�9900E0DD113EF931A2575BC0A9629C8B63.

McCall, G. J., & Simmons, J. L. 1978. Identities and interac-tions: An examination of associations in everyday life(revised ed.). New York: Free Press.

Mummendey, A., Klink, A., Mielke, R., Wenzel, M., & Blanz, M.1999. Social structural characteristics of intergroup rela-tions and identity management strategies: Results froma field study in East Germany. European Journal of So-cial Psychology, 29: 259–285.

Mummendey, A., & Wenzel, M. 1999. Social discriminationand tolerance in intergroup relations: Reactions to inter-group difference. Personality and Social Psychology Re-view, 3: 158–174.

Northrup, T. 1989. The dynamic of identity in personal andsocial conflict. In L. Kriesberg, T. Northrup, & S. Thorson(Eds.), Intractable conflicts and their transformation: 55–82. Syracuse, NY: Syracuse University Press.

O’Connor, E. J., & Annison, M. H. 2002. Building trust andcollaboration between physicians and administrators.The Physician Executive, 28: 48–52.

O’Connor, E. J., & Bujak, J. S. 2001. Trust, the strategic imper-ative: How to successfully build physician-system part-nerships. Leadership and Management Report, 9(5): 6–9.

O’Connor, E. J., Fiol, C. M., & Guthrie, M. 2006. Separatelytogether: Build unity by strengthening physician groups.The Physician Executive, 32: 16–21.

Ohlott, P., Chrobot-Mason, D., & Dalton, M. 2004. Collisioncourses: When social identity leads to conflict. Leader-ship in Action, 24(3): 8–12.

Pelham, B. W., & Hetts, J. J. 1999. Implicit and explicit per-sonal and social identity: Toward a more complete un-derstanding of the social self. In T. R. Tyler, R. M.Kramer, & O. P. John (Eds.), The psychology of the socialself: 115–143. London: Lawrence Erlbaum Associates.

Portilla, J. 2006. What exists is possible: Stories from conflictresolution professionals. Conflict Resolution Quarterly,24: 241–248.

Pratt, M. G. 2000. The good, the bad, and the ambivalent:Managing identification among Amway distributors.Administrative Science Quarterly, 45: 456–493.

Pratt, M. G., & Corley, K. 2007. Managing multiple organiza-tional identities: On identity ambiguity, identity conflict,and members’ reactions. In C. Bartel, S. Blader, &A. Wrzesniewski (Eds.), Identity and the modern organi-zation: 99–118. Mahwah, NJ: Lawrence Erlbaum Associ-ates.

Pratt, M. G., & Doucet, L. 2000. Ambivalent feelings in orga-nizational relationships. In S. Fineman (Ed.), Emotions inorganizations, vol. 2: 204–226. London: Sage.

Pratt, M. G., & Foreman, P. O. 2000. Classifying managerialresponses to multiple organizational identities. Acad-emy Management Review, 25: 18–42.

Pratt, M. G., & Rafaeli, A. 1997. Organizational dress as asymbol of multi-layered social identities. Academy ofManagement Journal, 40: 862–898.

Pruitt, D., & Olczak, P. 1995. Beyond hope: Approaches tosolving seemingly intractable conflicts. In B. Bunker &J. Rubin (Eds.), Conflict, cooperation, and justice: 59–92.San Francisco: Jossey-Bass.

Putnam, L. L., & Wondolleck, J. M. 2003. Intractability: Defi-nitions, dimensions, and distinctions. In R. J. Lewicki,B. Gray, & M. Elliott (Eds.), Making sense of intractableenvironmental conflicts: 35–62. Washington, DC: IslandPress.

Retzinger, S., & Scheff, S. 2000. Emotion, alienation, andnarratives: Resolving intractable conflict. MediationQuarterly, 18: 71–85.

Roccas, S., & Brewer, M. B. 2002. Social identity complexity.Personality and Social Psychology Review, 6: 88–106.


Rothman, J. 1997. Resolving identity-based conflict in na-tions, organizations, and communities. San Francisco:Jossey-Bass.

Rouhana, N. N., & Bar-Tal, D. 1998. Psychological dynamics ofintractable ethnonational conflicts: The Israeli-Palestin-ian case. American Psychologist, 53: 761–770.

Rouhana, N. N., & Fiske, S. T. 1995. Perceptions of power,threat, and conflict intensity in asymmetric intergroupconflict. Journal of Conflict Resolution, 39: 49–81.

Sen, A. 2006. Identity and violence: The illusion of destiny.New York: Norton.

Seull, J. 1999. “Ours is the way of God”: Religion, identity, andintergroup conflict. Journal of Peace Research, 36: 553–569.

Sherif, M., Harvey, O. J., White, B. J., Hood, W. R., & Sherif,C. W. 1961. Intergroup conflict and cooperation: The rob-ber’s cave experiment. Norman: University of OklahomaBook Exchange.

Shmueli, D., Elliott, M., & Kaufman, S. 2006. Frame changesand the management of intractable conflicts. ConflictResolution Quarterly, 24: 207–218.

Sidanius, J., Feshbach, S., Levin, S., & Pratto, F. 1997. Theinterface between ethnic and national attachment: Eth-nic pluralism or ethnic dominance? Public OpinionQuarterly, 61: 103–133.

Skevington, S. 1980. Intergroup relations and social changewithin a nursing context. British Journal of Social andClinical Psychology, 19: 201–213.

Spangler, B. 1993. Transformative mediation. http://www.beyondintractability.org/essay/transformative_mediation/?nid�1293.

Stephan, W. G., & Stephan, C. W. 1984. The role of ignorancein intergroup relations. In N. Miller & M. B. Brewer (Eds.),Groups in contact: The psychology of desegregation:229–255. Orlando, FL: Academic Press.

Stock, K. 2001. Peace . . . so close, yet so far away. www.cfcj-fcjc.org/full-text/2001_dra/kim_stock.html.

Sumner, W. G. 1906. Folkways. New York: Ginn.

Tajfel, H. 1978. Differentiation between social groups: Studiesin the social psychology of intergroup relations. NewYork: Academic Press.

Tajfel, H., & Turner, J. C. 1979. An integrative theory of inter-group conflict. In W. G. Austin & S. Worchel (Eds.), Thesocial psychology of intergroup relations: 33– 47.Monterey, CA: Brooks/Cole.

Thomas, D., & Ely, R. 1996. Making differences matter: A newparadigm for managing diversity. Harvard Business Re-view, 74(5): 1–15.

Tomlinson, E., & Lewicki, R. 2006. Managing distrust in intrac-table conflicts. Conflict Resolution Quarterly, 24: 219–228.

Turner, L. 2005. From transformation to revitalization: A newresearch agenda for a contested global economy. Workand Occupations, 32: 383–399.

Van Knippenberg, D., & Van Leeuwen, E. 2001. Organiza-tional identity after a merger: Sense of continuity as thekey to postmerger identification. In M. A. Hogg & D. J.Terry (Eds.), Social identity processes in organizationalcontexts: 249–264. Philadelphia: Psychology Press.

Wallace, J. E. 1995. Organizational and professional commit-ment in professional and nonprofessional organiza-tions. Administrative Science Quarterly, 40: 228–255.

Wang, L., & Pratt, M. G. 2008. An identity-based view of ambiv-alence and its management in organizations. In N. Ashka-nasy & C. L. Cooper (Eds.), Research companions to emo-tion in organizations: 589–604. Boston: Edward Elgar.

Wedge, B. 1987. Psychology of the self in social conflict. InE. Azar & J. Burton (Eds.), International conflict resolution:Theory and practice: 56–62. Sussex, UK: Wheatsheaf.

Weick, K. E., & Sutcliffe, K. M. 2001. Managing the unex-pected. San Francisco: Jossey-Bass.

Weick, K. E., Sutcliffe, K. M., & Obstfeld, D. 1999. Organizingfor high reliability: Processes of collective mindfulness.Research in Organizational Behavior, 21: 81–123.

Williams, K., & O’Reilly, C. 1998. The complexity of diversity:A review of forty years of research. Research in Organi-zational Behavior, 21: 77–140.

Wong, E. 2002. Bankruptcy hint at United Airlines. New YorkTimes, http://query.nytimes.com/gst/fullpage.html?res�9E07E2D7163DF936A2575BC0A9649C8B63.

Worchel, S. 1986. The role of cooperation in reducing inter-group conflict. In S. Worchel & W. Austin (Eds.), Psychol-ogy of intergroup relations (2nd ed.): 288–304. Chicago:Nelson-Hall.

Worchel, S., Andreoli, V. A., & Folger, R. 1977. Intergroupcooperation and intergroup attraction: The effect of pre-vious interaction and outcome of combined effort. Jour-nal of Experimental Social Psychology, 13: 131–140.

Yan, A., & Gray, B. 1994. Bargaining power, managementcontrol, and performance in United States–China jointventures: A comparative case study. Academy of Man-agement Journal, 37: 1478–1517.

Yin, R. 2003. Case study research (3rd ed.). Thousand Oaks,CA: Sage.

Zartman, I. W. 2005. Analyzing intractability. In C. Crocker,F. Hampson, & P. Aall (Eds.), Taming intractable con-flicts: 47–64. Washington, DC: United States Institute ofPeace Press.

Zartman, I. W., & Rubin, J. Z. (Eds.). 2000. Power and negoti-ation. Ann Arbor: University of Michigan Press.

C. Marlene Fiol ([email protected]) is a professor of strategic managementat the University of Colorado Denver. She received her Ph.D. in strategic managementfrom the University of Illinois at Urbana-Champaign. Her research interests includemanagerial and organizational cognition, organizational learning, and organization-al identity.


Michael G. Pratt ([email protected]) is a Winston Fellow and professor oforganizational studies at Boston College. He earned his Ph.D. from the Universityof Michigan. His current research interests include identity plurality, individual-organizational relationships, meaning of work, and intuition.

Edward J. O’Connor ([email protected]) is a professor of managementand health administration at the University of Colorado Denver. He received his Ph.D.in industrial/organizational psychology from the University of Akron. His researchinterests include physician-administration collaboration, leadership, and organiza-tional cognition.


MANAGING INTRACTABLE IDENTITY CONFLICTS · managing these conflicts. There is recognition that...

Documents

Transcript of MANAGING INTRACTABLE IDENTITY CONFLICTS · managing these conflicts. There is recognition that...