Download - THE SOCIAL PROCESSES OF ORGANIZATIONAL SENSEMAKING...and thus make it meaningful (Antaki, 1994; Potter & Wetherell, 1987). They can take a wide variety of forms, such as descriptions

THE SOCIAL PROCESSES OF ORGANIZATIONALSENSEMAKING

SALLY MAITLISUniversity of British Columbia

A longitudinal study of the social processes of organizational sensemaking suggeststhat they unfold in four distinct forms: guided, fragmented, restricted, and minimal.These forms result from the degree to which leaders and stakeholders engage in“sensegiving”—attempts to influence others’ understandings of an issue. Each of thefour forms of organizational sensemaking is associated with a distinct set of processcharacteristics that capture the dominant pattern of interaction. They also each resultin particular outcomes, specifically, the nature of the accounts and actions generated.

”Sensemaking” is a critical organizational activ-ity (Weick, 1995). For top managers, sensemakingactivities such as environmental scanning and is-sue interpretation are key tasks that significantlyinfluence organizational decisions and strategicchange (Gioia & Thomas, 1996; Smircich & Stub-bart, 1985; Thomas, Clark, & Gioia, 1993). For otherstakeholders, sensemaking powerfully affects howthey “construct” their identities (Pratt, 2000), pre-serve their organization’s image (Dutton & Duk-erich, 1991), and respond to organizational crises(Gephart, 1993). Sensemaking activities are partic-ularly critical in dynamic and turbulent contexts,where the need to create and maintain coherentunderstandings that sustain relationships and en-able collective action is especially important andchallenging (Weick, 1993).

Sensemaking occurs in organizations whenmembers confront events, issues, and actions thatare somehow surprising or confusing (Gioia &Thomas, 1996; Weick, 1993, 1995). As Weick ar-gued, “The basic idea of sensemaking is that realityis an ongoing accomplishment that emerges fromefforts to create order and make retrospective senseof what occurs” (1993: 635). Thus, sensemaking is aprocess of social construction (Berger & Luckmann,1967) in which individuals attempt to interpret andexplain sets of cues from their environments. Thishappens through the production of “accounts”—discursive constructions of reality that interpret orexplain (Antaki, 1994)—or through the “activa-tion” of existing accounts (Gioia & Thomas, 1996;Volkema, Farquhar, & Bergmann, 1996). In either

case, sensemaking allows people to deal with un-certainty and ambiguity by creating rational ac-counts of the world that enable action. Sensemak-ing thus both precedes decision making andfollows it: sensemaking provides the “clear ques-tions and clear answers” (Weick, 1993: 636) thatfeed decision making, and decision making oftenstimulates the surprises and confusion that createoccasions for sensemaking.

Organizational sensemaking is a fundamentallysocial process: organization members interprettheir environment in and through interactions withothers, constructing accounts that allow them tocomprehend the world and act collectively (Isa-bella, 1990; Sackman, 1991; Sandelands & Stablein,1987; Starbuck & Milliken, 1988; Weick & Roberts,1993). Recent sensemaking research, however, haseither tended to focus on its cognitive aspects (e.g.,Griffith, 1999; Thomas et al., 1993) or has examinedsocial processes that play out in extreme conditionsor crisis situations (e.g., Brown, 2000; Gephart,1993; Weick, 1993; Weick & Roberts, 1993). Lessattention has been paid to the sensemaking thatoccurs among large groups of diverse organiza-tional stakeholders as they address a range of issues(Weick, 1995). Such stakeholders engage in sense-making from a variety of organizational positions,histories, and personal backgrounds that create di-vergent frames of reference and lead them to takeon different roles in sensemaking processes (Dutton& Dukerich, 1991; Gephart, 1993; Weick, 1995).Despite the challenges raised by the inherently so-cial nature of sensemaking, the social processesthat underpin it at the organizational level remainrelatively underexamined (Eden, 1992). The pur-pose of this article, therefore, is to investigate thesocial processes of sensemaking among largegroups of diverse organizational stakeholders. Todo so, I draw on a longitudinal qualitative study of

I would like to thank Jean Bartunek, Chris Clegg, JaneDutton, Peter Frost, Richard Hackman, Tom Lawrence,Nick Turner, Editor Tom Lee, and the four anonymousreviewers for their support and helpful comments ondrafts of this paper.

� Academy of Management Journal2005, Vol. 48, No. 1, 21–49.

21

sensemaking in three organizations, across 27 issuedomains.

PREVIOUS RESEARCH

Research that has addressed the social processesof organizational sensemaking has tended to followone of two approaches. The first of these investi-gates how certain groups influence others’ under-standings of issues. Leaders have received particu-lar attention (Dutton & Jackson, 1987; Smircich &Stubbart, 1985). Gioia and Chittipeddi, for exam-ple, argued that a critical leader behavior duringstrategic change is “sensegiving”—which they de-fined as “the process of attempting to influence thesensemaking and meaning construction of otherstoward a preferred redefinition of organizationalreality” (1991: 442). Building on this work, otherstudies have examined the varieties of leader sense-giving. Gioia and Thomas (1996), for instance, sug-gested that leader sensegiving strategies may varydepending upon the strategic or political natureof an issue; Bartunek, Krim, Necochea, andHumphries (1999) identified a range of differentleader sensegiving strategies and found, for exam-ple, that when leaders felt personally threatened,they were more likely to focus their sensegivingefforts upon opportunities that strategic changepresented for stakeholders. All of this researchhighlights the importance of sensegiving as a fun-damental leadership activity within organizationalsensemaking.

Other studies have demonstrated how middlemanagers shape organizational accounts by sense-giving to their leaders. Dutton and colleagues, forexample, have examined how middle managersgain top management attention and influence orga-nizational action through “issue selling” (Dutton &Ashford, 1993; Dutton, Ashford, Wierba, O’Neill, &Hayes, 1997). Similarly, Westley (1990) showedhow middle managers could shape strategy throughtheir participation in “strategic conversations”with their bosses, influencing the way in which anissue was understood and “enacted.” These studiesdemonstrate that while leaders are uniquely placedto influence how issues are interpreted and under-stood in organizations, their interpretations can besignificantly shaped by the sensegiving efforts ofothers.

Although this research offers a valuable founda-tion for appreciating the social dynamics of organi-zational sensemaking, it largely ignores the interac-tion of different actors’ sensemaking behaviors andhow this interaction affects sensemaking processes.While previous studies provide insight into someof the strategies that leaders and stakeholders each

use to shape organizational understandings and ac-counts of issues, they have primarily focused onthe role played by just one party or the other. Thus,relatively little is known about the dynamics ofsensemaking when different parties engage simul-taneously or reciprocally in such activities, orabout the ways in which the accounts they generateare reconciled––or are not reconciled.

A second approach to organizational sensemak-ing has examined the social processes associatedwith sensemaking more holistically, but it hastended to do so in contexts that are marked by crisisor extreme circumstances. Weick’s (1993) analysisof the Mann Gulch disaster, for instance, suggests adialectical relationship between social structureand sensemaking: the accounts generated by sense-making facilitate the formation and reformation ofsocial structure (the social roles and relationshipsamong some group of actors), while social roles andrelationships provide a basis for sensemaking. Thus,Weick argued that without social roles and relation-ships in place, sensemaking can be difficult or impos-sible, leading to confusion and distress. In a studyof sensemaking on aircraft carriers, Weick and Rob-erts (1993) developed the concepts of “collectivemind” and “heedful interrelating” to describe howorganization members can generate reliable collec-tive action. Their study moved sensemaking awayfrom a strictly cognitive metaphor by associatingcollective mind not with a superordinate set ofcognitions, but with “a pattern of interrelated ac-tivities among many people” and by offering theargument that the intelligence of a collective minddepends upon the heedfulness with which peopleinterrelate.

Organizational disasters and crises have also pro-vided a fruitful context for research into sensemak-ing among multiple parties (Brown, 2000; Brown &Jones, 2000; Gephart, 1992, 1993). Emerging fromstudies of organizational inquiries is an image ofsensemaking in which participants work to findacceptable grounds for the assignment of responsi-bility. In Gephart’s 1993 study of a disaster inquiry,for instance, participants engaged in a multistagesensemaking process that drew on a range of localorganizational and individual attributes: formal or-ganizational schemes were used to identify poten-tially responsible individuals; these individualswere evaluated with respect to organizationalschemes and social norms; the evaluations werethen used to construct a subset of individuals with“uniquely problematic or laudable selves”(Gephart, 1993: 1507); and finally, responsibilitywas assigned on the basis of an emergent sense ofinterpretations shared by participants, with the ex-ception of those ultimately assigned responsibility,

22 FebruaryAcademy of Management Journal

who held divergent views. This research highlightsthe importance of attending to the process throughwhich agreement is achieved in sensemaking, andparticularly to the ways in which sensemaking pro-cesses interact with features of the settings inwhich they occur (Brown, 2000; Gephart, 1993).

From these studies of sensemaking in crisis situ-ations, three critical issues emerge. First, this re-search has focused on situations in which there issome pressure (and sometimes immense pressure)to make sense of the world quickly. This type ofsituation is not necessarily representative of orga-nizational sensemaking more generally, becauselong spans of time may pass in which organizationmembers remain confused by events and actionswithout developing sensible accounts. Second,these studies of crises have tended to examine rel-atively tightly coupled social systems, such as firecrews, flight deck teams, and industrial disasters,where members’ interpretations and actions typi-cally have direct and relatively immediate conse-quences. No such direct consequences may beobservable, however, when one considers sense-making across a diverse set of issues involvinglarger numbers of organizational stakeholders.Third, high-reliability environments, such as firefighting and aircraft carrier flight decks, may besomewhat distinct in their demands for sensemak-ing that provides members with shared accountsthat facilitate tightly coordinated collective action.In the more mundane world of most organizations,it may be that social processes of sensemaking oc-cur without resulting in broadly shared accounts.

The two sets of sensemaking research reviewedabove led to the research questions that guide thisstudy. The studies that have examined the socialprocesses of sensemaking in large organizationshave highlighted the critical roles played by leadersand stakeholders but have tended to focus on therole of only one party at a time. Most of whatresearchers know about interactive sensemakingamong organization members comes from researchon sensemaking in crises or under intense pressure,conducted over quite short time periods. Scholarstherefore understand relatively little about howheterogeneous sets of sensemaking parties interactin ongoing and quite ordinary sensemaking pro-cesses over extended periods of time. Further, al-though these studies have articulated insightful de-scriptions of important sensemaking processes,such as heedful interrelating (Weick & Roberts,1993), they have tended to provide singular de-scriptions of sensemaking processes. Relatively lit-tle attention has been paid to the potential variancein these processes, and especially to the heteroge-neous patterns of interaction that might be associ-

ated with sensemaking across a broader range ofsituations involving a diverse range of stakehold-ers. Together, these gaps lead to the first researchquestion of this study:

Research Question 1. Are there discernablepatterns of interaction that constitute the so-cial processes of sensemaking among diversestakeholders in complex organizations?

A second important aspect of sensemaking inorganizations is what it produces: accounts and theactions that are based on them. As discursive con-structions of reality that provide members with or-dered representations of previously unordered ex-ternal cues, accounts describe or explain the worldand thus make it meaningful (Antaki, 1994; Potter& Wetherell, 1987). They can take a wide variety offorms, such as descriptions of issues as political orstrategic, and explanations of the behavior of forestfires. Accounts are a critical resource that allowindividuals to accomplish work and negotiate theirday-to-day lives (Antaki, 1994; Boje, 1991; Gergen,1999), constructing ordered relationships amongsets of entities (events, people, actions, things) inways that enable people to act or at least to decideto act (Sackman, 1991; Starbuck & Milliken, 1988;Weick, 1993). The connection between accountsand action is particularly highlighted in smallgroup studies: a critical part of Weick’s explanationof the Mann Gulch disaster was the fire crew’sfailure to construct an account of the situation thatwould facilitate collective action. A similar dy-namic occurs during organizational change (Bar-tunek et al., 1999; Gioia & Chittipeddi, 1991), inwhich accounts are key to facilitating collectivetransformation. Although a range of studies havehighlighted the link between sensemaking and itsoutcomes, there has been no systematic examina-tion of this link. This gap leads to the presentstudy’s second research question:

Research Question 2. Are there patterns ofaccounts and action associated with the dif-ferent social processes of sensemaking inorganizations?

METHODS

This study used a qualitative methodology toaddress its two research questions. Qualitativemethods are well suited to the study of dynamicprocesses, especially where these processes areconstituted of individuals’ interpretations (Gioia &Thomas, 1996; Hinings, 1997). Because qualitativeresearch typically examines issues from the per-spective of the participant (rather than from that of

2005 23Maitlis

the researcher), it is especially appropriate, andtherefore frequently used, in the study of organiza-tion members’ constructions and accounts (e.g.,Dutton & Dukerich, 1991; Isabella, 1990). Further-more, because of its sensitivity to organizationalcontext and its potential for focusing upon activitysequences as they unfold, qualitative research is avaluable means of investigating dynamic processesin organizations (Pettigrew, 1992).

The aim of this study was theory elaboration,drawing on and extending important ideas fromresearch on organizational sensemaking. Theoryelaboration is often used when preexisting ideascan provide the foundation for a new study, obvi-ating the need for theory generation through apurely inductive, grounded analysis (Lee, Mitchell,& Sablynski, 1999). This study used a multiple casedesign (Lee, 1999; Yin, 1994), in which I tracedorganizational sensemaking processes in real time,as they unfolded over a two-year period in ninedifferent issue domains for each of three matchedorganizations. This design offered a strong founda-tion for elaborating theory: the similarity of thethree organizations allowed for meaningful com-parisons across the social processes and actors in-volved, while the diversity of issue domains andthe differences between the organizations provideda reasonable basis for generalizability.

Research Context

The study was carried out in three British sym-phony orchestras, a context well suited to sense-making research for several reasons. First, culturalindustries offer an environment in which sense-making is particularly critical. Because their prod-ucts are experiential, judgments about quality aresubjective and shifting, creating an uncertain, dy-namic environment in which members must en-gage in “interpretive enactment” to manage suc-cessfully (Hirsch, 1972; Lampel, Lant, & Shamsie,2000). Murnighan and Conlon (1991), for instance,showed that successful string quartets engaged incomplex and subtle sensemaking practices, ac-knowledging contradictions inherent to their work,but managing them implicitly, rather than trying toresolve them openly. The management of meaningis a central and critical activity in such organiza-tions, where products and processes are ambiguousand highly subjective.

Second, within the cultural industries, orches-tras offer an especially powerful site for a study ofsensemaking. They represent a very traditional or-ganizational form forced to operate in an increas-ingly challenging environment (Allmendinger &Hackman, 1996), one that is characterized by a

decline in government subsidy and corporate spon-sorship, and by growth in the number and varietyof competing leisure activities (Lebrecht, 1996;Robinson, 2000). Allmendinger and Hackman’s(1996) study of 78 orchestras in four nations high-lights the considerable impact of environmentalturbulence on these cultural institutions’ internalprocesses. Examining the effects of the fall of so-cialism on East German orchestras, they found sig-nificant differences in the orchestras’ abilities toadapt that were due in part to initiatives taken byboth leaders and musicians to deal with the chang-ing environmental conditions.

Third, the presence for orchestras of numerousand powerful stakeholders with frequently op-posing interests creates a context in which thesocial processes of organizational sensemakingare both important and visible (Elsbach, 1994).An orchestra is dependent not only on its musi-cian workforce and administrative team, but alsoon a wide range of other actors, such as conduc-tors, soloists, trustees, board members, and cus-tomers. These parties often have divergent inter-ests, a central tension of the business stemmingfrom conflicts between artistic goals and com-mercial pressures. Such a stakeholder-rich con-text makes the social processes of sensemakingparticularly significant. Moreover, the relativelysmall size of orchestras makes leaders and keystakeholders easily identifiable, facilitating theobservation of sensemaking processes.

Orchestras also provide a surprisingly high levelof generalizability to a broader population of organ-izations. In many respects, orchestras operate likemost medium-sized private enterprises: they haveproducts (concerts and recordings) that must becompetitively marketed and sold, and a wide rangeof customers to satisfy. Thus, as for any business,financial viability is a central issue. Their structureis also quite similar: orchestras are supported byadministrative teams that typically include fi-nance, marketing, operations, and human resourcesfunctions and are headed by CEOs who are ac-countable to boards of directors. Board directors’roles are unambiguous—they are not simply well-intentioned volunteers, but representatives of par-ticular community stakeholder groups who are le-gally accountable for the actions they take vis a vistheir orchestra. Key stakeholders include govern-ment, customers, employees, unions, and localcommunities. Because goals, structures, and stake-holder roles are comparable in orchestras and inother, similarily sized, organizations, one wouldexpect to find social processes of organizationalsensemaking in orchestras that are similar to thoseone would find in those other organizations. More-


over, many traditional industries are taking on thequalities of cultural industries as the products andservices they provide are increasingly sold on thebasis of their symbolic dimensions (Lawrence &Phillips, 2002). Orchestras thus provide a transpar-ent example of an increasingly common set of or-ganizational challenges.

Those in the orchestra sector group the 13 per-manent professional British symphony orchestrasinto three categories: provincial orchestras, BBCorchestras, and London orchestras (BBC/ArtsCouncil of England, 1994). Although all 13 work tosimilar objectives (concert giving, recording, andeducation/outreach), they differ in their primaryrevenue sources and governance structures. Provin-cial orchestras typically give at least three localconcerts per week, make occasional CD recordings,and tour overseas every one or two years. Musi-cians in these orchestras often feel they make animportant contribution to the cultural life of theirlocal communities. BBC orchestras, in contrast,spend most of their time in the studio, makingrecordings for radio broadcast. They give fewer liveconcerts and often play more unusual repertoire,making them much more dependent on publicfunding than on ticket sales. Musicians in the Lon-don orchestras are freelance rather than salaried,paid for each session they play with the orchestras.These orchestras tend to be the busiest, performingmost frequently, both in London and abroad, andoften recording for CD and film. The London or-chestras are proudly “self-governing,” each offi-cially run by a musician-dominated board that em-ploys an executive director and a team ofadministrators as managers.

The specific orchestras were chosen to meet thestudy’s aims. I sought a balanced sample of organi-zations in which issues and processes could becompared but in which there would also be somesignificant heterogeneity. Three orchestras, re-ferred to here as the provincial, broadcasting, andLondon orchestras, made up the sample; each wastaken from one of the three categories of Britishsymphony orchestra. Because the focus of thisstudy was on “ordinary” sensemaking taking placein “normal” organizations, I chose orchestras thatwere midlevel performers, neither the strongest northe weakest (artistically or financially) in their cat-egories. Doing this avoided my encountering pat-terns of sensemaking that might only be associatedwith unusually high or low performance and thusincreased the comparability of the contexts and thegeneralizability of the findings. In addition, be-cause data collection would be intensive, I choseorchestras that were not geographically remote

from my location in England (e.g., in Scotland orIreland).

Data Collection

Data collection was intensive over more than twoyears and involved interviews; observation of meet-ings, rehearsals, and orchestra tours; and extensivedocumentary analysis. Studying the organizationspart-time, as an overt nonparticipant observer, Igained access to each organization through an in-terview with its executive director and a subse-quent meeting with that individual in which I pro-posed my research. I then made a presentation tothe full orchestra to explain the study’s aim andapproach. All members understood me as an orga-nizational researcher and amateur musician with apassion for and fascination with symphony orches-tras. The data were largely analyzed after the studyperiod, but I honed my methods during the datacollection as I came to understand the issues, peo-ple, and contexts. For example, I did not begin byfocusing on nine specific issues or ten stakeholdergroups, but over many months identified these ascritical and common to the three orchestras’ sense-making.

In total, I conducted 120 formal interviews,carrying out repeat interviews with key infor-mants. Interviewees were selected using formal,snowball, and opportunistic sampling methods.In each orchestra, I interviewed the executivedirector and every member of the managementteam, as well as the primary nonmusician “over-seer,” and the principal conductor, where onewas in post at the time of study (for the broad-casting and provincial orchestras). I also inter-viewed all musicians who served as representa-tives on the boards of their orchestras (for theprovincial and London orchestras) or on theirorchestra committees (for the provincial andbroadcasting orchestras). In addition, I used asnowball sampling technique to identify otherinterviewees; for example, people were suggestedto me because they had been in the orchestra along time or had previously served as represen-tatives on a board or an orchestra committee. Aswell as using these formal and snowball samplingstrategies, I also interviewed any musician whoexpressed an interest. Because of the amount oftime I spent in each organization, and especiallytraveling on tour, there were many opportunitiesto talk to a very wide range of musicians.

The interviews were semistructured, and theybecame increasingly focused over the course of thestudy. The domains covered in early interviewsincluded the individuals’ organizational roles, the

2005 25Maitlis

main issues in which interviewees had been in-volved and those from which they had felt ex-cluded, and their perceptions of decisions and howthey had been made. Later in the study, and partic-ularly in repeat interviews with key informants(individuals who were involved in many of theissues), I modified my interview protocol to focuson sensemaking around specific issues and eventsthat related to each organization. In interviews con-ducted immediately after a significant meeting, Ioften asked certain questions relating specificallyto the discussions of that meeting, before broaden-ing out the questions to follow up on other matters.Table 1 summarizes types of interviewees and enu-merates them.

In addition to the recorded and transcribed inter-views, I conducted a large number of informal in-terviews throughout the study period; these oftentook place in buses, bars, and restaurants. Althoughthese generally covered the areas listed above, theyalso provided opportunities to pick up on thingsthat had just occurred (for instance, I might discusswith musicians over lunch what had just happenedin a meeting) and to observe groups of people in-formally talking, which offered insight into certaingroup dynamics (such as interactions betweenmanagers and musicians). At times I had the im-pression that the informal setting, and possibly thelack of tape recorder, led individuals to talk morecandidly than they might otherwise.

TABLE 1Summary of Data Sourcesa

Data Source

Orchestra

Provincial Broadcasting London

Formal interviews

Executive director 5 8 4Other managers 10 17 14Orchestra committee members 6 9 10b

Other players 4 3 10Player board members 10 n.a. 10b

Nonplayer board members/overseers 4 2 1Principal conductor 1 2 n.a.Total 40 41 39

Meetings observed

Senior management meetings 8 3 n.a.Orchestra committee meetings 2 2 1Orchestra committee and management meetings 2 4 n.a.Full orchestra meetings 4 5 –Management presentation to full orchestra 4 2 –Musicians union presentation to full orchestra 1 – –Office staff meetings 1 8 n.a.Board meetings 8 n.a. 7Board subcommittee meetings 10 n.a. n.a.Consultative committee meetings 5 n.a. n.a.Consultative committee subcommittee meetings 4 n.a. n.a.Management committee meetings n.a. n.a. 4Trustee meetings n.a. n.a. 2Senior management retreat 1 1 n.a.Board retreat 1 n.a. n.a.Team-building and leadership skills workshop n.a. 1 n.a.Education sessions – 2 –Foreign tours – 1 (7 days) 1 (4 days)Annual general meetings 2 n.a. –Meetings between orchestras or with other

external stakeholders7 9 1

Total 60 38 16

a An “n.a.” entry indicates an orchestra did not hold this type of meeting; an underscore, that the author did not attend any meeting ofthe indicated type for that orchestra.

b The London Orchestra Committee members were also its player board members; therefore, these ten interviews are only included oncein the total.


As well as conducting interviews, I observed1071 meetings, which included meetings of variousgroups within each organization (e.g., executiveteam, board meetings), and also meetings betweenorchestra leaders and those external to their organ-izations (e.g., funders, collaborating organizations).Discussion at these meetings addressed a wide va-riety of issues, including key conductor, player,and administration appointments, commercial is-sues such as ticket pricing and sponsorship, artisticstrategy and repertoire for key concerts, as well asmore mundane matters, such as the route an or-chestra bus should take on its way out of town.Throughout each meeting, I noted verbatim, as faras possible, what was said and by whom. Table 1also summarizes the types of meetings I observedand enumerates them.

Data Analysis

Data analysis comprised three main stages. First,I developed narratives that described the sense-making processes associated with a set of issuesthat arose in all three organizations during thestudy. In stage 2, I sought to answer the first re-search question: Are there discernable patterns ofinteraction that constitute the social processes ofsensemaking among diverse stakeholders in com-plex organizations? In stage 3, I addressed the sec-ond research question: Are there patterns of ac-counts and action associated with the differentsocial processes of organizational sensemaking?The analyses are described in detail below, andTable 2 summarizes the stages of analysis.

Stage 1: Creating narratives of the sensemak-ing processes. The first data analysis stage beganwith listing every organizational issue that arose inthe three orchestras during the study period (150–200 issues in each). I defined an organizationalissue as a topic of discussion that involved a ques-tion or concern connected in some way to the or-ganization as a whole, rather than to a small subsetof its members; for instance, a personal conflictbetween two players with no broader consequenceswould not be considered an organizational issue. Ithen reduced this list of all the issues in everyorchestra to a comprehensive set of “typical” issuesfacing a British orchestra. Four criteria for “typical-ity” were used. First, an issue had to have arisen inall three organizations. Meeting this criterion reliedon a process of data reduction (Strauss & Corbin,

1998) in which I grouped the issues thematicallyinto broader issue domains (for instance, I groupedinto one domain the issue of whether or not torenew the contract of the principal conductor,which faced two orchestras, with the issue of ap-pointing a new principal conductor, which facedthe third). The second criterion for inclusion in thesubset of typical issues was that data for the se-lected domains should be available from multiplesources and largely gathered as the issues unfoldedduring the study period, rather than retrospec-tively. The third criterion was that, together, thedomains should cover artistic, financial, and per-sonnel areas, and include both strategy and opera-tions. The fourth criterion was that organizationmembers should consider an issue to be of realsignificance. After developing a draft set of issues,I refined it through discussions with key respon-dents in each orchestra. All respondents stated thatthe final set of nine domains covered the majorissues that had arisen during the study period andthat they together captured the main categories ofissues normally faced by their organizations. Thenine issue domains were as follows: (1) program-ming of repertoire, (2) principal conductor appoint-ment/contract renewal, (3) re-engagement of guestconductors, (4) dealing with unsatisfactory playerperformance, (5) appointment of key players (e.g.,section principals), (6) players’ pay and contractreview, (7) identifying areas for cost cutting, (8)increasing income generation, and (9) collaborativeventures (with other organizations).

Following Eisenhardt and Bourgeois (1988), Ithen developed a narrative of between 5 and 25single-spaced pages that chronicled the sensemak-ing activities for each domain in each of the orches-tras. I constructed these 27 narratives by tracingthrough all the chronologically ordered raw datafor each orchestra to identify every time a givenissue arose, and then describing the organizationalsensemaking process through quotations from acombination of interviews, meeting transcripts, ar-chival documentation, and field notes. The narra-tives were thus composites made up of data fromall four sources.

Stage 2: Identifying forms of organizationalsensemaking. I next analyzed the 27 narratives inrelation to the first research question—ascertainingwhether there were discernable patterns of interac-tion that constituted the social processes of sense-making across the domains. In searching for pat-terns of interaction in organizational sensemaking,I was concerned with potential consistencies in, forexample, the ways in which stakeholders and lead-ers talked among themselves and to each other,how often different parties met and under what

1 Seven of these meetings were attended by membersof two orchestras; hence, the total number of meetingsobserved does not match the sum of the subtotals listedfor each orchestra in Table 1.

2005 27Maitlis

circumstances, and the rhythm of those interac-tions over time. This stage of the data analysis hadthree parts.

The first part involved identifying key stakehold-ers—actors who affected or were affected by one ofthe three studied organizations (Freeman, 1995)—and examining their contribution to organizationalsensemaking in each issue domain. During thefieldwork and analysis of the narratives, various

individuals and groups emerged as prominent, ei-ther playing major roles in virtually all issue do-mains or making critical contributions to two orthree. By combining certain groups and individuals(for instance, pooling all of an orchestra’s adminis-trators into one “administration” category), it waspossible to identify a set of stakeholders that werecomparable across issues and organizations. Afterrefining the set on the basis of feedback from key

TABLE 2Stages of Data Analysis

Data Analysis Stage Tasks Outputs

(1) Creating narratives of thesensemaking processes

1. Listing all issues arising in the three orchestras.2. Reducing issue lists by applying four criteria.3. Modifying and refining the list on the basis of

discussions with key informants.4. Constructing narratives for each of the key issue

domains from interview, meeting, archival datasources, and field notes.

1. Set of 9 key issue domains.2. Set of 27 narratives.

(2) Identifying forms oforganizational sensemaking

(a) Identifying stakeholders andanalyzing stakeholdersensegiving

1. Identifying all individuals and groups playing amajor role in virtually all 9 key issue domains ora very significant role in 2–3 domains.

2. Pooling certain groups and individuals to form aset of ten stakeholders whose activities werecomparable across issues and organizations.

3. Refining the list on the basis of discussion withkey respondents.

1. Set of ten key stakeholders.2. Classification of high or low overall

stakeholder sensegiving assigned toeach of the 27 issue domains.

4. Developing a list of stakeholder sensegivingactivities, drawing on existing theory.

5. Identifying instances of stakeholder sensegivingin each issue domain.

6. Determining overall level of stakeholder sense-giving in an issue domain by counting thenumber of stakeholders engaged in sensegivingactivities, and through a gestalt analysis of thefrequency and intensity of sensegiving activity.

(b) Analyzing leadersensegiving

1. Building on the stakeholder list to develop a listof leader sensegiving activities.

2. Identifying instances of leader sensegiving ineach issue domain.

1. Classification of high or low overallleader sensegiving assigned to eachof the 27 issue domains.

3. Classifying leaders’ roles as predominantlyshowing high or low sensegiving in an issuedomain, on the basis of a qualitative assessmentof the frequency and intensity with which theactivities were carried out.

(c) Identifying forms andprocess characteristics oforganizational sensemaking

1. Grouping the 27 issue domains into fourcategories (forms) representing intersections ofleader and stakeholder sensegiving roles.

1. Set of robust process descriptors foreach organizational sensemakingform.

2. Investigating the nature of leader-stakeholderinteraction for each form.

3. Tracing each organizational sensemaking process todetermine the process characteristics for each form.

4. Conducting iterative, cross-case analysis to buildmore abstract process descriptors.

(3) Identifying outcomes stemmingfrom each form

1. Identifying commonalties in the accounts andactions associated with each form.

2. Conducting iterative, cross-case analysis to buildup more abstract descriptions of sensemakingoutcomes.

1. Set of robust process outcomes foreach organizational sensemakingform.


informants in each orchestra, I had a set of tenstakeholder groups common to all three organiza-tions: orchestra administration, player representa-tives, section principals, players as a group, princi-pal conductor, overseers, customers, funders, guestartists, and the musicians’ union. For each organiza-tion, the leaders were the two people with greatestformal authority: in the provincial or broadcastingorchestras, these individuals were the executivedirector and the next most senior administrator; inthe London orchestra, the leaders were the execu-tive director and the chairman of the board.

Each of the 27 sensemaking processes was thenanalyzed to identify how stakeholders and leaderscontributed through various sensegiving activities.Stakeholder groups and leaders were identified as“high sensegiving” in a domain when they predom-inantly engaged in behaviors that attempted to in-fluence others’ sensemaking in that domain, and as“low sensegiving” when they exhibited very fewsensegiving behaviors. Included as sensegiving be-haviors were statements or activities that involvedproviding plausible descriptions and explanationsof extracted cues and constructing sensible envi-ronments for others (Weick, 1995). Drawing on thisconceptualization of sensegiving, as well as on areview of sensegiving activities identified in thesensemaking literature (e.g., Bartunek et al., 1999;Dunford & Jones, 2000; Gioia & Chittipeddi, 1991;Pratt, 2000), I examined a subset of the sensemak-ing narratives to develop a list of concrete sense-giving activities in the orchestras, which was thenused to identify instances of sensegiving in eachissue domain. Examples of sensegiving activitiesincluded contesting a proposal, calling a meeting,explaining a situation, issuing a warning, express-ing an opinion, writing a report, justifying a view,promoting a position, gossiping, and taking min-utes. Some sensegiving activities were unique toleaders, such as presenting an executive director’sreport to a Board, but the majority were common toleaders and stakeholders. For each sensemakingprocess, I determined the overall level of stake-holder sensegiving in two ways. The first was bycounting the number of stakeholders involved insensegiving for that process. The second wasthrough a more impressionistic “gestalt analysis”(Gioia & Thomas, 1996) of the frequency and inten-sity with which stakeholders engaged in the abovesensegiving activities, which provided a richer,more holistic assessment than number of stake-holders alone would have done. In a similar way, Idetermined leaders’ degree of sensegiving througha qualitative assessment of the frequency and in-tensity with which the activities identified in thenarratives were carried out.

The final part of stage 2 directly investigated thepossibility of identifiable social processes of organ-izational sensemaking based on the interaction ofleaders’ and stakeholders’ sensegiving behaviors. Ifirst grouped the 27 issue domains into four cate-gories that represented the intersection of leadersensegiving (low or high) and stakeholder sense-giving (low or high). I then examined the patternsof leader and stakeholder interaction that distin-guished each category, or what I refer to in the restof the article as forms of organizational sensemak-ing. To do this, I traced through each narrative,developing a set of process descriptors that drew onquotations from interviews, meetings, and fieldnotes. At this first level of analysis, these descrip-tors included terms such as “duplicitous,” “doingdeals,” and “left out of discussion.” Once I had setsof descriptors for all issue domains for a given formof sensemaking, I then engaged in an iterative,cross-case analysis, building up broader, more ab-stract categories (e.g., private meetings with keyparties) that captured the key characteristics of theorganizational sensemaking process for that form(Gioia & Thomas, 1996; Strauss & Corbin, 1998).The final part of this analysis involved combiningand refining these more abstract categories to de-velop core dimensions that captured the pattern ofsensemaking in each of the 27 instances and dis-tinguished among the four different forms. The endresult of stage 2 was thus a set of four forms oforganizational sensemaking, each described bycore process characteristics.

Stage 3: Identifying the outcomes of each formof organizational sensemaking. In the final dataanalysis stage, I explored potential patterns in theoutcomes associated with each form of organiza-tional sensemaking. My focus was on the accountsand actions generated (including proposals, stories,and decisions) and actors’ descriptions of theseaccounts and actions. I again used the constantcomparative method, gradually building up morerobust descriptive categories from the data. I beganby tracing through the sensemaking narratives foreach form to develop a set of descriptors that cap-tured the outcomes for that form. These descriptorswere based on quotations or observations from theraw data, such as “First we lay out key pieces of theprogram plan and then over time we fill in thespaces” and “Contract is renewed for a year to seewhat alternatives arise for conductor and orches-tra.” Then, through an iterative, cross-case analysis,I built up more abstract descriptions of the outcomesfor each form (e.g., “emergent series of actions”).

Tables 3–6, in the Results section, summarizethis analysis and show that the strength of evidencevaried for each process characteristic and outcome

2005 29Maitlis

in each issue domain. Evidence was consideredstrong where strong support (multiple sources,several examples) was found in the narrative forboth dimensions of the process or outcome (thatis, for sensemaking processes, evidence of how“animated” and controlled the process was). Ev-idence was considered moderate where fewersources and only a few examples were found in thenarrative for both dimensions, or where strong ev-idence was found for one dimension. Evidence wasconsidered weak when only moderate evidence ex-isted for just one dimension of the process or out-come in question. Although it would also havebeen interesting to examine the antecedents of thesensegiving behaviors, such an analysis lay beyondthe scope of this study.

While some researchers argue that the conven-tional criteria of reliability and validity cannot beapplied to interpretive research, it is neverthelessessential to show that the findings of a qualitativestudy are representative of the phenomenon of in-terest (Lee, 1999). There are several ways to in-crease the credibility of a naturalistic study. Theseinclude having “prolonged engagement” at one’sdata site and undertaking “persistent observation”(Lincoln & Guba, 1985); collecting comprehensive,descriptively rich data, ideally through “triangula-tion” in data sources and methods (Maxwell, 1996);and conducting “member checking,” a procedurein which “data, analytic categories, interpretationsand conclusions are tested with members of thosestakeholding groups from whom the data were orig-inally collected” (Lincoln & Guba, 1985: 314). Thisstudy used a longitudinal design, which involvedmy spending several hundred hours at the researchsites over a two-year period. Furthermore, the datawere gathered from numerous sources through a va-riety of methods. Data analysis was conducted on 27cases drawn from three different organizations. Theresults of this study were not based on a solitaryobservation or the casual comments of a few individ-uals, but were developed through a rigorous and iter-ative procedure that made full use of the wealth andcomplexity of the data collected. Throughout myanalysis, I also sought to identify discrepant cases andsought feedback on my interpretations and conclu-sions both from the study’s participants and fromseveral individuals external to the study.

RESULTS

Effects of Leader and Stakeholder Sensegiving onthe Process Characteristics of OrganizationalSensemaking

As described above, the degrees of leader andstakeholder sensegiving emerged as two important

dimensions that differentiated sensemaking acrossthe 27 issue domains. Examining these dimensionsled to two important findings. First, each of thesedimensions distinctively affected the process char-acteristics associated with organizational sense-making. High levels of leader sensegiving led tosensemaking processes that were highly controlled.One important aspect of these controlled processeswas that sensegiving by both leaders and stakehold-ers tended to occur in an organized, systematicfashion, rather than ad hoc: controlled sensemak-ing processes were dominated by scheduled meet-ings, formal committees, and planned events withrestricted attendance, rather than by informal, im-promptu meetings of self-organizing groups. Sense-making occurred in this controlled way both be-cause leaders drew on their formal authority toorganize sensegiving occasions in which issueswere discussed through formal channels, and be-cause stakeholders responded to leader sensegivingby participating in and supporting these organizedopportunities for sensegiving. A second key aspectof controlled processes was that a significantamount of sensegiving occurred in private meetingsbetween stakeholders and leaders, rather than inmore public, open forums. Leaders engaging inhigh levels of sensegiving were able to use keyresources available to them, such as time, space,and their personal networks, to create opportuni-ties to meet stakeholders one-on-one so that discus-sions could take place away from the scrutiny ofothers.

An example of sensemaking in which leadersengaged in high levels of sensegiving was the caseof programming in the broadcasting orchestra.Here, broadcasting leaders arranged to meet one-on-one with a range of stakeholders, each of whomheld views about suitable repertoire. In these pri-vate meetings, stakeholders such as the principalconductor and guest conductors, the BBC commis-sioning manager, local concert promoters, and arecording company each conveyed their prefer-ences and argued for certain pieces of music. Thesesensemaking processes were thus controlledthrough leader and stakeholder interaction thattook place in scheduled, private, meetings, whichallowed each party to more freely express needsand interests and offered them the opportunity toadjust and refine their accounts of the situationaway from the public eye.

In contrast, when leaders demonstrated lowerlevels of sensegiving, sensemaking processes wereless controlled. Low levels of leader sensegivingled to fewer mechanisms that might organize sense-giving activities. As well, sensegiving tended not tooccur in private conversations between leaders and


stakeholders but instead took place in open forumsthat included a large number and variety of parties.In the case of the provincial orchestra’s principalconductor reappointment issue, for instance, lead-ers engaged in very little sensegiving: when theydid, it primarily occurred in meetings at whichthere was little attempt to control either attendanceor what was discussed. Numerous discussionsabout the renewal and possible alternatives tookplace between leaders and the board, the ArtsCouncil of England, and musicians, but leaders didnot try to direct either the timing or content of thesediscussions.

The effect of high levels of stakeholder sensegiv-ing on sensemaking processes was distinctly differ-ent from that of leader sensegiving. When stake-holders engaged in extensive sensegiving, theycreated sensemaking processes that were highlyanimated. A central characteristic of this anima-tion was an intense flow of information: leadersroutinely reported back to their boards, executiveteams, and other stakeholders, and information wasalso regularly shared among stakeholder groups.The presence of stakeholders who were activelyengaged in shaping the interpretations of eventsand issues resulted in a greater circulation of infor-mation, both directly from those stakeholders andfrom leaders who were motivated to provide infor-mation in response to stakeholder activity. The is-sue of collaborative ventures in the provincial or-chestra, for example, involved a great deal ofstakeholder sensegiving and, consequently, highlevels of information flow. Sensemaking in thisdomain revolved around a potential joint venturethat was first discussed at both a board subcommit-tee and a senior management meeting and subse-quently conveyed by memo to all board members.Musician representatives, informed at a consulta-tive committee meeting, quickly shared news of theplan among their colleagues. In a short time, allmembers of the organization and its key stakehold-ers were aware of the proposal and, at least to somedegree, understood it.

The second characteristic of these animated pro-cesses was their continuous rhythm: sensemakingaround these issues remained active over an ex-tended period. The diversity of interests and per-spectives of the various stakeholders engaged insensegiving led to sensemaking processes that werenot resolved quickly or easily, processes in whichdifferent stakeholders engaged in the conversationat different times. Thus, sensemaking in animatedprocesses tended to occur in iterative discussionsthat continued over many months, as numerousstakeholders volunteered their opinions and statedtheir demands, and leaders worked to articulate

their own accounts of the issues of concern. In thecase of income generation in the London orchestra,for instance, stakeholders were highly engaged insensegiving: throughout the study period therewere many discussions between leaders and vari-ous stakeholders about income generation schemes,such as overseas tours, U.K. residencies, recordings,and government grants. Approaches to income gen-eration affected different stakeholders in differentways, which engendered significant sensegiving bymany groups and led to a sensemaking process thatwas active and ongoing for many months.

When stakeholders did not engage in significantlevels of sensegiving, sensemaking processes werenot animated. In these situations, there was a rela-tively light flow of information, few stakeholdersaware of an issue’s existence or importance, andstill fewer knowing its details. Moreover, low stake-holder sensegiving typically led to sensemakingprocesses with broken rhythms, rather than thecontinuity associated with highly animated pro-cesses: sensemaking primarily occurred in inter-mittent leader-stakeholder discussions that tookplace over an extended period. These discussionswere punctuated by occasional bursts of leadersensegiving to key stakeholder groups; most stake-holders, however, received leaders’ accounts onlyas sensemaking processes neared resolution. Thearea of collaborative ventures in the London orches-tra typified such sensemaking processes. Sense-making in this domain focused on the issue ofwinning a government grant for innovative newventures. The grant proposal was largely developedby the London orchestra’s executive director, withlittle input from the collaborating partners; discus-sions were kept quiet, with only those directly in-volved aware of the confidential plans. Only at keypoints in the unfolding process did the executivedirector relay updates and put forward his analysisof the situation to the orchestra’s board and otherkey stakeholder groups.

Leader and stakeholder sensegiving each fueledsensemaking processes with a particular form ofsocial energy: leader sensegiving produced con-trolled sensemaking with highly organized, system-atic processes in which there tended to be numerousprivate, one-on-one interactions between leaders andstakeholders; stakeholder sensegiving animatedsensemaking processes so that information flowedamong participants and the process remained contin-uously active over the life of the issues.

Four Forms of Organizational Sensemaking

The second important finding of this study con-cerned the interaction of the dimensions of anima-

2005 31Maitlis

tion and control. In each of the 27 issue domains,sensemaking processes were associated with acombination of animation and control that resultedin the emergence of four distinct forms of organi-zational sensemaking. When a sensemaking pro-cess was controlled and animated, organizationalsensemaking took on a guided form. Organizationalsensemaking was fragmented when the processwas animated but not controlled. When the processwas controlled but not animated, sensemakingemerged in a restricted form. Processes that wereneither controlled nor animated produced a mini-mal form of sensemaking. The four forms werequalitatively different not only in their proces-sual characteristics (research question 1) but alsoin their outcomes (research question 2).

The outcomes of the four forms were found tovary in the kinds of accounts and actions that were

generated. The accounts constituted descriptions ofan issue and its context and varied as to whetherthey were: (1) unitary or multiple and (2) rich ornarrow. The actions associated with these pro-cesses primarily involved organizational decisionsand varied in the degree to which they were: (1)enacted in emergent series or took the form of one-time actions (or planned sets of actions) and (2)consistent or inconsistent with other actions con-nected to the issue. Figure 1 summarizes the pro-cess characteristics and outcomes of each of thefour forms of sensemaking. In the remainder of thissection, I describe each of the four forms of organi-zational sensemaking, in each case first providingan example from the data and then discussing theform’s process characteristics and outcomes. Ta-bles 3–6 fully summarize process characteristicsand outcomes for each form by issue domain.

FIGURE 1Four Forms of Organizational Sensemaking


TABLE 3Guided Organizational Sensemakinga

Orchestra andIssue Domain Process: Animated and Controlled Accounts: Unitary and Rich Actions: Emergent and Consistent

1. Provincial:Incomegeneration

*** Numerous private meetings between theexecutive director and stakeholders todiscuss income generation, includingmeetings with local funders (e.g., citycouncil and corporate sponsors) to lobbyfor increased funding and convey theorchestra’s opportunities/challenges.Stakeholders responded by explainingtheir financial and philanthropicpositions. Musicians also contributedviews in a large number of meetings. Theexecutive director held meeting with anArts Council representative to establishwhy the orchestra was not receivingadditional funding. The executive directorcalled a meeting with the Board to givehis account of the discussion.

*** All parties came to understandthat the orchestra was in direneed of additional income andsupported the proposal to drawfunding from as many sources aspossible. The account of thesituation that emerged from theprocess emphasized theorchestra’s financial needs andthe roles that different funderscould play in meeting thoseneeds.

*** Over time, the executive directorworked with different stakeholdersto carry out a stream of incomegeneration activities (e.g.,corporate sponsorship, fund-raisers, increasing public funding,and building up the number of theorchestra’s out of townengagements). Each action wasintroduced separately but wascongruent with the others and theoverall aim of income generation.

2. Provincial:Collaborativeventure

** Provincial leaders organized a series ofseparate meetings with their partnerorchestra, subcommittees, and executiveteam to formulate, evaluate, and modifyproposals. After the story was leaked tothe press, the executive director sent amemo to inform board members of aproposal. Musicians were told of plans ata meeting of the Consultative Committee.

*** Leaders and stakeholdersdeveloped a sharedunderstanding of the nature ofthe collaboration and the reasonssupporting it: the collaborationwould first include sharedaccommodation and jointmarketing, and over time,perhaps more significantprojects. It was understood asadvantageous to the orchestras,with the potential to spawnadditional types of work forboth.

*** Initially, the Provincial sharedits accommodation with thepartner orchestra. Then some jointmarketing initiatives were started.Subsequently, discussions tookplace in which a proposal for jointmusic-making activities wereconsidered. Actors evaluated theimpact of each decision beforeventuring toward still greatercollaboration.

3. Broadcasting:Programming

*** Information about programming issueswas shared with a wide range ofstakeholders: the commissioningcontroller of Radio 3, orchestraadministration, concert promoters, andconductors. The senior producer set upseparate meetings with each party as hebuilt up the program. His vision guideddiscussions with the principal conductorand guest conductors, who proposedpossible repertoire, and with local concertpromoters and the orchestra’s maincommercial recording company, whoexplained their repertoire preferences.

***Leaders and stakeholderstogether came to anunderstanding of the orchestra’sprogramming needs andopportunities. This incorporatedthe BBC’s demands forcontemporary music, as well asthe repertoire suggestions ofguest conductors and soloists,and the need to include some“popular” pieces for theorchestra’s public performances.

***First an outline of major projectswas developed, but the programcontinued to be modified until ayear before it was played. Thesenior producer described theprogram as “constantly evolving”:it included a wide variety ofrecordings, concerts, tours, andactivities, all consistent with thevision of a national orchestra thatspecialized in contemporaryrepertoire and contributed to thecommunity.

4. Broadcasting:Principalconductorrenewal

*** In the year leading up to the contractrenewal date, discussions about theprincipal conductor took place regularly,as leaders considered their view of hisperformance and fit, and sounded out theopinions of others. Leaders evaluated theconductor’s performance and discussedreappointment with administration teamand BBC overseers. In regularmanagement team meetings, leadersdiscussed how the conductor’s recentconcerts went. The senior producergathered musicians’ views on theconductor.

*** Over time, a shared accountdeveloped of the existingprincipal conductor as a goodcandidate for reappointment.Musicians, managers and otherstakeholders perceived him asmusically powerful, andsomeone who related well tothose with whom he worked. Hewas widely understood to be thebest choice for the time.

*The orchestra offered the principalconductor a reappointmentcontract for another three years.

2005 33Maitlis

TABLE 3 Continued


5. Broadcasting:Players’ pay

*** Contract negotiations occurred jointlywith other BBC orchestras: the processinvolved regular meetings betweenBroadcasting leadership and other parties:the musicians’ union, BBC seniormanagement, and musician representatives.The content of unfolding negotiations wasreported back to musicians by unionrepresentatives, with much discussionamong musicians. Broadcasting’s leadersaddressed its musicians several times inorder to articulate their position and expresstheir strong desire for a resolution. Leadersasked the Orchestra Committee to design/distribute a confidential questionnaire togather musician views.

*** A unitary account emergedslowly, as the multipartynegotiations and relateddiscussions reached convergence.Musicians and leaders agreed thatcertain changes to the contractwould be necessary; negotiationfocused on the appropriate valueof various rights and conditions.

*** Elements of the new contractwere implemented over time. Forinstance, an agreement was reachedto increase the BBC’s acquisition ofrecording rights in exchange for asalary increase. Subsequent contractmodifications were implementedover the following year.

6. London:Programming

*** Information about programming wasshared among a wide range of stakeholders:recording companies, the orchestra’sresident concert hall, musicians, orchestraadministration, concert promoters, andconductors. The executive director arrangedongoing individual in-person and telephonemeetings to hear stakeholders’ demands andrequests. In these private meetings,stakeholders expressed their positions andinterests, and the executive director pressedfor repertoire he believed would be good forthe orchestra. The executive directordescribed the process as “juggling.”

*** A shared understanding of theorchestra’s programming needsevolved as the season plan wascomposed. Included was the ideaof the orchestra as performinghigh-profile concerts with bigname conductors and soloists, andas having good recordingopportunities, as well as goodplayer income generationopportunities.

*** The program plan served as theseason’s framework, and additionsand adaptations were made to itthroughout the year. These changeswere not predictable, but leadersworked to ensure that they wereconsistent with the orchestra’sartistic and financial goals.

7. London:Principalconductorappointment

*** The executive director sought out theprospective principal conductor and hisagent to express the orchestra’s interest andhear musicians’ feelings about the positionand conditions. This process continued overseveral months through parallel discussionsbetween the conductor and variousstakeholders. The executive directororganized a variety of private meetings withthe Arts Council, sponsors, and otherimportant groups; here, these partiesindicated their preferences and concernsabout the appointment. As discussionsunfolded, the chairman kept Boardmembers, musicians, and other stakeholdersinformed.

*** Parties gradually developed aconstruction of the proposedindividual as a strong candidatefor the position. A significantfigure in the international musicscene, he was both a great artistand able to bring the orchestravaluable contacts and reputation.All parties were aware of hisstrengths, but also of concernsrelating to the financial side of thecontract.

** The conductor was invited toconduct several engagements withthe orchestra. Over time, as thecontract was finalized andeventually agreed on, he acceptedthe position of principal conductor.

8. London: Costcutting

*** Cost cutting was discussed in regularmeetings of musicians and of the Board.These issues were also discussed in privatemanagement committee meetings withleaders; the purpose was to solicit the“objective view” of business people. Publiccost-cutting documents were made availableto broader orchestra stakeholder groups.Musicians engaged in a private ballot toindicate their preferences on extras’ pay cut.

*** Leaders and stakeholdersdeveloped a shared, richconstruction of the orchestra’sfinancial position: cost control hadalways been an issue andremained so. All partiesappreciated that there werenumerous ways to reducespending.

*** Many cost-cutting measures hadbeen carried out prior to the studyperiod, but these continued to beimplemented periodically, eachconsistent with the orchestra’soverall goal of minimizingunnecessary expenditure.


Guided Organizational Sensemaking

The first form of organizational sensemakingcomprised processes that were both highly con-trolled and highly animated. This type of sense-making was found in 9 of the 27 issue domains andoccurred in each of the three orchestras at leasttwice (see Table 3). In each instance, leaders werevery active as “sensegivers,” constructing and pro-moting understandings and explanations of eventsand of the process. At the same time, stakeholderswere also actively engaged in sensegiving, attempt-ing to shape beliefs about certain elements of theissues and their significance. These sensemakingprocesses were guided in that leaders, through thesystematic and confidential approach they broughtto their stakeholder interactions, were able to gather,coordinate, and shape stakeholder contributions.

Processes of guided sensemaking. The princi-pal conductor appointment in the London orches-tra offers a clear example of how guided organiza-tional sensemaking unfolds. Leaders carried outextensive sensegiving activities in private meetingswith different stakeholder groups, organized sys-tematically, and typically planned in advance. Atthe same time, the process was highly animated asa result of stakeholders who actively engaged insensegiving, pressing to take part in discussionsabout the issue and its resolution. Because of thelevel of stakeholder engagement, knowledge of theissue was widespread and discussed over an ex-tended period. The London leaders were keen toappoint a certain individual as principal conductorand worked hard, over several months, to convey to

different parties that the appointment could onlyhappen if the orchestra received sufficient fundingto support his salary. In a meeting with the Lon-don’s trustees, the chairman emphasized that “theorchestra could only proceed with this appoint-ment if adequate subsidy was available,” and onanother occasion, the executive director explainedto the board of directors that he would be meetingwith the Arts Council music director “to determinethe level of subsidy required.” In parallel, the ex-ecutive director explained privately to the cur-rently U.S.-based prospective principal conductorthat “the orchestra could not afford” to pay the kindof salary common in the USA. While leadersworked to shape stakeholder understandings,stakeholders constructed their own accounts of thesituation, speaking openly and with conviction.The Arts Council explained that “its support wasdirected towards the Orchestra’s overall program ofwork rather than to any one conductor” and arguedthat this appointment did not “merit special finan-cial assistance.” Directors in a board meetingclaimed that “orchestras in Britain go wrong whenthey enter into financial commitments that theycan’t afford.” In the face of these vehement argu-ments regarding the appointment’s financial viabil-ity, musicians spoke out regarding the conductor’sartistic strength. Overall, both the continuousrhythm and free information flow of highly ani-mated sensemaking and the systematic, organizedapproach of a highly controlled process character-ized this guided form of sensemaking.

TABLE 3 Continued


9. London:Incomegeneration

*** Discussions concerning incomegeneration schemes and projects continuedthroughout the study period. The executivedirector engaged a wide variety ofstakeholders (private sponsors, corporatesponsors, concert promoters, festivalorganizers, public granting agencies) inprivate one-on-one meetings. Each was oftenunaware of what the others were proposing.Leaders were proud of their successes inincome generation and took opportunities totell the Board, musicians, and externalstakeholders about new engagements,sponsors, and grants.

*** A shared, rich account of theorchestra’s needs and activitiesemerged that encompassed manystakeholders’ understandings ofhow income could be made.Fundamental to leaders’ and allstakeholders’ constructions of theLondon orchestra was the beliefthat its success rested onsuccessful income generationactivities.

*** Income generation activitiesinvolved distinct, discrete projectsthat were led by the executivedirector and moved gradually in aconsistent direction. These activitiescreated a coherent portfolio ofrecording and performance work thatgenerated sufficient income for theorchestra’s continued survival.

a The asterisks show the strength of evidence.*** Strong—Strong evidence for both dimensions of the process or outcome in this domain.** Moderate—Moderate evidence for both dimensions of the process or outcome, or strong evidence for one dimension in this domain.* Weak—Moderate evidence for one dimension of the process or outcome in this domain.

2005 35Maitlis

Outcomes of guided organizational sensemak-ing. Distinctive kinds of accounts and actions wereassociated with each form of sensemaking. Guidedsensemaking led to the production of accounts thatwere unitary and rich. Unitary accounts resultedfrom the systematic and controlled approach lead-ers adopted to give and gather constructions of theissue, which allowed them to guide the numerousand varied views of multiple stakeholders into anew, single, collective account. These accountswere rich because they incorporated the construc-tions of the many different parties engaged insensegiving: the new accounts generated throughthese processes tended to be cumulative and dense,as leaders built upon, rather than tore down, theaccounts and partial accounts generated by stake-holders along the way. While the issue of incomegeneration in the provincial orchestra, for instance,involved a wide range of internal and externalstakeholders, the executive director worked to pro-duce a single, coherent understanding of the or-chestra’s financial needs and of funders’ potentialroles in meeting those needs. He did so by drawingon stakeholders’ various accounts to construct arich, unitary understanding that took these multi-ple perspectives into consideration. The nature ofthe accounts arising from guided organizationalsensemaking processes is summarized in the fol-lowing proposition:

Proposition 1a. Organizational sensemaking inwhich both leaders and stakeholders are en-gaged in high levels of sensegiving will tend toproduce a unitary account (rather than multi-ple accounts) that is rich (rather than narrow).

An important effect of the production of accountsin sensemaking processes is that they facilitate ac-tion on the part of those involved (Weick, 1993,1995). An interesting finding in this study concernsthe impact of the type of account produced on thenature of subsequent actions. In guided sensemak-ing processes, the unitary, rich accounts enabledthe emergence of series of actions with consistentfoci. These accounts provided a common founda-tion for action, so that people’s activities werebased on accounts that captured a common under-standing of the situation and their goals. The pres-ence of multiple perspectives created accounts thatoffered a rich and enduring resource for the embry-onic emergence of a series of actions over time,while their unitary nature ensured a consistency inthe actions that emerged. For example, in the pro-vincial orchestra’s collaborative venture, the twoorchestras undertook a series of steps that cumula-tively contributed to the development of a partner-ship arrangement. Initially, the administration

team of a local chamber orchestra took up accom-modation in the provincial orchestra’s offices, andthen, over a period of months, the organizationsembarked on various joint fund-raising, marketing,and education initiatives. These activities werebased on a shared account of their competitivecontext and the value of collaboration, which facil-itated the consistency of their actions over time.The following proposition summarizes the natureof the actions associated with guided organizationalsensemaking.

Proposition 1b. Organizational sensemaking inwhich both leaders and stakeholders engage inhigh levels of sensegiving will tend to lead toan emergent series of actions (rather than aone-time action), that is internally consistent(rather than inconsistent).

Fragmented Organizational Sensemaking

The second form of organizational sensemakingresulted when processes were animated but notcontrolled. This form characterized sensemaking in7 of the 27 issue domains, occurring several timesin both the provincial and broadcasting orchestras(see Table 4). Stakeholders animated these frag-mented sensemaking processes by raising issues,generating and shaping accounts of the situations,and arguing for potential solutions. Leaders oftensought the views of stakeholders, but they did notattempt to organize or control discussions, nor didthey typically integrate stakeholder constructionsinto coherent collective accounts. This combina-tion of high stakeholder sensegiving and lowleader sensegiving produced organizational sense-making processes that were highly animated butuncontrolled.

Processes of fragmented sensemaking. A strik-ing example of fragmented sensemaking occurredin programming in the provincial orchestra. Theexecutive director described the plethora of partiesthat engaged in sensegiving on an ongoing basis:

[The principal conductor] sort of produced reper-toire by sparking off [the artistic advisor, the deputychief executive, and an external consultant]. . . .Emerging out of that process . . . is ultimately [theartistic advisor] producing rolling drafts of the rep-ertoire that responds to a strategy that was not wellarticulated, and now, we still don’t have a properlyarticulated artistic strategy and I’d like to see [thenew artistic director] looking at the one that [themarketing director] wrote when she arrived, and tryand get that fleshed out better.

In contrast to this abundant stakeholder sense-giving, sensegiving from the provincial orchestra’sleaders was minimal; these leaders neither pro-


TABLE 4Fragmented Organizational Sensemakinga

Orchestra andIssue Domain Process: Animated and Uncontrolled Account: Multiple and Narrow Actions: Emergent and Inconsistent

10. Provincial:Programming

*** Numerous multiparty meetings wereheld in which a variety of stakeholders,many times uninvited, put forward theirunderstandings of the issue andproposed ideas to shape its direction.Stakeholders outside the orchestra, suchas the Arts Council, criticized its poorartistic and financial performance,which they partly attributed to itsprogramming.

*** Leaders and stakeholdersstruggled to construct a unifiedaccount of the orchestra’smusical identity. Failing to dothis, they worked with numerouspartial and often conflictingaccounts of who the orchestrawas and the repertoire in whichit specialized.

*** Throughout the periodprogramming plans and proposalswere made and then changed.Different parties expressed confusionand frustration at the unclear andshifting plans.

11. Provincial:Principalconductorrenewal

*** Numerous ongoing discussions tookplace between leaders and the Board,among players, and between leaders andthe Arts Council about the renewal andpossible options. Stakeholders engagedin open exchange with each other, andwith leaders, about the strengths andweaknesses of the existing conductor.Leaders tried to direct or restrict thesediscussions. Discussions continued formore than a year, intensifying at certainpoints as action was required.

*** Accounts of the issue shiftedfrom one position to another, asleaders and stakeholders soughtresolution. Throughout the life ofthe issue, there remained a lackof agreement regarding the bestway to understand the situationand the “right” way to deal withit.

** As the contract deadline drewnear, the chairman met with theconductor to ascertain his wishes.The conductor was unwilling tocommit without hearing the views ofthe musicians. The chairman thenoffered him a short renewal whilethey tried to decide what to do.

12. Provincial:Guestconductors

*** The quality and style of differentconductors, as well as plans fordeveloping a “family” of guestconductors, were openly discussed in avariety of different meetings attended bythe leaders, Board, artistic advisorycommittee, and consultative committee,and at the Board retreat. Musiciansdiscussed the suitability of guestconductors in orchestra meetings.Musicians and board members putforward views about members of thefamily of conductors for more than ayear, as the leaders tried to specify itscomposition.

*** Multiple views of the issueand approaches to it prevailedthroughout the study period.While one account suggested theneed to develop a family ofconductors, the principalconductor, musicians, and otherstakeholders made suggestionsfor possible visiting conductorsthat were inconsistent with thisidea.

*** The artistic director reduced thenumber of different guest conductors,in discussion with the principalconductor. At the same time,commitments to conductors who hadalready been booked weremaintained. It was hard to detect acoherent strategy behind theconductor roster.

13. Provincial:Players’ pay

*** The leaders’ plan to introduce amore flexible players’ contract waswidely discussed among musicians,board members, and the musicians’union. This proposal was discussed inmeetings with management and unionrepresentatives, and was also raised infull orchestra meetings and Boardmeetings. Discussions continued overseveral months. Players, their chairman,and their Board representativesexpressed their views vehemently toleaders. The executive director listenedto stakeholders but was not forthcomingin his own view. The intensity ofdiscussions varied, but were ongoing,especially between musicians and theartistic director.

*** While there was a sharedunderstanding of the financialdifficulties, much disagreementand confusion remainedthroughout about how to addressthe problem. The executivedirector first argued for acontractual change, and thenlater for a pay freeze, pointing tothe principal conductor’s andmanagement team’s salaries andexpenses.

** Following his initial proposal forcontractual change, the executivedirector offered a pay freeze. Playersand others expressed surprise at thischange of direction and voted againstthe freeze. Despite this, the freezewas imposed as a temporarymeasure.

2005 37Maitlis

vided an overarching artistic vision, nor articulateda systematic process through which a long-termplan might be developed.

Outcomes of fragmented sensemaking. Sense-making processes that were both highly animatedand highly uncontrolled tended to produce multi-ple individualistic accounts. A high level of anima-tion ensured the discursive production of a varietyof stakeholder perspectives, but because control

was absent, these accounts were rarely integratedor reconciled. Stakeholders’ accounts thus tendedto accumulate over time: fragmented sensemakingprocesses remained active over the lives of issuesas stakeholders put forth more, diverse accounts ofsituations, accounts that no one worked to inte-grate. While, in principle, the integration of ac-counts might have been carried out not only by theleaders but also by key stakeholders or through a

TABLE 4 Continued

Orchestra andIssue Domain Process: Animated and Uncontrolled Account: Multiple and Narrow Actions: Emergent and Inconsistent

14. Provincial:Cost cutting

*** The need to cut costs was a salientissue for two years. Leaders andstakeholders discussed numerousapproaches, serially and in parallel. Forexample, the executive director’s mainidea—reducing the orchestra’s size—waswidely discussed by players, Boardmembers, the artistic director, the orchestramanager, and the Arts Council. Sectionprincipals met to appraise the plan.Musicians, outraged by the proposal, metseveral times to discuss an action plan andsubsequently presented alternativeproposals to the Board. The leaders madefew efforts to curtail these exchanges, nordid they attempt to shape the musicians’opinions.

*** Although the leaders wereconsistent in their account of theneed to cut costs, they did notpresent a coherent plan. Differentparties’ proposals were offered upand discussed, each groupmaintaining its construction of thesources of the problem andworkable solutions.

*** Different proposals wereadvocated and withdrawn inreaction to others’ responses. Overtime, a variety of cost-cuttingmeasures were introduced, but thesefailed to fit clearly into an overallbusiness plan.

15. Broadcasting:Playerperformance

** The concert master’s performance andfuture employment were discussed widelyover a period of several months, first amongleaders and the management team, and thenmore openly and with greater intensity andurgency as the deadline for contract renewalapproached. Section principals met withmanagement to express their feelings. Theprincipal conductor shared his opinion inmeetings with leaders. The Players’Committee discussed the issue amongthemselves and forwarded their views toleaders. Leaders gathered views buthesitated to present their own positions.

*** A wide variety of oftenconflicting accounts of thesituation circulated amongstakeholders. These accountsvariously addressed the concertmaster’s performance ability,leadership qualities, commitmentto the orchestra, the financial costsof the contract, the legalobligations, and alternativeindividuals for the position.

** Players told the executive directorthat they believed the concertmaster’s contract should not berenewed. After the leaders informedthe concert master, these samemusicians expressed their sorrowand regret at the decision to her.

16. Broadcasting:Playerappointments

*** A number of different musicians cameto try out for the position of second concertmaster, and their qualities were discussedfor several months. Players openlyexpressed their views of different candidatesin musician meetings and to leaders. Overtime, musicians became increasinglyimpatient and demanded to hear leaders’views and to see some action.

*** Disparate, conflicting accountswere generated regarding theposition and the candidates. Someaccounts suggested the musicianspreferred candidate was notinterested; others suggested he wasbut leaders had been clear enough inexpressing interest. Differingaccounts also emerged regardingwhether the position regardingwhether the position was for a“second” or a “first” concert master.

** As the leaders decided not torenew the contract of their existingconcert master, a “first” concertmaster position became availableand was offered to the preferredcandidate.

a The asterisks show the strength of evidence.*** Strong—Strong evidence for both dimensions of the process or outcome in the domain.** Moderate—Moderate evidence for both dimensions of the process or outcome, or strong evidence for one dimension in the domain.* Weak—Moderate evidence for one dimension of the process or outcome in the domain.


collective stakeholder effort, such stakeholder-driven integration was not evident in the presentstudy. Instead, these fragmented processes pro-duced multiple new accounts, each of whichtended to be quite narrow, representing the con-struction of a single individual or group. In the caseof the provincial’s players’ pay and contract issue,various members of the board, the orchestra com-mittee, the administration, and the musicians’union representative each put forward quite differ-ent and often contradictory arguments about thesituation and what should be done. At the end ofthis process, these accounts had failed to coalesce.The nature of the accounts produced through frag-mented organizational sensemaking processesleads to the following proposition:

Proposition 2a. Oganizational sensemaking inwhich leaders are engaged in low levels ofsensegiving and stakeholders are engaged inhigh levels of sensegiving will tend to producemultiple (rather than unitary) accounts thatare narrow (rather than rich).

The proliferation of individualistic accounts as-sociated with fragmented sensemaking had a sig-nificant impact on the forms of action produced.The existence of multiple, distinctive accountstended to generate series of inconsistent and con-tradictory actions. This was because the variousaccounts of the situation that were available toactors often differed so significantly that, over thelife of these issues, different groups and individu-als engaged in divergent actions based on their ownidiosyncratic interpretations of events. For exam-ple, in the provincial orchestra’s pay and contractissue discussed above, the leader first decided tointroduce a contractual change concerning thenumber of hours that musicians worked. In the faceof negative reaction from musicians and somemembers of his management team and board, heinstead offered the orchestra a pay freeze. Amidstsurprise and confusion at the unexpected change ofplan, the musicians, led by a union representative,rejected the pay freeze proposal. The executive di-rector ignored this rejection, imposing a temporarypay freeze without discussing it further. Musicians,although disgruntled, continued to work. The un-reconciled multiple accounts constructed in frag-mented processes resulted in actions that wererarely in harmony.

Fragmented sensemaking, where animated anduncontrolled processes lead to individualistic ac-counts and inconsistent actions, resembles “col-lapse of sensemaking” (Weick, 1993), in which ac-tors experience the disintegration of sharedmeaning. In both cases, we see individuals going

their own way to the detriment of collective sense-making processes and action. Weick associated thecollapse of sensemaking with crises that throw ac-tors into unfamiliar roles and discredit an existingrole system. The current study suggests, however,that less extraordinary circumstances can preventcollective meaning making. In the fragmented pro-cesses seen here, there was no precipitating crisis:instead, simple lack of control in sensemaking pro-cesses led to situations in which multiple accountsof issues were propagated and led to inconsistentactions. Moreover, whereas Weick described therapid collapse of sensemaking in a small team, hereone sees collapse occurring over extended periodsof time and across large, diverse groups of stake-holders. The following summarizes the nature ofthe action associated with fragmented organiza-tional sensemaking:

Proposition 2b. Organizational sensemaking inwhich leaders engage in low levels of sense-giving and stakeholders engage in high levelsof sensegiving will tend to lead to an emergentseries of actions (rather than a one-time action)that is internally inconsistent (rather than con-sistent).

Restricted Organizational Sensemaking

The third form of organizational sensemaking,the restricted form, occurred when processes werehighly controlled but not very animated. This formtypified sensemaking in 7 of the 27 issue domains,and it was found several times in the broadcastingand London orchestras (see Table 5). In each ofthese cases, leaders who engaged in high levels ofsensegiving promulgated overarching accounts ofissues they encountered, which stakeholderstended to accept with relatively few attempts toprovide alternative understandings. Leaders did,however, identify certain stakeholders as valuableto the process and sought them out to draw on theirconstructions of the issues at key points—largelythrough the careful, advanced scheduling of privatemeetings with individual groups.

Processes of restricted sensemaking. One exam-ple of restricted sensemaking was seen in thebroadcasting orchestra’s income generation area. Inan internal report to the BBC, the executive directoridentified several ways of “increasing the orches-tra’s income-earning potential.” Both in this reportand in private stakeholder meetings, he describedpossible ways in which the orchestra could gener-ate more revenue. Although every one of these,from making changes to the players’ contract, in-troducing a strand of lighter music to the orches-

2005 39Maitlis

TABLE 5Restricted Organizational Sensemaking

Orchestra andIssue Domain Process: Unanimated and Controlled Accounts: Unitary and Narrow

Actions: One-Time or Planned andConsistent

17. Broadcasting:Guestconductors

*** Few stakeholders were involved indiscussions of the issue. Main informationexchanges took place between the seniorproducer and potential guest conductors.The senior producer initiated meetings inwhich potential guest conductorsindividually talked with him about theirabilities, schedules, and preferences. Thesenior producer also informally gatheredmusicians’ preferences. Over time, heintegrated the proposals gathered.

*** The account that emerged wasan unambiguous description of theguest conductor roster for thecoming period and the seniorproducer’s rationale for it: theroster included known individualsas well as new and upcomingconductors that the orchestra wasinterested in trying out.

*** Once the engagements werefinalized, the orchestra was informedof the roster and guest conductorscame as scheduled.

18. Broadcasting:Cost cutting

*** Information was contained largelywithin the management team. Discussionstook place in the management teamintermittently, as budget constraints wereannounced. Little input from stakeholders; afew proposals for cutbacks were relayedprivately in closed discussions with leaders.The executive director informed musiciansof key changes once decisions had beenmade.

*** Leaders developed an accountthat built on the existing BBC-widenarrative about the budget cuts andcost control. Their description ofthe need for cost cutting in theorchestra was clear andunderstood by all.

*** The executive director drew upa set of cost-cutting initiatives, whichhe implemented in succession.

19. Broadcasting:Incomegeneration

*** Leaders worked to keep issues in thisarena quiet; they saw it as having politicalconsequences and did not want otherorchestras to hear about their plans ordiscussions. A few key stakeholders—concert promoters and one recordingcompany—expressed their interests andideas in private talks with leaders.Musicians and other internal stakeholdersknew little about the issue. As plans weredeveloped, leaders periodically madeannouncements, both internally and to thegeneral public.

*** Compared with the otherorchestras, the Broadcasting’saccount of its income generationwas quite narrow, and understoodby most stakeholders: theorchestra’s primary source offunding remained the BBC.

*** A set of new income generationactivities was implemented as part ofthe Broadcasting’s overall mission;these were consistent with itsidentity.

20. Broadcasting:Collaborativeventures

*** Discussions around the issue took placesporadically, depending on the prevailingpolitical and financial climate and on theleadership of the potential collaborators.Broadcasting leaders consideredcollaboration extremely sensitive and sokept information on the issue and relateddiscussions confidential. Private meetingswere held with a few key stakeholders:leaders of the other orchestra, the BBC, andlocal community figures.

*** The account, largelydeveloped and propagated by theexecutive director, described theconcept of a “superorchestra,” butwith few details as to how itwould be created, managed, orfunded.

* Although not enacted during thestudy period, the executive director’svision was for overnight creation of asuperorchestra.

21. London:Guestconductors

*** The executive director talkedindividually with potential guest conductorsand agents about their abilities, schedules,and demands, integrating these discussionsinto his construction of the overall shape ofthe season. Few other stakeholders werefocused on the issue, although somesuggestions were put forward fromindividual musicians.

** A shared understanding of theguest conductor roster for thecoming period emerged. The reasonsbehind each engagement were notalways clear to each stakeholder, butthere was general acknowledgementthat the executive director hired thebest set of people possible, givenconstraints.

** The executive director madearrangements and informed theorchestra periodically as they wereconfirmed. The roster was consistentwith the orchestra’s artistic/financialaims and desired image. Smallchanges occurred: cancellations andchanges in dates, rehearsals, andrepertoire.


tra’s repertoire, and repeating the same program ina number of different venues, would significantlyaffect the musicians, the director did not seek outplayers’ opinions. Indeed, he largely considereddoing so a waste of time, explaining: “I’m trying tosave the orchestra, and all they can talk about iswhether they get a 20-minute tea-break.” Musicianswere no more interested in discussing these finan-cial matters than the leader was in hearing theirviews. They did not talk about them much amongstthemselves, nor did they often come forward withtheir ideas or interpretations. A few key stakehold-ers, however—concert promoters and certain se-nior BBC executives—engaged in private meetingswith leaders through which they shaped the in-come generation issue and possible ways of ad-dressing it.

Outcomes of restricted organizational sense-making. From highly controlled and unanimatedsensemaking processes emerged accounts that weredistinct from those produced in guided and frag-mented sensemaking processes. As in the guidedprocesses, the controlled nature of restricted sense-making produced unitary accounts: a primaryproduct of sensemaking was a single, dominantinterpretation of the issue. In restricted sensemak-ing processes, however, this dominance resultednot from leaders working to integrate and synthe-size multiple perspectives, but from there being alack of alternatives to leaders’ constructions. Thus,

unlike the rich accounts produced in guided sense-making, the accounts emerging from restricted pro-cesses were relatively narrow, incorporating onlythe leaders’ perspectives. Although leaders maygenerally have broader understandings of some is-sues than do individual stakeholders (Hambrick &Mason, 1984), their perspectives cannot includethe variety of perspectives that exist across a rangeof interested stakeholders. In the case of the Lon-don orchestra’s players’ pay and contract issue, theleaders’ account of the situation was simple; theexecutive director told the orchestra, “We’ve doneall of this, we’re still not able to make the £150,000surplus we wanted; we’re doing this, and askingyou to do this.” He had chosen not to consultwidely, preferring to minimize discussion on thematter and believing there to be only one workablesolution. He knew that this account did not takeinto consideration each party’s interests, comment-ing in an interview, “We knew that it wasn’t goingto be unanimously approved,” but he defended it tothe musicians, board, and executive committee asthe only viable option. The nature of the accountsproduced through restricted sensemaking pro-cesses leads to the following:

Proposition 3a. Organizational sensemaking inwhich leaders engage in high levels of sense-giving and stakeholders engage in low levels ofsensegiving will tend to produce unitary

TABLE 5 Continued

Orchestra andIssue Domain Process: Unanimated and Controlled Accounts: Unitary and Narrow

Actions: One-Time or Planned andConsistent

22. London:Players’ pay

*** Leaders kept this sensitive issue quietuntil it was necessary to inform theorchestra at large. Discussions first tookplace only among leaders; the issue wasthen discussed with the Board in a series ofprivate, impromptu meetings. Discussionsaround the issue did not continue long: adecision was made within a few weeks.

*** Leaders and directorsconstructed an account that theyfelt was likely to be acceptable tothe majority of stakeholders: feesshould be reduced in a way thataffected the fewest musicians.

*** Pay was cut for a few keypositions, rather than across theentire orchestra; the pay cut wasimplemented and received withminimal resistance.

23. London:Collaborativeventure

*** In general, information on the issue andrelated discussions were kept confidential.Discussions between the executive directorand key stakeholders, such as thecollaborating orchestra and the collaboratingconcert hall, were held sporadically. AnArts Council representative was invited toattend a Board meeting to give herperspective. The venture plans progressedfitfully, when the executive director wasable to spend time on them.

*** The account of the situationwas dominated by the perspectivesof the executive director and theArts Council: the proposal had tobe innovative and clearly look likea collaboration.

*** A proposal for a collaborationbetween the London orchestra, thecollaborating orchestra, and theconcert hall was submitted to theArts Council.


2005 41Maitlis

(rather than multiple) accounts that are nar-row (rather than rich).

Typically comprising just single interpretationsof issues and single means of resolving them, thesenarrow, unitary accounts tended to lead to one-time actions to deal with the issues. On other oc-casions, when the issues could not be addressed bysingle actions, the accounts spawned planned setsof consistent actions intended to together tackle theissues. The narrowness of the accounts produced inrestricted sensemaking processes meant that theyprovided highly specific foundations for action:much like a play written around a linear plot thatcan lead to only one conclusion, an account pro-duced by restricted sensemaking processes doesnot provide the discursive resources necessary fororganization members to engage in any significantimprovisation or extension. For example, in thebroadcasting orchestra’s cost-cutting arena, the ex-ecutive director identified and prioritized a set ofrelated areas in which cutbacks could be made.Once this plan was laid out, leaders worked toensure each step was effectively enacted; neitherleaders nor stakeholders were driven to identifyalternative cost reduction models. These outcomesof restricted sensemaking processes lead to thefollowing:

Proposition 3b. Organizational sensemaking inwhich leaders engage in high levels of sense-giving and stakeholders engage in low levels ofsensegiving will tend to lead to a one-timeaction or a planned set of consistent actions(rather than an emergent series).

Minimal Organizational Sensemaking

The last form of organizational sensemaking ob-served in this study resulted from processes thatwere neither animated nor controlled. This mini-mal sensemaking was found in 4 of the 27 issuedomains, occurring twice each in the provincialand the London orchestras (see Table 6). In theseminimal sensemaking processes, each party tendedto await others’ interpretations of an issue, whichtypically came in response to some external trigger.Animation was low, with few stakeholders discuss-ing the issue or seeking to offer their constructionsof it. At the same time, leaders made little attemptto organize ways of promoting their interpretationsof it or to gather the views of their stakeholdergroups in any systematic way.

Processes of minimal sensemaking. In the areaof the London orchestra’s player appointments, forexample, leaders and orchestra members lookedprimarily to the relevant section principal for an

opinion about a job candidate. Few other stake-holders volunteered a view, and when other sec-tion principals were asked for theirs, they rarelyput forward alternative perspectives. As one boardmember explained, “Most of the principals aremore than happy to sit on the fence—they’ve got ahard enough job—they don’t want to put their oarin and stir things up.” Stakeholders did not discussthe issue widely, and months passed when no onepressed to address it. At the same time, leadersmade little attempt to shape understandings of anappropriate appointment or to influence how oth-ers saw a particular candidate. Nor did they try toorganize systematic ways of drawing out stake-holder opinion, or make any clear effort to encour-age leader-stakeholder discussion of the matter.Overall, this minimal form was characterized bylow levels of animation and discussion and by verylittle attempt to drive or organize the sensemakingprocess.

Outcomes of minimal sensemaking. Instancesof minimal sensemaking led to accounts and ac-tions that reflected the low levels of animation andcontrol associated with the process. As stakehold-ers failed to offer spontaneous constructions of is-sues, and as leaders neither encouraged them to doso nor put forward their own interpretations, sense-making in the minimal form produced only nomi-nal accounts of issues. Nominal accounts—onesthat provided only token understanding or inter-pretation—emerged as leaders and stakeholdersgrasped at any construction of issues that mightprovide some basis for resolution. Provincial or-chestra leaders and stakeholders, for example,struggled with the issue of a musician whose per-formance was in question; rather than actively en-gage in sensegiving with respect to the issue, how-ever, all parties avoided dealing with it. Aftermonths of uncertainty and procrastination, the or-chestra’s leaders suggested a “flexible arrange-ment” that would avoid terminating the player’semployment:

Maybe the right solution for him [the musician] is todo a much smaller percentage of the work, not severhis links entirely with the [provincial orchestra], butbe somebody who is brought in on a different con-tract. . . . Maybe that’s a solution.

I describe this interpretation of the situation as anominal account because it neither synthesizes theperspectives of multiple stakeholders, as do therich accounts described in the guided form of sense-making, nor does it articulate a well-developed in-terpretation based on a single perspective, as wasseen in the restricted form. The following proposi-


tion summarizes the nature of the new accountsarising from minimal organizational sensemaking:

Proposition 4a. Organizational sensemaking inwhich both leaders and stakeholders engage inlow levels of sensegiving will tend to producenominal accounts.

The nominal accounts produced in these pro-cesses provided very weak foundations for action.Accounts act as discursive resources for membersas they attempt to respond to an issue (Weick,

1993): a nominal account will thus do little to fostereither motivation (reasons for action) or imagina-tion (ways forward). Nevertheless, the nominal ac-counts produced in minimal sensemaking pro-cesses, while lacking richness and failing toincorporate multiple perspectives, seemed to freeleaders and stakeholders from their paralysis byallowing single compromise actions that providedtemporary relief. Issues that were associated withminimal sensemaking were not “nonissues”: theywere nonurgent matters of some concern for which

TABLE 6Minimal Organizational Sensemaking

Orchestra andIssue Domain Process: Unanimated and Uncontrolled Account: Nominal Action: One-Time Compromise

24. Provincial:Playerperformance

*** The principal conductor was unhappywith the performance of a certain player, butthis was not publicity discussed. Very littleinformation was exchanged about the issuefor several months. When the issue becameknown, musicians forcefully expressed theirviews about the ability of their colleague invarious meetings, and informal discussions.

*** For a long time, no accountexisted of the situation or ofappropriate ways of dealing withit. Over time, a tentative accountboth of the player’s ability and of aworkable solution emerged.

*** The executive director proposeda compromise solution: the playershould not be reauditioned or firedbut asked to do less work with theorchestra.

25. Provincial:Playerappointments

*** Despite key positions remaining open,little information was exchanged regardingplayer appointments. Although musicianswere brought in on trial, for long periodsthere was little discussion of who should beoffered a job. When the issue was discussed,it was by Board members and players, whobrought it up in larger meetings anddemanded some explanation and action.

*** Accounts of the issue wereoccasionally exchanged, but noone account stuck. There was littleagreement about the urgency of theneed to hire or about the bestcandidates for the positions.

** When it was understood thatthere was a threat of losing acandidate for one key position to acompetitor, the artistic directorcompelled the principal conductor toagree to make the appointment.

26. London:Playerperformance

*** Although some musicians expresseddiscomfort with one player’s performance,there was little open discussion about it forseveral months. People avoided talkingabout the issue because it involvedcolleagues and friends. When repeatedperformance problems made the issue verypressing, it was raised between the sectionprincipal and the chairman, and thendiscussed in Board meetings. Sectionprincipals also met to discuss the player’sperformance.

*** As constructions of the issueemerged, they differedconsiderably, depending onpeople’s positions and relationshipwith the target musician. Therewas no shared account of themusician’s problem or of the bestway to approach it.

*** After much delay and hesitation,the player was approached andasked to take some time off. Thiscompromise bought time for allparties.

27. London:Playerappointments

*** Sensemaking lay largely with thesection principal. Although musicians camein on trial, long periods passed with littlediscussion of whom to appoint. After a longtime and several trials, the section principalidentified a candidate for appointment andgathered views from the section and sectionprincipals. At a meeting between sectionprincipals and the Board, a case was madeto appoint the musician.

** When an account emergedabout the candidate, it was quitenarrow in its description and putforward by just a few interestedindividuals.

** After a period of years, theappointment was made when thesection principal deemed acandidate good for the position. Thisview was endorsed by fellowprincipals.


2005 43Maitlis

both leaders and stakeholders failed to engage inactive sensegiving. Although these issues oftenwent untackled for extended periods of time, theyultimately required some, at least temporary, reso-lution. With the provincial’s player performanceissue, a decision was made that was acceptable forthe time being but did not provide a workable long-term solution. Decreasing the musician’s amount oftime with the orchestra temporarily satisfied theprincipal conductor and his concerns for the or-chestra’s performance standard, but doing this alsomeant that the issue of the player’s performancewould not completely disappear. The followingproposition summarizes the nature of actions gen-erated through minimal organizational sensemak-ing processes:

Proposition 4b. Organizational sensemaking inwhich both leaders and stakeholders engage inlow levels of sensegiving will tend to lead toone-time, compromise actions.

CONCLUSIONS

Two research questions guided this study: (1)Are there discernable patterns of interaction thatconstitute the social processes of sensemakingamong diverse stakeholders in complex organiza-tions? and (2) Are there patterns of accounts andaction associated with the different social pro-cesses of sensemaking in organizations? I haveshown that different combinations of leader andstakeholder sensegiving produce four differentforms of organizational sensemaking—guided,fragmented, restricted, and minimal—each withdistinctive processual characteristics. I have alsoshown that each form produces a distinct set ofoutcomes in the kinds of accounts and actionsgenerated.

This study makes three important contributionsto the sensemaking literature. First, it identifies twokey dimensions that describe the social processesof organizational sensemaking—animation andcontrol. These dimensions provide an importantand missing element in sensemaking research: alanguage with which a variety of everyday sense-making processes can be described, compared, andcontrasted. Together, the concepts of animationand control can describe the different ways inwhich heterogeneous parties interact in ongoing,“ordinary” sensemaking processes over extendedperiods of time. This study thus complementsthose that have articulated dimensions of sense-making that, though rich and evocative, have beenless amenable to use in a wide range of sensemak-ing contexts. Second, this study identifies four dis-

tinct forms of organizational sensemaking that cap-ture the variance that exists when sensemakingoccurs among a pool of diverse actors addressing arange of organizational issues. This contribution isimportant for sensemaking research because itdemonstrates that sensemaking is neither a singu-lar, homogeneous process, nor a random, heteroge-neous set of processes. Moreover, these four formsconstitute an empirically grounded typology ofsensemaking processes upon which future researchcan be based. Third, this study contributes to thesensemaking literature by showing how these fourforms of sensemaking processes connect to differ-ent types of accounts and actions. For each form, Ihave developed propositions that relate the socialprocesses of sensemaking to specific types of ac-counts and action. These propositions go beyondprevious sensemaking research that has shown thatsensemaking is key to organizational accounts andaction but failed to examine heterogeneous forms ofsensemaking and to specify their relationships todifferent outcomes.

The study described in this article of course hasits limitations. One consideration is that it wasconducted in symphony orchestras. Orchestrashave goals, structures, and stakeholder roles simi-lar to those of many medium-sized private compa-nies, which is likely to mean that orchestras’ sense-making processes will be comparable to those insuch organizations. At the same time, orchestrasclearly also differ from private companies in thespecific products and services they provide and theenvironment in which they operate. Although it isunlikely that sensemaking processes in orchestrasdiffer fundamentally from those in other kinds oforganizations, the generalizability of the theorypresented here therefore remains to be tested.

It is also possible that the four forms of organiza-tional sensemaking identified here may not be ex-haustive: the forms could be supplemented by thediscovery of subtypes, perhaps through data col-lected from a larger and more heterogeneous set oforganizations. For the current study, however, Iargue that it would have been working beyond thepoint of diminishing analytic returns to seek suchsubtypes.

A further limitation here concerns the simplebifurcation of leader sensegiving and stakeholdersensegiving into predominantly high or low withineach issue domain. Because of the extensive dataset analyzed, it was necessary to synthesize anenormous amount of material to develop a coherentpicture of organizational sensemaking across issuedomains and organizations. Although previous re-search has shown that a party’s sensegiving behav-ior can vary at different points in a sensemaking


process (Bartunek et al., 1999; Gioia & Chittipeddi,1991), it lay outside the scope of this study toaddress the issue at this level of analysis. Despitethese limitations, this study has significant impli-cations for both research and practice.

Implications for Research

This study raises important questions for futureresearch on organizational sensemaking. The firstconcerns the antecedents of organizational sense-making, in particular, the role of the particularissue and organization involved in determining theform (the distinct pattern of leader-stakeholder in-teraction) of sensemaking enacted. Although in thisstudy neither issue nor organization alone deter-mined the form of sensemaking, each appeared tohave some impact. For instance, looking at the pat-tern of sensemaking across issues reveals that inmost cases two of the studied organizations shareda common form for a given issue. This patternsuggests that issue may affect the degree of sense-giving both leaders and stakeholders engage in andconsequently may affect the ensuing form of organ-izational sensemaking. With regard to the impact oforganization on sensemaking processes, we can seethat each orchestra in this study demonstrated onlythree of the four possible forms. This finding sug-gests that there may be organizations in which lead-ers inevitably become involved in sensegivingaround any active issue, and others in which thereare dynamic sets of stakeholders who almost al-ways seek to contribute to the formation of ac-counts, irrespective of the issue. There may also beorganizations in which both leaders and stakehold-ers are reluctant to let issues lie, so that one or bothparties will inevitably pick them up and attempt toconstruct accounts about them. The small numberof organizations in the present study precludes usfrom drawing stronger conclusions about organiza-tional antecedents of sensemaking, but it neverthe-less highlights interesting questions for futureresearch.

Another question concerns the distribution of thefour sensemaking forms that might be likely to beseen in different kinds of organizations and indus-tries. I might speculate, for example, that in high-velocity industries, where efficient organizationalprocesses are critical (Eisenhardt & Bourgeois,1988), there would be more instances of restrictedand fewer instances of guided sensemaking. Simi-larly, relatively few cases of fragmented or minimalorganizational sensemaking would seem likely innew start-ups, where highly energetic entrepre-neurs dominate sensemaking processes. In order todevelop understanding of organizational sense-

making, researchers need to acknowledge the dif-ferent forms it can take and examine the contextsmost conducive to each one.

A third question raised by the present researchinvolves the stability of the organizational sense-making forms and the ways in which they relate toone another. Although these dynamics were not afocus of this study, an interesting area for futureresearch would be to examine patterns of change inthe forms over time. The restricted form, for exam-ple, may be self-perpetuating, as the containmentand concealment of information become routinizedand formalized, and stakeholders become increas-ingly silenced (Morrison & Milliken, 2000). In con-trast, the fragmented form may be inherently unsta-ble because such frenetic activity is sustainableonly for a time and produces little unified action.Connected to the issue of stability is the question ofhow different forms of organizational sensemakingand their outcomes relate to one another over time.For instance, faced with a temporary solution of thekind generated through minimal organizationalsensemaking, do leaders, stakeholders, or both in-crease their sensegiving, perhaps achieving a moredurable outcome? Similarly, is fragmented sense-making, in which the process may exhaust thoseinvolved, typically followed by minimal sensemak-ing, or are leaders driven by the uncertainty toengage in more vigorous sensegiving activities?

A fourth research question concerns the relation-ship between the forms of organizational sense-making and important performance-related out-comes. Although this study examined certainoutcomes, it would be valuable to consider whetherparticular forms of organizational sensemaking arelikely to produce organizational results such as in-novation, efficiency, or financial performance. Forexample, the guided form, in which leaders drawon and integrate stakeholders’ accounts, may bemore likely to produce innovative proposals thanthe restricted form of organizational sensemaking,which relies primarily on leader sensegiving. Therestricted form, however, may be tied to fast, effi-cient action, and so might be most appropriatewhen there is a pressing need to respond to anissue. Equally, the fragmented form, which in thisstudy led to very little consistent action, may gen-erate large numbers of valuable ideas. Althoughthese relationships are speculative, they suggestthat the forms of organizational sensemaking foundin this study may have important effects on widerorganizational performance outcomes.

This study also connects and contributes to anumber of scholarly literatures not usually associ-ated with sensemaking research. One of these isresearch on multiparty stakeholder negotiation. Al-

2005 45Maitlis

though prior sensemaking research has noted thepotential for conflict among participants (Volkemaet al., 1996), the tendency to focus on single stake-holders (Westley, 1990), relatively small, homoge-nous groups (Weick, 1993), and tightly coupledsystems (Weick & Roberts, 1993) has resulted inrelatively little examination of the connections thatexist between negotiation processes and organiza-tional sensemaking. This study illustrates their in-terpenetration: sensemaking in organizations de-mands the negotiation of interpretations andexplanations among diverse actors; at the sametime, sensemaking and sensegiving processes areintegral to multiparty negotiations. An importantcontribution that the negotiation literature canmake to sensemaking research is highlighting theimportance of coalitions when parties anticipate anongoing relationship with each other (Polzer, Man-nix, & Neale, 1998). These dynamics were evi-denced in guided and restricted processes in whichleaders established coalitions with stakeholders inorder to construct common interests and, therefore,unitary accounts. In contrast, the fragmented formof sensemaking represented a process in which co-alitions—where they did exist—were typically toounstable to permit the construction of any sharedunderstanding.

This study also highlights the potential for sense-making research to contribute to the study of mul-tiparty negotiations. For instance, the multipartynegotiation literature discusses the common occur-rence of resolutions that are unstable (because it isalways in someone’s interest to shift alliances and“re-form” a proposed solution) or that involve de-ceit (where participants attempt to conceal theirpreferences in order to cast a deciding vote) (Gray &Clyman, 2003). The study described here identifiestwo ways in which sensemaking processes maylead to a resolution of such situations. First, guidedsensemaking may facilitate the construction ofoverarching accounts that help to integrate stake-holders’ views, potentially avoiding both unstablesolutions and necessary deceit. Alternatively,through fragmented sensemaking, stakeholdersmay resolve an impasse associated with an unsta-ble resolution by engaging in actions based on theirtemporarily acquiring the power or authority to act.Both possibilities point to the importance of con-sidering sensemaking processes when examiningmultiparty negotiations.

Together, these connections highlight the poten-tial for the development of theory and research thatintegrate research on negotiation and sensemaking.Although sensemaking and negotiation are distinctconcepts, this study illustrates the ways in whichthe two are interconnected in empirical contexts:

the social processes of organizational sensemakingobserved in the orchestras often included leadersand stakeholders engaging in negotiations that de-termined how accounts were produced and thekinds of decisions that flowed from those accounts.A potentially powerful framework for understand-ing the generation of collective action in organiza-tions might therefore be based on the relationshipsamong sensemaking, negotiation, and decisionmaking.

Another area connected to the present study isthe literature on procedural justice and deferenceto authority. A central finding in this area is thatpeople are more willing to accept explanations anddecisions when they perceive the processesthrough which they were reached as fair (Pruitt,Peirce, McGillicuddy, Welton, & Castrianno, 1993;Tyler, 2002). This idea is consistent with the out-comes of guided sensemaking processes: the uni-tary accounts and consistent actions guided sense-making produced were likely due at least in part tostakeholders’ perceptions that leaders were behav-ing in a procedurally fair manner, gathering andtaking into account many parties’ views about is-sues. This view does not explain, however, theunitary accounts and consistent actions associatedwith restricted sensemaking processes in whichleaders dominated sensegiving and often intention-ally excluded stakeholders. In these processes, itseems that stakeholders showed deference to lead-ers’ authority (Tyler, 2002; Tyler & Blader, 2000),perceiving leaders as legitimate and therefore buy-ing into their constructions of the issue and follow-ing their guidance. Thus, research on proceduraljustice and deference to authority may be useful inunderstanding sensemaking processes, and espe-cially their antecedents. Stakeholders may have de-ferred to leaders in the orchestras, for instance,because of the leaders’ evident expertise in certaindomains, which varied by orchestra (for instance,cost cutting in the broadcasting orchestra and col-laboration in the London orchestra). More gener-ally, this study highlights previously overlookedconnections between these literatures and organi-zational sensemaking by giving attention to a di-verse range of issues and stakeholders.

Implications for Practice

This study also has several important practicalimplications. The first stems from the connectionbetween sensemaking processes and outcomes.Each of the four forms is associated with distincttypes of accounts and actions; the optimal form ofsensemaking for an organization may therefore varywith the kind of outcome sought. For instance,


guided sensemaking processes may be particularlyvaluable in situations that require the developmentof a rich, multifaceted account that can be used asa resource for ongoing and spontaneous actions,such as establishing an organization’s core values.In contrast, restricted sensemaking processes mightbe advantageous when an issue (or its context) de-mands a narrowly defined account and is best ad-dressed by a single, decisive action. Fragmentedsensemaking processes may prove fruitful when anorganization would benefit from the constructionof a wide range of disparate accounts; such benefitsmight be associated with highly complex issues inwhich individual experimentation is needed. Al-though guided organizational sensemaking mightbe considered the most effective form, this studysuggests that leaders and stakeholders should try toengage in the form most closely aligned to the kindsof outcomes they hope to achieve.

A second practical implication of the study is formanagers seeking to encourage empowerment intheir organizations. The findings presented heresuggest a tension between the forces of animationand control in organizational sensemaking: anima-tion is a powerful element that brings diverse un-derstandings of an issue into discussion, but it ismost likely to result in consistent actions when thesensemaking processes are also relatively con-trolled. Thus, if consistent action is an importantgoal, organizations seeking employee empower-ment should take note of the importance of privateand systematically organized opportunities for em-ployees to contribute their views, as well as theimportance of leader sensegiving that providesstakeholders with an overarching framework fortheir empowered activities.

The third practical implication concerns the na-ture of animation in organizational sensemaking.The results of this study clearly show that in ananimated sensemaking process that involves nu-merous stakeholders in sensegiving roles, all par-ties need not be engaged simultaneously in theissue of concern. In the guided form of organiza-tional sensemaking, for example, the high level ofanimation resulted from the serial engagement ofstakeholders with an issue over time. This scenariocontrasts with some dominant images of stake-holder sensegiving that point to large-scale forums,such as multiparty retreats and conferences, as keyvehicles for increasing animation. This studyshows that organizations can achieve highly ani-mated sensemaking while reducing the cognitivecomplexity of the process when leaders connectsequentially with stakeholders in dyadic sense-giving exchanges.

REFERENCES

Allmendinger, J., & Hackman, J. R. 1996. Organizations inchanging environments: The case of East Germansymphony orchestras. Administrative ScienceQuarterly, 41: 337–369.

Antaki, C. 1994. Explaining and arguing: The socialorganization of accounts. London: Sage.

Bartunek, J., Krim, R., Necochea, R., & Humphries, M.1999. Sensemaking, sensegiving, and leadership instrategic organizational development. In J. Wagner(Ed.), Advances in qualitative organizational re-search, vol. 2: 37–71. Greenwich, CT: JAI Press.

BBC/Arts Council of England. 1994. BBC/Arts Councilreview of national orchestral provision consulta-tion document. London: BBC/Arts Council of En-gland.

Berger, P., & Luckmann, T. 1967. The social constructionof reality: A treatise in the sociology of knowledge.London: Penguin Books.

Boje, D. M. 1991. The storytelling organization: A studyof story performance in an office-supply firm. Ad-ministrative Science Quarterly, 36: 106–126.

Brown, A. D. 2000. Making sense of inquiry sensemak-ing. Journal of Management Studies, 37: 45–75.

Brown, A. D., & Jones, M. 2000. Honourable membersand dishonourable deeds: Sensemaking, impressionmanagement and legitimation. Human Relations,53: 655–689.

Dunford, R., & Jones, D. 2000. Narrative in strategicchange. Human Relations, 53: 1207–1226.

Dutton, J. E., & Ashford, S. J. 1993. Selling issues to topmanagement. Academy of Management Review,18: 397–428.

Dutton, J. E., Ashford, S. J., Wierba, E. E., O’Neill, R., &Hayes, E. 1997. Reading the wind: How middle man-agers assess the context for issue selling to top man-agers. Strategic Management Journal, 15: 407–425.

Dutton, J. E., & Dukerich J. M. 1991. Keeping an eye onthe mirror: Image and identity in organizational ad-aptation. Academy of Management Journal, 34:517–554.

Dutton, J. E., & Jackson, S. E. 1987. Categorizing strategicissues: Links to organizational action. Academy ofManagement Review, 12: 76–90.

Eden, C. 1992. Strategy development as a social process.Journal of Management Studies, 29: 799–811.

Eisenhardt, K. M., & Bourgeois, L. J. 1988. Politics ofstrategic decision making in high-velocity environ-ments: Toward a midrange theory. Academy ofManagement Journal, 31: 737–770.

Elsbach, K. 1994. Managing legitimacy in the Californiacattle industry: The construction and effectivenessof verbal accounts. Administrative Science Quar-terly, 39: 57–88.

2005 47Maitlis

Freeman, R. E. 1995. Stakeholder theory. In N. Nicholson(Ed.), The Blackwell encyclopaedic dictionary oforganizational behavior: 612–615. Cambridge, MA:Blackwell Business.

Gephart, R. P. 1992. Sensemaking, communicative dis-tortion and the logic of public inquiry legitimation.Industrial and Environmental Crisis Quarterly, 6:115–135.

Gephart, R. P. 1993. The textual approach: Risk andblame in disaster sensemaking. Academy of Man-agement Journal, 36: 1465–1514.

Gergen, K. J. 1999. An invitation to social construction.Thousand Oaks, CA: Sage.

Gioia, D. A., & Chittipeddi, K. 1991. Sensemaking andsensegiving in strategic change initiation. StrategicManagement Journal, 12: 433–448.

Gioia, D. A., & Thomas, J. B. 1996. Identity, image andissue interpretation: Sensemaking during strategicchange in academia. Administrative Science Quar-terly, 41: 370–403

Gray, B., & Clyman, D. R. 2003. Difficulties fosteringcooperative agreements in multiparty negotiations:Cognitive, procedural, structural and social. In M. A.West, D. Tjosvold, & K. G. Smith (Eds.), Interna-tional handbook of organizational teamwork andcooperative working: 381–400. London: Wiley.

Griffith, T. L. 1999. Technology features as triggers forsensemaking. Academy of Management Review,24: 472–488.

Hambrick, D. C., & Mason, P. 1984. Upper echelons: Theorganization as a reflection of its top managers.Academy of Management Review, 9: 193–206.

Hinings, C. R. 1997. Reflections on processual research.Scandinavian Journal of Management, 13: 493–503.

Hirsch, P. M. 1972. Processing fads and fashions: Anorganization set analysis of cultural industry sys-tems. American Journal of Sociology, 77: 639–659.

Isabella, L. A. 1990. Evolving interpretations as a changeunfolds: How managers construe key organizationalevents. Academy of Management Journal, 33:7–41.

Lampel, J., Lant, T., & Shamsie, J. 2000. Balancing act:Learning from organizing practices in cultural indus-tries. Organization Science, 11: 263–269.

Lawrence, T. B., & Phillips, N. 2002. Understanding cul-tural industries. Journal of Management Inquiry,11: 430–443.

Lebrecht, N. 1996. When the music stops: Managers,maestros and the corporate murder of classicalmusic. London: Simon & Schuster.

Lee, T. W. 1999. Using qualitative methods in organi-zational research. Thousand Oaks, CA: Sage.

Lee, T. W., Mitchell, T. R., & Sablynski, C. J. 1999. Qual-itative research in organizational and vocational psy-

chology: 1979–1999. Journal of Vocational Behav-ior, 55: 161–187.

Lincoln, Y. S., & Guba, E. G. 1985. Naturalistic inquiry.Newbury Park, CA: Sage.

Maxwell, J. A. 1996. Qualitative research design: Aninteractive approach. Thousand Oaks, CA: Sage.

Morrison, E. W., & Milliken, F. J. 2000. Organizationalsilence: A barrier to change and development in apluralistic world. Academy of Management Re-view, 25: 706–725.

Murnighan, J. K., & Conlon, D. E. 1991. The dynamics ofintense work groups: A study of British string quar-tets. Administrative Science Quarterly, 36: 165–186.

Pettigrew, A. M. 1992. The character and significance ofstrategy process research. Strategic ManagementJournal, 13: 5–16.

Polzer, J., Mannix, E. A., & Neale, M. A. 1998. Interestalignment and coalitions in multiparty negotiation.Academy of Management Journal, 41: 42–54.

Potter, J., & Wetherell, M. 1987. Discourse and socialpsychology: Beyond attitudes and behaviour. Lon-don: Sage.

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

Pruitt, D. G., Peirce, R. S., McGillicuddy, N. B., Welton,G. L., & Castrianno, L. M. 1993. Long-term success inmediation. Law and Human Behavior, 17: 313–330.

Robinson, J. 2000. Raising the demand curve for sym-phony orchestras. Harmony, 10: 1–10.

Sackmann, S. A. 1991. Cultural knowledge in organiza-tions: Exploring the collective mind. Newbury Park,CA: Sage.

Sandelands, L. E., & Stablein, R. E. 1987. The concept oforganizational mind. In S. Bacharach & N. DiTimaso(Eds.), Research in the sociology of organizations,vol. 5: 135–161. Greenwich, CT: JAI Press.

Smircich, L., & Stubbart, C. 1985. Strategic managementin an enacted world. Academy of Management Re-view, 10: 724–736.

Starbuck, W. H., & Milliken, F. J. 1988. Executives’ per-ceptual filters: What they notice and how they makesense. In D. C. Hambrick (Ed.), The executive effect:Concepts and methods for studying top managers:35–56. Greenwich, CT: JAI Press.

Strauss, A., & Corbin, J. 1998. Basics of qualitative re-search: Techniques and procedures for developinggrounded theory (2nd ed.). Thousand Oaks, CA:Sage

Thomas, J. B., Clark, S. M., & Gioia, D. A. 1993. Strategicsensemaking and organizational performance: Link-ages among scanning, interpretation, action and out-comes. Academy of Management Journal, 36: 239–270.


Tyler, T. R. 2002. Leadership and cooperation in groups.American Behavioral Scientist, 45: 769–782.

Tyler, T. R., & Blader, S. 2000. Cooperation in groups.London: Psychology Press.

Volkema, R. J., Farquhar, K., & Bergmann, T. 1996. Third-party sensemaking in interpersonal conflicts atwork: A theoretical framework. Human Relations,49: 1437–1454.

Weick, K. E. 1993. The collapse of sensemaking in orga-nizations: The Mann Gulch disaster. AdministrativeScience Quarterly, 38: 628–652.

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

Weick, K. E., & Roberts, K. H. 1993. Collective mind inorganizations: Heedful interrelating on flightdecks. Administrative Science Quarterly, 38:357–381.

Westley, F. 1990. Middle managers and strategy: Themicrodynamics of inclusion. Strategic Manage-ment Journal, 11: 337–351.

Yin, R. 1994. Case study research: Design and methods(2nd ed.). London: Sage.

Sally Maitlis ([email protected]) is an assistantprofessor of organizational behavior at the Sauder Schoolof Business, University of British Columbia. She receivedher Ph.D. from the University of Sheffield. Her researchinterests include the social and political aspects of organ-izational sensemaking and decision-making processes,and narrative and discursive approaches to the study ofemotion in organizations.

2005 49Maitlis