Dyadic Communication Relationships in Organizations: An

Dyadic Communication Relationshipsin Organizations: An Attribution/

Expectancy Approach

Bruce Barry • J. Michael CrantOwen Graduate School of Management, Vanderbilt University, 401 21st Avenue South, Nashville, Tennessee 37203,

bruce. barry @ owen. vanderbilt. eduMendoza College of Business, University of Notre Dame, Notre Dame, Indiana 46556, [email protected]

AbstractResearch in organizational communication has examined thestructure and content of interaction, but has paid little attentionto research traditions outside the organizational sciences thatexplore the social-psychological interconnections between re-lationship development and interaction. In this paper we drawupon and extend those traditions to develop a model of howcommunication relationships develop within organizationaldyads. The proposed model examines organization-based com-munication relationships through a synthesis of theoretical per-spectives on communication richness, relational communica-tion, interpersonal attribution, and social expectancy. We alsocall upon precepts of structuration theory to embed these mi-crolevel processes in an organizational context.

The relational outcome in the model is "interactional rich-ness,' ' a dyad-level construct that assesses the extent to whichcommunication within the dyad is high in shared meaning.Model antecedents are aspects of interaction through whichcommunicators reciprocally define their relationships, includingrelational message properties, message pattems that emergeover time, and relational perceptions. We propose that thesecommunication properties and behaviors give rise to relation-ship attributions. We then incorporate processes of expectancyconfirmation and violation to explain how specific communi-cation encounters lead individuals to reformulate attributionsregarding the status of a given relationship. Research proposi-tions articulate how attribution/expectancy processes mediatebetween relational communication behavior and relationshipdevelopment outcomes. We also develop propositions address-ing how relational communication behavior is influenced bymacrolevel factors, including hierarchy, structure, and culture.

In a concluding section we discuss the model's potential con-tribution to research and practice, address its limitations, andoffer recommendations for future research aimed at testing itsembedded hypotheses.(Relationships; Communication; Attribution; Richness)

It has long been recognized that managers spend a largeamount of their time engaged in communication activities(e.g., Mintzberg 1973), and many believe that successfulmanagers can be distinguished from less successful man-agers largely by their communication behavior (e.g.,Luthans et al. 1985). Researchers exploring the processesby which organizational members communicate interper-sonally have addressed both the structure and content ofinteraction, although typically as distinct lines of inquiry.Some examine organizational factors that underlie infor-mation sharing, and consider the pattems of communi-cation and organizationally relevant outcomes that ensue.An example is information richness theory (Daft andLengel 1984, 1986), which predicts that organizationalsuccess is related to the use of communication channelsthat are appropriate for a given task's information-processing requirements. Other approaches to organiza-tional communication are message centered, favoring anemphasis on the substance of information exchange overattention to the channels through which it occurs (seeStohl and Redding 1987, and Seibold et al. 1994, for re-views).

We focus in this paper on a critical aspect of commu-nication in organizational contexts that existing perspec-tives have largely ignored: the social-cognitive dynamicsof ongoing interpersonal relationships that actors bring toand define within dyadic interaction. For example, howmight the way a boss and subordinate perceive and makesense of their relationship affect their work-related inter-actions, such as problem-solving meetings or perfor-mance appraisal sessions? What social (e.g., previous en-counters) and cognitive (e.g., attributions of motives)variables and processes determine these perceptions ofworkplace relationships, and how are these processes andperceptions related to the quality of interaction thatensues? How are these dynamics shaped by macrolevel

1047-7039/00/1106/0648/$05.001526-5455 electronic ISSN

ORGANIZATION SCIENCE, © 2000 INFORMSVol. 11, No. 6, November-December 2000, pp. 648-664

BRUCE BARRY AND J. MICHAEL CRANT Dyadic Communication Relationships in Organizations

features of the organizational context, such as structureand culture?

Theoretical and empirical studies of the communica-tion aspects of interpersonal relationships are widely un-dertaken by communication theorists (e.g., Burgoon et al.1987, Millar and Rogers 1987) and social psychologists(e.g., Berscheid et al. 1989, Duck and Pittman 1994).Management researchers have examined narrow issues ofrelational exchange between leaders and followers (e.g.,Graen et al. 1982), and there is limited evidence thatworkplace friendships are related to work-related out-comes such as job satisfaction (Winstead et al. 1995).However, relationship dynamics have not been treated asa central feature within conceptualizations of organiza-tional communication, although a surge of research in-terest in this area is evident (e.g., Chang and Holt 1996,Greenhalgh and Chapman 1995, Jehn and Shah 1994,King and Hinson 1994)

Our specific objective is to propose a model of howdyadic relationships develop within organizational con-texts, and ultimately demonstrate why communicationsvary in efficiency and effectiveness based on perceptionsof the nature and depth of the relationship between theparties over time. Our approach is integrative in the sensethat it brings together disparate but related theoretical per-spectives on social interaction, namely communicationrichness (e.g.. Daft and Lengel 1986), relational com-munication (e.g., Stohl and Redding 1987), interpersonalattribution (e.g., Fletcher and Fincham 1991), and socialexpectancy (e.g., Jones 1990). Much has been empiricallyuncovered and written about the psychology of close re-lationships in everyday life, but virtually none of thiswork has been brought to bear on the special case oforganizationally bound relationships. In the pages thatfollow, we describe the contribution of these perspectivesto our approach to relationship dynamics in organiza-tional dyads, and develop propositions identifying factorsthat contribute to relational development. The paper con-cludes with a discussion of the implications of our anal-ysis for research and practice.

A Model of Relational DevelopmentA framework that outlines social-cognitive determinantsof relational development appears in Figure 1. Our ap-proach is informed by several distinct theoretical per-spectives on interpersonal behavior. First, the interac-tional richness construct, shown in the figure as theoutcome of relational development, is an adaptation andextension of theories of information richness (Daft andLengel 1984) and media richness (Sitkin et al. 1992).Second, antecedents on the left side of the model re-fiect the interplay between interaction and cognition that

is the hallmark of relational communication theory(e.g., Bateson 1972, Burgoon and Hale 1984, 1987;Montgomery 1992, Rogers and Millar 1988). Third, inthe center of the figure we invoke principles of socialattribution (Heider 1958) to analyze how dyad membersinterpret and respond to the motivations of interactionpartners. Fourth, we incorporate the perspective of socialexpectancy theory (e.g., Jones 1990), which focuses onhow interaction pattems confirm or violate an individ-ual's prior expectations. Expectancy processes are in-cluded as moderators that define the conditions underwhich relational pattems and attributions will influencethe development of communication relationships. Finally,adopting the perspective of structuration theory (Giddens1979, 1984), we analyze macrolevel aspects of the or-ganizational context that may influence the social-cognitive processes that lead to interactional richness.

Rousseau (1985) argued that theory about organiza-tional processes must explicitly address the role of levelof analysis, in part because organizational phenomena re-quire more emphasis on multilevel research. Our frame-work takes a meso-level approach (Capelli and Sherer1991) toward describing the processes that ultimatelylead to interactional richness. The ultimate outcome ofthe model, interactional richness, is a property of a dyadicrelationship. However, like many organizational phenom-ena, to fully understand its determinants requires movingbeyond one particular unit of analysis (the dyad) and in-corporating contextual variables. Our framework pointstoward a multilevel, complex process leading toward in-teractional richness. Its determinants come from both in-terpersonal and contextual variables, and from individual,dyadic, and organizational levels of analysis. Thus, theframework and the discussion that follows will incorpo-rate both micro and macro processes.

In the following sections, we discuss each of the ele-ments of the framework, beginning with the interactionalrichness constmct.

Richness Theory and The Interactional RichnessConstructAs originally formulated, information richness theory(Daft and Lengel 1984, 1986) described the linkages be-tween organizational context, the information-carryingcapacity of data, and managerial communication behav-ior. A communication event is relatively "rich" when theact of communication provides substantial new under-standing. Just as information can be more or less rich, sotoo can the communication channels through which in-formation is exchanged be compared on a richness con-tinuum. A hierarchy of media richness ranks communi-cation channels on the basis of feedback potential.

ORGANIZATION SCIENCEAôl. 11, No. 6, November-December 2000 649


Figure 1 The Determinants of interactionai Richness

Relational Contentof Messages

e.g.,

• informality* composure• conversation* metacommunication• empathy• affection* self-disclosure

Temporal Patternsof Messages

e.g.. frequencydiversitysymmetry

Relational Perceptions

e.g., ^ dependence• commitment* confidence* transferability

Macro-level Influenceshierarchy

• structureculture

Social AttributionsIndividual judgment that

relationship motivation is:

both expressive (+)

I imlrumental | ^ ^ • | both instrumental

ExpectancyDisconfirmation

' shared interaction nonnsattitudinal similarity

InteractionalRichness

the variety of cues that can be transmitted, the(im)personality of the message, and the extent of lan-guage variety. Face-to-face communication is the richestmedium because of its immediate feedback, combinationof audio and visual channels, potential for personal mes-sages, and variety of linguistic forms. Put a different way,the face-to-face medium has the highest level of "car-rying capacity," i.e., the best potential for transmittinginformation that is high in richness. Telephone commu-nication is next highest on the richness continuum, fol-lowed by personal written messages (e.g., letters andmemos), formal written messages (e.g., bulletins and re-ports), and finally numeric documents (e.g., computeroutput and statistical reports). It has been suggested thatthe richness of electronic mail falls between telephoneand written communication (Daft and Lengel 1986,Trevino et al. 1990, Trevino et al. 1987), althoughSchmitz and Fulk (1991) presented evidence that elec-tronic mail is lower on the richness continuum.

Central to richness theory is a matching hypothesis.

which holds that information richness varies positivelywith the complexity and equivocality of organizationalissues; thus, managers will employ rich media in responseto complex and changing circumstances—to cope withuncertainty. Less rich, or "lean" media (Trevino et al.1990) are more appropriately employed in response towell-understood issues where rapid feedback and high-variety information cues are less essential. An extensionof the matching hypothesis is the notion of media sensi-tivity: Managers who select media that match the richnesshypothesis (media-sensitive managers) are presumed toperform better than media-insensitive managers (Daft etal. 1987). Empirical support for these predictions of rich-ness theory has been tnixed (see Fulk and Boyd 1991 andMarkus 1994 for reviews). There is some evidence tosupport the notion that communicators choose media inaccord with the level of ambiguity of the communicationevent (e.g., Russ et al. 1990, Trevino et al. 1987), andthat doing so is related to managerial performance (Daftet al. 1987, Rice 1992).' Other research studies raise

650 ORGANIZATION SCIENCEA'OI. 11, No. 6, November-December 2000


doubts about the richness hypothesis, particularly in con-nection with "new" media such as electronic mail(Markus 1988, Rice and Shook 1990).

The critical element of richness theory for the presentanalysis is "carrying capacity"—the idea that commu-nication acts and channels vary according to their abilityto transmit information that enhances understanding (Daftand Lengel 1984) or conveys symbolic meaning (Trevinoet al. 1987). We argue that beyond information and me-dia, the notion of carrying capacity can be extended tocapture the dynamics of a developing communication re-lationship. Just as "information richness" describes theinformational depth of transmitted data, and ' 'media rich-ness" describes the information- and symbol-carrying ca-pacity of a communication channel, a dyadic communi-cation relationship may likewise be described in terms ofthe "richness" of informational flow between individ-uals.^ As communication relationships develop and socialdistance decreases, it follows that individuals within thedyad communicate more expertly and efficiently—sayingand meaning more, with more accurate reception andcomprehension, using fewer words and symbols. This ex-pertise and efficiency marks dyadic interaction as high ininformational and symbolic content; hence the dyad itselfis "interactionally rich."

In addition to communication efficiency, interactionalrichness may also reflect coordination within the dyad.Research by Bemieri and his colleagues (e.g., Bemieri etal. 1994) examining the coordination of movement withinsocial interaction—known as "interactional syn-chrony"—suggests that coordination is a basic elementof interpersonal rapport between two communicationpartners. Like rapport, interactional richness is an emer-gent social property at the dyad level of analysis that isdetermined by intradyadic processes of interaction andsocial perception. It is not some general notion of rela-tional closeness (e.g., Berscheid et al. 1989) per se thatdefines interactional richness; rather, we have in mind theevolution of shared systems of meaning through prior ex-perience, communication behavior, and the developmentof appropriate social-cognitive structures (e.g., relation-ally relevant perceptions, norms, attributions, and expec-tancies).

The notion of shared meaning as an important featureof relational development has been cultivated by scholarswho examine close relationships involving friends andintimates. Reviewing relevant empirical studies. Duck(1994) observed that "everyday talk manifests differ-ences in the meanings and understandings that areshared. . . .people are organizing and giving meaning torelationships by dealing with the variabilities that con-front them across time in dealing with their partners'' (p.

55). For organization-based relationships, we present"interactional richness" as a measurable dyad-level con-struct that rates not the global intensity of the relationshipitself, but rather the extent to which communicationwithin the dyad is high in shared meaning. We define theconstruct formally as the extent to which communicationwithin an interacting dyad at a given point in time isefficient (high in symbolic content), coordinated (char-acterized by synchronous interaction), and accurate(symbolic meaning is shared and appropriately inter-preted). Interactional richness is a characteristic of therelationship itself that is derived from the pattem of in-terpersonal experiences and communication characteris-tics that unfold over time.

Antecedents of Relational DevelopmentWithin the relational communication perspective, com-munication is a negotiated process through which com-municators reciprocally define their interpersonal rela-tionships (Stohl and Redding 1987). Relationalcommunication theorists thus regard messages and rela-tionships as interwoven (Millar and Rogers 1987, Mont-gomery 1992). Messages convey not only referentialmeaning, but also relational meaning through which in-dividuals interpret their relationships (Fairhurst et al.1987). Within Figure 1, research on relational commu-nication contributes behavioral and perceptual precursorsof relationship development, including (a) the relationalcontent of specific messages exchanged within the dyad,(b) pattems of messages that emerge over time and acrosscommunication encounters, and (c) perceptions by dyadmembers regarding the status of the nascent relationship.

Relational Content of Messages. Relational messageproperties are attributes of dyadic messages that defineand signal the progress of relational development. Rela-tional messages ' 'indicate how two or more people regardeach other, regard their relationship, or regard themselveswithin the context of their relationship" (Burgoon andHale 1984, p. 193). Properties of relational messages aredistinct continua along which relational communicationcan be classified. Consistent with social penetration the-ory (Taylor and Altman 1987) and with Gabarro's (1978,1990) model of working relationships, we assume thatrelational communication evolves from superficial todeeper levels as the affiliation develops. Over time, in-teractants move from tentative, exploratory forms of com-munication through stages where individuals exercise lesscaution, reveal more personality, become more friendlyand casual, and eventually come to predict and interpreteach other's behavior rapidly and accurately, with sen-sitivity to nuance.

Relational communication research provides a number

ORGANIZATION SCIENCEA'OI. 11, No. 6, November-December 2000 651


of measureabie constmcts addressing the relational con-tent of exchanged messages; we cite some examples hereand in Figure 1. Informality refers to the verbal tone ofexchanged messages, ranging from formal or decorous atone extreme to informal or casual on the other (Burgoonand Hale 1987). Composure describes the extent to whichmessages are imbued with (or lack) signs of apprehen-sion, anxiety, or worry about the communication encoun-ter (Burgoon and Hale 1987). Conversation is a measureof the volume of extraneous information, or ' 'small talk,"beyond purposive message content related to the actualpurpose of the communication encounter (Miller andSteinberg 1975). Metacommunication refers generally tomessages that have impact on how other messages areinterpreted (Montgomery 1988). More narrowly, it is"talk about talk"—the extent to which the exchange ofmessages includes explicit remarks about other messagescontained in the conversation (e.g., remarks about thetone or meaning of purposive messages) (Gottman 1979).Empathy is operationalized as the extent to which com-munication conveys emotional sensitivity and refiects ac-curate other-person perception (Buck 1989, Miller andSteinberg 1975). Affection is the extent to which mes-sages include explicit expressions of liking and attractionto the other dyad member (Burgoon and Hale 1987,Hinde 1979). Self-disclosure refers to the volume ofinformation that is communicated to which the otherparty would not otherwise have access (Greenhalgh andChapman 1993, Miller and Steinberg 1975). As relationalattributes of messages, these features specify how orga-nizational actors interpret and define their interpersonalrelationship.

Temporal Message Patterns. Beyond the specific,measurable attributes of individual messages, relationaldevelopment may also be analyzed in terms of aggrega-tions of messages and exchanges that take place over timeand across communication events (Wemer and Baxter1994). Such attributes take the form of emergent prop-erties that result from repeated interaction involving agiven dyad. Construct definitions of temporal propertieshave been articulated by researchers studying close rela-tionships (Berscheid et al. 1989), who regard them as de-fining characteristics of intimate associations, as well asby relational communication theorists (e.g., Millar andRogers 1987), and researchers examining social compar-ison processes (e.g., Erickson 1988). Gabarro (1990) ap-plied temporal issues to workplace dyads, arguing thatworking relationships proceed through four stages: ori-entation, exploration, testing, and stabilization. A key ad-vantage of developmental approaches is the ability to an-alyze emerging relationship norms and routines, which

over time accumulate and can become difficult to disen-tangle or change.

Operationalizable examples of temporal pattems drawnfrom these research perspectives include frequency—ameasure of how often interaction within the dyad takesplace; diversity—the variety of subjects that is the basisfor exchanges within the dyad over time; and symmetry—an assessment of the degree to which the mechanisms ofexchanges and distribution of communication outcomesare balanced, rather than skewed, between dyad members.Relationship development is facilitated to the extent thatdyad members interact often, communicate on a varietyof subjects, possess reasonably equivalent access to what-ever communication methods are employed, and experi-ence outcomes of interaction that do not notably favorone dyad member over the other.

Relational Perceptions. Basic to our approach is therole of social perceptions within a developing commu-nication relationship. The study of social perceptions indyadic contexts crosses the disciplinary boundaries of so-ciology, social psychology, and communication theory.Perception-centered approaches to social interaction as-sume that individuals treat others as they perceive them,not as they really are (Jones 1990). Of special interesthere is the relational perception—cognitive appraisals byindividual dyad members regarding the status and qualityof the interpersonal relationship between dyad members.Relational perceptions are products of one's perceivedsocial environment, and form the basis for decisions re-garding whether or not to participate in social situations,evaluations of others within those situations, and choicesregarding communication strategy (Miller et al. 1994).

In Figure 1, we identify four interpersonal perceptionexemplars that contribute to and are determined by rela-tional messages and pattems. Dependence is an individ-ual's perception regarding the relative dependence of selfon the other dyad member compared with the dependenceof the other on self (Emerson 1962, Burgoon and Hale1987). This perceived dependence is, naturally, infiu-enced by the broader climate of power and authority re-lations in the organization within which a particular dy-adic relationship is embedded. Commitment is anindividual's judgment regarding his or her level of psy-chological attachment to the other person and intentionto maintain the dyadic relationship (e.g., Rusbult 1980).Confidence is the perception that the other party will notbetray one's trust in future interaction (Burgoon and Hale1987, Millar and Rogers 1987). Transferability is the per-ception that altemative relationships are available, miti-gating the potential for exploitation within this affiliation(Millar and Rogers 1987). Thus, relational development

652 ORGANIZATION SCIENCEA/O1. 11, No. 6, November-December 2000


is influenced by the extent to which dyad members per-ceive equal levels of dependence, high levels of commit-ment, high levels of confidence in the trustworthiness ofthe other party, and the existence of few readily availablealtemative relationships.

Interrelationships Among Antecedents. Consistentwith the view that relational development entails "inter-locked cycles of messages, continually negotiated and co-defined" (Millar and Rogers 1987, p. 118), Figure 1 in-cludes reciprocal linkages among the three antecedentcategories. Relational messages are presumed to give riseover time to temporal patterns that signal closer affilia-tion; by the same token, the development of these pattemsover time increases the volume of relational message con-tent within subsequent individual exchanges. The con-ceptual link that ties message content and pattems withrelational perceptions is based on an uncertainty reduc-tion principle. Relational messages that are informativeabout the status of the relationship—such as conversationand metacommunication—should have an uncertainty-reducing impact (see Berger 1988 for a review). Accord-ingly, exchanges endowed with relational content mayyield positive relational perceptions. Over time, com-munication pattems accelerate relationship developmentthrough uncertainty reduction as dyad members com-municate more often (frequency), more equitably (sym-metry), and over a wider range of subjects (diversity).Reciprocally, relational content and temporal pattems areinfluenced by developing relational perceptions. As per-ceptions become more positive, incentives to imbue mes-sages with relational content are created. These percep-tions should also elicit changes in the temporal pattemsof exchange, i.e., create incentives and opportunities forcommunication that is more frequent and more diverse insubject matter.

Social Attribution Processes in OrganizationalRelationshipsIn the center of Figure 1, attributions of motivation forparticipating in and maintaining the relationship mediatebetween relational antecedents and the development ofinteractional richness within a given dyad. Attributiontheory emerged from Heider's (1958) insight that peopleare intuitive psychologists who perceive the causes ofothers' behavior. Antecedents and consequences ofcausal attributions have long been of interest to research-ers studying interpersonal relationships. We argue thatsocial attribution processes are central to understandingthe conditions under which relational communication be-havior translates into relational development.

Research suggests that causal attributions are ubiqui-tous; people make causal attributions regarding both their

own behavior and that of others (Nisbett and Valins1971). Information, beliefs, and motivation are the pri-mary inputs to the attribution process (see Kelley andMichela 1980 for a review). As information about peopleand events is obtained, the attributor relies on his or herbelief system to explain the reasons behind stimuli. Anumber of studies provide insight into the conditions un-der which one is motivated to attribute causality. For ex-ample, attributional reasoning is invoked when one en-counters unexpected events or fails to attain a goal(Weiner 1985), when people perceive dependence on an-other person for both positive and negative outcomes(Hastie 1984), and when one anticipates future interactionwith another person (Berscheid et al. 1976).

Classical attribution theory assumes a basic distinctionbetween causal factors that are located within the per-ceiver—internal attributions—and factors that are basedin the perceiver's environment—external attributions(Heider 1958). In dyadic communication, an intemal at-tribution places the locus of responsibility for an inter-action partner's verbal or nonverbal behavior within thepartner's stable disposition or attitude set (e.g., the bosswas forgiving about my mistake because he/she believesin the value of taking risks and leaming from errors),while an extemal attribution traces the behavior to anephemeral, situationally specific cause (e.g., the boss wasforgiving because he/she knows that at the moment I'mconsidering leaving the organization for a new opportu-nity).

The psychological literature on close relationships pro-vides substantial evidence that interacting dyads triggerthe causal attribution process as messages of varying re-lational content are exchanged over time, and relationalperceptions emerge (for a review see Fletcher andFincham 1991). This sequence of events elicits attribu-tional reasoning, particularly as multiple communicationevents unfold (Kelley 1973). Discussing intimate rela-tionships, Newman (1981) coined the term "interper-sonal attribution" to conceptualize the attribution pro-cesses of individuals in dyadic contexts. She argued thatattributions in dyadic relationships must move beyondtraditional dispositional/situational categories towardeach party and account for behavior with respect to in-teractive pattems occurring between the two persons inthe relationship. Howe (1987) provided evidence that in-terpersonal attributions are distinct from—and more com-plex than—attributions to intemal or extemal causes.

Thus classical attribution theory has offered substantialcontributions to understanding the causal reasoning ofparties within intimate relationships; it may, however, beincomplete as a framework for understanding such rea-soning in workplace relationships. This is because inti-mate relationships derive primarily from the strength of

ORGANIZATION SCIENCE/VOI. 11, No. 6, November-December 2000 653


emotional or expressive interpersonal ties between dyadmembers. Relationships among organization membersmay well feature such ties (e.g., Winstead et al. 1995),but also commonly revolve around instrumental ex-change in an economic sense. Put a different way, if thereis no intimacy, an intimate relationship scarcely exists. Aworkplace relationship, on the other hand, may be quitewell-developed in the absence of intimacy because of theinstrumental value of the relationship for the fulfillmentof organizational roles and objectives.

We propose, therefore, that relationship attributions inorganizational dyads may also be characterized along adimension contrasting expressive and instrumental attri-butions. These interpersonal attributions explain an inter-action partner's communication behavior in terms of per-ceptions of the parties' motivation for maintaining therelationship. Expressive attributions refer to beliefs thatthe relationship exists primarily because the parties deriveemotional satisfaction from it; instrumental attributionsare made when the parties are seen as motivated to main-tain the relationship to fulfill role demands, meet orga-nizational obligations, or otherwise make progress towardorganizationally relevant goals. Variations on this dis-tinction are common to a number of literatures, includingthe social network contrast between expressive and in-strumental ties (e.g., Ibarra 1993), social psychologicalreferences to exchange vs. communal relations (Mills andClark 1994) and intrinsic and extrinsic motivational ori-entations (Amabile et al. 1994), the distinction in soci-ology between gemeinschaft and gesellschaft relations(e.g., Marwell and Hage 1970), and discussions of work-ing vs. social relationships (Gabarro 1978).̂ Consistentwith Newman's (1981) ideas about interpersonal attri-bution, we argue for the existence of expressive and in-strumental attributions about the (aggregated) relation-ship rather than separate attributions regarding self andother party.

Communication Determinants of Relationship Attri-butions. A common assumption in the attribution liter-ature is that an individual's attributions will affect his orher subsequent actions (e.g., Heider 1958, Kelley 1973),yielding a variety of cognitive and behavioral outcomes.For example, attributing causes to actors rather than sit-uations affects liking for the actor, tmst, and perceptionsof the actor's persuasiveness (Kelley and Michela 1980)as well as conflict resolution and dysfunctional relation-ship behavior (Bradbury and Fincham 1992). Attributionsappear to mediate behavior and affective reactions by in-fiuencing our feelings about past events, our expectationsfor future events, and our attitudes toward others (Kelleyand Michela 1980). Thus, relational messages, pattems.

and perceptions should contribute to the formation of at-tributions that the relationship is expressively motivated,and diminish the likelihood of instmmental attributions.

PROPOSITION 1. The extent of relational message con-tent (e.g., informality, composure, conversation, empathy,affection, self-disclosure, and metacommunication) ispositively associated with the formation of expressive at-tributions, and negatively associated with the formationof instrumental attributions.

PROPOSITION 2. Temporal patterns of communicationthat signal relational development (e.g., frequency, di-versity, and symmetry) are positively associated with theformation of expressive attributions, and negatively as-sociated with the formation of instrumental attributions.

PROPOSITION 3. Favorable relational perceptions (e.g.,dependence, commitment, confidence, and transferabil-ity) are positively associated with the formation of ex-pressive attributions, and negatively associated with theformation of instrumental attributions.

Thus far, we have treated expressive and instmmentalmotivations as mutually exclusive. Such a conceptuali-zation is consistent with Clark and Mills' (1993) viewson communal and exchange relationships, whereby peo-ple are presumed to be motivated at any single point intime by only one factor. In contrast, other theoretical tra-ditions consider the case where constmcts similar to ourexpressive/instmmental distinction coexist. For example,Amabile et al. (1994) found that intrinsic and extrinsicmotivation are orthogonal, and one can be dually moti-vated by both. Ibarra (1993) made a similar point aboutnetwork relationships, arguing that some network rela-tionships can be both instrumental and expressive. In thecase of developing communication relationships, a mixedpattem of relational content and temporal pattems maylead a communicant to attribute both expressive and in-stmmental motives to the interpersonal relationship. Insuch a situation, we believe that expressive motivationswill dominate subsequent behavior. This view is consis-tent with Mills and Clark's (1982) argument that, for mostpeople, their communal relationships—those based on aconcem for the other party's needs rather than on a quidpro quo arrangement—are the relationships that are mostimportant to them. Thus,

PROPOSITION 4. Mixed patterns of relational messagecontent, temporal patterns of communication, and rela-tional perceptions will be associated with the formationof both expressive and instrumental attributions. In thiscase, expressive attributions will be determinative.

Attributional Shifts: The Role of Expectancies.

654 ORGANIZATION SCIENCE/VOI. 11, No. 6, November-December 2000


Propositions 1—4 are static in the sense that they treatattributions as more or less likely to be formed at a givenpoint in time depending, in part, on the stmcture of in-teraction that has taken place between the parties up tothat point. Of course, relationships are not static; indeed,at the foundation of the present analysis is a set of as-sumptions about the development of communication link-ages based on dynamic exchanges between dyad mem-bers. Even a relationship that we might regard as"stable" may still exhibit considerable change around abasic fulcmm of exchange (Dindia and Canary 1993). Asrelationships evolve over time, attributions reflect shiftsin a perceiver'sjudgments about the other party and aboutthe relationship. In our model, processes of social expec-tancy confirmation and disconflrmation form the basis forunderstanding how these shifts influence relationship de-velopment.

As deflned by attribution theorists, expectancies arejudgments and predictions about the behavior of othersand the social context within which that behavior occurs(Jones 1990, Trope 1986). Expectancies are part of thecognitive net of perceptions and schemas through whichinteractants filter their interpretation of and response toeach other's actions (Darley and Fazio 1980, Reis andShaver 1988). Violations of social expectancies (i.e., dis-confirmations) trigger cognitive evaluation processes thatprompt explanation-seeking (Planalp and Rivers 1994)and influence subsequent information-processing, per-ceptions, and behavior (Burgoon and LePoire 1993).Given a preexisting dyadic relationship, individuals cometo a communication encounter with expectations regard-ing the nature of the relationship and the appropriate tenorof interaction in light of that relationship (e.g.. Miller andSteinberg 1975). The behavior that actually occurs willeither confirm or violate those prior expectations.

Disconflrmed expectancies lead individuals to reinter-pret the origins of perceptions and behavior that werebehind the original expectancy. For example, in the realmof attitude change, when a listener hears a communicatoradopt an unexpected position, the listener revises his orher explanation for the communicator's message (Eaglyand Chaiken 1993). In close relationships, causal attri-butions are presumed to vary depending on the consis-tency or inconsistency of behaviors with expectationsbased on the status of the relationship (Fletcher andFincham 1991). Accordingly, in the present framework,expectancy violations are the mechanism by which attri-butional shifts regarding the parties' motivation to main-tain the relationship occur. For example, an individualwho regards a particular dyadic relationship as instru-mentally motivated will harbor certain expectations aboutthe nature of relational communication within the dyad

(e.g., messages low in informality, empathy, and affec-tion, with low diversity and symmetry over time). Anattributional shift (e.g., from instmmental to expressive)is triggered when the interaction partner engages in com-munication activity that violates those expectations.

For example, an auditing partner in an accounting firmmay view his or her relationship with a partner from theconsulting side of the business as being instmmentallymotivated. This partner would expect most contact to oc-cur through formal, infrequent communications contain-ing very little self-disclosure, such as brief memorandumsdescribing specific business issues. As the consultingpartner begins to initiate more frequent, informal com-munication revealing more personal information (such asengaging in regular hallway conversations about family-related issues), the auditing partner's expectations willchange regarding the communication relationship. Thisshift in expectations will be away from attributions ofinstmmental motivation and toward attributions of ex-pressive motivation. Thus,

PROPOSITION 5. Individuals adjust their attributions re-garding the motivational basis for the relationship whenrelational messages and patterns disconfirm an expec-tancy about the relationship.

But will expectancy disconfirmations always lead toattributional shifts? We argued above that adjustments toattributions about motivations for maintaining the rela-tionship result when expectancies regarding the relation-ship are violated. But it is also the case that expectanciesoften persist even in the presence of disconfirming infor-mation (Darley and Fazio 1980). We offer two lines ofargument addressing the circumstances under which at-tribution shifts are more or less likely to occur.

First, the persistence of expectancies is expected to bea function of the intemal versus extemal locus of causa-tion an individual assigns to communication behavior thatviolates those expectancies. Assume, for example, thatperson A is motivated expressively to maintain ties withperson B, and believes that B is similarly motivated. If aparticular communication from B runs counter to that be-lief—i.e., disconfirms A's social expectancies—then Amay be inclined to alter his/her perception of the parties'motivation to maintain the relationship. According toclassical attribution theory, A may perceive B's behavioras resulting from transient, context-specific causes origi-nating within the person's environment (an extemal at-tribution), or may judge B's behavior as contextually non-specific and originating within the person (an internalattribution). Intemal attributions reflect judgments thatthe other party's behavior is intentional and stable rather

ORGANIZATION SCIENCE/VOI. 11, No. 6, November-December 2000 655


than a product of fleeting situational conditions. Accord-ingly, an intemal attribution in response to behavior thatis inconsistent with prior expectations is more likely toprecipitate a change in the perceiver's view of the overallrelationship than an extemal attribution in response toinconsistent behavior. Thus,

PROPOSITION 6. A perceiver's attributions about theother party's motivation (instrumental vs. expressive) tomaintain a dyadic relationship are more likely to shiftwhen disconfirming communication behavior by the otherparty is attributed to intemal rather than external causes.

Second, preexisting relationship dynamics (i.e., thestrength of dyadic ties in advance of a given communi-cation encounter) may influence the persistence of inter-personal expectancies. We are unaware of research thatdirectly addresses this issue, but offer indirect reasoningin support of a prediction about the persistence of expec-tancies. Assume first that organization members are typ-ically motivated to maintain positive social ties with otherorganization members. It follows that expectancies willbe more persistent (less prone to disconfirmation) whenfollowed by an encounter that weakens social ties, andless persistent (more susceptible to disconfirmation) whenfollowed by an encounter that unexpectedly strengthenssocial ties. Put another way, disconfirming informationhas greater impact when it signals a closer relationshipthan the perceiver originally thought, but less impactwhen it signals a more distant relationship. Thus,

PROPOSITION 7A. The effect of disconfirmation on at-tributions and relational perceptions is greater when dis-confirming information indicates that the relationship ismore expressively motivated that expected.

PROPOSITION 7B. Expectancies are more persistent(and disconfirmations less infiuential) when disconfirm-ing information indicates that the relationship is moreinstrumentally motivated than expected.

Predicting Levels of Interactional RichnessOur framework takes the process of social attribution toa dyadic level of analysis, treating the match or mismatchbetween attributions made by parties to the relationshipas determinative. Psychodynamic approaches to intimacytreat close relationships as "a collaboration in which bothpartners reveal themselves, and seek and express valida-tion of each other's attributes and world views" (Reisand Shaver 1988, p. 369). The extent to which peoplehold similar attributions for an event has been studied ina variety of contexts, such as actor-observer biases(Bradley 1978), attributions for conflict in close relation-ships (Fincham and Bradbury 1987, Prager 1989), and

leader-member attributional differences (Bitter andGardner 1995). For example, one study found that satis-faction with a dyadic relationship was a function of sim-ilarity of attributions for role performance (Schaffer andKeith 1984). Among married dyads, experimental andlongitudinal data indicate that attributions influence mar-ital satisfaction and judgments of relationship quality(Bradbury and Fincham 1990).

Consistent with this reasoning, we hypothesize thatconvergence or divergence of instmmental or expressiveattributions within the dyad influences relationship de-velopment. Three scenarios are possible: (a) both partieswill make expressive attributions; (b) attributions will di-verge, with one party viewing the relationship as expres-sively motivated, and the other party viewing the rela-tionship as instmmentally motivated; or (c) both partieswill make instrumental attributions. We expect that thepotential for interactional richness is maximized whenboth parties make expressive attributions. In such cases,relational communication exists in good measure, andboth parties share a common motivational orientation thatwelcomes further interaction of a personal nature.

The potential for interactional richness is weakest un-der the second scenario, where one party views the rela-tionship primarily in expressive terms, and the other pri-marily in instrumental terms. In this situation, one party,but not the other, may be inclined to expand the use ofrelational communication. The result is a greater chanceof mismatch in expectations, norms, and judgments re-garding the relationship—a state of affairs that is incon-sistent with the development of an affiliation built onshared systems of communicated meaning. Thus,

PROPOSITION 8. Interactional richness within a dyad ishighest when both parties form expressive attributionsabout the relationship, and lowest when parties divergein their attributions.

Under the third scenario, agreement that the relation-ship is instmmentally motivated still allows for sharedmeaning systems, albeit on less personal levels. Dyadmembers may find such a relationship to be entirely sat-isfactory, with intensification of relational communica-tion not anticipated. In practical terms, when both dyadmembers judge the motivational basis for the relationshipto be instmmental, the possibility exists that both willperceive the relationship in similar ways and generallyobey common mles of interaction, even if underlying at-titudes and values remain independent. Accordingly, thepotential for interactional richness is uncertain, but stillapparent.



We propose that in these dyads with matched instm-mental attributions, other cognitive similarities or differ-ences come into play as moderating infiuences. First, re-gardless of whether a relationship is instmmentally orexpressively motivated, dyads will to varying degreescome to adopt and obey common norms of social inter-action (Hymes 1967). Dependence on another for the ful-fillment of needs (as in a well-established dyadic rela-tionship) elicits concem about that other person'sinterpersonal expectations and evaluations, leading to theformation of social norms (Jones and Gerard 1967). Thus,even where both parties regard their interaction as instm-mentally motivated, the presence of interaction normsthat both understand and to which both adhere shouldcontribute to interpretative acumen and communicationefficiency. We do not propose that common norms alonewill elevate the interactional richness of an instmmentallydriven relationship to the level of a situation where bothparties are expressively motivated. However, norms incommon are expected to yield some potential for en-hancing a dyad's communication relationship, even whenmotivations are essentially instmmental. Thus,

PROPOSITION 9. When both dyad members view the re-lationship as instrumentally motivated, the potential forinteractional, richness is enhanced to the extent that theparties' adhere to common interaction norms.

Second, we suggest a role for corresponding attitudeswithin the dyad. Given the balance theory postulate thatindividuals tend to agree with people they like (see Eaglyand Chaiken 1993), and the well-established finding thatindividuals tend to like people with whom they agree(Byme 1971), attitudinal similarity is a relevant aspect ofrelational development. Convergence of attitudes is to beexpected in most relationships that are expressively mo-tivated if the relationship is to intensify and endure. How-ever, in relationships motivated primarily by instmmen-tality, attitude similarity may vary widely from minimalto comprehensive. Pervasive similarity may compensatein part for the interaction deficiencies that otherwise ac-cme to a dyad whose exchanges are low in relationalcommunication, especially if attitudes converge on sub-jects that are germane to the substance of the parties'instmmental interests (e.g., job-related and organization-ally relevant issues).

PROPOSITION 10. When both dyad members view therelationship as instrumentally motivated, interactionalrichness is elevated to the extent that instrumentally rele-vant social attitudes correspond.

Macrolevel Influences: A Structurational ViewThe processes leading to (or constraining) the develop-ment of interactional richness described thus far are fo-cused on perception, attribution, and message construc-tion at the individual level of analysis. We have arguedthat interactional richness represents a dyad-level prop-erty of the interacting pair that refiects the pattem of in-terpersonal experiences and communication characteris-tics that unfold over time. Our focus, however, is notlimited to an abstract explanation of the communicationqualities of dyadic relationships in the universe of inter-actions. Rather, we are interested in the determinants anddevelopment of interactional richness among interactingpairs within formal organizations. The organizationalcontext brings stmctural and cultural influences to bearon the interplay of perceptions and actions at the heart ofour social-cognitive model. In this section, we discussmacrolevel factors that link organizational stmcture andculture to the process through which interactional rich-ness evolves.

We adopt stmcturation theory (Giddens 1979, 1984) asthe core theoretical framework guiding our analysis of theinfiuence of macrolevel forces on the microlevel pro-cesses contained within our model. The essence of struc-turation theory relevant here is its analysis of the interplayof systems, structure, and social interaction as an expla-nation for observable patterns of social behavior (Salazar1996). In stmcturation theory (see Giddens 1979, pp. 64—81; Giddens 1984, p. Ti-21), systems are the interde-pendent relations between actors or groups that take theform of ongoing or patterned social interactions. Struc-tures are sets of mles and resources that guide or governthese behavioral pattems or interactions—Giddens char-acterizes stmctures as "properties of social systems"(Giddens 1984, p. 25). Structuration refers to the ways inwhich social systems—pattems of relations—are repro-duced in interaction over time. Giddens describes struc-ture in terms of a critical duality: Structure is both a mech-anism for the organization of social systems (the mlesand resources guiding interaction) as well as an outcomeof these systems (new rules that emerge from interaction).Stmcturation theory has been applied to organizationalissues by researchers studying group processes (Poole etal. 1985, Salazar 1996), organizational climate (Bastienet al. 1995, Poole and McPhee 1983, Witmer 1997), andadvanced information technology (DeSanctis and Poole1994, Poole and DeSanctis 1990).

Structuration is an appealing tool for analyzing macro-level infiuences on the formation of interactional richnessbecause it theorizes direct links between stmcture, cul-ture, and interaction. Of particular interest here is macro-to microstmcturation that explains how dyadic interaction



is shaped by macrolevel forces. A key way this stmctur-ation occurs is when, as Giddens (1984) puts it, "the con-duct of individual actors reproduces the stmctural prop-erties of larger collectivities" (p. 24). However, ourinterest here lies not in the full sweep of macrolevel prop-erties that might elicit or constrain dyadic interaction, butin factors that specifically will influence communicationstyles, interpersonal sensemaking, symbolic understand-ing, attributional congmence, and other antecedents ofinteractional richness. We focus our attention on thestructurational influence of three clusters of macrolevelfactors: hierarchy, organizational stmcture, and culture.

Hierarchy. The relative position within an organiza-tional hierarchy of two individuals who comprise an in-teracting dyad is likely to affect communication processesthat give rise, over time, to interactional richness. As Con-rad (1983) observed, "Perceived power relationships sta-bilize and dismpt social stmctures. They are central tounderstanding the processes through which members ofsociety and formal organizations act" (p. 185). Empiricalresearch (e.g., Schilit and Locke 1982, Tjosvold et al.1992) has demonstrated that communication strategiesand interpersonal influence tactics vary depending on thehierarchical direction and other indicators of relative in-terpersonal power. An organization's formal hierarchy isthus stmcturational in the sense that individual commu-nicators will act in ways that reproduce hierarchical stmc-ture to the extent that they treat such stmctures as mlesthat constrain interaction.

The entry point for the effects of hierarchy into ourmodel of the determinants of interactional richness (seeFigure 1) is the relational content of messages constructedwithin the dyad. For example, communication directedupward within a superior-subordinate dyad is less likely,other things being equal, to be high in such attributes asinformality, affection, or self-disclosure, than lateral(peer-to-peer) interaction. Tangential conversation or me-tacommunication in a vertical (as opposed to lateral) dyadis presumably more likely to be initiated by the interactantwith higher hierarchical status than the individual withlower status. In general, we believe that the relationalcontent of intradyadic messages, which we previouslysuggested fosters expressive attributions about the dyadicrelationship, is constrained by salient hierarchical bound-aries that create mles for interaction.

PROPOSITION 11. The salience of hierarchical organi-zational roles that distinguish members of an interactingdyad inhibits the relational content of intradyadic mes-sages. Relational content will occur more frequentlyamong dyad members whose relationship is lateral orotherwise not defined by vertical hierarchical role differ-entiation.

Structure. Elements of organizational stmcture aresimilarly apt to present mles and resources that will,through the mechanism of stmcturation, influence pat-tems of communication at the dyadic level. In formulat-ing adaptive stmcturation theory, DeSanctis and Poole(1994) noted that the effects of organization stmcture ontask-level design and task priorities are likely to affecthow individuals and small groups use, interpret, and ap-ply their own locally available communication modali-ties. Research has also demonstrated that the methodsused by organization members to achieve interpersonalinfluence varies with the degree to which organizationalauthority stmctures are centralized (Krone 1992). Al-though the stmctures of departments, tasks, and technol-ogies may exhibit a multitude of effects on communica-tion activities, we assess formal stmcture as most clearlyrelated to the pattems of messages that emerge over time(the "temporal pattems of messages" in our model), andsecondarily related to the relational content of messages.

In the broadest terms, we draw on the classical contrastbetween organic stmctures, which feature low levels offormalization, low decision-making centrality, and nar-row spans of managerial control; and mechanistic stmc-tures, which are more formalized, more centralized, andfeature wider spans of control (Bums and Stalker 1961).Organic stmctures are associated with greater task vari-ety, and wider and more frequent communication activity(Daft 1992). To the extent that these stmctures also fea-ture less analyzable tasks, task-related interaction is notonly more frequent, but also more likely to take placethrough richer communication channels (Daft and Lengel1986). As a consequence, we contend that dyadic com-munication will, other things being equal, occur withgreater frequency and reflect a wider diversity of com-munication topics. Moreover, because of channel richnesseffects, organic structures may also facilitate relationalmessage content in the form of communication that isinformal, conversational, empathic, and imbued with me-tacommunication and self-disclosure.

PROPOSITION 12. The design elements of an organicdepartmental or organizational structure, such as lowformalization, low centralization, and narrow span ofcontrol, are positively associated with intradyadic mes-sage frequency and message diversity, as well as with therelational content of intradyadic messages.

Culture. With respect to communication in organi-zations, Bormann (1983) deflned culture as "the sum to-tal of ways of living, organizing, and communing builtup in a group of human beings and transmitted to new-comers by means of verbal and nonverbal communica-tion" (p. 100). A principal source of organizational cul-ture is found in what Denison (1990, p. 32) called



"perspectives"—shared rules and norms that suggest so-lutions to common problems and situations encounteredby members of the organization. These perspectives helpindividuals interpret situations and represent boundariesof acceptable behavior. From a structurational perspec-tive, collective organizational culture is reproducedthrough communication at microlevels as individualsadopt social practices that adhere to social rules generatedby macrolevel culture. In general, theories and ethnog-raphies of organizational culture highlight communica-tion roles and pattems as both manifestation and signifierof a given organization's cultural properties.

We propose that organizational culture influences com-munication processes leading to interactional richness tothe extent that behavioral norms foster or inhibit the de-velopment of close relationships. The schematic of ourmodel (Figure 1) shows this happening in two ways. First,culturally shared assumptions regarding acceptable or de-sirable social practices may facilitate or constrain the in-corporation of relational content (informality, conversa-tion, empathy, self-disclosure, etc.) into messagesexchanged within organizational boundaries. A strongculture oriented toward cooperative work and a blendingof personal and professional relationships increases thelikelihood that individuals will feel comfortable introduc-ing relational content into dyadic exchanges—even in dy-ads that have not advanced far beyond what Gabarro(1990) labeled the orientation and exploration stages of adeveloping relationship. Conversely, a culture orientedtoward crisp, exclusively task-focused interaction maystifle the incorporation of relational content.

Second, we hypothesize a role for culture as a factorin the social attribution processes at the heart of ourmodel. In propositions developed earlier, we proposedthat message patterns and relational perceptions influencethe formation of expressive versus instrumental attribu-tions regarding the motivation of dyad members to main-tain their relationship. Organizational (or suborganiza-tional) culture, we suggest here, has the ability tomoderate or bias individual tendencies to form attribu-tions in the face of evidence regarding relational status.For instance, a firm whose culture promotes isolated ca-reerism and the pursuit of radical self-interest may finditself populated with individuals who respond warily torelational message content, viewing such messages as in-strumental subversion rather than expressive overtures.On the other hand, in a high-involvement culture that fos-ters expressive activity, individuals may be inclined toperceive relational content as sincere and determinative,and hence be predisposed to undertake attributional shiftsin the presence of relatively little relational evidence. Onecan also imagine a setting where a culture of expressively

oriented communication is taken to an extreme, leadingindividuals to discount relational message content as cul-turally rather than individually constructed.

PROPOSITION 13. Organizational and/or subunit cul-ture infiuences the development of interactional richnesst>y (a) fostering or constraining the extent to which dyadicmessages are imbued with relational content, and (b) bi-asing how individual perceivers make relationship attri-butions given evidence of relational activity within thedyad.

DiscussionRecent literature on organizational communication andconflict signals an emerging consensus that relationshipissues are important but underexplored. At the same time,the popular press draws significant attention to the needfor effective organizational actors to advance their careersthrough careful management of interpersonal relation-ships (e.g., Famham 1996). Our aim in this paper was toadd theoretical value to the emerging study of relation-ship dynamics in organizations in a way that will stimu-late research and inform practice. We introduced the con-cept of interactional richness as an outcome of the processof relational development and proposed a conceptualmodel that analyzes relational and social-cognitive deter-minants of interactional richness.

An advantage of the approach taken here is that itmoves beyond a gestalt assessment of communication re-lationship quality to consider the underlying social-psychological dynamics of relational development. Ex-tant work on relationships in organizations has focusedprimarily on global and cross-sectional assessments of therelationship between an interacting pair: Is it close? Howclose? An exception is Gabarro (1978), although his per-spective on CEO-subordinate relationships was induc-tively drawn, not theoretically based. In our model, the-ories of social perception and attribution are extended tothe domain of relational development at work. The modelis dynamic, with interactional richness ultimately deter-mined by the commingling of dyad members' social-cognitive structures. As an extension of research on me-dia richness, interactional richness takes existing work onthe carrying capacity of communication features in a newdirection.

We also point to an integrative contribution. We set outto synthesize research on communication and the socialpsychology of dyadic relationships into a single frame-work directed specifically at the analysis of relationshipdynamics in organization-based interaction. Much of the

ORGANIZATION SCIENCEA'OI. 11, NO. 6, November-December 2000 659


research from which we have drawn was intended to de-scribe close personal relationships where heightened in-timacy is the measure of relational quality. Intimate re-lationships are much like other kinds of personalrelationships in that they involve high degrees of inter-dependence; in that sense many aspects of the theory andempirical findings that emerge from this body of researchare applicable to organizational settings. However, theclose relationships literature cannot fully account for theinstrumental aspects of workplace relationships that nat-urally occur with a hierarchical social structure and cul-ture of economic exchange. Thus, our aim was not simplyto transplant the study of interpersonal intimacy into anorganizational context, but rather to explore the com-munication dynamics that drive both close and arms-length relationships among organization members.

Knowing more about the psychological dynamics oforganizational communication is potentially of significantvalue to organizational participants. Interactional richnessis a desirable attribute of workplace relationships to theextent that organization members value communicationrelationships marked by efficiency of transmission andaccuracy of interpretation. The attribution/expectancymodel, while grounded in technical issues of social cog-nition, invites the individual communicator to analyze in-terpersonal connections in a plausible and accessible way:in terms of the match or mismatch of interpretations aboutwhy communicator and partner are inclined to maintaintheir association. Moreover, the model identifies numer-ous specific forms of communication behavior that giverise to these interpretations. It is likely that empirical stud-ies would reveal that some forms of relational commu-nication are more powerful than others at eliciting rele-vant relationship attributions. Armed with thatknowledge, organization members can learn to supple-ment their communication skills with a kind of "rela-tional literacy" that enables informed choices about howinterpersonal communication can foster desired forms ofrelationship development.

A key issue thus far unaddressed concems the desira-bility of interactional richness. Our treatment of the con-struct thus far has portrayed it as a desirable feature oforganizational communication relationships. Becausehigh levels of interaction richness ultimately yield moreefficient communications, this makes sense for most sit-uations. However, an argument can be made that inter-actionally rich dyads—especially when expressivelymotivated—might engage in excessive levels of com-munication seemingly unrelated to the task (e.g., smalltalk). While we believe that the most efficient dyadswould explicitly address this situation (e.g., through me-tacommunication), this potential downside of interac-tional richness must be acknowledged.

We also wish to mention three limitations of our anal-ysis that reflect tradeoffs between complexity and focus.First, our visual approach (Figure 1) to the processes con-sidered in this article oversimplifies the dynamic and re-cursive nature of relational development. This process is,of course, reciprocal and iterative, with outcomes a prod-uct of interactive give and take. Our decision to simplifythese processes—and omit the many conceivable feed-back mechanisms—was based on a desire to isolate andpresent the specific phenomena of interest in a manage-able way.

A second limitation is our focus on communication atthe dyadic level of analysis. Analytically, this orientationis appropriate because our approach is grounded in psy-chological theories of interpersonal process. However,our analysis is intended to apply to communication set-tings involving more than two participants, as well asamong broader social structures such as work groups andorganizations. By decontextualizing to the isolated dyad,we gain analytical clarity, but at the expense of a broadertreatment of relationship dynamics in multiparty inter-action.

Third, as presently formulated, the model excludes anyexplicit treatment of the role of individual differences. Anumber of psychological and biographical variables havebeen associated with verbal and nonverbal interpersonalcommunication (see Giles and Street 1994 for a review).For example, follow-up work on the information richnessmodel has found that cognitive style and communicationapprehension moderate the link between task require-ments and media choice (Alexander et al. 1991, Trevinoet al. 1990). Further, a communicator's competence to useparticular forms of communication presumably acts as aconstraint on one's choice of interaction strategy (Sitkinet al. 1992). It is reasonable to assume that some aspectsof relational development are determined in part by in-dividual differences. Incorporation of such differences ispotentially quite complex, given the need in a relationalmodel to consider personality and demography at the dy-adic level of analysis.

Potential individual differences that might be fruitfullystudied as moderators or predictors of interactional rich-ness include gender, race, self-monitoring, public self-consciousness, extraversion, and agreeableness. Researchon network relationships (e.g., Ibarra 1992, 1993) sug-gests that women and minorities have weaker networkties because of organizational factors (e.g., overall num-bers and hierarchical distributions), interaction dynamics(e.g., stereotyping), and constrained choices (e.g., diffi-culty forming cross-sex ties). Thus, there is reason to sus-pect that gender and race may play a role in the devel-opment of interactional richness within organizational

660 ORGANIZATION SCIENCEAÔI. 11, No. 6, November-December 2000


communication relationships. Self-monitoring reflectsone's awareness of and responsiveness to social andinterpersonal cues about situational appropriateness(Snyder and Gangestad 1982); therefore, self-monitoringmight be associated with one's sensitivity to relationalcontent and temporal pattems, and ultimately with rela-tional perceptions. A similar argument can be made forpublic self-conciousness (Fenigstein et al. 1975) whichrefers to one's social sensitivity and other-directedness.Two factors from the five-factor model of personality—agreeahleness and extraversion—also would presumablybe relevant to the model, because they reflect dispositionstoward behavior in social situations, such as organiza-tional communication relationships.

Constructing an agenda to test the ideas put forth hereraises issues of both measurement and research design.Most of the constructs we described have empirical pre-cedents in other literatures, with the notable exception ofinteractional richness. A first step, then, is confronting theissues related to the measurement of the interactionalrichness construct. As we have defined it, interactionalrichness captures a dyad's ability to communicate withefficiency (symbolic content), coordination, and accuracyof reception and interpretation. These are not attributesof communication events within a dyad; they are ele-ments addressing how communication events are con-structed, transmitted, and interpreted by dyad members.Accordingly, interactional richness cannot be measuredsolely by analyzing communication data. One must ex-amine both how an interacting pair communicates andhow individuals comprising the pair devise and perceivethose communications.

With this in mind, we envision interactional richnessas measureabie by both survey and behavioral ap-proaches. Surveys have been used to assess a variety ofelements of dyadic communication and relationship qual-ity, such as relational expectations (Kelly and Burgoon1991), self-disclosure (Jourard and Lasakow 1958), andrelational closeness (Berscheid et al. 1989). Followingthese traditions, it is reasonable to expect that perceptionsby dyad members of the information-carrying capacity oftheir relationship can be tapped by survey in a valid andreliable manner. Behavioral measures, such as contentanalysis of taped exchanges between interaction partners,also might be used to capture the richness of interaction.Such methods have been employed in the communicationliterature to measure variables related to dyadic interac-tion, such as self-disclosure (Prager 1989) and intimacy(Tesch and Whitbourne 1982). Behavioral methods allowthe researcher to assess characteristics of interaction frominteraction itself, rather than simply rely on self-reportedjudgments by interaction participants.

With respect to strategies for model testing, two dis-tinct avenues for empirical work are apparent. Conceiv-ably, the model could be considered in its entirety givenlongitudinal access to a sizeable pool of working dyadswith opportunities for both survey and observational datacollection. However, a logical first step would be to ex-amine modest research questions regarding the central at-tribution theory assumptions that guide our approach torelational development. It seems prudent to establish thebasic social-cognitive foundation of the model firstthrough selective empirical tests that are feasibly under-taken through survey and other less costly methods.

Ultimately, the goal should be to place a relationalmodel like the one presented here in the context ofbroader frameworks of organizational communication.The isolated study of interpersonal interaction is impor-tant in its own right, but understanding organizationalcommunication requires a broader perspective that incor-porates the structural and environmental influences onnetworks of information exchange. We see relationshipdynamics, conceptualized in terms of the development ofinteractional richness, as a promising avenue for improv-ing the predictive power of existing theories that adopt abroader perspective.

AcknowledgmentsThe authors acknowledge support for the first author from the Dean'sFund for Research at the Owen Graduate School of Management atVanderbilt University. A version of this paper was presented at the1997 annual meeting of the Academy of Management in Boston, MA.They thank Pat Calhoun for research assistance, and Bob Vecchio forhelpful comments on an earlier draft.

Endnotes'Although richness theory (Daft and Lengel 1984) originally addressedequivocality at the level of a communicator's job, empirical findingsindicate that the theory's predictive power is stronger at the level ofthe individual task or communication event (Fulk and Boyd 1991).Ôur adoption of the label richness potentially invites confusion givenexisting multiple uses of the term in organizational communicationresearch. We do, however, regard the term as appropriate on the basisof its precise meaning in the literature. Rather than limiting our analysisto global constructs describing relationship dynamics (e.g., "relationalcloseness" in the argot of social psychologists, or "intimacy" in theargot of communication theorists), we focus on information and sym-bolic content within dyadic interaction. Hence the notion of richnessas "carrying capacity" is directly relevant here, and so we employ theterm interactional richness to describe the carrying capacity of a dyadiccommunication relationship.

'The point of departure between social network studies and the presentanalysis lies within basic differences in the research questions of in-terest. In social network research, dyad-level relationship constructsare empirical tools that reveal broader patterns of social structure withinthe social system being investigated; these broader pattems are typi-cally the object of theoretical interest (Fombrun 1982). In contrast, our

ORGANIZATION SCIENCE Aôl. 11, No. 6, November-December 2000 661


theoretical focus is at the dyadic level on the psychology of relationaldevelopment through dyadic communication. Thus, social network re-search informs our understanding of the maintainance and measure-ment of interpersonal ties, but is not a window on how those ties de-velop through communication and social-cognitive processes.

ReferencesAlexander, E. R., L. E. Penley, E. Jemigan. 1991. The effect of indi-

vidual differences on managerial media choice. ManagementComm. Quart., 5 155-173.

Amabile, T. M., K. G. Hill, B. A. Hennessey, E. M. Tighe. 1994. Thework preference inventory: Assessing intrinsic and extrinsic mo-tivational orientations. / Personality and Social Psych. 66 950-967.

Bastien, D. T., R. D. McPhee, K. A. Bolton. 1995. A study and ex-tended theory of the structuration of climate. Commun. Mono-graphs. 62 87-109.

Bateson, G. 1972. Steps to an Ecology of Mind. Ballantine, New York.Berger, C. R. 1988. Uncertainty and information exchange in devel-

oping relationships. S. W. Duck, ed. Handbook of Personal Re-lationships. Wiley, Chichester, U.K.

Bemieri, F. J., J. M. Davis, R. Rosenthai, C. R. Knee. 1994. Interac-tional synchrony and rapport: Measuring synchrony in displaysdevoid of sound and facial affect. Personality and Social Psych.BMH., 20 303-311.

Berscheid, E., W. Graziano, T. Monson, M. Dermer. 1976. Outcomedependency: Attention, attribution, and attraction. / Personalityand Social Psych. 34 978-989., M. Snyder, A. M. Omoto. 1989. The relationship closeness in-ventory: Assessing the closeness of interpersonal relationships. /Personality and Social Psych. 57 792-807.

Bitter, M. E., W. L. Gardner III. 1995. A mid-range theory of theleader/member attribution process in professional service orga-nizations: The role of the organizational environment and im-pression management. M. J. Martinko, ed. Attribution Theory: AnOrganizational Perspective. St. Lucie Press, Delray Beach, FL.

Bormann, E. G. 1983. Symbolic convergence: Organizational com-munication and culture. L. L. Putnam, M. E. Pacanowsky, eds.Communication and Organizations: An tnterpretive Approach.Sage, Beverly Hills, CA.

Bradbury, T. N., F. D. Fincham. 1990. Attributions in marriage: Re-view and critique. Psych. Bull. 107 3-33., . 1992. Attributions and behavior in marital interaction. /Personality and Social Psych. 65 613-628.

Bradley, G. W. 1978. Self-serving biases in the attribution process: Areexamination of the fact or fiction question. J. Personality andSocial Psych. 36 56-7 \.

Buck, R. 1989. Emotional communication in personal relationships: Adevelopmental-interactionist view. C. Hendrick, ed. Close Rela-tionships. Sage, Newbury Park, CA.

Burgoon, J. K., J. L. Hale. 1984. The fundamental topic of relationalcommunication. Commun. Monographs 51 193-214., . 1987. Validation and measurement of the fundamentalthemes of relational communication. Commun. Monographs 5419-41., B. A. LePoire. 1993. Effects of communication expectancies.

actual communication, and expectancy disconfirmation on eval-uations of communicators and their communication behavior. Hu-man Commun. Res. 20 67-96., M. Pfau, R. Parrott, T. Birk, R. Coker, M. Burgoon. 1987. Re-lational communication, satisfaction, compliance-gaining strate-gies, and compliance in communication between physicians andpatients. Commun. Monographs 54 307—324.

Burns, T., G. M. Stalker. 1961. The Management of Innovation. Tav-istock, London, U.K.

Byme, D. E. 1971. The Attraction Paradigm. Academic Press, SanDiego, CA.

Cappelli, P., P. D. Sherer. 1991. The missing role of context in OB:The need for a meso-level approach. L. L. Cummings, B. M. Staw,eds. Research in Organizational Behavior. JAI Press, Greenwich,CT.

Chang, H. C, G. R. Holt. 1996. An exploration of interpersonal rela-tionships in two Taiwanese computer firms. Human Relations. 491489-1518.

Clark, M. S., J. Mills. 1993. The difference between communal andexchange relationships: What it is and is not. Personality andSocial Psych. Bull. 19 684-691.

Conrad, C. 1983. Organizational power: Faces and symbolic forms.L. L. Putnam, M. E. Pacanowsky, eds. Communication and Or-ganizations: An Interpretive Approach. Sage, Beverly Hills, CA.

Daft, R. L. 1992. Organization Theory and Design. 4th Ed. West, St.Paul, MN., R. H. Lengel. 1984. Information richness: A new approach tomanagerial behavior and organizational design. L. L. Cummings,B. M. Staw, eds. Research in Organizational Behavior. JAI Press,Greenwich, CT., . 1986. Organizational information requirements, mediarichness and structural design. Management Sci. 32 554-571., , L. K. Trevino. 1987. Message equivocality, media selec-tion, and manager performance: Implications for information sys-tems. MtS Quart. 11 355-366.

Darley, J. M., R. H. Fazio. 1980. Expectancy confirmation processesarising in the social interaction sequence. Amer. Psychologist 35867-881.

Denison, D. R. 1990. Corporate Culture and Organizational Effec-tiveness. Wiley, New York.

DeSanctis, G., M. S. Poole. 1994. Capturing the complexity in ad-vanced technology use: Adaptive structuration theory. Organ. Sci.5 121-147.

Dindia, K., D. J. Canary. 1993. Definitions and theoretical perspectiveson maintaining relationships. / Social and Personal Relation-ships 10 163-173.

Duck, S. 1994. Steady as she goes: Relational maintenance as a sharedmeaning system. D. J. Canary, L. Stafford, eds. Communicationand Relational Maintenance. Academic Press, San Diego, CA., G. Pittman. 1994. Social and personal relationships. M. L.Knapp, G. R. Miller, eds. Handbook of Interpersonal Communi-cation 2nd Ed. Sage, Thousand Oaks, CA.

Eagly, A. H., S. Chaiken. 1993. The Psychology of Attitudes. HarcourtBrace Jovanovich, Fort Worth, TX.

Emerson, R. M. 1962. Power-dependence relations. Amer. Soc. Rev.

Erickson, B. H. 1988. The relational basis of attitudes. B. Wellman,

662 ORGANIZATION SCIENCEA'OI. 11, No. 6, Novetnber-December 2000


S. D. Berkowitz, eds. Social Structures: a Network Approach.Cambridge University Press, Cambridge, MA.

Fairhurst, G. T., L. E. Rogers, R. A. Sarr. 1987. Manager-subordinatecontrol patterns and judgments about the relationship. M.L.McLaughlin, ed. Communication Yearbook 10. Sage, BeverlyHills, CA.

Farnham, A. 1996. Are you smart enough to keep your job? Fortune133(1)34-42.

Fenigstein, A., M. Scheier, A. Buss. 1975. Private and public self-consciousness: Assessment and theory. / Consulting and ClinicalPsych. 43 522-527.

Fincham, F. D., T. N. Bradbury. 1987. Cognitive processes and conflictin close relationships: An attribution-efficacy model. / Person-ality and Social Psych. 53 1106-1118.

Fletcher, G. J. O., F. D. Fincham. 1991. Attribution processes in closerelationships. G. J. O. Fletcher, F. D. Fincham, eds. Cognition inClose Retationships. Eribaum, Hillsdale, NJ.

Fombrun, C. J. 1982. Strategies for network research in organizations.Acad. Management Rev. 7 280-291.

Fulk, J., B. Boyd. 1991. Emerging theories of communication in or-ganizations. J. Management t7 407-446.

Gabarro, J. J. 1978. The development of trust, influence, and expec-tations. A. G. Athos, J. J. Gabarro, eds. Interpersonal Behavior:Communication and Understanding in Relationships. Prentice-Hall, Englewood Cliffs, NJ.. 1990. The development of working relationships. J. Gallegher,R. E. Kraut, C. Egido, eds. Intellectual Teamwork: Social andTechnological Foundations of Cooperative Work. Eribaum, Hills-dale, NJ.

Giddens, A. 1979. Central Problems in Social Theory: Action, Struc-ture, and Contradiction in Social Analysis. University of Califor-nia Press, Berkeley, CA.. 1984. The Constitution of Society: Outline of the Theory of Struc-turation. University of California Press, Berkeley, CA.

Giles, H., R. L. Street Jr. 1994. Communicator characteristics and be-havior. M. L. Knapp, G. R. Miller, eds. Handbook of Interper-sonal Communication, 2nd edition. Sage, Thousand Oaks, CA.

Gottman, J. M. 1979. Marital Interaction: Experimental Investigations.Academic Press, New York.

Graen, G. B., M. A. Novak, P. Sommerkamp. 1982. The effects ofleader-member exchange and job design on productivity and sat-isfaction: Testing a dual attachment model. Organ. Behavior andHuman Performance 30 109-131.

Greenhalgh, L., D. I. Chapman. 1993. The influence of negotiator re-lationships on the process and outcomes of business transactions.Working Paper, Dartmouth College, Hanover, NH., . 1995. Joint decision making: The inseparability of rela-tionships and negotiation. R. M. Kramer, D. M. Mesick, eds. Ne-gotiation as a Social Process. Sage, Thousand Oaks, CA.

Hastie, R. 1984. Causes and effects of causal attribution. J. Personalityand Social Psych. 46 44-56.

Heider, F. 1958. The Psychology of Interpersonal Relations. Wiley,New York.

Hinde, R. A. 1979. Towards Understanding Relationships. AcademicPress, New York.

Howe, G. W. 1987. Attributions of complex cause and the perception

of marital conflict. J. Personality and Social Psych. 53 1119-1128.

Hymes, D. H. 1967. Models of the interaction of language and socialsetting. J. Social Issues 23 8-28.

Ibarra, H. 1992. Homophily and differential retums: Sex differences innetwork structure and access in an advertising flrm. Admin. Sci.Quart. 37 422^147.. 1993. Personal networks of women and minorities in manage-ment: A conceptual framework. Acad. Management Rev. t8 56-87.

Jehn, K., P. Shah. 1994. Relationships at work: Are they worth it? Paperpresented at the 54th annual meeting of the Academy of Man-agement, Dallas, TX.

Jones, E. E. 1990. Interpersonal Perception. W.H. Freeman and Com-pany, New York., H. B. Gerard. 1967. Foundations of Social Psychology. Wiley,New York.

Jourard, S. M., P. Lasakow. 1958. Some factors in self-disclosure. /Abnormal and Social Psych. 56 91-98.

Kelley, H. H. 1973. The processes of causal attribution. Amer. Psy-chologist 2S 107-128., J. L. Michela. 1980. Attribution theory and research. Ann. Rev.Psych. 31457-501.

Kelly, D. L., J. K. Burgoon. 1991. Understanding marital satisfactionand couple type as functions of relational expectations. HumanCommun. Res. 18 40-69.

King, W. C. Jr., T. D. Hinson. 1994. The influence of sex and equitysensitivity on relationship preferences, assessment of opponent,and outcomes in a negotiation experiment. J. Management 20605-624.

Krone, K. J. 1992. A comparison of organizational, structural, and re-lationship effects on subordinates' upward influence choices.Commun. Quart. 40 1—15.

Luthans, F., S. A. Rosenkrantz, H. W. Hennessey. 1985. What do suc-cessful managers really do? An observational study of managerialactivities. J. Appl. Behavioral Sci. 21 255-270.

Markus, M. L. 1988. Infonnation richness theory, managers, and elec-tronic mail. Paper presented at the annual meeting of the Academyof Management, Anaheim, CA.. 1994. Electronic mail as the medium of managerial choice. Or-gan. Sci. 5 502-527.

Marwell, G., J. Hage. 1970. The organization of role-relationships: Asystematic description. Amer. Soc. Rev. 35 884-900.

Millar, F. E., L. E. Rogers. 1987. Relational dimensions of interper-sonal dynamics. M.E. Roloff, G.R. Miller, eds. Interpersonal Pro-cesses: New Directions in Communication Research. Sage, New-bury Park, CA.

Miller, G. R., M. Steinberg. 1975. Between People: a New Analysis ofInterpersonal Communication. Science Research Associates, PaloAlto, CA.

Miller, L. C, M. J. Cody, M. L. McLaughlin. 1994. Situations andgoals as fundamental constructs in interpersonal communicationresearch. M. L. Knapp, G. R. Miller, eds. Handbook of Interper-sonal Communication 2nd edition. Sage, Thousand Oaks, CA.

Mills, J., M. S. Clark. 1982. Communal and exchange relationships.Rev. Personality and Social Psych. 3 121-144.



, . 1994. Communal and exchange relationships: Controver-sies and research. R. Erber, R. Gilmour, eds. Theoretical Frame-works for Personal Relationships. Eribaum, Hillsdale, NJ.

Mintzberg, H. 1973. The Nature of Managerial Work. Harper and Row,New York.

Montgomery, B. M. 1988. Quality communication in personal rela-tionships. S. W. Duck, ed. Handbook of Personal Relationships,John Wiley & Sons, Chichester, U.K.. 1992. Communication as the interface between couples and cul-ture. S. A. Deetz, ed. Communication Yearbook 15. Sage, New-bury Park, CA.

Newman, H. 1981. Communication within ongoing intimate relation-ships: An attributional perspective. Personality and Social Psych.Bull. 7 59-70.

Nisbett, R. E., S. Valins. 1971. Perceiving the causes of one's ownbehavior. E. E. Jones, D. E. Kanouse, H. H. Kelley, R. E. Nisbett,S. Valins, B. Weiner, eds. Attribution: Perceiving the Causes ofBehavior. General Leaming Press, Morristown, NJ.

Planalp, S., M. Rivers. 1994. Changes in knowledge of personal rela-tionships. G. J. O. Fletcher, J. Fitness, eds. Knowledge Structuresin Close Relationships: A Social Psychological Approach. Eri-baum, Mahwah, NJ.

Poole, M. S., G. DeSanctis. 1990. Understanding the use of groupdecision support systems: The theory of adaptive structuration. J.Fulk, C. Steinfeld, eds. Organizations and Communication Tech-nology. Sage, Newbury Park, CA., R. D. McPhee. 1983. A structurational analysis of organizationalclimate. L. L. Putnam, M. E. Pacanowsky, eds. Communicationand Organizations: An Interpretive Approach. Sage, BeverlyHills, CA., D. R. Seibold, R. D. McPhee. 1985. Group decision-making asa structurational process. Quart. J. Speech 71 74—102.

Prager, K. J. 1989. Intimacy status and couple communication. J. Socialand Interpersonal Relationships 6 435-449.

Reis, H. T., P. Shaver. 1988. Intimacy as an interpersonal process.S. W. Duck, ed. Handbook of Personal Relationships. John Wiley& Sons, New York.

Rice, R. E. 1992. Task analyzability, use of new media, and effective-ness: A multi-site exploration of media richness. Organ. Sci. 3475-500., D. E. Shook. 1990. Relationships of job categories and organi-zational levels to use of communication channels, including elec-tronic mail: A meta-analysis and extension. J. Management Stud.27 195-229.

Rogers, L. E., F. E. Millar. 1988. Relational communication. S W.Duck, ed. Handbook of Personal Relationships. John Wiley &Sons, Chichester, U.K.

Rousseau, D. M. 1985. Issues of level in organizational research: Multi-level and cross-level perspectives. L. L. Cummings, B. M. Staw,eds. Research in Organizational Behavior. JAI Press, Greenwich,CT.

Rusbult, C. E. 1980. Commitment and satisfaction in romantic asso-ciations: A test of the investment model. J. Experimental SocialPsych. 16 172-186.

Accepted by Arie Y. Lewin; received September 19, 1999.

Russ, G. S., R. L. Daft, R. H. Lengel. 1990. Media selection and man-agerial characteristics in organizational communications. Man-agement Commun. Quart. 4 151-175.

Salazar, A. J. 1996. An analysis of the development and evolution ofroles in the small group. Small Group Res. 27 475-503.

Schaffer, R. B., P. M. Keith. 1984. A causal analysis of the relationshipbetween the self-concept and marital quality. J. Marriage and theFamily 46 909-914.

Schiht, W. K., E. A. Locke. 1982. A study of upward influence inorganizations. Admin. Sci. Quart. 27 304-316.

Schmitz, J., J. Fulk. 1991. Organizational colleagues, media richness,and electronic mail: A test of the social influence model of tech-nology use. Commun. Res. 18 487-523.

Seibold, D. R., J. G. Cantrill, R. A. Meyers. 1994. Communication andinterpersonal influence. M. L. Knapp, G. R. Miller, eds. Hand-book of Interpersonal Communication 2nd edition. Sage, Thou-sand Oaks, CA.

Sitkin, S. B., K. M. Sutcliffe, J. R. Barrios-Choplin. 1992. A dual-capacity model of communication media choice in organizations.Human Commun. Res. 18 563-598.

Snyder, M., S. Gangestad. 1982. Two investigations of self-monitoringprocesses. / Personality Soc. Psych. 43 123-135.

Stohl, C, W. C. Redding. 1987. Messages and message exchange pro-cesses. F. M. Jablin, L. L. Putnam, K. H. Roberts, L. W. Porter,eds. Handbook of Organizational Communication. Sage, New-bury Park, CA.

Taylor, D. A., I. Altman. 1987. Communication in interpersonal rela-tionships: Social penetration processes. M. E. Roloff, G. R.Miller, eds. Interpersonal Processes: New Directions in Com-munication Research. Sage, Newbury Park, CA.

Tesch, S. A., S. K. Whitboume. 1982. Intimacy and identity status inyoung adults. J. Personality and Social Psych. 43 1041-1051.

Tjosvold, D., I. R. Andrews, J. T. Stmthers. 1992. Leadership influ-ence: Goal interdependence and power. J. Social Psych., 132 39-50.

Trevino, L. K., R. H. Lengel, R. L. Daft. 1987. Media symbolism,media richness, and media choice in organizations. Commun. Res.14 553-574., , W. Bodensteiner, E. Gerloff, N. K. Muir. 1990. The rich-ness imperative and cognitive style. Management Commun.Quart 4 176-197.

Trope, Y. 1986. Identification and inference processes in dispositionalattribution. Psych. Rev. 93 239-257.

Weiner, B. 1985. Spontaneous causal thinking. Psych. Bull. 97 74-84.Wemer, C. M., L. A. Baxter. 1994. Temporal qualities of relationships:

Organismic, transactional, and dialectical views. M. L. Knapp,G. R. Miller, eds. Handbook of Interpersonal Communication 2ndedition. Sage, Thousand Oaks, CA.

Winstead, B. A., V. J. Derlega, M. J. Montgomery, C. Pilkington. 1995.The quality of friendships at work and job satisfaction. / Socialand Personal Relationships 12 199-215.

Witmer, D. F. 1997. Communication and recovery: Structuration as anontological approach to organizational climate. Commun. Mono-graphs 64 324-349.

664 ORGANIZATION SCIENCEAôl. 11, No. 6, November-December 2000

Documents

Dyadic Communication Relationships in Organizations: An