Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=misrc.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
Management Information Systems Research Center, University of Minnesota is collaborating with JSTOR to
digitize, preserve and extend access to MIS Quarterly.
http://www.jstor.org
Communicationand IT
Te'eni/Organizational
REVIEW:A COGNITIVE-AFFECTIVEMODEL OF
ORGANIZATIONALCOMMUNICATION FOR
DESIGNINGIT1 2
edge of the mechanisms that guide behavior is of the first two days center on the productionof
used to demonstratethe potentialfor developing two paper products,of which one is red and the
designprinciplesforfuturecommunicationsupport other blue (product #8123). A problem has
systems. occurred and it requires action and communi-
cation to solve it. We shall referto this example
Keywords: Organizational communication, throughoutthe article,butfor now, itwillsufficeto
communication complexity, cognition, affect, note that people have differentcommunication
organizationalmemory,design goals: they may request the next shift to take
action on product#8123, coordinate the team-
ISRLCategories: HA08,AC0401, HA10,AD0518 work,builda relationshipwithanotheremployee,
and motivateworkers. People also choose to use
different media for different goals. Moreover,
people choose differentformsof a message (e.g.,
Introduction the level of formality)butalso make cleveradapta-
tions to given situationsand media. Forexample,
Motivation, Scope and Contribution using a diaryto convey a happy greeting with a
smiling face effectively conveys an emotion
Nowadays, managers have at their disposal a through a medium that is usually expected to
wide varietyof communicationtechnologies from communicateshorttask-orientedmessages. (The
which to choose. A number of recent studies typo-product 1823-is intentionalto demonstrate
have reviewed and extended theories of how later how technology can help reduce errors in
managers choose a mediumfor a specific situa- communication.)
tion.3 Nevertheless, currenttechnology can also
affectwhat we communicate, as wellas how we The model developed below attemptsto explain
communicate it. Thus,the questionfordesigners how people choose the message form and the
has become broader: how should technology be
medium according to goals and situations.
designed to make communicationmore effective
Followingon fromthis model, if correct, it might
by changing not only the medium, but also
then be possible to design a computer-based
attributesof the message itself?
communicationsystem. For example, the diary,
A shortexampleof organizationalcommunication as partof a sophisticatedorganizationalmemory,
can illustratehow we choose the medium and could recognize that Jack is a new workerand
message form. Table 1 shows 10 messages supplement the message of the April3 morning
recorded on three consecutive days. Three shift with additionalcontext informationsuch as
messages are takenfroma diaryinthe production productname (ratherthanjust #8123) and details
room. The other seven use a variety of other about the blue dying color. More generally, a
media: e-mail,face-to-faceprivatemeeting,typed model of effective communication could be
memo, phone call, and voice mail.The messages incorporatedintocommunicationtechnologyso as
to adapt messages. This could be achieved by
automaticallyrecommendingto the sender the
3Webster optimal amount of context informationin the
(1998)providesa comprehensiveoverview of
theoriesthatdescribehowmanagerschoosea medium message.
fora specificsituation.
Amongthesetheoriesarethose
concerning mediarichness(DaftandLengel1984,1986) A modelof user behaviorforguidingdesign needs
andsocialpresence(Shortet al. 1976),theoriesrelated
to media richness(Rice et al. 1989; Straub1994; to satisfyseveral conditions.Itshoulddescribenot
Trevinoet al. 1987),furthertheoriessuch as those only the product, but also the process of com-
relatingto physicalaccessibilityof the mediumor munication,in order to identifyopportunitiesfor
inspace andtime(Markus1994a;Reinsch
availability
and Beswick1990;RiceandAydin1991),andindeed, computersupport.Forexample,the developersof
morerecentsuggestionssuchas a taskclosuremodel Colab (one of the more ambitiouscollaborative
(StraubandKarahanna 1998)anda combinedviewof
utilityandnorms(Kraut et al. 1998). supportsystems) providean insightfulanalysis of
252 MISQuarterly
Vol.25 No.2/June2001
andIT
Communication
Te'eni/Organizational
= 1 I I I~1 sl
- I re
L'4[(fJ~f= IIII C~
I(I4
Y.
I II I I Is E] I l . I II~
II 1I*II[e]mFI
I e imC'4 dI.I r.m~.
S S~~~=II
(9) April,4 (10) Voice mailfromCEOto all employees: Good morning. I want to
remindyou about the Europeanvisit tomorrow.We all need to be at
HappyBirthday,Jack! our best. You must believe me when I say there willbe no layoffs
as a resultof this merger. I have negotiated this issue to the very
(i)
and\ tlastdetailtellingthe newcomers that we have always been family
?-^ and that this is the way we stay!
tThis example follows scenarios of communication found in Saunders and Jones (1990) and Robinson et al. (2000).
MISQuarterly
Vol.25 No.2/June2001 253
and IT
Communication
Te'eni/Organizational
Within the perspective of choice, action, and Rasmussen 1986). The sources of communica-
relationship,we develop a model that has three tion complexitycan be categorized as cognitive
main factors, each of which includes several complexity, dynamic complexity, and affective
elements (shown in Figure1): complexity.5
II
II
II
II
complexity
Communication
II
*. . _ *0
- I I I
- M . .
A simple task can demonstrate these sources of screams "Right"and she answers slowly "OK,"he
communication complexity. Joe asks Rita to help is not sure that he can rely on her intent to follow
him bring his new boat to harbor through a narrow with appropriate action. This is a state of high
and long water passage. They decide to row the affective complexity.
boat and adjust the boat's course, where
necessary, by rowing faster on one side and The three factors and their elements are mapped
holding the oar steady on the side to which they in Figure 1 and described in Table 2. Later, the
wish to turn. They each take a side of the boat attributes or classes of the model are defined
and row in parallel, both maintaining eye contact more precisely. The paper proceeds as a journey
and each deciding on the rowing pace by from a more abstract discussion of communication
estimating the leeway between the boat and the goals rooted in philosophy, through to an analysis
bank. Now suppose that Joe cannot see what is of cognitive and affective strategies built on
happening on Rita's side and vice versa. They behavioral sciences, to the more concrete design
need to communicate continuously to inform one
implications with regard to information and com-
another of the leeway on each side in order to munication technologies. Figure 1 serves as a
work harmoniously. The intensity of the com-
map to keep on track. The criterion for choosing
munication is higher because Joe must rely on
the landmarks for the central path concerns how
information from Rita before he can act. Cognitive
to best uncover the process of communication so
complexity is, therefore, higher than in the that others can forge new paths along similar lines
previous scenario. Now let us suppose, further, for new conditions. For example, organizational
that it takes a few seconds to see the effect of
politics, which are left out of the analysis, undoub-
adjusting the speed of rowing on the change of
course. Joe says "Right"but sees no immediate tedly play a role in shaping communication, but
reaction (feedback) and is left uncertain as to they too could be investigated in the future, using
whether Rita heard or understood his message. the same rationale developed to analyze the effect
of culture. So while the review of the literature
Dynamic complexity is therefore high and Joe may
consequently ask Rita to confirm by saying "OK" attempts to be comprehensive within the boun
whenever she gets a message. Finally, let us say daries sketched out in Figure 1, the development
that Joe is not sure about how Rita feels toward of the model concentrates on representative
him today and suspects that she may not mind elements. The choices of these elements are
terribly if the boat gets scratched. So when he described in more detail later.
Each of the numberedarrowsin Figure 1 repre- cation process can affect it by determining a
sents a generalpropositiondeveloped below.The person's goal prioritiesand setting limitationson
expositionof the modelfollowsFigure1 fromright behavior (propositions 8 through 11). Special
to left, beginning with communication impact attention is devoted in this section to the inter-
(proposition1) in the next section of the paper. culturaleffects on the communicationprocess due
The heart of the discussion is presented in the to the growingimportanceand recent interest in
section that explains the elements of the com- global communication.The final section looks at
municationprocess, summarizesthe principlesof the model's implicationsfor research and design.
behavior assumed in this model, and proposes Ittakes the model's three factors (Figure1) as a
effective combinationsof goal, strategy,medium, specification of requiredfunctionality,adds the
and message attributes(propositions2 through7). notionof organizationalmemory as a necessary
We then examine how inputs into the communi- resource, and postulates some general design
guidelines for future communication support meant by communication. Of the many definitions
systems. of communication, we have sought one which
emphasizes goal driven behavior, which is later
The last introductorynote specifies the level of the shown to be the basis for choosing strategies so
theory (Klein et al. 1994). In Table 1, messages as to impact action and relationship. Gerald R.
1, 2, 3, and 8 are all associated with the task of Miller claims that, "In the main, communication
producing and delivering product #8123 with has its central interest in those behavioral
minimal delay. These four messages may be seen situations in which a source transmits a message
as a higher-level group of individual messages to a receiver(s) with conscious intent to affect the
that are all influenced by certain attributes of the latter's behavior" (1976, p.92). By analyzing the
common task. The model described below is a message, its communicative impact can be
mixed-determinants model in the sense that the assessed to the extent that the sender's intent
elements of a message (e.g., its medium) are (goal) has been understood and accepted once
determined by other elements of the message (its the message has been received. Furthermore, it
goal) but also by elements of the common task is only the immediate impact of the communicative
(e.g., the urgency to complete the task with act on the receiver's desire to react that is
minimal delay). The section on the communication investigated.7
process develops the relationships between ele-
ments of a message, assuming that the messages The model of organizational communication
are independent of group effects. The section on proposed here is a pragmatic theory of a rational
inputs to the communication process introduces communicator who uses resources to implement
the effects of inputs such as task attributes on communication goals. We build on the theory of
messages, assuming that messages are relatively communicative action (Habermas 1984), which
homogeneous with respect to the inputs.6 For similarly situates social communication within a
example, all messages associated with an urgent top-down hierarchy of goals and resources in the
task will usually be communicated by phone or e- context of social norms and cultural values.
mail but not by typed memos. The effects of both However, to be able to move from a theory of
levels may interact. For example, messages that communication to the design of systems that
are not only associated with an urgent task (a support communication, the top-down view must
group level) but also involve persuasion (a be complemented with a bottom-up view of how
message level) will be communicated by phone limited resources are used to achieve goals.
rather than by e-mail more frequently than urgent
messages not involving persuasion.
7An in-depthexaminationof communicationmakes it
necessary to concentrateon the directimpactof com-
municationto keep the discussion manageable. For
example,Computer-Supported CooperativeWorklitera-
Elements of Communication ture usuallytreats the impactof using communication
Impact systems on action and relationship.Hollingsheadand
McGrath(1995) provide several instances of such
impacts. Regarding action, communication affects
(1) the task product,e.g., time, numberand qualityof
The Theory of Communicative Action decision solutions, and (2) the users' reactions, e.g.,
satisfaction and rated effectiveness. Regardingrela-
tionship, communication affects (1) the relations
A definition of communication impact necessarily betweenactors,e.g., attractionand feelingof alienation,
begins with an explanation of what is actually and (2) the patternof interaction,e.g., totalor non-task
amountof participation. This paperexaminesthe impact
of the communicative act on the receiver's under-
standingof the message and on the receiver'srelation-
6We use the terminologyofferedby Kleinet al. (1994). ship withthe sender, butdoes not examineany aspects
In mixed-determinantsmodels, determinantsfrom a of organizationalimpact, such as decision quality.
varietyof levels may influencethe dependentvariables. Ideally, the individual and organizationallevels of
Homogeneitywithrespectto a constructimpliesthatthe analysis can be integrated (O'Reilly et al. 1987).
group members' values on a given construct are However,for practicalreasons, this paper is limitedin
sufficientlysimilarso that it is meaningfulto talkabout scope and assumes thatbettercommunicationwilllead
the groupconstruct. to betterdecision quality.
These two views explain, respectively,what and senders' obligationto providethe reasoning for
howpeople communicate,and,when examinedin the validityof theirclaims, and furthermore,these
the context of a particularsituation,they provide claims may be rooted in or regulated by the life-
the necessary knowledgefordesign. Forexample, world shared by the communicators.The com-
if we assume that senders exercise more control plexity of implementing communicative action
over interpersonalmessages when they feel the grows with the need for coordinationand type of
receiver'sreactionis less predictable,a designer action orientation (goals, norms, and values).
who knows this may build into the system both These twocharacteristicsare shown, respectively,
highand low levels of control(e.g., instantversus as a top-downand bottom-uparrowin Figure 2.
delayed feedback). Moreover,the system could The distinctionbetween goals and resources is
be designed to set the default level of control common in social (and organizational)analysis,
according to the level of uncertaintyabout the but the difference between norms and values
receiver'sreaction(e.g., accordingto how closely requires explanation. Norms are the orders of
the sender and receiver have workedtogether). interpersonalrelationshipsaccepted bysociety for
But such design decisions rely on knowledgeof the purposes of regulatingpractices and habitual
how people communicateand, therefore,extends behavior.Normsare expected to be validatednot
the scope of the theoryof communicativeaction. only against standards of rationalitybut also
Belowwe simplifythe frameworkused by Haber- against standards of relationships. In contrast,
mas and argue whythis frameworkis appropriate engrained in the cultureare values of objects of
for a model of organizationalcommunication, utilitythat are not usuallyputto tests of validity.
whilewe also indicatein general termswhat must
be adapted. To adopt communicativeactionfororganizational
communication,we evaluate whetherthe system
Habermas (1998) developed his theory of com- depictedin Figure2 is appropriateinthe organiza-
municative action as an element of a more tional setting. We regard an organizationas a
general theoryof society. social system that can be characterized by
resources, goals, norms, and culture.Moreover,
[The] concept of communicativeaction we assume thatorganizationsrelyon coordinated
refers to the interactionof at least two action and, indeed, act undernormsof rationality
subjects capable of speech and action in the sense that actors of the organizationare
whoestablishinterpersonalrelationships. guided by theirgoals, norms and cultureon how
The actors seek to reach an under- to act in order to produce desired outcomes
standing about the action situationand (Thompson 1967). Communicationcomplexity
their plans of action in order to coordi- has already been discussed. In other words,
nate their action by way of agreement looking at Figure 2, one can substitute"person"
(Habermas1984, p. 86). withan actor in an organization,replace "society"
with organization,and consider "culture'"asboth
Communicativeaction(a behavior)takes place in organizationaland nationalculture.Finally,in this
relation to three additional factors-culture, organizational setting, we take communicative
society, and person-that together constitutethe action to be an ideal formof organizationalcom-
"lifeworld,"which serves as the context for munication.Indeed,several studies of information
communication.These fourfactors are organized systems have recently used the theory of com-
in Figure2, alongside the correspondingproducts municative action to understand organizational
by which action is oriented: resources, values, communication(e.g., Ngwenyamaand Lee 1997;
norms, and goals. In order for goals to be Ngwenyamaand Lyytinen1997).8
achieved,coordinationbetween communicatorsis
necessary, as is the commitmentof communica-
tors to behave in certainways, whichis assumed 8Recentstudies have used Searle's(1969)theoryof
to be partof establishingrelationships. speech acts (e.g., Moore 1998; but see criticism by
Janson and Woo 1996), Habermas's(1984, 1987)
Habermasfurthertalks about two characteristics theoryof communicativeaction(e.g., Mingers1995;
Ngwenyamaand Lee 1997; but see criticismby
underlyingcommunicativeaction: rationalityand Sharrockand Button1997),anda combinationof both
complexity.Communicativeactionis based on the SearleandHabermas(e.g., Schoop1997).
Person
Goals
Culture A m Society
Values o ,A Norms
0)
E
o
Behavior
Resources
*-IsIIIIwM?u~1Iuu1oJIIirIsa'1[4'YKo1m~uIagII4'~'Ie]i[eflI mU.WIo]uul
I ilUII1r*1~I'LN?Te~4[o]pUe1VEi11 .I~1i..I*
The frameworkin Figure 2 must, however, be lifeworldthat is delimitedin relationto goals, but
adapted to formthe basis for design. First,while there is littleconcernwithhowthe situationaffects
the theory of communicative action has little behavior.Therefore,inourmodel,situationwillbe
connection with the physical aspects of com- characterized more specifically as task and
munication, a model leading to design must sender-receiverdistanceto demonstratehowthey
address the interactionbetween the communi- affect behavior.
cator and the media. The levels of context (in
Figure 2) are necessary but not sufficient for
designing informationtechnology; they must be Communication Impact Defined
complementedwithlowerlevelsof abstractionthat as Mutual Understanding and
describe the physical functions and form of
communication(Rasmussen 1986). Therefore,in Relationship
our model, behaviorwill be described in greater The first implicationof the theory of communi-
detailas communicationstrategies, medium,and cative action is a definitionof communication
message. Second, Habermas'categories of goals impact. Habermas(1984, 1987) claims that four
and actions, which he sees as universal,may be conditionsare necessary fora communicativeact
inadequate to capture the idiosyncrasies of a to take place:
specific settingsuch as an organization(Sharrock
and Button 1997). Therefore,we use the prin- (1) the act must be comprehensible,so that the
ciples of communicativeaction, but do not adopt receiver can understandthe sender;
the detailed classifications.Third,in the theoryof
communicativeaction, a situation represents a (2) the act must be true, so thatthe receivercan
temporallyand spatiallydefined segment of the share the sender's knowledge;
(3) intentions must be expressed truthfully,so receiver to purchase the product, then the
that the receivercan trustthe sender; and communicationmust notonly be comprehensible,
but also appropriate.
(4) the act must be appropriatewithin some
normativecontext so that the receiver can Relationship and mutual understandingare, of
agree with the sender within this value course, closely interrelated.Indeed, work rela-
system. tionships, and more specifically, mutual trust,
facilitate a more productiveflow of information
Communicationis at once an act of buildinga (Hartand Saunders 1997; Nelson and Cooprider
mutual understanding between sender and 1996). Moreover,relationshipsbuildtrust,which
receiverand buildinga relationshipbetweenthem. can be described as the confidence that the
The relationshipis necessary for gaining a com- receiverwillfulfillobligationsand behave in a fair
mitmentfrom the receiver to the sender or the and predictable manner (Anderson and Weitz
largergroupto whichthey belong. Infact, it is the 1989). Communicationis, therefore, more effec-
act of communicationmorethanthe informational tive when trust and commitmentare high (e.g.,
contentthatproducesthis commitment(Huffet al. Dore 1983; Williamson1975). Explanationsper-
1989). Althoughsome may regardthe thirdcon- ceived to be timely and sufficiently detailed to
ditionto be somewhat naive, truthfulexpressions allowforadequate understandingof the message
of intentions are the basis for the commitment are more likelyto lead to trust (Whitener et al.
necessary in social exchange (Blau 1964). 1998). In fact, any communication between
managers and subordinates that appears to be
These four conditions of validity allow us to open builds trust (Gabarro1978). On the other
develop two mirror perspectives: (1) defining hand, faulty communication and unsuccessful
mutualunderstandingand relationshipas the im- interactionmake it impossible to reduce psycho-
pact of successful communicationand (2) charac- logical distances between people (Schein 1996).
terizingimpedimentsto actionandto relationships Thus, mutualunderstandingand relationshipare
as poor communication.While Habermas com- intimatelyrelated.
bines understandingand relationship,we separate
the two, acknowledgingthat they are interwoven The four conditions of communicationinvalidity
yet assumingthatpeople can distinguishbetween also allow us to investigate poor communication
the two. Mutual understanding refers to the first as an impedimentto actionand relationship.Such
and second conditions (a comprehensible and a perspective is useful for two reasons: in prac-
true message) and is associated withthe action- tical terms, failures in communicationare often
orientedaspect of communication.Mutualunder- easier to measure than successful communi-
standing includes not only the receiver's under- cation, while conceptually,a focus on communi-
standing of the message, but also the sender's cation invaliditycan serve to explainbehaviorthat
awareness that the recipientof the message has attempts to overcome impedimentsto action. In
understoodit. Relationshiprefersto the thirdand linewiththe actionperspectiveadopted here, poor
fourth conditions (trustworthinessand appro- communicationcan be defined as an impediment
priateness)and is associated withthe dimension to action, that is to say, any exchange of infor-
of relationships within the communicative act mationthat leaves the receiver unable, unwilling,
(Habermas 1984). Successful communication or unsure of how to proceed with the sender's
necessitates both aspects, althoughthe marginal intended impact.Takinga similar approach, De
impact of a single communicativeact on rela- Bono (1976) considers thinkingfromthe point of
tionship may be smaller than that on mutual viewof actionand defines an effective explanation
understandingand may depend on the precise as one that allows an individualto decide on what
communicationgoal. Forexample, ifthe sender's to do next.
goal is to convey the price of a product, then
mutualunderstandingis the desired impact. On Highlevels of communicationcomplexitycan lead
the other hand, if the goal is to influence the to communicationfailures. Cognitive complexity
.j *J[U.J [e1.j.j iYZe] ii .13 I(*U IuE II3o] ml[*4E Le] U d IeI*I.
amimmill
264 MISQuarterly
Vol.25 No.2/June2001
Communicationand IT
Te'eni/Organizational
1161 I?X inei liii 111111[*I [.1 mUWIell IY1 ml.Eo] I 141 'I'] ml.liii' RTe1I IIOI4-Ee Eo] 'liii [I
notationof contextualizationis narrower:it seeks the process, and (2) testing and adjustingon the
to elucidate the situation in which the message basis of feedback during the process (online).
was created, detailing such issues as who is Indeed, people are capable of recognizing and
communicatingwithwhom,when, and underwhat adoptingspontaneous versus plannedcommuni-
conditions(Schwartzand Te'eni2000). Message cation, depending on whether the interdepen-
#3 inTable 1 is an instructionto postpone delivery dence betweenthem is parallelor sequential(Lea
(the core of the message) and message #2 (which 1991).
is threaded) gives the reason for the instruction
(thecontext).Additionalcontextinformationabout For planned control, one needs to consider
the message creation is the sender (contract whether the communication process is pre-
manager),receiver(logistics) and date (April2). determined,leaving the locus of controlwiththe
sender, or whetherit is flexible,leaving open the
progressionof communicationand lettingcontrol
Affectivity shift from one partnerto another. A particular
Affectivitymay be seen to be the inclusion of characteristicof controlthrough planningis the
affective components in the message that des- clear designation of who does what in the com-
cribe emotions and moods, not necessarily municationprocess and a distinctionbetweenplan
pleasant ones (Schwarz 1990). Emotions are and implementation.Inplannedcontrol,message
more intense, relativelyshort-lived,and usually redundancy,especially repetitionof key ideas, is
promptedby a clear trigger,such as excitement used to ensure successful communication(Mayer
about the prospects of success, an apology, and 1985). For example, senders sometimes copy
the pleasure of meeting someone. Moods, such messages to other people withthe sole intention
as the state of feeling good, are rather longer of pressuringthe receiverto take action (Philips
term affective states, usually with no salient and Eisenberg 1993). Additionally,the sender
cause. Schlosberg (1952) has mapped affect may send the same message several times,
accordingto two dimensions: attention-rejection rephrase messages, and summarize previous
and pleasantness-unpleasantness. Thiswas later communication.Importantcharacteristicsof this
reconfirmedfor communicationthat is both non- control behaviorare, therefore,redundancyand
verbal(Greenand Cliff1975) and verbal(Osgood repeated communication ("I'm sending this
1969). Affectivitycan be used to motivate,e.g., to message again to yourotheraddress,""Attached
sustain favorableattitudesand dispositions,and is a summary of our phone conversation")and
to inform, e.g., provide informationabout the multiplerecipients(includingmultiplecopies and
subjectiveevaluationof a product.Indeed,such a blindcopies).
quality is needed to cope with potential com-
municationproblemsdue to affectivecomplexity. Inseeking to achieve controlthroughtesting and
adjusting, the sender plays an active part in
ensuring that the process works well. Timely
Control by Testing and feedback is, of course, essential for effective
Control by Planning control (Te'eni 1992). For example, the sender
Controlis largely a matterof overseeing and, if
repeatedlyasks the receiverifthe communication
necessary, adjustingthe communicationprocess was successful and adjusts the message
to assure effective communication.Forexample,
accordingto the receiver'sreaction.Characteristic
Street and Cappella (1985) note the need for of this control behavior are online tests of the
maintaining coherence indiscourseand managing communicationsuch as "Didyou get the mes-
dominance and control,and Clarkand Brennan
sage?" and "Doyou understandthe message?"
(1991) emphasize the continuous need to coor-
dinate content and process. Moreover,different
levels of controlare needed for differenttypes of Perspective Taking
goals (Jordan1998). Controlcan take two basic Perspective takingis concernedwithwhetherthe
forms:(1) planningthe patternof communication, receiver's view and attitudeare a target of the
and if necessary contingency patterns,ahead of communicationor whether they are left outside
Goals Strategies
r --------------------I
Instruct action
Manage
interdependent action :\
Manage relationship
Influence
: - - -- - - - -- - - -
IT-orDm l(a2ty
0i *o 0w e
. *
gimTT1
* -
-.MI
l- -
i
Al- A71 Rui
Medium
Channel capacity
Strategie ^S 3 Interactivity
Adaptiveness
Control-testing and -
(ognitive complexity
adjusting
Control--planning 4/
Contextualization <
,
Dynamic complexity 7
Perspective taking
Affectivity
Affective comnplexity
Attention focusing
Size
Distribution
Organization
Formality
Message form
S - -s - A gA - A - -eg -e -
Medium
Attribute Action-Oriented Impact Relationship-Oriented Impact
Interactivity Immediatefeedback improvesunder- is importantfor affect in
Interactivity
standing(Clark1992; Dennis and Kinney CMC(Kiesleret al. 1985).
1998). Immediatefeedback speeds com-
munication,which in turn,improvesunder-
standing(Clarkand Brennan1991; Walther
1992). CMCwithblockedconcurrentinput
produces less non-redundantcommuni-
cation (Valacichet al. 1993).
Channel Multiplecues can improvebut also hinder Mixedresults on whether multiple
capacity understanding(Dennis and Kinney1998). cues seem less or more friendly
Higherchannel capacitycan speed but also (Fulkand Collins-Jarvisin press;
slow down communication(Chapanis 1988; Walther1992, 1995). Lowcapacity
Sproulland Kiesler1992). Highercapacity channels reduce social cues (Sproull
reduces explicitcontrol(Krautet al. 1998). and Kiesler1992) but not if com-
Video conferencingproduces more aware- municatorssense a social identity
ness and conversationalfluency than voice withthe communicatingparties (Lea
alone (Tangand Isaacs 1992), particularly and Spears 1991; Spears et al.
in largergroups (Daly-Joneset al. 1998). In 1990). Video conferencing is
comparisonto audio-only,video has no effective in promotingsocial activity
effect on mutualunderstanding(Gale 1990) (Fish et al. 1993).
or some improvementbut less than FtF, Video vs. FtFshows no effect on
when it is high quality(Doherty-Sneddonet initialtrust(Muhlfelderet al. 1999).
al. 1997).
Adaptiveness None found. Voice-mailseems more personal
than e-mail (Adamset al. 1993).
Combined FtF produces bettermutualunderstanding ComparedwithCMC,FtF is rated
(FtFversus than CMC(Strausand McGrath1994) and less relationshiporientedand less
CMC) only so for preferencetasks (Tan et al. expressive of affect (Hollingsheadet
1999). Mixedresults:FtF produces more al. 1993; Lea and Spears 1991;
validand novel argumentsthan CMC Walther1995). ComparedwithCMC,
(Kiesleret al. 1985; McGuireet al. 1987) FtF is rated less relationshiporiented
and equally validarguments(EI-Shinnawy but produces more total communi-
and Vinze 1998). CMCcauses information cation (Hiltzet al. 1986). However,
suppression (Hollingshead1996). CMC Siegel et al. (1986) found them to be
generates less communicationthan FtF in equallytask oriented. Use of e-mail
hierarchicalteams (Hedlundet al. 1998; decreases communicationand
Hightowerand Sayeed 1996). CMCgene- amount of greetings (Sarbaugh-
rates more productivebrainstormingby Thompsonand Feldman 1998).
reducingproductionblocking(Gallupeet al. CMCproduces less trustthan FtF
1994; Valacichet al. 1994) but also less (Rocco 1998) and poorersocial life
productivebrainstormingby increasingdis- (Markus1994b) and closer ties with-
tractionand complexity(Pinsonneaultet.al. in coalitionsbut more social unrest
1999). CMCgenerates more communica- overall(Rommand Pliskin1998).
tion in organizations(Schultze and Vanden- CMCover time increased social
bosch 1998). CMCgenerates more com- orientation,trust,and informality
municationbut processed inaccurately (Waltherand Burgoon1992) but
(Dennis 1996). CMCgenerates more biased slower than FtF (Chidambaram
discussions, especially when information 1996).
load is high (Hightowerand Sayeed 1995). A synchronousweb-base conference
Technology-performancerelationship produces less relationallinksthan
depends more on experience than type of FtF (Warkentinet al. 1997).
task (Hollingsheadet al. 1993).
of dynamic complexitybecause of its potentially cation often requires a wide varietyof signs that
spontaneous, unpredictableprogressionof com- can be transmitted only on channels of high
munication and the possibility of interruptions capacity (e.g., Carnevale and Isen 1986). The
(e.g., "chat" is more interactive and more media richness theory asserts that high channel
unpredictablethan asynchronouse-mail). Hence, capacity,e.g., face-to-face oralcommunication,is
unless used only when needed, interactivitymay necessary to enable social cues such as facial
prove to be a liability. expressions, body language, and tone of voice
that are absent in writtencommunicationor CMC
Proposition 3A: For controlby testing and such as e-mail. However, high channel capacity
adjusting,high,ratherthanlow,interactivity media have only the potentialfor enrichingcom-
is more effective. munication.They cannot ensure a richer com-
municativeact in reality (Ngwenyama and Lee
Although it may be hard to show when high 1997). Thereis, nevertheless, some evidence that
channelcapacityis counterproductive (Rice 1992; high channel capacity is perceived to be more
Tan et al. 1999b), it is possible to determinefor effective and more appropriatefor affectivecom-
which strategies it is especially useful. Contex- munication,primarilybecause of the complexityof
tualizationdeals with high cognitive complexity feelings and importanceof non-verbalmessages
throughthe provisionof multiplelayersof context, (Sproulland Kiesler1992; Westmyeret al. 1998).
multipleviews, and, in general, moretask-related Itis interestingto note thatthe use of audio-visual,
informationthan communicationwithoutcontex- in comparisonto audio only, has been shown to
tualization.It necessarily follows that such com- be effective in terms of informalcommunication,
munication requires greater channel capacity. which often includes social information(Blyet al.
Indeed, in this regard, media richness theory 1993; Fish et al. 1993).
predicts that communicationaimed at resolving
ambiguity and explaining interpretations will Proposition 3C: For affectivity,high, rather
requireinteractivemedia and high channel capa- than low, channel capacity is more effec-
city media (Daft and Lengel 1984). Furthermore, tive.
in a survey of a large petrochemicalcompany,
Russ et al. (1990) found that managers select Adaptiveness is necessary in strategies that
high channel capacity media for equivocal mes- attempt to tailor the message to a personal
sages and low channel capacity media for less perspective.Forexample, when comparedwithe-
equivocal messages, as did Daft and Lengel mail (higheradaptiveness) and face-to-face con-
(1986; Daft et al. 1987). Takinga differentpers- versation(highest adaptiveness), a bulletinboard
pective of video versus audio, Whittaker(1995) is unable to supporteffectivelyany formof adap-
sees video as providingdata aboutthe objects of tiveness. In practicalterms, it is sometimes hard
discussion rather than adding non-verbal cues to separate adaptiveness fromchannel capacity.
about the communicators, but nevertheless In a simple case, when the sender refers in the
regardsthis additionalchannel capacity as a tool message to the receiver's view too, then the
for buildinga shared context. channel should have the capacity to transmit
multipleviews. Ina more realisticcase, perspec-
Proposition 3B: Forcontextualization,high, tive takingincludes affective as well as cognitive
ratherthan low, channel capacity is more references to the receiver'sworld.Insuch cases,
effective. the channel must also have the capacityto sup-
port the personal touch by tone, pronunciation,
Similarly, affectivity copes with high affective and other non-verbalgestures. A useful example
complexityby adding emotions and dispositions is voice-mail,whicheliminatesthe synchronicityof
into the message. In comparison to cognitive a telephone conversationbut includesits channel
strategies, though,affectivityis relativelysensitive capacity and adaptiveness. E-mail,in compari-
to how the affect is transmittedand received son, can be adapted to the receiverbut has less
(Wallbottand Scherer 1986). Such communi- channel capacity. Adams et al. (1993) have
comparedthe two media and found voice-mailto channel capacity (Doherty-Sneddonet al. 1997).
be more personal than e-mail accordingto most For example, field workers used more explicit
people. Quitepossibly,this may be because most controlwhen channel capacitywas reduced from
people cannot separate adaptiveness from the video to audio (Krautet al. 1998). When inter-
capacityto transmita varietyof social cues. activityis low, however, senders may choose to
cope withreducedchannel capacityby increasing
Perspective takingusuallyrequiresthe sender to control through planning, e.g., by explaining
understand new viewpoints and adapt the procedures in different ways to increase the
message accordingly (Goldberg 1990). It also probabilitythat one of them willbe understood.
requiresthe sender to adaptthe message to make
itmorepersonal.Incontrast,managersconveying Proposition 4A: Senders will adapt to low
to a group the formalstructureof authorityand channel capacity coupled with high
code of behavior (e.g., setting procedures and interactivityby increasing control through
roles)willusuallyselect a writtenratherthanface- testing and adjusting.
to-face mode of communication.
Proposition 4B: Senders will adapt to low
Proposition 3D: For perspective taking, channel capacity coupled with low inter-
high,ratherthan low,adaptiveness is more activity by increasing control through
effective. planning.
that characterizes the configurationand style of The degree of message organization may be
the informationcommunicated.In contrastto the defined as the extent to which the message is
choice of media, the choice of the message form systematicallyordered to support mutualunder-
has received little attention in information systems standing, by explaining how message can be
research. Research in communication theory has understood (the word "structure"is avoided
also neglected it as evident by the call to "devote because it is usually associated with the task or
more energy than they have in the past to a close process). For example, the next paragraph
(startingwith "A highly organized message") is
study of messages themselves" (Stohl and divided into four ordered dimensions, creating a
Redding 1987, p. 494). Past characterizations of clear two-level hierarchy in the message that
message and the more recent work in CMC guides the readerfrom upperto lower level, and
suggest four attributes of a message: size, dis- proceeds from the first to the fourth item
tribution, degree of organization (structure), and sequentially.
degree of formality. Here, these attributes are
redefined when necessary to ensure they do not A highly organized message may, therefore, be
overlap with attributes of the medium.14 characterized in terms of several, but not
necessarily all, of the followingdimensions:
Message size is a function of the number of
semantic units such as words or sentences (Daft (1) An obvious set of ordered and clearly
and Lengel 1986). Distribution is the number of distinguished elements that can easily be
destinations to which the message is sent. Size differentiatedand discriminated(Schroderet
and distribution are the most popular measures of al. 1967) (e.g., paragraphswith an opening
communication in CMC (Rudy 1996), perhaps
that indicates the theme or sections with
subtitles or numbering);
because they are relatively easy to retrieve from
system logs with little need to code the material
(2) A clear allocation of tasks between sender
manually. and receiver so that the latter can imme-
diately understandthe action required(e.g.,
the sender provides informationand the
receiver is expected to take action).
14Formalityis an essential characteristic of com-
munication,but has received only casual attentionin (3) A clear structureof and access to different
empiricalresearch(Stohland Redding1987). Influential levels of contextto easily grasp and navigate
sources on formalityare Downs(1967) and Melcherand the macrostructure(van Dijk and Kintsch
Beller(1967). Both sources characterizeformalityas a
function of the organizationalsetting, namely the 1983) (e.g., explanations as footnotes,
capacity in which the sender acts. Stohl and Redding references to documents that provide more
compileda list of message classificationschemes, but details or a more complete rationale,hyper-
again, most of these schemes incorporateother con- text style access to more details);
structsof ourmodel (goals and inputfactors).However,
characterizationsof messages that do not clearly
distinguishbetween message and medium would be (4) A familiaror standardformatfor immediate
inappropriate for a model (such as ours) thatdoes. For recognition to avoid searching or learning
example,informationrichness has been definedas "the (Berlyne1960) (e.g., each of the diaryentries
abilityof informationto change understandingwithina inTable 1 begins withdate, shift,and author).
time interval. Communicationtransactions that can
overcome different frames of reference or clarify
ambiguousissues to change understandingin a timely The fourthattributeof message is formality,which
mannerare considered rich"(Daftand Lengel 1986, p. denotes interactiveclosuretowardsome organiza-
560). Incontrast,we seek constructsthatare evidentin tionallyaccepted representationof action(ledema
the message itself, ratherthan which remaina judge-
ment of its abilityto achieve a goal and, consequently, 1999).15 One scenario of such closure is the
blurthe distinctionbetween medium.Clearly,this issue
deserves more attention. Forexample, fromthe social
influence perspective (Fulket al. 1990) and symbolic
meaning perspectives (Sitkinet al. 1992; Trevinoet al. 15Thisdefinitiondoes not preculdeformalitiesdictated
1987), formalitycan be conveyed by the choice of by more general norms outside the organization(e.g.,
media. It may be that a separate constructof formality language), but for simplicitythey are subsumed under
could be defined in the contextof the mediumtoo. "organizationallyaccepted representationof action."
Proposition 5E: For control, high, rather sitions, which are more abstract and phrased in
than low,formalityis usuallymoreeffective. the accepted terminologyof the model, are of
greaterformalitythantheirprecedingparagraphs,
Proposition 5F: For contextualization,low, which should be more concrete and possibly
ratherthan high, formalityis usually more includespecific examples. One of the implications
effective. for design (discussed below) is that these shifts
fromone level to anothershould be supportedby
higher message organization but also with
Proposition 6: Adaptations to correspondingchanges of formality.
Non-compatible Message Forms
Havinglooked at how strategies dictatemessage
attributes, we now consider how message
attributes affect strategy selection. One direct Links Between Media and
result of larger messages is higher cognitive Message Form
complexitydue to informationoverload,whichcan
be reduced by attentionfocusing. Choice mechanisms involvingadmissibilityand
profitabilitycan also explain the interactions
Proposition 6A: Senders willadaptto a long between medium and message form. For
message by increasingattentionfocusing. example,e-mailfacilitatesan increase in message
distributionand a commensuratedecrease in dis-
Poor message organization also increases tributioncosts, and this makes distributionmore
cognitivecomplexity.Controlthroughtesting and attractiveto the user. Indeed, past research has
adjustingis a common reactionto a disorganized shown that CMC increases the distributionof
message, but can be accomplished only when messages (Palme 1985; Phillipsand Eisenberg
using interactivemedia.Forexample,insituations 1993; Sproull and Kiesler 1992). Similarly,
of stress, messages are often disorganizedand increases the time-relatedcost of long
interactivity
continuous control with feedback is usually the messages because of the online nature of the
only way to cope withthe complexity. dialog but not necessarily the velocity of mes-
sages (Jones et al. 1993). Infact, there is some
Proposition 6B: Senders will adapt to low evidence thatthis is what happens withthe use of
message organizationbyincreasingcontrol CMC(Trevinoet al. 1987). Social normsof using
through testing and adjusting, provided certaininteractivemedia, such as smalltalkat the
media interactivityis high. beginning of a face-to-face meeting or phone
conversation,may, however,moderatethis effect.
*6S~~~Ti
0. .301 i -i.ki ..-?=T1~f 0-T 30 .'0'~rr ~ T~~1
-g.m~
I1
Distance 9
Cognitive distance '?
Affective distance
LZi Dynamic complexi Strategies
Values i
Independence - ,/
Affective complex)
Interdependence
g - . 0
PTAiM- ii u tom10 -11
leagues began to use CMCbut also directlyby decision output(orgroup products)and qualityof
making CMC an acceptable norm of com- communicationsuch as accuracy and conver-
munication.Furthermore,in additionto the two gence (Smith and Vanecek 1990). Table 9 is
routes in which inputsaffect the process (shown limitedto recentstudies thatclearlylinkthe taskto
in Figure 5), inputs may also determine the the quality of communication. More general
feasibility of certain media, e.g., a physical reviewsof communicationmedia,task, and group
distance precludes an immediate face-to-face performancecan be found in Hwang(1998) and
meeting. Straus and McGrath(1994).
=II
1. I r.;t= I I II 11 I I I I I I I I II rn:l Ir
Situational Relationship-
Attributes Use and Action-Oriented Impact Oriented Impact
Task Immediatefeedback and multiplecues improveunder- None found.
analyzability standingparticularly for equivocaltasks (Dennis and Kinney
1998; Straus and McGrath 1994) but also asynchronous
communicationimprovesidea exchange in less equivocal
tasks (Shiraniet al. 1999). CMCimprovescommunication
only for low interdependencetasks (meta-analysisby Hwang
1998; Daly 1993; EI-Shinnawyand Vinze 1998). Group
supportsystems withFtFoutperformedthose without,only
for highlyequivocaltasks (Tanet al. 1999a) and reduced
differentials(Tanet al. 1999b). Richermedia are preferred
overallbut not as a functionof analyzability(D'Ambraet al.
1998). Video conference chosen for routinetasks (Webster
1998).
Task variety Varietyrequiresmore information(Daftand Macintosh1981). None found.
Use of inter-organizational
e-mail grows withuncertainty
(Kettingerand Grover1997).
Task Communicationundertime pressure is faulty: none found. Relationalcommuni-
temporality Use of phone relativeto e-mail increase with urgency cation is unsuccess-
(Wijayanayakeand Higa 1999). ful in short time
Communicationpatternschange over time vis-a-vis the task (Walther1995).
(Jones et al. 1994), and these temporaleffects interactwith
the type of medium(Saunders and Miranda1998) and with
communicationgenre (Orlikowskiand Yates 1994). The
perceivedappropriatenessof media grows withexperience
(Kingand Xia 1997).
information(whichwas defined as a special case In fact, people are reluctantto invest this setup
of "instructing action" goal) because more cost, often avoidingcommunicationor misunder-
informationis needed to describe highervariation standing. Further, communication over long
among instances. This is consistent with media periods of time may spiral out of controlunless
richness theory, which claims that high task effortis made to manage the communicationover
variety requires larger amounts of information time, and individualsremember when to initiate
(Daftand Lengel 1986). communication,rememberto respond,anddetect
problems in communicationwhere feedback is
Proposition 8A: Higher task variety often irregularand delayed. Saunders and Jones
increases the frequency of requesting (1990) propose a model of informationacquisition
information. indecision tasks inwhichcommunicationpatterns
change during the life cycle of the task. For
The effects on strategies is explainedthroughthe example, communicationabout the task may be
test inwhichcomplexityis the primary
profitability intense at some initialstage, stop fora while,and
cost and accurate goal achievement is the then resume sporadically,and not be sufficiently
effectiveness (Beachand Mitchell1987). Lowtask salient to regain the communicator'sattention.
analyzabilityincreases the probabilityof misunder- These changes of patternsare difficultto control.
standing how to proceed with action (Daft and Thus, in extremelylongtime spans, bothcognitive
Lengel 1984). This increases the cognitive com- complexity and dynamic complexity are high.
plexity of the communication because of the Typicalexamples are organizationalprocedures,
ambiguity and of
multiplicity meanings, which in which requirea careful planningof the message
turn increases the benefits of contextualization and what mistakes may arise as a resultof time
(Bolandet al. 1994; Gumperz1982). This too is in changes (similarto temporaldistances between
line withmedia richnesstheory,whichclaims that communicatorsdiscussed below). Furthermore,
lowertask analyzabilityrequiresricherinformation as control by planningis associated with higher
(Daftand Lengel 1986). message organization,this wouldalso explainthe
typicallyhigh organizationof procedures.
Proposition 8B: Lower task analyzability
increases the use of contextualization. Proposition 8C: A short time-to-complete
the task increases the use of control by
Of all temporalattributes,we concentrate on the testing and adjusting.
one most directly associated with the model,
namely, the time availableto complete the task. Proposition 8D: A very long time-to-
The relationshipbetween time-to-completeand complete the task increases the use of
communicationcomplexityis curve-linear:com- controlby planning.
municationin either very short or very long time
spans is more difficultthan in intermediatetime Futureworkcan develop more complex proposi-
spans. Undertime pressure, communicationwill tions about the interactionbetween task analyz-
be stressful as cognitivedemands exceed cogni- abilityand communicationgoals. When the goal
tive resources and feedback (which further of communicationis to influence or to manage
consumes cognitiveresources) may be infeasible interdependentaction for tasks of low analyz-
or uncleardue to time constraintsand, therefore, ability,the result is high cognitive and dynamic
dynamiccomplexitywillbe high. complexity(see the earlier section on assumed
principlesof behavior).Thus the interactionof the
On the other hand, when tasks stretch over long two sources of complexityproduces highercom-
periods of time, communicationis often out of plexityand, as a result, a higher chance of mis-
contextand out of control.To the receiver,a mes- understanding.This will result in a more intense
sage relatingto historicalevents will usually be use of communicationstrategies to cope withthe
out of context, makingit difficultto comprehend complexityand therebyplace additionaldemands
unless effortis made to get back intothe problem. on medium and message form. For example,
Sender-Receiver
Distance Use and Task-Oriented Impact Relationship-Oriented Impact
Sender-receiver Distance reduces amountof communica- Distance reduces non-task
physicaldistance: tion but moderatedby CMCavailability relatedcommunication
Space (Sarbaugh-Thompsonand Feldman1998; (Sarbaugh-Thompsonand
Sproulland Kiesler1991). However, Feldman1998).
distance has no effect (Valacichet al. CMCreduces impactof
1993). Distance is one of three major distance on buildingrela-
determinantsof media choice (Caldwellet tionships(reviewedin
al. 1995). McKennaand Bargh2000).
Distance affects media choice (Reinsch
and Beswick 1990; Webster and Trevino
1995).
Time CMCis particularlyuseful to communicate
between shiftworkers(Huffet al. 1989).
Temporalunavailabilityleads to less inter-
active media to promotetask closure
(Strauband Karahanna1998).
Sender-receiver Directionof communicationaffects media An awareness of sender-
psychological choice (Zmudet al. 1990). receiverrelationsis essential
distance: Richest exchange of informationis for successful relationships
organizational between supervisorand subordinates (Gabarro1990). Shared
(Allen and Griffeth1997). Rich media was knowledgemediates effect of
believed to improvecomprehensionwhen mutualtruston performance
seeking informationfrom outside the (Nelson and Cooprider1996).
organization(Lee and Heath 1999).
Conflictingresults on preferencefor e-mail
vs. phone as a functionof distance
(Wijayanayakeand Higa 1999). E-mailis
preferredwhen effective distance is high
(Markus1994b). Less shared information
leads to higherrate of relevantinformation
(Hightowerand Sayeed 1995; Stasser and
Titus 1987).t
Sender-receiver Incompatibleculturalpatternsof sharing None found.
psychological informationlead to less effective communi-
distance: cation (Brettand Okumura1998; Ohbuchi
intercultural and Takahashi1994). No supportfor effect
of intercultural
distance on frequencyof
informationseeking (reviewedin Berger
and Gudykunst1991).
tOlderbutunchallenged
observation:
direction
ofcommunication
affectsamountofinformation
transmitted and
(O'Reilly
Roberts 1974).
Specific
Proposition Input Component of Process Affected
Task 8A Highertask variety Morefrequentrequests of information
8B Lowertask analyzability Highercontextualization
8C Shorttime to complete task Highercontrolby testing and adjusting
8D Verylong time to complete task Highercontrolby planning
Distance 9A Greatercognitivedistance Highercontextualization
9B Greatercognitivedistance Morefrequentrequests of information
9C Greateraffectivedistance Less frequentrequests of information
Values 10A Greaterinterdependence Morefrequentmanagingrelationships
10B Greaterinterdependence Moreperspective taking
been found. This absence may be connected with designed to achieve goals determined by the
the widespread decrease in research into receiverand it is necessary to examine the same
individual differences in information systems strategy fromthe receiver's viewpoint,e.g., how
duringthe 1980s, aftera long stream of research does the receiverchoose to respond to a request
(Huber1983). Nevertheless, cognitivestyles that for information.
dictate communicationdetermine, by definition,
differentpreferences and capabilitiesfor different
A related perspective is that of privacy,which is
communicationstrategies.Thiswholearea seems
concerned with the rightof individualsto deter-
an untappedavenue forfutureresearchthatmay,
afterall, lead to individually
tailoredsystems. The minewhen, how, and to whatextent informationis
transmitted. Every act of communication dis-
generation of such systems may be especially
relevantto the new virtualorganizationin which closes something of the communicatorand often
the bulkof the communicationcannotrelyon face- this is regarded as a risky act (Goffman 1981).
to-face communication.16 The research directions recommended above
(particularlyrelationalgoals and affectivecontrol)
will demand a better understandingof privacy.
Exploring Other Perspectives: Furthermore, CMC intensifies disclosure by
Receiver, Privacy and Others makingrecordedinformationaccessible, and not
Several otherimportantissues have been omitted always in an obvious way. At a message level,
from the model to keep its complexitymanage- privacylinksdirectlyto the mediumand throughit
able. Perhapsthe most immediateneed is to add to organizational memory (a topic discussed
the receiver's perspective (c.f. Contractorand below). The sender's perceptionof the communi-
Eisenberg 1990; Rudy 1996). Communication cation's confidentialitydepends on media attri-
strategies have been describedfromthe sender's butes (Sillince 1997). It will be important to
perspective,since they have been formulatedas investigate how perceptions of confidentiality
means for achieving goals determined by the affect communicationbehavior,and to linkthem
sender. Yetthe active receiverbecomes a sender back to attributes of the media. For example,
herself the moment she responds, and following when do people compromiseand choose mediaof
this,the same strategiesare employed.Moreover, low confidentiality? Can privacy dictate low
the same strategies can be applied in a similar channel capacity?The privacyperspectivewillbe
fashion to the process of receiving information likelyto become a crucialaspect in understanding
(this assumes thatthe sender and receivershare organizationalcommunication,particularlyas the
the same communicationgoals). For example, boundaries between the workplace and home
the receiver can also use the attention-focusing (Venkatesh and Vitalari1992) and between the
and controlstrategies to improveunderstanding. organization and its suppliers and customers
Nevertheless, it is necessary to articulatenew continue to blur. Privacy may be an important
communication strategiesforreceivinginformation factor in generalizing our model across these
boundaries.
Support
communicationgoals
and strategies
r
Providemedium and
message form
Identify Feedback
situation r- .I I _ impact
\\-
,'
Organizationar,'\
\ memory
\%j
The diary in Table 1 is not only a medium for often less noticeable. The review of sender-
dialog but also partof an OM.Inotherwords, OM receiver distance and the discussion of proposi-
is not only used to supportcommunicationbut, in tion9 demonstratedhowthese distances are born
fact, can buildfromit.An importantchallenge is to and how they affect the communicationprocess.
distill the different types of context from the Technology's role in creating awareness has, in
originalmessage, store it in OM,and use it to re- the past, concentratedon the perceptualissues of
contextualizefuturecommunication(Schwartzand communication. Clark's contributiontheory of
Te'eni 2000). Furthermore,followingthe discus- discourse (Clarkand Brennan1991)suggests that
sion of formality,the OMwouldincludeinformation it is crucialto monitorthe receiver'sattentionand
that is organized at differentlevels of formality. understanding for effective communication,
Over time, some messages will be gradually resultinginseveral attemptsto design appropriate
abstracted to become increasingly formal. As supportsuch as Portholes(Dourishand Bly1992),
discussed above, itmay be importantto be able to ClearBoard (Ishii et al. 1992), and Peepholes
trace back from the formal to the original (Greenberg1996). Similarly,itshould be possible
message, when communicationbreaksdown.The to displaysemanticdistances between sender and
OMmay,therefore,be constructedthroughstories receiver,forexamplethose concerningdifference
(such as the one about Joey spillingtea), facts in terminology.Forinstance, kMail(Schwartzand
(such as product#8123 is incomplete),and more Te'eni 2000) uses previous knowledge to show
formalprinciples(such as no food in production). whether or not the same terms are shared by
(For some initialdirections on abstractingtexts, sender and receiver, assuming that a visual
see Crampeset al. 1998.) displayof differencesenhances the awareness of
theirexistence.
cation between different national cultures, structuresthat on the one hand enable effective
between different organizational cultures, or retrievaland, on the other hand, enable effective
between users with different profiles. In these presentationto the user as an accessible, multi-
examples, an organizationalmemory is needed layered knowledge structureof context such as
whichis able to characterizethe users and match hypertext.
them (e.g., DIRE,in French1994). Morecomplex
examples of such a process may involve The second problem(timing)calls formechanisms
differencesbetween people who have, inthe past, that detect the conditions under which contex-
ascribed different meanings to current terms. tualizationis needed to avoid informationover-
Furthermore,some errorsin communicationcan load. Spider (Boland et al. 1994) is designed to
be detected (or suspected) on the basis of the present context in a varietyof forms so that it can
immediateor relatedcontextstored in the OM. A lead more efficientlyto better, richercommuni-
trivialexample is thatthe mistakeinTable 1 about cation. The system displays the different
product1823 (shouldbe 8123) couldbe corrected rationalesbehindan issue in the formof cognitive
automaticallyfromits context. maps thathighlightthe similaritiesand differences
in the communicators'perspectives. Thus, mes-
sages are richerin context but (more importantly)
Communication Strategies and Goals are displayed in a fashion that is manageable.
Ourmodel identifiesa list of communicationstra- Maoet al. (1996) show howcontextualizedaccess
tegies, all of which are candidates for computer improves problemsolving with Hyper-Finalyzer.
support (see Table 13), but also identifies the The very same idea can be used in person to
conditions in which they should be activated in person communication. Indeed, Kock (1998)
response to communicationcomplexityinherentin looked at how expert systems technologies
the process (proposition2) and inducedby inputs supportcontextualizationin group-activities.
(propositions 8, 9, and 10). Contextualization
depends on informationretrievaland, as a result, Controlof communicationcan be enhanced by
is likely to be the most promisingdirectionfor several techniques.One is to emulatethe capacity
computer support. Nevertheless, this process builtintoface-to-facecommunication(whichwould
poses two mainproblems:providingcompleteand be more effective than the secondary reaction
relevant informationand providingit only when predicted by propositions4A and 4B). Effective
needed. The first problem implies knowledge eye contact, which helps maintain control by
informingthe sender what message the receivers cation of higher complexity,from more sources,
are sensing, can be emulated by advanced and on multiplemedia compete over the user's
displays that create and update a pictureof the limitedresources. One possible directionthatmay
spatial relationships between communicators be taken is to develop technologies that provide
duringthe session (Fussel and Benimoff 1995; more integrativesolutions, using combinationsof
Muhlbachet al. 1995). However,the qualityof message and medium, to grab more attention
media, with respect to channel capacity, inter- fromthe user and allocate it efficiently.
activity,and adaptiveness, has to be very high
(Doherty-Sneddonet al. 1997; see a review of Computersupportfor perspective taking can be
technical requirementsby Patrick1999). Another achieved indirectlyby presenting the receiver's
directionforsupportingcontrol(proposition4A) is perspective, therebytriggeringthe sender to use
to organize simultaneous feedback during the strategy of perspective taking. Presenting
message productionand manage itthroughoutthe your communicationpartner'scognitive map is
session (Herring1999). For example, messages one possibility (Boland et al. 1994). Designers
thatincludedynamicallycreated hypertextlinksto and scholars often use whiteboards to create
the organizationalmemoryshouldbe shownto the draftsof message and inviteothers to add their
sender beforetransmission(Schwartzand Te'eni perspectiveby moving,erasing, and addingideas
2000). The managementof feedback throughout aroundthe board. Flatlandis a computer-based
the session is particularlyimportantwhen there whiteboardthat groups ideas and presents them
are several participantsand each one may have in differentperspectives (Mynattet al. 1999).
his or her own space on the screen and when
there are several streams of conversation and Computersupportfor affectivityis likelyto be the
each streammust be associated withits particular more difficultto achieve, because much of our
feedback. Herring suggests that two different affective communicationin organizationallife is
views shouldbe possible, one linear(reflectingthe traditionallynon-verbal, instinctive, and, often,
time dimension), and the other non-linear intentionallynon-documented.Using information
(reflectingtopicalprogression). technologyto add social and emotionalmaterialto
messages is directlytied to how well we manage
A third direction is to support planned control to incorporate it in organizational memory as
(proposition4B). CMCgenerally provides more emotionalcontext.
control than other media with similar channel
capacitybecause it can store informationto allow Proposition 2 (about the appropriateness of
non-simultaneity(Hesse et al. 1988). Control communicationstrategies shown in Figure 2) is
throughplanningcannotguaranteeperfectimple- yet to be tested, but if shown to hold true, the
mentationof the plan (e.g., directingthe readerto propositioncan be used in designingsystems that
look for the explanations in the paragraphspre- automaticallyrecommend the use of communi-
ceding the proposition,but the reader may not cation strategies. If organizational memory is
search and find the explanations).Nevertheless, attributedwithcommunicativecharacteristicsthat
interactivetechnology,unlikea printedmessage, match those described in the model (e.g.,
can be designed to guide the reader when the AnswerGardenin Ackerman 1998), the system
situationarises. can use eitherstereotypicknowledgeor individual
knowledgeto assess when strategies are moreor
Information technologysupportsattentionfocusing less effective and recommendthem accordingly.
throughformattingeffects, multi-modalmessages Moreover,when communicationgoals are deter-
that includesynchronizedvoice and motion,and mined, it may be possible to match strategies to
pointingby remotecontrol(e.g., pcANYWHERE). goals. Some earlyworkon Coordinator(Winograd
Video conferencing,too, has provenan effective and Flores 1986) shows it is possible to assign to
technique to focus attentionor at least to create each message its purpose but it burdens the
an awareness of low attention(Daly-Jones et al. sender too much. Advances in technologies that
1998). However, these techniques may not determinethe user's communicationgoals could
suffice to drawand retainattentionas communi- be used to match appropriatestrategies, and
provide feedback to support control (see, for e.g., a simulationof probableerrorsdue to high
example, Ardissonoand Sestero 1996). Indeed, cognitive distance. Clearly,there is still much to
Collagen models collaborativediscourse on the do in terms of developingways of identifyingand
basis of goals thatare eithervoiced by the user or reporting on communication failures (see the
detected automaticallyby an intelligent agent discussion in the elements of communication
(Richand Sidner 1998). impactsection).
Vol.25 No.2/June2001
MISQuarterly 299
andIT
Communication
Te'eni/Organizational
300 MISQuarterly
Vol 25 No.2/June2001
andIT
Communication
Te'eni/Organizational
MISQuarterly
Vol.25 No.2/June2001 301
andIT
Communication
Te'eni/Organizational
Espinoza, J. A., and Garza, R. T. "SocialGroup Fulk, J., Schmitz, J. A., and Steinfield, C. "A
Salience and Inter-Ethnic Cooperation,"Joural Social InfluenceModelof TechnologyUse,"in
of Experimental Social Psychology (21:4), Organizations and CommunicationTechnology,
1985, pp. 380-392. J. Fulk and C. Steinfield (eds.), Sage Publi-
Fairclough, N. Discourse and Social Change, cations, NewburyPark,CA, 1990, pp. 117-140.
PolityPress, Cambridge,MA,1992. Fussell, S. R., and Benimoff,I. "Socialand Cog-
Fish, R. S., Kraut,R. E., and Root, R. W. "Video nitive Processes in InterpersonalCommuni-
as a Technologyfor InformalCommunication," cation: Implicationsfor Advanced Telecom-
Communicationsof the ACM(36:1), 1993, pp. munications Technologies," Human Factors
48-61. (37:2), 1995, 228-250.
Flores, F. "Information Gabarro, J. J. "The Development of Trust
Technologyand the Insti-
Influenceand Expectations,"in Inter-Personal
tution of Identity,"InformationTechnology &
Behavior: Communicationand Understanding
People (11:4), 1998, pp. 351-372. in Relationships,A. G. Athos and J. J. Gabarro
French, J. C. "DIRE:An Approachto Improving
Informal ScientificCommunication," Information (eds.), Prentice Hall, Englewood Cliffs, NJ,
1978, pp. 290-303,
and Decision Technologies(19), 1994, pp.527-
Gabarro, J. J. "The Development of Working
541.
Relationships," in Intellectual Teamwork:
Frost, P. J. "Power,Politics, and Influence,"in Social and Technical Bases of Collaborative
Handbook of OrganizationalCommunication: Work,J. Galegher, R. E. Kraut,and C. Egido
An Interdisciplinary Perspective,F. M.Jablin,L. (eds.), Erlbaum,Hillsdale,NJ, 1990, pp. 79-
L. Putnam, K. H. Roberts, and L. W. Porter 110.
(eds.), Sage Publications,Newbury,CA., 1987, Galbraith,J. R. OrganizationDesign, Addison-
pp. 503-547. Wesley, Reading, MA,1977.
Fulk,J. "SocialConstructionof Communication Gale, S. "HumanAspects of InteractiveMulti-
Technology,"Academyof ManagementJournal media Communication," InteractingwithCom-
(36:5), 1993, pp. 921-950. puters (2), 1990, pp. 175-189.
Fulk, J., and Boyd, B. "EmergingTheories of Gallupe,R. B., Cooper,W. H., Grise,M.,and Bas-
Communicationin Organizations,"Journal of tianutti,L. "BlockingElectronicBrainstorms,"
Management(17:2), 1991, pp. 407-446. JournalofAppliedPsychology (79:1),1994, pp.
FulkJ., and Collins-Jarvis,L. "WiredMeetings: 77-86.
Technological Mediation of Organizational Gefen, D., and Straub, D. W. "GenderDiffe-
Gatherings,"in New Handbookof Organiza- rences inthe Perceptionand Use of E-mail:An
tional Communication,F. M. Jablin and L. L. Extension to the Technology Acceptance
Putnam(eds.), Sage, NewburyPark,CA,2000, Model,"MIS Quarterly(21:4), 1997, pp. 389-
pp. 624-663. 400.
Fulk, J., and DeSanctis, G. "ElectronicCom- Gelfand,M.J., and Christakopoulou,S. "Culture
munication and Changing Organizational and NegotiatorCognition: Judgement Accu-
Forms,"OrganizationScience (6:4), 1995, pp. racy and Negotiation Processes in Indivi-
337-349. dualisticand CollectivisticCultures,"Organiza-
Fulk,J., and Ryu, D. "PerceivingElectronicMail tionalBehaviorand HumanDecisionProcesses
Systems: A PartialTest of the Social Informa- (79:3), 1999, pp. 248-269.
tionProcessing,"Paperpresentedto a meeting Ginsburg,M.,and Kambil,A. "Annotate!A Web-
of the InternationalCommunication Asso- based KnowledgeManagementSupportSys-
ciation, Dublin,1990. tem forDocumentCollections,"WorkingPaper
Fulk, J., Schmitz, J. A., and Schwarz, D. "The #IS-98-19, SternSchool of Business, NewYork
Dynamics of Context-BehaviorInteractionsin University,1998.
Computer-MediatedCommunication," in Con- Goffman,E. Formsof Talk,Universityof Pennsyl-
texts of Computer-MediatedCommunication, vania Press, Philadelphia,PA, 1981.
M. Lea (ed.), HarvesterWheatsheaf, London, Goldberg,D. T. "TheSocial Formationof Racist
1991, pp. 1-29. Discourse" in Anatomy of Racism, D. T.
Vol.25 No.2/June2001
MISQuarterly 303
andIT
Communication
Te'eni/Organizational
304 MISQuarterly
Vol.25 No.2/June2001
Communication
Te'eni/Organizational andIT
MISQuarterly
Vol.25 No.2/June2001 305
andIT
Communication
Te'eni/Organizational
306 MISQuarterly
Vol.25 No.2/June2001
andIT
Communication
Te'eni/Organizational
MISQuarterly
Vol.25 No.2/June2001 307
andIT
Communication
Te'eni/Organizational
308 MISQuarterly
Vol.25 No.2/June2001
andIT
Communication
Te'eni/Organizational
MISQuarterly
Vol.25 No.2/June2001 309
andIT
Communication
Te'eni/Organizational
Putnam, K. H. Roberts, and L. W. Porter(eds.), Tang, J. C., and Isaacs, E. A. "WhyDo Users Like
Sage Publications,Newbury,CA., 1987, pp. 451- Video? Studies of Multimedia-Supported Col-
502. laboration,"ComputerSupportedCooperative
Straub, D. "TheEffectof Cultureon ITDiffusion: Work(1), 1992, pp. 163-193.
E-mail and Fax in Japan and the U.S," Tatar, D. G., Foster, G., and Bobrow, D. G.
InformationSystems Research (4:3), 1994, pp. "Designfor Conversation:Lessons from Cog-
23-47. noter,"InternationalJournal of Man-Machine
Straub, D., and Karahanna, E. "Knowledge Studies (34;2), 1991, pp. 185-209.
WorkerCommunicationsand RecipientAvail- Te'eni, D. "Analysisand Design of Process Feed-
ability:Towarda Task Closure Explanationof back in InformationSystems: Old and New
Media Choice," OrganizationScience (9:2), Wine in New Bottles,"Accounting, Manage-
1998, pp. 160-175. ment and InformationTechnology(2:1), 1992,
Straus, S. G., and McGrath,J. E. "Does the pp. 1-18.
MediumMatter?The Interactionof Task Type Te'eni, D., Sagie, A., Zaidman,N., Schwartz,D.,
and Technology on Group Performance and and Hamburger,Y. "Organizational Communi-
Member Reactions," Journal of Applied cation: A Modeland FieldStudy,"IEEETrans-
Psychology (79:1), 1994, pp. 87-97. actions on Professional Communications
Street, R. L.,and Capella,J. H. (eds.). Sequence (44:1), March2001.
and Pattern in CommunicationBehaviour, Te'eni,D., and Schwartz,D. G. "Contextualization
EdwardArnold,London,1985 in Computer-MediatedCommunication," Infor-
Sussman, S. W., and Sproull,L. "StraightTalk: mationSystems Review (1), 2000, pp. 59-75.
Delivering Bad News Through Electronic Thompson,J. D. Organizationsin Action: Social
Communication," Information Systems Science Bases of Administrative Theory,
Research (10:2), 1999, pp. 150-166. McGraw-Hill, New York,1967.
Suzuki,S. "Cultural Transmissionin International Trevino, L. K., Daft, R. L., and Lengel, R. H.
Organizations:Impactof InterpersonalCom- "UnderstandingMedia Choices: A Symbolic
munication Patterns in IntergroupContext," InteractionistPerspective," in Organizations
HumanCommunication Research (24:1), 1997, and CommunicationTechnology,J. Fulkand C.
pp. 147-180. W. Steinfield (eds.), Sage Publications,
Tan, B. C. Y., Wei, K. K., Sia, C., and Raman, K. NewburyPark,CA, 1990, pp. 71-94.
S. "A Partial Test of the Task-MediumFit Trevino, L. K., Lengel, R. H., and Daft, R. L.
Proposition in a Group Support System "MediaSymbolism,MediaRichness,and Media
Environment,"ACMTransactionson Compu- Choice in Organizations:A Symbolic Interac-
ter-HumanInteraction(6:1), 1999a, pp. 47-66. tionistPerspective,"CommunicationResearch
Tan, B. C. Y., Wei, K .K.,and Watson, R. T. "The (14:5), 1987, pp. 553-574.
EqualizingImpactof a GroupSupportSystem Triandis,H. C. Individualismand Collectivism,
on Status Differentials,"ACMTransactionson Westview, Boulder,CO, 1995.
Information Systems (17:1), 1999b, pp.77-100. Triandis,H. C. "Cross-Cultural Studies of Indivi-
Tan, B. C. Y., Wei, K. K.,Watson, R. T., Clapper, dualismand Collectivism," inNebraskaSympo-
D. L., and McLean,E. R. "Computer-Mediated sium on Motivation,J. J. German(ed.), Univer-
Communication and Majority Influence: sity of NebraskaPress, Lincoln,NE, 1989, pp.
Assessing the Impactin an Individualisticand 41-133.
a CollectivisticCulture,"ManagementScience TrompenaarsF. Riding the Waves of Culture:
(44;9), 1998a, pp. 1263-1278. Understanding Cultural Diversity in Global
Tan, B. C. Y., Wei, K. K., Watson, R. T., and Business, McGrawHill,New York,1998.
Walczuch,R. M. "ReducingStatus Effectswith Turkle,S. TheSecond Self, Simonand Schuster,
Computer-Mediated Communication: Evidence New York,1984.
fromTwoDistinctNationalCultures,"Journalof Tyre, M.J ., and von Hippel, E. "TheSituated
Management Information Systems (15:1), Natureof AdaptiveLearningin Organizations,"
1998b, pp. 119-141. OrganizationScience (8:1), 1997, pp. 71-83.
310 MISQuarterly
Vol.25 No.2/June2001
andIT
Communication
Te'eni/Organizational
Valacich,J. S., Dennis,A., and Connolly,T. "Idea Walther, J. B., and Burgoon, J. K. "Relational
GenerationinComputer-BasedGroups:A New Communicationin Computer Mediated Inter-
Ending to an Old Story," Organizational action," Human Communication Research
Behavior and Human Decision Processes (19:1), 1992, pp. 50-88.
(57:3), 1994, pp. 448-467. Warkentin,M.E., Sayeed L.,and Hightower,R. T.
Valacich,J. S., Dennis, A., and Nunamaker,J. F., 'VirtualTeams versus Face-to-FaceTeams:An
Jr. "Electronic Meeting Support: The ExploratoryStudyof a Web-Based Conference
GroupSystemsConcept,"International Journal System," Decision Sciences (28:4), 1997, pp.
ofMan-MachineStudies (34:2), 1991, pp.261- 975-996.
282. Watson, R. T., Ho, T. H., and Raman, K. S.
"Culture: A Fourth Dimension of Group
Valacich, J. S., Paranka, D., George, J. F., and
Nunamaker,J. F. "CommunicationConcur- Support Systems," Communications of the
ACM(37:10), 1994, pp. 44-55.
rency and the New Media: A New Dimension
forMediaRichness,"Communication Research Webster, J. "DesktopVideoconferencing:Exper-
iences of Complete Users, Wary Users, and
(20:2), 1993, pp. 249-276.
Non-users," MIS Quarterly(22:3), 1998, pp.
Vallacher, R. R., and Wegner D. M. "WhatDo 257-286.
People Think They're Doing? Action Identi- Webster, J., and Trevino, L. K. "Rationaland
fication and Human Behavior,"Psychological Social Theories as ComplementaryExplana-
Review (94:1), 1987, pp. 3-15. tions of CommunicationMediaChoices: Two
Van de Ven, A. H., Delbecq, A., and Koenig, R.
Policy-Capturing Studies," Academy of
"Determinantsof CoordinationModes Within Manageent Journal (38;6), 1995, pp. 1544-
Organizations," AmericanSociologicalReview 1572.
(41:3), 1976, pp. 322-338. Weedman, J. "Taskand No-TaskFunctionsof a
Van Dijk,T. A., and Kintsch,W. Strategies of Computer Conference Used in Professional
Discourse Comprehension,Academic Press, Education: A Measure of Flexibility,"Inter-
New York,1983. national Journal of Man-Machine Studies
Venkatesh, A., and Vitalari,N. P. "AnEmerging (34:2), 1991, pp. 303-318.
Distributed Work Arrangement: An Weick, K. E. The Social Psychology of Orga-
Investigationof Computer-BasedSupplemental nizing,AddisonWesley, Reading, MA,1979.
Workat Home,"ManagementScience (38:12), Westmyer, S. A., DiCoccio, R. L., and Rubin,R.
1992, pp. 1687-1706. B. "Appropriateness and Effectivenessof Com-
Waldron, V. R., and Applegate, J. L. "Inter- munication Channels in Competent Inter-
personal Construct Differentiationand Con- personal Communication,"Journal of Com-
versational Planning: An Examinationof the munication(48:3), 1998, pp. 27-48.
TwoAccountsforthe Productionof Competent Whitener,E., Brout,S. E., Korgaurd,M. A., and
Verbal Disagreement Tactics,"Human Com- Werner,J. M. "Managersas Initiatorsof Trust:
municationResearch (21), 1994, pp. 3-35. An Exchange Relationship Framework for
Wallbott,H. G., and Scherer, K. R. "Cues and Understanding Managerial Trustworthy
Channels in EmotionRecognition,"Journalof Behavior,"Academy of Management Review
(23:3), 1998, pp. 330-349.
Personality and Social Psychology (51:4), Video as a Technology
Whittaker,S. "Rethinking
1986, pp. 690-699. for InterpersonalCommunications,"Interna-
Walther,J. B. "Interpersonal EffectsinComputer- tional Journal of Human-ComputerStudies
MediatedInteraction: A RelationalPerspec-
(42:5), 1995, pp. 501-529.
tive,"CommunicationResearch (19:1), 1992, Wijayanayake,J., and Higa, K. "Communication
pp. 52-90. MediaChoice by Workersin DistributedEnvi-
Walther,J. B. "RelationalAspects of Computer- ronment,"Information& Management (36:6),
MediatedCommunication: Experimental Obser- 1999, pp. 329-338.
vationsOverTime,"Organization Science (6:2), Williamson,O. E. Marketsand Hierarchies,The
1995, pp. 186-203. Free Press, New York,1975.
Vol.25 No.2/June2001
MISQuarterly 311
andIT
Communication
Te'eni/Organizational
Wilson,S. R. CoordinatingComplianceand Face Zmud,R. W., Lind,M. R., and Young, F. W. "An
Goals WithinPersuasive Messages, Unpub- AttributeSpace for OrganizationalCommuni-
lished DoctoralDissertation,PurdueUniversity, cation Channels," Information Systems
1989. Research (1:4), 1990, pp. 440-457.
Winograd, T., and Flores, F. Understanding
Computersand Cognition:A New Foundation
for Design, Addison Wesley, Reading, MA, About the Author
1986.
Wood, R. E. "TaskComplexity:Definitionof the Dov Te'eni is on the facultyat Bar-llanUniversity,
Construct," Organizational Behavior and Israel,where he serves as directorof the Univer-
HumanDecision Processes (37:1), 1986, pp. sity's e-learning center and as the founding
60-82. directorof the interdisciplinarycenter for Global
Yates, J., Orlikowski,W. J., and Okamura,K. Knowledge Management. Dov studies several
"Explicitand ImplicitStructuringof Genres in relatedareas of information systems inthe organi-
ElectronicCommunication:Reinforcementand zational context: human-computer interaction,
Change of Social Interaction,"Organization computer supportfor decision making and sys-
Science (10:1), 1999, pp. 83-103. tems design. Inaddition,he is interestedin infor-
Zack, M. H. "Interactivity and Communication mation systems for non-profitorganizations.His
Mode Choice in Ongoing Management researchusuallycombinesmodelbuilding,labora-
Groups,"InformationSystems Research (4:3), toryand fieldexperimentsand the developmentof
1993, pp. 207-239. prototypesystems such as Spider and kMail.He
Zigurs, I., and Buckland, B. K. "A Theory of is currentlyworkingon the design of adaptive
Task/TechnologyFit and GroupSupportSys- Web Articles and the use of informationtech-
tems Effectiveness," MIS Quarterly (22:3), nologyforsupportingcommunicationwithinmulti-
1998, pp. 313-334. nationalscorporations.