Perspective Making and Perspective Taking in Communities of Knowing.1995

Perspective Making and Perspective Taking in Communities of Knowing
Author(s): Richard J. Boland, Jr. and Ramkrishnan V. Tenkasi

Source: Organization Science, Vol. 6, No. 4 (Jul. - Aug., 1995), pp. 350-372
Published by: INFORMS
Stable URL: http://www.jstor.org/stable/2634993 .
Accessed: 19/04/2014 13:41
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
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.
.
INFORMS is collaborating with JSTOR to digitize, preserve and extend access to Organization Science.
http://www.jstor.org
This content downloaded from 131.123.1.227 on Sat, 19 Apr 2014 13:41:00 PM
All use subject to JSTOR Terms and Conditions
Perspective Making and Perspective
T aking in Communities of Knowing
Richard J. Boland , Jr. * Ramkrishnan V. T enkasi
Weatherhead School of Management, Case WestemReserve University, Cleveland , Ohio 44106-7235
School of Business Ad ministration, University of SouthemCalif omia, Los Angeles, Calif omia 90089-1421
Abstract
Knowled ge-intensive f irms are composed of multiple commu-
nities with specialized expertise, and are of ten characterized
by lateral rather than hierarchical organizational f orms. We
argue that prod ucing knowled ge to create innovative prod -
ucts and processes in such f irms requires the ability to make
strong perspectives within a community, as well as the ability
to take the perspective of another into account.
We present mod els of language, communication and cog-
nition that can assist in the d esign of electronic communica-
tion systems f or perspective making and perspective taking.
By appreciating how communication is both like a language
game played in a local community and also like a transmis-
sion of messages through a cond uit, and by appreciating how
cognition includ es a capacity to narrativize our experience as
well as a capacity to process inf ormation, we id entif y some
guid elines f or d esigning electronic communication systems to
support knowled ge work. T he communication systems we
propose emphasize that narratives can help construct strong
perspectives within a community of knowing, and that ref lect-
ing upon and representing that perspective can create bound -
ary objects which allow f or perspective taking between com-
munities.
We conclud e by d escribing our vision of an id ealized
knowled ge intensive f irmwith a strong culture of perspective
making and perspective taking, and by id entif ying some ele-
ments of the electronic communication systems we would
expect to see in such a f irm.
(Knowled ge Work; Organization Leaming; Distributed
Cognition; Communication Systems)
Introd uction
Organizations are d eveloping innovative prod ucts and
services on f aster cycle times (Purser and Pasmore
1992, Lawler 1992), causing an increase in knowled ge
work (Pava 1983) and a grad ual replacement of capital
and labor intensive f irms by knowled ge intensive f irms
(Starbuck 1992). Knowled ge work creates new und er-
stand ings of nature, organizations or markets and ap-
plies themin valued technologies, prod ucts, or pro-
cesses. Knowled ge intensive f irms are composed of
multiple communities with highly specialized technolo-
gies and knowled ge d omains (Purser et al. 1992). In the
pharmaceutical ind ustry, f or example, d eveloping new
prod ucts requires integration of knowled ge f roma
broad array of d isciplines such as molecular biology,
physiology, biochemistry, synthetic chemistry, pharma-
cology and even esoteric specialties such as molecular
kinetics (Hend erson, 1994). A similar pattern is observ-
able in other ind ustries. T he f irst generation of cellular
telephones used f ive sub technologies, but their third
generation incorporated f ourteen d istinct sub tech-
nologies (Granstrand et al. 1992). T he increasing pro-
lif eration of specialized and d istinct knowled ge com-
munities and the need f or their integration has also
resulted in the emergence of new organizational f orms,
among themthe lateral-f lexible f ormof organization
(Galbraith and Lawler 1993, Galbraith 1994). T he lat-
eral-f lexible organizational f ormrelies on peer-to-peer
collaboration (as opposed to a vertical hierarchy) in
achieving organizational objectives.
It is our contention that all organizations are becom-
ing more knowled ge intensive across the service, ind us-
trial and governmental sectors. It is easiest to see the
f und amental importance of knowled ge work in f irms
involved with new prod uct d evelopment in lead ing
ed ge technologies, but the relentless pace of change in
market expectations means that all organizations will
1047-7039/95/0604/0350/$01.25
Copyright ? 1995. Institute f or Operations Research
350 ORGANIZAT ION SCIENCE/VOl. 6, No. 4, July-August 1995 and the Management Sciences
RICHARD J. BOLAND, JR. AND RAMKRISHNAN V. T ENKASI Communities of Knowing
increasingly rely on creating new knowled ge and ad opt-
ing lateral organizational f orms. T he major issue f or
such f irms is to f ind creative ways f or representing and
integrating knowled ge across their lateral units (Weick
and Roberts 1993, Galbraith 1994).
Knowled ge prod uction involves communication
within and between a f irm's multiple communities of
knowing. We ref er to communication that strengthens
the unique knowled ge of a community as perspective
making, and communication that improves its ability to
take the knowled ge of other communities into account
as perspective taking. In this paper we employ mod els
of language, communication and cognition to propose
how electronic communication systems can be d esigned
to support perspective making and perspective taking
in knowled ge intensive f irms.
We argue that perspective making and perspective
taking are achieved by narrating our experience as well
as by rationally analyzing it. T hese processes are like
playing games with language as well as like transmit-
ting messages through a cond uit, and ' they involve
heightened levels of ref lexivity. T he narrating of expe-
rience is a critically important but of ten overlooked
element of knowled ge prod uction in knowled ge inten-
sive f irms, even though it is recognized that scientif ic
reasoning is of ten cond ucted through narratives and
that scientists' interpretive practices are embod ied in
their conversation (Knorr-Cetina 1981, Mulkay et al.
1983). Similarly, the importance of playf ul situated
action f or strengthening one's own perspective and the
importance of ref lexivity f or appreciating the perspec-
tive of another is not suf f iciently recognized in commu-
nication systemd esign. Af ter analyzing the processes
of perspective making and perspective taking, we d e-
scribe an id ealized knowled ge-intensive f irmto high-
light some of the f eatures of an electronic communica-
tion systemthat would support those processes.
We f irst present the concept of a community of
knowing as an open systemand provid e a brief overview
of mod els of language, communication and cognition
that can guid e the d esign of electronic communication
systems. T he language games mod el of Wittgenstein
(1974) and Bruner's (1986, 1990) mod el of narration as
a cognitive mod e are presented as supplements to the
d ominant organizational mod els of language as mes-
sage transmissions and cognition as inf ormation pro-
cessing. Science is used as an example of knowled ge
work to d raw implications f or applying these mod els to
the communication requirements in knowled ge-inten-
sive f irms. We then explore the d ynamics of perspec-
tive making and perspective taking and some potential
breakd owns in the perspective-taking process. T his al-
lows us to summarize the strengths and weaknesses of
the language game and the cond uit mod els of commu-
nication f or d esigning electronic communication sys-
tems, and to emphasize the importance of ref lexivity
and bound ary objects in perspective taking. Finally, we
present our ad mitted ly utopian vision of some f uture
applications of electronic communication f or support-
ing knowled ge work in a hypothetical f irmwith a
lateral-f lexible f orm.
Communities of Knowing as Open
Systems
Organizations are characterized by a process of d is-
tributed cognition in which multiple communities of
specialized knowled ge workers, each d ealing with a
part of an overall organizational problem, interact to
create the patterns of sense making and behavior d is-
played by the organization as a whole (Boland et al.
1994).
Organizations are necessarily characterized by d is-
tributed cognition because their critically important
processes and the d iversity of environments and tech-
nologies to be d ealt with are "too complex f or one
person to und erstand in its entirety" (Brehmer 1991,
p. 4; Nersessian 1992). T his problemis especially acute
in knowled ge-intensive f irms that rely.on multiple spe-
cialties and knowled ge d isciplines to achieve their ob-
jectives. Each such community of specialized knowl-
ed ge workers is what we terma "community of know-
ing."
A number of scholars such as Fish (1980), Fleck
(1979), Barnes (1983) and Brown and Duguid (1991)
have commented on the way that communities d evelop
unique social and cognitive repertoires which guid e
their interpretations of the world . Fleck's (1979) con-
cept of "thought collective" is one such notion that
emphasizes the unique interpretive repertoires of a
d istinct community of knowing. A thought world evolves
in a community of knowing as a "read iness f or d irected
perception". T hought world s with d if f erent f und s of
knowled ge and systems of meaning cannot easily share
id eas, and may view one another's central issues as
esoteric, if not meaningless. Other terms which echo
our concept of community of knowing includ e "inter-
pretive community" (Fish 1980), "context of learning"
(Barnes 1983), and "community of practice" (Lave and
Wenger 1990, Brown and Duguid 1991, Orr 1990).
However, given our f ocus on knowled ge-intensive f irms,
and our concern with the interaction of d if f erent ex-
pert knowled ge groups in the process of knowled ge
ORGANIZAT ION SCIENCE/Vol. 6, No. 4, July-August 1995 351
creation, we f eel "community of knowing" is the most
appropriate label f or our purposes.
T he multiple communities of knowing in
knowled ge-intensive f irms overlap in complex and
shif ting ways. T here is a rich structural hierarchy (Smith
1981) of communities of knowing within the f irm, and
between the f irmand its environment. Divisions, f unc-
tional areas, prod uct lines, prof essional specialties, pro-
ject teams, issue-based committees, and so on are all
possible sites f or communities of knowing that inter-
weave with each other across various levels of the
organization. Ind ivid uals will f ind themselves as mem-
bers of several communities of knowing operating
within a f irmand its environment.
In science, the interaction of communities of know-
ing has been viewed as an open systemby Star (1993)
and a similar notion has been expressed by Barnes
(1983) and Hesse (1974) who characterize each com-
munity as a knowled ge net within an institution or
culture. Even in settings where communication appears
unproblematic and knowled ge homogeneous, the nets
of ind ivid ual communities d if f er. It is through the
d ynamic interactions between such communities that
new conf igurations of the knowled ge net emerge by
creating new meanings, new linguistic routines, and
new knowled ge. Maintenance and ref inement of the
existing knowled ge in a community can be attributed to
f eed back processes operating within established rou-
tines and policies. T he creation of new knowled ge in
an organization, however, is of ten the result of an open
systemtransf ormation of that organization's communi-
ties of knowing as they question and revise routines
and create new processes and relationships among
themselves (Wiener 1954; von Bertalanf f y 1968, Argyris
and Sch6n 1978). We argue that perspective making
and perspective taking are the basis f or transf orma-
tions within and between communities of knowing and
thereby the basis f or open systemcontrol in knowled ge
work. Our principal contention is that d esigning elec-
tronic communication systems f or knowled ge intensive
f irms requires an appreciation of how they can med iate
the transf ormation and changing relationships among
communities of knowing by af f ecting perspective mak-
ing and perspective taking capabilities.
Assumptions about Language
Communication and Cognition
in Knowled ge Work
T he knowled ge work of perspective making and per-
spective taking requires ind ivid ual cognition and group
communication. Our und erstand ing of language, com-
munication and cognition are centrally important in
d esigning electronic communication systems, and be-
low we supplement the d ominant mod el f or each with
an alternative that we believe necessary f or creating
more ef f ective d esigns.
Supplementing a Cond uit Mod el of Communications
with a Language Games Mod el
In consid ering how electronic communication can be
d esigned to support knowled ge work in organizations,
two mod els of language and communication are impor-
tant to consid er. One is the cond uit mod el inspired by
the work of Shannon and Weaver (1949), which we see
as the d ominant mod el in management literature, the
other is the language game mod el of Wittgenstein
(1974). Each of these theoretical orientations is usef ul
f or thinking about language and communication in an
organization when it is viewed as an open systemof
communities of knowing. Each is good f or certain
purposes, but not f or all, and we will employ themf or
d if f erent need s in our analysis.
T he cond uit mod el is the most f amiliar in organiza-
tional stud ies, and portrays communication as a mes-
sage send ing and message receiving process through a
transmission channel with a limited channel capacity.
A cond uit mod el suggests that communication can be
improved by red ucing noise in the channel, with noise
d ef ined as the possibility f or error contaminating the
message on its route f romsend er to receiver. Noise can
be red uced by increasing the channel capacity; by ref in-
ing the proced ures f or encod ing and d ecod ing mes-
sages; by provid ing more reliable d ata storage and
retrieval f acilities; or by making the channel of commu-
nication more universally available. A central limita-
tion of the cond uit mod el is its unproblematic treat-
ment of a message (Red d ing 1972, Red d y 1979) in
which the symbolic or interpretive character of mes-
sages in language is not consid ered . T he encod ing and
d ecod ing activities are treated as d iscrete selections of
messages f roma pred ef ined set and the problemof
human meaning in language is avoid ed . In manage-
ment stud ies we see this mod el being used in organiza-
tion d esign (Galbraith 1973, T ushman 1978, Allen
1986), as well as research and d evelopment and innova-
tion (Davis and Wilkof 1988, Allen et al. 1980, Allen
and Cohen 1969). Attempts to expand the cond uit
mod el to ad d ress ambiguity in language can be seen in
stud ies of med ia richness (Daf t and Lengel 1984) and
the interpretation of noise (Ciborra et al. 1984), but
they remain within -a message transmitting f ramework.
352 ORGANIZAT ION SCIENCE/VO1. 6, No. 4, July-August 1995
As an alternative mod el we will consid er
Wittgenstein's image of language and communication
as games in f orms of lif e (Wittgenstein 1974, Astley
and Zammuto 1992). His image of communication ap-
preciates language as f und amentally and inexorably
embed d ed in the situated action of our immed iate
communities, or our "f orms of lif e". Action in our
immed iate communities is the locus f or language d evel-
opment and use. Conversations and activity in our
f orms of lif e are language games, and through our
language games we create the meaning of particular
word s and f orms of speech, and we continuously evolve
new ways of talking and acting together.
Wittgenstein spent the f irst part of his lif e trying to
d ef ine the essence of language, searching f or a stable,
id eal meaning f or word s and sentences, and the princi-
ples of logic that could be relied upon to provid e
unambiguous and coherent knowled ge (Wittgenstein
1961). He later rejected the notion of an id eal language
in which word s pictured objects, and meaning was
uniquely id entif iable and stable. Instead , he came to
see how language cannot be und erstood apart f romits
rooted ness in lif e experience, nor can word s stand
apart f romsituated use with unambiguous meaning.
Language games in f orms of lif e are the basis f or all we
know. T hrough action within communities of knowing
we make and remake both our language and our
knowled ge. Unlike the cond uit mod el, in a language
game there is no f ixed set of messages or meanings
f romwhich to choose in communicating.
23. But how many kind s of sentences are there? ... T here
are countless kind s: countless d if f erent kind s of use of what we
call "symbols", "word s", "sentences". And this multiplicity is
not something f ixed , given once f or all; but new types of
language, new language-games, as we may say, come into
existence, and others become obsolete and become
f orgotten. ...
Here the "language game" is meant tobring intopromi-
nence the f act that the speaking of language is part of an
activity, or a f ormof lif e. (Wittgenstein 1974, p. 11, as in
original)
Supplementing an Inf ormation Processing Mod e
of Cognition with a Narrative Mod e of Cognition
Bruner (1986, 1990) has proposed that we recognize
there are at least two d istinct mod es of cognition, the
inf ormation processing (or parad igmatic) mod e and the
narrative mod e. We believe that explicitly recognizing
the narrative mod e of cognition is important f or und er-
stand ing how perspective making and perspective tak-
ing occur. T he d ominant way of und erstand ing cogni-
tion tod ay is to emphasize its parad igmatic mod e, as
ref lected in inf ormation processing mod els of cognition
(Simon 1977). Bruner (1986, 1990) has proposed that
this inf ormation processing view of cognition, empha-
sizing the rational analysis of d ata in a mental problem
space and the construction of d ed uctive arguments, be
supplemented by recognizing that humans also have a
narrative cognitive capacity. We narrativize our experi-
ence almost continually as we recognize unusual or
unexpected events (the noncanonical) and construct
stories which make sense of them(restore canonical-
lity). Bruner argues persuasively that the narrative ca-
pability of humans is a f und amental cognitive process
through which our cultural world and sense of self are
constructed and maintained over time.
Good arguments and good stories are equally impor-
tant f or und erstand ing human cognition in knowled ge
work, but the two mod es are jud ged by d if f erent crite-
ria. Whereas an argument is jud ged to be good if it is
logical, coherent, consistent, and noncontrad ictory, a
narrative is jud ged to be good if it is interesting,
plausible, and believable. An argument proves some-
thing about the world to be true, but a narrative shows
how events or f eatures in the world are sensible and f it
within our shared cultural f ield . It is well recognized
that surf acing and challenging the of ten implicit as-
sumptions that und erlie a parad igmatic argument is an
important element f or innovative knowled ge work.
Narrative serves an important role in this regard . By
bringing the apparently noncanonical into relief along-
sid e the canonical, the narrative mod e of cognition
provid es access to the implicit assumptions and inter-
pretive structures that characterize a community of
knowing. We will give an example of such an analysis
later in the paper, when we d escribe some of the
electronic communication systems we envision.
T here are several points of similarity in the und erly-
ing assumptions of these mod els of communication and
cognition, particularly between the cond uit and
parad igmatic, and between the language game and
narrative (see T able 1). Both the cond uit mod el of
communication and the parad igmatic mod e of cogni-
tion are based on inf ormation processing images in
which word s point at things, meanings are not prob-
lematic, and the power of d ed uctive logic is empha-
sized . Similarly, both the language game mod el of
communication and the narrative mod e of cognition
are based on social constructionist images in which
word s gain sense only through actual use in a commu-
nity, meanings are symbolic and inherently ambiguous,
and the power of social processes, storytelling and
conversation is emphasized . In spite of this f amily
resemblance among the mod els of communication and
T able 1
Key assumptions behind cond uit mod el of communication and
parad igmatic mod e of cognition.
* T here is und erlying objective knowled ge in the world that has
universal applicability.
* Language can be a med iumf or representing objective knowl-
ed ge and word s have f ixed meaning.
* Human beings can achieve universality of und erstand ing since
f ixed meanings of word s can be communicated objectively f romone
person toanother.
* Realization of objective knowled ge is a rational process. Knowl-
ed ge evolves and progresses through the systematic application of
logic and principles of the scientif ic method .
Key assumptions behind language games mod el of communica-
tion and narrative mod e of cognition.
* Knowled ge as well as method s f or realizing knowled ge are
objective only tothe extent they are ratif ied as objective by a specif ic
community's interpretive conventions.
* Word s can have consensus of meaning only within a specif ic
community of knowing. However, even within a unique community,
the meaning of word s change and are never f ixed in time or space.
* Language is not a med iumf or representing our thoughts and
objective und erlying knowled ge but language is thought and knowl-
ed ge. T he limits of our language are the limits of our knowled ge
since we can explain the world only through language and narrative
f orms.
* Knowled ge evolves by inventing new language and narrative
f orms. Re-narrativizing the f amiliar or coming up with narratives that
explain the unf amiliar is the primary activity by which new knowled ge
comes about.
cognition being d rawn upon, they will not be com-
bined . Whereas the language games mod el is philo-
sophical and proposed as a more accurate d epiction of
human language and communication, the cond uit
mod el is technical and proposed as the necessary re-
quirements f or a communication system. One f ocuses
on human language, the other f ocuses on communica-
tion technology. T he narrative and parad igmatic mod es
of cognition, on the other hand , are meant by Bruner
(1990) as complementary f unctions of the same whole,
each being capacities of the human being. We will not
collapse them, and will treat all f our as having a
separate trad ition and use. We will employ each to
serve d if f erent purposes in und erstand ing the pro-
cesses of perspective making and perspective taking
among communities of knowing.
We will now look to science f or insights on how
communities of knowing d evelop and change through
communication. We will then d raw implications f or
d esigning electronic communication systems f or knowl-
ed ge work in organizations, and also consid er how
multiple communities of knowing interact.
Science as Knowled ge Work
in Communities of Knowing
Consid ering science as organized knowled ge work has
many insights to of f er f or und erstand ing perspective
making and perspective taking in knowled ge intensive
f irms and f or speculating on how electronic communi-
cation can be d esigned to support knowled ge work in
lateral organizational f orms. A central source f or these
insights is provid ed by T homas Kuhn (1970) as he
d escribes the historical process of scientif ic work.
Read ers are no d oubt f amiliar with Kuhn's argument of
how normal science within parad igms lead s to crisis
and revolution. For Kuhn, a parad igmis a shared sense
of what the metaphysical nature of the world is, what
problems are important, and what serve as good exem-
plars f or a d omain of concern.
T here are many d if f iculties with Kuhn's notion of
parad igm. It is of ten taken to be totalizing, unitary and
almost religiously held . As Masterman (1970) has noted ,
Kuhn (1962) used the termparad igmin many d if f erent
ways in the f irst ed ition of T he Structure of Scientif ic
Revolutions. In the revised ed ition Kuhn (1970) ac-
knowled ged the concept's ambiguity and ad d ed f urther
ref inements, but d ebates about just how a parad igmis
to be d ef ined or isolated f or f urther stud y in its own
right will not concern us here. We believe his basic
insight is valid , and is in keeping with Polanyi's (1967)
id ea of "tacit knowled ge", Bould ing's (1956) d iscussion
of "the image", Pepper's (1942) notion of "world hy-
potheses" and numerous others who point out that
perception is only accomplished through a perspective
(Burrell and Morgan 1979, Bartunek 1984).
Kuhn's (1970) insights are particularly relevant f or
und erstand ing how knowled ge is prod uced in a com-
munity of knowing by ref ining and clarif ying the per-
spective of the community. Development of knowled ge
in a community is a process of posing and solving
puzzles, thereby elaborating and ref ining the vocabu-
lary, instruments and theories that embod y the per-
spective. Agreement that knowled ge is progressing is
agreement that the perspective is strengthening. Unex-
pected events or f ind ings can only be recognized as
such f romwithin a perspective. Without a strong per-
spective the community cannot tell an anomaly f rom
noise; a challenge to their knowled ge f roman irrele-
vancy.
Collins (1983) makes some interesting observations
on the d ynamics of knowled ge d evelopment and the
354 ORGANIZAT ION SCIENCE/Vol. 6, No. 4, July-August 1995
d if f erent kind s of competence required of the scientist.
Working within a perspective has well established
method s f or externalizing its objects, and the scientist
should be competent in those respects. Collins terms
this "native competence". It is the kind of competence
that makes meanings, perceptions, and acts of the
native member f ollow naturally as a matter of course.
However, changing or overturning the taken f or granted
rules or replacing themwith a completely new set
requires "interpretive competence" on part of the sci-
entists. It lies in perspective taking: being able to
ref lect upon and renarrativize the f amiliar to open up
new insights and und erstand ings.
T he stronger and more well d eveloped a community's
perspective is, the more usef ul a cond uit mod el of
communication and f eed back becomes. As theories,
puzzles, measures and accepted results are clarif ied
and institutionalized within the community, the more
likely it is that messages can be thought of as selections
f roma pred ef ined set. T he process by which new
communities of knowing begin to f orm, however, and
the processes of questioning and changing perspectives
is not as well hand led by a cond uit mod el. Work that
questions a perspective is of a d if f erent logic type than
work within a perspective, and is primarily controlled
by the d ynamics of change in an open systemrather
than simple f eed back (Wiener 1954, von Bertalanf f y
1968, Bateson 1972). For this second -ord er knowled ge
work, the language games communication mod el is
more helpf ul than the cond uit mod el. Previously ac-
cepted und erstand ings, measurements, and logics are
in a sense "up f or grabs". T he perspectives behind
ways of knowing of the organizational communities are
being mad e in real time by the communities' members.
T he language of their communication is changing as
their practices in f orms of lif e are changing. Messages
cannot be separated f romthe evolving context of mak-
ing and using themas in the cond uit mod el.
T wo f inal themes f romKuhn that we will consid er
bef ore d rawing implications f or knowled ge work in
organizations are the incommensurability between per-
spectives and the emergence of new perspectives. If
members of a community create a strong perspective
and d o d istinctive and important knowled ge work, it
will of necessity approach becoming incommensurable
with other perspectives. T hey may use the same word s
as other communities of knowing, but they will use
themto see things in d if f erent ways (Knorr-Cetina
1981). T hey will look at the same phenomena as
another community, but will see d if f erent problems,
d if f erent opportunities, and d if f erent challenges
(Czarniawska-Joerges 1992). As Kuhn puts it, they will
live in a d if f erent world f romthose in other communi-
ties of knowing. Data important to one are irrelevant
to another, or are used f or entirely d if f erent purposes.
Arguments that persuad e convincingly in one commu-
nity of knowing have little or no weight in another.
And the more d eveloped and ref ined the community of
knowing becomes, with an increasingly elaborate and
d etailed perspective, the more nearly incommensurable
it becomes with others (Fleck 1979, Brown and Duguid
1991, Dougherty 1992). If the members' language games
within one community of knowing f ully und erstood and
appreciated the positions of another, they would not be
d if f erent communities and would not be d oing d istinct
knowled ge work.
Knorr-Cetina (1981) presents some ground ed exam-
ples of how local communities of knowing d evelop
their unique parad igmatic world s and are resistant to
changing them. In her sociological stud y of d if f erent
research units, she f ound that research laboratories
d eveloped local interpretations of method ical rules, or
a local know-how with regard to how to make things
work best in actual research practice. Criteria f or what
mattered and what d id not matter were neither f ully
d ef ined nor stand ard ized throughout the research com-
munity. Nor were the rules of of f icial science exempt
f romlocal interpretations. Many important selections
of the research process were locally d riven, includ ing
questions of ingred ients, instrumentation, and d uration
of experiments.
Implications of Kuhn and Knowled ge Work in Science
f or Und erstand ing Knowled ge-intensive Firms
A f irst implication of Kuhn f or thinking about knowl-
ed ge-intensive f irms is that the primary unit of analysis
should be the community of knowing. T he ind ivid ual
d oes not think in isolation and is not an autonomous
origin of knowled ge. A community of knowing is a
language game and neither the language nor the
knowled ge created within it comes f romthe actor
alone.
Second ly, a community of knowing requires perspec-
tive making in ord er to d o knowled ge work. Without a
strong perspective it cannot prod uce important knowl-
ed ge. A community's perspective d evelops by ref ining
its vocabulary, its method s, its theories and values and
its accepted logics through language and action within
the community of knowing. T his means that the com-
munity must, of necessity, have a space f or conversa-
tion and action isolated f romthe larger organization.
T hird ly, the ability of one community of knowing to
work jointly with another requires an ability to over-
come the d egree of incommensurability between them.
ORGANIZAT ION SCIENCE/VO1. 6, No. 4, July-August 1995 355
T his, of course, must be d one without sacrif icing the
integrity and d istinctiveness of their own perspective.
Below we will explore this process of perspective taking
in which the perspective of another can be taken into
account as part of a community's way of knowing.
Fourthly, the cond itions f or change in the perspec-
tive of a way of knowing come f romboth the insid e and
f romthe outsid e. Insid e the perspective, cond itions of
change come f romthe accumulation of anomalies as it
is tested and elaborated . Fromoutsid e the perspective,
pressure f or change comes f romad herents d rawn to a
promise of the aesthetics, power or excitement of a
new perspective. T his suggests that memories of errors
and anomalies are important to maintain and review
openly, and that the isolation of communities necessary
f or their d evelopment should be punctuated by period s
of interaction between communities.
Finally, new perspectives need to be nurtured and
given protection f rom strong d emand s f or perf or-
mance. Of necessity, they will not be able to compete
with an established perspective in another community's
way of knowing.
For a knowled ge-intensive f irm, then, we look to its
ecology of communities of knowing to und erstand its
possibilities f or d oing knowled ge work. Electronic com-
munication can med iate how the open systemof com-
munities emerge, d evelop, elaborate, suf f er crisis, and
transf ormwithin it. Electronic communication can also
med iate how communities of knowing interact and
their capacity f or perspective taking. It is to these
processes of perspective making and perspective taking
that we now turn.
T aking in Communities of Knowing
T he Process of Perspective Making
Perspective making is the process whereby a commu-
nity of knowing d evelops and strengthens its own
knowled ge d omain and practices. As a perspective
strengthens, it complexif ies and becomes better able to
d o knowled ge work. Complexif ication is achieved cog-
nitively through the use of parad igmatic analysis within
a narrative f raming of experience. It is a process of
d eveloping f iner language games, and f roma parad ig-
matic stand point, more precise causal laws. Complexi-
f ication signif ies a movement f roma global, und if -
f erentiated naming to a more precise explication of
constructs, where more coherent meaning structures
are d eveloped than preced ing ones (Wad d ington 1957,
Werner 1957). Knorr-Cetina (1981) proposed that sci-
entif ic conceptual systems have to progressively com-
plexif y themselves over a period of time to successf ully
solve scientif ic problems. T his implies the ability to
respond to shif ts and f luctuations in the novelty of the
scientif ic problemd omain by mod eling the shif ts them-
selves (Rubinstein et al. 1984).
A good example of complexif ication in perspective
making is presented by Brad shaw (1992) in his analysis
of the Wright brothers' invention of the airplane. He
also illustrates the interweaving of narrative f raming
and parad igmatic analysis in the perspective making
process. Brad shaw asks why were the Wright brothers
so successf ul in conquering the challenge of manned
f light, while many of their competitors with better
training and resources f ailed ? He answers that f irst,
the Wright brothers narratively f ramed the phe-
nomenon of f lying using a d if f erent metaphor than
their competitors, and second , they employed f iner
problemsolving proced ures. Whereas their competi-
tors narrated f light with a "chauf f eurs of the air"
metaphor, telling how f lying was akin to d riving a car
into the air, a group that includ ed the Wright brothers
narrated f light as being like "f lying a kite". Many of
the unsuccessf ul inventors had a propensity to con-
struct complete aircraf ts and then to test themby
measuring d istance and time in f light. T o these d esign-
ers, the airplane as a vehicle to be chauf f eured was an
assemblage of parts (wings, f uselage, propulsion, etc.)
and d eveloping an aircraf t meant exploring possible
d esigns f or conf iguring these parts.
However, f or the Wright brothers, the major concern
was to und erstand how a kite f lew, and to achieve its
f unctions (lateral control, suf f icient lif t, red uced d rag,
etc.) in the airplane. T hey f irst isolated these f unc-
tional problems and then proceed ed to solve themone
at a time. T he pattern in their work was to explore
solutions to subproblems using d irected experiments.
For example, a kite was built to explore lateral control
and wind tunnel experiments explored lif t and thrust.
T hrough extensive testing of mod els, the Wright broth-
ers "d iscovered an important error in aerod ynamics
overlooked by other investigators" (Brad shaw 1992,
pp. 246-247). Only when each separate problemwas
und erstood and solved d id the Wright brothers invest
time and energy in build ing a new craf t. T he Wright
brothers employed both narrative and parad igmatic
mod es of cognition in their perspective making, as they
mod eled and d eveloped more complex and f iner un-
d erstand ings of the workings of aerod ynamic laws. In
contrast, their competitors were exploring the possibil-
ity of f light with minimal und erstand ings of aerod y-
namic laws, and relied on trial and error, hoping one of
their mod els would f ly, without having any conception
of why. T hey lacked the strong perspective necessary to
d o important knowled ge work.
T he Importance of Narrative in Perspective Making
Perspective making within communities of knowing is a
social practice in a f ormof lif e. For insight into how
this process takes place in a community of knowing, we
will return to Jerome Bruner's work on the role of
narrative in constructing knowled ge of self and world
(Bruner 1986, 1990). Bruner argues that we must look
to how actors make meaning of their experience
through narrative if we are to und erstand the process
of perspective making. Bruner, synthesizing stud ies of
child d evelopment, language acquisition and concept
f ormation, proposes an innate narrative capacity as the
engine f or our cognitive activity. "T he typical f ormof
f raming experience (and our memory of it) is in narra-
tive f orm. What d oes not get structured narratively is
lost in memory." (Bruner 1990, p. 56) Parad igmatic
thinking is an important part of our cognitive reper-
toire, but only a part. Narrativizing our ref lexive moni-
toring and rationalization of cond uct is not ruled by an
abstracted logic. Within a community of knowing, a
narrative explanation works not only because it is logi-
cally acceptable, but also because it is lif elike and
plausible; it f its the culturally bound d emand s of a
f ormof lif e.
In parallel with Gid d ens' structuration theory
(Gid d ens 1976), Bruner emphasizes that when we nar-
rativize experience, we also construct and valid ate the
self . T he narrator's perspective as an essential element
in any story assures this. T he self is always at stake in
the ind ivid ual's narrativizing of experience, because the
self is at least the narrator (recognizing the canonical,
ind icating and explaining the anti-canonical, d etermin-
ing how the world should be) and of ten part of the
story (being herself d elineated as a causal agent with
motives, intentions and values).
T he importance of narrative has not gone unnoticed
in organizational research. Clark (1972) explored the
importance of sagas and Mitrof f and Kilmann (1976)
recognized the importance of myth. Myth and saga are
important, but they can d istract our attention f romthe
way that human cognition operates almost continu-
ously in a narrative, storytelling mod e. We wish to
emphasize that narrative is f ully equal to parad igmatic
analysis in the construction, maintenance and change
of perspectives in an organization. We see themin a
type of f igure-ground relation in which parad igmatic,
rational-analytic thought takes place in a context pro-
vid ed by narrative, and narratives are constructed
against a backd rop of parad igmatic und erstand ings in a
kind of "genuine union" of the two mod es (Boland and
Pond y 1983). T he rational analytic elements of a per-
spective in a community of knowing are a prod uct of
storytelling as much as they are a med iumf or it.
More recently, the role of stories and storytelling in
the d ay-to-d ay f unctioning of organizations has been
ad d ressed by Boje (1991). T he constructive, changing
quality of stories d ocumented by Boje in his f ocus on
situated practice is a major step toward the position we
argue f or here. He moves beyond the mythic view of
the story as an "object",f ound in Martin and Meyerson
(1988), McConkie and Boss (1986), and Gabriel (1991),
and turns our attention to the community d epend ent
process of prod ucing the story.
When scientists experience anomalies within a per-
spective they of ten turn to narrative in an attempt to
make sense of the noncanonical observation. Science,
and scientif ic papers d ocumenting experiments and
theories, in retrospect, always seemparad igmatic, lin-
ear and certain. T his is partly d ictated by the social
conventions of what good science is (Knorr-Cetina
1981). However, an examination of the inf ormal d is-
course of scientists presents another picture altogether.
Highly variable and inconsistent accounts of action and
belief are very much the norm. Actors continually
construct and reconstruct the meaning of their scien-
tif ic world through the f ormulation of d ivergent narra-
tive accounts. As Mulkay et al. (1983) summarize:
"Unless we und erstand how actors socially construct
their accounts of action and how actors constitute the
character of their actions primarily through the use of
language, we will continue to f ail ... to f urnish satisf ac-
tory answers to the long-stand ing questions about the
nature of action and belief in science" (pp. 195-196).
Others such as Nersessian (1992) and Eysenck and
Keane (1990) have also pointed out the important role
of narrative in scientif ic reasoning. T hought experi-
ments are a prevalent f ormof scientif ic reasoning in
which the scientist imagines a sequence of events and
then narrativizes the sequence in ord er to communi-
cate the experiment to others. Einstein is supposed to
have perf ormed thought experiments based on stories
about rid ing on a light beamand traveling in elevators.
Rutherf ord in his investigations of the structure of the
atomis reputed to have imagined the electrons as
revolving around the nucleus in the same way as plan-
ets revolve around the sun (Gentner 1983). Galileo
(Galilei 1638; cited in Nersessian 1992) likewise used a
thought experiment in arguing against the Aristotelian
theory that heavier bod ies f all f aster than lighter ones.
T he Process of Perspective T aking
In knowled ge-intensive f irms, competitive ad vantage
and prod uct success are a result of collaboration in
which d iverse ind ivid uals are able to appreciate and
synergistically utilize their d istinctive knowled ge
through a process of perspective taking (Dougherty
1992, Purser et al. 1992, Nonaka 1994, Hend erson
1994, Brown 1991). Duncan and Weiss (1979, p. 86)
summarize this process as one in which: "T he overall
organizational knowled ge base emerges out of the pro-
cess of exchange, evaluation, and integration of knowl-
ed ge. Like any other organizational process, ... [i]t is
comprised of the interactions of ind ivid uals and not
their isolated behavior." It requires a process of mutual
perspective taking where d istinctive ind ivid ual knowl-
ed ge is exchanged , evaluated , and integrated with that
of others in the organization (Nonaka and Johansson
1985, Shrivastava 1983).
Much of social behavior is pred icated upon assump-
tions an actor makes about the knowled ge, belief s and
motives of others. T his is the beginning of the process
of perspective taking, and is f und amental to communi-
cations. In any communication, the knowing of what
others know is a necessary component f or coord inated
action to take place (Bakhtin 1981, Clark 1985, Krauss
and Fussell 1991). As Brown (1981) observed , ef f ective
communicating requires that the point of view of the
other be realistically imagined . Others such as-
Rommetveit have af f irmed this point: "An essential
component of communicative competence in a pluralis-
tic social world ... is our capacity to ad opt the perspec-
tives of d if f erent others" (Rommetveit 1980, p. 126).
T he f und amental importance of taking the other's point
of view into account is seen in Mead (1934) who
ref erred to it as taking the attitud e of the other and
equated our ability to be f ully human with our ability
to maintain an inner conversation with a generalized
other.
In ord er f or perspective taking to proceed , the d i-
verse knowled ge held by ind ivid uals in the organization
must be represented in its uniqueness, and mad e avail-
able f or others to incorporate in a perspective-taking
process. Valuing d iversity of knowled ge by enabling
each type of expertise to make unique representations
of their und erstand ings, and assisting actors with d if -
f erent expertise to better recognize and accept the
d if f erent ways of knowing of others, is the f ound ation
f or perspective taking. It can be encouraged by com-
munication systems that includ e an emphasis on sup-
porting the d istinctive need s of separate communities
of knowing.
T he task of taking each other's knowled ge and back-
ground into account is a complex process, and can
f requently break d own. For example, Purser et al.
(1992) d id a comparative stud y of two knowled ge inten-
sive prod uct d evelopment projects of equal technical
complexity in a high-technology f irm. One project suc-
ceed ed while the other f ailed . T wo essential f actors
accounted f or the d if f erences in results between the
two projects. T he f irst was a higher incid ence of barri-
ers to knowled ge sharing among the members on the
f ailed project team. But behind this f irst f actor was a
second , causal f actor of f ailed perspective taking. T eam
members were unable to surf ace and reconcile d issimi-
larities in their knowled ge and cognitive f rames of
ref erence. Failure to achieve perspective taking through
d epicting and exchanging representations of their
unique und erstand ings d ramatically red uced their pos-
sibilities f or successf ul teamknowled ge work.
Perspective taking involves a variety of inf erential
and jud gmental processes. Ind ivid uals may utilize an
assortment of techniques includ ing stereotypes and in-
f erence heuristics to estimate what others know. Such
heuristics can ind uce systematic errors and biases
(Kahneman et al. 1982, Nisbett and Ross 1980). T he
read y availability of the actor's own perspective may
lead the actor to overestimate the likelihood that the
perspective will be shared by others (Steed man and
Johnson-Laird 1980). T his f alse consensus ef f ect, in
which subjects assume that others are more similar to
themselves than is actually the case (Ross et al. 1977) is
a f ormof bias particularly relevant to the perspective-
taking process. T his heuristic lead s to overestimates of
the extent to which a person's knowled ge is shared by
others, and stud ies support the existence of such a bias
(Dougherty 1992, Krauss and Fussell 1991).
Dougherty (1992) provid es an insightf ul analysis of
breakd owns in the perspective-taking process d ue to
actors' inability to surf ace and examine their d if f ering
interpretive schemes. She f ound that in unsuccessf ul
cases of new prod uct d evelopment, the key players
interpreted and und erstood issues around technology-
market linking and new prod ucts in qualitatively d if f er-
ent ways f romeach other and were not able to recon-
cile these d if f erences. T he d if f erences in interpretation
centered around three themes. T he f irst theme was
what people see when they look into the f uture, includ -
ing which issues are seen as most uncertain. What they
saw seemed uncertain, while what they d id not see, d id
not seemparticularly uncertain or even noteworthy.
T he business planner worried about positioning against
competition while the f ield person worried about id en-
tif ying the right potential customers. A second theme
characteristic of f ailed teams involved people's und er-
stand ing of the d evelopment process itself . People not
only ignored the activities of others and f ailed to argue
over relative priorities, they glossed over the concerns
of others, and tend ed not to appreciate their complexi-
ties. A third theme characteristic of f ailed teams
involved the d if f erent "thought world s" of teammem-
bers. For new prod uct d evelopment, d if f erent d epart-
mental thought world s were coherent and consistent
within themselves. T his red uced the possibility f or cre-
ative perspective taking, since members of a d epart-
ment thought that they alread y knew everything
(Dougherty 1992). As lucid ly word ed by Dougherty
(1992), "Nor is the problemlike the proverbial set of
blind men touching a d if f erent part of an elephant. It is
more like the tales of eye witnesses at an accid ent, or
of ind ivid uals in a troubled relationship-each tells us
a complete story, but tells a d if f erent one" (p. 191).
In summary then, the problemof integration of
knowled ge in knowled ge-intensive f irms is not a prob-
lemof simply combining, sharing or making d ata com-
monly available. It is a problemof "perspective taking in
which the unique thought world s of d if f erent communi-
ties of knowing are mad e visible and accessible to
others. Making explicit representations of one's knowl-
ed ge and und erstand ings to exchange with others en-
ables one to better appreciate the d istinct ways of
knowing that those others will attempt to communi-
cate. In ord er to integrate knowled ge through perspec-
tive taking, communication systems must f irst support
d iversity of knowled ge through the d if f erentiation pro-
vid ed by perspective making within communities of
knowing. Only af ter a perspective is d if f erentiated and
strengthened can it be ref lected upon and represented
so the actors in other communities of knowing have
something to integrate through a perspective taking
communication.
Implications f or Electronic
Communication Systems and Policies
T he d esign of electronic communication systems af -
f ects how organization members are able to engage in
perspective making and perspective taking and thus
build communities of knowing. In knowled ge-intensive
f irms, the problemof d esigning systems and policies
f or electronic communication is a problemof provid ing
an environment in which an ecology of communities of
knowing can d evelop through complexif ication over
time. In perspective making, a community of knowing
complexif ies by enriching and ref ining its d istinct per-
spective and way of knowing. Its categories f or parti-
tioning the world become more numerous and subtle;
the d istinctions it makes as to the appropriateness of
problemstatements, measures, tests and logics f or a
given situation become more esoteric and precise. In
perspective taking, complexif ication involves an in-
creased capacity f or communities of knowing to take
each other into account within their own language
games, and to construct new languag'e games f or their
interaction. T he d evelopment of complexif ied perspec-
tive taking represents the integrative capacity of the
ecology of communities.
T hese two d ynamics, perspective making and per-
spective taking, are instantiated only through speaking
and acting in a community. Electronic communication
med ia provid e an important part of the physical and
symbolic environment available f or engaging in the
f orms of lif e of the organization's communities of
knowing, but only a part. Other concerns such as task,
technology, structure, culture, reward systems and
lead ership style, all play a role in med iating the type of
language games that will emerge. Although the entire
set of these issues is beyond the scope of this paper we
will d iscuss some of the issues f urther in the next
section when we d escribe some examples of communi-
cation systems that would support these processes.
Here, we will concern ourselves with presenting a
certain sensibility as a way of thinking about how
electronic communication med ia provid e cond itions f or
the two d ynamics of perspective making and perspec-
tive taking.
A f irst element in the sensibility we propose is to
recognize the strengths and weaknesses of the mod els
of communication and cognition we are d rawing upon
in d esigning these systems. T he cond uit mod el, with its
assumption of messages that carry unambiguous mean-
ing if they are cod ed and d ecod ed error f ree, is a good
mod el f or thinking about the communication of well
established elements in a community of knowing's vo-
cabulary and method s of practice. Communication
within established community routines can and should
be ad d ressed with a cond uit mod el. T he knowled ge is
semi-f ixed and reliably interpretable within the com-
munity, so the assumptions of a cond uit mod el match
the communication need s well. T he organization-wid e
community where culture and id entity are acted out
and a sense of institution is d eveloped is also well
suited to a cond uit mod el. It is appropriate f or ques-
tions of broad cast band wid th and f or d evelopment of a
f irm-wid e vocabulary. Recent research in corporate
strategy emphasizing the importance of shared inter-
pretive schemes (Bartunek 1984, Ranson et al. 1980),
common visions (Collins and Porras 1991, Bennis and
Nanus 1985) or shared strategic image (Hamel and
Prahalad 1991, Bertad o 1990) are examples of this type
of communication at the level of the organization as a
whole.
T he symbolic quality of this culture-build ing commu-
nication, with its reliance on evocative images rather
than precise language is somewhat at od d s with the
cond uit mod el, but can generally be ad equately han-
d led by red und ancy or repetition. T he cond uit mod el
can support activities that broad cast and reinf orce im-
portant symbols, stories, and exemplars which become
commonly available to members of the community as a
whole and incorporated in their language games. Very
little in the way of d istinctive, organizational knowl-
ed ge work is accomplished at the cultural level of the
community as a whole. It is better thought of as a
backd rop against which the more esoteric language
games of more locally situated f orms of lif e are played
out.
T he cond uit mod el, however, d oes have some d is-
tinct weaknesses. T he perspective-making process re-
quires a nurturing of emergent communities of know-
ing, and requires a respect f or the uniqueness of a local
community's d istinctive f ormof lif e. T he cond uit mod el
stand s in opposition to this requirement with its em-
phasis on d eveloping d ata mod els, d ecision mod els and
communication f ormats that are common and shared
across the organization. Current research in inf orma-
tion technology of ten ref lects this inappropriate use of
the cond uit mod el with its emphasis on enterprise
mod eling and d ata architecture with a single, unif ied
d ata structure (Scheer 1992, Deng and Chaud hry 1992,
T argowski 1988, Richard son et al. 1990, Chen 1976).
Similarly, mod el management systems concern them-
selves with unif ying the d iversity of knowled ge in man-
agement d ecision mod els through a variety of meta-
level integrative techniques (Geof f rion 1987, Dolk 1988,
Elamand Konsynski 1987). Finally, it seems that a
principal concern with end -user computing is the re-
d uction of d iversity and the establishment of stand ard s
and common structures f or d ata and mod els (Brown
and Bostrom1989, Munro et al. 1987, Rivard and Huf f
1988). We d isagree with these calls f or commonality in
vocabulary and knowled ge practices, and call instead
f or recognizing the importance of strong perspective
making and d if f erentiation of knowled ge among a f irm's
communities of knowing.
Electronic med ia based on the wrong mod el of com-
munication can hind er perspective making and taking
in interactions among communities of knowing. An
example f romresearch on new prod uct d evelopment
processes will illustrate the point. T he task f or this new
prod uct d evelopment teamwas to choose a nonhuman
analogue such as a rat, rabbit or primate mod el with
which to cond uct tests of a new d rug compound they
were d eveloping f or certain af f lictions associated with
the human intestine. T he teamhad members repre-
senting d if f erent d isciplines such as lif e sciences, chem-
istry, toxicology and biopharmaceutics. T here were d if -
f erences of opinion as to the nonhuman analogue most
appropriate f or the task. As a result, the teammem-
bers resorted to a popular groupware prod uct and its
voting systemto reach a consensus. Based on the
voting proced ure, a rat analogue was chosen. Unf ortu-
nately, the rat was not suitable f or the task of repre-
senting the human intestine, but the teamonly f ound
that out at the human clinical trials. T he poor choice
had by then cost the company consid erable expense
and three years of d evelopment time. T he groupware
voting system, with its emphasis on f ind ing consensus,
hampered the teammembers f romf irst strengthening
and representing their own perspectives and then en-
gaging in a d ialogue of perspective taking with each
other. T he groupware helped red uce noise in the com-
munication and provid ed an illusion of certainty. What
was required , however, was a language games mod el of
communication to complexif y the unique und erstand -
ing of each through d ialogue within their community of
knowing. T hen, they should have employed a technol-
ogy that would support ref lexivity, and creation of a
visible representation of their unique knowled ge that
would have enabled perspective taking among them.
T he language games mod el also has its strengths and
weaknesses. One strength is helping us think through
issues of perspective making with its insistence on the
primacy of speaking and acting in a local community.
Electronic communication med ia may red uce bound s
of space and time f or such communities (Gid d ens
1991), but the language games mod el can help us to
recover the importance of enabling and protecting lo-
cal logics, practices and vocabularies (Jonsson 1992),
even within d ispersed communities. T he language
games mod el is also usef ul f or emphasizing the need
f or isolation to create id entity in a community of know-
ing. T ime f or participating in communities is limited ,
and id entif ied spaces f or members to engage the com-
munity's language games and d evelop its perspective
are an important cond ition f or its persistence and
d evelopment.
Schon (1979) provid es a vivid example of the need to
respect the importance of communication in local com-
munities f romthe history of town planning. When
town planners saw their task as a need to cure a
blighted area, they intervened with all manner of
planned renewals to tear d own and remake whole
sections of a city, of ten d isturbing the patterns of
communication within neighborhood s. But their ef f orts
went terribly wrong, again and again, until they came
to see such areas of town not as blighted , but as f olk
communities with a strong network of communication
and support that sustained themquite well in the f ace
of substantial d if f iculty. T he problemf or the town
planners then became how to d esign systems and poli-
cies that would enable that emergent capacity of the
local communities of knowing to strengthen and self -
organize. We hope to build such an awareness into our
approach to thinking about electronic communication
f romthe start.
T he language games mod el is also a good basis f or
thinking about narrative in a community of knowing. It
emphasizes that narrative is experientially ground ed
and that it is a search f or ways to make issues and
events of interest to the community sensible within its
way of knowing. T he causal implications and action
sequences in narrative are the source of perspective
making f or the community, as members ref lect upon
the und erlying logics, values and id entities of the com-
munity of knowing.
A major limitation of the language games mod el is
the "epistemic inhibitions of its own parad igm"
(Rubinstein et al. 1984). T he stronger a community of
knowing is supported by communication systems re-
f lecting a language game mod el of communication, the
stronger is perspective making complexities, and the
less able it may become to allow f or other ways of
seeing. A vivid example of this d ynamic is presented in
Dougherty's (1992) f ield stud y. T he various f unctions
involved in the prod uct d evelopment process agreed on
the need f or the prod uct to be market oriented . How-
ever, in the language games of the research and d evel-
opment group, market orientation meant prod uct spec-
if ications and technical f eatures: the market is what the
prod uct can d o. For the manuf acturing people, on the
other hand , a market-oriented prod uct was a d urable
and reliable one. Lowering the number of f eatures and
specif ications would improve its market orientation.
Further, the marketing group consid ered customer
need s on a customer by customer approach. For the
planning group, to be market oriented meant to posi-
tion the prod uct in the right market niche. T hey d id
not worry about prod uct f eatures, customer need s, or
reliable prod uct perf ormance. T his is where the reward
systems and culture of the organization become impor-
T able 2 T woMod els of Communication and T heir Relative
Merits f or Supporting Electronic Med ia in Systems
of Knowled ge Work
CONDUIT MODEL
Strengths
* Reliable and precise channel f or communicating well estab-
lished elements in the vocabulary of a community of knowing and
techniques of practice.
* Can f acilitate culture build ing, organization-wid e integration ac-
tivities through shared and common images.
Weaknesses
* Does not value d iversity; emphasis on unif ormd ata and d eci-
sion mod els and communication f ormat across the organization can
hamper the emergence of unique communities of knowing.
* Inappropriate f or supporting the narrative f orms of cognition that
are central tothe perspective making process.
* Common vocabulary and set of d ecision mod els d enies the
importance of perspective taking.
LANGUAGE GAME MODEL
Strengths
* Facilitates perspective making by virtue of its insistence on
primacy of speaking and action in a community of knowing.
* Und erscores the importance of enabling and protecting local
logics, local practices and local vocabularies.
* Implicates the importance of narrative in a community of know-
ing.
* Emphasis on narratives enables ref lection on und erlying logics,
values and id entities of the community of knowing.
Weaknesses
* Increasingly specialized language games results in epistemic
inhibitions (imposed by each community's unique parad igm) and
comes in the way of perspective taking.
* May heighten conf lict among communities.
tant in maintaining a balance between perspective
making and perspective taking. One important d esign
element in this regard is the establishment of an issue-
specif ic space f or perspective taking between strong
communities of knowing to take place. Isaacs (1993)
ref ers to this space as a container f or d ialogue, and we
will think of it as a f orumwithin an electronic commu-
nication system.
As we have seen, both mod els have strengths and
weaknesses that primarily relate to their role in per-
spective making, but both mod els alone have d istinct
weaknesses with respect to perspective taking (see
T able 2). T he cond uit mod el, with its emphasis on a
commonly available and exhaustive set of messages and
cod ing techniques d enies the importance of perspective
taking. A common vocabulary and set of d ecision mod -
els presumes that each member of the organization
participates in the same way of knowing and need s no
special support f or opening a space within the d ialogue
of their own local community f or taking the perspective
of another. T he language game mod el, as we have just
seen, also d oes not help in thinking about perspective
taking because of its emphasis on speaking and acting
within a f ormof lif e and its increasingly specialized
language games. Another aspect of communication
must be consid ered f or thinking about perspective tak-
ing, one that is absent or overlooked in the cond uit and
language games mod els. T his aspect concerns how the
richness of representations and the ref lexive capacity
of a communication systemenables the creation and
exchange of bound ary objects (Star 1989, 1993), which
we will d iscuss in the next section.
Ref lexivity, Bound ary Objects
and Perspective T aking
In our d iscussion of perspective making in communities
of knowing, we saw the ind ivid ual speaking and acting
within the community's f ormof lif e. For perspective
taking we need a shif t in emphasis, to f ocus on the
ind ivid ual's ability to make his or her own und erstand -
ing visible f or self -ref lection. Once a visible representa-
tion of an ind ivid ual's knowled ge is mad e available f or
analysis and communication, it becomes a bound ary
object and provid es a basis f or perspective taking.
Representations of ways of knowing f rommembers
in one community can then be exchanged with mem-
bers of another, who, having themselves engaged in an
ef f ort to make rich representations of their und er-
stand ings, can now engage in communication about the
perspectives of another. T his taking of the other into
account, in light of a ref lexive knowled ge of one's own
perspective, is the perspective-taking process.
Perspective taking is never a one-to-one mapping of
meanings. Members of the same community of know-
ing will not have f ull consensus, and members of d if f er-
ent communities cannot simply ad opt the meanings of
another. But as Star (1989, 1993) has observed , scien-
tists within and between communities d o f ind a way of
bringing their d istinctive perspectives into d ialogue
through the construction and d iscussion of bound ary
objects. An ind exed collection of items, a map, an
id ealized image, or a label can all serve as bound ary
objects around which sense making can take place.
Such bound ary objects d o not convey unambiguous
meaning, but have instead a kind of symbolic ad equacy
that enables conversation without enf orcing commonly
shared meanings. Bound ary objects can, of course, be a
center of intense conf lict as easily as one of coopera-
tive ef f ort. Creating and reshaping bound ary objects is
an exercise of power that can be collaborative or
unilateral. Nonetheless, in the absence of bound ary
objects, the possibility of perspective taking is limited
and the opportunity f or knowled ge work in the f irmis
red uced .
Ref lection on our own perspectives is d if f icult and
of ten not attempted . As Rubinstein et al. comment, "If
practicing scientists were more conscious of the pro-
cesses of science, it would go a long way toward cir-
cumventing the epistemological inhibitions imposed by
parad igms" (1984, p. 138). Collins (1983) also notes the
hid d en nature of such processes. He argues that many
times it is only when the rules go wrong that the
scientist questions the nature of his or her interpreta-
tion. "Otherwise, our giving of meaning to objects-our
interpretive practices are so automatic that we d o not
notice that any interpretation is involved "(Collins 1983,
p. 90). In Schutz's (1964) terms, ref lexivity is the ability
to period ically suspend our natural attitud e and notice
the matter-of -course, taken-f or-granted ways in which
our communities of knowing are constructed and inter-
preted , which can open possibilities to change them
(Collins 1983). Rubinstein et al. (1984) posit that be-
coming aware of , evaluating, and mod if ying perspec-
tives is required f or maintaining ad aptive knowled ge.
T here are many possible f orms f or bound ary objects
that can represent knowled ge f romone community f or
perspective taking by another, includ ing physical mod -
els, spread sheets, or d iagrams. We will present two
examples that could be incorporated in communication
systems: cognitive maps (Axelrod 1976, Huf f 1990,
Boland et al. 1994, Weick and Bougon 1986, Weick
1990, Ed en 1992), and narrative structures (T enkasi
and Boland 1993, Mulkay et al. 1983, Knorr-Cetina
1981).
A cognitive map is a d irected graph whose nod es
represent concepts or f actors in the actor's d ecision
d omain, and whose arcs represent cause-and -ef f ect re-
lationships between source and d estination nod es
(Boland et al. 1992, Burgess et al. 1992). Figure 1
presents an example of a cause map d epicting a physi-
cian's und erstand ing of quality in med ical care. Creat-
ing this map is an exercise in perspective making, and
exchanging it with actors f romother communities of
knowing within the hospital makes it a bound ary object
and opens the possibility f or perspective taking in the
search f or quality in med ical care. Build ing such a map
can be evocative f or the map creator, as well as inf or-
mative to its recipient. Creating cognitive maps can
reveal personal cause-and -ef f ect logic, which in turn
f orces the ind ivid ual to conf ront the reasonableness
and valid ity of previously tacit cause-ef f ect assumptions
Figure 1 A Physician's Map of Quality in Med ical Care
Med ical ~ ~ ~ ~ ~ civ
+_
T reatment
j m }d , uestl(-.)ues nmg
ancl ~ ~ Ba
c c~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~Lc
KO Med ice
ile
Ef f ectsPereNte
Arcs wo
+Ptieive
- r \ ()f Caref er -al s ompassi)
Xpectatio
Cureaiorehwniniats nerait of NI
c Sientif ic 5 \>
_/ <
of
Caref
OKnowle.lge y 1 363
/ / t
E~~~~~~~xpectation}
/ / \ ~~~~~ ~ ~~~(2) /
/ / Uni~~nque \ a-l
/ t ) ~~~~~~~~~~ ~
~ ~ ~~~~~f f i citseAC
Societal\ /ocietal
I xpectation
I
erceptio:n
Vof Meclicine, - f CareJ
Nc-)te: Arcs without
po,sitive
or
negative
relation,s sh(-.)wn inclicate,s uncertainty o)f
cau,sal ef f ect.
(Fiol and Huf f 1992, Weick and Bougon 1986). Creat-
ing maps of one's und erstand ing of a problemd omain
and ref lecting on themcan also f acilitate new and
more complex und erstand ings of the situation at hand ,
improving the chances f or scientif ic success (Weick
1990).
Cognitive maps are a good beginning f or making rich
representations of an und erstand ing within a perspec-
tive. But a key ingred ient f or communicative success is
a way to link elements and relations in a map, as well
as the map itself to unstated elements and assumptions
of the perspective. T hat is, the knowled ge representa-
tion grows richer as context is ad d ed , layer by layer, to
ind ivid ual elements in the cognitive map. T his suggests
a hypertext or hyper med ia communication environ-
ment in which actors f ind a self -ref lective space to
build rich knowled ge representations whereby f actors
in a cause map are linked to und erlying belief s, values
or assumptions in the f ormof spread sheets, notes, or
graphs or other cause maps (Boland et al. 1992, 1994).
Another kind of bound ary object that can serve as a
f ocal point f or perspective taking is a narrative struc-
ture. Narratives, if bracketed and approached f or anal-
ysis with an interpretive stance can also provid e ele-
ments of the ref lexive quality we see as necessary f or
perspective taking. Narrative analysis can ref lexively
give access to the implicit and unstated -assumptions
that are guid ing perspective making, and in so d oing
help enable a perspective taking process. We will
d emonstrate this by f irst presenting a narrative f roman
actual incid ent collected d uring f ield work in a phar-
maceutical company, and then analyzing and interpret-
ing its narrative structure.
T he Story of Norman, a Chemist
Norman stood up f romhis work bench in mid -morning and
went to the men's toilet where he used the urinal. Shortly
af ter returning tohis work bench, Norman f elt a numbness in
his penis. He was startled , but he immed iately thought that
trace amounts of the XV75 hypertension compound s he had
been working with that morning had been on his hand s and
may just be a powerf ul topical anesthetic.
He told two colleagues about this potential d iscovery and
created an inf ormal teamto explore its possibilities. Af ter
about six weeks, he obtained f ormal approval f romthe Assis-
tant Director f or this project and his team. At this stage, the
id ea was to go f or a topical application of the compound .
Af ter several weeks, Norman went tosee the Assistant Direc-
tor to inf ormhimof a metabolic stud y of the compound in a
cell culture that showed some ind ications of toxicity. He
learned f romthe Assistant Director that a market stud y had
just been completed showing that an oral f ormof the d rug
would be very successf ul and highly prof itable, whereas the
Figure 2 Narrative Structure of the Story of Norman
Ignore Focus on topical
iumibiiess versioii
T
Seoise ioew Forixi prof ret Develop oral
T ito
liscovei-y
team
inut versioi
- \ ~~~~~ ~ ~~~~~report
Coomplaixi of lax Discoolinue
proced ures peoject
Lepeod (:
d
= Keroel Everrt
( =Satellite Eveolt
topical version would actually have a very limited market
potential.
.T he Assistant Director told Norman that the toxicity report
was uncertain and that he should reorient his teamtoward an
oral f ormof the compound . "T hink positive," he told Norman.
"We have tomove on and we have to take risks if we expect
to reap reward s." Market projections of the proposed oral
f ormof the d rug were presented tothe Executive Committee
of the corporation, and were enthusiastically received .
Chatman (1978) presents an elaborate f ramework f or
d iagramming narrative structure, and we can use some
of his techniques in a simplif ied f ormto show how
narrative analysis can surf ace assumptions and aid re-
f lexivity in perspective making. In d iagramming the
structure of events in a story plot, Chatman (1978)
d istinguishes between major and minor events. He calls
major events kernels and shows themas a square in his
d iagrams. Chatman ref ers to minor events as satellites
(1978, p. 54) and shows themas circles in his d iagrams.
Satellites are events which enrich the story aestheti-
cally, but are not crucial to the plot. Satellite events
"necessarily imply the existence of kernels, but not vice
versa." (Chatman 1978, p. 54). Figure 2 is a partial
d iagramof the plot of Norman's story.
In d iagramming the f irst part of this story, we have
id entif ied two kernels, treating the other elements as
satellites. Other read ers might interpret the structure
d if f erently, but that is what keeps an interpretive con-
versation lively. T he two kernels we isolate are: the trip
to the toilet and the initial toxicity report. For purposes
of an example, we will provid e a brief analysis of the
two kernels.
First Kernel: A T rip to the T oilet. Fromthis kernel,
the story could have taken several d if f erent paths.
First, Norman could have simply returned to his work
station, and waited f or the numbness to go away.
Eventually it would have and this episod e would be
over. Or, Norman might have become enraged that no
one had warned himthat the XV75 compound s could
have this ef f ect. Af ter much f inger-pointing and the
establishment of stricter chemical hand ling policies,
this story would also eventually end without a new
project being instituted . Instead , Norman used the
event of the numbness to engage in perspective taking,
looking at XV75 and his own experience f roma per-
spective other than that of hypertension or personal
d iscomf ort. In so d oing, much about the canonicallity
of the world of the lab is revealed .
First, the way the kernel is resolved shows that it is
canonical to be open to the meaning of an unexpected
event, that science will take strange twists and the
seemingly irrelevant could be the basis f or an impor-
tant new d iscovery. It is canonical in this lab f or a
scientist to take any event, no matter how bizarre or
personal, and view it as a potential f or creating new
knowled ge. Second , we learn that it is canonical to see
the event of numbness as an experiment on oneself .
T his lesson of the narrative is supported by f ield work
which conf irmed that self -experimentation is a f re-
quent practice among the lab workers. T he f irst kernel,
then, can tell us much about the values and lab prac-
tices in this community.
Second Kernel: T he Preliminary T oxicity Report. In
this kernel, we can imagine several alternatives that d id
not happen; the project could have been f ocused on a
topical version only, because of the risk of toxicity, or
the project could have been d ropped altogether. In-
stead , canonicallity is restored by a call f or positive
thinking and the lure of a large prof itable market f or
an oral version. T he tension between the market/
prof it-seeking perspective in prod uct innovation and
the toxicology perspective is lopsid ed ly mad e canonical
in f avor of the market. In this kernel there is a d istinct
f ailure of perspective taking on the part of the Assis-
tant Director. As a result, the possibilities f or knowl-
ed ge creation in this network are d iminished , the f ram-
ing of the problemis constrained , and opportunities
f or a complex exploration of how risk, reward s, toxicity
and ef f icacy can become a topic of open d ialogue are
d iminished .
T here is obviously more that could be d one in read -
ing the canonicallity of the lab in these kernels, but
these examples should suf f ice. T he important point is
that the kernel is a "hinge" in the structure of the story
and interpretation of the kernel gives access to what is
canonical in a community that may be d if f icult to
surf ace otherwise.
Implications
Perspective taking through bound ary objects is a rela-
tively unexplored f rontier in electronic communication.
One can expect that tools and med ia to support ref lex-
ivity, representation of knowled ge structures and their
exchange with others in a perspective-taking process
will increase over time. Parad oxically, it is a kind of
communication with others that grows out of an im-
proved communication with self . Communication with
one's self is the basic stance of ref lexivity; an inner
conversation that build s and ref lects upon a represen-
tation of one's und erstand ing of a situation. Being able
to d o so implies that the perspective making in a
community of knowing has progressed f ar enough to
provid e a suf f iciently strong perspective to ref lect upon.
Having had this type of communication with one's self ,
the actor is equipped to enter into a new kind of
communication with others, that of perspective taking.
We now present some examples of the types of elec-
tronic communication systems suggested by our an-
alysis thus f ar, by d escribing an id ealized f irmthat
d isplays strong capacities f or perspective making and
perspective taking.
Some Examples of Designing
Communication Systems to Support
T aking
T he implications of using inf ormation technologies to
provid e support f or perspective making and perspec-
tive taking are best und erstood as the interrelationship
of organizational, cultural and technological elements.
T his insight was evid ent in the f irst experiences with
ind ustrial research laboratories (Marcson 1960, Carlson
1992), in the Manhattan project (Davis 1969) and also
in recent stud ies of new prod uct innovation (Law and
Callon 1992, Carlson 1992, Dougherty 1992). In keep-
ing with an emphasis on how a narrative and language
game orientation can be interweaved with parad ig-
matic reasoning, this section will present a plausible,
but ad mitted ly utopian f ormof a knowled ge intensive
f irm. In this id ealized f irm, a ref lexive hermeneutic
attitud e (Gad amer 1975, Boland 1993, Boland et al.
1994) and an open recognition of language games and
the process of perspective taking is assumed to be well
established . We will f irst d escribe the technological,
organizational and cultural backd rop f or such a hypo-
thetical knowled ge-intensive f irmof the near f uture.
We will then d escribe some applications of inf ormation
technologies that could be employed f or perspective
making and perspective taking by its communities of
knowing.
T echnologically, we expect to see that computing,
imaging and communication d evices have become ubiq-
uitous. T he inf ormation environment in this hypotheti-
cal f irmis a seamless integration of multimed ia d evices
f or collection, storage, processing and d isplay. T he
organization is replete with systems based on the con-
d uit mod el and language games mod el. Once certain
kind s of knowled ge are established and the perspective
of a community of knowing becomes mature, the d eci-
sion routines are embed d ed in project management
and other kind s of sof tware, although such d ecision
premises are always subject to question and revision.
Graphics, texts, mod els, aud io and vid eo applications
are all rad ically tailorable to a user's need s. Hyperlinks
f roman element in any one application to elements in
any other application are f ully supported , making con-
textually rich, complexly layered representations the
norm. Groupware is highly d eveloped , with multimed ia
meetings, and d iscussion groups in a wid e variety of
issue f orums. A sophisticated vocabulary of electronic
f orms f or initiating, replying or commenting on d eci-
sions mod els and d iscussion topics has emerged through
an open process of structuration (Gid d ens 1979).
Organizationally, the f irmis characterized by a criti-
cal d ensity of interd epend ent knowled ge communities.
T here is a post-mod em(Harvey 1989) quality to the
organization, and groupware communication processes
are marked by multiple voices with shif ting patterns of
interest, giving a sense of a f ragmented , almost chaotic
communication environment compared to the pred om-
inantly hierarchical one of the late 1980s. T he organi-
zation uses lateral teams extensively in which the verti-
cal authority structure plays a muted role while the
principle value ad d ing activities of knowled ge creation
and knowled ge application are carried out in a chang-
ing mosaic of lateral project teams. Because of the
f irm's strong lateral f ormand collaboration-based re-
ward structure, parochial interest groups and f ief d om-
like power bases which used to subvert ef f orts at f ree
and inf ormed communication have largely d isap-
peared . Ind ivid uals who play important liaison roles
between strong communities of knowing use their newly
d eveloped skills as "semiotic brokers" (Lyotard 1984)
to help f acilitate the perspective-taking process.
Culturally, the id ea that d oing work in a knowled ge-
intensive f irmmeans perspective making and perspec-
tive taking in communities of knowing has taken hold
and has shaped both ind ivid ual and group id entities.
Ind ivid uals have a ref lexive awareness of their parad ig-
matic as well as their narrative mod es of cognition. T he
culture reinf orces an awareness of the ind ivid ual's ca-
pacity to step outsid e of a message streamand engage
in meta communicative analysis (Bateson 1972). Mem-
bers of the f irmare used to taking an interpretive
stance, playing with possible meanings, searching f or
und erlying structures, questioning the social construc-
tion of new nouns and verbs in their language games.
T hey enter into and make read ings of communication
episod es with an open awareness of the hermeneutic
circle in which they tack back and f orth f roman
interpretation of the larger context of a perspective to
an interpretation of the d etailed elements of the mes-
sage at hand (Palmer 1969). T heir hermeneutic atti-
tud e means they avoid d ebate in f avor of d ialogue
unless compelling reasons call f or a d ialectic commu-
nicative process. T hey realize that d ebate is a win-lose
polarizing strategy that rarely results in true synthesis
or creative insights. Dialogue, in contrast, is a mutually
reinf orcing, working together through language. It is a
realization that we can assume a perspective-taking
orientation and benef it f romopening ourselves to the
horizon of another.
Within the organizational, cultural and technological
environment sketched above, communities of knowing
are using ad vanced groupware f acilities to cond uct
meetings, construct multi-author d ocuments, and coor-
d inate their promises and d ead lines, all with the capa-
bility to access d ata and knowled ge through a world -
wid e network of knowled ge repositories. As is true
tod ay, the groupware systems are composed of a series
of f orums which serve as "containers" f or d ialogue on
certain topics, issues, concerns, projects or tasks. Fo-
rums ref lect the way the knowled ge work is being
f ocused , and the kind s of knowled ge structures that are
emerging in the f irm, and are thus one avenue into its
communities of knowing. Ind ivid uals participate in
many f orums in an evolving pattern. T he lateral groups
to which they belong and their unique expertise d ef ines
the types and kind s of f orums in which they partici-
pate.
What takes place in these f orums are language
games. T he mod e of cognition is a mixed one in which
parad igmatic reasoning is interwoven with stories and
narration. T he applications of ad vanced inf ormation
technologies f or perspective making and perspective
taking that we d escribe below d epend upon there being
a higher level of ref lexivity in knowled ge intensive
f irms than is presently the case. T hese communication
systems d epend not only on talking about issues and
problems within a groupware environment, but on talk-
ing about how they are talking (Bateson 1972). It
d epend s upon a critical hermeneutic attitud e in which
the strangeness and multiple possibilities f or making
meaning in our conversations are constantly in our
awareness (Gad amer 1975, Ricoeur 1981, Boland 1993).
As groups f ormand ref ormin a knowled ge-intensive
f irmemploying a lateral organization structure, we
anticipate f ive new classes of electronic communication
f orums as examples of ones that would enhance the
processes of perspective making and perspective tak-
ing. Within each class there would be several d if f erent
types of f orums as we will d iscuss below. T he f ive new
classes of f orums we propose as examples are:
1. T ask Narrative Forums;
2. Knowled ge Representation Forums;
3. Interpretive Read ing Forums;
4. T heory Build ing Forums;
5. Intelligent Agent Forums.
T ask Narrative Forums
T his type of f orumhas been envisaged by Brown and
Duguid (1991) and Galbraith (1994) among others, and
is an implicit recognition of the importance of narra-
tivizing our experience and sharing the narratives with
others in our community. T hrough narrative, the com-
munity constructs its practices and its social world by
build ing and restoring its sense of the canonical. Nar-
rative, by making the implicit and the tacit inf erable to
the read er or the listener, is a critically important f irst
step in achieving perspective taking within and among
communities of knowing. Because these task narratives
would be multimed ia, and includ e vid eo and aud io,
they enable the benef its of learning by experience to
extend beyond normal constraints of space and time.
T ask-narrative f orums serve as perspective making
f or those creating the narratives and also serve as a
perspective-taking experience f or those read ing the
narratives. T he narrative is always incomplete and the
read er must "read into" the story in making it sensible.
Bruner (1990) ref ers to this "read ing into" as a sub-
junctive process and is a primary vehicle f or opening
oneself up to the perspective of another and making
real its possibilities f or seeing the world d if f erently.
Knowled ge Representation Forums
Current groupware enables linking f roma text d ocu-
ment to a spread sheet, d ecision mod el, graphic d epic-
tion, or picture. Once a d ocument is hyper-linked in
this way, the context it carries with itself is enriched
and its possibilities f or interpretation are increased . So
we are alread y used to seeing a message with other
d ocuments linked to or embed d ed within it. A knowl-
ed ge representational f orum, in contrast, is one which
f ocuses on the und erstand ing that lies behind such
Figure 3 Perspective Making and Perspective T aking
Community
of Cmuiyo
Knowing A' Knowing'B'
Perspective Makiing Perspective Making
- Narratives of experience - Narratives of experience
- Parad igmatic analysis - Parad igmatic analysis
- Ref lexivity and - Ref lexivity and
Representation Representation
Perspective T aking Perspective T aking
- Ref lexivity and - Ref lexivity and
nterpretive Read ing Interpretive Read i
Bound ary
Ob'jects
- Cause Maps
- Narrative miaps
- Mod els
- Classif ication
Schemes
complex d ocuments. It is a f orumthat captures a
community's cooperative ef f orts to ref lect upon, inter-
pret and d epict an und erstand ing of their situation to
themselves.
It is not a problem-solving or task-practice f orumso
much as a sense-making f orumin which the objects of
d iscussion are visual representations of their und er-
stand ing of a situation, a problemor an objective. It is
an openly ref lexive f orumin which communities of
knowing explicitly talk about their und erstand ings. Such
f orums could use storyboard s in which still or animated
pictures are assembled in a sequence, in a kind of
visual d epiction of an und erstand ing, or these f orums
could use cause maps as in Figure 1, or other d iagrams
and mod els f or representing an und erstand ing.
Like narrative f orums, representational f orums serve
as a perspective-making experience f or those construct-
ing, revising, or commenting on an emerging represen-
tation within a community of knowing. T hey also serve
as a perspective-taking experience f or those who read
themwith a hermeneutic attitud e of engaging the
horizons of another thought world (see Figure 3).
Interpretive Read ing Forums
Whereas representation f orums are overtly ref lexive in
that participants are trying to ref lect upon their current
state of und erstand ing of some issue, interpretive read -
ing f orums are a space f or ref lecting upon the assump-
tions and meanings revealed by the communications in
other f orums. In this f orum, participants are subjecting
other texts to re-read ings in hopes of portraying the
tacit and implicit meanings characterizing a community
of knowing, their own or others' (Czarniawska-Joerges
and Guillet d e Monthoux 1994).
Discussions in this f orumcould resemble d ialogues
of literary criticisms in which critical-ref lexive read ings
are mad e of the streams of entries in other f orums.
Eventually, re-read ings could even be mad e of the
d ialogues in the interpretive read ing f orumitself , as
layers of ref lexivity begin to compound (Ashmore 1989).
In ad d ition to interpretive essays, such a f orumwould
also be used f or d iscussing the narrative structures in
the task narratives f orumand in the narrativization
that is evid ent in all other f orums as well. Here, the
d iagramming of narrative structures, the isolation of
kernels, and the unpacking of how the canonical and
noncanonical are revealed f or d if f erent communities of
knowing would take place.
Another type of f orumwithin this class could be
f ocused on word s and might be known as a word -talk
f orum. T his would be another type of ref lexive analysis
in which word s (new word s, especially nouns and verbs,
as well as f amiliar ones) were systematically consid ered
as to their changing meanings and uses, their shif ting
contexts and connotations, and the implicit and tacit
assumptions they reveal.
T he interpretive read ing f orums d iscussed above are
the most explicitly hermeneutic-interpretive ones we
envision and have the greatest d epend ence on the
supportive organizational and cultural qualities d is-
cussed above. Without a wid ely shared sense of the
importance of perspective making and perspective tak-
ing in knowled ge creation, and a well established sense
of the value of a hermeneutic attitud e, these f orums
would not be possible.
T heory-build ing Forums
T hese f orums would most closely ref lect the d ialogue
of theory that is woven throughout scientif ic practice.
We envision this as a series of f orums in which d if f er-
ent communities of knowing articulate, critique, extend
and explore the theories that d o or should guid e their
work. T heory-build ing f orums are not just f or science
work, however, and we would anticipate that in a
knowled ge-intensive f irmthe ethos of perspective mak-
ing and perspective taking also would be held by the
f inancial, marketing and other nonscientif ic f ield s of
d iscourse within the organization, and between the
organization and its many environments.
In ad d ition to "T heory Corners", or f orums d ed i-
cated to d ialogue on theories within and across specif ic
communities of knowing, we would also expect theory-
build ing f orums to includ e "T hought Experiment" f o-
rums, where ind ivid uals played with theories and their
implications by narrativizing thought experiments.
T hought experiments entail the construction of mental
mod els by a scientist who imagines a sequence of
events, and then uses a "narrative" f ormto d escribe
the sequence in ord er to communicate the experiment
to others (Nersessian 1992). T hought-experiment f o-
rums would also includ e the construction and playf ul
exploration of simulation mod els, especially multime-
d ia simulations and virtual reality systems.
Once again, we see this class of f orums playing an
important role in perspective making and also perspec-
tive taking. Constructing theories and cond ucting
thought experiments in d ialogue within a community of
knowing is essential f or strong f orms of perspective
making. Participating in these f orums f romthe f ringes
of the community, or read ing and interpreting the
theory build ing f romoutsid e the community is a pow-
erf ul means of perspective taking.
Intelligent Agent and Expert SystemForums
T he f inal class of conversational f orums we will present
have to d o with intelligent agents and expert systems.
By intelligent agents we mean personal assistants in
the f ormof sof tware systems that can roamthe net-
work of f orums within a f irmas well as libraries,
repositories and inf ormation sources outsid e the f irm.
T hese agents help an ind ivid ual to d irect her attention
within the burgeoning f ield of f orums that could be of
importance and interest to her, and also help assemble
contextual materials f or build ing cross-d ocument links
in complexly layered representations.
We see these intelligent agents as well as expert
systems as another important f ormof ref lexivity within
the f irm. Both are classes of sof tware systems where
expertise and interests have been ref lected upon, mad e
visible and embod ied within these artif icial agents. In
the f orums we envisage, ind ivid uals would d evelop and
share insights into how these systems can best be
constructed and d eployed , and how their results can
best be interpreted . T he f orums would thus be d irectly
concerned with thinking about thinking, especially as
thought processes are embod ied in the active agents
and expert systems.
Summary
T he applications d escribed above are not intend ed as
an exhaustive listing, nor as a taxonomy of ways in
which perspective making and perspective taking can
be supported by electronic communication. Rather, we
intend merely to open a d iscussion of some possibilities
f or appreciating language games, narrative cognition
and ref lexivity in the d esign of electronic communica-
tion systems. No d oubt, many of these kind s of activi-
ties are alread y being explored in nascent f ormby
those organizations that are installing extensive group-
ware capabilities.
Conclud ing T houghts
Any d esign of an electronic communication system
implies a mod el of human communication and of hu-
man cognition. We have explored how principles and
policies f or the d esign of electronic communication
systems are af f ected by incorporating a language games
mod el of communication and an awareness* of the
narrative mod e of cognition. In so d oing, we have
argued that perspective making and perspective taking
in the science work of knowled ge-intensive f irms and in
f irms generally using lateral organizational f orms would
benef it f romsystems d esigned with this sensibility in
mind . We have also provid ed some examples of com-
puter-based applications which embod y these id eas
and the types of organizational, cultural and technolog-
ical precond itions f or themto succeed .
Behind all the arguments and examples we have
presented is an interest in strengthening the possibili-
ties f or perspective making and perspective taking in
communities of knowing. Making a strong perspective
and having the capacity to take another perspective
into account are the means by which more complexi-
f ied knowled ge and improved possibilities f or prod uct
or process innovation are achieved . Our d iscussion has
necessarily emphasized narrative cognition at the ex-
pense of parad igmatic, and ref lexivity at the expense of
action. T his was necessary, we believe, in ord er to open
a space f or d ialogue on these of ten overlooked f eatures
of social lif e that are none the less central to creating
knowled ge in communities and organizations.
Some of our assumptions about organization culture
and power as well as our proposed applications may
seemoptimistic about the chances of overcoming d eeply
entrenched f ears and resistance to change in organiza-
tions and their communities of knowing. We believe,
however, that d esigning electronic communication sys-
tems with a language game mod el of communication
that explicitly incorporates a narrative mod e of cogni-
tion and heightened levels of ref lexivity is an important
means to achieve just those organizational qualities we
are assuming.
Acknowled gements
T his work has been supported by National Science Found ation
Grant #IRI-9015526 and a grant f romthe T RW Found ation. T he
authors gratef ully acknowled ge the helpf ul suggestions mad e by
Katrine Kirk, Kalle Lyytinen, Ulrike Schultze and f our anonymous
reviewers on earlier d raf ts of this manuscript.
Ref erences
Allen, T . J. (1986), "Organizational Structure, Inf ormation T echnol-
ogy, and R & D Prod uctivity," IEEE T ransactions on Engineering
Management, EM-33, 4, 212-220.
and S. Cohen (1969), "Inf ormation Flows in R & D Laborato-
ries," Ad ministrative Science Quarterly, 12, 12-19.
-, L. T ushman and D. Lee (1980), "T echnology T ransf er as a
Function of Position in the SpectrumResearch through Devel-
opment toT echnical Services," Acad emy of Management Joumal,
22, 4, 694-708.
Argyris, C. and D. Sch6n (1978), Organizational Leaming: A T heory
of Action Perspective, Read ing, MA: Ad d ison-Wesley.
Ashmore, M. (1989), T he Ref lexive T hesis: Wrighting Sociology of
Scientif ic Knowled ge, Chicago, IL: University of ChicagoPress.
Astley, W. G. and R. F. Zammuto(1992), "Organization Science,
Managers, and Language Games," Organization Science, 3, 4,
443-460.
Axelrod , R. (1976), Structure of Decision, Princeton, NJ: Princeton
University Press.
Bakhtin, M. M. (1981), "Discourse in the Novel," in M. Holquis
(Ed .), T he Dialogic Imagination, Austin, T X: University of T exas
Press.
Barnes, B. (1983), "On the Conventional Character of Knowled ge
and Cognition," in K. D. Knorr-Cetina and M. Mulkay (Ed s.),
Science Observed , Lond on, UK: Sage Publications.
Bartunek, J. (1984), "Changing Interpretive Schemes and Organiza-
tional Restructuring: T he Example of a Religious Ord er," Ad -
ministrative Science Quarterly, 29, 355-372.
Bateson, G. (1972), Steps toan Ecology of Mind , New York: Rand om
House.
Bennis, W. and B. Nanus (1985), Lead ers: T he Strategies f or T aking
Charge, New York: Harper.
Bertad o, R. (1990), "Implementing a Strategic Vision," Long Range
Planning, 23, 5, 22-30.
Boje, D. M. (1991), "T he Storytelling Organization: A Stud y of Story
Perf ormance in an Of f ice-Supply Firm," Ad ministrative Science
Quarterly, 36, 106-126.
Boland , R. J. (1993), "Accounting and the Interpretive Act," Ac-
counting, Organizations and Society, 18, 2/3, 125-146.
and L. R. Pond y (1983), "Accounting in Organizations: T oward
a Union of Rational and Natural Perspectives," Accounting,
Organizations and Society, 8, 2/3, 223-234.
-, D. Schwartz, R. V. T enkasi, A. Maheshwari and D. T e'eni
(1992), "Sharing Perspectives in Distributed Decision Making,"
Association of Computing Machinery Conf erence on Computer
Supported Cooperative Work," T oronto, Ontario, Canad a,
306-313.
, R. V. T enkasi and D. T e'eni (1994), "Designing Inf ormation
T echnology toSupport Distributed Cognition," Organization Sci-
ence, 5, 3, 456-475.
Bould ing, K. E. (1956), T he Image: Knowled ge in Lif e and Society,
Ann Arbor, MI: University of Michigan Press.
Brad shaw, G. F. (1992), "T he Airplane and the Logic of Invention,"
in R. N. Giere (Ed .), Cognitive Mod els of Science: Minnesota
Stud ies in the Philosophy of Science, XV, Minneapolis, MN:
University of Minnesota Press.
Brehmer, B. (1991), "Distributed Decision Making: Some Notes on
the Literature," in Rasmussen et al. (Ed s.), Distributed Decision
Making: Cognitive Mod els in Cooperative Work, Blackwell, NY:
John Wiley.
Brown, C. V. and R. P. Bostrom(1989), "Ef f ective Management of
End -User Computing: A T otal Organization Perspective," Jour-
nal of MIS, 6, 2, Fall, 77-92.
Brown, J. S. (1991), "Research that Reinvents the Corporation,"
Harvard Business Review, Jan.-Feb., 102-111.
and P. Duguid (1991), "Organizational Learning and Commu-
nities-of -Practice: T oward a Unif ied View of Working, Learning,
and Innovation," Organization Science, 2, 1, 40-57.
Brown, R. (1981), Social Psychology, New York: T he Free Press.
Bruner, J. S. (1986), Actual Mind s, Possible World s, Cambrid ge, MA:
Harvard University Press.
(1990), Acts of Meaning, Cambrid ge, MA: Harvard University
Press.
Burgess, G. M., T . D. Clark, Jr., R. D. Hauser, Jr., and R. W. Zmud
(1992), "T he Application of Causal Maps toDevelop a Collec-
tive Und erstand ing of Complex Organizational Contexts in Re-
quirements Analysis," Accounting, Management and Inf ormation
T echnologies, 2, 3, 143-164.
Burrell, G. and G. Morgan (1979), Sociological Parad igms and Organi-
zationalAnalysis, Lond on, UK: Heinemann.
Carlson, W. B. (1992), "Artif acts and Frames of Meaning: T homas
A. Ed ison, His Managers and the Cultural Construction of
Motion Pictures," in W. Bijker and J. Law (Ed s.), Shaping
T echnology/Build ing Society, Cambrid ge, MA: MIT Press.
Chatman, S. (1978), Story and Discourse in Fiction and Film, Ithaca,
NY: Cornell University Press.
Chen, P. P. S. (1976), "T he Entity Relationship Diagram-T oward a
Unif ied View of Data," ACM T ransactions on Database Systems,
1, 1, 9-36.
Ciborra, C., P. Migliarese and P. Romano(1984), "A Method ological
Inquiry of Organizational Noise in Sociotechnical Systems," Hu-
man Relations, 37, 8, 565-588.
Clark, B. R. (1972), "T he Organizational Saga in Higher Ed ucation,"
Ad ministrative Science Quarterly, 17, 178-184.
Clark, H. H. (1985), "Language Use and Language Users," in
G. Lind zey and E. Aronson (Ed s.), Hand book of Social Psychol-
ogy, New York: Rand omHouse.
Collins, H. M. (1983), "An Empirical Relativist Programme in the
Sociology of Scientif ic Knowled ge," in K. D. Knorr-Cetina and
M. Mulkay (Ed s.), Science Observed , Lond on, UK: Sage Publica-
tions.
Collins, J. C. and J. I. Porras (1991), "Organizational Vision and
Visionary Organization," Calif ornia Management Review, 34,
30-42.
Czarniawska-Joerges, B. (1992), Exploring Complex Organizations,
Newbury Park, CA: Sage Publications, Inc.
and P. Guillet d e Monthoux (Ed s.) (1994), Good Novels, Better
Management: Read ing Organizational Realities, Switzerland : Har-
wood Acad emic Publishers.
Daf t, R. and R. Lengel (1984), "Inf ormation Richness: A New
Approach toManagerial Behavior and Organizational Design,"
in L. L. Cummings and B. M. Staw (Ed s.), Research in Organiza-
tional Behavior, 6, Greenwich, CT : JAI Press, 191-223.
Davis, N. P. (1969), Lawrence and Oppenheimer, Lond on, UK:
Jonathan Cape.
Davis, P. and M. Wilkof (1988), "Scientif ic and T echnical Inf orma-
tion T ransf er f or High T echnology: Keeping the Figure in Its
Ground ," Research and Development Management, 18, 1, 45-55.
Deng, P. S. and A. Chaud hry (1992), "A Conceptual Mod el of
Ad aptive Knowled ge-based System," Inf ormation Systems Re-
search, June, 127-149.
Dolk, D. R. (1988), "Mod el Management and Structured Mod eling:
T he Role of Inf ormation Resource Directory System," Commu-
nications of the ACM, June, 704-718.
Dougherty, D. (1992), "Interpretive Barriers toSuccessf ul Prod uct
Innovation in Large Firms," Organization Science, 3, 2, 179-202.
Duncan, R. and A. Weiss (1979), "Organizational Learning: Implica-
tions f or Organizational Design," in L. L. Cummings and B. M.
Staw (Ed s.), Research in Organizational Behavior, 1, Greenwich,
CT : JAI Press.
Ed en, C. (1992), "On the Nature of Cognitive Maps," Journal of
Management Stud ies, 29, 3, 261-265.
Elam, J. J. and B. Konsynski (1987), "Using Artif icial Intelligence
T echniques toEnhance the Capabilities of Mod el Management
Systems," Decision Sciences, 18, 487-501.
Eysenck, M. W. and M. T . Keane (1990), Cognitive Psychology: A
Stud ent's Hand book, Hove and Lond on, UK: Lawrence Erlbaum
Associates.
Fiol, C. M. and A. S. Huf f (1992), "Maps f or Managers: Where are
We? Where d o We Go FromHere?" Journal of Management
Stud ies, 29, 3.
Fish, S. (1980), Is T here a T ext in this Class? Cambrid ge, MA:
Harvard University Press.
Fleck, L. (1979), Genesis and Development of a Scientif ic Fact, in T . J.
T renn and R. K. Merton (Ed s.), Chicago, IL: University of
ChicagoPress (First published in German, 1935).
Gabriel, Y. (1991), "T urning Facts intoStories and Stories into
Facts: A Hermeneutic Exploration of Organizational Folklore,"
Human Relations, 44, 8, 857-876.
Gad amer, H. G. (1975), T ruth and Method , New York: Seabury.
Galbraith, J. (1973), Designing Complex Organizations, Read ing, MA:
Ad d ison-Wesley.
Galbraith, J. R. (1994), Competing with Flexible Lateral Organizations,
Read ing, MA: Ad d ison-Wesley.
and E. E. Lawler (1993), Organizing f or the Future, San Fran-
cisco, CA: Jossey-Bass Publishers.
Gentner, D. (1983), "Structure-Mapping: A T heoretical Framework
f or Analogy," Cognitive Science, 7, 155-170.
Geof f rion, A. M. (1987), "An Introd uction toStructural Mod eling,"
Management Science, 33, 5, 547-588.
Gid d ens, A. (1976), New Rules of Sociological Method , New York:
Basic Books.
(1979), Central Problems in Social T heory, Lond on, UK:
MacMillan.
(1991), Mod ernity and Self -Id entity, Stanf ord , CA: Stanf ord
University Press.
Granstrand , O., E. Bohlin, C. Oskarsson and N. Sjoberg (1992),
"External T echnology Acquisition in Large Multi-technology
Corporations," Research and Development Management, 22, 2,
111-133.
Hamel, G. and C. K. Prahalad (1991), "Corporate Imagination and
Exped itionary Marketing," Harvard Business Review, 69, 4,
July-Aug., 81-92.
Harvey, D. (1989), T he Cond ition of Postmod ernity, Oxf ord , UK: Basil
Blackwell.
Hend erson, R. (1994), "Managing Innovation in the Inf ormation
Age," Harvard Business Review, January-February, 100-105.
Hesse, M. (1974), T he Structure of Scientif ic Inf erence, Berkeley, CA:
University of Calif ornia Press.
Huf f , A. S. (1990), Mapping Strategic T hought, Chichester, UK: John
Wiley & Sons.
Isaacs, W. N. (1993), "T aking Flight: Dialogue, Collective T hinking,
and Organizational Learning," Organizational Dynamics, 22, 2,
24-39.
Jonsson, S. (1992), "Accounting f or Improvement: Action Research
on Local Management Support," Accounting, Management and
Inf ormation T echnologies, 2, 2, 99-116.
Kahneman, D., P. Slovic and A. T versky (1982), Jud gment und er
Uncertainty: Heuristics and Biases, New York: Cambrid ge Uni-
versity Press.
Knorr-Cetina, K. D. (1981), T he Manuf acture of Knowled ge: An Essay
on the Constructivist and Contextual Nature of Science, Oxf ord ,
UK: Pergamon Press.
Krauss, R. M. and S. R. Fussell (1991), "Perspective-T aking in
Communication Representation of Others' Knowled ge in Ref er-
ence," Social Cognition, 9, 1, 2-24.
Kuhn, T . S. (1962), T he Structure of Scientif ic Revolutions, Chicago,
IL: University of ChicagoPress.
(1970), T he Structure of Scientif ic Revolutions, (2nd ed .), Chicago,
IL: University of ChicagoPress.
Lave, J. and E. Wenger (1990), "Situated Learning: Legitimate
Peripheral Participation," IRL Report 90-0013, PaloAlto, CA:
Institute f or Research on Learning.
Law, J. and M. Callon (1992), "T he Lif e and Death of an Aircraf t: A
Network Analysis of T echnical Change," in W. Bijker and J. Law
(Ed s.), Shaping T echnology/Build ing Society, Cambrid ge, MA:
MIT Press.
Lawler, E. E. (1992), T he Ultimate Ad vantage, San Francisco, CA:
Jossey-Bass Publishers.
Lyotard , J. F. (1984), T he Postmod emCond ition: A Report on Knowl-
ed ge, Minneapolis, MN: University of Minnesota Press.
Marcson, S. (1960), T he Scientist in American Ind ustry, New York:
Harper & Brothers.
Martin, J. and D. Meyerson (1988), "Organizational Cultures and
the Denial, Channeling, and Acknowled gment of Ambiguity,"
in L. R. Pond y, R. J. Boland , and H. T homas (Ed s.), Managing
Ambiguity and Change, New York: Wiley, 93-125.
Masterman, M. (1970), "T he Nature of a Parad igm," in I. Lakatos
and A. Musgrave (Ed s.), Criticismand the Growth of Knowled ge,
Lond on, UK: Cambrid ge University Press.
McConkie, M. L. and W. R. Boss (1986), "Organizational Stories:
One Means of Moving the Inf ormal Organization During Change
Ef f orts," Public Ad ministration Quarterly, 10, 2, 189-205.
Mead , G. H. (1934), Mind , Self and Society, Chicago, IL: University
of ChicagoPress.
Mitrof f , I. I. and R. H. Kilmann (1976), "On Organization Stories:
An Approach to the Design and Analysis of Organizations
through Myth and Stories," in R. H. Kilmann, L. R. Pond y, and
D. P. Slevin (Ed s.), T he Management of Organization Design, New
York: North-Holland .
Mulkay, M., J. Potter and S. Yearley (1983), "Why an Analysis of
Scientif ic Discourse Is Need ed ," in K. D. Knorr-Cetina and M.
Mulkay (Ed s.), Science Observed , Lond on, UK: Sage Publica-
tions.
Munro, M. C., S. L. Huf f and G. Moore (1987), "Expansion and
Control of End -User Computing," Journal of MIS, 4, 3, Winter,
5-27.
Nersessian, N. J. (1992), "How DoScientists T hink? Capturing the
Dynamics of Conceptual Change in Science," in R. N. Giere
(Ed .), Cognitive Mod els of Science: Minnesota Stud ies in the
Philosophy of Science, XV. Minneapolis, MN: University of Min-
nesota Press.
Nisbett, R. E. and L. E. Ross (1980), Human Inf erence: Strategies and
Shortcomings of Social Jud gment, Englewood Clif f s, NJ:
Prentice-Hall.
Nonaka, I. (1994), "A Dynamic T heory of Organizational Knowled ge
Creation," Organization Science, 5, 1, 14-37.
and J. Johansson (1985), "Japanese Management: What about
the 'Hard ' Skills?" Acad emy of Management Review, 10, 2,
181-191.
Orr, J. (1990), "Sharing Knowled ge, Celebrating Id entity: War Sto-
ries and Community Memory in a Service Culture," in D. S.
Mid d leton and D. Ed ward s (Ed s.), Collective Remembering:
Memory in Society, Beverly Hills, CA: Sage Publications.
Palmer, R. E. (1969), Hermeneutics, Evanston, IL: Northwestern
University Press.
Pava, C. (1983), Managing New Of f ice T echnology, New York: Free
Press.
Pepper, S. C. (1942), World Hypotheses, Berkeley, CA: University of
Calif ornia Press.
Polanyi, M. (1967), T he T acit Dimension, Gard en City, NY: Double-
d ay.
Purser, R. E. and W. A. Pasmore (1992), "Organizing f or Learning,"
in R. Wood man and W. A. Pasmore (Ed s.), Research in Organi-
zational Change and Development, 6, Greenwich, CT : JAI Press.
_ and R. V. T enkasi (1992), "T he Inf luence of Delibera-
tions on Learning in New Prod uct Development T eams," Journal
of Engineering and T echnology Management, 9, 1-28.
Ranson, S., R. Hinings and R. Greenwood (1980), "T he Structuring
of Organizational Structures," Ad ministrative Science Quarterly,
25, 2, 1-17.
Red d ing, C. W. (1972), Communication within the Organization, New
York: Ind ustrial Communication Council.
Red d y, M. J. (1979), "T he Cond uit Metaphor," in A. Ortony (Ed .),
Metaphor and T hought, Cambrid ge, UK: Cambrid ge University
Press, 284-324.
Richard son, G. L., B. M. Jackson and G. W. Dickson (1990), "A
Principles Based Enterprise Architecture: Lessons f romT exaco
and Star Enterprise," MIS Quarterly, Dec., 385-403.
Ricoeur, P. (1981), Ed . and translated by J. T hompson, Hermeneutics
and the Human Sciences, Cambrid ge, UK: Cambrid ge University
Press.
Rivard , S., and S. L. Huf f (1988), Communications of the ACM, 31, 5,
552-561.
Rommetveit, R. (1980), "On 'Meanings' of Acts and What is Meant
By What is Said in a Pluralistic Social World ," in M. Brenner
(Ed .), T he Structure of Action, Oxf ord , UK: Blackwell and Mott,
108-149.
Ross, L., D. Greene and P. House (1977), "T he False Consensus
Phenomenon: An Attributional Bias in Self -Perception and So-
cial Perception Processes," Journal of Experimental Social Psy-
chology, 13, 279-301.
Rubinstein, R. A., C. D. Laughlin and D. McManus (1984), Science
as Cognitive Process: T oward an Empirical Philosophy of Science,
Philad elphia, PA: University of Pennsylvania Press.
Scheer, A.-W. (1992), Architecture of Integrated Inf ormation Systems:
Found ations of Enterprise-Mod elling, Berlin, Germany: Springer-
Verlag.
Sch6n, D. (1979), "Generative Metaphor," in A. Ortony (Ed .),
Metaphor and T hought, Cambrid ge, UK: Cambrid ge University
Press.
Schutz, A. (1964), Stud ies in Social T heory, Collected Papers, II, T he
Hague, T he Netherland s: Martinus Nijhof f .
Shannon, C. and W. Weaver (1949), T he Mathematical T heory of
Communication, Chicago, IL: University of Illinois Press.
Shrivastava, P. (1983), "A T ypology of Organizational Learning Sys-
tems," Journal of Management Stud ies, 20, 1, 9-25.
Simon, H. A. (1977), T he New Science of Management Decision, 2nd
rev., Englewood Clif f s, NJ: Prentice Hall.
Smith, C. S. (1981), "Structural Hierarchy in Science, Art and
History," in J. Wechsler (Ed .), On Aesthetics in Science, Cam-
brid ge, MA: MIT Press.
Star, S. L. (1989), "T he Structure of Ill-Structured Solutions: Bound -
ary Objects and Heterogeneous Distributed ProblemSolving,"
in M. Huhns and L. Gasser (Ed s.), Read ings in Distributed
Artif icial Intelligence 2, Menlo Park, CA: Morgan Kauf mann.
(1993), "Cooperation Without Consensus in Scientif ic Problem
Solving: Dynamics of Closure in Open Systems," in S. Easter-
brook (Ed .), CSCW: Cooperation or Conf lict, Lond on: UK:
Springer.
Starbuck, W. H. (1992), "Learning by Knowled ge-Intensive Firms,"
Journal of Management Stud ies, 29, 6, 713-740.
Steed man, M. J., and P. N. Johnson-Laird (1980), "T he Prod uctions
of Sentences, Utterances and Speech Acts: Have Computers
Anything toSay?" in B. Butterworth (Ed .), Language Prod uc-
tions: Speech and T alk, Lond on, UK: Acad emic Press.
T argowski, A. S. (1988), "Strategic Planning f or the Enterprise-Wid e
Inf ormation Management Complex: T he Architectural Ap-
proach," Journal of MIS, 5, 2, Fall.
T enkasi, R. V. and R. J. Boland (1993), "Locating Meaning Making
in Organizational Learning: T he Narrative Basis of Cognition,"
in R. W. Wood man and W. A. Pasmore (Ed s.), Research in
Organizational Change and Development, 7, Greenwich, CT : JAI
Press, 77-103.
T ushman, M. L. (1978), "Inf ormation Processing as an Integrating
Concept in Organizational Design," Acad emy of Management
Review, 3, 3, 613-624.
von Bertalanf f y, L. (1968), General SystemT heory, New York: George
Braziller Inc.
Wad d ington, C. H. (1957), T he Strategy of the Genes, Lond on, UK:
Allen & Unwin.
Weick, K. E. (1990), "Cognitive Processes in Organizations," in L. L.
Cummings and B. M. Staw (Ed s.), Inf ormation and Cognition in
Organizations, Greenwich, Cr: JAI Press.
and M. K. Bougon (1986), "Organizations as Cognitive Maps:
Charting Ways toSuccess and Failure," in H. Sims and D. Goia
(Ed s.), T he T hinking Organization, San Francisco, CA: Jossey-
Bass.
and K. H. Roberts (1993), "Collective Mind s in Organizations:
Heed f ul Interrelating on Flight Decks," Ad ministrative Science
Quarterly, 38, 357-381.
Werner, H. (1957), "T he Concept of Development f roma Compara-
tive and Organismic Point of View," in D. Harris (Ed .), T he
Concept of Development, Minneapolis, MN: University of Min-
nesota Press.
Wiener, N. (1954), T he Human Use of Human Beings, New York:
Avon Books.
Wittgenstein, L. (1961), T ractatus Logico-Philosophicus, Lond on, UK:
Routled ge & Kegan Paul (First English ed ition, with a transla-
tion, 1922).
(1974), Philosophical Investigations, New York: MacMillan (First
ed ition. 1953).
Accepted by Janet Fulk and Gerard ine DeSanctis; received September 1993. T his paper has been with the authors f or tworevisions.

Perspective Making and Perspective Taking in Communities of Knowing.1995

Diunggah oleh

Informasi Dokumen

Hak Cipta

Format Tersedia

Bagikan dokumen Ini

Bagikan atau Tanam Dokumen

Opsi Berbagi

Apakah menurut Anda dokumen ini bermanfaat?

Apakah konten ini tidak pantas?

Hak Cipta:

Format Tersedia

Perspective Making and Perspective Taking in Communities of Knowing.1995

Diunggah oleh

Hak Cipta:

Format Tersedia

Perspective Making and Perspective Taking in Communities of Knowing

Author(s): Richard J. Boland, Jr. and Ramkrishnan V. Tenkasi

Anda mungkin juga menyukai