Strategic Management Journal, Vol.

15, 53-71 (1994)

i- L AN ESSAY ON CORPORATE EPISTEMOLOGY

GEORG VON KROGH and JOHAN ROOS
Norwegian School of Management, Sandvika, Norway
KEN SLOCUM
SENCORP, Newpott, Kentucky, U.S.A.

The objective of this essay is to contribute to a new perspective of strategic management by

developing a new theory of organizational knowledge. The article focuses on how managers
can understand and guide knowledge development processes in organizations. Our
epistemology broadens strategic management to also include the advancement activities of
the organization. In addition to discussing development of organizational knowledge, the
essay also emphasises fundamental consequences for research methodology.

INTRODUCTION core questions. What is knowledge and how does

it develop as well as what are the conditions for
Dear reader, knowledge to develop? The objective of this
Please try to forget the reality you have previously article is to develop a new corporate epistemology
constructed and let yourself be open to the which can consequently contribute to a new
signals this article carries. These signals are truly perspective of strategic management.
distinct from those in previous articles and books Chakravarthy and Doz underscored that
within the strategic management field. The research within our field must become more
starting point is how managers understand and relevant to practice. They called for research
ensure knowledge development in organizations, that involves multidisciplinary team work, that
and the theoretical lens is autopoiesis theory. has focus on corporate strategy processes, and
In our opinion, rethinking the strategy para- that is action-research oriented (1992: 9-10).
digm implies rethinking how we view the organiza- The research presented in this paper not only
tion. In this article, rethinking the organization fulfills these criteria, it adds to the authors' line
means developing a new theory of organizational of thinking. It has been conceptualized and
knowledge, that is, a corporate epistemology. A thematized by one practitioner and two strategy
branch of one of the grand divisions of philosophy, professors, and, in addition to addressing an
methodology, epistemology deals with the ways important managerial issue and drawing on
of interpreting knowledge, i.e., the ways of theoretical lenses perhaps unknown to many
knowing (Montague, 1962).' 'Corporate episte- readers of the Strategic Management Journal, it
mology' is the theory on how and why organiza- is built on a new methodological approach.2
tions know. It therefore must deal with some Our theme is inspired by the tremendous

* Underlying our arguments is the notion of knowledge as

Key words: Autopoiesis, languaging, organizational being socially constructed; the environment which managers
knowledge, scaling, self-reference respond to is determined by previous experience, not by
observable 'objective' facts (e.g., Weick, 1979). Although
Montague (1962) defined epistemology as the extent '. . .the the setting described here is strategic management, we have
things and qualities of the world are dependent upon their chosen to include all sorts of purpositive organizations rather
being related as objects to a knower or subject' (p. 82). than restricting our discussions to firms.

CCC 0142-2095/94/O90053-19
@ 1994 by John Wiley & Sons, Ltd.
54 G. von Krogh, J . Roos and K . Slocum

transformations in contemporary society and the organization, as we have come to know it

economics and in the phenomenon we call the during the past century, is an inaccurate concept.
organization. Indicators of these transformations Likewise, our management practices and theories
are many. In education, knowledge has become need to be discarded, altered, or reinvented in
increasingly differentiated as a result of increasing order to give adequate descriptions and provide
efforts in research and development: there is no appropriate heuristics that can guide managers
longer a ‘right knowledge,’ but many coexisting in this new knowledge-intensive era. That is
conflicting pieces of knowledge. In turn, the exactly what this paper will address.
contest between different elements of knowledge The essay consists of three main sections.
continually increases the complexity of total First, our methodological approach is discussed.
knowledge conveyed through education (Lyotard, Second, our speculation on a new corporate
1984; Lawson and Appignanesi, 1989; Hage epistemology is presented. Building on this
and Powers, 1992). Further indicators may be epistemology, in the third section we discuss the
discovered in the arts, which, according to further advancement of the realm of strategic
Sakaiya (1992: 148), ‘. .is a field of human management.
endeavor that tends to be in the forefront of any
imminent social transformation.’ Contemporary
art presents itself in a way that is different from METHODOLOGY: A PROCESS OF
historical art; it combines elements from many MATCHING
epochs in new ways (Lyotard, 1984). In philos-
ophy, the coming of a postmodern age and its Not surprisingly, theorists in the field to an
implications for individuals and societies at large increasing extent seem to acknowledge that the
are being surveyed and debated (i.e., Habermas, process of strategic management research and
1992). Another indicator is the general shift in results of research are closely connected
the economy from diminishing to increasing (Chakravarthy and DOZ, 1992; Lorange et al.,
returns on resources in many industries (Arthur, 1993). This is in accordance with the well-known
1990). Principle of Indeterminism discovered by Werner
The most important signals to the strategic Heisenberg in the mid 1 9 2 0 ~ .In ~ short, this
management field include: the advent of infor- principle acknowledges that observation influ-
mation technology and new organizational forms ences what is seen and vice versa. Thus, any
(Sproull and Kiesler, 1991); corporate programs attempt to truly rethink the boundaries of
to fusion various types of technologies (Kodoma, strategic management must acknowledge this
1992);the emergence of new forms of manufactur- effect and, therefore, simultaneously cause a
ing (Drucker, 1990); the implementation of rethinking of the research approach. When we
competitive alliances, i.e., collaboration among wonder about something, we not only need to
competitors (Hamel, DOZ, and Prahalad, 1988); look for answers to questions never asked before,
the coming of new activity-based accounting we need to become inventive about ways of
systems (Johnson and Kaplan, 1987; Johnson, finding out things. Otherwise, methodology may
1992); and the availability of life-long learning become a severe constraint on the degree of
programs as a partnership between the individual novelty in the knowledge produced.
and the educational institution (Lorange, 1992; Because the process of finding new meaning
Ghoshal, Arnzen, and Brownfield, 1992). Conse- in the ideas of strategic management essentially
quences of these signals have been addressed by is a creative process, the theories and perspectives
various authors (Davidow and Malone, 1992; developed still require sufficient grounding in
Badaracco, 1991; Fombrun, 1992; Hamel and practical relevance. Our research approach tries
Prahalad, 1989,1993; Prahalad and Hamel, 1990;
Hamel, 1991).
The uncertainty principle, or principle of indeterminism,
An important conclusion to be drawn from was discovered (or rather, it was reaffirmed by the Greek
this work is that we need to develop a better Zen0 of Elea’s objection that an object cannot occupy a
understanding of the organization as a knowledge given place and be moving at the same time) when Heisenberg
was struggling to rethink the boundaries of the quantum
system. We suggest that the organization can be world that would accord with the new quantum mechanics
seen as a stream of knowledge. This means that (for a fuller discussion, see Heisenberg, 1971).
 Corporate Epistemology 55
to accomplish precisely this balance: while making constructed by the practitioner and academics.
use of both a new theoretical perspective and a The current article is a product of a period where
new methodology, we have simultaneously put every word is tested for the meaning it conveys
creative and practical relevance weight into this to both the practitioners and the researchers.
paper. It follows that our research is different from
Our approach synthesizes grounded and grand both traditional positivistic and anti-positivistic
theories, as opposed to the traditional approach research approaches. It is different from case-
of discussing implications of the one for the study research (Yin, 1984) in which the prac-
other (see Glaser and Strauss, 1967). The titioner is solely conveying a set of data for
approach assumes a joint effort by proponents research, and it is different from grounded theory
of grounded and grand theories, and the synthesis building (Glaser and Strauss, 1967; Eisenhardt,
is accomplished through numerous iterations and 1989) due to its focus on participation in theory
dialogues. construction.
The process of matching leads to theory This research is also distinct from other
construction. This is different from matching participant-oriented research processes, such as
as model validation, as discussed by logical participatory action research (e.g., Reason, 1988;
empiricists (e.g., Wold, 1969; Chronbach and Whyte, 1991) or action science (e.g., Argyris,
Meehl, 1955). Rather, ‘matching’ means unifying Putnam, and Smith, 1985).4 In such processes
languages, theoretical concepts and their inter- the researcher assumes that there are some
relationships. The matching that resulted in this ‘. . .questions that concern the researcher more
article invo.lved a two-step process: (1) theoretical than the participant. These are questions relating
discourse, and (2) inscription of theory. to the theorizing and knowledge accumulation
Unification of grand and grounded theory process itself. (Karlsen, 1991: 149; see also Elden
ideally happens through a phase of ‘theoretical and Levin, 1991). When epistemology is the
discourse’ in which conflicting theoretical claims, focus of study and the interest of the practitioner,
including concepts and their interrelationships, this assumption (as we discovered) does not
are put forth (see Habermas, 1984: 9). Through hold. Like any theorist, the practitioner is eager
frequent dialogue between the participants (two to see if ‘local theory’ also holds outside his/
strategy professors and one senior executive of her immediate action context, i.e., in another
a U.S.-based corporation), we unified our theor- organization. The heart of the difference between
ies on knowledge and knowledge development. our matching approach and other research
It follows from this that the academic language processes lies in the phase of inscription. This is
was not “victorious”, in the sense of being the the phase where the knowledge is made pre-
only legitimate language (Bordieu, 1991: 5 ) . In sentable such that it may inform other theory
this reasearch, the only legitimate language was building attempts and the knowledge develop-
the one that results from the matching process. ment of another context and another scale.
The matching process necessary to prepare this
paper took more than 1 year. During this time
period we were able to focus on concepts and TOWARDS AUTOPOIETIC
relationshipsof common interest, and discard those EPISTEMOLOGY
concepts and relationships that did not create a
basis for common understanding. The theoretical This section outlines our theoretical lens. Begin-
discourse included numerous activities, such as ning with the roots and applications of the
discussions of ideas and concepts in personal theory, we proceed to discuss the cognitivist and
meetings inside and outside each other’s organiza- strategic management notion of knowledge.
tions (both in Europe and in the U.S), and Further, we present a notion of knowledge
developing a joint frame of references by ordering from the new perspective, the notion of social
and reading each other’s reference literature. knowledge and, finally, the conditions for autopo-
The inscription process was the most critical iesis and knowledge development.
phase in the research project. Inscription implies
capturing and making an object (for further
study) out of the knowledge that has been co- See also the discussion in Chakravarty and Doz (1992).
56 G . von Krogh, J . Roos and K . Slocum

vation. The main requirement is that one must

Autopoiesis theory, its roots and applications
be able to give a precise description of the
Originally developed in the field of neurobiology component production processes (What is the
to characterize ‘livingsystems’, autopoiesis theory component and how do you guarantee its
suggests the composition and structure of individ- reproduction?) and systems (e.g., Varela, 1979).
ual cognitive systems (Maturana and Varela, Despite these restrictions and its relatively short
1980, 1987). Through its application in the social history, autopoiesis theory has had an impressive .
sciences, autopoiesis theory also emerges as a impact in many fields. For instance, in legal
new theory of knowledge of a social system (von theory and the sociology of law, the basic concept
Krogh and Vicari, 1993; Luhmann, 1986). of autopoiesis has created awareness as to the
Since its introduction, autopoiesis theory has legal system’s lack of renewal and resistance to
gradually evolved into a general theory of systems adapt to problems in the economy (Luhmann,
(Varela, 1979; Luhmann, 1986; van Twist and 1982, 1988; Teubner, 1988, 1991; Deggau, 1988).
Schaap, 1991). The theory’s main thesis is that In the debate on ecological consciousness and
the components of an autopoietic system are corporate responsivenessto environmental issues,
used to produce new components and their autopoiesis theory has helped increase the aware-
relations so as to recreate the ~ y s t e m .This
~ ness of communications problems (i.e., between
production of components does not depend environmentalists and corporate decision makers)
on an input-output relation with the system’s and advanced possible ways to overcome these
environment. Everything that the system needs problems (Luhmann, 1992). Autopoiesis theory
for its reproduction is already in the system. has also increased our understanding of how
Certain systems, then, and in particular cognitive computers and their function are related to the
systems, are created and recreated in a recursive, evolution of human language, thought, and action
self-generating, closed and autonomous manner, (Winograd and Flores, 1987). In the philosophy
hence the term auto-autopoiesis (grk. auto of science, autopoiesis theory has been used to
= self, poiein = to make, produce, remake point out the constitution of ‘everyday knowledge’
conceptualize).6 as opposed to ‘scientific knowledge’ (Maturana,
Autopoiesis theory is interesting to many 1991; Becker, 1991). In the field of management,
scientists in different fields. However, only in the concept of autopoiesis is used to address the
some instances can the label ‘autopoiesis’ be evolution of organizational knowledge (Vicari,
meaningfully applied to processes under obser- 1991; von Krogh and Vicari, 1993). It has also
formed a reference point for . understanding
(more in a metaphorical sense), evolutionary
The theory was originally meant for understanding cell organizational change (Morgan, 1986; Smith,
reproduction. In cell reproduction, not only are the cells 1982; Weathly, 1992). In this article, autopoiesis
reproducing themselves, but they are also reproducing
their own capacity to reproduce. Further, all metacellulars theory is used to articulate a corporate episte-
reproduce themselves through the couple cells that they are mology for strategic management.
composed of. Ongoing interaction between metacellulars,
i.e., individuals, has not been discussed by Maturana and
Varela in terms of autopoiesis. They speak only of structural The “cognitivist” notion of knowledge and the
coupling of autopoietic systems. However, other authors,
e.g., Luhmann, have argued that there are general principles heritage of strategic management research
of autopoietic organization in social systems. For a fuller
discussion of this see van Twist and Schaap (1992). One of the basic quests of epistemology concerns
(I Several important contributions towards a better understand- the notion of knowledge. Our primary concern
ing of organizational learning are based on cybernetics and is organizational knowledge, that is, knowledge
general systems theory (e.g., Argyris and Schon, 1978;
Senge, 1990). Hence, it may be of interest to some readers shared by organizational members. However,
to relate autopoiesis theory to more traditional general because autopoiesis theory had deep roots in
systems theories. This is done by Benseler (1980), Varela what is known as the cognitive sciences, and since
and Goguen (1979). and Varela (1979). Here, we should
mention that autopoietic systems can be described as we will later claim that knowledge development
complementary open (with respect to data) and closed (with at the individual scale resembles knowledge
respect to information). The systems theoretical notion of development at the social scale, this investigation
‘control’ exercised by the external environment, contingent
on the input-output of information, is supplemented with starts by briefly contrasting the notion of knowl-
the notion of ‘autonomy’ of the system. edge from autopoiesis theory with the more
 Corporate Epistemology 57

traditional notions of knowledge. This section both related to the social cognition of organiza-
also highlights how a traditional notion of tions and the cognition of individual managers.
knowledge has inspired researchers in the field At a very general level, several contributions
of strategic management to develop concepts on assume that managers and organizations create
the development of organizational knowledge. representations of their environment through
Since the mid-50s the ideas of Herbert Simon, processing information available to them in this
Noam Chomsky, Marvin Minsky, John McCarthy external environment (e.g., March and Simon,
and others have inspired growth of cognitive 1958; Argyris and Schon, 1978; Ginsberg, 1990;
science where human knowledge holds a particu- Gioia and Manz, 1985; Daft and Weick, 1984;
lar position (Gardner, 1985; Varela, 1992; see Weick, 1979; Huff, 1983; Hedberg, 1981.)8These
for example Newel1 and Simon, 1972; Minsky, representations are storable and retrievable in
1975; Simon, 1989). Varela calls the perspective, organization-wide knowledge structures that give
which builds on the ideas of these scientist, the organizational members a shared perception of
‘cognitivist’ perspective (Varela, 1992). Here, the world (Prahalad and Bettis, 1986; Lyles and
cognition is, to a large extent, seen as information Schwenk, 1992; Walsh and Ungson, 1991). The
processing and rule-based manipulation of sym- evolution of these structures is dependent on the
bols (like words). Knowledge is abstract, task- experiences gained.
specific and oriented towards problem solving. Further, the strategic management literature
At the very heart, the cognitivist perspective frequently assumes that organizations are prob-
assumes that the world is pre-given, and that the lem seekers and solvers, and that they develop
goal of any cognitive system is to create the most some task specific knowledge (Cyert and March,
accurate representation of this world (Varela, 1963; Lant and Mezias, 1990; Lant et al., 1992).
1992). Representations (e.g., of persons, things, In fact, Prahalad and Bettis (1986) take this
events) can be stored in and retrieved from argument a step forward by claiming that
schemata of the individual (Anderson, 1983; problem-solving behavior ingrained in knowledge
Bartlett, 1932; Schank and Abelson, 1977), and structures may be a potential source of a
if an event represented frequently reoccurs, their dominant general management logic; ‘the way in
representations are stored in scripts (Schank and which managers conceptualize the business and
Abelson, 1977). Nisbett and Ross (1980) coined make critical resource allocation decisions’
the term ‘knowledge structures’ to cover both (Prahalad and Bettis, 1986: 490). Such concep-
schemata and scripts. An important finding of tualizations and resource allocation decisions
Nisbett and Ross’s work is that the individual may endure in organizations and develop into
frequently develops rudimentary knowledge ‘cognitive rigidities’ due to conventional wisdom
structures by resolving ambiguity, making guesses and past experiences. The strategic management
about unobservable events, and inferring about literature here demonstrates that novel problems
causal relationships (Bruner, 1964; Nisbett and may be approached with old representations of
Ross, 1980; Tversky and Kahneman, 1973). the world. Several authors suggest that until a
Learning in the cognitivist perspective means major crisis occurs, or new top management
that the individual more accurately obtains replaces the old team, a change in rigid knowledge
representations of the world through assimilating structures can not be expected (e.g., Hedberg,
new experiences, for example, ’. . .a person Nystrom, and Starbuck, 1976; see also Grinyer,
actively constructs knowledge . . .by relating Mayes, and McKiernan, 1988). However, one of
incoming information to a previously acquired the central problems identified by Prahalad and
psychological frame of referen~e.’~ (Bruner and Bettis (1986) and Lyles and Schwenk (1992) is
Anglin, 1973: 397).
The cognitivist perspective has inspired substan-
tial theory development in strategic management, In these works knowledge has often been substituted by
other and less troublesome notions, like information, data,
resources, reputation, etc. This has also led to a view of
organizations as ‘nontrivial machines’ or computers (Kilduff,
This frame of reference, e.g., cognitive structure or theory, 1993), based on basic principles of information science (e.g.,
gives meaning and organization to the regularities in Resnikoff, 1989). The process by which firms develop,
experience, and allows the individual to ‘go beyond the sustain, improve or utilize knowledge has not been subject
information given.’ to extensive study.
58 G. von Krogh, J . Roos and K. Slocum

that little is known about how knowledge observation’ of the manager. Where you stand
structures actually develop. As previously stated, or what you know determines what you see or
this is one of the central questions in any what you choose to be relevant. Thus, in
corporate epistemology. autopoiesis theory, ‘knowledge’ and ‘observation’
In summary, two issues come forth from the are closely related.g To be more precise, in
cognitivist notion of knowledge: autopoiesis theory, distinctions and norms are
two central categories (Luhmann, 1986, 1988;
(a) The cognitivist perspective is concerned Varela, 1979). Knowledge is what makes man-
with how representations of the world are agers able to make distinctions in their obser-
created by information processing. In vations (for example between themselves and
turn, these representations are stored in others) and, based on their norms, determine
knowledge structures. what they see. The distinctions made reveal the
(b) Much literature in strategic management knowledge of the distinguisher. For example, in
builds on this information processing reading a management book, a manager first
assumption. isolates the book from the background, like a
table. Next, he has to make a distinction between
print and paper, or even between different fonts,
like Times or Geneva. In reading the book he can
The notion of knowledge in autopoiesis theory distinguish the book as a finance or strategy book
Evoking autopoiesis theory implies rethinking and, based on his set of norms, decide whether
some of the very basic assumptions behind the book is ‘good’ or ‘bad.’ He might even, if he
the previous contributions on how and why is skilled in the analysis of literature, isolate the
organizations know. We want to speculate on stream of thinking to which the book was intended.
what happens if one relaxes some of the This is all of his own doing. It would be difficult
assumptions of the cognitivist perspective and to predict exactly what kind of knowledge the
replaces them with the assumptions of autopoiesis manager would develop around this book, based
theory. In doing this we believe that autopoiesis on the mere act of giving him a book.
theory emerges as an important contribution to In turn, applying distinctions allows for new
these previous works in strategic management. knowledge to develop. By isolating a phenom-
Unlike the cognitivist perspective, autopoiesis enon, the manager can gain knowledge about it.
theory suggests that the world is not a pre-given The term used to describe this process of
state to be represented, but rather that cognition knowledge development is self-referentialify . Self-
is a creative act of bringing forth a world. referentiality means that new knowledge refers
Knowledge is a component of the autopoietic not only to past knowledge but also to potential
(self-protective) process, it is history-dependent , future knowledge (Luhmann, 1990; Varela, 1979).
context-sensitive and, rather than being oriented Managers use already established knowledge to
towards problem solutions, knowledge enables determine what they see, and they use what they
problem definition (Maturana and Varela, 1987; already know to choose what to look for in their
Varela 1992; Varela, Thompson, and Rosch environment. Knowledge is therefore highly
1992). Moreover, at the individual level, knowl- dynamic as managers make new observations,
edge is not abstract but rather is embodied in the talk, use their imaginations to envision possible
individual. How does this alter our conceptions of futures and courses of action. Increasing knowl-
managerial cognition? There are at least two edge enables managers to make finer and finer
important implications of claiming that cognition distinctions, and, in many ways, a kind of
is autopoietic: (1) knowledge is intimately knowledge structure evolves that resembles a tree.
connected to observation, and (2) the notion of The tree structure also implies that knowledge is
information becomes redefined. subject to scaling. Scaling means similarity
First, the proposition of ‘embodied knowledge’ transformations (when two objects have the same
suggests that all knowledge is dependent on the
manager, or everything known is known by This resembles very much the ideas of Piaget (1936) that
somebody. More importantly, however, knowl- the child constructs a reality through perceptual and
edge depends very much on the ‘point of conceptual experiences.
 Corporate Epistemology 59

shape regardless of their size), i.e., scales of manager who makes use of it to develop knowledge.
magnification. lo Here, scaling means moving Thus, he is closed with respect to knowledge (also
along the ‘distinction tree’ or knowledge struc- knowledge about the environment) but open with
ture, that is, adapting the distinction level to the respect to data from the outside. The only way to
circumstances. describe this process is to say that the manager is
To illustrate: By reading technical literature, a simultaneously open and closed.
R&D manager may use his knowledge to isolate Further, the manager is open with respect to
a technical innovation in telecommunications from data of different degrees of latency. A high
the bulk of nicely repackaged but old technical degree of latency means that the data are unclear
solutions. Moreover, he may distinguish a technical and ambiguous and may appear to be disturbances
innovation of high possible impact from an imitation or noise in the environment. Such data are
with no interest. Here the distinction might be not immediately presented to the manager as
‘innovation vs. imitation’ and the norm applied information. This data may indicate areas where
would be ‘a high degree of impact.’ In reading the manager has little or no knowledge. More
more about the innovation, the manager may manifest data, i.e., data with a low degree of
develop increasingly fine distinctions, for example latency, are lucid, clear, unambiguous, and
linear vs. nonlinear mathematics. In proceeding, meaningful to the system, and as such can easily
the resulting knowledge structure evolves into a be converted to information. The manager’s
distinction tree. In this example it is important to response to such data may be to make increasingly
note two features. First, the R&D manager, when fine distinctions.
talking to the CEO, may rescale using the In summary, in autopoietic theory:
distinction ‘innovation vs. imitation’ to arouse the
CEO’s interest. Second, the manager makes (a) The world is not pre-given to be represent-
distinctions even if the message he conveys does ed.
not make this explicit. Even if he only chooses to (b) Knowledge is connected to observation.
say: ‘Eureka, I have found a new innovation,’ (c) The notion of information is redefined.
rather than ‘I have found an innovation that is
distinct from imitation,’ his very use of the concept
of innovation implies something distinct from
The notion of organizational knowledge
imitation.
A second implication of saying that the cognition Previous contributions in organizational behavior
is autopoietic is that we need to distinguish between and strategic management literature have
data, information and knowledge. In autopoiesis attempted to bridge individual cognition with
theory, information is not a commodity or a social cognition of the organization (e.g., Pra-
substance, as is often assumed in the cognitivist halad and Bettis, 1986; Lyles and Schwenk, 1992;
perspective and the strategic management litera- Argyris and Schon, 1978; Weick and Bougnon,
ture. Rather, information is a process of interpret- 1986; Daft and Weick 1984; Ginsberg, 1990).
ation, or, to use the words of von Foerster (1984: Individuals have private knowledge that can be
193), ‘informationis theprocess by which knowledge a basis for organizational knowledge when
is acquired.’ Literally, information means ‘to put’ conveyed through speaking, gesturing, writing,
data ‘in form’ (lat. in = in, form = form, are = etc.’* Knowledge of the organization is shared
doing). Books, movies, lectures, papers, computer knowledge among organizational members. Organi-
programs, memos, etc. ,are data in the environment zational knowledge allows for shared distinction-
of the manager-not information. They are simply making in observations made by organizational
elements that may be vehicles for potential members of events, situations and objects that are
information. Information is dependent on the
l2 A considerable amount of this private knowledge may be
‘“The consequences of scaling in science were discussed difficult or impossible to convey linguistically or in writing.
already by Galileo (1638). See Bonner (1969) and Morrison This is what Polanyi (1958) called tacit knowledge. Neverthe-
and Momson (1982) for a discussion of scaling in nature less, tacit knowledge may find other ways of becoming social
and the universe respectively. like, for example, imitation of body language. For a fuller
I ’ In fact, here Derrida (1981) and Varela (1979) go so far treatment, see Spender (1993), Nonaka (1991). and Polanyi
as saying that no concept can exist without its opposite. (1958, 1%2).
60 G. von Krogh, J . Roos and K. Slocum

internal and external to the organization. These cess of creating information based on data requires
distinctions are created and maintained in conver- that organizational members not only read data,
sations between organizational members and hence but also discuss and file it for later use. A report
allow for new knowledge to develop in a self- written in one office by a certain team of specialists
referential manner. A prerequisite for organiza- may be sent by internal mail to another team of
tional knowledge to develop is the cardinal specialists out of the country. The team members
distinction between the organization and its environ- become informed through the data, and, by
ment, e.g., ‘What do we know about our environ- discussing their observations of the report, they
ment?’ Social norms are necessary to coordinate participate in developing organizational knowledge.
the opinions of organizational members as to what Some data have a manifest character and little
they observe. They also highlight conflict regarding conversation is needed for these data to become
observations and provide guidelines when organiza- a source of information and knowledge among
tional members need to negotiate on the content organizational members. Latent data, on the other
of observations (e.g., Daft and Weick, 1984). hand, may need extensive discussion before they
Like individual knowledge, organizational knowl- can be converted to information and thereby
edge is also scalable. In conversations, managers become a source of new knowledge. Organization’s
move up and down the distinction tree depending response to such data may be to adjust distinctions,
on the situation at hand. Perhaps this could be or alternatively make increasingly fine distinctions.
stated in the following way: an organization, when The whole process of organizational knowledge
developing knowledge for its own strategic decision development makes the organization simul-
making (and hence direction for action), is thinking taneously open, with respect to data, and closed,
at a scale that encompasses all other scales of with respect to knowledge.
knowledge that would be found in other parts of According to the traits of autopoiesis theory,
the corporation (the equivalent of flying at 30,000 history provides an important starting point for
feet). This level of thought, however, is useless knowledge development. For example, a company
unless it is logically linked to all the other lower in the industrial stapling business, which initially
scales of understanding (flying at 10,OOO feet; 4,000 grew from needs in the automotive industry in the
feet; 1,500 feet, etc.) so that, as decisions are 1940s, expanded the application of this knowledge
made, the corporate entity can be assured of the to medical wound closure technologies in the
reliability of the eventual implementation of 1980s. The knowledge needed for this new
decisions at lower scales of the organization, like technology stems from the profound knowledge of
marketing or manufacturing. This form of ‘high staple formation and the fastening of various
altitude’ thinking would indeed be very cumber- materials that had been developed since the 19&,
some if the corporate entity had to maintain for instance metallic, nonmetallic and natural
a contiguous structure of detailed information materials, and the appropriate tools for fastening
connecting all of its various activities throughout such materials. As seen from this example,
the corporation. Instead, the corporate entity needs organizational knowledge allowed for new types
only to deal with the understanding of the process of activities to develop in the ~rganization.’~
of subsequent distinction-making that may occur Distinctions like strategic vs. operational,
at each scale of knowledge development (seeing
the differences, not trying to manipulate the whole) l3 A final point; since knowledge of the social system is
and take these into account in its own high level shared knowledge, it is no longer entirely dependent on
knowledge development process. specific individuals (see also Argyris and Schon, 1978).
Autopoiesis theory also requires a rethinking of Individuals may leave the group (for example a physicist
may retire from his department and field) but the knowledge
what is seen as information available to the of the group does not (if there are knowledge connections,
organization. Normally memos, letters, reports, see below) vanish. The same is the case when parts of the
etc., produced by the organization are seen as social system spin off. For example, a medical venture
division of a pharmaceutical firm may sell off a high-tech
pieces of information (e.g. Sproull and Kiesler, organization that has taken years to establish. The knowledge
1991). A consequence of an autopoietic epistem- that resides in this organization, however, may not be lost
ology is that documentation, even if produced by but can form a basis for further corporate knowledge
development (given the knowledge connections). That is why
the organization, is data and these data fuel this first type of knowledge conversion process is so important
organizational knowledge development. The pro- to understand.
 Corporate Epistemology 61
production vs. marketing, and norms, like good It is shared among organizational mem-
vs. bad, new vs. old, are maintained in conver- bers.
sations among organizational members. But It is scalable and connected to the organiza-
because individuals differ in their knowledge tion’s history.
and observations, discussions frequently uncover It both demands and allows for languaging.
differences in distinctions, distinctions at diverse
scales, renamings, finer or broader distinctions,
etc. For example, the corporate strategist may
Conditions for autopoiesis
bristle when hearing the happy sales manager
talking about ‘the strategy of telephone sales’. If the manager’s cognition is the unit of analysis,
Here, we propose that a key concept to under- the conditions for autopoiesis are biologically
standing development of organizational knowl- given and relatively unproblematic; he must be
edge is languaging. For the social system it is by alive and the brain, as well as the senses, must
‘languaging’ that knowledge brings forth a world be functioning. However, when the organization
(Maturana and Varela, 1987; Becker, 1991). is the unit of analysis, organizational knowledge is
It is perhaps overly restrictive to conceptualize highly dynamic, ‘fragile,’and developed through a
organizational language as a static body of syntax, self-referential, simultaneously open and closed
signs and codes subject to consistent use over process. Given this, how do organizations ensure
time and place. As an alternative, languaging that the autopoietic process continues and, hence,
refers to the process in which language is not that knowledge develops? One answer is that
only maintained but is constantly being developed unless there are knowledge connections available,
based on previous language. Managers frequently knowledge at one point in time does not connect
discard distinctions, introduce new distinctions, with new. knowledge at a later point in time.
use old distinctions on new situations, put words Should this occur the autopoietic process, and
in new contexts, use distinctions in a metaphorical therefore knowledge development, stops.14
sense, etc. In the words of Maturana and Varela Knowledge connection is defined as the potential
(1987: 212) ‘because (he has) language there is for individuals to convey messages about their
no limit to what (the manager) can describe, observations.
imagine, and relate.’ In this process of languaging There are two conditions that need to be
some distinctions are maintained, discussed and satisfied for knowledge to connect in the organiza-
built on by others and thereby form the basis tion over time: (1) the availability of relationships,
for developing organizational knowledge and and (2) a self-description.l5 First, the organization
finer distinctions, while other distinctions, consists of a set of relationships that enable
because they are not understood, are forgotten, immediate knowledge connections. Organiza-
are disagreed on by others, or are discarded. tional members develop informal relationships
Maintaining distinctions can, however, become over time that can ensure that the distinctions
crucially important to the inventor of the distinc- they convey are further built on and developed
tion because it frequently relates to the role and by others. Organizational members are also
identity that person has within the organization. related to one another through organizational
Returning to the example of the R&D manager: structures and reporting relationships. These
when he has singled out an invention from the facilitate communication among individuals and
technical magazine, this is done in his position may therefore allow for organizational knowledge
as an R&D manager. It would represent a serious
threat to his position should an engineer from
l4 According to Maturana and Varela (1987) autopoiesis
production point out to him that the technical theory is a theory of life. When the individual organism no
solution was based on an imitation of old longer continues the knowledge creation process, its life
technology in an adjacent form. Further knowl- comes to an end. Thus, for the organism the most important
process to maintain over time is the autopoietic process
edge development and distinction-making based whereby knowledge is created and recreated.
on the isolation of the product as a technical This is different from the notion of ‘organizational memory’
innovation might also be constrained. for knowledge (see Walsh and Ungson, 1991; Weick. 1991),
including individual’s memory, documents, data and systems.
In summary, the notion of organizational From our autopoietic perspective, such an organizational
knowledge has the following properties; memory only stores and retrieves data, not knowledge.
 62 G. von Krogh, J . Roos and K . Slocum
to develop. For example, individuals working on (a) Relationships.
developing a new design on a graphical product (b) Self-description.
may be structurally related with those studying
new applications of multimedia. Product devel- The dimensions of our autopoietic corporate
opers may also be structurally related to the epistemology are summarized in Table 1.
marketing department. Moreover, the organiza-
tional structures function as a set of ‘expectations’
(Deggau, 1988). Individuals, groups, or subunits DISCUSSION
meet with structurally defined expectations to
develop the organizational knowledge, based on In a conventional theory-building paper impli-
what they know or what others in the organization cations of theoretical propositions would be
tell them. presented. As the reader has noted, we have
Second, knowledge connections require an chosen a presentation form which is more like
adequate self-description of the organization an essay. In the essay form the point is to give
(Luhmann, 1990). A self-description results from extensive discussions of surfacing issues. The
an ‘observation’ by the organization of itself. In corporate epistemology in this article does not
fact, a ‘self-description formulates the identity of constrain but merely broadens the tasks and
the (organization)’ (Luhmann, 1990: 253). This increases the importance of strategic manage-
provides criteria for selecting what passes for ment. The following is a list of questions that
’knowledge,’ and that, as such, should be further might fuel further discourse:
connected, as opposed to ‘noise’ that should not
be connected. In many organizations, descriptions
How does the corporate epistemology shed light
of the organization’s identity include business
on how knowledge develops in organizations?
ideas, mission statements, strategy documents,
vision statements, management principles, guid- As indicated above, a central issue raised by
ing values, etc. previous works in strategic management is the
For knowledge to connect with other knowl- problem-solving behavior of firms, that is, how
edge it requires, in general, that it passes for knowledge structures evolve and change. Two
organizational knowledge as defined by the self- issues seem to emerge from this work. First,
descriptions. Findings that are not related to there seems to be some consensus as to the role
current businesses may be discarded, customers of experiences in forming knowledge structures
that ask for other types of services than those (e.g., Prahalad and Bettis, 1986; Lyles and
provided by the organization may be ignored, Schwenk, 1992). Knowledge structures evolve
scientific findings that are not relevant for the and change as organizational members reach
line of business of the organization are not agreement on interpretations of their individual
interesting, and so on. Self-descriptions, there- and shared common experiences (Daft and
fore, have a legitimizing function in defining Weick, 1984; Hedberg, 1981; Weick, 1979).
adequate behavior. In the words of Maturana Second, as Lyles and Schwenk (1992) point out,
and Varela (1987: 174) ‘We admit knowledge only in rare instances will organizational members
whenever we observe an adequate behavior in a question core elements as opposed to the
given context, i.e., in a . . .domain we define by peripheral elements of the knowledge structure,
a question.’ The result of these self-descriptions like clients served, the scope of businesses, and
of the organization is a focused development on management philosophies. Eventually, however,
knowledge through knowledge connections. This massive criticism may also lead to changes in
, is in contrast to totally random knowledge the core of the knowledge structures of the
development constantly increasing in variety. organization.
Self-description prevents the organization from In our attempt to develop the corporate
drowning in ‘knowledge complexity.’ epistemology, we have put particular emphasis on
In summary, we have discussed the conditions languaging. As organizational members observe
for autopoiesis in terms of knowledge connec- events and situations, and as they engage in
tions. There are two prerequisites for knowledge languaging, that is, apply and invent distinctions,
connections, namely: phrases, sentences, etc. , they participate in
 Corporate Epistemology 63
Table 1. Dimensions of a corporate epistemology

The notion of organizational

The notion of knowledge knowledge Conditions of autopoiesis

Distinctions and norms Distinction-making Knowledge connection

Data -+information -+ Environment-organization - relationships
knowledge - self-description
Latent and manifest data Environment-organization
Scaling Scaling

developing organizational knowledge. Agreement even in the face of nonunderstanding will allow
and disagreement are apparent at many levels of for the development of new distinctions and
the organization at all times, and as organizational organizational knowledge over time. Self-descrip-
members strive towards agreement (or settle tion allows for ‘experimentation in the realm of
for disagreement) they continue to develop the unknown’ and organizational structure and
organizational knowledge, enabling finer and informal relationships bring people together.
finer distinctions. What Lyles and Schwenk Therefore, organizational members may recog-
(1992) call a core-set of knowledge structure, we nize new knowledge (distinguish it from old
would refer to as a set of fundamental distinctions. knowledge) even when they do not completely
Sometimes organizational members invent new understand it.
fundamental distinctions pertaining to organiza- ’’ This article has not sufficiently developed
tion vs. environment, strategic vs. operational, possible patterns of languaging as they pertain
innovation vs. imitation, etc. In other words, to the development of organizational knowledge.
they scale towards the ‘root’ of the distinction In the strategic management literature the role
tree. This scaling has seriously challenged existing of language and conversations has attracted
organizational knowledge and current distinction- relatively little interest, with some insightful
making. The reaction of other organizational exceptions (Fiol, 1991; Westley, 1990; see also
members is often apparent; they do not recognize Fletcher and Huff, 1990; Nonaka, 1988). How-
these new distinctions as advancing the knowledge ever, as Westley (1990) points out, conversations
of the organization. Why? Perhaps not so much are interesting to study for three reasons: They
because they disagree with the new distinctions, are discrete, ritualistic and observable events that
but rather because they do not understand the help to ground empirically strategic management
distinctions, i.e., they lack knowledge. research; they contain elements of authority; and
In the strategic management literature agree- they uncover cultural vocabularies. In addition
ment on knowledge structures typically assumes to this, our perspective stresses thai languaging is
a shared understanding. However, because people the nexus of organizational knowledge develop-
might say yes to something they do not understand ment.
and no to something they understand, this Future research in strategic management would
assumption does not always hold (Luhmann, benefit from studying languaging, the period
1990). Organizational members frequently intro- of time in which organizational knowledge is
duce new ideas, new concepts, and new experi- sustained, and the conditions under which the
ences. The key question is to what extent new autopoietic process is interrupted (knowledge
distinctions are ‘languaged’ in the organization, ceases to be a basis for further organizational
and how long they are sustained. New distinctions knowledge). The studies should attempt to map
often vanish simply because they are not under- distinction trees employed, as well as registering
stood or further debated. agreement, disagreement, understanding and
A key concept to understanding the sus- nonunderstanding. Such studies would uncover
tainability of organizational knowledge, there- both incremental drifts and drastic shifts in
fore, is knowledge connectivity. Consequently, organizational knowledge. Given the interest in
organizations that ensure knowledge connectivity knowledge structures and ‘dominant logic,’ we
64 G . von Krogh, J . Roos and K . Slocum

suggest that languaging research should isolate in some of these options. By linking this
the conceptualizations of businesses and critical fundamental distinction to the realm of strategic
resource allocation decisions. In doing this we management, an interesting picture emerges. The
would be able to find out how and why dominant list in Table 2 is not complete by any means,
logics may become an impediment. Of special but it illustrates the differences in strategic
interest is languaging pertaining to latent data. It management tasks.
would be interesting to understand the languaging The objectives of strategic management are
involved in making data manifest, the difference three-fold: It must ensure the survival, advance-
in languaging on latent and manifest data, and ment and balance between these two activities.
its possible impact on the dominant logic. The overall goal of survival activities, e.g.,
product-market positioning, planning, organizing,
routinization, and controlling, is to manage the
How does our corporate epistemology help us
input-output relationship between the firm and
rethink strategic management?
its environment. Survival activities pertain to
When it comes to rethinking strategic manage- traditional theories of strategic management
ment the following feature surfaces from the and organizing. There are many guidelines for
corporate epistemology presented in this article: methodologies to research them, and outlets to
A fundamental distinction between advancement publish the findings. The key is to manage the
and survival activities. firm’s survival in a way that makes it able to
There is a painting by Raphael in the Gallery engage in advancement activities.
of Vatican City that reveals to us fundamental Advancement activities, e.g., development
distinctions in Western civilization. The picture of distinctions and norms, scaling knowledge,
is dominated by two figures, Plato pointing to ensuring knowledge connectivity, languaging, and
the sky, giving us a symbol of speculative self-referencing, represents new activities for
thought,16 and Aristotle pointing to the ground, many organizations. There are few theories
providing a symbol of knowledge on matter, explaining them, limited knowledge on how to
substance, and nature. The mastery of nature is research them, and few outlets in which t o
necessary for the survival of the human race. publish. Consequently, advancement activities
Speculation, however, is needed for the advance- represent challenging research topics and method-
ment of human existence. ologies. Balance between the two activities
The corporate epistemology outlined in this pertain to both attention and time devoted.
article implies that the organization is simul- Why is this distinction important? If there is
taneously subject to survival and advancement a tremendous transformation in society, there
needs. Although not a novel distinction in
principle, l7 separating advancement and survival
Advancement and survival activities
activities may be critical for organizations, groups Table 2.
and individuals alike. Whereas advancement Advancement activities Survival activities
emerges from knowledge development, which in
turn can be seen as developing options for the Developing distinctions and Product-market
organization,18 survival activities mean engaging norms positioning
Scaling knowledge Planning
Ensuring knowledge Organizing
l6 In many of his dialogues, Plato often posed puzzles without
connectivity Routinization
solving them and the reader is left aware of his ignorance Self-referencing Controlling
of important issues.
l7 We are referring to the distinction between the natural
Languaging
world (body) and the spiritual world (soul).
I R From our perspective ‘new firm knowledge’ is knowledge
that extends beyond the ‘limits’ of existing firm knowledge.
The limits of knowledge are determined by the autopoietic epistemology of a (living) system. A social system cannot
process in the development of the corporate knowledge base see that it is unable to see what it cannot see. Luhmann
(due to self-reference). The critical distinction is between (1990) however suggests that an external observer can see
what the firm knows it knows, vs. what the firm knows it precisely this blind-spot, and help the system to realize its
does not know, and what the firm dues nor know it knows weakness. Latent data is of great value here because they
vs. what it does nor know it does not know. See also von indicate areas where the firm has a lack of knowledge (also
Foerster’s (1982) identification of the ‘blind-spots’ in the of its own knowledge).
 Corporate Epistemology 65

are also transformations on the corporate level. ‘scientific’ description of organizational knowl-
There are many conflicting pieces of knowledge edge implies the use of scientific methods. The
coexisting in organizations and people are being corporate epistemology highlights a fundamental
exposed to data originating in many epochs. dimension in research methodology, namely
Fundamental transformation is manifested in observer-dependency. An observer-independent
contemporary languaging. New perspectives and perspective implies that there is a domain of
solutions to managerial problems will emerge, consensus shared by the observers, that is, the
which are concerned with subjective knowledge individual observer’s role is not taken into
as opposed to objective truth; with unrepeatable account. The traditional description of scientific
and unique experiences rather than reoccurring, method (e.g., observation, model generation,
general lessons; with intertextual relations instead prediction, observation of results) presupposes an
of temporal causality; with richly personal com- observer-independent perspective (Hejl, 1980).
plexity over inert impersonal simplification. Dis- From our perspective, a scientific description of
tinguishing between advancement and survival organizational knowledge must always take the
activities allows researchers to develop ways for role of observer into account.I9 Such observer
the advancement of organizational knowledge, dependency only means that the reality of two
inspired by the languaging mentioned above. strategy professors and one senior executive is
A number of questions arise which need to be described in this article. Still, our construction
addressed by future research in the strategic of reality is not meant to be pushed through as
management field. The recent discussion about the ‘generally valid one’ in the field.
the ‘resource-based’ perspective of the firm Another important dimension in research
(e.g., Barney, 1991; Dierickx and Cool, 1989; methodology concerns the distance of observation.
Wernerfelt, 1984; Mahoney and Pandian, 1992) There are two principal ways of studying organiza-
poses the question of why firms differ (e.g., tion knowledge; systems-distant and system-
Nelson, 1991; Carroll, 1993; Schendel, 1991). close. As system-distant observers, researchers
Lyles and Schwenk (1992) suggest that firms differ draw distinctions that isolate streams of knowl-
due to their individually developed knowledge edge of the organization. Very often there is an
structures. The corporate epistemology goes implicit criteria for what passes for knowledge,
beyond this and suggests that firms can be less often related to a particular kind of behavior
different in terms of their survival activities and (Maturana and Varela, 1987). What is needed
more different in terms of their advancement for a study of organization knowledge to be
activities. valid is not only a definition of the generic
Future research should attempt to establish not characteristics of knowledge, e.g., naming a
only why, but where and how much firms differ knowledge base ‘core competence,’ but also a
with respect to advancement and survival activities. clarification of what kind of behavior is associated
What are the management characteristics in the with knowledge of the organization, e.g., that
survival and advancement realms? What is the core-competence is associated with consistent
balance between advancement and survival activi- investment across businesses (see Hamel and
ties? How are advancement activities measured? Prahalad, 1989). The main reason for this is that
What management structures are found in the in very few cases can you directly access the
advancement vs. the survival mode? What do knowledge of individuals or teams by asking
conversations in successful organizations center questions (Prahalad and Bettis 1986; Argyris and
around-survival or advancement activities? What Schon, 1978). The researcher has to rely on
is the linkage between “core competence” and observable behavior. For example, when do
“advancement activities”? researchers admit knowledge in a study of
corporate acquisitions and performance? Does
How does corporate epistemology help us to
l 9 An extreme position on this matter was taken by Hejl
rethink research methodologies in the strategic
(1980: 160), who stated that: Any description (of society)
management field? which does nor allow for the precise specification of the role
of rhe observer is not a scientific one. . .(a1)though it may be
In fact, corporate epistemology forces us to accepted by a great number of participants in the social
rethink the use of research methodologies. A process who call themselves “scientists”.’
66 G . von Krogh, J . Roos and K . Slocum

one admit that the management team of the take an active part in the traditional peer review
acquiring organization is knowledgeable of the process, including bringing them into editorial
acquired organization when they choose to boards. Another is for practicing managers to
intervene in managerial practices of the acquired provide commentaries on papers published within
organization? Does one admit that the acquiring the realm of strategic management, putting each
management team has knowledge only when potential contribution in ‘their’ form.
they are successful? If yes, would this not ignore Two methodologies surface as particularly
the knowledge gained by the experiences with appropriate for languaging: ethnographies and
the new company, and, in cases of diversification, action research. First, by doing organizational
a new business or industry? Such questions can ethnographies, researchers enter the organiza-
best be answered by conveying the criteria of tion, learn the distinctions and norms pertaining
what passes for organizational knowledge. to the knowledge of the organization, study self-
The corporate epistemology also helps us to descriptions in the organization, and establish and
address the recent concern about relevance among enter relationships necessary for the continuous
strategy (process) researchers (e.g., Chakravarthy knowledge development of the organization. The
and Doz, 1992; Van de Ven, 1992). Let us criteria of a valid study is whether or not the
assume that researchers, not only practicing researcher, in his descriptions of the organization,
managers, are encapsulated in an autopoietic uses a language (distinctions) that is meaningful
epistemology, which we think they (we) are. As to organizational members (see Joergensen, 1989;
systems-distant observers, perhaps even with an Geertz, 1973).20
observer-independent perspective, will all of our Second, action research seems to be a promising
theorizing be self-referent, running the risk of research strategy to increase validity and rel-
becoming less and less relevant to practice? From evance. In action research, both the researcher
our perspective the risk is substantial. We suggest and the practicing managers develop their mutual
that the closure of this epistemology may be knowledge about each other’s knowledge. How-
overcome through a system-closed and obser- ever, we are not advocating action research in
vation-dependent methodology, with a bearing which the researcher has questions which are
on the process of languaging. less interesting for the manager. As stated before,
Languaging in strategic management research we envision a type of action research in which
means that researchers and practicing managers both the researcher and manager jointly formulate
cospeculate, costudy, and cowrite. In doing questions about ‘how we know what we know’
this, new phenomena pertaining to strategic as well as jointly try to answer these questions.
management can be effectively isolated and A typical action research project (Susman and
named. Previous distinctions in strategic manage- Evered, 1978), would mean spending considerable
ment, on various scales, can also be reviewed and time on defining a common problem and
new distinctions developed. Lack of imagination extracting the general findings from the research.
might be the only impediment to the development All of this should contribute to unifying
(of finer and more rudimentary distinctions) of (dominant) languages from academia and practice
the field. into a sole legitimate language for the realm of
This article resulted from a theoretical discourse strategic management; a language that carries
between researchers and a practitioner. Results data that is more manifest to both parties than
followed from a deeply immersed process of to either one of them separately. In conventional
matching the practitioners’ grounded theory of terms, it will increase both validity and relevance.
knowledge development and the researchers’ We believe that languaging, in particular, will
grand theory of autopoiesis. This article illustrates result in strong signals that might bring together
a principle of developing knowledge through current autopoietic processes in individuals,
merging the two by the process of matching. groups, organizations and societies. Hopefully,
In addition to this particular approach to this may yield not only new shared agreement
languaging, at least two more ways of languaging between managers and strategic management
can be identified in the realm of strategic
management. One is journals and other research *” For examples of organizational ethnographies, see Gioia
outlets in the field asking practicing managers to and Chittipeddi (1991), Rosen (1991). and Pettigrew (1979).
 Corporate Epistemology 67

-
,
/

/ 2,2$ment
Cognitivism
/ Individual
cognition
Strategic

Strategic
Cognition
Individual / management
cognition
Strategic
Autopoiesis social / management
cognition

Figure 1. The quest for a new strategy paradigm

researchers, but also increase insights from the Perhaps the essay even could be written as a
perspective of both parties. dialogue.22 After all, it is our obligation to
present data as manifest as possible. Except in
books and special issues, where can essays and
A FINAL SELF-REFERENCE dialogues in the strategic management field be
published? Perhaps the Strategic Management
In our perspective the only way to rethink a Journal can experiment further with this form of
certain domain is to (1) move towards the roots presentation? ’

of the distinction tree, (2) make new rudimentary Finally, as noted by Calas and Smircich (1988),
distinctions, and (3) make increasingly finer the credibility and ‘goodness’ of research, and
distinctions referring to the rudimentary ones. thereby knowledge development, depends on
This is what we have tried to do in this article. both the writer and the activities of the reader.
We have examined some of the fundamental Therefore, dear reader, consistent with the autopo-
assumptions of the strategic management field ietic perspective, we urge you to begin to read
by moving towards the domain of cognition. Our this article again!
new rudimentary distinction was cognitivism vs.
autopoiesis. From here we were able to make a
finer distinction between social and individual ACKNOWLEDGEMENTS
cognition and subsequently indicate new domains
of strategic management (increasingly finer The authors gratefully acknowledge helpful com-
distinctions). The quest for a new strategy ments from Gary Hamel, C. K. Prahalad, Sam
paradigm is illustrated in Figure 1. Hariharan, Janet Ruggles, and two anonymous
This leads us to a brief discussion of the reviewers. Peggy S. Brcbnn and John Harald
presentation form of strategic management Aadne provided editorial assistance. An earlier
research. We believe that the conventional article version of this paper was presented at the SMJ
form found in most journals in the field is most Special Issue Conference ‘The Search for New
appropriate for research on what we have labeled Strategy Paradigms,’ Ann Arbor, September
survival activities. Here, the crucial task of the 16-17, 1993. This study was supported by NFW
claimant of new insights is to gain the acceptance NORAS (grant no. 215 1309) and NFR (program
of the scholarly community. For research regard- no. 28047).
ing advancement activities, i.e., knowledge devel-
opment, perhaps the essay form is more well-
suited to both readers and writers (e.g., Nelson, subjects in an easy way. Very often an essay represents a
1991; Rumelt, Schendel, and Teece, 1991).21 writer’s personal experiences and perspectives, kept in a
vivid and direct style. For a fuller discussion of research as
literature, stories and fiction, see Latour and Wolgar (1979).
21 Invented by Michel de Montaigne in the 16th century, the 22 As a literary form dialogue is an organized exposition of

essay is a particular literary composition dealing with a few contrasting philosophical or intellectual attitudes.
68 G. von Krogh, J. Roos and K . Slocum

REFERENCES Virtual Corporation. Harper Business Books, New

York.
Anderson, J. R. (1983). The Architecture of Cognition. Deggau, H. G. (1988). ‘The communicative autonomy
Harvard University Press, Cambridge, MA. of the legal system’. In G. Teubner (ed.), Autopo-
Argyris, C. and D. Schon (1978). Organizational ietic Law: A New Approach to Law and Society.
Learning. Addison-Wesley, Reading, MA. Walter de Gruyter, Berlin, pp. 128-151.
Argyris, C., R. Putnam and D. McLain Smith (1985). Derrida, J. (1981). Dissemination. University of
Action Science. Jossey-Bass, San Fransisco, CA. Chicago Press, Chicago, IL.
Arthur, W. B. (February 1990). ‘Positive feedback in Dierickx, I. and K . Cool (1989). ‘Asset stock
the economy’, Scientific American, pp. 92-99. accumulation and sustainability of competitive
Badaracco, J. L. (1991). Knowledge Link: How advantage’, Management Science, 35, pp. 1514-1530.
Firms Compete through Strategic Alliances. Harvard Drucker, P. F. (May-June 1990). ‘The emerging
Business School Press, Boston, MA. theory of manufacturing’, Harvard Business Review,
Barney, J. (1991). ‘Firm resources and sustained pp. 94-102.
competitive advantage’, Journal of Management. Eisenhardt, K. M. (1989). ‘Building theories from
17, pp. 99-120. case study research’, Academy of Management
Bartlett, S. C. (1932). Remembering: A Study in Review, 14(4), pp. 532-550.
Experimental and Social Psychology. Cambridge Elden, M. and M. Levin (1991). ‘Cogenerative
University Press, Cambridge. learning: Bringing participation into action
Becker, A. (1991). ‘A short essay on languaging’. In research’. In W. F. Whyte, (ed.), Participatory
F. Steier (ed.), Research and Reflexivity. Sage, Action Research. Sage Publications, Newbury Park,
London, pp. 226-234. CA, pp. 127-142.
Bender, F. (1980). ‘On the history of system thinking Fiol, C. M. (1991). ‘Managing culture as a competitive
in sociology’. In F. Bender, P. M. Hejl and W. resource: An identity-based view of sustainable
K. Kock, (eds.), Autopoiesis, Communication and competitive advantage’, Journal of Management,
Society: The Theory of Autopoietic System in 17, pp. 191-211.
the Social Sciences. Campus Verlag, Frankfurt, Fletcher, K. E. and A. S. Huff (1990). ‘Argument
pp. 33-43. mapping’. I n A. S . Huff (ed.), Mapping Strategic
Bonner, J. T. (1969). The Scale of Nature. Pegasus, Thought. John Wiley & Sons, Chichester,
New York. pp. 355-375.
Bordieu, P. (1991). Language and Symbolic Power. Fombrun, C. J. (1992). Turning Points: Creating
Harvard University Press, Cambridge, MA. Strategic Change in Corporations. McGraw-Hill,
Bruner, J. and J. M. Anglin (1973). Beyond the New York.
Information Given. W. W. Norton & Company, Galilei, G. L. (1638). Discorsi e Dimostrazioni
New York. Matematiche, Intorno d Due Nuoue Scienze.
Bruner, J. S. (1964). ‘Going beyond the information Appresso gli Elsevirii, Leida, Holland.
given’. In H. E. Gruber, K. R. Hammond, and Gardner, H. (1985). The Mind’s New Science: A
R. Jesser (eds.), Contemporary Approaches to History of the Cognitive Revolution. Basic Books,
Cognition. Harvard University Press, Cambridge, New York.
MA, pp. 41-69. Geertz, C. (1973). The Interpretation of Cultures.
Calas, M. B. and L. Smircich (1988). ‘Reading Basic Books, New York.
leadership as a form of cultural analysis’. In J. G. Ghoshal, S . , B. Arnzen and S. Brownfield (1992).
Hunt, B. R. Baliga, H. P. Dachler, and C. A. ‘Learning alliance between business and business
Schriesheim (eds.), Emerging Leadership Vistas. schools: Executive education as a platform for
Lexington, Lexington, MA, pp. 201-228. partnership’, Working paper, INSEAD, Fontaine-
Carroll, G. R. (1993). ‘A sociological view on why bleau.
firms differ’, Strategic Management Journal, 14 (4), Ginsberg, A. (1990). ‘Connecting diversification to
pp. 237-249. performance: A socio-cognitive approach’, Acad-
Chakravarthy, B. S . and Y. Doz (1992). ‘Strategy emy of Management Review, 15, pp. 514-535.
process research: Focusing on corporate self- Gioia, D. A. and C. C . Manz (1985). ‘Linking cognition
renewal’, Strategic Management Journal, Summer and behavior: A script processing interpretation
Special Issue, 13, pp. 5-14. of vicarious learning’, Academy of Management
Chronbach, L. J. and P. E. Meehl (1955). ‘Construct Review, 10(3), pp. 527-539.
validity in phychological testing’, Psychological Gioia, D. A. and K. Chittipeddi (1991). ‘Sensemaking
Bulletin, 52, pp. 281-302. and sensegiving in strategic change initiation’,
Cyert, R. M. and J. G. March (1963). A Behavioral Strategic Management Journal, 12(5), pp. 433448.
Theory of the Firm, Reprint 1992. Blackwell, Glaser, B. G. and A. L. Strauss (1967). The Discovery
London. of Grounded Theory. Aldine, Chicago, IL.
Daft, R. L. and K. E. Weick (1984). ‘Toward a model Goguen, J. A. and F. J. Varela (1979). ‘Systems and
of organizations as interpretation systems’, Academy distinctions: Duality and complementarity’, Inter-
of Management Review, 9(2), pp. 284-295. national Journal of General System, 5 , pp. 3143.
Davidow, W. H. and M. S . Malone (1992). The Grinyer, P., D. G. Mayes and P. McKiernan (1988).
Sharpbenders. Blackwell. Oxford.
 Corporate Epistemology 69
Habermas, J. (1984). The Theory of Communicative Lant, T. K., F. J. Milliken and B. Batra (1992). ‘The
Action. Beacon Press, Boston, MA. role of managerial learning and interpretation in
Habermas, J. (1992). The Philosophical Discourse of strategic persistence and reorientation: An empirical
Modernity. Polity Press, Cambridge. exploration’, Strategic Management Journal, 13(8),
Hage, J. and C. H. Powers (1992). Post-industrial pp. 585-608.
Lives: Roles and Relationships in the 21st Century. Latour, B. and S . Wolgar (1979). Laboratory Life:
Sage, Newbury Park, CA. The Social Construction of Scient$c Facts. Sage,
Hamel G. and C. K. Prahalad (May-June 1989). Beverly Hills, CA.
‘Strategic intent’, Harvard Business Review, pp. Lawson, H. and H. L. Appignanesi (1989). Dismantling
63-75. Truth: Reality in the Post-modern World. Weiden-
Hamel, G. (1991). ‘Competition for competence and feld and Nicolson, London.
interpartner learning within international alliances’, Lorange, P. (1992). Note regarding the strategy of
Strategic Management Journal, Summer Special Norwegian School of Management, Unpublished.
Issue, 12, pp. 83-103. Lorange, P., B. Chakravarthy, J. Roos and A. Van
Hamel, G. and C. K. Prahalad (March-April 1993). de Ven (1993). Implementing Strategic Processes:
‘Strategy as stretch and leverage’, Harvard Business Change, Learning and Co-operation. Blackwell,
Review, pp. 75-84. London.
Hamel, G., Y. L. Doz and C. K. Prahalad Luhmann, N. (1992). Ecological Communication.
(January-February 1988). ‘Collaborate with your Polity Press, Cambridge.
competitors-and win’, Harvard Business Review, Luhmann, N. (1982). The Differentiation of Society.
pp. 133-139. Columbia University Press, New York.
Hedberg, B. (1981). ‘How organizations learn and Luhmann, N. (1986). ‘The autopoiesis of social
unlearn’. In P. C. Nystrom and W. Starbuck (eds.), systems’. In F. Geyer and J. van der Zouwen
Handbook of Organizational Design, Vol. I. Oxford (eds.), Sociocybernetic Paradoxes. Sage, Beverly
University Press, New York, pp. 3-27. Hills, CA, pp. 172-192.
Hedberg, B., W. Starbuck and P. Nystrom (1976). Luhmann, N. (1988). ‘The unity of the legal system’.
‘Camping on the seesaws: Prescriptions for designing In G. Teubner (ed.), Autpoietic Law: A New
self-designing organizations’, Administrative Science Approach to Law and Society. Walter de Gruyter,
Quarterly, 21, pp. 41-65. Berlin, pp. 12-35.
Heisenberg, W. (1971). Physics and Beyond. Harper Luhmann, N. (1990). Essays on Self-reference. Colum-
& Row, New York. bia University Press, New York.
Hejl, P. M. (1980). ‘The problem of a scientific Lyies, M. A. and C. R. Schwenk (March 1992).
description of society’. In F. Benseler, P. M. Hejl ‘Top management, strategy and organizational
and W. K. Kock (eds.), Autopoiesis, Communi- knowledge structures’, Journal of Management
cation, and Society. Campus Verlag, Frankfurt, Studies, 29, pp. 155-174.
pp. 147-162. Lyotard, J. F. (1984). The Postmodern Condition: A
Huff, A. S. (1983). ‘Industry influence on strategy Report on Knowledge. University of Minnesota
reformulation’, Strategic Management Journal, 3(2), Press, Minneapolis, MN.
pp. 119-131. Mahoney, J. T. and J. R. Pandian (1992). ‘The
Johnson, H. T. (1992). Relevance Regained: From resource-based view within the conversation of
Top-down Control to Bottom-up Empowerment. strategic management’, Strategic Management Jour-
Free Press, New York. nal, 13(5), pp. 363-380.
Johnson, H. T. and R. S . Kaplan (1987). Relevance March, J. G. and H. A. Simon (1958). Organizations.
Lost: The Rise and Fall of Management Accounting. Wiley, New York.
Harvard Business School Press, Boston, MA. Maturana, H. (1991). ‘Science and daily life: The
Jorgensen, D. L. (1989). Participant Observation: A ontology of scientific explanation’. In F. Steier
Methodologyfor Human Studies. Sage Publications, (ed.), Research and Reflexivity. Sage, London,
Newbury Park, CA. pp. 30-52.
Karlsen, I. J. (1991). ‘Action research as method: Maturana, H. and F . Varela (1980). Autopoiesis and
Reflections from a program for developing methods Cognition: The Realization of the Living. Reidl,
and competence’. In W. F. Whyte (ed.), Partici- London.
patory Action Research. Sage Publications, Newbury Maturana, H. and F. Varela (1987). The Tree of
Park, CA, pp. 143-158. Knowledge (in Danish). Ask, Copenhagen.
Kilduff, M. (January 1993). ‘Deconstructing organiza- Minsky, M. A. (1975). ‘A framework for representing
tions’, Academy of Management Review, pp. 13-31. knowledge’. In P. H. Winston (ed.), The Psychology
Kodoma, F. (1991). Analyzing Japanese High Technol- of Computer Vision. McGraw-Hill, New York,
ogies: The Techno-paradigm Shift. Pinter Press, pp. 211-277.
London. Montague, W. P. (1962). The Ways of Knowing or
Lant, T. K. and S. J. Mezias (1990). ‘Managing the Methods of Philosophy, 6th impression. George
discontinuous change: A simulation study of organi- Allen & Unwin Ltd., London.
zational learning and entrepreneurship’, Strategic Morgan, G. (1986). Images of Organization. Sage,
Management Journal, Summer Special Issue, 11, Beverly Hills.
pp. 147-179. Morrison, P. and P. Morrison (1982). Powers of Ten:
70 G. von Krogh, J . Roos and K. Slocum
About the Relative Size of Things in the Universe. Spender, J. C. (April 15, 1993). Paper presented at
Scientific American Library, New York. the conference on ‘Distinctive competences and
Nelson, R. R. (1991). ‘Why do firms differ, and how tacit knowledge’, in Rome.
does it matter?’, Strategic Management Journal, Sproull, L. and S . Kiesler (1991). Connections: New
Winter Special Issue 12, pp. 61-74. Ways of Working in the Networked Organization.
Newell, A. and H. Simon (1972). Human Problem MIT Press, Cambridge, MA.
Solving. Prentice-Hall, Englewood Cliffs, NJ. Susman, G. I. and R. D. Evered (1978). ‘An
Nisbett, R. and L. Ross (1980). Human Inference: assessment of the scientific merits of action
Strategies and Shortcomings of Social Judgement. research’, Administrative Science Quarterly, 23,
Prentice-Hall, Englewood Cliffs, NJ. pp. 582-603.
Nonaka, I. (Spring 1988). ‘Creating organizational Teubner, G. (1988). ‘Evolution of autopoietic law’.
order out of chaos: Self-renewal in Japanese firms’, In G . Teubner (ed.), Autpoietic Law: A New
California Management Review, pp. 57-73. Approach to Law and Society. Walter de Gruyter,
Nonaka, I. (November-December 1991). ‘The knowl- Berlin, pp. 217-241.
edge-creating company’, Harvard Business Review, Teubner, G. (1991). ‘Autopoiesis and steering, how
pp. 96-104. politics profit from the normative surplus of capital’.
Pettigrew, A. (1979). ‘On studying organizational In R. J. In’t Veld, L. Schaap, C. J. A. M.
cultures’, Administrative Science Quarterly, 24(4), Termeer, and M. J. W. van Twist, Autopoiesis and
pp. 570-581. Configuration Theory: New Approaches to Societal
Piaget, J. (1936). La Naissance de lhellegence chez Steering. Kluwer Academic Publishers, Dordrecht,
llbzfants. Dechau et Nieste, Neuchatel et Paris, pp. 127-143.
France. Tversky, A. and D. Kahneman (1973). ‘Availability:
Polanyi, M. (1958). Personal Knowledge. University A heuristic for judging frequency and probability’,
of Chicago Press, Chicago, IL. Cognitive Psychology, 4, pp. 207-232.
Polanyi, M. (1962). Personal Knowledge: Towarh a Van de Ven, A. H. (1992). ‘Suggestions for studying
Post Critical Philosophy. Routledge, London. strategy process’, Strategic Management Journal,
Prahlad, C. K. and R. Bettis (1986). ‘The dominant Summer Special Issue, 13, pp, 169-188.
logic: A new linkage between diversity and perform- vanTwist, M. J. W. and L. Schaap (1991). ‘Introduction
ance’, Strategic Management Journal, 7(6), to autopoiesis theory and autopoietic steering’. In
pp. 485-501. R. J. In’t Veld, L. Schaap, C. J. A. M. Termeer, and
Prahlad, S. and G. Hamel (May-June 1990). ‘The M. J. W. van Twist, Autopoiesis and Configuration
core competence of the corporation’, Harvard Theory: New Approaches to Societal Steering.
Business Review, pp. 79-91. Kluwer Academic Publishers, Dordrecht,
Reason, P. (1988). Human Inquiry in Action. Sage pp. 3144.
Publications, London. Varela, F. J. (1979). Principals of Biological Autonomy.
Resnikoff, H. L. (1989). The Illusion of Reality. Elsevier North Holland, New York.
Springer-Verlag, New York. Varela, F. J. (1992). ‘Whence perceptual meaning? A
Rosen, M. (1991). ‘Coming to terms.with the field: cartography of current ideas’. In F. J. Varela
Understanding and doing organizational ethnogra- and J. P. Dupuy (ed.), Understanding Origins:
phy’, Journal of Management Studies, 28, pp. 1-24. Contemporary Views on the Origin of Life, Mind and
Rumelt, R. P., D. Schendel and D. J. Teece (1991). Society. Kluwer Academic Publishers, Dordrecht,
‘Strategic management and economics’, Strategic pp. 235-264.
Management Journal, Winter Special Issue, 2, Varela, F. J., E. Thompson and E. Rosch (1992).
pp. 5-29. The Embodied Mind. MIT Press, Cambridge, MA.
Sakaiya, T. (1992). The Knowledge-value Company. Vicari, S . (1991). The Living Firm (in Italian).
Bellew Publishing, London. EtasLibri, Milano.
Schank, R. and R. P . Albelson (1977). Scripts, Plans, von Foerster, H. (1982). Observing Systems. Intersys-
Goals, and Understanding: A n Inquiry into Human tems Publications, Seaside, CA.
Knowledge Structures. Lawrence Erlbaum, von Foerster, H. (1984). ‘Principles of self-organization
Hillsdale, NJ. in socio-managerial context’, In H. Ulrich and
Schendel, D. (1991). ‘Editor’s comments on the Winter G. J. B. Probst (eds.) Self-organization and
Special Issue’, Strategic Management Journal, Win- Management of Social Systems. Springer-Verlag,
ter Special Issue, 12, pp. 1-3. Berlin, pp. 2-24.
Senge, P. (1990). The Fifth Discipline. Doubleday von Krogh, G. and S. Vicari (1993). ‘An autopoiesis
Currency, New York. approach to experimental strategic learning’. In P.
Simon, H. A. (1989). Models of Thought, Vol. 2. Lorange, B. Chakravarthy, J. Roos and A. Van
Yale University Press, New Haven, CT. de Ven, Implementing Strategic Processes: Change,
Smith, K. K. (1982). ‘Philosophical problems in Learning and Co-operation. Blackwell, London,
thinking about organizational change’. In P. S . pp. 394-410.
Goodman (ed.), Change in Organizations: New Walsh, J. P. and G. R. Ungson (1991). ‘Organizational
Perspectives on Theory, Research, and Practice. memory’, Academy of Management Review, 16,
Jossey-Bass, San Francisco, CA, pp. 316-374. pp. 57-91.
 Corporate Epistemology 71
Weathly, M. J. (1992). Leadership and the New Science. Westley, F. R. (1990). ‘Middle managers and strategy:
Berrett-Koehler Publishers, San Francisco, CA. Microdynamics of inclusion’, Strategic Management
Weick, K. E. (1979). The Social Psychology of Journal, 11(5), pp. 337-351.
Organizing. Random House, New York. Whyte, W. F. (1991). Social Theory for Action: How
Weick, K. E. (February 1991). ‘The nontraditional Individuals and Organizations Learn to Change.
quality of organizational learning’, Organization Sage, Newbury Park, CA.
Science, pp. 116-124. Winograd, T. and F. Flores (1987). Understanding
Weick, K. E. and M. G . Bougnon (1986). ‘Organiza- Computers and Cognition. Ablex Publishing Corpor-
tions as cognitive maps’. In H. P. Sims, Jr. and ation, Norwood, NJ.
D. A. Gioia (eds.), The Thinking Organization. Wold, H. (1969). ‘Mergers of economics and philos-
Jossey-Bass, San Francisco, CA, pp. 102-135. ophy of science: A cruise in shallow waters and
Wernerfelt, B. (1984). ‘A resource-based view of deep seas’, Synthese, 20, pp. 427-482.
the firm’, Strategic Management Journal, 5(2), Yin, R. K. (1984). Case Study Research. Sage, Beverly
pp. 171-180. Hills, CA.