Content uploaded by David Stark
Author content
All content in this area was uploaded by David Stark on Mar 02, 2014
Content may be subject to copyright.
The Sense of Dissonance:
Accounts of Worth in Economic Life
David Stark
Princeton University Press
Princeton and Oxford
2
Copyright © 2009 by Princeton University Press
Requests for permission to reproduce material from this work should be sent to Permissions,
Princeton University Press
Published by Princeton University Press, 41 William Street, Princeton, New Jersey
08540
In the United Kingdom: Princeton University Press, 6 Oxford Street, Woodstock,
Oxfordshire OX20 1TW
All Rights Reserved
ISBN 978-0-691-13280-8
British Library Cataloging-in-Publication Data is available
This book has been composed in
Printed on acid-free paper. ∞
press.princeton.edu
Printed in the United States of America
1 3 5 7 9 10 8 6 4 2
10 9 8 7 6 5 4 3 2 1
3
for my parents
4
Non tener pure ad un loco la mente
Fix not thy mind
On one place only.
Dante Alighieri, The Divine Comedy: Purgatory, canto 10, ca. 1308-1321,
trans. Henry Cary
A good investigator doesn’t know what he’s looking for till he sees it.
Elmore Leonard, Mr. Paradise, 2004
Dissonance
(if you are interested)
leads to discovery
William Carlos Williams, Paterson IV, 1951
5
Contents______________________________________________
PREFACE 7
1. HETERARCHY: THE ORGANIZATION OF DISSONANCE 14
Searching Questions 14
For a Sociology of Worth 18
Entrepreneurship at the Overlap 24
Heterarchy 29
Worth in Contentious Situations 39
2. WORK, WORTH, AND JUSTICE IN A SOCIALIST FACTORY 42
The Partnership as Proof 43
Distributive Justice inside the Partnership 55
Maneuvering across Economies 66
Epilogue 74
3. CREATIVE FRICTION IN A NEW-MEDIA START-UP 80
An Ecology of Value 82
The Firm and the Project Form 89
Distributing Intelligence 93
Organizing Dissonance 98
Discursive Pragmatism and Bountiful Friction 102
Epilogue 104
4. THE COGNITIVE ECOLOGY OF AN ARBITRAGE TRADING ROOM 111
Studying Quantitative Finance 113
6
Arbitrage, or Quantitative Finance in the Search for Qualities 117
The Trading Room as a Space for Associations 121
The Trading Room as an Ecology 124
The Trading Room as a Laboratory 130
The Pursuit of New Properties 138
Epilogue 139
5. FROM FIELD RESEARCH TO THE FIELD OF RESEARCH 148
From Classification to Search 151
From Diversity of Organizations to the Organization of Diversity 158
From Unreflective Taken-for-Granteds to Reflective Cognition 165
From Shared Understandings to Coordination through Misunderstanding 171
From Single Ethnographies to the Broader Sites of Situations 175
REPRISE 182
ACKNOWLEDGMENTS 190
BIBLIOGRAPHY 193
7
Preface_______________________________________________
In retrospect, this book began with a dissertation that went unwritten. As a graduate
student at Harvard in the late 1970s working in the field of comparative sociology, I was
especially interested in understanding how industrial organization differed in capitalist
and socialist economies. While exploring how I might develop this interest into a
dissertation topic, I learned about the phenomenon of “peasant-workers” in socialist
Eastern Europe. I recognized that peasant-workers could be an analytically strategic
social group because their workdays spanned the world of socialist industry and the
world of privately held agricultural plots organized along entirely different principles. Of
course, the natural experiment would have been even better had the peasant-workers
moved on a daily basis from socialist to capitalist industrial forms. Still, the topic was a
good opportunity to explore what happens when people live in two social worlds
organized around very different modes of production.
With dissertation traveling fellowships, my wife Monique Girard, who was a graduate
student in anthropology at Harvard, and I left for research in Yugoslavia in 1979-80. We
had studied Serbo-Croatian together and, just married, we looked forward to sharing the
adventure of dissertation field research. A week after we arrived in Zagreb, Marshal Tito,
Communist leader of Yugoslavia, was hospitalized, and he lay dying throughout our
research visit. Although we were part of an exchange program that granted us permission
to conduct research, legal formalities were the least of our problems. With the political
situation completely uncertain, no academics, even those who advised us informally,
could offer official assistance. Moreover, with our phones tapped, our apartments
bugged, our landladies (three in nine months) harassed, and our friends reluctant to tell us
that they were being questioned, we could not jeopardize field research informants.
Yugoslavia, “socialism with a human face,” was a police state; and our dissertations were
the most minor victims of its eventual demise.
I returned to Harvard to consider my options. The independent trade union Solidarność
(Solidarity) had recently been recognized in Poland, and it made sense to convert my
knowledge of an Eastern European country and my background in a Slavic language to
study this new development. So I enrolled in an introductory Polish course taught by
Stanislav Barancsak, one of the founders of KOR (Committee to Defend the Workers)
and a personal friend of Lech Walesa. On the morning I was supposed to take the final
exam for the course, I was awakened by a Polish friend: “Martial law has just been
8
declared in Poland,” she said. “There’s no way you’ll be able to do dissertation research
there.” I did not take the exam.
With a supportive committee and a sympathetic department, I did complete a dissertation
on the organization of work under capitalism and socialism1 – but without carrying out
the ethnographic research that I had wanted to do on the collision of competing and
coexisting organizational principles. Several months before leaving Cambridge, I met the
eminent Hungarian economist János Kornai. Over many cups of coffee in a Mass Ave
café, we talked about several of his books, and I recounted the story of my dissertation
misadventures. “You’re a persistent young man,” János said upon hearing that I still
wanted to do fieldwork after my Yugoslav and Polish disappointments. “Come to
Hungary and we’ll do what we can to help you get access to firms.”
A year later, in the summer of 1983, I arrived in Budapest on the Orient Express from
Paris. Csaba Makó, a sociologist with whom I was well acquainted, met me at the Kelleti
train station. I can recall the moment as if it were yesterday. We were not even out of the
station when Csaba began talking excitedly about a new development in Hungarian labor
relations: through a measure initiated by the Politburo of the Hungarian Communist Party
a year earlier in 1982, workers had received the right to form “intraenterprise
partnerships.” In the partnerships, workers were running factory equipment on the "off-
hours" and on weekends, subcontracting to the parent enterprise and getting orders from
outside firms. Not entirely unlike the peasant-workers that I had wanted to study in
Yugoslavia, the partners were working in two forms of social organization. But the
partnership form was an even more extraordinary social laboratory for an organizational
sociologist: in the same factory, using the same technology, workers were moving on a
daily basis from bureaucratic to nonbureaucratic organizational forms as the selection of
supervisors, the organization of work, and methods of internal payment were left to the
discretion of the work partners. “From six to two we work for them,” Csaba told me was
the expression he had heard, and “from two to six we work for ourselves.” This was an
opportunity I could not fail to recognize. “I have to study this,” I told him as we got onto
the subway.
Csaba introduced me to his assistant, János Lukács, and we immediately set to work.
During that summer, we conducted dozens of interviews with workers and managers in a
number of factories to learn more about the work partnership “experiment,” always with
an eye to find one or more settings where we could do in-depth ethnographic work. We
found such an opportunity at “Minotaur”, a producer of tires and other rubber products
and one of the very largest state-owned enterprises in Hungary, where we were welcomed
by a work partnership in its machine-tool factory. The toolmakers had formed their
1 Based on library and archival research, my dissertation focused on a comparison of Taylorism and
Leninism – not in the conventional way that Lenin was fascinated by the prospects of introducing
Taylorism into state-owned firms, but on each as new class projects tied to “scientific” knowledge claims in
different domains. Whereas the scientific management of the firm claimed legitimacy on the basis of the
“laws” derived from “time and motion studies,” the attempt to manage an economy scientifically rested on
claims to knowledge of the “laws of motion of history.” The irony was that the attempt to scientifically
manage an economy through the budgetary instruments of central planning made it impossible to introduce
rationalized principles of scientific management within the firm.
9
partnership to gain recognition of their worth as highly skilled craftsmen, bringing them
into conflict with Minotaur management but also leading to new challenges of finding an
internal payment system to allocate their “entrepreneurial fee.”
Over the following years I returned frequently to Minotaur, supported by fellowships that
allowed me to take research leaves. In the fall of 1986 I traveled back and forth between
Budapest and Paris, where I was a visiting fellow at the Centre de Sociologie Politique et
Morale at the invitation of its founder, Luc Boltanski. I wrote the first draft of “Work,
Worth, and Justice” (chapter 2 of this book) in Paris while Luc and his collaborator,
Laurent Thévenot, were preparing the manuscript for their book De la Justification: Les
Economies de la Grandeur (discussed in chapter 1). Our conversations about their work
helped to clarify my ideas. Whereas the language of my graduate student days had been
one of “modes of production,” I now saw that economic sociology and organizational
theory could benefit from the vocabulary of “orders of worth” (my translation of the
French, les ordres de la grandeur). In place of the grand historic clash of modes of
production, I now heard another noisy clangor in the workplace: the clash of contending
principles of evaluation.
But while I was on a steep learning curve with my French colleagues, I was also
grappling with my field notes, for there I kept returning to something that I felt missing in
the drafts of De la Justification. Whereas Boltanski and Thévenot saw orders of worth as
conventions that made calculable action possible, the experiment in the machine-tool
factory presented considerably more ambiguity. In fact, at Minotaur, action was made
possible precisely because there was uncertainty about which order of worth was in
operation. Some actors, moreover, were attempting to benefit, not from asserting or
fixing their worth in one order, but by maintaining an ongoing ambiguity among the
coexisting principles. This ongoing rivalry produced an organizational reflexivity, grossly
distorted by the confines of the command economy and a closed political system but
incipient nonetheless, pointing to possibilities for the real entrepreneurial activity of
recombination.
After leaving Paris, I sent my “Work, Worth, and Justice” paper to Pierre Bourdieu, who
accepted it for publication in French in his journal Actes de la Recherche en Sciences
Sociales. At the encouragement of Bourdieu and Boltanski, I decided to write a book
based on my research at Minotaur and other Hungarian factories. A wonderful
opportunity to do so presented itself when I was invited to be a Visiting Fellow at Cornell
University’s Society for the Humanities starting in the fall of 1989. But some other
wonderful things also happened in the fall of 1989. There I was, in my office in Ithaca in
upstate New York, poring over my field notes for as long as I could before being pulled
away to follow fast-breaking political developments in Poland, Hungary, and East
Germany. The world I was writing about was undergoing a momentous change.
In early October my friend László Bruszt came through Ithaca brimming with news from
Eastern Europe. Throughout the summer of 1989, László had been a participant in the
“Roundtable” negotiations in which representatives of Hungary’s Communist
government and those of the political opposition hammered out an agreement on the
transition to free elections. As a delegate of the Hungarian League of Independent Trade
10
Unions (founded in his apartment in Budapest), he had been sent to consult with his
counterparts in Poland where similar Roundtable talks had taken place earlier in the year.
To say his story was exciting would be an understatement.
László proposed that I come to Hungary, that we interview all the key Communist and
opposition figures who had participated in the political negotiations, and that we do so
immediately, while their memories were fresh and before election outcomes selectively
modified their recollections. I seriously considered his proposal. On one hand, I was an
untenured assistant professor with a book to write and with the data to do so. But, on the
other, I was on leave from my teaching duties. Several days later the Berlin Wall was
toppled. I recognized that the opportunity to do real-time research on an epochal
transformation on a topic that my previous ten years had prepared me to understand
would not come more than once in a lifetime. The fascinating story of the Minotaur
toolmakers was now being played out on a much larger historical stage that I could study
firsthand. By December I was in Budapest.
In place of a book on workplace politics in the socialist period, my research with Bruszt
led to a quite different book, Postsocialist Pathways: Transforming Politics and Property
in East Central Europe, addressing the distinctive challenge of postsocialist politics:
Could the transformation of property rights and the expansion of citizenship rights be
achieved simultaneously?2 During this period I continued my research in economic
sociology, returning to several of the firms where I had studied the work partnerships
during the 1980s and augmenting the case studies by systematically collecting data on the
ownership structure of the two hundred largest Hungarian enterprises to chart network
ties among these firms. That research led me to question the notion of market transition
as a toggle switch from public to private property. I did find property transformation, but
I found that it took forms of “recombinant property” in which the boundaries of public
and private as well as the boundaries of firms were blurred in networks of intercorporate
ownership.3
Whether it occurs in politics or in the economy, I concluded that change, even
fundamental change, of the social world cannot be understood as the passage from one
order to another but should be seen as rearrangements in the patterns of how these orders
are interwoven. That is, instead of thinking about institutional change or organizational
innovation as replacement, I examined them as reconfigurations of institutional elements.
In short, I thought of organizational innovation as recombination, a theme that reappears
throughout the present book.
I completed my paper “Recombinant Property” as well as my book Postsocialist
Pathways while a Visiting Fellow at the Center for Advanced Study in the Behavioral
Sciences in Palo Alto in 1996-97. Near the end of my stay, I was watching my daughter’s
2 David Stark and László Bruszt, Postsocialist Pathways: Transforming Politics and Property in East
Central Europe, 1998.
3 David Stark, “Recombinant Property in East European Capitalism,” 1996.
11
soccer practice on a late afternoon and struck up a conversation with another parent. He
was curious about my research in Hungarian firms and asked me to describe what I had
been finding. “Well, I sometimes have difficulty knowing where one firm ends and
another one begins,” I began. He nodded an encouragement to continue. I went on to
mention the blurring of public and private and then how firms sometimes collaborate on
projects without getting all the property arrangements settled at the outset. At each step
he kept me going with an encouraging “Yeah, and . . . ?” After four or five such
promptings, he interrupted, “You’re not talking about Hungary. You’re talking about
Silicon Valley.”
The comic-strip version of this narrative would now show a big lightbulb glowing in a
bubble over my head. Indeed, I am aware that it seems almost comical that I could be
standing on the side of a playing field and suddenly realize that while I had been studying
a major social transformation elsewhere, another one was taking place in the society in
which I lived. I knew that the two processes could not be the same, but the possibility of
understanding that difference sparked my curiosity. If I had learned anything from my
research on the democratic revolutions in Eastern Europe, it was that if I wanted to study
the so-called digital revolution it would have to be by getting right into the middle of it.
How do you study a hurricane or a tornado? You fly into it and gather data while it is
happening.
In the fall of 1997 I joined the faculty of Columbia University, attracted by the prospect
of participating in the rebuilding of a renowned department and also by the opportunities
that New York City posed as a research site. It was not Silicon Valley, but it did have its
own version of that phenomenon called Silicon Alley, home to thousands of new start-up
companies in the field that came to be known as “new media.” With Monique Girard, I
conducted several years of ethnographic research among programmers and interactive
designers in one these new-media start-ups, a firm we call NetKnowHow. That
ethnography forms the basis of chapter 3 of this book. Toolmakers of a different sort
from my Minotaur machinists, the new-media workers at NetKnowHow were building
the tools of the digital economy. And, like the Minotaur toolmakers, the new-media
toolmakers were using the ambiguity of multiple evaluative principles to navigate
through uncharted territory – in this case the Internet land rush in what was then known
as the new economy.
While we were studying NetKnowHow, we were also learning that our predecessors at
Columbia University had put together a model research program on which we could
attempt to build. At midcentury, organizational analysts at Columbia, led by Robert
Merton and Paul Lazarsfeld, launched two ambitious research programs. On one track,
Merton and his graduate students examined the origins and functioning of bureaucracy
using various research methods (Peter Blau, small groups; Alvin Gouldner, ethnography;
James Coleman, survey research). On a second, parallel track, Merton and Lazarsfeld
established the Bureau of Radio Research to examine the dynamics of mass
communication, pioneering in the use of focus groups and methods to study the
demography of audience reception.
12
The idea of studying an organizational form and a form of communication in a period
when each was being reconfigured stimulated a series of conversations with my graduate
students. Whereas our Columbia predecessors had charted the structure of bureaucratic
organizations in the era of mass communication, we realized that our research challenge
would be to chart the emergence of collaborative organizational forms in an era of new
interactive technologies. But the difference between the earlier and the current Columbia
projects goes beyond the fact that our predecessors studied hierarchy and we study
heterarchy, or that they studied the social technologies of mass production and mass
communication and we study the social technologies of collaborative production and
collaborative communication. The important difference, as Bob Merton emphasized in
our conversations, is that the two tracks of research (organizational forms on one side,
communication technologies on the other) can no longer be conducted along parallel
lines. As we shall see in chapter 3 and especially in chapter 4, in our era the design of
heterarchical organization cannot be separated from design of the digital interface.
The newly founded Columbia Center on Organizational Innovation provided an
institutional platform for this research program. One of our first activities was a yearlong
speakers’ series and graduate seminar titled “Heterarchy.” With all the talk about
“multiple evaluative principles” in the seminar, we wondered whether it might be
interesting to establish a baseline by studying an organization in which there was no
ambiguity or disagreement about value. Such a study would provide a kind of standard
against which we could better understand the workings of heterarchical organizations
where value was in contention. With Daniel Beunza, one of the grad students in the
seminar, I took up this project and secured access to what we assumed would be the gold
standard of a single metric of value – the arbitrage trading room of a major international
investment bank on Wall Street, the scene of Chapter Four.
In ethnographic field research, as in any research method, nothing is more productive
than surprise. And the biggest and best surprise is the one that goes against the research
design. But precisely because it is research – that is, not a search for the already known –
you can never anticipate what it will be, and, on that account, you can never deliberately
design your project for that surprise. The best you can do is be prepared so you will be
able to recognize the opportunity for novel insight. Chapter 4 reports such an occasion of
surprise. Traders know that they are looking for value, for profits, but the specific
instances of such rewarding opportunities cannot be known in advance. Arbitrage
operates by making novel associations across highly abstracted qualities of securities and
their derivatives. But within arbitrage there are multiple principles for searching for
value, and the trading room is organized as a kind of cognitive ecology exploiting the
diversity of these principles to recognize (in fact, as we shall see, re-cognize) novel
associations.
* * *
As an epigraph, I quote a passage where Dante advises, “Fix not thy mind / On one place
only.” Certainly not advice that I consciously followed from the outset, it does account,
after the fact, for a process repeatedly at work in the zigs and zags, the false starts, the
seized opportunities, and the sudden turns in this intellectual journey. But if there were
13
abrupt turns, surely they could just as well be cast as returns: going to Hungary, yet
returning to the problem I had not been able to study in Yugoslavia; shifting from an
outdated, decaying industrial setting to high tech, yet returning to the notion of
entrepreneurial recombination that I had found only in incipient and distorted form in
Hungary; going downtown to the trading room, yet returning to the notion of diversity of
principles as key for recognition and re-cognition. So, yes, returns, many happy returns.
Because in these ritornellos (here the musical term), the theme is not simply repeated but
returns – recognizable, yet in a different form. And the more disconcerting the difference,
the more delightful the recognition. A curious, and I hope you will find a generative,
fixed idea this: to fix not thy mind on one place only.
The key insight, however, is not that sociologists should search in diverse settings. I bring
these three cases together to make a broader argument about search. The idea, in its most
simple form, is that organizations can see more, search better, with a sociological double
vision. Stated in greatest brevity: The Minotaur toolmakers recognize their worth and
their identities in the discrepancies between competing orders of the worth; the new-
media firm recognizes opportunities in the contention over what is a resource; the traders
recognize value in the diversity of principles of valuation. A society recognizes its
potential when it truly gives recognition to a multiplicity of ways of defining what is
valuable. Our wealth, no, even better, our worth is increased when there is open
disagreement about what is worthy.
Durham, England, December 2007
14
1 ____________________________________________________
Heterarchy: the Organization of Dissonance
SEARCHING QUESTIONS
Search is the watchword of the information age. Among the many new information
technologies that are reshaping work and daily life, perhaps none are more empowering
than the new technologies of search. With a few keywords at the toolbar, we can access
enormous databases to find an obscure article by a long-distant colleague, identify the
supplier of a critical component, read about the benefits and side effects of new
pharmaceutical products or medical procedures, or find the fact that immediately settles a
dispute about the performance of an opera, an athlete, or a mutual fund. Whereas the
steam engine, the electrical turbine, the internal combustion engine, and the jet engine
propelled the industrial economy, search engines power the economy of information.
Search is among the key concepts of this book because search is the process that best
exemplifies the challenges of contemporary organization. Ironically, those challenges
cannot be solved by the search technologies that are transforming how we work, how we
shop, and even how we locate ourselves in social and physical space. Certainly, new
search technologies have become invaluable for how organizations manage knowledge.
But the results they yield are of precisely the wrong kind to answer the more fundamental
problems confronting organizations today. The more challenging type of search does not
yield coordinates for a preidentified entity or category, as, for example, when I search for
an e-mail address or for a recent paper that I heard presented at a conference. Nor is it
even a search for solutions to clearly defined problems. The fundamental challenge is the
kind of search during which you do not know what you are looking for but will recognize
it when you find it.
Academics are familiar with the process. In fact, to distinguish it from the search for the
already known, we have a ready term: research.4 In other fields, the process goes by a
4 If you are a reader searching for a dissertation topic, you are familiar with this kind of search. If you
already knew precisely what you were looking for, chances are it has already been done. Innovative
research expands the problem field. The challenge therefore is to work enough outside the already known
while casting the research such that the new problem, concept, method, insight will be recognized by
others.
15
different name: innovation. John Dewey, one of the founders of the pragmatist school of
American philosophy, used another term: inquiry.5
Dewey was emphatic that inquiry, as a distinctive mode of search, should be
distinguished from problem solving. His clarification merits quoting at length
because it so nicely turns our attention from a well-defined problem to the more
interesting case of a perplexing situation:
[I]t is artificial, so far as thinking is concerned, to start with a ready-made
problem, a problem made out of whole cloth or arising from a vacuum. In reality
such a “problem” is simply an assigned task. There is not at first a situation and a
problem, much less just a problem and no situation. There is a troubled,
perplexed, trying situation, where the difficulty is, as it were, spread throughout
the entire situation, infecting it as a whole. If we knew just what the difficulty was
and where it lay, the job of reflection would be much easier than it is. . . . In fact,
we know what the problem exactly is simultaneously with finding a way out and
getting it resolved.6
Dewey’s evocation of perplexed and troubling situations will ring true to any reader who
has faced the challenge of knowing that sometimes you must search even when you do
not know what you are looking for. We grasp the difference between an assigned task, as
Dewey labels a simple search, and a challenging situation. We sense that there is a
difference between occasions when we look for solutions within a set of established
parameters and other occasions (Dewey would say situations) rife with uncertainty and
yet, precisely because of that, also ripe with possibilities.7 Life would be blessedly simple
if we could solve our searching questions with a few clicks at the toolbar. But it would be
neither interesting nor satisfying.
In their study of new-product development in cellular telephones, blue jeans, and medical
devices, Richard Lester and Michael Piore succinctly capture the difference between the
two types of search.8 In the analytic mode, the task of the good manager is to clearly
identify the problem, break it down into independent components, and organize a series
of decisions about how best to solve them. But Lester and Piore conclude that the most
important component of innovation is a process that is not directed toward the solution of
well-defined problems. This second mode is characterized by interpretation. Whereas
problem solving involves the precise exchange of information, the interpretive model
fosters open-ended, unpredictable conversation. Where the former seeks clarity, the latter
seeks spaces of ambiguity since the challenge is to integrate knowledge across
heterogeneous domains. Lester and Piore demonstrate that each of their cases of radical
innovation involves combinations across disparate fields: Fashion jeans are the marriage
5 Dewey was working in the pragmatist tradition that began with Charles Sanders Peirce’s idea of
communities of inquiry to account for the ways that people construct knowledge in collaboration with
others.
6 John Dewey, “Analysis of Reflective Thinking,” [1933] 1998, p. 140 (emphasis in the original).
7 Ann Mische and Harrison White, “Between Conversation and Situation: Public Switching Dynamics
across Networks,” 1998.
8 Richard K. Lester and Michael J. Piore, Innovation: The Missing Dimension, 2004.
16
of traditional workmen’s clothing and laundry technology borrowed from hospitals and
hotels. Medical devices draw on the basic life sciences as well as clinical practice. And
cellular phones recombine in novel form radio and telephone technologies. They
conclude that “without integration across the borders separating these different fields,
there would have been no new products at all.”9
Because innovation, in this view, involves bringing together incompatible traditions, we
should not expect that the process will be harmonious. With hindsight, it is easy to see
that high-fashion faded blue jeans are a recombination of workmen’s clothing and
laundry technology. If we can say that “of course!” cellular phones are the marriage of
the radio and the telephone, it is only because, as Lester and Piore show, the respective
communities worked from the starting point of their differences. In hindsight, we infer
that they must have known all along what they were looking for whereas, in fact, as
Dewey and the pragmatists argued, it was only in the conflictual process of attempting to
make a transformation in the world that the problem could even be formulated.10
Working broadly from within this same tradition, Lester and Piore observe
In many industries, innovations can be identified that did not, at least initially,
address a particular need or problem, or for which the problem became apparent
only after the product was in use. In such cases, the product developer frequently
starts out without really knowing what she is trying to create. (p. 41, emphasis
added)
The search problems that this book addresses are thus different from the everyday notion
of exploration, if that term calls to mind a process like exploring for petroleum or similar
searches for a good that is known in advance. Following James March, I shall use the
term exploration narrowly to refer to processes that break from successful, familiar
routines to search into the unknown.11 That is, if exploring for territory is your metaphor
of choice, the challenging searches would be efforts to recognize the terrae incognitae.
Stated as recognition of the unknown, the process of innovation is paradoxical, for it
involves a curious cognitive function of recognizing what is not yet formulated as a
category. It is one thing to recognize an already-identified pattern, but quite another to
make a new association. To take some mundane yet now ubiquitous examples: gas for
industrial lighting in the nineteenth century (recognizing a waste product of the process
of converting coal to coke as a valuable resource);12 the shopping cart (a basket on
wheels);13 the parking meter (a hitching post with a clock-type mainspring); the car radio
(pioneered by a family firm, now famously Motorola, that had made accessories for
carriages and sought a market in accessories for the new automobile); the airport
shopping mall (combining consumption and travel); and, more troubling, the megachurch
9 Lester and Piore, Innovation, pp. 14-15.
10 For a similar account of discovering the world through the conflictual process of attempting to transform
it, see Tracy Kidder, The Soul of a New Machine, 1981.
11 James G. March, “Exploration and Exploitation in Organizational Learning,” 1991.
12 Wolfgang Schivelbusch, Disenchanted Night: The Industrialization of Light in the Nineteenth Century,
1995, p. 18.
13 Catherine Grandclément, "Wheeling One's Groceries around the Store,” 2008.
17
of American exurbia (combining Wal-Mart architecture, televangelism, and highly
niched small groups or cells from the repertoire of underground movements to create a
new form of spirituality as mass-customized consumption). Each example of
recombination or repurposing involved a category switch, obvious now in retrospect
precisely because each could be recognized with little cognitive difficulty by the user.
Whether we refer to the process as research, innovation, exploration, or inquiry, the kind
of search that works through interpretation rather than simply managing information
requires reflective cognition. Whether in science, politics, civic associations, or business,
it is not enough just to embark on a search for an unknown breakthrough; you must also
be able to recognize it when you find it. And you must present the category-breaking
solutions in forms that are recognizable to other scientists, citizens, activists, investors, or
users. This is a tall challenge, for the more ambitious the project, the more deliberately ill
defined the initial process of search; and the more demanding the processes of eventual
recognition, the greater is the discomforting ambiguity facing the innovating
organization. Innovation, as Joseph Schumpeter observed, is recombination; but, as
Schumpeter argued as well, it is also deeply disruptive of cultural taken-for-granteds and
routines of organizational cognition.
We can now appreciate again Dewey’s characterization of inquiry as provoked by
“troubled, perplexed, trying situations.” Organizations facing such perplexing situations
have several options. The first temptation for the leaders of science projects, corporate
projects, or civic projects is to immediately address ambiguous situations pregnant with
interpretive search by using the clearly defined problem-solving strategy of analytic
search. But such a managerialist strategy of early top-down control entails the risk of
forgoing the big opportunities represented in innovations such as cellular phones, fashion
jeans, and breakthrough medical equipment. Although problem solving eventually came
into the picture, interpretation was the dominant mode of product development that led to
innovative success in each of these cases.14
The alternative strategy is more in line with John Dewey’s notion of inquiry as a guide
for innovation. Dewey’s attention to the productive possibilities of situations is the lesson
that I try to keep in mind throughout this book. Instead of avoiding perplexing situations,
organizations can embrace them. Even more radically, organizations can take the next
step: If perplexing situations provoke innovative inquiry, then why not build
organizations that generate such situations? Instead of merely responding to external
situations as they happen to present themselves, why not foster organizational forms that
regularly and recursively produce perplexing situations within the organization itself?
Organizations that adopt such forms will then be poised to undertake the challenging task
of ongoing innovation.
At the most elementary level, a perplexing situation is produced when there is principled
disagreement about what counts. Organizations that seek to generate productive,
perplexing situations can work from this basic starting point. Instead of enforcing a single
principle of evaluation as the only legitimate framework, they recognize that it is
14 Lester and Piore, Innovation.
18
legitimate to articulate alternative conceptions of what is valuable, what is worthy, what
counts. Such organizations have heterogeneous criteria of organizational “goods.” To
signal that this organizational form is a mode of governance that differs from a hierarchy
of command and a conceptual hierarchy of cognitive categories, I refer to it as a
heterarchy. As my case ethnographies demonstrate, heterarchies are cognitive ecologies
that facilitate the work of reflective cognition.
Such organizations, we shall see, are not frictionless. But friction is not something to be
avoided at all costs. We all prefer a smooth ride, but as you and your tire dealer know,
when taking a sharp curve, we count on friction to keep us on course. Friction can be
destructive. But, as the designers of the U.S. Constitution well understood when they
built the friction of checks and balances into our system of government, it can also be a
principled component of a functioning system with productive outcomes. That is, having
multiple performance criteria can produce a resourceful dissonance. If you are confident
that you know precisely what resources your organization will need in the indefinite
future to meet stable and predictable markets (or continue to get grants to meet your
unchanging mission as a nonprofit or a research operation), then dissonance is an
avoidable headache that you need not abide. But for many organizations the “foreseeable
future” is not long distant. Where the organizational environment is turbulent and there is
uncertainty about what might constitute a resource under changed conditions, contending
frameworks of value can themselves be a valuable organizational resource.
Entrepreneurship then, in this view, exploits uncertainty. Not the property of an
individual personality but, instead, the function of an organizational form,
entrepreneurship is the ability to keep multiple principles of evaluation in play and to
benefit from that productive friction.
FOR A SOCIOLOGY OF WORTH
What counts? Each of us confronts this question on a daily basis. Faced with decisions
involving incommensurable frameworks – work versus family life, career opportunities
versus loyalty to friends or attachment to a locality, vacations versus investments for
retirement, and so on – we ask ourselves what really counts. What is valuable, and by
what measures? As our lives are a search to find out what is really valuable, we try, we
fail, and we try again to learn from our mistakes.
In our roles as actors in organizations we face similar questions. In these organizational
settings we need to sift through a barrage of information – seemingly growing at an
exponential rate – to select what counts, what matters, what is of true relevance. More
fundamentally, organizations are engaged in a search for what is valuable. What new
products can be brought to market? What new technologies or production processes
should be pursued? Which will prove to be valuable and which will be a costly dead end?
And how should the performance of units, of teams, and of the individual employees
within them be evaluated? Nonprofits might be tax-exempt, but they are not exempt from
similar questions. Which campaigns and projects are worthy of pursuit? Will our
members, constituents, activists, targeted communities, and donors recognize their value,
perhaps quickly, or perhaps too late?
19
Within the sociological discipline, economic sociology is the specialization that deals
with societal and organizational questions of the valuable. The field’s founding moment
took place more than a half century ago at Harvard, where Talcott Parsons was
developing his grand designs for sociology. Parsons’s ambitions were imperial, with the
aim of reshaping much of the social sciences. But his instincts in academic politics led
him to be wary of economics as the discipline that could thwart his agenda if his program
was perceived as encroaching on its territory. Whereas sociology, psychology, and
anthropology could be claimed outright, economics would have to be maneuvered
around. To dispel any doubt about his intentions, Parsons walked down the hall in
Harvard’s Littauer Center to his colleagues in the Economics Department, alerting them
to his ambitious plans and assuring them that he had no designs on their terrain.15 Thus,
Parsons made a pact. In my gloss: You, economists, study value; we, the sociologists,
will study values. You will have claim on the economy; we will stake our claim on the
social relations in which economies are embedded.16
Although Parsons’s Pact suggests that we must choose a single vantage point – value or
values, economy or social relations – I adopt an analytic strategy of fusing across the
divide.17 The key concept in this fusion is the notion of worth. The polysemic character
of the term – worth – signals concern with fundamental problems of value while
recognizing that all economies have a moral component. Rather than the static fixtures of
value and values, it focuses instead on ongoing processes of valuation – whether in
assessing the value of firms under competing metrics of performance, or in studying the
incommensurable assessments made in everyday life. “What are you worth?” is a
question that can be unambiguous when constrained by context (as, for example, when
applying at a bank for a mortgage). But the same question in an art gallery—“Yes, but
what is it worth?”—already suggests that value might be different from price. And when
the question comes up among friends—“Honey, do you really think he’s worth it?”—we
know that several opposed evaluative criteria have been brought into play.
Worth is a wonderful word with deep roots (wort) in the old Anglo-Saxon tongue before
the Norman invasion brought the Latinate separation of value and values into the English
15 Charles Camic, “The Making of a Method: A Historical Reinterpretation of the Early Parsons,” 1987.
Although he characterizes it slightly differently, Velthuis similarly argues that in the mid-1930s Parsons
and the economist Lionel Robbins agreed on the terms of a disciplinary division of labor. Olav Velthuis,
“The Changing Relationship between Economic Sociology and Institutional Economics: From Talcott
Parsons to Mark Granovetter,” 1999.
16 Parsons’s Pact thus imposed a jurisdictional division of the social sciences that placed constraints on
sociology by limiting its range. Yet, by delimiting a legitimate object of study – society, though not the
economy – it ensured that the discipline would flourish in the great postwar expansion of the social
sciences.
17 Economic sociologists have adopted various strategies to break with Parsons’s Pact. In Markets from
Networks (2002) Harrison White basically turns the tables on the terms of the pact. Markets, he argues, are
not embedded in social relations; they are social relations. Instead of accepting the economists’ conception
of markets, White has developed a sociological theory of markets. As the counterpart to Harrison White,
Viviana Zelizer pointed out a way to escape from Parsons’s Pact along the value/values dimension. In
Pricing the Priceless Child (1985) Zelizer examines the interrelation between market or price and personal
or moral values in a rich historical study of child labor, adoption, and insurance. Zelizer’s later work on the
social meaning of money, on payment systems, and circuits of commerce boldly transgresses and
transcends the disciplinary divide.
20
language. With its double connotations of an economic good and a moral good, worth is
very difficult to translate into Italian, for example. None of the candidate terms has that
twinned salience, as each is heavily loaded toward either the value or the values side. On
the other hand, there is no such verb as “to worth” in English. We can “value something
as worth a great deal” or “judge someone as worthy,” but we cannot “worth” something
or someone. Meanwhile Italian has a perfectly apt verb, stimare. In this case, it was
English that separated it into “to estimate” (on the value side) and “to esteem” (on the
values side) – connotations that are equally salient in the Italian verb.18
Perhaps more than anyone on this topic, John Dewey was aware not only of how
everyday language constrains our thinking but also of how it can reveal insights about the
concepts we deploy. In his Theory of Valuation, Dewey explores the double meanings in
ordinary speech and points to words such as praise and appraise that parse in different
directions from a common root. After noting the twins estimate and esteem, Dewey
observes that it is suggestive “that praise, prize and price are all derived from the same
Latin word; that appreciate and appraise were once used interchangeably; and that ‘dear’
is still used as equivalent both to ‘precious’ and to ‘costly’ in monetary price.”19
With Dewey, I agree that we cannot appeal to everyday language to solve analytic
problems. But I also take his point that when we see some commonsense terms pulling
apart and others joining together, we should pay attention, for we will usually find a
problem worth studying. In particular, we can often see how ideas from ordinary
language become incorporated in the false dichotomies that we use in analysis—for
example, in viewing ends as values that are prized while regarding means as objects that
are appraised. For Dewey, it makes as much sense to see means as prized and ends
appraised. His pragmatic theory of inquiry as action shatters these dichotomies.
In the closing section of his Theory of Valuation Dewey provides a diagnosis of the crisis
of his time. Writing in 1939, he observes that emotional loyalties and attachments are not
linked to scientific debate, while ideas with their origin in scientific inquiry have not
succeeded in gaining emotional force.20 For Dewey this is a practical problem, and an
analytic one. In the penultimate paragraph he highlights this problem by returning to the
discussion of common speech with which he began his study.
In fact, and in net outcome, the previous discussion does not point in the least to
supersession of the emotive by the intellectual. Its only and complete import is the
18 I recently encountered this problem when giving a simultaneously translated public lecture at the
University of Modena and later when my essay “For a Sociology of Worth” was translated for an Italian
journal. On a more general note, writers who make words work very hard should give them due recognition
– or, at least, follow the lead of Humpty-Dumpty in Alice in Wonderland: “When I make a word do extra
work, I’m always sure to pay it very well.”
19 John Dewey, Theory of Valuation, 1939, pp. 5-6.
20 “We are living in a period in which emotional loyalties and attachments are centered on objects that no
longer command that intellectual loyalty which has the sanction of the methods which attain valid
conclusions in scientific inquiry, while ideas that have their origin in the rationale of inquiry have not as yet
succeeded in acquiring the force that only emotional ardor provides. The practical problem that has to be
faced is the establishment of cultural conditions that will support the kinds of behavior in which emotions
and ideas, desires and appraisals, are integrated.” Dewey, Theory of Valuation, p. 65.
21
need for their integration in behavior – behavior in which, according to common
speech, the head and the heart work together, in which, to use more technical
language, prizing and appraising unite in direction of action. (p. 65)
It is, then, with Dewey that we embark on an analysis of worth to develop tools for
understanding a richer calculus that integrates value and values, the intellectual and the
emotive, valuation and the evaluative. When we see that acts of estimation entail
practices of esteem, we see that payment systems are about recognition as well as about
monetary rewards. When we see inquiry as action, we see search less as a process of
finding what we already know to be valuable than as distributed practices for recognizing
opportunities by re-cognizing resources. When we see calculation as not separated from
judgment, we see that what counts in the processing of information is the capacity for
interpretation.
Following Dewey will require that economic sociology’s preoccupation with the analysis
of institutions should be augmented by close study of indeterminate situations. In making
this shift, economic sociology can draw lessons from developments in the field of science
and technology studies (STS). During its inaugural stage, the sociology of science, led by
Robert Merton, carved out a distinctive place for sociology by focusing on the institutions
of science<m->the structure of rewards and careers, patterns of citations, and the norms
of scientific life. Not rejecting this work but building on it, the next generation of STS
researchers moved into the laboratories to study scientists at work, observing the difficult
labor of stabilizing facts, the challenges of replicating experiments, and the ongoing
controversies of science in the making.21
Just as post-Mertonian studies of science moved from studying the institutions in which
scientists were embedded to analyzing the actual practices of scientists in the laboratory,
so an economic sociology can move from studying the institutions in which economic
activity is embedded to analyzing the actual evaluative and calculative practices of actors
at work.
In making this move, I draw on insights by Luc Boltanski and Laurent Thévenot, whose
book On Justification: The Economies of Worth, only recently translated, was originally
published in France in 1991.22 Boltanski, a sociologist, and Thévenot, an economist, are
part of a group of French economic sociologists23 whose work is collectively known as
21 Bruno Latour and Steve Woolgar, Laboratory Life: The Social Construction of Scientific Facts, 1979;
Trevor Pinch, Confronting Nature: The Sociology of Solar-Neutrino Detection, 1986; and Bruno Latour,
Science in Action: How to Follow Scientists and Engineers through Society, 1987.
22 Luc Boltanski and Laurent Thévenot, On Justification: The Economies of Worth, 2006. For an
accessible introduction to the major concepts in article form, see Boltanski and Thévenot, “The Sociology
of Critical Capacity,” 1999. Michèle Lamont together with Thévenot led an exciting project involving a set
of empirical studies, pairing French and American researchers, that demonstrates the fruitful application of
these ideas. See their edited collection, Rethinking Comparative Cultural Sociology: Repertoires of
Evaluation in France and the United States, 2000.
23 For a recent collection in English, see Conventions and Structures in Economic Organization: Markets,
Networks and Hierarchies, edited by Olivier Favereau and Emmanuel Lazega, 2002. Introductions to the
economics of conventions are provided in John Wilkinson, “A New Paradigm for Economic Analysis?”
1997; and Thierry Levy, “The Theory of Conventions and a New Theory of the Firm,” 2001.
22
“the economics of convention.”24 Just as Harrison White has developed a sociological
theory of markets, Boltanski and Thévenot are developing a sociological theory of value.
Their first move is to demonstrate that there is not just one way of making value but that
modern economies comprise multiple principles of evaluation. A modern economy (and
note that the word is not society but economy) is not a single social order but contains
multiple “orders of worth.”
One might object that this is not an escape from Parsons’s Pact. After all, as soon as you
make a plural out of value, you get values. But the orders of worth of the French school,
in fact, differ from the cultural systems of Parsonsian values and from the classificatory
codes of the new institutionalists. For my colleagues in American economic sociology,
values are counterposed to calculation; they are outside and distant from calculation.
More precisely, if cultural taken-for-granteds are the embeddings for value, they make
calculation possible precisely because they are a kind of antimatter to calculation.25 For
my French conventionalist colleagues, on the other hand, orders of worth are not values
counterposed to value but are constitutive of value. Orders of worth are the very fabric of
calculation, of rationality, of value.
Boltanski and Thévenot’s work refuses a dichotomy of value and values; instead, it fuses
them in the concept of worth. Although we are accustomed to thinking about “moral
economies” as opposed to market economies, for example, in the norms of close-knit
communities that embodied precapitalist traditions of the just and fair, 26 Boltanski and
Thévenot see all economies as moral economies. Each of the orders of worth operating in
the domain that we conventionally denominate as “the economy” is an economy. And, as
an economy, each is a moral order.
Boltanski and Thévenot delineate six discrete orders of worth, each epitomized by a
particular moral philosopher. From their perspective, I would be mistaken to say that I
live in a market economy. Markets are, indeed, one of the organizing principles of the
U.S. economy. But, as they show in their study of the domain of the corporation, in
addition to a market rationality (exemplified by the moral philosophy of Adam Smith), a
24 The French conventionalist school began with the idea that the qualities of labor were unknown prior to
hiring, but soon extended this idea to other commodities that suffered from deficiencies of “incomplete
contracts.” (The market for used cars is now a well-known example; see George A. Akerlof, “The Market
for ‘Lemons’: Quality Uncertainty and the Market Mechanism,” 1970.) Guidance systems and other
instrumentation in space vehicles provide a different kind of example in which the buyer cannot know in
advance how the qualities of the product will perform in extreme conditions. Of even greater interest are
cases in which the parties embark on complex collaborations in which the fundamental characteristics of
the joint product are not known in advance but are themselves the key aim of the collaboration. In this case,
the critical quality is the ability to collaborate. On discursive quality standards, see especially Charles Sabel
and Jane Prokop, "Stabilization through Reorganization?” 1996.
25 Paul J. DiMaggio and Walter W. Powell, “Introduction,” in The New Institutionalism in Organizational
Analysis, 1991. In this agenda-setting statement for the “new institutionalism” in economic sociology,
DiMaggio and Powell present a sharp critique of Parsons (pp. 15-22), making clear that whereas the old
institutionalism was about “values, norms, attitudes,” the new institutionalism analyzes “classifications,
routines, scripts, schema.” Emphasizing the importance of “unreflective activity,” DiMaggio and Powell
explicitly counterpose such cultural taken-for-granteds to calculative behavior (p. 22).
26 Social historian E. P. Thompson emphasized the force of such traditions in his pioneering article “The
Moral Economy of the English Crowd in the Eighteenth Century,” 1971.
23
modern economy also has an industrial or technological rationality (Saint Simon),
another organized around a civic logic (Rousseau), and still others arrayed according to
principles of loyalty (Bossuet), inspiration (Augustine), and renown or fame (Hobbes).
Boltanski and Thévenot are emphatic that their orders of worth do not map to separate
domains.27 Inspiration, for example, is not the special province of the world of art; nor
does a civic rationality correspond to the public sphere; and the market order can operate
as well in the domains of academia and religion. In the second part of On Justification,
Boltanski and Thévenot illustrate the operation of each of the orders of worth within a
single domain, that of the large corporation, through a content analysis of six best-selling
guidebooks to being a good manager – each written from the perspective of a different
order respectively.
As an example that each of the orders of worth is salient in the world of academia, take
letters of recommendation for faculty appointments. You do not need to read a great
many such letters to recognize that recommenders frequently refer to multiple principles
of evaluation. In fact, a given letter might include performance criteria from each of the
six orders of worth. We would not be surprised, for example, to read that a given
candidate is “very creative” (the order of inspiration); that she is incredibly “productive”
(the industrial); and that she is a “good citizen” (the civic). Moreover, the same letter
could note that her work is “frequently cited” (the order of fame or renown) and that she
is fiercely “loyal to her graduate students” (check off another). Has the letter writer
neglected the market order? We are not likely to hear about an academic as the author of
a “best-selling” book. Look through the letter again and you might find that the candidate
“has a strong record of getting grants.”
As coherent principles of evaluation, each of the orders of worth has distinctive and
incommensurable principles of equivalence. Each defines the good, the just, and the fair –
but according to different criteria of judgment. Each qualifies persons and objects with a
distinctive grammar or logic. As principles of evaluation, the orders involve systematic
associations of concepts; but the entities that populate an order of worth are not limited to
persons and ideas. On Justification shows in rich detail how the principles of evaluation
established in each order of worth entail discrete metrics, measuring “instruments,” and
proofs of worth objectified in artifacts and objects in the material world.
In this view, rational calculation is not opposed to moral judgment; instead, rationality
works within orders of worth. As such, I interpret Boltanski and Thévenot’s work as
casting new meaning on the term “bounded rationality.” Whereas we conventionally
27 Despite the similarity of a notion of multiple rationalities, Boltanski and Thévenot’s framework differs
markedly from that of Roger Friedland and Robert Alford, who identify several institutional domains, each
with its distinctive “logic of action” (“Bringing Society Back In: Symbols, Practices, and Institutional
Contradictions,” 1991). Whereas Friedland and Alford parse logics to domains (e.g., affective in the family,
cognitive in the market, etc.), Boltanski and Thévenot’s respective orders of worth are not isolated to
specific societal domains. Although it shares a similar intuition, their view also differs from that of Wendy
Espeland and Mitchell Stevens, who argue that “because societies are complexes of multiple institutions,
they are characterized by multiple modes of valuing” (“Commensuration as a Social Process,” 1998, p.
332). Because Boltanski and Thévenot’s orders of worth do not parse to separate institutions, all can be
operating in the economy.
24
think about bounded rationality as the cognitive limits on rationality (as, for example, in
the usage of the term by economist Oliver Williamson), in Boltanski and Thévenot’s
work rationality is possible only insofar as it takes place within the boundaries and
through the social technologies of particular orders of worth. In this latter sense we
should speak – and with a very different meaning – of bounded rationalities.
Drawing from Boltanski and Thévenot, as well as from Michel Callon and his
colleagues,28 in the framework that I adopt in this book, the familiar culturalist versus
materialist opposition becomes meaningless. All economic objects are thoroughly
cultural, and no moral order could operate without specific material objects. Moreover,
rationality is not something “above” the preconscious, nor is calculation somehow
“below” moral orderings. From my field research in Hungary, where I found a plurality
of economic forms operating in a single factory (see chapter 2), I was predisposed to the
idea that organizations are settings where multiple principles of evaluation are at play.
But because I do not confine these to the six moral orders of On Justification, I specify
the evaluative principles differently from one case to another, as is appropriate for each
case. Most importantly, my field research leads me to different conclusions from those of
Boltanski and Thévenot. As I shall argue in the next section and develop in the
subsequent substantive chapters, whereas they see orders of worth as making action
possible by resolving problems of uncertainty, my case ethnographies led me to see the
mix of evaluative principles as creating uncertainty and therefore as opening
opportunities for action.
ENTREPRENEURSHIP AT THE OVERLAP
Economic sociology, like many fields in the discipline, is populated with dualisms. In
addition to the dichotomy of value and values and the perennial “structure versus
agency,” we also find notions of calculation versus trust, and efficiency versus
legitimacy. One particularly productive distinction that continues to generate insights was
formulated by economist Frank Knight as the problem of risk versus uncertainty.29 For
Knight, uncertainty and risk are both shaped by the fact that the future is unknown. But
the two are not the same. In circumstances of risk, chances are calculable; that is, the
distribution of outcomes can be expressed in some probabilistic terms. Uncertainty,
however, lacks calculation: “All bets are off.”
The problem of uncertainty, it must be emphasized, is not a function of the limited
calculative power of the human actors confronting it. Instead it is a property of the
situation. The situation is indeterminate. John Dewey, writing about the same time as
Knight but in a different context, nicely expresses the problem of indeterminate
situations:
A variety of names serves to characterize indeterminate situations. They are
disturbed, troubled, ambiguous, confused, full of conflicting tendencies, obscure,
28 See especially Michel Callon and Fabian Muniesa, "Economic Markets as Calculative Collective
Devices," 2005.
29 Frank H. Knight, Risk, Uncertainty and Profit, 1921.
25
etc. It is the situation that has these traits. We are doubtful because the situation is
inherently doubtful.30
Santa Fe Institute economist David Lane succinctly summarizes the situation of
uncertainty: “the question is not what we do not know, but what cannot be known.”31
Although economists are now giving renewed attention to the problem of uncertainty,32
the typical view in the discipline, institutionalized in the neoclassical framework, was to
frame all economic action as cases of risk.33 Knight could see the direction that his
discipline was moving, and in his view the tendency to see all situations as those in which
the distribution of outcomes could be expressed in probabilistic terms would deprive
economists of the ability to grasp a problem that should be at the core of the discipline.
Knight argued that a world of generalized probabilistic knowledge of the future leaves no
place for profit and, as a consequence, no place for the entrepreneur. For Knight, what
defines profit is that it cannot be measured ex ante – as distinct from rents, which
constitute contractualizable residual revenue. In Knight’s framework, the entrepreneur,
properly speaking, is not rewarded for risk taking but, instead, is rewarded for an ability
to exploit uncertainty.
In Boltanski and Thévenot’s framework, there is little space for entrepreneurial activity.
For this French school of economic sociology, conventions (of which orders of worth are
a particularly well-elaborated variant) are a way of dealing with uncertainty. They are
engines for turning situations into calculative problems. Orders of worth can be
considered as social technologies to transform uncertainty into risk.34 The limitation of
this view – and here is my departure from Boltanski and Thévenot – is that it does not
give adequate attention to the problem that orders of worth cannot eliminate uncertainty.
In particular, they cannot eliminate the possibility of uncertainty about which order or
convention is operative in a given situation.
Taking this into account, we are in a position to restate the insight of Knight, but now in
new terms: it is precisely this uncertainty that entrepreneurship exploits.
Entrepreneurship is the ability to keep multiple evaluative principles in play and to
exploit the resulting friction of their interplay.
30 John Dewey, “The Pattern of Inquiry,” [1938] 1998, p. 171, emphasis in the original.
31 David Lane, “Models and Aphorisms,” 1995.
32 Adam Brandenburger, “The Power of Paradox: Some Recent Developments in Interactive
Epistemology,” 2007; Sheila Dow and John Hillard, eds., Keynes, Knowledge and Uncertainty, 1995; and
Edward Fullbrook, ed., Intersubjectivity in Economics: Agents and Structures, 2001.
33 Jens Beckert, “What Is Sociological about Economic Sociology? Uncertainty and the Embeddedness of
Economic Action,” 1996.
34 The coordination problems in Boltanski and Thévenot differ from Schelling’s case of a couple who get
separated in a large department store but who do not have a predefined meeting place (Thomas C.
Schelling, The Strategy of Conflict, 1960). The couple succeed in coordinating not despite the circular
specularity but because each knows that the other is trying to coordinate with him/her. Schelling’s case is
more like the common knowledge framework in Lewis’s notion of convention. (David K. Lewis,
Conventions: A Philosophical Study, 1969. For discussion see especially Jean-Pierre Dupuy, “Common
Knowledge, Common Sense,” 1989.) Boltanski and Thévenot’s orders of worth are not about the
application of rules and hence differ from “institutions” either in game theory or in the new
institutionalism.
26
In exploiting the uncertainty about which order of worth is operative, entrepreneurship
involves asset ambiguity.35 From ambiguity it makes an asset; and in creating assets that
can operate in more than one game, it makes assets that are ambiguous. In the subsequent
chapter, for example, we shall see how a group of highly skilled machinists, working in
Communist Hungary, exploited the ambiguity of the “economies” of redistribution,
market, and reciprocity that were operative in their factory. Their strategy was not
without limits and was not always successful, but it well-illustrates the possibilities and
the difficulties of playing in multiple games simultaneously. In chapter 3, we encounter a
new-media firm in Manhattan’s Silicon Alley that attempts to stay ahead of the curve of a
very rapidly changing market by benefiting from the friction between multiple,
incompatible principles for assessing the company’s products – sophisticated e-
commerce websites. In chapter 4, we shall see how a Wall Street trading room is
organized as a cognitive ecology in which the friction among competing principles of
arbitrage generates new ways of recognizing opportunities. That is, although very
different in their settings, the ethnographies will demonstrate how an entrepreneurial
rivalry of performance principles makes assets of ambiguity by keeping open multiple
ways of redefining, and hence recombining and redeploying, resources.
Entrepreneurship exploits the indeterminate situation by keeping open diverse
performance criteria rather than by creating consensus about one set of rules. As such, my
conception of entrepreneurship differs considerably from the strategic action of Neil
Fligstein’s “institutional entrepreneur.” For Fligstein, “Strategic action is the attempt by
social actors to create and maintain stable social worlds (i.e., organizational fields). This
involves the creation of rules to which disparate groups can adhere.”36 Rather than
involving the creation of rules for stability, my concept of entrepreneurship draws from
Harrison White, for whom the problem is not “how is there social order?” but that of
“getting action” in worlds that are already too ordered and rule governed.37
In more general terms, whereas the “new institutionalism” in economic sociology during
the 1980s developed concepts of classificatory rules, scripts, and cultural taken-for-
granteds to explain how organizations gain legitimacy to operate in stabilized
institutional environments, today organizations in rapidly changing environments face the
problem that their taken-for-granteds can soon be out-of-date. In this situation,
entrepreneurship is less about creating stability (building on success) than about creating
disruptions that prevent the path-dependent effects of locking into early successes.38 That
35 Asset ambiguity, thus, contrasts sharply with the concept of asset specificity developed by the economist
Oliver Williamson. By asset specificity Williamson referred to the extent to which investment in a given
asset was specific to a particular transaction. The degree of asset specificity was critical, Williamson
argued, in the decision to make or buy. (Oliver Williamson, “The Economics of Organization: The
Transaction Cost Approach,” 1981.) Charles Sabel and Bruce Kogut, by contrast, explored the problem of
asset interdependence, demonstrating that, under conditions of extraordinarily rapid technological change,
actors engage in hedging strategies vis-à-vis other organizations (partners or competitors) in their
organizational field. (Charles F. Sabel, “Moebius-Strip Organizations and Open Labor Markets,” 1990; and
Bruce Kogut, Weijan Shan, and Gordon Walker, “The Make-or-Cooperate Decision in the Context of an
Industry Network,” 1992.)
36 Neil Fligstein, “Social Skill and Institutional Theory,” 1997, p. 398.
37 See especially Harrison C. White, Identity and Control, 1992.
38 I elaborate these ideas further in chapter 5.
27
is, in fast-breaking fields, among the many challenges facing firms is the problem of
coping with success. Organizations that keep multiple evaluative principles in play, I
argue and demonstrate in my case studies, foster a generative friction39 that disrupts
received categories of business as usual and makes possible an ongoing recombination of
resources.
My perspective thus combines Knight’s notion that entrereneurship exploits uncertainty
with Schumpeter’s emphasis that entrepreneurship is disruptive and recombinatory. My
conception of entrepreneurship as keeping multiple evaluative principles in play and
exploiting the resulting dissonance thus differs from brokerage.
Brokerage, as Ron Burt powerfully demonstrates, exploits “structural holes” in the social
field, strategically locating gaps and profiting from the ability to broker among units that
are otherwise disconnected.40 Brokerage is frequently mistaken for entrepreneurship, but
the two roles and their corresponding social processes are distinct. Whereas the broker is
an insider to none and taxes flows, the entrepreneur is an insider to multiple games and
recombines assets.
For Burt, the key problem is access to information. Bridging ties provide access to new
ideas that are free-floating in the network environment – access that, in Burt’s view, is
not possible through redundant, cohesive ties. In my view, by contrast, the most
innovative ideas are not “out there” in the environment of the group. Instead of waiting to
be found, they must be generated.41 When the problem is the production of new
knowledge rather than simply access to information, the bridging ties of brokerage are
insufficient. Generating new knowledge of the Schumpeterian recombinant type requires
more intimate familiarity than can be produced by weak ties.
Recall Lester and Piore’s observations, mentioned at the outset of this chapter, about
cellular phones as a novel recombination of radio and telephone technologies: “without
integration across the borders separating these different fields, there would have been no
new products at all.”42 For me, the telling phrase in this passage is “integration across the
borders.” Lester and Piore do not refer to “contacts” across borders, for it is not enough
for different communities to be in contact. Recombinant innovation requires that they
interact. In network analytic terms, this suggests that entrepreneurship occurs at the
overlap of cohesive structures where different communities (defined by their cohesive
ties) intersect without dissolving their distinctive network identities.43 The network
39 On the notion of “creative abrasion,” see Dorothy Leonard-Barton, Wellsprings of Knowledge: Building
and Sustaining the Sources of Innovation, 1995; and John Seely Brown and Paul Duguid, “Knowledge and
Organization: A Social-Practice Perspective,” 2001.
40 Ronald Burt, Structural Holes: The Social Structure of Competition, 1995.
41 “When entry-deterring benefits are absent, competition switches from traditional elements of market
structure to the comparative capabilities of the firm to replicate and generate new knowledge.” Bruce
Kogut and Udo Zander, “Knowledge of the Firm, Combinative Capabilities, and the Replication of
Technology,” 1992.
42 Lester and Piore, Innovation, pp. 14-15.
43 To date, network analysts have typically defined cohesion as exclusive; that is, a given node can be a
member of only one cohesive structure. This view was driven more by methodological limitations than by
sociological insight. Georg Simmel, one of the founding figures of network analysis in the early decades of
28
diagrams in figure 1.1 represent these differences between brokerage and
entrepreneurship.
In addition to deep familiarity about resources, the work of recombinant innovation also
requires diversity. What is overlapping are not simply cohesive network structures but
also diverse, even disparate, evaluative principles. Thus, the diagram in figure 1.1 maps
diverse discursive frames as well as network ties.44 Within the same domain space, even
within the same organization, diverse performance criteria are colliding and competing.
Because there are multiple codes to evaluate performance, codified knowledge can be
broken up and recoded. With analogy to genetics, think of the friction of rivaling
principles as increasing the rate of mutation. But the dissonance of diverse evaluative
frameworks does more than simply speed up the production of novelty. The coexistence
of multiple, principled standpoints means that no standpoint can be taken for granted as
the natural order of things. Creative friction yields an organizational reflexivity.
From this perspective, entrepreneurship, as an enabling capacity, proves productive not
so much by encouraging the smooth flow of information or the confirmation of fixed
identities as by fostering a productive friction that disrupts organizational taken-for-
granteds, generates new knowledge, and makes possible the redefinition, redeployment,
and recombination of resources. In short, entrepreneurship occurs not at the gap but
through the generative friction at the overlap of evaluative frameworks.45
the twentieth century, had recognized that an individual could simultaneously participate in more than one
cohesive group. Balazs Vedres and I adopt new methods consistent with this insight to identify a distinctive
network position, “intercohesion,” at the intersection of cohesive group structures. Using historical network
analysis of the ties among the largest 1,800 enterprises in Hungary from 1987 to 2001, we demonstrate that
the entrepreneurial opportunities created by such overlap significantly contribute to high group
performance. Balazs Vedres and David Stark, “Opening Closure: Intercohesion and Entrepreneurial
Dynamics in Business Groups,” forthcoming.
44 As such, my notion of entrepreneurship draws on Ann Mische and Harrison White’s notions of
situations and publics (Mische and White, “Between Conversation and Situation”).
45 Espeland and Stevens offer a related perspective: “We suspect that claims about incommensurables are
likely to arise at the borderlands between institutions, where what counts as an ideal or normal mode of
valuing is uncertain, and where proponents of a particular mode are entrepreneurial” (Espeland and
Stevens, “Commensuration,” p. 332). My position has two points of similarity with this view, emphasizing,
Figure 1.1 Brokerage and Entrepreneurship
Brokerage
information flow
at the gap
Entrepreneurship
creative friction
at the overlap
29
As an ability to promote productive friction, entrepreneurship is not the property of an
individual – it is not, for example, the personality trait of tolerating ambiguity. Instead of
seeing entrepreneurs as individuals, I consider entrepreneurship as a property of
organizations. That is, organizational forms will differ in their capacity to sustain an
ongoing and productive rivalry among performance criteria making it possible to break
out of the lock-in of habituated, unreflective activity. I use the term “heterarchy” to refer
to the organizational forms with this capacity for reflective cognition.
HETERARCHY
Heterarchy46 represents an organizational form of distributed intelligence in which units
are laterally accountable according to diverse principles of evaluation. Two key features
are at work here. In contrast to the vertical authority of hierarchies, heterarchies are
characterized by more crosscutting network structures, reflecting the greater
interdependencies of complex collaboration. They are heterarchical, moreover, because
there is no hierarchical ordering of the competing evaluative principles. Here I first
discuss the feature of heterarchically distributed intelligence organized through lateral
accountability; then I turn to the second, related feature of the heterarchical organization
of diversity as rivalry among performance principles.
Distributing Intelligence
Heterarchy’s twinned features are a response to the increasing complexity of the firm’s
environment, in which it becomes difficult to project future states of the world from
current trends. Analysts at the Santa Fe Institute have several terms to refer to these
complexities. David Lane and Robert Maxfield denote them as “limited foresight
horizons,” in which the strategy horizon of the firm is so unpredictable that the firm
cannot even be certain about what product it will be producing in the near future.47 Stuart
Kauffman adopts the language of the irregular shape of “rugged fitness landscapes” with
multiple optimal solutions.48 A smooth fitness landscape is highly regular and single
peaked, reflecting a single optimal solution possessing a higher fitness value than any
other potential solution. A more complex or “rugged” fitness landscape, by contrast, is
not amenable to linear programming models (e.g., lower unit costs through economies of
first, uncertainty about principles of valuation and, second, that this occurs at the borderlands (especially if
we understand this not as boundary but overlap). But there are two very important points of difference:
First, because entrepreneurship is not between institutions but between principles of evaluation, it can take
place within an institution. As my cases demonstrate, it can take place within a single organization. Second,
entrepreneurs are not proponents of a particular mode of valuing but are exploiting the uncertainty of
multiple modes of valuing.
46As I discuss in more detail in the next section, the term heterarchy was first used by neurologist Warren
McCulloch in 1945. Gunnar Hedlund introduced the term to the social sciences with application to the
multinational corporation. See Gunnar Hedlund, “The Hypermodern MNC: A Heterarchy,” 1986; and
Gunnar Hedlund and Dag Rolander, “Action in Heterarchies: New Approaches to Managing the MNC,”
1990.
47 David Lane and Robert Maxfield, “Strategy under Complexity: Fostering Generative Relationships,”
1996.
48 Stuart Kauffman, “Adaptation on Rugged Fitness Landscapes,” 1989.
30
scale), because the topography is jagged and irregular, with multiple peaks corresponding
to multiple optimal solutions.49
As an example of such complexities, think of the scrambling that is taking place among
firms that are producing in fields that were once previously separated into the relatively
discrete categories of computers, telecommunications, software, media, or banking.
When a major computer electronics company markets songs and videos (Apple) or when
major software companies (Microsoft and Google) collide with newspapers and broadcast
giants over the delivery of news and entertainment, we know that competition is not
taking place within the boundaries of Standard Industrial Classification (SIC) categories.
In retrospect, we might say that the problem was simple: the industries listed above are
all converging. That retrospective view would suffer from the typical problem of 20/20
hindsight, assuming that what we see now could have been anticipated by the actors
involved. But it would also be incorrect; whatever else is happening, the rearrangements
have not been a case of simple convergence, as the key multimedia artifacts continue to
morph. Most critically, we cannot assume that our retrospective view, with its promise of
stabilization, offers guidance for the future. It does not. Because just when we have
figured out the intersecting paths among the list of industries above, along will come new
developments in such fields as genetics, linguistics, biophysics, mapping, and even social
network analysis to add to the mix of new recombinations.
The situation in “old” manufacturing sectors is scarcely different. Not so long ago, firms
like General Motors (GM) were easily categorizable. Then, the major materials were
steel, rubber, and plastic; the major costs were equipment and labor; and these firms
made automobiles and other vehicles. Today, an automobile can be viewed as an
entertainment system that we travel in;50 various computer components, taken together,
account for the greatest share of the value of a car; financing contributes the greatest
share of profits; and pension plans and medical insurance for retired employees are
among the highest costs. GM, doubtless, makes automobiles. But it could well be seen as
being in the computer business, the finance business, the insurance business, or even the
entertainment business.
Thus, in an increasing number of areas, many firms literally do not know what products
they will be producing in the not so distant future. To cope with these uncertainties,
instead of concentrating their resources for strategic planning among a narrow set of
senior executives or delegating that function to a specialized department, heterarchical
firms embark on a radical decentralization in which virtually every unit becomes engaged
in innovation. That is, in place of specialized search routines in which some departments
49 On the use of genetic algorithms designed to explore initially unpromising paths and thereby avoid the
danger of “climbing to the nearest peak,” which might simply be the highest point in a valley surrounded
by yet higher peaks, see John Holland, “Complex Adaptive Systems,” 1992; and Kauffman, “Adaptation.”
50 John Urry, “The ‘System’ of Automobility,” 2004.
31
are dedicated to exploration while others are confined to exploiting existing knowledge,
the functions of exploration are generalized throughout the organization.51
These developments increase interdependencies between divisions, departments, and
work teams within the firm. But because of the greater complexity of these feedback
loops, coordination cannot be engineered, controlled, or managed hierarchically. The
results of interdependence are to increase the autonomy of work units from central
management.52 Yet, at the same time, more complex interdependence heightens the need
for fine-grained coordination across the increasingly autonomous units.
These pressures are magnified by dramatic changes in the sequencing of activities within
production relations. As product cycles shorten from years to months, the race to new
markets calls into question the strict sequencing of design and execution.53 Because of
strong first-mover advantages, in which the first actor to introduce a new product
(especially one that establishes a new industry standard) captures inordinate market share
by reaping increasing returns, firms that wait to begin production until design is
completed will be jeopardized in competition. Like the production of “B movies” in
which filming begins before the script is completed, successful strategies integrate
conception and execution, with significant aspects of the production process beginning
even before design is finalized.
Production relations are even more radically altered in the processes analyzed by Sabel
and Dorf as simultaneous engineering.54 Conventional design is sequential, with
subsystems that are presumed to be central designed in detail first, setting the boundary
conditions for the design of lower-ranking components. In simultaneous engineering, by
contrast, project teams develop all the subsystems concurrently. In such concurrent
design, the various project teams engage in an ongoing mutual monitoring, as innovations
produce multiple, sometimes competing proposals for improving the overall design.
Thus, increasingly rugged fitness landscapes yield increasingly complex
interdependencies that in turn yield increasingly complex coordination challenges. Where
search is no longer departmentalized but is instead generalized and distributed throughout
the organization, and where design is no longer compartmentalized but deliberated and
51 The search for new markets, for example, is no longer the sole province of the marketing department, if
units responsible for purchase and supply are also scouting the possibilities for qualitatively new inputs that
can open up new product lines.
52 Luc Boltanski and Ève Chiapello, The New Spirit of Capitalism, 2005.
53 The still-definitive statement on the transformation from the long production runs of mass production to
the customized production of flexible specialization is Michael Piore and Charles F. Sabel, The Second
Industrial Divide, 1984. Whereas mass production uses specialized tools to make standardized products
(think of the dedicated tools of the Fordist car assembly line, replaced each year to make a new line of
nearly identical automobiles), flexible specialization uses standardized tools to make specialized products.
54 Michael C. Dorf and Charles F. Sabel, “A Constitution of Democratic
Experimentalism,” 1998.
32
distributed throughout the production process, the solution to the nonhierarchically
distributed intelligence of heterarchical firms is distributed authority.55
Under circumstances of simultaneous engineering where the very parameters of a project
are subject to deliberation and change across units, authority is no longer delegated
vertically but instead emerges laterally. As one symptom of these changes, managers
socialized in an earlier regime frequently express their puzzlement to researchers:
"There’s one thing I can’t figure out. Who’s my boss?" Under conditions of distributed
authority, managers might still "report to" their superiors, but increasingly they are
accountable to other work teams. A young interactive designer whom we shall meet in
chapter 3 expressed this succinctly: When asked to whom he was accountable, he replied,
“I report to [the project manager] but I’m accountable to everybody who counts on me.”
Thus, corresponding to the patterns of knowledge and communication that are
recombined laterally rather than flowing vertically, authority in the heterarchical firm
takes the form of lateral accountability.
Organizing Dissonance
Mid-twentieth century, there existed a general consensus about the ideal attributes of the
modern organization: it had a clear chain of command, with strategy and decisions made
by the organizational leadership; instructions were disseminated and information
gathered up and down the hierarchical ladder of authority; design preceded execution,
with the latter carried out with the time-management precision of a Taylorist
organizational machine. This consensus was still strong thirty years later when economist
Oliver Williamson published an article in the American Journal of Sociology confidently
assuming that he could embrace all economic activity within only two logics of
coordination – “markets and hierarchies.”56 By the end of the century, the main precepts
of that ideal organizational model would be challenged. The primacy of relations of
hierarchical dependence within the firm and relations of market independence between
firms was giving way to relations of interdependence among networks of firms and
among units within the firm.57
Heterarchical forms do not take the boundaries of the firm and the boundaries of its
internal units as fixed parameters. As Walter Powell and others show, the boundaries of
the firm, especially those in fast-breaking sectors, are crisscrossed by dense ties of
interlocking ownership and complex patterns of strategic alliances.58 Where the
environment is most volatile and uncertain, the real unit of economic action is
increasingly not the isolated firm but networks of firms. Turning to network ties inside
the firm, Peter Dodds, Duncan Watts, and Charles Sabel show that top-down patterns of
55 Walter W. Powell, “Inter-organizational Collaboration in the Biotechnology Industry,” 1996.
56 Williamson, “The Economics of Organization.”
57 Kogut and Zander, “Knowledge of the Firm”; Gernot Grabher and David Stark, “Organizing Diversity:
Evolutionary Theory, Network Analysis, and the Postsocialist Transformations,” 1997; and Paul DiMaggio,
ed., The Twenty-First Century Firm: Changing Economic Organization in International Perspective, 2001.
58 Kogut, Shan, and Walker, “The Make-or-Cooperate Decision”; Powell, “Inter-organizational
Collaboration”; and Walter W. Powell, Douglas R. White, Kenneth W. Koput, and Jason Owen-Smith,
“Network Dynamics and Field Evolution,” 2005.
33
organizational communication perform much more poorly than decentralized networks
on tasks of distributed problem solving. In a simulation of network perturbation
(comparable to an attack or other serious disruption), they further demonstrate that
“multi-scale networks” – with enough pockets of cohesion and enough random ties
among them – have the robust connectivity required to recover rapidly and respond
effectively in episodes of crisis.59 Networks dissolve boundaries external and internal to
the firm.
At this point, the reader is likely wondering why I am proposing another term –
heterarchy – to label the emerging organizational form. If these forms exhibit distinctive
network properties, then why not label them as “network organizations”? Similarly, if the
emerging organizational forms are characterized primarily by their nonbureaucratic
features, then terms such as “nonhierarchical” or “postbureaucratic” would come more
readily to hand.
Within the triplicate of markets, hierarchies, and networks, the term network stands for an
alternative coordinating mechanism.60 This alone would be enough to account for the
path dependency exhibited by the field in continuing to deploy “network” as a term to
denote changes in organizational form. But, however fruitful in stimulating more than a
decade of research, the problem of labeling these forms as “networks” conflates the name
for an organizational form with an analytic approach. That is, as the literature also
abundantly demonstrates, not only the emergent network form but also markets and
hierarchies can be analyzed in network terms.
But there is an even more important reason for not adopting the “network” label. In
economic sociology and organizational studies, social network analysis typically refers to
patterns of ties among persons (or anthropomorphized entities such as firms). But actors
in and across organizations do more than communicate with, or link to, others. They also
evaluate performance, justify their actions, and offer reasons to explain why things
should be done this way instead of that.61 When they do so, they refer either explicitly or
(more often) implicitly to principles of evaluation. Organizations can be seen as patterns
of ties, but they should also be seen as sites in which actors engage in practices of
justifying worth. Network ties are mechanisms of coordination but always alongside
performance criteria and the evaluative principles on which they are based.
A similar logic holds for rejecting the “postbureaucratic” label. Organizations can be
analyzed as patterns of authority; but all relations of authority, whether vertical or lateral,
must rest on principles of accountability. And the more lateral the patterns of authority,
the more diverse the principles of accountability.
When authority is distributed along lines of lateral accountability, we need to study those
who make and keep accounts (and who, most emphatically, are not simply the
accountants). To analyze the processes of evaluation that are central to the problems of
59 Peter Sheridan Dodds, Duncan J. Watts, and Charles F. Sabel, “Information Exchange and the
Robustness of Organizational Networks,” 2003.
60 Walter W. Powell, “Neither Market Nor Hierarchy: Network Forms of Organization,” 1990.
61 Charles Tilly, Why? 2006.
34
worth in organizations, we must thus first explore the concept of accounts.
Etymologically rich, the term simultaneously connotes bookkeeping and narration. Both
dimensions entail evaluative judgments, and each implies the other: Accountants prepare
story lines according to established formulas, and in the accounts given by a good
storyteller we know what counts.
In organizations, as in everyday life, we are all bookkeepers and storytellers. We keep
accounts and we give accounts, and, most importantly, we can be called to account for
our actions. It is always within accounts that we "size up the situation," for not every
form of worth can be made to apply and not every asset is in a form mobilizable for a
given situation. We evaluate the situation by maneuvering to use scales that measure
some types of worth and not others, thereby acting to validate some accounts and
discredit others. How am I accountable? What counts? Who counts? Can you be counted
on? Will you credit my account? By which accounting?
Heterarchies flatten hierarchy. But they are not simply nonhierarchical. The new
organizational forms are heterarchical not only because they have flattened reporting
structures but also because they are the sites of heterogeneous systems of accounting for
worth. A robust, lateral collaboration flattens hierarchy while promoting diversity of
evaluative principles. Heterarchies are complex adaptive systems because they
interweave a multiplicity of performance principles. They are heterarchies of worth.
Distributed authority implies not only that units will be accountable to each other but also
that each will be held to accountings in multiple registers. The greater interdependence of
increasingly autonomous work teams results in a proliferation of sometimes competing
performance criteria. Heterarchies are organizations with multiple worldviews and belief
systems such that products, processes, and properties carry multiple “tags” or
interpretations.62 Because resources are not fixed in one system of interpretation but can
exist in several, heterarchies make assets of ambiguity.
Organizational ecologists have long held that adaptability is promoted by the diversity of
organizations within a population.63 I extend and, in the process, modify64 this notion by
considering the problem of diversity for adaptability within an organization. In making the
shift from the societal to the organizational level, analysis moves from the ecologists’
diversity of organizations to the heterarchical organization of diversity. The adaptive
potential of organizational diversity may be most fully realized when diverse evaluative
principles coexist in an active rivalry within the enterprise. By rivalry, I refer not to
competing camps and factions but to coexisting logics and frames of action. The
organization of diversity is an active and sustained engagement in which there is more than
one way to organize, label, interpret, and evaluate the same or similar activities. It increases
62 Andy Clark, “Leadership and Influence: The Manager as Coach, Nanny, and Artificial DNA,” 1999; and
John H. Clippinger, “Tags: The Power of Labels in Shaping Markets and Organizations,” 1999.
63 “A system with greater organizational diversity has a higher probability of having in hand some solution
that is satisfactory under changed environmental conditions.” Michael T. Hannan, “Uncertainty, Diversity,
and Organizational Change,” 1986, p. 85.
64 I elaborate my theoretical discussion of these issues in the section “From Diversity of Organizations to
the Organization of Diversity” in chapter 5.
35
the possibilities of long-term adaptability by better search because the complexity that it
promotes and the lack of simple coherence that it tolerates increase the diversity of options.
As it shifts from specific search routines to a situation in which search is generalized, the
heterarchical firm is redrawing internal boundaries, regrouping assets, and perpetually
reinventing itself. Under circumstances of rapid technological change and volatility of
products and markets, it seems there is no single best solution. If one solution could be
rationally chosen and resources devoted to it alone, the benefits of its fleeting superiority
would not compensate for the costs of subsequent missed opportunities. Because
managers hedge against these uncertainties, the outcomes are hybrid forms.65 Good
managers do not simply commit themselves to the array that keeps the most options open;
instead, they create an organizational space open to the perpetual redefinition of what
might constitute an option. Rather than a rational choice among a set of known options,
we find practical action fluidly redefining what the options might be. Management
becomes the art of facilitating organizations that can reorganize themselves.
The challenge of the modern firm is the challenge of building organizations that are
capable of generating new knowledge. Flexibility requires an ability to redefine and
recombine assets: in short, a pragmatic reflexivity. To do so, heterarchies maintain and
support an active rivalry of evaluative principles. Rivalry is not competition among units
based on the same principles of evaluation. Neither is it compartmentalization, in which
different principles of worth map to separate departments or units, bounded and buffered
from contamination. It is not a replicative redundancy or slack (more of the same) but a
generative redundancy of difference.
I write of organizing dissonance because some forms of friction can be destructive. When
personalized, differences can be petty as opposed to productive. To be constructive,
rivalry must be principled, with the adherents of the contending frameworks offering
reasoned justifications. Moreover, where multiple evaluative principles collide in
heterarchical forms, the danger is that arguments displace action and nothing is
accomplished. Success requires attention to the structure of temporal processes. I refer to
a collective sense of rhythm and timing – of when to make temporary settlements to get
the job done, with the knowledge that this is not a once-and-for-all resolution of the
disagreements – as a discursive pragmatism. Heterarchy is neither harmony nor
cacophony but an organized dissonance.
Dissonance occurs when diverse, even antagonistic, performance principles overlap. The
manifest, or proximate, result of this rivalry is a noisy clash, as the proponents of
different conceptions of value contend with each other. The latent consequence of this
dissonance is that the diversity of value-frames generates new combinations of the firm’s
resources. Because there is not one best way or single metric but several mutually
coevolving yet not converging paths, the organization is systematically unable to take its
routines or its knowledge for granted. It is the friction at the interacting overlap of
multiple performance criteria that generates productive recombinations by sustaining a
65 Sabel, “Moebius-Strip Organizations”; and Charles F. Sabel and Jonathan Zeitlin, “Stories, Strategies,
Structures: Rethinking Historical Alternatives to Mass Production,” 1997.
36
pragmatic organizational reflexivity. Heterarchies create wealth by inviting more than
one way of evaluating worth.
A Metaphor for Organization in the Twenty-First Century
From where have we found metaphors for organization? The manufacture of pins served
as Adam Smith’s metaphor for the division of labor. Clocks have been ever popular; for
example, interrupted watchmakers served as Herbert Simon’s parable for the “nearly
decomposable” features of hierarchical systems.66 Where organizations, whether as
national economies or firms, were conceived as systems of planning, linear programming
served both as method and as metaphor. More recently, organizational ecology has
looked to biological systems for metaphors of evolution, selection, population, births, and
deaths.
But the dominant and long-lasting metaphor for organization, remarkably consistent over
fifteen centuries, comes from religion. The term hierarchy was originally coined by
Dionysius the Areopagite, a fifth-century medieval theologian, in two treatises on the
celestial and ecclesiastical hierarchies, respectively.67 In his Celestial Hierarchy we find
all the elements of the metaphor fully elaborated: nine distinct levels organized in three
tiers corresponding to senior executives, middle management, and lower-level
functionaries, with the angels (closest to humans) at the bottom and the seraphim (closest
to God) at the top. Each level supervises the level below and reports to the level above;
beings can advance through promotion up the ordering; information cannot bypass the
chain of command; and the structure is based on a strict hierarchy of knowledge, with the
literally all-knowing boss at the top.
The term heterarchy is not of such heavenly provenance. It was coined at the beginning
of the computer age, in 1945, by neurologist Warren McCulloch in an article published in
the (appropriately hybrid) Bulletin of Mathematical Biophysics. McCulloch titled his
elegant, five-page paper “A Heterarchy of Values Determined by the Topology of
Nervous Nets.”68 In place of Dionysius the Areopagite’s nine levels, McCulloch
simulates a network of six neurons. Several years earlier, with Walter Pitts, McCulloch
showed how to formalize the brain as a network of neurons viewed as logical processing
elements.69 In the “Heterarchy of Values” paper, he is simulating choice.
66 Herbert Simon, The Sciences of the Artificial, 1969.
67 See especially Gunnar Hedlund, “Assumptions of Hierarchy and Heterarchy, with Applications to the
Management of the Multinational Corporation,” 1993.
68 Warren S. McCulloch, “A Heterarchy of Values Determined by the Topology of Nervous Nets,” [1945]
1965.
69 Warren S. McCulloch and Walter H. Pitts, “A Logical Calculus of the Ideas Immanent in Nervous
Activity,” 1943. This work was critical in the definition of the classical computer architecture based on
stored programs devised by John von Neumann. It also laid the basis for the new field of “automata
theory.” Another collaboration (Pitts and McCulloch, “How We Know Universals: The Perception of
Auditory and Visual Forms,” 1947) was a pioneering paper on neural networks for pattern recognition
showing how visual input could control motor output through the distributed activity of a neural network
without the intervention of executive control. McCulloch and Pitts later collaborated with Lettvin and
Maturana on one of the classic papers on single-cell neurophysiology (“What the Frog’s Eye Tells the
37
In his simulation, McCulloch first maps the neuron circuits on a plane with no diallels, or
“crossovers.” He observes that the resulting structure is a hierarchy: “The order is such
that there is some end preferred to all others, and another such that all are preferred to it,
and that of any three if a first is preferred to a second and a second to a third, then the
first is preferred to the third” (p. 43).
McCulloch explicitly notes the similarity of such a hierarchical system to “the sacerdotal
structure of the church” and implicates the notion of a transitivity of values with “the
notion of the sacred or holy.” He points out that “to assert a hierarchy of values is to
assert that values are magnitudes of some one kind. Summarily, if values were
magnitudes of any one kind, the irreducible nervous net would map (without diallels) on
a plane” (p. 43).
Aware that extant theories of value assume that values can be treated as magnitudes of
some one kind, McCulloch argues to the contrary, stating explicitly that “for values there
can be no common scale.” The next step elegantly anticipates Kenneth Arrow’s
Impossibility Theorem on the intransitivity of preference orderings:
Consider the case of three choices, A or B, B or C, and A or C in which A is preferred to
B, B to C, and C to A. (p. 43)
To simulate intransitivity as the more realist problem in modeling choice, McCulloch
presents two solutions: introduce a diallel, a crossover, in the network (if represented on a
plane) or shift to the more complex topology of a torus. Either solution is
nonhierarchical:
An organism possessed of this nervous system – six neurons – is sufficiently
endowed to be unpredictable from any theory founded on a scale of values. It has
a heterarchy of values, and is thus interconnectively too rich to submit to a
summum bonum. (p. 44)
McCulloch’s highly original work led to the development of artificial networks as a new
computing technology, which, in turn, fed back to the computational modeling of the
brain.70 His idea of redundant network ties was important for the conception of reliable
organization built from unreliable parts, laid the basis for the new field of “automata
Frog’s Brain,” 1959). For an overview, see Michael A. Arbib, “Warren McCulloch’s Search for the Logic
of the Nervous System,” 2000.
70 After publishing “A Heterarchy of Values,” McCulloch chaired a series of ten meetings set up by the
Macy Foundation to explore what biologists could teach computer scientists about signal processing,
computation, and communication. The group involved biologists, technologists, and social scientists
including John von Neumann, Norbert Wiener, Gregory Bateson, and Paul Lazarsfeld. Its inaugural
meeting in New York, March 1946, was titled “Feedback Mechanisms and Circular Causal Systems in
Biological and Social Systems.” For summaries of the conferences and lists of participants see
www.asccybernetics.org/foundations/history/MacySummary.htm. For a lively discussion of the Macy
Conferences, see Jean-Pierre Dupuy, The Mechanization of the Mind: On the Origins of Cognitive Science,
2000.
38
theory,” and contributed to the fertile concept of “self-organization.”71 “A Heterarchy of
Values” is cited as an inspiration for non-Turing, or non-Euclidean, computing, most
recently in efforts to develop biology-based computing.
As one of the first efforts at network analysis – developed at the intersection of
neurology, computer science, mathematics, biophysics, and linguistics72 – McCulloch’s
pathbreaking paper is an appropriate source for a new metaphor for organization in the
twenty-first century. Metaphors matter. The field of organization studies will be enriched
if we adopt a concept that has applicability to the problem of “organization” inclusive of,
but also of wider generality than, the study of formal collectivities of human agents.
Biologists, for example, have recently rediscovered the problem of organization (of
which “the organism” is only the most apparent instance); levels of organization extend
down to the cellular, even molecular, level, and outward to speciation and processes of
coevolution.73 Life is organization. Similarly, to speak of information or knowledge is to
speak of organization. Work by colleagues in information science and the study of
cognition and learning74 suggests that hierarchy is not the only form of organization in
these fields.
Most revealing are changes in our conception of code. Formerly, the term evoked
procedures of codification in which elements were organized into a system of
encompassing and mutually exclusive categories. With language as the exemplar of
nonhierarchical structuring, code is now grasped in network terms. Researchers in genetic
code, for example, see two structural properties as critical to the evolution of evolvability.
The first is modularity, whereby elements retain their structure even as they are
recombined with other modules in higher levels of organization. The second, no less
important, is pleiotropy, whereby a sequence of genetic code is expressed in more than
one subsystem.75 In network terms, genetic code is tangled code. The term comes from
computer science, referring to the bane of the programmer dealing with crisscrossing
pieces of software. But where tangled code was to be avoided at all costs, work at the
forefront of software engineering, for example, in the qualitative shift from object-
oriented to aspect-oriented programming, is developing heterarchical software code in a
field that was once seen as quintessentially hierarchical.
71 John von Neumann, “Probabilistic Logics and the Synthesis of Reliable Organizations from Unreliable
Components,” 1956; and Warren S. McCulloch, “The Reliability of Biological Systems,” 1960.
72 McCulloch was involved in the design of a (graphical) triadic logic and was very interested in Charles
Sanders Peirce’s experiments with a triadic logic (see Arbib, “McCulloch’s Search”). Peirce, regarded as
the founder of philosophic pragmatism, argued that all cognition is irreducibly triadic. His triadic theory of
signs as icon, index, and symbol was a major contribution to modern linguistics.
73 See especially Walter Fontana and Leo Buss, “‘The Arrival of the Fittest’: Toward a Theory of
Biological Organization,” 1994; and Walter Fontana and Leo Buss, “The Barrier of Objects: From
Dynamical Systems to Bounded Organizations,” 1996.
74 Geoffrey Bowker and Susan Leigh Star, “Knowledge and Infrastructure in International Information
Management: Problems of Classification and Coding,” 1994; and Luis M. Rocha, “Adaptive Webs for
Heterarchies with Diverse Communities of Users,” 2001.
75 Gunter P. Wagner and Lee Altenberg, “Complex Adaptations and the Evolution of Evolvability,” 1996;
and Thomas F. Hansen, “Is Modularity Necessary for Evolvability? Remarks on the Relationship between
Pleiotropy and Evolvability,” 2003.
39
As a more general process, then, heterarchy refers to an organizational structure in which
a given element – a statement, a deal, an identity, an organizational building block, a
sequence of genetic code, a sequence of computer code, a sequence of legal code – is
simultaneously expressed in multiple crosscutting networks. “A program which has a
structure in which there is no single ‘highest level,’ or ‘monitor,’ is called a
heterarchy.”76
Thus, as a metaphor for organization in the twenty-first century, heterarchy has its
provenance at the intersection of extraordinarily generative sciences. It also has potential
for applicability across a wide set of domains including computer science, biology, and
informatics as well as organizational analysis in the social sciences. It does have one
drawback: it does not immediately trip off the tongue on first vocalization. But the terms
“bureaucrat” and “bureaucracy” – as amalgams of bureau and aristocrat/aristocracy – also
seemed peculiar when introduced to account for a new role and a new phenomenon.
Despite this drawback, heterarchy has a distinct advantage because, as a member of a
family of terms such as monarchy, anarchy, polyarchy, and hierarchy, the term
immediately denotes a form of governance. Indeed, perhaps the first exemplar of
heterarchical social organization was the U.S. Constitution, with its three branches of
government, each based on a distinctive principle of legitimation, none of which is
overarchingly superior.77 As a form of governance, heterarchy organizes dissonance. But
it is not a panacea. Just as the metaphor of heterarchy is not of heavenly provenance, so
the problems that the implementation of heterarchy creates are all too human.
WORTH IN CONTENTIOUS SITUATIONS
I follow John Dewey’s insights on problems of inquiry, worth, and uncertainty, I also
look to him for guidance on issues of methodology. In his Theory of Valuation, Dewey
insists repeatedly on the need to study processes of “actual valuation.” His remarks from
1939 remain on target today:
[T]he notion that an adequate theory of human behavior – including particularly
the phenomena of desire and purpose – can be formed by considering individuals
apart from the cultural setting in which they live, move, and have their being – a
theory which may justly be called metaphysical individualism – has united with
the metaphysical belief in a mentalist realm to keep valuation-phenomena in
subjection to unexamined traditions, conventions, and institutionalized customs.78
The case studies presented here adopt Dewey’s guidelines on both counts. First,
following Dewey’s injunction to study actual valuations in “cultural settings,” I further
specify the notion of setting, using ethnographic methods to study three very different
kinds of workplaces. I study situated cognition in situ. In each case the ethnographic site
is a single room – a factory workshop with about 100 manual workers, a former printing
76 Douglas R. Hoftstadter, Godel, Escher, Bach, 1979.
77 See László Bruszt, “Market Making as State Making: Constitutions and Economic Development in
Postcommunist Eastern Europe,” 2002; and Martin Landau, “Redundancy, Rationality, and the Problem of
Duplication and Overlap,” 1969.
78 Dewey, Theory of Valuation, p. 64.
40
loft converted to an open-plan layout housing about 80 new-media employees in
Manhattan’s Silicon Alley, and the Wall Street trading room of a major international
investment bank, similarly open plan, with about 160 traders.
Second, I follow Dewey’s advice that practices of evaluation should not taken as
“unexamined traditions, conventions, and institutionalized customs.” Methodologically,
the move is not simply to employ ethnography in specific settings but to shift from the
analysis of institutions to the study of indeterminate situations.79 As we shall see in the
following chapters, unsettling situations are special moments in which the researcher
discovers what is at stake because it is in such situations that the actors themselves
become cognizant of what had previously been taken-for-granted. By studying cases
involving the heterarchical rivalry of evaluative principles, we see that traditions,
conventions, and institutionalized customs are not left unexamined. Indeed, they are
opened up to reflective cognition by the actors themselves.
Because I examine situations in three distinctively different settings, the analytic lens for
studying worth – evaluative practices – changes focus as we move from case to case.
Correspondingly, the forms of indeterminate situations and the distinctive challenges of
recognition are specified as analytically appropriate for each case.
In the case of the Hungarian factory, we meet 18 highly skilled workers, operating
machine tools to build machine tools, who recognize an opportunity to win recognition of
their self-proclaimed worth. The cultural setting is state-socialist Hungary with its central
planning under one-party rule. More specifically, it is the exciting period of the mid- to
late 1980s, after the upheavals of the rise and later suppression of Solidarity in Poland but
before the collapse of communism in 1989. Yet more specifically, the machine shop of
about 100 workers is part of Minotaur,80 one of the largest state-owned enterprises in
Hungary, with more than 11,000 employees. The initial situations arise once Minotaur
recognizes the legal right for its employees to form “partnerships,” using the factory’s
equipment on the “off-hours,” during which the members of the partnership are free to
organize work on their own terms. If the routines of the shop floor had ever been “taken-
for-granted,” they certainly could be no longer. The parent company Minotaur exploits
the partnership form as a way to earn hard-currency revenues; meanwhile the members of
the partnership itself capitalize on the new form as a chance to demonstrate their worth.
But their success creates new situations in which the toolmakers, however unified in their
agreement that skill is the ultimate principle of value, face a series of perplexing
challenges about how to measure its performance. In the process, they come to recognize
new criteria of worth and new identities bound up with them. Later, with the collapse of
communism after 1989 and the privatization of their factory, they confront new situations
that challenge their worth, provoking them to articulate again their sense of justice.
79 On the rejection of both methodological collectivism and individualism in favor of “methodological
situationalism,” see Karin Knorr-Cetina, “Introduction: The Micro-sociological Challenge of Macro-
sociology,” 1981.
80 All names of firms and individuals are pseudonymous.
41
The new-media employees in my second ethnography are also, in their own way,
toolmakers building something – not operators of drills and lathes for cutting and boring
costly metals but software programmers and interactive designers using new-media tools
to build sophisticated online retail websites. The cultural setting is Manhattan, following
the recession of 1993 that lowered rents and left programmers as well as artists and
copywriters looking for work. More specifically, it is Silicon Alley at the end of the
1990s after the initial public offerings of Netscape and theGlobe.com but before the dot-
com boom went bust. Yet more specifically, it is in NetKnowHow, a start-up company
that grew from 15 to 150 employees during the several years we studied it. Here the
relevant situations are in projects where business strategists, interactive designers,
programmers, information architects, and merchandising specialists bring distinctive
disciplinary identities. Projects are sites of contention, not primarily about the worth of
the respective specialists but about the best criteria by which to evaluate the worth of the
websites they are building. It is this rivalry of evaluative principles that allows the firm to
never take its knowledge for granted. The collision of performance criteria yields a
distributed cognition capable of the kind of search in which you don’t know what you’re
looking for but will recognize it when you find it.
The arbitrage traders in the third ethnography would seem to be anything but toolmakers.
But, as we shall see, each trader skillfully customizes his tools of the trade. The setting is
Wall Street investment banking. More specifically, it is exactly at the turn of the century
in the period after the emergence of quantitative finance but before the Enron scandal.
Yet more specifically, it is the hedge fund of a firm we call International Securities, a
major international investment bank whose traders are engaged in sophisticated arbitrage.
Like the Hungarian toolmakers (uniformly highly skilled workers) and the new-media
workers (almost uniformly young and culturally hip), the traders are culturally
homogeneous. Even more than the Hungarian toolmakers, they share a common
definition of how to measure the worth of a trader, in this case by “the value of his book”
(the profitability – computed yearly, monthly, daily, hourly, literally minute by minute –
of a given trader’s deals). But this marked homogeneity belies a generative diversity, for
although the traders share a metric for evaluating one another, they differ on the most
salient dimension of their work: how to measure value in the games of arbitrage. As to
situations, it might seem at first glance that a trading room is a site for responding to
situations “out there” in the markets. But this is the nightly news version of markets with
stories of crises, surges, and swings. The actual problem for these arbitrage traders is less
how to respond to situations “out there” than how to recognize situations that their
competitors have not seen. As we shall see, the trading room is organized as a cognitive
ecology in which commitments to distinctive principles of arbitrage combine with
interactions across these principles to produce a situated cognition that not only
recognizes already-known kinds of opportunities but also re-cognizes situations as
opportunities. In the epilogue to this chapter, I examine how the traders responded to a
crisis situation, potentially a crisis of their identities, after their trading room was
destroyed in the September 11 attack on the World Trade Center.
