Dokumen - Pub Niklas Luhmann and Organization Studies 9788763003049 876300304x

SERIES EDITORS:
Stewart R. Clegg &

Ralph Stablein
Edited by
David Seidl and
A D VA N C E S I N O R G A N I Z AT I O N S T U D I E S
Kai Helge Becker
Niklas Luhmann and

Organization Studies
Liber & Copenhagen Business School Press

David Seidl and Kai Helge Becker (eds.)
Niklas Luhmann and Organization Studies
© CBS Press and the authors 2006
1st printed edition 2006

The e-book is published in 2013
Series editors: Stewart R. Clegg and Ralph Stablein

E-book production: PHi Business Solutions Ltd. (Chandigarh, India)
ISBN (e-book edition): 978-87-630-0304-9

ISBN (printed edition): 978-87-630-0162-5
CBS Press
Rosenoerns Allé 9
DK-1970 Frederiksberg C
Denmark
slforlagene@samfundslitteratur.dk
www.cbspress.dk
All rights reserved. No part of this publication may be reproduced or used in any
form or by any means – graphic, electronic or mechanical including photocopying,
recording, taping or information storage or retrieval system – without permission
in writing from the publisher.
01 Luhmann 05-11-27 21.55 Sida 3
Advances in Organization Studies

Series Editors:
Stewart Clegg
Professor, University of Technology, Sidney
Ralph E. Stablein
Professor, University of Otago, New Zealand
Advances in Organization Studies is a channel for cutting edge theoretical

and empirical works of high quality, that contributes to the field of organiza-
tional studies. The series welcomes thought-provoking ideas, new perspec-
tives and neglected topics from researchers within a wide range of disciplines
and geographical locations.
01 Luhmann 05-11-16 17.19 Sida 4
01 Luhmann 05-11-16 17.19 Sida 5
Table of Contents
Acknowledgements......................................................................... 7
Introduction: Niklas Luhmann and Organization Studies
David Seidl and Kai Helge Becker .................................................... 8
PART I: THE THEORY OF AUTOPOIETIC SOCIAL SYSTEMS

1. The Basic Concepts of Luhmann’s Theory of Social Systems
David Seidl ................................................................................. 21
2. The Concept of Autopoiesis
Niklas Luhmann ......................................................................... 54
3. The Autopoiesis of Social Systems
Niklas Luhmann ......................................................................... 64
PART II: ORGANIZATION, DECISION AND PARADOX

4. The Paradox of Decision Making
Niklas Luhmann ......................................................................... 85
5. Displacing the Paradox of Decision Making
Morten Knudsen .......................................................................... 107
6. On Gorgon Sisters: Organizational Action in the Face of
Paradox
Barbara Czarniawska.................................................................. 127
PART III: ORGANIZATION, INTERACTION AND SOCIETY

7. Organization and Interaction
David Seidl ................................................................................. 145
8. Organization and Society
Thomas Drepper ......................................................................... 171
9. The Design of Organization in Society
Dirk Baecker ............................................................................... 191
PART IV: LUHMANN’S THEORY IN THE CONTEXT

OF OTHER THEORIES
10. Luhmann’s Systems Theory and Theories of Social Practices
Kai Helge Becker ......................................................................... 215
5
01 Luhmann 05-11-16 17.19 Sida 6
11. Luhmann’s Systems Theory and the New Institutionalism

Raimund Hasse ........................................................................... 248
12. Luhmann’s Systems Theory and Postmodernism
Jochen Koch ................................................................................. 262
13. Luhmann’s Systems Theory and Network Theory
Michael Bommes and Veronika Tacke ....................................... 282
PART V: FORMS OF ORGANIZATION

14. Analysing Forms of Organization and Management:
Stock Companies vs. Family Businesses
Fritz B. Simon ............................................................................. 307
15. On Defining the Multinational Corporation. A Systems-
Theoretical Perspective
Darnell Hilliard .......................................................................... 324
PART VI: IMPLICATIONS FOR MANAGEMENT AND

CONSULTING
16. Communication Barriers in Management Consulting
Niklas Luhmann ......................................................................... 351
17. Strategic Management from a Systems-Theoretical
Perspective
Jan-Peter Vos ............................................................................... 365
18. Management Accounting from a Systems-Theoretical
Perspective
Tobias Scheytt ............................................................................. 386
PART VII: GLOSSARY AND BIBLIOGRAPHY

Glossary to Niklas Luhmann’s Terminology
David Seidl ....................................................................................... 405
Annotated Bibliography of Selected Works by Niklas Luhmann
Kai Helge Becker ............................................................................... 411
Contributors ..................................................................................... 423
References ........................................................................................ 424
Index ................................................................................................. 460
6
02 Acknowledgements 05-11-15 13.26 Sida 7
Acknowledgements
The idea for this book goes back to an international symposium on “Niklas
Luhmann and Organization Theory”, which was organised by the editors
and took place at the University of Munich in June 2002. Eight chapters of
the present volume are in some way or other based on presentations given
at the conference, whilst others have been written or translated particularly
for this book. Both the symposium and the present book would not have
been possible without the kind support of various organizations and people
who contributed time, money and other resources.
In particular, we would like to thank the University of Munich, Hamburg
University, Witten/Herdecke University, and the European Group for
Organizational Studies (EGOS) for supporting the symposium. Moreover,
we record our thanks to the Munich School of Management, the Münchner
Universitätsgesellschaft, the Verein zur Förderung der Führungslehre an der
Universität München e.V. and the Department of Economics and Business
Administration at Hamburg University, who provided the financial means
for the symposium and this book project. We are particularly grateful to Evi
Groher, Birgit Pemler, Claudia Lusch and Doris Eikhof for their invaluable
support in organising the symposium. We acknowledge the helpful critical
comments by Nils Brunsson, Alfred Kieser, Dirk Baecker, Werner Kirsch and
Günther Ortmann on the concept of this book and earlier versions of the
manuscript. To Steward Clegg, editor of the series Advances in Organization
Studies, we owe thanks for his confidence, good advice and patience. We are
grateful to Artemis Gause for her careful and thorough way of doing the
language editing. We thank Andreas Kuhn, Moritz Putzer and Andreas von
Ritter-Zahony for their patience and dilligence in checking the final manu-
script. Finally, we are thankful to our contributors for placing confidence in
the project and their positive spirit of cooperation in face of the numerous
and often substantial revisions that we suggested.
Munich and London, Autumn 2005
David Seidl and Kai Helge Becker
7
03 Introduction 05-11-11 16.15 Sida 8
Introduction: Luhmann’s
Organization Theory
David Seidl and Kai Helge Becker
Niklas Luhmann (1927–1998) was without doubt one of the most interest-
ing social thinkers of the twentieth century. Not only in German-speaking
countries, but, meanwhile, also among researchers in the English-speaking
academic world, Luhmann’s work is considered of equal rank and standard
to that by such prominent social theorists as Bourdieu, Giddens, Habermas
and Foucault. Drawing on very diverse strands of thinking within sociology
and philosophy, and combining them with research in cybernetics, chaos
theory, and biology, Niklas Luhmann developed a very distinctive and chal-
lenging new way of theorising about the social, which has stimulated
research in various academic fields such as media studies, the political sci-
ences, theology, philosophy, literature, pedagogics, sociology, and particu-
larly in organization studies.
Until very recently, most of this research was conducted almost exclusive-
ly in German-speaking countries while hardly any efforts were made any-
where else. In the last few years, however, there has been a growing interest
in Luhmann’s ideas among European organization-theorists: his works are
increasingly referenced also in international journals of organization studies.
Yet the gap between the various levels of research is still enormous. In view
of that, efforts to introduce Luhmann’s approach and the existing research
to the international community of organization scholars seem long overdue.
As a first step towards this objective, this book will explain the basic con-
cepts of Luhmann’s theory and will demonstrate its potential for studying
organizations.
Some important aspects of Luhmann’s theory

In view of the innovative character of Lumann’s work and the variety of new
concepts that he developed, it is simply not possible to provide a meaning-
ful summary of Luhmann’s achievements in this introduction. In fact, this
entire volume is intended to demonstrate the exuberance of ideas that can
be found in Luhmann’s writings and the novel perspectives they have to
offer, even on well-known phenomena. Nevertheless, we want to highlight
here at least some general aspects of his systems theory that we consider to
be of particular importance.
8
Probably the most striking among Luhmann’s conceptual achievements

– which is also the most controversial aspect of his work – is the develop-
ment of a theory that clearly differentiates between social and psychic
phenomena. The social and the psychic domains are conceptualised as two
clearly demarcated systems with no overlap of any kind: in other words,
“social systems” and “psychic systems” constitute “environments” for each
other. While at first it may seem completely counterintuitive, this conceptu-
alisation allows us to analyse the social – and the logic of the social (hence
also organizations) – in its own right, thereby shedding light on the genuine
dynamics that social phenomena possess, independently of individual psy-
chic processes. This does not mean, however – and this is where the contro-
versy started – that the human being (the psychic system) is considered irrel-
evant to the social system; on the contrary, as Luhmann stressed over and
over again, it only underlines the relevance of the two systems for each other.
A second aspect that deserves to be mentioned is Luhmann’s specific kind
of constructivist epistemology, which is closely linked with his concept of
social and psychic systems. In particular, it focusses on the schemes of mean-
ing, or “distinctions”, that social (and psychic) systems employ to make
sense of themselves and their surroundings. When it comes to analysing the
form and functioning of organizations and the processes of decision making,
this approach leads to a range of interesting new insights into the way in
which organizations construct themselves, establish their boundaries, and
observe their environments.
A third point that merits attention is that, on the basis of his general
theory of social systems, Luhmann has developed an extensive theory of
modern society. In short, his approach accounts for the different kinds of
rationality, or rather internal logics, that underlie the different domains of
modern society, such as the economy, politics, religion, science, or art. This
perspective provides a framework for analysing the various kinds of rela-
tionships between organizations and their societal environment as well as
between different parts of modern society and their specific types of organ-
ization.
Fourth, in conjunction with the aspects mentioned above, Luhmann has
also managed to develop his own innovative concepts to go beyond many of
the traditional dichotomies that have troubled social theorists for so long.
On the basis of his unique way of theorising social phenomena, Luhmann
provides new insights into theoretical issues such as micro/macro, struc-
ture/process, structure/action, continuity/change, and consensus/conflict.
A fifth aspect must be considered of particular significance in the context
of the present volume: Luhmann was not only a sociologist who has devel-
oped a unique perspective on social phenomena in general, and modern soci-
ety in particular, but, in contrast to other social theorists, he has also con-
tributed directly to organization studies. During the past decades, the classi-
9
cal notion of organization as static and hierarchical entities that coordinate

labour in a rational way has increasingly been seen as inadequate for
accounting for the variety of organizational phenomena. Building on his
general theory of social systems Luhmann has developed his own perspec-
tive on organizations that offers an original, inspiring and elaborated alter-
native to approaches that are meanwhile considered unsatisfactory.
A final achievement of the theory that we would like to stress here is its
self-referentiality. Since the theory aims at including all aspects of the social,
it consequently has to include also itself. In other words, the theory does not
claim the position of an extramundane observer, but treats itself also as a
social phenomenon in need of an explanation. Accordingly, the concepts of
Luhmann’s theory are such as to allow us to analyse – and relativise – the
very theory itself as a social phenomenon that takes place in the social
domain.
The reception of Luhmann’s work on

an international level
In view of all that, and given the fact that Luhmann’s theory has been a clas-
sical topos of even undergraduate courses in sociology within German-
speaking countries for more than a decade, it appears very surprising that
his works have received comparatively little serious attention within the
international field of organization studies. In addition to the language bar-
rier and the time necessary for translating his works into English, the main
reasons for the rather hesitant reception so far probably lie in the theory
itself. One reason is certainly the complexity of Luhmann’s works and the
enormous amount of topics and theoretical traditions they cover, which
make it very difficult for first-time readers to access his works unaided by
commentaries. Moreover, Luhmann developed a very distinctive terminolo-
gy to express his concepts, which presents an additional hurdle. Because of
that, it is often said that when starting to read Luhmann it takes a hundred
to two hundred pages before one actually understands anything. This is
quite a big investment in time and effort, considering that one can never
really know beforehand what one will get out of it.
As a first step towards removing the reservations towards Luhmann’s
works, this book aims at introducing Luhmann’s way of theorising and at
demonstrating its potential for organization and management research. For
this purpose the book provides, on the one hand, a detailed, step-by-step
introduction to Luhmann’s theory of social systems in general and his orga-
nization theory in particular, in which his central concepts and ideas will be
carefully outlined. On the other hand, it demonstrates how this theory might
be fruitfully applied to different areas and questions of organization and
10
management studies. In this way the potential of Luhmann’s theory for pro-
viding new perspectives on issues of organization and management and for
generating a rich variety of challenging new insights in these areas will hope-
fully be revealed.
The structure of Luhmann’s œuvre

Luhmann’s œuvre, which comprises more than seventy books and several
hundred articles, can be divided into two parts: the early Luhmann before,
and the late Luhmann after the so-called “autopoietic turn”, which took
place in the early 1980s. His early writings can be characterized as an adap-
tation and further development of the works of Talcott Parsons, whom he
met personally during his stay at Harvard University in 1960–1961. By con-
trast, in his late writings Luhmann developed his own “grand theory” with
its characteristic and often very bold way of theorising. Although he had
already been a widely recognized sociologist in German-speaking countries
since the beginning of the 1970s, Luhmann himself referred to his early
works as merely a “series of nils in [his] theory production” (Luhmann
1987a, p. 142). In line with that view, this book will mainly focus on the
later stage of his work.
The starting point of Luhmann’s late theory can be seen in his concept of
the autopoietic system. An autopoietic system is a system that reproduces its
own elements on the basis of its own elements. Luhmann took this original-
ly biological concept, modified it, and applied it to the social sciences. For
him the social domain has to be conceptualised as consisting of autopoietic
social systems, which reproduce themselves self-referentially on the basis of
their own specific logic. This concept allowed Luhmann to draw on the
existing body of social theory and to reorganise it in such a way as to gain
completely new insights into the functioning of the social. Luhmann wrote
about the concept of autopoiesis: “Autopoiesis, as a concept, has no empir-
ical explanatory value. Its potential lies rather in the fact that it forces other
concepts into adaptation” (Luhmann 2000c, p. 49; our translation). This
can be clearly seen in Luhmann’s organization theory. While Luhmann drew
heavily on the classical theories by Herbert Simon, James March and the
like, he produced a very innovative new theory, whose implications were at
times diametrically opposed to the conclusions of his predecessors.
With regard to its subject, Luhmann’s late oeuvre can be divided into two
parts: on the one hand, there is his theory of social systems in general, where
Luhmann developed his unique perspective on social phenomena. In those
works he elaborated the general notion of social systems as self-reproducing
systems consisting of communications. Here, his descriptions and
explanations deal with a variety of social phenomena on a rather abstract
11
level and are unspecific with regard to the different types of social systems.
The central book in this area is Social Systems (1995f), which was original-
ly published in German in 1984.
On the other hand, there are his theories that focus on the different types
of social systems. In these works the general theory of social systems is speci-
fied with regard to the different types of system: society, interaction and
organization. Of those three, the societal system occupies by far the greatest
part. Not only did Luhmann write about society as an autopoietic social sys-
tem – the main book here being his Gesellschaft der Gesellschaft (1997a)
– but he also wrote separate books on the different sub-systems of society,
which were themselves conceptualised as autopoietic systems. There are, for
example, works on the system of economy (Die Wirtschaft der Gesellschaft
1988a), on the system of science and humanities (Die Wissenschaft der
Gesellschaft 1990c), on the system of art (Art as a Social System 2000d), on
the political system (Die Politik der Gesellschaft 2000b), on the system of
religion (Die Religion der Gesellschaft 2000d), on the system of education
(Das Erziehungssystem der Gesellschaft 2002a), and on the legal system
(Law as a Social System 2004).
In contrast to the societal system, the interactional system, as the second
type of social system, received comparatively little attention. Luhmann
wrote merely a couple of articles on them. The most important book on
interaction systems, Kommunikation unter Anwesenden (1999), was even-
tually written by André Kieserling as a PhD thesis under Luhmann’s super-
vision.
The organization, as the third type of social system, occupies a special
place in Luhmann’s oeuvre: in his earlier career, Luhmann had worked in
public administration for several years and much of his early theory was
based on his own experiences there. In fact, he began his life as a sociologist
by publishing eight books on public administration and organization
between 1963 and 1969. Two of these, Funktionen und Folgen formaler
Organisation (“Functions and consequences of formal organization”), pub-
lished in 1964, and Zweckbegriff und Systemrationalität (“The concept of
ends and system rationality”), published in 1968, are considered milestones
in organization studies, as they anticipated many important issues that have
only recently started to receive appropriate attention. From the groundwork
of those early publications, Luhmann gradually shifted the focus of his
research to the project of his “grand theory” – however , without ever com-
pletely abandoning his interest in organization. It might be justified to say
that Luhmann’s general sociological approach is strongly influenced by his
detailed knowledge of organization on both the theoretical and the practical
level. Towards the end of his life Luhmann eventually revisited and rewrote
his former publications on organization theory, now on the basis of his
theory of autopoiesis, thereby integrating the former into the framework of
12
his grand theory. The result of this was published posthumously in 2000
under the title Organisation und Entscheidung (“Organization and deci-
sion”).
The contents of this book

As the content of the first chapter is taken for granted in the later parts of the
book, readers who are not familiar with Luhmann’s theory are strongly
advised to read that first before going to other parts.
Part I of this book is dedicated to Luhmann’s theory of social systems in

general. The first chapter, “The Basic Concepts of Luhmann’s Theory of
Social Systems” (David Seidl), provides a general introduction to and
overview of Luhmann’s theory. Here, his central concepts are introduced
and carefully explained. The chapter starts off with an account of the origi-
nally biological concept of autopoiesis and Luhmann’s modification of it as
a transdisciplinary concept, which is equally applicable to psychic and social
phenomena. The second section of the chapter focuses on the application of
the concept of autopoiesis to the social domain. For Luhmann, social
systems are communication systems which reproduce themselves self-refer-
entially. This conceptualisation is based on two other important concepts:
first, Luhmann has a very particular definition of communication. In con-
trast to the standard notion, communication is not conceptualised as the
transmission of meaning from a sender to a receiver but as an emergent phe-
nomenon arising from the interaction between different human beings. As
such, communications are not the “product” of individual human beings
but of their interaction. Second, Luhmann clearly distinguishes between
social systems and human beings (psychic systems): social systems reproduce
themselves on the basis of communications, and psychic systems on the basis
of thoughts. Both systems are operatively closed to each other and can mere-
ly cause mutual perturbations in each other. In the following two sections
the three types of social systems are described: society, as the all-encom-
passing communication system which includes the other two types of
system; (face-to-face) interaction, as the social system consisting of commu-
nications that are based on the perception of the physical presence of the
participants; and organization, as the system consisting of decision commu-
nications. The chapter closes with a description of Luhmann’s reading of the
calculus of distinction, which was developed by the British mathematician
George Spencer Brown, and on which Luhmann drew heavily in his latest
writings. It should be mentioned that this last section is not absolutely vital
for understanding the central aspects of Luhmann’s theory, and for that
purpose might be skipped; reading it, however, might help appreciate the
finer details of his writings.
13
The second chapter, “The Concept of Autopoiesis” (Niklas Luhmann), is

a translation of a chapter from Organisation und Entscheidung. In this text
Luhmann provides a very concise and clear description of his general con-
cept of autopoiesis. This text is generally considered as the best explanation
of that concept. In the third chapter, “The Autopoiesis of Social Systems”
(Niklas Luhmann), Luhmann shows how the concept of autopoiesis can be
applied to social systems.
After the general introduction of Luhmann’s systems theory in the first
part of the book, Part II focuses on Luhmann’s organization theory. Three
concepts form the core of this theory: decision, organization and paradox.
These concepts will be introduced and critically discussed, and their poten-
tial for empirical research will be demonstrated. According to Luhmann,
organizations have to be conceptualised as systems reproducing themselves
on the basis of decisions; they are nothing other than a network of decisions.
However, the notion of “decision”, as Luhmann has shown, is basically
paradoxical: decisions are ultimately undecidable. Thus, Luhmann places a
paradox at the heart of his organization theory. Consequently, all issues in
organization studies are somehow or other shown to be ways of handling,
or symptoms of, this paradox.
The first chapter of the second part, “The Paradox of Decision Making”,
is a paper by Niklas Luhmann himself that was published in German in
1993 and has never appeared in English before now. In this text Luhmann
develops his concept of “decision” as a paradoxical communication. This
serves as a starting point for unfolding his entire organization theory.
The following chapter, “Displacing the Paradox of Decision Making”
(Morten Knudsen), demonstrates how Luhmann’s concept of decision as
paradoxical communication can fruitfully be applied to empirical research
on organizations. On the basis of an in-depth case study of the changes in
the Danish health-care system, the chapter shows how the paradox of deci-
sion is handled differently under different circumstances. Apart from its
insights in organizational change, the text provides a host of examples of the
practical implications of Luhmann’s theory.
In the last chapter of this part, “On Gorgon Sisters: Organizational Action
in the Face of Paradox” (Barbara Czarniawska) Luhmann’s concepts of
(organizational) paradox and deparadoxification are discussed and com-
pared with those of other writers – in particular with those of Lyotard. The
chapter shows that paradoxes, if faced, lead to paralysis. The chapter dis-
cusses strategies for evading the paralysing effects of paradoxes. In order to
demonstrate the fruitfulness of this way of theorising, the concepts are
applied to an empirical case from the Swedish public sector.
Following the elucidation of Luhmann’s organization theory in the second
part, Part III focuses on the relation between organization and the other two
types of social systems: interaction and society. The first chapter,
14
“Organization and Interaction” (David Seidl) explores how the relation

between organization and interaction system can be conceptualised. On the
basis of Luhmann’s writings about the two types of system and his few
remarks on their interrelation, a systematic account of interactions within
and around organizations is developed. Apart from elaborating on an
almost neglected aspect of Luhmann’s theory, the chapter makes a valuable
contribution to research on organizational interactions in general.
The following chapter, “Organization and Society” (Thomas Drepper),
attempts to clarify the relationship between organization and society.
Proceeding from a critique of classical organization theory, numerous
aspects of the relationship between both kinds of system are discussed and
brought together from various parts of Luhmann’s oeuvre. In this way, the
chapter provides a comprehensive overview of the subject as covered in
Luhmann’s theory of social systems.
In the last chapter of this part, “The Design of Organization in Society”
(Dirk Baecker), the relation between organization and society is addressed a
second time. However, rather than presenting Luhmann’s own theorising
about their relation, the chapter presents an attempt to take Luhmann’s
organization theory a step further, first, by systematically basing it on the
calculus of distinctions by Spencer Brown, and second, by using it as a start-
ing point for a societal theory of organization. Organizational design is
thereby presented as a mechanism of structural coupling between organiza-
tional communication (social system) and perception (psychic system).
While this chapter might prove a difficult read for people without prior
knowledge of Luhmann’s theory, it nevertheless provides an excellent exam-
ple of the way in which certain scholars work on developing Luhmann’s
theory further. As the chapter draws on the calculus of distinctions, the read-
ers not familiar with it are strongly advised to read beforehand the relevant
sections of the first chapter in this volume, in which the calculus is briefly
introduced and explained.
In Part IV Luhmann’s organization theory is being confronted with and
compared to other strands of theorising in organization studies. By address-
ing the similarities and differences of systems theory with other approaches,
the chapters in this part aim at making transparent the particularities of
Luhmann’s theory. Moreover, this part addresses the strengths and weak-
nesses of the theory of social systems as well as possibilities for combining
it with other theoretical traditions.
In the first chapter of this part, “Luhmann’s Systems Theory and Theories
of Social Practices” (Kai Helge Becker), Luhmann’s approach is contrasted
with a strand of social theory that has received particular attention within
organization studies in the past few years: theories based on the notion of
“social practice”. At first, the chapter shows that systems theory and theo-
ries of social practices have two basic assumptions on the nature of social
15
phenomena in common: their focus on culture and the theoretical move of

the “de-centring of the subject”. Proceeding from these insights, it compares
the ways in which both strands of theorising conceptualise some social theo-
retical fundamentals, namely agency, structure, the role of material phe-
nomena and the body. In doing so the chapter aims at providing a better
understanding of the conceptual logic inherent in Luhmann’s theory.
Moreover, on the basis of assessing the particular strengths and weaknesses
of both systems theory and practice-based approaches, it is argued that these
two strands of theorising can be seen as complementary perspectives.
In the second chapter, “Systems Theory and New Institutionalism” (Rai-
mund Hasse), Luhmann’s theory is compared to another strand of theoriz-
ing that has become particularly influential in recent times: the New
(Sociological) Institutionalism. The chapter addresses in particular the
macro-sociological aspirations of both systems theory and the New
Institutionalism and shows the differences in perspective that characterise
the two approaches. Moreover, it is argued that systems theory offers a
micro-foundation of organizational processes that can complement the insti-
tutionalist approach.
The following chapter, “Luhmann’s Systems Theory and Postmodernism”
(Jochen Koch), shows that the widely held assumption, that Luhmann’s the-
ory is an exemplar of modernist (in contrast to postmodernist) thinking, is
inappropriate. Proceeding from the typical characteristics of postmodern
epistemology, it carefully elaborate that fundamental aspects of Luhmann’s
approach have strong parallels in postmodern theorizing. The chapter iden-
tifies areas of commonality and of difference between the logic underlying
Luhmann’s systems theory and postmodern theorising.
The last chapter of this part, “Luhmann’s Systems Theory and Network
Theory” (Veronika Tacke), confronts Luhmann’s systems theory with the
network approach in organization studies. The term “network” has become
one of the most widespread – and even most fashionable – terms within
organization research and is sometimes presented as being at odds with the
conceptualisation of organizations as “systems”. Drawing on some of the
insights of the network approach, the chapter takes up this issue and demon-
strates that Luhmann’s concept of systems theory not only does not contra-
dict the idea of networks, but can even serve as a common frame of refer-
ence that allows the integration of insights of both network theory and other
strands of organization theory.
While the first four parts of the book aimed at explaining Luhmann’s sys-
tems theory in general and his organization theory in particular and at
placing it in the context of other important strands of theorising, the fol-
lowing two parts will demonstrate the application of the theory to concrete
research questions. Part V shows how different forms of organization can be
analysed on the basis of Luhmann’s systems approach.
16
In the first chapter, “Analysing Forms of Organization and Management”

(Fritz Simon), Luhmann’s systems theory is used for comparing two differ-
ent types of companies: stock companies and family businesses are distin-
guished on the basis of the specific ways in which they observe (construct)
their relevant environments, i.e. the ways in which they represent the sys-
tem/environment distinction internally. The second chapter in this section,
“On Defining the Multinational Corporation” (Darnell Hilliard), discusses
the concept of the multinational corporation from a systems-theoretical per-
spective. The “multinationality” of an organization is redescribed as the
specific way in which an organization observes its environment. On the
basis of such an approach, it can be shown that the concept of “multina-
tionality” is itself ambiguous and, in view of current societal developments,
inappropriate. Instead of the “multinational corporation” the concept of the
“world corporation” is suggested to describe a corporation that observes
(constructs) its environment as a single horizon of worldwide business
opportunities and risks.
In Part VI the potential of Luhmann’s theory for questions of management
and consulting is demonstrated. Since Luhmann himself has written very
little on those issues, these chapters have to be read as possible applications
of Luhmann’s theory.
The first chapter of this part, “Communication Barriers in Management
Consulting”, is a text by Niklas Luhmann that was published in 1989 in
German and has been translated for the purposes of this book. In this text
Luhmann describes the implications of conceptualising the organization as
an autopoietically closed system for management consulting. According to
Luhmann’s theory, consulting firms and their client organizations have to be
conceptualised as operatively closed to each other. As a consequence, con-
sulting firms have no means of influencing any organizational processes
directly – despite what they claim. They can, however, cause perturbations
in their client systems, triggering internal processes in the systems them-
selves. Yet, the outcome of this process is beyond the consulting firms’ con-
trol.
The second chapter in this section, “Strategic Management from a
Systems-Theoretical Perspective” (Jan-Peter Vos), uses Luhmann’s theory to
analyse the self-referential logic of strategic management. It argues that the
existing approaches to strategy fail, as they are not aware of – or if they are,
deny – the circularity of their underlying logic: organizations are only
defined through their environment and the environment is only defined
through the organization. Since organization and environment recursively
constitute each other, organizations cannot find a starting point for defining
their strategies – ultimately, any starting point would necessarily be
arbitrary. In contrast to the existing approaches, Luhmann’s theory is shown
17
to offer the possibility of developing a theory of strategy, in which the recur-

sive constitution of organization and environment is acknowledged.
In the last chapter, “Management Accounting from a Systems-Theoretical
Perspective” (Tobias Scheytt), the perspective of Luhmann’s systems theory
is used for criticising the traditional view of management accounting as re-
presenting the organizational world in a neutral and objective way. Instead,
it is argued that management accounting produces context-dependent obser-
vations, which constitute constructions about the organization by the orga-
nization itself. In systems-theoretical terms, management accounting can be
conceptualised as an internal representation (re-entry) of the system/envi-
ronment distinction within the system. As such, these observations can be
shown to be paradoxical, requiring particular mechanisms for handling
them.
Finally, Part VII offers a useful glossary of the central terms in Luhmann’s
theory and an annotated bibliography of selected books and articles by
Niklas Luhmann.
18
04 Kap 1. Seidl 05-11-11 16.16 Sida 19
PART I
The Theory of Autopoietic

Social Systems
04 Kap 1. Seidl 05-11-11 16.16 Sida 20
04 Kap 1. Seidl 05-11-11 16.16 Sida 21
Chapter 1
The Basic Concepts of

Luhmann’s Theory of Social
Systems
David Seidl
The central concept around which the theory of social systems, as developed
by the later Niklas Luhmann, is built is the concept of autopoiesis, original-
ly developed by the two Chilean biologists Humberto Maturana and
Francisco Varela. Autopoiesis (< Greek: autos = self; poiein = to produce)
means self-(re)production. Thus, autopoietic systems are systems that repro-
duce themselves from within themselves, as for example a plant reproduces
its own cells with its own cells. Luhmann argued that the basic idea of
autopoiesis applied not only to biological but also to a large number of non-
biological systems. He thus appropriated the originally biological concept,
modified it and applied it to the social domain. In a similar way to biologi-
cal systems, social systems were thus conceptualised as systems that repro-
duced their own elements on the basis of their own elements.
In this chapter, Luhmann’s concept of autopoietic social systems will be
introduced, starting with the originally biological concept of autopoiesis by
Maturana/Varela and Luhmann’s modification of it as a general systems
concept (first section). Luhmann’s concept of social systems as a specific type
of autopoietic system will then be explained on that basis (second section).
The third and fourth sections will describe and explain the three existing
types of social systems: societal system, interaction system and organiza-
tional system. In the fifth section the mathematical calculus of distinction by
George Spencer Brown will be introduced, which Luhmann drew on exten-
sively in his later writings, and its relevance to Luhmann’s theory will be
shown. Readers who just want to gain a basic understanding of Luhmann’s
theory might skip this last section.
21
04 Kap 1. Seidl 05-11-11 16.16 Sida 22
The concept of autopoiesis

a. The original biological concept of autopoiesis
The theory of autopoiesis was developed by the two Chilean cognitive biol-
ogists Humberto Maturana and Francisco Varela in the sixties and early
seventies. They were trying to answer the question: What is life? Or: What
distinguishes the living from the non-living? Their answer was: A living sys-
tem reproduces itself. This self-reproduction they referred to as autopoiesis.
They defined the autopoietic system as a system that recursively reproduces
its elements through its own elements.
Central to the concept of autopoiesis is the idea that the different elements
of the system interact in such a way as to produce and reproduce the ele-
ments of the system. That is to say, through its elements, the system repro-
duces itself. A living cell, for example, reproduces its own elements.
Proteins, lipids etc. are not just imported from outside:
Consider for example the case of a cell: it is a network of reactions which pro-
duce molecules such that (i) through their interaction [they] generate and par-
ticipate recursively in the same network of reaction which produced them,
and (ii) realize the cell as a material unity. (Varela et al. 1974, p. 188)
In contrast to allopoietic systems (< Greek: allos = other; poiein = to pro-

duce), the elements of autopoietic systems are not produced by something
outside the system. All processes of autopoietic systems are produced by the
system itself and all processes of autopoietic systems are processes of self-
production. In this sense, one can say that autopoietic systems are opera-
tively closed: there are no operations entering the system from outside nor
vice versa.
A system’s operative closure, however, does not imply a closed system
model. It only implies a closure on the level of the operations of the system
in that no operations can enter or leave the system. Autopoietic systems are,
nevertheless, also open systems: all autopoietic systems have contact with
their environment (interactional openness). Living cells, for example,
depend on an exchange of energy and matter without which they could not
exist. The contact with the environment, however, is regulated by the
autopoietic system; the system determines when, through what channels,
and what type of energy or matter is exchanged with the environment (of
course there are certain external forces that might influence the system
directly – e.g. radioactive radiation might destroy parts of the system – but
these influences can never determine what operations come about).
This simultaneous (interactional) openness and (operative) closure of the
autopoietic system becomes particularly important when considering cogni-
tive processes. For Maturana and Varela the concept of living is directly
linked to the concept of cognition.
22
04 Kap 1. Seidl 05-11-11 16.16 Sida 23
Living systems are cognitive systems, and living as a process is a process of

cognition. (Maturana and Varela 1980, p. 13)
In this sense, the operations of an autopoietic system are defined as its

cognitions; life and cognition are one and the same. Hence, everything that
has been said about life applies equally to cognition: cognition is a self-
referential, autopoietic process. This stance is generally known as Radical
Constructivism (also: Operative Constructivism) and expresses the idea that
all cognitions (ideas) are constructs of the respective cognitive system and do
not in any way reflect any kind of external reality.
In the light of this, we might take a further look at the relation between
system and environment. The operative closure of the cognitive system
means that the environment cannot produce operations in the system.
Cognitions are only produced by other cognitions of the same system. The
operative closure does not, however, imply a solipsistic existence of the sys-
tem; on the contrary. As Maturana and Varela argue: operative closure is a
precondition for interactional openness. On the level of its operations, the
autopoietic system does not receive any inputs from the environment but
only perturbations (or “irritations”), which then might trigger internal oper-
ations in the system. In other words, external events may trigger internal
processes but cannot determine those processes. In this respect, Luhmann
(2000c, p. 401) speaks of a “trigger-causality” [Auslösekausalität] instead of
an “effect-causality” [Durchgriffskausalität]. For example, if one puts one’s
finger on the flame of a candle, the rapid movement of the atoms in the
flame will trigger an electric impulse in the nervous system, which will lead
to the cognition of “heat”. Thus, what can be seen very clearly in this exam-
ple is the clear distinction between cognitive system and environment. The
events in the environment do not enter the cognitive system; the rapid move-
ment of the atoms triggers operations in the system that are qualitatively
completely different. This triggering is only possible because the system has
produced specific structures, that is to say, nervous sensors, which can be
stimulated by the rapid movement of electrons (in the same way as they
would be stimulated e.g. by acid). From this the nervous system constructs
the sensation of heat – the heat does not exist in the flame; in the flame we
only have the rapid movement of atoms. If the eye, for example, was moved
into the direction of the flame, due to the specific structures of the nervous
system in the eye, the nervous system would construct the experience of light
and specific colours. Again, light and colours as such do not exist in the
flame; in the flame there are merely electromagnetic waves.
The theory of autopoiesis clearly distinguishes between, on the one hand,
the reproduction of the system as such, and on the other hand, the structures
according to which this reproduction takes place: in order to “survive”,
an autopoietic system has to produce constantly further elements. If this
23
04 Kap 1. Seidl 05-11-11 16.16 Sida 24
(re-)production stops, the system disappears. For instance, if a plant stops

producing its cells it is considered dead. It is irrelevant what concrete cells
are produced; whether the plant produces a new leaf, extends its roots or
grows a blossom does not matter – as long as any new elements are pro-
duced the plant is still alive. The fact of the reproduction as such – inde-
pendently of the concrete elements reproduced – is referred to as the
autopoiesis of the system. The likelihood of the continuation of reproduc-
tion, however, depends on the concrete elements reproduced. For example,
if a flower stops producing leaves and instead only extends its roots it loos-
es its viability; that is to say, its ability to produce any further elements at
all. What concrete elements are produced at any moment is determined by
the structures of the system (the system in this sense is structurally deter-
mined): for example, the stem of the plant restricts where new leafs can
grow. The structures themselves, however, are not pre-given in any sense, as
in structuralist theories, but are themselves the product of the autopoietic
system. In other words, in its reproduction the system produces and repro-
duces its very own structures of reproduction. This aspect, i.e. the self-
determination of its own structures, is referred to as self-organization. Thus,
while autopoiesis refers to the reproduction of elements as such, self-orga-
nization refers to the determination of structures (Luhmann 2000c, p. 47).
A central element within the theory of autopoiesis is the concept of struc-
tural coupling, which refers to the relation between systems and their envi-
ronments. As explained above, environmental events can trigger internal
processes in an autopoietic system but the concrete processes triggered (and
whether any processes are triggered at all) are determined by the structures
of the system. For example, some animals have certain neuronal structures
that allow certain electromagnetic waves in their environment to trigger
internally the sensation of certain colours; other animals, again, possessing
other structures might not be stimulated by such waves or might be stimu-
lated by them in other ways. A system is said to be structurally coupled to
its environment (or to other systems in its environment) if its structures are
in some way or other “adjusted” to the structures of the environment (or to
systems in the environment); in other words, if the structures of the system
allow for reactions to “important” environmental events. For example, ani-
mals living above ground are structurally adapted to a different environment
from those living underground. The former have structures that can be
stimulated by electromagnetic waves, which leads to different impressions of
colour, while the latter might have structures that can be more easily stimu-
lated by vibrations, which leads, respectively, to equally differentiated
impressions.
24
04 Kap 1. Seidl 05-11-11 16.16 Sida 25
b. Luhmann’s general, transdisciplinary concept of autopoiesis

There have been many attempts by social scientists to apply the concept of
autopoiesis to the social domain (for an overview of different applications
see Mingers 1995). Most of them, however, failed as they had tried to trans-
fer the original concept directly from one domain to the other. In contrast to
most others, Luhmann did not apply the original concept directly to the
social domain but (in line with the general systems tradition) tried to
abstract from the originally biological concept of autopoiesis a general,
transdisciplinary concept of autopoiesis. This transdisciplinary concept of
autopoiesis should then be open to re-specifications by the different disci-
plines, e.g. sociology, biology, psychology. In this respect, Luhmann wrote:
[I]f we abstract from life and define autopoiesis as a general form of system-
building using self-referential closure, we would have to admit that there are
non-living autopoietic systems, different modes of autopoietic reproduction,
and general principles of autopoietic organization which materialize as life,
but also in other modes of circularity and self-reproduction. In other words,
if we find non-living autopoietic systems in our world, then and only then
will we need a truly general theory of autopoiesis which carefully avoids ref-
erences which hold true only for living systems. (Luhmann 1986b, p. 172)
Luhmann suggests that we speak of autopoiesis whenever the elements of a

system are reproduced by the elements of the system itself. This criterion, as
he points out, is also met by non-biological systems. Apart from living sys-
tems, Luhmann identifies two additional types of autopoietic systems: social
systems and psychic systems. While living systems reproduce themselves on
the basis of life, social systems reproduce themselves on the basis of com-
munication, and psychic systems on the basis of consciousness or thoughts,
their elements are not physical substances but elements of meaning (for
explanations see below). Furthermore, social systems can be differentiated
into the three subtypes: societies, organizations and interactions (Figure 1).
1. Level Autopoietic Systems
2. Level Living Systems Psychic Systems Social Systems
3. Level Societies Organizations Interactions
Figure 1: Types of autopoietic systems.
25
04 Kap 1. Seidl 05-11-11 16.16 Sida 26
On the basis of this typology of systems one can derive a hierarchy of three
levels of analysis. On a first level we find statements which concern autopoi-
etic systems in general without reference to any particular mode of repro-
duction. On this level we can find the general concept of autopoiesis.
Statements on this level are equally valid for living as for psychological and
for social systems (and their subtypes). On a second level we find different
applications of the general theory of autopoiesis. There are three such areas:
research concerned with the particular characteristics of (1) living systems,
(2) psychic systems, and (3) social systems. Most of Maturana’s and Varela’s
research can be placed on the level of living systems. It produces general
statements concerning living systems, which, however, are not applicable to
social or psychic systems. Psychological research is concerned with the par-
ticularities of systems that are reproduced on the basis of consciousness.
Sociological research on this level is concerned with the particularities of
systems that reproduce themselves on the basis of communication.
Statements produced in this area concern all three types of social systems.
On a third level one can find research in the social field concerning the par-
ticularities of societies, organizations, and interactions. That is to say, for
each type of system the particular mode of reproduction has to be defined
and the consequences of the particular mode of reproduction analysed.
Thus, for social research in particular, one can find four different areas of
research: research on the general level of social systems (e.g. Luhmann
1995a) and research on the particular types of social systems – on societies
(e.g. Luhmann 1997f), on organizations (e.g. Luhmann 2000c), and on
interactions (e.g. Luhmann 1993j, pp. 81–100).
Against the backdrop of categorisation of analytical levels, the transfor-
mation of the original autopoiesis concept to a concept applicable to the
social domain becomes clear. Instead of being transferred directly from the
field of biology into the field of sociology, the concept is first abstracted to
a general concept on a transdisciplinary level, and then re-specified as social
autopoiesis and the autopoiesis of particular types of social systems. We can-
not examine the abstraction of the concept of autopoiesis in detail here, but
merely want to highlight two important modifications: the temporalisation
and de-ontologisation of the concept of element (if this modified, general
concept of autopoiesis were to be re-applied to the biological domain,
Maturana’s and Varela’s original theory would have to be modified accor-
dingly).
Luhmann’s general concept of autopoiesis radicalises the temporal aspect
of autopoiesis. While Maturana and Varela originally conceptualised the ele-
ments of their biological systems as relatively stable chemical molecules,
which have to be replaced “from time to time”, Luhmann conceptualises the
elements as momentary events without any duration. Events have no
26
04 Kap 1. Seidl 05-11-11 16.16 Sida 27
duration but vanish as soon as they come into being; they “are momentary
and immediately pass away” (Luhmann 1995f, p. 287).
Events are elements fixed as points in time. […] They occur only once and
only in the briefest period necessary for their appearance (the “specious pres-
ent”). (Luhmann 1995f, p. 67)
Through this shift from a reproduction of relatively stable elements, to a

reproduction of momentary events, Luhmann radicalises the concept of
autopoiesis. Because the elements of the system have no duration the system
is urged to produce constantly new elements. If the autopoiesis stops, the
system disappears immediately.
In addition to temporalisation, Luhmann deontologises the concept of ele-
ment. Elements are defined as elements merely through their integration into
the system. Outside or independently of the system they have no status as
elements; that is to say, they are “not ontically pre-given” (Luhmann 1995f,
p. 22). Elements can, of course, be made up of different components, which
could be analysed independently of the system, but as elementary units they
are only defined through their relation to other elements and (in this sense)
through the function they fulfil for the system as a whole. Luhmann writes:
[W]e have deontologized the concept of element. Events […] are not elements
without substrate. But their unity corresponds to no unity in the substrate; it
is created in the system through their connectivity. Elements are constituted
by the systems that are composed of them […]. (Luhmann 1995f, p. 215)
As a consequence of deontologising the concept of element, the concept of

“production” (as in “self-reproduction”) gets a functional meaning. “Pro-
duction” refers to the use of an element in the network of elements. The
important point in this conceptualisation is that the element and the use of
the element are not two different issues, but two sides of the same coin. It is
not that we first have the element and then the system makes use of it: only
by being used, i.e. by being related to other elements, does the element
become an element. Thus, one can say: the element is produced as a result
of being used (Luhmann 1997a, pp. 65–66). One can, of course, analyse the
substratum, on which an element rests, and find a whole range of causal fac-
tors that are involved in bringing the element about, but the particular unity,
as which the element functions in the system (i.e. the characteristics that
make it an element of the system), can only be produced by the system itself.
27
04 Kap 1. Seidl 05-11-11 16.16 Sida 28
Social systems
a. Communications as the elements of social systems
The first decision Luhmann as a theoretician had to make for constructing
his general sociological theory of autopoiesis (which on this level of analy-
sis is still unspecific with regard to the three types of social systems: society,
organization, interaction) was, what he should treat as the basic elements of
the social system. The sociological tradition suggests two alternatives: either
persons or actions. Luhmann rejected both as incompatible with the concept
of autopoietic social systems. Instead, he chose a completely different ele-
ment: communication (or more precisely: the communicative event), sug-
gesting a “conceptual revolution” (Luhmann 1986b, p. 178). He writes:
Social systems use communications as their particular mode of autopoietic

reproduction. Their elements are communications which are recursively pro-
duced and reproduced by a network of communications and which cannot
exist outside of such a network. (Luhmann 1986b, p. 174)
In order to understand this conception of social systems, we have to clarify

Luhmann’s concept of communication, which differs considerably from the
conventional notion of communication as an asymmetrical process of trans-
ferring meaning or information from a sender to a receiver. Building on the
speech theories of Karl Bühler (Bühler 1934), Luhmann conceives of com-
munication as a combination of three components: (1) information, (2)
utterance and (3) understanding, each of which Luhmann conceptualised as
a selection.
In accordance with Shannon and Weaver (1949) he defined information
as a selection from a repertoire of possibilities. Every communication selects
what is being communicated from everything that could have been commu-
nicated. With utterance Luhmann refers to the form of and reason for a
communication: how and why something is being said. One can say, the
utterance is the selection of a particular form and reason from all possible
forms and reasons. Understanding is conceptualised as the distinction
between information and utterance. For a communication to be understood,
the information has to be distinguished from the utterance: what is commu-
nicated must be distinguished from how and why it is communicated. For
example, if alter says to ego: “I am tired”, ego has to distinguish the infor-
mation (“I am tired” and not e.g.: “I am very energetic”) from the utterance
(the words alter is using and the reason why alter is saying it: for example,
here alter wants to indicate that ego should leave him alone; he is not say-
ing it in order to get any advice on what to do about his tiredness). Thus,
understanding can be defined as a selection of a particular distinction
between utterance and information.
28
04 Kap 1. Seidl 05-11-11 16.16 Sida 29
While most communication theories refer only to the first two elements
– information and utterance – in Luhmann’s concept, the third element
– understanding – plays a central role. Instead of approaching a communi-
cation from an “intended meaning” of the communication, Luhmann
reverses the perspective: (the meaning of) a communication is ultimately
determined through the understanding. Luhmann (1995, p. 143) writes:
“Communication is made possible, so to speak, from behind, contrary to the
temporal course of the process.” This is also called the “principle of
hermeneutics”:
[This principle states] that not the speaker but the listener decides on the
meaning of a message, since it is the latter whose understanding of the set of
possibilities constrains the possible meaning of the message, no matter what
the speaker may have had in mind. (Baecker 2001, p. 66)
A central point in Luhmann’s concept of communication is that the three

selections form an “insoluble unit”; undoubtedly, this unit can be divided
analytically into its three components (for example by other communica-
tions), but only as a unit does it constitute a communication. Because of
that, a communication – as this unity of the three selections – cannot be
attributed to any one individual (psychic system). Instead, communication
constitutes an emergent property of the interaction between many (at least
two) psychic systems. As Luhmann writes:
Communication is a genuinely social – and the only genuinely social – oper-

ation. It is genuinely social insofar as it presupposes the involvement of a
multitude of psychic systems but, or better: because of that, it cannot be
attributed as a unit to a single psychic system. (Luhmann 1997a, p. 81; my
translation)
Thus, although psychic systems are necessarily involved in bringing about

communication, the communication (as this unit) cannot be understood as
the product of any particular psychic system.
In order to render more precise Luhmann’s concept of communication we
have to take another, closer look at his concept of understanding. Under-
standing, as we said above, is the distinction between utterance and infor-
mation; but whose understanding is of relevance here? Again, for Luhmann
it is not the psychic system that is of interest. Instead, it is the understand-
ing implied by the ensuing communications – in the same way as the con-
crete meaning of a word in a text is only defined through the words follow-
ing it in the text. Thus, the meaning of a communication, i.e. what differ-
ence a communication makes for later communications, is only retrospec-
tively defined through the later communications. For example, whether a
“Yes” is understood as approval or as a question or as a neutral acknowl-
edgement of the given information is only determined through the reaction
29
04 Kap 1. Seidl 05-11-11 16.16 Sida 30
of the connecting communications; e.g. “I’m happy you agree”, “You don’t
believe me?”, or “What is your own opinion?” (again, the meaning of those
communications is itself only defined through the communications connect-
ing to them). In other words, Luhmann is not referring to any form of psy-
chic understanding, but to an understanding on the level of the communica-
tions. What the “involved” psychic systems think during the communication
processes, i.e. how the psychic systems understand the communication, is (at
first) completely irrelevant to the communication. For example, the psychic
systems might understand the “Yes” as a question, while the ensuing com-
munications might treat it as approval. Of course, what the psychic systems
think about the communications might ultimately influence the communi-
cations because of the structural coupling between the two systems: differ-
ent thoughts about the communications might lead to the psychic systems
causing different perturbations in the social system, and thus might ulti-
mately lead to different communications coming about. But it has to be
stressed again that the psychic systems cannot determine what communica-
tions come about.
This retrospective determination of the communication through ensuing
communications is connected with a fourth type of selection. With under-
standing, a communicative event, as the synthesis of the three selections
(utterance, information and understanding), is complete. However, if the
social system is not discontinued, a fourth type of selection will take place:
acceptance or rejection of the meaning of the communication. This fourth
selection is already part of the next communication. It is important not to
confuse the third and fourth selections: understanding does not imply
acceptance! For example, a pupil understands when the teacher says: “do
your homework”, but he might still reject the communication, answering:
“No, I won’t”. There might be communicative structures which make
acceptance more likely than rejection, but the concept of communication is
not focussed on acceptance – in contrast, for example, to Habermas’s (1987)
concept of communication. On the contrary, every communicative event
provokes the selection between acceptance and rejection. This distinction
between understanding (as part of the first communication) and the selec-
tion acceptance/rejection (as part of the ensuing communication) adds a
dynamic element which bridges the gap from one communicative event to
the next.
This leads to a very important point: the (re-)production of communica-
tions. In accordance with the general concept of autopoiesis, communica-
tions only “exist” as communications through their relation to other com-
munications; as explained above, a communication is only defined through
the ensuing communications. This does not mean that without the relation
there is nothing at all (there are, for example, words and sounds), but they
have no status as communications. In this sense one can say that it is the
30
04 Kap 1. Seidl 05-11-11 16.16 Sida 31
network of communications that “produces” the communications. In other

words, it is the context of other communications that makes a communica-
tion count as such at all. Luhmann thus famously said: “Only communica-
tions can communicate.”
So far we have explained the autopoiesis of social systems, i.e. the repro-
duction of communications through communications. We have explained
how communications produce communications but not what communica-
tions are produced, which is a question about the structures of social sys-
tems. Luhmann conceptualises social structures as expectations (Luhmann
1995f). In every situation certain communications are expected and not
others. For example, a question about one’s wellbeing is expected to be fol-
lowed by an answer on this issue and not by a statement about the latest
weather forecast. The expectation to a certain extent preselects the possibil-
ities for further communications: it makes certain communications more
likely than others (it does not, however, exclude any possibilities complete-
ly). These expectations are recursively reproduced through communications.
Whenever a specific expectation is met by an adequate communication, the
expectation is confirmed and thus likely to continue to function as a struc-
ture. However, if the expectation is repeatedly not met, it might be changed.
The topics of communication are an important form of social structure.
Topics provide preselections of all, in principle, possible communications:
certain possibilities of communication fit a specific topic and others do not.
For example, in a conversation about social theory one would not expect a
communication about cooking. However, if such a communication came
about, the topic of communication might be changed.
b. Interpenetration: the relation between social and

psychic systems
The relation between social system and “human being” is a very controver-
sial aspect of Luhmann’s theory; it is also the most misunderstood aspect.
For an adequate appreciation it is thus necessary to outline this relation
carefully.
In Luhmann’s theory the “human being” is not conceptualised as forming
a systemic unity. Instead, it has to be understood as a conglomerate of
organic and psychic systems. The former consists of biochemical elements,
the latter of thoughts. Both systems are operatively closed towards each
other: neither system can contribute elements to the other system. The sys-
tems are, however, structurally coupled; that is to say, their respective struc-
tures are adjusted to each other in such a way as to allow mutual perturba-
tions.
Although the “human being” does not constitute a systemic unity, the
social system treats it as such: it constructs it as a person. In other words,
“persons” do not exist as such: they are not systems but a construct of the
31
04 Kap 1. Seidl 05-11-11 16.16 Sida 32
social system with which this system refers to the conglomerate of organic
and psychic systems. A social system might, for example, construct the per-
son “John Smith”. Whenever the “corresponding” conglomerate of organic
and psychic systems causes perturbations in the social system, the social sys-
tem will refer to it as caused by “John Smith”. In the course of time a social
system will develop certain expectations about when and how this con-
glomerate might cause perturbations. These expectations become part of the
construct “John Smith”. Ultimately, we could say that a person is nothing
other than a complex of expectations that a system has vis-à-vis a specific
conglomerate of organic and psychic systems. Luhmann defines “person” in
this sense as the “social identification of a complex of expectations directed
toward an individual human being” (Luhmann 1995f, p. 210).
Particularly important for the social system is the psychic system. Like
social systems, psychic systems are meaning-constituted systems. However,
in contrast to social systems, the meaning events do not materialise as com-
munications but as thoughts. In other words, psychic systems reproduce
themselves on the basis of consciousness: only thoughts can produce
thoughts. Not even events in the brain, i.e. electric impulses, can take part
in the autopoiesis of psychic systems: a nerve impulse is not a thought.
Psychic systems are not only closed with regard to other types of systems but
also with regard to each other. No psychic system has direct access to an-
other psychic system; my thoughts can never enter your psychic system.
As operatively closed systems, psychic and social systems constitute envi-
ronments for each other: thoughts cannot become communications and
communications cannot become thoughts. Mutual influences are restricted
to the structural level. There merely exists a relation of structural coupling:
both types of systems are structurally adapted to each other in a way which
allows for mutual perturbation (see our explanations on structural coupling
above). Luhmann calls the specific structural coupling of social and psychic
systems interpenetration. Luhmann speaks of interpenetration if
an autopoietic system presupposes the complex achievements of the

autopoiesis of another system and can treat them like parts of the own sys-
tem. (Luhmann 1995g, p. 153; my translation)
The simultaneous (but separate) autopoieses of psychic systems is constitu-

tive for the autopoiesis of the social system. Without psychic systems social
systems are impossible – and probably vice versa. Every communicative
event presupposes “parallel” events in the psychic systems. For the percep-
tion of utterances, the social system depends at any rate on the psychic sys-
tem: the social system cannot hear spoken words, nor read letters.
Furthermore, psychic systems serve as a memory as they can remember com-
municative events beyond their momentary point of existence. Because of
their structural coupling, social systems can expect their communications to
32
04 Kap 1. Seidl 05-11-11 16.16 Sida 33
cause perturbations in the psychic systems and to receive perturbations from

the psychic systems when necessary. They can, for example, count on psy-
chic systems to trigger further communications after every communication.
Although psychic systems trigger communication processes and vice versa
– we repeat this point, since it is very important – the processes of the
psychic system and the social system do not overlap in any way.
The most important evolutionary achievement for the coupling of social
and psychic systems is language – this does not, of course, mean that com-
munication is possible only with language. Language ensures that psychic
systems are perturbed through the communication processes. Articulated
speech, for example, normally disturbs people who are not involved in the
communication more than mere noise does. As Luhmann writes:
The differentiation of specific objects of perception, which stand out and fas-
cinate as they have no resemblance at all with anything else perceptible, is
crucial [for the coupling between social and psychic systems] […]. Language
and writing fascinate and preoccupy consciousness and in this way ensure
that it comes along, although the dynamic of consciousness does not necessi-
tate this and always provides distractions. (Luhmann 1995g, p. 41; my trans-
lation)
Language is a purely social phenomenon (psychic systems do not think in

language) but thought processes can be structured in a way that is comple-
mentary to language, particularly during communication processes:
thoughts are broken down into chunks equivalent to those of sentences and
words. In other words, psychic processes are synchronised with communi-
cation processes and, in this way, they “know” when to contribute pertur-
bations to the communication process in order to make the reproduction of
the social system possible.
Although Luhmann’s strict distinction between social and psychic systems
runs counter to our everyday beliefs and almost all social and psychological
theories, it has one important theoretical advantage. It allows for a concept
of the social realm which is clearly distinguished from the psychological.
Consequently, social and psychic phenomena can be analysed in their own
right. This does not lead to a marginalisation of psychic systems for social
systems – as has often been criticised. On the contrary, through this differ-
entiation it can be clearly shown that, and in what way, both types of sys-
tem depend on each other. The treatment of human beings as environments
of the social system (and not as part of it), as Luhmann writes,
does not mean that the human being is estimated as less important than tra-
ditionally. Anyone who thinks so (and such an understanding underlies either
explicitly or implicitly all polemics against this proposal) has not understood
the paradigm change in systems theory.
33
04 Kap 1. Seidl 05-11-11 16.16 Sida 34
Systems theory begins with the unity of the difference between system and
environment. The environment is a constitutive feature of this difference, thus
it is no less important for the system than the system itself. (Luhmann 1995f,
p. 212)
c. Communication and action

While Luhmann suggests treating communications – and not actions – as the
elements of social systems, the concept of action does not become com-
pletely irrelevant. On the contrary, Luhmann assigns it an important role in
the reproduction of the system. The very fact that not only sociologists but
all social systems use the concept of action means that it cannot be ignored.
Often communication is treated as some kind of action; in this sense
Habermas (1987), for example, speaks of “communicative action”. But
Luhmann’s communication – and this is very important – is not a kind of
action. As explained above, communication is constituted as a synthesis of
a threefold selection of utterance, information and understanding. The con-
cept of action cannot account for all three selections. It might capture the
first two selections but certainly not the third: understanding.
[T]he perfection of communication implies understanding and understanding

is not part of the activity of the communicator and cannot be attributed to
him. (Luhmann 1986b, p. 178)
Thus, a central element of Luhmann’s concept of communication would be

missing if interpreted as action. Apart from that, the original intention of an
action is not important for the communication. For example, looking at
one’s watch might be understood as communicating one’s boredom,
although one only wants to know what time it is.
Luhmann suggests treating action as a (fictive) construct of social systems
for observing, and communicating about, their communications: social sys-
tems observe their communications not as communications but as actions,
which they causally attribute to “persons” (“actors”). As explained above
also the “person” is a construct of the social system, with which the system
refers to the human being as the conglomerate of psychic and social systems.
This attribution of communications to persons as actions is deeply ingrained
in language: our sentences are usually based on a subject-object logic
– “somebody is doing something”. In other words language forces the attri-
bution of communication to someone; it is almost impossible to speak about
communications as something that is not “done” by someone.
In this way, the social system constructs an image (self-description) of
itself as a nexus of actions. This self-description constitutes a simplification
of the system and this is also where its function lies. The simplified version
of itself serves as a means of orientation for its (re-)production, which has
several advantages: first, actions are easier to recognise and deal with than
34
04 Kap 1. Seidl 05-11-11 16.16 Sida 35
communications. While an orientation according to communications pre-

supposes a clear distinction between utterance, information, and under-
standing, an orientation according to action only has to deal with the spe-
cific rules of attribution.
The simplification lies in the fact that only actions and not fully communica-
tive events serve as connective points, in that an abstraction suffices to com-
municate action or simply connective behaviour, and in that one can to a
great extent omit the complexities of the complete communicative occur-
rence. The fact that one need not examine (or need examine only under very
specific conditions) which information an utterance referred to and who
understood it takes some of the load off. (Luhmann 1995f, p. 168)
Second, the description of communication processes as connections of

actions leads to clear-cut temporal relations between different elements.
While communications are completed only after understanding has taken
place, that is, the communicative occurrence is “held in suspense” between
utterance and understanding (Luhmann 1995f, p. 169), actions mark one
point in time (determined by the utterance).1 As a consequence, the different
communications are also much more clearly differentiated: while communi-
cations are heavily entangled with each other – with later communications
retrospectively defining the meaning of earlier ones – actions appear to be
self-defined and do not presuppose other actions.
Society and interaction

According to Luhmann we can distinguish three types of social systems:
society, face-to-face interaction and organization. All three systems are
social systems insofar as they reproduce themselves on the basis of commu-
nications. They are, however, different types of social systems insofar as they
reproduce different types of communications. In the following we will give
a brief description of Luhmann’s concept of society and interaction system;
the organization as the third type of social system will be dealt with after-
wards in a separate section.
a. Society
For Luhmann society is the system that encompasses all communications; all
communications that are produced are part of society and as such reproduce
it. Hence there are no communications outside society. The borders of socie-
ty are the borders of communication. Luhmann thus writes:
1 The phenomenon of “suspense” is particularly extreme in the case of communications by
letter where the utterance and the understanding are usually drawn far apart.
35
04 Kap 1. Seidl 05-11-11 16.16 Sida 36
[S]ociety is the all-encompassing social system that includes everything that is

social and therefore does not admit a social environment. If something social
emerges, if new kinds of communicative partners or themes appear, society
grows along with them. They enrich society. They cannot be externalized or
treated as environment, for everything that is communication is society.
Society is the only social system in which this special state of affairs occurs.
(Luhmann 1995f, p. 408)
As a consequence of this conceptualisation society only exists in singular:

there is only one world society. For Luhmann, society is thus “the auto-
poietic system par excellence”: all elements (communications) are produced
by the own elements (communications) and cannot get out of this network
of elements (society).
Amongst the three types of social systems, society is a very particular one
as it encompasses the other two systems – interaction and organization. As
the system including all communications, it also includes the specific inter-
actional and organizational communications. Or, the other way around, all
interactional and organizational communications always also reproduce
society.
In the course of its evolution society has undergone three major structur-
al changes; that is to say, changes of how societal communications were
structured (Luhmann 1997a). In archaic times society was differentiated
into equal subsystems (segmentation), e.g. different tribes, clans or families.
This was replaced later on by a differentiation according to the logic of cen-
tre and periphery: the differentiation between city and country. In late
medieval times a hierarchical form of differentiation emerged with different
social strata or classes (stratification). With the emergence of modern socie-
ty, around the 18th century, that was replaced by the current, functional dif-
ferentiation, where we find several societal subsystems specialised in serving
specific societal functions; for example, law, science, economy, art, religion.
Each of these primary forms of differentiation can be combined with the
other forms of differentiation on a secondary level: for instance, in stratified
society the various strata were often differentiated internally into equal sub-
systems (segmentation) or on the basis of centre vs. periphery. Similarly, the
various functional subsystems might be differentiated internally into equal
subsystems, into centre and periphery, or hierarchically.
While Luhmann has written extensively about all four types of differenti-
ation, the functionally differentiated society, as the present form of differen-
tiation, is of most interest to us here. It is characterised by the existence of
different functional systems. All of these systems are communication systems
that are themselves operatively closed on the basis of a specific binary cod-
ing. That is to say, all communications taking part in the reproduction of a
particular functional subsystem “carry” a specific code. For example, the
code of the legal system is legal/illegal; the code of the economic system is
36
04 Kap 1. Seidl 05-11-11 16.16 Sida 37
payment/non-payment; the code of the system of science is truth/untruth; the

code of the political system is power/non-power. Each of these systems com-
municates about itself and its environment according to its specific code: for
example, for the legal system something is either legal or illegal, or has no
relevance at all; for the economic system something is either a payment or a
non-payment, or has no relevance to it: that is to say, whether something is
legal or illegal is irrelevant to the economic system. Each communication of
a functional system relates to other communications of the same function
system on the basis of the function-specific coding. For example, a commu-
nication of the legal system relates to other legal communications as either
legal or illegal communication. A legal ruling refers to another legal ruling
(as a legal ruling) in order to substantiate itself – it cannot, however, refer to
payments being made (economic system). These functional systems are
operatively closed in the sense that only communications carrying the func-
tion-specific code can take part in the reproduction of the function system.
Thus, only legal communications can reproduce the legal system, while eco-
nomic, scientific, political etc. communications cannot; only scientific com-
munications can reproduce science, and so on.
The functionally differentiated society combines extreme inclusiveness
with extreme exclusiveness. On the one hand, each function system includes
all function-specific communications. Thus, all legal communications are
part of the legal system; all economic communications are part of the eco-
nomic system; all scientific communications are part of the scientific system,
etc. On the other hand, these societal functions are exclusively served by the
respective function system: only the legal system can provide justice (legali-
ty), while the economic or scientific system cannot; only the economic sys-
tem can produce payments; only the scientific system can produce truth.
Thus, each of these systems has an enormous reach as each refers to the
entire world society. At the same time, however, each system’s range is also
very narrow as each deals with only one function.
Functional systems constitute environments for each other. The systems
cannot “exchange” their communications: an economic communication
cannot take part in the scientific system, a political communication cannot
take part in the legal system etc. Each system reproduces itself self-
referentially and registers communications of other function systems merely
as perturbations, which it processes according to its own logic. The eco-
nomic system, for example, would register legal communications merely
with regard to their consequences on payments/non-payments. No function
system can control any other function system; no system – not even the
political system – is dominant over any other system. In this sense there is
no centre to the functionally differentiated society. The different systems are
merely structurally coupled to each other; that is to say, their structures are
adjusted to each other in such a way as to allow them to react to their
37
04 Kap 1. Seidl 05-11-11 16.16 Sida 38
respective operations. For example, the legal system and the economic sys-
tem are structurally coupled through sales contracts. For the legal system the
sales contract is a legal communication re-distributing legal rights and
duties; for the economic system it is an economic (i.e. different) communi-
cation re-distributing payments. In other words, the sales contract is two dif-
ferent communications for the two different function systems, but it allows
the two systems to somewhat “co-ordinate” their respective processes.
b. Interaction
Like all social systems (face-to-face) interactions are systems that reproduce
themselves on the basis of communications. In contrast to society, however,
these communications are of a particular kind; namely, communications that
are based on the perception of the physical presence of their participants.
There is no doubt that perception as such is clearly a psychic phenomenon
– communications cannot perceive. However, reflexive perception gives rise
to communication as Luhmann argues:
If alter perceives that alter is perceived and that this perception of being per-
ceived is perceived, alter must assume that alter’s behavior is interpreted as
communication whether this suits alter or not, and this forces alter to control
the behavior as communication. (Luhmann 1995f, p. 413)
Thus, every communication refers to the fact that all participants perceive
each other as present – a face-to-face contact is thus a precondition. How-
ever, not everyone who is physically present will also be treated as present
by the communication. For example, people at other tables in a restaurant,
although physically present, might not be considered present by the interac-
tional communication. Similarly, not all perceptible behaviour will necessar-
ily be treated as perceptible, i.e. treated as present, by the interaction; for
example blowing one’s nose. In other words, every interactional communi-
cation distinguishes between what to consider present and what to consider
absent. Making this distinction qualifies the communication as interaction-
al. One could also say, the interactional communications carry the code
“presence/absence” analogously to the function codes described above.
Like functional systems, interactional systems are operatively closed inso-
far as only communications carrying the code “presence/absence” take part
in the reproduction of the interaction system. Communications in an inter-
action can only connect to other communications that are treated as present
and not to those treated as absent (e.g. the communications of another com-
munication at the next table; unless those communications are treated as
present and thus as part of the same interaction system).
What communications are treated as present or absent depends to a
certain extent on the structures of the interaction. Like all social systems,
38
04 Kap 1. Seidl 05-11-11 16.16 Sida 39
interactional communications are structured through the topics of commu-

nication. These topics often make a pre-selection of who to treat as present.
For example, if the managers of a company sat in a pub to discuss the com-
pany’s strategy, the interaction would probably treat only the managers as
present and all other visitors of the pub, who might even happen to sit with
them at the same table, as absent. However, if the topic changed to football
new persons might be included in the interaction, while some of the initial
participants might be excluded again.
Organization
a. Decisions as the elements of organizations
Luhmann conceptualises organizations as social systems that reproduce
themselves on the basis of decisions. In other words, organizations are
systems that consist of decisions and that themselves produce the decisions of
which they consist, through the decisions of which they consist. (Luhmann
1992a, p. 166; my translation)
But what is a decision? Luhmann argues that the standard definitions of

“decision” are not very helpful. Mostly, a decision is defined as “choice”.
This, however, means that a decision is defined through a synonym that is
equally unclear. Sometimes the definition is specified somewhat more as “a
choice among alternatives”. This, however, does not add much as the con-
cept of alternative is itself only defined in relation to choice: alternatives are
those possibilities among which one can choose; in other words, the choice
defines the alternatives. Thus, one only finds tautological definitions.
Building on, and modifying, the existing definitions Luhmann suggests
conceptualising decision as a specific form of communication. It is not that
decisions are first made and then communicated; decisions are communica-
tions. As has been said about communications in general, decision commu-
nications too are not produced by “human beings” but by the social system,
the organization.
What is particular about decisions is that they are “compact communica-
tions” (Luhmann 2000c, p. 185), which communicate their own contin-
gency (“contingency” here in the sense of “also possible otherwise”). In con-
trast to an ordinary communication, which only communicates a specific
content that has been selected (e.g. “I love you”), a decision communication
communicates also – explicitly or implicitly – that there are alternatives that
could have been selected instead (e.g. “I am going to employ candidate A
and not candidate B”). As such, decision communications are always para-
doxical communications: the more they communicate that there are real
alternatives to the one that has been selected, the less the selected alternative
39
04 Kap 1. Seidl 05-11-11 16.16 Sida 40
will appear as justified and thus the less the decision will be accepted as
“decided”. Equally, the more the selected alternative is being justified as the
right selection, the less the other options will appear as alternatives and thus
the less the decision will appear as “decision”. Or, to put it in linguistic
terms, every decision communication contains a performative self-contra-
diction: the “report” aspect and the “command” aspect (Ruesch and
Bateson 1951) of the decision communication contradict each other. The
more clearly the decision is communicated as a selection among possible
alternatives (report aspect), the less the decision will be accepted by later
communications as a decision (command aspect).
Because of their paradoxical nature, decision communications are subtly
calling for their own deconstruction by the ensuing communications.
Without any other communicative provisions, decision communications
would have a very high “failure rate”. So, why does organizational com-
munication not break down all the time? Luhmann gives two answers to this
question. First, organizations totalise decisions as the organizational form of
communication – organizations are operatively closed on the basis of deci-
sions. Thus, even the deconstruction of a decision in an organization has to
be communicated as a decision. In other words, the rejection of a decision
can itself only be communicated as yet another decision, otherwise it would
not be part of the organizational autopoiesis (Luhmann 2000c, p. 145).
Furthermore, decision communications in organizations can usually refer to
other (successfully completed) decisions (“decision premises”; see below) to
stabilise the decision, i.e. decisions prohibiting the rejection of certain other
decisions (Luhmann 2000c, p. 142).
As Luhmann pointed out in his later writings (Luhmann 2000c), the
operative closure of organizations on the basis of decision communications
must not be misunderstood, in the sense that there are no other communi-
cations “in” organizations: there are, of course, also other communications,
such as gossip. These communications take place in the organization but
ultimately do not contribute to the autopoiesis of the organization.
Luhmann illustrates this idea with an example from biology:
In living cells there are also some minerals […] which do not take part in the
autopoiesis of the system, but which nevertheless serve important functions.
(Luhmann 2000c, p. 68; my translation)
b. Uncertainty absorption
Within organizations, decision communications are always integrated into a
process of connecting decisions – the actual autopoiesis of the organization.
Every decision is the product of earlier decisions and gives rise to ensuing
decisions. Luhmann describes this process of decisions connecting to each
40
04 Kap 1. Seidl 05-11-11 16.16 Sida 41
other using the concept of uncertainty absorption, the idea of which he takes
from March and Simon:
Uncertainty absorption takes place when inferences are drawn from a body
of evidence and the inferences, instead of the evidence itself, are then com-
municated. (March and Simon 1958, p. 165)
For a decision to be made, information is needed on the basis of which one

alternative can be chosen over the others. An investment decision, for
example, is based on information on the availability of financial resources,
on current interest rates, on current market demand etc. Formulating this
the other way around, one can say that a decision is “inferred” from the
given information. Yet, the important point is that no decision can rely on
complete information; some uncertainty inevitably remains. In our example,
there is uncertainty concerning future market demand, investment projects
of competing firms, future inflation figures etc. All this uncertainty, however,
is absorbed by the decision: all given information and all remaining uncer-
tainty is transformed into the selection of one alternative over the other
ones. Uncertainty absorption now takes place in the connection between
decisions. As decisions do not inform about the uncertainties involved in
making the decision – they merely inform about selected and excluded alter-
natives – ensuing decisions connecting to them cannot “see” the uncertain-
ties. That is to say, from the perspective of the connecting decisions orient-
ing themselves toward the first decision, the uncertainty of the first decision
is absorbed.
On the basis of such a processual understanding of decisions, we can dis-
tinguish between two “states” of a decision: before and after subsequent
decisions have connected to it. A decision is only completed when subse-
quent decisions connect to it. Before that, the decision is merely virtual
(Baecker 1999c 9, p. 138). The decision is virtual because the realisation of
the decision in subsequent decisions is expected but not yet realised. For
example, the organization decides to manufacture a particular new product
– in contrast to producing another new product or not producing anything
new at all. This decision is only virtual until subsequent decisions have com-
pleted it as a decision by orienting themselves according to it. The market-
ing division, for example, might decide on the advertisement of this new
product. This can be understood in analogy to the relation between differ-
ent communications, as described above, where a communication is only
completed once another communication connects to it by defining its mean-
ing retrospectively.
41
04 Kap 1. Seidl 05-11-11 16.16 Sida 42
c. Decision premises
A concept closely related to uncertainty absorption is that of decision prem-
ises, originally introduced by Herbert Simon (Simon 1957, p. 201). The con-
cept of decision premises refers to the structural preconditions that define –
or create – a decision situation; for example, the alternatives given, the
objectives of the decision, and so on. While one could include in the term
everything that influences the situation, Luhmann argues that such a concept
would not be very fruitful. Instead he restricts the term – in a first step – to
those structural preconditions that are themselves the “result” of other deci-
sions. In other words, a decision takes previous decisions as decision prem-
ises, or, formulated the other way around: every decision serves as a decision
premise for later decisions. With regard to the previous section we have
reversed our perspective: we are not looking at the transformation from the
viewpoint of the initially chosen situation towards the connection of subse-
quent decisions, but are looking “back” from the viewpoint of a decision
towards previous decisions and ask about their relevance to it. From this
viewpoint they serve as decision premises. To bring the concepts of uncer-
tainty absorption and decision premise together we can say: uncertainty
absorption takes place when a decision is used by subsequent decisions as a
decision premise.
An important aspect of the concept of decision premises is its double
function as both creating and restricting the decision situation. Decision
premises create the decision situation in the first place: they define the deci-
sion situation as such. Without decision premises there is no occasion for
decision making. At the same time, decision premises restrict the decision
situation by creating a particular decision situation and not a different one.
If decision premises define a decision situation as a choice between alterna-
tive A and alternative B, one cannot decide between X and Y.
The concept of decision premises becomes particularly interesting when
the concepts of decision and decision premise are applied recursively to each
other. Apart from the factuality of every decision becoming a decision prem-
ise for subsequent decisions, decisions can decide explicitly on decision
premises for other decisions, that is to say, they function as decisions on
decision premises. The crucial point of this is that a decision can decide on
decision premises which are not only binding for immediately succeeding
decisions, but for a multitude of later decisions. They serve as “a sort of
anticipated, generalised uncertainty absorption” (Luhmann 2000c, p. 261).
In this way decisions can influence other decisions that take place much later
in the decision process. Luhmann now suggests restricting the term decision
premise – in a second step – to those far-reaching decision premises. He dis-
tinguishes three types of such decision premises: programmes, communica-
tion channels and personnel.
Programmes are decision premises that define conditions for correct
42
04 Kap 1. Seidl 05-11-11 16.16 Sida 43
decision making; they are often also called “plans”. There are two different
kinds of programmes: conditional programmes and goal programmes.
Conditional programmes define correct decision making on the basis that
certain conditions are given. They generally have an “if-then” format – “if
this is the case, then do that”. Goal programmes, in contrast, define correct
decision making by defining specific goals that are to be achieved (e.g.
“profit maximisation”), and in this way structure the given decision possi-
bilities. Neither type of programme, however, removes the uncertainty from
the decisions that they bring forth – neither decides the decisions (otherwise
they would not be decisions). In the case of conditional programming there
is uncertainty about whether the conditions are actually met by the decision
situation – there is always some scope for interpretation. In the case of goal
programming the main uncertainty concerns the causal link between alter-
natives and the goal; for example, which alternative maximises profit. Apart
from that, there is in both cases uncertainty on whether the programmes
should actually be applied to the decision situation – reasons for making an
exception can always be found.
The decision premise of personnel concerns the recruitment and organiza-
tion of personnel. Organizations decide, on the one hand, on the com-
mencement and termination of membership and, on the other hand, on the
transfer of members to different positions within the organization, both with
and without promotion. Personnel is a decision premise insofar as it makes
a difference to the question of who is in charge of a decision. An experienced
manager is likely to “give rise” to different decisions from those of a new-
comer (this recognition of different individuals making a difference to the
organization does not contradict the concept of autopoiesis. Different indi-
viduals are only considered for the difference in perturbations that they
cause). In this sense, organizations have expectations about the behaviour of
different persons, which serve as a basis for selecting their personnel.
The decision premise communication channels concerns what can be
called the organization of the organization. Usually in an organization not
everybody can communicate with everybody at any one time, but the
communication is restricted to certain channels. The classic case is the hier-
archical structure, in which the communication channels only run vertically.
Decisions on one level only inform decisions on the next lower or next
higher level, but not decisions on the same level. That is to say, decisions can
only use other decisions on the vertical line as decision premises and not
ones on the horizontal line. Apart from the hierarchy, there exists a multi-
tude of other forms of communication channels – for example the matrix-
organization.
The three decision premises – programme, personnel and communication
channel – are coordinated through the creation of positions. Positions are
nodes at which the three decision premises meet and are specified with
43
04 Kap 1. Seidl 05-11-11 16.16 Sida 44
regard to each other. Every position executes a particular programme, is

filled by a particular person, and is located somewhere in the communica-
tion network. Positions coordinate decision premises in two respects: on the
one hand they coordinate them with regard to concrete decisions. On the
other hand they serve as an orientation for decisions on new decision prem-
ises: only such new decision premises can be integrated, which fit into the
existing structures of positions, or for which new positions can be created.
In his latest writings Luhmann (2000c) introduced another type of deci-
sion premise: the so-called undecidable decision premise. In contrast to the
decidable decision premise described above, these are premises that have not
been explicitly decided but are merely some sort of “by-product” of the deci-
sion process. These premises are undecidable since the organization does not
see their contingency and thus takes them as “necessary” and unchangeable.
There are two types of undecidable decision premises. The first one is the
organizational culture. Decision premises of this type refer to the way in
which an organization deals with its own processes of decision-making. For
example, if the organization always produces the same kind of decision (e.g.
recruiting merely male candidates) this might condense into an undecided
decision premise for future decisions – in the sense of “we have always done
it this way”. The second type of undecidable decision premise is the cogni-
tive routine, which refers to the way the environment is being conceptualised
by the organization. Cognitive routines, for example, inform about charac-
teristics of the customer.
d. The double closure of the organization

Like all autopoietic systems organizations can be said to be doubly closed
(on this point see particularly Baecker 1999c, p. 147 ff.): closed on the level
of their operations and closed on the level of their structures. The first clo-
sure refers to the conceptualisation of organizations as reproducing them-
selves exclusively on the basis of decisions. No external operations can take
part in the network of decisions nor can any decisions get out of this net-
work. In other words, on the basis of its operations, the organization has no
contact to its environment whatsoever. Decisions are only oriented accord-
ing to other decisions and nothing beyond the decision network. Thus, the
individual decisions are “blind” with regard to anything outside the organ-
ization. On this operative level the only thing that is important is the con-
tinuous reproduction of decisions out of decisions – regardless of which
decisions. As long as some decisions are produced, the autopoiesis of the
organization is continued.
The “blindness” of decisions is, however, compensated through the deci-
sion premises, i.e. the structures of the organization (structural level). These
decision premises (in particular: decision programmes, communication
channels, personnel) determine which decisions are produced. For example,
44
04 Kap 1. Seidl 05-11-11 16.16 Sida 45
a recruitment programme will lead to certain recruitment decisions and not

to decisions on the colour of wastepaper baskets. In other words, the deci-
sion premises channel the reproduction of decisions. In this sense, the orien-
tation according to decision premises serves the decisions as a substitute for
the orientation according to the environment. But even on this structural
level of decision premises the organization cannot get out of the decision
network; even there it does not get into direct contact with its environment.
This is the second closure. Decision premises are not in any way given from
outside but are themselves the product of decisions. Decisions and decision
premises are recursively reproduced – compare Giddens’s (1984) concept of
structuration. Thus, neither on the level of its operations (first closure) nor
on the level of its structures (second closure) is the environment directly
taken into account in the reproduction of decisions.
The double closure of the organization has a twofold implication. On the
one hand, double closure implies autonomy: it is the organization itself that
determines its own structures and operations. Without the ability to decide
on its own structures, the organization would be the mere continuation of
its environment. On the other hand, double closure implies unavailability.
As the organization can only operate on its “inside” and cannot distance
itself from itself (i.e. it has no other mode of operation than decisions), it is
captive of its own processes and thus does not have (complete) control over
itself.
e. The paradox of decision at the heart of Luhmann’s

organization theory
In this last section on organization I want to come back briefly to the
paradox of decision in order to highlight how crucial it is for Luhmann’s
way of theorizing about organizations. Particularly in his later writings,
(1993d; 2000c) Luhmann makes the paradox of decisions the starting point
for unfolding his entire organization theory. Independently of the con-
ceptualisation of decisions as communication, Luhmann argues that the very
idea of decisions is paradoxical. In this respect, he quotes Heinz von Foerster
(1992, p. 14), who famously wrote: “Only those questions that are in
principle undecidable, we can decide” – everything else would be mere
calculation. In other words, in a real decision situation the given alternatives
are all equally valid; there are no better or worse alternatives – otherwise
these would not be real alternatives. If the “alternatives” were of different
value (in which case they would not be real alternatives) there would be no
need to decide between them any more – the decision situation would have
already been decided. In a real decision situation with real alternatives, how-
ever, there is no valid reason for choosing one alternative over the other
– otherwise the alternatives would not be of equal value. Thus, at the heart
of every decision there is undecidability.
45
04 Kap 1. Seidl 05-11-11 16.16 Sida 46
In order to prevent a paralysis of decision situations, the paradox of deci-

sion has to be deparadoxified; i.e. the paradox has to be deferred to another
place. This deferral does not mean that the paradox disappears but that it is
just moved “out of sight”. For example, the undecidability of the decision
might be shifted to the selection of a decision rule; e.g. “choose the alterna-
tive that is least risky”. This decision rule might allow a clear ranking of the
different alternatives and in this way the decision rule can be said to
“decide” the original decision situation. In this case, the original decision
paradox has been deferred to the decision on which decision rule to choose.
There exist many different decision rules, which themselves constitute alter-
natives between which one needs to decide. Again, this decision situation is
itself undecidable and thus paradoxical. The paradox might thus have to be
deferred to yet another place, e.g. to the decision about a meta-decision rule.
Ultimately, the paradox can only be deferred but never solved. This might
lead to an infinite regress, unless the paradox ends up in a place where it is
not “noticed”. For Luhmann most organizational phenomena can in one
way or another be traced back to this undecidability of decisions: most
problems in organizations are directly or indirectly linked to the decision
paradox, and most structures and processes function as a means of depara-
doxification.
Luhmann’s theory of social systems as a theory

of distinction
In his writings – particularly the later ones (e.g. Luhmann 2000c) – Luh-
mann drew heavily on the calculus of distinctions, The Laws of Form, by the
British mathematician George Spencer Brown. This calculus allowed him to
describe the self-referential logic of autopoietic systems in an extremely
stringent and analytical way. While it is possible to comprehend Luhmann’s
theory of social systems also without Spencer Brown, a deeper appreciation
of it presupposes an at least rudimentary familiarity with Spencer Brown’s
basic ideas. Some of Luhmann’s followers have even tried to translate
Luhmann’s theory completely into the language of the calculus, aiming at
making the theory of social systems “calculable” (in particular Dirk
Baecker).
In the following, we want to introduce the central elements of Spencer
Brown’s Laws of Form and explain the way in which Luhmann used it in his
systems theory. It should, however, be noted that our explanations of
Spencer Brown are based on Luhmann’s own reading of him, which differs
from other readings.
46
04 Kap 1. Seidl 05-11-11 16.16 Sida 47
a. Observation as basic concept

Spencer Brown suggests treating observation as the most basic concept of
any analysis. As a concept it is supposed to be even more basic than e.g. that
of thing, event, thought, action or communication. This means, of course,
that the concept of observation is not restricted to its usual sense of optical
perception – optical perception is just one type of observation, the observa-
tion of psychic systems. Instead, its level of abstraction is such that it refers
to any operation from communications to thoughts and even to operations
of machines; even the observer himself is an observation.
Spencer Brown’s concept of observation does not focus on the object of
observation but on the observation itself as a selection of what to observe.
In this sense, the underlying question is not: what does an observer observe,
but how does an observer observe; how is it that an observer is observing
what he is observing, and not observing something else.
Every observation is constructed from two components: a distinction and
an indication. An observer chooses a distinction with which he demarcates
a space into two spaces (synonymously: “states” or “contents”). Of these
two states he has to choose one that he indicates. That is to say, he has to
focus on one state, while neglecting the other. It is not possible for him to
focus on both. In this way, the initially symmetrical relation between the two
states becomes asymmetrical. We get a marked state and an unmarked state.
Spencer Brown illustrates this rather abstract idea with an example: let us
imagine a uniform white piece of paper. On this paper we draw a circle. In
other words, we draw a distinction which creates an “inside” of the circle
and an “outside” of the circle. It is important to note that it is the act of
drawing the circle which establishes the two different states: without us
drawing the distinction, the two states as such do not exist. We can now
indicate one of the two states: either the inside or the outside. Let us choose
the inside. The inside becomes the marked state and the outside the
unmarked state. While we can see the marked state, the unmarked state
remains unseen. With the metaphor of figure and ground we can say: the
inside becomes figure and the outside ground.
Spencer Brown calls the distinction with both sides the form of the dis-
tinction. Thus, in contrast to the common use of the term, form does not
refer merely to the marked state. The form of something is not sufficiently
described by the defined – the marked state – but the unmarked state is a
constitutive part of it. The marked side cannot exist without its unmarked
side. In our example, the form of the circle is the inside together with the
outside of the circle. In this context, Spencer Brown declares:
Distinction is perfect continence. (Spencer Brown 1979, p. 1)
A distinction, thus, has a double function: like any boundary it both distin-
guishes and unites its two sides.
47
04 Kap 1. Seidl 05-11-11 16.16 Sida 48
Spencer Brown introduces a specific notation to refer to the distinction,

“the mark of distinction” or the “cross” (synonymously: token, sign, mark):
Marked state Unmarked state
Figure 2: The mark of distinction.
This sign symbolises the distinction separating the two sides. Connected
with this sign is the instruction to cross the boundary from the right to the
left side; a process by which the left side becomes the marked state and the
right side the unmarked state.
It is important to understand that the “cross” has two meanings: an oper-
ative and a descriptive meaning. Firstly, the cross stands for an instruction
to cross (!) the distinction from unmarked to marked state. Secondly, the
cross stands as a sign for the result of crossing; the marked state. In our
example, the cross can be meant as an instruction to draw a circle or stand
as a symbol of the result of drawing, i.e. stand for the circle itself. In this
context, Spencer Brown writes:
In the command
let the crossing be to the
state indicated by the token
we at once make the token doubly meaningful, first as an instruction to cross,
secondly as an indicator (and thus a name) of where the crossing has taken
us. (Spencer Brown 1979, p. 81)
In terms of the calculus, the cross is used both as operator and operand: on
the one hand, it gives instructions to calculate and, on the other hand, it is
the element that is calculated. This double meaning might be confusing, but
as Spencer Brown writes:
It is the condensation [of the two meanings into one symbol] which gives the
symbol its power. (Spencer Brown 1979, p. 81)
Another important element of the Laws of Form is the “unwritten” dis-

tinction (unwritten cross), which defines the space – context – within which
the distinction is drawn. In our example from above, the unwritten cross is
constituted by the border of the paper. As the border of the unmarked state,
it remains as unobserved as the unmarked space itself. Now we have a com-
plete unit of observation: a space defined by an unwritten cross is divided by
a distinction into two states; the relation between the two states becomes
asymmetrical by indicating one state as the marked state in contrast to the
other, which becomes the unmarked state (Figure 3):
48
04 Kap 1. Seidl 05-11-11 16.16 Sida 49
Unwritten cross
Marked state Unmarked state
Figure 3: Observation as distinction and indication.
The central point in this concept of observation is that once you have drawn
a distinction you cannot see the distinction that constitutes the observation
– you can only see one side of it. As Heinz von Foerster (1981, pp. 288–309)
pointed out, this can be referred to as the “blind spot” of observation. The
complete distinction with both its sides (the inside and the outside), can only
be seen from outside; if you are inside the distinction you cannot see the dis-
tinction.
We can now distinguish two orders of observation: first-order and second-
order observation (von Foerster 1981). So far we have been explaining the
operation of a first-order observer, who cannot observe the distinction he
uses in order to observe. The second-order observer is an observer who
observes another observer. He uses a different distinction from the first-
order observer: in order to observe the observer, he has to draw a distinction
that contains the distinction (the marked and the unmarked state) of the
first-order observer in his marked state. The second-order observer can see
the blind spot – the distinction – of the first-order observer. He can see what
the first-order observer cannot see and he can see that he cannot see.
Particularly, he can see that the first-order observer can see what he sees,
because he uses one particular distinction and not another. He sees that he
could also have used another distinction and, thus, that the observation is
contingent. In this sense, a second-order observation is more than a first-
order observation, because it not only sees its object – the first-order observ-
er – but it also sees what he sees, and how he sees; and it even sees, what he
does not see, and sees, that he does not see, that he does not see, what he
does not see (Luhmann 1993k, p. 16).
Since the second-order observer needs a distinction to observe the distinc-
tion of the first-order observer, he himself is a first-order observer, who
could be observed by another second-order observer. In this sense, every
second-order observation is only possible as a first-order observation and as
such knows as little about its own observation as every other first-order
observer.
For Luhmann, the most interesting element of Spencer Brown’s calculus of
form is the re-entry describing the operation of self-observation. As ex-
plained above, an observer can only observe the marked side, and not the
unmarked side or the distinction itself. In order to observe the other side he
49
04 Kap 1. Seidl 05-11-11 16.16 Sida 50
would have to “leave” the marked state and cross to the other side of the
distinction. This, however, would mean that it would no longer be possible
for the initially marked state to be observed; one can either observe the one
or the other side of the distinction but not both at the same time. As such,
self-observation, that is to say, observing one’s own observations, would be
impossible. Spencer Brown’s “solution” to the problem is the re-entry of the
distinction into the distinction; i.e. the original distinction contains a copy
of the distinction (with marked and unmarked states) in its marked state.
This, however, constitutes a paradox: the unmarked state is both unmarked
state and marked state (as it is contained in the marked state) and the
marked state is both marked state and unmarked state (as it contains the
unmarked state). In other words, the observer can see his blind spot, but
then, if he can see it, it is not his blind spot any more. Spencer Brown
unfolds this paradox claiming that the re-entered distinction is never exact-
ly the same as the original distinction.
b. Autopoietic systems as distinction processing systems

On the basis of the calculus of forms, Luhmann described autopoietic sys-
tems as distinction-processing systems. Every operation of an autopoietic
system constitutes an observation, i.e. a distinction and indication. Take, for
example, communications: every communication communicates something
(marked state) while at the same time having to leave everything else out – in
particular, all other possible communications (unmarked state). These other
possibilities of communication, however, are not just other communicative
options, which just happen not to have been realised, but they are consti-
tutive for the realised communication. That is to say, the meaning of the
communication depends to a large extent on what has not been communi-
cated. In other words, one needs to know what could have been communi-
cated (i.e. the context of the communication) in order to establish the mean-
ing of the communication. These other possibilities are on the unmarked
side of the communication, as they have not been communicated. Other
communications (second-order observation), however, can communicate
about the communication (first-order observation) and its unmarked state,
but only at the cost of producing yet another unmarked state. Thus, the
communication can never fully communicate about its own conditions of
communication.
Not only the operations but also the system itself can be conceptualised as
an observation, i.e. as a distinction and an indication. A system is consti-
tuted as a distinction between system and environment, of which the system
is the marked state and the environment the unmarked state. In accordance
with Spencer Brown’s concept of observation, the system and the environ-
ment are the two sides of the same distinction and as such are constitutive
for each other.
50
04 Kap 1. Seidl 05-11-11 16.16 Sida 51
While an observer can draw his distinctions where he likes and thus define
what to treat as a system and what as an environment, the concept of
autopoiesis assumes that the system/environment distinction is not drawn by
an external observer but by the system itself. Luhmann writes in this respect:
If we describe [something] as an autopoietic system, we are dealing with the

production and reproduction of a distinction (in terms of systems theory: the
distinction of system and environment), and the concept of autopoiesis says
that an observer using it assumes that the difference is produced and repro-
duced by the operations of the system itself. (Luhmann 2000c, p. 55; my
translation)
How are we to understand this reproduction of the system/environment dis-

tinction? Every operation of an autopoietic system constitutes a distinction
between that which it is, i.e. an operation of the system, and that which it is
not, i.e. an operation of the environment. Let us take the organization, for
example: the operations of that system are decisions (decision communica-
tions). Every decision constitutes a distinction between that which it is
(marked state), i.e. a decision and thus an element of the organization, and
that which it is not (unmarked state), e.g. a “normal” communication or a
thought. In this sense, every single decision (re-)draws the distinction
between system and environment. Thus, in actual fact, the reproduction of
decisions is the reproduction of the distinction decision/non-decision; that is
to say, of the distinction organization/environment. According to this con-
ceptualisation every single operation of a system reproduces the “boundary”
of the system. In this sense we do not distinguish between “boundary ele-
ments” and elements taking place “inside” the boundary, as the classical
notion of systems suggests.
This conceptualisation of the system’s boundary, as reproduced by every
single operation, implies an operative closure of the system: every operation
constitutes a distinction between the operation and everything else (i.e.
between system and environment). It can only be this operation as the one
side (marked state) of the distinction and not the other one (unmarked
state). For instance, a decision is only a decision (marked state) to the extent
that it is not something else (unmarked state). In this sense, operative clo-
sure of a system means that the system (i.e. system/environment distinction)
is only reproduced by operations that are themselves constituted as a
system/environment distinction. For example, the “decision system/environ-
ment” distinction (organization) can only be reproduced by operations con-
stituted as decision/non-decision distinctions and not by other distinctions
– for example thought/non-thought (decision is here included in the
unmarked space). The integration of other distinctions, in which decision is
included in the unmarked space, would dissolve the organization/environ-
ment distinction and thus dissolve the system. In other words, the system
51
04 Kap 1. Seidl 05-11-11 16.16 Sida 52
cannot enter its environment nor could the environment enter the system,
otherwise the distinction between system and environment would disappear.
While the system can only operate on the marked state of the system/envi-
ronment distinction, other observers outside the system might observe the
system/environment distinction by including this distinction in their marked
state. Consider, for example, an organization: the organization as a system
of decisions is constituted by the distinction “decision network/social envi-
ronment”. While the organization can only operate on its inside, – that is to
say, it can only produce and reproduce decisions and cannot enter its envi-
ronment, which consists of all kinds of other communications – the societal
system, which consists of the distinction “all-encompassing social system/
non-social environment”, contains the organization/environment distinction
in its marked state. Society can thus observe the distinction of the organiza-
tion and can thus see what the organization itself cannot see.
Although autopoietic systems can only operate on their inside (marked
state) and have no contact to their outside (unmarked state), the system/
environment distinction can re-enter the system. We can distinguish two re-
entries: first, every single operation distinguishes between other operations
of the same system and other events outside the system. In other words,
every operation has a self-referential aspect and an other-referential aspect.
Take, for example, communications as elements of a social system. Every
communication can be divided into, on the one hand, the utterance, i.e. how
and why something is expressed, which is (treated as) determined by the
communication system (self-reference), and on the other hand, the informa-
tion, i.e. what is expressed – (treated as) referring to events in the environ-
ment (other-reference). For example, A says to B: “My dog is dead”. Here
we can distinguish the utterance (i.e. the words A uses, what other commu-
nications this communication is referring to etc.) as the self-referential
aspect, and the information about a dog being dead as referring to some-
thing outside the communication network (other-reference). The important
point here is that the re-entered distinction is not identical with the distinc-
tion itself: (1) the utterance/information distinction is not the system/envi-
ronment distinction – a communication is not a system – and (2) the infor-
mation about the dog being dead is not the dead dog.
A second re-entry takes place on the structural level of the system.
Structures “represent” internally the system/environment distinction to the
system. As explained above with regard to organizations, the operations of
a system cannot observe their environment. Instead, they observe the
system’s programmes as a substitute for the environment and orient them-
selves according to them. Take, for example, a business programme of a
corporation. This programme refers, on the one hand, to the market situa-
tion, possible moves by competitors, characteristics of consumers, or some-
thing similar, and on the other hand, to the necessary decision processes in
52
04 Kap 1. Seidl 05-11-11 16.16 Sida 53
the organization. By taking the programme as a decision premise decisions

orient themselves according to the two aspects of the programme as if to the
organization/environment distinction itself. Again, here we have to note that
the programme represents the system/environment distinction but is not
identical to it.
These few comments on the application of Spencer Brown’s calculus of
distinctions to Luhmann’s systems theory have to suffice for now. While our
descriptions and explanations have been very selective, the basic ideas
should hopefully have become clear so far.
Conclusion
In this chapter we have tried to present the basic concepts and ideas of
Niklas Luhmann’s theory of social systems. We started with Luhmann’s
general concept of autopoiesis explaining how it was derived from the origi-
nally biological concept by Maturana and Varela. We went on to explain the
concept of the social system as an autopoietic system of communications,
where communications reproduce communications. We highlighted as one
of the central ideas in this context the clear distinction between social and
psychic systems. We have tried to clarify this often misunderstood idea.
From there we went on to describe the three types of social systems – socie-
ty, interaction and organization. In our last section we introduced the cal-
culus of distinction by Spencer Brown and demonstrated how it could be
and has been applied to the theory of social systems.
53
05 Kap 2.Luhmann 05-11-11 16.17 Sida 54
Chapter 2
The Concept of Autopoiesis

Niklas Luhmann
For several years now, there has been a lively discussion about
“autopoiesis”.1 Unlike in physics, however, in the realm of knowledge,
sound, as Jean Paul wrote, moves faster than light.2 Therefore, the word
“autopoiesis” is indeed familiar but the discussion is far from understand-
ing the concept sufficiently. And in turn, one encounters formulations with-
out the word “autopoiesis” that say the same thing but whose significance
is not recognized.3 In the following we want to outline briefly our under-
standing of the conceptual complex of self-reference, autopoiesis, and opera-
tive closure in general and with regard to organizations in particular.
The theory of self-referential systems abstains from determining its object
(in our case, organizations) by means of assumptions about its essence.
Experience shows that such assumptions lead to irresolvable differences of
opinion as soon as different observers offer different definitions of that,
which they take for the essence of the matter – regardless of the matter at
issue, be it the essence of law, of politics, of the family, of religion, or, actu-
ally, of organizations. Therefore, we begin with a circular definition: an
organization is a system that produces itself qua organization. Now, we only
have to define in what way this happens. However, this next step requires a
series of theory decisions that could be made differently if it were possible
to show how an alternative of the same quality would look.
1 For a topical survey, see especially Mingers (1995). Cf. also Robb (1989); Fischer (1991), in’t
Veld et al. (1991), Bardmann (1994), passim but esp. pp. 72 ff. and on the connection with
the discussions about “organization culture” pp. 365 ff.; Bailey (1994), pp. 285 ff. The pub-
lication of individual essays can barely be surveyed any longer. On the application to organi-
zations cf., e.g., Kirsch and zu Knyphausen (1991); Kickert (1993); Willke (1994a); Wollnik
(1994).
2 Paul (1961): “In the realm of knowledge – different from the physical realm – sound always
arrives earlier than light” (1014).

3 Thus one reads, “Any enterprise’s first product is itself,” in Bausor (1994), p. 181. There is
also talk about “autogenesis”; cf. Drazin and Sanderlands (1992). If one is referring to the
Greek sense of these words, however, then it is preferable not to proceed from an “origin” but
from the “product”. For a system is its own origin, only insofar as it is its own product. The
question concerning the origin is better left to theology.
54
05 Kap 2.Luhmann 05-11-11 16.17 Sida 55
Here, we are going to offer a concise synopsis that lists the most impor-
tant of these assumptions that have the effect of concretization and break
the circle:
1. The basal unit of an autopoietic system has the temporal form of an
event – that is, of an occurrence that makes a difference between
“before” and “after” and therefore can be observed only if one’s obser-
vations are based on the distinction before/after. Whenever we are con-
cerned with results, we also speak of the “operation”, and in the case of
organizations of the “decision”. In the context of a comparison of theo-
ries, it is important to keep in mind this foundation on events (and not
on substances).4 From this it follows that the theory proceeds from the
presumption of discontinuity, the presumption of a steady decay, and
takes continuity (thingness, substance, process) to be in need of an
explanation.5 A theory of autopoietic systems constructed in this man-
ner finds itself in radical opposition to all types of process theories,
including the dialectical ones. Such a theory rejects any kind of “essen-
tialism” and requires, on the contrary, that every event (or in our area:
every decision) leave all that follows to a subsequent event. Forms of
essence are but instructions for a repetition of the selection. The theory
of autopoiesis also stands in opposition to theories of action. For theo-
ries of action revert to the ideas (intentions, purposes) of an actor in
order to connect their “unit acts”. By contrast, events – e.g., communi-
cations – that constitute autopoietic systems produce surpluses of possi-
bilities so that in a further step something suitable may be selected. It is
not necessary that the selected possibility was anticipated; the decision
about this selection is made typically and better in retrospect, in light of
an event that has already taken place.
2. A system that has produced itself must be capable of observing itself
– that is to say, it must be capable of distinguishing itself from its envi-
ronment.6 Occasionally, this is disputed. But since “organization” can-
not mean the whole world, it is necessary to provide a criterion that
serves to delimit that which is designated as an organization. Under
these circumstances, the theoretically decisive question is whether this
delimitation is put into effect by the organization itself or not; in the
negative case, the question is, who or what else would put it into effect.
4 For a rather rare conception of this kind see Allport (1940; 1954; 1967). The relations to
Alfred North Whitehead’s cosmology are quite obvious.

5 Ingold (1986), p. 24 provides the appropriate formulation: “Process is to event as continu-
ity is to discontinuity.”
6 This requirement, taken by itself, need not lead to a theory of autopoiesis. Similar discus-
sions can be found in the context of a distinction between “matter” and “symbol”. See, e.g.,
Pattee (1982). But in such a case, these concepts must be clarified, especially in regard to the
concept of reference.
55
05 Kap 2.Luhmann 05-11-11 16.17 Sida 56
3. In the process of self-observation, an organization does not observe itself

as a stationary object whose qualities can be recognized. Instead, the
organization uses its own identity only for the purposes of continuously
attaching new determinations to it and subsequently giving them up
again.7 For this reason, autopoietic systems can also create variations in
their structures (this is called “self-organization”), insofar as such varia-
tions are compatible with the continuation of autopoiesis. All reflections
on identity that propose stable self-descriptions by means of content-
related properties must therefore proceed in a highly selective manner; in
the process, they commit themselves to exacting normative demands and
usually remain controversial.
4. The variability of the “self” that is introduced in each case as the fixed
point is guaranteed by the fact that the organization observes itself
observing. Already the organization system operates at the level of
second-order observation; it diagnoses its own observations continuous-
ly (albeit not in every individual case). The theory or organization must
therefore be placed at the level of third-order observation. Such a theo-
ry observes a system that observes itself; in consequence of this arrange-
ment, it is capable of extending its observations to circumstances that
are inaccessible to self-observation. Here, we touch on the classical
sociological problem of latent structures and functions.8
5. Accordingly, autopoiesis is possible only as long as the system finds itself
in a constant state of uncertainty about itself in relation to its environ-
ment, and as long as it can produce and control this uncertainty by
means of self-organization. The system cannot transform the built-in (we
will also say: the self-produced) uncertainty into certainty. The absorp-
tion of uncertainty can happen only as a transformation of the form of
uncertainty that is relevant in each moment. Such absorption happens as
an adjustment to the changing states of perturbation. Even reflection
and self-observation cannot change this. Every “transcendental” identi-
ty could endanger the continued self-reproduction of the system.
6. The best possibility of coming to terms with uncertainty is to stick with
all that has already happened. For this reason, organizations clarify the
meaning of their actions to a large extent retrospectively. This, however,
tempts them to pay little attention to the state of their environment at
any given moment. This differentiation at the operative level must be
balanced at the structural level. Preferably, decisions about the appro-
priateness of structures (such as decision programs or the typical length
7 For the same argument regarding the self-consciousness of psychic systems cf. Churchland
(1984), p. 73: “… self-consciousness involves the same kind of continuously updated knowl-
edge that one enjoys in one’s continuous perception of the external world.”
8 For a synopsis see Merton (1957), pp. 60 ff.
56
05 Kap 2.Luhmann 05-11-11 16.17 Sida 57
of time of operations) will therefore have to be made with regard to the

environment.
7. Concepts such as self-reference, self-observation, and self-description
presuppose operations that produce in reality that which they mean.
These operations must take place in the system (where else?). If one also
considers that this is not possible in the form of unconnected, singular
events, one comes up against the problem of the recursive interlacing of
these operations. In order to make itself possible, every operation pre-
supposes the recourse to and anticipation of other operations of the
same system. Only in this way can connections be identified and the
boundaries in relation to the environment be produced and reproduced.
It is for this very reason (aside from the fact that the biologist Humberto
Maturana gave this state of affairs this name) that one speaks of “auto-
poietic” systems.9 Therefore, a first application to the case of organiza-
tions was preventively and explicitly marked as a “metaphor”.10 A con-
cept of cognition that was too broad, and a tie to biochemical realiza-
tion that was too tight often influenced and misguided the subsequent
discussion.
8. In light of the complex and often confused discussion, a few explanato-
ry remarks on this topic are in order:
(a) As the concept of “poíesis” indicates, we are dealing with the pro-
duction of something: namely, the creation of the system as its own
product. Of course, this does not mean that the system itself has at
its disposal all of the causes that are necessary for self-production. A
causal theory simply has no place for such a combination of all
causes in a system (except in God). For concepts such as product,
production, and reproduction this already applies at the conceptual
level. In fact, only if a system does not simply exist but must repro-
duce itself by means of its own products can it depend on its envi-
ronment in precisely this sense. However, it is important that the sys-
tem has at its disposal a sufficient range of disponible causes (e.g., in
the case of an organization this would be members that are subject
to the instructions of the organization), so that it can secure its own
reproduction under regular circumstances.
(b) The concept “poíesis” does not emphasize the rule-bound, and even
less the absolute, certainty of production. Instead, it emphasizes
reproduction – that is to say, the production by means of its own
products. Following Heinz von Foerster, one can therefore also
speak of a “historical machine” – that is to say, of a system that
9 See especially Maturana and Varela (1975 and 1987). At this stage, the development of the
theory was strongly conditioned by epistemological questions and presupposed biochemical

processes insofar as the autopoietic operation itself was concerned.
10 In Morgan (1986).
57
05 Kap 2.Luhmann 05-11-11 16.17 Sida 58
produces its subsequent operations in the state in which it finds itself

due to its previous actions.
(c) According to all this, the concept of autopoiesis is defined formally.
For this reason, it leaves open qua concept the material operations
by which the autopoiesis is executed. These operations may be bio-
chemical or neurophysiological; they can also be conscious attentive
dispositions or communications. Here, we are neither arguing by
means of analogy nor metaphorically.11 Rather, it is a matter of dif-
ferent applications of a general theory.
(d) The mere concept of autopoiesis serves as the distinction and indica-
tion of a corresponding state of affairs. As a concept, it has no
empirical explanatory value. The effect of this concept consists first
and foremost in constraining and adjusting other concepts, e.g., the
concept of evolution or the understanding of the relationship
between system and environment. All further consequences depend
on the operations and the structures resulting from evolution and
learning, by means of which autopoiesis is materialized.
(e) The autopoiesis of the system is realized at the level of operations.
Therefore it is compatible with all structures that make a connection
from one operation to another possible. Here, the concept of struc-
ture is a concept that correlates with autopoiesis, not, as is usual in
another context, a concept that correlates with the division of labor.
Structures are produced (and reproduced or perhaps varied or even
forgotten) by operations for the use in operations.12 For this reason,
one cannot infer structural conservatism from the concept of
autopoiesis.13 On the contrary, it is precisely the closure vis-à-vis the
environment that offers the system opportunities for structural
11 Besides, it would cause no damage but be only of little use if we agreed that a metaphor is
at stake in our use of the term “autopoiesis”. For this would apply to all approaches to orga-
nization theory (see, e.g., Morgan 1980); and, since the concept of metaphor (from “meta-
pherein”) is a metaphor, it would merely amount to the requirement that every universalistic
theory must be reminded of the necessity of an autological self-foundation.
12 We find ourselves in full agreement here with Anthony Giddens’s theory of “structuration”
– with the single exception that Giddens rejects a systems-theoretical foundation of this con-
cept. See Giddens (1986).
13 As do, apparently, Kickert (1993) and many others who associate “autopoiesis” with a con-
servative ideology. Management consultants also tend to describe autopoietic systems as struc-
turally conservative in order to supply arguments in support of the function, if not necessity,
of specific interventions from the outside. See, e.g., Wollnik (1994). The thesis regarding struc-
tural conservatism does not teach us something about the theory of autopoietic systems but
rather about those entities that present this thesis: that is, about the autopoiesis of firms and
the educational institutions of management consulting; alternatively, we could say with
Maturana, loc. cit., p. 64: it does not tell us anything about the area that is being described
but rather about the observer who produces or uses such a description.
58
05 Kap 2.Luhmann 05-11-11 16.17 Sida 59
variation, which it could not have under the condition of immediate

linkages with the environment. Autopoietic modes of operations are
typical one-time-only inventions of evolution that have tended
towards structural diversification in the course of history. Thus
“autopoiesis” merely designates a barrier for structural variations.
However, as evolution extending over long times and expansive
spaces teaches us, it is precisely the difference between inner and
outer that leads to the acceleration of changes. Of course, it also
teaches us that these changes do not accord with the wishes of cer-
tain observers, which has the effect that to them, these systems may
appear as rigid and inflexible.
(f) The recursive interlacing of operations follows neither logical nor
rational rules. It only creates connections and holds out a prospect
of connectivity. Thus, sales figures may be taken as proof of success
and of the quality of a given organizational structure. Thus, infor-
mation may trigger the suspicion of distortions caused by specific
interests and motivate further attempts at confirming this suspicion.
Thus, in the international relations between organizations ecological
criteria for products may be interpreted as trade barriers. Hence
recursions actually secure the preservation and reproduction of
repressed paradoxes. Something determinate is always also some-
thing different.
9. Autopoietic systems are operatively closed and precisely in this sense
autonomous systems.14 The concept of operative closure does not
allow for any gradation; in other words, it does not permit that the
system operates in its environment or, vice versa, the environment in
the system. A system cannot be more or less autopoietic;15 but it can
be more or less complex. Operatively closed autopoietic systems can-
not be described via input/output functions; this much is clear for
mathematically demonstrable reasons.16 From this result descriptions
14 Cf. for a very narrowly circumscribed segment Varela (1979).

15 This is a contentious point. Especially Gunther Teubner (1987a; 1987b) has introduced the
opposing viewpoint into the discussion. For the application to organizations see Kirsch and zu
Knyphausen (1991). If one decided to take this path in the formation of concepts, one would,
however, need a conception of the unity of the system that can be produced without depend-
ing on the concept of autopoiesis. Moreover, this version would have to abandon the rigid con-
nection between autopoiesis and decision in the case of organizations. Then, “autopoietic”
systems would be systems in which autopoiesis also happens (among other occurrences). This,
however, would not explain the unity of the system. Perhaps in such cases it would be better
to replace the concept of autopoiesis by the old concept of circular causality. In any case, from
Maturana’s point of view, reflections on gradation pertain exclusively to the area of the struc-
tures of systems and precisely not to the area of autopoiesis itself.
16 On this topic and the reasons it provides for the necessity of a “blind spot”: that is, the
necessary opacity of the system to itself, cf. Heinz von Foerster (1993a, pp. 21 ff.; 1993b).
59
05 Kap 2.Luhmann 05-11-11 16.17 Sida 60
and impressions such as: freedom, arbitrariness, and opacity. The con-
cept of operative closure also abstracts from causal assumptions. In no
way does it claim a causal isolation (even if it were merely relative). It
is possible that a system (say, a brain) is operatively closed and at the
same time depends to an extreme degree on the continuous supply of
resources of a very specific kind (in our case this would be the circula-
tion of the blood). Operative closure merely means that the system can
operate only in the context of its own operations and that it depends
in this process on the structures that are being produced precisely by
these operations. It is in this sense that one speaks of self-organization
or, with reference to operations, of structural determination.
10. These theory stipulations have far-reaching consequences for the rela-
tion between system and environment. In this case, operative closure
does not mean that an organizational system cannot maintain any con-
tacts with its societal environment. Society, on the one hand, makes it
possible to communicate within society across the boundaries of sub-
systems. On the other hand, an organization cannot participate in
communication without observing itself as a participant. Qua recipi-
ents of communications, the organization’s own structures regulate the
information by which the system lets itself be perturbed and incited to
undertake its own information processing. Qua sender of communica-
tions, the organization decides what it wants to communicate and
what not. To this degree, the organization’s environment remains the
organization’s own construction, which is not to say that its reality is
denied. On this point, we agree with Karl Weick.17 Whatever is
observed in the organizational system as environment is always itself a
construction: that is, a filling-in of the external reference (other-
reference) of the system.18 In a manner of speaking, the environment
validates the decisions of the system by providing the context that
makes possible to determine retrospectively how one has decided
(Weick speaks of action). The environment makes possible the exter-
nalization of unpleasant causes of one’s own decisions – that is to say,
it makes possible a sort of “punctuation” of one’s own operations.
Thus, the environment is a collecting area for problems that allows the
system to disregard its own participation in the creation of these prob-
lems. To sum up, one might say that the environment provides the pos-
sibility of referring one’s own operations to a “niche” without posing
the question of how it is that the world and society in particular con-
tain such niches. Nothing else is expressed by the old concept “milieu”.
17 See Weick (1977a; 1979, pp. 147 ff). Cf. also Smircich (1983), pp. 229 ff.
18 “The ‘outside’ or ‘external’ world cannot be known,” one reads in Weick (1977), “The out-
side is a void, there is only the inside” (p. 273).
60
05 Kap 2.Luhmann 05-11-11 16.17 Sida 61
11. Although biology bestowed the concept of autopoiesis upon us, we can
safely leave open whether or not and in what way one could under-
stand the reproduction of relatively stable large molecules in cells as
“autopoietic.” Perhaps one might do so, because such reproduction is
possible only in cells, or perhaps because we are dealing here with
extremely unstable units that must be continuously replaced. In the
case of social systems, autopoiesis can be recognized much more easi-
ly, and it is structured quite differently, in any case. For we are not
dealing here with units that need to replicate and must be replaced con-
tinuously. Rather, social systems (and conscious systems as well) con-
sist merely of events that are about to vanish as soon as they have come
about and must be replaced not by the same but by different events.
The steady transition from one element to another, the steady repro-
duction of difference, can in fact be understood only in terms of
autopoiesis; for it presupposes the connectivity that is produced in and
by the system itself. No environment could inject something fitting for
the system as quickly as is necessary. Only the system can arrest its own
decay as it happens from one moment to the next. At the same time,
this situation makes very specific demands on structures; repetition is
precisely what they must not aim at; instead, they first and foremost
must regulate the transition from one element to the next. To this pur-
pose, a meaning that is rich in reference but nonetheless determinable
must provide the necessary orientation, as I have explained else-
where.19
12. Autopoiesis depends on the fact that a system is capable of producing
internal improbabilities and thereby deviating from the usual. In such
a case, structurally restricted contingencies function as information in
the system. In fact, they function as information that is not derived
from the environment, since the system cannot contact its environ-
ment. At best, they function as information about the environment
(and even this is not the case in biological systems such as cells,
immune systems, and brains but only in systems that can distinguish
between themselves and their environment in the medium of meaning).
Thus, an autopoietic system can only inform itself; and in the system,
information has the function of selectively restricting the possibilities
for the continuation of its own operations combined with the addi-
tional function of being able to decide relatively quickly about connec-
tive possibilities.
13. Closure in this operative sense is the condition of a system’s openness.
Especially in regard to the law of entropy in thermodynamics, the older
type of systems theory spoke of “open systems” in order to be able to
19 See Luhmann (1995f), pp. 59 ff.
61
05 Kap 2.Luhmann 05-11-11 16.17 Sida 62
explain how order is created and preserved through a contrary ten-

dency. But the question of what actually makes a system capable of
being open was not posed. In other words, one did not ask what degree
of systemic order had to be given so that a system could afford its own
openness and, if necessary, even increase the complexity of those
aspects that buttress its openness. This question was not asked because
empirical examples and the input/output model, respectively, provided
the orientation of this theory. The older type of cybernetics had already
spoken of systems that are closed in regard to information but open in
regard to energy. But only the more recent theory of self-referential sys-
tems states clearly that operative closure is the condition of openness.
14. The theory of autopoietic systems distinguishes strictly between the
continuation of autopoiesis and the preservation of certain structures
that serve the safeguarding of sufficient redundancy and connectivity
and thus, in one way or the other, make autopoiesis possible in the first
place. Thus, structures are presupposed as functional, as contingent,
and as also possible under different circumstances and in different
forms. Seen from the viewpoint of the theory construction, this allows
for an understanding of the ambiguity, the need for interpretation, and
the possible avoidance of specific structural arrangements. To put it
differently, one might say that the theory of autopoietic systems above
all points out to the observer that structures have meaning and there-
fore must be constituted within open horizons of referring to other
possibilities, while the system’s own autopoiesis is not under consider-
ation in autopoietic systems. With these insights, we find ourselves in
the vicinity of theories of “symbolic interactionism” or of theories of
the hermeneutic “interpretation” of reality – but without having to
resort to behaviorist (Mead) or subjectivist assumptions. In the case of
organizations, we see that uncertainties must be reduced and ambigui-
ties in the decision process must be clarified. At the same time, how-
ever, we also see that uncertainty and ambiguity must always also be
regenerated in the processing of meaning. Particularly the autopoiesis
of organizations is kept going precisely by the fact that uncertainty is
not only reduced but also renewed.
The impressive gain achieved by means of this complex maneuvering of con-

cepts consists in the transformation of the fundamental problem of systems
theory: the preservation of existence is turned into the preservation of a dif-
ference. This also means that we no longer speak of “existential” necessities
(e.g., an organization can exist only if …)20 but of the conditions of the pos-
20 Many – too many – items could be enumerated as such existential necessities, not just
“goals” or other “essential” structures (“functional requisites”) but also humans capable of
action, fresh air, constant laws of gravity, etc.
62
05 Kap 2.Luhmann 05-11-11 16.17 Sida 63
sibility of observing organizations. Whenever we describe organizations as

autopoietic systems, it is always a matter of the production and reproduc-
tion of a difference (in systems-theoretical terms: between system and envi-
ronment). The concept of autopoiesis means that an observer who uses it
presupposes that this difference is produced by the system itself and is repro-
duced through the system’s own operations.
– Translated by Peter Gilgen
Acknowledgements
This text is an edited and translated version of pp. 44–55 of Niklas Luh-
mann (2000) Organisation und Entscheidung (Opladen: Westdeutscher Ver-
lag).
63
06 Kap 3.Luhmann 05-11-11 16.18 Sida 64
Chapter 3
The Autopoiesis of
Social Systems
Niklas Luhmann
Meaning and life as different modes of

autopoietic organization
The term “autopoiesis” has been invented to define life. Its origin is clearly
biological. Its extension to other fields has been discussed, but rather
unsuccessfully and on the wrong premises. The problem may well be that we
use a questionable approach to the issue, “tangling” our “hierarchies” of
investigation.
At first sight it seems safe to say that psychic systems, and even social sys-
tems, are also living systems. Would there be consciousness or social life
without (biological) life? And then, if life is defined as autopoiesis, how
could one refuse to describe psychic systems and social systems as auto-
poietic systems? In this way we can retain the close relation between
autopoiesis and life and apply this concept to psychic systems and to social
systems as well. We are almost forced to do it by our conceptual approach
(Maturana 1980; Hejl 1982; Bunge 1979). However, we immediately get
into trouble in defining precisely what the “components” of psychic and
social systems are whose reproduction by the same components of the same
systems recursively defines the autopoietic unity of the system. And what
does “closure” mean in the case of psychic and social systems if our theo-
retical approach requires the inclusion of cells, neurophysiological systems,
immune systems, etc. of living bodies into the encompassing (?) psychologi-
cal or sociological realities?
Moreover, because it is tied to life as a mode of self-reproduction of
autopoietic systems, the theory of autopoiesis does not really attain the level
of general systems theory which includes brains and machines, psychic sys-
tems and social systems, societies and short-term interactions. From this
point of view, living systems are a special type of systems. However, if we
abstract from life and define autopoiesis as a general form of system-
building using self-referential closure, we would have to admit that there are
non-living autopoietic systems, different modes of autopoietic reproduction,
and general principles of autopoietic organization which materialize as life,
64
06 Kap 3.Luhmann 05-11-11 16.18 Sida 65
but also in other modes of circularity and self-reproduction. In other words,

if we find non-living autopoietic systems in our world, then and only then
will we need a truly general theory of autopoiesis which carefully avoids
references which hold true only for living systems. But which attributes of
autopoiesis will remain valid on this highest level, and which will have to be
dropped on behalf of their connection with life? The text that follows uses
this kind of multi-level approach. It distinguishes a general theory of self-
referential autopoietic systems and a more concrete level at which we may
distinguish living systems (cells, brains, organisms, etc.), psychic systems and
social systems (societies, organizations, interactions) as different kinds of
autopoietic systems (see Figure 1).
Self-referential autopoietic systems
Living Systems Psychic Systems Social Systems
Cells Brains Organisms Societies Organizations Interactions
Figure 1: Types of self-referential autopoietic systems.
This scheme does not describe an internal systems differentiation. It is a

scheme not for the operations of systems, but for their observation. It dif-
ferentiates different types of systems or different modes of realization of
autopoiesis.
This kind of approach is usable only if we are prepared to accept its anti-
Aristotelian premise that social systems, and even psychic systems, are not
living systems. The concept of autopoietic closure itself requires this theo-
retical decision, and leads to a sharp distinction between meaning and life as
different kinds of autopoietic organization; and meaning-using systems
again have to be distinguished according to whether they use consciousness
or communication as modes of meaning-based reproduction. On the one
hand, then, a psychological and a sociological theory have to be developed
which meet these requirements; on the other hand, the concept of auto-
poiesis has to be abstracted from biological connotations. Both tasks are
clearly interdependent. The general theory of autopoietic systems forms the
foundation of the theories of psychic and social systems; the general theory
itself, however, is meaningful only if this implementation succeeds, because
otherwise we would be unable to determine which kind of attributes are
truly general.
65
06 Kap 3.Luhmann 05-11-11 16.18 Sida 66
Communications as the basic elements of

social systems
To use ipsissima verba, autopoietic systems “are systems that are defined as
unities, as networks of productions of components, that recursively, through
their interactions, generate and realize the network that produces them and
constitute, in the space in which they exist, the boundaries of the network
as components that participate in the realization of the network” (Maturana
1981, p. 21). Autopoietic systems, then, are not only self-organizing sys-
tems. Not only do they produce and eventually change their own structures
but their self-reference applies to the production of other components as
well. This is the decisive conceptual innovation. It adds a turbocharger to
the already powerful engine of self-referential machines. Even elements, that
is last components (in-dividuals), which are, at least for the system itself,
undecomposable, are produced by the system itself. Thus, everything which
is used as a unit by the system is produced as a unit by the system itself. This
applies to elements, processes, boundaries and other structures, and last but
not least to the unity of the system itself. Autopoietic systems, of course,
exist within an environment. They cannot exist on their own. But there is no
input and no output of unity.
Autopoietic systems, then, are sovereign with respect to the constitution
of identities and differences. They do not create a material world of their
own. They presuppose other levels of reality. Human life, for example, pre-
supposes the small scope of temperature in which water exists as a liquid.
But whatever they use as identities and as differences is of their own mak-
ing. In other words, they cannot import identities and differences from the
outer world; these are forms about which they have to decide themselves.
Social systems use communication as their particular mode of autopoietic
reproduction. Their elements are communications which are recursively pro-
duced and reproduced by a network of communications and which cannot
exist outside of such a network. Communications are not “living” units,
they are not “conscious” units, they are not “actions”. Their unity requires
a synthesis of three selections: namely, information, utterance1 and under-
standing (including misunderstanding).2 This synthesis is produced by the
network of communication, not by some kind of inherent power of con-
sciousness, or by the inherent quality of the information. Also – and this
goes against all kinds of “structuralism” – communication is not produced
by language. Structuralists have never been able to show how a structure can
produce an event. At this point, the theory of autopoiesis offers a decisive
1In German I could use the untranslatable term, “Mitteilung”.
2 The source of this threefold distinction (which also has been used by Austin and Searle) is
Karl Bühler (1934). However, we modify the reference of this distinction. It refers not to
“functions”, and not to types of “acts”, but to selections.
66
06 Kap 3.Luhmann 05-11-11 16.18 Sida 67
advance. It is the network of events which reproduces itself, and structures

are required for the reproduction of events by events.
The synthesis of information, utterance and understanding cannot be pre-
programmed by language. It has to be recreated from situation to situation
by referring to previous communications and to possibilities of further com-
munications which are to be restricted by the actual event. This operation
requires self-reference. It can in no way use the environment. Information,
utterances and understandings are aspects which for the system cannot exist
independently of the system; they are co-created within the process of com-
munication. Even “information” is not something which the system takes in
from the environment. Pieces of information don’t exist “out there”, wait-
ing to be picked up by the system. As selections they are produced by the
system itself in comparison with something else (e.g., in comparison with
something which could have happened).
The communicative synthesis of information, utterance and understand-
ing is possible only as an elementary unit of an ongoing social system. As the
operating unit it is undecomposable, doing its autopoietic work only as an
element of the system. However, further units of the same system can dis-
tinguish between information and utterance and can use this distinction to
separate hetero-referentiality and self-referentiality. They can, being them-
selves undecomposable for the moment, refer primarily to the content of
previous communications, asking for further information about the infor-
mation; or they can question the “how” and the “why” of the communica-
tion, focusing on its utterance. In the first case, they will pursue hetero-
referentiality, in the second case self-referentiality. Using a terminology pro-
posed by Gotthard Günther (1979), we can say that the process of commu-
nication is not simply auto-referential in the sense that it is what it is. It is
forced by its own structure to separate and to recombine hetero-
referentiality and self-referentiality. Referring to itself, the process has to dis-
tinguish information and utterance and to indicate which side of the dis-
tinction is supposed to serve as the base for further communication.
Therefore, self-reference is nothing but reference to this distinction between
hetero-reference and self-reference. And, whereas auto-referentiality could
be seen as a one-value thing (it is what it is), and could be described by a
logic with two values only, namely, true and false, the case of social systems
is one of much greater complexity because its self-reference (1) is based on
an ongoing auto-referential (autopoietic) process, which refers to itself (2) as
processing the distinction between itself and (3) its topics. If such a system
did not have an environment, it would have to invent it as the horizon of its
hetero-referentiality.
The elementary, undecomposable units of the system are communications
of minimal size. This minimal size, again, cannot be determined independent
67
06 Kap 3.Luhmann 05-11-11 16.18 Sida 68
of the system.3 It is constituted by further communication or by the prospect

of further communication. An elementary unit has the minimal meaning
which is necessary for reference by further communication – for instance,
the minimal meaning which still can be negated. Further communication can
very well separate pieces of information, utterances and understandings and
discuss them separately, but this still would presuppose their synthesis in
previous communication. The system does not limit itself by using con-
straints for the constitution of its elementary units. If need be, it can com-
municate about everything and can decompose aspects of previous commu-
nication to satisfy actual desires. As an operating system, however, it will not
always do this to the extreme.
Communication includes understanding as a necessary part of the unity of
its operation. It does not include the acceptance of its content. It is not the
function of communication to produce a consensus as the favoured state of
mind. Communication always results in an open situation of either accept-
ance or rejection. It reproduces situations with a specified and enforced
choice. Such situations are not possible without communication; they do not
occur as natural happenings. Only communication itself is able to reach a
point which bifurcates further possibilities. The bifurcation itself is a reduc-
tion of complexity and, by this very fact, an enforcement of selection.
Automatically, the selection of further communication is either an accept-
ance or rejection of previous communication or a visible avoidance or
adjournment of the issue. Whatever its content and whatever its intention,
communication reacts within the framework of enforced choice. To take one
course is not to take the other. This highly artificial condition structures the
self-reference of the system; it makes it unavoidable to take other communi-
cations of the same system into account, and every communication renews
the same condition within a varied context. If the system were set up to pro-
duce consensus it soon would come to an end. It would never produce and
reproduce a society. In fact, however, it is designed to reproduce itself by
submitting itself to self-reproduced selectivity. Only this arrangement makes
social evolution possible, if evolution is seen as a kind of structural selection
superinduced on selectivity.
Societies and interactions as different types of social

systems
Social systems, then, are recursively closed systems with respect to commu-
nication. However, there are two different meanings of “closure” which
3This argument, of course, does not limit the analytical powers of an observer, who, however,
has to take into account the limitations of the system.
68
06 Kap 3.Luhmann 05-11-11 16.18 Sida 69
make it possible to distinguish between societies and interactions as differ-

ent types of social systems. Societies are encompassing systems in the sense
that they include all events which, for them, have the quality of communi-
cation. They cannot communicate with their environment because this
would mean including their understanding partner in the system, under-
standing being an essential aspect of the communication itself.4 By commu-
nication they extend and limit the societal system, deciding about whether
and what to communicate, and what to avoid. Interactions, on the other
hand, form their boundaries by the presence of people who are well aware
that communication goes on around them without having contact with their
own actual interaction. Interactions must take into account environmental
communication, and have to acknowledge the fact that persons who are
present and participate in the interaction have other roles and other obliga-
tions within systems which cannot be controlled here and now.
But interactions also are closed systems, in the sense that their own com-
munication can be motivated and understood only in the context of the sys-
tem. For example, if somebody approaches the interactional space and
begins to participate, he has to be introduced and the topics of conversation
eventually have to be adapted to the new situation. Interactions, moreover,
cannot import communication ready-made from their environment. They
communicate or they do not communicate, according to whether they decide
to reproduce or not to reproduce their own elements. They continue or dis-
continue their autopoiesis like living systems which continue as living sys-
tems or die. There are no third possibilities, neither for life nor for commu-
nication. All selections have to be adapted to the maintenance of autopoietic
reproduction. Something has to be said, or, at least, good and peaceful (or
bad and aggressive) intentions have to be shown if others are present.5
Everything else remains a matter of structured choice within the system.
Some of its structures, then, become specialized in assuring that communi-
cation goes on even if nothing of informative quality remains and even if the
communication becomes controversial and unpleasant (Malinowski 1960).
The relation between action and communication

Confronted with the question of elementary units, most sociologists would
come up with the answer: action. Sometimes “roles” or even human indi-
viduals are preferred, but since the time of Max Weber and Talcott Parsons,
4 For problems of religion, and particularly for problems of “communication with God” (reve-
lation, prayer, etc.), see Niklas Luhmann (1985).
5 This again is not a motive for action but a self-produced fact of the social system. If nobody
is motivated to say anything or to show his intentions, everybody would assume such com-
munications and they would be produced without regard to such a highly improbable psy-
chological environment.
69
06 Kap 3.Luhmann 05-11-11 16.18 Sida 70
action theory seems to offer the most advanced conceptualization.6

Communication is introduced as a kind of action – for example, as “kom-
munikatives Handeln” in the sense of Jürgen Habermas (1981). Usually this
conceptualization is taken for granted, and classical sociological theory finds
itself resumed under the title of “Theory of Action” (Münch 1982). Con-
troversies are fought over headings such as action versus system, or individ-
ualistic versus holistic approaches to social reality. There is no serious con-
ceptual discussion which treats the relation of actions and communications,
and the important question of whether action or communication should be
considered as the basic and undecomposable unit of social systems has not
been taken up.
For a theory of autopoietic systems, only communication is a serious can-
didate for the position of the elementary unit of the basic self-referential
process of social systems. Only communication is necessarily and inherently
social. Action is not. Moreover, social action already implies communica-
tion; it implies at least the communication of the meaning of the action or
the intent of the actor, and it also implies the communication of the defini-
tion of the situation, of the expectation of being understood and accepted,
and so on. Above all, communication is not a kind of action because it
always contains a far richer meaning than the utterance or transmittance of
messages alone. As we have seen, the perfection of communication implies
understanding, and understanding is not part of the activity of the commu-
nicator and cannot be attributed to him. Therefore, the theory of autopoietic
social systems requires a conceptual revolution within sociology: the
replacement of action theory by communication theory as the characteriza-
tion of the elementary operative level of the system.
The relation of action and communication has to be reversed. Social sys-
tems are not composed of actions of a special kind; they are not commu-
nicative actions, but require the attribution of actions to effectuate their own
autopoiesis. Neither psychological motivation, nor reasoning or capacity of
argumentation, constitutes action, but simply the attribution as such, that is,
the linking of selection and responsibility for the narrowing of choice.7 Only
by attributing the responsibility for selecting the communication can the
process of further communication be directed. One has to know who said
what to be able to decide about further contributions to the process. Only
by using this kind of simplifying localization of decision points can the
process return to itself and communicate about communication.
6 See the discussion of “The Unit of Action Systems” in Parsons 1937, pp. 43 ff., which had
a lasting impact on the whole theoretical framework of the later Parsons.
7 To elaborate on this point, of course, we would have to distinguish between “behaviour”
and “action”. A corresponding concept of “motive” as a symbolic device facilitating the attri-
bution of action has been used by Max Weber. See also Mills (1940); Burke (1945/1950); Blum
and McHugh (1971).
70
06 Kap 3.Luhmann 05-11-11 16.18 Sida 71
Reflexive communication is not only an occasional event, but also a con-

tinuing possibility being co-reproduced by the autopoiesis itself. Every com-
munication has to anticipate this kind of recursive elaboration, questioning,
denial or correction, and has to preadapt to these future possibilities. Only
in working out this kind of presumptive fitness can it become part of the
autopoietic process. This, however, requires the allocation and distribution
of responsibilities. And this function is fulfilled by accounting for action.
The process therefore produces a second version of itself as a chain of
actions. Contrary to the nature of communication itself, which includes the
selectivity of information and the selectivity of understanding, and thereby
constitutes its elements by overlapping and partial interpenetration, this
action chain consists of clear-cut elements which exclude each other.
Contrary to the underlying reality of communication, the chain of commu-
nicative actions can be seen and treated as asymmetric.
In this sense the constitution and attribution of actions serve as a simpli-
fying self-observation of the communicative system. The system processes
information but it takes responsibility only for the action part of this
process, not for the information. It is congruent with the world, universally
competent, including all exclusions, and at the same time it is a system
within the world, able to distinguish and observe and control itself. It is a
self-referential system and, thereby, a totalizing system. It cannot avoid
operating within a “world” of its own. Societies constitute worlds. Ob-
serving themselves, that is, communicating about themselves, societies can-
not avoid using distinctions which differentiate the observing system from
something else. Their communication observes itself within its world and
describes the limitation of its own competence. Communication never
becomes self-transcending.8 It never can use operations outside its own
boundaries. The boundaries themselves, however, are components of the
system and cannot be taken as given by a pre-constituted world.
All this sounds paradoxical, and rightly so. Social systems as seen by an
observer are paradoxical systems.9 They include self-referential operations,
not only as a condition of the possibility of their autopoiesis but also
because of their self-observation. The distinction of communication and
action and, as a result, the distinction of world and system are operative
requirements. The general theory of autopoietic systems postulates a clear
distinction between autopoiesis and observation. This condition is fulfilled
in the case of social systems as well. Without using this distinction, the
system could not accomplish the self-simplification necessary for self-
8 See the distinction between perceiving oneself and transcending oneself made by Hofstadter
(1979).
9 The term “paradox” refers to a logical collapse of a multi-level hierarchy, not to a simple
contradiction. See Wilden 1972, pp. 390 ff.; Hofstadter (1979); Barel (1979).
71
06 Kap 3.Luhmann 05-11-11 16.18 Sida 72
observation. Autopoiesis and observation, communication and attribution

of action are not the same and can never fuse. Nevertheless, self-observation
in this specific sense of describing itself as a chain of clear-cut and respon-
sible actions is a prerequisite of autopoiesis as such. Without this technique
of using a simplified model of itself, the system could not communicate
about communication and could not select its basic elements in view of their
capacity to adapt themselves to the requirements of autopoiesis. This par-
ticular constellation may not be universally valid for all autopoietic systems.
In view of the special case of social systems, however, the general theory has
to formulate the distinction of autopoiesis and observation in a way which
does not exclude cases in which self-observation is a necessary requirement
of autopoiesis as such.
Observing such systems under the special constraints of logical analysis,
we have to describe them as paradoxical systems or as “tangled hierar-
chies”. It is not the task of an external observation to de-paradoxize the sys-
tem and describe it in a way which is suitable for multi-level logical analy-
sis.10 The system de-paradoxizes itself. This requires “undecidable” deci-
sions. In the case of social systems these are decisions about the attribution
of action. If desired, these decisions themselves can be attributed as actions,
which again could be attributed as action, and so on in infinite regress.
Logically, actions are always unfounded and decisions are decisions precise-
ly because they contain an unavoidable moment of arbitrariness and unpre-
dictability. But this does not lead into lethal consequences. The system learns
its own habits of acting and deciding,11 accumulating experiences with itself
and consolidating, on the basis of previous actions, expectations concerning
future actions (structures). The autopoiesis does not stop in face of logical
contradictions: it jumps, provided that possibilities of further communica-
tion are close enough at hand.
Maintenance of social systems by self-referential

production of elements
The formal definition of autopoiesis gives no indication of the span of time
during which components exist. Autopoiesis presupposes a recurring need
for renewal. On the biological level, however, we tend to think about the
process of replacement of molecules within cells or the replacement of cells
within organisms, postponing for some time the final, inevitable decay. The
10 I do not comment on the possibility of a logical analysis of self-referential systems which

bypasses the Gödel limitations and avoids hierarchization.
11 “Learning” understood as aspect of autopoiesis, that is, as a change of structure within a
closed system (and not as adaptation to a changing environment): see Maturana (1983, pp.
60–71).
72
06 Kap 3.Luhmann 05-11-11 16.18 Sida 73
limited duration of life seems to be a way of paying the cost of evolutionary

improbability. All complex order seems to be wrested from decay.
This holds true for social systems as well, but with a characteristic differ-
ence. Conscious systems and social systems have to produce their own
decay. They produce their basic elements, that is, thoughts and communica-
tions, not as short-term states but as events which vanish as soon as they
appear. Events, too, occupy a minimal span of time, a specious present, but
their duration is a matter of definition and has to be regulated by the
autopoietic system itself: events cannot be accumulated. A conscious system
does not consist of a collection of all its past and present thoughts, nor does
a social system stockpile all its communications. After a very short time the
mass of elements would be intolerably large and its complexity would be so
great that the system would be unable to select a pattern of coordination and
would produce chaos. The solution is to renounce all stability at the opera-
tive level of elements and to use events only. Thereby, the continuing disso-
lution of the system becomes a necessary cause of its autopoietic reproduc-
tion. The system becomes dynamic in a very basic sense. It becomes inher-
ently restless. The instability of its elements is a condition of its duration.
All structures of social systems have to be based on this fundamental fact
of vanishing events, disappearing gestures or words that are dying away.12
Memory, and then writing, have their function in preserving not the events,
but their structure-generating power.13 The events themselves cannot be
saved, but their loss is the condition of their regeneration. Thus, time and
irreversibility are built into the system not only at the structural level, but
also at the level of its elements. Its elements are operations, and there is no
reasonable way to distinguish between “points” and “operations”. Dis-
integration and reintegration, disordering, and ordering require each other,
and reproduction comes about only by a recurring integration of disintegra-
tion and reintegration.
The theoretical shift from self-referential structural integration to self-
referential constitution of elements has important consequences for systems
maintenance. Maintenance is not simply a question of replication, of cul-
tural transmission, of reproducing the same patterns under similar circum-
stances, such as using forks and knives while eating and only while eating;14
its primary process is the production of next elements in the actual situation,
12 It is rare that social scientists have a sense for the radicality and the importance of this
insight; but see Allport (1940, 1954).
13 This explains that the invention of writing speeds up the evolution of complex societal sys-
tems, making it possible to preserve highly diversified structural information. This is, by now,
a well explored phenomenon which still lacks a sufficient foundation in theory. See Yates
(1966); Ong (1967); Havelock (1982).
14 This is the famous “latent pattern maintenance” of Parsons – “latent” because the system
cannot actualize all its patterns all the time but has to maintain them as largely unused possi-
bilities.
73
06 Kap 3.Luhmann 05-11-11 16.18 Sida 74
and these have to be different from the previous ones to be recognizable as

events. This does not exclude the relevance of preservable patterns; indeed,
it even requires them for a sufficiently quick recognition of next possibilities.
However, the system maintains itself not by storing patterns but by produc-
ing elements; not by transmitting “memes” (units of cultural transmission
analogous to “genes”)15, but by recursively using events for producing
events. Its stability is based on instability. This built-in requirement of dis-
continuity and newness amounts to a necessity to handle and process infor-
mation, whatever the environment or the state of the system offers as occa-
sions. Information is an internal change of state, a self-produced aspect of
communicative events and not something which exists in the environment of
the system and has to be exploited for adaptive or similar purposes.16
If autopoiesis bases itself on events, a description of the system needs not
just one, but two dichotomies: the dichotomy of system and environment,
and the dichotomy of event and situation.17 Both dichotomies are “world”
formulas: system-plus-environment is one way, and event-plus-situation is
another way, to describe the world. If the system (or its observer) uses the
event/situation dichotomy, it can see the difference between system and envi-
ronment as the structure of the situation, the situation containing not only
the system, but also its environment from the point of view of an event.
Processing information by producing events-in-situations, the system can
orient itself to the difference of internal and external relevances. As the hori-
zon (Husserl) of events, the situation refers to the system, to the environ-
ment and to its difference – but it does all this selectively, using the limited
possibilities to produce the next event as a guideline.18 Thus, the double
dichotomy describes the way in which the system performs the “re-entry” of
the difference between system and environment into the system. On the
other hand, the difference between system and environment structures the
limitation of choice which is needed to enable the system to proceed from
one event-in-situation to another event-in-situation.
Systems based on events need a more complex pattern of time. For them,
time cannot be given as an irreversibility alone. Events are happenings which
make a difference between a “before” and a “thereafter”. They can be iden-
tified and observed, anticipated and remembered only as such a difference.
Their identity is their difference. Their presence is a co-presence of the
before and the thereafter. They have, therefore, to present time within time
15 Dawkins’s term; see Dawkins (1976).

16 See also (for systems) von Foerster (1981) especially p. 263: The environment contains no
information; the environment is as it is.
17 Using this theoretical framework, it is not permitted to speak of “environment of events, of
actions, etc.”, or to speak of “situations of a system”.

18 See Markowitz (1979) for further elaborations using the method of phenomenological psy-
chology.
74
06 Kap 3.Luhmann 05-11-11 16.18 Sida 75
and to reconstruct temporality in terms of a shifting presence which has its

quality as presence only owing to the double horizons of past and future
which accompany the presence on its way into the future.19 On this basis
conscious time-binding can develop.20 The duality of horizons doubles as
soon as we think of a future present or a past present, both of which have
their own future and their own past. The temporal structure of time repeats
itself within itself, and only this reflexivity makes it possible to renounce a
stable and enduring presence (Luhmann 1982g). By a slow process of evo-
lution, the semantics of time has adapted to these conditions. For a long time
it used a religious reservation – aeternitas, aevum, or the co-presence of God
with all times – to avoid the complete historization of time. Only modern
society recognizes itself – and consequently all previous societies – as con-
stituting its own temporality (Luhmann 1980b). The structural differentia-
tion of society as an autonomous autopoietic system requires the co-
evolution of corresponding temporal structures with modern historicism as
the well-known result.
The contribution of the general theory of

autopoietic systems
These short remarks by no means exhaust the range of suggestions that the
theory of social systems can contribute to the abstraction and refinement of
the general theory of autopoietic systems (for a more extensive treatment,
see Luhmann 1984b). We can now return to the question, What is new
about it, given a long tradition of thinking about creatio continua, continu-
ance, duration, maintenance and so forth21 (Ebeling 1976)? Since the end of
the sixteenth century, the idea of self-maintenance has been used to displace
teleological reasoning, and to reintroduce teleology with the argument that
the maintenance of the system is the goal of the system or the function of its
structures and operations. It is no surprise, therefore, that the question of
what is added by the theory of autopoiesis to this well-known and rather
futile traditional conceptualization has been appended to this discussion.22
An easy answer would be to mention the sharp distinction between self-
reference on the level of structures (self-organization) and self-reference on
the level of basic operations, or elements. Moreover, we could point to the
19 One of the best analyses of this complicated temporal structure remains Husserl (1928). For
social systems see Bergmann (1981).
20 See Korzybski (1949). From an evolutionary point of view, see Stebbins (1982), pp. 363 ff.
21 See Ebeling (1976), and of course the extensive “functionalist” discussion about “systems
maintenance”.
22 See Jantsch (1981), who prefers the theory of thermodynamic disequilibrium and dissipa-
tive structures.
75
06 Kap 3.Luhmann 05-11-11 16.18 Sida 76
epistemological consequences of distinguishing autopoiesis and observation,

observing systems being themselves autopoietic systems. We have only to
look at the consequences of an “event-structure” approach for sociological
theory to be aware of new problems and new attempts at solution, com-
pared with the Malinowski/Radcliffe Brown/Parsons level of previous
controversies. There is, however, a further aspect which should be made
explicit.
The theory of autopoietic systems formulates a situation of binary choice.
A system either continues its autopoiesis or it does not. There are no in-
between states, no third states. A woman may be pregnant or not: she can-
not be a little pregnant. This is true, of course, for “systems maintenance”
as well. Superficial observers will find the same tautology. The theory of
autopoietic systems, however, has been invented for a situation in which the
theory of open systems has become generally accepted. Given this historical
context, the concept of autopoietic closure has to be understood as the
recursively closed organization of an open system. It does not return to the
old notion of “closed [versus open] systems” (Varela 1979). The problem,
then, is to see how autopoietic closure is possible in open systems. The new
insight postulates closure as a condition of openness, and in this sense the
theory formulates limiting conditions for the possibility of components of
the system. Components in general and basic elements in particular can be
reproduced only if they have the capacity to link closure and openness. For
biological systems this does not require an “awareness” of, or knowledge
about, the environment. For meaning-based conscious or social systems the
autopoietic mode of meaning gives the possibility of “re-entry”,23 that is, of
presenting the difference between system and environment within the sys-
tem. This re-entered distinction structures the elementary operations of these
systems. In social, that is, communicative, systems, the elementary operation
of communication comes about by an “understanding” distinction between
“information” and “utterance”. Information can refer to the environment of
the system. Utterance, which is attributed to an agent as action, is responsi-
ble for the autopoietic regeneration of the system itself. In this way infor-
mation and utterance are forced to cooperate, forced into unity. The emer-
gent level of communication presupposes this synthesis. Without the basic
distinction between information and utterance as different kinds of selec-
tion, the understanding would be not an aspect of communication, but a
simple perception.
Thus, a sufficiently differentiated analysis of communication can show
23 In the sense of Spencer Brown (1971). Gotthard Günther makes the same point in stating
that these systems of self-reflection with centers of their own could not behave as they do
unless they are capable of “drawing a line” between themselves and their environment. And
this leads Günther to the surprising conclusion that parts of the universe have a higher reflec-
tive power than the whole of it (Günther 1976, p. 319).
76
06 Kap 3.Luhmann 05-11-11 16.18 Sida 77
how the recurrent articulation of closure and openness comes about. It is a

constitutive necessity of an emergent level of communication. Without a
synthesis of three selections – information, utterance and understanding
– there would be no communication but simply perception. By this synthe-
sis, the system is forced into looking for possibilities of mediating closure
and openness. In other words, communication is an evolutionary potential
for building up systems which are able to maintain closure under the condi-
tion of openness and openness under the condition of closure. These systems
face the continuing necessity to select meanings which satisfy these con-
straints. The result is our society.
In addition, the concept of autopoietic closure makes it possible to under-
stand the function of enforced binary choices. The system can continue its
autopoiesis or it can stop it. It can continue to live, to produce conscious
states or to communicate, or it can choose the only alternative: to come to
an end.24 There are no third states. This is a powerful technical simplifica-
tion. On the other hand, the system lacks any self-transcending power. It
cannot enact operations from the outer world. A social system can only
communicate. A living system can only live. Its autopoiesis as seen by an
observer may have a causal impact on its environment. But autopoiesis is
production in the strict sense of a process which needs further causes, not
produced by itself, to attain its effect. The binary structure of autopoiesis
seems to compensate for this lack of totality. It substitutes this kind of
“internal totality”. To be or not to be, to continue the autopoiesis or not,
serves as an internal representation of the totality of possibilities. Everything
which can happen is reduced for the system to one of these two states. The
system emerges by inventing this choice, which does not exist without it.
The negative value is a value not of the world but of the system. But it helps
to simplify the totality of all conditions to one decisive question of how to
produce the next system state, the next element, the next communication
under the constraints of a given situation. Even unaware of the outer world,
the living system “knows” that it is still alive and chooses its operations in
using life for reproducing life. A communicative system too can continue to
communicate on the base of the ongoing communication. This requires no
reliable knowledge about outside conditions but simply the distinction
between system and environment as seen from the point of view of the sys-
tem. The unity of the autopoietic system is the recursive processing of this
difference of continuing or not which reproduces the difference as a condi-
tion of its own continuity. Every step has its own selectivity in choosing
autopoiesis instead of stopping it. This is not a question of preference, nor
a question of goal attainment. Rather, it has to be conceived as a “code” of
24 “End” in this theory, therefore, is not “telos” in the sense of the perfect state, but just the
contrary: the zero-state, which has to be avoided by reproducing imperfect and improbable
states. In a very fundamental way the theory has an anti-Aristotelian drift.
77
06 Kap 3.Luhmann 05-11-11 16.18 Sida 78
existence, if code is taken as an artificial duplication of possibilities with the

consequence that every element can be presented as a selection.
This may become more clear if we consider the case of social systems.
Autopoiesis in this case means “to continue to communicate”. This becomes
problematic in face of two different thresholds of discouragement. The first
tends to stop the process because the communication has not been under-
stood. The second tends to stop the process because the communication has
been rejected. These thresholds are related to each other because under-
standing increases the chances of rejection.25 It is possible to refrain from
communication in face of these difficulties, and this is a rather common
solution for interaction systems, particularly under modern conditions of
highly arbitrary interactions. Society however, the system of all communica-
tions, cannot simply capitulate in the face of these problems; it cannot stop
all communications at once and decide to avoid any renewal.26 The
autopoiesis of society has invented powerful mechanisms to guarantee its
continuity in the face of a lack of understanding or even open rejection. It
continues by changing the interactional context or by reflexive communica-
tion. The process of communication returns to itself and communicates its
own difficulties. It uses a kind of (rather superficial) self-control to become
aware of serious misunderstandings, and it has the ability to communicate
the rejection and restructure itself around this “no”. In other words, the
process is not obliged to follow the rules of logic. It can contradict itself. The
system which uses this technique does not finish its autopoiesis and does not
come to an end; it reorganizes itself as conflict to save its autopoiesis. When
faced with serious problems of understanding and apparent misunderstand-
ings, social systems very often tend to avoid the burden of argumentation
and reasoned discourse to reach consensus – very much to the dismay of
Habermas. Rather, they tend to favour the rejection of proposals and to
embark on a course of conflict.
However this may be, the communication of contradiction, controversy
and conflict seems to function as a kind of immune system of the social sys-
tem (Luhmann, 1984b). It saves autopoiesis by opening new modes of com-
munication outside normal constraints. The law records experiences and
rules for behaving under these abnormal conditions and, by some kind of
epigenesis, develops norms for everyday behaviour which help to anticipate
the conflict and to preadapt to its probable outcome (Luhmann, 1983a). In
highly developed society we even find a functionally differentiated legal sys-
tem which reproduces its own autopoietic unity. It controls the immune sys-
25 From an evolutionary point of view – see Luhmann (1981d).
26 That the physical destruction of the possibility of communication has become possible, and
that this destruction can be intended and produced by communication, is another question. In
the same sense, life cannot choose to put an end to itself, but conscious systems can decide to
kill their own bodies.
78
06 Kap 3.Luhmann 05-11-11 16.18 Sida 79
tem of the larger societal system by a highly specialized synthesis of norma-

tive (not-learning) closure and cognitive (learning) openness (Luhmann
1983b). At the same time, it increases the possibilities of conflict, makes
more complex the immune system and limits its consequences. It cannot, of
course, exclude conflicts outside the law, which may save the autopoiesis of
communication at even higher costs.27
The epistemological consequences of autopoietic

closure
A final point of importance remains: the epistemological consequences of
autopoietic closure. This problem also has to be discussed with respect to
the present situation of scientific evolution in which the theory of autopoi-
etic systems seems to offer advantages.
For many decades, scientific research has no longer operated under the
guidance of an undoubted orthodoxy – be it a theory of cognition or a the-
ory of science in particular. The universally accepted expedient is “pragma-
tism”; the results are the only criteria of truth and progressive knowledge?
This is clearly a self-referential, circular argument, based on a denial of cir-
cularity in theory and on its acceptance in practice. The avoidance of circu-
larity becomes an increasingly desperate stance – a paradox which seems to
indicate that the condemned solution, the paradox itself, is on the verge of
becoming accepted theory.
One way to cope with this ambiguous situation is to test whether method-
ologies have the capacity to survive the coming scientific revolution.
Functional analysis is one of them. It can be applied to all problems, includ-
ing the problem of paradox, circularity, undecidability, logical incomplete-
ness, etc. Stating such conditions as a problem of functional analysis invites
one to look for feasible solutions, for strategies of de-paradoxization, of
hierarchization (in the sense of the theory of types), of unfolding, of asym-
metrization and so on. Functional analysis, in other words, reformulates the
constitutive paradox as a “solved problem” (which is and is not a problem)
and then proceeds to compare problem solutions.28
27 Recent tendencies to recommend and to domesticate symbolic illegalities as a kind of com-
munication adapted to too high an integration of society and positive law seems to postulate
a second kind of immune system on the basis of a revived natural law, of careful choice of
topics and highly conscientious practice; see Guggenberger (1983).
28 In a way, the problem remains a problem by “oversolving” it, that is, by inventing several
solutions which are of unequal value and differ in their appropriateness according to varying
circumstances. This gives us, by a functional analysis of functional analysis, an example of
how de-paradoxization can proceed. The happy pragmatist, on the other hand, would be con-
tent with stating that a problem becomes a problem only by seeing a solution; see Laudan
(1977).
79
06 Kap 3.Luhmann 05-11-11 16.18 Sida 80
In addition to this kind of preadaptation in scientific evolution to an

expected change of the paradigm of the theory of cognition itself, the theo-
ry of autopoietic systems constructs the decisive argument. It is a theory of
self-referential systems, to be applied to “observing systems” as well
– “observing systems” in the double sense that Heinz von Foerster (1981)
used when he chose the phrase as the title of a collection of his essays. The
theory distinguishes autopoiesis and observation, but it accepts the fact that
observing systems themselves are autopoietic (at least, living) systems.
Observation comes about only as an operation of autopoietic systems, be it
life, consciousness or communication. If an autopoietic system observes
autopoietic systems, it finds itself constrained by the conditions of autopoi-
etic self-reproduction (again, respectively, life, consciousness and communi-
cation, e.g., language), and it includes itself in the fields of its objects,
because as an autopoietic system observing autopoietic systems, it cannot
avoid gaining information about itself.
In this way, the theory of autopoietic systems integrates two separate
developments of recent epistemological discussion. It uses a “natural” or
even “material” epistemology, clearly distinct from all transcendental aspi-
rations (Quine 1969) – transcendentalism being in fact a title for the analy-
sis of the autopoietic operations of conscious systems. In addition, it takes
into account the special epistemological problems of universal or “global”
theories, referring to a class of objects to which they themselves belong.29
Universal theories, logic being one of them, have the important advantage of
seeing and comparing themselves with other objects of the same type. In the
case of logic, this would require a many-valued structure and the corre-
sponding abstraction. Classical logic did not eliminate self-reference, but it
did not have enough space for its reflection. “The very fact that the tradi-
tional logic, in its capacity of a place-value structure, contains only itself as
a subsystem points to the specific and restricted role which reflection plays
in the Aristotelian formalism. In order to become a useful theory of reflec-
tion a logic has to encompass other sub-systems besides itself” (Günther,
1976, p. 310; emphasis added). Only under this condition does functional
analysis become useful as a technique of self-exploration of universal theo-
ries.
The usual objection can be formulated, following Nigel Howard, as the
“existential axiom”: knowing the theory of one’s own behaviour releases
one from its constraints (Howard 1971: pp. xx, 2 ff. and passim). For an
empirical theory of cognition, this is an empirical question. The freedom,
29 See Hooker (1975, pp. 152–79). Many examples: the theory of sublimation may itself be a
sublimation. Physical research uses physical processes. The theory of the Self has to take into
account that the theorist himself is a Self (a healthy Self, a divided Self). For this last example
see Holland (1977).
80
06 Kap 3.Luhmann 05-11-11 16.18 Sida 81
gained by self-reflection, can be used only if its constraints are sufficiently

close at hand. Otherwise the autopoietic system simply will not know what
to do next. It may know, for example, that it operates under the spell of an
“Oedipus complex” or a “Marxist” obsession, but it does not know what
else it can do.
A “new epistemology” will have to pay attention to at least two funda-
mental distinctions: the distinction between autopoiesis and observation on
the one hand, and the distinction between external observation and internal
observation (self-observation) on the other. To combine these two distinc-
tions is one of the unsolved tasks of systems theory.
Autopoiesis is the recursive production of the elements by the elements of
the system. Observation, being itself an autopoietic operation, applies a dis-
tinction and indicates which side is used as a basis of further operations
(including the operation of “crossing” and indicating the other side). Self-
observation is a special case of observation because it excludes other
observers. Only the system can self-observe itself; others are by definition
external observers. Therefore self-observation does not and cannot use cri-
teria. It cannot choose between different perspectives.30 It observes what it
observes, and31 can only change its focus and the distinction it applies. It is
always sure about itself. External observers, on the other hand, are always
a plurality. They have to presuppose other observers. They can observe other
observers and other observations. They can compare their observations with
others. They can be seduced into a reflexive observing of observations, and
they need criteria when different observations yield different results.32
Classical epistemology looks for a set of conditions under which external
observations yield identical results. It does not include self-observation (the
“subjective” or “transcendental” approach, meaning only that these condi-
tions can be found by introspection). This excludes, however, societies as
observed and as observing systems. For within societies all observations of
the society are self-observations. Societies cannot deny the fact that the
observation itself is an element of the system which is observed. It is possi-
ble to differentiate subsystems within the societal system, giving them the
special function of “observing society”. This still implies self-observation,
because the subsystem can only operate within the society. It can look at its
social environment, at political, economic, legal and religious affairs, but
this observation itself (1) is part of the autopoiesis of the societal system and
30 See, for the special case of conscious autopoietic systems (i.e. not for social systems!),
Shoemaker (1963, 1968).

31 It can, of course, distinguish between self-observation and external observation, and it can
observe external observation focusing on itself.

32 Not, of course, when different observations choose different objects; and the classical epis-
temology does not give useful criteria for the case in which different observations use differ-
ent distinctions.
81
06 Kap 3.Luhmann 05-11-11 16.18 Sida 82
(2) becomes self-observation as soon as it tries to observe and control its

epistemology. Somehow, a “Third Position” (Bråten 1984) would be re-
quired, one which contains the possibility of shifting between external
observation and self-observation and of defining rules which tackle the pa-
radox of being at the same time inside and outside of the system. It is with-
in this context that universal theories become relevant. They are designed
for external observation but are generalized in a way that includes the
observing system as well. The rules and methodologies of universal theories
may become the nucleus from which the development of a truly social epis-
temology can start. At least, they provide a field of experience in which we
are logically forced to oscillate between external observation and self-
observation.
We may leave as an unresolved question whether this kind of argument
can generate not only sociological33 but also biological (Maturana 1978)
and psychological (Campbell 1970, 1974, 1975) epistemologies. This de-
pends, last but not least, upon the possibility of applying the concept of self-
observation not only to social systems and conscious systems, but also to liv-
ing systems. In any case, advances in substantial theory may have side-
effects on the theories which are supposed to control the research. Until the
eighteenth century these problems were assigned to religion – the social sys-
tem specialized for tackling paradoxes.34 We have retained this possibility,
but the normalization of paradoxes in modern art and modern science seems
to indicate our desire eventually to get along without religion (Dupuy, 1982,
pp. 162 ff.). Apparently, our society offers the choice of trusting religion or
working off our own paradoxes without becoming aware that this is reli-
gion.
Acknowledgements
This text was originally published in Geyer, F./Zeuwen, J. v. d. (1986) (eds.),
Sociocybernetic Paradoxes: Observation, Control and Evolution of Self-
Steering Systems (London: Sage), pp. 172–192.
33 For a case study, using this mode of controlled self-reference, see Cole and Zuckermann
(1975).
34 See the impasse as formulated by Bishop Huet: “Mais lors que l’Entendement en vue de
cette Idée forme un jugement de l’objet extérieur, d’où cette Idée est partie, il ne peut pas savoir
très certainement et très clairement si ce jugement convient avec l’objet extérieur; et c’est dans
cette convenance que consiste la Vérité, comme je l’ai dit. De sorte qu’encore qu’il connoisse
la Vérité, il ne sçait pas qu’il connoît, et il ne peut être assuré de l’avoir connue” (Huet 1723,
p. 180).
82
07 Kap 4. Luhmann 05-11-11 16.18 Sida 83
PART II
Organization, Decision
and Paradox
07 Kap 4. Luhmann 05-11-11 16.18 Sida 84
07 Kap 4. Luhmann 05-11-11 16.18 Sida 85
Chapter 4
The Paradox of Decision

Making
Niklas Luhmann
What are “decisions”?

Decisions are often described as the causes of their effects. However, this is
simply the result of attribution, which drastically simplifies the vast number
of possible perceptions of causality. If we observe organizations as networks
of decision operations, we still have no access to meaningful causal analy-
ses; we only see what the organization itself reveals – highly selectively – for
purposes of exerting influence, planning, accounting, and controlling. The
progression of operations from decision to decision is thus a self-made arte-
fact. Like statistics, it is grounded in a fictional (and, in a narrower sense,
documented) reality. Decision making is, in other words, the way in which
the organization differentiates itself and, by doing so, recognises what it is
doing. No more, but no less.
This introductory reflection suggests that it is not so easy to define what
a decision actually “is”. When we speak of decisions, we normally think of
an act of selecting, which is characterised by a certain arbitrariness.
Something that has already been determined in all respects cannot be de-
cided. Therefore, a decision involves some degree of unpredictability, even
irrationality, which is precisely why people are tempted to exert influence.
This raises the question of whether efforts to rationalise decisions, or deci-
sion-making processes, make people less inclined to take decisions, under-
mine motives, and thus lead to the situation where nobody is particularly
committed to implementing a decision – once it has been taken1. The same
is true of risk acceptance2. The classical notion that good decisions are cor-
rect decisions and that correct decisions can be reached by rationally calcu-
lating means and ends is in the process of being dismantled. But what is tak-
ing its place?
The usual distinction of “decision”, as opposed to behaviour or simple
action by a moment of arbitrariness, was clearly related to the hierarchical
1 With regard to Swedish experience, see Brunsson (1985).
2 Cf. Japp (1992).
85
07 Kap 4. Luhmann 05-11-11 16.18 Sida 86
structuring of organizations. Not all behaviour in organizations is designat-

ed in those organizations as “decision”, and the decision-making burden
– and with it responsibility, experience, authority, and possibly wisdom
– increases, we assume, along the chain of command from bottom to top; or
at any rate this expectation is encouraged by the way decisions are talked
about in organizations. By resorting to whatever ruse possible, we let our
superiors take the decisions – and we let them enjoy doing so. They reinforce
their role with the expectation that has been linguistically institutionalised,
that it is only up to them (or especially up to them) to decide; even though
everybody (themselves included) knows that they are being controlled
through the preparation of the decisions. The use of language, because it
marks only few decisions as decisions, serves to “upgrade” decision making,
making certain decisions difficult and important. The mystery of decision
and the mystery of hierarchy support each other mutually. Both develop an
inexpressible (should we say divine?) moment, which makes them what they
are. Or what they appear to be?
Evidently, it is difficult to pin down the decision component “arbitrari-
ness” – both in practice and in theory. It is certainly not a psychological
moment but, rather like “intention”, it is a need to attribute decisions, a
need to localise the points of decision in the network of communications.
We should not ask, therefore, how the decision maker feels when he detects
signs of arbitrariness cropping up in him, or when he “has to pull himself
together”; rather, we should observe those points that attribute arbitrari-
ness3. Arbitrariness appears to be a fiction that serves attribution; or perhaps
even the impossibility of attribution is a precondition for attribution.
Whatever the case, the models of rational decision do not illuminate this
blind spot. Using the old poetic and theological term integumentum we
could say: the decision conceals what is decisive. But why?
We presume (and there are similar considerations; especially George
Spencer Brown’s calculus of form4 would confirm this) that the blind spot
serves to conceal a paradox which we would come upon if we pursued the
analysis further. Heinz von Foerster expresses this with great clarity: “Only
those questions that are in principle undecidable, we can decide”; everything
else is a matter of (more or less tedious) calculation5. An analysis that takes
this point seriously can go in two directions, depending upon whether it pro-
ceeds from a factual or a temporal description of the decision.
In the factual description, a decision is, according to general opinion, a
choice between alternatives. It is not the one or the other side of the
3 From the literature on “systems therapy” we learn that even this counts as a “systemic
approach”: the problem is not where the layperson would suppose, but elsewhere; and this is
why the system needs advice.
4 See Spencer Brown (1979); Luhmann (1993c).
5 Von Foerster (1992), p. 14.
86
07 Kap 4. Luhmann 05-11-11 16.18 Sida 87
alternative but precisely this “or” or this “between”. But what is this “or”,
what is the form of the alternativity6? It is not itself a possible object of
choice. You cannot decide on the “or”7. Evidently, the “or”, the form of
alternativity, is excluded from the possible decisional scope, although (or
because?) it is constitutive for it. Constructing alternativity, therefore, is all
about including the excluded, the “parasite” in the sense of Michel Serres8.
The parasite exploits the opportunity that arises from being excluded. The
decision maker himself is not an alternative, nor is he an option within the
alternative. He can only choose himself, and this he does unnoticed. The
more rational the choice, the more mysterious the parasite that evades
designation with ever more distinctions. With Jacques Derrida we can at
best follow the trail that the excluded (paradox) leaves behind; or to put it
more precisely: “la trace de l’effacement de la trace”9.
We find ourselves in more familiar company when we realise that semi-
otics, too, which is based upon the distinction of signifier and signified, has
to assume a third element, for example, the “interpretant” in the sense of
Peirce; and that legal experts are accustomed to the fact that they (1) have
to interpret the meaning (2) of a text (3). If we want to formulate this
assumption as a paradox, we talk about the unity of a distinction; about the
sameness of the different – in other words, the point at which Hegelians
experience the bliss of “sublation” [Aufhebung].
If we transfer this reflection from the factual dimension to the temporal
dimension, the same paradox of decision takes on a different form. All sys-
tems that reproduce themselves through their own operations can be said to
exist only in the form of recursive operation. In other words, they exist only
at the moment an operation is actually taking place. For this reason, the
world always exists concurrently with the actual operations – neither before
6 We should not forget that there have been attempts to logically reconstruct alternativity,
which aim at securing consistency and excluding paradox. See, e.g., Rödig (1969). But these
efforts simply shift the problem to further distinctions which themselves are not problema-
tised; thus, Rödig, for instance, distinguishes several possible worlds (e.g., p. 43) or (real) indi-
viduals and sets (p. 58), and, very typically in logic and linguistics, several “levels” of analy-
sis or language. All these endeavours have one weakness in common: they can only displace
but do not tackle the problem of (the unity of) distinction. If we are to represent this, we need
a clearly metalogical theory that encompasses logic, i.e., it can show that we can only avoid
paradox if we draw distinctions that are – for the moment – sufficiently plausible. “Draw a
distinction” is the first paradoxical rule of every calculus for avoiding paradox. See Spencer
Brown (1979), p. 3, on the need for a “creative” unfolding of paradoxes (cf. also Krippendorff
1984). There are also tendencies to recognise such procedures as mathematics if not as logic
(cf. Kauffmann 1998).
7 Of course, we can decide not to decide; but then we construct a new alternative with a new
“or”: to decide or not to decide.

8 Serres (1980). Derrida, too, uses the term in this sense.
9 Derrida (1972) pp. 76–77.
87
07 Kap 4. Luhmann 05-11-11 16.18 Sida 88
nor after.10 Insofar as such systems can use their operations for observing,
i.e. use them as distinction and indication, they can treat the present (in
which they alone can “be”, i.e., operate) as the boundary between a respec-
tive past and a respective future; in other words, they can make a temporal
distinction and thus in a sense de-temporalise the world. It then appears as
though the world exists from the beginning to the end of time (or for ever)
and in this world-time things exist for a longer or shorter duration. But that
is and will always be the construction of an observer, who can only ever
operate in the present and only ever concurrently with the world. Viewed
from the present moment, the world comprises the non-actual time horizons
of past and future, and the present disappears, as it were, in the world. It is
simply the distinction between past and future; or, to be more precise, it is
the unity of this distinction – in other words a paradox.
When we take a decision, we as observers need these non-actual time hori-
zons of past and future. It is only their non-actuality that enables us to con-
struct alternatives into the concurrently present world, which is always as it
is and never anything different. The actuality of the decision cannot be
reflected in the decision. It is and remains the boundary that renders dis-
tinction possible but cannot be incorporated into what is being distin-
guished. The decision itself is neither something in the past nor something in
the future, and it is neither the one side nor the other side of the alternative.
Undoubtedly, an observer can observe the decision; and another decision
can accept or reject the decision as a decision premise. But that would be a
different operation, for which the principle likewise holds that the unity of
the distinction that is made remains invisible for the distinguishing opera-
tion.
G.L.S. Shackle11 believes, therefore, that every decision can be construed
as “original”, as the “beginning” of something new and, accordingly, as
“subjective”. But this is just one observer’s schematisation of the problem;
there may well be other observers who, like ourselves, are happy with the
analysis from the perspective of an observation of observers, in other words
with second-order cybernetics. With this reformulation the insight remains
unchanged, that decision making actualises an – as it were – “unnatural”,
reverse relationship between past and future. From each present the past is
observed as no longer changeable, while the future is observed as still
changeable. Analogously, a decision cannot be determined by the past. It
constructs the alternativity of its alternative from the perspective of “what
might be”; and it constructs it in the present time. However, with regard to
future present times, the decision proceeds from the assumption that it will
10 This definition is also found in Shackle (1979), p. 20: “… the notion of the present, the
moment of actuality embracing all that is. All that is, is the present”.
11 See footnote 10.
88
07 Kap 4. Luhmann 05-11-11 16.18 Sida 89
make a difference whether and how a decision is taken. In other words:

there is no commitment to the (no longer changeable) past, but commitment
to the (still changeable) future. In this way, the decision appears motivated,
intentional, reasoned; and it can also be defended. It makes commitment vis-
ible, and by doing so, it contributes to uncertainty absorption in the
sequence of connecting decisions – even in cases, and there particularly,
where subsequent decisions can neither be predicted nor determined.
Accordingly, an observer (irrespective of which present he is in) sees the
decision differently, depending on whether he is considering the time before
or the time after the decision. For the observer the decision before the deci-
sion is a different decision from that after the decision. Before the decision
it is an open alternative; in other words, it is an open contingency. Several
decisions are possible, or so it is said.12 After the decision the contingency
becomes fixed; and we only see that the taken decision could have been dif-
ferent. The contingency (defined as neither necessary nor impossible) is now
fixed to a decision. Here, too, we can say: a decision is different things, “the
same is different”.13 Depending upon the approach, whether we are looking
at the alternativity or the temporal difference, the same paradox of decision
appears in different forms. But it is the same paradox; the paradox of the
unity of different things, the paradox of the unity of a form with two sides.
On the function of paradoxes

Where does this get us, you will probably ask. Practitioners manage per-
fectly well with a suitable mix of sobriety and mystification. Why push them
into the blind alley of a paradox and confront them with the closed self-
reference of a paradox: if A then not A; if not A then A.
At this point we need to find out more about the possibilities of handling
paradoxes. The term is rooted not in logic but in rhetoric. Logic has never
been on very comfortable terms with paradox. On the contrary: it has
ascribed it to the general category of contradiction and contented itself with
postulating consistency. In rhetoric, by contrast, the focus was less on con-
ditions of truth than on conditions of effective communication. Rhetoricians
were interested, for example, in the possibility of attracting attention and
encouraging people to reflect by making statements that contradicted gener-
ally accepted opinion, and which in this sense appeared “para-doxical”.
This technique enjoyed a revival in the wake of book printing, the human-
ism of the Renaissance, religious turmoil, wars centring on questions of
truth, and not least the demise in the 16th century of the method of scholas-
tic inquiry. Erasmus’s praise of stupidity and Thomas Moore’s Utopia gave
12 Why the plural, when we know that only one decision is possible.
13 To cite a title by Ranulph Glanville (1981).
89
07 Kap 4. Luhmann 05-11-11 16.18 Sida 90
the starting signals, and by the end of the century a body of literature, which
included drama and poetry, had been written.14 At first glance, this treat-
ment of paradox appears somewhat frivolous, especially the simple listing of
examples from antiquity. But obviously their intention was to disturb – for
example, a collection of paradoxes was published together with a volume
that confuted them.15 It can be assumed that theologians (John Donne, for
example) knew that the form of the paradox had succeeded the quaestio or
inquiry method: question, opinion, counter-opinion – and no authority to
decide the question!16
There can be no doubt, therefore, that the matter was serious. And yet,
the scientific movement that was emerging at the same time carried the vic-
tory, made history, and stifled rhetoric – along with its use of paradox,
which had fascinated readers. Although we still find many examples of such
works in the 17th and 18th centuries, which are aimed particularly at the
shallow morality and self-righteousness of contemporaries,17 the form had
become so hackneyed that in an age that was obsessed with progress it
brought no advancement (and indeed this had never been intended).
It was not until the 20th century that the problem – or rather the form – of
the paradox was rediscovered; partly in the problem of substantiating
mathematical logic, partly in the long drawn out death throes of ontological
metaphysics – to mention only Nietzsche, Heidegger and Derrida.18
However, we can content ourselves with a very much simpler analysis.
It is to be assumed – and who would suggest otherwise? – that all obser-
vation and description is an operation that is carried out in real terms and
in the here and now and, moreover, that this operation must be able to dis-
tinguish what it observes and describes. This applies not only to scientific
analysis but also to the internal processes of organizations. It applies both
to mental operations and to communications. It applies to experience and
action. Observation in this very general sense always requires that we
choose one thing as distinct from another; in other words, it requires the use
of a distinction, of which only the one side (marked side) and not the other
can be used for connecting subsequent operations.
Accordingly, observation is necessarily an inherently asymmetrical opera-
tion. For this reason, the need to mark both sides of the observational dis-
tinction simultaneously results in the paradox of a one-sided and two-sided
14 A comprehensive overview can be found in Colie (1966). Cf. also Lupi (1992); also the
hitorical accounts in Geyer/Hagenbüchle (1992).
15 A much cited and translated example is Lando (1545; n.d.).
16 See Malloch (1956).
17 Possibly the most famous example is Mandeville (1924 – originally 1714) – interesting also
because it illustrates the mode of solving the moral paradox with the distinction between pri-
vate and public. Cf. also Bernard (1759).
18 Lawson (1985) sees this as the distinctive feature of 20th century philosophy.
90
07 Kap 4. Luhmann 05-11-11 16.18 Sida 91
actualisation. On the other hand, a simultaneous functioning of the distinc-

tion has to be presupposed as the “form” of observation.19 The distinction
used by every observation is always, therefore, both two-sided and one-
sided. However, it need not (and cannot) observe this peculiarity itself. In
other words, it is rooted in a paradox, nevertheless it remains operational
because it conceals – can conceal, must conceal – its paradox with the mere
fact that the distinction is drawn. It does not see that it does not see what it
does not see, and this is, if we were again to formulate it in terms of tran-
scendental theory, a condition of its possibility. Only an observer of an
observer can indicate the paradox in which the observed observation is
grounded, but only with a likewise paradoxically grounded observation. On
the level of second-order observation, we see the distinction of first-order
observation as form. It is only here that we encounter the paradox underly-
ing all observation, by inquiring into the unity of the observational opera-
tion, i.e., the unity of its distinction, the sameness of what is being distin-
guished. The form of the paradox, therefore, only emerges on the level of
second-order observation. It simply serves to baffle the second-order observ-
er. It does not prevent what happens from happening; it does not prevent
operations and observational operations from actually taking place. Even
paradoxes, which is what we have been talking about all along, can be
observed and described through operations that are factually taking place.
The question is: why?
The answer is: there is no other means of ultimate substantiation, neither
for recognition nor for taking action and even less for decision making.
Unfolding the paradox

Questions of substantiation are best left to philosophy, where they are
addressed through esotericism, linguistic minutiae, fast metaphors, tran-
scendental pragmatism or in some other way. But even for the more mun-
dane analysis of organization studies and decision theory, the digression via
the paradox might reveal perspectives and issues which would otherwise go
unnoticed or which would not be seen in context. We recall that the “para-
dox shock” in traditional rhetoric was used to challenge common sense and
to break established habits. In order to clarify this aspect of paradox
management, we use the distinction (distinction!) of paradox and unfolding
of paradox (by analogy with problem and problem solving).20 We describe
how an observer who asks a question that may only be answered by a para-
dox copes with that; and the answer is intended as a theory that can be
19 The definition of the term “form” on the basis of the operative practice of an observer can
also be found in Spencer Brown’s calculus of form (see footnote 4).

20 Cf. apart from Krippendorff (1984); also Löfgren (1978; 1979).
91
07 Kap 4. Luhmann 05-11-11 16.18 Sida 92
empirically verified: he replaces it with another distinction that he considers

plausible.
To unfold a paradox is simply to shift the observer’s blind spot to a place
where it is less troublesome (in family therapy, which is based on construc-
tivism, we say: to a less painful place). Against one’s better judgement new,
stable identities are introduced, which can be maintained more easily. A typ-
ical example of this approach is Kant’s proof of the antinomies of tradition-
al metaphysics as points de départ for the search for a different metaphysics
grounded in the self-analysis of the mind.21 But identities can only be indi-
cated with the help of distinctions; however, these distinctions are also para-
doxical. This is something we have to accept whenever we unfold a first
paradox. The result can never, therefore, consist in a stable world construc-
tion that is without paradox, but in advantages in the way we organize our
cognition; or in greater historical plausibility of knowledge in situations in
which old distinctions have been exhausted and can be all too easily chal-
lenged.
We assume that we now have just such a case. Both in terms of how “deci-
sion” is generally understood and how hierarchy is generally perceived, we
face an integumentum, a moment that defies expression. In the case of deci-
sions, it is a moment of “subjective” arbitrariness which makes the decision
a decision. In the case of hierarchy, it is the authority of the position or of
superior knowledge, which we assume against our better judgement to pro-
ceed from bottom to top. We also presume that these two instances of “mys-
tification“ are connected, that they mutually support one another, that they
are anchored in language and that they secure normal operation insofar as
nobody objects to following that language rule and speaking that language
(whatever he or she thinks about it and whatever people say about it “off
duty”).
The question, therefore, is: can the paradox of decision be unfolded in any
other way? Are there other distinctions that are more plausible in the pres-
ent day, which achieve the same, which can likewise produce identities and
translate the infinite information burdens of patent paradox into finite infor-
mation burdens?
We know that we cannot expect such unfolding to take the form of a logi-
cally controllable step. We have to rely on creative “intuition” or, where
there is doubt, on the object of research in the respective field. In fact, it
appears that appropriate distinctions already exist. We simply have to find
them and present them based on a general scepticism towards rationality.
21 Here of course not yet: “against one’s better judgement”. Rather, his idea was that a mind
can discover identities in itself, which serve as necessary conditions for its own operations and
which can be supposed to have a greater stability than what has been handed down as the con-
ceptual framework of ontological metaphysics.
92
07 Kap 4. Luhmann 05-11-11 16.18 Sida 93
Specifically, they are (1) the distinction between decision premises and deci-
sion, (2) the distinction between decision and uncertainty absorption, and
(3) the distinction between rationality and motivation. We shall restrict our-
selves in the following sections to these, although it cannot be ruled out that
there may be further appropriate distinctions.
Decision premises
The distinction of decision premises and decision stems from Herbert Simon.
In older texts we read of “behavioural premises”.22 The purpose of that was
to draw attention to the facticity of the decision-making process, as distinct
from normative theories of rational decision making23. In other words, this
is about actual behaviour, which is observed and described as decision mak-
ing and which is possibly brought to account.
When we speak of “premises”, we have to think firstly of programmes as
prescribed ends and constraints (of a factual or normative nature) in the
selection of means. We also have to think, for example, of expectations con-
nected with social roles,24 of predetermined communication channels along
which a decision is to be prepared and validated, and not least of the qual-
ities (values, skills, social responsiveness) of the people who are involved in
decision making. Obviously, the term focuses attention first and foremost on
the structural constraints of the particular decisional scope25.
Initially the distinction of decision premises and decisions had a very lim-
ited meaning. It served to limit the possible scope of decision rationality.
This is expressed by the term “bounded rationality”. How large this scope
may be is a question of data processing and decision-making technology.
From the start, therefore, the theory aims only at sub-optimal problem solu-
tions – just as structural functionalism presents itself explicitly as a second-
best theory. If we wanted to break down the entire world into variables and
to correlate each one with all the others, or if we assumed that all prices on
the market influenced one another, then knowledge and decision capabilities
would obviously be overtaxed. And because this is so, we can always reckon
with a world, with a society and with organizations that are geared to
constraints on their own capabilities of observation. All markets of all
22 Cf. Simon/Smithburg/ Thompson (1950), especially pp. 57 ff.
23 This is, however, immediately qualified: “… we are concerned primarily with behavior that
is conscious and rational.” This is why he subsequently speaks synonymously of “decision
premise”. See, e.g. Simon (1957), p. 201. We go along with this terminology.
24 Simon (1957), p. 201.
25 Almost at the same time an epistemological structuralism emerges that limits the scope of
variation, which in itself could be relevant, to what is doable by means of a structural defini-
tion of objects. We should think here especially of Levi-Strauss and also of Parson’s version of
analytical structural functionalism.
93
07 Kap 4. Luhmann 05-11-11 16.18 Sida 94
organizations are from the outset markets with limited substitutive compe-
tition,26 in the same way as public opinion, towards which political organi-
zations gear themselves, is compartmentalised into different topics from the
outset.
This structuralist assumption of “bounded rationality” leaves two paths
open to account for a change even of structures. This question can be con-
sidered to be a problem of the evolution of a population of organizations,
and accordingly it can be left to evolutionary theory.27 This takes it out of
decision theory and simply postulates a certain non-coincidence of evolu-
tionary reward or penalty for decisions – whether they are rational or irra-
tional, though definitely not rationally determinable. However, the chances
of the evolutionary selection cannot be included in the decision process as
premises for decisions; these remain, in accordance with the self-conception
of evolutionary theory, dependent on chance.
The counter-concept is to be found in a theory of planning or steering
organizations on the basis of decisions on decision premises.28 Depending
upon the nature of the decision premises, this may be personnel planning,
programme planning, or organizational planning in the narrower sense, i.e.,
establishing communication channels in the organization. Seen from the
viewpoint of rational decision making, the interdependence of these differ-
ent types of decision premises is obvious. But that in turn implies an open
model, in which everything varies with everything else. The tendency in
practice is therefore to proceed from the results of prior decisions – espe-
cially from organizational objectives, but also from a given (and not readily
adjustable) staffing level or a given size of the system, which can only be
changed with major internal repercussions.
These constraints are manifested in the structural model of the “position”,
which is used particularly in public administration.29 A “position” is a prin-
ciple of identity that enables variation. One can change the incumbent, the
job profile or the reporting relationship – but not all at the same time. The
26 A consequence of this, by the way, is that the generalised readiness to accept money assumes
a paradoxical form in economic theory (cf. Orléan 1992). This is underlined by the individu-
alism of the (decision theories of) economic classics, but we would get the same result if we
omitted the external reference “individual” and merely considered the payment operations of
the economic system.
27 Thus Nelson/Winter (1982).
28 Cf. Luhmann (1971a), especially pp. 66 ff. Today, we are considerably more sceptical as
regards planning possibilities, which is reflected in the use of the word “steering” rather than
“planning” (or also regulative policy). See Willke (1983); Glagow/Willke (1987); Willke
(1992); Luhmann (1992b).
29 The private sector calculates more directly with financial resources that are available and/or
that could be meaningfully invested because here the financial limitations have an immediate
effect. At the same time, these tighter constraints make organizational planning more flexible
in those areas that they enable.
94
07 Kap 4. Luhmann 05-11-11 16.18 Sida 95
“position principle” therefore also serves to visualise constraints on change

by the status quo. Seen from this perspective, a principle of “bounded real-
ity” exists on the level of the decision on decision premises.
On the basis of the theory of autopoietic systems, i.e. of systems that
reproduce themselves through their own operations, we can give greater sig-
nificance to the term “decision premises”, which goes beyond the problems
relating to the rationality of decisions (but which, of course, does not rule
out that these problems continue to be addressed and researched). Every
decision is always in effect a decision premise for other decisions, whether
or not in the form of a rule, a programme, or a directive. Because once some-
thing has been decided, it need not normally be decided again. And if a deci-
sion has to be repeated, that decision will not be the same decision but is
forced by the very fact of the “procedure” into the form of a repetition. If
this applies, it follows that subsequent decisions interpret given premises, in
their recursive process of re-actualising their meaning, as decisions – irre-
spective of whether a choice had explicitly been made and what that choice
was. In this sense, we have an operational closure of the system on the basis
of its own decisions – regardless of what the psychic systems that are
involved experience as decisions.
Because decisions are premises for other decisions, the original paradox is
transformed into a new form, i.e., into the double meaning of enabling and
constraining subsequent decisions: future decisions are only rendered possi-
ble by the existence of past decisions. Or, in other words: in the communi-
cation, past decisions put future decisions under the strain of expectations,
which makes it inevitable that the subsequent decision is marked in the com-
munication as a decision.30 The reduction of complexity, which is achieved
each time a decision is communicated, also serves to open up the scope for
subsequent decisions. The system constantly oscillates between constraining
and extending decision possibilities and in this way secures its own
autopoiesis. This may be overlaid with objectives that serve the system as a
self-description. But these never lead to the situation where the system heads
funnel-like towards a natural end (telos) via increasingly narrower decision-
al scopes. If this were the case, it would not be an organization but a proj-
ect (which may be complex to a greater or lesser degree, long-term, and in
need of being organized).
In any case, the decision on decision premises has to be taken on the level
of second-order observation.31 Decision premises are planned for decision
makers; i.e., with a view to how these first-order observers observe their
decision situations. The same is true of every retrospective interpretation of
premises as decisions. In logical terms, this leads to considerable problems.
30 For a more detailed account see Luhmann (1984a).

31 For a fundamental account, see von Foerster (1981).
95
07 Kap 4. Luhmann 05-11-11 16.18 Sida 96
It requires a multi-valued logic, which is not yet available (or at least not in
the form of operational instructions).32 But this does not necessarily rule out
that we explicitly use the opportunities inherent on the level of second-order
observation. On this level we may especially take into account the con-
straints of “bounded rationality” (e.g., that the decision maker can neither
know nor change himself) and proceed generally from the assumption that
the observing observer cannot see that he cannot see what he cannot see.33
As a rule, we associate terms such as planning and steering with the notion
that this is done from above and is thus justifiable on the basis of a better
overview, though without precise knowledge of details. This may be the case
in some instances. However, the theory of the second-order observer is not
dependent on the mystifications of the hierarchy. It emphasises that this is a
very specific observation in very specific situations; and of course, an obser-
vation which has to base itself upon what it cannot see. We cannot evade
this fundamental paradox.
Uncertainty absorption
We have said that practice is always oriented according to the given results
of earlier decisions. This can be formulated in more general terms by distin-
guishing between decision and uncertainty absorption. Here again, the
source is Herbert Simon; but, as before, subject to considerable qualifica-
tion.34 “Uncertainty absorption takes place when inferences are drawn from
a body of evidence and the inferences, instead of the evidence itself, are then
communicated.”35 Now this is not exceptional; it happens whenever deci-
sions are communicated. The term “uncertainty absorption” expands the
term “decision premise”, taking it from the structural level to the processu-
al level. Uncertainty absorption takes place, we can therefore say, when deci-
sions are accepted as decision premises and taken as the basis for subsequent
decisions. In the style of Max Weber’s definition of power we can also add:
no matter what this acceptance is grounded in. And the whole point is: it is
not the decision operation itself, but a process that connects decisions.36
32 See as follow-up to George Spencer Brown and Gotthard Günther: Esposito (1992).
33 See also Luhmann (1991b).
34 See March/Simon(1958), pp. 164 ff. The qualification mentioned in the text is that with
“uncertainty absorption” only one specific variable is identified, which presumably correlates
with other variables (especially influence), but which is not explicitly related to the distinction
between decision and uncertainty absorption. The theory, therefore, does not sufficiently con-
sider the fact that uncertainty absorption, as the other side of the form of “decision” can never
be the subject of decision.
35 March/Simon (1958), p. 165.
36 See also Weick (1979), pp. 142 f. and passim, on the reduction of equivocality as an inter-
personal process.
96
07 Kap 4. Luhmann 05-11-11 16.18 Sida 97
This may be a question of authority in the strict, traditional sense, in other

words the assumption that the decision maker is able to adequately explain
and substantiate his decision. Authority in this sense depended in older
societal forms on (more or less) exclusive access to information, particular-
ly to information with normative relevance (religion and law). When these
barriers come down – and this happened with the development of the mass
media, which began with book printing – sources of authority have to be
replaced. Authority becomes increasingly pretentious, it needs to be “stage
managed”, but it is no longer able to prevail over opposing interests.37
Moreover, within organizations particularly there are innumerable func-
tionally equivalent mechanisms of uncertainty absorption which can com-
pensate loss of authority and which can lead to the uncontrolled acceptance
of the communicated reductions. It may also be feared that further questions
might be seen as a provocation and do more damage than good. The reasons
might lie in the responsibility structure and in mutual respect for another
person’s area of responsibility; or ultimately in the economy of labour
(others might call it the typical apathy) of the organization’s members,
because, after all, everybody has enough with his or her own workload.
In the field of organization theory there has been a switch from authority
to uncertainty absorption as the more general term – though not necessari-
ly using this terminology. On the structural level it is mirrored in a change
from hierarchy to heterarchy;38 or in a change from the principle of “unity
of leadership”, which is limited by the superior’s “span of control”,39 to the
principle of “redundancy of potential command”.40 This certainly does not
mean that the hierarchical order has been abandoned. It remains intact as an
emergency aggregate, a network to which it is possible to switch when prob-
lems cannot be resolved otherwise. But the normal functioning of an orga-
nization cannot, and here there is consensus, be described in this way. And
management consultants, too, apply different theories today to explain or to
make improvements, for example, the (in the meantime outdated) theory of
group dynamics which had led to the distinction between formal and infor-
mal organization and to “organizational development” programmes,41 or,
more in line with the above critique of the classical term decision rationali-
ty, the concept of the matrix organization.42 In this way, attempts are made
37 On the assumed effects of television cf. Meyrowitz (1985), esp. pp. 62 ff., 160 ff. With
regard to “opposing interests” we should consider particularly the loss of authority of the
experts in the ecological debate. Their statements are only believed when they appear on the
“right side”.
38 “Heterarchy” in the sense of the term coined by McCulloch (1965). It is inspired by aes-
thetics, neuropysiology and economics.

39 See, for example, Fayol (1925). Cf. also Gulick/Urwick (1937).
40 Thus Pask (1970), p. 2. Cf. also Baecker (1993): pp. 130–131.
41 Cf. for a critical account Wimmer (1991).
42 See Reber/Strehl (1988).
97
07 Kap 4. Luhmann 05-11-11 16.18 Sida 98
to bring the network of organized decision processes closer to the problems

and risks of uncertainty absorption. Put somewhat exaggeratedly, it is recog-
nised that the problem of rationality lies in its unattainability and that there
is no way of determining how to get to second-best solutions.
The inevitability (and thus organizability) of the simple fact of uncertain-
ty absorption cannot be challenged once we have abandoned the idea of
viewing the overall system under the maxim of optimal rationality with only
one correct decision in each case, which in the final analysis has to be
accounted for at the top of the organization. Obviously, no organization
could function if all past decisions had to be reviewed each time a decision
was taken. Instead, precedents crystallise, which serve as the basis for future
decision-making processes.43 This can then lead to subtle techniques of con-
sidered inattention with which to avoid the emergence of unwelcome prece-
dents.
These considerations also demonstrate that uncertainty absorption cannot
be the content of decisions. It is not the task of decisions to apply their deci-
sion programme and to absorb uncertainty as well. Uncertainty absorption
is neither a primary end nor a secondary end of decision making. It happens
automatically whenever decisions are taken in a communicative system.
If the theory holds thus far, we can say that the autopoiesis of the system
leaves its marks in uncertainty absorption. This means the production
(= reproduction) of decisions out of decisions is only possible in this form.
Structures, on the other hand, can be introduced and changed through deci-
sions (which from a processual perspective again leads to uncertainty
absorption, i.e., they can only be produced as part of the autopoietic
process). For this reason, uncertainty absorption is an invariable feature of
organizations, while structuring is a variable narrowing down of decision
contents.
Accordingly, uncertainty absorption takes on a particular significance in
relation to the other concepts of organization theory. It is not one of many
variables, but rather the organization system itself, as seen in a temporal
cross-section. This does not of course mean that no improvements are con-
ceivable here, but they must always be on the other side of the decision,
where the decision is actually taken. And, to return to what was said earli-
er, they come up against distinct boundaries in that they are limited to
“bounded rationality”.
We have already had the opportunity to point to parallels between orga-
nization theory and epistemology. Now, we are able to see that organiza-
43 The cult of precedents in the legal system can ultimately be ascribed to the impossibility of
re-opening closed decision processes, whereby the grossness of this solution in common law
has led to subsequent developments of detailed analysis of ratio decidendi and to possibilities
of distinguishing and overruling. On the attribution of this development to problems of deci-
sion making see especially Heiner (1986).
98
07 Kap 4. Luhmann 05-11-11 16.18 Sida 99
tions construct their own reality and how they go about doing so. It has to
take place in an unknown world, in a world which exists simultaneously
with all decisions and which changes concurrently with all decisions. This
can be understood in the sense of functionalist psychology (Brunswik,
Bruner) or in the sense of older cybernetics (Ashby) as being overtaxed by
environmental complexity, as the absence of “requisite variety”. But we can
also formulate it more radically: every organization operates in a world that
it cannot know. This world is transformed through uncertainty absorption
into a known world; it is replaced by a known world. This requires in ret-
rospect a first decision, which inscribes a distinction into the world, for
example, by defining a purpose, by forming a “coalition” of (future) mem-
bers with a corresponding clientele as environment,44 or simply by estab-
lishing another organization. As always, the first decision (and that means:
what in the recursive network of subsequent operations is taken to be the
first decision) brings about the unfolding of the paradox, obeying Spencer
Brown’s first rule: draw a distinction and, by doing so, mark out an area
within which decision making can take place. The rest is left to uncertainty
absorption.
Clearly, this course of action is unavoidable, but it has problematical
implications. There is a tendency to cling, so to speak, to the outcomes of
years of uncertainty absorption, which validate themselves as a familiar
world. This is particularly true where risks have been taken and those risks
have paid off;45 but also where people have grown accustomed to conflicts
and fail to recognise that those conflicts no longer exist.46 In this sense, there
are innumerable organizations that live of their failures because it is pre-
cisely these that provide a reliable basis for decisions. Earlier organization
theory saw an innovation problem here, carried out research on strategies
that promised success for the future, and banked on leadership.47 Today,
companies seem more likely to put their money on external consultants
who, because they are from outside, probably find it easier to restore uncer-
tainty.48
44 See Simon (1948), pp. 110 ff.

45 Cf. Japp (1992), with further references.
46 Examples can be found particularly in political organizations. Consider the handling of the
asylum problem in the Federal Republic of Germany or the education policy controversies in
North Rhine-Westphalia.
47 For a typical example see Selznick (1957).
48 Of course, I do not claim that the philosophies of consulting are sworn to this concept and
that they dispense with the offer of “superior knowledge” (which often fails because of inac-
curate knowledge of the internal organisational environment, or which understandably trig-
gers opposition.) After all, there are numerous connections between family therapy and
management consulting, which are inspired by system concepts, or by cognition-theoretical
constructivism, or maybe even by the paradox concept of the Milan school (see Selvini
Palazzoli et al., 1983). For more recent thoughts on the subject, see especially Wimmer (1992).
99
07 Kap 4. Luhmann 05-11-11 16.18 Sida 100
Rationality and motivation

The distinction between rationality and motivation (or rationality and
action), which Nils Brunsson suggests,49 reacts likewise to problems which
are brought to light by decisions that are expected to be rational.
Rationalisation requires decisions to be broken down into sub-decisions and
these in turn into sub-sub-decisions. It is easy to see in this a principle of
organizational growth, which at the same time reinforces the mode of con-
necting decisions through uncertainty absorption. The same applies, inci-
dentally – in the social dimension – to efforts to democratise decision mak-
ing.50
If we analyse such findings, we shatter the old notion that man is a ration-
al and social being who cherishes and enjoys reason and consensus. This
may well be, but the observation of decision making in organizations con-
veys a different impression, as if man were constantly operating on the other
side of his “nature”, in the domain of corruption and vice. But let us not
dwell upon such anthropologica – however much they might lend force to
the plea for a humane organization. It is sufficient for us to understand that
processes like rationalisation or democratisation do not take into account
the motivation needed for taking action. Somebody who grows weary of the
process of rationalisation (and democratisation, which we will leave aside in
the following) will hardly be prepared to commit himself to implementing a
decision that could have been taken differently on hundreds of counts. This
leads to efforts being made to establish a certain organizational culture, a
“corporate identity”, an ideologisation of corporate goals, which refer pre-
cisely to this juncture between rationality and motivation; but this effort
itself reveals the problem: it simply produces a new mystification at the
point where, out of habit, we suspect a paradox.
This paradox can be unfolded quite easily by differentiating between
social and psychical systems, and understanding such terms as “motive”,
“commitment”, “engagement”, and “conviction” psychologically. But this
just leads to a repetition of the same problem within the social system
“organization”, because “motives” also play a role first and foremost in
communication in that they convey, request or recall commitments. Motives
in this sense are stage directions in the dramaturgy of action.51 They are also
conditions of possible understanding and thus conditions of a possible anti-
cipation of understanding, or ultimately: themes of being prepared to render
account.52 And last but not least, we can say that “motives” are the point of
49 See Brunsson (1985). See also Brunsson (1989).

50 On this comparison: Luhmann (1981a), pp. 344 ff.
51 See Burke (1962).
52 Cf. Mills (1940); Blum/McHugh (1971); Zollschan/Overington (1976); Sarat/Felstiner
(1988).
100
07 Kap 4. Luhmann 05-11-11 16.18 Sida 101
connection for structural couplings with psychic systems; in other words,

they are those aspects with which communication indicates to itself that it
depends on psychic operations which remain alien to it; or vice versa: they
are the responses that a psychic system holds in store when it considers what
it would reply if it were asked why it does what it does.
These thoughts on the term “motivation” indicate that the communica-
tion of decisions comes to a bifurcation where it has to choose between the
language of rationality and the language of motivation. Rationality is con-
sidered to be self-motivating. We need not ask about motives when it is an
established fact (or it can be assumed to be an established fact) that the right
decision has been taken. It is only in the modern world, which always holds
a second-order observation in store, that the question of latent, “uncon-
scious” etc. motives is raised. It is only logical to keep the language of
motives on the same level as the language of rationality and to presuppose
this distinction as implicit in decision making.53 But in the decision practice
of organizations you would get some rather surprised looks if you were to
constantly question the decision makers about their motives. The motives
are carried along unmentioned, as it were, as an aspect of the mystery of
decision making. But the absence of motives becomes noticeable when diffi-
culties arise and people are required to do more than usual. The language of
motives is then overlaid in the form of a corporate ideology, not least as a
substitute for what failed to work in Taylorism or in group dynamics and
“organizational development”. In this way, therefore, the presence of a dis-
tinction, a bifurcation of rationality and motivation is simply obscured.
Here, too, unity is only to be had in the form of paradox.
Perception and communication

Terms like “decision rationality” or “motivation for action” are normally
taken to refer to achievements of the people participating in the organiza-
tion, people working “in it”. Consequently, we ought to subject the corre-
sponding phenomena to biological, neurophysiological or psychological
examinations. But this would probably not be very effective. We have, there-
fore, “deconstructed” the biological or psychical conceptualisations. Here it
is helpful to trace back what is commonly understood as decision to a para-
dox. To unfold this paradox, the social system “organization” employs spe-
cific terms: “rationality” means that the decision satisfies specific criteria
and “motivation” that the commitment and loyalty of individuals can be
53 This is my understanding of Spencer Brown. Under the definition “Distinction is perfect

continence” we read: “There can be no distinction without motive, and there can be no motive
unless contents are seen to differ in value” (Spencer Brown 1979, p. 1). But this terminology
plays no role in the unfolding of the calculus of indications.
101
07 Kap 4. Luhmann 05-11-11 16.18 Sida 102
won through their own behaviour. So far, so good. A by-product of this

analysis could be that it clears the vision for tackling afresh the question of
what exactly constitutes the mental operations that allow people to partici-
pate in an organization without becoming parts of that organization. It must
be something that the organization itself cannot be. Here we have to think
primarily of perceptions.
This may come as a surprise, because a long tradition sees the particular
achievement of the individual especially in the ability to think and in the
ability to subordinate his will to his ratio (intellect/reason). This emphasis
was a result of defining man in distinction to the animal, as a consequence
of which the instinctive features of his nature were classified as inferior.54
Regardless of our views on thinking: it is by no means as reliable and peace-
ful as perception. It is only through perception that the mind experiences the
directness of its relation to the world, whereby we mean “directness” in the
sense of not being conveyed through distinctions, nor even through the dis-
tinction of direct/indirect55. It is only perception that creates the medium
“world”, in which all distinctions, including perceived objects and events,
are inscribed. How can it be explained that such an important fact has been
virtually ignored in organization studies? Is it due to misdirected thinking
because of the humanistic tradition? Or is it because it was possible in this
way to unfold the paradox of decision by imputing a “subject” to it?
The least room for perception is given in public administration. What we
find here are core structures of the organized field of society where decision
is not only a process but also, and above all, a product of organization.
Participants have to be able to talk/listen and write/read. But because all of
this takes place in the spoken form, the scope for variation between differ-
ent individuals is narrow. What is perceived is based on an intention, and in
this respect it is a socially verifiable construct or, at least, recognisable as the
outcome of a decision.
But these are exceptional circumstances. We are able to see more clearly
from other organization forms how the directness of the perceived world
serves to filter decision operations. This is true, for example, of the eye/hand
54 However, it should be added that the law of nature included animals. Thus, e.g. Ulpian in
Digests 1.1.1.3: “Ius naturale est, quod natura omnia animalia docuit.” But the consequence
of this was that social development in its entirety had to be represented as a deviation from
natural law: marriage as a deviation from the natural reproduction instinct; property owner-
ship as a deviation from the natural accessibility of resources; slavery, servitude, employment
etc. as a deviation from natural freedom. 17th and 18th century contractual arrangements were
still based on this precondition. Today’s advocates of natural law appear to be unfamiliar with
these texts.
55 It should be noted that this directness is an illusion produced in the mind. The central nerv-
ous system, of which the mind is oblivious, operates with highly complex processes of dis-
crimination and selection.
102
07 Kap 4. Luhmann 05-11-11 16.18 Sida 103
coordination in industrial work, which only occasionally has to report

something. It is true above all of the vast, personnel-intensive areas of orga-
nization, which, by analogy with Anglo-Saxon usage, we could call “field
work”: the activity of teachers, of supervisors of all kinds.56 Here, the so-
ciologist observes a very particular organizational solution of the “inter-
face” problem of perception and communication, namely conceded autono-
my.57 Work of this kind is typically unsupervised work, carried out by an
individual in accordance with professional standards and – it is expected –
with commitment to one’s tasks. Perceptions cannot be repeated and can
only be checked with difficulty, by means of time-consuming reconstruction.
Directives are either heeded or neglected, followed or deliberately ignored,
depending on what the individual case requires due to its perceived com-
plexity. Relations with the main organization are established through
written reports or proposals, which seek to meet the requirements but which
provide only very selective accounts. In this way one “cools down” the para-
dox that the granted freedom of working without supervision induces cau-
tion with regard to (unexpected) supervisions. This can be avoided by draw-
ing up reports, i.e., by influencing the respective decision-making process in
the organization; this at least gives the senior people in administration the
chance to feel that they are performing their assigned tasks. There can be no
more than a loose coupling with the supervisory bodies, and in the concrete
decision situations we will typically find a combination of legality and ille-
gality. The distinction between perception and communication is trans-
formed into an ambiguity that cannot be standardised. In other words, there
is no technology for organizing such perception-dependent activities that is
sufficiently resistant to outside influences.58
Is it possible to deny the autonomy that resolves the problem and makes
it easier to sort out the decisions? This problem is currently being discussed
in connection with the supervision of high-risk, large-scale technical
plants.59 Alarm signals are installed here, which are coupled with the tech-
nical plant in such a way that they are perceptible whenever a breakdown
occurs. They are perceptible! But whether they are perceived and how they
are interpreted is decided in different systems and, more especially, in sys-
tems of a different type. The long evolutionary selection process of the
56 For an example from the field of water protection see Hawkins (1984), esp. pp. 57 ff. There
is a host of studies on unsupervised behaviour and self-protection through discrepancies in
reporting in the field of police work. See, for example, Rubinstein (1974); Brown (1981);
Aaronson/Dienes/Musheno (1984). Cf. also McCleary (1978), esp. pp. 145 ff.; Prottas (1979),
esp. pp. 26 ff., 61 ff.
57 On the relationship between unpredictability and autonomy see esp. Prottas (1979), pp.
111 ff.
58 See also Luhmann/Schorr (1982).
59 See, e.g., Halmann/Japp (1990).
103
07 Kap 4. Luhmann 05-11-11 16.18 Sida 104
central nervous system has programmed psychic systems to perceive irregu-

larities, movements, and changes, and they keep up the directness of their
relation to the world only as a distinction vis-à-vis such irregularities. But
irregularities are recognisable because they stand out from what is perceived
as normal, and it is in normality that deficiencies may escape observation.60
Moreover, human perceptions are always interpretable perceptions; they
construe, they create an abundance of references, they can mean “one thing
or the other”, and this makes them unreliable for the organization.
Organizations as decision-based, autopoietic systems are not capable of
perception. Accordingly, they cannot incorporate perceptions into their own
network of structure-determined operations. Perceptions create a different
world. Because humans are capable of perception, they tend to under-
estimate this discrepancy and, as it were, to see themselves as part of the
organization. This may be one of the reasons why the paradox of decision
has been unfolded towards responsible, rational subjects, who only occa-
sionally do the wrong thing. Today, the analytical tools available to systems
theory make such symbiosis appear questionable.
From principle to paradox

To sum up, we can say that the changeover from the traditional notions of
the ultimate foundation (principles, laws of nature) to paradox has the
advantage of allowing for a better systematisation in organization theory.
Paradoxically, it refers to the different ways of getting rid of the paradox
again. This non-logical, but non-arbitrary procedure can be described as a
search for suitable distinctions. We spoke earlier of “unfolding” the para-
dox. In doing so, the central problem of the organization – decisions – is
placed on one side of the distinction. With reference to Spencer Brown,61 we
could speak of the “inside” of the form, or with reference to Gotthard
Günther62 of the designation value of the binary scheme. The other side of
the distinction (the “outside”, the “reflection value”) would remain open to
further designations. These other sides need to be filled with designations in
order to render the distinction technique theoretically fruitful.
If we designate the other side as “decision premises”, we get a structural
understanding of the organization. If we designate it as “uncertainty absorp-
tion”, we get a processual understanding of the organization. And finally, if,
consistently with a normative-rational understanding of decisions, we see on
the other side a problem of “motivation”, we get to system/environment
60 Cf. Goffmann (1971). For the ethnological dimension of this distinction, see also Leach
(1982), pp. 111–112.

61 See footnote 4.
62 See, e.g., Günther (1980), pp. 140 ff.
104
07 Kap 4. Luhmann 05-11-11 16.18 Sida 105
problems relating to employee behaviour. We have used the term structural

coupling to mark these problems. On the other side of this system boundary
we will primarily find consciously directed perceptions; and the organization
has to find ways of accepting that this kind of operation will remain alien
and inaccessible to itself.
These dispositions build on existing research, but they also enable us to go
beyond it. They allow us to describe organizations as autopoietic systems
that reproduce decisions through a self-produced network of decisions.
Apart from characterising the individual operations as decisions, this
requires additional concepts that we already have at hand: (1) a term for
autopoiesis itself (= uncertainty absorption), (2) a term for the recursive use
of decision outcomes as structures for the selection of subsequent decisions
(= decision premise) and, finally, (3) a term that describes how the environ-
ment is involved in the internal operations without however being able to
contribute operations itself (= motivation).
Such a conceptualisation is relevant not only for organization theory. It
also opens up new perspectives for social theory and helps us to understand
better important societal function systems. This is only partly due to the fact
that these function systems are dependent upon organizations for their oper-
ations. It is also because in present-day society, relations between past and
present have to be explicitly coupled through decisions; they are no longer
the result of given necessities and impossibilities, or of history itself, or of
the imperatives of “practical reason”. But what does it mean that society has
to cope with decisions as paradoxes (which is unfolded in one way or
other)?
Both in the cognitive and the normative areas of societal communication,
i.e., both in science and law, there are signs of an increasing need to react to
definitions of situation that can only be described as self-made uncertainty.63
The original idea for this can be found in religion, in the attempt to achieve
social discipline among a population, which, in the Middle Ages, was still
mainly unconverted to Christianity, by dramatising the certainty of salva-
tion. Significantly, this attempt was based on creeds and confessions, in
other words on self-reference that had to be accepted.64 In later times this
model was copied and generalised, and today, after a phase of religious
enlightenment, we face an analogous societal enlightenment. In science, this
is reflected in the postulates of treating all scientifically proven (!) knowl-
edge merely as hypotheses and, more recently, of radically constructivist
epistemologies. In the legal system we find similar tendencies in a positivism
that is disconnected from all external sources of legitimation, in an
63 For the legal system, see Ladeur (1992), particularly the third part on the interpretation of
constitutional rights (pp. 176 ff.).

64 See the seminal article on the subject by Hahn (1982).
105
07 Kap 4. Luhmann 05-11-11 16.18 Sida 106
orientation towards interests and values, and in undefined legal concepts

that leave the outcome completely open. For both spheres the discourse
theory of Jürgen Habermas provides corresponding perspectives: discourses
about claims to the truth or norms no longer allow for prefiguration
through natural history, moral principles, or transcendental invariants of
practical reason;65 they relate to an as yet unknown future, in which a rea-
sonable decision that makes sense to everyone who is affected by it has to
be found, yet. The theory of self-referential, operationally closed systems
comes to the same result via a different route66. All information and, thus,
also the absence thereof, all certainty and uncertainty, is a construct of the
system itself and is determined by the distinctions with which it observes
the world. Innumerable other examples could be cited, for instance, from
the field of systems theory, from the research on risk, or from research on
the interdependencies between technology and ecology.67 The intellectual
climate of the 1990s seems to be resigned to the fact that the uncertainty of
the future is the only invariant on which all decisions will be grounded. But
if that is the case, it will be necessary to change decision theory itself from
principle to paradox. Because only then will we be able to avoid scenarios
of horror and gestures of despair and instead try to cope effectively on the
basis of the available insights.
A theory of this kind is therefore not restricted to specific areas of orga-
nization such as public administration or production companies, schools or
banks, political parties or courts of law. It also extends to the individual
function systems of society in which organizations have assumed very dif-
ferent roles. This requires a high level of abstraction. But it does not neces-
sarily mean that we have to work with unanalysed abstractions or with
vague terms. Rather, we should make the stipulations of our theory as trans-
parent as possible. Because only then can we control what needs to be
changed when parts of the theory fail to work.
– Translated by Karen Finney-Kellerhoff and David Seidl
Acknowledgements
This text was originally published in German in Verwaltungs-Archiv 84,
1993, pp. 287–310.
65 With all desirable clarity Habermas (1996).

66 See Luhmann (1990c; 2004). Cf. also Luhmann (1998) – especially pp. 75 ff.
67 Cf., e.g., Selvini Palazzoli et al. 1983; Simon (1990); Halfmann/Japp (1990).
106
08 Kap 5. Knudsen 05-11-11 16.20 Sida 107
Chapter 5
Displacing the Paradox of

Decision Making.
The management of contingency in
the modernization of a Danish county
Morten Knudsen
“When the unclean spirit is gone out of a man, he walketh through dry
places, seeking rest, and findeth none. Then he saith, I will return into my
house from whence I came out; and when he is come, he findeth it empty,
swept, and garnished. Then goeth he, and taketh with himself seven other
spirits more wicked than himself, and they enter in and dwell there.”
Matt. 12: 43–45
This chapter aims to demonstrate the usefulness of Niklas Luhmann’s sys-

tems theory for empirical research. It sets its point of departure in modern-
ization processes in the Danish public sector, where – as in many other
OECD countries – such processes led to major changes during the 1980s and
1990s. A steady stream of reforms has transformed the basic structures of
the public sector, new technologies of steering have emerged, ideals of
change and flexibility have been promoted, and new financial systems have
been implemented, just to mention some examples. This process of modern-
ization is further characterized by “decentralization” – the differentiation of
the public sector into multiple self-referential public organizations. This
chapter will describe some of the organizational dynamics in the modern-
ization process from a systems theoretical viewpoint while focusing on the
question of how the organization – in and through a process of self-
establishment – comes into being. The empirical basis for this study is pro-
vided by a Danish public sector organization, the Frederiksborg County
Health Authority. The chapter focuses on its emergence as an organization
and presents a number of observations, which are primarily generated by
Niklas Luhmann’s concept of the paradoxical character of decisions.
Further, two different types of “deparadoxization” are described and some
of their effects analysed.
107
08 Kap 5. Knudsen 05-11-11 16.20 Sida 108
In the first section, the key concepts used in this chapter are presented and
the strategy of analysis is explained. The following section shows how the
county’s self-referential Health Authority emerges and in the same process
manages the paradoxicality of its decisions. The emergence of the organiza-
tion and its subsequent deparadoxization is followed by increased uncer-
tainty and instability. This makes it relevant to look for complementary
strategies for deparadoxization. The following section demonstrates how
these strategies are also accompanied by uncertainty and subsequently dis-
cusses the relations between the analysed deparadoxization strategies and
the processes of reform in the public sector. The concluding section offers a
summary and an evaluation of the pertinence of Luhmanns theory to empir-
ical analysis.
Key concepts
In Luhmannian systems theory, social systems are seen as consisting of com-
munication. Decisions, whilst being the central operation of the organiza-
tion, are but one type of communication. The basic analytical movement of
systems theory is to proceed from the assumption that communication is
improbable. In so far as we experience the factual existence of organizations
on an everyday basis, the conceptualization of organizations as improbable
entities may strike one as odd. Nevertheless, this idea forms the point of
departure for Luhmann’s analytical strategy. The idea of the improbable
decision and the improbable organization is the basic assumption that gen-
erates observations. Empirical analysis involves showing how the presumed
improbability is overcome and the system is made possible. This section will
present the key concepts used here, in order to show what is meant by the
“improbability of the organization”.
According to Luhmann, social systems consist of recursively connected
elements. These elements are seen as communicative events – not as actions
or individuals. Communication, however, is only really communication if
– as Luhmann puts it – the communicative event is understood and “used as
the basis for connecting with further behaviours” (Luhmann 1995f, p. 141).
Communication comes into being as communication only when commu-
nicative events consecutively connect to each other. Singing in the bathroom
is thus not communication, unless, of course, somebody shouts “Quiet!”
Luhmann sees decisions as the basic elements of the organization and the
organization is thus defined as a network of recursively connected decisions.
This means that organizations simply consist of decisions referring to other
decisions. Without these connected decisions there is no organization. The
organization is self-referentially closed on the basis of its decisions. This
means that the organization cannot make decisions outside of itself, in its
108
08 Kap 5. Knudsen 05-11-11 16.20 Sida 109
environment, and that the environment cannot make decisions inside the
organization. For instance, IBM cannot decide the strategy of Macintosh,
and vice versa. In the framework of systems theory, the discussion of how
the organization emerges is posed as a question of organizational strategies
for the increase of probability for the connectivity between decisions.
The problem of connectivity is crucial for the emergence of the organiza-
tion, but no less for individual decisions. Decisions are “virtual” (Baecker
1999c, p. 139) in so far as decisions are not decisions until other decisions
connect to them. If decisions do not gain connectivity they were not deci-
sions in the first place, but – instead – organizational noise, as it were. The
connectivity of decisions is thus essential to the constitution of the organi-
zation. The necessity of this connectivity is an everyday experience in orga-
nizational life. What is really a decision can only be discerned in hindsight;
for instance, the approval of minutes at meetings: among the “decisions” of
the previous meeting, some are singled out as proper decisions at the fol-
lowing meeting. Further, most members of organizations have witnessed the
making of decisions that are quickly forgotten and erased or ignored in sub-
sequent organizational communication, which again means that they turn
out not to be proper decisions, but just “noise”. Decisions are continually in
danger of decay or failure of connectivity.
According to Luhmann, the decision’s paradoxical nature is what threat-
ens this connectivity and thus makes the organization improbable. Let us
take a closer look at the concept of decision to see what is meant. Luhmann
relates the concept of decision to the concept of observation and to the con-
cept of distinction (Luhmann 1993d, p. 289; 2000c, p. 126). To observe is
to indicate within the frames of a distinction, he says. “This is a profitable
investment” is, for instance, an indication within the frames of the distinc-
tion “profitable/non-profitable”. What transforms an indication into a deci-
sion is that it is possible to indicate both sides of the distinction, which
means that the distinction has the form of an alternative. Normally, the
question of whether to indicate “profitable” or “non-profitable” does not
constitute a decision. However, it does constitute a decision when an orga-
nization determines which investment to make, because it involves guessing
which investment will yield profit and which will not. The decision is only
a decision in so far as it has another side (an alternative decision could be
made). Thus, it is both the same (the indicated decision) and something dif-
ferent (the distinction). In the centre of this paradox (see Luhmann 1991a
and 1993c for a presentation of his general concept of paradox) lies an
undecidability (see also Derrida on “the undecidable” in Derrida 2002,
p. 31), which makes the decision contingent – that is, neither impossible nor
necessary. The reason for this undecidability hinges on the fact that either
side of the alternative may be indicated (Luhmann 2000c, p. 133). In other
words, a decision is not necessary, but contingent: it is possible to indicate
109
08 Kap 5. Knudsen 05-11-11 16.20 Sida 110
either side of the alternative. According to Luhmann “something is contin-

gent insofar as it is neither necessary nor impossible; it is just what it is (or
was or will be), though it could also be otherwise” (Luhmann 1995f, p.
106). The concept of contingency is used in the Aristotelian sense and des-
ignates that something is “neither necessary nor impossible”.
The decision must communicate itself as a decision, but by doing that it
also communicates its own alternative. A decision cannot help but commu-
nicate its own self-critique, i.e. communicate that it could also have been
made differently; which is central to Luhmann’s theory (Luhmann 2000c, p.
142). In the communication of a decision lies a self-destructive mechanism,
a tendency towards deconstruction. The present cannot help referring to the
absent – to use a deconstructive expression. The peculiar thing about deci-
sions is, on the one hand, that they cannot help but express their own con-
tingency, which precisely makes for a problem of connection: why connect
to a decision which could have been made differently – and which cannot
help but communicate this possibility? On the other hand, it is clear that the
factual existence of organizations is dependent on a factual connectivity of
decisions. The very existence of organizations testifies to the fact that they
evidently have found ways of handling the paradox and contingency of their
decisions. Studying how an organization comes into being, then, means con-
sidering the ways in which it handles the paradox and contingency of its
decisions. And the management of contingency is a central part of what
Luhmann terms “deparadoxization”.
In Luhmann’s theory, the question of conditions of possibility is central:
given their improbability, how do organizations nevertheless emerge and
come into being? Now, this general question can be formulated as a ques-
tion of how decisions are deparadoxized. Deparadoxization is the systems
theoretical expression for ways of managing the contingency inherent in
decisions, and for making sure that communication is not blocked, so that
the autopoietic reproduction of the organization may continue. Depara-
doxization can also be described as moving the paradox to a “less disturb-
ing place” (Luhmann 1993, p. 294). This will be demonstrated in our empir-
ical analysis.
In order to understand Luhmann – and Luhmannian analyses such as this
chapter – it is crucial to understand his basic analytical movement. Contrary
to most organization theories, Luhmann’s analytical move consists in recon-
structing – not primarily in revealing – the organization. The point can be
illustrated by contrasting Luhmann’s strategy of analysis with Mary Jo
Hatch’s “Spontaneous humour as an indicator of paradox and ambiguity”
(Hatch 1993). The title indicates that Hatch’s aim is to disclose something
hidden. But Luhmann takes what is hidden behind decisions and other
organizational activity – that is to say, paradoxes and ambiguities – as his
point of departure. Luhmann takes for granted that organizations are
110
08 Kap 5. Knudsen 05-11-11 16.20 Sida 111
paradoxical, contradictory, and contain elements of arbitrariness, and goes

on from there to ask how organizations avoid their own deconstruction.
This question is seen as a fundamentally empirical question. Here, it is
important to understand the manner in which the concept of the paradoxi-
cality of decisions functions as a means of generating observations, rather
than as a description of reality “out there”.
The central question of this chapter may therefore be formulated thus: in
the context of public sector modernization processes, how do organizations
handle the paradox and contingency of their decisions? This is a question of
current interest, since the modernization processes of the last 20–25 years
within the public sector in Denmark and other OECD countries have been
characterized by a tendency to form even more independent public organi-
zations making an increasing number of decisions. The basic question is thus
reformulated as one of how public organizations emerge, and specifically
how they increase the probability of decisions connecting to other decisions.
This question is approached by way of an analysis of the Frederiksborg
County Health Authority, the empirical case on which this chapter is based.
A self-referential organization emerges

The empirical focus of this chapter is the emergence of public organizations,
a dominant trend during the last 20–25 years of modernization of the public
sector. Words like autonomy, reorganization, and decentralization are fre-
quently used to characterize modernization processes in the public sector
(Baldersheim et al. 1995; Bentzon 1988; Brewer 2001; Condrey and
Maranto 2001; Petiteville 2000). Individual administrative units are made
responsible for their own modernization and efficiency (Corte-Real,
Nomden, Kelly and Petiteville 2000, p. 48). The new administrative units
themselves determine both the nature of their goals and how these goals are
best accomplished, while the individual units of administration are to act as
independent organizations. The aim is to make the public sector more effi-
cient and competitive, and this goal is – as it is claimed – best reached by
self-conscious and self-motivated public organizational entities. The term
“reorganization to reorganization” (Andersen and Born 2000) expresses the
pressure on these new organizations to continuously “modernize” them-
selves.
A main thread has been discernible in the indicated development: the dif-
ferentiation of the public sector into many autopoietic public organizations.
This is a complex development in that we witness not only a differentiation
of the public sector, meaning that organizations are differentiated into many,
smaller organizations, but also the fact that this process leads to the creation
of new organizations (sometimes by merging existing ones) making deci-
111
08 Kap 5. Knudsen 05-11-11 16.20 Sida 112
sions in new areas. These new organizations do not simply replace existing
organizations: they make yet more decisions about still more issues. The
result is a public sector with a significantly higher degree of complexity,
characterized by a plurality of self-referential organizations.
If we combine the systems theoretical assumption of the improbability of
decisions with the recent differentiation of the public sector into self-
referential organizations, we may bring into play the question of depara-
doxization: how do the emerging organizations deparadoxize their decision
making and what effects does this deparadoxization have? This question
guides our analysis of a single case: the Frederiksborg County Health
Authority.
Frederiksborg County is situated in Northern Zealand, Denmark. In
1999 it had approximately 365 000 inhabitants. There are four regular hos-
pitals in the county with 980 beds in total and a budget of approximately
300 000 000 Euro (1999). The Frederiksborg County Health Authority pro-
duced its first county hospital plan in 1980 under the name “SP80” (short
for “Health Plan 1980”). The plan features descriptions and decisions for
hospitals, with the county as its limit of validity. These and subsequent deci-
sions are identified as constituting the organization “Frederiksborg County
Health Authority”. The 1980 plan begins thus:
According to § 11, subsection 2 in law number 324 on hospital service of 19th

June 1974, responsibility for the working out of a plan for the hospital
service lies with the specific county, as does responsibility for necessary revi-
sion of the plan as dictated by development. (SP80, p. 9; my translation)
The similar health plan in 1997 (Sund97) begins in this way:
In connection with considerations of the 1996 budget of the 25th of

September 1995, the Frederiksborg County health committee decided to take
an initiative to start a revision of the health plan, so that such a revision could
take place through preliminary work by the health committee and the
involvement of all its collaborators. (Sund97, 1.1, my translation)
Seventeen years separate these two introductions – seventeen years and a

self-referential organization. SP80 legitimizes itself with reference to § 11,
subsection 2 in law no. 324, whereas Sund97 does the same with reference
to a decision taken by the county health committee. Comparing the first
pages of the two plans reveals the following differences:
112
08 Kap 5. Knudsen 05-11-11 16.20 Sida 113
SP80 Sund97
Time The timetable makes refe- The timetable makes
rence to the circular on time reference to a timetable
limits from the Ministry of proposal worked out by the
the Interior. Area Health Board in
Frederiksborg County.
Themes Refers to the circular from Refers to
the Ministry of the Interior a) the discussion of condi-
on schemes for registration tions and themes found in
and calculations. the plan itself (Sund97)
b) a number of work groups
that focus on a number of
specific themes determined
by the health committee.
Legislation Refers to laws and circulars The legislative frame is open
as given entities. to discussion.
Approving body The approval by the Ministry The acceptance of the
of Interior is mentioned. County Health Committee
(among others) is
mentioned.
Institutions Reference to Hovedstadsrå- Reference to a “broadly
referred to det (the Committee of the composed steering-group”,
Capital) and to the Ministry to the Area Health Board, to
of Interior. work groups and account-
ancy groups.
In the space of twenty years, we see an ever-increasing degree of reference to

the county’s own decisions and plans. Another notable change is a consid-
erable growth in possible areas for decision making in the county health
services. Frederiksborg County Health Authority makes increasingly more
decisions and thus emerges as an organization in its own right. These deci-
sions refer to other decisions made by Frederiksborg County Health
Authority itself, rather than to external premises such as laws or directives
coming from the central governmental authorities. Let me outline some of
the areas in which the organization expands. Here I refer to Luhmann’s
three types of decisions premises: programmes, personnel, and organiza-
tional structure (Luhmann 2000c, pp. 256 ff.).
From around 1990 onwards, the county increases its production of pro-
grammes. Programmes have the shape of plans (Health plans, Capacity
plans, plans for specialist practices, a general psychiatric plan, a plan for
physiotherapy etc.), but also consist of decisions concerning the placement
of specialist practices, the placing of the obstetrical wards, the number of
beds, and different kinds of policies (research policy, senior policy, staff pol-
icy, waiting-list policy, etc.).
In the 1990s, hospital staff also became an object of decision making.
113
08 Kap 5. Knudsen 05-11-11 16.20 Sida 114
Decisions were made concerning not only which persons were to be mem-
bers of the various committees, councils, clinics, and boards, but also the
staff’s more “personal” qualities. The county is concerned not only with its
staff’s formal qualifications, but also with “social qualifications”, and the
personalities of individual staff members are issues for decision making. The
County Health Authority also discusses desirable employee competences, as
well as which images of and views on the organization the staff should
preferably hold. Staff development and in-service training are discussed and
practiced.
From 1980 onwards, more and more extensive decisions about the orga-
nizational structure of the county are made. In 1980 structural planning was
primarily carried out by representatives of hospitals and hospital-related
professions. By the mid-1980s, the planning groups had assumed a more
permanent status, and from 1992 onwards, three “visible and competent
levels of management” are referred to: the county, the hospital, and the ward
(Sund92). In this way, the county as an independent entity gains in organi-
zational importance. From the 1992 (Sund92) Health Plan onwards, a com-
mon and united management of the county’s health service is mentioned.
Towards the end of the 1990s, the Health Plan of 1997 (Sund97) appears,
which fully accomplishes the move away from a principle of representation
towards the differentiation of a separate management: three directors, vice-
directors, and chiefs of staff are hired and work councils that fit the manage-
ment structure are established. Local hospital leadership is weakened and
administrative spheres moved to the County Health Authority. Speciality
medical committees are also established across the boundaries of the indi-
vidual hospitals. A common motto for the Health Plan 1997 is coined: “the
cooperative health services”. This new structure is the culmination of a
series of decisions, which have gradually built up the county’s ability to
make decisions about its own structure. At the same time, the new structure
enables a significant acceleration in the number of decisions made.
To sum up, over a period of less than twenty years, a County Health
Authority effectively creates itself as an organization in the shape of recur-
sively connected decisions and decisional premises. This may be described as
a double closure (cf. Chapter 1 in the present volume) of the organization:
the organization reproduces itself on the basis of its own decisions and in the
process also produces its own structures (in the shape of decision premises).
More and more issues are seen as relevant subjects to be determined by the
Health Authority, as issues that it can make decisions about. And to an
increasing extent these decisions refer to other decisions also taken by the
County. It is no longer sufficient to administer by pre-given premises – the
county now creates its own premises. Health Plans and similar County
decisions refer less and less to the outside, the other, and more and more to
self-made plans, goals, means, and decisions. The County no longer
114
08 Kap 5. Knudsen 05-11-11 16.20 Sida 115
administrates by given premises; the Health Authority itself decides the

premises. More and more areas that used to be on the “outside” are re-
located to the “inside” of the county’s boundaries, so that they may become
objects of decision making. General categories for county decisions include
capacity, aims, programmes, organizational structure, and staff, and the
county attempts to manage and control these by means of decision premis-
es. This means that one layer of decisions is built on top of the other; there
is a set of basic decisions, and on top of that a set of decision premises that
guide and limit decision making in the county. Let us have a closer look at
these decision-premises in order to relate this description to the problem of
deparadoxization.
Above, not only the emergence of an organization was described, but also
a way of managing the paradox of decision making. In Luhmannian theory,
the establishment of decision premises is one form of deparadoxization – a
way for the organization to manage its own contingency by imposing limi-
tations (Luhmann 2000c, p. 222). The limitations consist of decisions con-
cerning other decisions in the form of decision premises. These decision
premises impose limitations by demarcating the possible space for further
decisions. By delimiting the possible decisions, the probability for the con-
nection of decisions to other decisions is increased. The organization then
creates decision premises, limitations in the form of structures, programmes,
and plans for staff development. The decision premises function by increas-
ing the probability of acceptance of decisions by means of other decisions
(Luhmann 2000c, p. 145). One deparadoxization strategy is thus delimita-
tion. And since the decision premises are themselves decisions, the organi-
zation emerges from its own self-limitation.
If we relate the transformations in the County Health Authority to the
general tendencies in the modernization of the public sector in Denmark, we
may describe the emergence of the Frederiksborg County Health Authority
as a success. The establishment of a decentralized and autonomous public
organization has been accomplished in this specific case. And this process
was a leitmotif in the modernization programmes in the early 1980s and
after. This county may thus be seen as an example of the successful
implementation of central elements in the strategies of modernization. But,
as we shall see, this transformation is not without problems – problems
intimately connected with the deparadoxization of the decisions constituting
the organization. The next section will demonstrate some of the unintended
effects of the aforementioned deparadoxization strategy.
Deparadoxization and uncertainty (1)

In the time-dimension, decisions fixate contingency and absorb uncertainty
(Luhmann 1993d, pp. 291, 299). Before a decision is made, the alternative
115
08 Kap 5. Knudsen 05-11-11 16.20 Sida 116
is open and the decision is neither necessary nor impossible, but contingent.
After a decision has been made, contingency has been closed or fixated. The
emergence of the Health Authority as an organization means that uncer-
tainty is absorbed and contingency continuously fixated by means of deci-
sions. But as will be demonstrated in the following, our case shows that the
gradual establishment of new spheres of organizational decision making and
the increasing amount of decision premises also create increasing uncertain-
ty. The absorption of uncertainty takes the form of stabilization of expecta-
tions (Luhmann 2000c, p. 151), i.e. the establishment of decision premises,
and so we end up in a paradoxical situation: the creation of certainty itself
creates uncertainty. Let me present four empirically important ways in
which uncertainty is produced.
Increasing areas for decision making, increasing uncertainty

Decisions may open up contingency as they open up new areas for decision
making. As more spheres of organizational life are turned into areas for
decision making, as they become more plastic, less remains solid, static,
untouchable. What used to be given (as obvious goals, as rules, legislation,
routines etc.) becomes ever more contingent. This goes both for the Health
Authority as a whole, but also for hospitals and wards. Even individual staff
members must now see themselves as “strategic resources” whose compe-
tencies are to be developed, formed, and reshaped. A large number of deci-
sions now need to be made. Yet it is not clear what forms the premises of
the decisions – simply because these premises are also issues for decision
making. Contingency is limited by decision premises that establish structures
of expectations; but when new areas are opened up for decision making,
contingency is brought into play along with them.
Consider the following example: the County decided to establish a new
functional unit intended to cut across hospital borders. When its structure
was discussed by the County Health Authority, uncertainty was notoriously
present. I interviewed ward managers, representatives from the board of
directors, representatives from the unions, as well as the head clerk who
wrote the proposal. They all had different interpretations of the reasons why
the functional unit was set up, different views on how it was to be orga-
nized, different ideas of where the physicians were to work, which type of
relation to the various hospitals was to pertain, etc. This example shows not
only ways in which the structure – the explicit object for decision making –
was unclear, but also that the premises on which the decisions were to be
made were opaque. This uncertainty led to the circulation of a number of
rumours and conspiracy theories. These, in turn, only further increased the
general experience of uncertainty.
Furthermore, the simultaneous absorption and creation of uncertainty
calls attention to an important dynamic in the development of the organiza-
116
08 Kap 5. Knudsen 05-11-11 16.20 Sida 117
tion: the way decisions tend to call for each other. A decision made in 1995,
to ensure that the Health Authority hospitals were more cooperative and
used what was called “work sharing”, called for a new management struc-
ture and new types of cooperation. This again called for a decision to pro-
mote “staff development”. We see here how the organization – as it emerges
– needs to create the preconditions for decisions it has already made; the
organization emerges in a “stumbling” movement, as it were. The premise
for the decisions is incomplete and only determined after the decision has
been made. In this way, the organizational self becomes a future self. So,
confronted with its own contingency, the organization reacts with growth as
it tries to make more decisions, which may function as premises for subse-
quent decisions. To ensure the proper functioning of decisions already made,
further decisions must be made – and so on. The organization fails to stabi-
lize itself completely – it wavers.
Complexity
The decision premises delimit the space for possible decisions and increase
the probability that the decisions made actually connect to other decisions.
But as the number of premises grows, a new openness emerges in the form
of a surplus of decision premises. Selection becomes necessary and a choice
must be made between premises. The organization may react to this new
openness by making more decisions. A contradiction inherent in the estab-
lishment of decision premises is readily apparent: the decision premises –
intended to increase connectivity – evidently also increase the complexity of
the organization, making it uncertain which of the many limiting premises
the county is to connect to.
In an attempt to manage the growing complexity, a number of commit-
tees, councils and other minor units are established. The wards are upgrad-
ed. For instance, they get their own budgets and formal managers. This
development further increases the complexity of the organization – but now
on a structural level. The mere existence of committees etc. emphasizes the
need for further selections, choices, and decisions: an organizational carpet
interwoven with contingency. It becomes still more complex and unclear
who can make (or has made) particular decisions. The exact distribution of
responsibility for decision making between hospital managers, ward manag-
ers, the Health Authority board, the various teams and work groups, the
specialist committees, the councils etc., becomes unclear.
The simultaneous existence of opposite decision premises

The more complex the county’s decisions become, the greater the risk that
they do not all point in the same direction. Here, three examples of the
simultaneous existence of opposite premises will be presented:
117
08 Kap 5. Knudsen 05-11-11 16.20 Sida 118
1. Decentralization/centralization. On the one hand, we see a development

towards a centralized organization, inasmuch as an ever-increasing num-
ber of decisions are made that concern the county as a whole. However,
on the other hand, the wards are turned into self-managing entities.
These two principles do not necessarily exclude each other, but the simul-
taneous presence of two contrasting ideals leads to uncertainty, for,
which ideal are the employees to connect to?
2. Cooperation/competition. The formal motto in 1997 for the Frederiks-
borg County Health Authority was “the cooperative health-care system”.
Yet at the same time a particular hospital, Hørsholm, was made “inde-
pendent” and given a “separate board” with the explicit aim of becom-
ing “more competitive”. The subsystems were suddenly both to cooper-
ate and to compete. Further, these ideals have changed over time. In the
1980s, the subsystems were to cooperate, whereas from the beginning of
the 1990s, they were to compete. From the mid-1990s, they were again
to cooperate. The competitive ideal is at this point not clearly pointed out
in the plans from the period, but later plans mention it as a fact. Some of
the leading physicians interviewed remembered it as a verbally promoted
ideal, others believed that a plan of functional differentiation had been
agreed upon – but never really enforced.
3. Small/large organizational units. In relation to the establishment of
Hørsholm hospital as an independent unit there was a reference to “the
advantages that small organizational units have when staff members have
a sense of clarity, common aims, influence and a close management” (the
official newsletter: Nyt om sundhedsvæsenet no. 36-2001, my transla-
tion). At the same time, the official ideal advocated was that all the coun-
ty’s hospitals were to be run as one, uniform organization; that is, not as
small, individual organizational units.
Changing decisions
Rapid changes of decisions within a relatively short frame of time have also
contributed to the experience of uncertainty. For instance, Hørsholm
Hospital was merged with Helsingør Hospital in 1992. But it was already
made independent in 1998. In 2002, the county decided to sell Hørsholm
Hospital. This decision was cancelled in 2003, as it turned out that no one
wanted to buy the hospital. Another example (among many) is hip surgery:
before 1998, hip surgery was carried out at four different hospitals, but in
1998 it was assigned exclusively to Frederikssund Hospital. However, in
2000 it was assigned to Hørsholm Hospital. The department of obstetrics
was moved from Helsingør to Hørsholm in 1992, while in 1995 it was
moved from Hørsholm to Helsingør, and in 2003 it was moved to Hillerød.
In 1998, it was decided to merge three orthopaedic wards at three different
118
08 Kap 5. Knudsen 05-11-11 16.20 Sida 119
hospitals. This decision was followed by many decisions related to the place-
ment of functions, placement of staff, the organizational structure of the
new unit, new practices of visitation etc. In the spring of 2000, however, the
merger was cancelled. All of these rapid changes had significant impact on
the working life of hospital staff. Further, we may include the great number
of suggestions for changes, which were never realized, but which still rumble
on as possibilities. For instance, ideas were voiced in 1995 about turning
Frederikssund Hospital into a purely elective unit by closing the emergency
services section. The idea was formulated in a memorandum labelled “con-
fidential”. This episode hardly lessened the general feeling of uncertainty or
eliminated any conspiracy theories.
Summing up
In the section entitled “A self-referential organization emerges” we saw how
expectations in the shape of decision premises are one of the ways of limit-
ing contingency and thereby deparadoxizing decisions. In this part we have
seen another side of the story: the establishment of new spheres for organi-
zational decision making creates uncertainty, as does the increased com-
plexity, the simultaneous presence of opposite ideals, and changing deci-
sions. The limitations imposed increase contingency while managing it. The
organization emerges from the ongoing absorption of uncertainty, forming
a structure of decisions, which are recursively connected. But the decision
premises do not exclude contingency – on the contrary. The organization
encounters new problems while attempting to stabilize itself, and contin-
gency grows with the organization. Let us now change perspective for a
moment and see how the paradoxical nature of decisions is handled – in sup-
plementary displacement strategies.
Displacements
Let us recapitulate the problem of the paradoxical nature of decisions: the
contingency inherent in the decision means that it is possible to indicate
either side in the distinction. This openness makes it improbable that deci-
sions will connect to other decisions, for decisions cannot help but commu-
nicate their alternatives. This, then, raises the question of connectivity, for
why would we connect to a decision that could have been made differently
– and which also communicates this option? One way to increase the prob-
ability of connection is, as demonstrated above, the establishment of limita-
tions in the form of decision premises, which create and structure expecta-
tions. We may say that the paradox is displaced from the distinction of the
decision to the distinction between the decision and the decision premises.
This is what Luhmann calls “unfolding the paradox”.
119
08 Kap 5. Knudsen 05-11-11 16.20 Sida 120
The combination of the idea of deparadoxization as displacement with the

above observation – that the displacement of the paradox from the decision
to the decision-premises sometimes seems to allow uncertainty to grow
faster than it can fix it – shows us how this case study provides material for
a discussion of possible functional equivalents, i.e. other mechanisms of dis-
placement. Apart from displacement by decision premises, Luhmann men-
tions many other types of displacement or ways of “unfolding the paradox”;
e.g. attributing the decision to a decision-maker or to environmental de-
mands (cf. Becker and Haunschild, 2003). However, which types of dis-
placement can be found in an organization in the end is a question that can
only be answered with respect to a single empirical case.
The analysis presented here shows how the County Health Authority dis-
places contingency related to the factual dimension (what is decided and
why), the social dimension (who actually made the decision) and the tem-
poral dimension (when the decision was made). Now to some examples, fol-
lowed in the next section by a discussion of the effects and causes of such
displacements.
Deparadoxization by reasons
The health plans and the decision proposals in the 1990s share a common
trait: they get started by offering reasons for their own existence. This is
done partly by reference to goals to be accomplished, partly by listing exter-
nal challenges and demands that must be met (demographical changes,
medico-technological developments, growing expectations, political de-
mands for retrenchment etc.). These reasons displace contingency from the
decision itself to its reasons. Reasons, then, are simply ways to displace con-
tingency: the decision could be different, but when a reason is given, the con-
tingency is displaced to this reason. In Frederiksborg County, the reasons are
primarily presented within two topoi.
Decisions are legitimized by external demands, but they are also legit-
imized by aims, purposes, visions, and values. An example can show how
this works. The example is “Notat om fremtidig ledelses- og samarbejds-
struktur, udkast den 11/2 1997, Sundhedsforvaltningen” (“Memorandum
on the Future Management and Cooperation Structure”, sent out as a pro-
posal by the Health Authorities 2/11/1997). The memorandum proposes the
establishment of an independent management for the entire health service in
order to minimize hospital managements, increase the competencies and
responsibilities of the wards, and transfer the main administrative functions
from the hospitals to the County Health Authorities. The memorandum lists
more than thirty purposes and positive effects of the suggested restructuring;
however, some of the purposes are only very loosely coupled to the suggest-
ed structure. The overwhelming number of purposes indicates that contin-
gency is really a problem, because it shows that connectivity to the decision-
120
08 Kap 5. Knudsen 05-11-11 16.20 Sida 121
proposal cannot be taken for granted. Therefore, an abundance of reasons

is ascribed to it – ascriptions which are expected to increase the probability
of connectivity.
What happens, then, is that the contingency of the decision is displaced to
the purposes and the contingency of the purposes is displaced among all the
purposes, using a large number of chains of means and ends, which the
reader cannot survey. Yet a fiction of rationality is maintained by listing the
purposes in such a way as to indicate a hierarchy among them. Some of the
purposes and the proposals are coupled in a very loose way. The connections
between the proposals and the purposes are often so tenuous that the text
seems to be a juxtaposition of otherwise unrelated items gambling on the
fact that readers will add the necessary reasoning themselves.
Deparadoxization by hearings
A further characteristic of decisions is that they are first virtual (Baecker
1999c, pp. 132 f.), or proposals, as it were. Such a decision-proposal is a
kind of test-decision: will this decision gain connectivity? Hearings may be
seen as one way of testing such virtual decisions. Decision-proposals are typ-
ically sent to various groups in the organization for hearings. After the hear-
ings, the proposal is revised and sent to further hearings, until it is finally
confirmed. The hearings related to the general four-year plans, for instance,
prove to be very extensive undertakings. Groups involved in the hearings
are, among others, work councils, unions, wards, and municipalities. The
report of hearing-response résumés concerning the 1997 general plan con-
tains several hundred pages from eighty-eight different respondents.
Luhmann describes how the mystery of the decision is dissolved by per-
sonalization or consensus. The decision is ascribed to a person. So a tauto-
logy replaces the paradox of decision making: the decision-maker makes the
decisions (Luhmann 2000c, pp. 136 f.; Chapter 1 in the present volume).
This case study – especially in relation to hearings – shows supplementary
means for displacement. Contingency is displaced from any identifiable
actor or institution to “less disturbing places”. This, however, contrary to
what Luhmann appears to presume, does not mean that there is “a situation
of counterbalance” (Luhmann 2000c, p. 139). The decision – and its con-
tingency – is spread all over the organization, something that, naturally,
obscures its source, but does so without obtaining any consensus.
Deparadoxization by changing decision-proposals

A well-known phenomenon in the processes of writing, negotiating, and
rewriting proposals, is that the subject of the decision becomes increasingly
vague. In 1999, a head clerk and member of a small committee, who was to
make a draft structure for a new type of organizational unit, said:
121
08 Kap 5. Knudsen 05-11-11 16.20 Sida 122
The closer we get to the finishing line the less clear the papers become – in
order for people to connect to them. It is a well-known situation; it is often
like that. It doesn’t make it any easier for the rest of us when we have to fol-
low up on the decisions with the unions – for what exactly have they agreed
to? (Interview 10/11/1999, my translation)
The decision emerges, and its contingency is displaced by obscuring what

the decision is actually about. A decision is made, but what it means exact-
ly is left for the future to decide. In this way, contingency is displaced to the
future.
To sum up: In the present case the following types of displacement are
common: the why of the decision is displaced by the presentation of reasons,
the when and the who by hearings, and the what by emptying the decision
(obscuring what it is about). These types of displacements correspond to the
three dimensions mentioned above (the factual, the social, and the tempo-
ral), and – as we have seen – they are routinely built into the typical decision
processes. Now some of the effects of these kinds of displacements will be
described.
Deparadoxization and uncertainty (2)

“Displacement” refers to the various ways in which the contingency of deci-
sions is displaced to what Luhmann calls a less disturbing place. The section
above describes such displacement mechanisms: the ascription of causes,
which displaces contingency to a claimed system-exterior requirement or to
different types of reasons etc.; hearings, which displace the author(s) of a
decision to all participants in the hearing activities; time displacements,
which postpone decisions to the future or obscure the point in time when a
decision was made. The end result is a situation in which no one can deter-
mine who made a given decision, when it was made, or why it was made
– whether it was a means or an end – or, finally, what it was really about.
The different kinds of displacements displace indeed the paradox in so far
as they place the paradox out of sight. But in the process, the displacement
strategies create uncertainty as to what the decision is actually about, when
it was made, who made it and why it was made. The decisions just seem to
come into being without anybody knowing why or when or how. The most
widespread reaction to this phenomenon among the staff is the ascription of
hidden motives and secret plans. Someone must have a plan, and if we can-
not see it, it must be secret. The effect is, simply, a flourishing of conspiracy
theories. Secret motives are ascribed to top-managers, and the top-managers
themselves ascribe secret plans to other groups of employees. The effects of
displacements are felt at all levels of the organization.
One of the problems is that the hearings create expectations concerning
122
08 Kap 5. Knudsen 05-11-11 16.20 Sida 123
influence, which are, however, almost always disappointed, for it is simply

not possible to take all the comments into consideration. This, however,
gives the impression of pseudo-hearings, in which an unspecified group
already knows what it wants anyway, even though “they” do not say it. The
impression spreads that the board of directors wants something, but that
they nevertheless will not say straightforwardly what it is. Suggestions and
proposals are met and interpreted with the expectation that “someone is up
to something”, something unsaid, unclear, secret. All the reasons listed come
to act as blinds or smokescreens for the real reasons – seen in the light of the
current conspiracy theory. The mixture of openness and opaqueness (for
what is really the purpose?) leads to distrust. The overabundance of pur-
poses creates the suspicion that the real purpose is deliberately hidden
behind a smokescreen. Motives understood as intentions are always invisi-
ble and that is why they can be discussed forever, for they are not commu-
nicatively accessible.
Hearings, the ascription of causes, and the other strategies displace the
paradox, but the effect of the displacement is that the contingency is spread
rather than relocated to a “less disturbing place”. It seems that what we
have is a case of unstable deparadoxization, since the paradox is not “prop-
erly” displaced: the paradox is not deparadoxized in a stable way. The con-
tingency of the single decision is displaced to the whole organization, which
is thus permeated by uncertainty and suspicions of hidden motives. Thus,
mistrust is brought into play as a method for reducing complexity (Luh-
mann 1979) – no wonder, then, that in the case of Frederiksborg County the
board of directors introduced trust as a value in 1998 …
Uncertainty is a sine qua non for the organization as such. Luhmann puts
it this way: “Autopoiesis is therefore only possible if the system is in a con-
stant state of uncertainty about itself in relation to its environment and if the
system can produce and control this uncertainty by means of self-organiza-
tion” (Luhmann 2000c, p. 47, my translation). Nevertheless, the above
analysis points to three problems related to the uncertainty inherent in the
organization, problems that are specific to our case of a young organization
in a continuous process of reform. First, what was intended as a de-bureau-
cratization of the public sector appears to lead to re-bureaucratization; a re-
bureaucratization not in the form of formal rules, but in the form of activi-
ties related to decision making. As the organization grows it spends more
and more time on decision-related activities – compared to the amount of
time spent on actual health-care work. Second, the staff is fatigued by the
constant reforms, and a combination of resignation, frustration and con-
spiracy-theories proliferates as a reaction to increasing complexity, uncer-
tainty and constant changes, or, in other words, to decisions that do not
have connectivity. Third, it appears that those deparadoxization-strategies
that were intended to enable decision making in fact risk blocking the
123
08 Kap 5. Knudsen 05-11-11 16.20 Sida 124
decision-making process. The result is rapidly changing decisions (when a

decision is overruled, it has lost its connectivity) and decisions that are
dropped even before they are formally confirmed. Making decisions
becomes more and more difficult even as the organization increases the
amount of areas for decision making. Deparadoxization has paradoxical
effects: the various deparadoxization strategies, intended to aid the organi-
zation in its emergence, simultaneously hinder its emergence.
Why displacements?
In the present case, the displacement strategies only seemed to make matters
worse for the organization. Why, then, are they used? A more workable
strategy intuitively seems to be to “hierarchize”, i.e. to displace the paradox
“upwards”, to appeal to hierarchical structures, instead of attempting to
involve all parts of the organization in the decision-making process. The rea-
son why this strategy is not used is probably that at least some of the deci-
sions, which need deparadoxization, are decisions about the (hierarchical)
structure of the organization. This makes it difficult to use those same struc-
tures as deparadoxization strategies.
This point may be generalized. The reason why this particular organiza-
tion has a need for supplements to its decision premises is that the decision
premises are themselves issues for decision making. Since the decision prem-
ises are themselves decisions, they too involve a paradox – which needs
deparadoxization. Thus, the organization is forced to use supplementary
strategies of deparadoxization, since it cannot adequately handle its contin-
gency by recurring only to the network of its decisions.
Another line of argument is connected to the differentiation of the orga-
nization. We saw above how the system differentiates itself internally in
reaction to its increasing complexity. In the present process of reform, inter-
nal differentiation of the county effectively transforms conscientious offi-
cials and hierarchical organizational units into self-managing, flexible,
ready-to-change, enterprising subsystems. Significantly, such transforma-
tions do not happen as the result of explicit orders, for only when the sub-
systems rediscover themselves as enterprising systems full of independent
initiative do they become exactly that. When the aim is to create an organi-
zation differentiated into self-referential, self-modernizing subsystems, then
top-down management must limit itself in order not to overrule, and thus
paralyse, the self-reference of the subsystems. This is a fundamental contra-
diction in contemporary management. On the one hand, management gains
importance as a symbol of the organization’s unity and “self”. On the other
hand, management cannot fill this place authoritatively and issue hierarchy-
based orders, for that would block the subsystems’ self-reference. This is
also why management is inclined to use deparadoxization strategies such as
hearings and reasons – instead of hierarchical orders.
124
08 Kap 5. Knudsen 05-11-11 16.20 Sida 125
Reforms and self-reference

Luhmann’s notion of paradoxicality provides a framework of sufficient gen-
erality to analyze all types of organizations. According to this notion, all
organizations must deparadoxize the paradox of decision making in order
not to block the reproduction of the organization. Even though the present
case study attested to this general mechanism in organizations, it also
seemed to suggest that the insecurity described here is somehow specific.
What is special about the case? The answer to that question is that we are
witnessing a process of reform. The organization is fairly new and moreover
the analysis focused explicitly on its process of emergence. This is probably
part of the reason for the instability in its decision premises, or what we may
call “self-reference with an insufficient self”. This lack in the organizational
self only seems to consolidate the contingency inherent in the decisional
paradox. The emerging organization must, as it emerges, simultaneously
create its own self, its structures – while it must also presuppose itself as a
self. As the organization refers to itself, it refers to a self that does not exist
prior to this reference. The organization must presuppose itself in order to
create itself. This is true for all organizations, but in a situation of reform
where this self is not developed yet, the system’s self-reference is, of course,
very fragile. What is special about the present case is thus that contingency
is doubled: the organization faces the general contingency of decisions as
such, but also the contingency of the emerging organization which trans-
forms itself by a process of self-reference without having a very developed
self to refer to. The problem of self-referentiality combined with an insuffi-
cient self is repeated when sublevels of the system become self-referential.
The internal differentiation of the organization may be described as a mul-
tiplication of its self. This is also, however, an increase of the uncertainty of
the insufficient self, or better – the multiple insufficient selves. “All that is
solid melts into air …”.
Conclusion
Using the basic framework of Niklas Luhmann’s organization theory, this
chapter has analysed the emergence of a self-referential public organization
and discussed strategies of deparadoxization in relation to decision-making
processes. The Frederiksborg County Health Authority formed the empiri-
cal basis for the study, and different ways of deparadoxization were exam-
ined. The case-study pointed out a central mechanism in the emergence of
public organizations, namely that the very same strategies, by means of
which paradox and contingency are handled and uncertainty absorbed,
simultaneously co-produce even more contingency and uncertainty – a situ-
ation which the organization can only respond to with more decisions.
125
08 Kap 5. Knudsen 05-11-11 16.20 Sida 126
Key concepts in Luhmann’s organization theory functioned as guiding

lines for the analysis of this specific case. Based on relatively few concepts,
the theory proved effective as a means of generating observations. The
improbability or paradoxicality of decisions, which formed our point of
departure, allowed us to reconstruct concrete empirical phenomena related
to organizational reproduction. However, other deparadoxization strategies
than those described by Luhmann were found in the empirical material. This
does not show the inadequacy of Luhmann’s theory, since Luhmann never
intended to present a complete list of methods for deparadoxization.
Instead, it shows how his systems theoretical concepts and analytical strate-
gies may be used to generate empirical observations. And for that they have
proved fruitful. The present analysis has not resulted in abstract statements
of the type “the decisions made in Frederiksborg County are paradoxical”,
but instead in an analysis that has shed light upon well-known concrete phe-
nomena: the increasing amount of decisions, the proliferating conspiracy-
theories, the increasing uncertainty, the rapidly changing decisions etc.
These are phenomena often explained by their attribution to persons, not to
systemic mechanisms. In this way, Luhmann’s concepts generated a new per-
spective on decision-making processes in a public organization undergoing
a process of reform.
126
09 Kap 6. Czarniawska 05-11-11 16.20 Sida 127
Chapter 6
On Gorgon Sisters:
Organizational Action
in the Face of Paradox
Barbara Czarniawska
Homer adorned the shield of Agamemnon with “the blank-eyed face of the
Gorgo with her stare of horror” (Iliad 11.36). The later Greek poet, Hesiod
(8th–7th BC), increased the number of Gorgons to three – Stheno (the
Mighty, or the Strong One), Euryale (the Far Springer, or the Wide-
Stepping), and Medusa (the Queen, or the Guardian), and made them the
daughters of the sea god Phorcys and of his sister-wife Ceto.1 The Gorgons,
as Niklas Luhmann (1991a, p. 58) has pointed out, have the same effect as
paradoxes: staring at them results in immobility, and therefore in an inabil-
ity to act. Thus, one should avoid observation if action is intended. The rea-
son Luhmann brought the existence of the three, not one, Gorgons to the
attention of social scientists is that there is more than one problem with
paradoxes. Medusa was mortal, and could be killed by Perseus; but Stheno
and Euryale are immortal, and cannot be eliminated.
This chapter appropriates Luhmann’s idea and applies it to studies of
organizational practice. It is claimed that although the traditional mode of
operation in the Swedish public sector – the erstwhile model of welfare
organization – was characterized by an astute awareness of a need for con-
stant deparadoxization, the present mode is more and more influenced by
the private-sector mythologization of manager-heroes who resolve all para-
doxes once and for all – until the next Gorgon’s head is raised.
Protecting paradoxes
The usual criticism of paradoxes and the urge to “solve them” is connected
with the fact that they violate logic. This is correct: two opposing statements
within the same utterance are not logical, but paralogical. After all, logic is
1www.britannica.com/eb/article?eu=38185&tocid=0&query=gorgons;
www.theoi.com/Pontos/Gorgones.html; [Accessed April 13, 2002].
127
but a linear, one-dimensional set of rules, which came into being through the
praxis of the Indo-European languages. The fact that we adhere to and
roughly agree upon such a set of rules makes it easier to communicate with
one another, or so we believe. Logic is a conventional way of describing
one’s image of the world, which can then be discussed and negotiated with
other people. Logic is a device. Consequently, as Luhmann (1986b) points
out, paradoxes are not attributes of social systems but the result of using the
logical analysis as an observation tool.
A static two-dimensional picture like a map or a photo has serious defects
compared with the dynamic three-dimensional reality that it propounds to
represent. Yet we accept it without protest (its stands to reason that we did
not spend our three-week holidays sitting and smiling in front of a hotel),
because we realize the conventional character of such “representation”. We
start to rebel only when some smart constructivist tells us that it is not a re-
presentation, but itself a part of reality. Exposure to linguistic paradoxes
rouses a protest much more quickly. Words are, after all, devices with which
we shape our reality.2 In everyday life we take for granted that language
reflects reality, or at least that it represents reality as it is.
A pragmatist perspective reveals, however, that there are no grounds for
assuming correspondence between language and “reality”, be it iconic, sym-
bolic, or whatever. As the new pragmatist Richard Rorty points out, “no
linguistic items represent any nonlinguistic items” (1991a, p. 2). Paradoxes
are irritating, not because they go against the “essence of things”, but
because they ruin the conventional order of the surface that is our life.
Essentialists of various denominations are opposed to paradoxes because
reality (“as such” – a favorite essentialist expression) cannot possibly be
paradoxical. However, there are – and always have been – voices which
insist that a paradoxical image of reality is “truer” or simply more practical.
The paradox of the existence of humankind would then lie in the fact that a
social system can reflect upon itself only within its own frame of reference.
The existence of God was the traditional solution to this paradox; otherwise
people could not judge themselves objectively, because they are and remain
themselves even in the act of judgment. Science offered to replace God, but
being human-made, it cannot claim and maintain the same legitimacy.
If metaphysical solutions are unacceptable, then paradox itself must be
accepted. As Hofstadter (1980) pointed out in Gödel, Escher, Bach, mathe-
maticians, painters, and musicians were always aware of this paradox.
Instead of being threatened by it, they toyed with many variations on the
theme immortalized by Escher: a person who sees himself reflected in a mir-
ror that he holds, and so on, and so on … no resolutions, only further
2It is being claimed increasingly often, however, that images are beginning to prevail over
words (see, e.g., Godzich 1991) – that one language is encroaching upon the other.
128
reflections. The catastrophists in social science were traditionally worried

about this aporia: will these reflections never end? Certainly they will, says
Ashmore, in his brilliant self-reflective thesis (1989): the commonsense,
aesthetic criteria and the diminished “meaning added” set the limits.
Thus paradoxes can be seen as a property of life itself (an existential per-
spective), a perspective taken by an observer of autopoietic systems (a cyber-
netic perspective which can take both a metaphysical and a metaphorical
shape), or as a phenomenon occurring within language (a postmodern per-
spective).
While Niklas Luhmann represents the second take, Jean-Françoise
Lyotard stands for the third perspective. These two perspectives sometimes
coincide, and sometimes part ways. In general, Lyotard is more enthusiastic
about paradoxes than Luhmann is.
Logic is a language that is universal but not consistent, as it allows for-
mation of paradoxes, says Lyotard (1986). Indeed, says Luhmann (1995f, p.
95), the logic is possible only because the world of meaning encompasses all
the contradictions: “Otherwise, the minute one first encountered a contra-
diction, one would fall into a meaning gap and disappear.” But paradoxes
are the houseflies of logic: although they undoubtedly have a place in this
world, the prescription for dealing with them is clear. Paradoxes must be
eliminated. People refusing to do so must possess one of two special licences:
either they are mad or they are artists.
The urge to dissolve paradoxes does not come only from the aesthetic
unpleasantness of encountering a logical error. No matter how reflective
their attitude toward paradoxes, actors necessarily engage in the more or
less creative process of “deparadoxization”3 (Luhmann 1991a). Complain-
ing that “the paradoxization of civilization did nothing to civilize the para-
doxes” (which supposedly thrive in the wilderness; 1991a, p. 60), in his
essay on the three Gorgon sisters Luhmann criticized the postmodernists,
declaring that their call for a celebration of the paradox is like encouraging
us to look straight in the face of one of the Gorgons – which, as we know,
would turn us into stone. Reflectivity paralyzes.
Paradoxes paralyse an observer, continues Luhmann, but not an actor (in
his language, an autopoietic system). Buridan’s ass will survive, because it
will react, even if by stating that it cannot decide. “Contradiction puts a stop
to observation (…) precisely this can be sufficient grounds for doing some-
thing” (1995f, p. 360).
Lyotard focuses on the evolutionary possibilities inherent in paralogy
(1986). There are two kinds of “progress” in knowledge, he says. One
3I am not using Social Systems’ “deparadoxicalization” here, because it is impossible to pro-

nounce. I am also abandoning the term recommended to me by my English editor – “depara-
doxification” – in favor of “deparadoxization”, used in Luhmann, 1986b.
129
amounts to innovation – a new move, or a new argument within the estab-

lished rule. This move could be interpreted as a dialectic move, as the third
step, the synthesis, comes from the same repertoire as both the thesis and the
antithesis. The other is an invention, a move that introduces different rules,
thus turning the paradoxical into the accepted. “What used to pass as para-
dox, and even paralogism (…) can (…) acquire a new force of conviction
and win the acceptance of the community of experts” (ibid. pp. 43–44). This
kind of evolution is close to Hegel’s idea of dialectics, observes Luhmann
(1995f, p. 373). This might be correct, yet Lyotard does not see it as a play
of historical forces, but as the semantic work of human beings. He suggests
that the legitimization of knowledge should, or may be, based on paralogy.
Opposing Habermas, Lyotard says that it is not consensus but dissension
that moves knowledge forward. Here, Luhmann joins him again, indicating
the necessary instability of the systems.
Language, through the appearance of paradoxes, is forced to change its
rules; but because it permitted those paradoxes to appear in the first place,
the challenge and the change occur within the same system. Many attempts
have clearly demonstrated that language cannot be changed from the out-
side, either by order or by force. History has demonstrated equally well that
changes in language cannot be prevented either, no matter how many min-
istries of language the purists establish.
But a change within a language does not take place without battle. On the
contrary, it seems that language change must consist of a battle against rules
and principles that are taken for granted. As Thurman Arnold, a Harvard
lawyer, a New Dealer and a keen detector of paradoxes said, “[i]t seems to
be the eternal paradox of the human mind that principles and faiths which
are so essential to its comfort and to the orderly organization and transmis-
sion of ideas should at the same time always stand as the greatest obstacle
to discovery” (1935, p. 24).
The image of human beings as rational decision makers controlling the
environment requires a linear vision of the world, in which conflict and
ambiguity are temporary aberrations to be removed by the next rational
action. Much can be said about the practical advantages of such a vision of
the world; as usual, however, focusing on certain aspects forces us to gloss
over others. In this case, the inherent paradoxicality of human and social life
is the victim.
Organizing as deparadoxization
In organization studies, as Van de Ven and Scott Poole (1988) pointed out,
a quest for coherent and consistent theory led to the neglect of organiza-
tional paradoxes. The observed paradoxes in field studies, for example,
130
were taken to be cases of “anomalous communication” (Manning 1992).

And yet this anomalous communication lies at the heart of modern institu-
tions as we know them. “Having determined what action to take by logic of
appropriateness, in our culture we justify the action (appropriately) by a
logic of consequentiality”, observed March and Olsen (1989, p. 162) with
unerring acuteness. When facing action, we judge its kind and context from
the perspective of our own identity, and accordingly choose an appropriate
action from the accessible repertoire. When questioned, we justify the action
in terms of rationality (logic of consequentiality): motivated by personal
gains (premises that everyone can accept), we choose the best way of achiev-
ing them (steps that can be followed by anyone) and arrive at the best pos-
sible result (conclusions and evaluations that must be universally accepted
[Hanna Pitkin, 1972]). In other words, the two rhetorics (or two kinds of
logic, as March and Olsen call them) behind a reflective account and a jus-
tificatory account are usually at odds.
Usually, but not always: there are contexts in which the two accounts can
be safely separated. Also, the organizational narrative smoothly accommo-
dates both of them through the device of direct and indirect speech. In a
study by Jansson (1992), a decision about the kind of energy to be provid-
ed for a town was reached by using the logic of appropriateness, and was
both justified and questioned within the logic of consequentiality. The deci-
sion makers claimed that their decision was based on calculation of com-
parative costs, whereas a “serious counter-calculation” was demanded from
the critics of the decision.
Such tension, say March and Olsen, is typical of political institutions;
therefore, I might add, it is not surprising that we find it in public adminis-
tration organizations as well. “It can be seen as resulting in a kind of healthy
charade of hypocrisy in which reasons and actions are not tightly linked but
place pressure on each other in a way that strengthens each” (1989, p. 162)
– what Brunsson (1989) dubbed “a necessary organizational hypocrisy”.
In a study in the same tradition concentrating on Swedish public sector
organizations in the late 1980s (Czarniawska 1997), I noticed that the exis-
tence of paradoxes in everyday organizing seemed not only to paralyze
action, but also to stimulate it. My findings were corroborated by
Luhmann’s suggestion that “… because the paradox cripples observation, it
can be understood as an inducement, even as a compulsion to solution”
(1998, p. 112).
Thus it can be claimed that the existence of paradoxes incites action4
– at the minimum that of deparadoxization. There exist at least three
4 I am falling back on a more traditional dichotomy “action-reflection” because following
Luhmann’s differentiation between action and communication (1995f, Ch. 4) would unneces-
sarily complicate my reasoning.
131
deparadoxization strategies, well known and practiced in organizations:

temporization, spatialization, and relativization.
Temporization, says Gumbrecht, Luhmann’s follower and a comparative
literature theoretician, amounts to narrativization (1991a). The conflicting
elements are detached in time and the conflict is resolved in the future.
Thurman Arnold claimed that “[t]his technique is as old as the parables of
the New Testament. It is only its dialectical formulation that is modern”
(1935, p. 30). He noted also that such a temporized series of observations is
at odds with conventional science, understood as a logically formulated set
of principles that are valid at all times. And he added, aptly for my purpos-
es, that “[i]f this point of view must have a name we prefer to call it an
anthropological approach towards social ideas” (ibid., italics added).
Arnold also stressed the need to dramatize the reconciling narrative. In a
letter (quoted by Samuels, 1979), he presented his dramatist theory of socie-
ty:
A workable philosophy [of action] it seems to me, is necessarily a maze of

contradictions so hung together that the contradictions either are not appar-
ent or else are reconciled by a mystical ritual … the unity and the rhythm of
an institution requires people dancing in different directions and alternately
coming together and apart. (Arnold 1937, as quoted in Samuels 1979, p.
1001)
Conflicting issues can be decoupled not merely over time but also in space;
thus another strategy – spatialization (March 1988; Manning 1992): “This
committee deals with the technical problems, and that committee deals with
the psychosocial aspects of the proposed change.” The antitheses are simul-
taneously present, but not in the same place.5
Sometimes, however, neither temporization nor spatialization work. The
promise of a synthesis in the future is not convincing; the committees meet
in the corridor by mistake, and no longer stick to the issues in their domains.
Decentralization is perceived6 by the people subjected to it as centrally
ordered – a paradox that is a source of frustration and a cause of apathy for
them. The deparadoxization strategy used in such a context consists of
explaining different perceptions by the different levels of observation, where
the first-level observer is assumed to be blind to his or her own position and
role in the system (“… if you were in their place, you would see it different-
ly…”). It should be added that even this strategy fits the possibilities of the
narrative perfectly well. It is a matter of actorial shifting operations (Latour
5 Brunsson (1989) distinguished four strategies, all variations of temporization and spatializa-
tion: decoupling over time, between issues, between types of relevant environment, and
between suborganizational units.
6 “A paradox is of course always a problem of an observer”, Luhmann 1991a, p. 62.
132
1988), whereby the reader or listener can see the world through the eyes of
one or another first-level observer, but by virtue of being a second-level
observer, can also understand the limitations of that “native” point of view.
This means, in Umberto Eco’s (1981) vocabulary, combining the stances of
a semantic and a semiotic reader.
In a political context, a mediator can be called in and can claim the posi-
tion of second-level observer (a consultant, watching from outside, might be
able to see the points of view of the management and the unions at the same
time, and also understand why they are as they are). Thus, as Russian liter-
ary theoretician Mikhail Bakhtin pointed out, social science should aim not
for empathy but for extopy, an “outsidedness” (Roberts 1994, p. 250).7 In
a hierarchical context the higher level claims the position of a second-level
observer: “Down at the local office you cannot see the whole picture, and
so you fail to observe that what the management is doing is actually decen-
tralization!”
These variations of deparadoxization by relativization are also well
known in organizational life, but even they do not always work. Man-
agement and workers are locked into a stalemate, in which two points of
view cannot be reconciled, as neither party wants to acknowledge the rela-
tivity of their position.
When deparadoxization does not work,

or the tragedy of action
Gumbrecht (1991) points to the tragic aspect of paradoxicality, defining the
tragedy of action as a situation in which an actor, in the role of first-level
observer, perceives another actor as ambiguous (“What is management real-
ly trying to achieve?”). Such perception breaks the unity and causes a rup-
ture – sometimes just an “interruption” (Silverman and Torode 1980),
sometimes quite literally the breakdown of a person. In other words, one of
the first-order observers (an actor) suspects other actors of perceiving the
situation as non-paradoxical but communicating misleadingly.
Again, there is nothing peculiar about this situation – “… it is impossible
to imagine a reality constructed without paradoxes”, as Gumbrecht (1991,
p. 487) says, or as completely deparadoxized. And although it is easy to
agree with Luhmann that the contemplation of paradoxes excludes action,
it is impossible not to notice with Gumbrecht that deparadoxization strate-
gies frequently fail to work, and that action – individual and organizational
– reveals its tragic side. This tragic side is especially clear in cultural contexts
7 Bakhtinian exotopia, or outsidedness, is amazingly close to the notion of “observers of the
second order”.
133
of organizing that are far from the complacent pragmatism of the Swedish
public sector. My studies of public administration in Poland and Italy
(Czarniawska 2000; 2003) revealed situations in which paradoxes were per-
ceived as absurd, the efforts at deparadoxization did not work, and the
actors-cum-first-level-observers treated other actors not merely as ambigu-
ous, but as downright untrustworthy (“Whatever it is that they really want
to achieve, it will be at my expense”). One could almost say that public
administrators in those countries agreed with Luhmann’s opinion on the
incommensurability of systems;8 but it is the illusion of a possibility of com-
munication that prompts communication.
One reaction to such a situation can be an oscillation between the levels
of observation. After all, one way to see a blind spot metaphorically is
through self-reflection. True, this is like meeting a Gorgon by request, but
perhaps people do not turn to stone at the first encounter, but become more
“resistant” (which, after all, means “thing-like”). The mirror is directed not
at the Gorgon but at oneself; it does not kill the Gorgon, but neither does it
kill the watcher, who is not a Gorgon and is temporarily shielded from the
monster by the mirror. Again, having the examples of Polish and Italian pub-
lic administration in mind, one should add, with Luhmann, that “[t]he free-
dom, gained by self-reflection, can be used only if its constraints are suffi-
ciently close at hand. Otherwise the autopoietic system simply will not know
what to do next” (1986b, p. 187). Self-reflection, or “saving differences”
(Luhmann 1991a, p. 71) seems to be a way of dealing with the Gorgon sis-
ter Euryale, whom Luhmann associates with the rhetorical tradition. Thus
the way out of the vicious circle of public administration reforms in Poland
and Italy would lead through a reflection on actual practices constrained by
the requirements of the desired model (whereas at present the model is intro-
duced in blatant disregard of the actual practices).
Another, in fact opposite, reaction is to make a paradox invisible through
action. When deparadoxization strategies do not work, plunging blindly
into action might. Such a plunge requires the creation of a blind spot, a jump
into one part of the paradox, into a distinction, therefore losing sight of the
site on which the distinction must be made. This is, perhaps, the best way of
getting rid of the third Gorgon, Stheno, whom Luhmann associates with the
theological tradition, dealing with such distinctions as the difference
between Good and Evil. Acting might produce “a difference that makes dif-
ference” (Luhmann 1991a, p. 69). It is a reverse of relativization, a move
from the second-order observation to a first-order observation (and there-
fore accessible only to the second-order observers, the “observers of the sys-
tems”).
8 As systems are able to communicate only self-referentially, they cannot “understand” other
systems.
134
Many of the Swedish officials I talked to and observed in my study seemed

to have mastered the art of oscillation between action and reflection, and
between observation on the first level (an actor acting) and observation on
the second level (observers observing the actors acting, including them-
selves). Here is a striking example: a Municipal Finance Manager prepared
a thorough basis for a decision concerning a housing development in an area
that, as he knew through his role as a member of the local Emergency
Committee, was a part of a secret military plan.9 Baffled by this situation, I
asked whether the work did not feel like a waste of time? And wasn’t the
whole situation absurd?
“Not at all”, he countered. His role as public servant required a loyal fol-
lowing of political decisions, including those of which he did not personal-
ly approve, and he was often called to fulfill this requirement. His role as a
member of the Emergency Committee required the protection of confiden-
tial information. He regarded himself in a mirror and saw himself split. It
was then that he chose to create a blind spot by taking a point of view of
only one first-order observer: the public servant’s. The work did not, in fact,
feel absurd, he further explained, because in the time needed for the proper
implementation of the decision, the military might change their plans
– perhaps even as a result of a municipal decision. Not only might the
Gorgons go away while the mirror is directed at oneself, but a blind action
might contribute to the removal of the paradox. After all, autopoietic sys-
tems resolve their paradoxes themselves – it is a problem only for the observ-
er. “The autopoiesis does not stop when confronted by logical contradic-
tions: it jumps, provided that possibilities of further communication are
close enough at hand” (Luhmann 1986b, p. 180). Herein lies the difference
between this Swedish example and the Polish and Italian practice: in un-
stable political organizations, it is not certain when and with whom the next
opportunity to communicate will arise. Jumps might prove suicidal.
The example of the Municipal Finance Manager is not to be interpreted
as yet another optimistic managerial solution, however; it simply corrobo-
rates Gumbrecht’s observation that, outside of Greek tragedy, “there is no
general rule which posits than in certain conditions an experience of a para-
dox must lead to rupture” (1991, p. 492).
The common belief, however, is that it must, and the Swedish officials’
past mastery of “reflective action” has been, under the 1990s, submerged
under the heavy impact of another mode of dealing with paradoxes, more
typical of the action-prone private sector.
9Observe that it was not a simple coincidence. He was a member of the Emergency Committee
because he was the Municipal Finance Manager.
135
Slaying the Medusa, or the manager-hero

On Perseus’ feet were the flying sandals, and across his shoulders was slung
the black-bound sword, suspended on a sword-belt of bronze, and he hov-
ered like a thought in the mind, and all his back was covered with the head
of the monster, the dreaded Gorgo, and the bag floated about it, a wonder to
look at, done in silver, but the shining tassels fluttered, and they were gold,
and the temples of the lord Perseus were hooded over by the war-cap of
Hades, which confers terrible darkness. (Hesiod, Shield of Heracles
220–237)10
Perseus (by Fredrick Pomeroy 1898, Cardiff) Foto: B. Czarniawska.
This paragraph describes a triumphant hero, very much like the heroes
described in Fortune:
“There is, believe it or not, some academic literature that suggests that lead-
ership doesn’t matter,” we are told by the astonished Fortune writer. Well,
this academic is no less astonished: there are, believe it or not, some business
magazines so mesmerized with leadership that nothing else matters. “In four
10 www.theoi.com/Pontos/Gorgones.html; [Accessed April 13, 2002].
136
years Gestner has added more than $40 billion to IBM’s share value,” this
magazine wrote on April 14, 1997. Every penny of it! Nothing from the hun-
dreds of thousands of other IBM employees. No role for the complex web of
skills and relationships these people form. No contribution from luck. No
help from growing economy. Just Gestner. (Mintzberg, 1999)
Hesiod actually agrees with Mintzberg more than with Fortune. Continuing
with Shield of Heracles:
The son of Danae, Perseus himself, sped onward like one who goes in haste
or terror, as meanwhile the rest of the Gorgones tumbled along behind him,
unapproachable, indescribable, straining to catch and grab him, and on the
green of the steel surface gibbered the sound of their feet on the shield run-
ning with a sharp high noise, and on the belts of the Gorgones a pair of
snakes were suspended, but they reared and bent their heads forward and
flickered with their tongues. The teeth for their rage were made jagged and
their staring fierce, and over the dreaded heads of the Gorgones was great
Panic shivering. (ibid)
The second paragraph of Hesiod’s description was forgotten. Yet, beautiful

and triumphant as he was, Hesiod’s Perseus was flying in terror from the
remaining two Gorgons (for Percy Barnevik, the ex-CEO of ABB, one of
them must have had the Chairman of the Board, Martin Ebner’s, face).
Medusa can be killed, because she represents the logical tradition of dealing
with paradoxes.11 The other two Gorgons, standing for rhetoric and theol-
ogy, are immortal.12 But the myth of the paradox-slayer persists, and the
immortality of Gorgons – and paradoxes – is refused. Let me illustrate it
with an excerpt from my “shadowing” in the Finance & Accounting
Department of a Swedish municipality that I studied twice: once in the late
1980s, and again in the late 1990s (Czarniawska 1999). As the excerpt is
anonymized, I may permit myself a dose of irony and present the three inter-
locutors as The Hero, The Wise Guy, and The Young Guy.
11 In a feminist reading, Medusa stands for female rage and creativity (Pegasus was born from
her head). This reading is reinforced by the historical development – from the 5th century BC
onward, Medusa is represented with a beautiful face. It is also speculated that the Greeks bor-
rowed Medusa from Libyan religion, where she was one of three goddesses representing wis-
dom – thus the snakes around her head (www.arthistory.sbc.edu, accessed 14 April 2002).
12 In Luhmann’s opinion, “our society offers the choice of trusting religion or working off our
own paradoxes without becoming aware that this is religion” (1986b, p. 188).
137
Excerpt: A meeting of three managers in the Finance &

Accounting Department
The head of the department (Hero) and his two subordinates, the head
of the budget unit (Young Guy) and the head of the accounting unit
(Wise Guy) are editing a document concerning the budgeting and
budget follow-up procedures. Young Guy has written the draft of the
document, but Wise Guy produced a preparatory sketch that is often
mentioned now:
Hero: Let’s now talk about what Young Guy has written about
budgeting and budget follow-up procedures.
Young Guy: I think exactly the same as Wise Guy wrote on the first
page of his text – the really important question is the division of respon-
sibility …
Hero: When I look into Wise Guy’s notes, I see that he has written
about responsibility. This is right. This is how it should be. Young Guy,
can this part be written like the text Wise Guy wrote? With concrete
points?
Young Guy: But these are concrete points.
Wise Guy: It’s only bullets instead of numbers …
Hero: Well, I do not like the phrase “The group makes a decision …”
I cannot understand how a group can make a decision. Afterwards,
somebody is going to ask: “Who made this decision? Ah, it has been
decided …”
Young Guy: I intended to say that you, or whoever is the boss, makes
a decision. But I am prepared to clean it all out; I do not want to pro-
voke anybody.
Hero: We shall provoke! We can then retreat, but we must begin by
saying how it is. It is extremely dangerous to say that a group should
make a decision.
Young Guy: I shall try to avoid this in the future. (…)
Hero compares Wise Guy’s and Young Guy’s drafts: What does
“cooperate” mean in this context?
Wise Guy: That I operate, and you see to it that it works.
[They all laugh].
Wise Guy: I think that we should have some bit of paper from the
City Office that justifies what we are concocting here.
Hero: And then we can say directly: “As a task given to us by the
City Manager…” Because, you know, I do not believe in collective deci-
sions. … Wise Guy, we must stand for what we are saying. We cannot
accept anything that we know is stupid. If they decide otherwise, we
can always say afterwards: “Didn’t we say so?”
138
Wise Guy: Well, there are more of us than there is of them!

Young Guy: I do not want to live in a continuous conflict.
Hero: It’s the other way around. If we say nothing, the conflict will
continue. The City Manager must say it clearly: this is the way it’s
going to be.
Wise Guy: But will he?
Hero: I imagine so. He is the City Manager. He is the boss.
Wise Guy: Well, I mean – does he share our view?
Hero: No. But then he must say just that.
Wise Guy: I am afraid that it’s going to end with “we shall see”.
Hero: I for one am not going to sit quietly. We cannot wait. Our exis-
tence is at stake.
Young Guy: If we have responsibility we have to have authority.
Wise Guy: I do not think we have either.
Hero: I am in a learning situation. But I become very angry when this
kind of trash shows up. Perhaps I do not always show how angry I am,
but I am angry. I think that it is peculiar when my collaborators
– nothing personal against you, Young Guy – write a memo that I
know nothing about.
Young Guy: It is equally peculiar for me. I am torn between
Councilor X and you. I do not want sit on this fence any longer!
Wise Guy: I want this document to state clearly what my unit may
and can do. I know that various bits and pieces can be put together in
various ways, but I want to know in which.
The Hero represents, in a nutshell, the attitude typical of managers brought

from private sector to public sector organizations in the 1990s. He stands
for individual decisions, and is squarely facing his responsibilities. He wants
to provoke, and is not shy of conflicts: he is not going to sit quietly, and he
can become angry. He is also quite ignorant of most of the technical aspects
of his work. But the main point is that he is unable to entertain the possi-
bility that they are given an opportunity to formulate their own work rules,
which subsequently might or might not be considered valid. For him, ambi-
guity is an aberration that must be fought against.
His subordinates are pulling his leg, but gently – after all, he is their boss.
The Wise Guy reveals long experience and full awareness of many a para-
dox in municipal practice. His tactic is not to enter confrontation, but to
secure and deparadoxize those bits that can be so treated. The Young Guy
is nervous; he is on the side of the Wise Guy, but he does not want to sit on
a fence, of any kind. The times are a-changing, but in what direction?
The reader may think that this is an insignificant episode from a life in a
139
big organization, but it is precisely such insignificant episodes that eventu-

ally constitute great transformations. The early 2000s reveal, however, a pic-
ture that is even less clear than before. After a period during which the deci-
sive heroes celebrated their triumphs almost daily, the dot.com death and its
various concomitant financial scandals swept a chill over the initial wave of
enthusiasm. As usual in times of crises, a reflection is necessary, and action
paralysis is actually welcome. When in doubt, do nothing.
Such immobility is, however, always temporary; modern organizations are
action-oriented units. It is difficult to say what the future will bring. The
capacity for forgetting is enormous; after all, the financial sector was shaken
in the early 1990s by the junk bonds scandal (Corbridge et al. 1994), which
did not save it from falling in a similar way for dot.coms. Still, a compara-
tive study of General Directors (heads of public agencies that in Sweden are
autonomous from the respective Ministries) showed an interesting difference
between their views of the world in 1984 and in 1999 (Czarniawska 1985;
Genell 2000). In 1984, the GDs were fully aware of an identity crisis in the
Swedish public sector, which practically overnight turned from Savior to
Villain. Unwillingly, they admitted that the private sector was the obvious
source of inspiration and imitation. Consultants should help to imitate con-
crete solutions, but a direct import of managerial personnel was also neces-
sary (the Hero was one of the imported). In 1999, the GDs saw the public
sector as clearly and importantly different from the private sector. In their
opinion, the public sector was much more complex, and built around the
central concept of democracy, which cannot and must not be subordinated
to efficiency. Consequently, they saw no clear role models for their organi-
zations. They believed in the growing role of information technology and in
the development of locally defined competence.
Luhmann was right: practitioners abhor reflection and escape into action,
hoping for its deparadoxizing effects. But Lyotard was also right: no renew-
al without rupture, without a pause in routine action.
Observers of institutions: Paradoxes and

organization theory
Hans Ulrich Gumbrecht (1991) reconciled Luhmann’s caution and Lyotard’s
enthusiasm in the face of paradox. As a rupture of individual experience in
a real-life tragedy, he said, the paradox reveals a culture’s failure to solve its
problems (which is why the same tragedies on stage are so instructive, with-
out being threatening). As an object of reflective observation, the paradox
indicates the possibility for a culture to surpass itself, to create new solutions
and to build new institutions. The increased visibility of paradoxes signals
an epistemological crisis within a tradition (MacIntyre 1988), which in an
140
organizational context usually takes the form of a legitimacy crisis or an

identity crisis.
The direct experience of paradox is threatening to people and institutions;
as a topic for reflection – when the experience is indirect – it seems to lead
to renewal. May one extend it to a case of organized reflection (to wit:
organization theory) and see that it also opens possibilities for institutional
renewal? Ruptures in unity are culture’s windows on chaos – both chaos-
without-culture and chaos-as-prelude-to-cultural-renewal (Czarniawska
2001).
Paradox can thus be seen as an opportunity for the renewal of language
and the transformation of institutions. And as Jerome Bruner, who intro-
duced the narrative approach to psychology, pointed out, the force of the
narrative is that it can accommodate paradox: it serves the purpose of social
negotiations by positioning the exceptional alongside the conventional and
rupture alongside unity, and by offering a variety of linear plots (1990).
Paradox destroys the rational decision model, but only makes the narrative
more interesting. Paradox is an enemy of the paradigmatic mode of know-
ing, but at home with the narrative (syntagmatic) one.
Thus, if we, researchers, “catch” actors in contradictions – even collective
actors, like the public administration – it does not mean that we are
”unmasking them” due to the superiority of the logico-scientific mode over
the narrative. It may merely be that we have failed to observe how tempo-
rization has been employed – perhaps because it was done tacitly. Perhaps
we forgot to “relativize” – to relate ourselves to the actor’s standpoint – or,
in Latour’s words, we glossed over an actor’s meaning and imposed our cat-
egories instead of deploying the actor’s own (Latour 1993, p. 140). If an
actor does not regard something as a contradiction and we do, it is then our
task to see how this is possible, rather than solve our own paradox by
“pulling rank” of the ourselves-as-observers-of-the-second-order (the word
“meta” is usually rolled on scientific tongues like the sweetest candy). It is a
sobering thought – and one that should always be present – that in our own
practice, which is research, we remain blind to our own role and position,
like any other first-order observer. Making distinctions, we cannot see our-
selves making distinctions. This applies not only to us as individual
researchers, but also and above all to the disciplines we represent and obey.
In the ironic eye of that arch-pragmatist Thurman W. Arnold, researchers
were at the center of the business of deparadoxization. This is what he wrote
in Symbols of Government in 1935 (p. 18):
To accomplish this task of keeping alive faith that institutions are somehow
rational, the study of government is divided into three separate sciences in
order to conceal the fact that the symbols (…) are contradictory. Law proves
that good government is achieved by constantly refining and restating rules.
141
Economics convinces us that everything will work out all right if we only
leave it alone. Sociology, a loose and cloudy way of thinking, provides us a
shelf on which we may put the humanitarian ideals which run counter to the
eternal rule-making of the law and the eternal automatism of economics. This
makes the intellectuals happy because they can toss facts inconvenient to one
science over to another science for cataloguing and classification. Unthinking
people are comforted by the belief that somewhere in books which they never
have time to read there is an absolute proof of the rationality of their sym-
bols.
When the unthinking people – intellectuals included – start thinking, they

notice the paradox, and this can lead to frustration or to a new attempt at
deparadoxization, and consequently to innovation. Blurring the genres in
the social sciences (Geertz 1980) is one possible way of going beyond the
disciplinary division of labor in academia. This is certainly an innovative
move, although it causes frustration among young novices in the social sci-
ences, who want to know what is what and take notes.
One of the central paradoxes in the important cultural narratives is the
simultaneous presence of convention and exception, of tradition and
change. On a personal level, “the Self … stands both as a guardian of per-
manence and as a barometer responding to the local cultural weather”
(Bruner 1990, p. 110). On a societal level, “the culture … provides us with
guides and stratagems for finding a niche between stability and change”
(ibid.). Therefore, “… the viability of a culture inheres in its capacity for
resolving conflicts, for explicating differences and renegotiating communal
meanings” (Bruner 1990, p. 47). Bakhtin (1981) was thinking along these
lines when he pointed out the dangers of a monologic discourse, in which
some kind of authoritarianism – political, literary, scientific – smoothes out
tension and conflict. In his view, all knowledge development emerges from
a “sharpened dialogic relationship” between concepts and meanings that
already exist (Kelly 1992). Thus, the much-discussed “danger of incommen-
surability”13 pales in comparison with the danger of flying from reflection.
Not that it is possible, according to Luhmann, who saw modern society as
a society of observers of systems, a society that was forced, therefore, to
resign from authority and to espouse ignorance. As we cannot kill all the
Gorgon sisters, we may want to learn to live with them – in the theory as
well as in the practice of organizing.
13 See the special issue of Organization, 1998, 5(2).
142
10 Kap 7. Seidl 05-11-11 16.21 Sida 143
PART III
Organization, Interaction
and Society
10 Kap 7. Seidl 05-11-11 16.21 Sida 144
10 Kap 7. Seidl 05-11-11 16.21 Sida 145
Chapter 7
Organization and Interaction

David Seidl
Introduction
Niklas Luhmann distinguishes in his theory three types of social systems:
society, organization and interaction. While the functioning of each of the
systems (as an autopoietic system) has been described fairly well (for socie-
ty, see e.g. Luhmann 1993j, Part II; 1995f, Chapter 10; 1997a; for organi-
zations, see e.g. Luhmann 1992a; 1993j, Part III; 1993d; 2000c; and for
interactions, see e.g. Luhmann 1995f, pp. 405–436; 1999b; or Kieserling
1999), the relation between the systems has received comparatively little
attention. Particularly unclear remains the relation between organization
and interaction. Apart from a few remarks here and there (e.g. Luhmann
2000c, pp. 25, 255, 373) Luhmann himself has not written anything on that
issue. The only serious contribution comes from André Kieserling (1994;
1999, Chapter 11), who is not, however, very explicit on how to conceptu-
alise the relation in the “technical” terms of Luhmann’s theory.
Luhmann’s systems-theoretical perspective initially seems to oppose the
idea of organizational interactions: organization and interaction are con-
ceptualised as two systems that are operatively closed with regard to each
other; i.e. they cannot take part in the autopoiesis of each other, as, for
example, they take part in the autopoiesis of society. Thus, what can be
observed empirically – that there are interactions that make organizational
decisions and in this way contribute to the reproduction of the organization
– seems (at first) theoretically impossible. While the theory indeed excludes
the possibility of interactions becoming part of the organization, there are
other, more complex ways of conceptualising the contribution of interac-
tions to the organizational reproduction in terms of Luhmann’s theory,
which will be explored in this chapter.
In the following we will first explain Luhmann’s concepts of organization
and interaction as two autopoietic systems that reproduce themselves on the
basis of different operations. We will then explore ways of conceptualising
organizational interactions using the concepts of re-entry and interpenetra-
tion. This will be followed by an exploration of functions that interactions
might serve in organizations, and the ways in which organizations can con-
dition interactions for their purposes. We will close with a brief reflection on
145
10 Kap 7. Seidl 05-11-11 16.21 Sida 146
the advantages of the proposed conceptualisation of organizational interac-

tions.
Organization and interaction as two types of

social systems
According to Luhmann’s theory organizations and interactions can be con-
ceptualised as two types of social systems that reproduce themselves on the
basis of different types of communication: organizations on the basis of
decision communications, and interactions on the basis of communications
among people present. Both systems are perceived as operatively closed in
the sense that they can only operate in the context of their own operations
(Luhmann 2000e, pp. 51–52): organizations can only reproduce themselves
through (their own) decisions and interactions only through (their own)
communications among people present.
Organizations are conceptualised as autopoietic “systems that consist of
decisions and that themselves produce the decisions of which they consist,
through decisions of which they consist” (Luhmann 1992a, p. 166; my
translation). Three concepts are central to this understanding of organiza-
tion: the concept of decision, the concept of decision premise and the con-
cept of uncertainty absorption. Decisions constitute the elements of the
organization. They are a specific kind of communication: they are commu-
nications that communicate their own contingency (Luhmann 1993h, p.
339). That is, they communicate the decision as such and the alternatives to
the decision. In this sense decisions are paradoxical communications: every
decision communicates that there are alternatives to the decision – otherwise
it would not be a decision – and it simultaneously communicates that since
the decision has been made, there are no alternatives – otherwise, again, it
would not be a decision (Luhmann 2000c, p. 142). Because of their para-
doxical nature, decision communications at first may appear as extremely
improbable phenomena – they call for their own “deconstruction”
(Luhmann 2000c, p. 142). Empirically, however, one can observe that deci-
sion communications do succeed; in fact, in organizations decision commu-
nications are normality. How does the organization prevent the constant
deconstruction of its decisions? Firstly, this happens through the totalisation
of decision expectation (Luhmann 2000c, p. 145; Kieserling 1999, p. 352).
Any communication reproducing the organization is forced into the form of
a decision. Thus, even the deconstruction of a decision would itself have to
take the form of a decision. Secondly, an organizational decision takes place
in the context of other decisions which it refers to in order to stabilise itself
as a decision. In other words, every decision communicates also a “meta-
communication”, which says that the “decision maker” had the right,
authority or other good reasons for deciding the way he did (Luhmann
146
10 Kap 7. Seidl 05-11-11 16.21 Sida 147
2000c, p. 142). In general terms, decisions are stabilised through reference

to organizational structures.
With the concept of decision premise Luhmann captures the structural
aspect of organizations. Every decision constitutes a decision premise for
later decisions as it defines the structural conditions for it. Decision premises
here serve a double function in that they both create and restrict the deci-
sion. They create the decision situation in the first place: without decision
premises there is no occasion for decision making. At the same time, deci-
sion premises restrict the decision situation by restricting the decision possi-
bilities. Decisions not only serve as decision premises for succeeding deci-
sions, but can also decide on decision premises that are binding for a multi-
tude of later decisions. Luhmann distinguishes between three types of such
far-reaching decision premises: programmes, communication channels and
personnel (Luhmann 1992a; 1993d; 2000e).
With the concept of uncertainty absorption Luhmann captures the proces-
sual aspect of organizations; the question of how one decision leads to fur-
ther decisions and thus to a reproduction of the organization. That is, uncer-
tainty absorption describes the autopoiesis of the organization (Luhmann
1993d, p. 308). In the transition from one decision to the next one the
uncertainty involved in the original decision disappears. In other words, a
connecting decision takes the result of a preceding decision as given (the
decision was made!) and uses this result as a premise for itself. One could
also say, the first decision “in-forms” the decision following it only about
the result of the original decision but not about the uncertainty involved in
the decision making. To bring the two concepts together one can define:
uncertainty absorption takes place when earlier decisions are accepted as
decision premises for later decisions (Luhmann 1993d, p. 299; 2000c, p.
193).
The system of face-to-face interaction, in contrast to the system of
organization, does not consist of decision communications but of communi-
cations that are based on reflexive perception of the physical presence of its
participants. Analogously to the communications of organizations, interac-
tional communications contain a meta-communication communicating that
it is a communication amongst people present (Kieserling 1999, p. 67). In
other words, the communications carry the code presence/absence. With this
code, interactions refer to their system/environment distinction. Whoever is
considered present is treated as a participant in the communications (if not
as speaker then as listener) and whoever is considered absent is treated as
environment (Luhmann 1997a, p. 815). The difference of presence and
absence in this sense marks the difference of self-reference and hetero-refer-
ence for the operations of the system (Luhmann 1997a, pp. 815–816). Like
all social systems, interactions are complex systems, which means that they
are forced to make selections. Luhmann (1995f, p. 415) distinguishes
147
10 Kap 7. Seidl 05-11-11 16.21 Sida 148
between factual, temporal and social structures (on these three dimensions
of meaning generally see Luhmann 1995f, pp. 76–81): interactions usually
produce topics for communication; they place communications in a
sequence and, more particularly, they also take the end of the interaction
into account; they develop rules for taking turns in the communication – in
particular, when somebody speaks the others have to remain quiet.
Analogously to the organizational reproduction, where every decision cre-
ates the necessity for further decisions, every interactional communication
necessitates further interactional communications. Reflexive perception in
the interaction enforces a continuation of the communication. It is almost
impossible to evade communication in an interaction. As Luhmann writes:
“Even the communication of not wanting to communicate is communica-
tion. […] In practice, one cannot not communicate in an interaction system;
one must withdraw if one wants to avoid communication” (Luhmann,
1997a, p. 413; original emphasis).
In interaction systems the person (defined as a bundle of social expecta-
tions; see Luhmann 1995a) plays a very prominent role. While in organiza-
tions decisions are largely justified by reference to earlier decision premises
– mostly decisions are presented as a direct consequence of earlier decisions
– in interactions, communications are primarily attributed to persons; the
reasons for particular communications are primarily sought in the person
and his particular intentions. In interactions the person as such is held
responsible for what is communicated – more so than in organizations.
Ensuing communications take the personal aspect of earlier communications
into account and very often address that personal aspect explicitly; organi-
zational decision communications, in contrast, tend to focus on the result of
the decision and only in exceptional cases address the particular (personal)
circumstances of the earlier decision situation.
Conceptualising organizational interactions

Organization and interaction are two types of social systems that reproduce
themselves in very different ways. Nevertheless, the (theoretical and empiri-
cal) literature speaks of interactions taking place “within” organizations.
How is one to conceptualise the relation of organization and interaction in
such cases? “Within”, first of all, could be meant in a local sense: for
example, meetings take place in the offices; two workers at the conveyer-belt
talk about the new working hours. From a Luhmannian perspective the
locality of the interaction would not matter. Organization and interaction
are two different systems that constitute an environment for each other; that
is to say, the systems are operatively closed to each other. This inter-
pretation, however, is problematic, when we take into account that such
interactions might produce decisions for the organization; that is to say,
148
10 Kap 7. Seidl 05-11-11 16.21 Sida 149
interactions seem to play a role in the organizational reproduction – and in

this sense could seem to contradict the idea of operational closure.
An alternative solution to this problem could be to conceptualise the rela-
tion between organization and interaction analogously to that between
society and organization or society and interaction. Society constitutes the
all-encompassing system of communication in which organization and inter-
action differentiate themselves on the basis of specific communications
(Luhmann 1997a). The operations of organizations are decision communi-
cations, while those of interactions are interactional communications. Thus,
organization and interaction reproduce with every operation both them-
selves and the society, as their operations are decision communications or
interactional communications, both of which are communications. Society
can thus be found both inside and outside the organization or interaction
(Luhmann 1995f, p. 405 ff.; Luhmann 2000c, pp. 59, 383). Analogously,
one might try to conceptualise organizational interactions as social systems
that take place in organizations in the sense that they reproduce themselves
by “interactional decision communications”; that is to say, as social systems
that reproduce themselves by a specific kind of decision communications. In
this sense they would reproduce both themselves and the organization simul-
taneously. Such a conceptualisation, however, would not be compatible with
the empirical observation that decision communications within “organiza-
tional” interactions are relatively rare. Most communications – even in for-
mal meetings – are non-decision communications. In fact, it is particularly
the non-decision communication that is typical for (organizational) inter-
actions; decisions constitute a kind of “foreign matter” to the interaction (cf.
Kieserling 1994a; 1999, pp. 355 ff.). Often no decisions at all are made in
such interactions. Furthermore, sometimes part of an interaction is related
to the organization while other parts have nothing to do with the organiza-
tion: a business meeting ends in bed; during a board meeting some members
of the board discuss the latest results of the Football World Cup; after their
coffee break together colleagues make their way back to a meeting (cf.
Atkinson et al. 1978). In all these cases it is not that one interaction is ter-
minated and a new interaction starts but it is one and the same interaction.
Thus, the relation between organization and “organizational” interaction
has to be conceptualised differently from one of inclusion.
Before we attempt a conceptualisation of “organizational” interactions in
terms of Luhmann’s theory we will first give a short overview of the differ-
ent empirically observable forms of interactions “in” organizations. Firstly,
there are “deciding” interactions in which decisions for the organization are
made; one might think of board meetings, strategy meetings etc. Secondly,
there are “decision-preparatory” interactions in which the conditions for
later decisions are laid. That is to say, in the interaction itself no decisions
are made but later decisions are prepared: for instance, during interactions
149
10 Kap 7. Seidl 05-11-11 16.21 Sida 150
in which relevant information for later decisions is exchanged the partici-

pants may deliberate later decisions without making decisions. Thirdly,
there are “semi-detached” interactions with a merely loose relation to deci-
sion making; that is, interactions that communicate about decisions without
focusing directly on influencing decision making. One might think of infor-
mal interactions between members of an organization, such as gossiping
about organizational issues (cf. March and Sevón 1989; Drew and Heritage
1992; Boden 1994; Fuchs 1995; Kurland and Pelled 2000). Fourthly, there
are interactions unrelated to organizations, in which topics that are unrelat-
ed to the organization are communicated; e.g. members of an organization
talk during work about their night out together. These different forms of
interaction differ in the degree to which they contribute to the organiza-
tional reproduction: from a direct involvement in decision making to com-
plete detachment from the organization.
To start with we can exclude the fourth type of interaction from the list of
possible forms of “organizational” interactions. It can be treated as a nor-
mal interaction in the environment of the organization (cf. Kieserling 1997,
p. 360). Its location within the physical boundaries of the organization and
the involvement of members alone does not grant this type of interaction a
specific status with regard to the organizational reproduction. The other
three forms of interaction, however, have an organizational significance.
If one cannot conceptualise the relation between organization and
“organizational” interaction as one of inclusion, there remains only the pos-
sibility of treating it as a special kind of structural coupling; that is to say,
the structures of two systems are adjusted to each other in a way that allows
mutual influences on their respective reproductions (Luhmann 1995e).
Organizational interactions reproduce themselves on the basis of their inter-
actional communications but take the organizational operations into
account. That is, organizational interactions allow for a conditioning of
their own operations through the organization – in social, factual and tem-
poral respects. In other words, interactional communications orient them-
selves not only according to their interactional structures but also according
to organizational structures, that is to say, their environment. In terms of
systems theory this has to be conceptualised as a re-entry (Spencer Brown
1979) of the interaction/organization distinction (as the system/environment
distinction) into the interaction (as system). The interaction observes its own
communications with regard to their significance for the interaction and for
the organization. Of course, the interaction can only observe this distinction
according to its own logic, its own construction: the re-entered distinction is
not the original distinction; it is a “mark” and not a “cross”, as Spencer
Brown (1979) would put it. This means that the distinction is reconstructed
with regard to the interacting persons: as a distinction between their
organizational roles (and generally: membership) and interactional roles (cf.
150
10 Kap 7. Seidl 05-11-11 16.21 Sida 151
Kieserling 1999, pp. 360 ff.). In other words, interactional communications

take into account that the persons involved in the interaction have other
roles outside the interaction in the organization and that these roles have to
be taken into account in the interactional communications. Or, put the other
way around, the interaction treats the difference between the interactional
and organizational roles of its participants as the distinction between itself
and the organization. Interactional communications in this sense can be
observed with regard to interactional roles and with regard to organization-
al roles. In Heinz von Foerster’s terminology (1981, pp. 304–305; 1993b),
the interaction uses the (re-entered) interaction/environment distinction, in
the form of a difference in roles, as programme.
The re-entry of the organization into the interaction does not, however,
mean that organizational structures – the organizational logic – are intro-
duced into the interaction. First of all, due to the different forms of repro-
duction, organizational structures cannot serve as interactional structures –
they belong to different domains. The interaction, however, can reconstruct
the structures in interactional terms. For example, the hierarchical structure
of decision communications can be translated into a stratified order of
respect with regard to persons in different positions, a differential treatment
of persons with regard to their contributions to the communication, e.g. with
regard to turn-taking, granted speaking time, or with regard to possibilities
of criticism. Apart from that, the re-entry does not mean that the interaction
is trying to translate the organizational structures into interactional struc-
tures – attempting, that is, a representation of the organization. Even the
opposite could be the case: the interaction could counter the organizational
structures, e.g. by a perhaps explicit suspension of the hierarchical order of
positions. The important point, however, is that the interactional structures,
whether imitating or contrasting the organizational structures, refer to the
organizational structures. Without explicit reference to the organization, the
interaction could not be treated as “organizational”.
Although the re-entry of the organization into the interaction is accom-
plished through the reference to the participating persons (thus the social
dimension is ultimately central) the organization can be reflected in the
interaction in all three dimensions of meaning (Luhmann 1995, pp. 75–81).
In the social dimension – as already discussed above – the interaction takes
the participants’ personal experiences and actions as members of the
organization (which largely result from the organizational roles of the par-
ticipants) into account. In the factual dimension, the interaction structures
its communications according to organizational themes, e.g. the interaction
might communicate about new strategies for the organization or the official
organizational self-description (cf. Seidl 2005). Again, this does not mean
that these themes have to be dealt with in the same way as they are dealt
with in the organization. The participants might treat the relevant theme
151
10 Kap 7. Seidl 05-11-11 16.21 Sida 152
cynically, i.e. communicate about the “other side” of the theme; that, which
the official version of the theme disguises (cf. Luhmann 2000e, pp. 430 ff.).
What organizational themes are dealt with and how depends on the partici-
pating persons. As with all interactions, the personal relation to the topic is
decisive and contributions to a topic are attributed personally.
In the temporal dimension interactional communications might reflect the
temporal structures of the organization; for example, by scheduling the
interaction so as to coincide with external organizational events. The inter-
action might be terminated because the participants are needed elsewhere in
the organization, the participants in a meeting might wait for a latecomer
who is held up in another meeting, an important topic at the top of the
agenda might be moved to the end in order to allow for additional infor-
mation to be provided by the organization (e.g. the results of a parallel meet-
ing); or, vice versa, a topic might be dealt with early in a meeting in order
for its result to be available to other organizational communications. An
interaction might also be intentionally protracted in order to obstruct cer-
tain organizational processes. Again, more than anything else, the reflection
of temporal structures of the organization in the interaction is accomplished
through the focus on the persons participating and their organizational
roles.
The structuring of the interaction is very much oriented according to the
time available to the participants; the participants can bring the interaction
to an end by referring to other organizational obligations that coincide tem-
porally. In contrast, references to extra-organizational obligations – e.g. one
has to pick up one’s children from the kindergarten – are accepted as an ori-
entation for the temporal structuring of the interaction only in exceptional
cases (cf. Kieserling, 1999, p. 360). Of course, the “organizational” interac-
tion – to the extent that it takes place during “working hours” – can count
on the exemption of its participants from any other extra-organizational
obligations for the time being; that is to say, the availability of the partici-
pants is guaranteed by the organization and only organizational obligations
can thus count as reasons for participants’ terminating their presence.1 Even
if the temporal structuring is not oriented directly according to the tempo-
ral availability of participants, temporal structures of the organization are
still “transported” into the “organizational” interaction through the
1 In the case of “after-hour” meetings the situation is somewhat different. Whether or not one
can claim other obligations depends very much on the organizational “arrangements”. In
many contemporary organizations, for example consultancy firms, working hours are not
really pre-defined but depend on the particular tasks at any given time. In other words, mem-
bers are expected almost always to be available for the organization. This does not mean that
they will always be available, but that reasons for their non-availability have to be formulat-
ed as “organizational reasons” – e.g. one has to take a weekend off to keep a clear head for
an important meeting on Monday.
152
10 Kap 7. Seidl 05-11-11 16.21 Sida 153
participating persons. For example, an agreement has to be reached by a

certain time because one of the managers had promised so, or because the
career of one of the participants depends on it, or because it would make it
easier for the participants to continue their ordinary work. Or an informal
meeting in the corridor is terminated because the participants have to get
back to their desks to avoid being reprimanded for being lazy.
So far we have described some features of organizational interactions, in
particular the re-entry of the organization into the interaction. One could
also say that the re-entry of the organization into the interaction is the defin-
ing criterion – a re-entry pertaining to the interactional structures and not to
the autopoiesis of the organization (cf. Kieserling 1999, p. 79): to the extent
that such a re-entry takes place it is an “organizational” interaction. But
how can this re-entry be explained? Why would an interaction orient itself
according to an organization? An easy but somewhat unsatisfactory answer
would be that an interaction as an autonomous system – autopoiesis implies
autonomy (Luhmann 2000c, p. 51) – is free to select its own structures and
if it thus selects its structures with regard to an organization, then it does so
– it is its own selection; and if it does not, it does not. In this sense only the
interaction itself can determine whether or not to orient itself according to
the organization; in particular, according to the organizational roles of its
participants.2 It is the interaction itself that sees itself as “organizational”. It
has to be understood as a “self-attribution” (Kieserling, 1999, p. 79; see also
Luhmann 1999b, p. 64) of the interaction to the organization. In order to
ascertain whether an interaction attributes itself to the organization one has
to observe the interactional communications – whether or not the interac-
tion/organization distinction serves as a guiding difference for the commu-
nications.
What is of specific interest here is how the interaction describes itself, i.e.
what self-descriptions it uses in its communications. Interactions, in contrast
to organizations (Luhmann 2000c, Chapter 14; Seidl 2002, 2002b) or soci-
eties (Luhmann 1997a, Chapter 5), usually do not possess very elaborate
self-descriptions. Nevertheless, some (at least rudimentary) self-descriptions
can always be found. An interaction might describe itself as “strategy
review”, “crisis meeting”, “board meeting” or as “chance encounter”,
“flirt”, “class reunion” etc. The self-attribution of an interaction to an
organization would normally be reflected in the self-description, as is evi-
dent in several of the above examples.3
2 Or, put the other way around, whether to disregard any extra-organizational roles of its par-
ticipants. Alternatively, the interaction could include, for example, all the other roles of its par-
ticipants and in that way create an “intimate interaction” (cf. Luhmann 1995f, pp. 224 ff.),
rather than an organizational one.
3 One can also find explicit dissociations from the organizations in interactional self-
descriptions (e.g. “This is not a business meeting”) where the “vicinity” (in physical or tem-
poral respects) would otherwise suggest a self-attribution to the organization.
153
10 Kap 7. Seidl 05-11-11 16.21 Sida 154
The interactional self-attribution to an organization – as we indicated

above – need not refer to the entire interaction.4 In the course of their com-
munication processes interactions can change their focus. An interaction
might start as organizational and end as intimate (cf. Kieserling 1999,
p. 162). Any change of self-attribution – this cannot be emphasised enough
– is the interaction’s own “decision”. This, however, means that the interac-
tion first has to free itself from the organization-related structures (from its
guiding difference interaction / organization) before it can switch to the
structures of intimacy.5 Or an organizational interaction might suspend its
organizational self-attribution for a tea- or lunch-break (during which the
participants stay together) and resume its organization-related communica-
tions afterwards (cf. Atkinson, Cuff and Lee, 1978).
Beyond the mere establishment that an interaction attributes itself to an
organization, we can also try to answer why an interaction might attribute
itself to an organization. One reason is that for many interactions the refer-
ence to the organization is close at hand – in terms of Goffman’s concept
(1975) the organization offers a “frame” for the interaction. Often, partici-
pants of an interaction know each other only as members of an organiza-
tion. They might even just know that the other person is a member of the
same organization. As such the orientation according to the organizational
roles of the participants might be close at hand. This provides the interac-
tion with a point of reference for establishing its structures. Most impor-
tantly, it provides the interaction with topics of communication (cf.
Kieserling 1999, pp. 181, 205). Where participants who hardly know each
other would otherwise struggle to find topics for communication, the orien-
tation according to the organizational roles might provide an almost unlim-
ited source of possible themes.6 In this sense the self-attribution to the
organization might serve as a way of facilitating the reproduction of the
4 See, analogously, the change of self-attribution of interactions to functional systems of

society (Kieserling, 1999, p. 80 ff). There is, however, a difference in that in the case of a self-
attribution to a functional system of society through the specification of structures the inter-
action also uses the function-specific code as a guiding difference for its communications; that
is to say, the interactional communications (partly) take place in the functional system of
society. In contrast to the above, in the case of the organizational interaction, the organization
remains the environment of the interaction.
5 One of the greatest obstacles to such a switch is the coordination of the change of structures.
The changeover of structures has to take place simultaneously (in order to avoid a conflict
between structures) if a breakdown of the interaction and thus the end of its reproduction is
to be avoided. In order to accomplish the switch, the interactional communications might have
to use their reference to the organization; for example, one might refer to the end of working
hours in connection with the suggestion of having a drink together in a nearby pub.
6 Organizational interactions, however, mostly would present the situation as if the topic were
the reason for the interaction even where it was the other way around, i.e. where the interac-
tion came first and had to look for a topic (cf. Kieserling 1997, p. 195).
154
10 Kap 7. Seidl 05-11-11 16.21 Sida 155
interaction. In interactions among people who only know each other as

members of the same organization this self-attribution to the organization
might even serve as a kind of safety net when no other topics are close at
hand. For instance, one might start talking about the weather but after a
while this topic might be exhausted and a supply of other topics may not be
readily available when one does not know the other person. In this case, the
common organization (i.e. the respective memberships) might serve as an
orientation for the continuation of the communication (e.g. “What do you
think about the new CEO?”).
Often interactions form as a “response” to organization-related situations
of double contingency. For example, instead of asking their superior, two
workers who have to coordinate their interdependent tasks might resort to
a mutual interaction as a means of coordination. In such cases the reason for
the interaction seems to precede the interaction, although it ultimately
depends on the interaction (as it is an autonomous, autopoietic system)
whether it forms at all and what structures (in particular: what themes) of
communication, it selects. Nevertheless, the organization almost forces itself
onto the interaction. The self-attribution to the organization is so close at
hand that the interaction almost cannot ignore it and would have to “justi-
fy” not focusing on the organization. What one might find, rather, is that
such interactions may start as “organizational”, i.e. attribute themselves to
the organization, and may later change their self-attribution to something
else – with the difficulties described above.
Deciding interactions
In the following we want to focus on one of the forms of “organizational”
interactions described above: the deciding interaction. The important point
about such interactions is that they produce interactional communications
that can be treated by the organization as (organizational) decisions. That
is, the deciding interaction can contribute to the organizational reproduc-
tion. A particularly pressing question in this context is how to conceptualise
the status of these communications with regard to interaction and organiza-
tion respectively. Are these communications elements of both systems, in the
way interactional communications are also communications in the societal
system (Luhmann 1995f, Chapter 10)? Or are they something completely
different in each of the systems – for example, in the way “one and the
same” communication constitutes a legal communication (as a contract) in
the legal subsystem but a payment-related communication (as a transaction)
in the economic subsystem of society (Luhmann 1993a; Lieckweg 2001). In
order to answer this and other related questions we will apply to the rela-
155
10 Kap 7. Seidl 05-11-11 16.21 Sida 156
tion between organization and deciding interaction7 the concept of inter-

penetration, which Luhmann used for describing the relation between social
and psychic systems8 (1995f, Chapter 6; 1995j).
Luhmann speaks of interpenetration between autopoietic systems, if the
systems reciprocally presuppose the complex achievements of the
autopoiesis of the respective other system and treat them as part of the own
system (Luhmann 1995a, p. 153). If, that is to say, systems reciprocally
make their own “complexity (and with it indeterminacy, contingency, and
the pressure to select) available for constructing [the] other system”
(Luhmann 1995f, p. 213; original emphasis). Luhmann’s classical example
is the relationship between psychic and social system: social systems can rely
on psychic systems being stimulated by communication and processing
“parallel” psychic operations that allow them to “irritate” the social systems
at “suitable” moments in order to trigger further communications that
reproduce social systems.9 Vice versa, the psychic system presupposes com-
plex achievements of social systems, though it is somewhat less dependent
on social systems as it can also reproduce itself (for a certain time at least)
without social systems (Luhmann 1995j, p. 39). The important point in this
respect is that both systems are operatively closed with regard to each other.
It is not that the systems overlap in any way – psychic operations cannot
become part of social systems and vice versa. As Luhmann writes: “This
means that the complexity each system makes available is an incomprehen-
sible complexity […] that is, disorder – for the receiving system” (Luhmann
1995f, p. 214). In this sense, social systems can presuppose the contributions
of psychic systems to their autopoiesis without comprehending the func-
tioning of psychic systems – and vice versa.
Analogously, the relationship between organization and deciding interac-
tion can be described as one between two operatively closed systems, in
which the interaction makes its own complexity available to the organiza-
tion and vice versa. That is, the organization can rely on “its” interactions
to fulfil tasks necessary for the organizational reproduction, although the
organization cannot comprehend the logic of the interactional reproduction.
Luhmann writes, for example: “Work processes that cannot be ‘technisized’,
7 Another concept for analysing the relation between organization and organizational inter-
action is that of medium and form by Fritz Heider (1959). For an application see Seidl (2003a;
2005) and for applications to the relation between social and psychic system see Baecker
(1992a) and Luhmann (1995j).
8 Apart from that, he described also the relation between organizm and psychic system, the
brain and the nervous system (e.g. Luhmann 1995f, p. 214, endnote 7), but not the relation
between social systems.
9 One of the most important achievements that the social system presupposes the psychic sys-
tem will accomplish is perception – in particular the perception of uttered sounds (Luhmann
1995f, Chapter 6).
156
10 Kap 7. Seidl 05-11-11 16.21 Sida 157
for example lessons at school or preaching in church, […] force the organi-
zation to displace its performance conditions into interaction among people
present and to entrust it to inscrutable processes” (Luhmann 2000c, p. 373;
my emphasis; my translation).
In order to fully understand the concept of interpenetration in this context
we have to have a closer look at the decision communications that come
about in the course of deciding interactions. These communications (in some
way) take place in both systems – they are elements in the interactional com-
munications to the extent that further interactional communications connect
to them, and they are organizational communications to the extent that they
serve as decision premises for further decisions. Luhmann writes about inter-
penetration in general:
To be sure, interpenetrating systems converge in individual elements – that is,

they use the same ones – but they give each of them a different selectivity and
connectivity, different pasts and futures. Because temporalized elements
(events) are involved, the convergence is possible only in the present. The ele-
ments signify different things in the participating systems, although they are
identical as events: they select among different possibilities and lead to dif-
ferent consequences. Not least, this means that the convergence to occur next
is once again selection, that the difference of the systems is reproduced in the
process of interpenetration. (Luhmann 1995f, p. 215; original emphasis)
If we look at decisions in organizational interactions this is exactly the case.

We have a communication that is used in both systems, but has a different
meaning in each. In the interaction it has to be understood in the context of
other interactional communications: in reaction to what other interactional
communications did the communication come about, and what other inter-
actional communications can follow? In the interaction one is interested in
who said what, in response to whom, with what reactions from the others.
In this sense the “content” of the decision might be only of secondary
importance. For the organization, the communication has a very different
meaning. It has to be understood in the context of other decisions, which
might even lie outside the interaction: earlier decisions that served as
decision premises and later decisions for which this decision serves as a
decision premise. For the organization the meaning of this communication
results from its implications for connecting decisions, while for the inter-
action it results from the meaning for further interactional communications
that connect to it.10 As Huisman (2001, pp. 75–76) shows, what is taken as
a decision by the organization often corresponds to a whole series of
10 For Luhmann the elements of a system are only defined through their integration into the
system, i.e. their relation to other elements (Luhmann 1995f, pp. 20 ff.). In this sense the
“same” communication is for interaction and organization a different one, as it relates to
other elements of the respective systems in different ways.
157
10 Kap 7. Seidl 05-11-11 16.21 Sida 158
communications in an interaction, in which the decision as such is never

explicated.
We may illustrate this relation between organization and interaction with
an example: a university has decided to create a new position and in a later
staff meeting a particular candidate is chosen for it. For the organization the
decision on the candidate connects directly to the earlier decision to create
the position. In order to understand the relevance of the decision for the
organization it is enough to relate it to this decision as its decision premise
(and of course to other decision premises, e.g. who has the authority to
decide what etc.). For the continuation of the decision process (e.g. further
decisions on concrete curricula etc.) it is only relevant which candidate has
been chosen (and which ones have not). It is completely irrelevant who was
for or against the candidate (for what personal reasons), how long it took
the participants to reach the decision etc. What counts for the further deci-
sion process is the decided alternative, while the process culminating in the
decision, and the uncertainty involved in it are irrelevant or absorbed (cf.
Luhmann 2000c, p. 193). For the interaction, in contrast, it is exactly these
personal aspects of the communication that are important and in the con-
tinuation of the interactional communication this personal aspect will be
referred to (cf. Kieserling 1999, pp. 385–386).
The autopoiesis of organizations – as we explained above – can be
described with the concept of uncertainty absorption. In the transition from
one decision to the next decision the original uncertainty of the initial deci-
sion is absorbed – the ensuing decision only orients itself according to the
chosen alternative. With regard to our argument above, this means that the
“interactional circumstances” – the pros and cons, the doubts etc. – which
represent a great part of the uncertainty involved in the decision making are
left out in the ensuing decision communication. In other words, the decision
does not “in-form” later decisions about these issues – these issues, as
Bateson points out (1972, p. 315), do not constitute differences that make a
difference for the further reproduction of the organization. In contrast, with-
in the interaction, these issues are the differences that make a difference. The
interactional communication of a decision does “in-form” ensuing interac-
tional communications in these respects. Thus, in the interactional context
such interactional decision communications (usually) do not absorb the
uncertainty involved in the decision making. Even if the decision is not open-
ly challenged, the particular circumstances under which the decision was
made will have a great influence on further communications.
We have to distinguish between two different foci of “interactional” deci-
sions. “Interactional” decisions may refer exclusively to the organization
outside the interaction; for example, a decision on a new strategy. Or they
may refer to the interaction itself; that is, they may concern the structures of
the interactional communication: for example, decisions about the topics on
158
10 Kap 7. Seidl 05-11-11 16.21 Sida 159
the agenda and their order, decisions to close or interrupt a meeting. In the
first case the content of the decision communication (i.e. the chosen alter-
native) does not have any implications for the interaction. What is impor-
tant for the interaction is that the decision has been made, but not what deci-
sion has been made. In such interactions, as Kieserling (1999, p. 372) writes,
the decision is declared an “output category” that has to be delivered to the
interaction’s environment. In the second case the interaction treats its own
communication processes (partly) as decision processes. The communica-
tions are “stylised” as decisions and other communications connect to them
as if they treated them as decision premises. The interaction, however, can-
not but take the personal aspects of the communication into account as it is
the basis of its autopoiesis. Thus, even where the interaction presents its
communications as decision processes, the communication “in reality” is
reproduced differently. Thus, in order to understand the interactional com-
munications one has to understand the “interactional meanings” behind the
façade of decision communications (cf. Goffman’s distinction between front
and back stage; Goffman 1959): for example, who decided what against
whom and how does one react to it. The interaction puts its own communi-
cations into an “artificial” (for the interaction) form (Kieserling 1999,
p. 373), but this artificial form is only the surface behind which different
information is latently processed.11 Apart from that, interactions can never
present all their communications as decisions. A great part of the communi-
cations resist being put in the form of decisions – an example of this would
be the expression of surprise about particular opinions etc. (cf. Kieserling
1999, pp. 355–358; Luhmann 2000c, p. 255).
Interactions stylise (some of) their own communications as decisions in
order for the organization to integrate them into the organizational decision
process. That is to say, the interaction observes its own operations both with
regard to their implications for the interactional process and with regard to
the organization. This also means that, to a certain extent, the interaction
can regulate what kind of communications become available to the organi-
zation and what do not; or, in other words the interaction can (to some
extent) influence the organizational boundaries. In order to understand the
influence of organizational interactions on the organization it is important
to understand how the boundaries of interaction and organization relate to
each other. Luhmann writes about the boundaries of interpenetrating sys-
tems – referring in particular to the relationship between social and psychic
system, which can, however, be applied analogously to our question:
11 This does not mean that the interactional communication is “more” than a decision com-
munication, i.e. decision communication plus the interaction-specific communication. Rather,
as we said before, the reproduction of decisions follows a different logic from that of the inter-
actional communications. Or, as Kieserling (1999, p. 358) puts it, between organization and
“its” interactions there exist considerable differences in information.
159
10 Kap 7. Seidl 05-11-11 16.21 Sida 160
Decisively, boundaries of one system can be included in the operational

domain of the other. Thus the boundaries of social systems [analogously:
organizations] fall within the consciousness of psychic systems [analogously:
the communications of interaction]. Consciousness [analogously: interaction-
al communication] intervenes and thereby acquires the possibility of drawing
boundaries for social systems [analogously: organizations] precisely because
these boundaries aren’t, at the same time, boundaries of consciousness
[analogously: interactions]. The same holds conversely: the boundaries of
psychic systems [analogously: interactions] fall within the communicative
domain of social systems [analogously: organizations]. In the course of
orienting itself, communication [analogously: decision communication] is
constantly forced to use what psychic systems [analogously: interactions]
have already assimilated in their consciousness [analogously: interactional
communications] and what they have not. This is possible because the
boundaries of psychic systems [analogously: interactions] are not also
boundaries of communicative [analogously: decisional] possibilities. (Luh-
mann 1995f, p. 217; original emphasis)
Organizational interactions can deal with the organizational boundaries in

a differentiated manner. Not everything that is communicated in the inter-
action has to be put in the form of a decision and thus made available to the
organization. On the contrary, interactions very often intentionally prevent
communications from being interpreted as organizational decisions. In this
way interactions often distinguish between official communications (which
are presented as decisions) and unofficial communications, which often are
(explicitly) kept from being interpreted as decisions. The organizational
boundary often becomes an explicit topic of interactional communication:
what is to be made available to the organization and what not – this com-
munication is itself, of course, not intended to become an official communi-
cation. The boundary is often even drawn retrospectively; initially official
(decision) communications are retrospectively withdrawn as illegitimate
contributions – this is usual practice, for example, in courts (see Kieserling
1999, p. 373). Apart from interactions “offering” communications as deci-
sions to the organization there is also the possibility that the organization
interprets communications as decisions although they had not been origi-
nally intended as decisions. The interaction can only make “suggestions” to
the organization but cannot determine what communications are observed
and how they are observed.
In any relationship of interpenetration there have to be specific signs that
indicate to the systems (potential) points of coincidence. That is to say, the
“shared” elements have to be marked in some way in order for the respec-
tive other system to be able to recognise them. In the case of the relationship
between psychic and social system this function of coupling is served in the
first place by language (and writing; see Luhmann 1995f, p. 272; Luhmann
160
10 Kap 7. Seidl 05-11-11 16.21 Sida 161
1995j, p. 41). Articulated communication stands out from mere noise and
captures the attention of psychic systems (Luhmann 1995f, pp. 142–143).
And the other way around, if a psychic system wants to “contribute” to the
autopoiesis of a social system it tends to cast its utterance in linguistic form,
although there are, of course, also non-linguistic forms of communication
(Luhmann 1995f, p. 151). Such a marker for “shared” elements can also be
found in the case of the relationship between organization and organiza-
tional interaction. The interaction can highlight certain “culmination
points” (Luhmann 1993h, p. 339) in the flow of interactional communica-
tions in order to signal to the organization possible points of connection.
Such markers could be an explicit declaration of a communication as deci-
sion – “I thus conclude: we have reached the decision to…”. One can some-
times find that there exists a specific format for such “declarations”, as for
example the announcement of a verdict at court or awarding a doctorate at
university. One of the most prominent markers, however, is the record (cf.
Weber 1978); if a communication is put on record this is a strong signal for
the organization that the communication lends itself to being treated as an
organizational decision. Many organizations will only recognise something
as an organizational decision if it is put on record. For the interaction this
means that it can regulate its relation to the organization – or better: the
boundaries of the organization – by distinguishing between communications
“on record” and “off record”.12 What is to be put on record and what is to
be kept off record is often an explicit point of discussion in the interaction
(Boden 1994, p. 85; Kieserling 1999, p. 385). Sometimes specific points can
be erased from the records even retrospectively.
Functions of organizational interactions

Interactions, as we argued above, can make their complexity available to the
organization and in this way contribute to the fulfilment of organizational
functions. In the following we want to discuss three functions in particular:
complexity reduction, deparadoxification and memory.
(1) The possibility that a system may resort to the complexity of another
system – as the concept of interpenetration suggests – has implications for
the system’s way of handling its own complexity. At first, one could say that
this enables the system to increase its own given complexity by the com-
plexity that the other system makes available – in the sense that more ele-
ments and relations between elements are involved in the processing of
information (cf. Luhmann 1980a). On the basis of Ashby’s (1956, pp.
12The distinction “on record/off record” is the second order observation according to which
the interaction observes its relation to the organization.
161
10 Kap 7. Seidl 05-11-11 16.21 Sida 162
202 ff.) “law of requisite variety”13 one could expect organizations with –
for them – turbulent environments to increase their complexity by enlisting
the complexity of organizational interactions. Thus, the more complexity is
“required”, the more organizational interactions would be initiated by the
organization. One might interpret the empirical studies by Burns and Stalker
(1961) and Lawrence and Lorsch (1967) in that respect, if “organic struc-
ture” is understood as the use of interactional communications for coordi-
nating operations. Furthermore, one can find that organizations initiate
interactions when problems arise that cannot be handled on the basis of
organizational decision processes (cf. Schwartzman 1986, Schwartzman
1989; Weick 1995, p. 187; Kirsch 1992, p. 278). One could think, for exam-
ple, of crisis meetings and similar interactions. However, in order to under-
stand the mechanism by which the complexity of the organization is
increased through interpenetration with organizational interactions we have
to look at this issue from a slightly different angle.
On the basis of the notion of complexity as “pressure to select” (Luhmann
1993f; 1995f, pp. 23–28) one can say that the more a system can use the
complexity of another system for reducing its own complexity the less it has
to deal with that task itself.14 That is to say, a system that can instrumen-
talise the complexity of another system for making selections amongst its
possibilities does not need to make the selections itself; or, in other words, it
needs less structure. In this respect, Luhmann writes with regard to specific
cases of interpenetration between social and psychic systems:
[A]n empirically proven hypothesis fits these considerations: social systems

that can enlist more complex psychic systems need less structure. They can
cope with greater instabilities and quicker structural change. They can expose
themselves to chance and thereby relieve their internal regulation. This is
comprehensible if one correctly understands complexity and interpenetration,
namely, as a pressure to select that increases with size and as the ability to
condition this pressure in an open way. (Luhmann 1995f, p. 217; endnote
omitted)
This hypothesis can be applied analogously to the relation between organi-

zation and organizational interaction. The greater the complexity of
organizational interactions that an organization can enlist, the less does it
need decision premises. This line of argument finds support, for example, in
the literature on team coordination in contrast to hierarchical coordination,
which is also inherent in much of the literature on lean management or flat
hierarchies – if we understand teams as a succession of organizational
13 Variety can be understood as the number of possible system states, which is another expres-
sion for the number of elements and possible relations between them.
14 On the necessity of complexity for reducing complexity see Baecker (1999a).
162
10 Kap 7. Seidl 05-11-11 16.21 Sida 163
interactions.15 The more an organization can rely on interactions to “guide”

organizations in their decision making, the less does the organization need
to develop its own decision premises, or the less detailed do they have to be.
For instance, no elaborate rules about communication procedures, no
detailed decision programmes, no detailed personnel planning will be neces-
sary. Apart from that, one can often find that particularly at the apex of an
organization (in business organizations: among the top-management team)
decision processes are only to a very minor degree pre-structured in the form
of formal decision premises. Instead, the organization mostly relies on orga-
nizational interactions for “guiding” the organizational decision process.
On the basis of this second interpretation it becomes clear how the com-
plexity of the organization is increased by organizational interactions. The
complexity of an organization requires the organization to select decision
premises and in this way reduce its complexity. With the selection of struc-
tures the “variety”, i.e. the number of possible relations between elements,
is reduced. Reduction of complexity, however, is necessary in order for the
system to be able to operate – otherwise the system wouldn’t have enough
“internal guidance” to make self-reproduction possible (Luhmann 1995f, p.
283). By relying on organizational interactions to make the selections, the
organization can do without a pre-selection of possible relations in the form
of decision premises and in this sense allow for a greater variety. In other
words, organizational interactions increase the complexity of an organiza-
tion by relieving the organization of having to reduce its complexity. To go
back to our examples above, organizations with “organic structures” pos-
sess a greater variety as the possibilities of relations between elements is less
restricted by (formal) decision premises. Instead, the organization initially
allows for a greater variety of relations, which are limited only in the con-
crete decision situation through interactional processes.
(2) Organizations, more than any other types of social systems, have to
deal with paradoxes. As explained at the beginning of the chapter every
single operation of an organization constitutes an “explicit” paradox. In
order for a system not to get paralysed by its paradoxes the paradoxes have
to be deparadoxified. The main mechanism of deparadoxification of ordi-
nary decisions within the organization is the attribution to a “decision
maker” (Luhmann 2000c, pp. 136; 147). The paradox of decision is in that
way shifted from the operative level to the structural level. Instead of the
“undecidability of the decision” (Von Foerster 1992, p. 14), the structures
entitling the decision maker to make a specific decision come to the fore;
instead of asking about the form of “the alternativity of the decision”
(Luhmann 1993d, p. 289) one is led to ask about the decision premises that
15 To be precise, teams may be defined as a succession of organizational interactions with iden-

tical composition, which are perceived and expected as a unity (cf. Kieserling 1999, p. 221).
163
10 Kap 7. Seidl 05-11-11 16.21 Sida 164
justify the decision – in particular the hierarchical orderings of positions (cf.

Luhmann 2000c, p. 138). While any such question, if consequently investi-
gated, ultimately ends up again at the paradox of decision (i.e. the decision
to put certain structures in place16), usually the organizational complexity
inhibits such investigations – as Luhmann (2000c, p. 139; my translation)
writes: “the paradox is concealed by complexity”. Thus, this basic form of
deparadoxification in most cases is enough to prevent a paralysis.
There are, however, cases where the shift of the decision complexity onto
the level of structures is not enough as a means of deparadoxification; where
the shift does not prevent the exposure of the paradox. This could be the
case where the structures themselves are contradictory (e.g. where there are
unclear decision competencies); thus, instead of the operative paradox the
organization would have to deal with a structural paradox. It could also be
the case where no clear decision premises exist: for example, in connection
with decisions on “fundamental” decision premises (in the context of busi-
ness organization one also speaks of policies or strategies) or during major
turbulences in the organization. Closely related to this is the case where deci-
sions pertain to the very structures that serve as their decision premises.
These are primarily decisions on fundamental organizational change – this
could be a change of fundamental decision premises. In these cases the attri-
bution to decision makers and their respective positions (as a combination
of the three general decision premises) cannot serve as a means of depara-
doxification. In such cases organizations often use interactions as a means
of deparadoxification. Decisions are not attributed to individual decision
makers but to a “deciding” interaction. This could either mean that the
interaction is perceived directly as the “decision maker” (one could think of
a jury at court that announces its (!) verdict) or corporate boards announc-
ing their (!) decision on a new strategy. Alternatively, it could also mean that
one decision maker in an interaction is centrally held responsible for a deci-
sion but as other participants have assented, the decision ultimately is attrib-
uted to the interaction.
The attribution of decisions to interactions can serve as a means of
deparadoxification to the extent that the interactional processes – for the
organization – are incomprehensible. This means that the concrete process-
es leading to the decision making remain obscure – for the organization, not
the interaction. In this sense, the decision paradox is shifted to the
“mystery” (Luhmann 2000c, pp. 135–140) of an interaction: the interaction
has decided the undecidable but from outside one cannot say how (cf.
Schwartzman 1987, pp. 285–286). Usually one cannot even ask the
interaction, in contrast to individual decision makers, as it cannot be
16 And beyond that, the ultimate decision that creates the organization in the first place
– implying the paradox of its existence before its existence.
164
10 Kap 7. Seidl 05-11-11 16.21 Sida 165
addressed as a unity (Kieserling 1999, pp. 383–384) – the participants can,

of course, be asked but every participant might tell a different story.
Furthermore, interactions often actively contribute to the “mystery” by
withholding any information on the concrete processes involved in the deci-
sion making (cf. Kieserling 1999, p. 382; see also Goffman 1959 on
“secrets” of and in interactions in general). Again, one could think, for
example, of juries at court, which disclose at most the number of votes for
or against a particular judgement but no concrete reasons; or of corporate
boards that might disclose the reasons for a decision but keep their state-
ments so general that they do not really say anything about the factual deci-
sion making. Decisions defy in this way any criticism; they are stabilised by
reference to the incomprehensible complexity of the interaction. The only
points for potential criticism are the (starting) conditions of the interaction:
who participated in the interaction, what was the official agenda etc.; in
other words, criticism is redirected toward the decisions and decision
premises involved in “setting up” the interaction. This, again, can be evaded
by drawing the decision premises as much as possible into the interaction
(i.e. participants and topics are chosen from within the interaction) and by
presenting the interaction more or less as spontaneously formed, so that the
official justification for the initiation of the interaction can be provided
retrospectively from within the interaction.
Organizational interactions not only serve as a means of deparadoxifying
and thus stabilising decisions after they have been made, but they can also
serve in a similar capacity in the decision situations themselves. In other
words, interactions can provide an orientation in decision situations where
such guidance cannot be provided by the organization itself (this is just
another perspective on the point made above). Thus, where the organiza-
tional decision premises are ambiguous or conflicting, or where decisions
pertain exactly to the decision premises that could provide guidance, that
interaction can take over the lead; the decision process is pressed ahead by
the dynamic of the interaction. Decision making, in this sense, is less orient-
ed to organizational and more to interactional structures. Retrospectively,
however, the decision premises for the decisions can be – and probably have
to be – sought, reconstructed, reinterpreted or even invented. In other
words, decisions can be presented retrospectively as if they had been guided
by decision premises – decisions are “post-rationalised”. To some extent,
this line of argument runs counter to Kieserling’s hypothesis (1999, pp.
368–369) that, normally, interactions guide the organizational decision
processes, while in crisis situations the decision processes guide the inter-
actions. We would say, rather, that under crisis situations – in particular,
situations that lead to a fundamental ambiguity of the organizational
structures – interactions take the lead. Which system takes the lead under
normal conditions can be left open here.
165
10 Kap 7. Seidl 05-11-11 16.21 Sida 166
(3) “Organizational” interactions can function as a memory in the

organization. With “memory” Luhmann 1997a, pp. 567 ff., 2000c, pp.
192 ff.17 refers to the fact that systems discriminate between forgetting and
remembering: the memory selects what to forget and what to remember.18 In
contrast to traditional conceptualisations, memory is not understood as
“storage of information” (Walsh and Ungson 1991, p. 61) but as a function
that relates present situations to earlier operations of the system. One could
say that it defines the system-related “reality” of the concrete situation. In
this sense, the organizational memory determines in every concrete decision
situation what earlier decisions to relate to. In other words, it remembers
other decision situations and interprets the concrete decision situation in
their context. In order to understand why certain situations or aspects of sit-
uations are remembered and others forgotten one has to understand the con-
ditioning of the organizational memory. Luhmann writes about the organi-
zational memory:
When it comes to organizations the memory is linked to the uncertainty

absorption, which connects decisions with decisions. It forgets the generally
underlying uncertainty, unless it has become part of the decision in the form
of doubts or reservations. But it forgets also the numerous contributing deci-
sions (the invitation to a meeting, the unsuccessful attempts at pushing
through a request); it represses also most of what contributes to the
autopoiesis of the system. By and large […] it only retains what later deci-
sions draw on to form their decision premises. (Luhmann 2000c, p. 193, my
translation, footnote omitted)
In this sense organizations tend to remember only the results of decision

processes and to forget the concrete developments. This radical forgetting
– although a lot of potentially “useful” references to earlier situations get
lost – is necessary for the organization so that its capacity for information
processing does not get clogged up (Luhmann 2000c, p. 192). One could
also say that the memory serves as an important form of complexity reduc-
tion.
An organization cannot, however, dispense completely with all that the
organizational memory forgets. Often it needs to remember, for example,
information that had been gathered in connection with a decision, or the
concrete developments of a decision process, or the concrete arguments for
and against particular decisions by particular members of the organization
17 Luhmann here draws on the neurophysiological and general cybernetic studies of Von
Foerster (1949; 1950; 1969; 1985, pp. 133–172).

18 As Luhmann writes, the main achievement of the memory is forgetting; consequently,
“remembering” has to be conceptualised as the inhibition of forgetting (Luhmann 2000c,

p. 192).
166
10 Kap 7. Seidl 05-11-11 16.21 Sida 167
(and to remember that it has forgotten what it has forgotten). Such refer-
ences might be of particular importance in cases where the organization
experiences inconsistencies between its decision premises or where new deci-
sion situations are difficult to make sense of. Or it might just be that such
references make specific decision situations easier to handle. In such cases
– and it might indeed be so in almost all decision situations – the interaction
can “provide” the organization with an appropriate memory. This is the
case because the memory of (organizational) interactions is conditioned in a
different way. The interactional memory, in contrast to the organizational
memory, is not connected to the organizational uncertainty absorption.
Kieserling writes in this respect:
The memory of the organization remembers only decisions and forgets every-
thing else. In contrast, the memory of the interaction and its participants is
conditioned in a totally different way. They might remember the process
more than the result and the defeated more than the finally victorious candi-
dates. (Kieserling, 1999, pp. 385–386, my translation)
The interactional memory thus seems to be conditioned almost in the oppo-

site way to that of the organization. What the one tends to forget the other
tends to remember. Because of their relationship of interpenetration, the
interaction can make its own memory available to the organization. This
does not mean, of course, that the organizational memory is expanded by
the interactional memory. The interactional memory does not become part
of the organizational memory, in the same way that the interactional com-
plexity does not become part of the organizational complexity, as argued
above. The organization does not have direct access to the memory of the
interaction, but it can expose its decisions to the interactional memory by
having decisions take place within interactions. If a decision is discussed
within organizational interactions, the participants are likely to draw on
past experiences in connection with similar decision situations – for
instance, arguments they may have had previously about similar things.
Although these communications are inconceivable for the organization they
have an effect on the communication that will be ultimately interpreted by
the organization as its own decision and will thus have an effect on the
organizational autopoiesis. Organizations might, in this sense, instrumen-
talise interactions for remembering what they themselves have forgotten.
This memory function of organizational interactions can also be described
as a form of complexity reduction – to be precise: reduction of temporal
complexity. By making their memory available to the organizational repro-
duction, interactions serve a function of complexity reduction for the
organization. Analogously to our argument above, we can say that organi-
zations can use organizational interaction to select relations between the
present decision situation and earlier decision situations of the organization.
167
10 Kap 7. Seidl 05-11-11 16.21 Sida 168
To the extent that the organization can rely on interactions fulfilling this
task in concrete decision situations, interactions do not have to establish
structures that pre-select those relations. Apart from the fact that, due to
their particular mode of reproduction (i.e. uncertainty absorption), organi-
zations are always dependent on interactional memory (without that they
would not be able to reproduce themselves), there are different degrees to
which the interactional memory is drawn on. Analogously to our argument
above, we can say that the more an organization can rely on interactional
memory, the less elaborate are its own memory structures (e.g. written doc-
umentation etc.) and vice versa.
Possibilities for conditioning organizational

interactions
So far we have discussed functions that organizational interactions might
serve for “their” organization. We have not, however, asked how interac-
tions can be brought to serve these functions. It probably would not be
enough for organizations to rely on interactions to form spontaneously and
by chance at the “right” moment when they are “needed”. Rather, although
interactions are operatively closed vis-à-vis organizations, they can be never-
theless conditioned (cf. Luhmann 2000c, p. 373). That is to say, organiza-
tions can create certain context factors that make it likely for interactions to
reproduce themselves in ways that are of benefit to the organization. This
conditioning can be temporal, social and factual.
Conditioning in the time dimension is particularly focused on beginning
and ending an interaction. The organization can decide to start an interac-
tion and it can also decide on a certain time limit, after which the interac-
tion is to come to an end. For example, a meeting can be scheduled for a cer-
tain time – from three to four o’clock. The organization could also decide
on a sequence of interactions – for instance, every Monday from three to
four o’clock (cf. Luhmann 1997a, p. 818). However, as the organization has
no direct access to the interactional operations, it cannot start or terminate
them directly;19 only the interaction itself can produce its own beginning and
its own end. Yet, organizations can condition participation and thus indi-
rectly “stimulate” interactions to produce a beginning and an end. That is
to say, the organization can decide on a certain time span during which
members are to come together. The beginning of the interaction can be
conditioned insofar as the organization can arrange for members to come in
19Unless the organization decided, for example, to blow up the interaction with a bomb, in
which case the interaction would not be terminated by the interaction itself, but by the elimi-
nation of its (necessary) environment of psychic systems (cf. Luhmann 1995j, p. 45).
168
10 Kap 7. Seidl 05-11-27 21.57 Sida 169
face-to-face contact on the basis of which an interaction is likely to form.20

The end can be conditioned through the decision to withdraw the partici-
pants after the time has run out. The interaction can observe this prospect
of its participants being withdrawn after a certain time, and in expectation
of that produce its own end. When the time has run out, the participants of
the interaction say “good bye” to each other.
In the social dimension the organization can condition the interaction by
deciding on its participants. That is to say, the organization can influence the
interactional communications by deciding who is to participate and thus,
what perspectives become available in the interaction (cf. Kieserling 1999,
p. 378). Especially important in this context are the perspectives given
through the particular organizational roles of the participants. For example,
if the organization decides that senior managers are to participate, the inter-
action is likely to be different from a meeting between shop-floor workers
only.
Again, by deciding on the composition of the interaction the organization
can condition it, but it does not have any direct control over it. It is ulti-
mately left to the interaction to determine who is treated as present and who
as absent. In this sense, even people physically present might be considered
absent by the interaction. Similarly, the formal organizational roles of the
participants might not be referred to by the interactional communications.
Conditioning in the fact dimension concerns the organization’s influence
on the selection of topics. An explicit conditioning in this dimension would
be a decision on the agenda of a forthcoming interaction. The selection of
topics can also be indirectly influenced through decisions on the partici-
pants. For example, if many marketing managers are involved in the inter-
action, the communication is likely to focus on marketing issues. Another
way of conditioning the selection of topics is to set in advance certain out-
put categories for the interaction; state, for example, that certain decision
have to be made. An important form of conditioning concerns the con-
straining of the range of possible topics. Instead of influencing the interac-
tion in a way that makes the selection of a specific topic more likely, the
organization can also make the selection of certain topics more unlikely; for
example, by withholding information.
Although we have treated the different dimensions of conditioning sepa-
rately, here they are to a large extent interdependent. For example, by start-
ing an interaction for discussing recent sales figures, the organization might
condition the interaction in all three dimensions. In the object dimension the
interaction is conditioned because the topic of the communications is
preselected: the sales figures. In the social dimension the interaction is con-
20 In this respect, Kirsch 1992, pp. 271 ff.), speaks of the “creation of necessary and sufficient
initial and contextual conditions” for the emergence of interactions.
169
10 Kap 7. Seidl 05-11-11 16.21 Sida 170
ditioned because particular persons might have to participate in the interac-

tion – e.g. marketing managers. In the time dimension the interaction is con-
ditioned as a certain timing for the interaction is decided.
Conclusion
In this chapter we have suggested a specific conceptualisation of “organiza-
tional” interactions, based on Luhmann’s social systems theory. The basic
idea was to treat organization and interaction as two different types of sys-
tems that remain operatively closed with regard to each other. In other
words, “organizational” interactions take place in the environment of the
organization. Although this strict distinction runs counter to our intuitions,
which would see interactional communications as part of the organization,
it has important theoretical advantages. Firstly, it is a perspective that allows
perceiving (organizational) interactions whereas other theoretical perspec-
tives would not even have concepts for referring to them – which is proba-
bly one of the reasons why interactions in organizations so far have received
relatively little attention by social scientists (Schwarzman, 1986). Secondly,
the strict separation between organization and interaction allows analysing
organizational and interactional phenomena in their own right. This must
not be misinterpreted as disregard of the role of interactions in organiza-
tions. On the contrary – and this is our third point – through the differenti-
ation between organization and (organizational) interaction it can be clear-
ly shown that, and in what way, both systems depend on each other. This
conceptualisation forces the researcher to be explicit on the achievements of
each system and on the mechanisms through which they become available
to each other. In this sense this conceptualisation can also help discipline the
researcher.
170
11 Kap 8. Drepper 05-11-11 16.22 Sida 171
Chapter 8
Organization and Society.

On the desideratum of a society theory of
organizations in the work of Niklas Luhmann1
Thomas Drepper
Introduction and problem exposition

The interrelation of modern organizations and modern society has been a
major topic of interdisciplinary organization theory in recent years. Bringing
Society Back In (Friedland/Alford 1991) means bringing theory back in and
especially bringing theory of society and social theory back in. Hannan and
Freeman point out the missing connection between organizational theory
and sociological theory and stress the relevance of overcoming “the partial
intellectual isolation of the field of organizational sociology from the rest of
the discipline” (Hannan/Freeman 1989, p. 11) and of broadening the limit-
ed intra-organizational perspective. The following approaches are contribu-
tions to this theme:
• the New Institutionalism in organization theory (Scott 1995, Powell/Di-
Maggio 1991)
• organization theory applying the structuration theory of Anthony
Giddens
• the theory of corporate actors (Coleman 1982, 1990)
• some aspects of the population-ecology approach (e.g. Stinchcombe 1965,
1991)
All these approaches converge to some degree in the challenge of defining

the societal embedding of organizations without referring to a diffuse notion
of environment as a residual category. On the other hand, it seems that
theories of society do not spend much energy on conceptualizing the factor
termed “organization” in a systematic manner. To quote Hannan and Free-
man once again: “Although organizational processes figure prominently in
social change, most macrosociological theory and research make little
1The author would like to thank Kai Helge Becker and David Seidl for their helpful comments
and great support editing this text.
171
11 Kap 8. Drepper 05-11-11 16.22 Sida 172
reference to systematic research on organizations” (Hannan/Freeman 1989,

p. 10). How could this mismatch be interpreted?
Two dominant paradigms are standard references in the sociological diag-
nosis of modern society but stand rather unlinked beside each other. The
first consists in the concept of modern society as a highly differentiated unit
consisting of several societal sectors, fields, subsystems, value spheres or life
worlds with special patterns of meaning, normative and cognitive rules and
role expectations. This paradigm refers to the differentiation theory in
macro-sociology (as discussed by Durkheim, Simmel, Parsons, Eisenstadt,
Weber, and Luhmann). The second type of diagnosis focuses on the enor-
mous pervasion of modern society by formal organizations that characterize
the face of The Asymmetric Society (Coleman 1986) and dominate the life
courses of modern individuals. From this point of view the Society of
Organizations (Perrow 1991) is an organizational society and the modern
individual is The Organization Man (W. H. Whyte 1956). Both the argu-
ment on differentiation theory, and the argument on organizational society
in the tradition of Max Weber’s thesis on bureaucracy are part of sociolog-
ical mainstream. Taking these recent theories of modern society into
account, it becomes clear that organization theories examining the societal
embedding and the non-organizational preconditions of organizational per-
formance beg the question of whether a linkage between these basic analy-
ses is possible.
In the concert of social theories that offer the theoretical equipment for
systemizing the reciprocal relationship between modern society and modern
organizations, the sociological system theory of Niklas Luhmann seems to
be a good candidate for tackling this question. Three major parts of
Luhmann’s theory are most relevant for dealing with the problem of corre-
lating society and organization: on the level of a general social theory, the
theory of social systems introduces the distinction between interaction,
organization and society (Luhmann 1982c) as well as the late concepts of
autopoiesis and communication (Luhmann 1990a, 1990j, 1990f, 1995). The
general theory of organized social systems connects the sociological tradi-
tion of Talcott Parsons with decision theories that go back to the works of
Simon and March. Finally, the theory of society distinguishes various forms
of differentiation (segmentation, centre/periphery, stratification, functional
differentiation) as the primary structures of societies, and of society in the
process of sociocultural evolution (Luhmann 1977a, 1982d, 1992c).
Nevertheless, Luhmann’s efforts towards a systematization of the stated
topic are few, despite all those elaborated theoretical elements. Luhmann
addresses this problem explicitly only in some of his last publications. These
works deal with the connection between the internal differentiation of func-
tional subsystems on subordinated levels (segmentation, centre/periphery,
stratification) and the role of organizations within this process. In this
172
11 Kap 8. Drepper 05-11-11 16.22 Sida 173
context, Luhmann discusses not only new theoretical concepts (structural

coupling, external communicability, medium/form, centre/periphery), but
also rather classic notions (e.g. meaning, complexity, generalized media,
breaks of interdependency, congruent generalization of expectations).
Luhmann’s theoretical work is very promising with respect to the analysis
of the relationship between organizations and society, however, Luhmann
himself did not elaborate this issue much further. This chapter contributes to
filling precisely this gap and aims at discussing and bundling up the various
contributions that Luhmann made to this topic in his long academic career.
Furthermore, it will attempt to show that it is necessary to combine older
theoretical concepts with newer ones in order to elaborate some basic ele-
ments of a societal theory of organizations (see Drepper 2003).
The chapter is structured around seven topics that can be found within
Luhmann’s work and are central to a better understanding of the relation-
ship between organizations and society. The first section presents
Luhmann’s critique of the classical organization theory and the differentia-
tion concept of the ontological system model. Here, the dominance of the
means/end-scheme and the rationality-concept, as well as the system model
of a-whole-and-its-parts are examined critically. On the basis of that cri-
tique, Luhmann proposes a twofold theory of differentiation that combines
a theory of different types of social systems (interaction, organization, and
society) with a theory of structural differentiation of complex systems. The
second section puts forward the argument that this kind of twofold differ-
entiation theory serves as the theoretical background for interrelating orga-
nization and society. This line of argumentation can also be found in some
of Luhmann’s early writings, in which he formulates the theory of formal
organizations against the background of a functional-structural theory of
society. In the third section, an important topic of Luhmann’s societal theo-
ry will be discussed: the repraesentatio identitatis phenomenon of modern
society. The repraesentatio identitatis phenomenon refers to the impossibil-
ity of representing modern society, which has neither “centre” nor “top”,
from a superior standpoint within itself that coincides with this society’s
leading self-description. For that reason there are several possibilities for
describing and characterizing modern society, for example from a political,
scientific, pedagogical, technical, economical or juridical viewpoint. And
these descriptions and interpretations are so controversial that it is nearly
impossible to draw a consistent picture of the structure, processes, problems
and main issues pertaining to society. Every social sector claims to be most
relevant. Luhmann argues that organizations are able to compensate the
uncertainty that arises on the societal level, partially because they are based
on the modus of decision making, which absorbs uncertainty. But this uncer-
tainty absorption is of course limited to the level of organization systems
and cannot undo the authority deficit and uncertainty that exists on the soci-
173
11 Kap 8. Drepper 05-11-11 16.22 Sida 174
etal level. Further on in this section this argumentation will be linked to the
topic of corporate or collective actors. This section also includes some
methodological remarks on the phenomenon of collective or representation-
al action and the attribution of organizations as collective or corporate
actors (Dan-Cohen 1986). Some theoretical and methodological differences
between Luhmann and the action theory in the tradition of Max Weber and
Talcott Parsons will also be discussed in the same section.
The fourth section deals with Luhmann’s attempt to define the function of
organizations in relation to the various differentiation forms of modern soci-
ety, segmentation, centre/periphery stratification, and functional differentia-
tion. This conception presents his very first approach to work on the rela-
tionship of organization and society consistently. The theoretical concept of
interdependency break is the main theme in the fifth section. This very
abstract notion forms part of the cybernetics and systems-theory tradition
and marks another attempt to define the societal function of organizations.
Luhmann conceptualizes organizations as breaks of interdependencies with-
in broader systems; more particularly societal function systems. Structural
coupling is the topic of the fifth section. This notion goes back to Matu-
rana’s and Varela’s theory of living systems and refers to the structural
embedding of autopoietic systems within their environments. Even auto-
nomous systems are always environmentally embedded. Luhmann follows
this concept, uses it for supplementing the better-known term of interpene-
tration (Luhmann 1995f) and introduces it into his differentiation theory.
He analyses the structural couplings that link the autonomous function-sys-
tems of society, and – crucially in the context of this chapter – examines the
function of organizations with relation to the structural coupling of func-
tion-systems. This discussion will be elaborated in the sixth section.
All in all, this chapter tries to give a comprehensive overview of the pos-
sibilities of clarifying the relationship between organization and society from
the standpoint of Luhmann’s theory. To this end, various parts of Luhmann’s
oeuvre will be brought together and several aspects of the relationship
between these two types of social systems will be discussed.
Critique of the classical organization theory and

the differentiation concept of the ontological
system model
The critique of the classic theory of organizations and the concept of
instrumental rationality, or means/ends rationality of organizational func-
tioning, are important topics of Luhmann’s early organization theory (Luh-
mann 1968). Ever since his early writings, Luhmann has always been criti-
cal of the “overestimation of the means/ends orientation as a structural law
174
11 Kap 8. Drepper 05-11-11 16.22 Sida 175
of social systems” (1982, p. 25) and its idée directrice of organizations as

systems “that have the capacity for making rational decisions”, “pursue
objectives and are in control of their instruments” (Luhmann 1996a,
p. 341). In the classic model, organizations are treated as rationally per-
forming instruments of goal-attainment that seem to be a necessary feature
of the modern division of labour. Luhmann is critical of this concept because
it proposes the increase of societal rationality via organizational perform-
ance.
This stage in Luhmann’s work is characterised by two main theoretical
concepts: the displacement of the means/ends schema, as the dominant
model for describing organizations as a rational mechanism (Luhmann
1982b), and the turning away from the system model of a “whole-and-its-
parts” towards the system model of “systems and their environments”
(Luhmann 1995f, pp. 1–12, pp. 176–209). Luhmann combines these cri-
tiques and argues that the “whole-and-its-parts” system model is merged
with the concept of means/ends-rationality, and that this “joint venture”
promotes an ontologically inspired container metaphor of organizations:
“Organizations were seen as systems. A system was understood as an order-
ing of relationships through which parts are bound into a whole. And the
relations of the whole to its parts were interpreted through the means/ends
schema” (Luhmann 1982b, p. 25). Luhmann argues that the system model
of the whole-and-its-parts ignores the structural relevance of environments
to systems and creates a static picture of systems and their structures. He
rejects the idea of subsystem-formation as a process of decomposing a
broader system into smaller parts. Subsystem-formation is not “simply a
decomposition into smaller chunks“ (Luhmann 1977b, p. 31), but a multi-
plication of system/environment boundaries within an existing system/envi-
ronment-boundary. System differentiation is the reflexive form of system
building that is based on the reduplication of system/environment bound-
aries within systems. This process increases the internal complexity of a sys-
tem (Luhmann 1995f, pp. 176 ff.). Moreover, Luhmann argues that the
means/ends schema should not dominate the understanding of organiza-
tions, because goal- or end-orientation is only one possible structure of sys-
tem rationalization besides other possible functional structures in an organi-
zation – for example, the form of conditional programming as a decision-
premise (Luhmann 1982c).
On the basis of the above argument, Luhmann proposes a twofold theo-
ry of differentiation. On the one hand, we find the differentiation of the
various types of social systems: interaction, organization, and society (Luh-
mann 1982c). And on the other hand, we find the internal differentiation of
subsystems as a mechanism of reducing complexity within complex systems.
The following section will discuss both differentiation theories in Luhmann’s
approach and demonstrate how they interconnect. Combining these two
175
11 Kap 8. Drepper 05-11-11 16.22 Sida 176
theories is the crucial first step for conceptualizing the relationship of orga-
nization and society.
Types of social systems and the logic of internal

system differentiation
According to Luhmann’s concepts, interaction systems are social systems
that are based on the reciprocal perception of the physical presence of at
least two participants in a social situation. Interactions are structured
around topics of communications to which the participants contribute alter-
nately (for details see Chapter 7). Organizations are member-based, deci-
sion-processing communication systems that are structured around decision-
premises (programmes, personnel, and communication paths). In contrast to
interaction systems, organizations transcend the mode of successive alter-
nate contributions in interaction systems: a structural feature of many com-
munications is that although several communications with different subjects
can take place at the same time, it is not necessary that all organizational
members are present when a decision is made. For that reason, organiza-
tions are able to process more complexity than interaction systems, which
explains why Luhmann refers to the latter as “simple systems” in his early
writings.
Society is the communication system that encompasses all communica-
tions and includes all other forms of social systems. The different types of
social systems are nested within social reality. Every interaction and every
organization belongs to society and, furthermore, an interaction can, but
does not need to, belong to an organization. This context could be seen as
indicative of the differentiation of subsystems within broader systems. The
inclusive system offers structural premises and a pre-selected order to the
system it encompasses. In other words, it provides “specific structural prem-
ises on the basis of which a self-selective process can occur while limiting
this process in its possibilities” (Luhmann 1982c, p. 86). In doing so, it
imposes “order on the proximate environment of the subsystem. In this
double effect of incorporating on incorporated systems lies the condition for
the freedom of system developments” (Luhmann 1982c, p. 86). The func-
tion of society as the system that encompasses all other social systems can
be seen in enlarging and reducing “the complexity of external and internal
environments to the effect that other systems will find enough structure to
support boundaries and structures of higher selectivity. The process contin-
ues at the level of subsystems, repeating the same mechanism, and it arrives
at organizations and interactions of high specificity” (Luhmann 1977a,
p. 31).
According to this line of reasoning, organizations have to be conceptu-
176
11 Kap 8. Drepper 05-11-11 16.22 Sida 177
alised as substructures of broader systems. They cannot be understood as

goal-attaining agents or instruments of rational domination, in the way
Talcott Parsons understood them, for instance. Parsons defines the develop-
ment of organizations as “the principal mechanism by which, in a highly
differentiated society, it is possible to get things done, to achieve goals
beyond the reach of the individual and under conditions which provide a
relative maximization of effectiveness, in Chester Barnard’s sense” (Parsons
1960, p. 41). For Parsons, organizations are – as a type of collectivity – “the
essential operative units of social system”, the performing centres of goal-
attainment through which the universalistic values and norms of a society
become specific and concrete orientations of action. In this view organiza-
tions are the transformers of universalistic and normative patterns.
Luhmann does not follow this concept of collectivity, because it is based
on the methodological idea of social systems as a collective or corporative
aggregation consisting either of individuals or actions, which become cor-
poratively, consensually or territorially integrated (see Drepper 2003,
pp. 254–284). According to Luhmann – in contradiction to Parsons’
approach – organizations do not serve higher ends or interests, as they are
not the centres of societal rationality. Organizations are not simply depend-
ent subsystems of broader subsystems that fulfil the tasks commanded by a
higher level (Luhmann 1982c): they “only” use the possibility of building
boundaries in a self-selective way. Following this presupposition, organiza-
tions cannot be understood as performing agents of functional values,
norms, or role scripts. Organizations have to be conceptualized as self-
referential and self-determined systems with a specific logic of meaningful
reproduction (Luhmann 1990a and 1995f pp. 437–477). And this logic
cannot be understood in terms of a cause-and-effect relation that imposes
the expectation-patterns of a higher system level on a lower one in a unilat-
eral logic of hierarchical control. Organizations are too uncontrollable for
such a downward causation. On those grounds, Luhmann dismisses
Parsons’ LIGA-logic of structural components and subsystem nesting; he
dismisses, that is, the way Parsons conceptualizes the function of organiza-
tions within broader systems (Drepper 2003, pp. 262–272).
According to this theoretical conception, neither the relation between
society and organization, nor the relation between function-systems and
organization, must necessarily be thought of in a hierarchical way. The
means/end-schema, in concert with the ontological system model of a whole-
and-its-parts, can explain neither the mutual conditioning of society and its
function-systems nor the correlation between functional subsystems and
organizations. Function-systems are not the means for higher societal ends
or values, and organizations are not the means for attaining functional ends
or goals. Function-systems are not equipped with organizations which
enforce interests or realize ends in a top-down logic. Although organizations
177
11 Kap 8. Drepper 05-11-11 16.22 Sida 178
have been introduced into almost all functional domains, the central func-
tions of modern society cannot be realised as membership-based social
systems. Functionally differentiated societies “cannot be ruled by leading
parts (elites) as stratified societies could, to some extent, and […] cannot be
rationalized by chains of means and ends as a technocratic conception
would suggest. Their structural complexity can be formulated only by mod-
els that take into account several system/environment references at once”
(Luhmann 1977a, p. 37).
Organizations respecify the complexity of function-systems. This
“requires a displacement of problems from the level of the society to the
level of subsystems. This is not simply a process of delegation or decentral-
ization of responsibilities and not simply a factoring out of means for the
ends of society” (Luhmann 1977a, p. 39). Luhmann gives the following
examples for this theoretical argument: “Given our economic system, for
instance, production and consumption choices could never be yoked to-
gether into an organizationally unified decision-making process […]. Like-
wise the functions of education will remain divided up between the school
system and the family, although the emphasis may shift” (Luhmann 1982c,
p. 81). This leads to the conclusion “that societal function cannot be inte-
grally delegated to a single or unified organization. Instead, such functions
must be parcelized and respecified before being assigned to specific organi-
zations. […] The function of law is certainly not a norm. The function of
politics is not a standard of legitimacy. The ‘limits to economic growth’ is a
possible conference topic, but it cannot help entrepreneurs and enterprises
decide what they ought to do” (Luhmann 1982c, pp. 81–82).
Neither modern society as a whole nor the individual function-systems
can be realized through organizations. Luhmann states a very simple but
precise structural condition for this fact: the “motivational mechanism of an
organization presupposes voluntary and contingent membership – a realis-
tic chance for entrance and exit” (Luhmann 1982c, p. 77). In contrast to
this, modern society is based on the principle of total inclusion of almost
everyone in the functional domains: for example, everyone must participate
in economic activities and be educated, and everyone is influenced by polit-
ical decisions and is subject to the law. In other words, while the function-
systems of modern society are based on the principle of inclusion of the
whole population, modern organizations are grounded in the principle of
exclusion and the exchangeability of persons. The members of modern
organizations are recruited and selected by decision, and membership roles
embrace only a fraction of possible individual behaviour: “only on this con-
dition can the decision-making process of organizations be regulated in their
smallest details. The structurally enforced inclusion of the whole population
in the function-systems is therefore precluded by an equally indispensable
exclusion of almost everyone from all concrete organizations. […] Whereas
178
11 Kap 8. Drepper 05-11-11 16.22 Sida 179
the corporations of the old world were institutions complementing the strat-
ified household’s structures, modern organizations are parts of the function-
systems, but with the inclusion/exclusion relationship inverted” (Luhmann
1996b, p. 67). This argumentation is part of what Luhmann means when he
talks about the simultaneity of structural dependency and independency of
different social systems, and even of different types of social systems. And
this idea of the simultaneity of two seemingly contradictory processes – the
autonomy and discreteness of social systems and the structural dependency
as a result of environmental embedding – can already be found in some of
Luhmann’s early writings. Even Luhmann’s early concept of formal organi-
zation goes beyond the narrow analysis of internal organizational structures
to include the link of organization theory, theory of society and social system
theory. However, it is important to point out that he does not connect the
organization theory with the insights of a differentiation theory of society in
a particularly systematic manner. In his theory of formal organizations,
Luhmann draws on the functional-structural theory of society and merges a
functionalistic view with institution theory (Luhmann 1964, 1965, 1968).
For the early Luhmann one can state: Arnold Gehlen meets Talcott
Parsons! From this theoretical point of view, organizations appear as civiliz-
ing mechanisms, which formalize and stabilize improbable expectations. In
this early approach Luhmann conceptualized the organization as a formal
structure of social systems that “free the system from a concern with the
motivational structure of its members, and thus with all its own personal
and social conditions. Generalization and separation of roles are the social
mechanisms that allow bureaucratic systems to focus one-sidedly on their
demarcation from non-members” (1982b, p. 45). Formally organized social
systems are clear-cut social orders producing congruent expectations in the
social, temporal and material dimension of meaning (Luhmann 1990e).
Formal organizations are those social systems in complex social orders
which institutionalize continuing motivation.
A societal relevance of formal organizations arises from the separation of
motives and ends, and from the generalization of motives through member-
ship. Impersonal membership-rules make it possible “to stabilize highly arti-
ficial modes of behaviour over a long stretch of time” (Luhmann 1982c,
p. 75). This relevance refers to the course of sociocultural evolution, because
it “is beyond question, in any case, that only organizations are capable of
producing the motivational generalization and the behavioural specification
required in several of modern society’s most important functional domains”
(Luhmann 1982c, p. 76). At the same time, organization building is depend-
ent on specific societal conditioning structures that preselect the communi-
cational possibilities and make expectations probable, insofar that organi-
zations can attach the programming of specialized expectations which then
function as decision-premises (decision programmes, personnel, reporting
179
11 Kap 8. Drepper 05-11-11 16.22 Sida 180
lines). Those societal structures that are preconditions for the building of
organizations are the symbolic generalized media of meaning (Luhmann
1976b), the various forms of differentiation (segmentation, centre/periphery,
stratification, functional differentiation) and the form of modern citizenship
(Marshall 1950). These ideas will be examined more concretely later on in
this chapter. In modern society, the functional differentiation of meaning-
systems preselects and pre-patterns the specific spheres of functionally-
oriented, meaningful reproduction in such a way that the formation of orga-
nizations can be linked to the reduced complexity of the predefined func-
tional domains. Organizations are then able to define specific economical,
political, religious, educational, scientific ends and goals that refer to their
dominant societal environment.
The repraesentatio identitatis phenomenon of

modern society – Organizations as communicators,
communicative addresses and communicable
systems
The repraesentatio identitatis phenomenon is a concept that gives a good
example of the combination of organization theory with the theory of socie-
ty in Luhmann’s model. His thesis is that organizations are able to compen-
sate partially for the uncertainty and complexity of the societal level by
means of their specific modus of operation, i.e. their decision-making. In this
case, Luhmann borrows the concept of absorbing uncertainty from the
organization and decision theory and applies it to the theory of society. The
functionally differentiated society is characterized by a heterogeneous and
decentralized structure that makes it impossible to represent and describe it
in its entirety from an exclusive standpoint within society. The change from
the stratified and feudalistic societies, which were represented by elites at the
top or the centre, to the modern polycontextural society is what causes the
problem of representing society within society. Modern society cannot be
described or legitimized by a major self-description (Luhmann 1982h,
1990h). According to Luhmann, however, organizations are capable of fill-
ing this gap to a certain extent. By means of their structure they can partially
compensate for this deficit in authority and representation that is typical of
modern society. Organizational decisions allow and demand the attribution
of communications to visible roles and persons who act as decision-makers
and are embedded in manifest authority and hierarchy structures. Luhmann
posits that organizations function in the communication processes of mod-
ern society as relevant attribution points and addresses for uncertainty
absorption. But this uncertainty absorption is strictly temporal and limited
180
11 Kap 8. Drepper 05-11-11 16.22 Sida 181
to the level of organizational operations. Decisions are events and, as such,

they are transitory in a radical way. On the operational level of basic self-
reference they have to be reproduced permanently if the organization system
wants to “survive”. And on the structural level all organization structures
(programmes, persons and communication paths) are decision-premises and
as such can be changed by further decisions (see Luhmann 2000c and
Luhmann 1995f, p. 278 ff.). This implies that organizations cannot guaran-
tee stability on the societal level. Unlike society, which is subject to socio-
cultural evolution, organizations can be shaped and altered by decisions.
The topic of organizations as communication addresses refers to the dis-
cussion about corporate actors as action units. Luhmann adopts some
important arguments of evolution theory and differentiation theory, but dis-
misses fundamental methodological presuppositions of action theory in the
tradition of Max Weber and Talcott Parsons. The concept of an organiza-
tion’s ability to act collectively goes back to Weber’s ideal type of bureau-
cracy and Parsons’ concept of evolutionary universals – if not to earlier the-
ories – and here in particular, to the idea of official and representational
action “in the name of” a collectivity. Talcott Parsons connects the capacity
to act officially or “in the name of” with the concept of power, that is, “the
capacity of a unit in the social system, collective or individual, to establish
or activate commitments to performance that contributes to, or is in the
interest of, attainment of the goals of a collectivity” (Parsons 1964, p. 347).
Luhmann does not adopt this concept of collectivity, because it is based
on the idea of social systems as aggregates and containers consisting of con-
crete human beings or of their actions. Individuals are not the basic elements
of any social system (Luhmann 1990a, 1995f). Following this methodolog-
ical presupposition Luhmann rejects the argument that the capacity for col-
lective action is mainly needed for the internal coordination of a system, and
for producing binding forces and commitment. Instead, the capacity for col-
lective action has to be understood as a boundary function of complex sys-
tems with respect to their environment. From an evolutionary point of view
systems that are able to constitute and institutionalize the capacity for action
“in the name of” benefit from an increase in complexity. As Luhmann makes
clear, “the problem of specifying environmental contacts as constraints and
extensions of the system’s general location in the environment […] is a cen-
tral problem of all complex systems, a kind of threshold in the evolution of
greater complexity. On the level of social systems, this problem concentrates
on the capacity for collective action and the subsequent arrangements neces-
sary for it […]. The organization of the collective capacity for action must
be viewed as one of the most important early evolutionary achievements of
social systems, because it can improve decisively the external relationship of
these systems by internal restrictions” (Luhmann 1995f, pp. 198 ff.).
In his late works, after his so-called “autopoietic turn”, Luhmann replaces
181
11 Kap 8. Drepper 05-11-11 16.22 Sida 182
the term “action” with the term “communication” and changes his basic
theoretical concepts accordingly (Luhmann 1995f, p. 137). His central per-
spective on the collective “nature” of organizations, however, does not
change. In fact, from the point of view of his late theoretical framework,
Luhmann characterizes organizations as the only social systems that are
structurally equipped for external communication, that is, for crossing their
boundaries by communication. Neither the heterogeneous modern society as
a unit nor interactions as simple systems have the capacity to communicate
with their environments, so they cannot be addressed as collective actors.
Organizations are the only social systems in modern society that can be
addressed in communication processes as collective actors. This feature
makes organizations comparable to persons as authors and addresses of
communication. Persons are expectation-structures of and for communica-
tion. In Luhmann’s understanding they are not complete human beings in
the biological sense. They belong – in an order of various levels of general-
ization and identification – to the order of structural components of social
systems: person, role, programmes (norms and goals), and values (Luhmann
1995f, pp. 278 ff.; 1996b): “These different levels of identification can be
ordered on a dimension from abstract to concrete” (Luhmann 1977a, p. 46).
A person can be understood in its traditional sense “as a mask put on for
purposes of communication. A person is a speaker and an address required
for the reproduction of communication” (Luhmann 1996a, p. 343). The ref-
erence points for personalization vary historically, because what “can be
symbolized as a person depends on the societal system of communication
and its autopoiesis and not on specific ontological properties of its environ-
ment” (Luhmann 1996a, p. 344). From this point of view, the fact that
organizations are commonly referred to as “collective” or “corporate”
actors is an evolutionary effect of the development of the modern Western
legal tradition (Berman 1983). In the legal sense organizations are juridical
persons or corporate bodies (Dan-Cohen 1986). However, from the view-
point of sociocultural evolution this attribution routine is an “evolutionary
improbability”. Hierarchy is a feature of particular importance to external
communication and representation, because it allows attributing decisions
to programmes, roles and persons (deciders). And in modern society, hierar-
chy is predominantly institutionalized on the level of organizations.
It is a structural feature of modern society that it is flooded by represen-
tational communication. This always refers to organization systems and
their representational roles, but never to the function-systems or to the
entire society. What we hear every day are the voices of spokespersons who
hold key positions in organizations and whose boundary roles are heavily
promoted by the modern system of the mass media: representatives of mod-
ern states, representatives of trade and labor unions, representatives of med-
ical unions and so on. The process of becoming a communication address
182
11 Kap 8. Drepper 05-11-11 16.22 Sida 183
and of developing professional and organizational structures and standards

can often be noticed in cases of social movements aiming at political goals.
The anti-globalization movement ATTAC is a recent example for such a
development, as is the German party “Die Grünen” (The Green Party),
which started as an alternative, ecological movement in the seventies, and
was later on structurally transformed into a regular political party. The rel-
evant functional and organizational environment (states, parties, NGOs,
and corporations) “provokes the system to unified action. Then the question
is whether adequate preconditions for that action exist or whether they can
be developed quickly enough” (Luhmann 1995f, p. 199). The social move-
ment is forced to become a visible body in the political communication
process, if it wants to be accepted as a serious address in the world-wide sys-
tem of political communication. This suggests that adopting an organiza-
tional structure seems to be the common modern way of becoming a corpo-
rate address for communication.
Subordinate forms of differentiation and the role

of organizations
In this section the societal function of organizations will be discussed in rela-
tion to the various differentiation forms of modern society. This argumenta-
tion presents Luhmann’s first thorough attempt to conceptualize the rela-
tionship of organization and society on the basis of his advanced theory of
societal differentiation.
Primary forms of societal differentiation presuppose the conditions of all
further differentiations, in that they define their conditions and limitations
within a differentiated system. Whereas archaic societies selected communi-
cations under the conditions of equal segmentation, and aristocratic soci-
eties did so via the conditions of unequal strata, functional differentiation is
based on the selection of communications along specific functions (econo-
my, law, education, arts, politics, religion, sports etc). In this latter setting,
communications, contribute to the reproduction of those functional
domains. Structurally, it is always possible that primary differentiation
forms also include other differentiation forms. Temporal stratification and
rudimentary role-differentiation took place even within segmentary soci-
eties, and bureaucratic structures fulfilled important functions in the centres
of early high cultures and political empires (Eisenstadt 1963). Nevertheless,
such forms of differentiation should be regarded as subordinated forms,
which now and then appeared as pre-adaptive advances of further steps of
sociocultural evolution (Luhmann 1977a). This structural logic can also be
shown in the functional differentiation of modern society. Functional differ-
entiation as a primary structure is also compatible with all other forms of
183
11 Kap 8. Drepper 05-11-11 16.22 Sida 184
differentiation: “Societies are not able to dream up arbitrarily new forms of

differentiation” (ibid., p. 33), but evolve by decomposing and reconstruct-
ing already existing structures. This means that even in the functionally dif-
ferentiated society – though no longer on the primary level – there exist
segmentary, polarised (centre/periphery), as well as stratified differentia-
tions. The variety of internal differentiation within the overall setting of
functional differentiation counterbalances the complexity of modern socie-
ty.
As already mentioned, societal functions as such cannot be organized.
This, however, does not mean that organizations may not have a function
on subordinated levels of differentiation. In fact, Luhmann looks at the rela-
tionship between the internal differentiation of functional subsystems on
subordinate differentiation levels (segmentation, centre/periphery, and strat-
ification) and organizations within this internal differentiation. Organiza-
tions play an important role in the centre and in the periphery of some func-
tion-systems. Influential and important function-systems such as the econo-
my, the law and the political system display this pattern of differentiation.
The political system, for example, is segmented on its secondary level into
nation states that are in their turn differentiated between centre and peri-
phery on the third level of differentiation. The centre has a stratified struc-
ture that is dominated by the political administration, while the periphery
consists of the loose coupling between parties, interest unions, and NGOs.
This means that organizations are the dominant type of system that realizes
and stabilizes the level of centre/periphery-differentiation (Luhmann 1990i).
Put in cybernetic terms, the combination of multiple forms of differentiation
enables particular function-systems to combine a high degree of necessary
redundancy and unity with a high degree of necessary variety in order to
react to the perturbations of their societal environment. The redundancy
aspect is realized by organizations located in the centre, whereas the variety
aspect is realized by organizations and additional functional structures (e.g.
professions) in the periphery.
Organizations as interdependency breaks

The concept of interdependency breaks is another aspect of Luhmann’s
attempt to explain the societal function of organizations. Luhmann consid-
ers organizations as breaks of interdependencies within function-systems.
This argumentation follows the classical concept of interdependency
interruptions in cybernetics and general systems-theory, which is used to
analyse the processes through which organisms and systems achieve higher
levels of complexity by implementing step-functions (Ashby 1960). The
breaking of interdependencies is a mechanism that increases the internal
184
11 Kap 8. Drepper 05-11-11 16.22 Sida 185
complexity of systems. Internal system-differentiation is a kind of interde-

pendency interruption that equips the system with additional internal
selection-filters. Environmental demands can be selected and decomposed in
such a way as to affect only subsystems and not the whole system.
Environmental developments and occurrences do not affect the whole sys-
tem, but only those differentiated subsystems which are engaged in a partic-
ular environmental relationship. This kind of selectivity helps stabilize the
structure of the entire system.
The sociological relevance of this concept will be exposed later on in this
paragraph. Internal differentiation functions as a way of breaking up inter-
dependencies in social systems that have reached a certain level of complex-
ity. From the viewpoint of the theory of sociocultural evolution, introducing
breaks in interdependencies is a precondition for developing higher com-
plexity. In the case of modern society, there is a primary interdependency
break located on the first level of societal differentiation. This setting is spe-
cific to modern society and cannot be found in pre-modern societies, whose
strata were essentially based on interdependencies. In aristocratic societies
ascribed status and societal standing (Linton 1964) conditioned the access
to power, property, money, law and the arts. In contrast to that functional
differentiation is a differentiation form that is based on the simultaneous
independency and dependency of the different function-systems.
Function-systems process their specific functions autonomously. A partic-
ular function-system cannot substitute the function of the others. Modern
functional communications are conditioned by self-generated elements,
boundaries, processes and structures (Luhmann 1990f). Economic situations
and problems, for example, are topics specific to economic communications
but not to political, educational or legal communications. In economic com-
munications, the relevant criteria are economic: solvency, prices and con-
tracts, for instance. Similarly, the chance to accumulate property does not
depend on a specific type of social standing or class membership. To give a
further example, jurisdiction normally cannot be substituted by political
power or economic payment. Political power, educational success, legal deci-
sions “should” not be a matter of economic status and property in modern
society, just as political power should not be (ab)used to attain private goals,
make economic profit, and realize personal interests.
Modern function-systems work on their problems autonomously.
Therefore, modern society was obliged to invent structures that constitute
strong interdependencies within a subsystem, namely the symbolic general-
ized media of communication. These media (money, power, law, truth etc.)
bridge the difference of acceptance and refusal (Luhmann 1976b, 1990j)
and, in doing so, foster communications within the subsystems. However,
this implies that every function-system is dependent on the fulfilment of
other functions, as well as on the performances that the function-systems
185
11 Kap 8. Drepper 05-11-11 16.22 Sida 186
offer each other. The economy, for example, presupposes that the educa-
tional system “produces” adequate personnel that can be deployed and fur-
ther developed for the purposes of the working process. And the education-
al system presupposes that part of the money that is generated by economic
processes will be made available to education. In sum, this line of argument
does not suggest that there are no interdependencies at all in modern socie-
ty – corruption in politics and the economy are common knowledge – but
that such interdependencies are not stabilized on the primary level of differ-
entiation.
In fact, Luhmann’s point is that interdependencies occur on subordinated
levels of differentiation. He suggests that an organization, as a specific type
of social system, functions in wider function-systems as a structure that
breaks interdependencies. Organizations serve as substructures that limit
communicative connections. They block the direct correlation of every sys-
tem process. In the economic system, for example, specific corporations are
engaged in specific market segments, so that the ups and downs of car
prices, for instance, do not affect the price fluctuation of napkins and vice
versa. Similarly, in the political system certain parties stand for certain issues
and political orientations (conservatism, liberalism, socialism, ecology). This
differentiation on a subordinated level creates selectivity in the political sys-
tem and makes it adaptable to important societal issues. In science, to give
another example, universities and research institutes are often identified
with particular disciplinary paradigms. And these approaches often bear the
name of the university or institute: Harvard, Chicago, New York, Frankfurt,
and Bielefeld. The medium of truth and scientific disciplines alike are medi-
ated by and selected through organizational differences that function as
interdependency breaks.
In organizations departmentalism is a structure that functions as an inter-
dependency break. The benefit of this structural feature can be seen in the
stabilization of the system structure. Environmental perturbations that
touch one specific department or business unit and may cause a crisis do not
necessarily affect other departments directly. Every differentiated unit is a
segment of the organization that encompasses it, and has therefore its own
environmental relationships. This structurally legitimized indifference
enables the differentiated subsystems to concentrate on their specific tasks
and allows them to observe their specific environment for relevant informa-
tion. In organization theory, the need to counterbalance this process and to
reintegrate the differentiated subsystems is considered to be the complemen-
tary process of departmentalism. So the distinction between differentiation
and integration has become a well-accepted concept that describes the struc-
ture and dynamics of organizational systems (see Lawrence/Lorch 1967).
186
11 Kap 8. Drepper 05-11-11 16.22 Sida 187
Structural coupling and organizations

The concept of structural coupling can be understood as a complement of
the concept of interdependency breaks. The reverse of interrupted inter-
dependencies is structured interdependency between the autonomous
subsystems of society. The ability to break interdependencies among
function-systems is one important precondition for increasing societal com-
plexity. Structural coupling is basically needed as a counterpart to interde-
pendency breaks. It is a mechanism by which the operatively closed subsys-
tems of society are connected with relevant systems in their environment.
Luhmann’s notion of structural coupling originated in Maturana’s and
Varela’s theory of living systems. As he has often pointed out, the auto-
poiesis of systems is unthinkable without simultaneous structural couplings
between those systems and environmental resources and settings. A system
is structurally coupled to its environment, or to other systems in its envi-
ronment, if its structures are adapted or adjusted to the environmental cir-
cumstances or the structures of the surrounding systems. From an evolu-
tionary standpoint, one can say that modern society demands different
structural couplings, because functional differentiation increases the inde-
pendence and interdependency of function-systems simultaneously, as has
already been stressed in the previous chapter. In aristocratic societies the
unequal strata were coupled through households, which functioned as mul-
tifunctional segments. Households offered the possibility of communication
beyond the boundaries of strata, because in households people from differ-
ent estates interacted permanently. In contrast to that, modern society does
not predefine one particular form of structural coupling. A variety of differ-
ent forms is possible, because in principle every societal structure can func-
tion as a structural coupling. Symbols, texts (contracts, testimonies, and
reports), persons in specific roles (experts), and organizations (universities)
are some of the structures that Luhmann treats as structural couplings.
In his late writings, Luhmann discusses the relevance of organizations to
the various forms of coupling. According to Luhmann organizations con-
dense structural couplings and contribute to the structural couplings
between subsystems (Drepper 2003, pp. 237 ff.). Organizations are not
structural couplings in themselves, because structural couplings are institu-
tionalized on the level of society. Nevertheless, structural couplings would
not be able to achieve the necessary complexity for linking the autonomous
subsystems to each other without organizations that have the capability to
develop external communication, reap information, and bundle communi-
cations (Luhmann 2000c, p. 400).
The present thesis in this essay is that this abstract concept is not distinc-
tive enough and needs to be reformulated. There is a lot to be learned from
a structural comparison between function-systems that utilize organizations
187
11 Kap 8. Drepper 05-11-11 16.22 Sida 188
for structural couplings and function-systems that do not. Structural cou-

plings are based on existing societal structures and use the structural simi-
larities and compatibilities of the meaning-structures of subsystems. In struc-
tural couplings, the expectations that function-systems have from each other
condense and become institutionalized. To give an example, the structural
coupling between the legal and the political system rests on two well-
established symbolic generalized media with a clear-cut binary code: elabo-
rated programming and well-defined roles (politicians, lawyers, judges). On
this basis, the structural coupling can be established on a very abstract and
general level, namely that where the constitution is of interest in its capaci-
ty as a text.
With regard to coupled systems, structural couplings are expectation pat-
terns that condense the reciprocal expectations of those systems. If we look
at the constitution as a text that helps structure the coupling of the political
and legal system, we see that the political system relies on the legal legiti-
macy of the text, while the legal system presupposes the democratic legiti-
macy of the text. The legal system depends on the power (as symbolic gen-
eralized media) that has been embodied in the text, whereas the political sys-
tem depends on the law (as symbolic generalized media) that has also been
consolidated into the text. Through this circular process the constitution
becomes an institution and a valued address that is respected as a “para-
mount value” (Parsons) in both the political and the legal system. Structural
similarities to that can be seen in the coupling between the legal and the eco-
nomic system (contracts, property). In the case of contracts and property the
two symbolic generalized media – money and law – are institutionalized for
the purposes of structural coupling. If one signs a contract the economic and
legal conditions and consequences are clear: one incurs the obligation to pay,
and a breach of agreement can lead to legal action being taken.
In conclusion, one can say that when a system comes to fulfil its societal
function, a greater degree of structural elaboration reduces the necessity for
functionally equivalent structures that have to be stabilized on the level of a
lower system type (interaction or organization). If expectations cannot be
generalized permanently on the level of symbolic generalized media, formal-
ized expectations are a possible functionally equivalent structure. In an
empirical context, this refers to systems that cannot condition their opera-
tions through binary-coded and programmed symbolic generalized media
that bridge the communicative hiatus of acceptance and refusal. The educa-
tional and medical systems of modern society are examples of that. In those
cases formalized rules, professional roles, and personal authority are func-
tionally equivalent structures that alleviate the acceptance/refusal problem
of communication. And these structures are mainly stabilized on the orga-
nizational level (Drepper 2003, p. 240).
188
11 Kap 8. Drepper 05-11-11 16.22 Sida 189
Conclusion
Combining the theory of society with organization theory to form a societal
theory of organizations and an organizational theory of society is a current
topic in interdisciplinarily orientated organization theory (e.g. Tacke 2001,
Drepper 2003). The claim for bringing society back into organization theo-
ry is compounded by the demand for bringing organization back into the
theory of society.
What can be retained from the analysis of Luhmann’s theory is that the
societal embedding of organizations and the organizational side of society
have to be understood – in the tradition of Émile Durkheim – as a relation
of simultaneous enhancement. Society is the all-encompassing social system
that all further differentiations refer to. Because of that, organizations have
a double relation to their surrounding society: on the one hand, they repro-
duce societal communication – in this respect organizations are society – and
on the other hand, they distinguish themselves from their outer societal envi-
ronment – in that respect organizations are not society. Society exists on
both sides of the organization boundary. Thus organizations meet with
society in a double sense: in themselves and in their environment (Luhmann
2000c, p. 383). In this sense, the relation between the functionally differen-
tiated society and organizations must be understood as a “simultaneous
enhancement of both realities” (Luhmann 1982a, p. 12).
However, a systematic society theory of organizations is a desideratum in
Luhmann’s work. All in all, Luhmann’s contributions towards a systemati-
zation of the way in which organization and society interrelate are few and
dispersed. This text has tried to give a comprehensive overview of the
possibility to relate organization and society from the standpoint of Luh-
mann’s theory. The first section dealt with Luhmann’s critique of classical
organization theory, its rationality premises, and the ontological system
model. It expounded the thesis that Luhmann combines these critiques in his
early writings and that the different theoretical parts – in this case organi-
zation theory and general system theory – benefit from each other. Luhmann
uses the critique of classical-organization theory and the concept of organi-
zations and systems (whole-and-its-parts) that derives from that theory to
elaborate an alternative system and differentiation concept (open systems,
system/environment). In this period of Luhmann’s work the analysis of
organizations seems to be an excellent field for discussing and exemplifying
arguments with a basic theoretical relevance. Here, the theoretical path leads
from organization theory to social theory, so that the general theory of social
systems benefits from organization theory. The second section showed how
the basic theoretical differentiation of various types of social systems (inter-
action, organization, and society) could be combined with the theory of
internal structural differentiation of complex systems. That section aimed to
show that it is possible to focus on organizations as autonomous systems
189
11 Kap 8. Drepper 05-11-11 16.22 Sida 190
with their own logic of self-reproduction and at the same time examine them
as dependent substructures of a broader system. The second section also
highlighted some basic theoretical concepts regarding the relation between
formal organizations and functionally structured society. The formalization
of structures is a process that generalizes and civilizes complementary expec-
tations on the basis of membership rules. Furthermore, organization is the
primary system-level in modern society that stabilizes these expectation pat-
terns. Functional differentiation would not be possible without the system-
type of organization that provides the continuing motivation of individuals
and makes improbable expectations possible. The third section established
a link between the problem of repraesentatio identitatis of modern society
and two of the salient characteristics of organizations, namely, that organi-
zations absorb uncertainty and can be addressed as corporate actors in the
societal communication. Organizations compensate for the deficit in cer-
tainty particularly in the complex society, because they are visible addresses
that can be held responsible for certain actions and decisions.
The relation between subordinate differentiation-forms of modern society
and organizations was the main theme in the fourth section. This conception
presents Luhmann’s very first attempt to elaborate on the relationship
between organization and society systematically. Luhmann started working
explicitly on this issue in some of his last publications. In those works he
examined the connection between the internal differentiation of functional
subsystems on subordinate levels, and the role of organizations in this
process. Luhmann tried to define the function of organizations in relation to
the differentiation forms of segmentation, centre/periphery, and stratifica-
tion. This analysis has made clear that those differentiation levels cannot be
conceived of without organizational performance, and that the discourse on
modern society as an organizational society must take into account the rela-
tion between subordinate forms and organizational performance.
Further aspects of the societal functions of organizations were addressed
in the fifth and sixth section. The concept of interdependency break
described organizations as internal structures of broader systems that func-
tion as internal selection-filters for environmental demands. And the aspect
of structural coupling the seventh section discussed the complementary
process by which organizational performance contributes to the interrela-
tion of functional subsystems. In that section, the analysis of systems theory
stressed the functional equivalence of organizations to other social struc-
tures, such as symbolic generalized media and their materialisation (i.e.,
texts), professions, and roles.
The aim of this chapter was to present and discuss Luhmann’s diverse con-
tributions to the correlation between organizations and modern society.
There are a great many fruitful ideas in Luhmann’s work to draw on; never-
theless, the task of developing this topic and elaborating a societal theory of
organizations systematically remains to be done in the future.
190
12 Kap 9. Baecker 05-11-11 16.23 Sida 191
Chapter 9
The Design of Organization

in Society
Dirk Baecker
Which society?
One of the standard criteria of sociological organization research is that it
sees itself as a society theory of the organization and is able to deduce its
issues from hypotheses relating to the societal and thus the non-organized
elements of the organization. This applies on the level of the organization as
an “institution”, for which – depending on the particular social context
– there exists a clear idea of what can and what cannot be expected from
that organization in terms of regulatory activities. And it also applies on the
level of the organization as a communication, that is to say, the way in
which organized communication differs from non-organized communica-
tion. Typically, an organization’s institutionalized communication has to do
with “decisions” on “operations”. If you join an “organization”, you com-
municate and thus accept the expectation that your work will be the object
of the organization’s decisions. Organization means loss of autonomy for
the individual employee and a gain in autonomy for the organization. While
this is a precondition to all communication of decisions, a balance will
always have to be struck between how much loss of autonomy is bearable
and how much gain can be achieved.
The classical sociological theories of organization satisfy this precondi-
tion. Max Weber’s (1978) theory of bureaucratic organization describes the
filing system, the office, the civil servant, competency, qualification, profes-
sion, the form of financial compensation, separation of office and property,
and the idea of discipline, as structures in which the organization – in a tra-
ditional society on the threshold of modernity – was able to operate effec-
tively. Emile Durkheim’s (1984) theory of the division of labor suggests that
the feeling of solidarity with a different fellow being had to be accepted both
in the family and in other social groups so that the organization could con-
dition the degree of acceptable dissimilarity through the condition of com-
plementarity. Niklas Luhmann’s functional systems-theory states that the
individual organization can only be differentiated to the extent that it is able
191
12 Kap 9. Baecker 05-11-11 16.23 Sida 192
to relate its own programs, which have considerable degrees of freedom, to

the binary codes of a functional system (Luhmann 1997a, pp. 826–847).
It should be stressed that this sociological approach to organizational
research is only meaningful if we differentiate between organization and
society; in other words, if we do not assume that society itself can be orga-
nized or that everything that is possible or impossible in society is for that
reason possible or impossible in the organization as well. Society has no self-
evident bearing on decisions about labor; and what organizations consider
to be desirable and feasible is not identical with what society regards as
desirable and feasible (Udy 1970). It is only when we accept this difference
that Charles Perrow’s (1989, 2002) seemingly extreme position makes sense:
he claims that organizations in today’s society are increasingly “absorbing”
this society via the dependence of labor on earnings, the production and
handling of positive and negative external effects, and the bureaucratization
of access to factors of organization (e.g., human resources). It is only by
making this distinction between society and organization that we can
observe tendencies towards reducing or expanding it.
If we proceed from this distinction (see also Kieser 1989; Luhmann
1994a; Stern/Barley 1996), we do not find a neatly tied up social theory of
organization but rather a whole series of open questions which can be
ascribed to the fact that the term “society” is somewhat indefinite due to the
absence of a sociologically unequivocal social theory. Although the Marxist
notion of a “capitalist” society that has been the basis of industrial sociolo-
gy is gradually disappearing (Boltanski/Chiapello 1999; Deutschmann
2002), this does not mean that new notions are finding general acceptance.
In fact, of the various notions of society – from “risk society” to “informa-
tion society” to “knowledge society” – those that are currently holding out
reflect unresolved rather than resolved problems; that is to say, problems of
coping with responsibility, with the effects of too much information and
with knowledge deficiencies. Those who believe that the connection between
organizational research and societal theory means that the former is made
aware of the contextual conditions of world history, globalization, cultural
and religious differences, and the differentiation of functional systems which
permit every organization to define its reason for existence, its mission, and
its measure, will not be completely disappointed – but neither will they be
altogether satisfied.
The explanation is simple. Social theory today does not limit itself to
describing social macrostructures, which could define a more or less fixed
framework of conforming and deviating behavior for meso- and microstruc-
tures. If we accept the standard devised by Niklas Luhmann in his book Die
Gesellschaft der Gesellschaft, then social theory today is the theory of iso-
lating an elemental social operation referred to as “communication”, which
has to reproduce itself both microsociologically and macrosociologically
192
12 Kap 9. Baecker 05-11-11 16.23 Sida 193
(Luhmann 1997a; see also Willke 2001 and 2002). This suggests a depar-
ture from the broader social theories based on differentiation, evolution, and
media and a move towards the narrower issue of how we can reproduce a
communication that is assumed in principle to be improbable.
If we apply this interpretation of society, the question raised by the “risk
society” concept is whether and how the self-induced ecological risks to
society are or are not fed into the reproductive conditions of society, espe-
cially the distinction between individual and society (Beck 1992). The ques-
tion raised by the “information society” is whether and how communication
processes – in interactions between individuals, in organizations, and in
society – can cope with having to expect surprises constantly. Where the
“knowledge society” is concerned, the question is whether and to what
extent each position in society can cope with the attribution of knowledge
gaps and the need to acquire any missing knowledge. With the concept of
“globalization” the question is whether and how product, financial, and
labor markets can cope with the challenge of worldwide competition. And
the concept of the “world society” asks whether and how religion, culture,
intimacy, politics, economics, sport, the military, and education can be
equipped with a horizon of possibilities, and thus contingencies, which
covers the entire world of possibilities and contingencies, without seeking
refuge in the safe historical boundaries between religions, nations, epochs,
and ethnic groupings, which permitted the definition of more or less reliable
standards for what was to be expected in substantive, geographic, temporal
and social terms.
The reason for this ostensible “short circuiting” of micro- and macro-
sociological issues lies in the term “communication” which Luhmann coined
to define the “form” of communication. Following up G. Spencer Brown’s
(1972) notion of form, this means there is both an inside and an outside to
a communication, or in other words a communication comprises both the
aspects that are included in that communication and those that are excluded
(see Luhmann 1997a, pp. 36–43 and 78–91). Accordingly, a teacher’s strict-
ly local attempt to convey a basic knowledge of mathematics to his or her
student must not only reckon with the student either understanding or not
understanding mathematics. Teacher and student will also be assumed to
ask themselves constantly how the learning of math under the conditions of
school instruction relates to all kinds of matters of world society. Although
these are excluded from the actual communication of mathematics, they are
nevertheless included – as excluded elements – in the “form” of this com-
munication. The long-range implication is that exclusion per se is observed
in the communication (on both sides) and that the effectiveness of what the
communication actually includes or comprises (math instruction) will be in
whether this exclusion can or cannot be made plausible.
This interpretation of the term “communication” does not imply that all
193
12 Kap 9. Baecker 05-11-11 16.23 Sida 194
familiar differentiations “implode” in the interdependence of each and every

thing, as the holism of many a New Age philosophy would have us believe.
Rather, it suggests that every differentiation must always be prepared to be
made the issue that triggers disputes, and to have to be justified and/or nego-
tiated in its particular form. Of course, in order for such communications to
be possible via the exclusion and inclusion of the particular content of a
communication, distinctions have to be assumed and drawn which cannot
simultaneously be made the issue. In the society to which this definition of
communication reacts we have to deal not so much with processes of “de-
differentiation” but with processes of “oscillation” (a term coined by
Spencer Brown) in which the relationship between the inside and outside of
all kinds of distinctions is hampered by reflections on the quality of that dis-
tinction. The lines of distinction not only separate things from one another;
at the same time, they also relate things to one another, so that both sides of
a distinction occur on either side of the distinction, albeit in a different form
of operation (“cross”) and representation (“mark”). While the student is
being instructed by his teacher in mathematics, he or she is “operating”
within the frame of a school-class interaction, which revolves around the
subject of instruction, i.e., mathematics. Furthermore, he or she “marks”
whether and how this interaction is or is not effective in the light of politi-
cally motivated wars, economically modified career opportunities, cultural
education conventions, family-related expectations, and potential personal
relationships within the class community.
The simplicity and elegance of the concept of the form of communication
contrasts sharply with an enormous increase in the complexity of each single
communication which is made visible by this concept. Working with the
concept, we develop a feel for the “irritability” of every communication and
thus for the amount of work that has to be done on both sides of the com-
munication in order to reproduce it. It is in this context that Erving
Goffmann’s (1959) apparently micro-sociological, interactionist perspective
proves its far-reaching relevance to social theory because it is in fact thor-
oughly macro-sociological.1
Goffmann’s sociology not only reconciles – implicitly rather than explicit-
ly – micro- and macro-sociological issues (thus moving beyond the conven-
tional distinction), but he also works with the difference between communi-
cation and perception without drawing a dividing line between the two con-
cepts, as we would do today under the impression of the cognitive science-
based program of describing operationally closed mental and social systems
(Luhmann 1990g and 1995g). In fact, this is a further feature of the notion
1 See also Goffman 1974. This societal-theory dimension of his sociology is played down in
Goffmann’s (1983) own summary of his theoretical perspective. See also Rawls’ discussion
(1987) and Fuchs’s comment (1988).
194
12 Kap 9. Baecker 05-11-11 16.23 Sida 195
of the form of communication. Not only does it exclude objective aspects of

the communication or addressees or time horizons, whilst including certain
others, but does so under the condition that mental systems are in operation
in its environment, which – unlike communication – have the power of per-
ception and which, therefore, on the basis of their perceptions, to which
communication has no access, are likely to, or can both be discouraged and
encouraged, to take part in further communication. Communication
endeavors to transfer attractive or even compelling aspects of itself – warn-
ings to deviate, such as the audibility of the spoken language, the formal dis-
tinctness of script, the dramaturgy of body, space and time flows (Burke
1969). However, all this cannot alter the fact that mental systems – whose
capabilities are developed to a high level of sophistication exactly by taking
part in communication – are still free to direct their attention even when
they are susceptible to influence.
It is this form of communication that the social theory of organization has
to cope with today. It has the formidable task of describing the differentia-
tion of the organization both from society and within society under the con-
dition of drawing a distinction between communication and mental systems
that are capable of perception. The enormous complexity of this task is only
lessened by the fact that the organization has to use its distinction in socie-
ty vis-à-vis consciousnesses in order to remain attractive, and that it can
demonstrate to society this distinction vis-à-vis systems of consciousness in
order to maintain and expand its scope for differentiation. In other words,
the organization gains the commitment of consciousnesses through a “bet-
ter”, e.g., more reliable, “society” (namely: “community”) – or it has the
opposite effect because it is more authoritarian, more undemocratic and
more uncivilized than mental systems are prepared to accept. And organiza-
tion derives its legitimation from society by demonstrating that it is capable
of commitment vis-à-vis consciousness and possibly capable of different
degrees of commitment vis-à-vis the consciousnessses ascribed to employees,
customers and other observers.
Organizational design
In his book Organisation und Entscheidung, which concluded his work on
organization theory, Niklas Luhmann suggested using the category of
“design” to describe how an organization can secure the autopoiesis of the
communication of its decisions under the condition of fascination, orienta-
tion and commitment of mental systems in its environment (Luhmann
2000c, p. 148). Through its design, an organization makes itself perceptible
to these mental systems in such a way that they are not only impressed or
even fascinated; they also have to be able to recognize what the main con-
195
12 Kap 9. Baecker 05-11-11 16.23 Sida 196
cern of that organization is, not just as an aspect of perception but also as a
feature of the social system. And for this reason alone it will be vital not only
to spellbind, i.e., to constrain the mental systems, but rather to sustain their
interest, to encourage them to make a contribution, because the aim can not
only be to restrict their latitude, but also to give them some influence on the
decisions taken by the organization.
The complexity of these demands on the design of an organization is per-
haps nowhere more obvious than in a military organization, which under all
three possible conditions of operation, i.e., as training organization (in
peace), as a policing bureaucracy (in action), and as a bureaucracy in chaos
(at war; on the latter, see Keller 2002), has to be in a position to guarantee
coordinated action and maintain hierarchical distinctions. This means, the
military organization must resolve the problem of integrating highly specif-
ic perceptions on the one hand, with a far-reaching generalization of com-
mand structures on the other hand in such a way that the organization as a
whole is both effective and maneuverable. Each type of operation (training,
policing, war) must be enacted without the risk of one of its branches, army,
navy or air force, separating and becoming immune to commands.
The two major elements of military organizational design that contribute
much to resolving this problem are the weapon and the uniform. Both
describe the deployment conditions of the soldier in the field by defining his
scope for autonomy (the range of his weapon and competence of his rank)
and by conditioning this autonomy in terms of reinforcement and coordina-
tion (as regards the weapon) and superiority or inferiority (indicated by the
uniform) so that the individual’s relationship to his unit and its relative posi-
tion in the organization is manifest without further communication. Even
changes from one system state to another can be communicated and made
perceptible by modifying rather than substituting weapon and uniform.2
In other organizations the same function is performed differently. Officials
in all bureaucracies appear to adopt ultimately a specific set of gestures,
mimicry, and rhetoric which permits them to communicate without the need
for further thematization. They also clearly reflect the nature of the organi-
zation’s decision making and indicate the approach that would-be associates
should take when addressing this organization. The highly disciplined offi-
cial like the well-drilled soldier is not a fortuitous by-product of working
conditions in the military or in a government agency; rather, they are “fig-
ures in a design” through which uno actu the social distinction of each
organization and the structures of its autopoiesis are symbolically commu-
nicated and made manifest. It is for this reason, too, that a sociology of the
2 The “dress uniform” and non-bearing of a weapon, or its substitution by an ornamental

weapon, make clear to the off-duty soldier that his behavior vis-à-vis civilians must as a rule
be different from that vis-à-vis the enemy.
196
12 Kap 9. Baecker 05-11-11 16.23 Sida 197
bourgeois society was oriented towards the figure of the “substitute”, as

well as his behavior, prevalence and chances in life, in the belief that the two
figures of the “worker” and the “entrepreneur” and with them the condi-
tions of the “class society” were truly obsolete (Simmel 1989, chapt. 6;
Siemons 1997; Kracauer 1998); not to mention the field of literature that
emphasizes the claims laid upon perception by organized communication as
so unreasonable that the autopoiesis of the communication of decisions is
jeopardized.3
Just as the design of an organization can be described as a symbiotic
mechanism, to use Luhmann’s term, which permits every single communi-
cation to gain access to the psyche of a consciousness and the physis of a
body, this access can in turn be called into question as soon as the psyche
becomes aware – through other communications to which it is structurally
linked (for example, when it reads) – of what the implications are. A move
from the army to the air force can be an attractive proposition for the simple
reason that the shirt collar is not worn as tightly so the individual has entire-
ly different ways of proving his individuality.
Although the term “organizational design” is familiar in the literature on
organization, there has so far been no successful attempt to accurately
describe its problem reference. The distinction between social systems (com-
munication) and mental systems (consciousness), which has been developed
from the theory of self-referential systems, is – as far as the special case of
the organized social system is concerned – a proposition for describing the
problem reference to which the term reacts. Moreover, we can conceivably
interpret the notion of design – over and above organization – as a “struc-
tural coupling” mechanism (Luhmann 1995g, pp. 12–24) between con-
sciousness and communication. We cannot go into this here. It is more
important that even without an explicit definition the literature has com-
piled the structural features of “organizational design” which are consistent
with our interpretation of the problem and function of design.
First, the question of organizational design refers primarily to the separa-
tion and coordination of the various tasks of an organization: this means
that the boundaries around positions and jobs have to be drawn in such a
way that competencies are both distinguished from one another and related
to one another and that these boundaries are both drawn reliably and can
be observed (Thompson 1967; Galbraith 1973 and 1977). We should be
able to assume that an organizational design, which in the final analysis has
the job of presenting routines and contingencies not as two different worlds
but as something that is connected, will always symbolize every boundary
3 I am thinking here of course of Franz Kafka’s two novels The Castle and The Trial and more
especially of his short story “In the Penal Colony”.
197
12 Kap 9. Baecker 05-11-11 16.23 Sida 198
line in order to make the unity of distinction both communicable and per-
ceptible within the organization.4
The literature not only encompasses questions of distinction and reinte-
gration but also discusses questions relating to the type of communication
that permits us to differentiate and thus make autonomous the autopoiesis
of the communication of decisions within an organization and also to refer
back both to the conditions of their differentiation and to achievements in
the environment of the organized system. From this angle, it is worth con-
sidering the question of organizational design as a comparison of the effec-
tiveness of different communication genres: obviously, electronic, written,
and oral communication and the reference to documents versus discussion
notes, or telephone versus face-to-face communication offer quite different
possibilities both of differentiating and re-embedding a communication
(Daft/Lengel 1984; Yates 1989). Organizational design here means using the
communication channel as a variable for monitoring and control in the
management of the organization.
And finally, on the subject of organizational design we come to the so-
called context variables, which describe the advantages and disadvantages of
unequivocal versus equivocal decisions, unique versus recurring or compa-
rable decisions, variable versus stable decisions, heterogeneous versus homo-
geneous and independent versus autonomous decisions, depending on the
specific environments with which the organization is confronted (Butler
1991, p. 249). Here too, the organizational design must consider the extent
to which the “communicative” recruitment of what kind of environmental
contributions is involved or – at the other end of the scale – the “technical”
isolation and insulation of the organization from the environment. This
involves directing constant questions both at the internal environments of
the organization (positions, departments, hierarchical levels) and at the
external environments (mental: consciousness systems; social: other organi-
zations, functional systems, interaction; organic and physical: natural envi-
ronment). In doing so, the organization can only pursue those design
options which either have the seal of society’s approval or to which society
at least has no objection.
We could apply the term “regime” to describe socially approved options
for organizational design, if we take regime to mean the ability of a social
system to keep grasp on both its own sub-systems and systems in the
environment of this system. Regime, then, controls the risk of the control of
its sub-systems by controlling the risks of its control of environmental sys-
tems: “Institutions and the rhetorics they come to agree upon both are open
ended and can be extrapolated beyond present local disciplines and net-
works. Call the product, which is a native statement combining styles
4 For the “symbol” as “semiotic condensor” see Lotmann (1990).
198
12 Kap 9. Baecker 05-11-11 16.23 Sida 199
around institutions, a regime” (White 1992, p. 226). Typically, both these

risks are increased and controlled through differentiation and heterogeniza-
tion rather than through integration and homogenization, so that, for exam-
ple, a heterogeneous workforce is a reliable indicator that the level of inter-
nal and external differentiation of public administrations has increased
without a simultaneous decrease in the auto-control of these organizations.
This goes counter to the commonly held belief that the more homogeneous-
ly the reality produced by control presents itself, the greater is that control.
This is the reason Deleuze (1993) distinguishes “disciplinary societies” and
“control societies”.
Our thesis is that nowhere is the societal component of an organization
more impressively documented than in its design. Of course, this embraces
all forms of organizational culture much discussed since the 1980’s and with
whose help organizations have tried to take responsibility for defining their
societal preconditions in order both to be independent of these on the one
hand, and to be able to establish links with the societal environment on the
other hand (Ulrich 1984). For this reason, most organizational cultures aim
at a socially acceptable form of hierarchy; in other words coordination and
cooperation are designed to be both mentally acceptable and capable of re-
conciliation with individualization needs and to be in concert with the social
systems in the organization’s environment, i.e., other organizations or polit-
ical, economic, educational, and legal systems or the mass media.
At the same time, this social aspect of the organization is nowhere more
the subject of debate than in the way in which the organizational design is
not so much an “interface” to other social systems (including its own sub-
systems) but rather to the mental systems in its environment. There appear
to be at least two reasons for this, which are probable closely related,
although on first sight they seem to contradict one another. The first of these
is that as ecological risks grow, society is increasingly keen to adopt the per-
ceptibility of mental systems and no longer to simply dismiss them as the pri-
vate affair of individuals. It may be, and this is something that would need
to be researched further, that it is easier for organizations than for other
social systems both to focus and to control the excess “resonance”
(Luhmann 1989a) of the communication vis-à-vis perceived environmental
circumstances. The thematization of environmental behavior, which in func-
tional systems simply becomes caught up in the maze of the relevant binary
codings and produces little more than indignation and resignation in inter-
active systems, can be made the object of decisions provided these distinc-
tions obey at the same time demands of value-added product design and
quality with respect to markets and other publics.
The second reason is connected with the society’s growing interest in
“individualizing its individuals” (Luhmann 1989b). It seems to be in con-
tradiction to the first reason in as far as it deflects the attention of society
199
12 Kap 9. Baecker 05-11-11 16.23 Sida 200
from ecological problems and addresses the opportunities of individuals in

numerous relationships within society. On closer examination, however, it is
possibly one and the same thing. I say this because the interest in “individ-
ualizing individuals” increases as society attributes its “home-grown” prob-
lems to its own regulative standards and suggests that the solution lies in the
individuality of each member of society – in terms of politics, the economy,
education, the legal system, and religion. The resulting “self-conflict” has
permeated this society since Plato under different labels, the most current of
these probably still being the contradiction between “totalitarian” and “lib-
eral” social models, whereby “liberal” does not mean acquiescing to all the
demands of the individual simply because they come from the individual,
but to concede all an individual’s errors which, in the neighborly rivalry of
individuals, can be mutually neutralized (Hayek 1976).
Individualization means above all else to redefine the relationship between
communication and consciousness in each individual case, and thus to
increase the likely deviations of every single communication in the exchange
with each individual consciousness on the one hand, and to increase the
mutual orientation of communication and consciousness towards each other
because, as we know, there is nothing like uncertainty to provoke imitation
(including the imitation of deviations; see Alchian 1950, p. 219).
In other words, it may be that the ecologization and individualization of
society goes hand in hand with the attempt to equip society both with “irri-
tation potential” and with observation points in the light of its vulnerable
overall state. Whether this suggestion ascribes too much of Hegel’s “cunning
of reason” to society, whether these “measures” would be effective, given
the institutional inertia of quite different dimensions, and, finally, whether
indeed the state of society is vulnerable are questions that will have to
remain open here. It is more important for us that, at least on the organiza-
tional level, there are indicators that problems are not only being identified
and communicated but also that they are being tackled. Just as the now-
adays much maligned “bureaucracy” of the 18th and 19th centuries did much
to emancipate society from traditional orders of kinship, class, ethnic group-
ings and cliques (Stinchcombe 1974) – and given the enormity of the under-
taking, who would not expect there to be setbacks? – the organization could
play today an equally creditable role by providing society not only with the
still invaluable rationality of administrative structures (Luhmann 1977b;
Luhmann 2000c, chapt. 15), but also with the intelligence of perception-
sensitive forms of communication. To put it plainly: the aim is to give the
organization and thus society renewed access to the empirical worlds of
labor, which in the wake of “Taylorization” in factories, government offices,
schools, and universities, were replaced amid much dispute and unrest by
work concepts designed on scientific, pedagogical and didactic principles
(Edwards 1979; Malsch 1987).
200
12 Kap 9. Baecker 05-11-11 16.23 Sida 201
The attempt to access the intelligence of perception-sensitive forms of

communication is being made particularly in the frame of organizational
designs that are conceived as “self-designs” or at least that are receptive to
self-design (Hedberg/Nystrom/Starbuck 1976; Nystrom/Hedberg/Starbuck
1976). If intelligence can be interpreted as observing own knowledge defi-
ciencies with a view to drawing on the possible knowledge of others, and
accordingly both using and extending this knowledge (Baecker 2002), then
we are looking at organizations which – as “distributed systems” (Fox
1981) – can design not only beyond internal system differences but also
beyond system/environmental differences. We could also say that what we
see here are “networks” if by that we mean highly improbable reproduction
relations of heterogeneous elements (White 1992). In other words, an orga-
nization’s socially motivated but also socially representable access to per-
ception is not achieved by extending organization technology or by adopt-
ing more subtle forms of organized communication but by a combination of
both, in a strictly loose coupling, if that term is permissible. Perception can
be coupled to the organization if organizational technology poses commu-
nication tasks and if organizational communication poses technological
tasks, whereby technology is unable to resolve the communication tasks and
vice versa.
This is a new situation, because with this problem the dreams of “socio-
technical” systems – in which communication uses technology and technol-
ogy uses communication – appear to disintegrate (Bijker/Hughes/Pinch
1987). What we now see is a paradigm that is oriented more towards cer-
tain team sports or orchestral music: the specific techniques are laid down
in rules, and instruments with a certain latitude for variation are stipulated;
the level of communication (gestures, calls, signs from the conductor’s
baton) varies, and interaction between the players depends on perception.
To understand what is meant here, imagine for a moment a basket-ball game
or a symphony concert without mental systems (including the conscious
control of the body) being present.
With this approach, organizational design means relating technology and
communication – or causality and contingency – in such a way that maxi-
mum effectiveness in both technological and communication terms can only
be achieved by including perception. In the language of Neville Moray
(1984), we could also say that organizational design of this kind will neces-
sarily produce those “ill-defined” systems which can only be transformed
into “well-defined” systems through the non-predetermined but risk-
oriented and thus strictly choreographed and “playful” (Derrida 1970) use
of perception and corresponding reactions.
Currently the most convincing paradigm for what is meant here by orga-
nizational design are those high-reliability organizations (HROs) that Karl
E. Weick and Kathleen M. Sutcliffe (2001) have made the subject of their
201
12 Kap 9. Baecker 05-11-11 16.23 Sida 202
book on Managing the Unexpected. They study how work is organized on

aircraft carriers or in nuclear power stations to develop their concept of
“mindful management”, which in this context means the reintegration of
loosely coupled communication and perception. Management is “mindful”
when it is in a position to be aware of its own expectations, the limited hori-
zon of these expectations and the need for ongoing corrections and when it
draws on both communicative and perceptive data to check the soundness
of these expectations; in other words, when it directs its sights on ongoing
decision requirements and keeps its eyes (and mind) open. The competence
that is vital here may possibly be the “sensus communis” of an ongoing
assessment of a community’s historical situation, which Hans-Georg
Gadamer (1989) describes in Truth and Method as the essence of a globally
informed, communicatively prudent and wise, rather than theoretically
“correct”, behavior.5 Here too, the crucial point is that the objective is not
to install fixed couplings between individual perceptions and communica-
tion structures quasi-technologically, as these are imagined in the motivation
and management concepts of numerous management philosophies that are
erroneously believed to be grounded in the social sciences. On the contrary,
they should aim to reverse the “infantilization” (Argyris 1957, p. 66) of
employees in organizations by permitting loose couplings and defining
mature behavior expectations.
Routines
It should probably be pointed out that “organization” has always meant
relating causality and contingency, technology and communication in such a
way to one another that routines can be found, put firmly in place, and
monitored. There can be no doubt whatsoever that the exclusion and re-
embedding of individual, both cognitive and emotive achievements are at the
center of the social constitution of the organization. Establishing routines
has always meant and continues to mean safeguarding social processes,
whether these relate to filing, assembly line production, field service, uni-
versity research, classroom instruction, or treating the sick, against the errat-
ic effects of individual perceptions. At the same time, however, precisely
5 See also Gadamer’s comments on the essay “Sensus Communis, an Essay on the Freedom of
Wit and Humour in a Letter to a Friend” (1989 in: Shaftesbury 1999). Readers not wishing
to pursue these sources should study instead the typically antibureaucratic Hollywood film
and consider roles such as those by Nicolas Cage as Dr. Stanley Godspeed in “The Rock”
(USA, 1966) or Alec Baldwin as Jack Ryan in “The Hunt for Red October” (USA, 1990), not
to mention their respective co-actor Sean Connery in the roles of John Patrick Mason and
Captain Marko Ramius. Virtually every one of Silvester Stallone’s parts uses the same topos
of intelligent perception versus blind (and always highly organized) communication.
202
12 Kap 9. Baecker 05-11-11 16.23 Sida 203
those perceptions have to be “recruited” without which the routines will not
find their object and will not be cushioned by cooperation. The extent of the
closely linked de-motivation and re-motivation of individual perceptions
that are directed through the organization along different paths from, say,
family, play, intimacy, art and science, has throughout history never received
the acknowledgement it deserves.6
We cannot write that history here. We shall restrict ourselves in this final
section to naming three aspects from which it would make sense in social
terms to re-introduce the excluded individual perception into the organiza-
tion (Schimank 1986). In doing so, we shall focus on considerations that
center not on the abandonment of routines, but on their reformatting. Only
in this way can we justifiably claim to make a statement about “organiza-
tion” and not, say, about something entirely “other”, towards which we
need to “free” ourselves. We believe that the objective of an organization is
primarily to structure the communication of decisions in such a way that
decisions can absorb their uncertainty and not simply pass it on and allow
it to become the aggregated condition of the system (March/Simon 1994;
Luhmann 2000c).
Rather like the memory of a social system, routines are able to both “dis-
count” – from event to event – the difference between each of these events
and to “recount” the connection between each of them: in the face of the
complexity of these events, whose sheer variety makes any linearization and
sequentialization improbable, this means selectively forgetting what it is that
makes them individual and selectively remembering how individual they
have to remain in order to be coupled to past and future events. The refer-
ence to “society” in this context means being able to install loose couplings
between communication and perception, which the organization itself must
neither describe nor answer for and for each of which different options of
remembering and forgetting can be formulated. To operate “routinely”
means to be able to “depreciate” perceptions with a view to communication
and communication with a view to perceptions in order to be able to con-
tinue both.7
6 This is especially true if we concentrate on the difference between organization and society,
which remains to be introduced into the Foucault-type of analysis of the connection between
knowledge, power and discipline, which is particularly relevant to our context. To equate one
with another would mean not least to be unable to reflect on one’s own standpoint. The dis-
tinction between the closely coupled de-motivation and re-motivation derives from Saussure’s
language theory. The following quotation by Derrida (1974, p. 112) refers to this theory: “La
langue doit sa naissance à la démotivation, mais ne pourrait évoluer sans avoir recours en per-
manence à la rémotivation des signes et structures.”
7 The description of the problem may be unfamiliar, but the essence is well known. It has to
do with the above-mentioned “sensus communis” as well as with the political wisdoms of the
sixteenth and seventeenth centuries (Machiavelli, Baltasar Gracián).
203
12 Kap 9. Baecker 05-11-11 16.23 Sida 204
We shall concentrate on three routines in this sense: on the newly emerg-

ing cultural form of the organization under the conditions of the introduc-
tion of the computer into society, on risk-bearing in networks, and on the
significance of commitments in structuring the relationship between indi-
vidual and organization.
The most important of these aspects, which we can no more than outline
here, is probably that routines today have to do with a much changed “cul-
tural form” of society. If, like Niklas Luhmann (1997a, pp. 409 ff.), we may
interpret a cultural form of society as the way of dealing with the super-
abundance of meaning, which the major communication genres or media in
a particular society produce, then we can assume that today’s predominant
cultural form (predominant in the sense of giving orientation) is no longer
the teleological definition of a purpose, with which Aristotle restrained the
surpluses of meaning of the written word, nor the certainty-creating rest-
lessness with which Descartes reacted to the surpluses of meaning that
resulted from book printing, but the the idea of media of communication by
which Marshall McLuhan (1968) and Talcott Parsons (1977) considered
controlled the superabundance of meaning produced by the computer. These
media work their way via linking motivation to selection thereby reassuring
the society that there is no gain in degrees of freedom (motivation) without
a corresponding loss in degrees of freedom (selection).
Organization today still means – in best administrative science and busi-
ness economics tradition – the creation of a relationship between purpose
and conditions, in which sufficient tension can be built up for pursuing and
changing objectives (Luhmann 1977b, pp. 101 ff.; Luhmann 2002, pp.
256 ff.). Not least in the form of its traditional hierarchies, the organization
is conditioned to enact those routines which operate with a calculated quan-
tity of resources in pursuance of a particular purpose, or which allow those
purposes to be interchanged. It is not least for this reason that we distinguish
between labor on the one hand, which if at all possible should have a pur-
pose, and management on the other, which introduces that restlessness into
the organization that permits those purposes to be reviewed. In other words,
the cultural form of the organization is well prepared for both the written
word and book printing.
With the introduction of the computer, however, society has had to adapt
to yet another different cultural form which may be characterized as the
need to create an awareness of form that can cope with objective, social and
temporal decomposition – in the face of a complexity which has become
incalculable due to its ongoing and multifold computation. Where the intro-
duction of the written word had meant, among other things, that people had
to expect individual memories to deviate (and thus had to devise bookkeep-
ing) and where the introduction of book printing meant that increasing com-
plexity could only be managed by a change of classifications (and thus by
204
12 Kap 9. Baecker 05-11-11 16.23 Sida 205
specialists), the computer confronts us, as only art has done hitherto, with
the phenomenon of having to acknowledge each individual condition of a
system as the starting-point for different couplings and being able to control
(or rather: visualize) the resulting complexity only by constantly devising
new network-like “figures” (Gestalten).
The cultural form “medium” provides the appropriate background think-
ing, if we interpret “medium” as the observation of a range of possibilities
from which a selection has to be made and where selection can only be con-
tingent so that every ongoing selection is convincing from the viewpoint of
its materialization, and is adopted and examined from the viewpoint of its
effectiveness until a new convincing selection emerges.8 The medium moti-
vates, as Parsons suggests, through the fact of its selection, or, in other
words, both through selection, which takes place on a case-by-case basis,
and through the other possibilities, which, although excluded at the present
time, are (triggered by the fact of selection) kept in mind nevertheless.
Accordingly, the purposes and restlessness of an earlier age are not obso-
lete. On the contrary (Rosenblueth/Wiener/Bigelow 1943): they are overlaid
with a notion of “form” which considers the purposes to be secondary (rel-
ative to being on the move and in transition) and no longer accords to rest-
lessness the credit it once enjoyed, i.e., that it is simply the transition stage
on the way towards stabilization. The search now is for “disciplines”,
“regimes”, and indeed for “empires”, in which the precarious and thus
interdependent identities of institutions, ideologies, industries, and individ-
uals are configured in a frame that overarches heterogeneity. This can only
be described by a network theory which has recently gained currency and
which can handle the ambivalent phenomenon of identity and control
(White 1992).9
It is still too soon to predict with any certainty what the cultural form
“medium” will mean for organizations and their routines. What we can say
is that the shifting of the reference points “hierarchy” and “environment” to
“networks” and “projects” indicates a greater potential for deconstruction
and re-combination with regard to all production factors (labor, capital,
technology, real estate, organization and information) and thus confronts us
with a greater depth of focus both in innovation policy and in the analysis
of value-added chains, organizational processes and interface design (Peters
1993; Peters 1999; Heuskel 1999). Organizational designers, managers and
consultants no longer direct their sights just on criteria of technical and eco-
nomic efficiency and effectiveness; their approach could probably best be
8 If this thinking is accurate, we must accept that a “mathematically” stringent theory of com-
munication on the lines suggested by Shannon and Weaver (1963) might by of some future
prospect.
9 On the concept of “fractal integration” through the “control of differences” in the “empire”
see Hardt and Negri (2000, pp. 325 ff.).
205
12 Kap 9. Baecker 05-11-11 16.23 Sida 206
described rather as “cultivating improbable projects”. In doing so, they have

revived the hermeneutic practices advocated by Gadamer (1989) as well as
systemic approaches which are based on Heinz von Foerster’s, Humberto R.
Maturana’s and Francisco J. Varela’s (Maturana/Varela 1980; Von Foerster
1981; Varela 1999) concept of “structure-determined systems”: according to
these, every organization, every corporation, every government authority
has its historical state, which is a product of the states through which that
organization has passed and which has to be recognized (which does not
necessarily mean: presupposed) as crucial to everything that is to come. This
state, as the term “medium” indicates, is nontransparent in all significant
respects, and we can grasp it only by using “operational research” (Ashby
1958, p. 97) based on current information. In the medium of an organiza-
tion thus interpreted, the cultivation of a project means to realize the
improbable by comparing available and non-available production process-
es. The organization can verify at all times whether it is in a position to do
so by keeping a close watch on the relation between communication and
perception. Only if it has routines for controlling this relation, based on the
concept of loose coupling, can the organization hope to make the difference
between communication and perception the paradigm for combining labor,
resources and results.
In the context of the cultural form “medium” with which society is react-
ing to the introduction of the computer, the organization and its routines are
also assumed to be able to cope with a far higher level of nontransparency,
with regard not only to social and mental, but also to ecological and tech-
nological relationships than would hitherto have seemed healthy to all
visions of control (Kelly 1990). The concept of “form”, as developed by G.
Spencer Brown (1972), can be usefully applied in this situation because it
incorporates the “other”, “unmarked”, side of the distinction in the form of
differentiation so that no transparency of self-evidence is to be expected at
any point of operations.
The reverse side of this concept – and this is our second aspect in consid-
ering how re-formatted routines might appear today – is that the two sides
of the distinction are not only separated from one another, but that they are
also related to one another. We suggest using the term “risk” – which has
been effectively used in the social sciences for several decades – to describe
the connection between the two sides of the distinction and thus to describe
the constitution of a form (Douglas/Wildavsky 1982; Short 1984; Wildav-
sky 1988; Beck 1992; Luhmann 1993i). To use a differentiation suggested
by Niklas Luhmann (1990g, pp. 131–169; 1993i), “danger” denotes the
connection – visible only to the outside observer – between the two sides of
a distinction from the viewpoint of a threat to this distinction; “risk”
denotes the re-introduction of this observation on the inside of the distinc-
tion for the purposes of an inside reflection of that distinction, from the
206
12 Kap 9. Baecker 05-11-11 16.23 Sida 207
viewpoint not only of a threat to it but also of coping with this threat or, if
you like, from the viewpoint of risk management.
Taking up one of Kenneth J. Arrow’s lines of thought – “the competitive
allocation of risk-bearing is guaranteed to be viable only if the individuals
have attitudes of risk-aversion” (Arrow 1964, p. 91) – we shall assume that
all network connections only emerge when risks are taken in order to avoid
them and when this paradox enfolds as it spreads to the different positions
in the network. Again, we are faced with the question of distributed systems,
but in this context we pose it in a strictly sociological sense, i.e., how can a
connection be made between two or more elements when this connection
has no other motive than itself, or in other words, when the corresponding
objective, social, and temporal motives have to be created after the fact and
in order to maintain the connection? The answer to the question is that the
connection will only materialize when all involved positions can assess with
an adequate degree of reliability the risks that each one of them will take,
the extent to which it recognizes and can cope with this risk, and the extent
to which this risk makes each position selectively dependent on all other
positions. Only then is it true to say that risks are taken in order to avert
them.
It is safe to assume that from the viewpoint of risk-bearing through risk
assessment, organizational routines have been over-taxed for several decades
now because the increased complexity and turbulence not only of the social
environment but also of mental, cultural, technological and ecological envi-
ronments has forced risk perceptions on them for which they have no cou-
pling modus. To recognize, on one hand, and to fear mental, cultural, tech-
nological and ecological risks, on the other, is one and the same thing to an
organization when it can no longer trust the regimes of authority, hierarchy,
expert know-how and social indifference, which in the past afforded pro-
tection from these risks (and transformed them into ecological hazards). If
these regimes fall victim to the transformation through which society adapts
to ecological hazards in the broadest sense of the word, organizations are
rendered helpless unless they can transform these same risks, from which
they were hitherto protected, into options. These in turn have to be tailored
to the organization’s different environments which are now transformed
into network partners.
In other words, organizations must begin to increase, present and cope
with mental, cultural, technological and ecological risks – alongside the con-
tinuing social risks – in order to expand their networking capability in all
these dimensions (Handy 1994; Handy 1996; Sennett 1998). Careers must
be broadened by options for self-fulfillment – and the concomitant draw-
backs of imposed “flexibilization” and forced aging. The familiar social
milieus of work must be extended by training and development programs –
and the drawback of emancipation from the “milieu of origin”. The
207
12 Kap 9. Baecker 05-11-11 16.23 Sida 208
conventional buyer’s or self-developer’s attitude towards technologies has to

be replaced by a willingness to share the risks of the technology develop-
ments of partners who need to have different options in order to stay calcul-
able. An organization should no longer ignore external effects but must
internalize these in order to maintain demonstrably its sustainability with
the ecological environment. And finally, the organization must broaden its
hitherto more or less monopolistic connection to its elected functional sys-
tem (the economy for corporations, politics for government authorities
including the military, the law for courts and solicitors, education for
schools, science and universities, art and culture for museums, galleries and
theaters etc.). It must consider other, if not all, functional systems in order
to demonstrate its established position in society or, in other words, its risk
acceptance, both to the inside and the outside.
Although the degree to which this takes place will be highly selective, it
does not mean that it could just as well not take place. It means, however,
that all organizational routines would be hopelessly over-taxed unless the
organization were able to “take refuge” communicatively in taking account
of certain risks only by perception and to take refuge perceptively in taking
account of certain risks (possibly the same ones) by communication. “Talk”
and “action” in which an organization’s decision-making behavior was hith-
erto embedded according to Nils Brunsson (Brunsson 1985; Brunsson 1989;
Brunsson/Olsen 1993), are supplemented through “narration” and “experi-
ence” (Guillet de Monthoux 1983; Czarniawska 1997; Rifkin 2000), so that
those dimensions of perception can be addressed which are not currently
accommodated in communication; which, in other words, are neglected in
the decision-making process. Virtually every manager is aware of these new
dimensions of concern; the organization must do everything in its power to
develop a networking capability in areas which only recently lay far beyond
its horizon of interest. And this networking capability, thus my hypothesis,
will depend on the organization’s willingness to accept risk whatever the
cost of the acceptance may be.
A third aspect to be considered when identifying which routines enable an
organization to capitalize on its difference vis-à-vis society to organize its
handling of communication and perception can be derived from the exten-
sive literature on issues relating to “commitments”. The concept of “com-
mitment” was introduced particularly into sociological literature to describe
ties within an organization which, in terms of rationality, considered means
to be interchangeable with other means with regard to their effectiveness for
achieving ends. According to Philipp Selznick’s (1966, pp. 255 ff.) famous
study of organizations, commitments refer to means and goals and to the
connection between these; in other words they transform rational loose cou-
plings into traditional fixed couplings, or freedom of scope into networks
which – like immune systems – rebel against their dissolution. Commitments
208
12 Kap 9. Baecker 05-11-11 16.23 Sida 209
of this kind evolve out of the autopoiesis of the organization (Selznick

speaks of the task of preserving the organization); out of the social and pro-
fessional backgrounds of the organization’s employees, especially when
these are coupled with demands that are not identical with those of the
organization; out of the institutionalization of the organization’s values and
goals, which restrict options and dictate behavior guidelines (Selznick under-
lines the commitment effect of so-called precedents); out of the organiza-
tion’s social and cultural environment whenever this comes into conflict
with the organization and to which the organization can only react with cer-
tain “claims to power”, which in turn trigger ties on the inside; and, finally,
out of interests which crystallize in the course of the organization’s history
of delegation practices.
Today, we tend to see commitments of this kind not as a problem but as
a solution to the problem. Provided that the risks of “overcommitment” can
be eliminated – these are seen in a lack of creativity, opposition to change,
and exaggerated conformity – commitments can be used as routines with
which to tackle motivation problems (Scholl 1981; Randall 1987). This
means we demand the chance to take on commitments on condition that
these are viewed by all involved as something they can renege on. In this
form, we might assume, commitments are a means of developing small,
compact milieus – given the above-described demands for complexity both
on the cultural form of the organization and its incorporation in risk net-
works. These milieus derive a degree of momentum and energy – that in
some cases can be quite considerable – from the question of how long they
will succeed in reproducing themselves and from the efforts required of them
to liberate themselves from the commitment pressure they create.
Commitments are social artifacts, but not of the kind in which individu-
als are actually committed to organizations or even vice-versa, but in which
standards or routines define the intensity of communicative and perceptive
relationships an organization expects or does not expect from its members.
They are the result of a usually implicit, seldom explicit negotiation process
in which the organization and its members provide each other with zones of
indifference (Barnard 1968, p. 167), to which behavioral expectations can
be attributed. These zones of indifference give an up-to-date reflection of the
risk networks in which the organization is embedded. Accordingly, by keep-
ing the various positions of these networks and their specific risks under
close surveillance, consensus can be reached that everything that happens
outside these networks should be ignored, at least for the present time, but
– thanks to the “form” of communication – that it should continue never-
theless to be kept under observation. Only in this way is an organization
capable of taking action and making decisions. And only in this way can it
determine the price of this capability and how long it is willing to pay that
price.
209
12 Kap 9. Baecker 05-11-11 16.23 Sida 210
In all three routines – the cultural form of the organization, risk-bearing

networks, and the introduction of dissolvable commitments – the organiza-
tion seeks its coupling to society, under the condition of re-integrating hith-
erto excluded communication and perception potential. In each routine, a
bureaucracy that has been prematurely declared dead is reproduced and cre-
ates worlds that are no longer orientated towards the models of the worker,
the civil servant, the blue-collar employee or the manager, but which amal-
gamate all these in a new model. This model is yet without a name in the
social sciences, but can be dubbed “the hitchhiker” in honor of probably its
most accurate description to date (Adams 1979). We take the hitchhiker to
mean someone who, under the condition of his almost complete exclusion,
is nevertheless included from time to time and, what is more, in a parasiti-
cal symbiosis. That is, a hitchhiker practices inclusion in exceptional cases.10
The hitchhiker is that social figure who is no longer disappointed by the
failure of the Critical Theory to get universal inclusion accepted in organi-
zations (Luhmann 2000c, p. 392), but who accepts it as a challenge and
despite his admittedly weak, indeed powerless, position refuses to be
deprived of the opportunity to exclude all organizations from enlisting him
– and to exclude from that exclusion only a small number of these and only
in exceptional cases.
The figure of the hitchhiker, who is free to get in and out and who is at
home both at the wheel and in a free-rider position defines a counter-factual
and therefore normative version of an elite whose only proof of identity is
that it is able to elect itself, and whose only chance is to select a combina-
tion of communicative and perceptive competencies by means of a kind of
self-fulfilling prophecy, which can be coupled to different levels and hori-
zons of organization.11 The “rule of the office”, of which Max Weber (1978)
spoke in connection with the rationalization tendencies of the modern age,
and from which he believed corporations still to be free because of their
locally superior handling of knowledge, has spread to all organizations
without exception in the age of “knowledge management” and, in the midst
of the Götterdämmerung of capitalism (Deutschmann 2002, p. 253), is
about to create space for an “intelligence” which, with sound theoretical
know-how, practical experience, and reflective capability, relies no longer on
the expert but on the “networker” who is able to bond with the cadre and
open up to a watchful public (Bach 1999; Willke 2002, p. 116).
10 Luhmann (2000c, pp. 390 ff.), found that the organization completely reverses modernity’s
social norm of inclusion as the normal case and exclusion as the exceptional case: its princi-
ple of membership excludes everyone and includes only very few, based on strict selection.
11 It remains to be seen whether the picture of the clover-leaf organization drawn be Handy
(1990) – first leaf: professional core, second leaf: outsourced work, third leaf: part-time work-
ers – will survive the leap into the present century, bypassing the current structural and eco-
nomic crises.
210
12 Kap 9. Baecker 05-11-11 16.23 Sida 211
The figure of the hitchhiker can be used both to illustrate that no organi-
zation is conceivable today without reference to the society from which the
hitchhiker comes and into which he or she can return at any time; at the
same time, it tells us nothing about the competencies of this hitchhiker, the
structure of the organization, and the quality of the society. While every
organization develops its specific organizational design, other things are
going on at the same time in different places. “Society” means to bring this
insight to bear when establishing an organizational form.
Conclusion
These thoughts are all directed towards re-introducing organizational prac-
tice into organization theory, but at the same time explaining what this prac-
tice can actually accomplish. It would be wrong to assume that organiza-
tional practice in government authorities, corporations, laboratories,
armies, schools, and theaters has in the past failed to separate and re-
integrate communication and perception (the opposite is in fact true); nor
should we assume that these organizations are aware of what they have
achieved in this field. The function of theory is to create awareness of such
achievements. And the re-introduction of practice into theory consists in
making corrections in the crucial area of the organization’s reference to
society in order to be able to “inform” practice about itself. What we have
here is a complementary relationship between theory and practice in which
theory has the benefit of knowing the problems and practice has the benefit
of knowing the solutions.
This bears out the experience in dealing with a type of theory as it has
been developed in the 20th century under the various labels of pragmatism,
operationalism or constructivism. This experience suggests that theories can
never be abstract enough to be able to reach the concrete level of practice
that they endeavor to describe. Hardly any thought is more abstract than
that underlying the “form of distinction” devised by G. Spencer Brown
(1972). Hardly any thought is more abstract than that of the “operational
closure” of empirical systems which Heinz von Foerster (1981), Humberto
R. Maturana, Francisco J. Varela (1980), Niklas Luhmann (1995f) and
others have pursued. And hardly any thought is more abstract than that
underlying the operational separation and structural coupling of social
systems (communication) and psychical systems (consciousness) put
forward by Niklas Luhmann (1990g and 1995g). And yet all three ideas are
based on the possibility of reaching the concrete level of practice that theory
endeavors to describe.
The fact that this practice has a societal reference precisely at the place
where it is typical of a specific organization and where it makes this
211
12 Kap 9. Baecker 05-11-11 16.23 Sida 212
organization unique – within which decisions on how to deal with percep-

tion are taken but also where certain perception-related forms of dealing
with communication are suggested or obstructed – is the initial abstract
assumption and concrete discovery of a societal theory of organization,
which this text proposes should be pursued. Where in the past we have con-
centrated on observing individual organizations either from the perspective
of the ends they achieve or fail to achieve or under the perspective of the
restlessness which they attempt to cope with and to create, in times to come
we will have to observe how an organization handles one of its key
resources, the perceptibility of its members, its customers, and other players,
and what communication practices it develops in order both to utilize and
limit these resources.
212
13. Kap 10. Becker 05-11-11 16.24 Sida 213
PART IV
Luhmann’s Theory in the

Context of other Theories
213
13. Kap 10. Becker 05-11-11 16.24 Sida 214
214
13. Kap 10. Becker 05-11-11 16.24 Sida 215
Chapter 10
Luhmann’s Systems Theory

and Theories of Social
Practices1
Kai Helge Becker
Introduction
Various new conceptual developments have shaped the discourse of socio-
logical and organizational theorizing during the past decades. Apart from
Niklas Luhmann’s theory of social systems, some of the most innovative the-
oretical improvements in understanding the social dimensions of human
behaviour have been achieved by a branch of social theories that can be des-
ignated as “theories of social practices”. The concept of “practice” attract-
ed authors from diverse strands of theorising, as it is reflected in the works
of social theorists such as Pierre Bourdieu (1984; 2000) and Anthony
Giddens (1979; 1984), cultural theorists such as the late Michel Foucault
(1984a; 1984b), philosophers such as Theodore Schatzki (1996; 1997) and
Charles Taylor (1985; 1997), and ethnomethodologists such as Michael
Lynch (1993) and Laurent Thévenot (2001) – just to name a few. Drawing
on various concepts from different theoretical origins, above all particularly
concepts from phenomenology, hermeneutics, structuralism and Wittgen-
steinian philosophy (cf. Reckwitz 2000, for an overview), they all have
structured – partly by emphasizing quite different aspects – their work to a
significant extent around the notion of (social) practices.
Also in management and organization studies, practice-based approaches
have proven to be highly innovative: whether it comes to human resource
management (e.g. Townley, 1993), creative industries (e.g. Eikhof and
Haunschild 2005), information technology (e.g. Orliowski 2000), organiza-
tional culture (e.g. Knights and Willmott 1987), accounting (e.g. Macintosh
and Scapens 1990), career (e.g. Iellachitch et al. 2003), bureaucracy (e.g.
Dandeker 1990), technical work (e.g. Orr 1996), or to health management
1The author is thankful to Axel Haunschild and David Seidl for comments on a previous draft
of this chapter.
215
13. Kap 10. Becker 05-11-11 16.24 Sida 216
(e.g. Haunschild 2003), just to mention some of the topics covered so far,
“practice theoretical” approaches have had an enormous impact during the
past decades. Among these, especially the writings of Bourdieu, the late
Foucault, and Giddens have proved to be of particular importance.
Surely, scholars in organization studies, when drawing on these authors,
have not always addressed all aspects of the concept of practices. Also, it has
sometimes not even been the specific mode of practice-related theorising that
has received the most attention when referring to those writings. Never-
theless, the idea of practice has become an integral part of management and
organization studies. Therefore a closer look at the relation of theories of
social practices and systems theory seems to be of a certain significance.
Moreover, what makes a comparison of both strands of theorizing particu-
larly attractive is the fact that – despite their various differences – both the-
oretical approaches share two common basic assumptions on the “nature”
of the social: first, both are closely related to what has been called the “cul-
tural turn” (e.g. Alexander 1988) in social theory and hence have a “con-
structivistic” stance towards the social. They both emphasize the importance
of interpretive schemes, symbolic codes, or cognitive routines in order to
understand how we construct and process collectively shared patterns of
meaning when making sense of ourselves and the world we are engaged in.
Second, both modes of theorizing share a common interest in “de-centring
the subject”. This theoretical move, which has been put on the agenda by
(post-)structuralism (cf. e.g. Giddens 1979, pp. 9–48), has proven to be par-
ticularly significant for the development of social theory in the second half
of the twentieth century. It stresses the fact that it is not adequate to con-
sider the subject as the independent origin of social phenomena. Instead, any
analysis of the social has to take into account its collective, inter-subjective
“nature” beyond anything that subjects, agents, or actors could determine.
Clearly, the theoretical problems and the kind of reasoning that led to
these two common basic perspectives must be considered to have been part-
ly different in origin and intent. Also, obviously, both kinds of theories make
use of different vocabularies and terminologies. However, theories of social
practices, as well as Luhmann’s theory of social systems, eventually arrived
at notions of the social that take into account these basic assumptions in a
comparable way and on a level very fundamental to their modes of theoris-
ing. And yet: in the face of these common ideas about the “nature” of the
social, both strands of the social sciences ended up in developing very dis-
tinct concepts of the social itself, namely “social practices” and “social sys-
tems”. This begs the question of what made the authors opt for their spe-
cific concepts and which advantages and disadvantages result from their
decisions. Which particular strengths and weaknesses do these two strands
of social theory imply, given the fact that they have in common both a cul-
tural and a “non-subjective” thrust in their mode of theorizing?
216
13. Kap 10. Becker 05-11-11 16.24 Sida 217
In the remainder of the chapter, this issue will be addressed as following:

in the second section, our point of departure will be to characterise impor-
tant features of the notion of social practices. While, quite naturally, it will
not be possible to give a detailed account of all particularities within this dis-
course, this section provides the foundations for our comparative analysis in
the following.
The subsequent third and fourth sections compare both theoretical
approaches with respect to the two underlying assumptions on the “nature”
of the social that both strands of theorizing have in common. The third sec-
tion addresses the “constructivist” perspective shared by both approaches,
and in the fourth section we will elaborate in detail on the theoretical move
of de-centring the subject and its impact on systems theory and theories of
social practices. Based on this, the fifth and sixth sections explore how the
distinct perspectives of both social theoretical approaches translate into
partly similar, partly different concepts of some important social theoretical
fundamentals. In particular, we will analyse the relation of structure and
agency in the fifth section and address the status of the “material” and the
role of the body in the sixth section. The chapter ends up with a conclusion
in the seventh section, which provides a summary of the similarities and dif-
ferences between systems theory and theories of social practices and con-
trasts the particular strengths and weaknesses of the two approaches.
In sum, the aim of this chapter is threefold: first, to provide a better under-
standing of the conceptual logic inherent in Luhmann’s theory by relating
some of his most important conceptual decisions to the concepts developed
within the discourse on social practices; second, on that basis, to assess the
particular strengths and weaknesses of Luhmann’s approach compared with
practice-related ones; and finally, to demonstrate how both approaches can
complement each other’s perspectives for further enhancing our under-
standing of organizations.
Basic aspects of the concept of “social practices”

The aim of this section is to provide an overview of important theoretical
leitmotivs in the development of theories of social practices. This calls for
some precautionary remarks: clearly, it cannot be stated that all those dif-
ferent writers on social practices have contributed to one single coherent
body of theory. Rather on the contrary: due to the diversity of theoretical
backgrounds that have influenced practice-based theorizing, the writings of
Bourdieu, Foucault, Giddens, Lynch, Schatzki and others accentuate quite
different aspects and also differ with respect to the theoretical vocabulary
developed. Nevertheless, they have much more in common than the mere
use of the word “practices”. In fact, one can make out a set of shared theo-
retical problems, interests, and developments such that it makes sense to
217
13. Kap 10. Becker 05-11-11 16.24 Sida 218
consider these theoretical approaches a family of closely related theories.

And despite the conceptual differences that can be found in these works, one
can note a distinct shift in theorizing towards quite parallel notions of prac-
tice in the writings of these authors, which justifies speaking of a paradigm
of social practices – and probably even of a “practice turn” (Schatzki et al.
2001) in contemporary social theory.
In view of this state of affairs in practice-based theorizing, it must be
stressed that here it will not be possible to give a comprehensive account on
this family of theories that does justice to the details and subtleties of each
individual writer on social practices. Rather, we will describe the features of
practice-based approaches in an “ideal, typical” way to have a basis for fur-
ther analysis in the following sections. In doing so, we will mainly refer to
the work of Bourdieu, Giddens and the late Foucault, taking into account
their particular significance for the development of organization studies.
More detailed analyses on the concept of social practices, covering also
other theorists on social practices, can be found especially in the writings of
Reckwitz (e.g 2000; 2002a; 2003a) and Schatzki (e.g. 1996; 2001), which
are partly intended to provide an overview of the theoretical development in
the field, and partly written with a rather programmatic intention.
Given these preliminary notes, what justifies speaking of a strand of
“practice-based” theorizing? What constitutes the basic notion of practice
that the different writings on “social practices” have in common, despite
their diverse and heterogenous backgrounds?
As a point of departure, in an attempt to provide a general first approach
towards a definition, one may characterize a social practice as a “temporal-
ly unfolding and spatially dispersed nexus of doings and sayings” (Schatzki
1996, p. 89). Examples for practices in this sense are academic practices
such as writing and publishing (e.g. Bourdieu 1988), practices of sports (e.g.
Bourdieu 1984), practices of the self and practices of living a relationship
(e.g. Foucault 1984b), or practices of using information systems (e.g.
Orliowsky 1992, following Giddens). From such a point of view, it is obvi-
ous that the concept of practices refers to more than mere single actions;
rather, practices consist of a bundle of interwoven strands of behaviour. In
fact, one can make out at least three aspects that link “doings and sayings”
so as to form a “nexus”, i.e. a practice, and which must be considered fun-
damental to the specific perspective on human behaviour and the social that
the notion of social practices involves: knowledge, the role of the body, and
the material components of practices (cf. Reckwitz 2002a).
1. Practices are carried out on the basis of knowledge, on the basis of
“schemes of perception, thought and action” (Bourdieu 1990b, p. 54).
Being “knowledgeable agents” (Giddens 1984, p. 15), people interpret
themselves and the world and are engaged in processes of sense-making.
218
13. Kap 10. Becker 05-11-11 16.24 Sida 219
However, as practice theorists emphasize, the kind of knowledge

involved in practices is not mainly a “know-that”, but rather a know-
how, a practical understanding of the situation and a capability of cop-
ing with it practically. The “feel for the game” (Bourdieu 1990a, p. 9)
does not primarily consist of discursive, explicit rules of how to act,
which could be verbalized and explained. Instead, practices are rou-
tinized behaviour, rooted in the “practical consciousness” of the agents
(Giddens 1984), and drawing on “the pre-verbal taking-for-granted of
the world that flows from practical sense” (Bourdieu 1990b, p. 68).
Moreover, this knowledge is not an individual mental phenomenon, but
a collectively shared understanding, fundamentally social in “nature”,
which, for example, lays the foundation for analysing management and
work practices as a cultural phenomenon related to modern society (e.g.
Townley 2002; Orr 1996; Shenhav 1999).
2. Another important characteristic of practice-based approaches is the
emphasis on the fact that practices are closely intertwined with bodily
experiences. The knowledge involved in carrying out a practice also con-
sists of “motor schemes and body automatisms” (Bourdieu 1990b, p. 69)
and constitutes a person’s “habitus” as an embodied phenomenon.
Whether Bourdieu (1984) describes the “hexis” (body control and poise)
of the upper class, or whether Foucault (1984a) investigates pleasure and
desire, the practice mode of theorizing highlights “that ‘the body’ is ‘con-
stituted’ within some social matrix” (Schatzki 2001, pp. 2–3). From such
a point of view, Bourdieu (1990b, p. 68), drawing on the concept of
“practical understanding” addressed in the previous paragraph, states
that “practical belief is not a ‘state of mind’, … but rather a state of the
body”. And in a similar fashion – though generally putting less emphasis
on the role of the body – Giddens notes that “structure is not ‘external’
to individuals” (Giddens 1984, p. 25) and “(f)undamental to social life is
the positioning of the body” (Giddens 1984, p. xxiv). Within the dis-
course of organization studies, this aspect of practice-based theories has
contributed to both general reflections on the role of the body in an orga-
nizational context (e.g. Dale 2001) and analyses of particular organiza-
tional phenomena, such as health management (e.g. Haunschild 2003).
3. A third integral part of the concept of practices is the notion that prac-
tices also have a material component; that is to say, carrying out a prac-
tice implies making use of “things”. In addition to the mental and bodi-
ly aspects of practices and to analysing human interactions, practice
theorists also take into account the relation of bodies and minds to
“objects”, “artefacts”, and (allocative) “resources” (cf. Reckwitz
2002b). Prominent examples of this perspective can be found in
Bourdieu’s seminal analysis of tastes (1984) where he – among others –
describes the eating practices of different classes (e.g. eating fish
219
13. Kap 10. Becker 05-11-11 16.24 Sida 220
containing bones in the upper class in contrast to the eating habits of the
working class, where fish dishes without bones are preferred), or in
Foucault’s widely influential work on practices of disciplining, punish-
ment and imprisonment (1979). In management studies, the development
of practice-based approaches to understanding the use of information
technology in organizations, for example (e.g. Orliowski 2000), has
fruitfully drawn on this particularity of theories of social practices.
Summing up these three aspects, we can state that practice theorists conceive
of “practices as embodied, materially mediated arrays of human activity
centrally organized around shared practical understanding” (Schatzki 2001,
p. 2).
Systems theory and theories of social practices as

cultural theories
Having described some characteristics of practice-based social theories as a
point of departure, we now turn to the first of the two fundamental assump-
tions that systems theory and theories of social practices share: their empha-
sis on the cultural aspects of social phenomena. We will start with a look at
theories of social practice and then examine Luhmann’s theoretical position.
The “cultural turn” and theories of social practices

Since the 1970s, roughly speaking, an “interpretive” or “cultural” turn in
social theory has been under way (Reckwitz 2005a). This theoretical devel-
opment took place against the background of social sciences that used to
explain social phenomena – following Durkheim’s or Parsons’ path, for
example – in terms of shared norms and values. Also, it must be seen in con-
trast to rational-choice theories – whether of economic, or rather sociologi-
cal, origin – which analyse social order as a result of choices based on
means, ends, and utility. Against this background, researchers focussing on
“cultural” phenomena have increasingly emphasized that the social can not
be understood in full unless one pays attention to the processes of sense-
making that organize our perceptions, provide interpretations of ourselves
and impart meaning to our surroundings. In anthropology, cultural studies,
gender studies, sociology of science, history, as well as the political sciences,
more and more scholars have stressed the fact that a thorough analysis of
social phenomena must also comprise the symbolic codes, interpretive
schemes or cognitive routines that underlie and shape human behaviour. In
the field of management and organization studies, the new paradigm that
emerged has translated into organizational symbolism, narrative approach-
es, Weick’s theory of enactment, (sociological) neoinstitutionalism and
Foucauldian discourse analysis – just to name a few.
220
13. Kap 10. Becker 05-11-11 16.24 Sida 221
In view of the outline of practice-based approaches in the last section, it is

evident that also practice-based theories must be considered a part of the
cultural branch of social theories. Particularly the emphasis on “practical
understanding”, “shared knowledge” and “schemes of thinking, perceiving
and acting” clearly situates theories of social practices in the field of cultur-
al theorizing (Reckwitz 2002a). Accordingly, practice-based approaches
within the discourse of management studies have often explicitly, but at least
implicitly, taken a “social constructivist” stance towards organizational phe-
nomena. Macintosh (1994, pp. 190–192), for example, shows how Alfred
Sloan’s implementation of accounting practices at General Motors was
rooted in the context of a specific social construction of individualism. In
adopting a similar perspective, Orliowsky (1992) emphasizes the impor-
tance of social patterns of meaning for using technology, and therefore high-
lights the “interpretive flexibility of technology”. Further examples of such
a “constructivist” perspective can be found in Orr’s (1996) “ethnography”
of modern work practices and Townley’s (1993) description of how HRM
practices construct predictable and calculable employees.
Systems theory as cultural theorizing

In contrast to this, it may be not so obvious at first sight to understand also
Luhmann’s theory as a branch of cultural theorizing. How could a theory
that places the rather “technical” term “system” at its conceptual core ever
be a cultural mode of theorizing? And in fact, the concept of “culture” itself
does not play an important role within Luhmann’s theoretical framework.
But, when starting to look beyond the mere use of the term “culture”,
Luhmann’s systems theory appears to be so genuinely cultural an approach
that one could even be inclined to ask if it is about anything other than cul-
tural phenomena at all. Although here we do not intend to go that far, the
point at stake is that an understanding of the social as a cultural phenome-
non lies at the heart of Luhmann’s way of thinking.
The theoretical concept that is crucial for imparting a cultural perspective
to the theory of social systems is Luhmann’s notion of distinction. In fact, it
is exactly this theoretical building block in systems theory that gives rise to
a conceptual parallel to what other cultural theories describe as “patterns of
meaning”, “cognitive codes” or “schemes of thought, perception and
action” (see also Reckwitz 2005b). To make this point, it is helpful to recall
some basic concepts of the theory of social systems: situating distinctions at
their core (cf. also Seidl and Becker 2006), Luhmann’s theoretical enterpris-
es start off with the distinction of system and environment, and – on the
basis of this first distinction – later introduce the distinction of social and
psychic systems. One of the fundamental aspects that these two types of sys-
tem – and only these two types of system – have in common is the fact that
they process meaningful information (Luhmann 1995f, p. 59). Strongly
221
13. Kap 10. Becker 05-11-11 16.24 Sida 222
drawing on phenomenology, Luhmann emphasizes that all elements of psy-

chic and social systems take the form of meaning (Luhmann 1995f, p. 63)
and permanently generate new “horizons of meaning” (Husserl) to orient
themselves. On its part, each element of psychic and social systems is an
observation made up by a distinction (and an indication). In the case of
social systems, these distinctions constitute the topic of what is communi-
cated (the selection “information” in terms of Luhmann’s theory of com-
munication). In the case of psychic systems, they make up the content of the
perceptions, thoughts, and emotions that “take place” in the minds of the
participants of a social system. In other words: social and psychic systems
process meaningful information by applying distinctions. On the basis of
these distinctions, they construct, i.e. “enact” (Weick), themselves, their
boundaries, their identities, their environments, and their communications
(and thoughts/emotions/perceptions, respectively) during the process of
autopoiesis.
While these aspects already show that Luhmann’s perspective obviously is
a specific version of constructivism (cf. Luhmann 2002c), we have to be
more precise when speaking of “culture” in a strict sense. In fact, culture,
made up by interpretive schemes or patterns of meaning, is primarily a struc-
tural phenomenon, that is, a phenomenon beyond single, individual appli-
cations of certain distinctions. Therefore, the mere fact that psychic and
social systems permanently apply distinctions when observing does not qual-
ify for the designation “culture” yet. However, there are, of course, distinc-
tions that psychic and social systems apply repeatedly and make use of fre-
quently. In this case, a – more or less stable – social structure evolves, which
basically consists of what one can call patterns of meaning, interpretive
schemes or cognitive codes.
In view of this theoretical perspective, it does not come as a surprise that
large parts of Luhmann’s œuvre are actually cultural analyses, i.e. – in sys-
tems theoretical terms – they are devoted to the study of those distinctions
that play an important role in modern society. Two fields of his research
especially stand out here: first, historical investigations on what Luhmann
calls the “semantics” of society. With this term he refers to the stock of pat-
terns of meaning, the hoard of concepts (distinctions) that society holds
ready for frequent use. Among Luhmann’s writings, there are, for example,
investigations on how the different specific notions of love have shaped inti-
mate relationships in the past centuries (Luhmann 1986a), on the interpre-
tive scheme we refer to as “nation-state” (Luhmann 1997a, pp. 1045–1054),
on the notion of nature and how it relates to both the scientific, technical
discourse and the ecological movement (Luhmann 1999a pp. 9–30), and on
how specific forms of semantics structured interactions in the upper class
during the 17th and 18th centuries (Luhmann 1993e, pp. 72–161).
222
13. Kap 10. Becker 05-11-11 16.24 Sida 223
A second important example of the cultural perspective inherent in Luh-

mann’s works can be found at the core of his theory of society. According to
Luhmann, society is mainly structured into functional sub-systems, such as
economy, academia, politics, arts, the legal system, and religion. The com-
munications within each of these systems are based on a distinct binary
“code”. In the case of the economic system, for example, the code is “pay-
ment/non-payment”, in the case of the academic system it is “true/false”,
and the legal system is organized around the code “legal/illegal”. It is not
possible to go into the details here (cf. Luhmann 1989; or Chapter 1 in the
present volume), but, generally speaking, it is on the basis of these codes that
Luhmann explains the patterns of meaning, the interpretive schemes shap-
ing and structuring the communications in the different systems. In doing so,
Luhmann imparts to his sociological theory a genuinely cultural thrust.
Moreover, during the past years especially the systems theorist Dirk
Baecker (cf. also his chapter in this book) emphasized the fact that distinc-
tions lie at the heart of systems theory. He considers the particular strength
of a systems theoretical point of view to be a focus on how, i.e. by means of
which types of distinctions, social systems come into being and differentiate
themselves from an environment. And Wil Martens (forthcoming 2006) has
recently elaborated on an analysis of organizations as a cultural phenome-
non from Luhmann’s perspective. Finally, even Luhmann, who – for reasons
beyond the scope of this chapter (cf. Luhmann 1999a) – normally does not
prefer to use the term “culture” in this context, after all states that patterns
of meaning, when being preserved and reapplied, “condense into what could
be called culture in a comprehensive term” (Luhmann 1997a, p. 409; my
translation, italics in the original).
In sum, despite the differences of both modes of theorizing, it can be said
that theories of social practices as well as systems theory agree on the fact
that social phenomena must be analysed primarily as cultural phenomena.
However, it must be noted that theories of social practice stress the practi-
cal character of the shared understanding on the basis of which we relate to
ourselves and get along with our surroundings. In contrast to this, Luhmann
rather emphasizes that these patterns of meaning have a dynamics of their
own, somehow detached from the necessities of everyday life. This particu-
larity of systems theory will be further addressed below.
The de-centring of the subject

Perceiving the social primarily as a cultural phenomenon immediately calls
for thorough concepts of where the patterns of meaning, symbolic codes and
cognitive routines are located. When it comes to addressing this question,
one line of reasoning has proven to be especially influential in the past
223
13. Kap 10. Becker 05-11-11 16.24 Sida 224
decades: the theoretical move of “de-centring the subject”. In the following,

we will first address the particularities of this specific perspective on theo-
rizing the social and consider its impact on theories of social practices. In
contrast to this, we will then examine which theoretical accentuations made
Luhmann choose his specific way of de-centring the subject, and what the
consequences of his choice are. It will turn out in the end that Luhmann’s
social theoretical concepts must be regarded as being both more radical and
more traditional than their practice-based counterparts.
The de-centring of the subject and its impact on theories of

social practices
Since Descartes’ momentous distinction between an “inner”, mental sphere
(res cogitans), and an “outer”, material sphere (res extensa), a line of
thought emerged, which went via Kant’s critique of pure reason to Husserl’s
phenomenological concept of the mind and Sartre’s notion of a subject free
to choose itself. Having finally translated into phenomenological, interpre-
tive and ethnomethodological sociology, this tradition of theorizing consid-
ered the subject as the – more or less – independently given origin of social
phenomena. Consequently, these approaches emphasized the idea that
analysing the meaning the actors themselves ascribe to a situation, and
investigating the conscious mental processes involved lie at the heart of soci-
ological concern.
While these notions provided the foundations of the cultural mode of the-
orizing described above, they also attracted criticism. Especially structural-
ism, evolving from Ferdinand de Saussure’s seminal analysis of language
(1966) and particularly his distinction of langue and parole, made a strong
point against the concept of a pre-given subject that is – more or less – trans-
parent to itself. Particularly influential in this respect were Lévi-Strauss’
structural analyses of myths (e.g. 1968), which aimed at revealing the
“unconscious psychical structures”, the “unconscious teleology of the
mind”. Lévi-Strauss focussed his analyses on a non-conscious, supra-indi-
vidual, collective level of cognitive codes (“langue” in de Saussure’s terms)
that shape human behaviour, so to speak, “behind the subjects’ back”. By
abandoning a subject-centred perspective, he intended “to show not how
men think in myths, but how myths think themselves in men, and without
their being aware of it” (Lévi-Strauss 1964, p. 20; my translation) and there-
by challenged the notion of a self-conscious subject.
Michel Foucault, in the famous last chapters of his investigation on The
Order of Things (1966), even went a step further. From his (early) point of
view, the entire concept of the subject must be considered a mere effect of
discourse, i.e. a result of a specific collective structure of knowledge (“epis-
teme”) that has been underlying the modern academic discourse since the
19th century. Accordingly, Foucault’s “archaeological” method shifts the
224
13. Kap 10. Becker 05-11-11 16.24 Sida 225
attention to “non-mental”, somehow material, discursive statements

(“énoncés”) “outside” the mind (1969). In doing so, it opens up the way to
analysing – without reference to “the subjective”, and in their own right –
those collectively shared patterns that make up human culture. Similarly,
around that time, other writers, such as Jacques Derrida (e.g. 1979), Jacques
Lacan (e.g. 1977) or Paul Ricoeur (e.g. 1974), also criticised the concept of
the cogito and underlined the need for a “de-centring of the subject” in
social theory.
The writers elaborating on a theory of social practices reacted to these
– quite controversial – reflections in their own way, which was partly criti-
cal, partly affirmative. Following structuralism, theories of social practices
stress the significance of collectively shared interpretive schemes that basi-
cally shape what makes up the subject. Instead of assuming a free, pre-given
subject, they focus on the unconscious, primarily non-discursive, supra-
individual structures, i.e. the “practical consciousness” (Giddens) or the
“habitus” (Bourdieu). Socially constituting the subject, these structures con-
sist of tacit, implicit knowledge and lay the very foundations of agents’ per-
ceptions, interpretations and even bodily behaviour. Moreover, these pat-
terns of meaning are not only “structures structurantes”, i.e. structuring
structures, but also “structures structurées”, i.e. structured structures
(Bourdieu). Therefore, in reflecting class-related social differences and gen-
erating social fields, for example, they possess a logic of their own, beyond
anything agents could determine.
However, practice-based theories, in contrast to structuralism, acknowl-
edge that agents can partly become aware of and reflect on the patterns of
meaning shaping their behaviour. They possess a “discursive conscious-
ness”, speaking in Giddens’ terms (1984), and can engage in the “care of the
self” (Foucault 1984b). Yet, the routinized, embodied character of social
practices prevents them from being entirely transparent to the agents. As the
late Foucault (1984a; 1984b) demonstrates, their desires and pleasures are
socially shaped on a very fundamental level. Similarly, Bourdieu (1990b,
p. 69) emphasizes that agents, having acquired “social necessity turned into
nature, converted into motor schemes and body automatisms, … never
know completely what they are doing … [and therefore] what they do has
more sense than they know”.
Going back to Descartes’ distinction of the “subjective” res cogitans and
the “objective” res extensa, it is possible to sum up the theoretical strategy
theories that practice-based approaches use in de-centring the subject.
Instead of situating social phenomena in an inner, mental sphere (as phe-
nomenologists would do), or situating them in an outer, material sphere (as
Foucault’s – early – archaeological approach suggests), they analyse the
social-cultural world as being located in the social practices themselves. In
doing so, they make the Cartesian distinction collapse (Reckwitz 2002a;
225
13. Kap 10. Becker 05-11-11 16.24 Sida 226
2003b): social practices, as a “nexus of doings and sayings” (Schatzki 1996,

p. 89), are collectively shared patterns of both meaning and bodily behav-
iour, which are neither transparent to nor made up by self-conscious sub-
jects. Instead, being embodied routines, they are non-discoursive, primarily
tacit, “unconscious” and supra-individual patterns of meaning and bodily
behaviour which “mould” and constitute individual agents and lie at the
foundation of their particular “doings and sayings”.
Niklas Luhmann’s move to de-centring the subject

Niklas Luhmann’s theoretical œuvre takes up the issue of the de-centring of
the subject in a different way, which must be considered to be partly more
radical, partly less radical than the mode of theorizing in the practice-based
approaches. Luhmann also comes to the conclusion that social phenomena
cannot be located in a pre-given subject. However, in drawing strongly on
the legacy of Husserlian phenomenology, Luhmann’s point of departure and
main concern is the question of whether it is possible to conceptualize the
social as an intersubjective phenomenon.
Edmund Husserl’s famous problem of intersubjectivity (e.g. 1913) arrives
when one intends to develop a theory of the social in proceeding from the
assumption of pre-given subjects. In order to shed light on what made
Luhmann choose his perspective on social phenomena, this problem will be
briefly sketched according to Alfred Schütz’s (1975) criticism of Husserl.
Also Alfred Schütz (e.g. 1974), in drawing on the concept of “life-world”,
tried to solve the problem of intersubjectivity inherited from Husserl. It
would be beyond the scope of this chapter, however, to elaborate on the
respect in which Luhmann (cf. 1995f, pp. 80–81, 145–146; 2002f) was not
satisfied by the concepts proposed by a phenomenological sociology.
Proceeding from the assumption of a pre-given subject which, in view of
phenomenological analysis, is well aware that the way it perceives the world
is based on the constructions of its own mind, the first theoretical difficulty
is as follows: how can a subject, from its purely “inner”, mental point of
view, conclude that there are other subjects experiencing the world in a sim-
ilar way? Husserl suggested a solution that mainly consists in the idea that
the subject could become aware that its body is essential for its perception
of the world, i.e. the subject could have a “kinaesthetic consciousness”. On
that basis, it could make out other bodies in the “world outside its mind”
and conclude from the fact that their behaviour is similar to its own body’s
behaviour, that there may also be other subjects in the world apart from
itself. Second: at best, this is how all other subjects would experience their
surroundings, i.e. they all would experience each other as an “alter ego”.
Still, this is not sufficient to conceptualize social phenomena, namely collec-
tively shared, supra-individual patterns of meaning, which are relevant to all
subjects. Even if, in a third step, all subjects concluded that all other
226
13. Kap 10. Becker 05-11-11 16.24 Sida 227
“subjects/bodies” base their behaviour on patterns of meaning they all col-

lectively share, one would still not come closer to a theory of the social. For,
in proceeding in this way of theorizing, the world of social phenomena
would still be the perception or “fiction” of single individual subjects and
not an intersubjectively shared sphere of collective patterns of meaning.
In short, it is from such a line of reasoning that Luhmann arrived at the
conviction that a programme of research that attempts to develop a theory
of social phenomena “out of” the inner mental sphere of pre-given subjects
must be considered infeasible. According to Luhmann, the problem of inter-
subjectivity in a strict sense cannot be solved.
Therefore, his systems theoretical approach proceeds in a different way.
Developing the concept of the psychic system, Luhmann somehow preserves
and takes up the legacy of Husserl’s notion of the mind. However, Luhmann
is convinced that it is impossible to overcome the problem of intersubjectiv-
ity when trying to situate social phenomena inside the mental sphere of the
subject. This is why he conceives of the world of social phenomena as a gen-
uinely social sphere in its own right, as realm of phenomena sui generis, and
introduces the concept of the social system accordingly. Being detached from
the subjective mental sphere of psychic systems, social systems possess a
dynamics of their own and account for what could be called the “inter” of
intersubjectivity.
In order to fully appreciate what Luhmann’s theoretical endeavour is
aimed at, it is helpful to recur to what was stated above about the Cartesian
distinction of an inner, mental, and an outer, material sphere. Theories of
social practices react to this distinction by making it collapse into the con-
cept of social practices. In contrast to this, it can be stated that Luhmann’s
bold concept of the social system attempts to surpass this distinction by
introducing a third sphere beyond the mental and the material sphere: the
sphere of the social itself.
Consequences of Luhmann’s way of de-centring the subject

In order to further illustrate Luhmann’s way of de-centring the subject, three
consequences of the systems theoretical perspective will be addressed. First,
social phenomena are independent from mental phenomena and are not
conceived as originating in them. Being autopoietic, i.e. self-(re)producing
systems, social systems themselves, instead of individual subjects, are the
origin of social phenomena. “Only communications can communicate”, as
Luhmann’s famous statement goes (2002g, p. 156). In other words: what
counts as a social phenomenon, i.e. a communication, cannot be determined
by the minds of individuals, but is “decided” by the context of other social
phenomena. Moving a piece on a chessboard, for example, becomes a com-
munication in the interaction system “playing chess” – the communication
“checkmate”, for instance – only with reference to the rules, other moves
227
13. Kap 10. Becker 05-11-11 16.24 Sida 228
etc. within the game, no matter what the individual player had in mind
about the move. In organizations, for example, there are situations where an
organizational member has done just business as usual and finds out later
that the organization treats this behaviour as a deliberate decision, attribut-
ing responsibility, rewarding merit or calling for consequences. It can be said
that Luhmann, in choosing such a perspective on the social, takes up the
Durkheimian maxim that social phenomena can only be explained ade-
quately by being related to other social phenomena.
Second, social phenomena also possess a dynamics of their own. Not only
can they be traced back to other social phenomena, but they can also follow
their own “rules”, their own “logic”. Psychic systems can only “irritate”
social systems, but not determine how a social system evolves. As experience
in the ups and downs of organizational life shows, organizations in particu-
lar have a certain degree of freedom to attribute communications (i.e. deci-
sions) to different persons, groups or committees, thereby electing them as
heroes or scapegoats. On the basis of this, specific systems theoretical focus
on the internal dynamics of social systems, management researchers (e.g.
Wimmer 2004; see also Luhmann in this book) have developed a systems
theoretical approach to management consulting. Taking into account that
consultants (i.e. psychic systems) cannot directly influence and control the
client system (the organization), this perspective allows, for example, for
analysing the well-known problems of translating consultants’ recommen-
dations into organizational action.
Third, what is more, as a result of their internal dynamics, social systems
develop their own structures. Common distinctions in academia, such as
“formal” versus “informal organization”, or “subjectivism” versus “objec-
tivism”, as well as the social conventions about when and how they are to
be used, possess a “life of their own”. Being manifest in the communica-
tions, i.e. publications of the academic system, they are independent of the
appreciation of single individual authors and, according to systems theory,
“exist” as structures, i.e. patterns of meaning, on a genuinely social level.
Similarly, organizational structures (decision programmes, channels of com-
munication, etc.) must be attributed to the level of the social, since they will
not (necessarily) change when an organizational position is filled with a dif-
ferent person.
Luhmann’s perspective as both radical and traditional

Having arrived at this point of view, we could say, in the first instance, that
Luhmann’s approach to social phenomena is more radical than the practice-
based approach: in introducing the social as an autopoietic nexus of com-
munication, in the way that it is clearly separated from psychic systems, he
has de-centred the subject much further. At the core of social phenomena,
there is the autonomous self-reproduction of social systems, which can only
228
13. Kap 10. Becker 05-11-11 16.24 Sida 229
be “irritated” by psychic systems on the basis of “interpenetration” (cf. the

introductory chapter), but not determined. Social systems are operationally
closed: they “decide” on their own about what counts as a social phenome-
non, possess a dynamics of their own, and develop their own structures. In
contrast to this, practice-based approaches do not consider social phenome-
na to have a dynamics of their own: having also performed the move of de-
centring the subject, theories of social practices agree with Luhmann that
social phenomena lie beyond what individuals could determine. Social phe-
nomena, however, come into being only by actors’ carrying out certain prac-
tices.
Moreover, even compared to Lévi-Strauss’ famous dictum that he intends
to show “how myths think themselves in men” (see above), Luhmann’s per-
spective must be seen as a radical one: while Luhmann agrees that commu-
nications “communicate themselves”, he does not share the perspective that
they communicate “in men”, for he has placed communications in a sphere
on their own, entirely outside the sphere of the mental. However, it may be
advisable to note here that Luhmann – in contrast to Lévi-Strauss – does not
claim that communications are caused by other communications in social
systems. Instead, his perspective indicates a functional approach (Luhmann
1995f, pp. 12–58). That is to say, what constitutes a communication from
Luhmann’s point of view is the fact that it fulfils a communicative function.
In other words: communications are communications insofar as they func-
tion as communications in relation to other communications.
In the end, with respect to this perspective of an autonomous sphere of the
social, Luhmann comes very close to the genealogical period of Foucault’s
works. By further developing the archaeological mode of analysis, which
aimed at investigating the (rather static) structures that underlie discourses,
Foucault’s genealogical approach emphasizes how discourses follow their
own internal dynamics and logic. On the basis of his non-subjective concept
of power, Foucault conceives of social phenomena as a result of non-
personal relations of forces, which evolve out of the specific “rationality”
underlying discourses. His famous investigations on the Birth of the Prison
(1975) or the Will to Knowledge (1976), for example, explore how dis-
courses are shaped by “great anonymous strategies”, which do not “result
from the choice or decision of an individual subject” (Foucault 1976, p. 95).
This radical aspect of Luhmann’s way of de-centring the subject, however,
is only one side of the coin. For, from a different point of view, Luhmann
does not carry the de-centring of the subject as far as theories of social prac-
tices do. Still adhering to the legacy of Husserl’s concept of the mind (cf.
Luhmann 2002f), Luhmann introduces the concept of the psychic system in
order not to lose sight of the important role mental phenomena play for the
emergence of the social. According to Luhmann, social phenomena cannot
take place without being stimulated or triggered by psychic systems, which
229
13. Kap 10. Becker 05-11-11 16.24 Sida 230
are operationally closed, autopoietic systems in their own right, detached

from the realm of the social. Certainly, this should not be interpreted as if
Luhmann had taken up the notion of a subject that is free to choose by itself
and entirely independent of social phenomena. In fact, social and psychic
systems coevolve with each other and permanently restrict each other’s
degrees of freedom by being “coupled” on a structural level, by means of
language, for example. Nevertheless, we cannot help stating that in preserv-
ing a distinct subjective, mental sphere – i.e. the psychic system – Luhmann’s
social theory appears to be less radical than practice-based theorizing – one
could even say: more traditional.
Let us summarize what has been said in the preceding sub-sections about
the role the subject plays in systems theory and practice-based theories. In
the end, both theories agree that social phenomena are not mental phenom-
ena. Both have “externalized” the social and placed it beyond a mental, sub-
jective sphere, thereby taking into account that it is not adequate to consid-
er independent, pre-given, self-conscious, transparent subjects as the origin
of social phenomena. In de-centring the subject, they both agree that the
mental sphere is not independent of social phenomena, but, instead, that the
social strongly moulds and shapes what used to be perceived as a free sub-
ject. In view of these common grounds, also the difference of both approach-
es has been elaborated on, especially with regard to the question of how they
deal with the Cartesian dualism between the inner, mental, and the outer,
material sphere. Given this background, the following section turns to the
question of how this difference in perspective affects the way in which both
theories conceptualize some basic vocabulary of social theory, namely struc-
ture and agency, the status of the material, and the role of the body.
Structure and agency

One of the most fundamental aspects of any social theory is the question of
how it addresses the topics of structure and agency, and how these topics are
related (Walsh 1998). Especially this issue is often exploited as an occasion
for highlighting rather polemically a certain approach as the only one capa-
ble of adequately describing social phenomena. This section will not follow
this path and will instead explore the notions of structure and agency in a
symmetrical way that does justice to both modes of theorizing. Therefore,
after some preliminary remarks, in the following we will first emphasize the
similarities of theories of social practice and systems theory and then turn to
what distinguishes both modes of theorizing.
230
13. Kap 10. Becker 05-11-11 16.24 Sida 231
Preliminary note on the structure/agency dichotomy in both

strands of theorizing
When addressing the issue of structure, agency, and their relation, one
immediately encounters the differences in concepts and vocabularies
between systems theory and theories of social practices. Regarding the lat-
ter, examining the issue of “structure versus agency” starts in a quite
straightforward way: social practices are carried out by agents. Therefore, in
addressing the question of how practices are carried out, practice theorists
have elaborated on a distinct notion of human agency and its relation to
structures.
In contrast to this, the social theoretical dichotomy of structure versus
agency does not translate directly into the concepts of the theory of social
systems. In fact, in proposing the concept of the social system, Luhmann
explicitly intended to overcome the concept of human agency (cf. Luhmann
1995c): agents are not the basic unit of action anymore. Therefore, in order
to be able to adequately compare both strands of theorizing, it is necessary
to look out for concepts that take the place of the idea of agency. In other
words: we will have to find concepts that – with respect to the theoretical
logic inherent in the framework of systems theory – function as equivalents
of the structure/agency relationship.
For this purpose, it is important to recall Luhmann’s way of de-centring
the subject. It has been stated above that Luhmann has abandoned the
notion of an independent subject as the origin of social phenomena by intro-
ducing the concept of the social system while keeping the traditional idea of
a mental sphere, i.e. the psychic system: the notion of the subject has been
abandoned in favour of the distinction between social and psychic systems.
This means that the concept of the subject is, so to speak, divided into two
parts and replaced by the inner, mental sphere (the psychic system), and the
outer, social sphere (the social system). Therefore, structures appear in sys-
tems theory twice: as (mental) structures of psychic systems and as (social)
structures of social systems. Accordingly, what comes closest to the concept
of action (in contrast to structures) within Luhmann’s framework is the
single operations of social and systems; that is, communications in social
systems, and thoughts, emotions, and perceptions in psychic systems, respec-
tively. From this point of view, it seems to be justified for the purpose of our
analysis to consider both psychic and social systems as “agents” and com-
pare the relation of operations and structure in systems theory with the rela-
tion of actions and structure in theories of social practices. In fact, once one
has got acquainted with this rather unusual perspective, it is possible to
make out a concept of structure versus operations in systems theory that is
analogous to Giddens’ and Bourdieu’s notions of structure versus agency.
The similarities of both strands of theorizing that become visible in this per-
spective will be addressed in the following.
231
13. Kap 10. Becker 05-11-11 16.24 Sida 232
Similarities in the concepts of structure and action/operation

Of the three main exponents of practice-based theorizing on which this
chapter draws, Anthony Giddens (1984) probably presented the most trans-
parent account of agency. According to his notion, agents act primarily on
the basis of a “practical consciousness” and a “discursive consciousness”
(the concept of the “unconsciousness” will be skipped here, because – from
a social theoretical point of view – it does not play a central role in practice-
based theorizing nor in systems theory). The concept of practical conscious-
ness refers to the “reflexive monitoring of action” (Giddens 1979, p. 56)
that takes place when agents skilfully carry out the routinized behaviour
lying at the heart of the notion of social practices. In contrast to this, the
concept of discursive consciousness expresses the fact that agents can partly
“rationalize”, i.e. examine their behaviour. This leads to a theoretical setting
where agents mainly carry out routinized behaviour, which is only partly
open to reflection and deliberate change.
Similarly, Pierre Bourdieu’s (1977) perspective takes into account that
agents act on the basis of routines rooted in their “habitus”. Understanding
“le sens du jeu”, agents – in creative, skilful improvisation – apply these rou-
tines to meet the necessities of daily life within the economic constraints of
a class-based society. In doing so, they reproduce the very social structures
which inform their behaviour. It must be noted, however, that Giddens
places more emphasis on the intelligence and competence of agents, while
Bourdieu stresses more the social, embodied constraints imposed by the
habitus (cf. Tucker 1998, p. 71).
In any case, the practice theoretical concept of routines that are skilfully
applied by agents gives rise to a notion of structure that explicitly takes a
stance against voluntarism as well as structural determinism: from the prac-
tice theoretical perspective of a “duality of structure” (Giddens 1979, p. 69),
structures both enable and constrain actions and are considered both a
medium and an outcome of action. Analogously, Bourdieu (1977, p. 83)
states that “structures are themselves the product of historical practices and
are constantly reproduced and transformed by historical practices whose
productive principle is itself the product of the structures which it conse-
quently tends to reproduce”. Moreover, structures, being “virtual” in
“nature”, do not possess an ontological status on their own, but exist only
“in its instantiations in … practices” (Giddens 1984, p. 17), that is to say,
only insofar as they are applied in the course of actions.
In a theoretical move parallel to Giddens’ distinction between the practi-
cal and the discursive consciousness, Luhmann also distinguishes between
different levels of how systems examine or “observe” themselves. However,
because Luhmann’s conceptual efforts are much more detailed and complex
(1995f, pp. 437–477), it is not possible here to give a comprehensive
overview of all the concepts related to systems’ self-observations, such as
232
13. Kap 10. Becker 05-11-11 16.24 Sida 233
“self-description”, “reflexivity”, “reflection” and “rationality”. Yet, it may

be interesting to mention two “extreme points” within the range of concepts
Luhmann offers.
On a very rudimentary level, each social system performs a basic kind of
self-observation, for “every communication declares, whether consciously
or not, thematically or not, that it belongs to a system” (Luhmann 1995f, p.
456). This basic self-observation just addresses the fact that communications
cannot be communications without referring to the communicative context
that constitutes them as communications. In contrast to this, other ways of
self-observations are more “demanding”, in terms of what is required to
make them happen. The most complex and demanding kind of self-obser-
vation Luhmann distinguishes is rationality. “(T)his idea decrees that a sys-
tem must control its effects on the environment by checking their repercus-
sions upon itself if it wants to behave rationally” (Luhmann 1995f, p. 475).
While interaction systems can normally be expected not to examine them-
selves so thoroughly – often they do not even describe themselves in relation
to their environment (“we”/“they”) – one may assume that organizations
sometimes at least try to be rational in the sense of this definition.
In any case, however, it holds that the mere complexity and internal
dynamics of a social system prevents any psychic system from being able to
determine what is happening in the social system. Moreover, psychic systems
can observe social systems only with respect to their own distinctions.
Consequently, they are never able to see through and entirely comprehend a
social system’s internal “logic”. Whatever the extent of rationality involved
may be, psychic systems can only “irritate” the social system and never be
in control of how the process of communication will be going on. Therefore,
from the point of view of a psychic system as an “agent”, Luhmann – for
different reasons, however – would also agree to Bourdieu’s conclusion that
“agents never know completely what they are doing (and) … what they do
has more sense than they know” (Bourdieu 1990b, p. 69).
Besides the fact that systems theory – similarly to Giddens’ notion of
agency – distinguishes also different degrees of self-examination, there is
another similarity in both strands of theorizing, concerning the notion of
action/operations and structure: the relation between structure and action in
theories of social practices is equivalent to the relation between operations
and structure in systems theories. Structures are structures only insofar as
they function as structures. They are virtual in existence and do not have any
ontological status beyond the mere operations of the system they structure
(Luhmann 1995f, pp. 278–356). Moreover, Luhmann considers the relation
of structure and operations a relation of mutual dependence. By explicitly
referring to Giddens, Luhmann perceives the relationship of structure and
“agency” (operations) as “one of reciprocal enabling” (Luhmann 1995f,
233
13. Kap 10. Becker 05-11-11 16.24 Sida 234
p. 293): also the structures of psychic and social systems are both a medium
and an outcome of the operations taking place in the systems.
Differences in the concepts of structure and action/operation

However, despite all these similarities against the background of distinct
concepts and theoretical vocabularies, there are four aspects that especially
mark a difference between both strands of theorizing. Again, these four
aspects must be considered a direct result of Luhmann’s genuine way of de-
centring the subject. Therefore, they all must be understood in the context
of Luhmann’s decision to conceptualize the social world as a phenomenon
in its own right.
First: it must be stressed that there is a difference between systems theory
and practice-based approaches with regard to the question of what counts
as a social phenomenon. In the perspective of theories of social practices,
each single action is social in a very fundamental sense. Whenever an agent
carries out a practice, she or he draws on knowledge and bodily routines
that are not just individual characteristics or “properties”. Instead, the very
notion of practices involves the idea that the schemes and principles gener-
ating actions (the “habitus” in Bourdieu’s terms) are collectively shared.
Hence, all practices are conceived as inherently social in “nature”.
In contrast to this, Luhmann’s concept of the social system leads to a clear
distinction between social and non-social operations. Only those phenome-
na that occur in the communicative context of a social system are social phe-
nomena, according to Luhmann. Brushing one’s teeth in the bathroom
alone, for example, is not considered a social phenomenon from the per-
spective of systems theory (Luhmann 2002b), while it is clearly understood
as carrying out a certain social (!) practice from the point of view of prac-
tice-based approaches. By definition, in systems theory, brushing one’s teeth
is social only if it is part of a communicative process: when parents explain
it to their children, for example, when people share a bathroom and com-
municate via brushing their teeth (e.g. by making fun of each other about it),
or when it is shown on TV. With regard to this perspective, Luhmann comes
close to Max Weber (1978, p. 4; italics added according to the German orig-
inal), who defined a social action as an action whose “subjective meaning
takes account of the behaviour of others and is thereby oriented in its
course.”
However, it would be misleading to over-emphasize this difference in per-
spective between the two strands of theorizing. Also, Luhmann would not
deny that practices carried out in solitude are shaped or constituted by social
phenomena: since psychic and social systems coevolve with each other, the
structures of psychic systems (including their memories) can be expected to
reflect strongly the social conditions under which they evolved (cf. Luhmann
1990c, p. 49). Therefore, it is rather only a matter of definition whether one
234
13. Kap 10. Becker 05-11-11 16.24 Sida 235
considers this kind of action as a social phenomenon or a phenomenon that

is brought about by social phenomena. Yet, this difference in perspective
further illustrates Luhmann’s efforts to conceptualize a social theory that
provides the social with a distinct, somehow “pure”, level of analysis of its
own.
Second: so far, this section has only focussed on the systems theoretical
notion of structure/agency with respect to the relation of operations and
structures in psychic and social systems. While this has proved to be a fruit-
ful perspective to discover similarities of both strands of theorizing, there is
something more Luhmann’s theory can offer when it comes to the notion of
agency. Apart from the operations of psychic systems (thoughts, perceptions
and emotions) and social systems (communications), Luhmann himself has
also developed a unique concept of action.
According to this concept, actions are not certain activities carried out by
actors, agents, or subjects, but a way of how social systems describe their
own processes of communication: while the process of autopoiesis is under
way, social systems create bundles of behavioural expectations towards the
psychic systems participating in a communication. These bundles of expec-
tation, called persons, function as addresses for communications. In doing
so, they facilitate the (re-)production of social systems by providing “sup-
port points” (Luhmann 1995f, p. 167) for further connective communica-
tions. In this perspective, “persons” are social, i.e. communicative construc-
tions by social systems. They are images or “fictions” that “condense” dur-
ing a communicative process and are used by the social system as points of
reference for communications. Actions, in Luhmann’s terms, occur when-
ever communications are attributed to these persons, by using pronouns and
expressions such as “I mean” or “you said”, for example. In other words:
communications usually describe themselves as being uttered by a person;
they “pose” as actions in order to facilitate and steer the process of a social
system’s autopoietical (re-)production.
By conceptualizing persons as constructions by social systems, Luhmann
comes close to Foucault’s (1966) perspective on the subject as an effect of
discourse. In Luhmann’s theoretical setting, however, the mutual interplay of
mental operations by psychic systems, communications by social systems,
and the communicative construction of persons, gives rise to a more com-
plex perspective in the end: whenever communications or actions take place,
they must be considered a result of both (a) mental contributions by psychic
systems, which “irritate” a social system and “trigger” communications,
and (b) the autopoiesis of the social sphere, which constitutes communica-
tions as communications and ascribes them to persons. From this point of
view, one could state that, in the end, it is the mutual interplay of social and
psychic systems that Luhmann has put in place of the traditional idea of the
subject, and that makes up his genuine concept of agency.
235
13. Kap 10. Becker 05-11-11 16.24 Sida 236
Such a perspective opens up a variety of options for – both theoretical and

empirical – research. Among other things, it provides a general framework
for investigating the relation of individual mental phenomena and the world
of the social, which could be stimulating for theories of socialization, for
example, or theories of organizational membership. Moreover, Luhmann’s
genuine concept of action lays the foundation for analysing, for example,
under which circumstances and by means of which constructions social sys-
tems (organizations) attribute communications (decisions) to whom and
with which effect (cf. also Knudsen in this book on decision-making in
reform committees).
Third: the analytical advantages provided by separating psychic and social
systems and conceptualizing the world of the social as a sphere in its own
right, also give rise to a perspective that is partly questionable: the interplay
of psychic systems is always mediated by the internal dynamics of social sys-
tems. From this point of view, human interaction is never an immediate one.
Although this notion of interaction obviously has its benefits, as seen above
(cf. also Chapter 7 of the present volume, for example), it downplays an
aspect highlighted by theories of social practices: every interaction takes
place against the background of collectively shared understanding and
bodily routines, i.e. the stock of social practices provided by society. Such a
perspective sees action as embedded in a common “life-world” (Schutz and
Luckmann 1974) and emphasizes that communications necessarily presup-
pose – to put it in Wittgensteinian terms (e.g. Wittgenstein 1953, § 206, 241;
von Savigny 1991) – a certain shared “life-form”. Especially the Wittgen-
steinian influence has strongly shaped the framework of practice-based the-
orizing, namely the concept of collectively shared rules (e.g. Giddens 1984;
Taylor 1999). This concept of rules, in its turn, fostered empirical analyses
of class structures (e.g. Bourdieu 1984) and diverse social environments (e.g.
Schulze 1992) – in other words: analyses of shared “life-forms”.
In contrast to this, Luhmann offers the idea of a coevolution of psychic
and social systems (cf. the first aspect above) and the notion of a stock of
distinctions that society provides (semantics) as a basis for communication.
Also, he emphasizes that communication always includes misunderstanding
and is possible only because of the mediating and stabilizing achievements
of social systems. For – during any interaction – the participants’ minds
always remain fundamentally nontransparent to each other (Luhmann
1995f, pp. 103–136). Surely, there is empirical evidence for Luhmann’s basic
assumption of the mutual inaccessibility of psychic systems and for his
observation that interactions often work on rather unsecured grounds.
Nevertheless, such a perspective must give rise to doubts whether
Luhmann’s theoretical setting can fully comprehend the directness and
immediacy characterising everyday interactions, and, hence, whether it can
entirely compensate for the ideas related to concepts such as “life-form” or
236
13. Kap 10. Becker 05-11-11 16.24 Sida 237
“life-world”. And yet, in view of Luhmann’s reasons for his way of de-cen-
tring the subject (see above), it must also be questioned whether the concepts
offered by the practice-based discourse can provide an adequate answer to
Husserl’s problem of intersubjectivity.
Fourth: it has been argued above that the theory of social systems and the-
ories of social practice share a notion of structure that is directed against
both voluntarism and structural determinism, and that they state that agents
and systems have some room to reflect on their actions and operations
respectively. Yet, there is an important aspect of structures in practice-based
approaches an equivalent of which cannot be found in systems theory: the
routinized character of social practices. Structures in theories of social prac-
tices not only both constrain and enable behaviour, but they also reflect the
fact that action is primarily based on behavioural routines. This idea gives a
specific twist to the notion of structures: they are fundamentally interwoven
with the body. While both strands of theorizing agree on the significance of
cognitive and normative schemes (cf. Luhmann 1995f, pp. 319–325), theo-
ries of social practices, especially Bourdieu’s and Foucault’s approaches, also
account for the significance of motor schemes and body automatism. This
difference in perspective will be further elaborated on in the following sec-
tion.
The status of the “material” and the body

The question of how theories of social practices and systems theory relate
material things and the body with the world of the social is another impor-
tant aspect of a discussion that sheds light on similarities and differences of
both strands of theorizing. Again, these topics must be considered a result
of how both theories have performed their specific kind of de-centring the
subject. In the following, this section will start with addressing two basic
assumptions regarding the physical, material world that are shared by both
strands of theorizing. We will then turn to the question of how material
“things” (i.e. objects, artefacts, or resources) are conceived in systems theo-
ry and practice-based approaches, and finally, analyse the way in which both
modes of theorizing account for the role of the body.
Two basic assumptions common to both strands of theorizing

First: as a direct result of theorizing the social from a cultural perspective (cf.
the third section of this chapter), both strands of theorizing stress the fact
that the realm of the material (including the human body) is not an objec-
tive reality independent of social phenomena. Instead, they emphasize that
the way in which human agents or social systems deal with material entities
strongly depends on how they interpret them and impart meaning to them
237
13. Kap 10. Becker 05-11-11 16.24 Sida 238
(cf. Luhmann 2002c; Giddens 1976). Both material things and the body are
always encountered as already interpreted. The way agents and social sys-
tems perceive and refer to the material world is fundamentally based on the
interpretive schemes, symbolic codes, distinctions, or cognitive routines that
society provides, and hence a matter of social construction.
Second: the world of the material is a general necessary precondition for
the existence and evolution of social practices as well as social (and psychic)
systems. Both theories agree that there could not be any social phenomenon
at all without a material background, such as molecules, cells, sound waves,
paper to write something down on, food to eat, or air to breathe. Of course,
this is a rather trivial statement within the framework of most social theo-
ries and also from the point of view of practice-based approaches. Yet, when
it comes to the theory of social systems this aspect may call for some clari-
fication with respect to the concept of autopoiesis. In what way does this
position not contradict the notion of self-(re)production of social systems?
As Luhmann (1990, p. 32; my translation) writes: “Autopoiesis does not
mean that the system could exist merely by itself, on its own resources, with-
out any contribution by the environment. Rather, it is about [the fact that]
… all elements the system consists of are produced by the system itself. Of
course, this is only possible on the basis of the continuum of matter given by
the physically constituted reality.” In producing or constructing a commu-
nication, social systems draw on the material world (the ink of the letters
when something is communicated in a book, for example) and impart mean-
ing to it. As a result, the ink forming the letters obtains a (social) function
in a communicative context. In this way, not the ink of the letters them-
selves, but the communicative aspect that makes them function as a com-
munication becomes an element of the social system. In other words: the
concept of the autopoiesis of social systems highlights that social phenome-
na possess a dynamics that makes them capable of “organizing” material
phenomena according to structures of meaning – to the effect that commu-
nications come into being. Therefore, also in a perspective that conceptual-
izes the autopoietical (re-)production of social systems, material phenomena
are an indispensable precondition of social phenomena.
Apart from this general stance towards material phenomena, practice-
based approaches and systems theory differ very much in the way they con-
ceive material objects and the body. Of course, again it will not be possible
to elaborate in the following on all the subtleties that can be found in the
writings of Luhmann and theorists of social practices; however, the topics
addressed here should be sufficient to depict the main differences.
The status of the material in systems theory

Though Luhmann has not systematically elaborated on a theory of the mate-
rial, it is possible to make out in his works two central aspects of how
238
13. Kap 10. Becker 05-11-11 16.24 Sida 239
material things are related to social phenomena: as a topic of communica-

tion in social systems, and as a means of orientation via the perception of
psychic systems.
First: material things can be a topic of communication in social systems.
The distinctions and indications used by a communication to observe some-
thing can refer to a thing in the physical world that is part of the system’s
environment. In doing so, organizations can communicate about a computer
system or production equipment, for example. Due to the constructivist
stance of systems theory, these material things are not objectively pre-given,
but only become relevant to a social system if they are constructed as a part
of its environment. In particular, Luhmann mentions two ways in which
social systems can refer to material things. In the case where the social sys-
tem experiences the relationship to material things as one of dependency,
Luhmann (1995f, p. 184) calls material things resources of the social system.
A second way in which social systems can impart meaning to the material
world is via objects (Luhmann 1997a, pp. 99, 585). When a social system
treats something as an object it draws on structures of meaning that define
the “proper” shape and “proper” name for certain things. These structures
allow for repeatedly identifying things and referring to them without giving
rise to doubts as to what is meant and how to deal with those things. Of
course, these two ways in which social systems can deal with material things
are not mutually exclusive. In fact, most resources will be treated as objects
by social systems. In any case, material things as such have no relevance to
social systems, unless they are observed, i.e. enacted as a part of their envi-
ronment.
Second: material things can serve social systems as a means of orientation.
Though social systems can observe via the distinctions they apply when
communicating, they cannot perceive. Perception is a unique feature
achieved by psychic systems through structural coupling with the human
brain (Luhmann 2000a, Chapter 1). From this perspective, social systems
are blind and strongly depend on the capacity of psychic systems to perceive.
In fact, Luhmann (cf. the chapter on the paradox of decision making in this
book) considers the ability of organizational members to perceive as their
genuine contribution to facilitating the autopoietical reproduction of the
organization. By perceiving the building where the organization is located,
making use of written notes, electronic communication technology, or signs,
for example, psychic systems can facilitate organizational communications
and support the autopoiesis of the organization.
However, it must be emphasized here again that, in any case, the social
system itself decides what counts as a communication and how much per-
ceptions influence organizational processes. “It is characteristic, especially
for organizations, that they standardize the perceptive field of individuals,
by means of documents, for example, and in doing so regulate which
239
13. Kap 10. Becker 05-11-11 16.24 Sida 240
perceptions are given a chance to be transformed into communication”

(Luhmann 2000c, p. 188; my translation).
In making this point, Luhmann can provide a fresh perspective on Herbert
A. Simon’s concept (1965, pp. 79–109) of the “psychological environment”
of a decision. When analysing the bounded rationality involved in decision
making, Simon discovered the significance of how organizations influence
individual choices by directing the decision-makers’ attention to certain
alternatives, means, and ends while making them ignore others. In Simon’s
framework this idea directly led to the concept of decision premises, which
constitute the psychological environment of a decision. Luhmann’s theory
now allows to distinguish two levels involved in a decision: first the level of
the interplay between psychic and social systems, which accounts for the
way in which organizations regulate human perceptions. And second, the
level of the social system “organization” itself, where – on a purely social
level – uncertainty absorption takes place and decisions are justified with
respect to decision premises.
Coming back to the general role of material objects in systems theory, it
must be noted that the two aspects mentioned above, i.e. objects as a topic
of communication and as a means of orientation via perception, have one
important point in common: the material world is not considered a part of
any communication, i.e. of any social phenomenon. From the perspective of
the theory of social systems, the realm of the social is systematically
detached from the material world. Apart from being a general precondition
for social systems, material things are relevant only insofar as they are
objects of interpretation within the social systems’ collective structures of
meaning. They only occur as “carriers” of meaning with respect to the dis-
tinctions that social (and psychic) systems apply.
The status of the material in theories of social practices

In contrast to the above, theories of social practices put more emphasis on
the social relevance of the materiality of objects, artefacts, or resources. Of
course, also practice-based approaches agree that objects can be an impor-
tant topic of communication and that the perception of human agents plays
a crucial role when dealing with objects. However, they make a stronger
point: material entities are not only perceived, interpreted or communicated,
they are also used, handled and applied. Their application requires certain
motoric skills and know-how (“practical understanding”), and they shape
the way in which social practices are carried out. Already the notion of
social practices includes, on a very fundamental level, the idea that practices
must be considered “materially mediated” (Schatzki 2001, p. 2). In this per-
spective, material objects in their very materiality make up an integral part
of social practices. “(T)heir social significance does not only consist in their
being ‘interpreted’ in certain ways, but also in their being ‘handled’ in
240
13. Kap 10. Becker 05-11-11 16.24 Sida 241
certain ways and being constitutive, effective elements of social practices”

(Reckwitz 2002b, p. 210).
A prominent example of the effects of material entities beyond interpreta-
tion and communication can be found in Foucault’s seminal work on disci-
plining and punishing. In analysing Bentham’s panopticon, Foucault (1975
pp. 195–217) demonstrates how architecture gives rise to certain practices
of surveillance. The panopticon is not only a building that can be perceived
or communicated about, but also implies in its materiality specific practices
of how prisoners are disciplined, pupils taught, and workers controlled.
Giddens (1984) has made a similar point about material objects by arguing
that not only rules, but also resources must be considered an important part
of social structures. By constituting power, only certain allocative resources
(e.g. raw material, technology, goods etc.) enable agents to carry out prac-
tices. From this point of view, Reckwitz (2002b, p. 212) notes that the
“things handled in a social practice must be treated as necessary components
for a practice to be ‘practiced’. […] Certain things act, so to speak, as
‘resources’ which enable and constrain the specificity of a practice.”
An example may illustrate this particular perspective. Drawing on
Giddens’ concepts of resources and rules, among concepts of other theorists,
there is a range of practice-based research that investigates the use of infor-
mation technology in organizations (e.g. Orliowski 1992; DeSanctis and
Poole 1994). Proceeding from the idea that technological artefacts are
inscribed with the developers’ assumptions and knowledge of the world,
researchers in this field analyse how the structure inherent in IT “shapes
action by facilitating certain outcomes and constraining others” (Orliowski
and Robey 1991, p. 148). In developing a “practice lens” for studying the
impact of IT on social practices, Orliowski (2000) further elaborates on
such a practice-based understanding of technology. Particularly, she intro-
duces the distinction between technological artefacts and technology-in-
practice, i.e. the patterns of social practices that actually evolve during the
use of a certain technology. In doing so, Orliowski opens up a way of study-
ing how people implement or “enact” a technological device through the
daily practices of using it against the (structural) background of the options
and constraints the technological artefact provides.
To sum up, theories of social practices offer a more direct approach to
investigating the social role of material objects, as resources or artefacts
make up an integral part of social practices. In contrast to that, systems
theory looks at the material world in a rather indirect way since objects
become socially relevant only via the perception of psychic systems and as a
topic of communication. Therefore, in having its primary focus on the social
systems themselves, systems theory – and this can also be observed in
Luhmann’s writings – basically runs a certain risk of underestimating the sig-
nificance of the material world. However, it is important not to overlook
241
13. Kap 10. Becker 05-11-11 16.24 Sida 242
another aspect addressed above: Luhmann’s perspective also opens up a

genuine way of analysing how organizations, in possessing a dynamics of
their own, can support or block the impact of material entities on their net-
work of communications. This provides a basis for research that takes into
account that organizations have certain degrees of freedom to regulate – on
the purely social level of their communications – the relevance of certain
artefacts, objects, or resources and how these influence organizational deci-
sion making.
The role of the body in systems theory and theories of

social practices
As also Luhmann (1995f, p. 245) complains, the role of the human body
certainly is a neglected and under-researched topic in the social sciences (cf.
Hassard et al. 2000). From this point of view, it is particularly interesting to
ask what systems theory, compared with theories of social practices, could
offer to rectify this unsatisfactory state of affairs.
Since the body is also a material phenomenon, all aspects mentioned
above about the status of the material in both strands of theorizing basical-
ly apply also to the status of the body. Especially, this implies a perspective
according to which the human body is a necessary precondition for social
phenomena and must always be seen as already interpreted and subject to
patterns of meaning and cultural schemes. Similarly, when it comes to sys-
tems theory, it follows from this general point that the body can be a topic
of communication and a means of orientation via perception. Concerning
the latter aspect, it may be worth adding (cf. Luhmann 1995f, pp. 246–247)
that for social systems the body is central in locating psychic systems and
attributing communications to them.
Given the details in the preceding subsections, it does not seem to be nec-
essary to address all these aspects again, this time with explicit reference to
the human body. Consequently, in the following, we will only highlight the
focal point of the particular perspective the two theoretical approaches pro-
vide.
As already mentioned in the second section of this chapter, as in the case
of material objects, here too the body plays an integral role in the very
notion of social practices. In fact, “a practice can be understood as the reg-
ular, skillful ‘performance’ of human bodies” (Reckwitz 2002a, p. 26).
Accordingly, Giddens (1984, p. 3) states that action should “not … be dis-
cussed in separation from the body”. Putting even more emphasis on this
aspect, Bourdieu (2000, p. 168) notes that “the social order is merely the
order of bodies” and has coined the concept of the habitus to account for
the fact that the social is essentially connected to the body. And finally, in
attaching a comparable importance to the role of the body, Foucault – since
242
13. Kap 10. Becker 05-11-11 16.24 Sida 243
the early genealogical period – has engraved in his writings the famous
leitmotif to “expose a body totally imprinted by history and the process of
history’s destruction of the body” (Foucault 1977, p. 83).
It has been shown in the fourth section of this chapter that it is not with-
out a reason that this emphasis on the human body emerged within the prac-
tice-based strand of theorizing. Because theories of social practices explicit-
ly aim at overcoming the Cartesian dichotomy of “mental versus material”
phenomena, they have been able to take into account the relevance of the
body to the analysis of social phenomena ab initio. Therefore, it would not
be an exaggeration to state that especially the theoretical efforts in the prac-
tice-based strand of theorizing have strongly shaped and enriched the agen-
da of research on the social significance of the body during the past two
decades.
Given this positive record, what would be the specific contribution that
systems theory has to offer? Though it is not possible to address all details
here (cf. Luhmann 1981c; 1995f, pp. 244–251), the crucial aspect is near at
hand: as is the case with Luhmann’s perspective on material things, we
encounter again a theoretical setting that clearly separates the social sphere
from the sphere of the body. While Bourdieu or Foucault, for example, make
the strong point that the social is “inscribed” in the body, Luhmann would
insist on the fact that the social is not inscribed anywhere at all. It can shape
psychic systems, and in turn the body, during a process of coevolution
(Luhmann 1995f, p. 241); social phenomena themselves, however, can never
leave the sphere of the social system – as thoughts as such can never leave
the mind. This particular theoretical position imparts to Luhmann’s writings
a tendency not to devote enough attention to the bodily constraints of
human behaviour and the impact of social phenomena on the body (how-
ever cf. Luhmann 1986a, on the semantics of love). Certainly, when com-
paring Luhmann’s and say Foucault’s œuvres, one could not consider Luh-
mann a “social theorist of the body”, while Foucault’s writings surely justi-
fy such a description.
Nevertheless, it is advisable not to stop here and instead ask again what
Luhmann’s notion of an autonomous social sphere implies for the analysis
of the role of the body. One concept particularly deserves being mentioned
here: the idea of symbiotic mechanisms, which Luhmann (1981c) has coined
to account for the way in which society relates to the body. Symbiotic mech-
anisms, being considered a purely social phenomenon, regulate how social
systems activate, control, direct or suppress bodily behaviour, and how they
react to or deal with certain (sometimes also communicatively disturbing)
movements of the body. According to Luhmann, specific social systems
– and one may also consider organizations here – have developed their own
kinds of symbiotic mechanisms. When it comes to interaction systems, for
example, society holds ready certain more or less tactless (or discreet) forms
243
13. Kap 10. Becker 05-11-11 16.24 Sida 244
of avoiding or coping with bodily behaviour (e.g. with yawning during a

business presentation). Concerning systems of intimate relationships, love
has been coded as a “passion” (Luhmann 1986a): sexual activities are close-
ly linked to – more or less stable – relationships and can serve as a proof of
love. And the political system, for example, has developed its own specific
methods of dealing with physical force and violence, primarily by ceding the
use of force to the government and subjecting it to the law.
Certainly, against the background of Bourdieu’s concept of the habitus
and Foucault’s radical writings on how the social constitutes the body,
Luhmann’s writings on the relation of the social to the body are, at least not
at first sight, not among the particular strengths of his theory. However, as
has been explored in the previous sections, Luhmann’s notion of the social
as an autonomous phenomenon in its own right also opens up new per-
spectives for research. Actually, the idea that the social is systematically sep-
arated from the psychic, material and bodily spheres calls for more investi-
gations on the mutual interrelations between these levels. The mere fact that
the concept of social practices allows for a rather direct approach to
analysing the impact of the social on bodily behaviour should not be seen as
a substitute for more thorough reflections from a different angle. With
respect to organizations, such reflections, for example, would include
analysing the symbiotic mechanisms that organizations use when producing
decisions autopoietically. The key question then would be: which mecha-
nism do organizations employ to activate, regulate, discipline, and suppress
the influence of human bodies – for the sake of maintaining their auto-
nomous processes of decision making? Therefore, in view of issues of this
kind, there is good reason to suppose that, even at its most disturbing point,
Niklas Luhmann’s theory of social systems would be able to stimulate and
enrich research on the generally neglected topic of the human body.
Summary and conclusion

This chapter has aimed at comparing two strands of theorizing that are
among the most innovative contributions to the current discourse in the field
of organization studies: theories of social practices and the theory of social
systems. On the basis of a close look at the social theoretical background of
both modes of theorizing, it has been possible to compare the fundamental
theoretical assumptions that shape the perspective of both strands of theo-
rizing. In particular, it has been demonstrated that both theoretical
approaches share a cultural, “constructivist” approach, which focusses on
the patterns of meaning, interpretive schemes, and distinctions that consti-
tute the way in which agents or systems relate to the world. Moreover, it has
been shown that this cultural perspective is closely connected to the theo-
244
13. Kap 10. Becker 05-11-11 16.24 Sida 245
retical move of de-centring the subject. In fact, both strands of theorizing

agree on overcoming the notion of a self-centred subject as the independent
origin of social phenomena in order to take into account the collective, inter-
subjective “nature” of the social.
Despite this common theoretical interest, the two modes of theorizing fol-
low a different course: the concept of social practices aims at making
Descartes’ distinction of an “inner”, mental, and an “outer”, material world
collapse into the notion of practices as a bundle in which bodily activities,
material objects, and shared practical understanding are tightly interwoven.
In contrast to this, Niklas Luhmann – strongly influenced by Husserl’s
reflections on the problem of intersubjectivity – prefers to consider the col-
lective “nature” of the social by introducing a social “realm” in its own
right. Hence he clearly separates the spheres of the mental, the social, the
material, and the body. It has been demonstrated that the notion of the
social resulting from Luhmann’s seemingly unusual conceptual decision has
parallels with certain (post-)structuralist perspectives on the social.
Moreover, it has been shown that Luhmann’s approach must be considered
both more radical and traditional than practice theoretical concepts. And
also, it has turned out that such a concept of the social allows especially for
theorizing the idea that social phenomena – among others: organizations –
possess a dynamics of their own, which cannot simply be traced back to the
behaviour of individual agents.
On the basis of those observations, the subsequent sections elaborated on
the way in which the different theoretical agendas of both strands of theo-
rizing translate into consequences for the notion of structure and agency, the
status of the material, and the role of the body. First of all, it turned out that
– despite their fundamental differences – systems theory and theories of
social practices still share many aspects when it comes to theorizing the
social. With respect to the topic of structure and agency, they both distin-
guish different degrees to which agents or systems can examine or rational-
ize their operations and they both are aware of the duality of structure. With
respect to the material world and the body, both agree particularly that
material and bodily phenomena are always mediated by interpretive
schemes or distinctions and are therefore always encountered as already
interpreted.
However, it has also been shown in which respect the specific conceptual
strategies involved in both modes of theorizing amount to significant differ-
ences. Concerning the topic of agency, the practice theoretical notion,
according to which the mental and the social level coincide, emphasizes the
common grounds of shared “life-forms” provided by social practices. In
contrast, its specific perspective of distinguishing psychic from social sys-
tems enables systems theory to provide a unique way for analysing the
mutual interplay between mental and social phenomena. With respect to
245
13. Kap 10. Becker 05-11-11 16.24 Sida 246
organizations, Luhmann’s approach is therefore capable of examining the

question of how organizations construct decision makers, and under which
circumstances and by which mechanisms decisions are attributed to whom.
Concerning the status of the material and the role of the body, the gener-
al setting is similar: while bodily and material aspects form an integral part
of social practices, systems theory systematically excludes material and bodi-
ly phenomena from the operations of social systems. The particular strength
of practice-based approaches consists in their ability to reflect the rou-
tinized, bodily aspect of human behaviour. Additionally, they can adequate-
ly theorize the significance of the very materiality of resources, which comes
into play when practices are constituted and shaped by dealing with the
material world and handling objects. In turn, however, systems theory opens
up new ways for analysing how social processes relate to material and bodi-
ly phenomena. This is because Luhmann provides a perspective to analyse
how the social system “organization” disposes of how material objects
relate to the organizational network of decisions. And with respect to the
body, Luhmann’s concept of symbiotic mechanisms could stimulate research
on the question of how organizations activate, discipline, or suppress bodi-
ly movements during their internal processes of autopoietical reproduction.
In sum, we can state that theories of social practices provide a perspective
on social phenomena that, in comprising mental, social, bodily, and materi-
al aspects, allows for an entirely integrated analysis of the social conditions
that influence human behaviour. However, compared with systems theory,
they appear rather weak when it comes to analysing the internal dynamics
of social systems. As a mirror image to this, systems theory separates the
social from the mental, the bodily, and the material, can therefore fruitfully
emphasize the genuine dynamics of social phenomena, and is able to con-
tribute to an understanding of the mutual interplay of these spheres. Yet,
systems theoretical analyses can barely reflect entirely the direct significance
of the material world and the bodily aspects of human behaviour. Especially
with respect to organization studies, this result has an important implication
for the relation of the two approaches, whichever of these two approaches
one may prefer in the end.
First: as organizations certainly do possess an internal dynamics of their
own, it may be advisable to researchers with a cultural “constructivist” per-
spective who are interested in the functioning of organizations not to ignore
the perspective of systems theory. Apart from Giddens’ sketchy note on
organizations (1984) – which, to the best of my knowledge, has not proved
to have had a significant impact on organizational studies – one can barely
make out any practice-based concepts that explicitly relate to the internal
form and functioning of organizations as a specific sphere in its own right.
At least for the moment, practice-based approaches are in need of a com-
plementary view that can compensate for this particular weakness.
246
13. Kap 10. Becker 05-11-11 16.24 Sida 247
Second: it has been demonstrated by various findings since the days of

Max Weber that organizational phenomena can certainly not be entirely
explained on the strength of their own dynamics alone. Whether it is the
particular life-forms that are typical for modern society, the role of the body,
or the impact of technology on organizations: at least for the moment, also
systems theorists need a complementary view to illuminate the restrictions
involved in their particular approach.
These two aspects imply that it may be advisable to perceive systems the-
ory and theories of social practices not as competitors for truth in the dis-
course of organization theory, but rather as complementary approaches that
could enrich each other’s perspectives. Especially the common theoretical
roots in the cultural mode of theorizing may provide sufficient foundations
for a fruitful process of mutual stimulation.
In concluding this chapter let us turn to the French écrivain André Gide.
It was his conviction that “the importance of a proposed new system, of a
new explanation of certain phenomena, cannot be measured solely by its
exactitude, but also, above all, rather, by the impulse it imparts to the spirit
to make new discoveries, new observations … by the ways it opens, the
obstacles it removes, and the weapons it provides” (Gide 1947 p. 116; my
translation). After all, it is precisely in the sense of these words that systems
theory and theories of social practices must be considered equal partners in
investigating the social phenomenon we commonly refer to as “organiza-
tion”.
247
14 Kap 11. Hasse 05-11-11 16.25 Sida 248
Chapter 11

and the New Institutionalism
Raimund Hasse
Introduction
There is no neglect of social theories in organization research worth speak-
ing of. Theories from other fields have always been taken into consideration
– sociological, as well as others; for example, those of institutional econom-
ics (Williamson/Ouchi 1981) or those of cognitive psychology (Weick 1979,
1995; Sims/Goia 1986). Such openness may be considered to be a strength
of organization research. With respect to sociology, researchers have incor-
porated interactionism (Silverman 1970), cultural theory (Hassard/Parker
1993) and postmodern thinking (Boje et al. 1996; Cooper/Burrell 1988) into
the interdisciplinary field of organization studies. Furthermore, most of the
methodological debates in the social sciences have been reflected upon (e.g.
Weaver/Gioia 1994). As a result, there are diverse perspectives on organiza-
tions (Morgan 1986) and comparative evaluation of those theories is under-
developed.
Since the late 1980s grand theories and meta-theoretical concepts have
been discussed animatedly. Additionally, there has been a revised interest in
macro-sociological parameters such as social norms and expectations, insti-
tutional and legal contexts, as well as cultural phenomena. In this context,
Niklas Luhmann has attracted a great deal of attention – initially in
German-speaking Europe and more recently in the international debate.
Compared with other grand theorists such as Michel Foucault and Anthony
Giddens, Luhmann’s work differs in explicitly focusing on issues of formal
organization. Thus, there are not only meta-theoretical concepts and macro-
sociological insights that can be related to issues of organization, there has
additionally always been a distinct organization sociology which deserves a
closer look (Luhmann 1964, 1973, 2000c).
In the US, interest in combining organizational issues with a macro-
sociological perspective has directed attention towards the new institution-
alism, as represented by John Meyer and his collaborators (Jepperson 2001;
Hasse/Krücken 1999). The new institutionalism has dealt with the spread
248
14 Kap 11. Hasse 05-11-11 16.25 Sida 249
and growing significance of formal organizations (Boli/Thomas 1997) and,

not unlike Luhmann, it has considered formal organizations to be a charac-
teristic feature of modern society. Researchers in this context have also
focused on the societal imprinting of formal organizations. In so doing, they
have incorporated macro-sociological parameters and dealt with the impacts
of a so-called world polity which predominantly consists of universal norms
and expectations of rationalization and democratic principles (Meyer et al.
1994). What is often labeled “new institutionalism for organizational analy-
sis” (Powell/DiMaggio 1991; Scott 1995) is thus based on a much broader
sociological perspective.
Though being based on different – and incommensurable – theoretical and
methodological roots, both systems theory and new institutionalism are
macro-sociological approaches which have dealt intensively with the links
between an organization and society. This is the starting point of the fol-
lowing investigation which focuses on two issues. Firstly, a striking contrast
will be outlined: While the new institutionalism is primarily interested in
processes of change towards convergence (related to the diffusion of norms
and expectations), the systems theory predominantly draws a distinction
between types of organizations (assumed to be deeply influenced by process-
es of functional differentiation). Thus the systems theory may contribute to
a compensation for what is currently being criticized by the new institution-
alists themselves: the bias towards over-emphasizing the convergence of
organizational forms. Secondly, I will argue that the systems theory is a
promising candidate for organization analysis because insights into links
between an organization and society are accompanied by a strand of
research which deals with the micro-dynamics within a given organization.
This micro-perspective is necessary in order to focus on the internal dynam-
ics which determine the sensitivity to external parameters. The course of
argument will be developed in three steps:
In the first section, I will refer to the new institutionalism by highlighting
its macro-sociological aspirations as particularly spelled out in the world
polity-concept. It will be shown that – according to this view – any organi-
zation is assumed to be influenced by the diffusion of norms and expecta-
tions which are predominantly transmitted by international organizations,
consultants and professional associations. It will be pointed out that funda-
mental differences between organizations and types thereof are neglected.
The new institutionalism evaluates such differences as residuals, and change
is expected to materialize as isomorphism. Critical remarks will also be
made as to the neglect of micro-dynamics in organizations – crucial to the
perception and processing of societal impacts. It will also be shown that, due
to this limitation, the new institutionalism tends to outline an over-socialized
and deterministic understanding of organizations.
In the second section, I will describe Luhmann’s approach by emphasizing
249
14 Kap 11. Hasse 05-11-11 16.25 Sida 250
his insights into the societal effects and overall functions of formal organi-
zations. It will be shown that the systems theory in this respect is as macro-
sociological as the new institutionalism. Organizations appear to be require-
ments for the reproduction of modern society, and they are assumed to have
an enormous impact upon social evolution. As far as the reverse impact on
organizations is concerned, it will be shown that the systems theory sheds
light on fundamental differences between certain types of organizations, all
of which are strictly related to functional differentiation as the characteris-
tic structure of modern society. In doing so, the systems theory offers a per-
spective for the compensation of the strict focus on isomorphism which
characterizes the new institutionalism.
In the third section, it will be argued that interactions between organiza-
tions and society do not cover the entire spectrum of Luhmann’s contribu-
tions to organization theory. Instead, systems theory offers a sophisticated
concept of organizational processes. Here, emphasis is placed on internal
dynamics, otherwise neglected in the new institutionalism. Luhmann’s core
assumption is that any organization is a distinct social system: From this
point of view, organizations are systems which are characterized by a spe-
cific mode of communication – decision making. These decisions are related
to earlier decisions made by the organization itself and societal impacts on
organizations tend to be devaluated in this concept.
Organizational analysis, new institutionalism,

and world polity
Organization sociology does not need to be reminded that organizations are
embedded in a social context which deeply influences any given organiza-
tion. Instead, the basic distinction between an organization and its social
environment is an established scheme (Aldrich/Marsden 1994). External
parameters have even been considered to be determinants of organizational
forms – in particular in contingency theory, where environmental factors
such as market demand or available technologies have the status of inde-
pendent variables determining the formal structure of any organization
(Mintzberg 1979; Scott 1984). Similarly, the so called resource dependence
theory emphasized the dependency of organizations on a variety of resources
such as capital, knowledge and technology (Pfeffer/Salancik 1978). Finally,
the new institutionalism of organizational analysis has re-addressed
Weberian issues of legitimacy, and in doing so has focused attention on
external norms and expectations (Jepperson 1991; DiMaggio/Powell 1991;
Scott 1995).
Another line of demarcation from both contingency theory and resource
dependence theory is the fact that the new institutionalism is a macro-
250
14 Kap 11. Hasse 05-11-11 16.25 Sida 251
sociological approach which does not restrict itself to the level of formal
organizations and their interfaces (Jepperson 2001). Instead, organizational
issues are embedded in a research program that aims to identify processes of
rationalization as a comprehensive worldwide cultural project (Dobbin
1994). In focusing on the so-called world polity as its driving force, a cul-
tural variant of modernization theory is offered (Meyer et al. 1997; Meyer
et al. 1994).
In order to avoid any misinterpretation, two core assumptions of the new
institutionalism deserve a closer look. Instead of arguing in a functional
mode, rationalizing principles are considered to be modern myths. From this
point of view, means of rationalization such as technology, education and
accounting diffuse on a global scale because they are believed to increase
efficiency. No claims regarding material effects are made. By contrast, it has
been shown that the relationship between an increase in efficiency and such
means is often difficult to measure or, if it can be evaluated, may be rather
weak or even absent (Meyer/Zucker 1989). Secondly, instead of referring to
rational actors as a starting point of the analysis, these actors are considered
to be the outcome – or materialization – of the cultural script in which the
world polity emphasizes rationality, self control and utility maximization
(Meyer/Jepperson 2000).
Organizations are considered to be mediating agencies of the world poli-
ty. It has been shown, for example, that technical standards or curricula dif-
fuse on a global scale due to the activities of international organizations,
consultants and professional associations. Where the grid of such organiza-
tions is dense, world polity is implemented, and this shapes or even consti-
tutes the concepts of individuals, states, and organizations, all of which are
considered to be modern actors (Thomas et al. 1987; Meyer/Jepperson
2000). The implementation of the world polity is thus a function of organi-
zations (Boli/Thomas 1997).
From this point of view, some organizations, such as the UN, the Inter-
national Labor Organization, Amnesty International or Greenpeace, can
shape states, individuals and organizations, all of which are expected to act
and change in accordance with the world polity. As a consequence, organi-
zations (as well as states and individuals) are the recipients of world polity
– and this applies to any organization, though empirically differing in
degree. The result is that organizations such as firms, administrative agen-
cies, schools, universities, hospitals and so on are expected to incorporate
means of rationalization – e.g. accounting practices, technologies and
reforms etc. Furthermore, all organizations are expected to adhere to social
and legal norms (for example, equal opportunities) and to reflect upon eth-
ical standards (such as sustainable development). The core hypothesis is that
organizations are generally expected to adapt to these prescriptions of world
polity and its institutionalized agents (Meyer/Scott 1983).
251
14 Kap 11. Hasse 05-11-11 16.25 Sida 252
The basic principles of the research program were spelled out as early as
1977. John Meyer and Brian Rowan’s article in the American Journal of
Sociology can be considered to be the starting point of the new institution-
alism. It has stimulated controversial debates and served as a frame of ref-
erence for the bulk of empirical investigations (cf. Scott 1995; Hasse/
Krücken 1999; Schneiberg/Clemens 2004). Its core message may be summa-
rized on the basis of a threefold question/answer scheme:
Basic question Answer
Why do organizations act and struc- Predominantly in order to achieve
ture themselves in accordance with legitimacy; partly due to conditions of
the expectations of their institution- complexity and uncertainty making
alized environment? rational decisions and calculable con-
sequences difficult.
How do organizations adapt to the Norms and scripts are formally repre-
institutionalized expectations of their sented and ritually enacted in order to
environment? demonstrate conformity and commit-
ment.
What are the material effects of these Only a few consequences are to be
processes of adaptation? expected because organizations tend
to decouple their practices; i.e. core
activities and decision making from
formal representation and ritual
enactment.
The notion of a decoupling of internal practices from formal structures and

processes is crucial to the new institutionalism. Decoupling implies that a
sharp line of demarcation is drawn between two dimensions of organiza-
tional life. Otherwise organizations would lack any selectivity with respect
to their institutionalized environment. Without decoupling organizations
would be confronted with the risk of not being able to protect their core
activities and decision making processes from external prescriptions beyond
their control. Otherwise an organization would lose its capacity to develop
and adhere to its own routines.
From an organization sociology point of view, it should be noted that this
model is closely related to James D. Thompson’s classical consideration of
formal structures as buffers to protect core activities from environmental
disturbances (Thompson 1967). Whereas Thompson referred to material
buffers as, e.g., the storage of supply products important to avoid disconti-
nuities in assembly line production, the new institutionalism focuses on
legitimacy. With respect to institutionalized norms and expectations, the
assumption is that no filter can be established on the organization/environ-
ment-interface because organizations need to demonstrate conformity and
commitment. In order to avoid societal norms and expectations threatening
core activities, however, selectivity emerges inside an organization – between
252
14 Kap 11. Hasse 05-11-11 16.25 Sida 253
organizational practices and routines, on the one hand, and formal struc-
tures and processes, on the other. The material effects of formal structures
and processes on organizational core activities may thus be rather weak.
Another form of selectivity may result from mediating agencies. This is
DiMaggio & Powell’s perspective (DiMaggio/Powell 1983). The authors are
not interested in macro-sociological issues of the world polity, nor do they
emphasize decoupling mechanisms at the level of a given organization.
Instead they consider organizational fields to be independent variables.
Fields are constituted by powerful professions, regulatory agencies and other
significant organizations (i.e. competitors and partners). Powell/DiMaggio
(1991) assume that organizations in a given organizational field are affect-
ed by the same professions and regulations. Equally important is the fact
that organizations in any given field affect each other – either mutually or
unilaterally. Organizations tend to imitate each other’s structures and
processes in order to learn from and minimize the risk of post-decisional
regret (Harrison/March 1984). The result is that within organizational
fields, or societal sectors, to use John Meyer and Richard Scott’s term
(Meyer/Scott 1983), isomorphism is to be found (Kondra/Hinings 1998).
DiMaggio & Powell (1983) exclusively refer to processes of isomorphism,
noting that the lack of interest in complimentary differences between fields
has led to sharp criticism (Schneiberg/Clemens 2004). First of all, it does not
acknowledge that the idea of distinct fields is an analytical construction
which neglects self-evident overlaps between, and differences within, real
fields. Secondly, it confronts the fallacy of contingency theory which
assumes that environmental factors plainly determine an organization’s
structure. The paradigm shift from focusing on functional requisites (con-
tingency theory) to issues of legitimacy (new institutionalism) does not
resolve this theoretical problem. Thirdly, strictly referring to organizational
fields is predominantly an inter-organizational perspective which is fading
from the macro-sociological dimension of the world polity approach. To
summarize, the focus upon organizational fields may not be an appropriate
means of investigating the interplay of organizational and societal change
from a broader perspective.
An alternative and more promising path may be to put more emphasis on
the inherently multi-dimensional character of the world polity, which is
assumed to shape modern organizations (Friedland/Alfaord 1991). Here, the
potentially contradictory and mutually exclusive character of institutional-
ized norms and expectations has to be taken into consideration.
Rationalization principles, for example, may collide with issues of democra-
tization and deliberation when reforming public services. Likewise, short-
term competitive advantages and first-mover advantages may be at risk
when industrial production is aiming toward principles of sustainable
development. These contradictory relations hint at the fact that it may be
253
14 Kap 11. Hasse 05-11-11 16.25 Sida 254
impossible to follow all institutionalized norms and expectations to the same

degree.
Without explicitly referring to the world polity-program this argument
has been put forward by Nils Brunsson (2002), and his contribution has
been quite convincing in arguing that organizations have no alternative to
the selection and interpretation of such norms. As a consequence of such
selection and internal processing, it may be concluded that institutionalized
norms and expectations do not always have the same effect (Zucker 1977).
Decoupling may thus not be a universal mode but a variable which empiri-
cally ranges from loose to tight coupling. As a consequence, the formal
incorporation of institutionalized norms and expectations may or may not
make a difference in terms of internal practices – and the material impacts
of such incorporation may vary according to the organizational processing
of these expectations. In this respect, a micro-foundation which highlights
the internal dynamics is necessary for analyzing material effects on organi-
zations and the corresponding repercussions on society. Unfortunately, such
a micro-foundation is mostly absent in the new institutionalism.
Thus far it has been argued that the new institutionalism may serve as a
starting point for an evaluation of systems theory, because – and not unlike
Luhmann – it has contributed to an understanding of links between organi-
zational forms and societal context. Two characteristic features have been
critically reflected upon: (1) the neglect of differences with respect to socie-
tal impacts and corresponding organizational forms, and (2) a missing
micro-foundation which needs to be overcome in order to shed light on
organizational processes and their effects. As will be argued in the next two
sections, Luhmann’s theory deserves a closer look because it offers insights
which may contribute to a solution to both research problems.
Organization/society – Links in Luhmann’s theory

of social systems
Unlike most sociologists who have put forward a grand theory, Niklas
Luhman addressed organizational issues throughout his career. Referring to
his biography one may even argue that organization research was his entry
into the social sciences – and it may be noted that in the famous controver-
sy between systems theory and critical theory Habermas accused Luhmann
of transferring a technocratic perspective from administrative organizations
to the whole of society (Habermas/Luhmann 1971b). Regardless, organiza-
tional and administrative topics dominated Luhmann’s agenda in the 1960s
and early 1970s when he was heavily influenced by organization research
from the US, in particular by the work of Herbert Simon and James March.
In his later publications, Luhmann readdressed organizational issues with-
254
14 Kap 11. Hasse 05-11-11 16.25 Sida 255
in his advanced theoretical framework. Not only does this apply to “Die
Gesellschaft der Gesellschaft” (1997a) where the spread of formal organiza-
tions is related to the evolution of modern society and functional differenti-
ation as its primary mode, but also to his outline of an autopoietic theory of
organization as spelled out in “Organisation und Entscheidung” (2000c, pp.
39 ff.). Therefore it would not be an exaggeration to argue that Luhmann
was a genuine organization sociologist in touch with the research field and
heavily interested in related single issues.
Luhmann’s development of a general social theory may be characterized
by paradigmatic changes (as from cybernetics/functionalism in the beginning
via evolution theory/autopoiesis to a formal theory of form/theory of obser-
vation). By contrast, Luhmann’s perspective on organizations has been more
consistent over the years. Nevertheless, two approaches to organizations
may be distinguished. On the one hand, and not unlike the new institution-
alism, Luhmann emphasized the interrelationship between an organization
and society. On the other hand, organizations are considered to be distinct
social entities – a level of reference which is only loosely coupled with its
social environment. In this section I focus on interrelationships. Then I will
deal with internal dynamics.
General remarks on the societal effects of formal organizations can be
found already in the beginning of Luhmann’s contribution to organization
sociology: In the 1960s and 1970s organizations were considered to be con-
flict regulators and communication channels. They were characterized with
respect to their capacity to process different issues simultaneously – some-
times 24 hours a day, as is the case with hospitals, for example, and for
decades or even for centuries, as is the case with churches. Furthermore,
Luhmann argued that organizations were preconditions for the differentia-
tion of labor, and could thus trigger forms of action which would otherwise
have been rather unlikely. Highlighting differentiation as the core principle
of social evolution, Luhmann concluded that “… a high degree of … speci-
fication of behavior can only be achieved by the mechanism of organization
which modern society requires in many of its most important functional
domains” (Luhmann 1975b, p. 13, translated by the author). It thus seems
appropriate to state that, according to Luhmann, modern society needs to
be based on formal organizations, and in that sense modern society is an
organization society (Schimank 2001).
Comparable weight has been put on the societal functions of organiza-
tions in later writings on social evolution and functional differentiation
(Luhmann 1987b). With respect to contemporary society, the significance of
organizations has even been compared to generalized communication media
– such as truth (science), power (politics) and property (economy) – which
is clearly at the heart of his theory of functional differentiation. Referring to
the differentiation of the modern economy, Luhmann noted: “Media and
255
14 Kap 11. Hasse 05-11-11 16.25 Sida 256
organizations are two forms of the formation of unlikely, sophisticated

expectations … In this respect, they are functionally equivalent means, and
all functional systems … are dependent on their interplay” (Luhmann
1987b, p. 41, translated by the author).
Profound reflections on the social impacts of the emergence and diffusion
of formal organizations are also to be found in historical descriptions. Here,
Luhmann stated that “… organizations originally emerge in response to an
obvious need for decision making about collective action … as a conse-
quence of money economy or under conditions of religious pluralism”
(Luhmann 1981a, p. 361, translated by the author). This interpretation is
based on the observation that organizations emerged in medieval cities and
proved to be rather influential as regards processes of specialization. In this
respect, the differentiation of politics, science, economy etc. appears to be
tightly coupled with corresponding organizations like public administration
offices, universities, banks, etc., all of which are considered to be so-called
primary organizations of “their” functional systems (Luhmann 1997a, pp.
840 f.). Emphasis is thus placed on close links between the functional dif-
ferentiation of modern society and the emergence of specialized organiza-
tions.
Currently, links between organizations and society are one of the main
foci within systems theory (cf. Kneer 2001). In this context, organizations
are assumed to incorporate codes and entire programs of functional systems
and such incorporation is considered to determine the identity of most
organizations (Tacke 2001). Insofar as organizations give sole priority to
codes and programs of certain systems, a categorical distinction between
types of organizations, like economic, scientific or political, can be drawn.
Organizations, then, may even be considered to be sub-systems of the func-
tional systems whose logic they have incorporated (Willke 1994b).
This “organizations as sub-systems” perspective is of paramount im-
portance within the framework of Luhmann’s theory because it allows the
re-addressing of macro-sociological issues of integration and steering
(Kneer 2001). The reason for this is that organizations are considered to be
the only social systems that can communicate with each other.1 In so doing
they can setup inter-organizational relationships with any type of organiza-
tion.2 Insofar as organizations are considered to represent societal systems
1 Theoretically, such inter-organizational communication is possible because organizations
have a dual status. They are social systems (which have internalized codes of programs of
functional systems), and they are actors (which means within the framework of Luhmann’s
theory that communication can be attributed to them just as it can be attributed to people).
2 Such inter-organizational arrangements have gained enormous attention through the dis-
course on social networks. Strictly in accordance with Luhmann’s distinction of only three
societal levels (interaction, society, organization) these network structures need to be re-
conceptualized as being either interaction (among organizations) or organization (with single
organizations as members, as in the case of associations or neo-corporate arrangements and
social pacts) (cf. Kneer 2001).
256
14 Kap 11. Hasse 05-11-11 16.25 Sida 257
such as politics or economy, communication between them offers the poten-

tial for processes of mutual adjustment or even for regulation which would
otherwise have to be questioned from a systems theory point of view
(Lieckweg 2001).
Thus far, emphasis has been put on close relationships between organiza-
tions and their societal context. It has been shown that the systems theory
assumes that organizations fulfill social functions which are crucial for the
reproduction of modern society. Furthermore, a trend has been outlined
which predominantly characterizes organizations with respect to functional
differentiation. Organizations are assumed to copy codes and programs of
those functional systems, with which they are expected to identify. From this
point of view, societal functions and relationships with the social environ-
ment have been at the center of Luhmann’s reflection upon organizations.
This perspective might easily be compared with the new institutionalism.
The result of such a focused comparison is that both theories are based on
different concepts of modern society, which has a profound impact on the
applied perspective on organizations. According to the new institutionalism,
society is predominantly made up of norms and expectations which affect
any organization. Thus, processes of change toward isomorphism are
expected. The systems theory, by contrast, highlights the differentiation of
society into functional systems. According to this view, processes of differ-
entiation into politics, economy, science, and so on are considered to be the
prevailing pattern of modern society. As most organizations are considered
to be closely related to one of these functional systems, the systems theory
highlights different types of organization. This may serve as a starting point
for questioning the new institutionalism’s emphasis on isomorphism. In the
next section it will be argued that this is not the only insight the systems
theory provides. Instead, Luhmann’s contribution to organization sociology
offers a micro-sociological perspective.
Micro-foundation: Organizations as distinct systems

Luhmann’s reflection on the relationship between society and organizations
have been accompanied by a strand of research that exclusively focuses on
organizations as distinct social systems. Initially this was indicated by the
fact that his functional analysis predominantly referred to the focal organi-
zation itself. Luhmann was interested in the impacts of certain organiza-
tional forms on the reproduction and functioning of a given organization
(Luhmann 1964). Later as he developed his theory of social systems to a
more advanced level, organizations were conceptualized as autopoietic sys-
tems which relate their decisions to earlier decisions in a self-referential
manner (Luhmann 1988c; 2000c).
257
14 Kap 11. Hasse 05-11-11 16.25 Sida 258
In what follows, I will argue that Luhmann has focused on organizations

with respect to organizational effects throughout his career. The starting
point was a sharp criticism of Max Weber’s ideal type of bureaucracy as an
efficient and legitimate form of modern organization (Weber 1972;
Luhmann 1971b). Such a criticism was very much in line with the main-
stream of organization sociology at that time (cf. Hasse 2003a, pp. 55 ff.).
In contrast with many others, however, Luhmann did not restrict his analy-
sis to an empirical falsification of Weber’s thesis about the bureaucratic form
as the materialization of modernization processes. Instead, and in contrast
to Weber’s ideal type, he called the means/end character of organizations
into question and argued that ends fulfill functions other than regulating
action on the basis of decision outcomes (Luhmann 1973). It should be
noted, however, that these functions are strictly related to the focal organi-
zation, not to the whole of society or the functional systems thereof. The
main research question then became: In what respect do certain structural
features of an organization affect the reproduction of this organization?
(Luhmann 1964)
This preference for analyzing the functions of organizational forms with
respect to the organization itself can best be illustrated by focusing upon a
fundamental requirement of any organization – but probably of the whole
of society as well: The need to absorb complexity and uncertainty. In orga-
nizational analysis, addressing issues of uncertainty absorption has a long
tradition. Key figures like James March (March/Simon 1958) and James
Thompson (1967) have dealt with uncertainty absorption, and Luhmann
related his findings to the research field in general. From a macro-sociolog-
ical point of view, however, it is striking that uncertainty absorption was
strictly considered to be a requirement of the organization. Uncertainty
absorption was expected to protect the focal organization from risk and
information overload. Implications for the whole of society or for non-
organized parts thereof were explicitly neglected, and in this respect no dif-
ference is to be found between Luhmann, on the one hand, and organization
researchers such as March or Thompson, on the other.3
When emphasizing the quasi-sovereign status of organizations in
Luhmann’s contributions to organization sociology, his basic distinction
between interaction, society, and organization as the only levels of reference
offers important insights (Luhmann 1975b). Organizations are considered
here to be much more important conceptually than other social entities such
as groups, classes, or networks, none of which are considered to be such
3 Against the backdrop of those characteristics of contemporary life which are emphasized in
interpretations of so-called information societies (Castells 1996; Webster 1995; Duff 2000) it
may be more evident than ever that uncertainty absorption is required not only by organiza-
tions but by the whole of society as well. For such a reflection on organizations as “uncer-
tainty absorbers”, cf. Hasse (2003b).
258
14 Kap 11. Hasse 05-11-11 16.25 Sida 259
basic principles of modern society. In “Interaction, Organization, and

Society” (Luhmann 1975b) the coupling between an organization and its
wider societal context is explicitly neglected. Instead, the emergence and dif-
fusion of formal organizations is treated as a “completely independent
development which embodies a new principle of boundary spanning and self
selection, and which can be deduced neither from interaction nor from
society …” (Luhmann 1975b, p. 12). Furthermore, it was even assumed that
organizations may decide to shift their orientation as, for example, a
research organization may decide to relate its decision making to political
issues. This implies that the notion of a fixed relationship between an orga-
nization and its specific societal context is called into question. Instead, the
identity of an organization is considered to be open to change on the basis
of decision making, as are any other parameters of an organization (cf.
Hasse/Japp 1997).4
Since the end of the 1980s, the core concept of organizations as decision
making entities has been outlined in greater detail. Organizations are con-
ceptualized as autopoietic systems which relate decisions to their own past
decisions. They are considered to be “systems which are composed of deci-
sions, and which produce the decisions they are composed of … by them-
selves” (Luhmann 1988c, p. 166, in italics, translated by the author). From
this point of view it is not surprising that an embeddedness of organizations
is not the main argument in his masterpiece “Organisation und
Entscheidung”. Instead, organizations are supposed to be self referential
entities (Luhmann 2000c, pp. 39 ff.) implying that organizations or types
thereof are not considered to be subsystems – or agents – of functional sys-
tems.
Summing up, it can be concluded that Luhmann’s contribution to organi-
zation sociology consists of two perspectives. Organizations, on the one
hand, are considered to be a distinct frame of reference. In early contribu-
tions, theoretically based on principles of functional analysis, this is spelled
out by emphasizing the impact of an organization’s structural features on the
organization itself. Then, and as the systems theory developed in the direc-
tion of self-reproduction, organizations were conceptualized as self-
referential systems. As a consequence, organizations appear to be only loose-
ly coupled with their societal environment. On the other hand, Luhmann
has argued that the emergence and diffusion of formal organizations is tight-
ly coupled with societal evolution. Initially this was hinted at by the societal
functions of formal organizations. More recently, the relationship between
organizations and functional systems, such as politics, economy, and
4 With respect to the historical development of Luhmann’s systems theory it may be noted that
this indicates that the basic concept of social systems as self-referential entities has been
applied to organizations even at an early phase of the theory’s development.
259
14 Kap 11. Hasse 05-11-11 16.25 Sida 260
science, has been at issue. In this context, organizations appear to be

preconditions of functional differentiation, and organizations are described
as parts thereof. Occasionally, organizations are even assumed to mediate
between and within functional systems. In this respect, the significance of
organizations to macro-sociological processes can hardly be over-estimated.
Conclusion
The starting point was the observation that the incorporation of theories
from other fields is a characteristic feature of organization research.
Recently, systems theory as developed by Niklas Luhmann and his scholars
has attracted a significant amount of attention. In addition to meta-
theoretical and methodological concepts Luhmann’s contribution is con-
sidered to be a unique macro-sociological approach which extensively
reflects upon the significance of formal organizations.
I have focused on Luhmann’s contribution to organization sociology by
showing that the systems theory is not the only macro-sociological approach
which theorizes on the links between organizations and their wider societal
context. The new institutionalism offers a particularly strong macro-
sociological alternative. As a consequence, the potential exists for a com-
parative evaluation which may be important for bringing Luhmann’s theory
further toward the core of organization sociology. Additionally, it has been
shown that a micro-foundation of organizational processes has accompa-
nied Luhmann’s macro-sociological focus on organizations.
Both perspectives offer important insights which can be explored on the
basis of a focused comparison. According to the new institutionalism,
organizations are a characteristic feature of modern society. It has been
argued in this context that organizations such as international organiza-
tions, consultants and professional associations support the diffusion of
western culture and the prevailing forms of rationality on a global scale. In
so doing, organizations are having a profound impact on states, individuals
and organizations, all of which are considered to be modern actors deeply
influenced by world polity. This implies that organizations (as well as states
and individuals) are recipients of world polity. They are expected to change
in accordance with pre-given and organizationally mediated scripts. Because
of its emphasis on the wider societal context, the macro-sociological char-
acter of the new institutionalism cannot be questioned. As its weakness, the
strict focus on isomorphism has been criticized, and a top-down perspective
has been identified which fails to acknowledge an organization’s selective
and idiosyncratic processing of norms and expectations.
The systems theory, by contrast, has offered insights into both organiza-
tion/society links and internal dynamics. Links are spelled out when formal
organizations are supposed to be a precondition for the evolution of mod-
260
14 Kap 11. Hasse 05-11-11 16.25 Sida 261
ern society and when considering organizations as deeply influenced by their

societal context. According to this point of view, organizations tend to iden-
tify themselves (and are identified), for example, as economic, political or
scientific organizations. The systems theory thus distinguishes between types
of organizations. Rather than over-emphasizing changes in the direction of
convergence, such types are assumed to be a persistent feature of modern
society.
The strict focus on internal dynamics, by definition, devaluates any links
with the societal context. Instead, information processing and decision mak-
ing are at issue. In this line of research, Luhmann considered organizations
to be self-reproducing systems, and it is worth noting that the basic idea of
self-reference was applied to organization sociology quite early on.
Furthermore, decision making as an autopoietic process is clearly at the
heart of “Organisation und Entscheidung” (2000c) which was Luhmann’s
most important contribution to organization theory. From this point of
view, the systems theory is characterized by a micro-foundation which can-
not to be found at all in the new institutionalism.
In summary, Luhmann has offered insights which may be used as a
resource pool for the advancement of our understanding of both internal
processes and external links. However, there is still a need for clarification
as regards how these perspectives can be theoretically integrated.
Empirically, issues of both organizational change over time (as an ongoing
process and with structural impacts as, for example, discussed by Chandler
(1977) and Fligstein (1990)) and societal effects of certain types and forms
of organization (which may or may not fit the scheme of functional differ-
entiation) seem to have been neglected. As a consequence, and in striking
contrast to historical observations, contemporary impacts on society which
go beyond the mere reproduction of basic structural features of modern
society cannot be addressed appropriately – but this may be considered as
another parallel to the new institutionalism.
261
15. Kap 12. Koch 05-11-11 16.25 Sida 262
Chapter 12

and Postmodernism
Jochen Koch
Introduction
When Robert Cooper and Gibson Burrell outlined their program of organi-
zational postmodernism in 1988, they were very clear about splitting the
world of social theory into two opposing schools of thought: “Modernism
with its belief in the essential capacity of humanity to perfect itself through
the power of rational thought” and “Postmodernism with its critical ques-
tioning, and often outright rejection, of the ethnocentric rationalism cham-
pioned by modernism” (Cooper/Burrell 1988, p. 92). This diametrical con-
cept – contrasting apparently not only different theoretical perspectives but
also deep-rooted worldviews or even ideologies – was accompanied by a def-
inite assignment with regard to the relevant discourse participants: the mod-
ern side should be seen as represented basically by Habermas and Luhmann,
the postmodern side mainly by Derrida, Foucault and Lyotard. Not only did
Cooper and Burrell make an obvious distinction between the two camps on
the modernist side, namely a critical perspective held by Habermas and a so-
called functionalist perspective viewed as represented by Luhmann, but they
also presented a clear assessment of Luhmann’s systems theory: “Luhmann’s
work represents a formalization as well as a justification of the develop-
ments charted by Bell. In what is sometimes called the ‘new systems theory’,
Luhmann spells out the inexorable rationality of systemic modernism in
which Kant’s notion of the critical, rational subject is completely repressed
in the interests of a machine-like system of social functionality. Society itself
becomes a gigantic organization” (Cooper/Burrell 1988, p. 96).
Such an estimation and pigeonholing of Luhmann’s work must be irritat-
ing at least to those who have been taking a closer look at the significant
developments in systems theory in the last thirty years. Nevertheless, not
only is this assignment embedded in a strong tradition corresponding to the
(self-)interpretation of modernity in the line of Weber (1979) and Habermas
(1984), with regard to the development of modern society in a conflictive
manner between reason and rationalization, but this estimation of Luh-
262
15. Kap 12. Koch 05-11-11 16.25 Sida 263
mann’s work also approaches the postmodern critique of modernity, as for-

mulated by Lyotard (1984), who characterizes this attribution of systems
theory as a problematic and affirmative modernist approach. And last but
not least, Luhmann himself has shown a great affinity to the notion of
“modern” and in several statements has rejected the idea of a postmodern
society, calling it a non-workable description of the present society (Luh-
mann 1998).
Hence the relationship between Luhmann’s theoretical work and the
numerous approaches to postmodernism seems more complex and ambigu-
ous than previously thought, and for that reason it should be re-considered
and examined in a more detailed way. It is the purpose of this paper to resi-
tuate Luhmann in the context of the ongoing debate on postmodernism,
accentuating and clarifying the similarities between systems theory and the
main theoretical ideas of various postmodern positions without losing sight
of the evident differences.
Part 1 and 2 of this chapter will offer a short overview of the ideas cen-
tral to postmodernism, referring to the conceptual framework introduced by
Lyotard and supplementing it with some ideas of Derrida and Foucault. Part
3 will then focus on the different components of Luhmann’s theoretical
work by highlighting possible parallels to postmodern thinking. In part 4
three pitfalls of an exclusively postmodern reading of Luhmann’s systems
theory will be discussed, and in part 5 a comparative perspective will be pro-
vided by drawing on both differences and similarities between systems the-
ory and postmodern approaches. Finally, a short summary in the form of
five elements of a theory of difference will be presented, showing the central
affinity between systems theory and postmodern thinking.
The impetus of postmodern thinking

Postmodern approaches are introduced into social science in order to criti-
cize, reject and overcome the foundational premises of modernism. For the
most part, the underlying premise of modernism conceptualizes the devel-
opment of modern society as an inherently ambiguous process of conflict
between the “origin” idea of the Enlightenment (liberation of the life-world)
and its degenerated systemic version in the form of a mechanization of social
order. While these two faces of modernism might seem incongruent and
antagonistic, postmodern critique posits that they both rely on the same
assumptions: that (1) discourses represent both reason and reality and that
(2) there is a conscious subject. The idea of a (direct) representation and the
idea of a (totally) conscious subject are fundamental to the self-understand-
ing of modernism, and they both reflect the idea of a fundamental unity.
In contrast, postmodernism can be understood as a general “assault on
unity” (Power 1990). Instead of unity one can find in the centre of any
263
15. Kap 12. Koch 05-11-11 16.25 Sida 264
discourse a web of contradictions or “discursive gaps”, and this could be

interpreted also as the locus of power reconceptualized as a subtle and
covert web of prior structuring. Instead of unity a postmodern understand-
ing tries to emphasize the idea of difference, and this motive can be found in
all postmodern approaches: “Foucault’s genealogical method is therefore
similar to Lyotard’s agonistics and Derrida’s deconstruction: all deny the
concept of a perfect origin and substitute for it a process of differential con-
testation” (Cooper/Burrell 1988, p. 101).
A framework of postmodern theory

Central to the idea of postmodernism is the work of the French philosopher
Jean-François Lyotard (1984; 1988) and his all too famous claim of the end
of meta-narratives. However, beyond the mere proclamation of the end of
meta-narratives, one can find in his philosophy an important and profound
framework for a discursive understanding of recent society that focuses on
three different problems of communication, knowledge and social integra-
tion: (1) the problem of legitimation, (2) the problem of chaining and (3) the
problem of meta-narratives.
The problem of legitimation

Generally speaking, “legitimation is a process by which a legislator is
authorized to promulgate (…) a law as a norm” (Lyotard 1984, p. 8).
Lyotard draws on scientific discourse to illustrate this problem, suggesting
that any statement has to fulfill a given set of conditions in order to be
accepted. The rules of this scientific discourse are therefore the domain of a
scientific “legislator” who prescribes the relevant conditions (for instance,
internal consistency and experimental verification) and determines whether
a statement is considered scientific or not.
At this point the problem of legitimation becomes evident. Scientific legit-
imation can also be required of the rules inherent to scientific discourse, or
even of the goal of such a discourse. If it is the goal of scientific discourse to
distinguish between true and false, what is the status of this goal itself? In
other words: is the distinction between true and false itself true or false?
This simple question leads scientific communication to fundamental and
virtually unsolvable problems. Yet such a reflection of a discourse on its own
rules and goal cannot be prevented unless rules are introduced which them-
selves do not fulfill the scientific standards. “This is exactly what Nietzsche
is doing, though with a different terminology, when he shows that ‘Euro-
pean nihilism’ resulted from the truth requirement of science being turned
back against itself” (Lyotard 1984, p. 39). The undermining effects are in-
evitable.
264
15. Kap 12. Koch 05-11-11 16.25 Sida 265
At this point Lyotard introduces the idea of narrative knowledge and thus
the distinction between scientific and narrative knowledge, which can be
seen as a specific means of unfolding the problem of legitimation. According
to Lyotard, the main difference between these two types of knowledge lies
in their different approaches to legitimation. While scientific knowledge is
always confronted with the problem of legitimation, narrative knowledge is
not, because the process of generating and transmitting narrative knowledge
is presumed to be self-legitimating. In this sense, scientific knowledge is sup-
plemented by narrative knowledge in order to close the gap of legitimation,
and therefore narrative knowledge has an important relevance to the possi-
bility of making scientific statements.
However, the legitimating function attributed to narrative knowledge in
this relationship is not obvious, nor can it be made explicit without losing
its function of legitimation. Consequently, to resolve the problem of legiti-
mation a specific genre of discourse cannot refer explicitly to a simple nar-
ration. “A science that has not legitimated itself is not a true science; if the
discourse that was meant to legitimate it seems to belong to a pre-scientific
form of knowledge, like a ‘vulgar’ narrative, it is demoted to the lowest
rank, that of an ideology or instrument of power” (Lyotard 1984, p. 38).
The problem of chaining

While the problem of legitimation in a narrow sense can be regarded as a
“discursively enclosed” one because of the complex relationship between
included and excluded parts of narratives, the problem of chaining is in the
first place clearly an open one. It deals with the possibilities and limits of
linking different genres of discourse (their phrases or their knowledge claims
respectively) with one another. Whereas different phrases or knowledge
claims of one genre of discourse could be linked following the rules of that
discourse, there are no rules for linking different genres of discourse. In the
concept of Lyotard, this problem is presented in a clearly antagonistic per-
spective: “By its rule, a genre of discourse supplies a set of possible phrases,
each arising from some phrase regimen. Another genre of discourse supplies
another set of other possible phrases. There is a differend between these two
sets (…) because they are heterogeneous. (…)” (Lyotard 1988, p. xii).
This problem of linking, or (as proposed here) chaining, is presumed to be
fundamental, induced by the premises of the genres of discourse themselves.
The conflict between different genres of discourse is inevitable and therefore
called a “differend” to mark the intended antagonistic relationship. In this
sense there are no unified rules for chaining ethical and scientific claims, for
example, because ethics and science follow their own rules of discourse. It is
presumed that this problem cannot be resolved in general, so that any “solu-
tion” or any claim of a solution is suspected to explicitly or implicitly
265
15. Kap 12. Koch 05-11-11 16.25 Sida 266
suppress potential differends. This does not mean that there cannot be any
solution at all, but that there is no general solution. In this discursive under-
standing, and due to the impossibility of having general and unifying rules,
such a general idea of chaining is called a grand narrative. In this sense the
problem of chaining refers also to a narrative, but on a deeper level, which
then can be understood as a grand narrative or even as a meta-narrative.
The problem of meta-narratives

Meta-narratives are no simple narrations but narrations that seem to be
based on discursive knowledge but in fact are not. Lyotard has identified
two major forms of such narratives, one more political and the other more
philosophical (Lyotard 1984, p. 31). Both forms could be seen as represent-
ed in the concept of Enlightenment. “The explicit appeal to narrative in the
problematic of knowledge is concomitant with the liberation of the bour-
geois classes from traditional authorities” (Lyotard 1984, p. 30). Further-
more, such a meta-narrative also resolves the problem of legitimation for
each genre of discourse. This can be explained by drawing again on scien-
tific discourse. In this sense the goal of science to distinguish between true
and false is itself stabilized against auto-logical reflection by referring to the
idea of a grand narrative; for instance, to the idea of Enlightenment. True
are those claims that deserve the idea of Enlightenment, and it is the idea of
Enlightenment that is true in itself. Beyond solving the problem of legitima-
tion, this narrative can be considered a grand narrative, because it also rec-
onciles different genres of discourse.
This is the point of departure for all postmodern critiques, as it shows that
the idea of a meta-narrative or even a grand narrative has to be considered
the pseudo-solution of both problems mentioned above. The problem of a
meta-narrative in this interpretation becomes a problem of the invisible,
subconscious and emergent structures of the knowledge development
process, and therefore it also entails a linkage to power and power struc-
tures.
Lyotard (1988) has analysed this problem from an ethical perspective
under the notion of “differend”, demonstrating in an instructive way how
certain types of claims are systematically excluded. A similar conceptualisa-
tion can be found in the work of Foucault (1972; 1973) under the notion of
episteme and in his later work (1979) under the notion of apparatus (dis-
positif).
In this sense casting doubt on a meta-narrative requires a highly sophisti-
cated framework, which offers an analytical perspective for grasping the
“subconscious structure” of the conceptual ideas behind society and there-
fore of society too. While Foucault is certainly more sceptical, this is the
266
15. Kap 12. Koch 05-11-11 16.25 Sida 267
point where the concept of deconstruction developed by Derrida (1977;

1978) fits in.
One of the central outcomes of the deconstructive work developed by
Derrida is the idea of de-centering the subject. On both an epistemological
and a societal level, modernism is basically built on the idea of a subject as
a unity and as a fundament. This may be best represented in the Cartesian
cogito ergo sum embodying the concept that the ultimate grounding for all
thought and theorizing should be the process of human self-consciousness
and therefore the subject. For postmodern approaches this idea of an ulti-
mate and foundational subject is instead highly suspicious, and one can find
this questioning of the subject not only in the work of Derrida, but also in
the work of Foucault and Lyotard. Common to these postmodern critics is
the attempt to think and to theorize without attributing a central role to an
idea called “subject” in the sense of an unquestionable reference point.
Instead of focusing on the subject, the postmodern approaches look at the
role of language and at conceptualizing philosophy, science and society from
a discursive viewpoint, as was delineated above.
This short framing of central postmodern ideas serves as a background for
pointing out the similarities and differences with regard to Luhmann’s
theory.
Postmodern elements in the theory of Luhmann

One of the most appealing aspects of a comparison between Luhmann’s the-
ory and the different postmodern approaches could perhaps be detected in
the previously mentioned idea of de-centering the subject. Like postmod-
ernists, with their approaches for undermining the modern idea of the sub-
ject, Luhmann has definitely not reserved a central position for the subject
in either his theory or his structural description of modern society. Basing his
social theory on the main element of communication, Luhmann places the
subject (or in a sociological sense the individual) not at the centre but in the
environment of the society (see Chapter 1 in the present volume). At first
sight, and following a confrontation on this level, one is tempted to see a
clear parallel between the postmodern idea of de-centering the subject and
Luhmann’s approach of placing the individual in the environment. At sec-
ond sight, however, the idea of de-centering the subject in a postmodern
sense has a clear critical impetus and a clear critical implication, whereas
Luhmann’s “de-centering” is due to theoretical decisions to understand
society on the basis of communications. Put this way, it leads us to state that
these are quite different backgrounds and thus also potentially quite differ-
ent ideas. But on the other hand, and perhaps at third glance, both the post-
modern and the Luhmannian “de-centering” converge on a more
267
15. Kap 12. Koch 05-11-11 16.25 Sida 268
fundamental, or even better, on an epistemological understanding of the

problem. In this sense both the idea of de-centering the subject and
Luhmann’s idea of thinking society “without” individuals at centre stage are
“embedded” in the epistemological idea of difference.
The idea of difference in the work of Luhmann

Luhmann’s theory is a constructionist and difference-theoretical approach,
in the sense that it attempts to treat neither the objects of the world nor the
terms of world description as simply given (Luhmann 2002e). The starting
point for all reflections (for all theory- or system-building processes) is not
identities but distinctions, to which other distinctions can be connected. For
that reason identity is conceived only as a difference, namely as a difference
between identity and difference (Luhmann 2002a). This form of radical de-
ontologisation of objects goes along with the idea that there is no unity with-
out a reference, but every reference leads to another distinction. In this sense
the unity of a system is not representable without referring to its environ-
ment, and that means referring to the distinction between system and envi-
ronment. Thus, the unity of the distinction between system and environment
is another difference – for example the difference that constitutes an observ-
er observing by distinguishing system and environment. Therefore, any con-
ception of unity in Luhmann’s theory has to be seen as referring to another
difference; and there is no final and ultimate unity, neither in the structure
of modern society nor in its semantic description. This idea obviously leads
to the paradoxical “foundation” of systems theory: every unity is a differ-
ence, or, to put it another way: “The same is different” (Luhmann 1993g).
Therefore it is “foundational” to the theory of Luhmann that there is no
foundation in terms of a final and ultimate unity, for example consensus, cri-
tique, subjectivity or even intersubjectivity, because every foundation leads
only to another distinction and therefore to a paradoxical statement.
Focusing on such paradoxical statements is fundamental to Luhmann’s the-
oretical work, and it is also an important issue in postmodern thinking.
System and environment

The central distinction in Luhmann’s sociological theory is that between sys-
tem and environment. It constitutes the paradigm of recent systems theory
(Luhmann 1995f, p. 176). Every system is constituted by a specific relation-
ship to its environment. Therefore, and in the interest of emphasizing the
difference-theoretical impetus, it would be more accurate to speak of sys-
tem-environment-theory rather than systems theory (cf. Luhmann 2006).
This implies that neither should the concept of environment be misunder-
stood as a kind of category residual to the system, nor should the concept of
268
15. Kap 12. Koch 05-11-11 16.25 Sida 269
system be interpreted as a kind of “perfect origin” that is independent of the

environment. Both – system and environment – “are what they are only in
reference to each other” (Luhmann 1995f, p. 177).
This reference is constituted by a boundary and at the same time consti-
tutes that boundary between system and environment. The concept of
boundary, therefore, is important for understanding the idea of difference
between system and environment (but also of every other distinction in use),
because “(b)oundaries cannot be conceived without something ‘beyond’;
thus they presuppose the reality of a beyond and the possibility of transcen-
dence” (Luhmann 1995f, p. 28). In this sense the boundary forms the sys-
tem and its environment; it both constitutes the distinction between them
and at the same time is constituted by them. This obviously leads to the idea
of self-reference.
The notion of self-reference

It is one of the basic ideas of Luhmann that every system constitutes its
boundaries in a self-referential manner, meaning that the operation which
draws a distinction between system and environment has to be seen as an
operation of the system itself. This leads to another fundamental paradox:
how can it be assumed that a system is constituted self-referentially when the
boundary between system and environment is fundamental to the constitu-
tion of the system? Without such a boundary there is no system, and with-
out a system there is no self-referential constitution.
Obviously this self-referential statement is not an invention of Luhmann’s,
but, as such, one of the central epistemological problems of modern science
(Rasch 2002). However, as shown above, it also appears as a crucial aspect
of postmodern theorizing. The idea of paradoxical statements is central to
elucidating the difference between modern and postmodern approaches, and
the interesting point is how this problem has been treated in a specific theo-
retical setting – or to use Luhmann’s language: how it is “unfolded” (Luh-
mann 2002f). The recognition and the unfolding of this problem indicate a
certain postmodern understanding of the constitution of knowledge, and
reflections in this regard show why the idea of self-reference is one of the
most interesting concepts in the theoretical work of Luhmann at least with
relation to the postmodern debate.
With Luhmann one can distinguish at least two important and different
forms of self-reference, which relate to different forms of self-referential
operations: reflection and basal self-reference. According to Luhmann, the
operation of constituting a boundary constitutes the unity of a system but
not the identity of that system. The difference between unity and identity is
therefore important and is related to the position of an observer. While the
unity of a system can be observed only by an external observer (an observ-
er observing with another difference), the identity of a system is constituted
269
15. Kap 12. Koch 05-11-11 16.25 Sida 270
by a self-observation of that system building upon the difference that con-

stitutes the self-observing system.
When Luhmann (1990a; 1995f) describes social systems not only as self-
referential but also as autopoietic systems, he focuses on that insight: the
operation of constituting a boundary of a system (which is the system) can-
not be observed and therefore cannot be reflected in that moment of consti-
tution or by the distinction which is constituted in that moment. The idea of
autopoiesis is indeed very helpful in indicating that proposition: it also
stands for a temporalized concept of systems in stating that the process of
boundary constitution is not a nonrecurring operation but has to be repeat-
ed as long as the system exists, and the system exists as long it is able to con-
nect element to element.
In this sense Luhmann’s theory is not only temporalized but also contains
an interesting parallel to the idea of différance proposed by Derrida (1984).
In this relation, the notion of autopoiesis indicates something that Derrida’s
model could not indicate in a strong sense because any indication is simply
deferred. The process of autopoiesis itself has to be presumed unobservable
because any description of this process could only be provided by the use of
another distinction. However, observation and even self-observation are
possible, but not in the sense of a direct representation.
The logic of form

For grasping and providing a theoretical description of this highly sophisti-
cated idea of self-referential systems, Luhmann refers to the laws of form
proposed by the mathematician George Spencer Brown (1979), who has
developed an abstract calculus of distinction. This calculus starts with the
instruction “Draw a distinction!” and this is meant as both, constitute a
boundary and indicate one side of that boundary.
For Luhmann the process of drawing a distinction is an operation that
constitutes a double-sided form by indicating one side. For example, draw-
ing the distinction between system and environment constitutes the differ-
ence (the boundary) between system and environment by indicating the side
“system”. In this sense Luhmann introduces the distinction between opera-
tion and observation by stating that every observation is itself built on an
operation, and to operate means to draw a distinction between two sides by
indicating one of them.
Operations are always blind, meaning that any observation has a “blind
spot” in the sense of an unobserved aspect because it has to draw a distinc-
tion. The blind spot lies in the operation itself and in the necessity that any
observation has to draw a distinction in order to observe. Therefore, the
process of observing an operation requires at least two different operations:
an observed and an observing operation, which cannot be the same. Thus,
the distinction between operation and observation can only be observed by
270
15. Kap 12. Koch 05-11-11 16.25 Sida 271
another observer constituted by another distinction, which itself is blind to

its own constituting operation. Therefore, one can state that there is basal
self-reference but no possibility for basal self-observation. Self-observation
is possible, however, but only by using other differences for observing and
therefore by building on other distinctions.
On the basis of the logic of form, this idea of self-observation can be
described as a re-entry, which means, for example, the re-entry of the dis-
tinction between system and environment in the system (Luhmann 1993g).
In this sense any self-observing system operates with the difference between
system and environment for further observations. These are operations “in”
the system, built on the difference between system and environment.
For Luhmann the notion of re-entry is very helpful because it describes the
paradox of self-observation and shows how such paradoxes are unfolded in
a productive way (Luhmann 1988c; 1995h). Therefore, the notion of re-
entry stands not only for the idea of fundamental paradoxes, but also for the
unfolding of paradoxes. This is always the case when a system is able to
embrace a specific distinction as both the same and different. For the pur-
poses of self-observation any observing system has to distinguish between
itself as observing and itself as observed, and therefore it treats itself as dif-
ferent, whereas the process of self-observation has to neglect the distinction
between observing and observed in order to observe “itself”. Every system
that is capable of self-observation is capable of handling re-entries. And the
criterion for the ability of handling re-entries depends on the distinction in
use. Therefore, the distinction between system and environment is for Luh-
mann a very “performant” one because it provides the operation of obser-
vation with a high ability to re-enter, and thus with the ability to unfold
paradoxes.
In consequence it is not exaggerated to state at this point that Luhmann’s
conceptualisations are similar to postmodern thinking in many respects. On
the other hand, Luhmann’s idea of responding to the question of who is
drawing a distinction with the concept of a self-referential observing system,
which can be observed by other (self-referential) observing systems and so
on, may give the impression that, in the end, Luhmann formulates a theory
that tries to offer the possibility of grasping and representing society by
building on multiple but consciously connected observers. Even if it is not
Luhmann’s idea to represent a complete representation of recent society by
building on the idea of “observing observers”, and therefore of second-order
observation, his conceptualisations nevertheless entail at least three further
aspects that could be regarded as problematic from a postmodern perspec-
tive, as discussed in the next chapter.
271
15. Kap 12. Koch 05-11-11 16.25 Sida 272
Obstacles to a postmodern annexation of systems

theory
As it has been argued until now, there are numerous similarities between the
main tenets of postmodernism and the theoretical approach developed by
Luhmann. However, at least three aspects of Luhmann’s theory seem prob-
lematic from a postmodern standpoint, which are nevertheless also central
to the epistemological positioning of systems theory. These premises are (1)
that there are systems, (2) the statement of a higher performativity of sys-
tems theory and (3) Luhmann’s claim to have formulated a super-theory
with universal validity. These three aspects should not be reconceived in the
form of underlying assumptions (or of an underlying meta-narrative)
because they are not unconscious premises of systems theory, but rather
explicit statements and therefore reflected dispositions of Luhmann’s
approach. All three aspects obviously provoke postmodern opposition – not
only at first sight – and in the end they seem to justify the evaluation of
Luhmann’s theory as an essentially modernist approach. But this is definite-
ly not the case and therefore a closer examination is required.
“There are systems”

As mentioned above, Luhmann’s theory is developed as a constructionist
approach paired with a radical de-ontologisation. Therefore, one might be
puzzled by the fact that systems theory also allows for the statement that
there are systems, meaning that the notion of system in Luhmann’s theory
not only is a conceptual idea, but also should be viewed as indicative of an
object in the (social) world (Luhmann 1995f, p. 2). This statement seems to
contradict the main constructionist and also postmodern insight that neither
language nor any other symbolic system can provide us with a direct repre-
sentation of the world or of any object in the world. Therefore, Luhmann
seems to undermine his own approach with regard to the reification of sys-
tems by giving them an ontological status. But this interpretation does not
go far enough to explain Luhmann’s statement.
For Luhmann it is important to distinguish between a conceptual, seman-
tic level and a structural level. “Thus the statement ‘there are systems’ says
only that there are objects of research that exhibit features justifying the use
of the concept of system, just as, conversely, this concept serves to abstract
facts that from this viewpoint can be compared with each other and with
other kind of facts within the perspective of same/different” (Luhmann
1995f, p. 2).
The use of the notion “system” means an abstraction at both a semantic
and a structural level. But one has to distinguish carefully between a con-
ceptual abstraction (on a semantic level) and a self-abstraction (of a system
itself and therefore on a structural level). In this sense the statement there are
272
15. Kap 12. Koch 05-11-11 16.25 Sida 273
systems suggests a comparison of the extent to which structural forms of

self-abstraction are built upon conceptual abstractions (Luhmann 1995,
p. 2) but it does not suggest the idea that there is a direct representation of
reality.
No term, not even the notion of system, is able to represent reality in a
direct way. Therefore, the statement there are systems does not undermine
this basic epistemological insight, because it merely provides the assumption
that there are self-abstractions in the social world constituted by using (or
better: observing with) the difference between system and environment.
Furthermore, abstraction (or self-abstraction) is an epistemological necessi-
ty, and the application of the notion of system is also for Luhmann only one
possible means of doing so. But it is also Luhmann’s conviction that the
paradigm system/environment is one of the most performant for this pur-
pose, if not the most performant at all. And this leads to a second problem-
atic aspect.
Legitimation by performance
As Lyotard has shown, the problem of legitimation is central to postmodern
critique. In this sense legitimation does not refer simply to political dis-
courses, but has in addition, or even basically, an epistemological impact,
homing in on the blind spots or latencies of any theoretical statement in
order to uncover its unreflected and problematic premises or assumptions.
Even if Lyotard argues that in the end the problem of legitimation could not
be resolved but was always revealed (by statements based on the so-called
narrative knowledge), the search for legitimation still remains central: the
insight that no statement could attain an ultimate legitimation and therefore
has to be treated as deconstructible, does not lead to relinquishing the idea
of legitimation at all.
From this perspective, Luhmann’s claim that systems theory is the better
theory because it provides a more adequate or even powerful description of
recent society (Luhmann 1997a, p. 1095), has to be regarded as highly prob-
lematic because it presumes that this legitimation by performance necessar-
ily leads to the identification of an adequate description with an affirmative
attitude. It is known that this is the main starting point for the meanwhile
classic Habermasian (1990) critique, which deplores the lack of critical
potential of systems theory by drawing on the shortcoming that it has no
normative foundation. For Habermas, Luhmann’s approach is only an affir-
mative one, which in specific respects probably shows what society is but
which could never show what society should be. For Luhmann, on the other
hand, the distinction between a critical and an affirmative position stems
from the old-fashioned idea of the nineteenth century, when theories were
observed by distinguishing reality from ideality. It is Luhmann’s conviction
that for a theory to be adequate in today’s time, this contrast has to be
273
15. Kap 12. Koch 05-11-11 16.25 Sida 274
replaced with the twin questions of “What is the case?” and “What is
behind this?” (Luhmann 1978).
In this sense systems theory is neither critical nor affirmative, but rather
both, and in the end it is more critical than critical theory could ever be,
because trying to cope with the second question “What is behind this?”
implies a permanent self-questioning of any distinction in use. Therefore,
Luhmann’s approach fits right into the discursive setting designed by
Lyotard.
Lyotard himself states that every genre of discourse has a specific end, and
every genre of discourse is chaining phrases in reference to this end. Nothing
else is stated by Luhmann when he evaluates different theoretical approach-
es with regard to their capacity to fulfill a specific goal of a specific system.
In this sense it is the goal of scientific discourse to distinguish between true
and false, and that means showing what is the case and showing what is
behind this.
Analysing Luhmann’s approach from this perspective therefore leads to
the insight that not only the critique of Habermas (1990) can be rejected as
outdated, but also that the postmodern critique of Lyotard (1984) has to be
seen in this respect as a misleading one. The idea of performance and the
form of legitimation Lyotard is aiming at is obviously not the performance
that Luhmann has in mind when he states that systems theory is the better
approach.
Therefore, Luhmann evaluates different theoretical approaches in com-
parison to systems theory by drawing on the distinction adequate/inade-
quate, but for Luhmann an adequate description always has the potential to
answer both questions. Referring to this highly sophisticated requirement,
Luhmann is not stating that systems theory is the only imaginable approach
able to cope, but that it is certainly one of the best, and certainly better than
any approach coming from Frankfurt (Luhmann 2002c). In this respect the
evaluation of better vs. worse has its point of reference in the necessity of
coping with those twin questions, and this leads to the third and final prob-
lematic aspect of a postmodern understanding of systems theory.
A super-theory with universal validity

The third aspect of Luhmann’s approach, which must provoke at least mild
irritation in the postmodern camp, is the claim that systems theory is a
super-theory with universal validity (Luhmann 1995f). Both the idea of a
“super-theory” and the claim of universality are at first sight incompatible
with the postmodern ideas represented above. But even from the perspective
of other components of systems theory, the claim of a super-theory and uni-
versality seems highly problematic. For example, if we focus on Luhmann’s
conceptualisation of complexity as irreducible (1990b), and the consequence
that all necessary reductions of complexity are always selective and therefore
274
15. Kap 12. Koch 05-11-11 16.25 Sida 275
incomplete, then the claim of a universal super-theory (heading obviously

for completeness) also seems incompatible with systems theory. However, in
order to avoid a hasty judgment of incongruity and inconsistency, what
Luhmann may have precisely intended by referring to “super” and “univer-
sal” deserves a closer examination.
There is no doubt that it would be problematic if systems theory claimed
to be universally valid, but Luhmann relates and reserves the claim for uni-
versality for a very specific aspect. It is not his intention to formulate an
exclusive claim of truth with regard to other theoretical approaches. It is
neither his intention to state that systems theory could reflect the complete
reality of an object, nor to claim that it could exhaust all possible forms of
recognizing an object. None of these claims are intended, but systems theo-
ry does claim universality for its grasp of its objects (Luhmann 1995f, p.
xlvii). In this sense systems theory as a theory of social systems has to
include all aspects of the social world.
This may appear as a very subtle difference, but it is nevertheless of great
significance. According to Luhmann, of course, a super-theory claiming uni-
versality is also selective with respect to its objects, and does not imply that
the chosen selection is the best and only one. But super-theories with a claim
to universality are constructed as self-referential theories, meaning that they
appear as their own objects. Furthermore, they are constructed in such a
way that they do not exclude, but rather include, other competing approach-
es. Therefore, super-theories are totalizing not in the sense of exclusion, but
in an including manner.
In this respect it is imperative for totalizing theories to have the potential
to reconstruct other theoretical approaches (even those with contradictory
statements) in their own terminology. The basic principle of this attribute
can be clarified by contrasting systems theory with another approach, for
example critical theory. In this sense one can reconstruct the critical theory
of Habermas (1984) in terms of systems theory by drawing on the main dis-
tinctions in use (e.g. life-world vs. system; critical vs. affirmative and so on)
and showing what can be made visible and what has to remain opaque by
doing so. On the other hand, starting with critical theory one can accept cer-
tain elements or statements of systems theory but one has to reject all the
others. While Luhmann’s approach enables us to see, observe, reconstruct
and at the very least deconstruct critical theory, Habermas forces us to
choose a side in the hope that we have picked the “right” one.
Certainly this is a slightly exaggerated contrast, but it reflects an impor-
tant difference in the theoretical self-understandings of each approach con-
cerning the difference between inclusion and exclusion. In this respect, how-
ever, it seems very plausible that the idea of a universal super-theory as for-
mulated by Luhmann realises basic issues of postmodern thinking.
275
15. Kap 12. Koch 05-11-11 16.25 Sida 276
Considering the aspects discussed above, systems theory, which at first

sight seems to entail various problematic statements with regard to post-
modern theorizing, at second sight appears to be a connectable variation
and also concretization of main postmodern ideas. However, this evaluation
does not imply that systems theory should automatically be seen as a post-
modern approach.
(Not) drawing a distinction: between Luhmann and

the postmodern ideas of difference
A comparison between Luhmann’s theory and the approaches of Foucault,
Lyotard and Derrida as the main representatives of postmodernism could be
accomplished on many different levels. The following consideration will
focus primarily on the conceptualisation of difference in the respective the-
ories.
Luhmann has developed his constructionist and difference-theoretical
approach by referring to many interdisciplinary ideas that do not stem from
the same philosophical background that constitutes the focus of the debate
on modernism and postmodernism (Luhmann 1993a). Therefore, to resitu-
ate Luhmann in this debate and to focus on the similarities between Luh-
mann’s approach and the main postmodern frame of theorizing, one has to
pay attention also to the differences in this regard. In so doing, however, one
starts with the insight that Luhmann’s theoretical self-understanding is basi-
cally not a modern one but entails a specific “consciousness of distance”
with relation to the main modern assumptions and in this respect also
reflects the postmodern critique of modernism. Therefore, the difference
between Luhmann’s approach and critical theory is not only one of systemic
modernism vs. critical modernism, but also, or perhaps even primarily, this
difference points to a position at a distance from modern epistemologies.
The general assumption of complexity (which at the same time undermines
the Habermasian ideas of both life-world and discourse), the assumption of
inevitable blind spots inherent to any observation or communication (which
undermines the idea of communicative self-transparency) and so on – all of
this constitutes a theoretical framework which makes a clear difference with
respect to a modern understanding. But whether Luhmann’s approach could
or should, therefore, be treated as postmodern is another question to be dis-
cussed.
However clear and evident the similarities between certain elements of
Luhmann’s theory and the frame of postmodern thinking presented above
may be, some differences are also apparent. In this sense the relationship
between Luhmann’s conceptualisation of an adequate description of recent
society and the Foucauldian idea of the modern episteme “man” (Foucault
276
15. Kap 12. Koch 05-11-11 16.25 Sida 277
1973) and the consequences for every knowledge claim to be pseudo-knowl-

edge, seems to be the most appealing. In Foucault’s view, the condition of
modern society and modern science brings about an inevitable oscillation
between different oppositions (“doublets”), such as the empirical and the
transcendental, which cannot be unfolded and do not lead to positive
knowledge (Dreyfus/Rabinow 1983, pp. 25–43). It is certainly not Luh-
mann’s conviction that there is something like truth in itself, but Luhmann
would not describe this oscillation between two different sides of a distinc-
tion as a hopeless search for positive knowledge. Instead, and to Luhmann’s
way of thinking, this modern condition leads to the necessity of unfolding
the paradoxes of modern knowledge, whereas for Foucault the answer to
these paradoxes seems to lie in his concept of power.
In this sense, redescribing Luhmann’s theory in terms of Foucault’s idea of
power (1979) might produce interesting insights, but the assumption that
every blind spot or latency could be regarded as an indicator of the existence
of (unobtrusive) power has to be rejected on the grounds of its one-
sidedness. In terms of a theory of distinction, the centrality of power has to
be replaced at the very least with the idea that even power must have
another side; power therefore refers to another difference, and the other side
of that difference could not be called power again.
With regard to the work of Lyotard, there are more and far-reaching sim-
ilarities one can grasp by paralleling the constitution of genres of discourse
and the idea of functional differentiated systems. In this sense Luhmann’s
conceptualisation of societal subsystems, based on a specific binary coding
(like truth/untruth for the system of science), fits into the idea of genres of
discourse offered by Lyotard. This does not mean that Luhmann’s and
Lyotard’s theories bear identical descriptions of modern society, but on a
conceptual level they offer a very similar idea of difference. Therefore, how-
ever different these theoretical backgrounds may be, one can find similar
ideas suggesting an analogy by referencing the problem of legitimation and
the problem of chaining.
While Lyotard (1984) has unfortunately situated Luhmann’s approach in
a more or less unbroken tradition of older ideas of systems theory, which
basically refer to procedural forms of legitimation, Luhmann (1997a, p. 90)
himself has emphasized the common elements in both approaches – for
example, their similar positioning of the subject. However, Luhmann also
criticizes Lyotard’s presumed neglect of the fact that every chaining opera-
tion constitutes a difference between system and environment, which has to
be reflected in the system or the discourse (Luhmann 1997a, p. 90). Here,
Luhmann refers to Lyotard’s concept of the differend, for which the prob-
lem of chaining is central, but Lyotard is also engaged – as in The
Postmodern Condition – in legitimation problems and problems of self-
277
15. Kap 12. Koch 05-11-11 16.25 Sida 278
reference. The aspect that Luhmann presumes to be neglected by Lyotard

indeed refers to the problem of legitimation, which is also central for
Lyotard.
Examining both distinctions, the problem of legitimation and the problem
of chaining therefore reveal a more similar conceptualisation of the discur-
sive description and the systemic description as and of an adequate idea of
recent society. And above all, Lyotard’s proclamation of the end of meta-
narratives fits in perfectly with Luhmann’s idea that the modern society has
lost its capacity to represent the whole through one of its parts. In the end,
the primary source of friction may be that Luhmann would not call this
(structural) situation “postmodern”, but the situation is qualified and
described in nearly the same way by postmodern semantics and systems the-
ory (Luhmann 1990h).
Last but not least, with regard to the relationship between systems theory
and deconstruction, one can again grasp a deeper and more significant dif-
ference. While the basic idea of deconstruction can be applied to the prob-
lem of legitimation, the deconstructive program of Derrida is certainly not
reserved only for revealing paradoxes of self-reference in the sense of the
problem of legitimation. For that reason also differences constituting genres
of discourse (Lyotard) or the functional differentiation of society (Luhmann)
are presumed to be deconstructible and are therefore potential objects of
deconstruction. This poses a problem for a sociological theory such as
Luhmann’s, because so-called proven forms of observation are elementary
for theorizing, and therefore notions like “society”, “system”, “environ-
ment” are not called into question at a certain level of theory-building, even
if they are not taken for granted as assumptions in that theory. For Luhmann
it makes no sense to examine and deconstruct the notion of system, for
example, whereas for Derrida this is certainly a highly problematic concept.
But before focusing on this different level of deconstruction, one has to
state that there are clear similarities or analogies between Luhmann’s and
Derrida’s theories, mostly with regard to the problem of legitimation and,
therefore, of self-reference. Luhmann tried to work out these parallels by
reconstructing the theory of deconstruction as a second-order observation
(Luhmann 1993b). At the same time, however, Luhmann tends to build up
a critical distance to deconstruction by questioning its theoretical potential
to construct. “Deconstruction seems to recommend the reading of forms as
differences, to look at distinctions without the hope of regaining unity at a
higher (or later) level, or without even assuming the position of an ‘inter-
pretant’ in the sense of Peirce. Deconstruction seems to be adequate for an
intellectual climate heading toward cultural diversity. But are there any firm
rules and any hope for results in the deconstruction business – that is, are
there any framings that are not themselves deconstructible? Or would apply-
ing deconstruction lead only to reflexivity, recursively, and self-reference
278
15. Kap 12. Koch 05-11-11 16.25 Sida 279
resulting in stable meanings, objects, or what mathematicians call eigenval-

ues? It seems that there is only différance” (Luhmann 1993b).
In this sense Luhmann clearly perceives the similarities between his own
theoretical approach and deconstruction, but on the other hand he clearly
states:
1. That the focus of deconstruction is all too narrow. “It is almost a one-
word discussion, or a text/context discussion, where deconstruction is
the text and the history and the usages of the word the context”
(Luhmann 1993b, p. 767).
2. That the deconstructive rejection of logocentrism cannot be paired with
a general rejection of all formalisms and logics. Here Luhmann refers to
Gotthard Günther. “Günther is not satisfied with the fuzziness of verbal
acoustics and paradoxical formulations and tries, whether successful or
not, to find logical structures of higher complexity, capable of fixing new
levels for the integration of ontology (for more than one subject) and
logics (with more than two values)” (Luhmann 1993b, p. 768).
3. That it is not the task to attempt to clarify what Derrida possibly means.
“Derrida himself distinguishes le sens de sens (the meaning of meaning)
and signalisation, which could mean ‘making and deferring distinctions’
on the one hand and ‘indications’ on the other. His writings are écriture
in his one sense (that is, also ‘inscriptions’). It remains open what they
indicate” (Luhmann 1993b, p. 768).
In the end, it is this openness and indeterminacy that is irritating for

Luhmann as he considers deconstruction. Luhmann can understand why the
writing of Derrida has to be self-deconstructive in order to be able to show
how deconstruction works and that this idea of deconstruction fits right into
the philosophical project of Derrida. But for Luhmann this does not mean
that there could not be further steps built on distinctions, which may be
regarded as deconstructible, but which do not have to be deconstructed
because it makes no sense to deconstruct everything. On the contrary, it is
Luhmann’s idea that the ongoing processes of observing constitute so-called
eigenvalues, which represent successful, performant or even proven forms of
observation, and which are therefore interesting for purposes of being con-
nected with further distinctions. Therefore, Luhmann wants to reconstruct
(and integrate) deconstruction as a specific form of second-order observa-
tion. With this idea he goes beyond deconstruction – or he falls behind it,
depending on whether one endorses the perspective of Luhmann or the per-
spective of Derrida.
Whereas Derrida is fundamentally sceptical about any idea of appropria-
tion and instrumentalisation of deconstruction, for Luhmann, Derrida is
permanently engaged in obscuring the preconditions for the use of
279
15. Kap 12. Koch 05-11-11 16.25 Sida 280
distinctions. “But also one can pursue the contrary objective of making the
architecture of theories as clear as possible so that an observer may decide
whether to follow their suggestions or to choose at certain points an alter-
native path” (Luhmann 1993b, p. 768 f.). One can attribute this difference
between Luhmann and Derrida to their different theoretical projects, with
Derrida (1978) philosophically engaged in deconstructing the tradition of
metaphysical thinking and Luhmann (1997a) engaged in a sociological the-
ory of modern society. However, there will remain a basic difference that
cannot be resolved by drawing on the different perspectives induced by a
discipline’s interests.
While Luhmann conceptualizes his theory of distinction, which confirms
the emergence of proven eigenvalues, by stating that there is a specific form
of societal differentiation which cannot be negated (and this is similar to the
discursive setting stated by Lyotard) without losing sight of any sociological
idea of stating what the case is, for Derrida any claim of whatever the case
may be is deconstructible. However important the second part of the above-
mentioned twin question is for Luhmann, it is not acceptable for him to
“destruct” the first part by the second.
Fitting elements of different theories of difference

The discussion presented here has tried, despite all the evident differences,
to demonstrate that there are indeed many similarities between postmodern
conceptualisations and Luhmann’s difference-theoretical foundation of sys-
tems theory. Therefore, the classification of Luhmann’s approach as a tradi-
tionally modernist one, providing only affirmative descriptions of a “cold
and mechanical process” of rationalisation, has to be rejected. However
great the differences between the primary postmodern ideas and Luhmann’s
systems theory are, there are more significant indicators, which justify the
view that systems theory has to be positioned at a clear postmodern distance
from modernism. In this sense, and as a summary, the following five ele-
ments could be seen as a framework for such an approach forming a dis-
tance to modern epistemologies by drawing on the idea of difference.
1. There are two fundamentally different forms of difference: a) one be-
tween different discourses (or systems), b) the other between a discourse
and its otherness (or between system and environment). The first one
constitutes and is constituted by a problem of chaining; the second one
constitutes and is constituted by a problem of legitimation.
2. A theory of difference has to deal with both problems and their relation-
ship without claiming or referring to a meta-narrative. The relationship
between discourses (or systems) cannot be described without involving
the self-relation of a discourse (or a system) to its otherness (or environ-
ment).
280
15. Kap 12. Koch 05-11-11 16.25 Sida 281
3. A theory of difference claims a basal form of self-reference. On a con-

ceptual level this claim leads to the recognition of paradoxes and to the
insight that every knowledge claim is built upon the unfolding of a par-
adox.
4. Such a theory of difference has the status of a universal super-theory in
the sense that it tries to include and not to exclude itself and competing
theoretical approaches as an object of itself.
5. Such a theory of difference is not just a meta-theory that only decon-
structs all claims in terms of differences, but one that also constructs all
its objects and concepts as differences to which further differences can be
connected.
In the end one may state that there are differences that make differences, but
that there are perhaps more differences that “unify” postmodern thinking
and Luhmann’s systems theory.
281
16. Kap 13. Bommes 05-11-11 16.26 Sida 282
Chapter 13

and Network Theory
Michael Bommes & Veronika Tacke
Introduction
Networks have been gaining remarkably in empirical relevance in a variety
of societal realms during recent years. Accordingly, the term network has
become widespread and even fashionable within the social sciences. This is
true, not least for organizational research, to which various social sciences
contribute.
Scholars of political organization have identified a huge number and vari-
ety of “policy networks” on different levels of policy formation. In view of
this, they have described the state as a “network state” (Teubner 1999).
Parallel to this, the diversity and growing relevance of so-called “business
networks” exposed by an impressive number of empirical studies suggested
the advent of a “network economy” (Hessinger et al. 2000). Beyond this, the
“network revolution” (Teubner, ibid.) seems to have reached nearly all
societal realms of organization.
Although scholars of organizational network analysis noted, at an early
stage, the conceptual challenges that network phenomena provide in con-
sideration of current sociological knowledge regarding organizations
(Mayntz 1992), efforts to bring together theoretically elaborated concepts of
organization, on the one hand, and general concepts of the network, on the
other, remained few and far between. Instead, the respective research and
discussions primarily centred around the idea of networks as a third form of
coordination: While economists referred to economic exchange and a
respective contractual continuum along which networks are a “discrete
structural alternative”, i.e. positioned “between” markets and hierarchies
(Williamson 1991), sociologists tended to emphasize the role of trust and
solidarity, thus denoting that networks emerge “beyond” (ibid.) market
price and organizational command. Networks seem to be “more social”
(Powell 1990) than other social forms of co-ordination. In search of theo-
retical tools, researchers who followed this thread of discussion turned to a
sociological network approach that is well known for its concept of
282
16. Kap 13. Bommes 05-11-11 16.26 Sida 283
“embeddedness” (Granovetter 1985). Though the concept was not devel-

oped particularly for this context, it became famous, if not fashionable, even
in the context of organization research.
Concerning the difference between basic theoretical concepts and empiri-
cal evidence, the reference (if not fallback) of organization research to net-
work concepts has sometimes resulted in confusion. The so-called “dissolu-
tion of organizational boundaries”, which has become a common observa-
tion and parlance during recent years, exemplifies this. Whereas describing
organizations means referring to system boundaries, network theory
inevitably dissolves the organization, as well as any other social structure,
into nothing but individual network relationships. Ignoring the constitutive
distinction between members and non-members, organizational boundaries
disappear. Thus, the dissolution effect is, primarily, a necessary effect of the-
oretical changeover. As a consequence, the question of whether or not some
kind of structural change (i.e. the dissolution of boundaries or the emer-
gence of networks) has taken place is impossible to resolve with regard to
empirical evidence.
It is not surprising that scholars of systems theory are more sceptical and
hesitant regarding the idea of dissolving system boundaries. At the same
time, this may be one reason why systems theorists remained hesitant
regarding the theoretical challenges that the network phenomenon poses.
Besides a lot of rather loose and uneven terminological uses of the network
concept (see Kämper/Schmidt 2000, pp. 219 ff.), only two theoretical sug-
gestions have been made. They both refer to the particular case of organi-
zational networking. Whereas Gunther Teubner (1992, 1996) suggested de-
scribing networks as a “real emergence-phenomenon” that emanates from a
“re-entry” of contract and organization, Kämper and Schmidt (2000), in-
stead, reconstructed networks as a “structural coupling” of organization
systems which makes use of interaction systems. Both suggestions are high-
ly (and all too) specific. They empirically encompass very different circum-
stances (business networks and policy networks), at the same time lacking
an angle for theoretical generalizations going beyond these particular cases.
Regarding the state of the art, in any further attempt to develop a systems-
theoretical concept of networks, at least two interrelated requirements
would have to be considered. In order to grasp the huge variety of empirical
network cases appearing within and between differentiated societal con-
texts, an adequate network concept must primarily be formulated as a suf-
ficiently general concept, which can then be re-specified in order to cover
this variety. Secondly, in order to meet this challenge, the concept must take
into account basic insights and eligible claims from the theory of society
which, in the case of systems theory, is a theory of functional differentiation.
In view of the state of the art, it makes sense to turn to the theory of society
283
16. Kap 13. Bommes 05-11-11 16.26 Sida 284
as a framework, and thus take the long way round in order to arrive at the
particular case of so-called organizational networks.
The chapter starts with some major differences between the sociological
network approach and sociological systems theory. Picking up some basic
ideas from the sociological network approach and reformulating these in
terms of systems theory, the first part of the chapter aims to develop a sys-
tems-theoretical concept of networks (see Tacke 2000). With regard to pri-
mary structures of societal system differentiation and respective modes of
inclusion, the chapter argues, on the one hand, that networks are a second-
ary form of order in modern society: Networks presuppose social systems
(i.e. organizations and function systems) as far as it can be shown that net-
works assume their form from a reflexive combination of options that are
represented by (individual or organizational) addresses, and that these
options are due to the social addresses’ particular profiles of inclusion and
exclusion in various social systems. At the same time, networks supplement
social systems: Whereas modern societies’ basic systems rest upon a “pri-
macy of problems” (which then guides the search for the relevant social
addresses), networks follow a “primacy of addresses” (which allows the cre-
ation of particular new problems and possibilities which the underlying sys-
tems cannot provide).
Against this theoretical background, the second part of the chapter reveals
aspects of coupling networks and organizations. Depending on the social
addresses involved, which can be personal or organizational, “personal net-
works within the organizational context” are distinguished from “organiza-
tional networks”. Whereas the former can either remain “silent” or be
acknowledged through organizational decisions, the latter are necessarily
brought about by organizational decision.
Social-theoretical network approach and theory of

societal differentiation
Theoretical divergences
Systems theory is a social theory which at the same time comprises a socio-
logical theory of (modern) society. This means that the theory not only
describes its objects of research on the level of basic concepts (e.g. system,
observation, autopoiesis, communication), but further specifies these basic
concepts with regard to modern society. By explicitly taking into account the
primary structures of societal differentiation, especially function systems,
organizations, and systems of interaction, systems theory can analyse their
interplay and, in doing so, describe the characteristics of modern society.
Compared to this, the sociological network approach is a social theory with-
out a theory of society. It does not differentiate modern society from its
284
16. Kap 13. Bommes 05-11-11 16.26 Sida 285
historical predecessors. Instead, the respective distinctions are explicitly

denied by scholars of network theory (see Granovetter 1985).
There is no dissent between scholars of systems theory and those of the
sociological network approach regarding the assumption that networks had
already existed in pre-modern societies. Instead, it would be mistaken from
both points of view to regard networks as a specifically modern phenome-
non. However, in contrast to network theorists, systems theory emphasizes
that the conditions of emergence and reproduction and the societal meaning
of networks (as well as other structural forms) have radically changed with
the transition from societies ordered by segmentation and the stratification
of families and groups to modern society which is primarily differentiated
into function systems, organizations and interactions. From this point of
view, the weakness of purely socio-theoretical approaches is that they neg-
lect the transition from one form of societal differentiation to another, as
well as its structural implication for networks. Therefore, they overestimate
historical continuities.
At first glance, it seems evident that the idea and concept of the network
inherently accounts for societal discontinuity, which scholars of the network
approach ignore. On the one hand, the network concept is founded on basic
assumptions, which are assumed to be universal, but which are actually only
of primary structural significance in the case of stratificatorily differentiated
societies. The older stratified societies conditioned the individual’s partici-
pation in communication according to status and affiliation, and actually
defined social relationships according to “status-role-sets” (Burt 1982), as
scholars of network theory phrase it. Network theorists’ actual reference to
conditions of stratificatory differentiation becomes even more apparent
when they circumscribe networks using a “causal primacy” of social-
structural “positions” and “social bonds” (Hessinger et al. 2000, p. 31;
Thorelli 1986).
On the other hand, network concepts obviously presuppose social condi-
tions which have emerged for the first time in modern society, since, within
this society, social relationships are constitutively established in a reflexive
way. Network theory will only apply and make sense on the precondition
that one’s position is no longer conditioned and bound by the societal struc-
ture, as was the case in stratified societies. In other words, network theory’s
empirical subject is precisely the multiplicity of occasions which allows for
the building of networks – in modern society.
The primacy of the societal differentiation of meaning

Looking at it this way, network theory seems to reflect on a particular
modern experience and view: former social bonds become fluid – and this is
the precondition for the emergence of new, reflexive bonds. This fluidisation
abolishes the “causal primacy” of socio-structural positions and status-
285
16. Kap 13. Bommes 05-11-11 16.26 Sida 286
bonds while replacing it with another societal primacy, henceforth condi-

tioning the individual’s options regarding social participation and, as a con-
sequence, the social contexts of network building.
The transition from pre-modern to modern society is accompanied by the
detachment of individuals from social positions assigned by societal struc-
ture. Instead, new forms of conditioning develop which selectively specify
the social processes of communication and the conditions of individual par-
ticipation. Theorists of social differentiation – from Marx, Durkheim, or
Simmel to Weber and from Parsons and Habermas to Luhmann – have
demonstrated that the forms of conditioning of the social process are forms
of social meaning driven by the logic of social differentiation. Max Weber
has already emphasized that social relationships are based on the content of
meaning which mutually adjusts people to each other. He describes modern
society’s primary form of differentiation as a “differentiation of value
spheres,” which is the result of a process of social rationalisation that took
off in Europe during Early Modern Times (Weber 1976). Later, Niklas
Luhmann conceptualised this differentiation process more strictly theoreti-
cally and methodologically as an evolution of functional systems (Bommes
1999, p. 66). He argued that social relationships – together with the differ-
entiation of meaning – are dragged into a certain “maelstrom of inclusion”
of functional systems (Luhmann 1997a, p. 738). That is to say, relationships
are no longer determined, once-and-for-all, by the individual’s affiliations to
stratified families and groups but become selectively specified and condi-
tioned by different social contexts of meaning – whether these are function-
al (structured by e.g. economic, scientific, legal, political, educational, or
religious codes of communication), organizational (structured by decision
and membership) or interactive (based on reflexive perceptions of co-present
participants).
The sociological network approach ignores the processes of societal dif-
ferentiation although these imply radically different modes of the individ-
ual’s inclusion in society and transform the conditions of his or her commu-
nicative participation in society. The structural form called “functional dif-
ferentiation” is only possible when individuals no longer assume only one
position in social structure that defines the horizon of their social options
(affiliation, total inclusion), but are able and forced to participate in diverse
contexts of communication, each of which operates under its own restricted
criteria of individuals’ relevance (partial inclusion). The structural precondi-
tion of this multi-contextual form of inclusion is the individual’s exclusion,
i.e. his or her detachment from feudal bonds. In modern society, then, the
contexts of the various systems and the respective aspects of meaning deter-
mine whether or not individuals are addressed in communication, and
whether or not they are treated as relevant (inclusion) or irrelevant (exclu-
sion) as regards further communication (Luhmann 1995b, p. 241). The
286
16. Kap 13. Bommes 05-11-11 16.26 Sida 287
systems themselves regulate what topics of communication they take up and

what kinds of positions these particular communications assign to people
(Luhmann 1997a, pp. 738 ff.). These positions change with the systemic
contexts of meaning, and they are not transferable arbitrarily from one con-
text to another: A professor has to stay in the hospital waiting room until
the physician has finished treating a student. If there is a causal primacy at
all, it is in principle not concerned with positions and bindings, but refer-
ences of meaning, and thus social systems.
Systems and networks as complementary social structures

Taken together, the point where network theory and systems theory disagree
concerns the question of how the individual’s participation in the social
world is regulated. Nevertheless, the two theories do not diverge accidental-
ly. Instead, they show distinct complementarities – at least from a systems-
theoretical perspective.
Systems theory emphasizes that modern society’s primary structure con-
sists of differentiated social meaning, i.e. social systems dealing with specific
(functional, organizational or interactive) problems. Individuals become rel-
evant within the scope of those systemic problems. This means that they are
used as addresses for communication according to the specific meaning and
criteria of the systems’ operation. The primacy of problems, by which sys-
tems are defined, determines a person’s relevance as an address for commu-
nication, thereby conditioning “social relationships” due to the specific
social meaning (e.g. the problem of illness and the roles of patients and doc-
tors in the health service). Compared to this, network approaches assume a
primacy of addresses, thereby ignoring the specific meaning-related condi-
tioning of social relationships by social systems in modern society.
Theoretically they leave open the question of what kind of meaning social
relationships process and what kinds of problems they deal with or afford.
The difference between network theory and systems theory could be
understood as a theoretical incommensurateness – and left at that. Indeed,
the difference and the complementary deficits can also be seen as a challenge
and turned into sociological productivity. The following sections try to take
an initial step in this direction and will be conducted from the perspective of
systems theory. Functional and organizational systems, on the one hand, and
networks, on the other, will be considered to be complementary forms of
social structure that are based on different structural primacies (problems,
addresses).
As mentioned before, the establishment of functional and organizational
systems is due to a primacy of problems, while network constitution features
a primacy of addresses. The crucial and initial point of network emergence
is the observation that addresses – the capabilities they represent in commu-
nications – can be combined. Networks thus emerge from the consideration
287
16. Kap 13. Bommes 05-11-11 16.26 Sida 288
of options and indeterminate potentials which become accessible through

the social addresses at hand and the social possibility of combining them.
From this perspective, networks are actually a form of structure that is
based on a primacy of addresses, as is assumed by network theory. However,
our initial assumption, i.e. that the constitution of functional and organiza-
tional systems and networks is complementary, does not imply the aban-
donment of the societal primacy of functional differentiation, which systems
theory denotes, but rather presupposes it. This can be seen when consider-
ing more closely the process of building networks. Formally, the driving
force for network building can be found in those social potentials which
become accessible by linking the possibilities stemming from heterogeneous
system contexts of meaning. This accessibility is based on the inclusion of
the networks’ members in these systems. It entails two presuppositions: the
formal freedom to combine social addresses and the existence of social sys-
tems that provide the kind of resources used in social networks and made
accessible to their members.
Against this backdrop, we assume an asymmetric relationship of func-
tional and organizational systems, on the one hand, and networks, on the
other. If networks, structurally speaking, follow a primacy of addresses, then
theoretically this needs to be conceived of as a secondary form of order in
modern society. The structural reason for this is to be found in the mode of
generating social addresses in modern society. This mode forms the basis for
the freedom to combine these addresses in networks, and it is a product of
system differentiation – and thus presupposes these primary societal struc-
tures. Social addresses result from the modes of inclusion and exclusion of
individuals in a functionally differentiated society and in particular in its
functional and organizational systems. Networks certainly cannot be causal-
ly explained by structures of societal differentiation. Nevertheless their
emergence and growing importance is due to the primary differentiation
structures of modern society.
Network emergence as reflexive combinatorics of

addresses
In order to meet a great variety of network phenomena theoretically and to
grasp even very fluid empirical forms of social networks, the network con-
cept proposed here defines networks in a minimalist and formal way.
Therefore, it is necessary to distinguish the network concept from mere
address connectivities (Stichweh 2000a), which have misleadingly been
characterised as “networks” in the so-called “small world” research
(Milgram 1967; Watts 1999) as well as in research into scientific citations
(e.g. Baldi 1998). The simple fact that scientists cite other scientists’ articles
288
16. Kap 13. Bommes 05-11-11 16.26 Sida 289
and that individuals can name more than 1,000 people as acquaintances
does not yet match the network concept as it is defined here. In order to dis-
tinguish clearly between address connectivities and networks, the next sec-
tion further examines what can be defined as a social address, what kind of
social addresses are considered during network building, and how networks
can emerge and stabilise by combining social addresses.
Social addresses – individuality and ubiquity

A social address is a more or less elaborated profile of attributes and forms
of behaviour that personalised others are provided with and identified by
during communication. Addresses are attributional artefacts produced with-
in communication for communication (Luhmann 1995a; Fuchs 1997a).
While proper nouns basically indicate social addresses, they become “com-
plex addresses” when accumulating further distinctions (Stichweh 2000a).
Pertaining to this, there are, on the one hand, those components subsumed
under the idea of addresses in everyday life. Their horizon of meaning refers
to communicative accessibility, at least in terms of a postal address, a tele-
phone number and increasingly also an e-mail address. However, addresses
only gain a social contour and individuality from an individual’s current
inclusions and exclusions. His or her profile of inclusions and exclusions
refers back to the individual history and career of participation in differen-
tiated system contexts of society. At the same time, it indicates the horizon
of the individual’s potential relevance to particular kinds of communication
in the future.
Concerning the individual profile of addresses, it is not only important
that individuals participate to varying degrees in economic, political, legal,
educational, familial or scientific communication. It seems even more impor-
tant that participation can take place in 1. different functional roles
(achievement/audience), 2. particular organizations (with reputation in
some fields but not in others), and 3. different parts of the world (where
individuals may be citizens or foreigners). Thus, according to its primary
structure of differentiation, modern society constitutes social addresses in a
poly-contextual way (Fuchs 1997).
The implications are twofold: On the one hand, this means that address-
es emerge as highly individual profiles of inclusion and exclusion which
open up a more or less large variety of options for communication (and
exclude others at the same time). On the other hand, the poly-contextual
character of individual addresses implies that they usually do not become
activated completely or in all their particulars. Instead they become selec-
tively relevant, depending on the respective contexts of meaning (the family,
the firm, the sports club, the hospital, the court, the parliament etc.) which
determine the aspects remembered and those either not recalled or un-
known.
289
16. Kap 13. Bommes 05-11-11 16.26 Sida 290
Besides the poly-contextual character of social addresses, the ubiquitous

relevance of their emergence also has to be noted. Social addresses come into
play whenever communicative meaning is attributed to someone whom the
communication can conceive of as a unitary instance of the production and
reception of utterances (Fuchs 1997; Stichweh 2000a). This has at least
three important consequences for a concept of networks set up in a systems-
theoretical frame:
1. There are two instances that communication can treat as “personalised
others” in modern society: Not only individuals but also organizations
are units of communication that utterances can be unitarily ascribed to.
To put it more simply, organizations can communicate (outward) to
others in their environment. This neither holds for function systems nor
for systems of interaction. For networks, according to their core activity
of combining addresses, this implies that they can rely on individuals as
well as on organizations.
2. If any participation in communication is inextricably associated with a
social address, and if the construction of social addresses bears upon the
communicative necessity of attributions, social addresses will be consti-
tutively linked with the differentiation of communicative meaning and
the respective modes of inclusion. Thus, provided that communication
constantly and inevitably produces addresses, it seems that networks are
dependent on the structures of system differentiation, at least if we
assume that the main reasons for initiating the formal core activity of net-
works (i.e. combining addresses) can be found in the social options re-
presented by these addresses.
3. Network building relies on the difference between the basic and the
reflexive use of addresses in communication. The basic relevance of
addresses is clear: As soon as communication starts, inclusion takes place
– somebody has to be treated as relevant – and as soon as inclusion takes
place, communication utilises and disposes of addresses – this somebody
has to be addressed in one way or another. The reverse does not neces-
sarily apply: Communication can be concerned with addresses as its topic
without including the “somebody” being talked about. Empirically, this
is a trivial issue. Addresses and their specificities can be inquired about
during communication, and can thus also be provided for further com-
munication: “Concerning X’s problem, I suggest contacting Y in Z. I
have attached his e-mail address.” Indeed, it is this kind of reflexive
handling of addresses that seems especially crucial in network building.
290
16. Kap 13. Bommes 05-11-11 16.26 Sida 291
Individual address books as the potential for network-building

Modern society has abolished principal restrictions regarding communica-
tive addressability and, moreover, developed media of communication with
a global range and scope. Based on that, addresses can be activated quite
suddenly and unexpectedly in time and space. Nevertheless, the reflexive use
and activation of addresses does not happen completely accidentally. A par-
ticular form of social memory has to be taken into account here: the social
phenomenon of address books. Regarding network-building, two forms can
clearly be distinguished:
Directories are one form of address book, and are structured according to
categories. They are more or less like public address registers, and serve a
more technical aspect of communicative accessibility, although, mostly with-
out uncovering particular aspects of social addressability or at least reduc-
ing these to a minimum. Unless one knows whom to call regarding what
matter, a local telephone directory with alphabetic sorting is rather useless.
Compared to this, yellow pages are sorted by category as well, but on the
basis of functional aspects of addressability. The categories (healthcare,
lawyers, firms) and subcategories (e.g. internist, ear specialist, cosmetic sur-
geon) give access to specific communications, while at the same time sig-
nalling that addressability cannot be relied on in other respects.
Differing from public address registers, individual address books save
addresses for one certain address, private or organizational. Individual
address books are unique. This is due to their individual address composi-
tion, but more importantly, because every registered address refers to a par-
ticular history of contacts which adds meaning to each address and marks
the exclusive respects of addressability that further communication can
resort to and – as could be the case – can count on. Besides functional
addresses saved for retrieval (doctor, lawyer, hairdresser), individual address
books comprise addresses characterised by their mobilisability which makes
them an important resource underlying networks. For example, the address
of the doctor I rely on from time to time differs in this respect from the
address of my acquaintance who, among other things, is also a physician.
The horizon of meaning called acquaintanceship allows mobilising him or
her even with regard to questions and communicative expectations normal-
ly excluded from the framework of professionally organized medical care
and respective asymmetric roles (doctor/patient).
The difference between functional and mobilisable addresses, however, is
neither fixed nor literally codified. Obviously, this difference cannot be read
from address book letters, but can at most be inferred from its written sub-
strate. Sociologically, this inference does not refer to psychic cognition but
is a social achievement. The peculiar mutual meaning of addresses is stored
in social memory and must be restored by communication. Accordingly,
291
16. Kap 13. Bommes 05-11-11 16.26 Sida 292
even the distinction between functional and mobilisable addresses must

always take place anew by means of communication – and can likewise fail.
Network constitution
The individual address book is not a network in itself, but rather the result
of the reflexive use made of addresses. It keeps at the ready the address pool
for networking and is in many respects also a document of networking, i.e.
of the addresses implied. Network-building starts out from a reflexive han-
dling and restoring of addresses which are expected to be mobilisable. The
addresses are observed selectively due to the possible approachable options
that their profiles of inclusions and exclusions announce. The pool of mobil-
isable addresses is characterized by a range of diffuse, indeterminate hori-
zons of possibilities. Although the specific poly-contextual profile of each
address also constrains possibilities, new options can suddenly be discovered
in its frame: One’s former colleague lives in a city, where one is looking for
a flat; the student of media studies is looking for an internship and the com-
munication just remembers that the professor’s brother works for a broad-
casting company, and so on.
Access to perceived options and the first steps towards networking are
precarious. Is it still possible today to ask one’s former colleague whether or
not he or she can help with house hunting? The initial position, the take-off,
is precarious, because mutuality is generally bound to the context of mean-
ing which defines it: Fellow students mutually help each other with their
studies, but do not naturally also act as housing agents for each other. Is it
possible to ask one’s former colleague to help in the provision of an intern-
ship for one’s daughter or will he give advice concerning investment using
his inside knowledge as a banker? The attempt to utilize discovered context-
spanning options is not covered socially, because the different contexts of
meaning (e.g. studies, housing, organizational membership, financial expert-
ise) are factually in no way related to each other. The chance of being per-
ceived as illegitimately connecting heterogeneous contexts of meaning and,
moreover, the anticipated denial of an unreasonable demand are just other
words for the risks inherent in a communicative attempt at networking.
Provided that the risk of such an illegitimate attempt is not only taken but
also meets with success, the communications may discover and mutually ask
for more and other possibilities that the involved addresses could provide
for each other. Due to the self-energizing effects of communication (Japp
1992) which result from successes in uncertainty absorption (March/Simon
1958; Luhmann 2000c), the unlikelihood of network emergence can be
transformed step by step into a likeliness of its maintenance.
292
16. Kap 13. Bommes 05-11-11 16.26 Sida 293
Stabilizing networks
Although, in the first instance, heterogeneity is a particular barrier to the ini-
tiation and establishment of networks, in the next step it becomes the par-
ticular basis and promoter of their stability. This is due to the necessity of
solving the problem of how members of networks can compensate for the
heterogeneous options made accessible to them by other network members.
Modalities of accounting for heterogeneous assistances, accesses, complai-
sances and provisions are needed. The solution lies in the dimension of time:
Social compensation for rendered services is inevitably postponed to a later
point in time as a “credit” for an undefined future service in return. The nec-
essarily unspecified and not beforehand clearable proportion of those het-
erogeneous services is always combined with the permanent creation of a
“remaining obligation” to a return service as well (Luhmann 1997a, p. 635).
That is to say, the poly-contextuality of the arrangement supports the estab-
lishment of a generalised reciprocity rule. Reciprocity is able to stabilise a
network and, in addition, to ensure its unquestionable repute as well as its
high flexibility in different circumstances (ibid.). Social networks may prac-
tice the rules of reciprocity silently, but they may also simplify these condi-
tions by relying on network-specific myths. They may underline “the
brotherhood” or “deep bonds of friendship” or refer to kinship or ethnicity
as “primary bonds”. Those secondary forms of stabilisation may then
broaden or restrict the flexibility of these networks by providing them with
the structures to account for their borders.
The principle of generalised reciprocity is significant for the stabilisation
of networks. It is not their constitutive principle (as in Powell 1990;
Mahnkopf 1994), but it allows the establishment of their operational conti-
nuity in time. Therefore, the question of stabilisation mechanisms in the
dimension of time must be distinguished from the question of network con-
stitution.
Organizations and networks

So far, it has been shown that networks presuppose the differentiated social
systems of modern society since they constitute the social addresses which
networks are built upon. These social systems provide addresses containing
individual profiles of inclusions and exclusions, thus making them attractive
and relevant to the reflexive combination in highly specialized networks.
In this section we turn to the relationship between organization systems
and networks, a structural distinction which the general network approach
can neither grasp nor give reasons for. As we have argued before, the net-
work approach systematically underexposes the difference between person-
al and organizational addresses, as well as the distinction between the social
293
16. Kap 13. Bommes 05-11-11 16.26 Sida 294
and factual aspects of inclusion. In fact, both individuals and organizations

can be activated as network-addresses. Against this backdrop, formal orga-
nizational networks have to be distinguished from personal networks in the
organizational context. This distinction will be taken as a framework here,
in which aspects of coupling organizations and networks are considered. We
will draw on various empirical examples, but the aim is neither to examine
empirical cases in detail nor to establish a theoretical typology of networks
in relation to organizations. Instead, we aim to highlight aspects of coupling
the two underlying structural forms (organizations and networks), thereby
exposing some structural conditions, inherent constraints, driving forces and
possible effects.
Personal networks in the organizational context

In view of the huge variety of networks that can be found in modern socie-
ty, only dealing with networks in relation to organizations seems to be an
extensive restriction. In fact, this restriction seems minor given that, firstly,
organizations occur in almost all function systems, secondly, organizations
typically procure access to various functional provisions, and, thirdly, indi-
viduals spend most of their lifetimes in and around organizations. Thus, the
case of a network to which organizational accesses have no relevance is
probably a curiosity.
With regard to networks among scientists it has, for example, been argued
that the aspect of organization is “almost totally irrelevant” (Mullins 1968,
pp. 795 f.). However, this result is misleading without detailed examination.
Respondents were asked to name the addresses of other scientists with
whom they “communicated information regarding research.” In fact, the
respective network addresses mostly did not belong to the same department
or organization but instead to “distant” organizations. This is not surpris-
ing, since the more sub-disciplines are specified, the less probable it will be
to find addresses for scientific communication in the same location
(Stichweh 1999a). While the argument underlines the relevance of (global)
networks in modern science, it simultaneously underestimates the relevance
of organizations to these networks of highly functional communication. In
his study, Mullin has taken for granted that respondents (as well as their net-
work partners) were employed at universities or other research organiza-
tions which provide their scientists with economic income, with functional-
ly necessary resources for scientific work, and, last but not least, with an
organizational address that signals to others their inclusion in a scientific
achievement role. In contrast, it is well known that scientists without a post
or organizational position tend to be excluded from scientific collaboration
as well, i.e. they are communicated as less relevant to scientific networks, if
not “almost totally irrelevant”. As a consequence of this, these scientists
294
16. Kap 13. Bommes 05-11-11 16.26 Sida 295
may however become relevant to professional support networks, which do

not primarily transmit scientific information but establish “access to
resources necessary for scientific work” (regarding the case of Russian immi-
grants in Israel, see Toren 1994, p. 726). As far as these resources (research
funding, positions and the like) are organizationally bound, professional
support networks presuppose the access of at least some addresses to the
respective organizations.
As much as organizations are relevant to scientists’ networks, scientific
organizations largely rely on these forms of network building. In order to
comparatively analyse how organizations formally deal with and capitalize
on personal networks which they find in their own context, or even initiate,
we now turn to further cases of personal networking before the particular
case of organizational networks combining organizational addresses will be
considered.
The silent character of personal networks in the organizational context

The following case exemplifies the general idea that personal networks in an
organizational context exist in the “darkness of informality,” as Teubner
(1996) put it. The case is taken from a study of the recruitment of appren-
tices at a major company in the German car industry (Bommes 1996). Con-
cerning the nearly 200 apprentices who were finally recruited in one age-
group, it was the case for about 60% of that group that one or more of their
parents had already been employed by the same organization. This propor-
tion was only 30% concerning the approximately 450 applicants for
apprenticeships. This difference in proportion was surprising even to mem-
bers of the company’s personnel department who, in principle, knew about
kinship-related tradition-building at the firm. As the study foreshadows, this
surprising effect resulted from personal networking: Applicants’ parents (as
organizational members) asked their shop stewards (who had access to the
respective organizational decision-making arena) to put in a good word for
their sons and daughters, possibly by reminding the shop stewards of past
collegial support in union issues or by raising future expectations.
The study does not refer to the idea of “informality” but characterises
these circumstances of organizational inclusion appropriately as “silent”.
Such networks are not officially announced within organizations, even if all
persons involved know about them. The formal acknowledgement of per-
sonal networks’ influence on organizational recruitments is improbable pri-
marily in the semantic framework of so-called “human resource manage-
ment,” which demonstrates its organizational relevance and even societal
modernity through factual notions of individuals’ achievements (as opposed
to social ascriptions; Parsons). However, the difference between “knowing”
and “saying” does not reveal the whole story, pointing to a second, struc-
tural condition producing this “silence”. Since organizations are social
295
16. Kap 13. Bommes 05-11-11 16.26 Sida 296
systems reproducing themselves purely through decisions (Luhmann 2000c),

they can only attribute final results concerning the recruitment of members
or apprentices to their own decisions, even if a personal network has pro-
duced this organizational result anonymously (Luhmann 2000c, p. 243). An
organization’s self-reference thus implies that nobody ever becomes an orga-
nizational member without the organization having decided on this.
The aspect of silent coupling refers to what Luhmann called “non-
decisional decision premises” (Luhmann 2000e, p. 145): The network
belongs to the organizational context in that it comes into existence by rea-
son of organizational decisions. Nevertheless, the network itself is not pro-
duced by organizational decision, even if it acts selectively as a premise for
decisions, allocates or branches off organizational resources, and eventually
participates even in the systems’ structural drift.
Following Luhmann’s observations of regional circumstances in Southern
Italy, the personal networks’ capability of exploiting systems seems to turn
out to be a principal weakness of the latter (Luhmann 1995c). However, the
systems’ principal openness to the establishment of secondary social struc-
tures is not synonymous with decreasing standards of achievement.
Regarding conditions of accomplished and prevailing functional differentia-
tion, organization systems also profit from personal networks. The reason
for this can be found in different modes of including individuals.
The logic of addresses, which networks are built upon, signifies that they
follow a social, particularistic mode of inclusion. Systems instead, due to
their factual problem-related orientation, have a universalistic mode of
inclusion in common. That is to say, they are based on inclusion criteria
defined by the respective problems and programs that the system deals with.
Although deciding highly selectively as regards membership, organizations
do not consider questions of participation in view of social aspects of the
participants’ addresses but with respect to their capability to solve factual
problems. Since organizational decision criteria are factual from the outset,
they typically leave margins of decision contingency open. Even though con-
straining and affording personnel decisions, organizational decision criteria
do not positively select exactly one individual for one organizational posi-
tion. Accordingly, personal networks, due to their address-related social ori-
entation, can close residual decision contingencies even without essentially
violating organizational decision criteria. Personal networks are parasites in
this case: They live on the organizations which feed them, while caring for
their host by releasing it from residual decisional contingencies.
The personal network as an organizational solution

In view of their own problems and programs, organizations can appreciate
or even reckon with existing personal networks in their environments. The
silent character of organizational inclusion via personal networks which is
296
16. Kap 13. Bommes 05-11-11 16.26 Sida 297
denoted above raises the question of whether organizations make personal

networks an issue of their own decisions turning the social orientation that
defines a network into a factual solution for organizational problems.
Examples are easy to find. It is well known that members of sales forces
are recruited not only according to personal sales and marketing skills, as
testified by career certificates, but also in view of the length of the customer
lists they bring with them from former employments at other organizations.
In a similar way, editors are often recruited on the basis of contacts with
authors they are likely to bring with them (Levitt/Nass 1989, p. 195), and
universities ask professorship applicants about their international scientific
contacts. Thus, organizations capitalize on these networks which seemingly
exploit the organization in the way discussed above. In all these cases orga-
nizations take advantage of existing individual address books and couple
them with personal networks that have emerged around new members’
addresses in other organizational contexts.
The cases dealt with so far have in common that the coupling of personal
networks and organizations rests upon at least one individual’s double inclu-
sion – as an organizational member and as a participant in a network. This
double-inclusion turns out to be a precondition for coupling personal net-
works and organizations. Apparently different examples can demonstrate
this:
For example, alumni networks, whether founded by graduates or initiat-
ed by university administrations, exclusively connect addresses of erstwhile
participants. In this case, the relation of network and organization is not
realised by double inclusion (even though this is possible too). Rather, the
connection is formally assigned to an organizational position which the
organization expects to serve the alumni-network as a whole not least in
view of accessible organizational advantages (e.g. career services for current
students, advertisements for the university, donations from alumni).
Accordingly, the network’s point of contact at the university is addressed by
network members not primarily as an individual person but based on his or
her organizational competencies. Therefore, he or she can be replaced by
another person without causing too much trouble – either for the network
or for the organization.
A closer look at the alumni case, however, reveals another important
aspect: Essentially, alumni-networks are not networks in a strict sense.
Insofar as everyone who was a student of the respective university can par-
ticipate in those networks, participation is conditioned by a universal crite-
rion (even though participation is open exclusively to former members only).
That is to say, alumni-networks are actually associations which, in view of
the diffuse opportunities concerning the participants’ careers, nevertheless
serve purposes of networking. Compared to the particularism of networks,
the associational principle is based on a universal criterion of participation.
297
16. Kap 13. Bommes 05-11-11 16.26 Sida 298
This precisely legitimises the organization to encourage personal networking

as well. Conversely, the absence of such a general criterion seems to be the
reason why universities typically do not formally acknowledge so-called stu-
dents’ societies that select participants according to highly particularistic cri-
teria, and sometimes simply by ascription.
Temporary substitution of personal networks by organization

Another form of coupling organizations and networks of former organiza-
tional participants, to be briefly discussed here, refers to the organizational
trend towards vertical disaggregation. The example serves to pave the way
for considering the case of formal organizational networks afterwards.
Among other things, vertical disaggregation means that organizations out-
source personnel resources which they – depending on demand – pull in
afterwards on a project basis again, either in view of costs or as a conse-
quence of a political decision. Under these circumstances some organiza-
tions, as has been shown for the U.K. television industry for instance, do not
hire project co-workers individually. Instead, they buy in performing artists,
technicians, writers, directors, and other freelancers as complete “configu-
rations of key actors” who “know and trust each other” (Starkey et al.
2000, pp. 299 f.). Faced with the high level of uncertainty concerning the
success of new TV programmes, the organizations resort to existing net-
works of professionals which seem to guarantee high quality output, if not
because of success, due to their well-rehearsed and proven knowledge bases.
Correspondingly, these networks are dependent on organizations which
recurrently highlight the combined knowledge potential of the latter and,
moreover, procure income for network participants. Since organizations
seem to temporarily turn these personal networks into relationships between
organizational members, the authors’ characterisation as “latent organiza-
tion” (ibid.) may be reasonably appropriate. However, this description tends
to ignore that – as a consequence of the temporary appointments – the
“latent organization” usually becomes activated by different organizations.
That is to say, the personal network structure tends, in fact, to become inde-
pendent of the individual organization that induced its emergence in the first
place by turning to vertical disaggregation and rehiring.
The different constellations of the relationship between organizations and
personal networks discussed so far all turn out to revolve around aspects of
organizational membership, whether referring to double inclusion (social
dimension), to organizational task and position (factual dimension), or to
temporary substitution (dimension of time). In this way, double inclusion is
a precondition, in fact, of coupling organizations and personal networks, if
coupling means that networks condition organizations – in as much as
organizations condition networks. The relationship of coupling means that
both contexts – organizations and networks – make use of the structural
298
16. Kap 13. Bommes 05-11-11 16.26 Sida 299
options provided by the respective other under its own operational premis-
es. Indeed, both contexts tend to rely on these external options for their own
operational reproduction, thus mutually feeding each other.
Not only has the form of silent coupling demonstrated, which is quite
common in various organizational contexts, that organizations, whenever
acknowledging and using personal networks, do so according to their own
specific premises and problems, but also the other examples. From the out-
set, these premises and problems are not defined in the social but in the fac-
tual dimension. When organizations couple with personal networks and
make use of the structural options they have provided (i.e. knowledge, cus-
tomer relations, reduction of decisional contingencies, outsourcing of risks
etc.), they do this under the premises of their own defined problems. The
networks as such remain outside of the organization. That is to say, they are
treated as a relevant structure that belongs to the organizational environ-
ment – or they are substituted through organization.
Organizations do not formally acknowledge being conditioned by the par-
ticularism of personal networks – even if they allow for conditionings
through selected environments. They transform the particularistic elements
penetrating their decision process – concerning, for instance, personnel
recruitment, sales practices or contracting out – into a universalistic form in
an ad hoc manner. They claim criteria of competence and competitiveness,
efficiency and effectiveness, austerity and success as being the final guiding
principles of their decisions. In this way, they ironically feed and at the same
time protect the particularism of networks that they make themselves
dependent on.
Formal organizational networks

So far, we have discussed the relationship between personal networks and
organizations. Ultimately, we also want to discuss the relationship between
organizations and organizational networks. Organizations start to combine
organizational addresses and the related reciprocal options. In principle, the
mechanism of network constitution is the same as in the case of personal
networks: Here too, the reflexive combination of addresses and respective
options creates new options.
Since organizations consist of decisions and since decisions characterise
their mode of reproduction, (Luhmann 2000c), they are able to act and com-
municate in their own name and responsibility. On this basis, they are also
potential candidates to be addressed in communication and to build net-
works. Organizational networks should, therefore, be expected to be as flex-
ible or stable, bi- or multilateral, as any other network. Empirical research
seems to underpin this. It proves that the “complementary strength” (Powell
1990) which makes organizational networks attractive to organizations is
based on the reflexive observation of addresses and the “intelligent
299
16. Kap 13. Bommes 05-11-11 16.26 Sida 300
combination” (Teubner 1992) of respective options. Again, this indicates the

formal similarity with personal networks.
Organizational networks, however, can make specific use of the reflexive
combination of addresses that have been shown to be the formal constitu-
tive mechanism of network building above. This is due to the fact that
organizations have the capacity to also decide on their own boundaries. For
example, the strategic “outsourcing” of entrepreneurial functions demon-
strates that organizations cannot only check their address book for attrac-
tive addresses; moreover, they can (co-)create new addresses for networking
by using their own (certainly more or less restricted) resources. Based on
decisions, new organizations are founded and equipped with a particular
profile (e.g. small size, high innovation potential, regional accesses) that can
be linked to the profile of the focal organization (e.g. large size, high finan-
cial strength, global access) within a specific organizational network. The
driving force seems to be the expectation that new options and accesses may
emerge through this particular way of combining addresses.
The structural shift in demand in the context of globalisation that has
taken place during recent decades has resulted in, among other things,
tremendous growth and variety in organizational networks, mainly in poli-
tics and the economy. In respect of this, it may be appropriate to speak of a
“network revolution” (Teubner 2000). Beyond structural conditions, this
organizational “network revolution” at the same time seems to go back to
the fact that the reflexive combination of organizational addresses has
stepped out of the “darkness of informality” (Teubner 1996). Organiza-
tional networks have become the officially announced formula not only
regarding organizational self descriptions but also the (self-)reflection of
societal function systems (for the “network state,” see Teubner (1999); for
the “network paradigm of industrial development,” see Hessinger et al.
2000). Thus, regarding the relevance of social addresses, the institutionali-
sation of “cultures of observation” (Fuchs 1992) takes place on a structur-
al as well as on a semantic level.
In both respects, it affects the inter-organizational as well as the intra-
organizational relationships of observation. This has been discussed in
organization research under the headings of the “decentralisation of organi-
zational structures” and the “levelling off of hierarchies”. Resources and
options regarding organizational units, departments and positions are no
longer coordinated according to a logic of serial interdependence but by a
logic of reciprocity. Subsystems become equipped with their own and par-
ticular addresses (that can be approached), and are expected to combine and
expand their options by way of reflexive modes of observation.
At the same time, this seems to make intra- and inter-organizational
networks constitutively ambivalent, and indicates their structural point of
weakness: Communicative events in organizational networks are, in
300
16. Kap 13. Bommes 05-11-11 16.26 Sida 301
principle, organizational events (they refer to member roles). Nevertheless,

the organization remains latent, because organizations cannot process reci-
procity and the particularism linked with it. The latency can always, how-
ever, be made organizationally manifest. This particularly entails the danger
that the dynamics of organizational decisions tend to destroy the subtle
structures of reciprocity. The same communicative event always has a
double meaning – it is part of the reproduction of one or more of the par-
ticipating organizations and of the network existing in between. If this is
true, the capacity of organizations to create networks through decisions
taken at the same time implies a high potential for instability in the respec-
tive boundaries. The reason for this is that organizations can directly inter-
vene in these networks. For the organizational staff this includes a high level
of insecurity: They are expected to behave in a way that allows communi-
cation as members of the organization, even though these communications
are formally not covered by decisions. At the same time, they are expected
to produce social options based on structures of reciprocity. If these efforts
finally come to be seen as failures, they will be treated as a decisional out-
come that was not formally covered by the organization. Responsibility falls
back on the individual member. In other words: Organizations communicate
trust in order to motivate their members to build intra- and inter-organiza-
tional networks, and they claim the possibility to renounce this trust – while
leaving their members alone finally in the case of failure. In this case, they
have no possibility of proving that their present cause of action rests on deci-
sions taken in the past.
Taken together, organizations have the particular advantage of being able
to decide on their own boundaries and even to create attractive organiza-
tional address-profiles for combination in organizational networks. How-
ever, this advantage simultaneously implies the high risk of the networks col-
lapsing, since it remains too close to the decisional mechanism of organiza-
tional reproduction. That is to say, organizations are in permanent danger
of destroying the subtle mechanism of reproduction upon which networks
rely (i.e. reciprocity and respective informality). Thus, organizational net-
works seem to be based on highly fragile processes of coupling that create a
need for sensitive forms which allow the maintenance and smooth crossing
of boundaries.
Conclusion
This chapter discusses the relationship between social systems and networks.
The actual sociological prominence of social network approaches poses a
challenge to sociological systems theory which describes modern society as
a functionally differentiated society. Social networks have empirically gained
301
16. Kap 13. Bommes 05-11-11 16.26 Sida 302
more and more relevance in the context of the recent globalisation process-
es. This has been the backdrop to actual sociological tendencies of redefin-
ing all kinds of social structures as social network structures, especially in
the context of organization sociology. The chapter initially discusses the
strengths and weaknesses of both the social network and the system theo-
retical approach and identifies complementary deficits. This is taken as a
potentially productive starting point for a more constructive exercise; i.e.
clarification of the relationship between the social systems of modern socie-
ty which systems theory primarily deals with, and social networks, which
are the primary subject of network approaches.
This clarification effort, however, is undertaken on the basis of a system
theoretical perspective. The reasons for this are at least twofold: There is
principally no possibility of taking a neutral third position of observation
that would be able to avoid starting out from its own theoretical premises.
A theoretical decision right from the beginning cannot be circumvented. The
decision taken here – to start out from systems theory – is based on two
assumptions: a) The systems-theoretical combination of a general social
theory and a theory of society allows the making explicit of some of the
implicit premises of network theory. It allows understanding of the peculiar
character of social networks in modern society; b) Systems theory provides
the theoretical means of specifying the structural conditions for the emer-
gence and reproduction of networks based on the general conceptual
approach. In doing so, it allows the avoidance of both: on the one hand, the
danger of merely adding an external “emergency wheel” to the general
approach and, on the other hand, over-generalising the network concept by
using it as a catch-all term. The chapter elaborates some aspects of this in
two steps. It first introduces a concept of social networks that is general and
formal enough to account for the rich empirical variety of networks in the
various social realms of modern society (like franchising networks, networks
of illegal immigrants, networks of scientists, women or neighbours) together
with their modes of emergence and reproductive stability, which is the com-
bination of addresses and the reliance on reciprocity. In the second step the
productivity of this concept is exemplified by discussing the relationship
between organizations and social networks. Only the results, some of which
are finally summarized here, can justify the unavoidable bias towards start-
ing out from one or another specific theoretical frame.
In view of the so-called “network revolution,” it has been argued again
and again that the distinction between organizations and networks is
increasingly becoming blurred. In fact, communication forms shift con-
stantly, changing between organizational and network communication, even
within the organizational context. Nevertheless, there is no reason to aban-
don the difference in theoretical terms, since attributions of utterances can
generally be expected to be unambiguous in order to fulfil their simplifying
302
16. Kap 13. Bommes 05-11-11 16.26 Sida 303
function within communication. However, communications can, in fact,

fluctuate back and forth between organizational and personal addresses.
They become multiple events that are operationally relevant to both, orga-
nizations and networks, even if this happens in different ways and by pro-
ducing different structures and linkages. Against this backdrop, attributions
may have to be analysed in more detail in further empirical research. How
do organizations that formally deal with networks avoid notions of “parti-
cularism” (which tends to be observed as corruption), and how do they con-
strain “universalism” (which does not allow the use of the advantages that
networks provide)? How do organizations balance out the subtle relation-
ship between decision and reciprocity in intra- and inter-organizational net-
works? Any analysis of the relationship between networks and organiza-
tions should be able to reconstruct even those structural tensions – instead
of simply replacing them with clear cut assumptions about “actors” and
thereby making important differences and empirical questions disappear as
if by magic.
The systems-theoretical approach opens up a fresh and arm’s length view
of networks which – due to aspects of trust and reciprocity – has too often
been described as a “worthwhile” or “better” social form. Networks are not
“better” but different. On the one hand, they are pre-modern: Whereas the
mode of inclusion of function systems or organizations is universalistic, net-
works are particularistic from the outset, as their modes of operation and
inclusion can demonstrate. On the other hand, they are innovative, and can
thus be described as modern: They create new and unpredictable possibili-
ties by combining options produced on the basis of systems differentiation.
Nevertheless, concerning any normative evaluation, one should bear in mind
that this “innovativeness” also holds true for modern networks of co-
optation, corruption and terror.
Systems theory allows an account of the plurality and specificity of net-
works. It makes visible the fact that they rely on the plurality of functional
and organizational systems, providing them with highly individual profiles
of addresses due to differing histories of inclusion and exclusion. This is pre-
cisely one of the preconditions for the variety and specificity of social net-
works. The possibility of combining the plurality of systems and their het-
erogeneous provisions, based on the option to link social addresses (indi-
viduals and organizations) freely, provides the basis for the emergence of an
irreducible variety of networks. Furthermore, and due to the permanent cre-
ation of “remaining obligations,” reciprocity seems to be the social form
that provides the conditions for the operational continuity of networks in
time.
The analysis of the relationships between organizations and social net-
works makes visible how they mutually feed each other and rely on the
structural features of the respective other. This has been discussed in relation
303
16. Kap 13. Bommes 05-11-11 16.26 Sida 304
to the mechanisms of “double inclusion” and identified as a precondition

for coupling between both. Moreover, we have discussed the similarities and
differences of relating organizations to personal and organizational net-
works. In particular, the analysis of couplings between different structures
leads to insights which the network approach ignores by reason of its exclu-
sive reliance on the concept of social networks. One insight we assume to
follow from our analysis is that the emergence of global structures (e.g. glob-
al structures of migration) cannot be adequately explained by referring
either to organizations or to networks alone. Instead, the mutual interde-
pendence of organizations and networks has to be taken into account.
Whereas the systems-theoretical approach leads to new insights hitherto
missing from network approaches, it admittedly fails to provide the elegant
tools of formal analysis which are well-known from network approaches.
However, it remains an open question whether or not this difference
between the approaches has to be understood as a failure of systems theory
or whether this is due to a reductionist network concept that excludes some
key aspects from the analytic approach, even though only the latter enables
consideration of the salience of networks in modern society. It seems that the
well known critical objections that have been raised regarding the peculiar
formalism of the network approach gain their substance particularly from
those shortcomings which we tried to elaborate on here within the systems-
theoretical framework, thereby denoting, among other things, the distinc-
tion between address connectivities and network structures.
304
17 Kap 14. Simon 05-11-11 16.27 Sida 305
PART V
Forms of Organization
17 Kap 14. Simon 05-11-11 16.27 Sida 306
17 Kap 14. Simon 05-11-11 16.27 Sida 307
Chapter 14
Analysing Forms of Organiza-

tion and Management:
Stock Companies vs. Family
Businesses
Fritz B. Simon
Introduction
Drawing on Von Foerster (1981) and Spencer Brown (1979) Luhmann con-
ceptualises organizations as observing systems. Every organization observes
itself and its environment and on the basis of these observations reproduces
itself. We, as organization theorists, in turn, can observe and analyse these
organizational observations. Depending on the particular observations we
might distinguish between different forms of organization and different
management styles; and we can examine to what extent the two fit together.
The successful management of a family business, for example, requires a dif-
ferent mode of observation from that of a stock company. A useful tool for
such an analysis can be found in Spencer Brown’s calculus of forms, which
allows us to represent these complex constellations of observations in simple
and comprehensible format. In the following I want to demonstrate the
fruitfulness of this approach.
The chapter is structured into nine sections. In the first section I will intro-
duce Spencer Brown’s concept of observation, which lies at the heart of
Luhmann’s conceptualisation of organizations as observing systems. This
will also provide us with the formal apparatus for analysing organizational
observations. In the second section I briefly explain the concept of auto-
poiesis. In the third section I will discuss the ways in which organizations
may be observed and analysed. The fourth section is concerned with the dis-
tinction between different types of coupling between organizations and their
relevant environments. In the fifth and sixth sections I examine how organi-
zations observe themselves and their relation to the environment. In the
seventh section I point out some unavoidable conflicts in the organization’s
handling of its relation to its environments. In the eighth section two types
307
17 Kap 14. Simon 05-11-11 16.27 Sida 308
of organization will be compared: stock company and family business. In

the ninth section I try to analyse the consequences of introducing stock
options into businesses as an incentive for the management. The chapter
concludes with a brief reflection on the fruitfulness of the suggested
approach to organizations.
Observation
In the context of Niklas Luhmann’s systems theory (2000c) the term obser-
vation is used according to the definitions George Spencer Brown provides
in his book Laws of Form (1979). He defines the operation of observation
as the coupling of two different distinctions in two different spaces. These
two spaces are called “space of the first distinction” and “space of the
second distinction”. The first distinction is called “distinction”, the second
distinction is called “indication”.1
Outside their mathematical context those terms can be rendered in more
simple, everyday language as follows: in order to observe, every observer has
to make two distinctions and couple them: (1) distinctions between different
phenomena in a first phenomenal domain, and (2) distinctions between
other phenomena in a second phenomenal domain, which function as indi-
cators for the first distinctions. The process of observation thus results from
coupling the two operations: distinction and indication. The sense or mean-
ing of the indication is the first distinction.
The act of drawing a distinction separates a space, state, or content on the
inside of a boundary from a space, state, or content on the outside of the
boundary (Spencer Brown 1969, p. 4). The two sides can be indicated by
distinctions in the domain of language, i.e. by giving the two sides of the dis-
tinction different names (for example “inside”/“outside”). For instance,
observers of biological systems can distinguish a living organism as an enti-
ty that is distinct from other living systems and/or the non-living environ-
ment by using their senses (perception/first distinction). Furthermore, they
can give it a name – for example “my cat” – to indicate what they are aware
of (indication/second distinction). By calling the inside of the distinction
“my cat” the rest of the universe can be called “not my cat” (Figure 1).
Usually we don’t realize that by making distinctions we always create a form
with two sides. Because we don’t give the outside of the distinctions (envi-
ronment) a special name we are not usually conscious of it: it is not indicat-
ed and therefore it is not observed. Where there is no indication there is no
observation.
1 Note that this is a slightly different interpretation of Spencer Brown’s concept of observation
from that provided in Chapter 1 in this volume.
308
17 Kap 14. Simon 05-11-11 16.27 Sida 309
Inside,
Outside, Not My-Cat My Cat
Figure 1.
To illustrate distinctions Spencer Brown draws a circle as a “mark of dis-

tinction” on a plane surface (as shown in Figure 1). In this way, he illustrates
that distinctions are always characterised by the creation of a boundary. One
cannot reach the inside from the outside without crossing the boundary and
vice versa. Applying this principle to distinctions in three-dimensional space
will cause these distinctions to create boundaries between inside and outside,
i.e. so called “entities” or “objects” and their environment.
Spencer Brown offers a second way to symbolize distinctions, which can
be used for mathematical purposes and for analysing complex structures and
relationships of more complex structures of distinctions. It consists of so-
called “tokens for the mark of distinction”, which he refers to as “crosses”.
He calls these tokens “crosses” because of the double meaning of the term.
“Cross” is not only a name for a special symbol but at the same time an
injunction to cross the boundary from the outside to the inside of the dis-
tinction, that is to say, it induces a dynamic that brings forth the distinction
(Figure 2).
Inside,
My Cat
Figure 2.
When we apply these different types of symbol-systems we have two options

for representing or describing the distinctions made by an observer: either
by using circles or by using crosses. My suggestion for the following text is
to use circles to symbolize first distinctions. This presupposes that these first
distinctions are generated by causal processes that are different from the
processes of the observation from an outside perspective (e.g. the biochemi-
cal processes that generate the body of a cat or the communicational
processes that generate an organization). In contrast, crosses will be used to
symbolize second distinctions (indications) in order to show how the former
processes (or first distinctions) are described or represented or categorized in
309
17 Kap 14. Simon 05-11-11 16.27 Sida 310
communicational processes. Whenever circles are used to represent commu-

nicational processes (as happens in organizations or other social systems) the
crosses describing these processes can be considered as meta-communica-
tion. In organizations the meta-communication can become an integral part
of the communication that keeps the organization going.
Since all meta-communication is communication one has to be careful not
to mistake the token of a mark of distinction for the mark of distinction as
such. In other words: we have to avoid mistaking a symbol for the object it
symbolizes; mistaking the communication about a phenomenon for the phe-
nomenon. We have to avoid making the epistemological error of mistaking
the characteristics of observations for characteristics of the observed phe-
nomenon; mistaking, that is, the second distinction (indication) for the first
distinction – e.g. eating the menu.
Let’s have another look at an organism as a distinction – our cat – the
better to illustrate the difference between first and second distinctions, i.e.
between marks of distinctions and tokens of the mark of distinction. In
Figure 1 the circle represents the mark of distinction. In the context of this
paper, it represents the first distinction, i.e. the body of the cat. But in real
life things are different: what stands for a mark of distinction/first distinc-
tion is the body of the cat. We resort to using circles in this paper because
of the obvious difficulty of introducing real cats into a paper. The cross, in
contrast, represents the second distinction, i.e. it indicates the body of the
cat. It represents the name of an organism called “cat” which is being
observed (either by an external observer or by itself). But let us be clear that
the indication of “a cat” is “no cat”.
This distinction is relatively easy when we refer to biological processes
(cats) and the communication about biological processes (speaking about
cats). It becomes much more problematic when we refer to social systems as
communicational systems and the communication about communicational
systems. The danger of mixing up the logical levels and creating epistemo-
logical errors increases immensely.
Autopoietic systems
As outside observers we can distinguish organizations from their environ-
ments. From a constructivist perspective we have to decide how to concep-
tualise the causality (generating mechanism) of the creation of the distinc-
tion called “organization” (cf. Varela 1984). In the context of Luhmann’s
systems theory the answer is clear: the processes that generate organizations
as distinctions are communicational processes that delimit an organization
as a social system from its environment (i.e. communication as a phenome-
non in the space of first distinction). The logic of these dynamic processes
310
17 Kap 14. Simon 05-11-11 16.27 Sida 311
by which an organization is separated from its environment can be charac-

terised as autopoietic.
The notion of “autopoietic systems” was developed by Humberto
Maturana (1975; Maturana/Varela 1980) to explain the specific character-
istic of living systems. As a biologist, he focused his attention on biological
processes and the organization of their internal dynamics. As “autopoietic”
he characterizes a structure of processes in which a network of interactions
between the elements of a system creates these very elements through their
own activities. In this way the boundary between inside and outside of an
organism is generated and maintained. This is a self-organized and auto-
nomous process without any creator outside of the system. To apply the
notion of Spencer Brown: the first distinction is generated by physiologi-
cal/biochemical processes which create a boundary between the inside and
outside of a body. Maturana usually symbolizes autopoietic systems as cir-
cles with an arrow to illustrate that the distinction is actively created and
autonomously maintained; it ceases to exist if these generating and defining
processes are terminated.
This principle of autopoietic reproduction can be regarded as characteris-
tic for organizations as well. They too are created and kept “alive” as long
as certain processes continue or are repeated. In this case, however, these
autopoietic processes are not biochemical but communicational. Figure 3
shows two ways to illustrate this dynamic: (1) as a circle with an arrow, and
(2) as a repetition of crosses:
▲
Organization
or
X X = Organization
X = communicational processes that generate the distinction organization/environments
Figure 3.
There is always a need for repetitive patterns of communication that keep

the organization as an entity distinct from other social systems and environ-
ments. Only if those processes are realized will the organization emerge and
only as long as they are realized will the organization survive. As soon as
they stop, the existence of the organization comes to an end.
311
17 Kap 14. Simon 05-11-11 16.27 Sida 312
The organization as observed system

We now can use circles or crosses to represent organizations and their envi-
ronments as observed phenomena in order to analyse conditions and rela-
tionships that are relevant to their survival. From the viewpoint of systems
theory the unity that an organization and the characteristic and for its sur-
vival essential environments represent has to be conceptualised as the co-
evolutionary unit of survival (Figure 4).
▲
▲
Environment Environment
a c
▲
Organization
▲
▲
Environment Environment
b …n
Figure 4.
If we look exclusively at autopoietic systems as environments of organiza-

tions, then the circles in Figure 4 represent communicational, psychic or bio-
logical systems as environments. For the purposes of organizational theory,
biological systems can be regarded as given, so long as psychic systems or
other social systems are given. We can now focus our attention on the latter:
if the respective circle represents a psychic (autopoietic) system, then the
processes represented are psychic processes. If the respective circle represents
a social system, then the processes represented are communicational
(autopoietic) processes.
To describe the organization and its relevant environments (i.e. other sys-
tems) that we sketched above we can use crosses (Figure 5).
Unit of = Organization Environment Environment Environment Environment

Survival a b c d
Figure 5.
If we examine an organization and its relevant environments from an out-

side perspective – for example as sociologists – we may consider other com-
municational processes to be of relevance to the survival of the organization
than the organization itself does.
312
17 Kap 14. Simon 05-11-11 16.27 Sida 313
Strict and loose coupling

Up to now we have provided notations to describe distinctions between
“system” and “environment”. If these two together build a unit of survival,
we can consider them to be coupled with each other. It is possible to distin-
guish different types of coupling. Karl Weick (1979) introduced the concepts
of strict and/or loose coupling into organizational theory. If the elements of
a system or the elements of a system and its environment are coupled in a
strict way, a rather inflexible structure or structure of relationships emerges.
If they are coupled in a loose way, the structure becomes more flexible. Both
terms have to be seen in relation to each other, i.e. “strict coupling” and
“loose coupling” are relative concepts: couplings are more or less strict,
more or less loose.
If we focus our attention on the coupling between an organization and its
environment, we see that the various environments can be considered to be
exchangeable to a greater or lesser extent. If we consider, for example, the
relationship between a firm and a single customer, the coupling between the
firm and the customer is relatively loose, because the individual customers
are exchangeable. If we look at the coupling between the same firm and the
market in which it sells its products, the coupling between the firm as a
social system and the market as a social system is relatively strict, because
there are not that many different markets in which a company can sell its
products. An organization will possess greater or lesser degrees of autonomy
depending on the type of coupling between it and its relevant environments.
The management or leadership of this organization has to take the degree of
coupling into account, if it intends to develop survival strategies for the
organization. The less strict a coupling is between an organization (x) and
an environment (y) (which can also be an autopoietic system, i.e. a psychic
or social system), the less can the unity they represent be considered a unit
of survival.
Figure 6 shows how an outside observer can witness those different types
of coupling as the emergence of larger or smaller unities:
x x
y y
Strict coupling Loose coupling

Figure 6.
313
17 Kap 14. Simon 05-11-11 16.27 Sida 314
Represented as crosses, both types of couplings can be symbolized as follows

(Figure 7):
x y x y
Strict coupling Loose coupling

Figure 7.
To illustrate the impact of loose vs. strict couplings let’s have a look at the
different relationships a company has with its employees and customers.
Since there are contracts that bind company and employee together over a
longer time span, the coupling between those can be considered relatively
strict (Figure 8).
Coupling company/employee: x y
x = company, y = employee
Figure 8.
What we have here is a cross that creates a bigger unit made up of compa-
ny and employee. These two build (at least temporarily) a bigger unit of sur-
vival. Nevertheless, both coupled systems have the possibility of ending the
contract and consequently the coupling. So, in this sense, the relationship
between a company and an employee would not be necessarily described as
strict coupling. The reason for this is that there are markets in which com-
panies compete for employees and potential employees have to make their
choice between different job offers. However, if we compare the coupling of
a company and an employee with the coupling of a company and a customer
who buys a product of this company just once, it becomes clear that the cou-
pling of the latter is relatively loose (Figure 9).
Coupling company/customer: x y
x = company, y = customer
Figure 9.
In the scheme of crosses we can see that the two crosses of company and
customer are not united by a further cross. Both are independent of each
other, that is to say, both are free to couple themselves for the next
314
17 Kap 14. Simon 05-11-11 16.27 Sida 315
transaction with other partners. Clearly, there is no co-evolution of the two

systems, so, creating customer loyalty is a sensible goal for many companies.
Observation of a firm
In an economic context, the survival of a firm can be defined as the ability
of that firm to cover its expenses (e.g. bills, wages, interest on loans etc.).
This is only possible if a firm manages to generate income. To that end, eco-
nomic units usually operate in a way that is commonly referred to as “doing
business”, which in most cases means producing goods or services that can
be sold on a market. A firm, however, is active in markets not only as a seller
but also as a buyer in that it buys products, ideas, labour and so on. To sur-
vive as an autonomously acting unit, in the long run it must be able to gen-
erate at least as much money as it needs in order to pay different creditors
or stakeholders.
As already mentioned, other social systems (e.g. markets) can be consid-
ered relevant environments for the survival of a business. However, not only
social systems are relevant: since every company needs employees to fulfil
typical functions, it must also take into account the multitude of psychic sys-
tems of its employees as a second type of environment that is relevant to its
survival. In order to function independently of individual workers or
employees, a company has to make sure its employees are exchangeable. A
company can find enough employees on the job market to secure its sur-
vival.
From an outside perspective the relationship between a company and its
relevant environments can be illustrated by the following arrangement of
circles (Figure 10):
▲
▲
▲
▲
▲
…
▲
▲
Employees as
▲
▲
Observers
Market of Market of
Products and Employees
Services
▲
▲
▲
Capital-
Market/ Firm Market of
Family of States
Owners
Figure 10.
315
17 Kap 14. Simon 05-11-11 16.27 Sida 316
Since the processes that keep a firm running are communicational process-
es, the crucial question is whether a company would apply the same dis-
tinctions in its self-description as an outside observer would apply in
analysing the environments relevant to the survival of the company. Usually
the majority of employees are not and need not be aware of the relevant
environments. This kind of observation of the company (in both senses) is
part of the responsibility of the management (Figure 11).
a : b c d e f g
a = observation of the management e = market of employees

b = firm f = market of products/services
c = employees as units of survival g = market of states/political systems/tax systems
d = capital market
Figure 11.
The management has to take into account all these systems and environ-
ments, if it is to come to decisions that enhance the chances of the firm’s sur-
vival. If the management observes according to the structure of crosses
shown in Figure 11 (which it does not need do) it sees the coupling of the
company (b) with the employees (c) as relatively strict, and the coupling
with the different markets (d, e, f, g) as relatively loose. If it applies a glob-
al perspective it might perceive the relationship between workers or employ-
ees and company also as loose.
In the process of progressive globalization the degree to which the cou-
pling with a certain state has to be considered strict changes almost from
month to month. There are certainly companies for which the exchange-
ability of sites which could host their production is very limited, while others
are free to move their plants around the world in a rather short time.
Nevertheless, every management has to assess the extent to which the envi-
ronments of the company are exchangeable. The management has to include
in its reality-construction the respective system/environment-distinctions as
well as the types of coupling between them.
Management and re-entry

From Luhmann’s systems perspective the management – in contrast to the
“manager” as a psychic system – has to be conceptualised as part of the
organizational communication. Its main function is to introduce the distinc-
tions between the company and the environments relevant to its survival
316
17 Kap 14. Simon 05-11-11 16.27 Sida 317
into the communication within the company (cf. Baecker 1993). In other
words: the management has to enforce the “re-entry” (Spencer Brown 1969,
p. 56) of the firm/environments-distinctions into one side of the distinction,
i.e. the firm (Figure 12).
h : b c d e f g a
a = reality-construction/-observation of e = market of employees

the management f = market of products/services
b = firm g = market of states/political systems/tax systems
c = employees as units of survival h = interventions of the management
d = capital market
Figure 12.
In order to be able to observe itself, the organization has to make distinc-

tions and indications in the domain of communication. The management
can/must introduce specific distinctions into the internal communication of
the firm (Simon 2002b). This way, the management enables the organization
to observe itself, i.e. to realise a re-entry of the distinction organization/envi-
ronment(s) into the organization. Therefore, the first, and probably in the
long run most important function of the management is to focus the obser-
vation of the company on specific system/environment differences and inter-
actions (h). The second function of the management is to influence the inter-
nal communicational and interactive structures of the company accordingly,
in order to enhance the firm’s chances of survival.
Figure 12 illustrates an example of intervention (h) of management, i.e.
introducing its reality-construction (a) into the communicational processes
of the firm. The management’s construction of reality (a) is structured
according to specific distinctions: the firm (b) and the employees (c) are con-
ceptualized as a co-evolutionary unit of survival, i.e. as strictly coupled. The
different types of markets that build the relevant environments of this unit
(d, e, f, g) are relatively loosely coupled with each other. The reality-
construction/observation (a) of the management couples the unity of the
company and the employees strictly with the different markets, and the
intervention of the management (h) re-introduces (“re-entry”) this observa-
tion into the communication that keeps the firm going.
Unavoidable conflicts
There exist some fundamental contradictions and conflicts between systems
and environments. More specifically, in the case of an autopoietic system,
317
17 Kap 14. Simon 05-11-11 16.27 Sida 318
the more exchangeable its relevant environments, the more independent

from, or indeed the less dependent on, those environments it is. For exam-
ple, the larger the number of competent employees available on the market,
the less dependent is the company on any particular employee. This is the
reason why unions were created in the first place: for a company the single
worker might be exchangeable as relevant environment, however, the entire
workforce of a country or territory is not exchangeable. In theory the unity
of all workers has more power to negotiate with the company than the single
worker does. This works both ways: the higher the number of companies
that show an interest in hiring a particular person, the more exchangeable is
every company for this person and the less the power individual companies
have as relevant environments for the survival of this employee. So, both
communicational participants can preserve their relative independence by
keeping their respective environments exchangeable (Simon 2004, p. 90).
The same structure of conflict of interests can be witnessed in the rela-
tionship and coupling between company and investor. Strong competition
among companies for the allocation of capital will increase the dependence
of companies on individual investors and vice versa: strong competition
among investors keen to invest in a particular company will increase that
company’s power to define the conditions of the investment.
In a nutshell, we can say that capitalism today is not dealing with the
dyadic contradiction between capital and work as Marx defined it (a view
in which the company as a unit of survival played no role at all). If we look
at the conditions of company survival from the perspective of systems theo-
ry, we see that there are always unavoidable conflicts between autopoietic
systems and their relevant environments. Basically, the management respon-
sible for the well-being of a firm has to deal with two different conflicts:
first, a conflict between the company and the workforce, both of which
compete for money, liquidity etc. The same principle – i.e. that power is
always on the side of the partner who is less exchangeable for the other –
shapes the second conflict, i.e. the relationship between the owners/share-
holders (capital) and the company. So the management always has to deal
with a triangular conflict: company vs. workforce and company vs. capital.
Stock companies vs. family businesses

The concept of exchangeability is of great explanative value in the analysis
of the different relationships between companies and their respective stake-
holders. We will demonstrate this with regard to the relation between
investors and stock companies on the one hand and family members and
family businesses on the other.
In order to make decisions, the management of a firm has to take into
account the various loose or strict types of coupling between the company
318
17 Kap 14. Simon 05-11-11 16.27 Sida 319
and its environments. A stock company, for example, has to deal with a
relevant environment of investors and financial markets; a family business
with a family of owners. The coupling between company and investors in
the stock market is less strict than the coupling of a family business and the
owner’s family or its members. These different types of coupling have to be
observed and taken into account in the process of managerial decision-
making. It follows that the management’s observation/construction of a
stock company will be very different from that of a family business.
If we look at stock companies, top-managers have limited contracts like
other employees. Because of that, the managers (as psychic systems) are like-
ly to possess distinct personal interests that are ultimately also reflected in
the managerial communications. In this sense the management can be said
to represent different concerns from those of the organization as a whole.
The management has to take this into account in its own observations.
Another aspect that the management has to observe is its own interventions
in the company’s communication and the particular impact these have. All
these aspects have to be represented in the management’s observation of the
firm and its environments. Again, this construction of reality can be sym-
bolized by an arrangement of crosses (in order to reduce the complexity of
the figure we will abstract from the market of states, as it does not mark any
principal differences between family firms and stock companies). Figure 13
represents the observation of the management (a) of a stock company, which
explicitly takes the interests (i) and the interventions (h) of the management
into account, but which, in this case, does not see the employees (c) and the
company (b) as strictly coupled:
a : b c i d e f h

d = capital market i = interests of the management
Figure 13.
The basic difference between a family firm and a stock company concerns
the coupling between the company and its shareholders (see Figure 14).
319
17 Kap 14. Simon 05-11-11 16.27 Sida 320
a : b j c i e f h
a = observation of the management f = market of products/services

b = firm h = interventions of the management
c = employees as units of survival i = management
e = market of employees j = family
Figure 14.
The figure shows that there exists a rather strict coupling between the owner
family (j) and the firm (b). It is much more difficult for family members to
disassociate themselves from the company than it is for an investor of a
stock company. Usually they are not allowed to sell their shares on the mar-
ket, because many owner families have rules, contracts, and by-laws that do
not allow the trading of shares outside the family. This serves to secure the
family as an acting unit with respect to the company. For the company, the
owners are not exchangeable in the same manner as investors are for a stock
company and vice versa. Families and their businesses can be characterised
as “partners” involved in co-evolution (Simon 2001; 2002a, p. 17).
Dealing with an owner family and/or its members means that the manage-
ment can count on a more reliable relationship than it would with a (capi-
tal) market. This allows for longer planning perspectives and affects, among
other things, the way in which the management deals with insider informa-
tion and presents the company to the public. In other words, a different style
of management can be applied, and decisions can be made according to
different values, which are based on the rules and goals of the family rather
than the rules, goals, and evaluations of investors and financial markets. Fig.
14 shows a rather strict coupling of employees (c) and individual managers
(i) to the unity consisting of family and firm (b, j). Strict coupling is often,
though not always, the case with family firms.
A second aspect we need to underline is that in a family firm the manage-
ment must be aware that its actions (h) influence not only the company but
also the family. In these cases there is always a double re-entry, i.e. a re-entry
into two social systems: the family (j) and the firm (b). The management is
observed by the company and by the family, as it too observes the company
and the family. Ideally they accomplish the coupling of both. This task is
quite different from the task of keeping investors in a good mood, i.e. man-
aging investor relations in a stock company. It takes different skills to gain
the trust of an old aunt in a family, for instance, from those it takes to please
a young analyst in the stock market.
320
17 Kap 14. Simon 05-11-11 16.27 Sida 321
Effects of stock options

If a company offers stock options to its top management, the consequences
for the company can be far-reaching (Figure 15). The reason for this is that
the reality-construction of the management changes tremendously (at any
rate it is very likely that it will). In this case the management has to enforce
a double re-entry both into the company (b) and into the stock market (d).
What’s more, the management has to consider how the stock market views
the company. It must intervene in the stock market in order to get a higher
value for its own stock options. Its decision making will be oriented to the
presumable interest of the investors, regardless of whether this is in the long-
term interest of the company or indeed good for the survival of the compa-
ny. In other words, the management becomes strictly coupled with the stock
market (i, d). The negative side effect is that the management may be
seduced into trying to trick the markets by faking balances for instance, or
succumb to the by now well-known, deviant kinds of behaviour witnessed
in the Worldcom and Enron cases.
a : b c i d e f h

d = capital market i = management
Figure 15.
The situation becomes more difficult, indeed paradoxical, if a family busi-

ness goes public and becomes a stock company (Figure 16). In such a situa-
tion the management has to deal with family members as well as investors,
i.e. with a family (j) and the capital market (d). Both can be considered re-
levant environments, and both have to be taken into account when crucial
managerial decisions are made. The situation becomes even more compli-
cated if the management is offered stock options. Since this is one of the
most difficult situations possible, let’s examine it briefly.
321
17 Kap 14. Simon 05-11-11 16.27 Sida 322
a : b j c i i d e f h
a = observation of the management f = market of products/services

b = firm h = interventions of the management
c = employees as units of survival i = management
d = capital market j = family
e = market of employees
Figure 16.
As Figure 16 demonstrates, a conflict may arise between two different kinds

of corporate cultures. The management (i) is strictly coupled with the unity
consisting of firm and family (b, j) as well as with the stock market (d). This
probably creates a conflict of values: if the management acts in the interest
of the investors (an orientation that the stock options particularly promote)
it runs the risk of contradicting the values of the family and vice versa. If the
family judges that the actions and decisions of the management (i) are
directed against the interest of the family or against the goal of the long-term
survival of the firm, it will fight against the management. If in the view of
the stock market (i.e. the analysts) the actions and decisions of the manage-
ment (i) are directed against the standards expected from a stock company
its shares will fall. The management is in a paradoxical position: inside and
outside the strict coupling with firm and family, and, at the same time, inside
and outside the strict coupling with the capital market. It has to manage a
threefold re-entry in that it has to introduce (h) this contradictory con-
struction of strictly coupled unities not only into the company (b) but also
into the family (j) and into the stock market (d). Making the paradoxical
logic of the chosen strategy and the patterns of managerial decisions appear
coherent and plausible to both investors and family members is unavoidably
a difficult task.
Concluding remarks
The aim of this paper was to demonstrate the fruitfulness of applying
Spencer Brown’s calculus of forms to some of the concepts of Niklas
Luhmann’s systems theory for analysing different forms of organization and
management. This formal apparatus has proven particularly helpful in the
analysis of very complex relationships as it allows representing them in a
simple and transparent way. It has already been used successfully in other
areas, such as the analysis of the implicit (dis-)organization of schizophrenic
322
17 Kap 14. Simon 05-11-11 16.27 Sida 323
thought (see Simon 1988, pp. 307–314). In this chapter it has helped us
examine the very complex relationships between different organizations or
organizational units and their relevant environments. The application of this
model to stock companies and family businesses demonstrates that it can
even be used to analyse planned structural decisions in order to anticipate
problems and conflicts that can arise from the creation of new, strictly cou-
pled units of survival.
323
18 Kap 15. Hilliard 05-11-11 16.28 Sida 324
Chapter 15
On Defining the Multi-

national Corporation.
A Systems-Theoretical
Perspective1
Darnell Hilliard
Since the 1960s there has been an economic debate on the so-called multi-
national corporation. Observing an increasing internationalisation of pro-
ductive activities and direct investments since the 1950s, economic theorists
thought to have identified a new form of corporation that seemed to show
new qualities that were clearly different from the national corporation and
the historical forms of international business (for an overview see Dunning
1971; 1974). Therefore, economic theorists started to analyse the multina-
tional corporation by trying to find a general definition of their object. But
the classical as well as the current theoretical debate on defining the present
multinational corporation is characterised by a deep disagreement in form
and content (for a critical discussion on classical definitions see Aharoni
1971). On the one hand, there exists confusion about the type of operation
to which the term “multinational” should refer, as well as about the degree
of foreign operations that have to be performed by a corporation in order to
speak of a “multinational” corporation. On the other hand, all definitions
characterise present multinational corporations as corporations that operate
in two or more national environments; that is to say, they define the multi-
national corporation with regard to the structuring of society into different
nation-states. However, in view of increasing economic, political, legal, and
cultural globalisation, is it really appropriate to describe the present envi-
ronment of large corporations as a plurality of national environments?
This chapter argues that Niklas Luhmann’s sociological systems theory
contains several starting points for developing a more general definition of
the multinational corporation, which, first, does not restrict the multina-
tionality of a corporation to one specific type of operation (or: corporate
1I want to thank Kai Helge Becker and David Seidl for very fruitful comments on earlier drafts
of this chapter.
324
division), and, second, takes into account that the globalised environment of
present corporations can no longer be described as a plurality of national
environments. In other words: concerning the first aspect, we refer to
Luhmann’s systems theory because it conceptualises organizations (corpora-
tions) as social systems reproducing themselves by communicating decisions
(and not by productive, financial, or technological operations). This gener-
al understanding of an organization (corporation) enables one to look at all
divisions of a business organization. With regard to the second aspect,
Luhmann’s systems theory conceptualises the social environment of present
organizations not as a plurality of national environments, such as national
markets, national economies, nation-states or national societies, but as a
single world society that is differentiated into globalised function systems. In
sum, these features of systems theory open up a perspective to re-define the
present multinational corporation by describing the way in which this cor-
poration form relates its decisions to the worldwide communicative contexts
of the world society, thereby offering an alternative to the shortcomings of
the economic debate.
The chapter is organized in three sections. The first section outlines the
two fundamental problems of classical and current definitions mentioned
above. The second section introduces three central concepts of Luhmann’s
systems theory: organizations are conceptualised as social systems that
reproduce themselves as networks of decision communications. Corpora-
tions are conceptualised as a specific organizational form that reproduces
itself by relating decisions to business opportunities. Finally, the present
society is conceptualised as a single world society whose functional differ-
entiation into different societal subsystems is no longer limited through any
territorial boundaries. Proceeding from these systems-theoretical considera-
tions, the third section outlines a sociological concept of the present “world
corporation”. Here, the present world corporation is defined as a specific
type of corporation that relates its decisions to a single business environment
of business opportunities linked worldwide. In contrast to the existing con-
ceptualisations of the multinational corporation the main attribute of pres-
ent multinational corporations is no longer seen in the ability of operating
in two or more national environments, but in the ability of maximising prof-
it on a worldwide basis. This will be demonstrated with regard to three dif-
ferent corporate divisions: finance, production, and marketing. The final
section summarises the main insights of the previous sections and points out
some additional advantages of a systems-theoretical definition that have to
be explored by future research.
325
Classical and current approaches to

the multinational corporation
In this first section we will outline six prominent economic and organiza-
tion-theoretical approaches to the multinational corporation. In doing so,
we concentrate on pointing out the theoretical shortcomings that these def-
initions have in common. In particular, there are two problematic aspects of
classical and current approaches to defining the multinational corporation:
firstly, all approaches define the “multinationality” of a corporation by
relating it – a priori – to one specific type of operation. And secondly, even
in times of a globalised world society, each approach sees the main attribute
of the multinational corporation in its ability to cross two or more national
environments in one or another way.
Classical approaches
Stephen H. Hymer’s thoughts on foreign direct investments are generally
considered a starting point of the economic theory of the multinational cor-
poration (Hymer 1977 [1960]). In Hymer’s model, the “multinationality” of
a corporation refers to its basic operations, which are called direct foreign
investments. The operation of direct foreign investment is distinguished
from the traditional portfolio investment by the aspect of “control”. On the
one hand, “control” may refer to the extent to which the decisions of one
corporation are affected directly by the decisions of another corporation; on
the other, to the legal ownership; that is, the percentage of the equity of a
corporation owned by another corporation (Hymer 1977 [1960],
pp. 32–33). How significant the extent to which the decisions of one corpo-
ration are affected directly by the decisions of another corporation should
be (i.e. what percent of the equity of a corporation has to be owned by
another corporation in order to speak of “control”), is not explicitly de-
fined. However, in this perspective, the multinational corporation comes
into existence by direct foreign investments. The operation of direct foreign
investment may take different modes. “The form will vary; there may be col-
lusion, tacit or overt; the enterprises may merge and become one firm; they
may have a profit-sharing agreement (through minority interests, for
example); for a time the enterprise may even compete; but if there is inter-
dependence and if there are only a few firms so that they can recognize this
interdependence”, a multinational corporation comes into existence (Hymer
1977 [1960], pp. 91–92). In Hymer’s view the “multinationality” of the
multinational corporation refers to the crossing of two or more nation-states
by direct foreign investments. Here, the main attribute of the multinational
corporation is seen in the ability to face different nation-state based factors,
such as different governments, different laws, different languages, and dif-
ferent economic and cultural conditions (Hymer 1977 [1960], e.g. p. 28).
326
Robert Z. Aliber’s view on defining the multinational corporation marks

another prominent approach within the theory of direct foreign investment
(Aliber 1970, 1971). According to Aliber, the multinational corporation is
defined by the fact that its financial capital (utilised for the acquisition of
plant and production equipment) is moved across different national curren-
cy areas. Therefore, the distinction between domestic and foreign investment
depends on the existence of different currency areas. In other words, the
operational boundaries of the multinational corporation converge with the
boundaries of different (national) currency areas. “The ‘foreignness’ of the
investment reflects the movement across the boundaries … between curren-
cy areas. In the absence of such boundaries, the distinction between foreign
investment and domestic investment disappears. Now the boundaries … of
currency areas tend to be congruent with national boundaries” (Aliber
1970, p. 21). The significance of the explanatory variable “currency areas”
is explained by the microeconomic relevance of different interest rates
(caused by different factor endowments of capital) and different costs of
capital (caused by different exchange-rate risks) within the international
environment. The multinational corporation has, therefore, to consider a
multiplicity of specific risks and opportunities when making investment
decisions (Aliber 1971, p. 54). According to these arguments, the multina-
tional corporation is characterised by its ability to move financial capital
across two or more national or regional “currency areas”.
John H. Dunning’s “eclectic” theory of the multinational corporation pos-
tulates that the multinational enterprise distinguishes itself from historical
international transactions in two respects (Dunning 1971, p. 16). Firstly, it
embraces the international transfer of separate, but complementary, factor
inputs (viz. equity capital, knowledge, and entrepreneurship, but also goods
and services) usually controlled by a single institution. That is, a multina-
tional corporation not only exports singular factors of production to other
countries, but also exports package deals, thereby controlling their utilisa-
tion. The second unique quality of this corporation form, Dunning argues,
is that the factors of production that are transferred between countries are
not traded, but rather moved and controlled within the corporation. Thus,
the “multinationality” refers to the transfer of factors of production across
at least two nations. In this regard, the following remark is central: “As a
firm widens its territorial horizons, it changes its character to a certain
extent” (Dunning 1971, p. 31). Dunning argues that a firm that graduates
from supplying a regional market to supplying a national market, finds itself
in competition with new firms and is confronted with expanded market
structures. Similarly, a company with plants in more than one part of the
country would have more flexibility in the organization of its production
activities than a single-plant company. Finally, a company with production
units in more than one country is faced with environmental conditions that
327
differ from both of those forms. That is, John H. Dunning sees the specific
quality of the “multinational” in the fact
that multinational enterprises directly control the deployment of resources in

two or more countries and the distribution of the resulting output generated
between these countries. (Dunning 1974, pp. 14–15)
In contrast to the national corporation, the multinational corporation con-

trols its production of goods and services in different nation-states. Thus,
facing different national environments – that is, economic, legal, and politi-
cal differences between nation-states regarding the conditions of production
of goods and services – seems to be the specific feature of the multinational
corporation, because “[n]one of these problems really face the multi-plant
domestic firm. Within a country there is usually unrestricted (though not
costless) mobility of factor inputs and goods; there is a single currency and
rarely will there be any substantial differences in the rate of profits tax
between the regions” (Dunning 1971, p. 32).
Current approaches
John Cantwell’s evolutionary approach defines the basic operation of the
multinational corporation as technology (Cantwell 1989 and 1995). In con-
trast to the traditional meaning of technology as an engineering concept that
describes the mechanics of production processes and the physical character-
istics of the products made, Cantwell’s definition relates “technology” to the
entire process of production, i.e. to scientific as well as to organizational
aspects of a given corporation (excluding financial and marketing innova-
tions). It is primarily the “tacit” element of technological operations by
which a single multinational corporation accomplishes its specific reproduc-
tion. The uniqueness of technological experiences, embodied in the existing
organizational routines and skills, represents the specific aspect by which the
multinational corporation distinguishes and differentiates itself from its
environment (e.g. from competing corporations): “This is the part of tech-
nology which differentiates firms or MNCs, and which cannot be exchanged
between them as it is derived from and tied to the localised and collective
learning experience of the teams of a given company through their own
development of production” (Cantwell 1995, p. 24). In contrast to the
exchangeable public elements of technology, only its tacit and embodied ele-
ments constitute the “essence” of corporation-specific competitive advan-
tages (Cantwell 1995, p. 25). Cantwell describes the “multinationality” of a
given corporation by referring to the international interdependence of its
local technology-based subsidiaries (Cantwell 1995, pp. 37–45). Thus, local
technological innovations at one position are affected by the innovations of
another position within the multinational corporation. This reproduction of
328
technological operations within an “international” network of other tech-

nological operations of the corporation allows the corporation to distin-
guish itself from other (forms of) corporations: “MNCs are distinguished
from other firms through their deployment of international networks for
innovation” (Cantwell 1995, pp. 50–51). It is the fact that “tacit capability
and hence innovation is differentiated across countries”, which constitutes
the specific quality of this corporation form (Cantwell 1995, p. 48). The dif-
ferentiation of the multinational corporation is described as a differentiation
of technological capabilities, which is reinforced by increasing feedback
between the corporation and its different environmental conditions; as a
growing embedment in different national systems of innovations (Cantwell
1992, p. 77 passim). Thus, from Cantwell’s point of view, the “multina-
tionality” of a corporation refers to its ability to transfer technological capa-
bilities across different national systems of innovations.
Drawing on the theory of interorganizational networks (Benson 1975;
Cook 1977) Sumantra Ghoshal and Christopher A. Bartlett outline a net-
work theory that defines the multinational corporation as a “network” of
specific exchange relationships (for example, concerning the transfer of pro-
duction equipment, marketing skills or management capabilities) between
organizational units (including headquarters and different national sub-
sidiaries) that are embedded in a structured context of different national
environments:
We propose here a framework that conceptualizes the multinational as a net-

work of exchange relationships among different organizational units, includ-
ing the headquarters and the different national subsidiaries that are embed-
ded in … a structured context. (Ghoshal/Bartlett 1993, p. 79)
Ghoshal and Bartlett define the environmental reference of the multination-

al corporation as a structural network of different national organization-sets
of suppliers, customers, partners, regulatory agencies, competitors, etc. The
different members of a national organization-set may themselves be con-
nected through exchange ties. Further, the different national organization-
sets in which the corporate subsidiaries are embedded can also be connect-
ed through exchange relationships. “Due to such linkages among different
local organization-sets, all members of all the organization-sets of the dif-
ferent units of M [= MNC] collectively constitute what we shall call the
external network within which the multinational network is embedded”
(Ghoshal/Bartlett 1993, p. 87). Each single national subsidiary reproduces
the whole corporation not only by internal resource exchange with other
subsidiaries, but also by its exchange relationships with an external network
of economic, political and legal actors. That is why Ghoshal and Bartlett
define “the MNC as an interorganizational system rather than as an organi-
zation” (Ghoshal/Bartlett 1993, p. 100). Thereby, according to Ghoshal and
329
Bartlett the “multinationality” of a corporation refers to its internal as well

as external exchange relationships across (and within) different national
organization-sets.
According to the previous analyses we can summarise two theoretical
shortcomings these classical and current definitions on the multinational
corporation have in common:
1. All approaches define the “multinationality” of a corporation by relating
it to one specific type of operation. Hymer, for example, defines the
multinational corporation with reference to direct foreign investments,
Aliber refers to financial operations, Dunning to production operations,
Cantwell to technological operations, and Ghoshal and Bartlett to oper-
ations of specific resource exchange. Obviously, there exists a theoretical
confusion about the type of operation to which a definition of the multi-
national corporation should refer. If we want to decide whether a corpo-
ration is a multinational corporation or not, do we have to look at its
financial or productive operations, at its technological operations or at
another type of operations? How should we deal with corporations pro-
ducing only in one country, but taking advantage of global financial mar-
kets; or transferring technology within one country, but marketing their
products in many countries of the world; or recruiting management
employees only within the home country, but having locations of re-
search and development worldwide? Because of these unanswered ques-
tions we need a more general definition of the multinational corporation,
which does not restrict the multinationality of a corporation to one spe-
cific type of operations (i.e. to a specific corporate division such as
finance, production, marketing, etc.). We need a general definition in the
sense that it considers that a modern corporation might operate in a very
limited (e.g. local) manner in one respect and at the same time in a more
or less unlimited (e.g. global) manner in another.
2. In each approach, multinational corporations are characterised with
regard to their ability to operate in two or more national environments.
In contrast to national corporations, multinational corporations are con-
ceptualised as operating across different – more or less separated –
national environments such as nation-states (Hymer), national currency
areas (Aliber), national systems of innovations (Cantwell), or national
organization-sets (Bartlett/Ghoshal). But in times of increasing econom-
ic, political, legal, scientific, and cultural globalisation of the present
world society (see for example: Robertson 1992; Ohmae 1995; Meyer et
al. 1997; Beck 1998a, 1998b; Dicken 1999) it seems anachronistic to
consider nation-state based units the primary environmental reference of
large corporations. We assume that the social environment of present
corporations can no longer be described as a plurality of specific national
330
environments to which the corporate operations refer, but as a world-

wide societal environment whose boundaries have to be defined in func-
tional, and not in geographical terms. “The adjective societal emphasizes
that organizational sectors in modern societies are likely to stretch from
local to national or even international actors. The boundaries of societal
sectors are defined in functional, not geographical terms: sectors are
comprised of units that are functionally interrelated even though they
may be geographically remote” (Scott/Meyer 1991, pp. 117–118). There-
fore, a definition is necessary that enables a description of the multina-
tional corporation that takes into consideration these globalised struc-
tures of the present world society.
In the following sections we react to these theoretical problems by drawing

particularly on three concepts of Luhmann’s sociological systems theory:
organization, corporation and world society. Luhmann conceptualises
organizations as a specific form of social systems reproducing themselves as
networks of decisions. Corporations are regarded as a specific type of
organization reproducing itself by relating its decisions to business opportu-
nities. This operational understanding of the corporation should enable us
to develop a broader definition of the multinational corporation that is no
longer restricted to the operations of a specific corporate division (financial
decisions, productive decisions, marketing decisions, etc.). Further, Luh-
mann conceptualises the present environment of organizations not as a plu-
rality of nation-state based units, separated through territorial boundaries,
but as a single world society whose boundaries can no longer be seen as
restricted to nation-state based boundaries. This understanding of the envi-
ronment of organizations has far-reaching consequences for conceptualising
the so-called “multinational” corporation.
Luhmann’s concepts of organization, corporation

and society
Organization and corporation
In this section we want to outline four central points of Luhmann’s systems-
theoretical concept of organizations (corporations) which are crucial for the
argumentation presented in this chapter: (a) decisions as elementary opera-
tions of the organization, (b) two levels of organizational reproduction
(double closure), (c) corporation as a specific type of organization, and (d)
society as the organization’s environment.
a) Luhmann conceptualises organizations as social systems, which repro-
duce themselves on the basis of decisions (Luhmann 1976a; 2000c; 2003).
In other words, organizations are “made up of decisions, and capable of
331
completing the decisions that make them up, through the decisions that
make them up” (Luhmann 2003, p. 32). Being a specific form of communi-
cation, the elementary operations of organizations – decisions – are not
understood as psychological operations of an individual, nor as an internal
event within a human consciousness, but as a social phenomenon. Decision
communications are not produced by “human beings” but by the organiza-
tion itself. What is specific about decisions in contrast to other forms of
communication is that they are “compact communications” which commu-
nicate their own selectivity (Luhmann 2000c, p. 185). In contrast to ordi-
nary communications, which only communicate a specific content that has
been selected, a decision communication communicates also – explicitly or
implicitly – that there are alternatives that could have been selected instead.
Any decision communicates a selection between the alternative preferred
and the alternatives rejected. Decisions are decisions, and not only mere
communications, because they communicate also the context of possible
alternatives from which the alternative selected is chosen. These alternatives,
however, are not merely alternatives which just happen not to have been
selected, but they are constitutive for the alternatives selected: the meaning
of every single decision depends to a large extent on what has not been
selected, i.e. the context of the decision (Luhmann 2000c, p. 64).
b) The autopoietic reproduction of organizations by communicating deci-
sions implies a double closure of organized systems (Baecker 1999c, pp.
126–168; Luhmann 2000c, pp. 61–74): the first closure is realised by the
fact that decisions are produced by the network of other decisions.
Organizations reproduce themselves exclusively on the basis of decisions.
No external operations can be part of the network of decisions nor can any
decisions break out of this network. On the basis of its decisions, the
organization has no contact to its environment. Decisions are only connect-
ed to other decisions and nothing beyond the decision network. Thus, in
actual fact, the reproduction of decisions is the reproduction of the distinc-
tion decision/non-decision, i.e. of the distinction organization/environment.
According to this conceptualisation every single decision constitutes and
reproduces the “boundary” of the organization. The boundary reproduced
by every single decision implies an operative closure in the sense that the
organization can only be reproduced by operations constituted as deci-
sion/non-decision distinctions – other distinctions, for example, thought/
non-thought, cannot. On this operative level of merely reproducing deci-
sions, the organization is “blind” with regard to the complete distinction
with both its sides of decisions (system) and non-decisions (environment).
On this level, organizations “see” only one side of the distinction that con-
stitutes their operative boundary; that is to say, organizational decisions and
not environmental non-decisions. The only thing that is important here is
the continuous reproduction of decisions out of the network of decisions
332
– regardless of which decisions are selected and which decisions are exclud-
ed; regardless of the meaning of every single decision depending on envi-
ronmental criteria. But as long as it is communicated that any decision (i.e.
any selection between alternatives) has been made, the autopoietic repro-
duction of the organization can be continued.
The second closure of the organization is realised according to how deci-
sions are made and are related to the organization’s environment. The
“blindness” of decisions on the operative level is compensated through the
re-entry of the distinction decision/non-decision, i.e. the distinction organi-
zation/environment, into the decision process of the organization. This
second level of autopoietic reproduction of organizations is called self-
observation (see Luhmann 1986b for this term in general). Only on the level
of self-observation regarding its basic operations is an organization able to
observe the complete unit of each single decision constituted by the distinc-
tion decision/non-decision, i.e. the distinction organization/environment.
Only on this second level of autopoietic reproduction can an organization
see both sides of the distinction between decisions and non-decisions, i.e. the
organization itself and its environment. The re-entry of this distinction into
the decision process of the organization implies a self-referential aspect and
an environment-referential aspect. While the self-referential aspect refers to
decisions made out of decisions, the environment-referential aspect relates
this process of decision-making to the environment of the organization,
which determines the criteria on which the self-referential reproduction of
decisions depends (e.g. market size, consumer preferences, competition
structure, political and legal regulations, scientific knowledge, cultural
values, etc.). By observing its own decisions through the distinction self-
reference/environment-reference – which is not identical with the distinction
decision/non-decision itself – the organization system is able to inform itself
about its current state (the alternatives selected), its possible states (alterna-
tives rejected), and necessary changes due to environmental factors. Which
self-observations of its decisions are made depends not on the environment,
but only on the internal conditions, i.e. structures, of the organization itself.
Thus, neither on the level of its operations (first closure) nor on the level of
its self-observation (second closure) does the organization get into direct
contact with its environment.
c) The concept of double closure enables Luhmann to distinguish differ-
ent types of organization. On the operative level (first closure) all organiza-
tions are equal insofar as they reproduce decisions through decisions
– regardless of which decisions. Organizations represent a specific form of
social system because they reproduce themselves by decision communica-
tions – in contrast to ordinary communications. On the level of self-
observation (second closure), however, the general form of organization can
be differentiated into different types: organizations can be different
333
regarding which decisions are communicated and how these decisions are
related to the environment-referential aspect of an organization; in other
words, organizations can be different regarding the way in which the dis-
tinction organization/environment (reproduced by every single decision) is
observed within the organization. For example, social services observe this
distinction with regard to helping defaulters; schools observe it with regard
to imparting knowledge to students; and churches with regard to proselytis-
ing irreligious people. While the self-referential aspect of any type of
organization refers to decisions made out of decisions, the environment-ref-
erential aspect can imply very different criteria on which the self-referential
reproduction of decisions depends (e.g. “welfare”, “teaching”, “prose-
lytism”). Now we can address the question of what is – in Luhmann’s theo-
retical perspective – specific to the type of organization which is called “cor-
poration”. The organizational type of the corporation comes into existence
if, and only if, an organization observes the unit of its operative distinction
decision/non-decision, i.e. organization/environment, with regard to exploit-
ing business chances, i.e. profit maximisation (Baecker 1999c, pp. 237–242;
see also Sombart 1919, pp. 101–103; Sombart 1928, p. 321; Gutenberg
1958, pp. 43–44). The environment-referential aspect of the corporation
refers to those, and only those business opportunities that determine the
risks and chances of profit maximisation (e.g. market size, market growth,
competition structure, etc.) on which the self-referential reproduction of
decisions depends. In other words political, economic, legal, scientific, eco-
logical, and other societal aspects become relevant to the corporation’s deci-
sion-process only insofar as they imply chances or risks for profit maximi-
sation. Thus, the corporation can be characterised as a specific type of
organization as it not only reproduces decisions (at the operative level) but
also relates these decisions to business opportunities in order to realise and
maximise individual profit through accounting for cost and benefits (at the
level of self-observation).
d) A last point concerns the concept of the organization’s environment:
Luhmann’s organization theory is a societal theory of organizations, because
it conceptualises the organization’s environment as society. Society is the sys-
tem that encompasses all communications. All communications that are
produced are part of society and as such reproduce it. Hence there are no
communications outside society; the borders of society are the borders of
communication (Luhmann 1995f, p. 408). Here, the important point is
Luhmann’s assumption that the opportunities for the autopoietic reproduc-
tion of organizations (e.g. the opportunities for maximising profit by busi-
ness decisions) depend on the communication borders of society. In the
course of societal evolution, society’s communication borders have expand-
ed from local to territorial, national, and worldwide communication
borders. Each of these communication borders has specific implications for
334
the opportunities for the autopoietic reproduction of organizations: those of

the medieval township (e.g. restrictive town regulations on organizational
business behaviour – town-protectionism) are different from those of the
early modern territorial state (e.g. uniform regulations on production and
trade within a territory – mercantilism). Equally, those of the nation-state
based society of the 19th century (e.g. nationalised rules on organizing busi-
ness – liberalism, or rather, protectionism) are different from those of the
present world society (e.g. transnational economic rules – neo-liberalism, or
rather, neo-protectionism). According to these considerations, we can say
that Luhmann’s organization theory examines the question of how an
organization represents its environment, which is constituted by the com-
munication borders of the surrounding society (all non-decisions) within the
organization, and how the internally represented societal borders are relat-
ed to the process of making decisions out of decisions. In contrast to eco-
nomic, organizational, and managerial models of organizations, Luhmann’s
theory does not focus primarily on specific relationships between a given
organization and other organizations, between an organization and mar-
kets, between an organization and a nation-state, or between an organiza-
tion and the economic system. Instead, Luhmann suggests that we analyse
the very different relationships between an organization and society, i.e. the
– historically changing – opportunities of organizations for reproducing
themselves within society, against society (Luhmann 2000c, pp. 380–416;
see also Chapter 9 in this volume).
From pre-modern societies to the present world society

Luhmann distinguishes between different stages in the development of socie-
ty (see Luhmann 1977a; 1982f., and 1997a, pp. 609–776): archaic societies
were differentiated into equal subsystems, e.g. different tribes, clans or fam-
ilies. As the primary form of differentiation, it is the tribe, the clan or the
family that constitutes a society and not vice versa: a society is constituted
by a plurality of equal social units. Medieval societies were differentiated
into unequal subsystems (e.g. different social strata or classes) on a primary
level, and into equal relationships between town and land on a secondary
level. In contrast to both these forms of societal differentiation, modern soci-
ety is differentiated into different subsystems specialised in serving specific
societal functions; for example, producing consensus regarding different
interest groups (political system), distributing scarce goods (economic sys-
tem), providing justice (legal system), etc. All of these functional systems of
modern society are characterised as communication systems that reproduce
not only themselves but also the all-encompassing system of society by spe-
cific communications (political decisions, payments, legal ruling, etc.). Thus,
every single differentiation of the functional systems differentiates, i.e.
expands, the communication borders of society. The primary form of
335
functional differentiation of the modern society can be combined with other

forms of differentiation on the secondary level. The early modern society of
the 16th century is characterised by the combination of functional differen-
tiation as the primary form of differentiation on the one hand, and segmen-
tation into territorial states as the secondary form of differentiation on the
other. The modern society of the 17th and 18th centuries realises the seg-
mentation into mercantile states as the secondary form of differentiation;
and the functional differentiation of the modern society of the 19th century
is combined with the segmentation into different nation-states. Each of these
combinations has restricted the expansion of modern society’s communica-
tion borders in one or another way.
In contrast to these historical forms of (modern) society, Luhmann
describes the present society as a society whose communication borders are
no longer limited by any territorial borders. Luhmann argues that the pres-
ent society has to be conceptualised as a world society in the double sense
that it constitutes (a) one single, and only one single, structural network of
communications covering the globe and (b) one single, and only one single,
phenomenological world including all communication possibilities
(Luhmann 1975a; 1990m; 1997a, pp. 145–171; Stichweh 2000b):
Under modern conditions … and as a consequence of functional differentia-

tion, only one societal system can exist. Its communicative network spreads
over the globe. It includes all human (i.e., meaningful) communication.
Modern [present, DH] society is … a world society in a double sense. It pro-
vides one world for one system; and integrates all world horizons as horizons
of one communicative system. The phenomenological and the structural
meanings converge. A plurality of possible worlds has become inconceivable.
The worldwide communicative system constitutes one world that includes all
possibilities. (Luhmann 1990m, p. 178. See also Stichweh 1998, p. 215 and
1999b, p. 292.)
(a) Concerning the first aspect, present society is characterised by the fact
that there is only one single structural network of communications consti-
tuting only one single society. All previous societies were surrounded by
other societies. All previous societies had been built up by geographically
limited networks of communications, which were surrounded by other com-
munication networks of other societies. In other words, there was a plural-
ity of (local, regional or even national) societal systems that had only mar-
ginal (or even no) communicative contact to each other. In contrast, the
globalised communicative network of present society is no longer limited to
any territorial boundaries. Of course there are still territorial boundaries,
but they no longer define the communication borders of the societal system
(Luhmann 1982e, pp. 240–242). As a consequence of unlimited functional
differentiation, each communication relates to other communications on the
336
basis of a function-specific, symbolically generalised communication medi-

um (e.g. money, power, law, etc.), which is indifferent with regard to terri-
torial boundaries. Each of these generalised symbols of communication
media (Giddens calls them “symbolic tokens”) disembeds communications
from local contexts and relates them to the global context of world society;
economic communications relate to other economic communications, polit-
ical communications relate to other political communications, legal com-
munications relate to other legal communications, etc., wherever they take
place.
(b) Concerning the second aspect, Luhmann characterises the present
society as a single phenomenological world of communicative possibilities.
Understanding the world in a “phenomenological” sense (Husserl 1962; see
Farber 1967 for an introduction to Husserl’s phenomenological theory) does
not mean that the world is just another word for the empirical globe on
which social life takes place. It does not refer to some sort of physical reali-
ty, but to an emergent, namely social, reality that is constituted by commu-
nications, and only by communications. What is specific to the world of the
present society is the point that there is – as a second consequence of unlim-
ited functional differentiation – only one single social, i.e. meaningful,
world, which includes all communicative possibilities, that is to say, all pos-
sibilities of drawing and observing social distinctions. Each symbolically
generalised communication medium produces a universal world of endless
communicative possibilities, which is no longer restricted by religious or
moral considerations: the individual world-perspectives of economic
observers are integrated by a general social understanding of “money” that
opens up very different possibilities of making more money. Similarly, the
individual world-perspectives of political observers are integrated by a gen-
eral social understanding of “power” that spans very different possibilities
of accumulating political power; and the individual world-perspectives of
legal observers are integrated by a general social understanding of “law”
that spans very different possibilities of ruling social conflicts. Every single
world-societal communication not only produces and reproduces a single
communicative system, but also a single context of social meaning, which
includes all possibilities of dealing with economic, political, legal or other
function-specific problems respectively (i.e. there are no other possibilities of
social understanding outside such a communicative context constituted by a
function-specific communication medium).
Following this line of argument, Luhmann uses the term “world horizon”
in order to highlight the point that the social context of each world-societal
communication can no longer be thought as the unity of all things
(aggregatio corporum) or the unity of all observable and unobservable ideas
(universitas rerum). The social world of the present society, so Luhmann,
can only be represented as an endless horizon of other communicative
337
possibilities that is reproduced by every single communication on the struc-

tural level. The world itself as the unity of all actual and potential meaning
of communication cannot be represented by any single communication.
That is to say, each attempt to represent the unity of all actual and potential
meaningful communication just produces other possibilities of constructing
meaningful communication (e.g. of using and earning money, of maintain-
ing and accumulating political power; of ruling social conflicts, etc.). In
other words: the world itself has became unobservable for each social
observer because there is no longer any meaningful environment from which
this meaningful world could be distinguished. The world appears rather as
a meaningful horizon within which possibilities of constructing meaningful
communication refer to other possibilities of constructing meaningful com-
munication (and so on), and beyond which are no other possibilities of con-
structing meaningful communication.
While this general form of world society started to emerge around the 18th
century, an entirely globalised world society did not come into existence
before the second half of the 20th century (see Willke 2003 for a systems-
theoretical analysis based on Luhmann’s concept of world society). The
present world society is no longer characterised by the combination of func-
tional differentiation into various societal subsystems, and segmentation
into different territorial units. Rather, its functional differentiation is com-
bined with its differentiation into globalised structures of expectation with-
in each of these societal subsystems: for example, global economic relation-
ships, political agreements, legal rules, etc. These are caused by the transna-
tional activities of a historically new level of global organizations such as
WTO, GATT, OECD, World Bank, or INGOs (see Luhmann 1997a,
pp. 165–166). Only because of the emergence of such globalised structures
of expectation, which are described in more detail in the next section, has
one single structural network of communications as well as one single phe-
nomenological world become possible. Only these organized structures of
expectation have ensured that communications can be connected continu-
ously to other communications worldwide, on the basis of a similar social
understanding of economic, political, or legal behaviour. Now, we want to
explore some theoretical consequences for (re-)defining corporations that
reproduce themselves within such a globalised world society.
From the medieval business organization to

the present world corporation
In this section we will outline a sociological concept of the world corpora-
tion (as a particular form of corporation) that is based on Luhmann’s sys-
tems-theoretical insights. Let us start with Luhmann’s assumption that the
338
form of a corporation depends on the communication borders of its sur-

rounding society and the way in which these borders are represented within
the corporation.
Historical forms of business organization

Medieval business organizations of the 14th and 15th centuries (e.g. craft
organizations, trade organizations, family organizations) came into exis-
tence by distinguishing themselves from the communication borders of a
local township within these communication borders; that is to say, by
observing the distinction organization/township. They reproduced them-
selves by relating their decisions (self-referential aspect) to opportunities for
covering the essential requirements (e.g. food, housing, clothing) of their
members (environmental aspect). These opportunities for covering essential
requirements were limited strictly through several regulations of the
medieval town: for example, political and legal rules concerning the import
and export of raw material and goods, as well as the production of goods;
market rules on buying and selling goods; theological and moral rules on
social and economic behaviour, etc. (Sombart 1919, pp. 70–91).
Early modern corporations of the 16th century came into existence by dis-
tinguishing themselves from the communication borders of a territorial state
within these communication borders. They no longer reproduced themselves
by making decisions on covering the essential requirements of their members
but by making decisions on maximising profit (Sombart 1919, pp. 101–172;
Gutenberg 1958, pp. 43–44). In other words, their reproduction was char-
acterised by relating self-referential decisions to environmental business
opportunities. Those business opportunities were enabled through several
societal developments that were connected to the combination of the func-
tional differentiation of society into autonomous subsystems as the primary
form of differentiation, and the segmentation into territorial states as the
secondary form of differentiation (e.g. the replacement of local economies
by territorial economies, which were connected to each other through
“international” trade relations based on the symbolically generalised com-
munication medium of “money”; the replacement of local political systems
by political systems which were specialised in serving territorial consensus
between different social groups by implementing unified social rules on the
basis of symbolically generalised “power”; the substitution of local legal sys-
tems with legal systems which were specialised in serving territorial peace by
implementing unified legal norms on the basis of symbolically generalised
“law”, etc.). On the other hand, corporations were faced with several limi-
tations regarding the expansion of their business environments beyond ter-
ritorial borders (e.g. import and export rules made by their home states).
339
Trade organizations of the 17th and 18th centuries (e.g. English or Dutch
East-Indian companies) came into existence by distinguishing themselves
from the communication borders of a mercantile society within these com-
munication borders. Here, the self-referential aspect of the distinction
organization/mercantile society re-entering the organization by every single
decision is related to an environmental aspect that is imagined as a society
in which different (European) states struggle for political and economic
resources. What is specific to these trade organizations – which were used
by their home states as one central instrument of mercantile foreign policy –
is the fact that they reproduced decisions on maximising profit as well as on
accumulating (political) power. In contrast to the early modern corporations
of the 16th century, their business environments were no longer restricted to
the territorial borders of their home state. On the other hand, however, their
business opportunities were limited to certain trade routes, trade partners,
branches of trade, commodities, etc., besides which other disregarded and
unexploited business opportunities had existed (Chaudhuri 1978; Gaastra/
Bruijn 1993).
Large corporations of the late 19th century (e.g. Singer, Standard Oil
Company, Bell Telephone Company, Edison Electric Light Company, Bayer,
BASF, Hoechst) came into existence by distinguishing themselves from the
communication borders of a nation-state based society within these com-
munication borders. They reproduced themselves by relating their self-
referential decisions to expanded (environmental) business opportunities
that were enabled through various processes of nationalisation, which at the
time were taking place in most states (nation-state building) of modern soci-
ety: e.g. the abolishment of social inequalities, which was connected with
hierarchical societal strata through the performance principle; the transfor-
mation of feudal property into private property; the abolishment of all non-
economic barriers of market entry, such as privileges and monopolies guar-
anteed by the state; unified rules of business procedures, nationwide busi-
ness jurisdiction; unified structures of communication and traffic; general
freedom of production and trade; legal acceptance of foreign corporations,
etc. The modern society was represented within those corporations as a col-
lection of – more or less – similar national business environments allowing
them to capture new opportunities for maximising profit on the basis of
mass production and mass distribution (Wilkins 1970; Teichova et al. 1986;
Wilkins 1991). But, for several reasons we cannot discuss here, firstly, the
business opportunities of each – European as well as American – corpora-
tion were limited to a small number of nation-states, and secondly, those
“multinational” business opportunities were not observed as interrelated
business opportunities constituting a single, worldwide business environ-
ment.
340
The present world corporation

In contrast to the different forms of corporation described above, during the
20th century a specific form of corporation came into existence by distin-
guishing itself from the communication borders of a globalised world socie-
ty within these communication borders. In other words, since the middle of
the last century we have been able to observe a form of corporation that
comes into existence by the re-entry of the distinction organization/world
society into its decision process. What is specific to this form of corporation
is the fact that the self-referential aspect of the distinction re-entering the
organization by every single business decision is related to an environmen-
tal aspect that is represented as, and only as, an unlimited “world horizon”
of business opportunities. In the last section we pointed out that Luhmann
himself introduced the concept of the “world horizon” in order to charac-
terise the present society. Here, we want to use this concept in order to char-
acterise a particular form of corporation. We want to call this form “world”
corporation. We suggest that if, and only if, a corporation relates (i.e.
observes) its decisions (with regard) to an endless world horizon of business
opportunities, should we speak of a world corporation. Any historical form
of corporation has related its decisions to a business environment, which has
had itself an environment of other business opportunities that had not been
taken into account by the decision-process of the corporation: the local busi-
ness environment of medieval business organizations was surrounded by
other local business environments; the territorial business environment of
early modern corporations was surrounded by other territorial business
environments; the European-Asian business environment of East-Indian
trade organizations was surrounded by other business opportunities refer-
ring, for example, to transatlantic or African trade; and the few national
business environments of the large corporations of the late 19th century were
surrounded by other local and national business opportunities somewhere
on the globe. In contrast to these historical forms of corporation, present
world corporations relate their decisions to a business environment that
includes all business opportunities existing on the entire globe (see Hilliard
2005 for a more detailed analysis of the evolution of present world corpo-
rations). Here, the term “world horizon” refers to the fact that the decisions
of the corporation are related to a business environment which itself has no
other business environment(s).
In other words, world corporations can no longer rely on protected mar-
kets, national contexts, or territorially bounded political and legal power,
nor on the implicitness of a national consumer culture. Rather, these corpo-
rations have to take into account several world societal developments such
as the globalised structures of (a) a world economic system (e.g. more and
more globalised commercial relations because of which it has become mean-
ingless to speak of “national” economies), (b) a world economic policy (e.g.
341
worldwide political agreements on market and trade liberalisation, support-

ed by different global organizations such as WTO, GATT, OECD, or the
World Bank), (c) a world economic law (e.g. worldwide harmonisations of
national laws referring to the import and export of goods and services), or
(d) a worldwide consumer culture (e.g. globalised preference structures).
Globalised structures of this kind open up an endless horizon of business
opportunities, which cannot be reduced to specific national or regional
boundaries: worldwide possibilities of buying and selling goods and services,
worldwide possibilities of sourcing and distributing, worldwide possibilities
of financing and producing, worldwide possibilities of recruiting human and
technical resources, etc.
These world-societal developments have created a business environment
for corporations in which a corporation’s competitive position in one coun-
try significantly depends on its position in other countries or vice versa (see
also Porter 1986). Therefore, the corporation’s world horizon of business
opportunities refers not merely to an addition of more or less isolated
(national or regional) business opportunities but to a single environment of
interrelated business opportunities in which the rivals compete against each
other on a worldwide basis. Consequently, a world corporation must – in
some way – make its decisions on a worldwide basis in order to exploit the
interrelations among these various business opportunities. Following this
line of theoretical argumentation we can now define more precisely what
makes up the new quality of the corporation form that is represented by
present world corporations: rather than being constituted primarily by an
increase of possibilities of maximising profit, it refers to the nature of profit
maximisation itself. With regard to world corporations, the nature of prof-
it maximisation refers to a specific range of economies of scale and scope
that is are realised by decisions that take into account the differences and
similarities between all the different business environments existing on the
globe. In the next section we will illustrate these general considerations by
applying them to three corporate divisions: finance, production and mar-
keting.
Finance, production and marketing of the world corporation

Concerning the division of finance, a world corporation might come into
existence if decisions (on raising and investing capital) are made in view of
a single world horizon of financial opportunities; in other words, if decisions
are not primarily related to (a plurality of separated) national financial envi-
ronments, but orientated to the worldwide interrelations between different
local financial environments as well as local and global financial markets.
This type of interrelations concerns interest rates, inflation rates and tax
burdens, currency risks, investment conditions such as investment incen-
tives, protection of investment, investment-income benefits, etc. (Lessard
342
1986). Here, the specific quality of profit maximisation can arise, firstly,
from arbitrage of cross-border rate differences between different financial
environments via internal financial transactions (e.g. advancing of funds to
a subsidiary through an equity or a loan, a transfer of goods or intangibles
at less than an arm’s-length price, or a guarantee that enables it to borrow
locally); secondly, it can arise from arbitrage across different tax regimes
through shifting income into jurisdictions with relatively low rates or rela-
tively favourable definitions of income (e.g. via the pricing of interaffiliate
financial transactions or transfer prices of real inputs and outputs); thirdly,
it can arise from exploitation of exchange rate-volatility by shifting this risk
to hedging transactions such as currency futures, swaps, foreign-currency
borrowing, or to suppliers and customers by the choice of invoicing
currency.
Concerning the division of production, a world corporation might come
into existence if decisions are related to a single world horizon of linked pro-
duction opportunities. The specific quality of a world corporation’s profit
maximisation can arise, firstly, from relating the decisions on locating its
diverse production activities (number, size, and location of plants; choice of
technology and equipment; assigning the production of materials, compo-
nents, and products to specific manufacturing facilities; vertical span of the
manufacturing process) to a worldwide production environment, which
includes all possibilities for production (Flaherty 1986). These decisions
define the set of plants and their manufacturing processes, as well as the
physical flows of material and products among them. Besides decisions on
centralising production within one local factory, which can manufacture suf-
ficient volume to supply the world market with a single product design at
low cost, because of huge economies of scale, specific profit can be gained
by decisions on decentralising manufacturing (of similar or dissimilar prod-
ucts) across different production environments. For example: first, by
exploiting (a) differences in location-specific risks and chances across
worldwide location options, (b) differences in operating costs across world-
wide technology options (e.g. taking advantage of short-run price move-
ments), (c) differences in capital costs and exchange rates across worldwide
financial options, (d) differences in product stages across worldwide market
options (e.g. taking advantage of the introduction of new products by build-
ing on an existing base, or of fine-tuning subsequent moves in response to
changing market conditions through downstream investment). Second, (a)
by taking advantages based on intangible (technology) assets that can be
transferred easily within a corporation, but only with difficulty outside a
corporation, or (b) by taking advantage of using one’s own technology in
different foreign factories (e.g. gaining a specific return on a corporation’s
technology that would not be possible through an arm’s-length sale of this
technology).
343
In addition to that, the specific quality of a world corporation’s profit

maximisation can arise from relating the decisions on the coordination of its
dispersed production activities (i.e. the sharing and use of information about
manufacturing and technology support by different facilities) to a world-
wide environment of possibilities for production. For example: (a) by deci-
sions on procurement, i.e. on obtaining material, subassemblies, compo-
nents, and equipment needed for manufacturing from outside suppliers (e.g.
profit by lowering cost and improving quality through sending procurement
specialists – who help their counterparts identify and negotiate with local
suppliers – from one local plant to another); (b) by decisions on aggregate
production planning, i.e. on the specification of what volume of which prod-
ucts will be produced on what date in a plant for a defined period (e.g. prof-
it by (i) equalising production requirements in one local environment with
free capacity elsewhere, (ii) by shifting orders among different business envi-
ronments to hedge exchange rate fluctuations, or (iii) by transferring signif-
icant information about products and their material content, as well as new
parts and products across different business environments through close
coordination between the monthly plans of the various plants); (c) by deci-
sions on manufacturing engineering, i.e. on all engineering activities related
to different local business environments (e.g. profit by saving manufacture
costs by transferring technological improvements developed at one local
plant to other similar local plants, transferring manufacture responsibility
and capability for one product from one local plant to another, etc.).
Concerning the division of marketing, a world corporation might come
into existence if decisions are related to a single world horizon of linked
marketing opportunities (Takeuchi/Porter 1986; Dahringer/Mühlbacher
1991; Jeannet/Hennessey 1995). Here, the specific quality of a world cor-
poration’s profit maximisation can arise from relating the decisions on locat-
ing its various marketing activities to a worldwide business environment
which include all possibilities for marketing. For example, centralised pro-
duction of advertisements and sales promotion materials can produce
economies of scale and faster accumulation of learning in both the produc-
tion and development of goods; a centralised sales force stationed in the
home market or in regional headquarters can provide support to local sub-
sidiaries by sending them highly skilled sales specialists; centralised service
support can generate economies of scale in complex repair functions, or
analyse field data from around the world and establish appropriate servic-
ing procedures; centralised advertising can produce economies of scale by
selecting one advertising agency to handle its worldwide advertising oppor-
tunities (economising in needed coordination) or by using media with rela-
tively high global spillovers. Further, new ways of maximising profit can be
seen in decisions on decentralising the corporation’s marketing activities to
each national or regional business environment in which a corporation
344
competes on a worldwide basis, i.e. by being responsive to national or

regional business opportunities in performing its diverse activities (pricing,
distribution, advertising, sales promotion) market by market.
The specific quality of a world corporation’s profit maximisation also can
arise from decisions on coordination are related to a single world horizon of
linked marketing opportunities: firstly, from on decisions on performing
marketing activities using similar methods worldwide (reinforcement of the
corporation’s reputation or image when either buyers or information are
globally mobile; reinforcement of the corporation’s differentiation when
customers receive the same treatment by the sales service anywhere;
economies in training and in purchasing the marketing inputs). Secondly,
from decisions on transferring marketing know-how and skills from one
business environment to another (e.g. a shared market-entry approach can
reduce costs because the most successful practice can be applied to a new
business environment; shared marketing information regarding shifts in
buyer-purchasing patterns, life-style changes, new merchandising ideas,
early market signals by competitors etc., can imply specific opportunities for
maximising profit through knowledge-based competitive advantages).
Thirdly, from decisions on sequencing new products or marketing practices
across different business environments (e.g. cost reduction by employing
already developed products or programmes). Fourthly, from decisions on
the integration of the activities of various marketing groups across national
business environments (e.g. economies in the utilisation of the sales service
because of avoiding duplication in selling efforts; advantages of differentia-
tion through offering a single contact for global buyers or an exclusive after-
sale service).
Conclusion
In this chapter we have argued for a new conceptualisation of the multina-
tional corporation. We started off with a discussion on the existing
approaches, pointing out two central shortcomings: firstly, all approaches
define the “multinationality” of a corporation by relating it to just one spe-
cific type of operation. Each writer, however, sees a different type of opera-
tion as crucial: direct foreign investments (Hymer), financial operations
(Aliber), production operations (Dunning), technological operations (Cant-
well), or operations of specific resource exchange (Ghoshal/Bartlett).
Secondly, all approaches characterise the multinational corporation with
regard to the crossing of two or more nation-state based environments: the
crossing of national currency areas (Aliber), national systems of innovations
(Cantwell), or national organization-sets (Bartlett/Ghoshal). This reference
to a collection of nation-state based environments, however, seems to blind
345
out the “real” quality of such corporations: in times of increasing globalisa-

tion of the present world society and its various function systems, the main
attribute of multinational corporations can no longer be seen in the ability
of crossing different national boundaries and facing them as a plurality of
more or less separate national environments, but in the ability of accounting
for cost and benefits regarding a single business environment of business
opportunities linked worldwide.
On the basis of this criticism on economic and organizational approach-
es, we introduced the sociological concept of the “world corporation” refer-
ring to various insights of Luhmann’s systems theory. In contrast to histori-
cal forms of corporations, we defined present corporations as world corpo-
rations, if – and only if – they relate their decisions to a single business envi-
ronment of business opportunities linked worldwide and enabled through
differentiation into globalised structures of expectation; that is to say, if they
maximise profit on the basis of a specific range of economies of scale and
scope, which are realised by decisions that take into account the differences
and similarities between all the different business environments existing
within a globalised world society. We suggest that this systems-theoretical
definition of the world corporation represents a fruitful strategy for facing
the two theoretical problems mentioned above: firstly, it does not restrict the
“multinationality” of a corporation to a specific type of operations such as
financial operations, productive operations, marketing operations, etc. If we
take (business) decisions as the basic operation of corporations, it depends
on the given empirical case, in which respect a corporation reproduces itself
as a world corporation. Thus, a corporation might be described as a world
corporation, if it realises and maximises profit by relating its financial deci-
sions, its productive decisions, or its marketing decisions (and so on) to a
worldwide business environment. In other words, our definition enables one
to think of the world corporation as a world corporation with relation to
very different types of operations, or rather corporate divisions, by doing
away with any bias towards any specific type of operation or corporate divi-
sion. Secondly, this systems-theoretical definition takes into account that the
social environment of present corporations can no longer be regarded as a
plurality of separate nation-state based units to which the corporation’s
operations refer, but only as a single world-societal environment, which is
differentiated into globalised function systems. By assuming a single world
horizon of interconnected business opportunities, the specific quality of cor-
porations can be re-conceptualised as the ability to exploit a specific range
of economies of scale and scope, through decisions that take into account
the various differences and similarities between all the different business
environments existing within the present world society.
In addition to these two advantages of a systems-theoretical definition of
the world corporation we see several other advantages, which have to be
346
explored by further research in detail: first, conceptualising the primary

environmental reference of large corporations as a single world society that
is differentiated into various globalised function systems enables one to
analyse not only corporation/market-relations or corporation/economy-
relations, but very different societal preconditions for the realisation and
reproduction of the corporation form we called world corporation.
Historical studies have to explore in detail (a) which economic, political,
legal, scientific, and cultural developments had enabled a single world hori-
zon of interconnected business opportunities for corporations; (b) which
societal structures (for example, caused by a specific form of societal differ-
entiation or a low level of functional differentiation) had made it difficult for
corporations to maximise profit on a worldwide basis; (c) which societal
structures of expectation seem to be necessary to stabilise the reproduction
of world corporations. Second, conceptualising the basic operations of
world corporations as decisions that are related to a worldwide business
environment enables one to ask (a) why corporations take into account a
world horizon of interconnected business opportunities within one or more
corporate divisions, and not within other divisions; (b) which operative pre-
conditions within the non-globalised divisions must be given so that a cor-
poration can act within one or more corporate divisions as a world corpo-
ration; (c) which operative aspects within the globalised division(s) seem to
promote the globalisation of decision-making within the non-globalised
divisions. Third, conceptualising the new quality of corporate being as the
ability of maximising profit on a worldwide basis through decisions that
take into account the various differences and similarities between all the dif-
ferent business environments existing within the present world society
enables one (a) to avoid any reduction of the world corporation’s new qual-
ity to a simple growth of isolated opportunities of maximising profit; (b) to
explore the complex range of known and unknown economies of scale and
scope, i.e. to become aware of very different motives for maximising profit
on a worldwide basis; (c) to study the specific techniques of accounting for
cost and benefits, which are necessary to maximise profit on a worldwide
basis.
347
19 Kap 16. Luhmann 05-11-11 16.29 Sida 349
PART VI
Implications for Management

and Consulting
19 Kap 16. Luhmann 05-11-11 16.29 Sida 350
19 Kap 16. Luhmann 05-11-11 16.29 Sida 351
Chapter 16
Communication Barriers in
Management Consulting
Niklas Luhmann
I
Whenever management consulting is under discussion, the layman (a cate-
gory that includes practically all scholars) first of all thinks of applied sci-
ence. It is debatable whether microeconomics or financial science, social
psychology or perhaps even sociology ought to be the paradigmatic science
on which consulting can be based. The answer may well depend on the
problems posed by each individual case; for complex problems that touch
on several disciplines at once, an “interdisciplinary” orientation is consid-
ered advisable. In all of the disciplines, the basis of consulting consists of a
specific competence derived from empirically established generalizations.
Known statistical probabilities are applied to each individual case, even
though they are meaningless in individual cases. One makes such mistakes
discreetly, only to compensate for them by means of detailed analyses of the
object in question. In certain rare cases, these analyses are subsequently eval-
uated for the purpose of controlling and correcting the scholarly hypothesis.
In this model of the relationship between consultants and companies,
there are no deep-seated problems of communication. For whatever reasons,
the consultants may be inclined to present their knowledge as certain and
their proposals as based on careful work. In so doing, they surpass reality in
a way that cannot itself be included in their communication. As is generally
the case with application-oriented research, management consultants too are
the perpetrators, as well as the victims of a certain “rhetoric of applica-
tion”.1 At precisely the same time that they are active for economic or other
reasons, they fall victim to the difficulty of addressing the specific commu-
nicative relationship in which they participate within this relationship (it is
well known that the Palo Alto School has identified this problem as the
cause of paradoxical communication). However, once one has begun to trust
the concept of applied science, this difficulty does not constitute a
fundamental problem; at most, it causes disturbances and derailments of
1 See Mulkay (1987).
351
19 Kap 16. Luhmann 05-11-11 16.29 Sida 352
communication. If one trusts in the concept of applied science, one also

trusts in the communication that comes with it. For what would science be,
if it could not be communicated?
For a long time, the resulting practical difficulties were written off as
being due to “complexity”. Ever since the publication of a famous essay by
Warren Weaver – an essay from which, in the meantime, an entire literature
has sprung2 – it has been known that for large and complex systems, among
which business corporations must doubtless be counted, there are no ade-
quate scientific methods and findings; for science has methods only for sys-
tems that are small enough to be described with a few variables, or for large,
but homogenous, quantities that can be processed statistically. However, in
view of the tight corner in which research found itself, the only advice to be
given was to continue trying anyway and to gradually expand scientific
competence concerning this problem of complexity. In this process, one
guideline for the reduction of complexity could be to focus on the particu-
larities of the individual cases and avoid generalizations. To this purpose,
management-consulting companies cultivate the “experiences” that
inevitably accumulate with a large number of individual projects. Now, the
problem is how these experiences can be recorded, publicized, and discussed
within the consulting group – in seclusion, as it were, and apart from day-
to-day business. However, this merely returns us to the problem of induction
and the problem of illegitimate generalizations; in addition, the stories and
the very convincing “practical examples” acquire a questionable predomi-
nance.
Some communication problems can be settled, if one assumes that the
consultants possess scientific competence and have to include communica-
tion concerning their competence in their communication with the compa-
ny. Certain expressive means can accomplish this – for instance, a white lab-
oratory coat and a knowing countenance. However, these means will be suf-
ficient only if the advice given can be implemented strictly on technical
grounds. “Strictly technical” means, in this context, “without knowledge of
the theoretical connections”. In the field of application under discussion,
such a reduction to the strictly technical is not useful, precisely because of
its complexity. In such cases, theories must be “understood” in the process
of their application. Application tests whether the implied marginal condi-
tions actually hold true and continue to hold true, or whether they have to
be modified and even abandoned when new information affects their pre-
suppositions. If the consultants want to avoid being incorporated perma-
nently into the action network of the company, thereby being recruited and
“employed”,3 they must transfer a sufficient amount of information so that
2 See Weaver (1948). Examples of more recent work and conference volumes are: Lanzara and
Pardi (1980); Université des Nations-Unies (1986); Bocchi and Ceruti (1985).
3 The currently practiced compromise is temporally limited employment for consultants.
352
19 Kap 16. Luhmann 05-11-11 16.29 Sida 353
the company can act not just according to their suggestions but according to
the theory that lies behind them. However, the more embedded a theory is
in further theoretical connections – that is to say, the more its scientific qual-
ity increases – the less such a procedure can be implemented.4 Like the eco-
nomic system itself, the consultants who attempt to establish a scientific
basis for their proposals are forced to simplify and popularize their external
communication – that is, to change the language. Furthermore, it is notice-
able that in the field of research dedicated to the theory of the firm, fashions
change constantly. The meaningless names of these fashions (such as “orga-
nizational development” or “strategic planning,” etc.), at the same time, do
suggest a completely new orientation of research and consulting and thus
can serve as self-congratulatory labels in the scientific, as well as the eco-
nomic, competition between consulting firms.5
Undoubtedly, it is not a matter of giving up the model of applied science
altogether because of these weaknesses, and of adopting a counter-model
while keeping the flag flying. There is no counter-model available for such a
switch. We only want to pose the question of whether it is indeed correct
that everything a consultant knows or finds out can actually be communi-
cated in the same way as science can be communicated. In addition, we
would like to examine whether there might be conceptions that would
encourage precisely the cultivation of such non-communicable knowledge.
II
The model of applied science suggests that we should aim for the production
of the highest possible degree of agreement; it also suggests that communi-
cation is the means to achieve this goal. It acknowledges that communica-
tion may be difficult, but insists that it can be improved from the point of
4 Nowadays, however, the theory of science assumes that the limitation to “adjacent” theo-
retical connections is precisely typical and that it would be impossible in science to bring about
a change in theory without this limitation. Cf. Polanyi (1962), p. 59, and (1983), pp. 72 f.
Also, Philippe van Parijs (1981), p. 50. This merely means that science itself has a problem
with its own complexity.
5 Even systems theory is often perceived in this sense, as a “scientific program” or a program
for applying science. Of course, nothing is to be said against that. However, to conclude, there-
fore, that systems are nothing really given but merely analytical constructions or models,
would be a serious, undue reduction. This is the modus operandi of a theory of science whose
method derives from a naïve conception of the theory of science (see e.g. Lenk and Ropohl
1978, esp. Ropohl’s introductory contribution; concerning the application of the theory to
microeconomic research; see Lenk, Maring, and Fulda 1985). The naïveté that no theory of
science should allow itself consists of a lack of reflection: namely, the reflection that the sys-
tem concept is also applicable to the person who uses it for his analyses, and therefore cannot
very well be understood as a construct without empirical reference. In other words, one ought
not to confuse the relativity of all statements in regard to a particular observer (system refer-
ences), and the specific case of the system of science functioning itself as the observer.
353
19 Kap 16. Luhmann 05-11-11 16.29 Sida 354
view of its own problems. By contrast, a systems-theoretical analysis may

urge an entirely different thought upon us: namely, that consulting firms are
interested in the preservation of their own practices.6 They value system
boundaries, which they reproduce in their own operations. They must pre-
vent their best employees from being recruited by their client firms. They
must be able to end consulting relationships that have lasted too long, even
when their clients do not require it or when it explicitly goes against a
client’s wishes. Often, they will view the multiplicity of tasks, the similarities
and differences of problem areas in client firms, and the methodology of
their own modus operandi as learning conditions that allow them, in the
long term, to improve not only their market position but also the interest
and motivation of their employees.
While the demand for management consulting is often created through
internal conflicts in a firm (this is equally true for successful and less suc-
cessful, or even almost ruined, firms), the consultants must set a high value
on being able to stay neutral in the face of internal conflicts. If they agree
with one side, it has to appear like a “coincidence”, and should, whenever
possible, be justified by reasons that are unique to the internal dispute.
Often, conflicts within a firm will make it advisable to search for a third
alternative that will not be perceived as taking the already established point
of view of one of the quarreling factions. Thus, consulting often has the
function of providing open or latent conflict therapy. How its effect express-
es itself on the firm’s balance-sheets may then become a side issue.
Seen from a systems-theoretical perspective, a consulting group thus pre-
serves and reproduces itself in an environment that emerged as a conse-
quence of the group’s own differentiation and is given for it. Its preservation
and reproduction as a system-in-an-environment cannot be reduced to the
sum of all system-to-system relations, not even if the sum of these individ-
ual relations could be expressed by means of a monetary amount per time
unit, the profit or loss of the consulting firm. In distinguishing system-
environment relations from system-system relations, we acquire a tool for
posing the problem, one which could gain importance in the treatment of
communication problems; for no system can communicate with its environ-
ment. Communication always requires an addressee: that is, another system,
one which must be understood in turn as a system-in-its-own-environment
if one intends to communicate with it.7
6 There is no need to understand this principle in the narrow sense: that is, merely as the
preservation of the economic profitability of the firm. It also applies when a consulting firm,
perceiving more opportunities than it can exploit during a boom period of its particular mar-
ket, tries to confirm its “identity” even when turning down lucrative inquiries.
7 In this sense, our case study concerning consulting firms is merely an example for a much
more general problem; for no system can communicate with its environment, although every
system must reproduce itself in its environment – or it ceases to exist.
354
19 Kap 16. Luhmann 05-11-11 16.29 Sida 355
This means that the unity of a system participating in communication and

the unity of its environment can neither be included in the communication
nor expressed by means of communication. The unity of the difference
between system and environment is merely being reproduced in communi-
cation. If we proceed from the distinction between system and environment
and thus from systems theory, we gain a certain insight into the problem of
incommunicability (we are not obliged to proceed in this manner; to do so
is and remains the decision of the observer). That a group of consultants, as
in our case, cannot communicate itself completely (but is nonetheless
capable of communicating internally about this impossibility of complete
communication) is due to the fact that communication is the operation by
means of which the group carries on its own autopoiesis, and thus the means
by which it regenerates its own unity, as well as the difference between this
unity and its environment.
III
In the process of applying this general insight to the case of management
consulting, I intend to explicate this insight in several steps. There will be
some obvious parallels with other forms of therapy (psychotherapy, family
therapy, systems therapy, etc.); such parallels have become elements of the
self-understanding of many management consultants because of regular con-
tacts with these forms of therapy that are initiated by communications under
the influence of science. However, in the following we will not pursue these
possible expansions of our inquiry. Instead, we will limit ourselves to
management consulting.
Thanks to our earlier treatment of complexity, we can now recognize the
positive function of a difficulty that we hitherto looked at from a rather neg-
ative angle. The complexity of the theory nexus at the base of the consult-
ants’ judgments and proposals is not just an obstacle on the way to success.
It also ensures the continued non-identity of the systems. By accepting
demanding theoretical foundations, the consulting group gains the possibil-
ity of holding its own language aloof and thus avoids being drawn into the
jockeying for position and the factions that occur within the client system.
In this sense, incomprehensibility can serve as protection. “Assumptions
guided by theory enable consultants to remain in the position of observers
and to avoid assimilation to the role of participants.”8 Whether such theo-
ries, according to scientific criteria, are true or untrue (or, in most cases,
merely promising and connective), and whether or not they prove effective,
is only of secondary importance in regard to this problem. They mark and
8 This point is made by Exner et al. (1987), pp. 267 f. with reference to Kind (1986).
355
19 Kap 16. Luhmann 05-11-11 16.29 Sida 356
protect the unity of the system that employs them. The functions of marking
and protecting the unity of the system may remain latent for both sides of
the consulting relationship; in fact, they must remain latent for the client
side. If one tried to communicate about these functions within the relation-
ship, communication would become “paradoxical” in the precise sense of
the term, as it has been defined by systems therapy.
Once we have accepted this, additional prospects open up. Thus, the con-
sultants may well search for and employ theoretical principles that cannot
be accessible to the companies that hired them, at least as far as the struc-
ture of these principles is concerned. The consultants might be particularly
interested in “latent” structures and functions. As soon as they begin to
observe how a company observes or, rather, how observation is enacted in a
company – in other words, how distinctions are drawn and indications
made – they can assign a meaning to the problems of this system that is not
at the disposal of the system itself. And even if it were possible to transmit
this meaning, the transmission would affect the meaning semantically and
turn it into a different meaning; and even this meaning would be observed
differently depending on the system from which it is observed. Thus, the
insight into latent problems, functions, and structures does not necessarily
exclude the possibility of trying to include this latency in the communica-
tion; of making the “unconscious” conscious, very much as in psychoanaly-
sis; and of thereby achieving certain effects. However, the meaning of the
schema “latent/manifest” reaches much farther than this enactment of con-
sulting strategies. Consultants who work with this schema of observation
have to make the choice of whether they want to expose the latent functions
and structures or whether such a course of action is not advisable. In pre-
cisely those cases in which they notice that latency itself has a function – for
instance, the repression of unsolvable problems – they will go about expos-
ing such latency hesitantly as long as they can neither predict nor control its
effects. At the very least, they will suspect that exposing the “fundamental
lie” of the observed system might lead to catastrophe. “Catastrophe”, in this
context, is to be understood as another principle of stability and thus, in all
likelihood, as another “fundamental lie”.
Latency is not just any sort of absence. It is not a normal side effect of
structural or operational selections, arising from the fact that something in
particular is marked instead of something else. Latency is a specific mode of
absence; it is a “sustaining” – one might almost say a “subjective” – mode
of absence on the basis of which structural selections really become possible.
It should also be said that it is not a matter of the “unmarked space” – the
colorlessness of light, the noiselessness of air, the whiteness of paper – in
other words, the medium that makes the tighter couplings of specific forms
really possible. Rather, it is a matter of a specific type of prohibition that
itself remains inaccessible. Different theoretical interpretations of this fact
356
19 Kap 16. Luhmann 05-11-11 16.29 Sida 357
are possible – for instance, in Sigmund Freud’s or perhaps in René Girard’s

sense. If we look at the proposed theory on which I base my argument (a
theory which itself is sustained by such a prohibition), we see that every dis-
tinction that is used operationally can itself be distinguished; however, at
some point we can no longer ask: distinguished from what? And why?9
Now that the perfume of psychoanalysis has lost much of its flashiness
and fascination, we can put the problem under consideration into a more
general form with the help of the cybernetics of second-order observation.
The central issue is that the observer of an observer has and can use (or not
use) the possibility to see that which the observed observer cannot see. All
observation (including second-order observation) deals naïvely with its own
distinctions. It must presuppose the reference of such distinctions: that is, it
must assume that the distinction (let’s say: cause/effect; guilty/not guilty;
profitable/not profitable; protecting the environment/destroying the envi-
ronment) designates something real. In other words, it uses its own distinc-
tion as a blind spot that organizes the possibility of observation. This blind
spot can be replaced only if it is exchanged for another blind spot. Thus,
every system’s own basic distinction becomes a natural necessity, while a dis-
tinction used by others becomes merely an artificial, contingent, chosen, and
replaceable schema. But it is precisely this differentiation that can be seen
and communicated only if one has confidence in the distinction “natural/
artificial” and uses it naïvely, namely by presupposing it.10
This turn towards second-order cybernetics can be clarified if we compare
such cybernetics to classical theories of reflection. If we formulate the basic
problem as the reflection of reflection, we end up with an infinite regression
and, even worse, with a hierarchical architecture that signals our own
superiority. The reflection of reflection takes place at a “higher” level and
promises better insights. However, we also know that an infinite regression
cannot amount to an infinite improvement. Instead, all too soon the point
of minimal utility will be reached, to speak in economic terms. Second-order
cybernetics dispenses with this comprehensive architecture and thus avoids
any gesture of superiority.11 The relations of observation remain strictly
horizontal relations, no matter how intricate, demanding, and recursively
ordered they may be. Precisely in regard to the salient point – that is, the
9 This might well be the reason why George Spencer Brown (1971) bases his logical calculus
on the injunction: “Draw a distinction!”.
10 When we speak of “presupposing” something, we explicitly refer to that which once upon
a time was called “the subject”; however, we no longer assign a substantive meaning to this
concept, but an operational meaning, a meaning related to events.
11 On the critique of this pretension of theories of reflection and on the demand for a strictly
horizontal arrangement in the post-modern style, see also: Latour (1988a), pp. 168 f. How-
ever, Latour then (pp. 175 f.) turns explicitly against the (incorrectly presented) language of
observers observing observers.
357
19 Kap 16. Luhmann 05-11-11 16.29 Sida 358
operationally necessary naïveté concerning one’s own basic distinctions

– there is no hierarchy. At most, there are the following questions: What is
productive for which purpose? And how much complexity can be built up
on the basis of the connectivity of one’s own distinctions and indications?
By now, plenty of theoretical concepts have emerged that overtly or
covertly take advantage of the possibility of observing that which observed
observers cannot observe. Management consultants, for instance, can rec-
ognize (or believe to recognize) that a high level of personal attribution of
successes and failures may, at the same time, serve as an indicator of con-
cealed structural problems that nobody can touch without burning their fin-
gers. In such cases, one speaks of “designated patients”. “Symptoms are
readily attached to persons (and groups); among other things, such localiza-
tion serves to protect the structure.”12 This view corresponds to the general
insight of systems therapy, which often is called “systems” therapy precise-
ly because of this very insight: namely, that problems, for the most part, can-
not be cured at the place where the client localizes them. One variant of this
theory can be found in the theory of mimetic conflict and the violence it pro-
vokes, violence that does not affect the causes of the conflict but instead
affects certain chosen victims.13 According to this theory, the problem is pro-
duced by rivalry: that is, the unbearable equality of needs and desires that
leads to the establishment of violent differentiations (conversely, the so-
called Palo Alto School of systems therapy took the collapse of differences
as the starting point of its method of problem identification). In turn, such
diagnostic theories are distinctions – distinctions, however, that cannot be
transferred to the client whose problems they help to shift. They merely
serve as the consultant’s instrument for the construction of functional
equivalences, or as the background assumption for the construction of
“injunctions”, by means of which the system is to be moved to another, less
painful, view of the problem.
We are not questioning the scientific value of using such concepts in the
service of intervention. Also, we leave open the question of the extent to
which they have been used successfully in practice to date. We are only inter-
ested in the assumption on which they are based. This assumption implies
that a system could be overstrained through insight into (or communication
about) its own structures. The old concept of “latent” structures applied
merely to regularities that could be recognized only by means of specific
data collection and statistical analysis.14 This poses merely methodical prob-
lems and perhaps problems of complexity, but not communication prob-
lems. The other version of latency, derived from psychoanalysis, is more
12 Exner et al. (1987), p. 268.
13 See Girard (1982).
14 Empirical social scientists prefer to use the term “latent structure analysis”. Following Paul
Lazarsfeld they thereby refer to a specific measuring method with dichotomized indicators.
358
19 Kap 16. Luhmann 05-11-11 16.29 Sida 359
likely to admit the question of whether enlightenment is a responsible option

or might have destructive consequences, for which reason it might be avert-
ed. Second-order cybernetics provides yet another variant of this concept. In
this theory, specific latencies – namely, the non-distinction of the distinction
that is cognitively operational in each case – appear as the absolute condi-
tion of the possibility of cognition. Thus, the difference of distinctions itself
becomes problematic; distinguishing distinctions is of no help, since such a
procedure merely reiterates the same problem.
IV
If one splits up that which is given as “unmarked space” (in Spencer Brown’s
sense), as “matter” (as it is used in Aristotelian philosophy), or as “world”
(in Husserl’s sense) by means of the distinction between system and envi-
ronment, a strange effect ensues. The two sides of this distinction are given
only in the distinction itself. That is to say, they are given only for those who
use this distinction: namely, the observers. The unity of the system can be
found neither in the system nor in the system’s environment. The unity of
the environment can be found neither in the environment nor in the system.
The unity of one side as well as the other is, in the strict sense, nothing but
the unity of the distinction that separates and thereby unites these two sides.
Only the observers can see the unity of the system, since only they can dis-
tinguish the system from the environment. An observer need not be an
external observer; in fact, in the case of the system “society”, there is no suf-
ficiently complex external observer. The system itself can be the observer.
But in that case, the observer must conduct the observation of the system
within the system itself and use some of the system’s operations in the
process. This amounts to the drawing of a boundary between operations of
reflection, which use the distinction between system and environment, and
other operations that produce and reproduce the difference of system and
environment (according to the way the observer can see it at that time).
This state of affairs, including the problem of reflection it contains, makes
evident the advantages that can be gained by separating external and inter-
nal observation, other- and self-observation. Self-observation is always
bound to include reference to itself and thus, in the course of its execution,
to change that which it intends to observe. External observation can objec-
tify the system it observes. However, this objectivization also happens only
by means of its own operations of observation (that is to say, by means of
operations of observation that depend on the underlying structure and dis-
tinction).
Thus, possibilities of observation that could not be realized merely by self-
observation within a system can be gained from the difference between
internal and external observation. This difference presupposes system
359
19 Kap 16. Luhmann 05-11-11 16.29 Sida 360
difference: that is, in our case, the separation of firms that provide manage-
ment consulting from those that receive it. It can be observed from both
sides by means of the proper distinction; this is true even for a third observ-
er who might be interested in a theory of management consulting (as are we
right now). The choice of distinction that must be made operationally in any
given system is always based on an irreducible arbitrariness. However, this
only means that one has to rely on observing who operates with what kind
of distinction.
This somewhat complicated interlude was necessary in order to make
clear that the differentiation between external and internal observation gives
rise to communication barriers that cannot subsequently be dissolved. This
is especially true when both sides internalize the distinctions and then are in
a position to observe the difference between the two sides, the two systems,
the two perspectives, etc. Since every observation qua operation needs to be
based in the system, the separation can never be “sublated.”15 The advan-
tages of this separation can be exploited only if one accepts the separation.
Even if the systems communicate intensely, this does not change anything,
since the difference is and remains presupposed in the communication. A
new system arises out of communication: the contact system of the consult-
ing relationship. Now, this system can also be observed with regard to its
own states, its own developments, and its own selections. Communication
does complicate the system/environment relations; however, it does not
change anything about the fact that everything said is said by an observer,16
and that every determination must be determined by a system relying on a
distinction.17
V
It might not be a good idea to ask believers or even theologians whether it
makes sense to distinguish between God and man. They will answer with the
distinction “believing/unbelieving” – that is, a distinction that permits them
to designate their system. Entrepreneurs who are asked why they distinguish
between profit and loss will, in all likelihood, refer the questioner to the psy-
chiatric care unit of their company. Committed environmentalists would
gasp for air if one contested the usefulness of categorizing their information
by means of the distinction “protecting the environment/destroying the envi-
ronment”. Systems theorists may feel much the same way if they are
15 In light of this concept of “sublation,” let us only remark that Hegel made such an attempt
in his Logic and marked it with the concept “Spirit”. However, ever since then nobody has
tried again.
16 This is actually a quotation! See Maturana (1975), p. 324.
17 Evidence for the increased acceptance of this view in organizational science can be found in
Probst (1987), esp. pp. 68 ff. and passim.
360
19 Kap 16. Luhmann 05-11-11 16.29 Sida 361
prevented from conducting their analyses by means of the distinction

“system/environment”. After all, in all these cases, we might say: Let’s do it
differently; however, this might at the same time mean: From now on, this
is no longer a concern of mine.
In everyday communication, things and events, states and changes, are
talked about. The designation of that, about which one talks, is possible
only by means of distinctions. For the most part, these distinctions remain
merely implicit; they are a side effect of the designation of something as
“this” (and nothing else). Language is fully adjusted to this mode of com-
munication. Among other things, language is able to designate distinctions
(as in the example above); but even then, the context of distinctions within
which the distinctions in question are thematized, and from which they are
distinguished in turn, remains merely implicit. Let’s test this assumption. If
one poses the question of the distinction between God and man to theolo-
gians, it is not clear from which distinction this particular distinction is dis-
tinguished: from the distinction between God and the Devil? Or God and
creation? Or God and Christ? Or perhaps even man and animal? Or man
and system? – etc. If the question is directed towards the distinction that is
fundamental in each case, the burdens of communication increase, but the
problem does not disappear. In the final analysis, operational positivity – the
autopoiesis of communication – remains preponderant. It does what it does,
always by means of particular distinctions.
It can be very useful for the practice of management consulting to return
to the question of the distinctions by means of which communication mana-
ges to designate something: that is, the means by which communication
manages to choose and determine specific topics. The consulting group may
also develop theory preferences for certain distinctions with which it works:
for instance manifest/latent, structure/function, system/environment, perpe-
trator/victim (Girard), first-order observer/second-order observer, etc. As
soon as one’s own distinctions are distinguished from those of the client sys-
tem (formulating the problem in this way is itself a distinction of distinc-
tions), the presupposition of a common world vanishes. By means of this
distinction of distinctions one becomes aware of the fact that criteria, as well
as learning processes, are different in each case; whenever they are put into
action, they depend on a certain schema and they are used in relation to a
specific distinction. Now, here, a choice must be made (once again by means
of a distinction): namely, whether one prefers to reconstitute a common
world or to insist on the differences and continue working with incongruent
perspectives. In the former case, communication becomes difficult and even-
tually amounts to a refined illusion. The latter possibility could be called a
refined incongruence.18
18It hardly needs to be said again that in this case, as in previous cases, a problem is decom-
posed by means of a distinction.
361
19 Kap 16. Luhmann 05-11-11 16.29 Sida 362
This analysis goes far beyond all theories that were invented centuries ago
using such concepts as simulation and dissimulation. It barely has anything
in common with psychoanalysis, and even less with the therapeutic methods
and goals of psychoanalysis. However, it hints at the possibility that the
logic of distinctions is linked to the process of differentiation. In order to dif-
ferentiate themselves, systems use blind spots specific to themselves; that is,
they use distinctions with which they identify blindly. This is not likely or
even possible in every society. However, modern (or, as some would prefer
to say in reference to this issue, postmodern) society is expansive enough to
make this form of differentiation possible; in fact, one might even expect
such differentiation. Indeed, one counts on communicatively unbridgeable
differences, on a pluralism of “values”, “ideologies”, or “discourses”, to use
a recent term.
A study of management consulting can show that the problem of differ-
entiation is not at all limited to intellectual and emotional mega-concepts
and, furthermore, that it is not merely a problem of society in its entirety, of
political goals, social movements, and the organization of excitement and
alarm. On the contrary, it also makes everyday distinctions possible in areas
where a common world-view and common goals have hitherto been
assumed as given or at least as intended.
The more theoretical and conceptually precise such an analysis is, the
more likely is its empirical verification. Evidently, management consulting
moves from strictly microeconomic goals and a strictly microeconomic
analysis (which merely attempt to copy the world of the entrepreneurs and
seek to improve their position) to a mode of observation and description
that, while not giving up this orientation, attempts to reconstruct unity
through difference. In the relationship between consultants and company
this leads to communication barriers – to incommunicabilities that arise not
only on the tactical (i.e., simulation!) but also on the structural level. To put
it differently, the consultants’ imaginations become dependent on distinc-
tions that they can, by themselves, neither distinguish nor communicate any
longer; instead, they must use them operationally (the alternative, therefore,
cannot be yet another distinction; it must be the cessation of further dis-
tinctions). This move does not take place at the cost of communication. It
does not lead to a situation that would no longer allow the communication
of all that has been communicable hitherto. A more likely conjecture might
state that an increase in communication, an intensification of system-to-
system relations above a certain threshold, is possible, but only by means of
a single difference; that is to say, it is possible as the increase of communi-
cation by means of incommunicabilities. In fact, the experience of this phe-
nomenon is nothing new. It had already been noticed in the seventeenth
century in matters of love and religious conduct. In both cases, the increase
of (socially reflected) communication and the issue of sincerity were central.
362
19 Kap 16. Luhmann 05-11-11 16.29 Sida 363
VI
Finally, the critique of the model of applied science is also of importance for
the following questions: How are the consultants observed from the view-
point of the client company? How are they selected? And, one might almost
say, what kind of consulting do they receive? If we stick with the classical
model, the only question of concern is how well consultants know that
which they can know. This question presupposes a reality that is independ-
ent of observation. However, this condition is by no means a given, as even
the theory of science now admits.19 The description of reality depends on the
observer. This leads to the question: Which observer do we choose if we
want to know what is going on (undoubtedly, most often we will choose
ourselves, of course)? In the best-case scenario, we might be able to neutral-
ize the differences between observers by using an “injunctive” logic that for-
mulates operational injunctions (that is, an operational calculus) that lead to
the same result for everyone, if they are heeded.20 However, as soon as it is
no longer merely a matter of formally (logically) processing distinctions
(forms), but rather of processing a content-related choice of specific distinc-
tions, the difference between the observers – and that means the difference
between the consultants – becomes relevant. They can be distinguished from
each other according to which distinction each one of them employs as that
distinction which is not put into question. In other words, they are observed
in regard to that which cannot be observed by them. They are selected
because of their specific blind spots.
Of course, normally, such a degree of circumspection and lucidity can
hardly be assumed. Instead, names provide some orientation: personal
names, names of companies, names of theories. Reputation is condensed
into names. Differences in reputation, as well as the history of already exist-
ing contacts, direct the selection. However, if this is indeed the case, then it
is not possible for the company in search of management consulting to fully
rationalize the choice of consultants and subsequently to fully rationalize
how each should be treated. Indeed, for these purposes, communication bar-
riers would once again be indispensable. It would be necessary that the form
of the observation (that is, the distinctions by means of which consultants
construct that which is reality for them) could be chosen. Observers can
neither see themselves nor the distinction they employ, because it guides
them invisibly, very much as if it were a particular perspective. This simply
means that consultants are chosen and subsequently observed in regard to
that which they cannot see. Furthermore, communication about this blind
spot is not possible.
19 See von Foerster (1981). On the application to the relation between knowledge and inter-
vention see also Sgritta (1988).
20 Spencer Brown (1971). On the application to problems of therapy, cf. Simon (1988).
363
19 Kap 16. Luhmann 05-11-11 16.29 Sida 364
The problem is not, first and foremost, a problem of deliberate suppres-

sion. Neither is it primarily due to the impossibility in practical communi-
cation of saying everything all at once, or the subsequent problem, which is
that the sequence of communication does not allow the pursuit of every hint
of a potential meaning. The real problem lies in the difference between the
levels of first-order and second-order observation: that is, in the difference
between the observation of things, events, objects, and the observation of
observers. Strategy and the restriction of communication for practical rea-
sons can be used in order to preserve this difference. But even when com-
munication is not controlled in such a way (such control could in any case
happen only in a very limited fashion and only by means of psychic systems),
one recognizes on the level of second-order observation that it is impossible
to integrate everything that can be seen from this level into the first-order
level. If there is a second-order observation at all, goals such as “enlighten-
ment” or “the application of knowledge” have no bearing any longer; the
autological component of this second-order level of observation forces this
insight upon us. Consequently, “knowing better”, gestures of superiority,
leadership pretensions, and the hierarchization of the difference between
levels must be avoided.21 In the absence of such things, we gain the chance
of letting difference qua difference become irritating, stimulating, and even-
tually productive. And this, to be sure, would be true even if the consultant
were able to convince the client company to institutionalize such a difference
of first- and second-order levels within the company itself – namely, as the
difference of levels of communication about communication about commu-
nication.
– Translated by Peter Gilgen
Acknowledgements
This text was originally published in German in Luhmann, N. and Fuchs, P.
(1989) Reden und Schweigen (Frankfurt am Main: Suhrkamp), 209–227.
21 Laudan’s study (1984) is an attempt to use this same hierarchical interpretation of the dif-
ferences between levels of communication in order to explain how, in science, highly probable
dissension is turned into consensus. But Laudan’s theory rests on a purposive concept of action
and would fail if the conceptualization of the difference between levels were detached from
this base.
364
20 Kap 17. Vos 05-11-11 16.30 Sida 365
Chapter 17
Strategic Management
from a Systems-Theoretical
Perspective
Jan-Peter Vos
Introduction
This is certainly not the first study that has questioned the reasoning behind
the strategic management approaches that have gained attention in the past.
Several authors have criticised strategic management approaches for being
overly rational (e.g. Daft and Weick 1984; Mintzberg and Waters 1985;
Weick 1987; Pettigrew 1988; Knights and Morgan 1991; Whittington 1993;
Rajagopalan and Spreitzer 1996; Barry and Elmes 1997; Calori 1998).
Others have questioned the lack of evidence regarding their assumptions by
emphasising that the environment cannot be observed independently of the
organizations with which it interacts (e.g. Child 1972; Weick 1987; Knights
1992; Stacey 2000; Baecker 2003). Each of these scholars, in his own words,
has stated that organizations need to interpret themselves rather than their
environment. Within organization studies, Karl Weick extended this line of
reasoning the most by suggesting that strategic sensemaking is self-
referential (Weick 1995, p. 23). Until now, unfortunately, the insight that the
environment can only be observed internally has not received much atten-
tion within mainstream organization studies. To stress this point further:
strategy researchers prefer to obscure rather than acknowledge the self-
referential circularity between organizations and their environments.
Extending Weick’s argument that strategic management is self-referential,
this chapter aims to offer a better understanding of the role played by self-
reference in strategic sensemaking. However, acknowledging that strategic
management appears to be self-referential is one thing, making use of it on
a theoretical level is another, and this should not be underestimated. An
approach ideally suited to this is the theory of social systems, as developed
by Niklas Luhmann. After all, Luhmann’s theory is explicitly focussed on
the self-referential constitution of social systems (e.g. Luhmann 1984b;
1995f, and 1997a).
365
20 Kap 17. Vos 05-11-11 16.30 Sida 366
This chapter is structured into five sections. Before we embark on a jour-

ney into the theoretical labyrinth of social systems theory, we will show how
the traditional approaches to strategic management obscure the role of self-
reference by making either the environment or the capabilities of organiza-
tions the point of reference for the definition of successful strategies (first
section). After that, the shortcomings of the common denominator of these
either/or approaches, i.e. the paradigm of adaptation, will be discussed (sec-
ond section). Subsequently, social systems theory is used to shed new light
on the relationship between organizations and their environments (third sec-
tion). These insights are then used to outline an approach to strategy
research that solves the circularity problem by treating strategic manage-
ment as self-referential (fourth section). Finally, some conclusions are drawn
with respect to the theoretical contributions of a systems-theoretical per-
spective on strategic management.
The fallacy of either/or approaches to strategy

In the strategy literature, the self-referential circularity between organiza-
tions and their environment in defining strategies is obscured by making
either the environment (“outside-in”) or the capabilities of organizations
(“inside-out”) the point of reference for a definition of successful strategies.
By means of the “outside-in” approach of Porter (1985) and the “inside-
out” approach of Prahalad and Hamel (1990), it will be illustrated that such
either/or approaches to strategy are self-defeating.
Porter (1985), for instance, states that creating and sustaining a competi-
tive advantage means reckoning with competitive forces within a sector of
industry in order to distinguish oneself from competitors. The competitive
forces determine the rules of the game concerning business in a sector of
industry. According to Porter, organizations are wise to the extent that they
obey these strategic rules. In Porter’s approach to strategic management,
organizations should not change the strategic rules, as this would lead to a
stuck-in-the-middle position. The only way to become distinctive is to either
adopt a “cost leadership” strategy or a “strategy of differentiation”. Since
the strategic rules within a sector of industry are supposed to be objective,
all competitors can observe the same strategic rules and accordingly choose
their strategy, which is supposed to make them distinctive. Ironically, this
results in a situation in which strategy no longer means doing things differ-
ently, but doing things the same way as your competitors. After all, if all
organizations adopt either a strategy of “cost leadership” or “differen-
tiation”, the only way to distinguish yourself from your competitors is to
adhere to a “stuck-in-the-middle” strategy, which should be avoided at all
costs according to Porter. As a consequence, this “outside-in” strategy is
366
20 Kap 17. Vos 05-11-11 16.30 Sida 367
undermined by the paradox that organizations become indistinct by trying

to become distinct from their competitors.
The most popular strategic management movement of the 1990s is not
free of self-defeating reasoning either. In recent years, the two most used
buzzwords in strategic-management literature were “core competence” and
“core capability”. According to Prahalad and Hamel (1990), the founding
fathers of this concept, who heavily disputed the “competitive advantage”
concept of Porter because it only led to short-term strategic success, core
competences offer organizations a sustainable competitive advantage in the
long run. To sustain their development, organizations should use their core
competences in diverse and independent markets. Thus, in contrast to the
“outside-in” approach it is the internal competences that are to determine
how the organization acts on the external markets (“inside-out”). However,
in order to identify its core competences, the company ultimately has to take
the perspective of the customers; it is the customer who determines which
competencies lead to distinctive competitive advantages. In their book
Competing for the Future Hamel and Prahalad (1994) state that the ques-
tion of which capabilities of an organization should be considered core com-
petences, eventually boils down to the question of which capabilities, in the
view of customers, lead to distinctive competitive advantages. This implies
that short-term strategic success is a key issue for the positive development
of sustainable long-term competitive advantages. Ironically, it also implies
that organizations can only start to define strategies “inside-out” by observ-
ing themselves from the outside. As a consequence, this “inside-out” strate-
gy is undermined by the paradox that organizations become dependent on
customers by trying to become independent of them.
The fact that paradoxes can be brought to light within the two most suc-
cessful (!) strategic-management approaches of the past two decades should
not be interpreted as a shortcoming of these approaches per se. As will be
illustrated later, paradoxes appear to be omnipresent in organizations due to
the self-reference involved in dealing with their environment. Strategic
management approaches can only be criticised for denying their paradoxical
foundation. This denial has forced both the “outside-in” approach of Porter
and the “inside-out” approach of Prahalad and Hamel to treat either the
environment or the capabilities of organizations as the ultimate point of ref-
erence for explaining strategic success. As we have seen, this leads to self-
defeating reasoning, which makes the approaches practically irrelevant. As
a result, either/or approaches to strategy fail to grasp the specifics of the way
in which members of organizations deal with strategy. In the next section an
attempt is made to offer an explanation for the perseverance of the either/or
approaches to strategic management.
367
20 Kap 17. Vos 05-11-11 16.30 Sida 368
The paradigm of adaptation

The problem with either/or approaches to strategy can be traced back to the
paradigm underlying strategic thinking in general (see Mintzberg 1990 for
an overview): the paradigm of adaptation. Igor Ansoff, the founding father
of modern strategic thinking, wrote: “[The paradigm] concerns the logic
which guides the process by which an organization adapts to its external
environment” (Ansoff 1987, p. 501; italics added). In order to reveal the
problem with this logic, we will first examine the problem of strategic
management, as defined within the paradigm of adaptation, and how
researchers try to solve it within this paradigm. After that, we focus on the
paradox that is obscured by the paradigm of adaptation, making it impos-
sible for organizations to use it.
Within the paradigm of adaptation, strategic management is related to the
problem of defining successful strategies for dealing with an environment
that is ever changing. In the Strategic Management Journal – the non plus
ultra outlet for strategy research – studies attempting to find such successful
measures abound. The hidden assumption within these studies is that
organizations trying to observe their environment can conceptualise it as
something that exists independently of their own existence. In other words,
the environment is out there for all to see. Following this line of reasoning,
the solution to the adaptation problem is rather straightforward: observe
what is going on and take measures accordingly. From a systems-theoretical
point of view, this solution is conceptually based on “The Law of Requisite
Variety” or “Ashby’s Law”1. This law states that in order to be in control,
a system needs at least as many control measures as there are external vari-
ations (Ashby 1956, pp. 206–207). This “requisite variety” ultimately
implies that organizations need to establish a point-to-point correspondence
to their environment (Luhmann 1995f, p. 25). However, since the environ-
ment is much more complex than any organization can be, such a point-to-
point correspondence is impossible (Schreyögg and Steinmann 1987, p. 94;
Luhmann 1995f, p. 24). As a consequence, an organization can never cap-
ture the full complexity of its environment; every observation of the envi-
ronment is ultimately just a simplified construction. This leads to the crucial
question: what are companies ultimately supposed to adapt to? It cannot be
the environment as such because of its incomprehensibility, and it cannot be
the organizational construction of the environment because then organiza-
tions would just adapt to themselves. Thus, by obeying the “Law of
Requisite Variety” one stumbles upon indecision: we cannot decide whether
to adapt to the external environment or not. This is paradoxical: while
adapting to the external environment one cannot decide whether one is actu-
1To the credit of Ashby it should be mentioned that he never intended cybernetics in general
and his law in particular to be of empirical relevance (see Ashby 1956, pp. 2–3).
368
20 Kap 17. Vos 05-11-11 16.30 Sida 369
ally adapting to it or not. This paradoxical indecision is obscured by the

“Law of Requisite Variety” – and this is a crucial point – because the obser-
vation of the environment is not regarded as being problematic.
Notwithstanding this basic assumption, strategic management appears to be
obstructed by self-reference: the outside of an organization appears to be an
internal projection. The either/or approaches presented in the previous sec-
tion fail to acknowledge the self-reference involved in strategic management:
Porter states that the strategic rules of an industry sector can objectively be
determined, and Prahalad and Hamel state that organizations can objective-
ly decide upon their core competences by observing their customers. As
such, both approaches find their conceptual base in adaptation to the envi-
ronment.
The inability of strategy researchers to comprehend the self-referential
foundation of strategic management can probably be best illustrated by
means of a recent discussion with respect to the dynamic capabilities view
(Teece and Pisano 1994; Teece et al. 1997). This dynamic capabilities view
is the scientific counterpart of the management-oriented, core-competence
approach. Priem and Butler (2001, p. 28) have argued that the dynamic
capabilities view is undermined by the tautology that “competitive advan-
tage is defined in terms of value and rarity, and the resource characteristics
argued to lead to competitive advantage are value and rarity”. Barney
(2001, pp. 41–42), and Eisenhardt and Martin (2000, p. 1116) reacted by
stating that all strategic management theories are tautological in this sense.
According to these scholars, the question should not be whether theories can
be restated in a way that makes them tautological, but whether theories are
able to generate hypotheses that can be tested empirically. From this reac-
tion it is clear that the authors acknowledge that they, as organization
researchers, construct their object of investigation but at the same time have
missed the crucial point, i.e. that the hidden tautology is exactly the tautol-
ogy organizations face while deciding upon their valuable and scarce
resources. In other words, the authors fail to see that self-reference con-
strains strategic decision making. Perhaps we need to embrace the notion of
self-reference rather than obscure it, as the paradigm of adaptation does.
Such a stance is exactly what social systems theory offers us, as will be
described next.
Self-reference and social systems theory

To illustrate the implications of the turn towards self-reference, we will first
present a new paradigmatic stance towards adaptation. This stance will then
be used to describe how organizations can deal with the impossibility of
directly observing their environment. Then, we shall describe various ways
in which organizations need to deal with self-reference in order to maintain
369
20 Kap 17. Vos 05-11-11 16.30 Sida 370
themselves throughout time. Lastly, a methodology is presented for studying

this self-maintenance empirically.
The system/environment-distinction
Embracing the notion that organizations can be observed as self-referential
systems sheds new light on the relationship between social systems and their
environment (Luhmann 1995f, pp. 181–185). Within the theory of self-
referential systems, each system has its own environment. This is a different
conception of the system/environment distinction from that of open systems
theory (which the paradigm of adaptation is based on), where the environ-
ment is conceptualised as including the system (Figure 1).
The main implication of considering system and environment as exclusive
of each other is that the environment is specific to a self-referential system.
As such, the totality of all existence that is conceptualised within open sys-
tems theory as “the” environment is something that has no meaning within
the theory of self-referential systems. According to this theory, systems are
conceptualised as the unity of the distinction between system and environ-
ment (Luhmann, 2006). This unity is regarded as the world (“Welt”) of the
system and relates to the ultimate form of complexity that social systems
need to deal with in their existence (Luhmann 1995f, p. 208). The concept
of world does not refer to an all-embracing ontological concept of an objec-
tive social reality (Luhmann 1995f, p. 208). Rather, world relates to the
totality of existence that is specific to a single system, because a social sys-
tem cannot observe its world independently of itself due to the self-reference
involved (Luhmann 1988b, p. 42). To put it differently, a social system is
part of the reality it observes. As a consequence, a system cannot adapt to
its environment as something independently given.
Environment “Welt”
Environment System
System
Open Systems Theory Self-Referential Systems Theory
Figure 1: The system/environment-distinction in various disguises.
370
20 Kap 17. Vos 05-11-11 16.30 Sida 371
The paradigm of self-adaptation

Luhmann accepts the paradox that adaptation towards the environment is
only possible as self-adaptation. For social systems theory, the paradigm of
adaptation should therefore be substituted by a paradigm of self-adaptation.
According to Luhmann, the move to self-reference implies that the concept
of adaptation recedes to a secondary rank, without losing its significance
(Luhmann 1995f, p. 349). The paradigm of self-adaptation contains the
paradigm of environmental adaptation in the sense that a self-referential sys-
tem can naïvely decide to treat its environment as independent of itself. This
implies that the “Law of Requisite Variety” is only wrong from the per-
spective of an observer of a social system and not necessarily “wrong” from
the perspective of the social system itself. It is only a contingent, i.e. a neither
necessary nor impossible, way of reducing complexity. For a self-referential
system, the primary question in dealing with environmental change relates
to the way in which the system distinguishes itself from its environment
(Luhmann 1995f, p. 349). The paradigm of self-adaptation has a serious
implication for the way the interaction between organizations and their
environment should be conceived. All change, whether an adaptation to the
environment or not, is regarded as self-change despite the fact that the envi-
ronment remains a stimulus to change (Luhmann 1995f, p. 350). Because of
the paradigm of self-adaptation, Ashby’s “Law of Requisite Variety” should
be reformulated as the “Law of Requisite Reflexivity” (Vos 2002, p. 50): in
order to stay in control, an organization needs to be able to deal with its
inabilities by means of self-observation. This law states that social systems
should be able to develop new ways of observing their environment and
themselves, depending on the situation at hand.
Dealing with self-reference

The reflection of a social system upon its own existence can be described as
self-observation. During such self-observations the system/environment
distinction reappears within itself. There is a “re-entry” (Spencer-Brown
1972; Luhmann 1994b) of the system/environment distinction into the sys-
tem. This re-entry implies that the system/environment distinction is used to
observe what is observed by means of the system/environment distinction,
i.e. a distinction is used to observe the same distinction. Self-observation is
therefore clearly self-referential, and as such paradoxical, because the same
distinction is used twice to observe something different. That is, the distinc-
tion is both the same and different at the same time (Luhmann 1990c, p.
379). Due to the self-reference involved, a social system cannot observe
what it is observing and how it is observing at the same time. That is to say,
during self-observations one cannot simultaneously observe the self and the
act of self-observation. This temporal problem indicates the paradox of self-
371
20 Kap 17. Vos 05-11-11 16.30 Sida 372
observation: while observing itself, a social system does not observe itself. In
other words, during self-observation the “self” is its own “blind spot”.
Broadly speaking, the self-reference involved in self-observation is displayed
in the form of tautological and paradoxical reasoning (Luhmann 1990h): a
system is respectively what it is, i.e. a self-observing system observing itself,
or it is what it is not, i.e. a self-observing system that is not observing itself.
The fact that self-observation becomes trapped in tautological and para-
doxical reasoning, indicates that there is a limit to the knowledge social sys-
tems can obtain about themselves and their environment: social systems can-
not get to the bottom of their existence. Note that this “unbearable lightness
of being” that is characteristic of social systems only comes to the fore
because of the act of self-observation: if the system had refrained from self-
observations, it would never have been tempted to describe its existence as
an existential problem. It is therefore quite ironic that, during self-
observations, a social system stumbles upon a problem that it has already
solved, i.e. its existence (see also Czarniawska in this volume).
The fact that social systems fail to see through their existence does not
imply that self-knowledge is impossible. It merely indicates that self-
observation is a highly contingent affair, in the sense that the identity of a
social system is something that appears to be entirely dependent on the way
the system identifies itself with respect to what it is and what it is not (i.e.
its environment). While identifying its identity, a social system becomes
trapped within the chicken-and-egg problem that the system is what its envi-
ronment is not, and the environment is what the system is not (Figure 2).
Making sense of this chicken-and-egg problem is similar to the problem of
the Baron of Münchhausen who tried to pull himself out of the swamp by
his own hair. While the system distinguishes itself from the environment it
finds out that its environment is an internal construction. As a consequence,
the system does not import information from the environment but self-
produces information about the environment. In the words of Von Foerster
(1981, p. 263): “The environment contains no information; the environment
is as it is”. As such, during self-observations, making sense of the environ-
ment involves “asymmetrising” the tautology that the environment is what
it is. Likewise, due to the self-reference involved in self-observation, the sys-
tem needs to “asymmetrise” itself in order to produce information about
itself.
The chicken-and-egg problem that undermines self-observation can be
solved when the system “de-tautologises” either itself or its environment. By
just doing it, by just carrying out a self-observation this “asymmetrisation”
takes place. That is, by carrying out a self-observation, a social system is
able to identify something about itself or its environment. Depending on the
starting point of the self-observation, either the tautology that the system is
what it is, or the tautology that the environment is what it is escapes its self-
372
20 Kap 17. Vos 05-11-11 16.30 Sida 373
The system
is what it is System
The system is what
the environment is not
Self-Observation
The system is what

the organization is not
The environment
Environment is what it is
Figure 2: The self-referential problem of self-observation.
referential closure, i.e. becomes “de-tautologised”, and offers a clue for

subsequent “informed” actions (Figure 3). In terms of strategic manage-
ment, the first solution depicted in Figure 3 relates to “outside-in” self-
observations and the second to “inside-out” self-observations. An “outside-
in” solution, for instance, could start with the observation that the environ-
ment is hostile or “hypercompetitive”. This would enable the social system
to learn something about itself, i.e. whether the system is offensive or defen-
sive.
Solution I
“Let us assume the
environment is …” Environment Organization “… then we are …”
t=1 t=2
Solution II
“Let us assume “… then our
that we are …” Organization Environment environment is …”
t=1 t=2
Time
Figure 3: Solving the self-referential problem.
Dealing with self-reference thus requires acting naïvely in the sense of acting
first and thinking later. Only by being naïve is a social system able to make
373
20 Kap 17. Vos 05-11-11 16.30 Sida 374
itself existent or to “temporalise” complexity and in the meantime create its

world (Luhmann 1995f, p. 33; pp. 46–47). As we have seen, this world is
“imperfect” because it is impossible to grasp its unity from within: “it is the
world after the fall from grace” (Luhmann 1995f, p. 208). The “original
sin” of self-referential systems, so to speak, relates to the paradoxical oper-
ation of them ignoring their ignorance.
Self-reference and self-reproduction

One may wonder at this point what exactly the benefits of conceptualising
strategic management as a self-referential phenomenon are. After all, first
the validity of “outside-in” and “inside-out” approaches to strategy is ques-
tioned and subsequently these two approaches prove to be the only valid
solutions possible in dealing with self-reference. The point is that by
focussing on the self-referential aspects of strategic management, it is possi-
ble to observe both approaches as two functional equivalent solutions to the
adaptation problem. By this manoeuvre it becomes possible to explore
strategic management in a descriptive manner rather than in the prescriptive
manner of mainstream strategy research. Consequently, the way organiza-
tions deal with strategy and the consequences of their strategies can be
described in more complex ways. As will be explained next, this description
can take place on the level of operations, processes and systems.
Within social systems theory, three types of self-reference can be distin-
guished that are linked to these three levels of systemic aggregation
(Luhmann 1995f, pp. 443–444). According to Luhmann, the elements of
social systems are communications (see Chapter 1 in the present volume).
On the level of operations, self-reference relates to basal self-reference,
which involves the recursive relation between communications as such. This
recursion relates to the self-reproduction or autopoiesis of communications
based upon communications. Take, for instance, a discussion between two
managers about the undesired consequences of a decision they made in the
past. This discussion enables the managers to decide upon measures to
counteract the undesired consequences. The operational self-reference in this
example relates to the recursive relationship between the past and the pres-
ent decision; in other words, the past decision has “produced” the present
decision.
The second type of self-reference is called reflexivity and is accounted for
on the level of processes. In a process, communication takes place in the
light of an expected reaction to it. If, for instance, a manager informs the
other members of the management team about a serious sales problem,
which he is unable to solve, the manager expects that the other managers
will help him to come up with possible solutions. Reflexivity happens when
during a communication process the communication refers to the communi-
cation process as such: for example, a member of the management team
374
20 Kap 17. Vos 05-11-11 16.30 Sida 375
observes that the management team is not communicating very effectively

about the sales problem. The operations of organizations become structured
throughout time because of processual expectations, even to the extent that
expectations may determine what organizations are able to communicate
about. As such, the reflexive expectations both enable and constrain the
operations of organizations.
Finally, on the systemic level self-reference is referred to as reflection.
Reflection involves communication about the system/environment distinc-
tion. As such, it is an operation by which an organization identifies itself in
contrast to its environment. As we have seen, this self-observation relates to
the re-entry of the system/environment distinction into the system. Through
reflections made, organizations make sense of what their environment has
to expect from them, e.g. the expectation that shareholders do not want to
be surprised when it comes to company results, which may lead manage-
ment to plan shareholder meetings frequently to render the company more
credible as a source of investment. Naturally, these expectations may differ
from the “real” expectations of those in the environment.
Strictly speaking, organizations do not need any form of self-referential
reflection to reproduce themselves, i.e. they “only” need to continue oper-
ating. Processual reflexivity and systemic reflection, however, are necessary
operations for coping with organizational and environmental change that
may obstruct the self-reproduction of organizations. For instance, by ignor-
ing persistent sales losses, a company may put its continuation at risk.
Scientific observation of social systems

The starting point of social systems theory is a problem; that is, the ultimate
complexity brought forth by the distinction between system and environ-
ment. Social systems need to reduce this complexity in order to be able to
reproduce themselves. Luhmann’s functional method can be seen as a means
of analysing the way in which social systems solve this problem, and of iden-
tifying functionally equivalent solutions to it. For example, if inflation solves
problems of distribution in a relatively conflict-free way (with whatever side
effects), inflation is a functional equivalent to national planning, which is
politically riskier, because it is richer in conflict (Luhmann 1995f, p. 54).
The goal of functional analysis is to compare such functional equivalents to
find out their risks or dysfunctionalities. As such, the achievement of the
functional method resides in the broadening and limiting of what is possible
by opening up better and, above all, more complex possibilities of analysis
and comparison (Luhmann 1995f, pp. 54–55). Accordingly, the theoretical
achievement provided by the functional method relates to the construction
of problems. Functionally equivalent solutions to problems are of theoreti-
cal relevance because they offer a means of describing how and explaining
375
20 Kap 17. Vos 05-11-11 16.30 Sida 376
why social systems evolve as they do. To facilitate this, Luhmann uses the
distinction between first and second-order observation, which he borrowed
from Heinz Von Foerster’s second-order cybernetics (1981).
First-order observation relates to the act of observation as such, e.g. the
observation of an organization’s competitors by the organization itself.
Likewise, second-order observation relates to the observation of other
observations; for example, a strategy consultant may observe how the
organization observes its competitors. We can use this distinction between
first- and second-order observation as a guide for our research on organiza-
tions: if one observes in the first-order mode, one puts oneself in the posi-
tion of the organization and tries to observe what it observes while observ-
ing. In contrast to that, the mode of second-order observation implies a crit-
ically distanced position towards the organizational observations. The
researcher observes the way in which the observational schemas of the
organization influence how it observes. The aim is to observe what social
systems cannot observe due to the specific way in which they observe; in
other words, the blind spots of social systems (Luhmann 1990h, p. 139).
The act of self-observation by organizations is, in fact, a second-order
observation of themselves. In order to prevent observational confusion, we
need to distinguish between the first- and second-order observations of
organizations by themselves, and by strategic-management researchers
(Table 1). The first-order observation by organizations relates to their strate-
gies as such; their second-order observations relate to their strategic self-
descriptions. The first-order observation by strategic-management
researchers of the first-order observations by organizations relates to the
description of the way in which organizations have “de-tautologised” the
chicken-and-egg problem of Figure 2 to become operational. The second-
order observation of this by strategic-management researchers involves the
description of the blind spots of the first-order observations by organiza-
tions. The first-order observation of the second-order observations by
organizations relates to the way in which organizations have “de-
paradoxised” themselves in order to observe or describe themselves. The
second-order scientific observation of this relates the description of the blind
spots of these self-descriptions of organizations.
376
20 Kap 17. Vos 05-11-11 16.30 Sida 377
It should be clear that the observation of the way in which organizations

deal with the self-referential problems of strategic management backfires on
the strategic management researchers. That is, to enable the scientific obser-
vation of observing organizations, strategic-management researchers need to
“de-tautologise” and “de-paradoxise” also themselves. Due to the self-
reference involved, a systems-theoretical perspective on strategic manage-
ment is of a radically constructivist nature. Only by acknowledging this are
we able to explain that the “outside-in” and “inside-out” approaches to
strategic management are constructed by scholars to observe the strategic
management of organizations, i.e. these approaches to strategic manage-
ment do not necessarily relate to the way in which organizations perceive
their own strategies. By linking the concepts of first- and second-order
observation and the self-reference involved with this it is possible to outline
a research approach to strategic management that unifies the existing
“outside-in” and “inside-out” doctrines of strategy. This will be done in the
next section.
Table 1: Self-reference and levels of observation.

First-order observations Second-order observa-
by organization tions by organization
researchers researchers
(i.e. to observe what (i.e. to observe what
organizations observe) organizations cannot
observe)
First-order observations Observing how organi- Observing the blind
by organizations (i.e. zations “de-tautolog- spots of operational
their strategies as such) ise” themselves in order organizations, i.e.
to operate strategically observing what organi-
zations cannot observe
due to they way they
perceive their world
Second-order observa- Observing how organi- Observing the blind
tion by organizations zations “de-paradoxise” spots of reflexive organi-
(i.e. their strategic self- themselves in order to zations, i.e. observing
descriptions) reflect upon themselves what organizations can-
strategically not reflect upon due to
way they reflect upon
their world
A self-referential framework for strategy research

At the start of this chapter, we indicated that the existing approaches to
strategic management obscure their self-referential grounding. In this sec-
tion, the insights of social systems theory will be used to outline a research
377
20 Kap 17. Vos 05-11-11 16.30 Sida 378
framework for strategic management, which acknowledges that neither the

environment nor the capabilities of the organization offer exclusively the
starting points for defining strategies. First, we shall explore how the func-
tional method may aid in examining the strategies of organizations on the
level of operations, processes and systems. In addition, the role of first- and
second-order observation in this respect will be made clear. Subsequently,
the developed framework will be related to existing strategy research to
illustrate its potential.
In the previous section it appeared that self-observation is a highly con-
tingent affair, in the sense that an organization is able to describe itself in a
multitude of ways. As such, each self-description is a reduction. Social sys-
tems theory uses the self-referential forms of tautology and paradox to char-
acterise theoretically these contingent reductions. The move to self-reference
implies that strategy research needs to be able to analyse both “outside-in”
and “inside-out” self-descriptions of organizations – as opposed to the exist-
ing either/or approaches. Each of these alternatives is a functionally equiva-
lent solution to the self-referential problem of adaptation to the environment
being only possible as self-adaptation.
When analysing the way in which organizations deal with this problem
from a scientific point of view, the focus is on the decisions made by
organizations to enable their self-reproduction (cf. Luhmann 2000c). The
focus on strategic decisions should not be interpreted as an indication that
decision making is considered to be subjectivist by nature, or the “essence”
of managerial work (cf. Simon 1960), but rather, that strategic decisions are
symbolically enacted upon (cf. Weick 1987) by means of communication
that transcends the level of individual members of the organization
(Luhmann 2000c)2. The functional method aids in exploring the way in
which organizations “de-tautologise” and “de-paradoxise” (Luhmann
1990h, p. 139) their strategies in order to develop theories that make a dif-
ference (cf. Weick 1999). As indicated above, this can be done on three levels
of aggregation. Note that these levels of aggregation correspond to the wide-
spread categorisation used in strategy research between the strategy content,
process and context (Pettigrew 1987; De Wit and Meyer 1994).
• On the level of operations, it is possible to analyse how organizations
decide upon the content of their strategies by enacting strategic concepts
(e.g. added values or core competences) that enable and constrain them to
formulate strategies (i.e. to asymmetrise themselves).
• On the level of processes, it is possible to analyse how organizations
decide upon the expected results of their strategies by enacting strategic
2 As such, the members of an organization remain in the environment of the organization,

which only means that the organization cannot determine the psyche of their individual mem-
bers and vice versa.
378
20 Kap 17. Vos 05-11-11 16.30 Sida 379
routines (e.g. management team meetings) that enable and constrain them
to reflect upon the expected and unexpected results of their strategies (i.e.
to structure themselves).
• On the systemic level, it is possible to analyse how organizations decide
upon their strategic context by enacting strategic roles (e.g. employer and
competitor) that enable and constrain them in the reflection on the dis-
tinction between themselves and their environment (i.e. to identify them-
selves).
The three types of self-reference (i.e. basal self-reference, reflexivity and

reflection) and corresponding levels of aggregation are not linked by
Luhmann to the methodological concepts of first- and second-order obser-
vation, at least not explicitly. Nonetheless, by doing this, some useful indi-
cations can be obtained for the kind of knowledge that systems theory and
the functional method can contribute to strategy research (Table 2).
Table 2: Functional analysis of strategic management.

First-order observations Second-order observations
Strategic Describing specific functional Comparing various functional
content equivalents to make sense of the equivalents to formulate the
content of strategies by observ- content of strategies and thus
ing the way strategic concepts find out their functionalities and
are enacted dysfunctionalities
process equivalents to make sense of the equivalents to structure the
results of strategies by observing strategic process and thus find
the way strategic routines are out their functionalities and
enacted dysfunctionalities
context equivalents to make sense of the equivalents to identify the
strategic context by observing strategic context and thus find
the way strategic roles are out their functionalities and
enacted dysfunctionalities
The functional analysis of strategic management by means of first-order

observation aims to explore the way in which members of organizations give
meaning to their organization’s strategic content, process and context. In
addition, by means of second-order observation, the functional analysis of
strategic management is aimed at observing how organizations may or may
not jeopardise their existence because of their strategic manoeuvres. The
theoretical aim of the framework is to describe how and to explain why
strategic management takes place in organizations as it does. By adopting
this stance, strategy researchers can analyse strategic management as being
both intentional and unintentional. In the remainder of this section, the
framework will be described in more detail.
379
20 Kap 17. Vos 05-11-11 16.30 Sida 380
Making sense of strategic content by means of

strategic concepts
Strategic content, as it is used here, does not refer to the planning of strate-
gies as such but to the concepts used within the process. This implies that
when we focus on the content of strategies by means of first-order observa-
tion, the use of management concepts or symbolic generalisations such as
“core business”, “core competence”, “added value” and “leverage” can be
uncovered. While the rise and fall of management concepts such as these is
a phenomenon worthy of further critical exploration (e.g. Johnson 1990;
Collins 2000; Ortmann and Salzman 2002), it cannot be denied that they
benefit organizational sensemaking processes (e.g. Duimering and Safayeni
1998).
Apparently, it is not important what strategic management concepts
mean, it is what organizations and strategists make them mean. Take, for
example the notion of “core competence”. Throughout their entire book,
Prahalad and Hamel (1994) remain vague about what core competences
actually are. Notwithstanding this, since the launch of the core competence
concept, several authors have made checklists that enable managers to deter-
mine whether their organization has core competences (e.g. Stalk, Evans and
Shulman 1992; Bartness and Cerny 1993). In management-oriented “guru”
literature, many strategic management concepts can be found that should
aid organizations to define strategies. While it is easy to question the scien-
tific validation of strategic concepts, it is less easy to question their validity.
After all, strategic concepts are used by managers, consultants and
researchers to highlight issues about organizational life that would remain
underexposed without them. Strategic concepts function as means by which
organizations “de-tautologise” and “de-paradoxise” themselves in order to
become operational. In other words, strategic concepts “asymmetrise”
chicken-and-egg problems: for instance, that the markets to be served
depend on the products offered, and the products to be offered depend on
the markets served. Theoretically, each strategic concept is regarded as con-
tingent, which implies that no strategic concept can claim superiority above
others, since that would contradict its lack of necessity. Therefore, strategy
gurus seem right and wrong at the same time. They are right in formulating
various ways of becoming competitive and wrong in their one-sided prefer-
ence for highlighting specific ways of becoming competitive.
The second-order observation of the content of strategies is aimed at com-
paring and evaluating functional equivalents, in the way organizational
members make sense of the content of their organization’s strategies by the
strategic concepts they enact or wish to enact. For instance, what are the dif-
ferences between the strategies of organizations that define strategies “out-
side-in”, and those of organizations that define them “inside-out”? The the-
oretical relevance of this perspective is to indicate the functions of strategic
380
20 Kap 17. Vos 05-11-11 16.30 Sida 381
concepts by defining problems in such a way that more complex ways to

analyse and compare the content of strategies arise as a result. This stance
includes the problematic of the strategy concept as such.
Strategy as a concept “only” dates back to the 1960s (Rumelt et al. 1994).
Knights and Morgan (1991) highlight an interesting paradox in this respect:
if the concept of strategy is considered to be so important for the survival of
organizations, how did they survive so long without having a concept of
strategy? The authors argue that the concept of strategy constitutes the
problems to which it claims to be a solution (Knights and Morgan 1991,
p. 255). In line with this, the second-order observation of strategic content
should focus on the problems that are solved by deciding upon strategic con-
cepts and the problems that result from these decisions. For instance, the
concept of strategy as such provides both organizations and strategy
researchers with a means of rationalising organizational successes and fail-
ures. The problem with this rational model is that it fails to describe the
specifics of the way in which strategic decision-making takes place (Mintz-
berg and Waters 1985). Strategies appear to have emergent qualities that
evolve from operations taking place throughout an organization (Quinn
1980). That is, strategies can be both intentional and unintentional and
strategic concepts function as a solution to the problem of how to rationalise
strategies a priori and a posteriori.
Theories on strategy content may explain the ability and inability of
organizations to redefine their strategy. Weick, for instance, stresses that
“loose coupling” between strategic decisions made throughout time may
well increase the potential of organizations to reproduce themselves (Weick
1977). That is, loose coupling increases an organization’s potential to
produce unexpected, surprising strategic operations (Luhmann 2000c). The
focus on such operational variations, with respect to the second-order
observation of the content of strategies, enables strategy researchers to
explain why the enduring enactment of strategic concepts such as the core-
competence concept may prevent deviating strategic sensemaking, which
results in the transformation of a “core competence” into “core rigidity”
(Leonard-Barton 1995).
Making sense of strategic process by means of

strategic routines
The strategy process relates to decisions concerning the use of strategy rou-
tines by organizations for structuring their expectations or aspirations. Just
like the strategy content, the strategy process has gained significant attention
in the past (e.g. Quinn 1980; Pettigrew 1992). Regarded first as a rational
process by the likes of Chandler (1962), Cyert and March (1963), Ansoff
(1965), and Hofer and Schendel (1978), the concept of emergent strategies
arose later on (e.g. Quinn 1978; Mintzberg and Waters 1985; Weick 1987;
381
20 Kap 17. Vos 05-11-11 16.30 Sida 382
Johnson 1988), and now strategies appear to have chaotic and self-
organizing qualities (e.g. Brown and Eisenhardt 1998; Stacey 2000). The
cited authors predominantly focus on the decision processes concerning
strategy. Others have adopted a postmodern perspective, focussing on
power and political issues that influence the strategy process (e.g. Knights
and Morgan 1991; Knights 1992; Barry and Elmes 1997).
The notion of the strategic process as it is used here combines both per-
spectives, and aims at discovering the expectations of expectations (struc-
tures) that keep the self-reproduction or autopoiesis of strategy formulation
going. The first-order observations aim at uncovering the routines that
structure the communication processes concerning the way strategies are
formulated, implemented, evaluated, etc. (cf. Hendry and Seidl 2003) and
that may lead to a “dominant strategic logic” (Prahalad and Bettis 1986;
and Bettis and Prahalad 1995) that enables and at the same time constrains
organizations to formulate strategies.
The second-order observation of the strategy process is aimed at compar-
ing functionally equivalent ways in which organizational members make
sense of the strategy process by means of the strategic routines they enact or
wish to enact. The theoretical relevance of this perspective is to indicate the
functions of strategic routines. Take, for example, the strategic importance
of a routine enacted by Naskapi Indians in Labrador to decide on the most
profitable direction for hunting (Weick 1977, p. 203). These Indians answer
the question by holding dried caribou shoulder bones over a fire. As the
bones become heated they develop cracks and smudges that are then “read”
by an expert. These cracks indicate the direction in which the hunters should
look for game. The Naskapi believe that this practice allows the gods to
intervene in their hunting directions. This bone-reading routine functions as
a means of complicating the decision process, in order to solve the problem
of fixed patterns of hunting activity that may result in the local extinction of
game. As such, the routine is a functional equivalent to a table of random
numbers (Weick 1977, p. 204).
Theories on the strategy process may explain the ability and inability of
organizations to live up to their own expectations by means of the way in
which they structure their self-reproduction. For example, by using an evo-
lutionary model on the strategy process, Burgelman (1996) found that Intel
failed to make a success of their DRAM business because of the various
departmental routines that led to the favouring of other businesses.
Making sense of strategic context by means of strategic roles

There is a vast amount of literature on the way in which the strategic con-
text influences organizations. Take for instance, contingency theory (Burns
and Stalker 1961), resource dependence theory (Pfeffer and Salancik 1978),
population ecology (Hannan and Freeman 1977), and institutional theory
382
20 Kap 17. Vos 05-11-11 16.30 Sida 383
(Selznick 1957). Unfortunately, these approaches leave little room for

organizations to decide on their own course of action. A more management-
oriented research area relates to “corporate governance”, or “stakeholder
theory”, which focusses upon the way in which organizations observe them-
selves in relation to their environment. As such, it is closely related to the
concept of strategic context as presented here. Since the publication of the
landmark book by Freeman (1984), the idea that organizations have stake-
holders has become commonplace in both organization studies (e.g.
Alkhafaji 1989; Brummer 1991; Donaldson and Preston 1995; Jones 1995;
Frooman 1999; Henriques and Sadorsky 1999; Jones and Wicks 1999; Scott
and Lane 2000; McWilliams and Siegel 2001; Tirole 2001) and management
literature (recently e.g. Cummings and Doh 2000, and Waddock and Smith
2000). According to Freeman, the stakeholder approach is about groups and
individuals who can affect an organization, and about managerial action
taken in response to those groups and individuals (Freeman 1984, p. 48).
The analysis of stakeholders concerns three questions: “Who are they?”,
“What do they want?” and “How are they going to try to get it?” (Frooman
1999, p. 191). The first-order observations of the strategic context of
organizations could focus on the way in which organizations self-referen-
tially deal with the expectations of stakeholders, and how these expectations
are used as a measure of corporate identity by organizations (Gioia and
Thomas 1996; Scott and Lane 2000). This conception of corporate identity
highlights the fact that organizations have to make sense of several distinct
and probably contradictory demands, depending on the stakeholders
thought to be of relevance. Thus, identity should be viewed as a multidi-
mensional incoherent construct.
The second-order observation of the strategy context is aimed at compar-
ing and evaluating functional equivalents, in the way organizational mem-
bers make sense of the strategy context by means of the strategic roles they
enact or wish to enact. The theoretical relevance of this perspective is to
indicate the functions of strategic roles by defining problems in such a way
that more complex ways to analyse and compare the strategic context arise
as a result. For instance, as Seidl (2003b; 2005) has pointed out, strategic
self-descriptions have an integrative and an operative function. The integra-
tive function of a self-description of the strategic context is to represent the
organization to the organization, which gives the organization a sense of
unity when comparing its different parts to each other. As such, the “out-
side-in” stakeholder approach and the “inside-out” mission statement are
functionally equivalent solutions to the problem of presenting the organiza-
tion to the organization. The operative function of self-descriptions is to give
direction to the decision making regarding the strategic roles to be enacted.
Theories on the strategy context may explain the ability and inability of
organizations to adapt to their environment by means of self-adaptation.
383
20 Kap 17. Vos 05-11-11 16.30 Sida 384
Gioia et al. (2000, p. 67), for instance, use the concept of adaptive instabil-
ity to explain how the discrepancy between the way in which Shell saw itself
and how others perceived the company caused the major controversy over
Shell’s plan to dispose of the Brent Spar offshore storage and loading plat-
form in the Atlantic.
Conclusions
The main argument of this contribution has been that existing approaches
to strategic management obscure the self-referential relationship between
organizations and their environment. By acknowledging this self-referential
relation, it becomes possible to unify existing strategic-management
approaches and to observe strategic management in more complex ways. We
started out by showing that existing approaches to strategic management are
self-defeating. It was argued that this was due to the paradigm of adapta-
tion, which is based upon the basic assumption that organizations can
observe their environment as something existing independently of them-
selves. Social systems theory offers a way out by conceptualising the relation
between organizations and their environment as a self-referential one; that
is to say, organizations can only perceive their environment as an internal
projection. Within this conception, strategic management means dealing
with the paradoxical problem that adaptation to the environment is only
possible as self-adaptation. It appeared that by being naïve, or ignorant of
their ignorance, organizations are able to escape their self-referential im-
prisonment. As such, ignorance is a conditio sine qua non of organizations
reproducing themselves. Subsequently, a framework was presented that
enables the conceptualisation of strategic management as a self-referential
phenomenon. With respect to mainstream strategy research, the application
of social systems theory, and the proposed turn to self-reference for strategy
research the following conclusions can be drawn.
It is surprising that in the book Fundamental Issues in Strategy: A
Research Agenda, edited by Rumelt, Schendel and Teece (1994), no contri-
butions could be found that addressed the self-referential problems con-
cerning the paradigm of adaptation in the same or in other terms. No refer-
ence whatsoever is made within the book that strategic management
involves making sense of self-referential phenomena such as tautology and
paradox when it comes to observing the environment. This leads us to the
conclusion that the distinction between organizations and their environ-
ment, as conceptualised within mainstream strategy research, is superseded.
After all, within general systems theory, ideas about self-referential observa-
tion date well back to the previous century.
Unfortunately, the current state of the theory of self-referential systems
does not give detailed guidelines for applying this theory to the empirical
384
20 Kap 17. Vos 05-11-11 16.30 Sida 385
exploration of the role of self-reference. The fact that so little empirical

research is carried out on the basis of self-referential systems theory is still a
serious shortcoming of this theory because, as they say, the proof of the pud-
ding is in the eating3. Following on from this observation, it seems rather
ironic that the theory of self-referential systems, with all its idiosyncratic
terms, is necessary to explain how social systems such as organizations make
themselves possible by means of ignorance. In spite of this observation, it
was found that the theory of self-referential systems is quite straightforward
in its application. Therefore, it is surprising that it is so hard to find detailed
instructions for the functional analysis of social phenomena such as organi-
zations. Hopefully, this contribution will lead to more discussion in this
respect.
Lastly, with respect to the application of the self-referential framework to
strategy research as presented here, it should be stressed that the assumption
that organizations need to deal with self-reference backfires on strategic
management research as a whole. If we take the self-referential closure of
both organizations and ourselves seriously, it seems that the only knowledge
we have to offer to practitioners takes the form of “ironic compassion” on
their attempts to make sense of their environment and themselves. After all,
now we know that also strategy research has its blind spots, a “sense of
modesty” seems appropriate towards both organizations and ourselves.
Based on the arguments presented in this chapter, the suggestion is to go
ahead with the aim to restrict ourselves to describing how organizations
make sense of the strategy phenomenon (cf. Nicolai 2000, pp. 301–302). By
doing this we can extend our “learned ignorance” (Cusa 1981) with respect
to strategic management. For now, however, we should ignore our ignorance
of how to observe the potential of the move to self-reference.
3 Note that due to the radical constructivist nature of social systems theory, empirical research
can never function as the ultimate test with respect to the truth of it. After all, empirical find-
ings are self-produced “facts” that do not speak for themselves but only to the system that has
produced them. Notwithstanding this observation, more empirical research, which has been
inspired by systems theory, may persuade more scholars to take social systems theory more
seriously.
385
21 Kap 18. Scheytt 05-11-11 16.31 Sida 386
Chapter 18
Management Accounting
from a Systems-Theoretical
Perspective1
Tobias Scheytt
Introduction
Within positive accounting theory, systems of management accounting and
control are mostly interpreted as instruments that are aimed at improving
business. From this perspective, such systems help to depict organizational
reality in a neutral and objective way. Systematically, the purpose of such
management accounting systems is recognised as a means of reducing com-
plexity in organizations to a level that can be handled. In relation to prac-
tice, they provide the basis of “proper” information for managerial decisions
and the resulting possibility for coordination of different managerial func-
tions (like human resource management, operations, finance, strategy mak-
ing). This forms the general basis for translating complex issues into
“rational” decisions – on the basis of “professional” methods for planning
and controlling (Belkaoui 1997; Emmanuel, Otley, and Merchant 1999;
Kaplan, Young, and Atkinson 2003).
This view is both unchallenged and unquestioned as are the presupposi-
tions it is based on. However, like all theories that focus on organizations
adopt a specific epistemological and ontological position (Burrell and
Morgan 1979), this positive accounting theory is also based on a specific
world-view. When analysing the epistemological and ontological back-
ground, it becomes obvious that approaches to management accounting of
that kind are implicitly rooted in an idealistic worldview. Positive or
“rational” accounting theories (“RATS”, as labelled by Jönsson and Mac-
intosh 1997) see the management of organizations as a matter of rational
choice and agency. From this point of view, it is consistent to argue that
management accounting systems have to provide representations that repli-
cate the organizational reality as exactly as possible. Critical accounting
1 I am grateful to Kai Helge Becker, David Seidl, and particularly Kim Soin for their helpful
comments on earlier versions of the chapter.
386
21 Kap 18. Scheytt 05-11-11 16.31 Sida 387
theory (“CATS” in the words of Jönsson and Macintosh 1997) is usually

held as being the direct opposite to rational accounting theory. Critical
approaches state that management accounting is far from being neutral and
objective but serves particular interests, has distorting effects, puts psychic
pressure on the workforce in organizations and is a tool that can be utilised
to support existing power structures. However, critical accounting theory
also systematically assumes that the effects caused by the use of management
accounting systems are clearly identifiable and can be consistently described.
Both the rational and the critical tradition therefore embody the ideal of a
strong correspondence between what is observed and the observation deliv-
ered by management accounting systems.
The following reflection is based on the argument that the ontological pre-
supposition common to rational and critical approaches leads to two short-
comings: first, it simplifies the interrelation between observation and the
observed. Second, it disregards the main function of management account-
ing systems namely, to construct a framework that fundamentally shapes
communication patterns within organizations. This chapter aims at explor-
ing an alternative viewpoint for the epistemological and conceptual under-
pinning of management accounting theory. It argues for a constructivist –
and more specifically: a social systems theory – approach in order to con-
ceptualise the function of management accounting. It outlines that manage-
ment accounting plays an important role in the formation of the self-
observations of organizations in that it specifically pre-defines decisions in
organizations. The chapter is structured as follows. First, a summary of the
ontological and epistemological positions of both, positive and critical
management accounting theory is given. For the positive strand, the chapter
draws on management accounting theory in the German-speaking area as
this tradition exemplifies a rather unquestioned position that management
accounting is a tool for representing the world “outside”. For the critical
strand, it refers to the Anglo-Saxon discussion with its profound and criti-
cal examination of the diverse effects that are caused by the application of
management accounting systems. Second, an alternative understanding of
the role management accounting plays in organizations is outlined. The
intricate interrelation of management accounting systems and communica-
tion is described with respect to social systems theory. Third, the function of
management accounting is discussed in relation to social systems theory’s
notion of decision. The self-referential character of management accounting
is emphasised as “organized” self-observation of an organization. Finally, a
brief conclusion is drawn which indicates the consequences from this frame-
work for the theorising of management accounting.
387
21 Kap 18. Scheytt 05-11-11 16.31 Sida 388
Management accounting and the objectification

of reality
For an outline of positive accounting theory it is useful to relate to the
management accounting discourse in the German-speaking area as it exem-
plifies in a clear manner the specific underpinnings of this theoretical
approach. The scientific discourse on management accounting and control
is – for specific historical reasons that cannot be explained here in detail –
different to the international discussion. In brief, it is one specific epistemo-
logical aspect that is of particular relevance to management accounting
research in the German-speaking area: management accounting theory usu-
ally has a normative orientation, in that the main aim of theory building is
to improve praxis by making recommendations on how to design processes
of controlling and managing business. In the German-speaking discussion,
this notion has led to a variety of contributions that strive for a positive
description of the role of management accounting techniques (Jacobs and
Haeracleous 2002): Management accounting tools have to support manage-
ment in grasping “what is really going on” in an organization and hence to
take control of business processes. In general, the basis for these conceptu-
alisations is a realistic ontology: accounting systems are seen as neutral tools
for “imaging” (not imagining!) an organization which neither distort an
organization’s reality nor intervene in the context in which they are applied.
In order to understand the assumptions underlying management account-
ing theory in the German-speaking area, one has to consider a further
important characteristic that stems from the broader discourse on business
administration. Since the early stages of its development, German business
administration theory is coined by a notion of an over-arching theoretical
framework, namely that of Betriebswirtschaftslehre (general business
administration theory) that encompasses several sub-disciplines (like mar-
keting, logistics, finance, leadership, etc.). Although management account-
ing practices are widespread in organizations’ reality, its status as an
autonomous sub-discipline within Betriebswirtschaftslehre is generally
doubted. As it is a relatively young part of general business administration
theory, the differentiation of management accounting from other sub-
disciplines, that were already “on the scene”, like for example, cost account-
ing or information systems management, is seen as problematic. In this
sense, advocates of other sub-disciplines regularly question what the inher-
ent theme and specific significance of management accounting theory could
be. As a consequence, the management accounting discourse is, since its
infancy, coined by a quest for the “core” of management accounting that
can be identified as a distinctive characteristic (Weber/Schaeffer 1999).
With respect to the claim for a specific scientific object, the mainstream
focused in the last decade primarily on the notion of coordination.
Coordination as the main function of management accounting denotes prac-
388
21 Kap 18. Scheytt 05-11-11 16.31 Sida 389
tices that enable or support goal-oriented interventions. As decisions about

the concrete design of processes and structures within organizations form
the main task of business management, it is the coordination of these
processes and structures that is the core task of management accounting
(Horváth 2003; Küpper 1997; Weber 2004). Management accounting theo-
ry, as the “science of coordination”, has therefore to provide solutions that
support management, since these coordination problems cannot be effi-
ciently handled by management itself, due to its diverse tasks and duties.
Weber (1992) distinguishes two different types of coordination applicable
in organizations: “coordination through personal communication by order
or self-organization” and “coordination through technocratic communica-
tion by programs or plans”. Personal communication, on the other hand, is
a suitable coordination mechanism where the complexity of the “con-
trolled” organization is rather low – e.g. in small and medium-sized enter-
prises. As for larger organizations, their complexity, their dynamic environ-
ment and the pressure to develop strategies and goals for the operational
systems enforces “coordination by plans” which is therefore in this view the
most efficient way of coordinating. This interpretation is based on an under-
standing of the functions of management that relates to the (first-order)
cybernetics of the “classical” management cycle. Management is seen as
being carried out within different sub-systems, that is, the planning system,
the information system, the control (namely, monitoring) system, the per-
sonnel management system, and the organizing system (Küpper 1987). The
coordination of these management sub-systems is identified as “secondary”
coordination and is the domain of management accounting practice where-
as the planning and coordinating of operational processes (“primary” coor-
dination) remains the domain of management practice (Weber 1992;
Horváth 2003). However, not only in terms of this structural differentiation,
but also from a processual perspective the coordination of management
activities is seen as necessary. The progress in different phases of the man-
agement process is understood as an object of coordination practices.
Management accounting, from this perspective, is to support the manage-
ment in the “cybernetic circle” of planning (definition of goals, formulation
of plans), implementation of plans (breaking down strategic plans into oper-
ational programs), realisation (without the possibility of any discretion on
behalf of the task performing actors), and control (with the result of a feed-
back towards plan definitions or towards a search for possibilities to
enhance performance).
The analytical differentiation of the management function into domains
or sub-systems and its analytical split into several phases is understood as a
necessary response to the increasing complexity within and around organi-
zations. Management accounting has therefore to ensure the “coordination
of the whole management system to achieve goal-directed management
389
21 Kap 18. Scheytt 05-11-11 16.31 Sida 390
action” (Küpper, Weber, and Zünd 1990, p. 283; transl. by the author).
Planning methods, budgeting, goal hierarchies, management standards, and
the determination of transfer prices are the “generic management account-
ing tools”, by which the “rational” coordination of the management sub-
systems has to be executed (cf. Küpper 1997, pp. 25–29). Rationality in this
context means “to proceed consciously and goal-oriented” (Küpper 1997,
p. 59; transl. by the author). Or, as two of the most influential exponents of
the German-speaking discourse stated (Weber and Schaeffer 1999), theories
of management accounting and control systems have to have the objective
of “assuring” the rationality of management.
The notion of rationality that underlies this conceptualisation is the very
basic means-end rationality in Max Weber’s (1968) sense. Accordingly, the
notion of the organization is that of a mechanism that acts like the ideal type
of a Weberian bureaucracy, or, to use Morgan’s (1986) metaphors, as a
“machine-like” functional mechanism. Undoubtedly, this approach
acknowledges that the reality of organizations is coined by dynamics, con-
flicts, inefficiencies, micro-politics, personal interests etc. From its normative
point of view however, these phenomena are just distortions from an imag-
inable optimum of a means-end rationality. And, it is management account-
ing whose main task is to pursue the idea of a perfect rational organization
and to eliminate any disturbances that pulls the organization away from this
path.
This very short description of the basic features of German-speaking
management accounting theory already indicates some conceptual charac-
teristics that are incompatible with the state of the art of current manage-
ment and organization theory. For example, the very basic insight that all
organizations are shaped by the diverse and often contradicting individual
interests and that organizational actors tend to look after their own interest
in decision-making situations (for example, March 1996) is neglected by the
notion of an ideal type of rationality whose realisation is at least seen as pos-
sible. Furthermore, the notion that there is one rationality standard that has
to be realised contrasts with the basic insight that organizational actors fol-
low diverse rationality concepts and that therefore talk and action in orga-
nizations necessarily differ (Brunsson 1989). Additionally, it is disregarded
that rationality is in organization theory conceptualised as a construct orga-
nizational actors have to make sense of in each situation within the ongoing
organizational processes (cf. Weick 1995) and that therefore no rationality
standard can be “implanted” into an organization, even if it is the very basic
standard of a means-end rationality.
But it is not only the rationality standard which is a critical feature of
German management accounting theory; it is also the notion of management
that shows conceptual shortcomings. This is largely because its under-
standing of management refers to modernist, Taylorism-like conceptions.
390
21 Kap 18. Scheytt 05-11-11 16.31 Sida 391
Consequently, the notion of coordination adopts the same conceptual nar-

rowness. And, the view that planning precedes action (and for example that
structure follows strategy), is in contrast to the more complex views of the
management process, which are sceptical about the “standards” of rational-
ity of management (Cohen, March, and Olsen 1972). The idea that mana-
gerial and operational (e.g. the management accountant’s) tasks, while sys-
tematically overlapping, can be individually assigned, suggests a problemat-
ic view of power relations as it disregards the processual character of task
assignment. On a more technical level, it remains unclear how the micro-
processes of coordination are (or should be) actually formed, i.e. what the
specific contents, procedures and outcomes of such processes are. The issue
of communication that is crucial for the actual process of coordination tends
to be neglected. It is not surprising that the interpretation of the organiza-
tion as a neutral body, and of management as a “rational” activity, is
methodologically followed by a positivist research orientation. The useful-
ness of a theory is therefore judged with respect to its abilities to copy the
reality of organizations.
One could argue that critical accounting theory contradicts the epistemo-
logical and ontological suppositions of positive accounting theory and is
therefore an alternative way of conceptualising the function of management
accounting. Although this tradition encompasses different theoretical per-
spectives, the overarching view is that positive accounting research is not
value-free but serves certain ideologies (see, for example, Tinker, Neimark,
and Merino 1982; Chua 1986; Armstrong 1987; Laughlin 1987; Dillard
1991; Roslender 1993; Miller 1994; Roslender and Dillard 2003). It is
argued that in positive accounting theory the intricate interrelation between
management accounting and the social and organization context in which it
operates (Hopwood 1983; Puxty 1993; Miller, 1994) is simplified via an
abstract distinction between the organization and the ways in which it is
observed, measured and controlled. This led to the development of the so-
called “new accounting research” which is “distinguishable by its (fuller)
recognition of accounting as constitutive of, as well as constituted by, the
social and organizational relations through which it travels, and with which
it engages” (Morgan and Willmott 1993, p. 4).
However, one aspect of critical accounting theory corresponds to posi-
tivist approaches and is particularly important for the conceptualisation of
the function of management accounting. As Jönsson and Macintosh (1997)
explain, critical accounting theorists try to reveal the conflicts under the sur-
face of a seemingly stable social world and to identify the class and group
struggles that are kept hidden behind the façade. In order to establish this
critical attitude, critical accounting theory must, in a way similar to positive
accounting theory, presuppose an ontological constant, from which the
coercive and distorting nature of accounting practices can be identified. As
391
21 Kap 18. Scheytt 05-11-11 16.31 Sida 392
Jönsson and Macintosh (1997, p. 375) highlight, critical accounting theory

is therefore
Foundationalist in that it asserts that analysis of a social order must have a

firm theoretical grounding and in that it posits ex ante the presence of ubiqui-
tous deep structures running below the surface of daily social existence. […]
The ultimate goal is to enlighten academics, practitioners and students alike
about this underside of the accounting world. Such an idealistic impulse
holds out the hope of a more democratic, humanistic and less coercive world.
Despite the diversity of the different approaches it is thus the taken for
granted ontological basis that is common to both, positive and critical
accounting theory. It is the notion of an ideal (organizational) world in the
case of positive accounting theory, in which problems are solved in the man-
ner of a perfect means-end rationality. Similarly, critical accounting theory
relates to theoretical constructs (or normative ideas) that identify the orga-
nizational world as distorted and exploiting the individual. Hence, although
the background ontologies are clearly opposed, it is their function in the
process of theory building that is identical in both approaches.
Management accounting and systems theory:

coordination, communication, and control
We can conclude from the brief discussion of positive and critical manage-
ment accounting theory that the character and the outcome of management
accounting processes are more complicated than the simplifying view of
management accounting as a neutral and objective tool (or, from a negative
perspective, as a means for distortion and coercion) indicates. Hence,
management accounting practices cannot be fully conceptualised within a
theoretical framework that is based on an idealistic (Hegelian) world-view,
which underlies these approaches. This is also indicated by the manifold and
sometimes ambiguous functions that management accounting has in organi-
zational life, for example, in corporate scandals, fraud, embezzlement and
mismanagement and also by the nature of reactions to this by auditing and
risk management practices (Power 1997; 2004). Hence, with respect to this
tangle of diverse intended and unintended effects, the function of manage-
ment accounting cannot be explained by approaches, which are based on the
above mentioned dualistic configuration of management accounting system
vs. organizational reality – whether it appears in the form of positive or in
the form of critical accounting theory. The sometimes surprising outcomes
of the implementation and application of management accounting systems
would, from this perspective, indicate that decisions and actions would be
irrational in relation to what the systems prescribe. Hence, on the basis of a
392
21 Kap 18. Scheytt 05-11-11 16.31 Sida 393
dichotomy of a focal organization and a management accounting system as

a (mis-)representation of this organization, it would be impossible to ade-
quately consider the multi-faceted functions management accounting sys-
tems have.
To overcome the dichotomous configuration, it is useful to refer to the
basic fact that management accounting systems are designed and used by an
organization in correspondence solely to its own requirements, but not to
any outside references. This corresponds with the situation when a manage-
ment accounting system is implemented in an organization. In simple terms,
a management accounting system is developed within and by the organiza-
tion that implements it for its own purpose. In this process, the systems has
to be established, formed and used by the organization or, in the words of
Karl Weick (1995), the organization has to “make sense” of the system and
the information provided by it. The essential question is then how manage-
ment accounting can be described as a recursive process – without the need
to relate to an outside reference, whether a rationality standard, an ideolog-
ical presupposition, or a notion of an ideal world.
Logically, such an approach has to argue that the function of management
accounting systems cannot be identified a priori and without reference to the
focal organization. Rather, the meaning of management accounting is “con-
stituted in use” (Hedlin 1996, p. 193). Expressed in ontological terms, such
an interpretation has a de-ontic orientation in that it explains the use of
management accounting as a process without specific presuppositions, or as
an operation that an organization applies to itself as a means for self-
observation and self-description. By this conceptual rearrangement that
philosophers would call a “linguistic turn” (Wittgenstein 1961) manage-
ment accounting systems adopt a prominent role in the organizational lan-
guage game (cf. Mauws and Phillips 1995). Against this backdrop, manage-
ment accounting is an operation, rather than a task, a tool for, or a process
of representation. This operation can be understood as a medium for shap-
ing the ongoing process of communication in organizations or as a self-
deconstruction of an organization (see, for a similar argumentation, Chia
1997). And, if we take into consideration that communication is a medium
in social life by which coordination takes place, we can argue that coordi-
nation is brought about by communication. This contrasts with positive
accounting theory, where management accounting systems exert the coordi-
nation function via technical adaptations, for example by coordinating dif-
ferent management tasks or phases in the management process. Hence, one
can conclude that management accounting systems shape communication
and thereby constitute coordination.
But how are we to understand the exact functional mechanisms by which
management accounting systems inform communication, and hence coordi-
nation? In the following, it is suggested that social systems theory (Luhmann
393
21 Kap 18. Scheytt 05-11-11 16.31 Sida 394
1990d; 1995f; 1997a; 2000c; Baecker 2001b) provides a framework by

which the functions of management accounting can be conceptualised on the
basis of a very simple operation, namely, the distinction between that what
is (or can be) communicated – and that what is not (or cannot be) commu-
nicated. I do not have the space here to give a detailed introduction into the
concept of social systems theory (see for that purpose Luhmann 2003). In
the following, however, those basic configurations of social systems theory
which are relevant for a conceptualisation of management accounting as
operation of self-observation shall be sketched.
In systems theory words, organizations are complex and intransparent
systems (Luhmann 1997b). Certainly, this insight is anything but new for
management accounting theory. Already Anthony’s groundbreaking work
on “management control systems” (Anthony 1965) emphasises that organi-
zations are complex systems which need specific techniques and practices for
control. Also, the German term for management accounting, Controlling,
indicates this system-oriented view. However, while Anthony sees the system
of an organization as an entity that can be controlled from the outside (e.g.
by a manager controlling a business unit), systems theory states that organi-
zations are autopoietic systems which control themselves in a self-referential
mode. The notion of autopoiesis of social systems was derived from the
work of Maturana and Varela (1980) on the self-reproduction of natural
systems. The main consequences of this theoretical switch are twofold. First,
while a system is structurally coupled with its environment, it is opera-
tionally closed. That is to say, while material (information, resources etc.) is
mutually exchanged between a system and its environment, the ways in
which systems (re-)produce themselves are solely based on their own
resources (material, processes, information, cognition etc.). Second, to main-
tain the border between system and environment is crucial for the existence
of the system. A confusing complexity of information is offered to the sys-
tem by the outside world, thus the system has to reduce this complexity and
thereby to separate itself from its environment to protect its own function-
ality. In contrary to the theory of bio-systems (Maturana and Varela 1980),
where physical and chemical processes are understood as the type of infor-
mation that forms the basis of self-reproduction, it is, in the case of social
systems, communication that is the basic operation of the system.
These characteristics of communication in social systems theory are the
reason, why this framework is particularly useful for a conceptualisation of
management accounting. Therefore, the role and function of communica-
tion for social systems shall be briefly explained. Social systems are defined
as consisting of communications, which are the “elements” of the system,
and references between communications, which are the “relations” between
the “elements”. Individuals (or “psychic systems” in the terminology of
Luhmann 1995f) are therefore seen as part of the environment of the social
394
21 Kap 18. Scheytt 05-11-11 16.31 Sida 395
system. Hence, communication is not to be understood in the common sense

of discussion, conversation, talk, or discourse, carried out by individuals or
groups. Rather, a social system is formed by a network of ephemeral phe-
nomena of communication acts, appearing and immediately disappearing
within the system. Therefore, communications obtain their relevance only
through following communications that refer to them (Luhmann 1995f).
This construction forms the de-ontic basis of social systems theory: commu-
nications are not referred to facts or deeds, but solely to previous commu-
nications. Hence, communications are analogous to self-observations: a
communication reacts to previous communications and is therefore an
observation of an “element” of the system, and will be addressed by subse-
quent communications, hence is object of future self-observations.
One could argue, that this interpretation of a social context replicates a
mechanistic world-view, as the concept of system – particularly with refer-
ence to the first generation of systems theory (Wiener 1961; Bertalanffy
1955) – is generally associated with a notion of control, in which social rela-
tions are seen to be steered in a quasi-technical or strictly functional man-
ner. However, what distinguishes social and technical systems is the fact that
communication is an operation that is shaped by a fundamental contin-
gency. A communication cannot fully determine how an ensuing communi-
cation will relate to it. Rather, the meaning of a communication can be
understood only in retrospect, in its relevance for future communications
which relate to it. Hence, while the sequencing and chaining of communica-
tions as the sole operation of the system seemingly is a socio-technical mech-
anism, it is nevertheless a priori contingent, whether and how future com-
munication will link to present communication.
The reference to communication in social systems theory is related to the
position of epistemological constructivism (Luhmann 1990i). Accordingly,
social systems are, in this view, understood as black boxes. This under-
standing refers to the observation that factors which cannot be fully identi-
fied influence the functional mechanisms of the system. Hence, a social sys-
tem is a so-called “non-trivial machine” (von Foerster 1984), in contrast to
“trivial machines” which can be controlled in a technical manner: the sys-
tem is non-predictable, due to the complexity, which exceeds its own “pro-
cessing capacity” for the production of self-knowledge. Therefore, the sys-
tem is, to a certain extent, intransparent even for its own self-observation:
“The system cannot produce a complete description of itself” (Luhmann
1997b, p. 359; transl. by the author). However, this intransparency does not
imply that systems are generally uncontrollable. Rather, a system is able to,
and has to, stabilise itself recursively by the ongoing flow of communica-
tions carried out within itself Luhmann (1997a, pp.190 ff.).
The contingency of communication is also relevant for understanding the
role of management accounting. If we make the “linguistic turn”, we have
395
21 Kap 18. Scheytt 05-11-11 16.31 Sida 396
to accept that controlling a business as the main goal of management

accounting is not a technical process, but is also contingent and fuzzy, as it
is carried out by and in communication. As Baecker (2000, pp. 59 f.) high-
lights:
Any control is an act of communication and can only be successful to the

extent that communication is successful. […] Any insight into the communi-
cation dynamics of control is a deconstructive one. As soon as we enter the
concept of communication, out goes the concept of causality. Communica-
tion means to be unsure about cause and effect. We have to wait for causes
to effect their effects, and for effects to be caused by causes as soon as com-
munication comes into play. Communication is the concept that steps in as
soon as Gregory Bateson’s statement that “the cause is not the effect” is taken
seriously. This means that “something” is going to happen for a cause to
effect its effect, and for an effect to be caused by a cause. Causes have to
select their effects, and effects have to select their causes in a world that is
characterized by both over-determination and under-determination, i.e. by
there being lots of causes and lots of effects floating around, with no definite
relationship between them.
If we consider this “fuzzy” nature of communication (and hence control),

we face two consequences. First, we realise that management accounting
systems, even though well-known as systems for steering and controlling
organizations, are neither instruments for copying the organizational
“world” in representational forms, nor instruments from which one-to-one
conclusions on future operations can be drawn. Second, the role communi-
cation plays within the social system is not only that of a transmitter of
information from a sender to a recipient, but also a medium in the cognitive
process of the whole social system. This also holds for management account-
ing systems: They influence organizations via communication, but not by
representing economic measures.
Despite the fact that management accounting is a “fuzzy” practice – be-
cause it is carried out via “fuzzy” communication – one can learn from
observations of management accounting praxis, that it is nevertheless under-
stood as a powerful tool having a potential for controlling strategic as well
as operational issues in a variety of organizational types and settings. This
apparent contradiction can be solved if one refers to the form of communi-
cation specific to organizations. In the next chapter, the notion of decision
as the form of communication specific to organizations shall be therefore
further explored.
396
21 Kap 18. Scheytt 05-11-11 16.31 Sida 397
The function of management accounting:

delivering distinctions, defining decision premises
According to Luhmann (2000c), the basic operation of organizations is deci-
sions. Hence, decisions and communications are analogical in the case of
organizations: a decision communicates what has been decided. As is the
case with communications in all social systems, the meaning of a decision in
an organization results from its reference to previous decisions and ensuing
decisions that refer to it. Hence, the web of decisions is the fabric of an
organization.
But how, then, does management accounting intervene in this network of
decisions? At first sight, one could argue that management accounting
affects decisions by providing facts and figures or, in systems theory words,
by performing operations of self-observation that are validated when a deci-
sion refers to them. However, the systems which deliver the facts and figures
don’t come out of the blue; rather, these tools and techniques have specific
requirements insofar as the organization has to choose the scope and the
design of these tools and techniques. Thus, the formation of a management
accounting system also originates in a decision situation. One can conclude
that establishing a management accounting system means to make a decision
that pre-forms the basis to which ensuing decisions can refer to.
For understanding this recursive structure it is useful to relate to two
further elements of social systems theory namely, the interpretation of
Spencer Brown’s (1979) notion of distinction on which Luhmann’s notion of
decision is based on, and, second, the meaning of decision premises in
Luhmann’s concept of the organization (2000c, pp. 222 ff.). As already indi-
cated, it is the very fact in social systems theory that the “point from which
all further investigations must begin (is) not identity but difference”
(Luhmann 1995f, p. 1). Hence, communication is an explication of an
observation, and the way in which an observation is performed is that one
draws a distinction. The discussion of observations as distinctions (cf.
Luhmann 1997a, pp. 120 ff.; 2006) draws on Spencer Brown’s work on the
“calculus of form” (Spencer Brown 1979). The basic proposition is that
“only by drawing a distinction one is able to indicate something and thus
know what to refer to” (Baecker 1999b, p. 7). But this distinction should
not be understood as a definition in the sense of a “classical” representa-
tional epistemology. Rather, a distinction contains more than just an indica-
tive act, by which certain characteristics of an entity are categorised. Spencer
Brown (1979) states that the distinction includes three elements: the indica-
tion of “something” (in Spencer Brown’s words: the “marked state”);
furthermore, the non-indicated rest (“unmarked state”); and finally, the dis-
tinction itself, by which the marked and the unmarked state are separated.
To relate this to the function of management accounting, we can argue
that management accounting systems pre-form specific distinctions that are
397
21 Kap 18. Scheytt 05-11-11 16.31 Sida 398
used in further operations of the organization. In this sense, implementing a

management accounting system is a process of drawing a distinction
between that, what can cognitively be addressed by ensuing decisions and
that what cannot be cognitively addressed. In designing its management
accounting systems an organization therefore distinguishes “patterns” for
future self-observation. For example, if a company implements a Balanced
Scorecard (cf. Kaplan and Norton 1996) for strategic control, it has to
define objectives which are related to strategic goals and to define measures
on which the focus of (management) attention is directed. This is a process,
however, of distinguishing those objectives and measures which are seen as
relevant from those that are – from this moment on – irrelevant.
It is obvious from this description that the implementation of a manage-
ment accounting system in an organization is significant for future decisions.
Such a system aligns the focus of observations in relation to what is
indicated by it and fades out all the rest, that is, what is not indicated by the
drawn distinction. If, then, a decision has to be made, it has to be related to
the information provided by the management accounting system. For
example, if a Balanced Scorecard is once implemented as the central means
for managing performance, no other mode or type of observation will be
any longer relevant for the measurement of units or processes. For example,
a company might start on the basis of a Balanced Scorecard to identify those
business divisions that do not contribute much in pursuing its visions, or
operations which do not contribute to the overall performance – and trans-
form these observations into a decision on, for example, outsourcing of a
unit or a restructuring of processes. Therefore, economic quantities can only
be assessed (and get thereby relevant) through the lenses of this particular
management accounting system once it is implemented.
The logic of these mechanisms correspond to what Luhmann (2000c,
pp. 222 ff.) names “decision premises”. With this concept, borrowed from
the work of Simon (1951), Luhmann denotes those factors that pre-define a
decision situation, but are themselves the result of previous decisions. The
premises serve in future decision situations as a means for orientation. This
does not mean, however, that the content of future decisions is completely
fixed by decision premises. But regardless of whether conformity to, or
deviance from, the premises is the result of a decision, the references a deci-
sion relates to are in any case delivered by the premises. As Luhmann states,
decision premises “focus the communication on the distinctions that are laid
down in the premises” and are utilised in that one does not have to “repeat-
edly unwrap the full complexity of (decision) situations” (Luhmann 2000c,
p. 224; transl. by the author).
Although not explicitly mentioned by systems theory, one could argue that
management accounting systems are important examples of decision prem-
ises, considering their significance in decision situations. The influential role
398
21 Kap 18. Scheytt 05-11-11 16.31 Sida 399
management accounting systems play in decision situations, though, should

not lead to the conclusion that an organization is completely “blind” and
dependent on the distinctions provided by the management accounting sys-
tem. Certainly, from “within” the distinction, for example in a specific deci-
sion situation, the distinction is the “blind spot” for any observer (von
Foerster 1981). However, as is the case with all distinctions, the implemen-
tation of a management accounting system results in more than the indica-
tion of a “marked state”: the act of distinguishing also constitutes the com-
plete form of the distinction which includes both sides of the distinction
– the marked and the unmarked state. This complete form of the distinction,
however, can only be observed from the outside. For example, an external
consultant can examine a management accounting system in order to figure
out whether the system is suitable for the purposes of a company. But this is
obviously another mode of operation than using information provided by
the system in decision situations. In systems theory words, the observation
of the complete form of a distinction comes about by a “re-entry” (Spencer
Brown 1979, pp. 113 ff.) of the distinction into the realm it distinguishes.
Hence, as soon as a distinction is drawn, its functions can be grasped only
by means of a second-order observation. On this level, the complete scene
of a decision situation becomes visible: an observer who observes a distinc-
tion in a second-order observation, for example an investment decision, does
not only observe the two sides of the distinction – that means to agree or to
disagree with the decision taken; the observer also observes the “form” that
make such a choice possible, for example, the figures and ratios that are rel-
evant in the decision process and led to the selection of a particular invest-
ment alternative – and no other; and it is possible to observe that what
“hides” (Baecker 2001b) behind the distinction, for example, the debates
and political processes that led to the selection of one particular alternative
in the decision situation. However, it is clear from the analysis of this “dis-
tinction theory”, that such a second-order observation is an exceptional
form of observations and that this operation is difficult to perform within
an organization that “sticks” to its first-order observations delivered by the
implemented management accounting systems.
Conclusion
The aim of this chapter is to suggest social systems theory as a framework
for analysing the significance of management accounting for the self-obser-
vation of organizations. With respect to the notion of the organization as an
autopoietic, operationally closed system, management accounting was
depicted as a source for distinctions which serve as observations in the com-
munication processes within organizations. Furthermore it was explained
that the tools and measures that are the used for the production of these dis-
399
21 Kap 18. Scheytt 05-11-11 16.31 Sida 400
tinctions are, reciprocally, a result of the ongoing communication process

within the organization.
Two main conclusions can be drawn on the basis of this brief outline of a
systems theory approach to management accounting. First, this conceptual-
isation allows integrating management accounting theory with basic con-
cepts of advanced organization theory. For example, the function of
management accounting can be related to sense-making as the basic require-
ment in processes of organizing (cf. Weick 1995). In this perspective,
management accounting intervenes into the processes of sense-making in
that it penetrates the cognitive patterns to which communication acts in an
organization have to relate. Basic notions (or: distinctions) enacted in the
organization are “defined” by management accounting, for example what is
related to as costs, prices, accountability, success, and failure. Management
accounting can thus be defined as the function, by which the semantic prin-
ciples underlying the communication processes within the organization are
constituted.
This leads us into the second conclusion which concerns the epistemolog-
ical and ontological status of management accounting theory. On the basis
of a social systems theory approach it becomes obvious, that the function of
management accounting is not representing the world outside, but con-
structing the “reality” in an organization. This perspective responds to the
claims for a de-ontic conceptualisation of management accounting without
the need to relate to an outside reference, whether a rationality standard, an
ideological presupposition, or a notion of an ideal world. As Luhmann
(1997b, p. 369) emphasises, one of the consequence of a systems theory
approach is that
One will have to refrain from representational epistemology at all, if one

acknowledges that systems do not operate outside their borders in their envi-
ronment, and therefore cannot communicate their states at all. Operationally
closed systems can only operate within their borders, even if the meaning of
border suggests to them that there has to be another side. With their own
operations they are unable to pierce their borders and therefore they are
unable to compare internal and external states of affairs.
Consequently, if management accounting is conceptualised on the basis of

social systems theory, its function cannot be defined as the representation of
organizational realities. Seen from the level of a second-order observation, it
is also obvious that management accounting systems are not a stable means
for the observation of organizational realities; indeed, these systems are tem-
poral and transient phenomena. Therefore, their potential does not depend
on the rightness or falseness of the data they provide but in their potential
to form the distinctions to which one has to refer to in processes of self-
observation.
400
21 Kap 18. Scheytt 05-11-11 16.31 Sida 401
To sum it up, the analysis of management accounting systems based on the

social systems theory approach is not merely a speculative reflection. It can
be stated that this approach provides comprehensive insights in how
management accounting systems affect patterns of communication and
observation in organizations and hence, what the “real” function of
management accounting is. Thus, a systems theory perspective makes
management accounting accessible to an interpretation that goes beyond an
instrumental description of management accounting techniques and incor-
porates technical aspects into the cognitive effects that are the result of the
application of such instruments. The “outcomes” of management account-
ing systems are the pre-definition and validation of the “language” used in
a focal organization (cf. Baecker 1992b). They distinguish via their applica-
tion what can become a communicable “fact” – and what cannot. It is this
specific function that is peculiar to management accounting – and it is this
specific function that constitutes the influential role management accounting
plays in today’s organizations.
401
21 Kap 18. Scheytt 05-11-11 16.31 Sida 402
22 Glossary 05-11-11 16.31 Sida 403
PART VII
Glossary and Bibliography

22 Glossary 05-11-11 16.31 Sida 404
22 Glossary 05-11-11 16.31 Sida 405
Glossary to Niklas Luhmann’s

Terminology
David Seidl
Action: actions are (fictive) constructs of social systems for observing, and commu-
nicating about, their communications. While communications are the unity of three
selections (utterance, information and understanding), the concept of action cap-
tures merely the first two (utterance and information). The observation of commu-
nication as an action is thus a simplification. When social systems observe commu-
nications as actions, they treat them as the products of persons (“actors”) rather
than of the social system itself.
Autopoiesis: autopoiesis means self-reproduction. A system is autopoietic if its ele-
ments are reproduced through its own network of elements. This does not mean that
the system itself has at its disposal all of the causes necessary for self-reproduction.
It merely means that the system has at its disposal a sufficient range of disponible
causes, so that it can secure its own reproduction under normal circumstances.
Code: a code is a two-sided distinction, where the two sides are the exact opposite
of each other; for example true/false, just/unjust, present/absent. Many social sys-
tems reproduce themselves on the basis of a communication coded in a particular
way. The communications of the legal system [functional differentiation], for exam-
ple, carry the code legal/illegal. That is to say, each legal communication communi-
cates something as being legal or illegal. Similarly, the communications of the inter-
action system carry the code present/absent.
Cognitive routine: cognitive routines are particular types of decision premises. They
refer to the way in which the environment is conceptualised by the organization.
Cognitive routines, for example, inform about characteristics of the customer.
Communication channel: communication channels are particular types of decision
premises. Luhmann refers to them also as the “organization of the organization”.
They determine who can communicate with whom about what.
Communication: in contrast to the classical notion of communication as the trans-
fer of information from a sender to a receiver, according to Luhmann’s concept,
communication is understood as the synthesis of three selections: information (what
is communicated), utterance (how and why it is communicated) and understanding
(the distinction between utterance and information). Communication as this unity
of the three selections is an emergent phenomenon that is not attributable to a single
individual: it presupposes at least two individuals.
Complexity: complexity refers to a situation where there are more elements than can
be related to each other. Complexity in this sense implies the necessity of selection.
405
22 Glossary 05-11-11 16.31 Sida 406
Contingency: contingency is formally defined as the negation of necessity and

impossibility. Contingency means a situation where different selections are possible.
Luhmann speaks of “open contingency”, when the selection is still to be made, in
contrast to “closed contingency”, when the selection has already been made, i.e.
when a different selection could have been made. Decisions transform situations of
open contingency into ones of closed contingency.
Decision premise: The concept of decision premise denotes the structural precondi-
tions for decision making. Luhmann distinguishes between decidable and undecid-
able decision premises. Decidable decision premises refer to those decision premises
that are the result of individual decisions. Every decision is a (decidable) decision
premise for further decisions in the sense that further decisions can take it as a given
point of reference for their own selection. A particular type of decidable decision
premises consists of decision premises that are explicitly decided on, of which there
are three forms: decision programmes, communication channels and personnel.
Undecidable decision premises are decision premises that are not produced by indi-
vidual decisions but emerge as a by-product of the decision-making process. There
are two types of undecidable decision premises: organizational culture and cognitive
routines.
Decision programme: decision programmes are a particular type of decision prem-
ise that define conditions for correct decision making; they are often also called
“plans”. There are two different kinds of programmes: conditional programmes and
goal programmes. Conditional programmes define correct decision making on the
basis that certain conditions are given. They generally have an “if-then” format – “if
this is the case, then do that”. Goal programmes, in contrast, define correct decision
making by defining specific goals that are to be achieved, e.g. “profit maximisa-
tion”, and in this way structure the given decision possibilities.
Decision: decisions are a particular type of communication. In contrast to ordinary
communications, which only communicate a specific content that has been selected
(e.g. “I love you”), a decision communication communicates also – explicitly or
implicitly – that there are alternatives that could have been selected instead (e.g. “I
am going to employ candidate A and not candidate B”). Because of that, decisions
are paradoxical [paradox] communications: the more they communicate that there
are real alternatives to what has been selected, the less the selection will appear as
justified and thus the less the decision will be accepted as “decided”. The more the
selected alternative is justified as the right selection, the less the other options will
appear as alternatives and thus the less the decision will appear as a “decision”.
De-paradoxization (synonymously: de-paradoxification): the concept of de-
paradoxization, or de-paradoxification, refers to the transformation of paradoxes
into forms in which the paradoxicality is disguised. Paradoxes cannot be solved or
eliminated; they can only be shifted “out of sight”, i.e. postponed, suspended, trans-
formed or replaced by using other distinctions.
Element/operation: the elements of autopoietic systems [autopoiesis] are the sys-
tem’s operations. In the case of social systems, for example, the elements/operations
are communications, and in the case of psychic systems they are thoughts. These ele-
ments/operations are momentary events, which occur only once and only in the
briefest period necessary for their appearance. As a consequence, autopoietic sys-
406
22 Glossary 05-11-11 16.31 Sida 407
tems are forced to incessant reproduction if they are not to disappear. The system’s
elements are defined as elements merely through their integration into the system.
Outside or independently of the system they have no status as elements; that is to
say, they are “not ontically pre-given”. Elements can, of course, be composed of dif-
ferent components, which can be analysed independently of the system, but as ele-
mentary units they are only defined through the functions they serve for the system
as a whole.
Environment: the environment is not a kind of pre-given world outside the system,
but it is co-produced together with the system. The environment is as much a prod-
uct of the system as the system itself. In this sense, the environment is the organiza-
tion’s construction, which is not to say that its reality is denied.
Functional differentiation: differentiation refers to the formation of systems within
systems. The differentiation is functional if the subsystem acquires its identity
through the fulfilment of a particular function for the system as a whole. Luhmann
speaks of functional differentiation particularly with regard to society. Modern
society is differentiated into different functional systems. Each of these systems
serves exclusively one particular societal function. There is, for example, the legal
system, the economic system, the system of education, the system of art. All of these
systems are autopoietic systems [autopoiesis] that are merely structurally coupled
[structural coupling] to each other. Most of these systems reproduce themselves on
the basis of a particular code.
Human being: human beings are conglomerates of several autopoietic systems: psy-
chic system, cells, brain, and organism. This conglomerate does not form a systemic
unity. Its different parts are merely structurally coupled.
Interaction system: interaction systems are a particular type of social system, which
produces itself on the basis of particular communications: communications among
people present. They presuppose the participants’ reflexive perception of their phys-
ical presence.
Interpenetration: interpenetration is a particular type of structural coupling. The
prime example of interpenetration is the relation between social system and psychic
system. Luhmann speaks of interpenetration if an autopoietic system presupposes
the complex achievements of the autopoiesis of another system and can treat them
as parts of the own system.
Meaning: meaning is formally defined as the difference between actuality and poten-
tiality. A momentarily actual thought or communication refers to other possible
thoughts or communications. The meaning of a particular communication or
thought is its surplus of references to other thoughts and communications. Social
systems and psychic systems both process meaning; the first in the form of commu-
nication, the second in the form of thoughts. Luhmann distinguishes three dimen-
sions of meaning. The fact dimension of meaning distinguishes references into “this”
and “something else”. Something is “this” and not “something else”. A knife is a
knife and not a spoon, or fork. The temporal dimension of meaning divides refer-
ences into “before” and “after”, as well as into the two horizons of past and future.
The social dimension of meaning refers to the differentiation of references according
to the distinction of “ego” and “alter”. It is a distinction between different social
407
22 Glossary 05-11-11 16.31 Sida 408
perspectives. In every meaningful event the three dimensions can only take place in
combination.
Observation: drawing on Spencer Brown, Luhmann uses the term “observation” in
a very abstract and broad sense, which goes beyond its usual meaning as optical per-
ception. An observation is any type of operation that makes a distinction in order to
indicate either side of the distinction. A communication, for example, is an obser-
vation as it communicates a particular meaning out of all possible meanings; i.e. it
selects a particular meaning from other possible meanings.
Operative closure: operative closure means that a system can only operate in the
context of its own operations and that it depends in this process on the structures
that are being produced precisely by these operations [self-organization]. No opera-
tions from outside can become operations of the system, nor can the system operate
in its environment. A thought, for example, cannot become an operation in a social
system, nor can a communication become an operation in a psychic system.
Operative closure does not mean that the system’s operations are not affected by
external influences. On the contrary, operatively closed systems usually depend on
external influences; the operative closure is a precondition for a system’s interac-
tional openness. All autopoietic systems [autopoiesis] are operatively closed.
Organization: organization is a particular type of social system that reproduces itself
on the basis of decisions.
Organizational culture: organizational culture is a particular type of undecidable
decision premise. It refers to the way in which an organization deals with its own
processes of decision making. For example, if the organization always produces the
same kind of decision (e.g. recruiting merely male candidates) this might condense
into an undecidable decision premise for future decisions – in the sense of “we have
always done it this way”.
Other-reference (synonymously: hetero-reference, environmental reference, external
reference): the concept of other-reference describes the reference of operations to the
environment of the system. Social systems, for example, communicate [communica-
tion] about phenomena of their environment.
Paradox: a paradox occurs when the conditions of the possibility of an operation are
at the same time the conditions of its impossibility. At the centre of Luhmann’s
organization theory is the paradox of decision: if the decision communicates that
there are real alternatives to the decision, the decision will not be accepted as “decid-
ed”. However, if the selected alternative is justified as the right selection, to which
there are no real alternatives, the decision will not appear as a real “decision”.
Formally, paradoxes are conceptualised as a re-entry of a distinction into itself, i.e.
the outside of the distinction enters its inside. As a result, the inside is both inside
and outside of the distinction; the distinction starts to oscillate: if one is inside, one
is outside and thus is inside… Due to this oscillation paradoxes provide no clear
point of connection for ensuing operations and thus tend to lead to paralysis.
Person: persons are a construct of social systems. The person is a bundle of social
expectations directed toward individual human beings.
Personnel: personnel is a particular type of decision premise that concerns the
recruitment and organization of personnel. Organizations, on the one hand, decide
408
22 Glossary 05-11-11 16.31 Sida 409
on the commencement and termination of membership and, on the other hand, on

the transfer of members to different positions within the organization – both with
and without promotion.
Position: positions are nodes at which the three (decidable) decision premises (deci-
sion programme, communication channel, personnel) meet and are specified with
regard to each other. Every position executes a particular programme, is filled by a
particular person, and is located somewhere in the communication network.
Psychic system (synonymously: mind, mental system): psychic systems are autopoi-
etic systems [autopoiesis] that reproduce themselves on the basis of consciousness or
thoughts. Their elements are thoughts that produce further thoughts.
Re-entry: drawing on Spencer Brown, Luhmann speaks of a re-entry when a dis-
tinction is applied to itself. This would occur, for example, if one analysed whether
the distinction just/unjust were itself just, or if a system reflected (inside the system)
its system/environment distinction. Formally, the distinction re-enters the distinction
and the outside of the distinction appears on its inside. As a consequence, the inside
of the distinction becomes both inside and outside. Thus, the value of the distinction
becomes ambiguous; it oscillates between inside and outside: if one is inside, one is
outside and thus inside … [see paradox]. The re-entry is also described as a self-
observation [observation].
Self-organization: in contrast to the concept of autopoiesis the concept of self-
organization does not refer to the (re)production of elements but to that of struc-
tures. Autopoietic systems are always self-organizing as they determine their own
structures.
Self-reference: self-reference designates every operation that refers in some way or
other back to itself. Autopoietic systems [autopoiesis] are self-referential in the sense
that all their operations refer to other operations of the same system.
Social system: social systems are autopoietic systems that reproduce themselves on
the basis of communications (communication). Their elements are communications
that produce further communications. Luhmann distinguishes three types of social
systems: society, organization and interaction.
Societal differentiation: in the course of its evolution, society displayed different
forms of differentiation. In archaic times society was differentiated into equal sub-
systems (segmentation), e.g. different tribes, clans, or families. This was replaced
later on by differentiation according to the logic of centre and periphery, i.e. the
differentiation between city and country. In late medieval times a hierarchical form
of differentiation emerged, which was characterised by different social strata or
classes (stratification). With the emergence of the modern society, around the 18th
century, this was replaced by the current, functional differentiation, where we find
several societal subsystems specialised in serving specific societal functions; e.g. law,
science, economy, art, religion. Each of these primary forms of differentiation can be
combined with the other forms of differentiation on a secondary level; e.g. in strat-
ified society the different strata were often differentiated internally into equal sub-
systems (segmentation), or according to the difference centre/periphery. Similarly,
the different functional subsystems might be differentiated internally either into
equal subsystems, or into centre/periphery, or hierarchically.
409
22 Glossary 05-11-11 16.31 Sida 410
Society: Society is a particular type of social system, which includes all meaningful
communication and is always formed when communication refers to other commu-
nication. All other types of social systems take place within society. Besides repro-
ducing themselves, they always also reproduce society.
Structural coupling: two systems are said to be structurally coupled if their respec-
tive structures are adjusted to each other in such a way as to allow for mutual influ-
ences. Social systems and psychic systems, for example, are structurally coupled
through language. Structural coupling can explain why systems, despite their oper-
ative closure (i.e. in spite of the absence of any operative coupling), remain respon-
sive towards other systems in their environment. Interpenetration is a particular
form of structural coupling.
Structure: structure refers to the selection of relations between the elements of a sys-
tem. The structures of organizations are decision premises. Structures and opera-
tions are recursively related: structures enable and restrict the operations that then
reproduce or change the structures for further operations.
Uncertainty absorption: uncertainty absorption describes the processual aspect of
organization. In the transition from one decision to the next the original uncertain-
ty involved in the decision is absorbed. The ensuing decision takes the result of the
earlier decision as given and does not have to consider the uncertainty involved in
that decision. In other words, uncertainty absorption takes place when decisions are
accepted as decision premises and taken as the basis for subsequent decisions.
410
23 Bibliography 05-11-11 16.32 Sida 411
Annotated Bibliography
of Selected Works by
Niklas Luhmann
Kai Helge Becker
The enormous œuvre of Niklas Luhmann during his life as a scholar, from
1958 to 1998, comprises some hundred different publications, including
many dozens of monographs. Giving a full, or indeed an accurate, account
of his works is not an easy task, as Luhmann was not particularly meticu-
lous about keeping a complete list of his publications. The most compre-
hensive catalogue of Luhmann’s writings by far, is Niklas Luhmann’s
Complete Works 1958–2005, scheduled to be published simultaneously in
the USA and Germany in 2006 under the editorship of Klaus Dammann
(University of Bielefeld) and a team of several editors whose aim is to cata-
logue Luhmann’s publications in different languages. This CD-ROM will
contain a list of Luhmann’s works in altogether about twenty languages and
counts ninety-three publications in English. Among the latter, there are
seventeen books, fifty-five essays, including lectures and contributions to
books, as well as several reviews and interviews. Thanks to Klaus Dam-
mann’s kind support, the bibliography of the present volume already bene-
fits from his editorial work.
The present annotated bibliography is intended to facilitate further read-
ing by providing an overview of Luhmann’s œuvre, as well as some basic
orientation to those interested in his work. The texts listed here are a selec-
tion of Luhmann’s English publications and have been chosen according to
the criteria of accessibility (regarding both the content of the text and the
source where the text has been published) and representativeness for
Luhmann’s later works after his “autopoietic turn”. The selection has been
guided also by a certain bias towards what may be considered especially rel-
evant to the scholar of organization studies. In nine cases, so far untranslat-
ed German texts have been included, due to their outstanding significance
among Luhmann’s works. In sum, the forty-eight publications chosen here
cover approximately half of Luhmann’s books and essays written in English.
Due to the internal structure of Luhmann’s theory, his writings strongly
overlap in content. Therefore, in conjunction with Luhmann’s papers in this
volume, this necessarily subjective selection of English (and additional
411
German) publications will serve, despite its limited scope, as a helpful guide
to the vast landscape of Luhmann’s writings.
In view of the exuberance of ideas that can be found even in the smallest
among Luhmann’s writings, the annotations presented below are as concise
as possible. Primarily, they focus on information that may be especially rel-
evant for a first insight into the main topics and the basic structure of his
oeuvre. Occasionally, they also present some background information on
the publications. The selected books and papers have been grouped accord-
ing to subject areas, instead of the usual order according to the year of pub-
lication. This way of organizing the material may further support the aim of
providing an overview of Luhmann’s writings and their interrelations, and
will hopefully facilitate identifying those publications that are most valuable
to the reader. Many of the selected texts have been published more than
once. In these cases, the most accessible source has been chosen, and both
the year of the first publication and the language in which the text original-
ly appeared have been added in brackets.
Epistemology (theory of observation)

The concept of observation as a (paradoxical) operation of distinction and
indication lies at the heart of Luhmann’s late works. In particular, it lays the
foundations for his “difference-oriented” way of theory-building in general,
his notion of how systems observe and can be observed, and, as a conse-
quence of this, his constructivist concept of epistemology.
[1] “Identity – What or How?”, in N. Luhmann (2002), Theories of
Distinction: Redescribing the Descriptions of Modernity (Stanford,
CA: Stanford University Press), 113–127 (first pub. in German 1990).
In this text, Luhmann distances himself from the classical concept of ontology,
which is understood as a mode of theorizing that is based on the distinction of
“being/non-being”. Against this background, he explains his concept of observa-
tion and how this shifts the focus from the question of what a certain identity is,
to the question of how identities are constructed.
[2] “The Cognitive Program of Constructivism and the Reality That Re-
mains Unknown,” in N. Luhmann (2002), Theories of Distinction:
Redescribing the Descriptions of Modernity (Stanford, CA: Stanford
University Press), 128–154 (first pub. in German 1990).
Here Luhmann takes up the discourse of “radical constructivism”. Aiming at clar-
ifying the debate, he elaborates on his concept of constructivism and describes in
what way it is an empirically based approach presupposing a pre-given “real”
world that must be considered cognitively unapproachable.
[3] “System as Difference”, Organization, 13/1, [forthcoming 2006] (in

German: tape recording 1992; printed edn. 2002).
412
This text is based on a transcription of a lecture on his social-theoretical perspec-

tive, which Luhmann delivered to students at the university of Bielefeld in the win-
ter term 1991/1992. It is a fairly accessible explanation of the “difference-oriented”
approach underlying his way of theorising, and of how it translates into his con-
cept of the system/environment distinction.
[4] “The Paradox of Form”, in D. Baecker (1999) (ed.), Problems of Form

(Stanford, CA: Stanford University Press), 15–26 (first pub. in German
1993).
Proceeding from Spencer Brown’s concept of form as a distinction and indication,
Luhmann describes here his concept of the paradox that lies at the heart of every
observation, and shows how it can and must be de-paradoxized for an observation
to take place.
[5] “The Paradox of Observing Systems”, in N. Luhmann (2002), Theories

of Distinction: Redescribing the Descriptions of Modernity (Stanford,
CA: Stanford University Press), 79–93 (first pub. in English 1995).
In this text, Luhmann addresses the paradox of observation with regard to the
question of what it means to observe meaningful systems that are themselves
observing. Moreover, he sketches how such a perspective can be applied to observe
the system of modern society.
[6] “Deconstruction as Second-Order Observing”, in N. Luhmann (2002),

Theories of Distinction: Redescribing the Descriptions of Modernity
(Stanford, CA: Stanford University Press), 94–112 (first pub. in English
1993).
In this text Luhmann starts with the example of the debate on gays in the U.S. mil-
itary, in order to contrast his concept of observation with the discourse on decon-
structivism, as based on Derrida’s works. He shows how the idea of deconstruction
can be reconstructed and further developed by the concept of second-order obser-
vation.
[7] “Why Systems Theory?”, Cybernetics and Human Knowing, 3/2

(1995), 3–10 (first pub. in German 1994).
In this short text, Luhmann sketches his concept of autopoietic systems and its epis-
temological consequences in the context of other theoretical approaches. He argues
that systems theory is an adequate way of observing society for it can account for
the experience of contingency that characterises modernity.
Social theory (general theory of social systems)

The publications listed here all refer to Luhmann’s general concepts of var-
ious social phenomena, i.e. to what other theorists would describe with con-
cepts such as “social action”, “practices”, “discourse” or “acting self”. In
Luhmann’s systems-theoretical framework, these issues are addressed in the
framework of a general theory of social systems that is based on his notion
of communication and includes the questions of how social systems are
related to the mind and how they evolve over time.
413
[8] “Meaning as Sociology’s Basic Concept”, in N. Luhmann (1990),

Essays on Self-Reference (New York: Columbia University Press),
21–79 (first pub. in German 1971).
This essay originates in the prominent Habermas/Luhmann debate on systems-
based research in the early 1970s. A lot of concepts that will later turn out to play
a major role in his general theory of social systems are already addressed here in
the context of a theory of meaning.
[9] “What is Communication?”, in N. Luhmann (2002), Theories of

This rather short publication is one of Luhmann’s most salient texts on his concept
of communication. It is structured along several theses on the “nature” of commu-
nication which are then briefly explained.
[10] “How Can the Mind Participate in Communication?”, in N. Luhmann

(2002), Theories of Distinction: Redescribing the Descriptions of
Modernity (Stanford, CA: Stanford University Press), 169–186 (first
pub. in German 1988).
In this text, Luhmann describes the relationship of psychic and social systems dur-
ing the process of communication. It may serve as a quite accessible presentation
of Luhmann’s often criticised idea that only communication, not humans, can
communicate.
[11] “The Concept of Society”, Thesis Eleven, 31 (1992), 67–80.

This paper first gives a brief overview of Luhmann’s reasons for some of his basic
social-theoretical concepts. Moreover, the consequences of his approach on the
analysis of language, the issue of rationality, and the relation between individual
and society are demonstrated.
[12] “The Self-Description of Society: Crisis Fashion and Sociological

Theory”, International Journal of Comparative Sociology, 25 (1984),
59–72.
Here Luhmann sketches three of the basic elements of his social theory: the con-
cepts of evolution, system differentiation, and self-referentiality. The discussion is
embedded in Luhmann’s criticism that society, including sociology, tends to be too
quick to judge critically certain social phenomena instead of analysing more thor-
oughly the situation at hand.
[13] “The Individuality of the Individual: Historical Meanings and

Contemporary Problems”, in N. Luhmann (1990), Essays on Self-
Reference (New York: Columbia University Press), 107–122 (first pub.
in English 1986).
Luhmann’s distinction between psychic and social systems begs the question of
how he reconstructs the concepts of individuality and the subject. This paper
argues that the concept of the subject is historically rooted in the transition from
pre-modern to modern society. With the “subject” being relativized in this way, the
concept of autopoietic systems turns out to be its legitimate successor regarding the
question of individuality.
414
[14] “The Evolutionary Differentiation between Society and Interaction”, in

J.C. Alexander et al. (1987) (eds.), The Micro-Macro Link (Berkeley:
University of California Press), 112–131.
This publication provides Luhmann’s perspective on the well-known micro/macro
problem in social theory. Instead of seeing it as a primarily theoretical issue,
Luhmann makes the point that this dichotomy is mainly empirical in nature, as it
is based on the historical process of differentiation between society and interaction.
[15] “The Paradox of System Differentiation and the Evolution of Society”,

in J.C. Alexander and P. Colomy (1990) (eds.), Differentiation Theory
and Social Change: Comparative and Historical Perspectives (New
York: Columbia University Press), 409–440.
Luhmann describes here how he conceives of social differentiation. After sketching
the historical background of the concept, Luhmann presents the four kinds of sys-
tem differentiation that he distinguishes. In the remainder, he applies his concept to
analyse the evolutionary process of the transition to modern society.
[16] “The Improbability of Communication”, in N. Luhmann (1990),

Essays on Self-Reference (New York: Columbia University Press),
This publication is one of Luhmann’s earlier papers introducing his concept of
“media of communication” as a means by which society deals with the unlikeliness
of successful communication. This theory is fully elaborated in [32].
[17] Social Systems (Stanford: Stanford University Press, 1995) (first pub. in
German 1984).
This book is Luhmann’s groundbreaking opus magnum on social theory. Starting
from a general theory of autopoietic systems, Luhmann elaborates on his theory of
social and psychic systems and reconstructs many social-theoretical fundamentals,
such as structure, time, conflict, and rationality, from his systems-theoretical per-
spective. The final chapter sketches the consequences that his approach had on his
take on epistemology and theory of science (as presented in full in [21]) some years
later. It is worth noting that the English edition contains an additional preface on
the concepts of subject and action.
Theory of society (theory of modernity and

functional differentiation)
This section comprises publications on how Luhmann conceives of modern
society. At the core of his theory lies the description of the different func-
tional subsystems that characterise modern society and – as a result of this
– the idea that modern society is multi-perspectival with no centre that could
represent it as a whole.
415
[18] Ecological Communication (Chicago: University of Chicago Press,

1989) (first pub. in German 1986).
Here Luhmann describes how the different functional subsystems of society react
to today’s ecological problems. Based on a talk given at the Rhein-Westphalian
Academy of Science, this book probably is by far the most accessible introduction
to the way in which Luhmann conceives of modern society as consisting of several
functional subsystems.
[19] Law as a Social System (Oxford: Oxford University Press, 2004) (first
This is one of the two books (among altogether eight) on different functional sub-
systems that have already been translated into English. Apart from Luhmann’s
comprehensive analysis of the system of law, it may be of interest to English-
speaking readers also because it covers the basic concepts for analysing functional
subsystems (i.e. code, programme, function etc.) in a fairly detailed way.
[20] Art as a Social System (Stanford: Stanford University Press, 2000) (first
This exploration of the subsystem of art is the second translated book among those
on the functional subsystems. On a more general level, this book features a detailed
account of the relationship between communication and perception, and of the role
of second-order observation in the context of functional subsystems.
[21] Die Wissenschaft der Gesellschaft [“The Science of Society” or

“Science as a Social System”] (Frankfurt am Main: Suhrkamp, 1990).
Apart from its description of the academic system of society – which comprises
both the sciences and the humanities – this book presents the most elaborated
account of Luhmann’s epistemological position and of how he situates his own aca-
demic work. The book has not been translated yet; however, the following two
papers may provide international readers with an insight into Luhmann’s concepts.
[22] “The Differentiation of Advances in Knowledge: The Genesis of

Science”, in N. Stehr and V. Meja (1984) (eds.), Society and
Knowledge: Contemporary Perspectives in the Sociology of Knowledge
(London: Transaction), 103–148 (first pub. in German 1981).
Though this essay was written some years before The Wissenschaft der Gesellschaft
[21] and is therefore not based that extensively on Luhmann’s later concepts, it
already presents in a rather detailed way many of his main ideas on the genesis and
functioning of the academic system, which were further developed in subsequent
publications.
[23] “The Modernity of Science”, in N. Luhmann (2002), Theories of

This paper is the translation of the final chapter of [21]. In qualifying science as a
social phenomenon that has evolved with modern society, Luhmann finally rela-
tivizes his entire theory as a communication that adopts the particular perspective
of only one specific functional system and is inseparably bound to the particular
conditions of modernity.
416
[24] Die Wirtschaft der Gesellschaft [“The Economy of Society” or

“Economy as a Social System”] (Frankfurt am Main: Suhrkamp,
1988).
This publication covers the economic subsystem. The English-speaking reader will
find some general remarks on this system in other publications, such as [18], [19]
and [39], for example. [43] is a translation of the final chapter. Some of Luhmann’s
very early ideas on the economic system can be found in N. Luhmann (1982), The
Differentiation of Society (New York: Columbia University Press), 190–225 (first
[25] Die Religion der Gesellschaft [“The Religion of Society” or “Religion

as a Social System”] (Frankfurt am Main: Suhrkamp, 2000).
Luhmann’s book on the religious system has been published posthumously on the
basis of an already rather elaborated manuscript. It may be worth noting that this
book also contains a chapter on religious organizations. Until a translation is avail-
able, Luhmann’s early ideas on religion, as presented in N. Luhmann (1984),
Religious Dogmatics and the Evolution of Society (New York: Mellen; first pub. in
German 1977), may give a first insight into the topic, including Luhmann’s per-
spective on religious organizations.
[26] Die Politik der Gesellschaft [“The Politics of Society” or “Politics as a

Social System”] (Frankfurt am Main: Suhrkamp, 2000).
Published posthumously on a similar basis as [25], this book describes the political
subsystem, including one chapter on the organizations of the political system.
Luhmann’s very early ideas on this topic have been published in N. Luhmann
(1982), The Differentiation of Society (New York: Columbia University Press),
138–189 (first pub. in German 1968 and 1974).
[27] Das Erziehungssystem der Gesellschaft [“The Educational System of

Society” or “Education as a Social System”] (Frankfurt am Main: Suhr-
kamp, 2002).
Also this book is a posthumous publication. The probably most important early
publication on the educational system is N. Luhmann and K.E. Schorr (2002),
Problems of Reflection in the System of Education (Oxford: Berghan; first pub. in
German 1979, rev. edn. 1988), which also includes a chapter devoted to educa-
tional organizations.
[28] The Reality of the Mass Media (Cambridge: Polity Press, 2000) (first
Though Luhmann considers also the mass media a societal subsystem, he has not
placed it in the series of the other seven books on subsystems. From his point of
view, this system is not fully differentiated, due to the way in which it communi-
cates with the other parts of society. This book has attracted much attention
because of its thought-provoking thesis that the things we know about the world
are a construct of the mass media.
417
[29] Love as Passion (Cambridge: Polity Press, 1986) (first pub. in German
1982).
An important part of Luhmann’s works are historical analyses of how structural
changes in society entail semantical changes. This book presents a detailed analy-
sis of how, as a consequence of the transition from a stratified to a functionally dif-
ferentiated society, the modern concept of love evolved from the 17th century on.
Further analyses of this kind have been published in the four-volume Gesellschafts-
struktur und Semantik [“The Structure of Society and Semantics”] (Frankfurt am
Main: Suhrkamp, 1980–1995). [13] is an excerpt from a chapter of the third
volume.
[30] Observations on Modernity (Stanford: Stanford University Press,

1998) (first pub. in German 1992).
This collection of papers on modern society reacts, among other topics, to charac-
terisations of today’s society as “postmodern”: Luhmann takes up the postmodern
insight of the impossibility of métarécits (Lyotard) and explains the claim of a post-
modern condition by the fact that modern society is functionally differentiated. In
this way, the postmodern semantics turns out to make sense only in the context of
a society that, structurally, is still modern.
[31] “Globalization or World Society: How to Conceive of Modern

Society?”, International Review of Sociology, 7/1 (1997), 67–79.
Here, Luhmann elaborates on his concept of modern society as a functionally dif-
ferentiated world society that cannot be adequately described in terms of regional,
ethnic or cultural differentiation. In relating this concept to the discourse on “glob-
alisation”, this paper shows the critical potential of systems-theoretical analyses.
[32] Die Gesellschaft der Gesellschaft [“The Society of Society” or “Society

as a Social System”], 2 vols. (Frankfurt am Main: Suhrkamp, 1997).
These two volumes contain Luhmann’s summa sociologica concerning the theory
of society as a whole, in contrast to the different functional subsystems it contains.
They comprise an account of society as the encompassing social system, a fully
elaborated theory of media of communications (cf. [16]), the final stage of his the-
ories of evolution and system differentiation (cf. [15]), and several chapters on
societal self-descriptions. There is also a brief chapter on how organizations relate
to society.
Organization theory
The publications in this section make up, or rather mark the core of
Luhmann’s theoretical work on organizations, which spanned his entire aca-
demic life. In the early works listed here, Luhmann critically examines the
heritage of the classical concept of rational organizations coordinating
action on the basis of division of labour and by means of hierarchy. In the
later works, Luhmann has fully developed his alternative perspective on
organizations as a specific kind of social system whose characteristic ele-
ments are decisions.
418
[33] “Ends, Domination, and System: Fundamental Concepts and Premises

in the Work of Max Weber”, in N. Luhmann (1982), The
Differentiation of Society (New York: Columbia University Press),
Referring to results of empirical research until the 1960s, this paper criticises
Weber’s notion of organizations as entities coordinating labour on the basis of
authority and means-end relations. In addition, Luhmann sketches a possible alter-
native from his systems-theoretical point of view.
[34] Funktionen und Folgen formaler Organisation [“Functions and

Consequences of Formal Organization”] (Berlin: Duncker & Humblot,
1964).
Combining the distinctions of “system/environment” and “formal/informal”,
Luhmann analyses the effects of formal organizational structures by focusing on
the possible conflicts that arise when organizations have to develop formal struc-
tures that must both integrate their members and allow for adaption to environ-
mental demands. Through this perspective, the book succeeds in integrating the
findings of the empirical research at that time and raises many issues that were put
on the agenda of organization studies only decades later.
[35] Zweckbegriff und Systemrationalität: Über die Funktion von Zwecken

in sozialen Systemen [“The Concept of Ends and System Rationality:
About the Function of Ends in Social Systems”] (Tübingen: Mohr/
Siebeck, 1968).
This book presents a fundamental systems-theoretical critique of the classical con-
cept of organizational means and ends (cf. [33]). In developing Simon’s criticism
much further, Luhmann argues that organizations cannot be understood in terms
of a final end or purpose, but instead that the concept of ends is only one among
many functionally equivalent solutions to the problem of how organizations can
deal with a complex environment. Many of the concepts in this text still are of
importance in Luhmann’s late works on organizations.
[36] “Interaction, Organization, and Society”, in N. Luhmann (1982), The

Differentiation of Society (New York: Columbia University Press),
69–89, 372–376 (first pub. in German 1975).
This early paper introduces Luhmann’s central distinction of the three types of
social systems: interaction systems, organizations and society. Having yet to devel-
op his concept of autopoiesis, Luhmann sees organizations as systems that draw a
boundary within society by means of the distinction “member/non-member”, with
membership being subject to certain conditions imposed by the organization.
[37] “A General Theory of Organized Social Systems”, in G. Hofsteede and

M.S. Kassem (1976) (eds.), European Contributions to Organization
Theory (Assen: Van Gorcum), 96–113 (first pub. in German 1975).
This paper aims at presenting more thorough reflections on organizations than
those of the contingency approach in the mid 1970s. Organizations are conceived
of as self-referential systems that experience their structures as contingent, there-
fore put themselves under strong pressure to select, and in this situation refer to
419
decision premises. This notion of organization is later reconstructed by Luhmann

on the basis of the concept of de-paradoxisation.
[38] “Organization”, in T. Bakken and T. Hernes (2003) (eds.), Autopoietic

Organization Theory: Drawing on Niklas Luhmann’s Social Systems
Perspective (Copenhagen: Copenhagen Business School Press), 31–52
(first pub. in German 1988).
This is the first text where Luhmann elaborates on his theory of organizations as
autopoietic systems with decisions as their basic elements. From this perspective, he
addresses the self-referential closure of organizations, the concept of decisions as
events in time, the role of membership, organizational structures (including deci-
sion premises), the relation of organizations to their environment on the basis of
the concepts of redundancy and variety, and, finally, rationality as a re-entry of the
“system/environment” distinction into the system. Although rather short, this text,
read in combination with Luhmann’s chapter on the paradox of decision making
in the present volume (chapter 4), covers many of the basic concepts of Luhmann’s
organization theory.
[39] “Membership and Motives in Social Systems”, Systems Research: The

Official Journal of the International Federation for Systems Research,
13/3 (1996), 341–348.
This short paper provides a brief overview of some concepts central to Luhmann’s
late organization theory. After introducing his notion of the social system,
Luhmann presents the concepts of person and motives, and addresses the function
of organizational membership. In the final section, he sketches his notion of the
economy as an autopoietic system.
[40] Organisation und Entscheidung [“Organization and Decision”]

(Opladen: Westdeutscher Verlag, 2000).
This book, published posthumously on the basis of an almost finished manuscript,
contains Luhmann’s fully developed organization theory on the basis of the con-
cepts of organization as an autopoietic system (cf. [38]), and of the paradox of
decisions (cf. Luhmann’s chapter in the present volume). In this perspective,
Luhmann addresses the different kinds of decision premises that he distinguishes,
the topics of structural change, technology, and rationality, and the ways in which
organizations relate to both themselves via self-descriptions and their societal envi-
ronment.
Miscellaneous
This final section comprises some publications that are not central to
Luhmann’s theory, but may nevertheless be of interest to the scholar of
organization studies. The books and essays listed here address the topics of
trust, power and risk, the question of steering and control regarding social
systems, Luhmann’s relation to phenomenology and the Frankfurt school of
sociology, his concept of sociological theory, and, finally, his stance towards
ethics.
420
[41] Trust and Power (Chichester: Wiley, 1979) (first pub. in German 1973
and 1975).
The two shorter books collected in this publication, address two concepts that do
not play a major role within the general framework of Luhmann’s theory (though
the concept of power is fairly central to Luhmann’s theory on politics as presented
in [26]). However, they are a very accessible read and offer in nuce an insight into
Luhmann’s style of theory construction. The part on power contains also a chapter
on organizations.
[42] Risk: A Sociological Theory (Berlin: de Gruyter), 1993 (first pub. in

German 1991).
This book can be seen as Luhmann’s equivalent to the sociological diagnosis of a
“risk society” as advocated by other authors. Especially Luhmann’s distinction
between risk (referring to damage that may result from one’s own decisions) and
danger (referring to damage stemming from external sources) has attracted quite
some attention. Moreover, as [40] has not been published yet in English, this book
may offer an interesting read regarding Luhmann’s perspective on technology. Also,
this book includes a short chapter on organizations.
[43] “Limits of Steering”, Theory, Culture & Society, 14/1 (1997), 41–57
(first pub. German 1988).
This paper is the translation of the final chapter of [24]. Though it has a certain
focus on the economic system, the theory of steering (in the sense of a reduction of
a difference) that Luhmann develops here is more general in scope and must there-
fore be considered of relevance to organization studies and a theory of manage-
ment.
[44] “The Control of Intransparency”, Systems Research and Behavioral

Science, 14 (1997), 359–371.
Given this bibliography’s focus on the social sciences (which corresponds to the fact
that Luhmann always saw himself as a sociologist), it should not be overlooked
that Luhmann has innovatively contributed also to General Systems-Theory. This
paper elaborates on the role of time and memory with relation to self-referential,
operationally closed systems.
[45] “The Modern Sciences and Phenomenology”, in N. Luhmann (2002),

Theories of Distinction: Redescribing the Descriptions of Modernity
(Stanford, CA: Stanford University Press), 33–60 (first pub. in German
1996).
This essay is the published version of a lecture that Luhmann delivered in Vienna
in commemoration of Husserl’s Vienna lectures in 1935. It sheds light on the rele-
vance of Husserl’s phenomenology to Luhmann’s thought, and the way in which
Luhmann reconstructs Husserl’s ideas in the framework of his systems theory.
[46] “I See Something You Don’t See”, in N. Luhmann (2002), Theories of

This paper is based on a talk Luhmann gave at a conference on the significance of
the Frankfurt School. Here Luhmann (rather polemically) highlights his main
421
arguments against the type of critical sociology that was developed by this school
of thought in post-war Germany.
[47] “‘What Is the Case?’ and ‘What Lies behind It?’: The Two Sociologies
and the Theory of Society”, Sociological Theory, 12/2 (1994), 126–139
(first pub. in German 1993).
This paper is a translation of Luhmann’s final lecture at the University of Bielefeld
in 1993. In relation to what he considers to be the two basic questions of sociolo-
gy, Luhmann characterises his own concept of sociological theory as a way in
which society describes itself within society.
[48] “Paradigm Lost: On the Ethical Reflection of Morality”, Thesis Eleven,

29 (1991), 82–94 (first pub. in German 1988).
This is the published version of a talk Luhmann gave on the occasion of being
awarded the Hegel-Preis in 1989. It presents Luhmann’s sceptical perspective on
moral communication in view of the conditions characterising modern society, and
assesses the conditions of the possibility of an adequate ethical theory.
422
Contributors
Dirk Baecker is Professor of Sociology at the University of Witten/Herdecke.
Kai Helge Becker is a doctoral student at the Department of Operational
Research, London School of Economics.
Michael Bommes is Professor of Sociology at the Institute for Migration
Research and Intercultural Studies, University of Osnabrück.
Barbara Czarniawska is Professor of Management Studies, GRI-Research
Institute at the School of Business, Economics and Law, Göteborg Uni-
versity.
Thomas Drepper is Postdoctoral Researcher at the Nijmegen School of
Management, Radboud University Nijmegen.
Raimund Hasse is Professor of Sociology at Luzern University.
Darnell Hilliard is a researcher at the Institute for World Society Studies,
Bielefeld University.
Morten Knudsen is Assistant Professor at the Department of Organization
and Industrial Sociology, Copenhagen Business School.
Jochen Koch is Assistant Professor of Management at the Institute of
Management, Freie Universität Berlin.
Niklas Luhmann (†) was Professor of Sociology at Bielefeld University.
Tobias Scheytt is Assistant professor of Controlling at the Department of
Organization and Learning, University of Innsbruck.
David Seidl is Assistant Professor of Organization and Strategy at the In-
stitute of Business Policy and Strategic Management, University of Munich.
Fritz Simon holds a chair for Leadership and Organization in Family
Businesses at the University of Witten/Herdecke.
Veronika Tacke is Professor of Sociology at Bielefeld University.
Jan-Peter Vos is Assistant Professor of Organization Studies at Eindhoven
University of Technology.
423
24 Referenser 05-11-11 16.33 Sida 424
References
Aaronson, D.E., Dienes, C.T., and Musheno, M.C. (1984), Public Policy and Police
Discretion: Processes of Decriminalization (New York: Clark Boardman
Callaghan).
Adams, D. (1979), The Hitch Hiker’s Guide to the Galaxy (London: Pan Books).
Aharoni, Y. (1971), “On the Definition of a Multinational Corporation”, The
Quarterly Review of Economics and Business, 11/3: 27–37.
Alchian, A.A. (1950), “Uncertainty, Evolution, and Economic Theory”, Journal of
Political Economy, 58: 211–221.
Aldrich, H.E., and Marsden, P.V. (1994), “Environments and Organizations” in N.J.
Smelser and R. Swedberg (1994) (eds.), The Handbook of Economic Sociology
(Princeton, NJ: Princeton University Press), 361–392.
Alexander, J.C. (1988), “The New Theoretical Movement”, in N.J. Smelser (1988)
(ed.), Handbook of Sociology (London: Sage), 77–101.
Alhafaji, A.F. (1989), A Stakeholder Approach to Corporate Governance: Managing
in a Dynamic Environment (New York: Quorum Books).
Aliber, R.Z. (1970), “A Theory of Direct Foreign Investment”, in C.P. Kindleberger
(1970) (ed.), The International Corporation (Cambridge, MA: MIT Press),
17–34.
Aliber, R.Z. (1971), “The Multinational Enterprise in a Multiple Currency World”,
in J. H. Dunning (1971) (eds.), The Multinational Enterprise (London: Allen &
Unwin), 49–56.
Allport, F.H. (1940), “An Event-System Theory of Collective Action: With
Illustrations from Economic and Political Phenomena and the Production of
War”, The Journal of Social Psychology, 11: 417–445.
Allport, F.H. (1954), “The Structuring of Events: Outline of a General Theory with
Applications to Psychology”, Psychological Review, 61: 281–303.
Allport, F.H. (1967), “The Theory of Einestructure (Event-Structure): Report of
Progress”, American Psychologist, 22: 1–24.
Andersen, N.Å., and Born, A. (2000), Kærlighed og omstilling (Copenhagen:
Samfundslitteratur).
Ansoff, H.I. (1965), Corporate Strategy: An Analytic Approach to Business Policy
for Growth and Expansion (London: McGraw-Hill).
Ansoff, H.I. (1987), “The Emerging Paradigm of Strategic Behavior”, Strategic
Management Journal, 8/6: 501–516.
Anthony, R.N. (1965), Planning and Control Systems: A Framework for Analysis
(Boston, MA: Harvard Business School Press).
Argyris, C. (1957), Personality and Organization (New York: Harper & Brothers).
Armstrong, P. (1987), “The Rise of Accounting Controls in British Capitalist Enter-
prises”, Accounting, Organizations and Society, 10/2: 129–148.
Arnold, Thurman W. (1935), The Symbols of Government (New Haven: Yale
University Press).
Arrow, K.J. (1964), “The Role of Securities in the Optimal Allocation of Risk-
Bearing”, Review of Economic Studies, 31: 91–96.
Ashby, W.R. (1956), An Introduction to Cybernetics (London: Chapman & Hall).
424
24 Referenser 05-11-11 16.33 Sida 425
Ashby, W.R. (1958), “Requisite Variety and Its Implications for the Control of
Complex Systems”, Cybernetica 1: 83–99.
Ashby, W.R. (1960), Design for a Brain: The Origin of Adaptive Behaviour (2nd
edn., London: Chapman & Hall).
Ashmore, M. (1989), The Reflexive Thesis: Wrighting Sociology of Scientific
Knowledge (Chicago: The University of Chicago Press).
Atkinson, M.A., Cuff, E.C., and Lee, J.R. (1978), “The Recommencement of a
Meeting As a Member’s Accomplishment”, in J. Schenkein (1978) (ed.), Studies in
the Organization of Conversational Interaction (New York: Academy Press),
133–275.
Bach, M. (1999), Die Bürokratisierung Europas: Verwaltungseliten, Experten und
politische Legitimation in Europa (Frankfurt am Main: Campus).
Baecker, D. (1988), Information und Risiko in der Marktwirtschaft (Frankfurt am
Main: Suhrkamp).
Baecker, D. (1992a), “Die Unterscheidung zwischen Kommunikation und Bewußt-
sein”, in W. Krohn and G. Küppers (1992) (eds.), Emergenz: Die Entstehung von
Ordnung, Organisation und Bedeutung (Frankfurt am Main: Suhrkamp).
Baecker, D. (1992b), “The Writing of Accounting”, Stanford Literature Review, 9/2:
157–178.
Baecker, D. (1993), Die Form des Unternehmens (Frankfurt am Main: Suhrkamp).
Baecker, D. (1999a), “Fehldiagnose ‘Überkomplexität’”, in D. Baecker (1999),
Organisation als System (Frankfurt am Main: Suhrkamp), 27–37.
Baecker, D. (1999b), “Introduction”, in D. Baecker (1999) (ed.), Problems of Form
(Stanford: Stanford University Press), 1–14.
Baecker, D. (1999c), Organisation als System (Frankfurt am Main: Suhrkamp).
Baecker, D. (2000a), “Ausgangspunkte einer soziologischen Managementlehre”,
Soziale Systeme, 6: 137–168.
Baecker, D. (2000b), “Die Strategie der Organisation”, Wittener Diskussions-
papiere, 74 (Universität Witten/Herdecke).
Baecker, D. (2001a), “Editorial: Begriff und Phänomen der Strukturellen
Kopplung”, Soziale Systeme, 7/2: 219–221.
Baecker, D. (2001b), “Why Systems?”, Theory, Culture & Society, 18: 59–74.
Baecker, D. (2002), Wozu Systeme? (Berlin: Kulturverlag Kadmos).
Baecker, D. (2003), “Die Strategie der Organisation”, in D. Baecker (2003) (ed.),
Organisation und Management: Aufsätze (Frankfurt am Main: Suhrkamp),
152–178.
Baecker, D. et al. (1987) (eds.), Theorie als Passion (Frankfurt: Suhrkamp),
423–446.
Bailey, K.F. (1994), Sociology and the New Systems Theory: Towards a Theoretical
Synthesis (New York: State University of New York).
Bakhtin, M.M. (1981), The Dialogic Imagination: Four Essays (Austin, TX:
University of Texas Press).
Baldersheim, H., and Ståhlberg, K. (1995) (eds.), Towards the Self-regulating
Municipality: Free Communes and Administrative Modernization in Scandinavia
(Dartmouth: Aldershot).
Baldi, S. (1998), “Normative versus Social Constructivist Processes in the Allocation
of Citations: A Network-Analytic Model”, American Sociological Review, 63/6:
829–846.
425
24 Referenser 05-11-11 16.33 Sida 426
Bardmann, T.M. (1994), Wenn aus Arbeit Abfall wird: Aufbau und Abbau orga-
nisatorischer Realitäten (Frankfurt am Main: Suhrkamp).
Barel, Y. (1979), Le Paradoxe et le Systeme: Essai sur le Fantastique Social
(Grenoble: Presses Universitaires).
Barnard, C.I. (1968), The Functions of the Executive (repr., Cambridge, MA:
Harvard University Press).
Barney, J.B. (2001), “Is the Resource-based ‘View’ a Useful Perspective for Strategic
Management Research? Yes”, Academy of Management Review, 26/1: 41–56.
Barry D., M. and Elmes (1997), “Strategy Retold: Toward a Narrative View of
Strategic Discourse”, Academy of Management Review, 22/2: 429–452.
Bartness, A., and Cerny, K. (1993), “Building Competitive Advantage Through a
Global Network of Capabilities”, California Management Review, 35/2: 78–103.
Bateson, G. (1972), Steps to an Ecology of Mind (New York: Ballantine).
Bausor, R. (1994), “Entrepreneurial Imagination, Information, and the Evolution of
the Firm”, in R.W. England (1994) (ed.), Evolutionary Concepts in Contemporary
Economics (Ann Arbor, MI: University of Michigan Press), 179–189.
Beck, U. (1992), Risk Society: Towards a New Modernity (London: Sage).
Beck, U. (1998a) (ed.), Perspektiven der Weltgesellschaft (Frankfurt am Main:
Suhrkamp).
Beck, U. (1998b) (ed.), Politik der Globalisierung (Frankfurt am Main: Suhrkamp).
Becker, K.H., and Haunschild, A. (2003), “The Impact of Boundaryless Careers on
Organizational Decision Making: An Analysis from the Perspective of Luhmann’s
Theory of Social Systems”, International Journal of Human Resource Manage-
ment, 14/5: 713–727.
Belkaoui, A. (1997), Accounting Theory (London: Dryden Press).
Benson, J.K. (1975), “The Interorganizational Network as a Political Economy”,
Administrative Science Quarterly, 20: 229–249.
Bentzon, K.H. (1988), Fra vækst til omstilling – moderniseringen af den offentlige
sektor (Copenhagen: Nyt fra samfundsvidenskaberne).
Bergmann, W. (1981), Die Zeitstrukturen sozialer Systeme: Eine systemtheoretische
Analyse (Berlin: Duncker and Humblot).
Berman, H.J. (1983), Law and Revolution: The Formation of the Western Legal
Tradition (New York: The President and Fellows of Harvard College).
Bernard, J.-F. (1759), Eloge d’Enfer: ouvrage critique, historique et moral (La
Haye).
Bertalanffy, L. (1955), “General Systems Theory”, in G. Chen et al. (1955) (eds.),
The General Theory of Systems Applied to Management and Organization, vol. 2
(Seaside, CA: Intersystems), 9–18.
Bettis, R.A., and Prahalad, C.K. (1995), “The Dominant Logic: Retrospective and
Extension”, Strategic Management Journal, 16/1: 5–14.
Bijker, W.E., Hughes, T.P., and Pinch, T. J. (1987) (eds.), The Social Construction of
Technological Systems: New Direction in the Sociology and History of
Technology (Cambridge, MA: MIT Press).
Blum, A.F., and McHugh, P. (1971), “The Social Ascription of Motives”, American
Sociological Review, 36: 98–109.
Bocchi, G., and Ceruti, M. (1985) (eds.), La sfida della complessità (Milan:
Feltrinelli).
426
24 Referenser 05-11-11 16.33 Sida 427
Boden, D. (1994), The Business of Talk: Organizations in Action (Cambridge:

Polity).
Boje, D.M., Gephart, R.P., and Thatchenkery, T.J. (1996) (eds.), Postmodern
Management and Organization Theory (Thousand Oaks, CA: Sage).
Boli, J., and Thomas, G.M. (1997), “World Culture in the World Polity: A Century
of International Non-Governmental Organization”, American Sociological
Review, 62: 171–190.
Boltanski, L., and Chiapello, È. (1999), Le nouvel esprit du capitalism (Paris: Galli-
mard).
Bommes, M. (1996), “Ausbildung in Großbetrieben; Einige Gründe, warum aus-
ländische Jugendliche weniger Berücksichtigung finden”, in R. Kersten, D. Kiesel,
and S. Sargut (1996) (eds.), Ausbilden statt Ausgrenzen: Jugendliche ausländi-
scher Herkunft in Schule, Ausbildung und Beruf (Frankfurt am Main: Haag und
Herchen), 31–44.
Bommes, M. (1999), Migration und nationaler Wohlfahrtsstaat: Eine differen-
zierungstheoretische Analyse (Opladen: Westdeutscher Verlag).
Bourdieu, P. (1977), Outline of a Theory of Practice (Cambridge: Cambridge Uni-
versity Press).
Bourdieu, P. (1984), Distinction: A Social Critique of the Judgement of Taste
(London: Routledge & Kegan Paul).
Bourdieu, P. (1988), Homo Academicus (Stanford: Stanford University Press).
Bourdieu, P. (1990a), “Fieldwork in Philosophy”, in P. Bourdieu (1990), In other
Words: Essays Towards a Reflexive Sociology (Cambridge: Polity Press), 3–33.
Bourdieu, P. (1990b), The Logic of Practice (Cambridge: Polity Press).
Bourdieu, P. (2000), Pascalian Meditations (Stanford: Stanford University Press).
Bråten, S. (1984), “The Third Position – Beyond Artificial and Autopoietic Reduc-
tion”, Kybernetes, 13: 157–163.
Brewer, G.A. (2001), “The Great Transformation? Administrative and Civil Service
Reform in Western Democracies” in S.E. Condrey and R. Maranto (2001) (eds.),
Radical Reform of the Civil Service (Lanham: Lexington Books).
Brown, M.K. (1981), Working the Street: Police Discretion and the Dilemma of
Reform (New York: Russell Sage Foundation).
Brown, S.L., and Eisenhardt, K.M. (1998), Competing on the Edge: Strategy as
Structured Chaos (Boston, MA: Harvard Business School Press).
Brummer, J.J. (1991), Corporate Responsibility and Legitimacy: An Inter-
disciplinary Analysis (New York: Greenwood Press).
Bruner, J. (1990), Acts of Meaning (Cambridge, MA: Harvard University Press).
Brunsson, N. (1985), The Irrationality of Organization: Irrationality as a Basis for
Organizational Action and Change (Chichester: Wiley).
Brunsson, N., and Olsen, J.P. (1993), The Reforming Organization (London:
Routledge).
Brunsson, N. (2002), The Organization of Hypocrisy: Talk, Decisions and Actions
in Organizations (Oslo: Copenhagen Business School Press).
Buckley, W. (1967), Sociology and Modern System Theory (Englewood Cliffs, NJ:
Prentice Hall).
Bunge, M. (1979), “A Systems Concept of Society: Beyond Individualism and
Holism”, Theory and Decision, 10: 13–30.
427
24 Referenser 05-11-11 16.33 Sida 428
Burgelman, R.A. (1996), “Process Model of Strategic Business Exit”, Strategic

Management Journal, 17/S1: 193–214.
Burke, K. (1945), A Grammar of Motives (Reprinted together with A Rhetoric of
Motives 1950, Cleveland, Ohio: World Publishing Company).
Burke, K. (1962), A Grammar of Motives and A Rhetoric of Motives (Cleveland,
World Pub. Co.).
Burke, K. (1969), A Grammar of Motives (repr., Berkeley: California University
Press).
Burns, T., and Stalker, G.M. (1961), The Management of Innovation (London:
Tavistock).
Burrell, G., and Morgan, G. (1979), Sociological Paradigms and Organizational
Analysis: Elements of the Sociology of Corporate Life (Aldershot Hants: Gower).
Burt, R.S. (1982), Towards a Structural Theory of Action: Network Models of
Social Structure, Perceptions, and Action (New York: Academic Press).
Butler, R. (1991), Designing Organizations: A Decision-Making Perspective (Lon-
don: Routledge).
Bühler, K. (1934), Sprachtheorie: Die Darstellungsfunktion der Sprache, (Jena:
Fischer).
Calori, R. (1998), “Philosophizing on Strategic Management Models”, Organiza-
tion Studies, 19/2: 281–306.
Campbell, D.T. (1970), “Natural Selection as an Epistemological Model”, in R.
Naroll and R. Cohen (1970) (eds.), A Handbook of Method in Cultural
Anthropology (Garden City, NY: Natural History Press), 51–85.
Campbell, D.T. (1974), “Evolutionary Epistemology”, in P.A. Schilpp (ed.), The
Philosophy of Karl Popper (La Salle, HI.: Open Court Publications), 413–465.
Campbell, D.T. (1975), “On the Conflicts between Biological and Social Evolution
and between Psychological and Moral Tradition”, American Psychologist, 30:
1103–1126.
Cantwell, J. (1989), Technological Innovation and Multinational Corporations
(New York: Basil Blackwell Publishers).
Cantwell, J. (1992), “The Internationalisation of Technology Activity and its
Implications for Competitiveness”, in O. Granstrand et al. (1992) (eds.),
Technology Management and International Business (Chichester [u. a.]: Wiley),
75–95.
Cantwell, J. (1995), “Multinational Corporations and Innovatory Activities: To-
wards a New, Evolutionary Approach”, in J. Molero et al. (1995) (eds.),
Technological Innovation, Multinational Corporations and New International
Competitiveness (Reading, UK: Hardwood Academic Publishers), 21–57.
Castells, M. (1996), The Rise of the Network Society (Malden, CA: Blackwell).
Chandler, A.D. (1962), Strategy and Structure: Chapters in the History of the
Industrial Enterprise (New York: Anchor Books).
Chandler, A.D. (1977), The Visible Hand: The Managerial Revolution in American
Business (Cambridge, MA: Harvard University Press).
Chaudhuri, K.N. (1978), The Trading World of Asia and the English East India
Company, 1660–1760 (Cambridge: Univ. Press).
Chia, R. (1997), Organizational Analysis as Deconstructive Practice (Berlin/New
York: de Gruyter).
Child, J. (1972), “Organizational Structure, Environment and Performance: The
Role of Strategic Choice”, Sociology, 6: 1–22.
428
24 Referenser 05-11-11 16.33 Sida 429
Chua, W.F. (1986), “Radical Developments in Accounting Thought”, The

Accounting Review, 61/4: 601–632.
Churchland, P.M. (1984), Matter and Consciousness: A Contemporary Introduction
to the Philosophy of Mind (Cambridge, MA: The MIT Press).
Cohen, M.D., March, J.G., and Olsen, J.P. (1972), “A Garbage Can Model of
Organizational Choice”, Administrative Science Quarterly, 17/1: 1–25.
Cole, J.R., and Zuckerman, H. (1975), “The Emergence of a Scientific Specialty:
The Self-exemplifying Case of the Sociology of Science”, in L.A. Coser (1975)
(ed.), The Idea of Social Structure: Papers in Honor of Robert K. Merton (New
York: Harcourt Brace Jovanovich), 139–174.
Coleman, J.S. (1982), The Asymmetric Society (Syracuse, NY: Syracuse University
Press).
Coleman, J.S. (1990), Foundations of Social Theory (Cambridge, MA and London:
The Belknap Press).
Colie, R.L. (1966), Paradoxia Epidemica: The Renaissance Tradition of Paradox
(Princeton: Archon Books).
Collins, D. (2000), Management Fads and Buzzwords: Critical-Practical Per-
spectives (London: Routledge).
Condrey, S.E., and Maranto, R. (2001) (eds.), Radical Reform of the Civil Service
(Lanham: Lexington Books).
Cook, K.S. (1977), “Exchange and Power in Networks of Interorganizational
Relations”, Sociological Quarterly, 18: 62–82.
Cooper, R., and Burrell, G. (1988), “Modernism, Postmodernism and Organiza-
tional Analysis: An introduction”, Organization Studies, 9/1: 91–112.
Corbridge, S., Martin, R., and Thrift, N. (1994) (eds.), Money, Power and Space
(Oxford: Blackwell).
Corte-Real, N., and Kelly, P. (2000) (eds.), Administrations in Transition.
Modernisation of Public Administration in Four Countries: Portugal, the
Netherlands, Ireland and France (Maastricht: European Institute of Public
Administration).
Cummings, J.L., and Doh, J.P. (2000), “Identifying Who Matters: Mapping Key
Players in Multiple Environments”, California Management Review, 42/2:
83–104.
Cusa, N. of (1981), Nicholas of Cusa on Learned Ignorance: a Translation and an
Appraisal of De Docta Ignorantia by Jasper Hopkins (Minneapolis: A.J. Benning
Press).
Cyert, R.M., and March, J.G. (1963), A Behavioural Theory of the Firm
(Englewood Cliffs: Prentice-Hall).
Czarniawska, B. (1985), Public Sector Executives: Managers or Leaders (Stockholm:
EFI Research Report).
Czarniawska, B. (1997), Narrating the Organization: Dramas of Institutional
Identity (Chicago: The University of Chicago Press).
Czarniawska, B. (2000), A City Reframed: Managing Warsaw in the 1990s
(Reading, UK: Harwood Academic Publishers).
Czarniawska, B. (2001), “Having Hope in Paralogy”, Human Relations, 54/1:
13–21.
Czarniawska, B. (2003), A Tale of Three Cities, or the Glocalization of City
Management (Oxford, UK: Oxford University Press).
429
24 Referenser 05-11-11 16.33 Sida 430
Daft, R.L., and Weick, K.E. (1984), “Toward a Model of Organizations as

Interpretation Systems”, Academy of Management Review, 9/2: 284–295.
Dahringer, L.D., and Mühlbacher, H. (1991), International Marketing: A Global
Perspective (Reading, MA: Addison-Wesley).
Dale, K. (2001), Anatomising Embodiment and Organization Theory (New York:
Palgrave).
Dan-Cohen, M. (1986), Rights, Persons, and Organizations: A Legal Theory for
Bureaucratic Society (Berkeley and Los Angeles: University of California Press).
Dandeker, C. (1990), Surveillance, Power and Modernity: Bureaucracy and
Discipline from 1700 to the Present Day (Cambridge: Polity Press).
Dawkins, R. (1976), The Selfish Gene (New York: Oxford University Press).
Deleuze, G. (1995), Negotiations, 1972–1999 (New York: Columbia University
Press).
Derrida, J. (1970), “Structure, Sign, and Play in the Discourse of the Human
Sciences” in R. Macksey and E. Donato (eds.), The Languages of Criticism and
the Sciences of Man: The Structuralist Controversy (Baltimore: The Johns Hop-
kins Press), 247–265.
Derrida, J. (1972), Marques de la philosophie (Paris: Critique).
Derrida, J. (1974), Glas (Paris: Galilée).
Derrida, J. (1977), Of Grammatology (Baltimore MD: Johns Hopkins University
Press).
Derrida, J. (1978), Writing and Difference (Chicago: University of Chicago Press).
Derrida, J. (1979), Writing and Difference (London: Routledge).
Derrida, J. (1984), Margins of Philosophy (Chicago: University of Chicago Press).
Derrida, J. (2002), Negotiations: Interventions and Interviews, 1971–2001 (Cali-
fornia: Stanford University Press).
DeSanctis, G., and Poole, M.S. (1994), “Capturing the Complexity in Advanced
Technology Use: Adaptive Structuration Theory”, Organization Science, 5/2:
121–147.
Deutschmann, C. (2002), Postindustrielle Industriesoziologie; Theoretische Grund-
lagen, Arbeitsverhältnisse und soziale Identitäten (Weinheim: Juventa).
Dicken, P. (1999), Global Shift: Transforming the World Economy (3rd edn.,
London: Chapman).
Dillard, J. (1991), “Accounting as a Critical Social Science”, Accounting, Auditing
and Accountability Journal, 4/1: 8–28.
DiMaggio, P.J., and Powell, W.W. (1983), “The Iron Cage Revisited: Institutional
Isomorphism and Collective Rationality in Organizational Fields”, American
Dobbin, F.R. (1994), “Cultural Models of Organization: The Social Construction of
Rational Organizing Principles”, in D. Crane (1994) (ed.), The Sociology of
Culture (Oxford: Blackwell), 117–142.
Donaldson, T., and Preston, L.E. (1995), “The Stakeholder Theory of the Corpora-
tion: Concepts, Evidence and Implications”, Academy of Management Review,
20/1: 65–91.
Douglas, M., and Wldavsky, A. (1982), Risk and Culture: An Essay on the Selection
of Technical and Environmental Dangers (Berkeley: California University Press).
Draft, R.L., and Lengel, R.H. (1984), “Information Richness: A New Approach to
Managerial Behavior and Organization Design”, Research in Organizational
Behavior, 6: 191–233.
430
24 Referenser 05-11-11 16.33 Sida 431
Drazin, R., and Sanderlands, L. (1992), “Autogenesis: A Perspective on the Process

of Organizing”, Organization Science, 3: 230–241.
Drepper, T. (2003), Organisationen der Gesellschaft – Gesellschaft und Organisa-
tion in der Systemtheorie Niklas Luhmanns (Wiesbaden: Westdeutscher Verlag).
Drew, P., and Heritage, J. (1992) (eds.), Talk at Work: Interaction in Institutional
Settings (Cambridge, MA: Cambridge University Press).
Dreyfus, H.L., and Rabinow, P. (1983), Michel Foucault: Beyond Structuralism and
Hermeneutics (2nd edn., Chicago: The University of Chicago Press).
Duff, A. (2000), Information Society Studies (London: Routledge).
Duimering, P.R., and Safayeni, F. (1998), “The Role of Language and Formal
Structure in the Construction and Maintenance of Organizational Images”,
International Studies of Management & Organization, 28/3: 57–85.
Dunning, J.H. (1971), “The Multinational Enterprise: The Background”, in J.H.
Dunning (ed.), The Multinational Enterprise (London: Allen & Unwin), 15–48.
Dunning, J.H. (1974), “The Distinctive Nature of the Multinational Enterprise”, in
J.H. Dunning (1974) (ed.), Economic Analysis and the Multinational Enterprise
(London: Allen & Unwin), 13–30.
Dupuy, J.-P. (1982), Ordres et désordres: Enquête sur un nouveau paradigme (Paris:
Seuil).
Durkheim, E. (1984), The Division of Labor in Society (New York: Free Press).
Ebeling, H. (1976) (ed.), Subjektivität und Selbsterhaltung: Beiträge zur Diagnose
der Moderne (Frankfurt: Suhrkamp).
Eco, U. (1981), The Role of the Reader: Explorations in the Semiotics of Texts
(London: Hutchinson).
Edwards, R. (1979), Contested Terrain: The Transformation of the Workplace in the
Twentieth Century (New York: Basic Books).
Eikhof, D.R., and Haunschild, A. (2005), “For Art’s Sake! Managing Artistic and
Economic Logics in Creative Production”, paper presented at the 11th APROS
Conference, Melbourne, Australia, 04/12–07/12/2005.
Eisenhardt, K.M., and Martin, J.A. (2000), “Dynamic Capabilities: What Are
They?”, Strategic Management Journal, 21/10–11: 1105–1121.
Eisenstadt, S.N. (1963), The Political System of Empires (London and New York:
The Free Press Glencoe).
Emmanuel, C., Otley, D., and Merchant, K. (1999), Accounting for Management
Control (3rd edn., London: Thomson Business Press).
Esposito, E. (1992), L’operazione di osservazione: Costruttivismo e teoria dei siste-
mi sociali (Milano: F. Angeli).
Etzioni, A. (1968), The Active Society: A Theory of Societal and Political Processes
(New York: The Free Press).
‘Evaluering om samarbejdsudvalgenes oplevelse af information, medbestemmelse-
medinflydelse, samarbejde. Evalueringsgruppen af den nye ledelsesstruktur’ 1999,
Sundhedsvæsenet, Frederiksborg amt (Evaluation of the Works Councils’
Experience of Information, Participation, and Cooperation).
Exner, A., Königswieser, R., and Titscher, S. (1987), “Unternehmensberatung-
systemisch: Theoretische Annahmen und Interventionen im Vergleich zu anderen
Ansätzen”, Die Betriebswirtschaft, 47: 265–284.
431
24 Referenser 05-11-11 16.33 Sida 432
Farber, M. (1967), The Foundation of Phenomenology: Edmund Husserl and the

Quest for a Rigorous Science of Philosophy (2nd rev. edn., New York: State Univ.
of New York Press).
Fayol, H. (1925), Administration industrielle et générale (Paris: Dunod).
Fischer, H.R. (1991) (ed.), Autopoiesis: Eine Theorie im Brennpunkt der Kritik
(Heidelberg: Auer).
Flaherty, M.T. (1986), “Coordinating International Manufacturing and
Technology”, in M.E. Porter (1986) (ed.), Competition in Global Industries
(Boston, MA: Harvard Business School Pr.), 83–109.
Fligstein, N. (1990), The Transformation of Corporate Control (Cambridge, MA:
Harvard University Press).
Foerster, H. von (1949), Das Gedächtnis: Eine Quantenphysikalische Untersuchung
(Vienna: Deuticke).
Foerster, H. von (1950), “Quantum Mechanical Theory of Memory”, in H. Von
Foerster (1950) (ed.), Cybernetics: Transactions of the Sixth Conference (New
York: Josiah Macy Jr. Foundation), 112–145.
Foerster, H. von (1969), “What is Memory that It may Have Hindsight and
Foresight as Well?”, in S. Bogoch (1969) (ed.), The Future of Brain Sciences:
Proceedings of a Conference Held at New York Academy of Medicine (New York:
Plenum Press), 19–64.
Foerster, H. von (1981), Observing Systems (Seaside, CA: Intersystems
Publications).
Foerster, H. von (1984), “Principles of Self-Organization – in a Socio-Managerial
Context”, in H.U. Gumbrecht and G.J.B. Probst (1984) (eds.), Self-Organization
and Management of Social Systems: Insights, Promises, Doubts, and Questions
(Berlin: Springer), 2–24.
Foerster, H. von (1985), Sicht und Einsicht: Versuche zu einer Operativen
Erkenntnistheorie (Braunschweig: Vieweg).
Foerster, H. von (1992), “Ethics and Second-Order Cybernetics”, Cybernetics and
Human Knowing, 1: 9–19, 61–88.
Foerster, H. von (1993a), “Das Gleichnis vom blinden Fleck: Über das Sehen im all-
gemeinen” in G.J. Lischka (1993) (ed.), Der entfesselte Blick: Symposion, Work-
shops, Ausstellung (Bern: Benteli Verlag), 15–47.
Foerster, H. von (1993b), “Für Niklas Luhmann: Wie Rekursiv Ist Kommunika-
tion?”, Teoria Sociologica, 2: 61–88.
Foucault, M. (1966), The Order of Things: Archaeology of the Human Sciences
(1970, London: Tavistock Publications).
Foucault, M. (1969), The Archaeology of Knowledge, London: Routledge).
Foucault, M. (1972), The Archaeology of Knowledge (New York: Harper Colo-
phon).
Foucault, M. (1973), The Order of Things: An Archaeology of Human Sciences
(New York: Vintage/Random House).
Foucault, M. (1975), Discipline and Punish: The Birth of the Prison (1979,
Harmondsworth: Penguin).
Foucault, M. (1976), An Introduction: The Will to Knowledge, vol. 1 of
M. Foucault (1981), The History of Sexuality (Harmondsworth: Penguin).
Foucault, M. (1977), “Nietzsche, Genealogy, History”, in P. Rabinow (1984) (ed.),
The Foucault Reader (New York: Pantheon), 76–100.
432
24 Referenser 05-11-11 16.33 Sida 433
Foucault, M. (1979), Discipline and Punish: The Birth of the Prison (New York:
Vintage/Random House).
Foucault, M. (1984a), The Use of Pleasure, vol. 2 of M. Foucault (1988), The
History of Sexuality (Harmondsworth: Penguin).
Foucault, M. (1984b), The Care of the Self, vol. 3 of M. Foucault (1990), The
History of Sexuality (1990, Harmondsworth: Penguin).
Fox, M. (1981), “An Organizational View of Distributed Systems”, IEEE Trans-
actions on Systems, Man, and Cybernetics, 11: 70–80.
Freeman, R.E. (1984), Strategic Management: A Stakeholder Approach (Boston,
MA: Pitman).
Friedland, R., and Alford, R.R. (1991), Bringing Society Back in: Symbols, Practices
and Institutional Contradictions, in W.W. Powell and P.J. DiMaggio (1991) (eds.),
New Institutionalism in Organizational Analysis (Chicago and London: The
University of Chicago Press), 232–266.
Frooman, J. (1999), “Stakeholder Influence Strategies”, Academy of Management
Review, 24/2: 191–205.
Fuchs, P. (1992), Die Erreichbarkeit der Gesellschaft (Frankfurt am Main: Suhr-
kamp).
Fuchs, P. (1997), “Adressabilität als Grundbegriff der soziologischen System-
theorie”, Soziale Systeme: Zeitschrift für soziologische Theorie, 3: 57–79.
Fuchs, S. (1988), “The Constitution of Emergent Interaction Orders: A Comment on
Rawls”, Sociological Theory, 6: 122–124.
Fuchs, S. (1995), “The Stratified Order of Gossip”, Soziale Systeme, 1: 47–73.
Gaastra, F.S., and Bruijn, J.R. (1993), “The Dutch East India Company’s Shipping,
1602–1795, in a Comparative Perspective”, in F.S. Gaastra and J.R. Bruijn (1993)
(eds.), Ships, Sailors and Spices: East India Companies and their Shipping in 16th,
17th and 18th Centuries (Amsterdam: NEHA), 177–208.
Gadamer, H.-G. (1989), Truth and Method (2nd rev. edn., New York: Crossroad).
Galbraith, J. (1973), Designing Complex Organizations (Reading, MA: Addison-
Wesley).
Galbraith, J. (1977), Organization Design (Reading, MA: Addison-Wesley).
Geertz, C. (1980a), “Blurred Genres: The Refiguration of Social Thought”,
American Scholar, 29/2: 165–179.
Genell, K. (2000), Skilda Världar – Att vara Direktör i Staten (Stockholm, SCORE
Research Papers, 2000/2).
Geyer, P., and Hagenbüchle, R. (1992) (eds.), Das Paradox: Eine Herausforderung
des abendländischen Denkens (Tübingen: Königshausen & Neumann).
Ghoshal, S., and Bartlett, C.A. (1993), “The Multinational Corporation as an Inter-
organizational Network”, in S. Ghoshal and E. Westney (1993) (eds.),
Organisation Theory and the Multinational Corporation (Basingstoke [u.a.]:
Macmillan St. Martin’s Pr.), 77–104.
Giddens, A. (1976), New Rules of Sociological Methods (London: Hutchinson).
Giddens, A. (1979), Central Problems in Social Theory: Action, Structure and
Contradiction in Social Analysis (Berkeley: University of California Press).
Giddens, A. (1984), The Constitution of Society: Outline of the Theory of Structura-
tion (Cambridge: Polity Press).
Giddens, A. (1986), The Constitution of Society: Outline of the Theory of Structura-
tion (Berkeley, CA: University of California Press).
433
24 Referenser 05-11-11 16.33 Sida 434
Gide, A. (1947), Corydon (3rd edn., Paris: Gallimard).

Gioia, D.A., and Thomas, J.B. (1996), “Institutional Identity, Image, and Issue Inter-
pretation: Sensemaking During Strategic Change in Academia”, Administrative
Science Quarterly, 41/3: 370–403.
Gioia, D.A., Schultz, M., and Corley, K.G. (2000), “Organizational Identity, Image,
and Adaptive Instability”, Academy of Management Review, 25/1: 63–81.
Girard, R. (1982), Le bouc émissaire (Paris: LGF – Livre de Poche).
Glagow, M., and Willke, H. (1987) (ed.), Dezentrale Gesellschaftssteuerung:
Probleme der Integration polyzentrischer Gesellschaft (Pfaffenweiler: Centaurus-
Verlags-Gesellschaft).
Glanville, R. (1981), “The Same is Different”, in M. Zeleny (1981) (ed.), Auto-
poiesis: A Theory of Living Organization (New York: Elsevier Science).
Godzich, W. (1991), “Vom Paradox der Sprache zur Dissonanz des Bildes”, in
H.U. Gumbrecht and K.L. Pfeiffer (1991) (eds.), Paradoxien, Dissonanzen, Zu-
sammenbrüche: Situationen offener Epistemologie (Frankfurt am Main: Suhr-
kamp), 747–758.
Goffman, E. (1959), The Presentation of Self in Everyday Life (New York: Anchor
Books).
Goffmann, E. (1971), Relations in Public: Microstudies of the Public Order (Har-
mondsworth: Penguin).
Goffman, E. (1974), Frame Analysis: An Essay on the Organization of Experience
(Cambridge, MA: Harvard University Press).
Goffman, E. (1975), Frame Analysis: An Essay on the Organization of Experience
(Harmondsworth: Penguin).
Goffmann, E. (1983), “The Interaction Order”, American Sociological Review, 48:
1–17.
Granovetter, M. (1985), Economic Action and Social Structure: The Problem of
Embeddedness, American Journal of Sociology, 91: 481–510.
Guggenberger, B. (1983), “An den Grenzen der Verfassung”, Frankfurter Allgemeine
Zeitung, no. 281, 3 December.
Guillet de Monthoux, P. (1983), Action and Existence: Anarchism for Business
Administration (Chichester: Wiley).
Gulick, L., and Urwick, L. (1937) (eds.), Papers on the Science of Administration
(New York: Institute of Public Administration).
Gumbrecht, H.U. (1991), “Inszenierte Zusammenbrüche oder: Tragödie und
Paradox”, in H.U. Gumbrecht and K.L. Pfeiffer (1991) (eds.), Paradoxien,
Dissonanzen, Zusammenbrüche: Situationen offener Epistemologie (Frankfurt
am Main: Suhrkamp), 471–494.
Gutenberg, E. (1958), Einführung in die Betriebswirtschaftslehre (Wiesbaden:
Gabler).
Günther, G. (1976), “Cybernetic Ontology and Transjunctional Operations”, in
G. Günther (ed.), Beiträge zur Grundlegung einer operationsfähigen Dialektik,
vol. 1 (Hamburg: Meiner), 249–328 (esp. 319).
Günther, G. (1979), “Natural Numbers in Trans-classic Systems”, in G. Günther
(1979), Beiträge zur Grundlegung einer operationsfähigen Dialektik, vol. 2
(Hamburg: Meiner), 241–264.
Günther, G. (1980), Beiträge zur Grundlegung einer operationsfähigen Dialektik,
vol. 3 (Hamburg: Felix Meiner Verlag).
434
24 Referenser 05-11-11 16.33 Sida 435
Habermas, J. (1981), Theorie des kommunikativen Handelns (Frankfurt: Suhr-

kamp).
Habermas, J. (1984), Reason and the Rationalization of Society, vol. 1 of The
Theory of Communicative Action (Boston, MA: Beacon Press).
Habermas, J. (1987), The Theory of Communicative Action (Cambridge: Polity).
Habermas, J. (1990), The Philosophical Discourse of Modernity, 12 Lectures
(Cambridge, MA: MIT Press).
Habermas, J. (1996), Facts and Norms: Contributions to a Discourse Theory of
Law and Democracy, trans. William Rehg (Cambridge: Polity).
Habermas, J., and Luhmann, N. (1971), Theorie der Gesellschaft oder Sozial-
technologie (Frankfurt am Main: Suhrkamp).
Hahn, A. (1982), “Zur Soziologie der Beichte und anderer Formen institutionali-
sierter Bekenntnisse: Selbsthematisierung und Zivilisationsprozeß”, Kölner Zeit-
schrift für Soziologie und Sozialpsychologie, 34: 408–434.
Halfmann, J., and Japp, K.P. (1990) (eds.), Riskante Entscheidungen und Kata-
strophenpotentiale: Elemente einer soziologischen Risikoforschung (Opladen:
Westdeutscher Verlag).
Hamel G, and Prahalad, C.K. (1994), Competing for the Future (Boston, MA:
Harvard Business School Press).
Handy, C. (1990), The Age of Unreason (Boston: Harvard Business School).
Handy, C. (1994), The Age of Paradox (Boston: Harvard Business School).
Handy, C. (1996), Beyond Certainty: The Changing Worlds of Organizations
(Boston: Harvard Business School).
Hannan, M.T., and Freeman, J.H. (1977), “The Population Ecology of
Organizations”, American Journal of Sociology, 82: 929–964.
Hannan, M.T., and Freeman, J.H. (1989), Organizational Ecology (Cambridge,
MA: Harvard University Press).
Hardt, M., and Negri, A. (2000), Empire (Cambridge, MA: Harvard University
Press).
Harrison, J.R., and March, J.G. (1984), “Decision Making and Post-Decision
Surprises”, Administrative Science Quarterly, 29: 26–42.
Hassard, J., and Parker, M. (1993) (eds.), Postmodernism and Organizations
(London: Sage).
Hassard, J., et al. (2000), Body and Organization (London: Sage).
Hasse, R. (2003a), Die Innovationsfähigkeit der Organisationsgesellschaft:
Organisation, Wettbewerb und sozialer Wandel (Opladen: Westdeutscher Verlag).
Hasse, R. (2003b), “Organization and Change in Information Societies – Net-
working, Fragmentation and the Problem of Uncertainty Absorption”, in
G. Bechmann et al. (2003) (eds.), Across the Divide (Berlin: Edition Sigma),
177–196.
Hasse, R., and Japp, K.P. (1997), “Dynamik symbolischer Organisationspolitik”, in
M. Birke et al. (1997) (eds.), Handbuch Umweltschutz und Organisation
(München: Oldenbourg), 134–162.
Hasse, R., and Krücken, G. (1999), Neo-Institutionalismus (Bielefeld: Transcript).
Hatch, M.J. (1993): “Spontaneous Humour as an Indicator of Paradox and
Ambiguity”, Organization Studies, 14/4 (Berlin).
Haunschild, A. (2003), “Humanization Through Discipline? Foucault and the
Goodness of Employee Health Programmes”, Tamara: The Journal of Critical
Postmodern Organization Science, 3/1: 46–59.
435
24 Referenser 05-11-11 16.33 Sida 436
Havelock, E.A. (1982), The Literate Revolution in Greece and its Cultural
Consequences (Princeton: Princeton University Press).
Hawkins, K. (1984), Environment and Enforcement: Regulation and the Social
Definition of Pollution (Oxford: Oxford University Press).
Hayek, F.A. (1980), Individualism and Economic Order (Chicago: Chicago
University Press).
Hedberg, B.L.T., Nystrom, P.C., and Starbuck, W.H. (1976), “Camping on Seesaws:
Prescriptions for a Self-Designing Organization”, Administrative Science
Quarterly, 21: 41–65.
Hedlin, P. (1996), Accounting Investigations (Doctoral Thesis, University of
Stockholm).
Heider, F. (1959), “Thing and Medium”, Psychological Issues, 1: 1–34.
Heiner, R.A. (1986), “Imperfect Decisions and the Law: On the Evolution of Legal
Precedents and Rules”, The Journal of Legal Studies, 15: 227–261.
Hejl, P.M. (1982), Sozialwissenschaft als Theorie selbstreferentieller Systeme
(Frankfurt: Campus).
Hendry J., and Seidl, D. (2003), “The Structure and Significance of Strategic
Episodes: Social Systems Theory and the Routine Practices of Strategic Change”,
Journal of Management Studies, 40/1: 175–196.
Henriques, I., and Sadorsky, P. (1999), “The Relationship Between Environmental
Commitment and Managerial Perceptions of Stakeholder Importance”, Academy
of Management Review, 42/1: 87–99.
Hessinger, P., et al. (2000), Fokus und Balance: Aufbau und Wachstum industrieller
Netzwerke (Opladen: Westdeutscher Verlag).
Heuskel, D. (1999), Wettbewerb jenseits von Industriegrenzen: Aufbruch zu neuen
Wachstumsstrategien (Frankfurt am Main: Campus).
Hilliard, D. (2005), Vom Unternehmen zum Weltunternehmen: Soziologische
Analysen, unpubl. Ph.D. thesis (Bielefeld).
Hofer, C.W., and Schendel, D. (1978), Strategy Formulation: Analytical Concepts
(New York: West).
Hofstadter, D.R. (1979), Gödel, Escher, Bach: An Eternal Golden Braid (Hassocks,
Sussex: Harvester Press).
Hofstadter, D.R. (1980), Gödel, Escher, Bach: An Eternal Golden Braid (New York:
Vintage Books).
Holland, R. (1977), Self and Social Context (New York: St Martin’s Press).
Hooker, C.A. (1975), “On Global Theories”, Philosophy of Science, 42: 152–79.
Hopwood, A. (1983), “On Trying to Study Accounting in the Context in Which it
Operates”, Accounting, Organizations and Society, 8/2–3: 287–305.
Horváth, P. (2003), Controlling (9th edn., München: Vahlen).
Howard, N. (1971), Paradoxes of Rationality: Theory of Metagames and Political
Behaviour (Cambridge, Mass.: MIT Press).
Huet, P.D. (1723), Traité Philosophique de la Faiblesse de l’Esprit Humain (Amster-
dam: Du Sauzet).
Huisman, M. (2001), “Decision Making in Meetings as Talk-in-Interaction”,
International Studies of Management and Organization, 31: 69–90.
Husserl, E. (1913), Ideas Pertaining to a Pure Phenomenology and to a Phenom-
enological Philosophy: Book 1 (English trans., 1990, New York: Springer).
436
24 Referenser 05-11-11 16.33 Sida 437
Husserl, E. (1928), “Vorlesungen zur Phänomenologie des inneren Zeitbewußt-

seins”, Jahrbuch für Philosophie und phänomenologische Forschung, 9: 367–496.
Husserl, E. (1936), Die Krisis der europäischen Wissenschaften und die transzen-
dentale Phänomenologie: Eine Einleitung in die phänomenologische Philosophie,
ed. W. Biemel (1962) (2nd edn., Den Haag: M. Nijhoff).
Hymer, S.H. (1977), The International Operations of National Firms: A Study of
Direct Investment (Cambridge, MA: MIT Pr.).
Iellatchitch et al. (2003), “Career Fields: a Small Step Towards a Grand Career
Theory?”, International Journal of Human Resource Management, 14/5:
728–750.
Ingold, T. (1986), Evolution and Social Life (Cambridge, UK: Cambridge University
Press).
Jacobs, C., and Haeracleous, L. (2001), “Seeing without Being Seen: Towards an
Archaeology of Controlling Science”, International Studies of Management and
Organization, 31/3: 13–135.
Jansson, D. (1989), “The Pragmatic Uses of What is Taken for Granted: Project
Leaders’ Applications of Investment Calculus”, International Studies of
Management & Organization, 19/3: 49–63.
Jantsch, E. (1981), “Autopoiesis: A Central Aspect of Dissipative Self-organization”,
in M. Zeleny (1981) (ed.), Autopoiesis: A Theory of Living Organization (New
York: North Holland), 65–88.
Japp, K.P. (1992), “Selbstverstärkungseffekte riskanter Entscheidungen – Zur
Unterscheidung von Risiko und Rationalität”, Zeitschrift für Soziologie, 21:
31–48.
Jeannet, J.-P. Hennessey, and David, H. (1995), Global Marketing Strategies (3rd
edn., Boston, MA: Houghton Mifflin).
Jepperson, R.L. (1991), “Institutions, Institutional Effects, and Institutionalization”,
in W.W. Powell and DiMaggio, P.J. (1999) (eds.), The New Institutionalism in
Organizational Analysis (Chicago, IL: University of Chicago Press), 143–163.
Jepperson, R.L. (2001), The Development and Application of Sociological Neo-
institutionalism (Florence: EUI Working Paper RSC No. 2001/5).
Jönsson, S., and Macintosh, N.B. (1997), “CATS, RATS and EARS: Making the
Case for Ethnographic Accounting Research”, Accounting, Organizations and
Society, 22/4: 367–386.
Johnson, G. (1988), “Rethinking Incrementalism”, Strategic Management Journal,
9/1: 75–91.
Johnson, M. (1990), Business Buzzwords: The Tough New Jargon of Modern
Business (Oxford: Blackwell).
Jones, T.M. (1995), “Instrumental Stakeholder Theory: A Synthesis of Ethics and
Economics”, Academy of Management Review, 20/2: 404–437.
Jones, T.M., and Wicks, A.C. (1999), “Convergent Stakeholder Theory”, Academy
of Management Review, 24/2: 206–221.
Kämper, E., and Schmidt, J. (2000), “Netzwerke als strukturelle Kopplung:
Systemtheoretische Überlegungen zum Netzwerkbegriff”, in J. Weyer (2000) (ed.),
Soziale Netzwerke: Konzepte und Methoden der sozialwissenschaftlichen
Netzwerkforschung (München: Oldenbourg), 211–235.
Kaplan, R.S., and Norton, D.P. (1996), The Balanced Scorecard: Translating
Strategy into Action (Boston: Harvard Business School Press).
437
24 Referenser 05-11-11 16.33 Sida 438
Kaplan, R.S., Young, M.S., and Atkinson, A.A. (2003), Management Accounting
(New York: Prentice Hall).
Kauffmann, L.H. (1988), “Self-reference and Recursive Forms”, Journal of Social
and Biological Structures, 10: 53–72.
Keller, J. (2002), “Die Streitkräfte als Organisation: Was macht sie im Vergleich so
einzigartig?”, presentation given at workshop “Konsequenzen aus der
Wissensgesellschaft” at the Zentrum für Analysen und Studien der Bundeswehr,
Waldbröl, Germany, Feb. 2002: 19–20.
Kelly, A. (1992), “Revealing Bakhtin”, New York Review of Books, 24 [September
issue]: 44–48.
Kelly, K. (1990), Out of Control: The New Biology of Machines, Social Systems,
and the Economic World (Redwood City, CA: Addison-Wesley).
Kickert, W.J.M. (1993), “Autopoiesis and the Science of (Public) Administration:
Essence, Sense and Nonsense”, Organization Studies, 14: 261–278.
Kieser, A. (1989), “Organization, Institutions, and Societal Evolution: Medieval
Craft Guilds and the Genesis of Formal Organization”, Administrative Science
Quarterly, 34: 540–564.
Kieserling, A. (1994), “Interaktion in Organisationen”, in K. Dammann, D.
Grunow, and K.P. Japp (eds.), Die Verwaltung des politischen Systems: Neue sys-
temtheoretische Zugriffe auf ein altes Thema (Opladen: Westdeutscher Verlag),
168–182.
Kieserling, A. (1999), Kommunikation unter Anwesenden: Studien über
Interaktionssysteme (Frankfurt am Main: Suhrkamp).
Kind, H. (1986), “Gibt es spezifische Faktoren in der Psychotherapie?”, Praxis der
Psychotherapie und der Psychosomatik, 31: 191–196.
Kirsch, W. (1992), Kommunikatives Handeln, Autopoiese, Rationalität: Sondier-
ungen zu einer evolutonären Führungslehre (München: Kirsch).
Kirsch, W., and Knyphausen, D. zu (1991), “Unternehmungen als ‘autopoietische’
Systeme?”, in W.H. Staehle and J. Sydow (1991) (eds.), Managementforschung I
(Berlin: de Gruyter), 75–101.
Kneer, G. (2001), “Organisation und Gesellschaft: Zum ungeklärten Verhältnis von
Organisations- und Funktionssystemen in Luhmanns Theorie sozialer Systeme”,
Zeitschrift für Soziologie, 30: 407–428.
Knights, D. (1992), “Changing Spaces: The Disruptive Impact of a New
Epistemological Location for the Study of Management”, Academy of
Management Review, 17/3: 514–536.
Knights, D., and Morgan, G. (1991), “Corporate Strategy, Organizations, and
Subjectivity: A Critique”, Organization Studies, 12/2: 251–273.
Knights, D., and Willmott, H.C. (1987), “Organizational Culture as Management
Strategy”, International Studies of Management and Organization, 17/3: 40–63.
Kondra, A.Z., and Hinings, C.R. (1998), “Organizational Diversity and Change in
Institutional Theory”, Organization Studies, 19: 743–767.
Korzybski, A. (1949), Science and Sanity: An Introduction to Non-Aristotelian
Systems and General Semantics (3rd ed., Lakeville, Conn.: Institute of General
Semantics).
Kracauer, S. (1998), The Salaried Masses: Duty and Distraction in Weimar Germany
(New York: Verso).
438
24 Referenser 05-11-11 16.33 Sida 439
Krippendorff, K. (1984), “Paradox and Information”, in B. Derwin and M.J. Voigt

(1984) (eds.), Progress in Communication, vol. 5, 45–71.
Kurland, N.B., and Pelled, L.H. (2000), “Passing the Word: Toward a Model of
Gossip and Power in the Workplace”, Academy of Management Review, 25:
428–438.
Küpper, H.-U. (1987), “Konzeption des Controlling aus betriebswirtschaftlicher
Sicht”, in A.-W. Scheer (1987) (ed.), Rechnungswesen und EDV: 8. Saarbrücker
Arbeitstagung (Heidelberg: Physica), 82–116.
Küpper, H.-U. (1997), Controlling: Konzeption, Aufgaben und Instrumente (2nd
edn., Stuttgart: Schaeffer-Poeschel).
Küpper, H.-U., Weber, J., and Zünd, A. (1990), “Zum Verständnis und Selbstver-
ständnis des Controlling: Thesen zur Konsensbildung”, Zeitschrift für Betriebs-
wirtschaft, 60/2: 281–293.
Lacan, J. (1977), Écrits: A Selection (London: Tavistock).
Ladeur, K.-H. (1992), Postmoderne Rechtstheorie: Selbstreferenz – Selbstorganisa-
tion – Prozeduralisierung (Berlin: Duncker und Humblot).
Lando, O. (1545), Paradossi, cioe sententie fuori del commun parere (Vinegia:
Bernardino Bindoni).
Lando, O. (n.d.), Confutatione del libro de paradossi nuovamente composta, in tre
orationi distinta (n.p.).
Lanzara, G.F., and Pardi, F. (1980), L’interpretazione della complessità: Metodo sis-
temico e scienze sociali (Naples: Guida).
Latour, B. (1988a), “The Politics of Explanation: An Alternative”, in S. Woolgar
(1988) (ed.), Knowledge and Reflexivity: New Frontiers in the Sociology of
Knowledge (London: Sage), 155–176.
Latour, B. (1988b), “A Relativistic Account of Einstein’s Relativity”, Social Studies
of Science, 18: 3–44.
Latour, B. (1993), “Pasteur on Lactic Acid Yeast: A Partial Semiotic Analysis”,
Configurations, 1/1: 129–146.
Laudan, L. (1977), Progress and its Problems: Toward a Theory of Scientific
Growth (London: Routledge & Kegan Paul).
Laudan, L. (1984), Science and Values: The Aims of Science and their Role in
Scientific Debate (Berkeley: University of California Press).
Laughlin, R.C. (1987), “Accounting Systems in Organizational Contexts: A Case for
Critical Theory”, Accounting, Organizations and Society, 12/5: 479–502.
Lawrence, P.R., and Lorch, J.W. (1967), Organization and Environment: Managing
Differentiation and Integration (Boston, MA: Harvard University Press).
Lawson, H. (1985), Reflexivity: The Post-Modern Predicament (London: Open
Court Pub Co.).
Leach, E. (1982), Social Anthropology (Glasgow: Oxford University Press).
Lenk, H., and Ropohl, G. (1978), Systemtheorie als Wissenschaftsprogramm
(Königstein: Athenäum-Verlag) [esp. introduction].
Lenk, H., Maring, M., and Fulda, E. (1985), “Wissenschaftstheoretische Aspekte
einer anwendungsorientierten sytemtheoretischen Betriebswirtschaftlehre”, in
G.J.B. Probst and H. Siegwart (1985) (eds.), Integriertes Management: Bausteine
des systemorientierten Managements (Berne: Haupt), 165–178.
Leonard-Barton, D. (1995), Wellsprings of Knowledge: Building and Sustaining the
Sources of Innovation (Boston, MA: Harvard Business School Press).
439
24 Referenser 05-11-11 16.33 Sida 440
Lessard, D.R. (1986), “Finance and Global Competition: Exploiting Financial Scope
and Coping with Volatile Exchange Rates”, in M.E. Porter (1986) (ed.),
Competition in Global Industries (Boston, MA: Harvard Business School Pr.),
147–184.
Lévi-Strauss, C. (1964), Mythologiques: Le Cru et le Cuit (Paris: Plon).
Lévi-Strauss, C. (1968), Structural Anthropology (London: A Lane).
Levitt, B., and Nass, C. (1989), “The Lid on the Garbage Can: Institutional
Constraints on Decision Making in the Technical Core of College-Text-
Publishers”, Administrative Science Quarterly, 34/2: 190–207.
Lieckweg, T. (2001), “Strukturelle Kopplung von Funktionssystemen ‘über’ Orga-
nisation”, Soziale Systeme, 7/2: 267–289.
Linton, R. (1964), The Study of Man (Englewood Cliffs: Prentice Hall).
Löfgren, L. (1978), “Some Foundational Views on General Systems and the Hempel
Paradox”, International Journal of General Systems, 4: 243–253.
Löfgren, L. (1979), “Unfoldment of Self-Reference in Logic and Computer Science”,
in: Finn V. Jensen, Brian H. Mayoh, and Karen K. Møller (1979) (eds.),
Proceedings of the 5th Scandinavian Logic Symposium, Aalborg, 17–19 January
1979 (Aalborg: Institut for Elektroniske Systemer) 205–229.
Lotmann, J. (1990), “The Symbol in the Cultural System”, in J. Lotmann (1990),
Universe of the Mind: A Semiotic Theory of Culture (Bloomington: Indiana
University Press), 102–111.
Luhmann, N. (1964), Funktionen und Folgen formaler Organisation (Berlin:
Duncker & Humblot).
Luhmann, N. (1965), Grundrechte als Institution: Ein Beitrag zur politischen
Soziologie (Berlin: Duncker & Humblot).
Luhmann, N. (1968), Zweckbegriff und Systemrationalität: Über die Funktion von
Zwecken in sozialen Systemen (Tübingen: Mohr/Siebeck).
Luhmann, N. (1970), “Gesellschaft”, in vol. 1 of N. Luhmann (1970), Soziologische
Aufklärung (Opladen: Westdeutscher Verlag), 137–153.
Luhmann, N. (1971a), Politische Planung: Aufsätze zur Soziologie von Politik und
Verwaltung (Opladen: VS Verlag für Sozialwissenschaften).
Luhmann, N. (1971b), “Zweck-Herrschaft-System: Grundbegriffe und Prämissen
Max Webers”, in R. Mayntz (1971) (ed.), Bürokratische Organisation (Köln:
Kiepenheuer & Witsch), 36–55.
Luhmann, N. (1973), Zweckbegriff und Systemrationalität: Über die Funktion von
Zwecken in sozialen Systemen (Frankfurt am Main: Suhrkamp).
Luhmann, N. (1974), Soziologische Aufklärung: Aufsätze zur Theorie sozialer
Systeme, Band 1, 4. Auflage ([vol. 1, 4th edn.] Opladen: Westdeutscher Verlag).
Luhmann, N. (1975a), “Die Weltgesellschaft”, in vol. 2 of N. Luhmann (1970),
Soziologische Aufklärung (Opladen: Westdeutscher Verlag), 51–71.
Luhmann, N. (1975b), “Interaktion, Organisation, Gesellschaft”, in N. Luhmann
(1975), Soziologische Aufklärung 2: Aufsätze zur Theorie der Gesellschaft
(Opladen: Westdeutscher Verlag), 9–20.
Luhmann, N. (1976a), “A General Theory of Organized Social Systems”, in G. Hof-
stede and M.S. Kassem (1976) (eds.), European Contributions to Organization
Theory (Assen: Van Gorcum), 96–113.
440
24 Referenser 05-11-11 16.33 Sida 441
Luhmann, N. (1976b), “Generalized Media and the Problem of Contingency”, in

J.J. Loubser et al. (1976) (eds.), Explorations in General Theory in Social Science:
Essays in Honor of Talcott Parsons, vol. 2 (New York: The Free Press), 507–532.
Luhmann, N. (1977a), “Differentiation of Society”, Canadian Journal of Sociology,
2: 29–53.
Luhmann, N. (1977b), Zweckbegriff und Systemrationalität: Über die Funktion von
Zwecken in sozialen Systemen (repr., Frankfurt am Main: Suhrkamp).
Luhmann, N. (1978), “Soziologie der Moral”, in N. Luhmann and St. H. Pfürtner,
(1978), Theorietechnik und Moral (Frankfurt am Main: Suhrkamp), 8–116.
Luhmann, N. (1979), Trust and Power: Two Works by Niklas Luhmann
(Chichester: Wiley).
Luhmann, N. (1980a), “Komplexität”, in E. Grochla (1980) (ed.), Handwörterbuch
der Organisation (Stuttgart: Poeschel), col. 1064–1070.
Luhmann, N. (1980b), “Temporalisierung von Komplexitat: Zur Semantik neuzeit-
licher Zeitbegriffe”, in N. Luhmann (1980), Gesellschaftsstruktur und Semantik,
vol. I (Frankfurt: Suhrkamp), 235–300.
Luhmann, N. (1981a), “Organisation und Entscheidung”, in N. Luhmann (1981)
(ed.), Soziologische Aufklärung 3 (Opladen: VS Verlag für Sozialwissenschaften),
335–389.
Luhmann, N. (1981b), Soziologische Aufklärung 3 (Opladen: Westdeutscher
Verlag), 335–414.
Luhmann, N. (1981c), “Symbiotische Mechanismen”, in N. Luhmann (1981),
Soziologische Aufklärung 3 (Opladen: Westdeutscher Verlag), 228–244.
Luhmann, N. (1981d), “The Improbability of Communication”, International
Social Science Journal, 23(1): 122–132.
Luhmann, N. (1982a), “Durkheim on Morality and the Division of Labor”, in N.
Luhmann (1982), The Differentiation of Society, trans. S. Holmes and C.
Larmore (New York: Columbia University Press), 3–19.
Luhmann, N. (1982b), “Ends, Domination and System”, in N. Luhmann (1982),
The Differentiation of Society, trans. S. Holmes and C. Larmore (New York:
Columbia University Press), 20–46.
Luhmann, N. (1982c), “Interaction, Organization, and Society”, in N. Luhmann
(1982), The Differentiation of Society, trans. S. Holmes and C. Larmore (New
York: Columbia University Press), 69–89.
Luhmann, N. (1982d), “Systems Theory, Evolutionary Theory, and Communication
Theory”, in N. Luhmann (1982), The Differentiation of Society, trans. S. Holmes
and C. Larmore (New York: Columbia University Press), 255–270.
Luhmann, N. (1982e), “Territorial Borders as Systems Boundaries”, in R. Strassoldo
and G.D. Zotti (1982) (eds.), Cooperation and Conflict in Border Areas (Milano:
Angeli), 235–244.
Luhmann, N. (1982f), The Differentiation of Society (New York: Columbia Univ.
Pr.).
Luhmann, N. (1982g), “The Future Cannot Begin: Temporal Structures in Modern
Society”, in N. Luhmann (1982), The Differentiation of Society (New York:
Luhmann, N. (1982h), “The Self-Thematization of Society”, in N. Luhmann (1982),
The Differentiation of Society, trans. S. Holmes and C. Larmore (New York:
441
24 Referenser 05-11-11 16.33 Sida 442
Luhmann, N. (1982i), “World-Time and System History: Interrelations between

Temporal Horizons and Social Structures”, in N. Luhmann (1982), The
Differentiation of Society (New York: Columbia University Press), 289–323.
Luhmann, N. (1983a), Rechtssoziologie (2nd ed., Opladen: Westdeutscher Verlag).
Luhmann, N. (1983b), “Die Einheit des Rechtssystems”, Rechtstheorie, 14: 129–54.
Luhmann, N. (1984a), “Soziologische Aspekte des Entscheidungsverhaltens”, Die
Betriebswirtschaft, 44: 591–603.
Luhmann, N. (1984b), Soziale Systeme: Grundriß einer allgemeinen Theorie, 2.
Auflage ([2nd edn.] Frankfurt am Main: Suhrkamp).
Luhmann, N. (1985), “Society, Meaning, Religion – Based on Self-Reference”,
Sociological Analysis, 46: 5–20.
Luhmann, N. (1986a), Love as Passion: The Codification of Intimacy (Cambridge:
Polity Press).
Luhmann, N. (1986b), “The Autopoiesis of Social Systems”, in F. Geyer and J. Van
der Zouwen (eds.), Sociocybernetic Paradoxes: Observation, Control and
Evolution of Self-steering Systems (London: Sage), 172–192 [reprinted in the pres-
ent volume].
Luhmann, N. (1987a), “Biographie, Attitüden, Zettelkasten”, in N. Luhmann
(1987) Archimedes und wir (Berlin: Merve Verlag), 125–155.
Luhmann, N. (1987b), “Die Differenzierung von Politik und Wirtschaft und ihre
gesellschaftlichen Grundlagen”, in N. Luhmann (1987), Soziologische Aufklärung
4 (Opladen: Westdeutscher Verlag), 32–48.
Luhmann, N. (1988a), Die Wirtschaft der Gesellschaft (Frankfurt am Main: Suhr-
kamp).
Luhmann, N. (1988b), Erkenntnis als Konstruktion (Bern: Benteli).
Luhmann, N. (1988c), “Organisation”, in W. Küpper and G. Ortmann (1988)
(eds.), Mikropolitik: Rationalität, Macht und Spiele in Organisationen (Opladen:
Westdeutscher Verlag), 165–185 [Trans. in T. Hernes and T. Bakken (2003) (eds.),
Autopoietic Organization Theory: Drawing on Niklas Luhmann’s Social System
Perspective (Oslo: Abstrakt Forlag)].
Luhmann, N. (1988d), “The Third Question: The Creative Use of the Paradoxes in
Law and Legal Theory”, Law and Society Review, 15: 153–165.
Luhmann, N. (1989a), Ecological Communication (Cambridge: Polity Press).
Luhmann, N. (1989b), “Individuum, Individualität, Individualismus”, in N. Luh-
mann (1989), Gesellshaftsstruktur und Semantik: Studien zur Wissenssoziologie
der modernen Gesellschaft, vol. 3 (Frankfurt am Main: Suhrkamp), 149–258.
Luhmann, N. (1990a), “Autopoiesis of Social Systems”, in N. Luhmann (1990)
(ed.), Essays on Self-Reference (New York and Oxford: Oxford University Press),
1–20 [reprinted in the present volume].
Luhmann, N. (1990b), Complexity and Meaning, in N. Luhmann (1990) (ed.),
Essays on Self-Reference (New York and Oxford: Oxford University Press),
80–85.
Luhmann, N. (1990c), Die Wissenschaft der Gesellschaft (Frankfurt am Main:
Suhrkamp).
Luhmann, N. (1990d), Essays on Self-Reference (New York: Columbia University
Press).
Luhmann, N. (1990e), “Meaning as Sociology’s Basic Concept”, in N. Luhmann
(1990), Essays on Self-Reference (New York: Columbia University Press), 21–79.
442
24 Referenser 05-11-11 16.33 Sida 443
Luhmann, N. (1990f), “Modes of Communication and Society”, in N. Luhmann

(1990), Essays on Self-Reference (New York: Columbia University Press),
99–106.
Luhmann, N. (1990g), Soziologische Aufklärung 5: Konstruktivistische
Perspektiven (Opladen: Westdeutscher Verlag).
Luhmann, N. (1990h), “Tautology and Paradox in the Self-Descriptions of Modern
Society”, in N. Luhmann (1990) (ed.), Essays on Self-Reference (New York and
Oxford: Oxford University Press), 123–143.
Luhmann, N. (1990i), “The Cognitive Program of Constructivism and a Reality that
Remains Unknown”, in W. Krohn, G. Küppers, and H. Nowotny (1990) (eds.),
Selforganization: Portrait of a Scientific Revolution (Dordrecht: Kluwer), 64–85.
Luhmann, N. (1990j), “The Improbability of Communication”, in N. Luhmann
(1990), Essays on Self-Reference (New York: Columbia University Press), 86–98.
Luhmann, N. (1990k), “The Medium of Art”, in N. Luhmann (1990), Essays on
Self-Reference (New York: Columbia University Press), 215–226.
Luhmann, N. (1990l), “The State of the Political System”, in N. Luhmann (1990),
Essays on Self-Reference (New York: Columbia University Press), 165–174.
Luhmann, N. (1990m), “The World Society as a Social System”, in N. Luhmann
(1990), Essays on Self-Reference (New York and Oxford: Columbia Univ. Press),
175–190.
Luhmann, N. (1991a), “Sthenographie und Eurialistik” in H.U. Gumbrecht and
K.L. Pfeiffer (1991) (eds.), Paradoxien, Dissonanzen, Zusammenbrüche:
Situationen offener Epistemologie (Frankfurt am Main: Suhrkamp), 58–82.
Luhmann, N. (1991b), “Wie lassen sich latente Strukturen beobachten?”, in
P. Watzlawick and P. Krieg (1991) (eds.), Das Auge des Betrachters – Beiträge
zum Konstruktivismus: Festschrift für Heinz von Foerster (Munich: Carl-Auer-
Systeme), 61–74.
Luhmann, N. (1992a), “Organisation”, in W. Küpper and G. Ortmann (1992)
(eds.), Mikropolitik: Rationalität, Macht und Spiele in Organisationen (2nd edn.,
Opladen: Westdeutscher Verlag), 165–185.
Luhmann, N. (1992b), “Politische Steuerungsfähigkeit eines Gemeinwesens”,
unpublished manuscript, Universität Bielefeld.
Luhmann, N. (1992c), “The Concept of Society”, Thesis Eleven, 31: 67–80.
Luhmann, N. (1993a), Das Recht der Gesellschaft (Frankfurt am Main: Suhrkamp).
Luhmann, N. (1993b), “Deconstruction as Second-Order Observing”, New Literary
History, 24: 763–782.
Luhmann, N. (1993c), “Die Paradoxie der Form”, in Baecker, D. (1993) (ed.),
Kalkül der Form (Frankfurt am Main: Suhrkamp).
Luhmann, N. (1993d), “Die Paradoxie des Entscheidens”, Verwaltungs-Archiv:
Zeitschrift für Verwaltungslehre, Verwaltungsrecht und Verwaltungspolitik, 84:
287–310 [see Chapter 4 of the present volume for a translation of the article].
Luhmann, N. (1993e), Gesellschaftsstruktur und Semantik, vol. 3 (Frankfurt am
Main: Suhrkamp).
Luhmann, N. (1993f), “Haltlose Komplexität”, in N. Luhmann (1993) (ed.),
Soziologische Aufklärung 5: Konstruktivistische Perspektiven (2nd edn., Opladen:
Westdeutscher Verlag), 59–76.
Luhmann, N. (1993g), “Observing Re-Entries”, Graduate Faculty Philosophy
Journal, 16/2: 485–498.
443
24 Referenser 05-11-11 16.33 Sida 444
Luhmann, N. (1993h), “Organisation und Entscheidung” in N. Luhmann (1993)

(ed.), Soziologische Aufklärung 3: Soziales System, Gesellschaft, Organisation
(3rd edn., Opladen: Westdeutscher Verlag), 335–389.
Luhmann, N. (1993i), Risk: A Sociological Theory (New York: de Gruyter).
Luhmann, N. (1993j), Soziologische Aufklärung 3: Soziales System, Gesellschaft,
Organisation (3rd edn., Opladen: Westdeutscher Verlag).
Luhmann, N. (1993k), Soziologische Aufklärung 5: Konstruktivistische Perspek-
tiven (Opladen: Westdeutscher Verlag).
Luhmann, N. (1994a), “Die Gesellschaft und ihre Organisationen”, in H.-U. Derlien
et al. (eds.), Systemrationalität und Partialinteresse: Festschrift für Renate Mayntz
(Baden-Baden: Nomos), 189–201.
Luhmann, N. (1994b), “Observing Re-Entries”, Protosoziologie, 6: 4–13.
Luhmann, N. (1995a), “Die Form Person” in N. Luhmann (1995), Soziologische
Aufklärung 6: Die Soziologie und der Mensch (Opladen: Westdeutscher Verlag),
142–154.
Luhmann, N. (1995b), “Inklusion und Exklusion”, in N. Luhmann (1995), Sozio-
logische Aufklärung 6: Die Soziologie und der Mensch (Opladen: Westdeutscher
Verlag), 237–264.
Luhmann, N. (1995c), “Instead of a Preface to the English Edition: On the Concepts
‘Subject’ and ‘Action’”, in N. Luhmann (1995), Social Systems (Stanford: Stan-
ford University Press), xxxvii–xliv.
Luhmann, N. (1995d), “Kausalität im Süden: Soziale Systeme”, Zeitschrift für
Soziologische Theorie, 1: 7–28.
Luhmann, N. (1995e), “Probleme mit operativer Schließung”, in N. Luhmann
(1995) (ed.), Soziologische Aufklärung 6: Die Soziologie und der Mensch
(Opladen: Westdeutscher Verlag), 12–24.
Luhmann, N. (1995f), Social Systems, trans. J. Bednarz, Jr. and D. Baecker (Stan-
ford: Stanford University Press).
Luhmann, N. (1995g), Soziologische Aufklärung 6: Die Soziologie und der Mensch
(Opladen: Westdeutscher Verlag).
Luhmann, N. (1995h), “The Paradox of Observing Systems”, Cultural Critique, 30
[Fall issue]: 37–55.
Luhmann, N. (1995i), “Why Does Society Describe Itself as Postmodern?”, Cultural
Critique, 30 [Spring issue]: 171–186.
Luhmann, N. (1995j), “Wie ist Bewusstsein an Kommunikation beteiligt?”, in
N. Luhmann (1995) (ed.), Soziologische Aufklärung 6: Die Soziologie und der
Mensch (Opladen: Westdeutscher Verlag), 37–54.
Luhmann, N. (1996a), “Membership and Motives in Social Systems”, Systems
Research, 13: 341–348.
Luhmann, N. (1996b), “Complexity, Structural Contingencies and Value Conflicts”,
in P. Heelas et al. (1996) (eds.), Detraditionalization (Oxford: Blackwell), 59–71.
Luhmann, N. (1997a), Die Gesellschaft der Gesellschaft, vols. 1 and 2 (Frankfurt
am Main: Suhrkamp).
Luhmann, N. (1997b), “The Control of Intransparency”, System Research and
Behavioral Science, 14/4: 358–371.
Luhmann, N. (1998), Observations on Modernity, trans. W. Whobrey (Stanford,
CA: Stanford University Press).
444
24 Referenser 05-11-11 16.33 Sida 445
Luhmann, N. (1999a), Gesellschaftsstruktur und Semantik, vol. 4 (Frankfurt am

Main: Suhrkamp).
Luhmann, N. (1999b), “Kommunikation über Recht in Interaktionssystemen”, in
N. Luhmann (1999) (ed.), Ausdifferenzierung des Rechts: Beiträge zur Rechts-
soziologie und Rechtstheorie (Frankfurt am Main: Suhrkamp).
Luhmann, N. (2000a), Art as Social System (Stanford: Stanford University Press).
Luhmann, N. (2000b), Die Politik der Gesellschaft (Frankfurt am Main:
Suhrkamp).
Luhmann, N. (2000c), Organisation und Entscheidung (Opladen and Wiesbaden:
Westdeutscher Verlag).
Luhmann, N. (2000d), Die Religion der Gesellschaft (Frankfurt am Main:
Suhrkamp).
Luhmann, N. (2002a), Das Erziehungssystem der Gesellschaft (Frankfurt am Main:
Suhrkamp).
Luhmann, N. (2002b), Einführung in die Systemtheorie (Heidelberg: Carl-Auer-
Systeme Verlag).
Luhmann, N. (2002c), “I See Something You Don’t See”, in W. Rasch (2002) (ed.),
Theories of Distinction: Redescribing the Descriptions of Modernity (Stanford,
CA: Stanford University Press), 187–193.
Luhmann, N. (2002d), “Identity – What or How?”, in W. Rasch (2002) (ed.),
Theories of Distinction: Redescribing the Descriptions of Modernity (Stanford,
CA: Stanford University Press), 113–127.
Luhmann, N. (2002e), “The Cognitive Program of Constructivism and the Reality
that Remains Unknown”, in W. Rasch (2002) (ed.), Theories of Distinction:
Redescribing the Descriptions of Modernity (Stanford, CA: Stanford University
Press), 128–154.
Luhmann, N. (2002f), “The Modern Sciences and Phenomenology”, in W. Rasch
(2002) (ed.), Theories of Distinction: Redescribing the Descriptions of Modernity
(Stanford: Stanford University Press), 33–75.
Luhmann, N. (2002g), “What is communication?”, in W. Rasch (2002) (ed.),
Theories of Distinction: Redescribing the Descriptions of Modernity (Stanford:
Stanford University Press), 155–168.
Luhmann, N. (2003), “Organization”, in T. Bakken and T. Hernes (2003) (eds.),
Autopoietic Organization Theory: Drawing on Niklas Luhmann’s Social Systems
Perspective (Oslo/Malmö/Copenhagen: Abstrakt/Liber/Copenhagen Business
School Press), 31–52.
Luhmann, N. (2004), Law as a Social System, trans. K.A. Ziegert (Oxford: Oxford
University Press) (first pub. in German 1993).
Luhmann, N. (2006), “System as Difference”, Organization, 13/1 [forthcoming].
Luhmann, N., and Schorr, K.E. (1982), “Das Technologiedefizit der Erziehung und
die Pädagogik”, in N. Luhmann and K.E. Schorr (1982) (eds.), Zwischen
Technologie und Selbstreferenz: Fragen an die Pädagogik (Frankfurt am Main:
Suhrkamp), 11–40.
Lupi, W.F. (1992), “Ars Perplexitatis: Etica e retorica del discorso paradossale”, in
R. Genovese (1992) (ed.), Figure del Paradosso (Naples: Liguori), 29–59.
Lynch, M. (1993), Scientific Practice and Ordinary Action: Ethnomethodology and
Social Studies of Science (Cambridge: Cambridge University Press).
445
24 Referenser 05-11-11 16.33 Sida 446
Lyotard, J.-F. (1979/1986), The Postmodern Condition: A Report on Knowledge

(Manchester: Manchester University Press).
Lyotard, J.-F. (1984), The Postmodern Condition: A Report on Knowledge
(Minneapolis: University of Minnesota Press).
Lyotard, J.-F. (1988), The Differend: Phrases in Dispute (Minneapolis: University of
Minnesota Press).
Lüscher, R.M. (1988), Henry und die Krümelmonster: Versuch über den fordistis-
chen Sozialscharakter (Tübingen: Gehrke).
Macintosh, N.B. (1994), Management Accounting and Control Systems: An
Organizational and Behavioral Approach (Chichester: John Wiley & Sons).
Macintosh, N.B., and Scapens, A.W. (1990), “Structuration Theory in Management
Accounting”, Accounting, Organizations and Society, 15/5: 455–477.
MacIntyre, A. (1981/1990), After Virtue (London: Duckworth Press).
Mahnkopf, B. (1994), “Markt, Hierarchie und soziale Beziehungen: Zur Bedeutung
reziproker Marktgesellschaften”, in N. Beckenbach and W. van Treeck (1994)
(eds.), Umbrüche gesellschaftlicher Arbeit (Göttingen: Schwartz) [Soziale Welt,
Special Issue 9: 65–84].
Malinowski, B. (1960), “The Problem of Meaning in Primitive Language”, in C. K.
Ogden and J. A. Richards (1960) (eds.), The Meaning of Meaning: A Study of the
Influence of Language upon Thought and the Science of Symbolism (10th ed.,
London: Routledge & Kegan Paul), 296–336.
Malloch, A.E. (1956), “The Technique and Function of the Renaissance Paradox”,
Studies in Philology, 52: 191–203.
Malsch, T. (1987), “Die Informationsierung des betrieblichen Erfahrungswissens
und der Imperialismusder instrumentellen Vernunft”, Zeitschrift für Soziologie,
16: 77–91.
Mandeville, B. (1924), The Fable of the Bees: Or, Private Vices, Publick Benefits
(Oxford : Clarendon Press) (first pub. 1714).
Manning, P.K. (1992), Organizational Communication (New York: Aldine de
Gruyter).
March, J.G. (1988), Decisions and Organizations (Oxford, UK: Blackwell).
March, J.G. (1996), “Understanding How Decisions Happen in Organizations”, in
J.G. March (1996), The Pursuit of Organizational Intelligence (London:
Blackwell), 13–38.
March, J.G., and Olsen, J. (1989), Rediscovering Institutions: The Organizational
Basis of Politics (New York: The Free Press).
March, J.G., and Sevón G. (1989), “Gossip, Information and Decision-Making”, in
J.G. March (1989) (ed.), Decision and Organization (Oxford: Blackwell),
429–442.
March, J.G., and Simon, H.A. (1958), Organizations (New York: Blackwell
Publishers).
March, J.G., and Simon, H.A. (1993), Organizations (2nd edn., Cambridge, MA:
Blackwell).
Markowitz, J. (1979), Die Soziale Situation (Frankfurt: Suhrkamp).
Marshall, T.H. (1950), Citizenship and Social Class – and Other Essays (Cambridge:
Cambridge University Press).
Martens, W. (2006), “The Distinctions within Organizations: Luhmann from a
Cultural Perspective”, Organization, 13/1 [forthcoming].
446
24 Referenser 05-11-11 16.33 Sida 447
Maturana, H. (1975), “The Organization of the Living: A Theory of the Living

Organization”, International Journal of Man-Machine Studies, 7: 313–332.
Maturana, H.R. (1978), “Cognition”, in P.M. Hejl et al. (1978) (eds.),
Wahrnehmung und Kommunikation (Frankfurt: Lang), 29–49.
Maturana, H.R. (1980), “Man and Society”, in F. Benseler, P.M. Hejl, and W. Köck
(1980) (eds.), Autopoiesis, Communication and Society: The Theory of Auto-
poietic System in the Social Sciences (Frankfurt: Campus), 11–31.
Maturana, H.R. (1981), “Autopoiesis”, in M. Zeleny (1981) (ed.), Autopoiesis: A
Theory of Living Organization (New York: North Holland), 21–30.
Maturana, H.R. (1983), “Reflexionen: Lernen oder ontogenetischer Drift”, Delfin
II: 60–72.
Maturana, H.R., and Varela, F.J. (1975), Autopoietic Systems: A Characterization
of the Living Organization (Urbana, IL: BCL Research Report).
Maturana, H.R., and Varela, F.J. (1980), Autopoiesis and Cognition: The
Realization of the Living (Dordrecht: Reidel).
Maturana, H.R., and Varela, F.J. (1987), The Tree of Knowledge: The Biological
Roots of Human Understanding (Boston: Shambhala).
Mauws, M.K., and Phillips, N. (1995), “Understanding Language Games”, Organi-
zation Science, 6/3: 322–334.
Mayntz, R. (1992), “Modernisierung und die Logik von interorganisatorischen
Netzwerken”, Journal für Sozialforschung, 32/1: 19–32.
McCleary, R. (1978), Dangerous Men: The Sociology of Parole (Beverly Hills:
Harrow and Heston).
McCulloch, W.S. (1965), Embodiments of Mind (Cambridge, MA: MIT Press).
McLuhan, M. (1964), Understanding Media (New York: McGraw-Hill).
McWilliams A., and Siegel, D. (2001), “Corporate Social Responsibility: A Theory
of the Firm Perspective”, Academy of Management Review, 26/1: 117–127.
Merton, R.K. (1957), Social Theory and Social Structure (2nd edn., Glencoe, IL: Free
Press).
Meyer, J.W., Boli, J., and Thomas, G.M. (1994), “Ontology and Rationalization in
the Western Cultural Account”, in W.R. Scott and J.W. Meyer (1994) (eds.),
Institutional Environments and Organizations: Structural Complexity and
Individualism (Thousand Oaks, CA: Sage), 9–26.
Meyer, J.W., et al. (1997), “World Society and the Nation-State”, American Journal
of Sociology, 103/1: 144–181.
Meyer, J.W., and Jepperson, R. (2000), “The ‘Actors’ of Modern Society: The
Cultural Construction of Social Agency”, Sociological Theory, 18: 100–120.
Meyer, J.W., and Rowan, B. (1977), “Institutionalized Organizations: Formal
Structures as Myth and Ceremony”, American Journal of Sociology, 83: 340–363.
Meyer, J.W., and Scott, W.R. (1983), Organizational Environments: Ritual and
Rationality (Beverly Hills, CA: Sage).
Meyer, M., and Zucker, L.G. (1989), Permanently Failing Organizations (Newbury
Park, CA: Sage).
Meyrowitz, J. (1985), No Sense of Place: The Impact of Electronic Media on Social
Behavior (New York: Oxford University Press).
Milgram, S. (1967), “The Small World Problem”, Psychology Today, 1: 61–67.
447
24 Referenser 05-11-11 16.33 Sida 448
Miller, P. (1994), “Accounting as Social and Institutional Practice: An Introduction”,

in A.G. Hopwood and P. Miller (1994) (eds.), Accounting as Social and
Institutional Practice (Cambridge: Cambridge University Press), 1–41.
Mills, C.W. (1940), “Situated Actions and Vocabularies of Motive”, American
Mingers, J. (1995), Self-Producing Systems: Implications and Applications of
Autopoiesis (New York: Plenum).
Mintzberg, H. (1979), The Structuring of Organizations: A Synthesis of the
Research (Englewood Cliffs, NJ: Prentice-Hall).
Mintzberg, H. (1990), “Strategy Formation: Schools of Thought”, in J. Frederickson
(1990) (ed.), Perspectives on Strategic Management (Boston: Ballinger).
Mintzberg, H. (1999), “Managing Quietly”, Leader to Leader, 12.
Mintzberg, H., and Waters, J.A. (1985), “Of Strategies, Deliberate and Emergent”,
Strategic Management Journal, 6/3: 257–272.
Moray, N. (1984), “Humans and Their Relations to Ill-Defined Systems”, in
O.G. Selfridge, E.L. Rissland, and M.A. Arbib (1984) (eds.), Adaptive Control of
Ill-Defined Systems (New York: Plenum Press), 11–20.
Morgan, G. (1980), “Paradigms, Metaphors, and Puzzle Solving in Organization
Theory”, Administrative Science Quarterly, 25: 605–622.
Morgan, G. (1986), Images of Organization (Beverly Hills, CA: Sage).
Morgan, G., and Willmott, H. (1993), “The ‘New’ Accounting Research: On
Making Accounting More Visible”, Accounting, Auditing and Accountability
Journal, 6/1: 3–36.
Mulkay, M., Pinch, T., and Ashmore, M. (1987), “Colonizing the Mind: Dilemmas
in the Application of Social Science”, Social Studies of Science, 17: 231–256.
Mullins, N.C. (1968), “The Distribution of Social and Cultural Properties in
Informal Communication Networks Among Biological Scientists”, American
Sociological Review, 33/5: 786–797.
Münch, R. (1982), Theorie des Handelns: Zur Rekonstruktion der Beiträge von
Talcott Parsons, Émile Durkheim und Max Weber (Frankfurt: Suhrkamp).
Nelson, R., and Winter, S. (1982), An Evolutionary Theory of Economic Change
(Cambridge, MA: Belknap Press).
Nicolai, A.T. (2000), Die Strategie-Industrie: Systemtheoretische Analyse des Zu-
sammenspiels von Wissenschaft, Praxis und Unternehmensberatung (Wiesbaden:
Deutscher Universitäts-Verlag).
“Notat om fremtidig ledelses- og samarbejdsstruktur, udkast den 11/2 1997”, Sund-
hedsforvaltningen (Memorandum on the Future Management and Cooperation
Structure sent out as a proposal by the health authorities 11/2/1997).
Nystrom, P.C., Hedberg, B.L.T., and Starbuck, W.H. (1976), “Interacting Processes
as Organization Designs” in R.H. Kilmann, L.R. Pondy, and D.P. Sleven (1976)
(eds.), The Management of Organization Design, vol. 1 (New York: North
Holland), 209–230.
“Nyt om Sundhedsvæsenet” no. 36-2001 (An official newsletter dated 2001, cata-
logue number 36).
Ohmae, K. (1995) (ed.), The Evolving Global Economy: Making Sense of the New
World Order (Boston, MA: Harvard Business School Press).
Ong, W.J. (1967), The Presence of the Word: Some Prolegomena for Cultural and
Religious History (New Haven, Conn.: Yale University Press).
448
24 Referenser 05-11-11 16.33 Sida 449
Orléan, A. (1992), “La monnaie et des paradoxes de l’individualisme”, Stanford

French Review, 15: 271–295.
Orliowski, W. (1992), “The Duality of Technology: Rethinking the Concept of
Technology in Organizations”, Organization Science, 3/3: 398–427.
Orliowski, W. (2000), “Using Technology and Constituting Structures: A Practice
Lens for Studying Technology in Organizations”, Organization Science, 11/4:
404–428.
Orliowski, W., and Robey, D. (1991), “Information Technology and the Structuring
of Organizations”, Information Systems Research, 2/2: 143–169.
Orr, J. (1996), Talking about Machines: An Ethnography of a Modern Job (Ithaca,
NY: Cornell University Press).
Ortmann, G., and Salzman, H. (2002), “Stumbling Giants: The Emptiness, Fullness,
and Recursiveness of Strategic Management”, Soziale Systeme, 8/2: 205–230.
Ortmann, G., Sydow, J., and Windeler, A. (1997), “Organisation als reflexive
Strukturation”, in G. Ortmann, J. Sydow, and K. Türk (1997) (eds.), Theorien der
Organisation: Die Rückkehr der Gesellschaft (Opladen: Westdeutscher Verlag),
315–354.
Parijs, P. van (1981), Evolutionary Explanation in the Social Sciences: An Emerging
Paradigm (London: Tavistock), 50.
Parsons, T. (1937), The Structure of Social Action (New York: McGraw-Hill).
Parsons, T. (1960), “A Sociological Approach to the Theory of Organizations”, in
T. Parsons (1960), Structure and Process in Modern Societies (New York: The
Free Press), 16–58.
Parsons, T. (1964), “Evolutionary Universals in Society”, American Sociological
Review, 29: 339–357.
Parsons, T. (1977), “Social Structure and the Symbolic Media of Interchange”, in
T. Parsons (1977), Social Systems and the Evolution of Action Theory (New York:
Free Press), 204–228.
Pask, G. (1970), “The Meaning of Cybernetics in the Behavioural Sciences (The
Cybernetics of Behaviour and Cognition; Extending the Meaning of ‘Goal’)”, in
J. Rose (1970) (ed.), Progress in Cybernetics, vol. 1 (London: Gordon & Breach),
15–44.
Pattee, H.H. (1982), “Cell Psychology: An Evolutionary Approach to the Symbol-
Matter Problem”, Cognition and Brain Theory, 5: 325–341.
Paul, J. (1961), “Clavis Fichtiana seu Leibgeberiana”, in vol. 3 of J. Paul (1961),
Werke (Munich: Saur), 1011–1056.
Perrow, C. (1989), “A Society of Organizations”, in M. Haller, H.-J. Hoffmann-
Nowotny, and W. Zapf (1989) (eds.), Kultur und Gesellschaft: Verhandlungen des
24. Deutschen Soziologentags, Zürich 1988 (Frankfurt am Main: Campus),
265–276.
Perrow, C. (1991), “A Society of Organizations”, Theory and Society, 20: 725–762.
Perrow, C. (2002), Organizing America: Wealth, Power, and the Origins of
Corporate Capitalism (Princeton: Princeton University Press).
Peters, T. (1993), Liberation Management: Necessary Disorganization for the
Nanosecond Nineties (London: Pan Books).
Peters, T. (1999), The Circle of Innovation: You Can’t Shrink Your Way to
Greatness (New York: Vintage).
449
24 Referenser 05-11-11 16.33 Sida 450
Petiteville, F. (2000), “Modernising the French Administration: Four Reform

Projects”, in Corte-Real, N., and Kelly, P. (2000) (eds.), Administrations in
Transition.
Pettigrew, A.M. (1987), “Context and Action in the Transformation of the Firm”,
Journal of Management Studies, 24/6: 649–670.
Pettigrew, A.M. (1988) (ed.), The Management of Strategic Change (Oxford: Black-
well).
Pettigrew, A.M. (1992), “The Character and Significance of Strategy Process
Research”, Strategic Management Journal, 13 [Special Issue, Winter]: 5–16.
Pfeffer, J., and Salancik, G.R. (1978), The External Control of Organizations: A
Resource Dependence Perspective (New York: Harper & Row).
Pitkin, F.H. (1972), Wittgenstein and Justice (Berkeley: The University of California
Press).
Polanyi, M. (1962), “The Republic of Science: Its Political and Economic Theory”,
Minerva, 1: 54–73.
Polanyi, M. (1983), The Tacit Dimension (Gloucester, MA: Peter Smith Publisher
Inc), 72 f.
Porter, M.E. (1985), Competitive Advantage: Creating and Sustaining Superior
Performance (London: Collier Macmillan).
Porter, M.E. (1986) (ed.), Competition in Global Industries (Boston, MA: Harvard
Business School Pr.).
Powell, W.W. (1990), “Neither Market nor Hierarchy: Network Forms of
Organization”, Research in Organizational Behaviour, 12: 295–336.
Powell, W.W., and DiMaggio, P.J. (1991) (eds.), The New Institutionalism in
Organizational Analysis (Chicago and London: The University of Chicago Press).
Power, M. (1990), “Modernism, Postmodernism and Organization”, in J. Hassard
and D. Pym (1990) (eds.), The Theory and Philosophy of Organization: Critical
Issues and New Perspectives (London: Sage), 109–124.
Power, M. (1997), The Audit Society: Rituals of Verification (Oxford: Oxford
University Press).
Power, M. (2004), The Risk Management of Everything: Rethinking the Politics of
Uncertainty (London: Demos).
Prahalad, C.K., and Bettis, R.A. (1986), “The Dominant Logic: A New Linkage
Between Diversity and Performance”, Strategic Management Journal, 7/6:
485–501.
Prahalad, C.K., and Hamel, G. (1990), “The Core Competency of the Corpora-
tion”, Harvard Business Review, 68/3: 79–82.
Prahalad, C.K., and Hamel, G. (1994), Competing for the Future (Boston, MA:
Harvard Business School Press).
Priem, R.L., and Butler, J.E. (2001), “Is the Resource-based ‘View’ a Useful Per-
spective for Strategic Management Research?”, Academy of Management Review,
26/1: 22–40.
Probst, G.J. (1987), Selbstorganisation: Ordnungsprozesse in sozialen Systemen aus
ganzheitlicher Sicht (Berlin: Blackwell Wissenschaft), 68 ff.
Prottas, J.M. (1979), People Processing: The Street-Level Bureaucrats in Public
Service Bureaucracies (Lexington, MA: Lexington Books).
Puxty, A. (1993), The Social and Organizational Context of Management
Accounting (London: Academic Press).
450
24 Referenser 05-11-11 16.33 Sida 451
Quine, W. v. O. (1969), “Epistemology Naturalized”, in W. v. O. Quine (1969) (ed.),

Ontological Relativity and Other Essays (New York), 69–90.
Quinn, J.B. (1980), Strategies for Change: Logical Incrementalism (Homewood:
Irwin).
Rajagopalan, N., and Spreitzer, G.M. (1996), “Toward a Theory of Strategic
Change: A Multi-Lens Perspective and Integrative Framework”, Academy of
Management Review, 22/1: 48–79.
Randall, D.M. (1987), “Commitment and the Organization: The Organization Man
Revisited”, Academy of Management Review, 12: 460–471.
Rasch, W. (2002), “Introduction: The Self-positing Society”, in W. Rasch (2002)
(ed.), Theories of Distinction: Redescribing the Descriptions of Modernity
(Stanford, CA: Stanford University Press), 1–30.
Rawls, A.W. (1987), “The Interaction Order Sui Generis: Goffmann’s Contribution
to Social Theory” Sociological Theory 5: 136–149.
Reber, G., and Strehl, F. (eds.) (1988), Matrix-Organisation: Klassische Beiträge zu
mehrdimensionalen Organisationsstrukturen (Stuttgart: Schäffer).
Reckwitz, A. (2000), Die Transformation der Kulturtheorie: Zur Entwicklung eines
Theorieprogramms (Weilerswist: Velbrück).
Reckwitz, A. (2002a), “Toward a Theory of Social Practices: A Development in
Culturalist Theorizing”, European Journal of Social Theory, 5/2: 245–265.
Reckwitz, A. (2002b), “The Status of the ‘Material’ in Theories of Culture: From
‘Social Structure’ to ‘Artefacts’”, Journal for the Theory of Social Behaviour,
32/2: 195–217.
Reckwitz, A. (2003a), “Grundelemente einer Theorie sozialer Praktiken: Eine
sozialtheoretische Perspektive”, Zeitschrift für Soziologie, 32/4: 282–301.
Reckwitz, A. (2003b), “Die Grenze des Sozialen und die Grenzen der Moderne:
Niklas Luhmann, die Kulturtheorien und ihre normativen Motive”, Mittelweg 36,
12/4: 61–79.
Reckwitz, A. (2005a), “Cultural Turn”, in A. Harrington et al. (eds.), Encyclopedia
of Social Theory (London: Routledge [forthcoming]).
Reckwitz, A. (2005b), “Culture”, in A. Harrington et al. (eds.), Encyclopedia of
Social Theory (London: Routledge [forthcoming]).
Ricœur, P. (1974), The Conflict of Interpretations: Essays on Hermeneutics
(Evanston: Northwestern University Press).
Rifkin, J. (2000), The Age of Access. The New Culture of Hypercapitalism: Where
All of Life Is a Paid-For Experience (New York: Tarcher/Putnam).
Robb, F. (1989), “Cybernetics and Supra-Human Autopoietic Systems”, Systems
Practice, 2: 47–74.
Roberts, G. (1994), “A Glossary of Key Terms” in P. Morris (1994) (ed.), The
Bakhtin Reader (London: Edward Arnold).
Robertson R. (1992), Globalization: Social Theory and Global Culture (London:
Sage).
Rödig, J. (1969), Die Denkform der Alternative in der Jurisprudenz (Berlin:
Springer).
Rorty, R. (1991), “Inquiry as Recontextualization: An Anti-Dualist Account of
Interpretation”, in vol. 1 of Objectivity, Relativism and Truth: Philosophical
Papers (New York: Cambridge University Press), 93–110.
451
24 Referenser 05-11-11 16.33 Sida 452
Rosenblueth, A., Wiener, N., and Bigelow, J. (1943), “Behavior, Purpose and
Teleology”, Philosophy of Science, 10: 18–24.
Roslender, R. (1992), Sociological Perspectives on Modern Accountancy (London:
Routledge).
Roslender, R., and Dillard, J. (2003), “Reflections on the Interdisciplinary Per-
spectives on Accounting Project”, Critical Perspectives on Accounting, 14/3:
325–351.
Rubinstein, J. (1974), City Police (New York: Hill and Wang).
Ruesch, J., and Bateson, G. (1951), Communication: The Social Matrix of
Psychiatry (New York: Norton & Co.).
Rumelt, R.P., Schendel, D.E., and Teece, D.J. “Fundamental Issues in Strategy”, in
R.P. Rumelt, D.E. Schendel, and D.J. Teece (1994) (eds.), Fundamental Issues in
Strategy: A Research Agenda (Boston, MA: Harvard Business School Press), 9–24.
Samuels, W.J. (1979), “Legal Realism and the Burden of Symbolism: The
Correspondence of Thurman Arnold”, Law & Society Review, 134: 997–1011.
Sarat, A., and Felstiner, W.L.F. (1988), “Law and Social Relations: Vocabulary of
Motive in Lawyer/Client Interaction”, Law and Society Review, 22: 737–769.
Saussure, F. de (1966), Course in General Linguistics (New York: McGraw-Hill).
Savigny, E. von (1991), “Common Behaviour of Many a Kind: Philosophical
Investigations Section 206”, in R. Arrington and H.-J. Glock (1991) (eds.),
Wittgenstein’s Philosophical Investigations: Text and Context (London:
Routledge), 105–119.
Schatzki, Th. (1996), Social Practices: A Wittgensteinian Approach to Human
Activity and the Social (Cambridge: Cambridge University Press).
Schatzki, Th. (1997), “Practices and Actions: A Wittgensteinian Critique of
Bourdieu and Giddens”, Philosophy of the Social Sciences, 27/3: 283–308.
Schatzki, Th. (2001), “Introduction: Practice Theory”, in Th. Schatzki et al. (2001),
The Practice Turn in Contemporary Theory (London: Routledge), 1–14.
Schatzki, Th., et al. (2001), The Practice Turn in Contemporary Theory (London:
Routledge).
Schimank, U. (1986), “Technik, Subjektivität und Kontrolle in formalen Organisa-
tionen: Eine Theorieperspektive”, in R. Seltz, U. Mill and E. Hildebrandt (1986)
(eds.), Organisation als soziales System: Kontrolle und Kommunikationstechno-
logie in Arbeitsorganisationen (Berlin: edition sigma), 71–91.
Schimank, U. (2001), Funktionale Differenzierung, Durchorganisierung und Inte-
gration der modernen Gesellschaft, in V. Tacke (2001) (ed.), Organisation und
gesellschaftliche Differenzierung (Opladen: Westdeutscher Verlag), 19–38.
Schneiberg, M., and Clemens, E.S. (2004), “The Typical Tools for the Job: Research
Strategies in Institutional Analysis” [Forthcoming in: W.W. Powell and D.L. Jones
(eds.), How Institutions Change (Chicago, IL: Chicago University Press)].
Scholl, R.W. (1981), “Differentiating Organizational Commitment from Expectancy
as a Motivating Force”, Academy of Management Review, 6: 589–599.
Schreyögg, G., and Steinmann H. (1987), “Strategic Control: A New Perspective”,
Academy of Management Review, 12/1: 91–103.
Schulze, G. (1992), Die Erlebnisgesellschaft: Kultursoziologie der Gegenwart
(Frankfurt am Main: Campus).
452
24 Referenser 05-11-11 16.33 Sida 453
Schutz, A. (1975), “The Problem of Transcendental Intersubjectivity in Husserl’s

Works”, in I. Schutz (1975) (ed.), Collected Papers: Studies in Phenomenological
Philosophy, vol. 3 (The Hague: Martinus Nijhoff), 51–91.
Schutz, A., and Luckmann, T. (1974), Structures of the Life-World (London:
Heinemann).
Schwartzman, H.B. (1986), “The Meeting As a Neglected Social Form of Organiza-
tional Studies”, in B.M. Straw and L.L. Cummings (1986) (eds.), Research in
Organizational Behavior, vol. 8 (Greenwich, CT: JAI Press), 233–258.
Schwartzman, H.B. (1987), “The Significance of Meetings in an American Health
Care Centre”, American Ethnologist, 14: 271–294.
Schwartzman, H.B. (1989), The Meeting – Gatherings in Organizations and Com-
munities (New York: Plenum Press).
Scott, S.G., and Lane, V.R. (2000), “A Stakeholder Approach to Organizational
Identity”, Academy of Management Review, 25/1: 43–62.
Scott, W.R. (1984), Organizations: Rational, Natural, and Open Systems (Engle-
wood Cliffs, NJ: Prentice Hall).
Scott, W.R. (1995), Institutions and Organizations (Thousand Oaks, CA: Sage
Publications).
Scott, W.R., and Meyer, J.W. (1991), “The Organization of Societal Sectors:
Propositions and Early Evidence”, in W.W. Powell and P.J. DiMaggio (1991)
(eds.), The New Institutionalism in Organizational Analysis (Chicago: Univ. of
Chicago Press), 108–140.
Seidl, D. (2002), “Metaphorical Self-descriptions of Organizations”, in A. Kieser
and A. Müller (2002) (eds.), Communication in Organizations (New York: Lang).
Seidl, D. (2003a), “Organisation im Medium der Kommunikation: Zum Verhältnis
von Organisation, Gesellschaft und Interaktion in Luhmanns Theorie sozialer
Systeme” (Munich Business Research Paper No. 2003–8, Ludwig-Maximilians-
Universität).
Seidl, D. (2003b), “Organizational Identity in Luhmann’s Theory of Social
Systems”, in T. Bakken and T. Hernes (2003) (eds.), Autopoietic Organization
Theory: Drawing on Niklas Luhmann’s Social Systems Perspective (Oslo: Liber
Abstrakt), 123–150.
Seidl, D. (2005), Organizational Identity and Self-Transformation: An Autopoietic
Perspective (Aldershot: Ashgate).
Seidl, D., and Becker, K.H. (2006), “Organizations as Distinction Generating and
Processing Systems”, Organization, 13/1 [forthcoming].
Selvini Palazzoli, M., Boscolo, L., and Cecchin, G. (1983), Paradox and Counter-
paradox: A New Model in the Therapy of the Family in Schizophrenic Trans-
action, trans. E.V. Burt (New York: Aronson).
Selznick, P. (1957), Leadership in Administration: A Sociological Interpretation
(Evanston, IL: Harper & Row).
Selznick, P. (1966), TVA and the Grass Roots: A Study in the Sociology of Formal
Organization (repr., New York: Harper & Row).
Sennett, R. (1998), The Corrosion of Character: The Personal Consequences of
Work in the New Capitalism (New York: W.W. Norton & Co).
Serres, M. (1980), Le parasite (Paris: Grasset).
Sgritta, G.B. (1988), “Conoscenza e intervento: verso approccio interattivo”,
Rassegna Italiana di Sociologia, 29: 537–562.
453
24 Referenser 05-11-11 16.33 Sida 454
Shackle, G.L.S. (1979), “Information, Formalism, and Choice”, in M.J. Rizzo

(1979) (ed.), Time, Uncertainty, and Disequilibrium: Explorations on Austrian
Themes (Lexington, MA: Lexington Books), 19–31.
Shaftesbury, [Anthony Ashley Cooper], Earl of (1999), Characteristics of Men,
Manners, Opinions, Times (new edn., Cambridge: Cambridge University Press).
Shannon, C.E., and Weaver, W. (1963), The Mathematical Theory of Communica-
tion (Urbana, IL: Illinois University Press).
Shenhav, Y. (1999), Manufacturing Rationality: The Engineering Foundations of the
Managerial Revolution (Oxford: Oxford University Press).
Shoemaker, S. (1963), Self-Knowledge and Self-Identity (Ithaca, NY: Cornell
University Press).
Shoemaker, S. (1968), “Self-Reference and Self-Awareness”, Journal of Philosophy,
65: 555–567.
Short, Jr., J.F. (1984), “The Social Fabric at Risk: Toward the Social Transformation
of Risk Analysis”, American Sociological Review, 49: 711–725.
Siemons, M. (1997), Jenseits des Aktenkoffers: Vom Wesen des neuen Angestellten
(München: Hanser).
Silverman, D. (1970), The Theory of Organizations (New York: Basic Books).
Silverman, D., and Brian, T. (1980), The Material Word: Some Theories of
Language and its Limits (London: Routledge and Kegan Paul).
Simmel, G. (1978), Philosophy of Money (Boston: Routledge & Kegan Paul).
Simon, F.B. (1988), Unterschiede, die Unterschiede machen. Klinische Epistemo-
logie: Grundlage einer systemischen Psychiatrie und Psychosomatik (4th edn.,
1999, Frankfurt: Suhrkamp).
Simon, F.B. (1990), Meine Psychose, mein Fahrrad und ich: Zur Selbstorganisation
der Verrücktheit (Heidelberg: Carl-Auer-Systeme).
Simon, F.B. (2001), The Family in Family Business: Organizational Theory and
Practice, Scandinavian Journal of Organizational Psychology, 2: 41–52.
Simon, F.B. (2002a) (ed.), Die Familie des Familienunternehmens: Ein System zwi-
schen Gefühl und Geschäft (Heidelberg: Carl-Auer-Verlag).
Simon, F.B. (2002b), “The De-Construction and Re-Construction of Authority and
the Role of Management and Consulting”, Soziale Systeme, 8: 283–293.
Simon, F.B. (2004), The Organization of Selforganization: Foundations of Systemic
Management (Heidelberg: Carl-Auer-Verlag).
Simon, H.A. (1948), Administrative Behavior: A Study of the Decision-making
Processes in Administrative Organization (New York: Macmillan).
Simon, H.A. (1957), Models of Man – Social and Rational: Mathematical Essays on
Rational Human Behavior in a Social Setting (New York: Wiley & Son).
Simon, H.A. (1960), The New Science of Management Decision (New York: Harper
and Row).
Simon, H.A. (1965), Administrative Behavior: A Study of Decision-Making Process
in Administrative Organization (2nd edn., New York: The Free Press).
Simon, H.A., Smithburg, D.W., and Thompson, V.A. (1950), Public Administration
(New York: Wiley & Son).
Sims, H.P., and Gioia, D.A. (1986) (eds.), The Thinking Organization (San
Francisco: Jossey-Bass Publishers).
454
24 Referenser 05-11-11 16.33 Sida 455
Smircich, L. (1983), “Implications for Management Theory”, in L.L. Putnam and

M.E. Pacanowsky (1983) (eds.), Communication and Organizations: An Inter-
pretive Approach (Beverly Hills, CA: Sage), 221–241.
Sombart, W. (1919), Der moderne Kapitalismus, vol. 2 (3rd edn., Munich and
Leipzig: Duncker & Humblot).
Sombart, W. (1928), Der moderne Kapitalismus, vol. 1 (6th edn., Munich and
Leipzig: Duncker & Humblot).
Spencer Brown, G. (1971), Laws of Form (2nd edn., London: Crown pub.).
Spencer Brown, G. (1972), Laws of Form (1st American edn., New York: Julian
Press).
Spencer Brown, G. (1979), Laws of Form (2nd edn., New York: Penguin).
Stacey, R.D. (2000), Strategic Management & Organizational Dynamics: The
Challenge of Complexity (3rd edn., Harlow: Prentice Hall).
Stalk, G., Evans, P., and Lawrence, L.E. (1992), “Competing on Capabilities: The
New Rules of Corporate Strategy”, Harvard Business Review, 70/2: 57–69.
Starkey, K., Barnatt, C., and Tempest, S. (2000), “Beyond Networks and
Hierarchies: Latent Organizations in the U.K. Television Industry”, Organization
Science, 11/3: 299–305.
Stebbins, G.L. (1982), Darwin to DNA, Molecules to Humanity (San Francisco:
Freeman).
Stein, I. (1994), “Die Theorien der Multinationalen Unternehmung”, in S. Schoppe
(1994) (ed.), Kompendium der Internationalen Betriebswirtschaftslehre (3rd edn.,
Munich and Vienna: Oldenbourg), 49–151.
Stern, R.N., and Barley, S.R. (1996), “Organizations and Social Systems: Organiza-
tion Theory’s Neglected Mandate”, Administrative Quarterly, 41: 146–162.
Stichweh, R. (1998), “Konstruktivismus und die Theorie der Weltgesellschaft”, in A.
Reckwitz and J. Sievert (1998) (eds.), Konstruktion, Interpretation, Kultur: Ein
Paradigmenwechsel in den Sozialwissenschaften (Opladen: Westdeutscher Ver-
lag), 208–218.
Stichweh, R. (1999a), “Globalisierung von Wirtschaft und Wissenschaft:
Produktion und Transfer wissenschaftlichen Wissens in zwei Funktionssystemen
der modernen Gesellschaft”, Soziale Systeme: Zeitschrift für Soziologische
Theorie, 5: 27–39.
Stichweh, R. (1999b), “Zur Genese der Weltgesellschaft: Innovationen und
Mechanismen”, in M. Bauschulte et al. (1999) (eds.), Wege – Bilder – Spiele
(Bielefeld: Aisthesis), 289–302.
Stichweh, R. (2000a), “Adresse und Lokalisierung in einem globalen Kommunika-
tionssystem”, in R. Stichweh (2000), Die Weltgesellschaft: Soziologische Analysen
(Frankfurt am Main: Suhrkamp), 220–231.
Stichweh, R. (2000b), Die Weltgesellschaf: Soziologische Analysen (Frankfurt am
Main: Suhrkamp).
Stinchcombe, A.L. (1965), “Social Structure and Organizations”, in J. March (1965)
(ed.), Handbook of Organizations (Chicago: Rand MacNally & Company),
142–193.
Stinchcombe, A.L. (1974), Creating Efficient Industrial Administrations (New York:
Academic Press).
Stinchcombe, A.L. (1990), Information and Organizations (Berkeley and Los
Angeles: University of California Press).
455
24 Referenser 05-11-11 16.33 Sida 456
Tacke, V. (2000), “Netzwerk und Adresse”, Soziale Systeme: Zeitschrift für

Soziologische Theorie, 6/2: 291–320.
Tacke, V. (2001) (ed.), Organisation und gesellschaftliche Differenzierung
(Wiesbaden: Westdeutscher Verlag).
Takeuchi, H., and Porter, M. E. (1986), “Three Roles of International Marketing in
Global Strategy”, in M. E. Porter (ed.), Competition in Global Industries (Boston,
MA: Harvard Business School Pr.), 111–146.
Taylor, C. (1985), Philosophy and the Human Sciences: Philosophical Papers,
vols. 1 and 2 (Cambridge: Cambridge University Press).
Taylor, C. (1997), Philosophical Arguments (Cambridge, MA: Harvard University
Press).
Taylor, C. (1999), “To Follow a Rule…”, in R. Shusterman (1999) (ed.), Bourdieu:
A Critical Reader (Oxford: Blackwell), 29–44.
Teece, D.J., and Pisano, G., (1994), “The Dynamic Capabilities of Firms: An Intro-
duction”, Industrial and Corporate Change, 3/3: 537–556.
Teece D.J., G. Pisano, and Shuen, A. (1997), “Dynamic Capabilities and Strategic
Management”, Strategic Management Journal, 18/7: 509–533.
Teichova, A., et al. (1986) (eds.), Multinational Enterprise in Historical Perspective
(Cambridge: Cambridge Univ. Press).
Teubner, G. (1987a), “Hyperzyklus in Recht und Organisation”, in H. Haverkamp
and M. Schmid (1987) (eds.), Sinn, Kommunikation und soziale Differenzierung:
Beiträge zu Luhmanns Theorie sozialer Systeme (Frankfurt am Main: Suhrkamp),
89–128.
Teubner, G. (1987b), “Episodenverknüpfung: Zur Steigerung von Selbstreferenz im
Recht”, in D. Baecker et al. (1987) (eds.), Theorie als Passion (Frankfurt am
Main: Suhrkamp), 423–446.
Teubner, G. (1992), “Die vielköpfige Hydra: Netzwerke als kollektive Akteure
höherer Ordnung”, in W. Krohn and G. Küppers (1992) (eds.), Emergenz: Die
Entstehung von Ordnung, Organisation und Bedeutung (Frankfurt am Main:
Suhrkamp), 189–216.
Teubner, G. (1996), “Double Bind: Hybrid Arrangements as De-Paradoxifiers”,
Journal of Institutional and Theoretical Economics, 152: 59–64.
Teubner, G. (1999), “Polykorporatismus: Der Staat als Netzwerk öffentlicher und
privater Kollektivakteure”, in H. Brunkhorst and P. Niesen (1999) (eds.), Das
Recht der Republik (Frankfurt am Main: Suhrkamp), 346–372.
Teubner, G. (2000), “Hybrid Laws: Constitutionalizing Private Governance
Networks”, in R. Kagan and K. Winston (2000) (eds.), Legality and Community
(California University Press).
Thévenot, L. (2001), “Pragmatic Regimes Governing the Engagement with the
World”, in Th. Schatzki et al. (2001), The Practice Turn in Contemporary Theory
(London: Routledge), 56–73.
Thomas, G.M., et al. (1987) (eds.), Institutional Structure: Constituting State,
Society, and the Individual (Newbury Park: Sage), 242–260.
Thompson, J.D. (1967), Organizations in Action: Social Science Bases of
Administrative Theory (New York: MacGraw-Hill).
Thorelli, H.B. (1986), “Networks: Between Markets and Hierarchies”, Strategic
Management Journal, 7: 37–51.
456
24 Referenser 05-11-11 16.33 Sida 457
Tinker, A.M., Merino, B., and Neimark, M.D. (1982), “The Normative Origins of
Positive Theories: Ideology and Accounting Thought”, Accounting, Organiza-
tions and Society, 7/2: 167–200.
Tirole, J. (2001), “Corporate Governance”, Econometrica, 69/1: 1–35.
Toren, N. (1994), “Professional-Support and Intellectual-Influence Networks of
Russian Immigrant Scientists in Israel”, Social Studies of Science, 24/4: 725–743.
Townley, B. (1993), “Foucault, Power/Knowledge and Its Relevance for Human
Resource Management”, The Academy of Management Review, 18/3: 518–545.
Townley, B. (2002), “Managing with Modernity”, Organization, 9/4: 549–573.
Toyne, B., and Walters, P.G.P. (1989), Global Marketing: A Strategic Perspective
(Boston MA: Allyn and Bacon).
Tucker, K.H. (1998), Anthony Giddens and Modern Social Theory (London: Sage).
Udy, Jr., and Stanley H. (1970), Work in Traditional and Modern Society (Engle-
wood Cliffs, NJ: Prentice Hall).
Ulrich, P. (1984), “Systemsteuerung und Kulturentwicklung: Auf der Suche nach
einem ganzheitlichen Paradigma der Managementlehre”, Die Unternehmung, 38:
303–325.
Université des Nations-Unies (1986), Science et pratique de la complexité: Actes du
Colloque de Montpellier Mai 1984 (Paris: La documetation).
Waddock, S., and Smith, N. (2000), “Corporate Responsibility Audits: Doing Well
by Doing Good”, Sloan Management Review, 41/2: 75–83.
Walsch, J.P., and Ungson, G.R. (1991), “Organizational Memory”, Academy of
Management Review, 16: 57–91.
Walsh, D.F. (1998), “Structure/Agency”, in C. Jenks (1988) (ed.), Core Sociological
Dichotomies (London: Sage), 8–33.
Van de Ven, A.H., and Scott Poole, M. (1988), “Paradoxical Requirements for a
Theory of Change”, in R.E. Quinn and K.S. Cameron (1988) (eds.), Paradox and
Transformation: Toward a Theory of Change in Organization and Management
(Cambridge, MA: Ballinger Publishing Company), 19–64.
Varela, F.J. (1979), Principles of Biological Autonomy (New York: Prentice Hall
PTR).
Varela, F.J. (1984), “Two Principles of Self-Organization”, in H. Ulrich and G.J.B.
Probst (1984) (eds.), Selforganization and Management of Social Systems: Insides,
Promises, Doubts and Questions (Bern: Haupt), 25–32.
Varela, F.J. (1999), Ethical Know-How: Action, Wisdom, and Cognition (Stanford,
CA: Stanford University Press).
Varela, F.J., Maturana, H., and Uribe, R. (1974), “Autopoiesis: The Organization of
Living Systems, its Characterization and a Model”, Biosystems 5: 187–196.
Watts, D.J. (1999), “Networks, Dynamics, and the Small World Phenomenon”,
American Journal of Sociology, 105: 493–527.
Weaver, G.R., and Gioia, D.A. (1994), “Paradigms Lost: Incommensurability vs
Structurationist Inquiry”, Organization Studies, 15: 565–590.
Weaver, W., (1948), “Science and Complexity”, American Scientist, 36: 536–544.
Weber, J. (1992), “Die Koordinationssicht des Controlling”, in K. Spreemann and
E. Zur (eds.), Controlling: Grundlagen – Informationssysteme – Anwendungen
(Wiesbaden: Gabler), 169–183.
Weber, J. (2004), Einführung in das Controlling (10th edn., Stuttgart: Schaeffer-
Poeschel).
457
24 Referenser 05-11-11 16.33 Sida 458
Weber, J., and Schaeffer, U. (1999), Sicherstellung der Rationalität von Führung als
Funktion des Controlling, Die Betriebswirtschaft, 59/6: 731–746.
Weber, M. (1956), Wirtschaft und Gesellschaft (Tübingen: Mohr).
Weber, M. (1968), Economy and Society: an Outline of Interpretive Sociology (New
York: Bedminster Press).
Weber, M. (1972), Wirtschaft und Gesellschaft, ed. J. Winckelmann (5th edn.,
Tübingen: Mohr Siebeck) (first pub. 1922).
Weber, M. (1978), Economy and Society: An Outline of Interpretive Sociology
(Berkeley, CA: University of California Press).
Webster, F. (1995), Theories of the Information Society (London: Routledge).
Weick, K. (1995), Sensemaking in Organizations (London: Sage).
Weick, K.E. (1977a), “Enactment Processes in Organizations”, in B.M. Shaw and
G.R. Salancik (eds.), New Directions in Organizational Behavior (Chicago:
Krieger Pub. Co.), 267–300.
Weick, K.E. (1977b), “Re-Punctuating the Problem”, in P.S. Goodman,
J.M. Pennings, and associates (1977), New Perspectives on Organizational
Effectiveness (San Francisco: Jossey-Bass Limited) [the Jossey-Bass Social and
Behavioral Science Series], 193–225.
Weick, K.E. (1979), The Social Psychology of Organizing (2nd edn., Reading, MA:
Addison-Wesley).
Weick, K.E. (1987), “Substitutes for Corporate Strategy”, in D.J. Teece (1987) (ed.),
The Competitive Challenge: Strategies for Industrial Innovation and Renewal
(Cambridge, MA: Balling Publishing Organization).
Weick, K.E. (1995), Sensemaking in Organizations (Thousand Oaks, CA: Sage.
Weick, K.E. (1999), “That’s Moving: Theories That Matter”, Journal of Manage-
ment Inquiry, 8/2: 134–142.
Weick, K.E., and Sutcliffe, K.M. (2001), Managing the Unexpected: Assuring High-
Performance in an Age of Complexity (San Francisco: Jossey-Bass).
Veld, R.J. in’t, et al. (1991) (eds.), Autopoiesis and Configuration Theory: New
Approaches to Societal Steering (Dordrecht: Kluwer Academic).
White, H.C. (1992), Identity and Control: A Structural Theory of Action (Princeton,
NJ: Princeton University Press).
Whittington, R. (1993), What is Strategy – And Does It Matter? (London:
Routledge).
Whyte, W.H. (1956), The Organization Man (New York: Simon and Schuster).
Wiener, N. (1961), Cybernetics – or Control and Communication in the Animal and
the Machine (2nd edn., Cambridge, MA: MIT Press).
Wildavsky, A. (1988), Searching for Safety (New Brunswick: Transaction).
Wilden, A. (1972), System and Structure: Essays in Communication and Exchange
(London: Tavistock Press).
Wilkins, M. (1970), The Emergence of Multinational Enterprise: American Business
Abroad from the Colonial Era to 1914 (Cambridge, MA: Harvard Univ.Pr.).
Wilkins, M. (1991) (ed.), The Growth of Multinationals (Aldershot: Elgar).
Williamson, O.E. (1991), “Comparative Economic Organization: The Analysis of
Discrete Structural Alternatives”, Administrative Science Quarterly, 36: 269–296.
Williamson, O.E., and Ouchi, W.G. (1981), “The Markets and Hierarchies Program
of Research”, in A. Van de Ven and W.F. Joyce (1981) (eds.), Perspectives on
Organization Design and Behavior (New York: Wiley), 347–370.
458
24 Referenser 05-11-11 16.33 Sida 459
Willke, H. (1983), Entzauberung des Staates: Überlegungen zu einer sozialen

Steuerungstheorie (Königstein/Ts.: Athenäum).
Willke, H. (1992), “Prinzipien politischer Supervision”, in H. Bußhoff (1992) (ed.),
Politische Steuerung: Steuerbarkeit und Steuerungsfähigkeit. Beiträge zur
Grundlagendiskussion (Baden-Baden: Nomos), 51–80.
Willke, H. (1994a), “Systemtheoretische Strategien des Erkennens: Wirklichkeit als
interessierte Konstruktion”, in K. Götz (1994) (ed.), Theoretische Zumutungen:
Vom Nutzen der systemischen Theorie für die Managementpraxis (Heidelberg:
Carl-Auer-Systeme), 97–116.
Willke, H. (1994b), Systemtheorie II: Interventionstheorie. Einführung in die
Theorie der Intervention in komplexe Sozialsysteme (Stuttgart: UTB).
Willke, H. (2001), Atopia: Studien zur atopischen Gesellschaft (Frankfurt am Main:
Suhrkamp).
Willke, H. (2002), Dystopia: Studien zur Krisis des Wissens in der modernen
Gesellschaft (Frankfurt am Main: Suhrkamp).
Willke, H. (2003), Heterotopia: Studien zur Krisis der Ordnung moderner Gesell-
schaften (Frankfurt am Main: Suhrkamp).
Wimmer, R. (1991), “Organisationsberatung: Eine Wachstumsbranche ohne profes-
sionelles Selbstverständnis”, in M. Hofmann (1991) (ed.), Management Forum
(Heidelberg: Physica), 46–136.
Wimmer, R. (1992) (ed.), Organisationsberatung: Neue Wege und Konzepte (Wies-
baden: Gabler).
Wimmer, R. (2004), Organisation und Beratung (Heidelberg: Carl-Auer-Systeme
Verlag).
Wit B. de, and Meyer, R. (1994), Strategy – Process, Content, Context: An Inter-
national Perspective (St. Paul, MN: West).
Wittgenstein, L. (1953), Philosophical Investigations (Oxford: Blackwell).
Wittgenstein, L. (1961), Tractatus Logico-Philosophicus, trans. D.F. Pears and
B.F. McGuinness (London: Routledge & Kegan Paul).
Wollnik, M. (1994), “Interventionschancen bei autopoietischen Systemen”, in
K. Götz (1994) (ed.), Theoretische Zumutungen: Vom Nutzen der systemischen
Theorie für die Managementpraxis (Heidelberg: Carl-Auer-Systeme), 118–159.
von Foerster, H. (1981), Observing Systems (Seaside, Cal.: Intersystems Publica-
tions).
Vos, J.P. (2002), The Making of Strategic Realities: An Application of the Social
Systems Theory of Niklas Luhmann (Eindhoven: Eindhoven University Press)
[ECIS Thesis Series].
Yates, F.A. (1966), The Art of Memory (London: Routledge & Kegan Paul).
Yates, J.A. (1989), Control through Communication: The Rise of System in
American Management (Baltimore: John Hopkins University Press).
Zollschan, G.K., and Overington, M.A. (1976), “Reasons for Conduct and the
Conduct of Reason: The Eightfold Route to Motivational Ascription”, in
G.K. Zollschan and W. Hirsch (1976) (eds.), Social Change: Explorations,
Diagnoses, and Conjectures (New York: Wiley), 270–317.
Zucker, L.G. (1977), “The Role of Institutionalization in Cultural Persistence”,
American Sociological Review, 42: 726–743.
459
25 Index 05-11-11 16.33 Sida 460
Index
A boundary 47–51, 88, 160, 161–67, 175–81,
accounting 137, 221, 235, 175, 181–2, 259, 184–90, 194–6, 203–9,
251, 386–401 269–70, 308–11, 332 274, 355, 358–9,
action bureaucracy 181, 196, 368–75, 386, 389,
– concept of 28, 34–5, 66, 200, 258, 390 394–95, 398, 405
70–72, 108, 181, 218, complexity reduction 68,
231–237, 405 C 95, 161, 163, 166–7,
– collective 174, 181, calculus of form 46–53, 352
196, 256 86–7, 270, 307, 397 connectivity 27, 59,
– theory of 55, 69–70, centre/periphery 36, 61–62, 109–110, 117,
181 172–74, 184, 409 119–21, 123–4, 157,
actor 55, 70, 121, 133–5 chaining 264–6, 277–80 358
– collective 141, 177, change 9, 66, 94–7, 107, consciousness 25–6, 32–3,
181–82, 223 130, 132, 142, 154–5, 58, 64–6, 73, 76–8,
– corporate 171, 174, 249, 253, 259–261, 371 80–82, 195, 197–8,
181–183, 190 classical organization the- 200, 224–6, 232–3,
adaptation 69–80, 186–7, ory 10–11, 85, 97, 263, 266–7, 409
251–52, 368–71, 374, 173–4, 189, 191, 389, constructivism 9, 92, 211,
378, 383–84 418 216–7, 221–22, 239,
address, social 284, code 36–8, 147, 188, 192, 310, 387, 393, 395, 412
288–291, 293, 300–303 223, 256–7, 286, 405, – radical 23, 105, 377,
addressability 291 416 385, 412
agency, agent 76, 176–7, co-evolution 75, 230, 234, consulting 97, 99, 133,
218–9, 225–6, 230–45, 236, 243, 312–20 205, 228, 249, 251,
259, 386 cognitive routine 44, 216, 351–64, 376, 380, 399
asymmetry 47–48, 71, 79, 220, 405 contingency 39, 44, 89,
90, 372, 378–80 commitment 89, 100–101, 107–126, 146, 155,
attribution 29, 34–5, 181, 195, 208–9, 252 201–202, 296, 395,
70–72, 85–6, 120, 126, communication 406, 413
152–155, 163–4, 180, – acceptance 30, 68, contingency theory 250,
182, 228, 235–6, 246, 96–7, 115, 185, 188 253, 382, 419
289–90 – communicative address convergence 157, 249,
autonomy 45, 59, 75, 180, 183, 235, 261
115, 153, 155, 174, 290–291, 294 corporation, historical
179, 185–9, 198, – channel 42–5, 93–4, forms of 324, 339, 341,
228–9, 243–4, 311–15, 147, 198, 242, 405 346
339 – information 28–30, coupling
autopoiesis 11–12, 21–35, 34–5, 52, 66–8, 71, – strict 313–314,
54–83, 95, 98, 105, 77–7, 221–22, 405 317–323
110–11, 123, 129, – media of 185, 204, 253, – structural 24, 30–32,
134–5, 145–7, 153, 291, 337, 415 37–38, 101, 105, 150,
166, 174, 187, 195–8, – meta-communication 173–174, 186–188,
227–30, 235, 238, 255, 146, 147, 310 197, 211, 239, 283,
270, 310–13, 333–5, – understanding 28–30, 394, 410
405 34–35, 66–71, 76–8, – loose 103, 120, 184,
405 201–203, 206, 208,
B – utterance 28–35, 52, 255, 259, 313–314,
blind spot 49–50, 86, 66–70, 76–7, 161, 290, 317, 381
132–135, 270, 357 302, 405 cultural turn 216, 220
body 197, 201, 218–219, complexity 67–8, 73, 112,
237–247 117, 123–4, 156,
460
25 Index 05-11-11 16.33 Sida 461
cybernetics 8, 62, 88, 99, – secondary 36, 184, 284, 108, 157, 180, 203,
174, 184, 255, 288, 296, 335–336, 300–1, 406
357–359, 368, 389 339, 409 evolution 33, 36, 58–9,
– societal 183, 185, 68, 73, 75, 77–9, 94,
D 284–288, 335, 347, 409 129–30, 172, 179,
de-centring the subject – stratificatory 36, 172, 181–83, 185, 187, 230,
217, 223–231, 234, 174, 180, 183–84, 190, 234, 236, 255, 259–60,
237, 267–268 285, 409 286, 312, 317, 328,
decentralization 107, 111, – forms of 36, 172, 179, 334, 341, 409, 414–5
118, 132–133, 178, 300 183–184, 206, 280, evolutionary universals
decision 39–46, 85–106, 286, 335–36, 362, 409 181
108–11, 331–335, 406, diffusion 249, 256, evolution theory 94, 181,
420 259–260 255
decision premise displacement 119–124 exclusion/inclusion 39, 71,
– communication channel distinction 9, 46–53, 55, 178, 193–194, 202,
42–44, 93–94, 147 198, 67, 71–72, 76–7, 81, 210, 284, 286,
255, 405 85, 91–93, 99–104, 288–289, 292–293
– decision programme 44, 109, 134, 150–53, expectation 31–32, 43,
98, 163, 179, 228, 406 194–9, 206, 221–23, 70, 72, 86, 93, 95, 116,
– cognitive routine 44, 268–71, 274, 307–23, 119–20, 122–3, 148,
405 332–4, 356–63, 370–1, 179, 182, 167–88,
– organisational culture 397–9, 405–9, 412 190–91, 202, 248–54,
44, 100, 199, 408 division of labour 58, 256–7, 260, 338,
– personnel 42–44, 94, 175, 191, 418 346–7, 375, 381–83,
113, 140, 147, 176, double closure 44–45, 408
179, 296, 299, 408 114, 331–333 external communicability
deconstruction 40, duality of structure 232, 173, 182, 187
110–111, 146, 264, 245
267, 273, 275, F
278–281, 393, 396, 413 E financial market 319–320,
decoupling 132, 252–254 ecologization 200 330, 342
de-ontologisation 26–27, economy 12, 36–8, 97, firm 41, 251, 289, 291,
268–272 155, 173, 178, 185–6, 313, 315–322,
de-paradoxization 46, 72, 188, 208, 255–9, 282, 326–329, 353–354, 360
79, 107–12, 115, 315, 325–30, 335–47, form 46–53, 86–7, 270,
123–7, 129–34, 353–4, 398, 407, 417 307, 397
139–42, 163–65, effciency 111, 140, 205, formal/informal 12, 97,
376–8, 406, 413 251, 258, 299, 389 123, 150, 153, 163,
design 91–213, 388–389, eigenvalues 279–280 169, 179, 188–9, 250,
397 element 11, 21–31, 34–6, 252–5, 294–5,
differend 265–266, 277 39–40, 51, 61, 66–78, 297–301, 419
differentiation 81, 108, 155, 157, function
– theory of 172–175, 179, 160–63, 185, 222, 238, – functionalism 93, 222
181 270, 311, 313, 405–7, – functional differentia-
– functional 36–37, 78, 409 tion 36–37, 78,
172–190, 249–50, enactment 77, 196, 204, 172–190, 249–250,
255–257, 278, 283, 220, 222, 239, 241, 255–257, 278, 283,
288, 296, 336–339, 252, 356, 378–83, 400 288, 296, 336–339,
407, 415 epistemology 9, 79–82, 407, 415
– internal 124–125, 172, 98, 105, 140, 267–9, – functional equivalence
175, 184–85, 190 272–3, 276, 280, 97, 120, 188, 190, 256,
– primary 36, 172, 386–391, 395, 397, 412 358, 374–375,
183–186, 190, event 26–8, 30, 32, 55, 378–380, 382–383, 419
255–256, 284–289, 61, 66–7, 71, 73–4, 76, – functional method 375,
335–336, 339, 409 378–379
461
25 Index 05-11-11 16.33 Sida 462
– functional-structural 225–36, 294–5, 372, M

theory 93, 173, 179 396, 405 management 110, 124,
– functional subsystems institution 86, 131–132, 133, 162–3, 198, 202,
12, 36, 60, 155, 172, 140–141, 178–182, 204, 307, 316–22,
177, 184–187, 190, 187–188, 191, 205, 365–385, 386–401
223, 256, 277, 325, 209, 220, 248–261, marked state/unmarked
335, 338–339, 407, 300, 364 state 47–52, 397, 399
409, 415–417 interaction 35–9, 69, 78, meaning 9, 28–30, 32, 35,
145–70, 175–6, 188, 41, 50, 61–62, 64–5,
G 194, 227, 232, 236, 68, 76–7, 95, 129,
goal-attainment 77, 175, 243, 258–9, 405, 407, 141–42, 151, 157,
177 415, 419 180–88, 187, 204,
Gorgon sisters 127–143 interdepedency breaks 220–28, 237–40,
group 97, 101, 113–114, 184–187 242–4, 279, 285–92,
121, 123, 138, 191, interpenetration 31–32, 301, 308, 332–3,
228, 258, 286, 335, 71, 156–157, 160–162, 336–8, 356–7, 379,
339, 354–355, 383, 391 167, 174, 229, 407 397, 407, 414
intersubjectivity 226–227, medium/form 61, 102,
H 237, 244, 268 173, 186, 205–6, 327,
habitus 219, 225, 232, isomorphism 249–250, 339, 356
234, 244 253, 257, 260 member(ship) 43, 114,
heterogeneity 180, 182, 150–51, 154–5, 168–9,
198, 201, 205, 288, K 176–9, 190, 209–10,
292–93 knowledge 236, 239, 283, 288,
hierarchy 10, 36, 43, 79, – claim 276, 281 293, 296–8, 301, 329,
85–86, 92, 96–97, 124, – deficiency of 192–3, 409
133, 151, 162, 164, 201 memory 32, 73, 166–8,
177, 182, 196, – development 142, 266 203, 291, 421
198–199, 204–205, – narrative 265, 273 meta-narrative 264,
207, 282, 300, 340, – schemes of 218–9, 266–7, 272, 280, 418
390, 409, 418 224–5, 234, 241 modernity, modernization
human being 9, 13, – scientific 79, 92, 105, 191, 251, 258, 262–3,
31–34, 39, 130, 181, 265–6 267, 276, 280, 295, 415
182, 332, 407 motive 70, 85, 100–1,
hypocrisy 131 L 104–5, 122–3, 131,
language 33–4, 66–7, 80, 178–9, 190, 209, 301,
I 86, 101, 128–30, 160, 354, 421
ideal type 181, 218, 258, 195, 224, 267, 272,
390 361, 393, 410 N
identity 56, 66, 74, 92, latent/manifest 56, 73, network 198, 201, 204–5,
94, 140–41, 222, 256, 101, 159, 273, 277, 207–10, 282–305, 329
259, 268–9, 355, 372, 301, 354, 356, 358–9, new institutionalism
383, 397, 412 361 248–61
immune system 61, 64, Law of requisite variety
78–79, 208 162, 368–9, 371 O
indication 47–50, 58, 88, legitimation 105, 128, observation
109, 222, 239, 270, 141, 188, 195, 219–20, – concept of 46–53,
279, 308–310, 317, 264–6, 273–4, 277–8, 79–82, 270–1, 308–18,
356, 358, 397, 399, 412 292 408, 412
individualization 199–200 life-form 236, 245, 247 – of first order 48–50, 91,
information 28–30, 34–5, life-world 172, 226, 236, 95, 133–5, 141, 361,
52, 59–62, 66–8, 74–7, 263, 275–6 364, 376–83, 399
97, 106, 165–6,
462
25 Index 05-11-11 16.33 Sida 463
– of second order 49–50, oscillation 134–5, 194, – decision programme

56, 88, 91, 95–6, 101, 277, 408 42–5, 52–3, 56, 93–5,
133–4, 141, 271, other- (hetero-, external-, 97–8, 113, 115, 147,
278–9, 357, 359, 361, environmental-) refer- 163, 175–6, 179,
364, 376–82, 399–400, ence 52, 60, 67, 94, 181–82, 192, 256–7,
412 147, 329–30, 347, 408 242, 296, 389, 406
– paradox of 45–50, – conditional 43, 175,
79–82, 85–105, P 406
107–25, 127–42, paradox 45–6, 50, 71, 79, – goal 43, 406
163–4, 367–8, 371, 82, 85–105, 107–25,
378, 384, 406, 408, 412 127–42, 163–4, 207, R
– self-observation 49–50, 269, 271, 281, 367–8, rational 10, 12, 59, 85–7,
56–7, 71–2, 81–2, 371, 378, 384, 406, 93–101, 103, 121,
246–7, 270–1, 333–4, 408, 412 130–1, 141–2, 165,
371–6, 387, 393–5, paralogy 127, 129–30 173–8, 189, 200, 208,
397–8 paradigm of adaptation 210, 220, 229–33, 249,
ontological 90, 173–5, 368–71, 384 251–53, 262, 381,
177, 182, 189, 232–3, perception 32–3, 38, 47, 386–7, 390–1, 420
272, 279, 370, 386–8, 76–7, 101–105, 132, redundancy 62, 97, 184,
391–2, 400, 412 147–8, 194–7, 420
open systems theory 22, 200–203, 206–12, re-entry 49–52, 74, 76,
61, 76, 370 239–42, 407 150–53, 271, 283,
operation 22–3, 47, person 31–34, 43–4, 69, 316–7, 320–22, 333,
50–52, 55, 57–61, 121, 126, 148, 150–53, 341, 371, 375, 399, 409
67–8, 71, 73, 75–7, 170, 178, 180–182, reflection 56, 59, 80–1,
80–1, 85–92, 101–105, 187, 228, 235, 358, 134–5, 140–42, 194,
116–7, 149–50, 156, 405, 408 208, 232–3, 266, 269,
231–37, 270–1, 308, personnel 42–4, 113, 147, 300, 357, 359, 375, 379
325–33, 359, 374–5, 163, 176, 295–6, reflexivity 78, 175, 232–3,
378, 396–8, 406 298–9, 406, 408 288–93, 299–300, 371,
operative closure 22–3, phenomenology 222, 374–5, 377
25, 31–32, 36–8, 40, 224–6, 336–8, 420 resource 60, 116, 187,
44–5, 51, 53, 58–65, position 44, 92, 94–5, 204, 219, 237–242,
68–9, 72, 76–9, 95, 158, 196, 207, 228, 288, 291, 296, 298,
108, 145–146, 148–9, 285–6, 294, 296–8, 409 300, 328–30, 345, 369,
156, 168–70, 187, 211, post-decisional regret 253 394
229–30, 331–33, 385, postmodernism 129, resource dependence
394, 408 262–81, 418 theory 250, 382
organization(al) post-structuralism 216, role 86, 93, 132, 140,
– concept of 39–46, 245 150–54, 169–72,
85–106, 108–11, power 37, 73, 77, 181, 178–80, 182–3, 187–8,
331–335 185, 229, 241, 253, 190, 285, 287, 289,
– culture 44, 100, 199, 255, 264–6, 273, 277, 291, 301, 379, 382–3
215, 322, 408 318, 337–41, 387, 391, routine 44, 116, 140, 182,
– forms of 9, 211, 420 197, 202–211, 216,
249–50, 254, 256–8, practice 94, 96, 101, 127, 220, 232, 234, 236–7,
261, 307–22, 331, 141, 148, 160, 211, 252–3, 328, 378–82,
333–4 215–47, 396 405
– performance 172, 175, pragmatism 79, 91, 134,
190, 389, 398 211 S
– society 172, 190, 255 profession 103, 114, 182, science(s)
– sociology 171, 248, 184, 188, 190–1, – applied 351–53, 363–4
250, 252, 255, 257–61, 209–10, 249, 251, 253, – modern 82, 264–7, 277
302 295, 386 – social 129, 133, 142,
programme 202, 220
463
25 Index 05-11-11 16.33 Sida 464
– system of 9, 12, 36–7, – theory of 132, 171–73, – subsystem 12, 36,

105, 186, 208, 255–7, 179–80, 188–92, 194, 80–81, 155, 175–178,
259, 409, 416 212, 223, 283–4, 302, 184–187, 198–199,
– theory of 79, 264–7, 334, 415 223, 256, 277, 300,
277 – world society 36–7, 325, 335, 338–339,
segmentation 36, 172, 193, 325–6, 330–31, 389–390, 407, 415–417
174, 179, 183–4, 190, 335–8, 341, 346–7, 418
285, 336, 338–9, 409 stratification 36, 172, T
selection, selectivity 174, 180, 183–4, 190, tautology 39, 76, 121,
28–31, 34, 39–41, 285, 409 369, 372, 376–8, 380,
55–6, 61, 66–71, 74, structural coupling 24, 384
76–8, 85, 93–4, 103, 30–32, 37–38, 101, trust 123, 134, 207, 282,
105, 117, 147, 153, 105, 150, 173–174, 298, 301, 303, 320,
157, 162–3, 169, 186–188, 197, 211, 351–2, 421
176–7, 183–6, 190, 239, 283, 394, 410 truth 37, 79, 89, 106,
203–5, 207–8, 210, structural level 32, 44–5, 185–6, 222, 264, 275,
252–254, 259–60, 52, 56, 73, 96–7, 117, 277
274–5, 289, 292, 296, 163, 181, 338, 362
332–3, 344, 356, 360, structuralism 66, 93, U
363, 399, 405 215–6, 224–5 uncertainty (absorption)
self-description 34, 56–7, structure/agency 9, 40–43, 56, 62, 89, 93,
95, 151, 153, 173, 180, 230–37 96–100, 104–6, 108,
233, 316, 376, 378, subconscious structure 115–6, 118–20, 122–3,
383, 393, 414, 418, 420 224, 266 125–6, 146–7, 158,
self-organization 24, 56, subject 102, 104, 216, 166–8, 173, 180, 190,
60, 75, 123, 389, 409 223–31, 234–5, 237, 200, 203, 240, 252,
self-reference, self-referen- 262–3, 267–8, 414 258, 292, 298, 410
tial symbiotic mechanisms
– basal 269, 271, 374, 197, 243–4, 246 V
379, 409 symbolic generalized values 93, 106, 120, 177,
– systems 11, 23, 37, 46, media 173, 179, 185, 182, 209, 220, 320,
52, 57, 62, 64–75, 80, 188, 190 322, 333, 362
89, 107–8, 111–12, system variety 99, 162–3, 184,
124–5, 147, 177, 259, – autopoietic 11, 22–31, 203, 368–9, 371, 420
269–71, 296, 333–4, 54–63, 64–82, 310–11,
339–41, 365–79, 405, 409, 413–15 W
383–5, 394, 409, 414 – biological 21–26, 61, world horizon 337,
– theories 10, 275, 278, 64–5, 69, 74, 76–7, 80, 341–47
281 82, 174, 187, 308, world polity 249–51,
semantics 75, 222, 243, 311–312 253–4, 260
418 – system/environment dis- world society 36–7, 193,
social order 179, 220, tinction 9, 22–7, 51–53, 325–6, 330–31, 335–8,
242, 263, 392 104, 147, 150, 175, 341, 346–7, 418
social practices 215–47 178, 189, 201, 268–9,
society 273, 316–7, 354,
– modern 9, 36, 75, 142, 360–61, 370–71, 375,
172–4, 177–188, 190, 407, 413
222, 249–50, 255–7, – mental, psychic 25–7,
259–61, 267–8, 277–8, 31–34, 64–5, 101, 156,
284–91, 293–4, 335–6, 159–62, 221–22,
340, 407, 415 227–30, 235, 406–9,
– system of 12–13, 36, 414
52, 69, 79, 81, 149, – social 9, 11–13, 28–35,
155, 182, 189, 335–6, 54–63, 64–82, 226–30,
418 409, 413–5
464

Dokumen - Pub Niklas Luhmann and Organization Studies 9788763003049 876300304x

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Dokumen - Pub Niklas Luhmann and Organization Studies 9788763003049 876300304x

Uploaded by

Copyright:

Available Formats

SERIES EDITORS:

Stewart R. Clegg &

Kai Helge Becker

Niklas Luhmann and

Liber & Copenhagen Business School Press

© CBS Press and the authors 2006

1st printed edition 2006

Series editors: Stewart R. Clegg and Ralph Stablein

ISBN (e-book edition): 978-87-630-0304-9

Advances in Organization Studies

Advances in Organization Studies is a channel for cutting edge theoretical

PART I: THE THEORY OF AUTOPOIETIC SOCIAL SYSTEMS

PART II: ORGANIZATION, DECISION AND PARADOX

PART III: ORGANIZATION, INTERACTION AND SOCIETY

PART IV: LUHMANN’S THEORY IN THE CONTEXT

11. Luhmann’s Systems Theory and the New Institutionalism

PART V: FORMS OF ORGANIZATION

PART VI: IMPLICATIONS FOR MANAGEMENT AND

PART VII: GLOSSARY AND BIBLIOGRAPHY

Munich and London, Autumn 2005

David Seidl and Kai Helge Becker

Some important aspects of Luhmann’s theory

Probably the most striking among Luhmann’s conceptual achievements

cal notion of organization as static and hierarchical entities that coordinate

The reception of Luhmann’s work on

The structure of Luhmann’s œuvre

The contents of this book

Part I of this book is dedicated to Luhmann’s theory of social systems in

The second chapter, “The Concept of Autopoiesis” (Niklas Luhmann), is

“Organization and Interaction” (David Seidl) explores how the relation

phenomena in common: their focus on culture and the theoretical move of

In the first chapter, “Analysing Forms of Organization and Management”

to offer the possibility of developing a theory of strategy, in which the recur-

The Theory of Autopoietic

The Basic Concepts of

The concept of autopoiesis

In contrast to allopoietic systems (< Greek: allos = other; poiein = to pro-

Living systems are cognitive systems, and living as a process is a process of

In this sense, the operations of an autopoietic system are defined as its

(re-)production stops, the system disappears. For instance, if a plant stops

b. Luhmann’s general, transdisciplinary concept of autopoiesis

Luhmann suggests that we speak of autopoiesis whenever the elements of a

1. Level Autopoietic Systems

2. Level Living Systems Psychic Systems Social Systems

3. Level Societies Organizations Interactions

Figure 1: Types of autopoietic systems.

Through this shift from a reproduction of relatively stable elements, to a

As a consequence of deontologising the concept of element, the concept of

Social systems use communications as their particular mode of autopoietic

In order to understand this conception of social systems, we have to clarify

A central point in Luhmann’s concept of communication is that the three

Communication is a genuinely social – and the only genuinely social – oper-

Thus, although psychic systems are necessarily involved in bringing about

network of communications that “produces” the communications. In other

b. Interpenetration: the relation between social and

an autopoietic system presupposes the complex achievements of the

The simultaneous (but separate) autopoieses of psychic systems is constitu-

cause perturbations in the psychic systems and to receive perturbations from

Language is a purely social phenomenon (psychic systems do not think in

c. Communication and action

[T]he perfection of communication implies understanding and understanding

Thus, a central element of Luhmann’s concept of communication would be