Discursive Analytical Strategies: Understanding Foucault, Koselleck, Laclau, Luhmann

Discursive analytical PB.
A-W (reprint) 6/10/08 10:11 Page 1
The creation of knowledge and discourse is integral to modern society and is no longer
DISCURSIVE ANALYTICAL STRATEGIES

the preserve of scientific institutions such as universities. Political institutions are
being established which have the explicit purpose of producing meaning and language,
spin-doctors shape ‘knowledge’ to sway public opinion, while business and industry is
taking on the ‘management of meaning’. All influence how we understand our society
and shape political agendas.
More than ever, therefore, we need to be able to step back and question the production
of meaning. This exciting and innovative book fills a gap in the growing area of discourse
analysis within the social sciences. It provides the analytical tools with which students
and their teachers can understand the complex and often conflicting discourses across
a range of social science disciplines.
Examining the theories of Foucault, Koselleck, Laclau and Luhmann, the book:
• focuses on the political and social aspects of their writing;

• discusses and combines their theories to suggest new analytical strategies for
understanding society;
• combines theory with practical illustrations.
A bestseller in Denmark, this English edition is vital reading for anyone with an interest
in discourse analysis. It will also be invaluable to anyone looking at the analytical works
of Foucault, Koselleck, Laclau and Luhmann. Students will find the clear exposition of
the theories and strategies, supported by an easy-to-digest, easy-to-read layout, which
includes summaries and boxed examples highlighting the relevance of analytical
DISCURSIVE
strategies to social and policy research.
Niels Åkerstrøm Andersen is Professor of Political Management at the Department

of Management, Politics and Philosophy at the Copenhagen Business School.
ANALYTICAL
STRATEGIES
Understanding Foucault,
Andersen
Koselleck, Laclau, Luhmann
Niels Åkerstrøm Andersen
BT001_cover_Bibliovault.indd 1 9/5/2013 11:41:10 AM

Discursive analytical strategies
Understanding Foucault, Koselleck,

Laclau, Luhmann
Niels Åkerstrøm Andersen

First published in Great Britain in January 2003 by
Policy Press North American office:
University of Bristol Policy Press
1st Floor c/o The University of Chicago Press
1-9 Old Park Hill 1427 East 60th Street
Chicago, IL 60637, USA
Clifton
t: +1 773 702 7700
Bristol BS2 8BB
f: +1 773-702-9756
UK
Tel +44 (0)117 954 5940 e:[email protected]
www.press.uchicago.edu
e-mail [email protected]
www.policypress.co.uk
© The Policy Press 2003

Transferred to Digital Print 2008
British Library Cataloguing in Publication Data
A catalogue record for this book is available from the British Library
ISBN 978 1 4473 4220 5 EPDF

Niels Åkerstrøm Andersen is Professor of Political Management at the Department of
Management, Politics and Philosophy, Copenhagen Business School, Denmark.
Cover design by Qube Design Associates, Bristol.
Front cover: photograph supplied by kind permission of Mary Shaw.
The right of Niels Åkerstrøm Andersen to be identified as author of this work has been asserted by
him in accordance with the 1988 Copyright, Designs and Patents Act.
All rights reserved: no part of this publication may be reproduced, stored in a retrieval system, or
transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise
without the prior permission of The Policy Press.
The statements and opinions contained within this publication are solely those of the author and
contributors and not of The University of Bristol or The Policy Press. The University of Bristol and The
Policy Press disclaim responsibility for any injury to persons or property resulting from any material
published in this publication.
The Policy Press works to counter discrimination on grounds of gender, race, disability, age and sexuality.
Contents
List of tables and figures IV
Acknowledgements V
Glossary VI
Introduction IX
1 The discourse analysis of Michel Foucault 1
2 Reinhart Koselleck’s history of concepts 33
3 The discourse theory of Ernesto Laclau 49
4 Niklas Luhmann’s systems theory 63
5 A hall of mirrors or a pool of analytical strategies 93
References 119
Appendix A: Examples of other analytical strategies 125
Appendix B: Further reading 127
Index 131
III
List of tables and figures
Tables
0.1 Method versus analytical strategy XIII
1.1 Foucault’s analytical strategies 31
2.1 Koselleck’s analytical strategies 48
3.1 Laclau’s analytical strategies 62
4.1 First- versus second-order observation 71
4.2 The differentiation of society 83
4.3 Luhmann’s analytical strategies 92
5.1 Analytical strategies compared 97
5.2 Problems of conditioning related to the analytical strategies 114
Figures
1.1 The genealogy of psychoanalysis 21
1.2 The elements of self-technology 25
1.3 Dispositive analysis 28
1.4 The double movement of dispositive analysis 29
2.1 Synchronous versus diachronic 47
3.1 Chains of difference and equivalence 55
3.2 The relationship between deconstruction and discourse analysis 58
4.1 The sign of difference 65
4.2 The marked difference 65
4.3 A difference observed through a difference 66
4.4 The distinction system/environment re-entered as a part of itself 67
4.5 The sign of re-entry 68
4.6 Re-entry of the distinction system/environment 81
4.7 The calculus of form 84
4.8 Media/form staircase 86
4.9 The relationship between form analysis and semantic analysis 89
4.10 The relationship between differentiation and semantics 90
4.11 The relationship between systems analysis and media analysis 91
5.1 Analytical strategy 117
IV
Acknowledgements
This book presents Michel Foucault, Reinhart Koselleck, Ernesto Laclau and
Niklas Luhmann as analytical strategists. There is a reason for this. I have for
some years been studying the more fundamental changes in European societies.
Trying to capture these fundamental changes, it is easy to become a victim of
current self-descriptions and future images in society. Instead of grasping the
changes you become a prisoner of the discourse producers of the day and
their strategically constructed future images. It is easy to confuse the actual
changes with the images and, in so doing, one may become an instrument in
confirming the discourse producers’ political predictions of trends: globalisation,
the knowledge society, the network society and the dream society. Studying
change, it is essentially difficult to maintain the necessary distance to the object;
to the society that should be studied.
In my terminology, this problem is one of analytical strategy: how can you
critically analyse a coherence of meaning of which you are a part? Foucault,
Koselleck, Laclau and Luhmann explored this problem and it is the four very
different strategies of handling it that form the theme of this book.
The book has been written in a very inspiring environment at the
Department of Management, Politics and Philosophy at Copenhagen Business
School. I would like to thank my colleagues for their challenging discussions
and fruitful comments. In particular I would like to thank Christina Thyssen,
Asmund Born and Hanne Knudsen.
V
Glossary
An analytical strategy is a second-order strategy for the observation of how
‘the social’ emerges in observations (or enunciations and articulations). The
elaboration of an analytical strategy involves shaping a specific gaze that allows
the environment to appear as consisting of the observations of other people or
systems.
Communication is a selection process, consisting of a synthesis of three selections:

(1) selection of information, that is, what is to be communicated; (2) selection
of form of message, that is, how the information is to be communicated; and
finally, (3) selection of understanding, that is, what should be understood about
the message.
Concept (Koselleck) is a word condensing a wide range of social and political

meanings. Concepts comprise an undecided abundance of meaning, a
concentration of meaning, which makes them ambiguous. Precisely through
its ambiguity, the concept creates a space of signification, which is open to
interpretation and can become a semantic battlefield.
Deconstruction is showing how differences are contingent, that is,

deconstruction is about retracting or unpacking differences to show that they
are not differences at all – that the ‘bar’ (/) between two opposing elements,
which isolates one from the other, cannot be maintained. In short, to
deconstruct is to demonstrate the impossibility of a distinction.
Discourse (Foucault), or more precisely a discursive formation, is a system of

dispersion for statements. It is not a structure existing on a level different from
statements; discursive formation is simply the regularity of the irregular
distribution of statements.
Discourse (Laclau) is a structural totality of differences that is a result of an

articulatory practice. The totality is, however, never fully achieved. Discourse
is a never-completed fixation process that takes place through articulation
within a field of discoursivity with drifting relations.
Form (Luhmann) is the unity of a difference.
A guiding distinction is the distinction that can define the frame for second-
order observations. For the second-order observer, the guiding distinction
divides the world and dictates how the world can be observed.
VI
Glossary
Hegemony is only possible when something exists that can be hegemonised.

This only occurs when a discourse lacks final fixation, when the discursive
elements hold a surplus of meaning, and when the signifiers are not irreversibly
linked to the signified. Consequently, hegemony signifies the never-concluded
attempts to produce a fixed point of discourse, to which there will always be
a threat.
Information is a difference that makes a difference to a system. It is the system

itself that selects information from pure noise and irritation on the basis of its
own operation of distinction. That is, the system needs expectations to be
surprised.
Method is the rules and procedures required to produce scientific knowledge.
Nodal points are privileged discursive points, which serve to arrest the flow of
relationships without ever becoming a real centre of the discourse. The
discursive struggle about the construction of nodal points is, so to speak, a
struggle about the conditions of conflicts within a specific discourse.
Re-entry is the operation by which a difference is copied into itself, thus

becoming a part of itself.
Second-order observation is an observation of an observation as an observation.

That is, not reducing an observation to something else, for example, to an
ideology, individual intentions, subjectivity and so on. An observation is an
indication within a frame of difference.
Semantics (Luhmann) are characterised as the accumulated amount of

generalised forms of differences (for example, concepts, ideas, images and
symbols) available for the selection of meaning within systems of
communication.
Social system is an autopoietic system of communication, which defines itself

in the construction of its environment through communicative descriptions.
A social system is a system only in relation to its environment, and the
environment, in turn, exists only in relation to a system. Both system and
environment are internal structures of communication. However, environment
is not ‘reality’ as such. Environment consists of that which is defined by the
communication as its relevant surroundings. Any system therefore is identical
to itself in its difference (and only in its difference) from the internal
environment construction.
VII
Statement is a function of existence that enables groups of signs to exist. The

statement is the smallest unit, which brings forth phenomenon through
enunciation. We are therefore able to recognise the statement by its momentary
creation rather than by its appearance as sign, sentence, book or argument.
Statements are positive events that produce existence through enunciation.
This function of existence contains at least four aspects: object, subject,
conceptual network and strategy.
Subject positions are the spaces from which one speaks and observes in a
discursive formation. Subject positions have rules for the acceptance of certain
individuals into the spaces, rules for acceptance regarding in what situation
the subject position can be used as a platform for speaking and observing, and
rules for the formation of statements once one has assumed a specific position.
Subjecting indicates that an individual or a collective is proclaimed to be a

subject within a specific discourse. The individual or the collective is offered
a particular position in the discourse from which they can speak and act in a
meaningful way. Subjecting thus signifies the space in which the discursive
individual receives itself.
Subjectivation happens when the individual or the collective is not only formed
as a subject but also wishes to be the subject. Subjectivation signifies the space
in which the individual gives itself to itself.
VIII
Introduction
From method to analytical strategy

The social sciences currently exist in the light of constructivism. A number of
social scientists see themselves as different types of constructivists, and the
constructivist spectrum is broad – from Peter Berger and Thomas Luckmann
via Pierre Bourdieu to Michel Foucault and Niklas Luhmann. So, why is
constructivism so popular now? In my view it has simply become too difficult
not to be a constructivist. Regardless of the field of social science one focuses
on, the most noticeable thing is change, and changes often touch upon and
challenge fundamental values, raising questions about the constituent character
of what we see.
If we look at politics, the European nation states are being integrated in the
European Union and the World Trade Organisation – a process that challenges
the value of the sovereign state. If we look at medicine, new technologies
(such as prenatal diagnosis) question the value of individuality. The welfare
state is experiencing a growing number of reforms based on spending politics,
which questions the value of solidarity. We frequently find that our categories
do not suffice; they seem inadequate. They appear to point to a former order
of society that no longer exists. This is the case with, for example, the notion
of state sovereignty. We use the ‘conventional concept’ to evaluate the concept’s
change. We evaluate European integration by using categories belonging to a
dated world order. We analyse the future by the standards of the past.
Our experiences of the insufficiency of categories encourage us to take a
step back in order to look at the categories themselves – their construction,
their history and their position within the fields of our focus. Rather than
analysing European integration from the perspective of ‘state sovereignty’, we
inquire about the concept of sovereignty itself and the historical conditions of
its existence and transformation. We inquire into how the idea of the sovereign
state has been shaped and how it may be threatened by the European Union.
We look at which new concepts could be emerging, which new meanings the
old concept of sovereignty is given, how a new institutionalised framework is
built around a new understanding of sovereignty and so on. Rather than
forming an immediate judgement of new technologies, we make inquiries
about the evolution of technology and of individual values. In what way do
new technologies change our understanding of life, individuality, destiny,
responsibility and freedom? To what extent do new technologies not only
create new ideas, but also revive old ideas and reopen old questions, for example
IX
the question of social hygiene dating back to the social reforms of the 1920s
and 1930s?
At the same time, knowledge has become an intrinsic part of the organisation
of society in a much more strategic way than before. Today sociologists speak
of the knowledge society, the corporate community talks about knowledge
management. Generating knowledge is no longer exclusive to independent
scientific institutions such as universities. In the corporate community,
knowledge-intensive businesses have been established, such as the big
international consulting firms PLS Consult, Deloitte & Touche and McKinsey.
They do not simply operate and utilise organisation theories developed by
the universities. They are themselves generators of theories and concepts, and
their concept development not only serves the purpose of scientific knowledge,
it also aims to seduce and sell.
Within government, similar discursive institutions have been established in
order to develop scientific discourses and to diagnose the condition of society,
with the intention of controlling the political agenda, defining the framework
for negotiation and installing a sense of responsibility in organisations, political
parties and individuals. One European example is the European Environment
Agency, which was established in order to collect data and make objective
decisions, and to function as a creator of environmental knowledge or discourses
and as a campaigning organisation. Other European examples are the Group
of Policy Advisers and the Joint Research Centre, both of which were created
under the auspices of the European Commission, and the European Monitoring
Centre on Racism and Xenophobia.
In relation to this, some people speak of knowledge politics. Today’s researcher
faces numerous investigations, concepts, problems, solutions, theories,
descriptions and explanations, embedded in the spirit of science, but also with
a scientific standing and political, strategic and administrative function that
remains obscure. Once again, this calls for research that takes a step back and
questions these investigations, concepts, problems, solutions, theories,
descriptions and explanations:
• How have they come into being?

• Which strategies and policies have shaped them?
For example...
Are expressions such as ‘the complete colleague’,‘lifelong learning’ or ‘dedicated

employees’ innocent indications? Or are they a sign of a new form of training,
which means that as an employee or citizen you are not only required to
perform your duty but to invest your whole life, your soul and your
involvement in your organisation or in being a ‘good’ citizen?
X
Introduction
Also, it appears that the subdivision of the social sciences into branches of
knowledge related to particular functions in society has become increasingly
problematic. Economic science applies to economy; jurisprudence applies to
the courts; media studies to the mass media, and so on. Today, more than ever,
it is evident that the different fields are helping to invalidate each other. Each
field has its own discourse and its own concepts, its own limited resonance;
each field can only communicate with itself without regard for the other
fields. It becomes difficult to identify one’s research with one area of research
exclusively, knowing that what looks like progress in one field might very well
be detrimental to others. For example, the development within the media of
the combination of ‘hidden cameras’ with ‘political journalism’ may have created
a ‘new genre’, but this development fundamentally disturbs the political process.
Methods have been developed to put a price on ‘care’ and ‘unhealthy lifestyles’,
but this produces totally unpredictable events within other fields – both positive
and negative – possibly without anyone realising it and consequently without
the possibility for self-correction. Once again, it therefore seems obvious that
we need to take a step back and to question the evolution of the different
fields, their communicative closure on their own functions, the limited reflective
ability of the individual fields, and their attachment with and detachment
from other fields.
Thus, we see the outlines of a new form of questioning appear, which does
not merely question actions within a field but which also questions the way
questions are asked in the field; questions the emergence of the categories, the
problems, the arguments, the themes and the interests. This form of questioning
entails a theoretical shift from the primacy of ontology to the primacy of
epistemology (Pedersen, 1983), from first-order observations of ‘what is out
there’ to second-order observations of the point that we are watching from
when we observe ‘what is out there’. From being to becoming.
Ontology and epistemology

Every scientific position entails ontology and epistemology. They point to
fundamental questions in every science. Ontology is concerned with the
question of basic assumptions about the world and the being of the world;
epistemology is concerned with the question of basic assumptions about the
precondition of cognition of the world. But not only are there different
answers to the two questions; there are also different tensions and hierarchies
between them. It makes a fundamental difference whether one begins by
answering the question of ontology or by answering the question of
epistemology.
An ontologically over-determined theory is one that starts with the question of
being and asks:
• What does it mean that something exists?

• What are the fundamental possibilities for deciding whether the statement
of the theory is true, objective, or scientific?
XI
It is characteristic of ontologically founded thinking that it moves rather swiftly

away from these more fundamental questions to the question of method
(Habermas, 1972, pp 71-91). The principal methodological question is:
• Which procedures and rules are necessary in order to obtain theoretical

knowledge?
For example...
When a survey of the public administration is conducted by sending

questionnaires to all leaders of institutions, the method determines the ontology
of the administration as establishments represented solely by their leaders.
One could say that the method’s priority within an ontologically founded thinking
nullifies the question of the essential nature of the administration. By its choice
of method, it suspends inquiries about the construction of the administration
as a particular social phenomenon, the specific characteristics of the
administration, the administration’s dependency on history, and its function
and particulars in relation to other systems. It generally suspends inquiries
about the unity of the administration. In conclusion, an ontologically founded
thinking produces a presupposition of the object. What you see is taken for
granted. It ontologises the administration by reducing it to a certain form of
reality that is unquestionable.
Within ontologically over-determined science, methodology holds ultimate

priority. The method decides “what exists, or what reality is” (Pedersen, 1983,
p 35).
Epistemologically over-determined thinking, on the other hand, is in its nature of
second order. It does not primarily ask what but how. It asks:
• In which forms and under which conditions has a certain system of meaning
(such as a discourse, a semantics or a system of communication) come into
being?
• What are the obstacles to understanding the possibilities of thinking within
– but also critically in relation to – an already established system of meaning?
• How and by which analytical strategies can we obtain knowledge critically
different from the already established system of meaning? (Pedersen, 1983)
Whereas ontologically over-determined thinking ontologises the object,

epistemology de-ontologises its object.
Of course, one cannot simply escape ontology by beginning in epistemology
and posing the question as one of historical and social conditions of cognition.
Every epistemology entails ontology, even in the ‘French’ version of this theory.
But, when you give priority to epistemology, you can work with an empty
ontology. Empty ontology does not mean that you do not have ontology;
rather, it gives an ontological subscription of emptiness to being. It is an
XII
Introduction
ontology that is restricted in its approach to reality, to only saying ‘reality is’.
The object, however, is not presupposed. In this way, epistemology is concerned
with the observation of how the world comes into being as a direct result of
the specific perspectives held by individuals, organisations, or systems. It also
asks how this causes the world – in the broadest sense – to emerge in specific
ways (while the observers themselves also emerge as individuals or
organisations). Therefore, an epistemological starting point poses not a question
of method, but a question of analytical strategies.
Analytical strategy
Analytical strategy does not consist of methodological rules but rather of a
strategy that addresses how the epistemologist will construct the observations
of others – organisations or systems – to be the object of his own observations
in order to describe the space from which he describes. From an epistemological
point of view the perspective constructs both the observer and the observed.
Hence analytical strategy as a way to stress the deliberate choice and its
implications, and to highlight that this choice could be made differently with
different implications in respect of the emerging object. The problem of the
epistemological restriction to ‘how’ questions and not ‘what’ or ‘why’ questions,
is that it constructs the researcher as a ‘purist’ (that is, as one who does not
assume anything in advance about the object to be studied). However, one
needs to assume something in order to recognise and observe the object. This
is the basic condition of working with analytical strategies.
Simply put, the difference between method and analytical strategy can be
viewed as follows:
Table 0.1: Method versus analytical strategy1

Method Analytical strategy
Observation of an object Observation of observations as observations

The goal is to produce true knowledge The goal is to question
about a given object presuppositions, to de-ontologise
What rules and procedures are needed Which analytical strategies will enable us to
to produce scientific knowledge? obtain knowledge, critically different from
the existing system of meaning?
This shift from method to analytical strategy raises a number of questions that
still have not been answered, and possibly not even conceived of in a satisfactory
manner. Within analytical strategy, the question of scientific knowledge poses
itself in a different way. Other questions appear and become essential, while
certain methodological questions become irrelevant. It is a problem that many
constructivist studies do not realise this. The stringent methodological question
often fades or disappears because it is not compatible with second-order
observations, but far too often it is not superseded by analytical-strategic self-
discipline. Instead of method, we often see a pragmatic examination of
XIII
procedures; the result is often sloppy and inept. Many constructivist studies
lack scientific meticulousness in the shape of thorough accounts of their
analytical strategies. Often, it is extremely difficult to identify the basis of
scientific studies and, unfortunately more often than not, the criticism raised
by more mainstream-positivist positions is well justified.
Let me provide just two examples of analytical-strategic difficulties raised
by the epistemological turn.
The first example looks at analyses directed at the construction of social

identities, such as the construction of an administration, a social movement or
a new discourse. Here, the analytical-strategic question of when something
can be seen as constructed promptly suggests itself. This is not a methodological
problem – it cannot be solved using methodological standards. It is an analytical-
strategic problem of establishing the lens through which the evolution of a
social movement can be seen.
Without an accurate identification of the conditions necessary for a social

movement to be seen as a social movement (and not just as a group or an
organisation trying to present itself as a movement), it is virtually impossible
to study the construction of a specific social movement and equally impossible
to criticise the study of its construction. The more exact the identification,
the more sensitive to the empirical is the analysis. Unfortunately, it is not
simply a question of choosing a definition, because ‘social movements’, both
in the shape of a social system and as a concept, have an historical evolution
of their own to which an analytical strategy must be sensitive.
The other example looks at studies of change. A methodological approach to

change would generally be interested in explaining changes – organisational
changes, political reforms and so on. In the search for explanations and in the
formulation of methods able to examine what causes change, change is typically
ontologised. The way change occurs is a given. Whether or not change is seen
at all or, if so, in what respect, is ignored.
Studies of change within epistemology, on the other hand, formulate the question
of when change can be considered a change. From an epistemological
perspective one immediately stumbles across the analytical-strategic difficulty
that any formulation of change is based on the observer’s construction. Change
can only be characterised within the framework of specific differences: a
change must have a beginning and an end. Prior to a beginning one must
assume an end – whether or not the end provides actual closure. Thus, the
nature of any formulation of change is teleological, and the analytical strategy
of the epistemological observer must reflect this fact through inquiries and
analytical-strategic decisions about the position from which one describes
change and, consequently, the distinctions that determine what appears as
change to the observer.
XIV
Introduction
The problem holds different definitions in various epistemological programmes.

Some are concerned with defining the conditions of an epistemological
breakdown, that is, the criteria for defining that point when the observed
system of meaning is no longer the same but new (see for example, Regnault,
1983). Luhmann’s evolutionary theory focuses on, for one thing, the moment
when one principle of differentiation for the construction of new systems of
communication is superseded by another (Luhmann, 1990a). In Laclau’s
discourse analysis, the focus is on the moment when a new nodal point takes
over the function of fixating the decentralised elements in the discourse.
About the book

The aim of this book is to contribute to the development of an analytical-
strategic language. I attempt to do so by introducing four constructivists with
a communication angle – I present them here as analytical strategists and ask
the following questions:
• What characterises the way they ask?

• How do their epistemological programmes apply to the observation of
observations, the description of descriptions, the signifying of signs?
• How does each of them construct their eye to second-order observations?
This book also juxtaposes four different theor ies about society as
communication or discourse in regard to their analytical strategies. The four
theories are Michel Foucault’s discourse analysis, Reinhart Koselleck’s concept
analysis, Ernesto Laclau’s discourse theory and Niklas Luhmann’s systems theory.
This book is not a broad introduction to the four writers. Rather, it
presupposes a certain familiarity with or interest in some form of discourse
analysis. Neither does the book aim to weigh the writers’ powers of explanation
against each other, or to provide some kind of theoretical synthesis. Rather, it
attempts to to invoke an analytical-strategic discussion of different ways of
defining society as communication. It also aims to see what possibilities for
observation unfold when the concern is no longer given objects but, instead,
the question of how problems, individuals, interests – all kinds of social identities
– come into existence as and within communication. Which analytical
difficulties do we encounter when the innocence of the empirical collapses –
when we can no longer pretend that ‘the object out there’ discloses how it
wants to be observed, when we know that it is our ‘eye’ that makes the object
appear in a particular way?
Once again, I wish to stress the fact that this book is a contribution to the
development of an analytical-strategic discourse. I am not certain that my
way of presenting the problem is the ultimate one – my argument that the
analytical-strategic problem is different from the methodological problem might
be incorrect. I do not wish to give the answers to all possible analytical-
strategic problems. Neither do I wish to present a manual for discourse analysts.
XV
Rather, I wish to create an outline for the conception and discussion of

analytical-strategic problems and, in that respect, the book ought to pose
more questions than answers.
The four writers

In this book, I have chosen four ways of seeing – the discourse-analytical, the
concept-historical, the discourse-theoretical and the systems-theoretical –
because, within my approach, they generate four very different epistemological
programmes for second-order observation. Foucault is discourse about
discourse, Koselleck conception of concepts, Laclau significations of signifiers
and Luhmann observation of observations. The differences between them are
used to show the abundance of possible analytical strategies and the abundance
of ways in which analytical-strategic problems can be conceived. Traditionally,
one is either a systems theorist or discourse analyst. In my approach, the four
perspectives fundamentally agree about the emphasis on a new epistemology.
The disagreements are over how.
Despite their differences, all four share the following notions:
1. Their theories are all programmes for second-order observations.

2. All four theoretical programs are anti-essentialist. Reality contains no essence
that requires observation and examination of any specific order. They all
presuppose an empty ontology without ascribing anything definitive to
reality.
3. They all reject philosophy of consciousness, each in their own way. Luhmann
does so by defining society as communication in which different forms of
consciousness cannot communicate with each other, but can only exist as
the surroundings for communication. Foucault, by a de-centring of the
subject into numerous discursive subject positions. Koselleck, by
presupposing the forming of ideas to the forming of subjects. Finally,
Laclau, by locating the subject in the gap between the uncompleted structure
and undecidability.
4. They all reject an ontology of action, that is, the notion that at least actions
are real and therefore able to serve as an objective point of reference for
examinations within the social sciences. On the contrary, actions are seen
as discursive, semantic, or communicative attributes. Actions can be
attributed to subjects or systems, or they can attribute them to themselves
in the process of constructing a responsible or liable self.
5. Finally, they all settle with the notion of criticism. None of them employs
a distinction between uncritical positivism or mainstream thinking on the
one hand and themselves as critical on the other. Rather, they are critical
of any thinking that claims to be critical! None of them believes that there
is a place, an argumentative platform, from which one can be critical in any
universal sense. Even so, the rejection of the idea of criticism, and particularly
critical theory, plays itself out differently in the four studies.
XVI
Introduction
To elaborate the final point, Luhmann’s criticism of critical theory is first and
foremost a critique of the notion that there could be a place from which the
critic can describe society as a whole. According to Luhmann, there are only
systems-relative descriptions of society, and these can never be absolute.
Furthermore, Luhmann denounces critical theory for focusing more on how
the world is not (that is, how the world disappoints the norms of the critic),
than on how the world is (Luhmann, 1991). Instead, Luhmann’s ideal is the
unsentimental gaze. Laclau rejects the notion of criticism through a
deconstruction of the figure of emancipation as the domicile of the critic, that
is, the notion of freedom without power (Laclau, 1996a). Koselleck rejects the
notion of criticism through a conceptual-historical analysis of the concept of
criticism (Koselleck, 1988). He examines the conditions of its origin and its
decline as hypocrisy. Finally, Foucault insists that any truth is always founded
on an injustice, which also pertains to his own discursive analyses. Foucault’s
alternative to the critical researcher is:
an intellectual, who dissolves self-evident truths and universal explanations,

someone who, in the midst of the inertness and restraint of the present,
detects and points out the weak points, the openings and the fields of force,
someone who is constantly moving and who is not too precisely aware of
where he will find himself or what he will be thinking a little into the
future, since he is much too preoccupied with the present.... (Schmidt and
Kristensen, 1985, p 130)
This book contains four chapters about Foucault’s discourse analysis, Koselleck’s
history of ideas, Laclau’s discourse theory and Luhmann’s systems theory
respectively. The four chapters inquire about the enquiry of these four
epistemologists:
• By which fundamental distinctions is the analytical eye shaped?

• How do they create their object?
• Which analytical strategies do they infer?
The output of each of the four epistemologists is very different. Some have
produced reasonably few, but concentrated, theoretical studies, others have
primarily conducted concrete analyses and their analytical-strategic
contributions stem from these works, others have an output that is vast in size,
theoretically as well as analytically. Although I ask the same questions in
regard to the different epistemologists, the approach is necessarily different
with each one, hence the different representations of the four epistemologists
both in size and style. The benefit of this is that it consolidates the width of
the analytical-strategic problems that emerge. In the concluding chapter I
offer a more explicit formulation of what I mean by analytical strategy.
Moreover, in the conclusion I conduct a more direct comparison of the four
analytical strategies, although knowing that this is an impossible venture, I
attempt to line them up and let them reflect each other.
XVII
Delimiting marks
It is obvious to ask, why not four other writers? Why not include, for example,
Habermas, Bourdieu and Fairclough? It could be argued that Jürgen Habermas
really ought to be considered in a book that is founded on the notion of
society as communication, as, since the 1960s, Habermas has observed the
linguistic turn of the social sciences. In my view, however, Habermas represents
a transitional figure between ideological criticism and discourse analysis. He
sees the unavoidable centrality of language and communication, but he does
not fully live out the consequences.
As early as 1965, in his book Knowledge and human interests (Habermas,
1972), Habermas realises that it is necessary to settle with ideological criticism
and the distinction it creates between true and false consciousness. However,
he refuses to give up the universal status of the concept of criticism; rather, he
aimed to find a new way to base criticism. His first proposal for this can be
found in an article about universal pragmatism (that is, dialogue without
hegemony directed at understanding) (Habermas, 1991). Later, he displaces
and reformulates his universalistic concept of criticism several times, for example
in his ethics of discourse (Habermas, 1992) and, most recently, in the weighty
tome Between facts and norms (Habermas, 1996). But, the fact remains that he
never establishes a programme for second-order observation, which
simultaneously admits to its status as first-order observation. His insistence on
a universalistic foundation implies that, when observing communication
through his theoretical programme, it is always by means of an ideal about
communication. This means that Habermas cannot observe observations as
observations but always has to observe them as deviations from a universal
regularity. Habermas has committed himself to observing the way the world
is not rather than the way it is. Thus, Habermas’s programme remains in the
quagmire of ideological criticism, even if it is a very advanced version.
I could also have included Pierre Bourdieu in this book and, indeed,
contemplated it for a long time. Many of Bourdieu’s works contain reflections
of analytical strategy. In my own empirical discourse analysis I have, on several
occasions, employed elements from Bourdieu’s analytics. When comparing
Bourdieu and Habermas, they very often read the same texts but from
completely different perspectives. They both read the works of Searl and
Austin, for example, about different types of speech actions. But while Habermas
inquires about the universal conditions of speech actions directed at
understanding, Bourdieu inquires about the historical institutional conditions
that authorise a specific person as the generator of specific speech actions
(Bourdieu, 1992). It is my understanding that Bourdieu pursues a double
strategy. On one hand, he is without doubt an epistemologist and thinks,
therefore, in terms of analytical strategy – he always inquires about the origins
of problems, fields and speech actions. On the other hand, he appears to have
trouble accepting the cool gaze of epistemology and often resorts, therefore,
to metaphysics of suppression as if the distinction between over and under
were universally designed. He seems to be in constant oscillation between a
XVIII
Introduction
first- and second-order programme for observation; between letting go of

ideological criticism and falling back on it.
Finally, it is pertinent to ask, why not Norman Fairclough? I believe that
the wide base that discourse analysis gained in the 1990s is largely the
achievement of Norman Fairclough (see for example, Fairclough, 1995). In
particular, he has made discourse analysis accessible to linguistically inclined
researchers, but also to certain parts of the social sciences. One of his
achievements is the linking of discourse analysis with text analysis (Fairclough,
1992), and with his tight empirical analysis he has lifted discourse analysis out
of a postmodern theoretical fetishism. In many ways, his ambitions are close
to those of this book. I do believe, however, that he remains within a relatively
more classical methodological thinking, with the result that the epistemological
objective in discourse analysis does not appear clearly in his writings. I find it
hard, in other words, to recognise the unique discourse-analytical eye in
Fairclough. This does not mean that I fail to appreciate his many competent
analyses – it might also just point to my own narrow concept of analytical
strategy (see Chouliaraki and Fairclough, 1999).
My own path to the analytical-strategic question

Let me give a brief, and I hope helpful, account of my own relationship with
analytical strategy. This book has not come into being purely as a result of
theoretical deskwork (although that has also been necessary). I have not
invented the analytical-strategic question out of the blue; it has presented
itself to me in a number of concrete empirical studies in which I have applied
some kind of epistemological analysis. As time has gone by, I have simply
needed to formulate a more concise definition of the analytical-strategic
question. And there is more to come.
My own work has always played itself out in the space between theory and
empirical studies. It has focused on analysing how different ideas have shaped
the world, and the analytical-strategic problem of observing these ideas without
becoming a slave to them and without simply reducing them and judging
them from the perspective of a different idea has always presented itself.
In the book Selvskabt Forvaltning (The autopoiesis of the administration)
(Andersen, 1995), I asked two questions:
1. How is a certain political and administrative discourse invented and

propagated?
2. And what is the effect of the propagation of this discourse on the constituent
character of the administration?
The first question was initiated by the rapid modernisation of the public
sector and by the fact that administrative policies appeared to establish a new
discursive regime. I wanted to describe this regime and its shaping, but I was
not interested in criticising it from a normative position, such as, for instance,
‘the defence of the welfare state’. The first question addresses how something
XIX
is constructed; hence, the analytical-strategic question asks how it is at all

possible to study a discourse under construction. In order to be able to study
the construction or the origin of something, one has to be able to modify the
origin and ascribe unambiguous criteria to the moment when one states that
something has come into being. Here, I adopted and maintained Pedersen’s
distinction between ideal, discourse and institution. The analytical-strategic
problem primarily consisted of defining clear and observable measures for
when something was an ideal, when this ideal could be said to have developed
into a discourse and, finally, when this discourse could be said to have become
institutionalised.
The second question about the constituent effects on the administration
was initiated by the distinctive characteristics of administrative policies. First,
they are policies of the second order: they are policies about the administrative
framework for politics. Second, they are policies that cover the boundaries of
the administration, including the relationships between public and private,
administration and citizen, administration and employee, and also between
administration and politics. Administrative policies appear not only to alter
conditions within the public sector, but also to change the entire conception
of what can be understood as the public sector.
Consequently, the second question addressed systemic disruptions, and so the
analytical-strategic question became: How it is possible to locate systemic
disruptions as such? The dilemma presents itself as a question of when a given
system – here the public administration – is no longer the same system but a new
one. To understand this further, I used Luhmann’s systems theory and defined the
problem as a question of how displacements happen within the internal
differentiation forms of the administrative system. Any administration contains
numerous sub-systems; the differentiation form is the conformity in the sub-
system’s way of being different from each other and from each of their respective
surroundings. If I could demonstrate that the differentiation form was no longer
the same, I would have shown that the administration was no longer the same.
That is to say that a displacement of the administration’s internal differentiation
form means that sub-administrations are no longer formed and developed in the
same way – that their constitution is radically different.
My work with these two questions led me within my dissertation on the
development of administrative politics and the administration in 20th-century
Denmark), to the formulation of a particular analytical strategy which I have
named institutional history (Andersen, 1994). One of the focal points of my
dissertation was to show how the boundaries of the administration were put
at stake and changed through administrative policies. This opened up new
theoretical questions about how the connections between systems are changed
when their boundaries are threatened. Until this point I have dealt with the
boundary between public/private and politics/administration.
In reference to the public/private boundary, I raised the question about the
constitution of the private sector in relation to the redefinition of boundaries
on the part of the public sector (Andersen, 1996). My thesis was that
outsourcing does not simply lead to more market and less politics, but that
XX
Introduction
private corporations that orient themselves towards public markets are forced
to subscribe to the discursive codes of the political system, that is, they are
forced to constitute themselves anew. Thus, the boundary between politics
and economy is no longer more or less identical to the difference between
public organisations and private corporations. Within public markets the
boundary between politics and economy penetrates the internal communication
of private corporations. Here, the question focuses on the relationship between
organisation and function system and on how private organisations can associate
themselves with different function systems, including the political and the
economic systems. The analytical-strategic question is concerned with how
to study the emergence of a private corporation whose identity is not necessarily
consistent but is relative to the association with either the political code or the
economic code. This required the formulation of an entirely new analytical
strategy in regard to institutional history – an analytical strategy capable of
pertaining to the organisational level, separating organisational communication
from other forms of communication.
In a book analysing the history of the outsourcing debate (Andersen, 1997),
again, I required a new analytical strategy. Unlike the study of the origin of
administrative politics, there is no single discourse on outsourcing. Rather
than being an autonomous discourse, outsourcing is and has been a drifting
concept whose meaning is interjected simultaneously in numerous mutually
conflicting discourses. Hence the predicament was not (as in the question of
administrative politics) the examination of the constr uction and
institutionalisation of one discourse, but rather the unfolding of interdiscursive
relationships within time and space surrounding the concept of outsourcing.
The analytical-strategic task involved defining the criteria for when and how
a relationship exists, which also means when and how a discursive separation
exists. For this purpose, I developed a genealogical analytical strategy based
on the works of Michel Foucault.
My latest book, Kærlighed og omstilling – italesættelsen af den offentligt ansatte
(Love and reorganisation – the articulation of the public employee), makes inquiries
into the boundary between public administration and employee (Andersen
and Born, 2001). The book examines the articulation of ‘the employee’ over
the past 150 years in Denmark. In the past, the relationship between employee
and organisation was discussed in an impersonal fashion, in which duties and
rights were the focus. Today, however, the employee is also required to ‘love’
the employing organisation and to share its outlook, otherwise he or she ought
to leave before being racked by feelings of guilt. In the language of employment
policies, this is termed ‘involvement’, ‘initiative’ and the ‘responsible employee’.
The analytical strategy used in this book is inspired by the work of Luhmann
and Koselleck, with a Foucauldian slant. The book combines three analytical
strategies: a semantic historical-analytical strategy, a differentiation analysis and
a form/medium analysis of the coding of employee communication. The
semantic analysis examines the conceptual history of the employee – we look
at how the employee is created in different semantic regimes. The differentiation
analysis relates the semantic development to the development in the ‘unity of
XXI
the administration’ over 100 years during which the public institution changes
from the ‘innocent institution’ via the ‘professionally responsible institution’
into the ‘strategic organisation’. In the strategic organisation, the codes of the
conversation become fleeting. Subsequently, the form/medium analysis pertains
to the codes that become available to the communication about the relationship
between employee and organisation. The focus is on three codes: the legal
code (right/wrong), the educational code (better/worse learning competencies)
and the code of love (loved/not loved). The code chosen for the specific
communication determines who can communicate about what, including what
can be considered as a good argument. Finally, the focus shifts to the way in
which inclusion and exclusion takes place. With love’s codification of the
communication it becomes the responsibility of the employee to be included
in the workplace. Exclusion becomes self-exclusion.
In the encounter with empirical studies, I have, over time, been able to
locate different analytical-strategic problems. I have defined various analytical
strategies pertaining to certain epistemological research questions. But one of
the difficulties I have encountered, which this book addresses, is the lack of a
general, well-defined concept of analytical strategy capable of describing what
an analytical strategy consists of. I have always been of the opinion that there
is not, and cannot, be one analytical strategy capable of answering all the
questions that a creative epistemologist could possibly ask. I have always
encouraged students and others to define their own analytical strategy in
respect to their field of interest and their questions. In many ways, this seems
like an unfair request when one is not able to clearly explicate what an analytical
strategy is or when an analytical strategy is a good strategy; when there is not
even a single book or article that addresses the question in detail.
Since there is not one true analytical strategy, it makes no sense to grant
priority to any theoretical trend (for example that of Foucault or Luhmann).
I have always skipped between different epistemological schools and have
been happy to combine elements if they fit the specific analytical strategy.
Hence, this book should be regarded as an attempt to speak of analytical
strategy in general terms, and to create a field that enables analytical-strategic
discussions across the boundaries of different epistemological trends. The
book may make difficult reading to some people; the ambition, nevertheless,
is not theory for the sake of theory but rather the improvement of empirical
analyses within the epistemological realm. What is the purpose of professing
to discourse analysis if one does not conduct actual, empirically founded
discourse analyses?
Note
1
The distinction should not be understood as a normative regulation against the
use of methods for discourse analysts. The central question is whether a
methodological or a discourse-analytical perspective is primary in the research
design. Naturally, within one analytical strategy different methods can be
introduced which the analytical strategy must then question.
XXII
1
The discourse analysis of
Michel Foucault
ore than anyone, Michel Foucault has developed and created an agenda
M for discourse analysis. Moreover, he has received the widest recognition

within the social sciences. At present, he functions as nothing short
of a guru within research on economic regulation, and has inspired, for example,
analyses of the genealogy of financial management and calculation.
Although I have studied Foucault longer than any other writer, he is probably
still the one I have the most difficulty in presenting. There are several reasons
for this.
First, for the more subjective reason that, having dealt with a particular
work over a long period of time, it is difficult to maintain a distance from it;
gradually you invent your own Foucault that might be more telling of yourself
than of Foucault. If anyone feels that my readings abuse Foucault, this could
be why.
Second, Foucault has fundamentally failed in one respect. He has founded
a school of thought even though it was his explicit ambition not to do so.
Today, there are researchers of Foucault in the same way that there are researchers
of Kierkegaard or Benjamin. His works have turned into philosophical
monuments. A vast number of books have been written about him: Foucault
the philosopher, the person, the historian, the political theorist, the aesthetic,
the ethicist, the sociologist, and, in this case, the analytical strategist. As with
Marx in the 1970s, it is now almost impossible to refer to Foucault without
being held accountable for various readings of Foucault. The worst thing to
do is to allude to Dreifuss and Rabinow; then, one is characterised as an
‘Anglo-Saxon’ (which is not a distinguished trait). In what follows, Foucault
will be read as a discourse analyst, which he was, although not exclusively.
Generally, Foucault can be defined as a formation analyst, who, for certain
periods, ascribed a principal position to discourse analysis, yet, at other times,
granted it a more secondary position in relation to, for example, institutional
analysis.
Third, Foucault’s particular way of creating concepts makes it difficult to
present them without depleting them in relation to his own definitions. All
concepts are created through distinctions. In the work of Laclau and Luhmann,
all concepts are bivalent, meaning that they possess two sides – this makes
them more easy to manage. Foucault’s concepts, on the other hand, are
polyvalent – they are many-sided and often come into being through countless
negative delimitations and very few, if any, positive definitions. In a presentation
such as this, it is very difficult not to reduce Foucault’s valencies.
1
Fourth, Foucault’s work is rather unsystematic. It is not possible, therefore,

to draw out a coherent discourse theory from his work. Again, this is related
to the fact that he did not want to create a school of thought. In the words of
Lars Henrik Schmidt: “His analytics is simply too consciously unsystematic
for it to develop into an actual theory; he himself does not even cohere to the
more programmatic proposals which nevertheless exist. It is primarily a
particular analytics, a practice which beckons meditation and imitation without
possible repetition” (Schmidt and Kristensen, 1985, p 5).
Rather than theory, the outcome of Foucault’s works consists of a number
of analytical strategies and analytical-strategic examinations of, for example,
periodisations, delimitation of discourses, monuments rather than documents
and demonstrations of rupture. This is where Foucault’s strength lies, but
these analytical strategies are always defined in relation to a specific research
question, a specific problem, and hence the question is whether they thereby
allow for simple generalisation. That, however, is precisely my goal.
Foucault as a structuralist
Foucault is often called structuralist or poststructuralist. That makes sense
given the fact that he developed his discourse analysis in an environment of
structuralists, especially his teacher, the Marxist and ideology theorist Louis
Althusser. However, he has, on several occasions, rejected both those labels.
In this book Foucault will be read as a phenomenologist, without consciousness
as the origin of meaning – a subjectless phenomenology.
Structuralists work from a distinction between the manifest surface and the
latent structure. Structuralist analysis reconstructs the hidden and latent structure
based on logical breaches on the manifest level. In other words, structuralists
work from the idea that, underneath the visible, directly accessible text, lays a
slightly displaced invisible text that controls the questions and answers posed
by the visible text. The invisible text amounts to a regime that condemns
specific questions. By directing attention toward those moments in the manifest
text when meaning breaks down – when answers are given without questions
and when questions are posed without answers – structuralism can raise
questions regarding the underlying text, which must exist so that the apparently
illogical and meaningless makes sense after all. In Althusser, it is called
symptomatic reading, that is, a reading of the logical breaches of the visible
text as symptoms of an underlying and controlling structure (Althusser and
Balibar, 1970). Poststructuralists maintain the notion of structure but see the
structure as much more loose and undecided. Foucault rejects both positions.
To Foucault there is only one level, which is that of appearance. Foucault
focuses on the statements as they emerge, as they come into being. It is crucial
to him never to reduce them to something else. Consequently, discourse is
not a structure to Foucault; a point to which we will return later in this
chapter.
However, if we were nevertheless to associate Foucault with structuralism
(which does in some respects seems fair, since his entire position is developed
2
The discourse analysis of Michel Foucault
within a structuralist environment), he could be called a transformation structuralist.

First, he would not define the notion of structure within a synchronous
perspective as, for example, in the works of de Saussure, Lévi-Strauss and
Althusser. Structure would then be transferred into a diachronic analytics,
precisely as a transformational structure, asserting itself at moments of change
and rupture. Hence, structures would not be an attribute of the world but of
the diachronic story. Consequently, transformational structures are not
structures that are uncovered, but rather the writer’s construction of the
historical relationship between discourse and institution.
Foucault as a discourse analyst

Foucault’s fundamental concern is the questioning of discursive assumptions.
He challenges individual will and reason by showing how every utterance is
an utterance within a specific discourse to which certain rules of acceptability
apply. Foucault challenges knowledge as a neutral speech position by showing
that humanities and the social sciences in particular are inseparable from
moralising projects; that humanities and the social sciences do not simply
elucidate the world but establish regimes of knowledge and truth that regulate
our approach to ourselves, each other and our surroundings respectively.
Foucault wants to show how any discourse involves excluding procedures,
which not only exclude themes, arguments and speech positions from the
discourse, but also produce outsiders, denounce groups of people as sick,
abnormal or irrational, and grant other groups the right and legitimacy to
treat these people (for example by imprisonment or therapy). Foucault wants
to show how power in society cannot be pinpointed, and thus separated and
isolated from, for example, the social sciences and public welfare institutions
such as schools and hospitals, and that power is ubiquitous as a productive
factor. Power is present in our approach to things insofar as the objects we
relate to are always discursive objects, produced by and in discourse. Power is
present in our approach to ourselves insofar as our self-relation is a product of
power. Power is present in our approach to others insofar as, for example,
‘criminals’, ‘mad people’ or ‘sick people’ are not in and of themselves criminal,
mad or sick. Conversely, criminality and illness are discursive positions, which
are established with the intent to control.
Madness and mental illness
In Madness and civilization, Foucault (1971) seeks to examine the emergence

of the empty negativity of modern reason – folly. His thesis is that madness
participates in producing reason by being the other side of reason. There can
be no reason without reason recognising itself as such. That requires reason to
be able to exclude and separate itself from folly. The genealogy of madness is
thus not a story of the conquests of psychology and psychiatry; it is not a
study of the language of reason but of the silence of folly and madness (Foucault,
1971). To a certain extent, it is a study of the conditions of the origins of
3
modern discoursivity, that is, of that which fundamentally defines discourse as

discourse within modernity. Guided by the discursive boundary between
reason and madness, Foucault examines the constituent relations between reason
and madness. He examines the excluding procedures and forms of reason,
and hence of society, including inclusive forms of exclusion.
Illness is one example of an inclusive form of exclusion. Illness is a position
in society: by defining ‘madness’ as illness, that which is excluded from society
can be kept under control within society.
The discursive history of the reason/madness boundary can be roughly
divided into four periods. In the Middle Ages, ‘mad people’ were looked on as
a kind of ideal, from whom one could learn something. Mad people were
believed to possess a more immediate relationship with God; they were closer
to God than other people. In the Christian search for salvation, mad people
therefore possessed something generally desirable. In the second period (the
17th-18th centuries), the perception of mad people changed from one of
consecration to one of ‘sickness’. Mad people were feared and locked up
together with the poor and the criminal, not in order to help and treat them
but in order to ‘dispose’ of them. Mad people were thought to set a bad
example in the face of the new work ethic, in which it was believed that
idleness was the root of all evil. Mad people represented idleness and lacked
the ability to actively partake in production. They were therefore a disgrace.
The third period (the 19th century) defined differences between madness,
criminality and poverty. Mad people were locked up in hospitals, where
attempts were made to ‘cure’ them (in this context, curing means reinstalling
in the mad person a sense of dependency, humility and gratitude towards
society). One of the methods used in this treatment was cold showers and
baths, which were meant to revive the spirits of the mad person; another was
to inject new blood into the veins of the patient in order to boost the blood
circulation. There were no distinctions made between the psychological and
the physical. Not until the fourth period (the 20th-21st centuries) does such
a distinction occur. Now, we attempt to understand mental illness and to
perceive ‘madness’ based on ‘mad people’s’ experiences and their distortion of
reality. We now aim to restore communication – not only between the mad
and the sane person, but also between the mad and sane aspects of the adapted
personality.
As a study of the transformational conditions of modern discoursivity, Madness
and civilization is, furthermore, a study of that which modern society recognises
as its history. In the words of Vincent Descombes:
Foucault intends to sound the limits of what we can recognize as our history.
At the interior of this history of ours, as of all history, identity presides;
within it, a single culture enables a number of human beings to articulate a
collective ‘we’. This identity – here is what must now be demonstrated – is
constituted through a series of exclusions. [...] Foucault goes further and
asserts that the history of madness is the history of the possibility of history.
‘History’, as we understand it, implies in effect the accomplishment of works
4
and the transmission of words endowed with meaning. Now madness,

according to Foucault, is defined by ‘the absence of works’. The mad person’s
gestures culminate in nothing, his delirious talk refers to nothing: his life is
fundamentally workless and inoperative. The possibility of history rests upon
the decision that all gestures and words which afford no positive significance
be rejected as unreason. Madness surrounds history on all sides: it is there
before history and is still there after history. (Descombes, 1980)
In Mental illness and psychology, Foucault (1976) examines the conditions within
psychology of speaking of illness. How is the notion of illness construed?
Where does psychology find its concepts of illness? And how do these concepts
of illness place psychology in a particular relationship with the patient? Foucault
defines four questions:
1. Under what conditions can one speak of illness in the psychological domain?
2. What relations can one define between the facts of mental pathology and
those of organic pathology?
3. How did our culture come to give mental illness the meaning of deviancy
and, to the patient, a status that excludes him?
4. And how, despite that fact, does our society express itself in those morbid
forms in which it refuses to recognise itself? (Foucault, 1976)
The first two questions address the way in which the question of illness operates
within the pathological discourse. Foucault explores the shaping of the psyche
through the creation of various forms of analysis in psychology, primarily
historical and phenomenological analysis. He observes the way psychology
and psychoanalysis observe their patients. He observes, for example, how
mental illnesses occur when psychoanalysis views its patients through the
concept of the individual story. He thus inquires about the discursive figures
that cause particular individual behaviour to emerge as illness. By answering
the first two questions, Foucault has presented the forms of appearance of
illness, but not yet the conditions of its appearance. He seeks to uncover the
latter through the last two questions, which address the construction of madness
outside psychology. Once again, the boundary between reason and folly
becomes central to the analysis. Foucault’s thesis is that those psychological
analyses that define the ill person as an outsider are projections of specific
cultural and discursive themes. Foucault shows how man has become “a
‘psychologizable species’ only when his relation to madness made a psychology
possible, that is to say, when his relation to madness was defined by the external
dimensions of exclusion and punishment and by the internal dimensions of
moral assignment and guilt” (Foucault, 1976, p 73). Foucault concludes:
Psychology can never tell the truth about madness because it is madness
that holds the truth of psychology. [...] There is a very good reason why
psychology can never master madness; it is because psychology became
5
possible in our world only when madness had already been mastered and
excluded from the drama. (Foucault, 1976, pp 74, 87)
History and humanity
In The order of things, Foucault (1974) shifts the focus onto his own scientific
discourse and its historical conditions. He sets out to examine the historical
conditions of structuralism1. He raises the question of why structuralism
becomes a dominant societal phenomenon, manifesting itself in a large number
of humanist sciences: in structuralist psychoanalysis with Lacan as the main
figure (for example, Lacan, 2001), in structuralist ethnology and anthropology
with Lévi-Strauss (for example, Lévi-Strauss, 1996) and in a structuralist literary
criticism with writers like Barthes and Kristeva (Kristeva, 1989; Barthes, 1990).
In all three cases, Foucault particularly subscribes to structuralist linguistics
with Fernando de Saussure as the primary point of reference (de Saussure,
1990). Foucault investigates how language replaces man as the object of
knowledge. Foucault simultaneously studies the origin and disappearance of
the humanities as a result of the repression of human beings by language. In
Classicism, ‘man’ is considered the centre of the world – man is defined as the
subject in an double extension of, on one hand, a transcendental being charged
with will and reason, and, on the other hand, an empirical being (as specific
people that can be observed). This defines the potential conditions of the
humanities as humans observing humans. However, through structuralism,
language becomes subject by a similar division into both a transcendent and
an empirical being. It is no longer humans but structures that speak – ‘it’
speaks as Lacan formulates it; the conditions of the humanities recede. Instead,
we are faced with language, speaking of itself through itself. Foucault says,
“where there is a sign, there man cannot be, and where one makes signs speak,
there man must fall silent” (Foucault, 1998a, p 266). In other words, discourse
analysis becomes central.
With The order of things, Foucault makes structuralism the object of discourse
analysis and seeks to distance himself from it by analysing structuralism and its
genealogy. He begins in structuralism and writes his way out of it through
discourse analysis. In Discipline and punish, Foucault (1977) asks:
Where does this strange practice and the peculiar project of locking someone
up in order to rediscipline them, which is implied by the criminal law of the
modern age, stem from? [...] Behind the insight into people and behind the
humanity of the punishments we rediscover a particular disciplinary
investment in the human bodies, a mixture of submission and objectification,
one and the same relation between ‘knowledge–power’. Is it possible to
define the genealogy of modern morality based on a political history of the
bodies? (Krause-Jensen, 1978, author’s translation)
6
Discipline and punish
Discipline and punish traces the history of prison and describes how the types
of surveillance and punishment change fundamentally from the beginning of
the 19th century. The change is outlined as a shift from the torment and
absolute monarchy of the Middle Ages, to the modern prisons, in which the
apprehension of punishment, rather than the appalling theatre of torment and
torture, is supposed to deter people from criminal acts. In the modern prison,
“the convicted body is no longer displayed, now it is hidden. It is no longer
destroyed, but instead detained and isolated because it carries something, which
can be tamed and trained” (Krause-Jensen, 1978, p 124, author’s translation).
This is not a simple movement from barbarism to humanism, but rather a
movement defined by the rise of disciplining as a new and effective form of
power. An expression of this displacement is Bentham’s famous Panopticon –
a prison in which prisoners do not know if or when they are being observed;
in which the observer cannot himself be observed, forcing the prisoners to
discipline themselves based on the notion of constant surveillance: “The
Panopticon is a machine for dissociating the see/being seen dyad: in the
peripheric ring, one is totally seen, without ever seeing; in the central tower,
one sees everything without ever being seen” (Foucault, 1977). The movement
is from public torment to hidden surveillance, where the soul, not the body, is
to be punished.
However, Discipline and punish is not only a book about prisons. It is a book
about how a particular power of normalisation – disciplining and surveillance
– is shaped in conjunction with the rise of the prison but is later propagated
and generalised in society. In Foucault’s understanding, the prison is not
simply an institution subordinate to the courts of justice. Conversely, prison
is a form that subjects itself to the courts and also prevails in a large number of
relationships in society at large. Foucault speaks of the relative carceral
continuity as the many connections that exist between, on the one hand, the
surveillance and disciplining forms and, on the other hand, institutions such
as education, social services and workplaces with organised hierarchy:
The judges of normality are present everywhere. We are in the society of

the teacher-judge, the doctor-judge, the educator-judge, the ‘social worker’-
judge; it is on them that the universal reign of the normative is based; and
each individual, wherever he may find himself [sic], subjects to it his body,
his gestures, his behaviour, his aptitudes, his achievements. The carceral
networks, in its compact or disseminated forms, with its systems of insertion,
distribution, surveillance, observation, has been the greatest support, in
modern society, of the normalizing power. (Foucault, 1977, p 304)
Foucault’s analytical strategy develops parallel with this shift in his questioning.
In this chapter, I make a distinction between four analytical strategies: the
archaeology of knowledge, genealogy, self-technology analysis and dispositive
analysis. In certain readings of Foucault, the analytical strategies successively
7
replace each other. In my reading, they are constructed on top of one another.
My outlines of Foucault’s analytical strategies are not complete, however; it is
possible to trace other strategies in his work, for example an aesthetic analysis.
Archaeological discourse analysis

The archaeology of knowledge (1986a) is Foucault’s first major attempt at describing
his efforts, and is a rationalisation and systematisation of his prior works rather
than a programme for his future work. Foucault never consistently adhered to
his own knowledge-archaeological analytical strategy, so there is no reason for
the rest of us to be literal in our reading of it. The archaeology of knowledge was
never intended to be a methodical description for systematised repetition and
imitation; however, it is useful as a catalogue of the analytical-strategic questions
that arise in attempts to invoke discourse without taking to structuralism or
other forms of reductionism.
The knowledge archaeology installs a distinction between statement,
discourse and discursive formation. These three concepts provide a joint
foundation for discourse analysis:
• statement is the atom of discourse – its smallest unit;

• discourse is the final, actually demarcated body of formulated statements – it
is the archive of the discourse analyst;
• discursive formation is a system of dispersion for statements; it is the regularity
in the dispersion of statements.
However, discourse is not a structure and does not exist on a level different
from statements. Statements do not manifest themselves as a discursive structure.
Discursive formation simply consists of the regularity of the irregular
distribution of statements. In other words, the fundamental guiding difference
in Foucault’s knowledge archaeology is regularity/dispersion of statements.
This is the basic difference in Foucault’s analyses of discourse:
Whenever one can describe, between a number of statements, such a system

of dispersion, whenever, between objects, types of statements, concepts, or
thematic choices, one can define a regularity, we will say, for the sake of
convenience, that we are dealing with a discursive formation. (Foucault,
1986a, p 38)
The difference entails an ontological asymmetry. Whereas, according to

Foucault, the statements actually exist, regularity is a construction that is created
through discourse analysis. Hence, it is the discourse analyst who constructs
the regularity of the dispersion, that is the discursive formation. My questions
are:
8
• How is this conducted?

• When is a statement a statement?
• When is regularity a regularity that can be defined as discursive formation?
These are the fundamental analytical-strategic problems. What the discourse

analyst is concerned with – that which is, according to Foucault, the object of
discourse analysis – is statements (énoncé). As the building blocks of discourse,
the way a statement is demarcated is of the utmost significance for the formation
of discourse analysis. Like Foucault, I will first outline what statements are
not, that is, what discourse analysis does not analyse.
Discourse analysis is not textual analysis
According to Foucault, a discourse does not consist of texts and discourse

analysis is therefore not textual analysis. Texts, as such, are much too boundless
for them to function as the basis for discourse analysis. Although books serve
a particular purpose and possess certain economic value, they are not
independent discursive units:
The outlines of a book are never clearly and stringently defined: no book
can exist by its own powers; it always exists due to its conditioning and
conditional relations to other books; it is a point in a network; it carries a
system of references – explicitly or not – to other books, other texts, or
other sentences; and the structure of reference, and thereby the entire complex
system of autonomy and heteronomy, depends on whether we are dealing
with a dissertation on physics, a collection of political speeches, or a science
fiction novel. It is true that the book presents itself as a tangible object; it
clings to the tiny parallelepiped surrounding it: but its unity is variable and
relative, does not let itself be constructed or stated and therefore cannot be
described outside a discursive field. (Foucault, 1970, p 152, author’s
translation)
Discourse analysis is not literary analysis
Like texts, the works of authors are not bound wholes: “Apparently it indicates
the sum of texts that can be denoted by a proper name” (Foucault, 1970,
p 153). However, says Foucault, how about texts that are signed under a
pseudonym? How about hasty notes or sketches that are discovered
posthumously? What, for example, should be considered as belonging to the
works of Nietzsche? Only the books? Or also letters and postcards and texts
signed Kaiser Nietzsche? The fundamental problem, however, consists in the
general tendency to perceive works, as heterogeneous as they might be, as the
embodiment of one writer’s thoughts, experiences or unconscious. In other
words, to view texts as the indication of a whole, which is not visible in the
textual fragments but must be ascribed through interpretation. Consequently,
9
Foucault delivers a critique not only of actual literary analysis but of any
textual analysis, which professes to refer the statements of the text back to the
author and his intentions, concerns, unconscious, circumstances and so on.
Discourse analysis is not structural analysis

Statements do not express an unspoken structure that secretly animates every
statement. The discourse at hand is, in the end, not simply the disturbing
presence of the unspoken (Foucault, 1970, pp 154-5).
Discourse analysis is not a form of discourse commentary

To comment on discourses means to inquire about the statement, meaning
and intention of discourses. Discourse commentary attempts to expose the
underlying meaning of the stated, based on the assumption that the deeper
meaning comes closer to an essential truth. According to Foucault, however,
commenting means an admittance of access to the signified over the signifier.
His ensuing question is whether it is possible, conversely, to construct a discourse
analysis capable of escaping the fate of commentary by presupposing that the
stated only exists in its historical rise and emergence (Foucault, 1986b, p 17).
Subsequently discourse analysis becomes an analysis of statements in their

positivity:
Every discursive moment (that is, every statement) must be observed in its
positive suddenness, in this punctuality which it enters and in this temporal
dispersion which makes it possible to repeat it, realise, forget, transform,
efface it until the last traces, hide it, far from all eyes, in the dust of books.
(Foucault, 1970, p 155, author’s translation)
Foucault calls discourse analysis “pure description of discursive facts” (Foucault,

1972, p 234). He even maintains that the spirit of discourse analysis is a
felicitous positivism (p 234).
Analysis of statements
In short, a statement needs to be analysed in its appearance, as it emerges, and
cannot be reduced to expressing anything other than itself, such as, for example,
the intention of the statement, the context, the concern or the meaning of the
statement. Foucault rejects any reductive or interpretative statement
descriptions. He consistently avoids questions of what or why in relation to
statements but only asks how. The question ‘What is the meaning of the
statement?’ instantly ontologises the statement – it is reduced to a given,
containing a secret. The question ‘Why this statement?’ reduces the statement
to its cause. Only questions of how the statement appears grant full attention
to the statement, only how does not immediately shift the attention away from
10
the statement itself to that which could possibly provide the statement with a
meaning or an explanation.
• When, then, is a statement a statement?

• How do we recognise a statement?
To Foucault, a statement is a function of existence that enables groups of signs to

exist (Foucault, 1986a, pp 86-8). What does that mean? It means that the
statement is the smallest unit, which brings forth phenomenon through enunciation.
We are thus able to recognise the statement by its momentary creation rather
than by its appearance as a sign, sentence, book or argument. Statements are
positive events that produce existence through enunciation. This function of
existence contains at least four aspects: object, subject, conceptual network
and strategy.
Discursive objects
A statement is only a statement if it creates objects. Those objects referred to

by a statement are not ‘objects in themselves’ or ‘objects out there’, they are
discursive objects constructed, classified and identified by the statement itself.
The statement creates the object to which it refers through enunciation. The
enunciation of the object implies that it is brought to life as a social and
discursive fact and can therefore be articulated (Foucault, 1986a, pp 88-92).
Subjects
A statement can only be regarded as a statement if it creates subject positions

that can be signed over to individuals, that is, if the statement creates discursive
spaces from which something can be stated. The question is:
• In which positions can, and must, the statement be held by an individual,

for the individual to become the subject of the statement?
Again, subjects do not stand outside of the statement; conversely, the statement
articulates the space and possibility of subjects.
Conceptual network
A statement can only be regarded as a statement if it situates its elements of

signification in a space in which they can breed and multiply. Any statement
subscribes to certain concepts, and yet it is not simply a variation of conceptual
combinations. The status of the concept is determined by the statement,
through the statement’s endowment of an associated field that consists of all
the formulations that the statement implicitly or explicitly refers to, either by
repeating them, modifying them, adapting them, opposing them or commenting
on them. According to Foucault, all statements re-actualise other statements
11
in some way. Linguistic elements such as signs and sentences are only statements
if they are immersed in an associated field, in which they simultaneously
appear as unique elements. Foucault summarises it thus:
There is no statement in general, no free, neutral, independent statement;

but a statement always belongs to a series or a whole, always plays a role
among other statements, deriving support from and distinguishing itself
from them: it is always part of a network of statements, in which it has a role,
however minimal it may be, to play. (Foucault, 1986a, p 99)
This should not only be understood retrospectively as the fact that the statement
becomes a statement by relating to prior statements, but also understood
progressively as the way in which the statement paves the way for potential
future statements. Hence, a statement becomes a statement only if it both re-
actualises and extends other statements. This enunciative function, by the way,
has a striking resemblance to Luhmann’s notion of meaning, which is discussed
in Chapter Four.
According to Luhmann, meaning is merely the unity of actualisation and
potentialisation. Conversely, the enunciative function differs significantly from
Laclau’s more poststructuralist theory of signification, which is concerned
with structural relations between elements. I propose that perhaps Foucault’s
discourse analysis and Luhmann’s systems theory are closer to one another
than Foucault’s discourse analysis and Laclau’s discourse theory because of
their phenomenological kinship.
Strategy
A statement can be defined as a statement only if it is integrated into operations

or strategies in which the identity of the statement is maintained or effaced
(Foucault, 1986a, pp 100-5). A statement is not simply that which is stated
independently of time, place and materiality. A statement always chooses a
materiality, at least in the form of a medium for its creation, for example
speech, writing, report, arrangement or image. It always seeks support in a
context; it appears with a status derived from the strategic context of its origin.
It is always strategic in the sense that the statement as re-actualisation emerges
as one choice among other possible actualisations. Consequently, it cannot
easily be displaced in time and space.
For example...
Regardless of the attempted analogies, the readaptation for the screen of the
books of Morten Korck is not an identical transcription of the ‘original’ statements.
The statements are repeated, but their materiality and strategic standing is different.
The fact that Lars von Trier is responsible for the screen version is an enunciative
event, which exceeds the film itself.
12
Therefore, discourse analysis consists of an analysis of statements, in which

statements exist as an event, constantly enunciating subject positions, discursive
objects, conceptual relations and strategies. The world comes into being, so to
speak, through the statement as event.
Constructing the archive

Statements are always statements in a discourse. The field of discourse analysis
is “the compilation of all actual statements (spoken or written) in their historical
dispersion and in their specific momentary value” (Foucault, 1970, p 155,
author’s translation). Discourse is “the always final and actually delimited
body of precisely those linguistic sequences that have been formulated”
(Foucault, 1970, p 156, author’s translation). However, the discourse as this
body of statements is not self-evident. It is the first task of the discourse
analyst carefully to outline this body, in order to construct the archive as that
which ultimately regulates what has been said and not been said in a given
society. Naturally, it is impossible to decide in advance which discursive
formation regulates the dispersion of particular statements. As a discourse
analyst, it is necessary to travel the long and cautious road via the archive in
order to approach the question of the shaping of specific discursive formations:
“Only through a description of the archive of a discourse does it become
possible to see the discursive formation as well as its transformation” (Ifversen,
1997, p 486). In Foucault’s words: “One ought to read everything, study
everything. In other words, one must have at one’s disposal the general archive
of a period at a given moment. And archaeology is, in a strict sense, the
science of this archive” (Foucault, 1998a, p 263). Reading “everything” means
a number of things to Foucault:
• First, it means that, since it is not possible to define the discursive formation
beforehand, one cannot limit one’s reading to a theme such as, for example,
madness. Themes can relate to each other in unpredictable ways, which,
moreover, can change over time and between spaces. We therefore have to
follow the references of the statement and the references of the references
in time and space in the broadest sense, until they appear to form a completed
whole. There is no shortcut without consequences, as with the fate of the
discourse commentator.
• Second, Foucault maintains that it does not suffice to read the canonical
works pointed out by the history of ideas. It is crucial that the reading also
includes the statements of the institutions, statements that demonstrate
practice. Unravelling the history of madness includes readings of
philosophical works as well as scientific dissertations and the statements,
regulations and accounts of the institutions themselves.
• Third, and finally, we must be careful not to install a preconceived distinction
between official and more private and individual sources, as if the private
and personal sources exist outside the discourse.
13
For example...
In the studies by Foucault of the self-relation of the self, diaries and personal
notes from Rome in the 1st and 2nd centuries play an important role in the
analysis. Likewise, novels, paintings and personal documents appear as frequently
used material. And, understandably, we cannot weigh up the types of statement
against each other in advance, as if a text by Voltaire is of greater importance
than documents written by an historically anonymous person, which, in turn, are
more important than the diaries of a doctor, and so on.
Not until the archive has been established is it possible to inquire about the
discursive formations. Not until the entire body of statements has been pieced
together can we start to ask questions about the way one or more regularities
appear in the irregular dispersion of statements, in other words, how the
dispersion of statements over time seems to be regulated by different discursive
formations. For Foucault it is about “seizing the statement in its momentary
conciseness and total singularity, determining its exact boundaries, establishing
the connections to other statements to which it can be linked, and showing
which other categories of statements are thus excluded” (Foucault, 1970,
p 156, author’s translation). The fundamental question to the statements posed
by discourse analysis in the attempt to create discursive formations is: ‘Why
did this and no other statement happen here?’ (Foucault, 1970, p 156, author’s
translation). The aim is to detect the rules that govern the way different
statements come into being in discursive formations. Rules in this context
mean rules of acceptability, that is, rules about when a statement is accepted as
a reasonable statement.
The formation of statements

Foucault distinguishes between four levels in a discursive formation, that is,
four bodies of rules for the formation of statements (Foucault, 1986a, pp
21-71).
The formation of objects
The question here is why statements shape their objects the way they do.
• What is the regularity of the dispersed formation of objects by the

statements?
• According to which rules are the objects created, ordered and classified?
• Which relations (for example, cause–effect relations) are established between
the discursive objects?
• Which hierarchy of objects does the individual object form a part of?
• How are the objects specified and characterised?
14
These types of questions serve the purpose of isolating specific discursive

formations and their rules for the formation of objects.
The formation of subjects
A discursive formation is a regulation of the dispersion of subject positions as

they come into being in an individual statement. Here, the fundamental
question is why the statements create subject positions the way they do.
Discursive objects are always enunciated from a particular place – objects are
always objects of a subject. Consequently, the question of subject positions is
a question about the place from which the objects are enunciated.
• From which subject positions do the objects appear the way they do in the
discourse?
This pertains to the rules of acceptability for the shaping of the spaces from
which one can speak and observe in the discursive formation, but it also
pertains to the existing rules for the acceptance of entering certain individuals
into the spaces that are being created, and when this can happen. That is:
• Which qualities relate to subject positions?

• In which situation can the subject position be used as a platform for speaking
and observing?
• What are the rules for observation and for the formation of statements
when one assumes a specific position?
The formation of concepts
The enunciation of the subject positions and the discursive objects furthermore
implies the connection to concepts. The questions are:
• Why does the statement actualise particular concepts and not others?
• How do concepts organise and connect statements?
• What are the rules for conceptualisation and how do specific discursive
formations draw on concepts from other formations, including the rules of
transcription, which seem to exist between different discursive formations?
The formation of strategies
A specific discursive formation is thus provisionally characterised by a particular

creation of objects, subject positions and concepts. However, even in regard
to the different levels of rules for the definition of formations, other objects
could be enunciated, other subject positions could be established, other concepts
could be actualised and linked together. Strategy is about rules of selection
for the completion or actualisation of the rules of acceptability. Jan Ifversen
attempts a formulation of these questions thus: “The potential possibilities of
15
the discursive formation is narrowed down by virtue of ‘strategic decisions’.

In order to illustrate the appearance of a discursive formation, one needs to
identify the limitations expressed by the strategic decision” (Ifversen, 1997,
p 483). This question is by definition interdiscursive – it addresses how
individual discursive formations always come into being in relation to other
discursive formations.
This process involves a battle and competition, but also a mutual constitution
between different discursive formations – a constellation that defines the
conditions of the distribution of statements by the individual discursive
formation. The inter-discursive relations need to be defined as they appear.
They are defined through the individual discursive formations and their
statements. Again, it is important to avoid any form of reductionism such as,
for example, referring the inter-discursive competition to structures or
conditions that do not appear explicitly in the competition, for example,
economy, power, international regimes of production. The formulation of the
rules of formation for strategies concerns the formulation of the unity of the
mutual exclusion of the discursive formations. Foucault formulates it thus:
One may descr ibe between several discourses, relations of mutual

delimitation, each giving the other distinctive marks of its singularity to be
the differentiation of its domain of application. [...] This whole group of
relations forms a principle of determination that permits or excludes, within
a given discourse, a certain number of statements. These are conceptual
systematizations, enunciative series, groups and organizations of objects that
might have been possible (and of which nothing can justify the absence at
the level of their own rules of formation), but which are excluded by a
discursive constellation at a higher level and in a broader space. A discursive
formation does not occupy therefore all the possible volume that is opened
up to it right by the systems of formation of its object, its enunciations and
its concepts; it is essentially incomplete, owing to the system of formation
of its strategic choices. (Foucault, 1986a, p 67)
Conclusion
Accordingly, the regulating difference in the analytical strategy of the knowledge

archaeology is regularity/dispersion of statements. The archaeological eye
divides the world into dispersed statements and the regularity of the dispersion.
In fact, that is all there is. The remaining concepts serve the sole purpose of
defining when a statement can be regarded as a statement and when regularity
can be regarded as regular. Discursive formations are analysed as regularities
of dispersion of statements in which statements are seen as events possessing
certain functions of existence, which produce objects, subjects, conceptual
networks and strategies.
16
Genealogy
Genealogical analysis cannot be separated from archaeological analysis. The
relation between archaeology and genealogy is frequently portrayed, in Foucault,
as two periods in his writings in which the genealogical breakthrough is
summoned with the article ‘Nietzsche – genealogy and history’ (Foucault,
1991). However, before this publication Foucault had already used the term
‘genealogy’, at times together with the word ‘archaeology’. Moreover, the
majority of the questions addressed in the Neitzsche article had already been
discussed before in The archaeology of knowledge (Foucault, 1986a), in the chapter
‘Discourse and discontinuity’, and in different introductory problem
formulations in his historical works. In this section, I will construe the
genealogical analytical strategy as the historical dimension of the knowledge
archaeology. Consequently it is impossible, in my understanding, to conduct
a knowledge-archaeological analysis without combining it with a genealogical
analysis. In my exposition, the following is chosen as my starting point.
Framework for genealogical analysis
Whereas the framework for the eye of knowledge archaeology is the difference
regularity/dispersion of statements, the framework for the eye of genealogy is
the difference continuity/discontinuity.
The designation of genealogy as a particular historical analytical strategy was

initially developed by Friedrich Nietzsche, primarily in his dissertation On the
genealogy of morality (Nietzsche, 1998). In this text, Nietzsche explains how he
began by looking into the origins of good and evil but how he gradually realised
that this way of asking was tremendously problematic because of the question’s
implied supposition about an original essential moral that would appear increasingly
clear as one approached its source on tracing it back through history.
Nietzsche redefines his question and thus ascribes to it the characteristics of

genealogy. He asks:“Under what conditions did man invent those value judgements
good and evil? And what value do they themselves have?” (Nietzsche, 1998, pp 2-
3). The genealogy of morals subsequently revolves around “the conditions and
circumstances out of which [these values] have grown, under which they have
developed and shifted” (p 5). The aim becomes to question the value of values
in all facets. The genealogy of morals is also therefore a critique of morality.
The genealogical analytical strategy is defined in opposition to traditional

historiography. In the genealogical analytical strategy, Foucault assumes
Nietzsche’s critique of history in which Nietzsche distinguishes between three
methods of historiography.
17
The monumental method
This historiography cultivates the connections and continuity of greatness of

all times. Monumental history harmonises heterogeneity, generalises and,
conclusively, analogises. It is all but critical: it deceives and seduces through
analogies and leads to rashness (Nietzsche, 1988). National history is a classic
example of this. Other, more current, examples could be various histories
about our path toward globalisation, or books about the formation of the idea
of Europe, which is now unfolding through the European Union. But it
might also be books about the historical progress of medical science or the
dissemination of modern humanism and its repression of the barbaric means
of punishment of earlier times.
The antiquarian method
Antiquarian historiography cultivates the past for the sake of the past. It
becomes a blinded collection mania and a restless amassing of things of the
past. Such a historiography mummifies life; it is incapable of breeding life and
always underestimates the future (Nietzsche, 1988).
The critical method
This historiography stands in the service of life. Critical historiography has,

as its starting point, the notion that the past must be broken up and annulled
in order to allow the living to exist. The past must be held accountable,
submitted to awkward interrogations and convicted – not in order to achieve
justice or mercy, but in order to allow for the emergence of life and force
(Nietzsche, 1988).
The fundamental distinction in Nietzsche’s critique of history is whether
the historiography stands in the service of life or death. The antiquarian
operates in the service of death, exclusively through his mummified
historiography. Monumental historiography is more ambiguous: it provides
people with a future prospect that enables them to act and perform, but it
leads to rashness because it presents the future in the singular, as necessary and
inevitable. Critical historiography alone is in the service of life, by providing
an opening in which the past locks us up in presuppositions and morals.
Knowledge about the past is not in itself beneficial; on the contrary, says
Nietzsche, we need to forget in order to live. Robert Scharff sums it up thus:
What Nietzsche’s first generation opponent of history chooses to do is

reject the conception of past, future, and present that inform historical science:
at some decisive point, he must forget rather than remember more and
more of the past. He must turn away from rather than survey the whole
process, so that the future looks open instead of finished in anticipation.
And the less he presently knows of the whole, the younger he feels. (Scharff,
1974, p 74)
18
Regularity and discontinuity
According to Allan Megill, Foucault maintains this distinction between the

historiography of life and death. Megill refers to the tension between Apollo
(the god of knowledge, the arts, order and civilisation) and Dionysus (the god
of wine, mysteries, darkness and death) as a fundamental tension in Foucault’s
analytical-strategic development (Megill, 1979, p 459). In Foucault’s work,
this tension appears as the tension between archaeology as the systematic
analytical strategy (Apollonian) concerned with the regularity of the irregular,
and genealogy as an analytical strategy, concerned with discontinuity, which
brings on life and undermines presuppositions (Dionysian). Foucault hence
tries to extend the Dionysian thinking in his genealogies. According to
Foucault, genealogy comprises three critical forms of application that serve
life:
1. Reality-destructive use, which opposes the historical motif recollection-

recognition. Foucault’s historiography is reality-destructive when, as in
Madness and civilization, it challenges the way the present recognises itself in
its historical texts.
2. Identity-destructive use, which opposes the historical motif continuity-tradition.
Foucault’s historiography is identity-destructive, when, as in Discipline and
punish, it denounces the humanist identity by showing that the history of
punishment does not consist in one unbroken movement towards a
humanisation of the penalty system since the Middle Ages, but, conversely,
that modern means of punishment indicate an intensification of punishment
because the aim is no longer simply to torment the body but also to control
the psyche. Foucault thus employs history in order to undermine the
humanist complacency of the present. He wants the professional
practitioners of humanism in the institutions to feel uncomfortable. Foucault
says of the identity-destructive application of genealogy: “The purpose of
history, guided by genealogy, is not to discover the roots of our identity, but
to commit itself to its dissipation” (Foucault, 1991, p 95).
3. Truth-destructive use, which opposes the historical motif of knowledge. This
applies to Foucault’s critique of psychology and psychoanalysis when he
demonstrates, for example, how the truth about insanity can never be found
in psychology, for the simple reason that the truth about insanity is located
in the very society that is responsible for creating the conditions of the
origin of insanity and madness. Foucault says, in regard to the truth-
destructive use of genealogy, “all knowledge rests upon injustice” (Foucault,
1991). He concludes that the goal is to create in history a counter-memory
(Foucault, 1991, p 93).
The purpose of genealogy, for Foucault, is not therefore a description of actual

events. Genealogy is a history of the present designed to outline the historical
conflicts and strategies of control by which knowledge and discourses are
constituted and operate, and to use these descriptions as a counter-memory.
19
Foucault is concerned with the redescription of “not the anticipatory power

of meaning, but the hazardous play of dominations” (Foucault, 1991, p 83).
Genealogy does not pretend to go back in time to restore an unbroken

continuity that operates beyond the dispersion of broken things; its duty is
not to demonstrate that the past actively exists in the present, that it continues
secretly to animate the present, having imposed a predetermined form on
all its vicissitudes. (Foucault, 1991, p 81)
Rather, the aim is to query the discourses and practices of the present by
referring them back to the hegemonic conditions under which they have
been established, which also includes pointing out ruptures in the grounds on
which strategies, institutions and practices are shaped. The presuppositions of
the present are to be dissolved by means of history2. Mats Beronius sums up
the main task of genealogy:
Instead of analysing an event, a phenomenon, or the history of an idea by

looking for the governing idea and leitmotif which should link together the
source and result of history, the analyst should attempt to draw the different
and multifaceted branches of a genealogical tree. [...] To trace the origin of
a social phenomenon (an institution, a practice, an idea etc) does not imply
the establishing of the birth of the phenomenon, but rather the tracing of
its line of descent. (Beronius, 1991, pp 50-2, author’s translation)
Foucault himself formulates it thus:
The isolation of different points of emergence does not conform to the

successive configurations of an identical meaning; rather, they result from
substitutions, displacements, disguised conquests, and systematic reversals. If
interpretation were the slow exposure of the meaning hidden in an origin,
then only metaphysics could interpret the development of humanity. But if
interpretation is the violent or surreptitious appropriation of a system of
rules, which in itself has no essential meaning, in order to impose a direction,
to bend it to a new will, to force its participation in a different game, and to
subject it to secondary rules, then the development of humanity is a series
of interpretations. The role of genealogy is to record its history: the history
of morals, ideals, and metaphysical concepts, the history of the concept of
liberty or of the ascetic life; as they stand for the emergence of different
interpretations, they must be made to appear as events on the stage of
historical process. (Foucault, 1991, p 86)
Here, the continuity and discontinuity of difference is a tool for observation,

employed in order to distinguish discontinuity in that which presents itself as
continuity and to examine possible continuities in that which presents itself as
new, different or unique. The genealogical method is concerned with the
continued openness of the object. Instead of a preliminary definition of the
20
object of our examination, we need to investigate how our object has been
construed historically in different ways and in different settings. We are not
only to look for those events that stand out clearly as seen from the present,
but also for those constructions, strategies and practices that, for some reason,
never distinguished themselves, disintegrated or changed into something else.
Hence, we should not just trace that which became history, but also very
much that which has been defined as mistakes, antiquated, unrealistic and so
on. In this respect, the approach needs to be wide rather than deep. The way
of seeing that the difference continuity/discontinuity constitutes is a glance of
dissociation,
a glance that distinguishes, separates, and disperses; that is capable of liberating

divergence and marginal elements – the kind of dissociating view that is
capable of decomposing itself, capable of shattering the unity of man’s being
through which it was thought that he could extend his sovereignty to the
events of his past. (Foucault, 1991, p 87)
Figure 1.1 attempts to illustrate the genealogical analytical strategy.

Here, the objective is to trace the lines of descent of psychoanalysis. Each
circle expresses a discursive strategy or an institutional practice. Each letter
expresses discursive or institutional elements, which are part of a strategic or
institutional order. The lines illustrate the relationships between discursive
formations and practices. A relationship requires at least one element to be
carried on in a subsequent formation or practice. As discursive formation and
practice, psychoanalysis has thus come into being through a transformation of
elements from a multitude of formations and practices, including the Catholic
practice of confession, the hospital and the internment institution. Genealogy
can be used to illustrate, for one thing, how psychoanalysis both repeats and
Figure 1.1: The genealogy of psychoanalysis
Confession
a, b, i
Treatment Psychology of the 18th century
ø, å
c, d, k
Mental institution
a, c, æ,
e, w,
c, æ, o, å
e, z Psychoanalysis
æ, q
The hospital
Internment
e, w, x
Source: This figure is a slightly altered version of a figure from Noujain, 1987.
21
renews a control issue pertaining to the handling of deviants, but which has
obtained a specific definition and elaboration within the individual formations
and practices.
For example...
One example of these specific definitions and elaborations of the control issue
is the Poor Law internment institution, in which poor people, criminals and
people with mental illness were locked up together and excluded from ordinary
life and from the rest of society, to hide these examples of unwanted, non-
industrious laziness. Later examples are the mental institutions, in which mad
people were distinguished as a particular group and notions of societal dependency
were impressed on them in their treatment, as discussed above. Finally, there is
psychoanalysis, which introduces the distinction between the body and the
consciousness of the ‘madman’ and aims to restore his communication with the
outside world (Foucault, 1971).
Genealogy provides no explanations of causality. The lines of descent are not

causal, for example ‘at first, there were hospitals, which were further developed
in the mental institution, which led to the development of psychoanalysis’. The
lines of descent simply imply an originating affiliation, that is, that psychoanalysis
ties together and transforms elements from previous discursive strategies and
practices in its own process of construction. Moreover, there is no simple
seriality in which discursive strategies successively supersede each other.
Conversely, a new strategy can easily emanate through the transformation of
elements from a specific discursive formation while this strategy is sustained. It
is even probable that the ‘first’ discursive strategy survives the ‘offspring’.
The endpoint of genealogy determines which discourses and discursive kinships

one discovers. A minor displacement of the endpoint is therefore able to alter
history quite significantly. The focus could, for example, be on hospitals rather
than psychoanalysis and this could identify new and central moments that might
also involve significant relationships to psychoanalysis, which would otherwise
have been overlooked, for example the medical way of seeing.
This does not mean, however, that discursive histories are fictional in the sense
that it is entirely open which stories can be told. Rather, it means that no
history can be described without being rooted in problem and perspective.
History is a constructed reality of a perspective and is real and observable as
such. This question is unfolded frequently in Foucault’s work, for example in
relation to periodisation:
Every periodization carves out in history a certain level of events, and,

conversely, each layer of events calls for its own periodization. This is a set
of delicate problems, since, depending on the level that one selects, one will
22
have to delimit different periodizations, and depending on the periodization

one provides, one will reach different levels. In this way one arrives at the
complex methodology of discontinuity. (Foucault, 1998a, p 284)
This, of course, has implications for the reading and analysis of the different
monuments of history, such as texts, images, legal reports and so on. The
displacement of one’s perspective and problem changes not only what one
sees, but also the way of seeing employed when reading and analysing. It
changes the statements to which one’s attention is drawn as well as the
connections one sees between statements:
The texts that I spoke of could easily be taken up again, along with the
very material that I treated, in a description that would have a different
periodization and would be situated at a different level. For example, when
the archaeology of historical knowledge is done, obviously it will be necessary
to again use the texts on language, and it will be necessary to relate them to
the techniques of exegesis, of the criticism of sources, and to all the knowledge
concerning sacred scripture and the historical tradition. Their description
will be different then. But if they are exact, these descriptions should be
such that one can define the transformations that make it possible to go
from one to the other. (Foucault, 1998a, p 284)
Again, as Foucault points out, this does not lead to relativism but rather perspectivism.
Given a specific perspective and problem, genealogy becomes sensitive to its
material in a specific way. However, genealogical descriptions can be criticised
for their perspective. Naturally, it is always possible to take a step back and
investigate the relevance and probability of a particular perspective and problem,
thus questioning all criteria of selection and the validity of the examination.
In the same way that archaeology is concerned with statements as they
appear, avoiding any form of discourse commentary, a similar phenomenological
antireductionism is stressed in genealogy. Foucault maintains that genealogy
must be unobtrusive and:
record the singularity of events outside of any monotonous finality; it must

seek them in the most unpromising places, in what we tend to feel is without
history – in sentiments, love, conscience, instincts; it must be sensitive to
their recurrence, not in order to trace the gradual curve of their evolution,
but to isolate the different scenes where they engaged in different roles.
Finally, genealogy must define even those instances when they are absent,
the moment when they remained unrealised (Plato, at Syracuse, did not
become Mohammed). (Foucault, 1991, p 76)
He emphasises that genealogy is “gray, meticulous, and patiently documentary.

It operates on a field of entangled and confused parchments, on documents
that have been scratched over and recopied many times” (Foucault, 1991,
p 76).
23
Self-technology analysis
Knowledge archaeology and genealogy are among the most clearly described
analytical strategies in the writings of Foucault, but there are a number of
others. Self-technology analysis concerns the analysis of the technologies
available to an individual’s manifestation of itself as subject. Whereas knowledge-
archaeological discourse analysis allows for studies of the way in which subject
positions are created, the analysis of self-technologies permits studies of the
practical staging of the relationship between individual and subject position.
At an early stage in his writings, Foucault spoke of practice as something
that does not unambiguously coincide with discourse. His understandings of
practice, however, remain unclear, although he seems to border on directions
for practice. In relation to the history of sexuality, the concept of practice
appears to be replaced by the concept of technology, although this replacement
is not unambiguous. What is unambiguous, in turn, is the question posed by
means of the concept of technology – it pertains to the self-relation of the
subject to its self-care.
In his sociology, Simmel defines a distinction between position and vocation
as two different ways in which a person can become an individual. Similarly,
Foucault distinguishes between subjection and subjectivation (for example,
Foucault, 1997, 1998b; and Balibar, 1994). To Foucault, subjection means that
an individual or collective is proclaimed subject within a specific discourse.
The individual or collective is offered a specific position in the discourse from
which one can speak and act meaningfully in a specific way. Foucault speaks
of subjectivation when the individual or collective has not only been made the
subject but also wishes to be so. Subjection, thus, signifies the space where one
receives oneself, whereas subjectivation signifies the space where one gives oneself
to oneself (Schmidt, 1990).
This distinction is not merely theoretical. It is also a distinction between
two forms of discursive demands on the person who is to become the subject:
a distinction between two modes of subjection. The appropriation by the
individual of a subject position is not observable in terms of discourse analysis.
It is, however, observable whether and how the discourse demands active self-
appropriation of a subject position. This is exactly what Foucault discovered
in the history of sexuality – the fact that the individual is not only required to
fill out a particular subject position but also to care for her/himself
independently. The interesting point is that the discourse itself, through the
subjecting of the individual, makes a distinction between subjecting and
subjectivation in which the former forms a counter-concept in relation to the
latter. As it is, demands are made on the individual not to simply receive
oneself passively but actively to give oneself to oneself. In that sense, the
aforementioned mode of subjecting is a mode of transformation – it invokes the
passively receiving and subjected so that s/he may cross the line from subjection
to subjectivation, thereby making her/himself actively sovereign in her/his
own self-creation. It is an invocation to the individual to invoke her/himself.
24
The question is then how this self-transformation comes about. Foucault

opens up the question in his study of self-care by introducing technology as
mediation. Foucault distinguishes between four types of technology: production
technologies, sign technologies, power technologies and self-technologies. He
sees the latter as technologies that allow for individuals to influence operations
that concern their own body, soul, thoughts, control and mode of existence,
so that they are able to transform themselves and achieve a specific state of
happiness, purity, wisdom, perfection or immortality (Foucault, 1997). In
other words, self-technologies are procedures that prescribe how the individual
is to define, maintain and develop her/his identity with a view to self-control
and self-awareness (Foucault, 1997). The purpose of these technologies is for
the self to address itself.
Foucault gracefully skips over how to perceive of technology generally. He
defines four criteria in relation to self-technology:
1. The transformational mode of subjecting.

2. The objectification of the self.
3. Self-activating activity.
4. Telos.
These criteria also define the basic question through which we will examine
the technologies (see also, Davidson, 1986; Hacking, 1986).
Figure 1.2: The elements of self-technology

1. The transformational mode of subjecting
Self-technology 2. The objectification of self
3. Self-activating activity
4. Telos
1. The transformational mode of subjecting
The first criterion of technology is the fact that it is a transformational mode

of subjecting, that is, that the individual is called on to give itself to itself. The
question is:
• How is the individual subjected with a view to crossing the boundary

between subjecting/subjectivation?
25
2. The objectification of the self
The second criterion is the objectification of the self, which concerns the
form of knowledge that the individual can establish about itself as self. The
question is:
• As what is the self to master itself? Is the self to master itself as feeling, as
desire, as will, or as culture?
3. Self-activating activity
The third criterion is the stipulations for self-activating activities, for example,
the diary as self-activity that emerged in the 2nd century in the Roman Empire
(Foucault, 1997), or the confessional practice of the Catholic Church. Examples
of modern technologies could be the ‘time manager’ of the 1980s or competence
interviews in the workplace.
4. Telos
Finally, the fourth criterion is the fact that the self-activity needs to have a
direction or an aim that extends beyond the mere activity. The question is:
• In what way does the self-technology provide the individual with a particular
telos for life?
Conclusion
Self-technology can therefore be understood as prescriptions for operations

through which the individual, having received itself through subjecting, is
able to reach a point of transformation so that it can give itself to itself in
order to obtain a particular personal goal or condition. In that way, self-
technologies are defined as technologies through which the individual can
transform itself from a state of having responsibilities into taking on
responsibilities, which means that the individual puts her/his own development
on the agenda and accepts responsibility for it.
This understanding of self-technology analysis thus divides the world into
subjecting and subjectivation, and constructs a sensitivity to the practices
through which the self can summon itself and activate itself in order to master
its own creation. The concept of ‘subjecting’, which was a subconcept in
archaeological discourse analysis, increases in importance and becomes one
aspect of the fundamental analytical way of seeing.
26
Dispositive analysis
The final analytical strategy of Foucault, which I will briefly discuss, is dispositive
analysis. Dispositive analysis is quite difficult to describe and Foucault’s
systematic descriptions of the analysis are very limited. Nevertheless, it is a
central analytical strategy, particularly in Discipline and punish (1977) and History
of sexuality (1978). Most works on Foucault have disregarded dispositive analysis
and have focused on archaeology and genealogy; those who have addressed
dispositive analysis have described it very differently. Gilles Deleuze describes
dispositive as a structuring of light composed by lines of different nature
(Deleuze, 1992), Neil Brenner as a functional imperative (Brenner, 1994), and
Mitchell Dean as a regime of practices (Dean, 1999, p 21).
What makes dispositive analysis difficult to grasp is the fact that it can only
be perceived as a complementary analytical strategy, which succeeds the
archaeology, genealogy and self-technology analyses. These analytical strategies
are preconditions of dispositive analysis because the focus of dispositive analysis
is precisely the interconnections between different discourses, institutions,
practices, self-technologies, tactics and so on, within a particular period.
Foucault defines dispositive as:
1. “a thoroughly heterogeneous ensemble consisting of discourses, institutions,

architectural forms, regulatory decisions, laws, administrative measures,
scientific statements, philosophical, moral and philanthropic propositions”;
2. “the nature of connections that can exist between these heterogeneous
elements”’;
3. “a strategic imperative” (Foucault, 1980, pp 194-5).
Whereas archaeology divides the world into the regulation and dispersion of
statements, geneaology into continuity and discontinuity, and self-technology
analysis into subjection and subjectivation, dispositive analysis divides the world
into apparatus on one hand and strategic logic on the other hand. The distinction
between apparatus and strategic logic is the ‘eye’ of dispositive analysis.
The apparatus is the ‘heterogeneous ensemble’; it is a system of elements
between which there exists a functional connection. The strategic imperative
or logic is a generalised schematic that brings about a particular logic. These
are always relative in relation to one another. There is no apparatus without
the apparatus acting as an apparatisation and, thus, a function of a strategic
logic. In turn, there is no strategic logic except through the effects it defines
through an apparatus. I have attempted to outline dispositive analysis in Figure
1.3.
Generalisation
This analysis is in fact a ‘two-in-one’ analytical strategy, since Foucault oscillates

between viewing the problem from the side of the apparatus and the side of
the strategy respectively of the guiding distinction. When indicating the side
27
Figure 1.3: Dispositive analysis
Dispositive
Apparatus Strategic logic
Discourse For example:

Self-technology legal/illegal
Architecture desired/undesired behaviour

Institutions security/insecurity
........ confession/perversion
‘strategy’ of the distinction, Foucault examines the way in which a discursive

element is taken out of the specific context in which it was created, and is
then generalised and given schematic qualities so that it is made available as
logic to a multiplicity of discourses, self-technologies, institutions and so on.
For example, Foucault examines how specific laws are generalised within the
schematic of legal/illegal, which deals with its surrounding environment
according to very specific procedures. Or the way discipline is generalised
within the schematic of desired/undesired with a functionality that concerns
the prevention of the undesired before it occurs, subjecting its environment to
a particular disciplining. Or how specific actions in the army are generalised
within the schematic of security/insecurity with a functionality about being
prepared for the unexpected, and preventing negative effects and maybe even
turning them to an advantage (Raffnsøe, 2000, p 58).
These generalisations always mean that the specific technology or discursive
element is taken out of its specific context and becomes a strategic logic that
can be installed into many more contexts than just the one that produced it.
For example...
Foucault speaks of the ‘prisonisation’ of society, thus pointing to how the logic
of discipline, which is born out of a context of surveillance and punishment, is
generalised so that it can subsequently obtain meaning in educational systems,
workplaces, and so on.
Apparatisation
On the other hand, when indicating the side of the apparatus in the guiding
distinction, the question is how a particular strategic logic is brought about
through apparatisation, in which forms, such as discourses, technologies and
architectures, are linked as functional elements in a system.
28
For example...
One possible question is:
• How are elements linked together to form an educational system in a

particular way when the logic of discipline is formed?
We have schools that divide students according to age, divide space into classrooms
and divide time into lessons. We have architectural forms that construct classrooms
with peepholes so that the class does not know if or when it is being observed by
the principal.
The two analytical movements are illustrated in Figure 1.4.

The study of generalisation is therefore about becoming aware of the strategic
logics brought about by technologies, discourses, self-technologies, rites,
decisions, practical arrangements and so on. It is important not to perceive
the strategic logics teleologically: they serve no purpose that extends beyond
themselves. They establish a kind of intention and direction without concrete
content and programme, and without a particular will or subjectivity. The
logic is an abstract intentionality, defined solely by a schematic (for example,
desired+/undesired–), with a basic motivation and strategy without subject
and content. What is lacking in the strategic logics, with respect to content,
will and programme, must be provided by the apparatus.
By contrast, the study of the apparatisation of the logic concerns the way in
which these logics are brought about through linking elements in an apparatus,
and how the actual creation of the apparatus specifies and programmes the
unfolding of the logics. In the study of the apparatisation, it is important to be
aware of at least four conditions:
Figure 1.4: The double movement of dispositive analysis
Strategic logic Apparatus How are discursive or technological elements

generalised into a schematic, which forms a specific
strategic logic?
How are forms linked as functional elements in an

Apparatus Strategic logic apparatus that brings about a specific strategic logic?
29
• First, the elements of discourse and self-technology must have been

constructed as objects in preceding archaeological discourse analysis and
self-technology analysis before they can be studied as ‘functional elements-
in-an-apparatus’. Thus, it is the discourse analysis and not the dispositive
analysis that constructs the individual discourse through its awareness of
the regularity in the dispersion of statements. Because of this, I find it
difficult to perceive of dispositive analysis as more than an extension of a
number of logically preceding analyses.
• Second, an apparatisation of a particular strategic logic is not constructed
‘from scratch’. Self-technologies, knowledge, institutions and so on will
often have previously formed part of other apparatuses. Apparatisation,
thus, is often about the way elements are reinstalled with new relational
functions in the bringing about of other logics.
• Third, it is possible for several logics to be involved in the creation of an
apparatus. It is an element of the analysis of an apparatisation to examine
both how the logic is deflected by the functional elements that are linked
together in a specific apparatus and also how different logics deflect each
other within the same apparatus. If we again take the school as an example,
it does not only become an apparatisation of the logic of discipline; the
logic of law is also imprinted in the school and deflects the logic of discipline
by permitting and prohibiting certain disciplining technologies.
• Fourth, Foucault is highly aware of the fact that no apparatisation takes
place without counter-power also moving into the apparatuses in the shape
of forms and elements that cannot be absorbed by strategic logics. Neil
Brenner even suggests an understanding of the question of counter-power
as a Foucauldian version of dysfunctionality (Brenner, 1994).
Conclusion
In conclusion, we see in Foucault at least four different analytical strategies,

each constituted on their guiding distinction. Genealogical analysis seeks,
by a gaze of disruption, to open up the discursive field through tracing
practices, discourses and institutional lines of descent, including the lines of
connection to different historical conflicts and strategies of control.
Archaeological analytical strategy, in turn, seeks to describe discourses as
regularities in the dispersion of statements, including the displacement over
time of these regularities. Together they form an analytical strategy that is
designed to evoke formation and transformation. Self-technology analysis
attempts to capture the ways in which an individual can give itself to itself.
Finally, there is dispositive analysis, which inquires, as a layer on top of the
other analytical strategies, about how discursive and extra-discursive elements
are linked together in an apparatus in the bringing about of a particular
strategic logic.
I have attempted to encapsulate these analytical strategies in Table 1.1. The
left-hand column indicates the analytical strategies; the centre column tries to
30
capture the particular way of inquiring of each strategy; the right-hand column
provides a concrete example of how a specific social issue can be discussed,
based on Foucault, and what the differences are depending on which of
Foucault’s analytical strategies is employed.
The example used here pertains to the development of the relationship
between employee and organisation. Over the past 20 years, the construction
and definition of the employee has been put on the agenda. The relationships
between individual and role has been addressed. Employees’ personal lives
have been integrated into the workplace. Staff policies have become life policies
with management speak including terms such as ‘lifelong learning’,‘flexibility’,
‘personal competencies’, the ‘complete human being’. New practices, such as
performance and competence reviews, have come to prevail.
Table 1.1: Foucault’s analytical strategies

Analytical strategy General question Example
Archaeological discourse analysis Why did this and not another In what way has a new discourse
statement occur in this place? been established that articulates the
employee as a complete and
responsible human being with
responsibility for his or her own
development?
Genealogy How are different discursive In what way does the new discourse
formations and discursive about the complete employee not
strategies shaped and transformed? simply express a humanisation of the
workplace, but also an
internationalization of previous forms
of discipline
Self-technology analysis How have self-technologies been How are performance and
created and how do they prescribe competence reviews developed and
the way an individual can give itself carried out as self-technologies,
to itself? through which the employee can learn
to master himself as strategic resource
in the workplace?
Dispositive analysis How are forms linked together as In what way does the educational way
functional elements of an apparatus? of seeing become a general schematic
How are discursive or technological and strategy for organising, and how is
elements generalised in a schematic the educational scheme propagated
that develops a strategic logic? and apparatised in employee
discourses, self-technologies, office
architecture, and so on?
31
As it is hopefully clear from Table 1.1, the choice of analytical strategy makes
an important difference to one’s empirical and problem-related sensitivity:
• Archaeology illuminates the emergence and regimental character of the

discourse.
• Genealogy gives insight into the displacement of disciplinary forms.
• Self-technology analysis highlights the ways of ordering of the new practices.
• Dispositive analysis makes one aware of the logics that unfold through the
new technologies, practices and institutions, and how these are linked
together in an apparatus.
Notes
1
My exposition of The order of things draws on Raffnsøe, 1999.
2
For a further discussion of Foucault’s genealogy, see for example, Megill, 1979;
Roth, 1981; Shiner, 1982; Kent, 1986; Mahon, 1992.
32
2
Reinhart Koselleck’s history
of concepts
einhart Koselleck’s history of concepts probably represents the clearest
R breakthrough of the linguistic turn within historical science. Koselleck’s

history of concepts was developed at end of the 1950s and, since then,
has become an extensive programme for the study of the rise of modernity.
The programme studies the origins of a great number of concepts, particularly
political and administrative ones, in connection with the effects of their social
history. This has resulted in a comprehensive encyclopaedia in eight volumes
of the history of concepts entitled Geschichtliche Grundbegriffe, Historisches Lexikon
zur politisch-sozialen Sprache in Deutschland (Brunner et al, 1972). The
encyclopaedia includes concepts such as ‘politics’, ‘citizen’, ‘public’, ‘nature’,
‘interest’, ‘representation’, ‘administration’ and ‘tyranny’.
Jan Ifversen (1997) characterises the history of concepts as a linguistic turn
within the discipline of social history. Although the history of concepts has
developed as an indication of limitations and problems in traditional social
history, it has never proclaimed itself to be a new, radical, historical science in
opposition to all traditional historiography. The history of concepts has
therefore been able to develop without a destructive conflict with the less
theoretical historical sciences. On the contrary, the history of concepts has
commanded respect from traditional historians, while Foucault’s discourse-
analytical historiography has never obtained recognition within traditional
historical disciplines. The history of concepts has been seen and has seen itself
as a niche within historical science, but not until recently has conceptual
historical thinking extended beyond the encyclopaedia to which it is rooted.
This shift is particularly visible in Scandinavia in the work of historians such
as Bo Stråht, Henrik Sidenius and Kari Palonnen (see for example, Stråht,
1990).
This chapter focuses only on Koselleck’s history of concepts. There have
been other attempts at writing the history of concepts, but from the discourse-
analysis perspective, Koselleck’s contribution is of the greatest interest (see for
example, Pocock, 1987; Tully, 1988; Richter, 1990, 1995). As Ifversen points
out (1997), the other efforts often do not reach beyond assumptions in respect
of conceptual history. For example, Quentin Skinner (1984), despite an
intention to do the opposite, remains located in the history of ideas, focusing
on canonical works without regard for more anonymous texts (such as
33
administrative texts) and without regard for those semantic or discursive fields
of which the canonical works form a part (Ifversen and Østergård, 1996,
p 25).
The relevance of the history of concepts can be summed up in three
characteristics:
1. It presents a serious, anti-essentialist and constructivist challenge to traditional

historiography.
2. It is founded on a reasonably simple theoretical apparatus that has not yet
been unfolded, but which, in turn, has proved to be extremely sensitive to
the empirical, exceedingly operational and analytically powerful.
3. It has succeeded in functioning on two different levels at once by writing
the history of individual concepts, their origins, transformation and effects
on social and political practices, as well as combining these individual
histories in a joined description of the transformation of political semantics
in the notion of neuzeit (neuzeit is the designation by historians of concepts
of the transition from premodernity to modernity).
Origins of the history of concepts
The fundamental premise for the history of concepts is the idea that concepts
are central to the constitution of society, including the constitution of action
as well as agents of action. The shaping of concepts is hence not a simple
surface phenomenon. If historical science is indeed seeking to raise critical
constitutional questions, it must do so by defining the shaping of concepts as
the central issue. Without concepts, notes Koselleck, there is no society and
no political fields of action (Koselleck, 1982, p 410). This does not imply that
everything can be reduced to concepts. According to Koselleck, the history of
concepts must include linguistic as well as sociohistorical data – any semantics
entail non-linguistic content (Koselleck, 1982, p 414). (We will return to this
point in connection with the notion of concept later in this chapter.)
The constitution of society can therefore be examined as a semantic battle
about the political and social; a battle about the definition, defence and
occupation of conceptually composed positions. Concepts must be perceived
of as reaching into the future: “Concepts no longer merely serve to define
given states of affairs, they reach into the future. Increasingly concepts of the
future were created, positions that were to be won had to be first linguistically
formulated before it was possible to even enter or permanently occupy them”
(Koselleck, 1982, pp 413-14).
Included in this proposition is the notion that semantics change at a slower
pace than the events themselves. There are several factors related to this point:
• First, the fact that semantics, as a repository of articulated experience, contains

the conditions of possible events. Semantics anticipate possible events but
not their necessity.
34
Reinhart Koselleck’s history of concepts
• Second, the fact that the possible ways of semantic expression are limited
gives it a more enduring stability. The semantic units (concepts) outlive
the occasions comprised by the history of events:
When new experience becomes, as it were, a part of language’s inventory,
as in the case of the centuries-old but ever more nuanced debate about
the constitutions, or in the perennially recurring conflict between different
notions of might and right, then the semantics has a slower rate of
change than the events themselves. (Koselleck, 1989, p 657)
• Third, precisely because the concepts possess a different inner time structure
than the events they participate in inciting or conceiving of, they are marked
by perseverance in respect to their epoch, which turns semantics into an
obvious historical object.
Knowledge interests
The purpose of the history of concepts is not merely to add yet another
object to the historical sciences. The purpose is to examine the way in which
the shaping of concepts and the transformation of the semantic fields move
history and reach into our future. Koselleck criticises the kind of historical
science that devotes itself exclusively to the study of individual incidents on
the assumption that detailed studies of incidents are less problematic than
studies of the structural history of longitudinal lines. Koselleck demonstrates
how the narrative content of structures and events is fundamentally the same.
Incidents are always construed narratively as events and, moreover, any event-
oriented historical science is too meagre and is unable to contribute to a
historical diagnosis of potential futures (Koselleck, 1985, pp 105-15).
The history of concepts also distinguishes between diachronic and
synchronous analysis. Diachronic analysis consists of an analysis of the historical
origins and transformation of individual concepts, and thus suggests a notion
of concept. Synchronous analysis consists of an analysis of the semantic field
in which concepts appear and connect with other concepts. Synchronous
analysis frequently refers to concepts such as the ‘semantic field’, ‘semantic
structure’ and ‘counter-concepts’. I will further address these two conceptual
sets below; however, initially I wish to stress that, to Koselleck, there can be no
qualified history of concepts without both of them. The concept-historical
analysis must continually move back and forth between the diachronic and
the synchronous in order neither to fall back on a barren history of words nor
forwards onto the pitiless social history of concepts.
The notion of concept

What is, then, Koselleck’s notion of concept? First, it must be emphasised
that, to Koselleck, concepts hold the very key to the unfolding of the space of
signification of which experience and expectations form a part (Ifversen, 1997,
p 448). The central concept is not discourse, statement or sentence, but concept
35
alone, and this indicates one of the main differences between the notion of
concept and many other discursive traditions such as those of Foucault and
Laclau, for example.
Accordingly, concepts provide the portal to studies of the historical space of
signification. But,
• What, then, is a concept?
Koselleck highlights the importance of distinguishing between word and

concept in order not to depart into a history of mere words. Subsequently, he
unfolds his notion of concept in a triad composed by word (that is, designation),
meaning (that is, concept) and object (that is, fact) (Koselleck, 1972):
Each concept is associated with a word, but not every word is a social and
political concept. Social and political concepts possess a substantial claim of
generality and always have many meanings. (Koselleck, 1982, p 418)
The distinctions between words, concepts and facts serve to define an

autonomous space for the concept. Concepts are tied to words, but they are
not one with the words; concepts refer to facts, but they are not identical to
the facts, because the concepts provide the surroundings with meaning – not
the reverse.
The difference between words and concepts revolves around the ability of
words to work unambiguously in practical usage.
For example...
There are several meanings to the word ‘tramp’, but when used it is unambiguous.
If someone gives advice to a potential boyfriend – “Stay clear of Sheila. She is a
tramp” – the meaning is clear.
Conversely, a concept has to remain ambiguous, even when applied, in order

to continue to be a concept – concepts are always fundamentally ambiguous.
For example...
One example is the concept of ‘equality’, which is never completely defined and
can signify many things, from equal pay, minority issues and ethnic equality, to the
struggle against the patriarch or the concept of ‘the state’ (which involves a
range of elements such as territory, power, tax collection, jurisdiction). Or it
may signify the concept of ‘ecological sustainability’ (which may include eco-
balance or the relationship between economic growth and the environment).
This concept functions as a point of reference to environmental bureaucrats as
well as a utopia to dedicated environmentalists; it signifies both the good life and
a pragmatic balancing of ecology and economy.
36
Koselleck specifies the construction in this way: “The meaning of a word can
be determined exactly through definition, concepts can only be interpreted”
(Koselleck, 1972, p XXIII, author’s translation).
According to Koselleck, words become concepts through the condensation
of a wide range of social and political meanings. Thus, concepts comprise an
undecided abundance of meaning, a concentration of meaning, which makes
them ambiguous. Without this ambiguity, concepts would not be capable of
reaching into the future but, precisely because of its ambiguity, the concept
reveals a space of signification that is open to interpretation and can become
a semantic battlefield. Precisely through its ambiguity, the concept can create
positions for later occupation and conquest, create time and space and so on.
Without the ambiguity of concepts, there are simply no conditions of
conceptual disagreement. If we were able to define ‘equality’, ‘ecological
sustainability’, ‘freedom’, ‘responsibility’, ‘representation’, there would be no
politics, no semantic battle and, consequently, no history. It is the concept of
equality that organises the fight for equality as a struggle about reducing
ambiguity and establishing positions from which equality can be represented.
Historically, we can observe how various meanings have been condensed into
the concept of equality and, thus, have organised the struggle for equality.
Equality initially was taken to mean equality between the sexes. Gradually, it
became identified with equality in the job market and eventually the conceptual
field was extended to society more broadly so that, today, the conflict is about
the equalisation of equality between the sexes with equality among ethnic
minorities or between disabled and able-bodied people. Slowly, the concept
of the politics of equality is taken over by politics of multiplicity. We are no
longer able to define employment policies only in respect of equality between
the sexes, but must also have regard to a flexibility and multiplicity that includes
a variety of people: men, women, young people, old people, Muslims, Christians,
weak people, strong people and so on. In this way, the concepts reach into the
future: they define a space of signification and they define the meaning of the
struggle for meaning.
Conclusion
The history of concepts thereby suggests an analytical strategy, which analyses
history as a semantic struggle about turning words into concepts through the
condensing a wide range of meaning into the concepts. Of course, this is not
a linear history. A word can remain the same even though its meaning changes.
A concept can remain the same even though its linguistic designation changes.
A concept can survive even though its content changes and so on.
This raises a question that is parallel with (but probably not identical to)
Laclau’s question of hegemony and floating signifiers. According to Laclau, a
hegemonic struggle can only take place if the structure is open – if the signifiers
float above the signified – and the hegemonic struggle is therefore about
arresting the floating. The concept of the floating signifier appears to function
in a similar way to Koselleck’s notion of concept. The ambiguity of the
37
concept indicates that the concept does not possess a fixed relation to the fact
to which it refers, in the same way that Koselleck’s definition implies that the
concept moves into the fact by adding meaning to it. Both characteristics of
the concept (its floating and creative qualities) form the conditions for a semantic
battle.
Semantic field analysis and counter-concepts

According to Koselleck, the analysis of the shaping, endurance and
sociohistorical effects of individual concepts should happen in conjunction
with an analysis of the way concepts appear in relation to other concepts –
what he calls semantic fields. The analytical division of concept and field can
simultaneously be seen as a distinction between diachronic and synchronous
analytical strategies. Only through a continuous oscillation between the two
perspectives “does the political and social process of change emerge” (Ifversen,
1997, p 451, author’s translation).
Unfortunately, neither the definition of the semantic field nor the strategy
for the analysis of semantic fields is sufficiently developed. Koselleck thematises
the semantic field by means of the concept of ‘counter-concepts’ in the article
‘The historical–political semantics of asymmetric counter-concepts’ (1985).
Counter-concept is an analytical concept, which can be brought to bear on
the relationships between concepts within a field. Counter-concept is defined
by Koselleck as a concept that is in constitutive opposition to another, for
example man/woman, public/secret, tolerance/intolerance. Any concept
therefore obtains meaning from its counter-concept, and it is therefore crucial
to study the shaping of concepts in relation to the shaping of counter-concepts.
It is not completely clear whether Koselleck believes that all conceptual
relations contains the characteristics of concept/counter-concept or whether
‘counter-concept’ is one among many possible analytical concepts that can be
used when analysing concepts within their semantic fields. Consequently, it is
unclear whether Kosellek’s article provides a general or specific example for
the analyses of semantic fields. With this reservation in mind, the idea behind
the analysis of counter-concepts is briefly outlined below.
Counter-concepts
Koselleck links the notion of counter-concept with the question of the
construction of collective identities, in particular the construction of political
subjects. He asserts that the construction of identities always involves asymmetric
classification – identities are always shaped in asymmetric relations between ‘us’
and ‘them’. The simple identification of the ‘us’ offers far from adequate
conditions of giving ‘us’ the capacity to act:
But a ‘we’ group can become a politically effective and active unity only
through concepts which are more than just names or typifications. A political
or social agency is first constituted through concepts by means of which it
38
circumscribes itself and hence excludes others; and therefore, by means of

which it defines itself. A group may empirically develop on the basis of
command or consent, of contract or propaganda, of necessity or kinship,
and so forth; but however constituted, concepts are needed within which
the group can recognize itself as a functioning agency. In the sense used
here, a concept does not merely denote such an agency, it marks and creates
the unity. The concept is not merely a sign for, but also a factor in, political
or social groupings. (Koselleck, 1985, p 160)
Accordingly, political identities from movements, parties, interest groupings

and so on are always constructed across the classification ‘us’/‘them’, since
there is no ‘us’ without ‘them’ and the exclusion of ‘them’ from ‘us’ is constituent
for the identity of ‘us’. It is important to emphasise the asymmetric character
of the distinction in the same way that it is critical to realise that ‘them’ is not
an ontological category. ‘They’ do not refer to an objective fact – in Lacanian
terms, one could say that ‘they’ speak in ‘us’.
Subsequently, concepts are necessary in order for a group to define and
recognise itself as a functioning subject. Concepts not only classify but also
mark and create unity. Again, it is important to bear in mind that concepts
exist in a concentration of ambiguous meaning that ensues as a fundamental
quality of the constitution of the group. Thus, the semantic struggle about
concepts becomes central to the shaping of political identities.
In effect, this paves the way for studies of how political identities are shaped
in conjunction with the designation of concepts and counter-concepts to
specific identities: the construction of the bourgeois (as opposed to the
proletarian), of the socialist (as opposed to the liberal), of the West (as opposed
to the East), of the Protestant (as opposed to the Catholic). Moreover, it
allows for studies of the effects that the shaping of concepts has on the
possibilities for the shaping of identity. Often, a concept will prevail whereas
its counter-concept will be replaced, so that, for example, the opposition of
the West contra the East transforms into the West contra Communism, which
changes again into the West contra the Muslim world.
Singularity/generality
Still, the creation of identity through counter-concepts cannot occur without
tension in the relationship between, in Koselleck’s terms, singularity and generality.
In many ways, this separation coincides with Laclau’s distinction between
particular and universal (a point to which I return in Chapter Three). On the
one hand, we can ascertain that concepts such as ‘movement’, ‘party’ and
‘interest group’ can be exercised on equal terms in the self-construction of a
range of different identities. The concepts of ‘movement’, ‘party’ and ‘interest
group’ are transferable in the sense that they can be subsumed, applied and
transcribed by many different groups in a large number of contexts. Similarly,
the identities that are constructed in association with the concepts of
39
‘movement’, ‘party’ and ‘interest group’ are reciprocal, in the sense that they do
not exclude and incapacitate each other.
For example...
In the current Danish Parliament there are 10 parties that all shape their identities
by referring to the concept of ‘party’, which, in itself, does not present a problem.
The Social Democrats and Venstre (Liberal Party) are different, but reciprocal,
identities that do not incapacitate each other.
On the other hand, Koselleck notes that there tends to be singularisation, that
is, for the general to be subjected to the singular: “Historical agencies tend to
establish their singularity by means of general concepts, claiming them as
their own” (Koselleck, 1985, p 160). This is the case when a religious community
lays claim to the Church, or when a political party claims to the represent the
people. The Communist Party in the Soviet Union was an extreme example
of the singularisation of the general designation of ‘people’ and ‘party’. However,
the same thing is true when generality is being restricted, for example through
prohibiting certain political parties (such as the Communist Party or the
National Socialist Party), or not allowing a religious community to be
recognised as such. Generally, this applies whenever specific, that is, singular,
requirements exist for affiliating oneself with the general. These requirements
are always present, and hence the general is never truly general. There is
always a limit to transferability, and reciprocal identities are therefore only
reciprocal in relation to the identities they collectively exclude. The singular
and the general never appear as pure differences, they always contaminate
each other.
The problem of the singularisation of the general requires an analysis of
how identities are constructed in relation to the shaping of concepts to focus
always on how the tension between singularity and generality is defined.
For example...
When studying the formation of an organisation, however proper it may seem

(and probably in particular when studying good organisations), we need to explore
how this organisation singularises certain concepts. How, for one thing, is the
concept of ‘help’ singularised as volunteer organisations subsume this concept in
an increasing number of areas? For example, help provided by organisations
such as Danchurchaid, Livsliniens and the Samaritans for suicidal people, patients’
organisations and their counselling groups for people with cancer. How does
‘help’ increasingly become something that professional organisations induce and
something, in effect, that is detached from the personal relationships of everyday
living?
40
Furthermore, the question of singularisation impels the analyst to look at

conceptual connections, because singularisation provokes counter-concepts
from those groups that have been defined as the ‘other’. Without awareness of
how the origin of a concepts is constitutively related to its counter-concept,
one loses sensibility to the sociohistorical effects of the shaping of concepts.
Without the analysis of the semantic field, one ends up, as previously noted, in
the barren history of words.
However, the counter-concept also possesses a history of its own. In other
words, the conditions of the shaping of counter-concepts change historically and hence
the conditions of the analysis of semantic fields also change by means of the
opposition of concept/counter-concept. The tension between concept and
counter-concept is not the same in a modern and a premodern context.
Koselleck demonstrates this through an analysis of the conceptual history of
counter-concepts and, in so doing, simultaneously analyses the conceptual–
historical conditions of his own history of concepts!
Somewhat simplified, the history of counter-concepts is outlined below.
The history of counter-concepts
In antiquity, the opposition of the Hellenes and the barbarians demonstrates

how counter-concepts can evolve. Until the 6th century BC, all non-Greeks
were regarded as barbarians (those who spoke a strange language: ‘bar-bar’).
Subsequently, the barbarian became a counter-concept to the Hellenes. The Greeks
began to describe themselves as Hellenes, and the notion of Hellenes versus
barbarian became a universal structure that included all human beings in two
groups. The structure was asymmetric, since the difference was always articulated
by the Hellenes. The barbarian became symbolic of strangers, who were defined
negatively and were attributed with qualities such as cowardice, brutality and
ignorance. They were everything the Hellenes were not.
This distinction served different functions at different times. It was used to

secure the unity of the Hellenes, to guard against civil war, as justification of the
exploitation of the barbarians as slaves, and so forth. Initially, Hellenic and barbarian
were mutually excluding concepts, in which the barbarians as strangers were
demarcated in negative terms, but recognised as such. Later on, the concepts
become territorialised so not only did they differentiate the stranger, they also
demarcated the Hellenic territories from those of other countries.
Christianity introduced the next phase. Rather than maintaining the distinction
between Hellenes and barbarians, Christians recruited members from both sides.
They did not define themselves in opposition to Romans, Jews, Hellenes and so
on; the concept ‘Christian’ was created potentially to encompass all human beings,
regardless of race, class and so on. However, the distinction between Hellenic and
barbarian was gradually superseded by the distinction between Christian and
heathen, as a separation of religious and non-religious people. The distinction was
41
construed temporarily, according to the prevailing notion that all heathens would
eventually become Christians. This temporality installed a dynamics in Christianity.
Gradually, however, through the crusades, for example, becoming a Christian
was not just an option. Exclusion and expulsion became the other side of
Christianity.
The third phase is represented by humanisation. Until this point, beginning with
the renaisance, ‘man’ signified the unity of differences. Humankind was the sum
of the Hellenes and the barbarians, and, later, the sum of the Christians and the
heathens. Today, however, humankind and humanity have developed into an all-
encompassing counter-concept that outdoes the opposition between Christians
and heathens. This has happened in conjunction with the discovery of America
and general globalisation, which means it is no longer possible to discover and
annex new land ad infinitum; and the dispersal of the Christians into different
religious communities, which means that the concept of ‘Christian’ in itself is no
longer singular. Humanity thus emerges as a concept that produces a minimum
definition uniting the divided Christians.
At the same time, we see a gradual repression by political theory of theology

with God as the Creator. Humanity is generalised as a concept in relation to
which all other concepts become particular. The generalised concept of humanity
obtains a critical function – it serves as the foundation for the criticism of things
as inhuman. As the all-encompassing concept of humanity is subsumed by the
political, it attains totalitarian effects. Humanity changes side from being a counter-
concept to becoming a concept, but this process turns it into an empty category
that must constantly be imbued with new meaning. Accordingly, this necessitates
an internal differentiation of humanity into superior and inferior people in an
attempt to charge the empty category with meaning. At first, the difference
superior/inferior person is transcribed from the societal hierarchy of superior/
inferior differences, and refers the superior people to the leading classes of
society. Then the distinction is liberated from the concepts of class and hierarchy.
A superior person is no longer someone who belongs to a particular class, but
someone who is ‘more than human’ and thus incarnates humanity as such. In
the writings of Karl Marx, the relationship between superior and inferior people
is reversed and, in effect, superiority becomes a concept laden with negativity,
which functions as a critique of leading people who think too much of themselves.
The rest of the history is well known.
Thus, not only do the definitions of concept and counter-concept change,

the tension between them also changes. At first, the ‘other’ is defined negatively
but recognised as such. Then the concepts become territorialised and grow to
demarcate control of space. Subsequently, the relationship between concept
and counter-concept is temporalised so that, in effect, people associated with
the counter-concept can surpass the difference from counter-concept to
concept. Finally, the concept is totalised and generalised, which turns it into
42
an empty category that needs to be continually imbued with new meaning. It

is therefore impossible to possess any preconceived knowledge of the precise
characteristics of a counter-concept, which means that the construction of
the analytical concepts becomes an integral part of the concept-historical
analysis.
Five fundamental prelinguistic distinctions

As we have seen, Koselleck forms a distinction between the diachronic (in
which studies of the rise, duration and transformation of the individual concepts
are conducted) and the synchronous (which addresses the structural history
of the extended lines, focusing on the construction and displacement of semantic
fields). However, Koselleck actually employs yet another level, a metahistorical
or prelinguistic level, which consists of couples of opposition or two-sided differences
that precede the historical articulation of the concepts. They are differences
or antitheses “without which no history can come to be, regardless of the
forms they take on in particular cases – economic, religious, political, or social,
or something involving all these empirical factors” (Koselleck, 1989, p 651).
Evidently, all these prelinguistic couples of opposition are:
grasped by man [sic] linguistically: by means of language they are reshaped

socially or regulated politically. There is no acting human community that
does not determine itself linguistically. It is almost always membership of a
certain linguistic community that determines whether one is included or
excluded. It is almost always the capacity to master certain modes of speech
or specialized languages that decides whether one will move higher or lower,
up or down, in a society. Almost always there are linguistic norms that are
generation-specific and that sort out, diachronically, the experiences and
hopes of the old from those of the young man, as a linguistic being, simply
cannot avoid transforming the metahistorical givens linguistically in order
to regulate and direct them, so far as he can. (Koselleck, 1989, p 652)
Koselleck’s aim is to connect the history of concepts to Heidegger’s

existentialism, and the couples of opposition below are originally taken from
Heidegger’s Being and time (1978). Koselleck distinguishes between three
prelinguistic couples of opposition, and my understanding is that these three
couples form a sketch for later elaboration. The three couples are:
• before/after;
• inside/outside;
• up/down.
(They are sometimes portrayed as five couples: before/after; birth/death; inside/
outside; friend/enemy; up/down.)
43
Before/after
Any history is, by its nature, poised between a before and an after, and this
distinction is prelinguistic insofar as we are always living in the space between
birth and death. This means not only that the lifespan is limited, but also that
we live with death as the horizon and know that murder as well as suicide is
possible. The distinction is furthermore prelinguistic in the sense that
generations succeed generations, which results in a range of overlapping fields
of experience that exclude each other in layers. Here, Koselleck speaks of
diachronic conflicts (Koselleck, 1987, pp 11-14, 1989).
Koselleck extended the difference before/after in his distinction between
space of experience and horizon of expectation (Koselleck, 1985, pp 226-88). ‘Space
of experience’ and ‘horizon of expectation’ should be understood as
epistemological categories. There can be no time–space relationship and thus
no history without a conceptual formation of a space of experience and a
horizon of expectation. The present does not exist in itself but only in the
tension between a space of experience and a horizon of expectation, and all
forms of political and social action must therefore ascribe to this tension. It is
of great significance that it is a couple of opposition, that is, a two-sided difference:
there is no experience without expectation and vice versa. The space of
experience and the horizon of expectation thus corrupt one another; no side
of the difference holds a preliminary status of privilege. The space of experience
is seen as present past (that is, as the events that have been incorporated and
remembered), while expectations designate present future (which directs itself
towards the forthcoming, towards that which has not yet been experienced,
towards that which will be prompted later). The horizon of experience is the
line behind which new experience will emerge that cannot yet be seen
(Koselleck, 1985, p 273).
To Koselleck, historical time is created, in continuously changing patterns
and relations, by the mere tension between expectation and experience.
Concept–historical analysis should uncover that which applies over time as
expectation and experience, but also the tensions and relationships between
expectation and experience, which are historical and must be exposed by the
semantic analysis (Koselleck, 1985, p 275).
Inside/outside
No society or human action exists without a distinction between inside and
outside. The distinction inside/outside is constitutive for the spaciousness of
history:
In this formal opposition, too, lie the seeds of varies potential histories.
Whether it be the embattled retreat into a cave, or the forcible enclosure of
a house, whether it be the drawing of a border that occasions, or concludes,
a conflict, or rites of initiation, whether we are talking about grants of
44
asylum, or secret societies, or the examination systems and admissions

qualifications that create modern social entities, or membership in a political
community into which one normally is born – in all these cases the
difference between ‘inner’ and ‘outer’ remains fundamental for the conflicts
that arise and for their resolutions, fundamental, in short, for the histories
with which we are all familiar. (Koselleck, 1989, p 651)
The distinction inside/outside is pushed to its extremes analogous to the

friend/enemy distinction (at times, however, Koselleck completely separates
the two distinctions). Inside we stand shoulder to shoulder in respect of our
common welfare; outside is that which threatens our welfare:
We have to understand that the couple of opposition of friend and enemy

formally addresses mortality, which emerges behind the scene of all human
self-organization. Regardless of whether in actual history Greeks and
Barbarians or Greeks and Greeks fight each other, whether Christians and
heathens have fought or Christians among themselves, whether the modern
unities of action establish themselves in the name of humanity and treat
the opponent as a brute, or the unities of action consider themselves subjects
of class in order to actually overrule the classes – the empirical extension in
its diachronic succession presupposes the couple of opposition of friend
and enemy. (Koselleck, 1987, pp 14-15, author’s translation)
Up/down
There are also different positions within a community. On the inside, we are
joined against the things or people on the outside, but inside this community
one can be up or down. The distinction between up and down can therefore
define the internal pecking order. Nevertheless, the distinction can be
articulated in countless ways, for example as a master/slave relationship, as in
Greek and old-European terminology, or as the democratic political system of
modern times that provides the standards for a continuous exchange of up/
down positions (Koselleck, 1989, p 651). In politics, for example, the ruling
party might be in opposition after the elections; new parties are established
and voted into Parliament while other parties lose their support and dissolve.
There can be no political self-organisation, no distribution of relationships of
dependency and hence no history without the broadest definition of the
distinction up/down.
Collectively, the three (or five) couples of opposition form the metahistorical
conditions of the constitution of history: “According to Heidegger, we are
dealing with existential definitions, that is, a particular kind of transcendental
categories which indicate the possibility of histories without, however,
describing specific histories” (Koselleck, 1987, p 20). In my opinion, Koselleck
wishes to introduce three points.
45
First, he wants to install a distinction between history and language, which

simultaneously inserts a distinction between the occurrence and the articulation of
the occurrence as an event. Articulation, however, only becomes intelligible
in relation to history. The distinction between history as a chain of occurrences
and the articulation of these chains is itself constituent of the relationship
between them:
It is this difference between history in the actual process of its occurrence

and history in its linguistic elaboration that remains, in any case, fundamental
for their relationship. In any case, this difference between history as it takes
place and its adaptation in language is constituent of the relationship between
them. (Koselleck, 1989, p 652)
The same applies to the language that precedes the events and contributes, in
spoken or written form, to triggering the events. Moreover, the articulation is
a process of selection (my choice of words!). Not all occurrences become
articulated as events and those that do become events could equally have been
articulated as different events. Finally, some occurrences simply do not allow
for articulation; language fails and takes the form of a boisterous silence:
When the fluctuating distinction between ‘inner’ and ‘outer’ hardens into
passionate conflict between friend and foe, when the inevitability of death
is pre-empted by killing or by self-sacrifice, when the relation between ‘above’
and ‘below’ leads to enslavement and permanent subjugation or to
exploitation and class struggle, or when the tension between the sexes leads
to degradation – in all these cases there will occur events, or chains of
events, or even cataracts of events, which are beyond the pale of language,
and to which words, all sentences, all speech can only react. There are
events for which words fail us, which leave us dumb, and to which, perhaps,
we can only react with silence. (Koselleck, 1989, p 652)
Second, by employing the couples of opposition, Koselleck seeks to establish

a space outside of language (although Koselleck is clearly aware that he has
defined the five couples within language) that can define some fundamental
guidelines for the history of concepts. This space outside language enables
him to raise questions about how the five couples are active through history,
how they are conceptualised over time, unfolded within different semantic
fields and, in general, how, across the five distinctions, new semantic structures
that constitute society and its players are continuously created.
Third, the metahistorical couples of opposition serve as a point of reflection
for historiography, since historiography is subject to the same prelinguistic
distinctions. Accordingly, Koselleck conducts a rather extensive critique of
the historical-philological criticism:
• because it disregards the fact that historical reality only exists in shapes of
language;
46
• because it disregards the fact that a selection of that which has been deemed
worthy of reminiscence has already taken place;
• because it disregards that it is always the linguistically-fixed events that
imbue unique events with their meaning of relative duration or specific
signification;
• most importantly, because it fails to explain why formerly reliable and, for
that reason, abandoned histories have to be rewritten at all (Koselleck,
1989, pp 661-6).
Conclusion
Koselleck’s history of concepts consists of a combination of two analytics,
between which one is continuously going back and forth. One is the diachronic
analysis that focuses on single objects and their origins and transformation.
These studies concern the shaping and consolidation of meaning into words,
which are thereby transformed into objects that reach into the future by means
of the constituent effect they have on, for example, the shaping of political
agents, their identity and their ability to act. The study of the history of
concepts is simultaneously a study of the semantic battle about concepts,
including a battle about occupying concepts, about generalising concepts and
about singularising the general.
On the other hand, synchronous analysis studies the way concepts always
come into being in relation to other concepts in a semantic field. A semantic
field is organised as the relationship between concepts and their counter-
concepts. Prelinguistic couples of opposition provide guiding differences for
the analysis of counter-concepts.
Koselleck’s concept-historical analytical strategy is summarised in Figure
2.1.
Figure 2.1: Synchronous versus diachronic
Synchronous analysis
Diachronic analysis History of single concepts
Concept/meaning
Singularisation/generalisation
Displacements in the semantic fields
Concept/counter-concept
Before/after
Inside/outside
Us/them
Up/down
47
More schematically, the concept-historical analytical strategy is outlined in

Table 2.1. The left-hand column indicates the analytical strategy; the centre
column indicates the general inquiry of the analytical strategy; the right-hand
column provides a concrete example of a possible conceptual history. The
example used in this table is the abnormalisation of eating.
In Western Europe, we have no real lack of food. On the contrary, obesity
seems to have been conceptualised as the most significant nutritional problem.
However, if we take a closer look at the articulation of issues concerning food,
the problem is not simply an excess of fat and the solution is not simply good
nutritional information. Obesity appears to have been conceptualised, not
only as a symptom of malnutrition, but as a signal of an unhealthy lifestyle.
Nutritional experts no longer speak only of our food; they believe they are
able to tell us how to live our lives. Nutritional policies have become life
policies. Why is it that so many heterogeneous meanings – such as ‘healthy
living’, ‘nutriments’, ‘body-consciousness’, ‘prevention’, ‘assessments of lifestyle
risks’,‘individual responsibility’ – have been condensed into our present concepts
of eating? This is a typical concept-historical analysis.
Table 2.1: Koselleck’s analytical strategies

Conceptual history How is meaning condensed How are ‘healthy living’, ‘nutrients’,
into concepts, thus ‘body-consciousness’, etc
constituting the space of condensed into the concept of
possibility of the semantic ‘eating’, thus constituting a conflict
battle? of life policies?
Semantic field analysis How do concepts appear How does the concept of eating
in relation to counter- appear in relation to counter-
concepts, thus creating concepts, thus creating a
a semantic field? semantic field that pertains to
nutritional policies?
48
3
The discourse theory of
Ernesto Laclau
he Argentinian Ernesto Laclau has conducted one of the most
T comprehensive rewritings of Foucault’s discourse analysis. Ernesto Laclau

cur rently works in the Depar tment of Gover nment, Essex
University, UK; throughout his career he has been interested in questions of
politics and the state. He was associated with the Marxist criticism of the
1970s and had a similar perspective to those of Louis Althusser and Antonius
Gramsci. In 1985, he published the book Hegemony and socialist strategy, written
together with Chantal Mouffe (1985). In this book, he formulates his final
departure from Marxism and defines a new critical project, constructed around
a discourse-analytical reconstruction of the concept of hegemony. With Chantal
Mouffe, he conducts a genealogical analysis of the concept of hegemony and
reaches an entirely new definition of the concept detached from the Marxist
figures. At the same time, he reconstructs Foucault’s discourse analysis in
order to define a general, pure discourse analysis in which all the non-discursive
elements found in Foucault’s work have been removed. By defining the concept
of hegemony as central to discourse theory, it is restored as a political theory.
As far as I am aware, Ernesto Laclau is alone in defining discourse analysis as
a political theory. I will return to this point later in this chapter.
Ernesto Laclau has maintained his political discourse theory since 1985.
His work is exclusively theoretical and, to my knowledge, he has conducted
no empirical discourse analyses, although he has performed certain
deconstructions of concepts, which are of empirical importance (for example,
the concept of representation; Laclau, 1993b). His theoretical work has primarily
consisted of clarifying and systematising discourse theory; he has not, however,
emphasised the development of analytical tools or operational strategies. Rather,
his aim has always been a general, political discourse theory. In addition to
numerous conceptual definitions, one can trace a certain shift in the central
focus from discourse analysis towards deconstruction. Deconstruction has
always clearly accompanied discourse theory but, at present, Ernesto Laclau
writes more about deconstruction than about discourse analysis, without having
left discourse analysis behind. Similarly, his current inspiration stems less from
Foucault and Althusser, and increasingly from Derrida, Lacan and Zizek1. To
a large extent it is possible to see in Laclau conceptual sets that refer back to
each of these influences respectively. The concepts ‘discourse’, ‘discourse
analysis’,‘moment’,‘genealogy’,‘articulation’ and ‘regulated dispersion’ all derive
from Foucault. Concepts such as ‘floating signifier’,‘empty signifier’ and ‘nodal
point’ derive from Lacan and Zizek. Finally, the concepts ‘undecidability’,
49
‘deconstruction’, ‘logic of supplementation’ and ‘never fully closed structures’

derive from Derrida.
This chapter presents selected key concepts and discusses the aim of Laclau’s
discourse analysis, including the particular relationship between deconstruction
and discourse analysis that I believe suggests itself in Laclau.
Discourse
Let us begin with the concept of discourse. This, of course, is taken from
Foucault, but in this context it embodies a more unambiguous and general
definition. Laclau defines discourse as a structural totality of differences (Laclau
and Mouffe, 1985, pp 105-14): “When, as a result of an articulatory practice,
one has become capable of configuring a system of exact different locations,
this system of different locations is called discourse” (Laclau, 1985, p 113).
Discourse, articulation, discoursivity

There are two important distinctions in respect to the further development of
discourse theory in Laclau. These are different from the other theories presented
in this book. The first distinction is between articulation and discourse, congruent
with the relationship between practice and system. Articulation is designated
by any practice that establishes relationships, that is, differences and similarities
between elements (Laclau and Mouffe, 1985, p 105). Discourse is not in itself
a practice but differs exactly by being the result of a practice, which, from the
observing position of the discourse analyst, can be charged with the
characteristics of an orderly pattern of differences. Contrary to Luhmann’s
notion of system, discourse as a system contains no self that is capable of
differentiating itself from its surroundings. Hence, practice is not located in
the separate discourses, although it exists in a discursive setting, without,
however, having to admit the symmetrical character of the discourse.
This leads to the second distinction between discoursivity and discourse. Within
this distinction, discoursivity signifies the fact that identities (objects, subjects,
technologies, problems and so on) inevitably appear relationally, for example,
only in relation to something else do social identities take on meaning. But
these relationships do not necessarily possess the systemic character of discourse
– the different locations of the identities can be imprecise and floating. Thus,
the division between discoursivity and discourse is simultaneously a division
between floating and fixed (or partially fixed) relationships. This brings us to
the question of emergence, that is:
• How does a discourse emanate from and ascribe systemic properties to

relationships?
Using the distinction between discoursivity and discourse, discourse can be

observed as a never-completed fixation process that takes place through
articulation within a field of discoursivity with drifting relations:“Any discourse
50
The discourse theory of Ernesto Laclau
is constituted as an attempt to dominate the field of discoursivity, to arrest the

flow of differences, to construct a centre” (Laclau and Mouffe, 1985, p 112).
This also means that the relationship between articulation and discourse is
a dynamic one. The ever-floating and hence undecided structure of discourse
prompts articulation as that which attempts to arrest the flow of differences
and determine the undetermined without, however, being able to conclusively
do so. Consequently, articulation can never be presupposed based on discourse.
Articulation is more than, and must be more than, the realisation of the structure.
Discursive (nodal) points

The centre of the discourse, striving to arrest the flow of relations without
ever becoming a centre, consists of privileged discursive points. As Derrida
has shown and as Laclau maintains, a centre must, in order to be a centre in
the classical sense, exist simultaneously within and without the system whose
relationships are being fixed by the centre. However, since the centre of the
discourse itself is defined discursively (that is, defined within), the centre can
never be present outside the system of differences, extending the discursive
game infinitely (Derrida, 1976). More simply: there is no place outside the
discourse that can remain untouched by it and is able to define its order. As a
result, the discursive battle is capable of constantly making the centre pivotal
to that battle in such a way that it changes the rules of the discursive game.
Hence no differences are fixed; at most they can obtain partial fixation. Laclau
identifies these centres that never quite become centres as nodal points (which
is a translation from Lacan of the French point de capition). The relationship
between articulation, discoursivity, discourse and nodal points is summarised
thus:
The practice of articulation, therefore, consists in the construction of nodal

points which partially fix meaning; and the partial character of this fixation
proceeds from the openness of the social, a result, in its turn, of the constant
overflowing of every discourse to the infinitude of the field of discoursivity.
(Laclau and Mouffe, 1985, p 113)
The floating surplus of meaning and undecidability

As already noted, Laclau’s point of departure is a definition of discourse as a
structure, but he stresses on several occasions that the structure never reaches
full closure. On the contrary, a structure tends to collapse from within because
it is never able to completely fix its elements into steady relationships, because
it cannot irrefutably locate something outside the structure to define and
maintain itself.
This is a central point and one that extends to all parts of discourse theory.
The necessary incompleteness of the structure is, regardless of the specific
focus or perspective, a recurrent theme throughout discourse theory, and,
51
depending on the perspective, the problem is given different definitions. It is

therefore necessary to further elaborate this point.
The political
The first detail pertains to the political. Politics happen precisely because
structures are never complete. If a structure were able to reach full closure, it
would exclude the political. Because of the eternal undecidability of the
discursive structures, however, politics acquire a central role within all structures
as that moment when undecidable structures, demanding a conclusion or
closure, become partially fixed. Discourse analysis is a political analysis of the
way contingent relations become fixed in one way, but could have been fixed
in many others.
S/subjects
Furthermore, this is the most substantial critique of Althusser’s ideology theory.
In Althusser, there is no room for the political because the structures are seen
as complete. This is illustrated by the problem of the subject. Althusser makes a
distinction between the small subject and the big Subject. The big Subject is
the model for the creation of all smaller subjects. When an individual or
collective is defined as subject it is granted a steady position in relation to the
big Subject (or the master subject); a steady position in the structure of the
ideology. The individual becomes subject to the Court, to God, to the Economy,
or whatever the big Subject happens to be in the proclaiming ideology in
question. In relation to the declaration, the individual faces a simple choice
between reason and madness, between accepting her/his place and becoming
a ‘sensible’ subject, or being excluded from reason without the possibility of
defining oneself as a subject outside the ideology. If, on the other hand, the
structure is seen as incomplete, the need for politics emerges. The individual
cannot fully identify with the subject position provided by the discourse but
is forced into filling in the structural gaps through identification, and this
identification process is precisely political because it requires a choice that
cannot be explained away – a choice about the subject’s self-constitution.
Signs and signifiers

Moreover, the question of the structure’s incompleteness appears in relation
to the notion of signs. The point of departure is de Saussure’s concept of signs
(1990). A sign in this context consists of the unity of the difference between
signifier and signified, between, for example, a sound-image and the concept
to which the sound-image refers (for example, c-a-t/cat). For de Saussure, it
is the relationship between the signs that constitute the meaning of the sign
(for example between the signs c-a-t/cat, d-o-g/dog, h-o-u-s-e/house,
d-o-g-l-e-a-s-h/dog leash). Language exists as the system of relations between
signs, and language is perceived of as a complete system of fixed relations. But
52
what happens if that structure is seen as incomplete? That would mean that
signifiers and their signified were not irrefutably tied to each other. Instead
they would be partially free floating layers on top of each other; the signifiers
would be floating signifiers.
The expression ‘floating signifier’ was first used by Claude Lévi-Strauss,
although it was the psychoanalyst Jaques Lacan who unfolded the concept
theoretically (Macey, 1988, p 137; Lacan, 2001, pp 146-78). Lacan demonstrates
how the sliding of the signifier across the signified forces the signifier to step
into, or down onto, the level of the signified. Hence, the signifiers influence
that which they signify. This means that signifiers receive a particular status in
relation to signification and to the signified.
For example...
Lacan’s favourite example is that of two identical doors next to each other. The
doors (the signified) are not signified in the same way, however – one door is
signified ‘Gentlemen’ and the other door is signified ‘Ladies’. The effect is obvious:
the exchange of signifiers cuts through the bar (/) and into the signified. The
incomplete structure thus appears, not only in the sliding relationships between
the signs, but also insofar as the bar that separates signifier from signified proves
to be less watertight than assumed by de Saussure.
Laclau transfers these concepts to discourse analysis, which focuses on:
• how signifiers and the signified do not relate to each other in a predetermined
relationship;
• how discursive signification happens precisely in relation to displacements
in the signification of the signified;
• how which signifier is fixed above a given signified is very much a political
issue.
The discursive battle is therefore a conflict or struggle over which signifiers

are to be tied to which signified.
For example...
A dead seal lands on the beach. How should we signify this event? There are still
seals off the shore of xville; pollution takes new victims; the ruthless over-fishing of
herring is starving the seal population; or another flu epidemic ravages the seals
off the sound? The signifier steps down into the signified and transforms the seal
(the signified) into a signifier of its own – the dead seal signifies the diversity of
nature, pollution, a flu epidemic, and so on. The signifier is also obviously political:
should we take measures against fishing or pollution, or simply wonder about
the whim of nature?
53
This battle over fixing sliding signifiers is, as I have noted before, a battle over
the definition of nodal points (or, to use Lacan’s term, privileged signifiers),
which can arrest the sliding of the many signifiers across the signified.
Besides the application of these concepts, Laclau introduces yet another
concept – the notion of the empty signifier. (I believe this is Laclau’s invention,
although, here too, he is heavily indebted to Lacan.) The question of the
signifier that signifies nothing emerges in Lacan’s famous article on the non-
existence of the woman: “This the is a signifier characterized by being the only
signifier which cannot signify anything” (Lacan, 1982, p 144). Such a signifier
is the only signifier capable of ending and defining a limit for the chain of
signification. The analogue question in Laclau addresses the partial finality of
the discursive game. The empty signifier is used to signify that which does
not allow for signification, that is, the limit of the discursive signification: “An
empty signifier can consequently only emerge if there is a structural impossibility
in signification as such, and only if this impossibility can signify itself as an
interruption of the structure of sign” (Laclau, 1996a, p 37).
Summary
A discourse consists of different elements of signification, which only obtain
identity through their mutual differences in the discourse. The condition of
mutual differences, however, is that the elements are identical or equivalent in
respect to belonging to the discourse and existing within the boundaries of
the discourse. On other words, on the one hand, elements can only form an
identity through their mutual differences, while on the other hand, the
differences are cancelled out by the equivalent relation provided by the elements’
attachment to the discursive structure. This is only possible if there are different
types of differences; if a difference exists that is radically constituent for the
differences of the system. Laclau identifies this difference as the excluding
boundary. Within the boundary exists a system of relational elements; outside
the boundary exists only pure, indifferent being in relation to which every
element of the system is equally different. Thus, outside the boundary only
radical indifference exists (which is why the distinction inside/outside the
system should not be confused with Luhmann’s distinction between
surroundings/system).
The empty signifier occurs, in effect, as a possibility for the signification of
“the pure cancellation of all difference” (Laclau, 1996a, p 38). Laclau summarises
the argument thus:
There can be empty signifiers within the field of signification because any
system of signification is structured around an empty place resulting from
the impossibility of producing an object which, nonetheless, is required to
be the systematicity of the system. (Laclau, 1996a, p 40)
An empty signifier is that which signifies the indifferent and the cancellation
of difference. All differences must be equally different in relation to it, while
54
also being different from each other. Locating and analysing the empty signifier
entails a signification of the ultimate limit of the discourse but, since such a
signifier will always be inside the discourse, it will never be possible to represent
it fully.
Laclau illustrates the relationship between equivalence and difference as
portrayed in Figure 3.1.
Hegemonic analysis and the battle of fixating

The formulation of the problem of floating elements of signification opens up
a redefinition of the question of hegemony or supremacy in society. The
concept of hegemony in Laclau’s discourse theory functions, in short, by
focusing the aim of discourse analysis. It is not the main purpose of discourse
analysis to produce individual analyses of conceptual displacement or individual
battles over the fixating of signifiers, but rather to uncover the general
hegemonic relationships in society, and conditions for the transformation of
hegemony.
With the discourse-analytical reconstruction of the concept of hegemony,
other traditionally Marxist concepts find new formulations as well, including,
first and foremost, the concepts ‘opposition’ and ‘antagonism’. However, first,
we will only indicate the direction of the conceptual turn.
The notion of hegemony is linked directly to the argument about the
incompleteness of structures and to the continually fully or partially floating
elements of discourse. The basic understanding is that hegemony is only
possible when something exists that can be hegemonised, and that this is only
the case when discourse lacks final fixation, when the discursive elements
hold a surplus of meaning and when the signifiers are not irreversibly linked
to the signified. Consequently, hegemony signifies the never-concluded
attempts to produce a fixation, to which there will always be a threat. As
Laclau metaphorically notes, hegemony “is like writing in water. It is something
impossible, unstable, and vulnerable, but to a certain extent still something
Figure 3.1: Chains of difference and equivalence
Chain of difference between Chain of equivalence between the elements

elements of the left side through a common
constitutive difference to ‘other’
...X/X/X/X/X/X/X/X/... X
X equals an element
55
that can be accomplished” (Laclau, 1985, p 107). Hegemony creates the space
for “a politics of signifiers” (Laclau, 1983).
Hegemonisation subsequently consists in the imposition onto elements of a
certain way of relating to each other. This also means that hegemonisation
brings elements together that have not previously been brought together (Laclau
and Mouffe, 1985, pp 134-45). The economic hegemonisation in society
appears in elements that were not earlier considered socioeconomic being
suddenly recognised as such, and, furthermore, in these elements now relating
to themselves as social economy. One example could be the emergence of
policies for older people from the mid-1980s.
For example...
Until the mid-1980s the elements of residential homes, older people, home care,
and older people’s housing in Denmark were not associated with elements such
as social–economic balance, inflation, finance and so on. From the mid-1980s,
however, policies for older people became a central socioeconomic and political
issue. Subsequently, a number of policy works are commissioned. This results in
two things. First, residential homes, older people’s housing, home care and other
issues are brought together under the same policy umbrella through the joint
formulation of policies for older people (which has never happened before).
Second, policies for older people establish themselves through the formulation
of a political expenditure issue about the social–economic coordination of public
services for older people. Hence, the hegemonisation of policies for older people
within the socioeconomic conception appears not as a hierarchy of relations of
superiority and inferiority, but as the investment of a particular logic in the
construction and shaping of policies for older people.
Discourse analysis versus deconstruction
As highlighted in the introduction to this chapter, there appears to be a traceable

shift of focus in Laclau’s writings, from discourse analysis to deconstruction.
In this section, I will argue that discourse analysis and deconstruction provide
two completely different perspectives, but that Laclau brings them into a
complementary relationship in which the output of one becomes the input of
the other. The bridge between them appears as that which Laclau calls ‘logics’.
Discourse analysis is, as already noted, an analysis of the discursive system of
dispersion – of the way discursive elements are dispersed and placed in relation
to each other. Deconstruction is in no way synonymous with discourse analysis.
Neither is it one ‘technique’ among several techniques within discourse analysis,
but is in fact in clear opposition to discourse analysis.
56
Origins and definition of deconstruction
The concept of deconstruction derives from Derrida. In the article ‘Letter to

a Japanese friend’ (Derrida, 1988), Derrida replies to a letter he has received in
which he is asked the question: what, really, is deconstruction? Derrida, however,
refuses to answer the question on the assumption that deconstruction in itself
should be anything definite. Instead, Derrida chooses to elaborate on what
deconstruction ought not to be, and in relation to discourse analysis the answer
is rather interesting. Derrida emphasises that deconstruction is neither analysis
nor criticism (Derrida, 1988): deconstruction is not analysis because it refuses
to be reduced to simple principles and it is not criticism because there exists no
place from which criticism can be conducted. Conversely, discourse analysis is
precisely analysis because discourse analysis reduces and refers the many
articulations to a particular system of dispersion. Even though discourse analysis
does not claim, as does structuralism, to be able to refer to one simple principle
but rather maintains the structural openness, the act of conducting a discourse
analysis still requires a reductive description. I will leave whether or not discourse
analysis is a form of criticism open to discussion. But I am of the definite
opinion that deconstruction cannot be incorporated in a discourse analysis.
But then what is deconstruction? Derrida gives the answer to that question
in a subordinate clause:“‘deconstruction’ is precisely the delimiting of ontology”
(Derrida, 1988, p 4). But, since deconstruction consists of such ‘delimiting’,
Derrida must also insist that deconstruction cannot be reduced in this way
since it would then stand outside itself. The result is thus:“What deconstruction
is not? Everything of course! What is deconstruction? Nothing of course!”
(Derrida, 1988, p 5).
Naturally, this is a completely satisfactory definition, so I will endeavour to
answer the question even if Derrida refuses to. My understanding of it is that
deconstruction is about showing how differences are contingent, that is,
deconstruction is about retracting or deconstructing differences by showing
that they are not differences at all; that the ‘bar’ between two elements, which
isolates one from the other, cannot be maintained. Even after the difference
has been established, it is still contingent what it consists in, if at all. It is not
a question of locating differences within a system, or a discourse, but a question
of addressing the individual difference in order to prove that it is not valid.
One example of this is the difference between speech and writing, which
Derrida deconstructs in Of grammatology (1980). Here the issue is the dissolution
of the difference between speech and writing, first by showing that writing
precedes speech and second by posing the argument that speech and writing
are both writing – one phonetic, the other graphic.
Deconstructing difference
The deconstruction of a difference does not end, however, once it has been
shown that the difference is not valid. The illustration of the non-difference
of the difference (so to speak) is simultaneously an illustration of those
57
mechanisms or games that are present in each differentiation. The point is to

see the difference as a mechanism in a game of signification. Usually, there
appears to be a hierarchical relationship between the two elements in a difference
– one tries to subjugate the other. By deconstructing differences, the relationship
is usually reversed in such a way that what appears to be the norm is recognised
as a game of dominance. This could, for instance, pertain to the concept of
sign, which I have previously mentioned in this chapter. It would appear
obvious to assume that the signified would hold the dominant side in the
relation between signifier and signified. Without something to refer to, how
can there be reference? But the deconstruction of the sign puts holes in the
bar dividing the signifier from the signified and the signifier appears on the
same level as the signified. Subsequently, the hierarchy of difference is reversed
– the signifier becomes the dominant element and suddenly our attention is
drawn to a central mechanism of signification. This type of mechanism is
often referred to by Laclau as ‘logic’, and we are thus able to talk about a logic
of signification as the mechanism present in signification and in the displacement
of signification onto the signified.
The relationship between deconstruction and discourse analysis

Deconstruction cannot be incorporated into discourse analysis, but what, then,
is the relationship between them when Laclau insists on using both. The
answer is that the relationship is circular (see Figure 3.2). Deconstruction
pinpoints mechanisms or logics whose unfolding within the discursive battles
of history can be studied in discourse analysis. Discourse analysis, in return, is
able to provide deconstruction with politically central concepts and dualities
(two-sided differences) and so on. The logic of signification functions as a
point of departure for a discourse analysis of, for example, the ecological
discourse. What we examine here is the battle over how floating signifiers are
to be fixed above the signified. We examine how names are interchanged;
how some signifiers are sought out as privileged in order for them partially to
fixate other signifiers (for example, nitrate pollution of drinking water); and
finally, whether there is a breakdown in the signification logic in the form, for
example, of an empty signifier, which causes all meaning to implode. This
could focus on the notion of ‘the ecological latitude’.
Figure 3.2: The relationship between deconstruction and discourse analysis

logics
Opening of the
political
through the Demonstration
deconstruction discourse analysis
illustration of hegemony
of the
undecidability
of a difference
dualities
58
Logics therefore become clues or points of reference for discourse analyses.

Discourse analysis analyses the hegemonic constellations within which logics
play themselves out. Without logics as deconstructive input, discourse analysis
could not obtain the same sensitivity in relation to the central mechanisms or
to structural incompleteness. Laclau describes it thus:
Deconstruction and hegemony are the two essential dimensions of a single

theoretico-practical operation. Hegemony requires deconstruction: without
the radical structural undecidability that the deconstructive intervention
brings about, many strata of social relations appear as essentially linked by
necessary logics and there would be nothing to hegemonise. But
deconstruction also requires hegemony, that is, a theory of the decision
taken in an undecidable terrain: without a theory of decision, that distance
between structural undecidability and actuality would remain untheorised.
(Laclau, 1996b, 59-60)
Logics
In recent years Laclau has conducted a number of deconstructions of different

important political conceptual couples. Some of the logics that result from
these are described below.
Logic of signification
We have already touched on the logic of signification. It is important to note

that a number of Laclau’s deconstructions in relation to the logics discussed
below are simply applications of the logic of signification on the relationships
of specific signifiers to their specific signified. Therefore, once one has a better
understanding of the logic of signification, it is possible to explore much
further.
Logic of representation
The designation of the logic of representation happens as a result of a

deconstruction of the difference representative/represented (Laclau, 1993b).
It is an assumption of prevailing political theory that the represented side of
the difference is the dominant side. Representation only makes sense if it
actually represents something. Furthermore, prevailing political theory discusses
how representation can be distorted in relation to what it represents, which
leads to a democracy-theoretical discussion about which institutions ensure
the most accurate representation. Deconstruction leads to a reversal of the
difference and to a displacement of the theoretical question of representation.
Not only does the representative distort the represented, it also indispensably
partakes in constructing the represented and can therefore never stay neutral
in relation to it: “The relation representative–represented has to be privileged
59
as the very condition of a democratic participation and mobilization” (Laclau,

1996b, p 49). This opens up the question of how the struggle of representation
is always a struggle over the construction of the represented.
Logic of tolerance
The logic of tolerance indicates the mechanism that is present in the difference
tolerance/intolerance. Here, too, the difference is reversed through
deconstruction which shows that tolerance cannot be explained in itself without
turning into its opposite. Consequently, intolerance is simultaneously the
condition of the possibility and the impossibility of tolerance; the undecidability
of the distinction extends the possibility of both poles. With the logic of
tolerance we are able to study the hegemonic battle over what should be
tolerated and what should not be tolerated (Laclau, 1996b, pp 50-2).
Logic of power
The logic of power indicates the mechanism present in the difference power/
liberation. Once again, the difference is reversed. Jurisprudence assumes that
freedom is the condition of power; power is the restriction of freedom and
therefore earns its potential through freedom. Laclau reverses the difference
and points out that what restricts freedom is also what makes it possible – a
society free from power is an impossibility. Once again, this provides the basis
for a discourse-analytical study of the continued negotiation and displacement
of the power/freedom boundary (Laclau, 1993b, 1996b, pp 52-3).
Logic of equivalence; logic of difference
Moreover, Laclau configures what he calls logic of equivalence and logic of

difference. It is not clear whether these logics are used in the same way as in
the above examples. If so, they could be understood thus. Laclau chooses as
his point of departure a deconstruction of the very difference difference/
equivalence in order to show how differentiation holds implications in relation
to the articulation of equivalence and, conversely, how, at the same time,
equivalence affects differentiation. The two logics are central in Laclau’s
discourse analysis.
The logic of equivalence is the logic of simplification of the political sphere.
Through the articulation of equivalence between elements, the possibility of
an interchangeability of elements is increased. At the same time, the number
of subject positions is reduced. For example, by making almost everything
equivalent to social economy – from tax issues and the environment to equality
– more and more elements can be included and interchanged within the same
chain of values. Meanwhile, the number of positions through which one is
able to participate in the conflict over taxes, the environment and equality is
reduced.
60
Conversely, the logic of difference signifies the logic through which the
political sphere is widened and increases its internal complexity. The elements
do not become particularly interchangeable, but the number of subject positions
(that is, the positions from which one can be political) is increased (Laclau and
Mouffe, 1985, p 130).
Logic of universalisation
Finally, I wish to mention the logic of universalisation, which is based on the

deconstruction of the difference particular/universal (Laclau, 1996b, p 59).
Precisely because the universal is universal it ought to be able to stand alone.
Deconstruction shows, however, that this is not possible: “The conclusion
seems to be that the universality is incommensurable with any particularity
yet cannot exist apart from the particular” (Laclau, 1992, p 90). This forms the
basis for studies of how attempts are made to universalise the particular, and
how universalisation defines what can be articulated as particular.
Conclusion
Laclau also configures a few other logics that are usually presented as aspects
of almost all logics, including the logic of supplement and the logic of impossibility.
The concept of logic does not therefore appear to be exercised completely
consistently. The logic of supplement and the logic of impossibility involve a
metalogic that might also apply to the logic of difference and the logic
equivalence, but not to the logic of representation, which, one must assume, is
bound by a particular historicity. We can therefore observe a division of
duties between deconstruction and discourse analysis, but also a lack of
theorisation of their mutual relationships, which appears as an inconsequent
and insufficiently defined concept of logic. This raises a number of questions,
particularly about universality:
• Are some logics more universal than others?

• And if they are not universal, what is their field of validity?
• What is the relationship between history and logic?
• What is the theory behind the definition of the central dualities in
deconstruction?
• Is there a hierarchy of logics with, for example, the logic of universality
being the most universal?
• If so, can the hierarchy of logics itself be deconstructed?
However, these questions do not change the fact that, analytically, the linking
of deconstruction and discourse analysis is exceedingly powerful.
61
Conclusion
Laclau’s two analytical strategies and the way they pose questions have been
summarised in Table 3.1. In the right-hand column I have attempted to
provide an example of a possible concrete analysis. The example used in this
table is the issue of animal ethics that has taken place in most EU countries, in
particular in the UK. A deconstructive analysis could address the discursive
infinity of the debate about animal ethics. Ethics always claims to be universal,
to be unavoidable, and to not let itself be deflected by pragmatic and specific
circumstances. Ethics are only ethics if they raise themselves above concrete
circumstances. A deconstructive analytical strategy, therefore, could address
the duality of universal/particular in order to demonstrate that any claim for
universality is always particular, and that the two sides of the duality must
incessantly pollute and contaminate one another. Deconstruction could point
to the conditions of impossibility of animal ethics. Subsequently, a hegemonic
analysis would be able to analyse how the conditions of impossibility of ethics
are unfolded in specific discursive battles about trying nevertheless to fixate
animal ethics. Hegemonic analysis could point out different discursive attempts
to fixate and how different attempts must exclude something from their
discoursing in order to form a coherent discourse. The more militant discourses
define production and financial prioritising as antagonistic opposites to ethics.
These discourses would not be able to maintain their strong sense of unity
without rejecting any kind of pragmatism, which, in turn, prevents them from
ever gaining any form of hegemony. By contrast, production and economically
oriented discourses of agriculture might involve ethics, but on behalf of the
purity of ethics as precisely universal ethics. Ethics become temporalised. It
is something that should be achieved in the long term and a consideration
that, because of its temporal distance, must be weighed against other
considerations in the present, such as the economy and production.
Table 3.1: Laclau’s analytical strategies

Deconstruction Which infinite logic In what way does any discourse

is installed with a about animal ethics rely on the
specific duality? duality of universal/particular?
Which infinite logic does any animal
ethics therefore have to unfold?
Hegemonic How are discourses In what way is the signifier ‘animal
analysis established through never- ethics’ sought, given value and fixated
concluded battles about in conflicts between different
the fixation of floating discourses with different strategies
elements of signification? of equivalency and difference?
Note
1
In the article ‘Discourse’, Laclau even distinguishes between two fundamental
notions of discourse analysis – his own poststructuralist discourse analysis and
Foucault’s second-order phenomenology (Laclau, 1993b).
62
4
Niklas Luhmann’s systems theory
f the writers included in this book, Niklas Luhmann is unquestionably,
O and in many ways, the most comprehensive. He has written more than
50 books and over 400 articles. One of his goals has been to test
the general stating power of systems theory by writing at least one book
about each of the function systems of modern society. Hence, he has conducted
extensive historical analyses of the scientific system, the political system, the
system of arts, the educational system, the system of justice, the system of
regions, love and family as a system, the system of mass media, and the system
of economics. But his work is also the most comprehensive in another respect.
Of the authors in this volume, Luhmann’s work probably involves the highest
number of different theoretical questions, provides the widest scope of
theoretical approaches and possesses the highest degree of flexibility for the
elaboration of projects within theory. However, in this book, we are not
concerned with the many analyses of different social systems but only with
the Luhmanian ‘eye’ for social systems as communication.
Luhmann’s path towards the ‘communication-theoretical turn’ is completely
different from that of Laclau and Foucault. The communication-theoretical
turn did not appear until around 1980. Prior to this, Luhmann had suggested
that meaning was the fundamental concept of sociology (Luhmann, 1990b,
pp 21-79), but it was not until the beginning of the 1980s that he changed his
notion of social systems as systems of action to an understanding of these
systems as autopoietic systems of communication, consisting in and by
communication. To a certain extent, Luhmann’s systems theory can be viewed
as a communication-theoretical rewriting of the systems theory of Talcot
Parsons. In Parsons’ work, the communication-theoretical inspiration stems
not from French structuralism, but from a unique patchwork of the calculus
of form from the mathematician Spencer-Brown, the theory of life as a system
of autopoiesis from the biologist Maturana, the theory of information as
differences that make a difference from the autodidact Bateson, and the theory
of meaning from the phenomenologist Husserl.
Luhmann’s theoretical architecture has often been compared to a labyrinth.
This is not only because it is difficult to get in and even more difficult to get
out, but particularly because it is not construed deductively. It is inductive in
a very particular way and always allows for the observer to step back and ask:
‘Why this particular concept?’, through which all concepts suddenly appear
in a new light. I will return to this point later in this chapter.
63
As there are so many different points of entry to the theory, it is by no

means a given how to introduce Luhmann – there is no obvious first concept.
Moreover, it appears that the concepts change ‘colour’ depending on the concept
one chooses to begin with. Already, there are several conceptual dictionaries
of Luhmann’s systems theory (for example, Krause, 1996), one of which suggests
specific routes and specific chronologies, by which one can approach Luhmann’s
concepts. In one sense, each route establishes its own systems theory. Here, I
have chosen to begin with the concept of observation, which means that I
prioritise the Spencer-Brownian Luhmann. Two alternatives are the Parsonian
Luhmann (where the relation of system/evolution is central) and the
Maturanian Luhmann (where the central concept is that of autopoiesis).
Systems theory as second-order observation

By choosing to define the concept of observation as a starting point, systems
theory is thus defined as a theory about second-order observation. The concept
of observation immediately doubles as, on the one hand, a general concept of
observation and, on the other hand, as concrete and specific observations,
which we, as systems theorists, can observe as observations precisely by means
of a general concept of observation.
Form and difference

Luhmann’s theory of observation is founded on a particular definition of
form and difference, particularly inspired by Spencer-Brown (1969), von
Foerster (1981) and Günther (1976). The basic premise is to view observations
as operations that do not refer to conscious subjects but to differences. Luhmann
simply defines observation as a specific operation of creating distinctions: to
observe is to indicate something within the boundaries of a distinction. In the words
of Spencer-Brown: “We take as given the idea of distinction and the idea of
indication, and that we cannot make an indication without drawing a
distinction” (Spencer-Brown, 1969, p 1).
For example...
We might, for example, fasten upon something artistic. ‘Art’ is then indicated.
But art can only be indicated within the boundaries of a distinction. The opposite
side could be ugliness, unsightliness, disharmony and so on, and this other side of
the distinction makes a difference to the way art appears as an object to the
observer. Therefore, what we observe is above all dependent on the distinction
that defines the framework for what is indicated in the world at large.
Spencer-Brown indicates the difference as shown in Figure 4.1.
64
Figure 4.1: The sign of difference
Indication/distinction
Like anything else, the concept of observation must be grounded in a distinction.

In Luhmann, it is the distinction indication/distinction. More precisely:
observation is the unity of the distinction indication/distinction. To demarcate
is to indicate or even name something in the world and, as previously noted,
that demarcation always takes place as an indication of one side of a distinction.
In Luhmann’s universe, distinctions are always two-sided. A distinction possesses
an inner side as well as an outer side; the inner side being the indicated side.
The inner side is designated ‘the marked space’, while the outer side is designated
‘the unmarked space’. Whenever there is observation, one side in the difference
is marked and the other remains unmarked.
Such a distinction isolates the marked from the unmarked and it is therefore
only possible for one side of the distinction to be indicated at a time. If both
sides are marked, the distinction is cancelled out. It is not possible
simultaneously to observe an object as beautiful and ugly – except over time
– but then it is no longer the same observation. Hence, distinctions are always
asymmetrical because only one side is marked (see Figure 4.2).
Through the observation of something, the blind spot of observation is
shaped. The blind spot is the unity of the distinction that constitutes the
framework for the observation. In Luhmann’s theory, the unity of the distinction
is defined as form. The blind spot of observation exists because of the fact that
observation cannot see that it cannot see that which it cannot see.
Figure 4.2: The marked difference
Second-order observations
We are now closing in on systems theory. Luhmann attempts to establish his

systems theory precisely as second-order observation. If first-order observation
is the indication of something within a distinction, second-order observation
is observation directed at first-order observation and its blind spot. Systems
theory thus inquires about the blind spots of society and of the systems of
society, about the distinctions that fundamentally decide what can appear in
society and how.
65
Obviously, all second-order observations are also, at the same time, first-
order observations, since they indicate first-order observation within a
distinction. There exists, therefore, no privileged position for observation.
This is illustrated in Figure 4.3.
Figure 4.3: A difference observed through a difference
{ m
Reference
An observation does not merely indicate within a distinction. The operation

of observation furthermore establishes a distinction between self-reference and
external reference, between what is observed and the observing system. The
observing system comes into being through observation, together with what is
observed. The system does not precede the world it observes: both are
constructed through the observation that separates them. The system is thus
not constituent of the observation. On the contrary, it is the observation and
the distinction actualised by it that decides how the world appears to which
system. In this sense, reality is always systems relative and the operation of
observation divides the world into system and environment. Thus, systems
theory is based on three distinctions: indication/distinction, system/
environment and first-/second-order observation.
First- and second-order observations
Let me clarify the conditions of first- and second-order observations. First-

order observation is the observation by a system of something in the environment.
First-order observations thus use external reference: they refer to the
environment. Second-order observations, however, are observations of the
observing system itself – not in any way but precisely as observer. Second-
order observations are thus self-referential.
66
Distinctions
Observations of the first as well as the second order are observations within a
distinction, but not all ways of distinguishing allow for second-order
observation. Luhmann distinguishes between three ways of making a
distinction:
• First, by distinguishing something from something else without specifying

the other side of the distinction (for example, horse/not horse). What
occurs in this distinguishing operation Luhmann simply refers to as object
(Luhmann, 1993a, p 15).
• Second, by indicating in a way that restricts the other side of the distinction
(for example, plus/minus, man/woman, warm/cold, Danish/foreigner). This
type of distinction is referred to as concept and always implies a counter-
concept (Luhmann, 1993a, p 16).
• Third, a particular variant of concepts, in which distinctions are made by
copying a conceptual distinction and re-entering it into the inside or outside
of the concept itself, to thereby indicate certain aspects of the concept.
These are therefore concepts capable of conceptualising themselves. Such
concepts are referred to by Luhmann as second-order concepts. Concepts of
the second order are thus restrictive distinctions, which can be re-entered
into or re-enter themselves. For example, the distinction of government/
opposition can be re-entered into itself in the sense that both government
and opposition are able to conceive of itself as having a government (a
deciding fraction) and an opposition (an opposing minority).
A system is only able to observe itself if it indicates itself within the framework
of a second-order concept. If a system is to observe itself as observer, the
system is required to divide itself in two: the observer and the observed. Hence,
Luhmann proposes that a system, constituted as the unity of the distinction
system/environment, is only able to observe itself as observer if it can copy its
guiding distinction and re-enter it into itself, that is, its ability to divide itself
by entering the distinction system/environment into the system itself (Luhmann,
1995a, pp 37-55). This process of re-entry is illustrated in Figure 4.4.
Figure 4.4: The distinction system/environment re-entered as a part of

itself
system/environment
re-entry copying
system/environment
67
In the mathematical tradition following Spencer-Brown, such re-entry is

indicated as shown in Figure 4.5.
Figure 4.5: The sign of re-entry
Source: Kaufmann, 1987, pp 53-72
Paradox
This way of thinking has a number of implications. On the level of first-order

observation, observations are paradoxical, since the observer must distinguish
without being able to choose her distinction. An observer of the first order
cannot see the distinction on which her observation is based, and yet she can
make distinctions. That is a paradox. Conversely, the observer of the second
order sees that the observing observer can only see that which his distinction
lets him see. He is thus able to see how the first-order paradox is removed,
becomes invisible, or, in Luhmannian terms, is de-paradoxified. An observer
observing an observer is not, however, able to simultaneously observe himself
as observer and therefore has his own blind spot. This displaces the paradox of
observation, but at the same time the nature of the paradox changes to ‘a
paradox of re-entry’. The paradox on the level of the second order is the fact
that the distinction system/environment is, at the same time the same and not
the same once the sub-system has separated itself (by copying the distinction
system/environment and re-entering it into the system) in order to observe
the system as observer (Luhmann, 1993b, pp 763-82). The re-entry implies
that a part of the system obtains a higher reflexive capacity than the rest of it.
Perspective
Even though an observer of the second order is simultaneously an observer of
the first order, the outlook is nevertheless different:
• On the level of the first order, the outlook is mono-contextual. The observer
sees what he sees. He makes use of a distinction without being able to
distinguish.
• On the level of the second order, the outlook is poly-contextual. Although
the observer still observes within the framework of a distinction, the observer
of observers knows that she cannot see that which she cannot see. She
knows that reality depends on the observer – that the observed is contingent
with the difference that defines the boundaries of the observation. She
uses a distinction but she is also able to distinguish.
68
• The observer of observers as observers is able to see that his observations of

observations do not lead to random results and that the choice of a particular
distinction has implications.
In other words, the theory of observation holds autological implications. The

theory’s statements about observation are true of the theory itself. In the
words of Elena Esposito:
[Luhmann’s] approach is not simple self-referential but auto-logic, where

with ontology one indicates the condition by which the knowing systems
is itself one of the objects it has to know: when it describes its object it
then also describes itself, and the description modifies the object to be
described. (Esposito, 1996)
Luhmann speaks of a radical constructivism that includes itself (Luhmann,

1986, pp 129-34, 1990c).
Conducting systems-theoretical analysis

This requires of the systems-theoretical observer to meticulously substantiate
and account for the way she constructs and thus observes her reality of
observations (Luhmann, 1988a). As a minimum, a systems theorist must do
the following.
Account for and substantiate his choice of guiding distinction
Guiding distinction is understood as that distinction which is defined as the

framework for observation of observations. It has great implications for viewing
observing observers through the distinction system/environment. The moment
one defines the distinction system/environment as the basis of second-order
observation, reality is always constructed as either system or environment. We
will then be viewing a system’s observations as either referring to the system
itself or to the environment of the system. We are encouraged to always see
how the observing system divides the world into the system itself and its
environment when it observes, that is, how the observing system comes into
being through its distinction between system and environment. It is not
presupposed that a second-order observer observes with the distinction system/
environment. In principle, any difference with the possibility of re-entry into
itself can exist as the framework for observations of the second order, and any
of these differences hold implications for the construction of reality.
Account for the conditioning of the chosen guiding distinction
In this context, conditioning means the definition of conditions of indication.

If, for example, we take as our starting point the guiding distinction system/
environment, conditioning means the designation of the conditions of when
69
that which we observe is accepted as a system or the environment of a system

respectively. One condition could be that the system itself is able to distinguish
between itself and its environment. To reiterate, the systems of the world do
no let us know how they wish to be observed; only we can be held accountable
for our observations through the explication of the conditions of observation.
Point out, substantiate and account for the implications of the exact
observation point
Operating as an observer of the second order with distinctions that can re-
enter themselves, it is not self-evident what is defined as object. Even when
observing observers through, for example, the distinction system/environment,
it remains for the observer to decide which system is defined as the observation
point, since all systems simultaneously constitute the environment of other
systems. The moment one system has been selected as the observation point,
all other systems can only be seen as environment, and only then to the extent
that they are constructed as such by the particular system. In the words of
Luhmann, any theory about observed systems must designate its systems reference:
which it uses as its starting point in order to designate which system which
things exist as the environment of. [...] However, if one wants to know the
systems reference upon which an observer is based [...] one must observe
the observer. The world does not disclose the way it wants it. With the
choice of systems reference, one has simultaneously designated the system
which draws its own boundaries and thus divides the world into system and
environment. (Luhmann, 1995b, p 46)
For example...
We might be interested in observing observations on national financial

management. If we define the Agency of the Environment as our observation
point, we will be able to observe, for one thing, the way financial management
emerges as an organisational programme for economic decisions within the
Agency of the Environment. We will also be able to see how the Agency of the
Environment constructs itself as ‘environment’ in relation to the state, in a
distinction between economic factors under its control and economic factors
defined by the environment. However, by defining the political system as the
observation point, another distinction becomes central – namely politics/
administration – and financial management becomes a political battlefield for the
control of an administration (including the Agency of the Environment) that is
viewed as increasingly ungovernable. Hence, the observation point constitutes
what we see and which questions emerge.
70
This demands self-restriction and precision of gaze, which is implied in

Luhmann’s systems theory, because of theory about observation as the unity
of indication and distinction nevertheless allows for a particular analytical
flexibility.
The observer of the second order can always ask:
• Why this particular guiding distinction?

• Why not a different distinction, which could lead to the replacement of
the chosen guiding distinction by a different distinction capable of re-
entry?
In systems theory there is no concept of essence forcing a second-order observer

to view reality in a particular way. Reality, as such, is not observable anyway
and does not therefore demand anything specific from the observer. Reality is
observer-dependent and in that respect it is a construction and also a reality.
The observer of the second order can also always inquire about the
conditioning of the guiding distinction:
• Is it possible to condition the particular guiding distinction differently and

more productively?
For example, in his earlier writings Luhmann conditioned social systems as

systems of action, but later found it more productive to condition social systems
as systems of communication.
Finally, there is always the option of moving the observation point. It is
always possible to move backwards in the process of re-entry so that that
which was seen as system now appears as sub-system, or move forward in the
process of re-entry so that that which was seen as system now appears as
environment. This applies, not only to the distinction system/environment,
but to all concepts capable of re-entry (that is, all second-order concepts).
I have tried to summarise these notions of observation and observation of
the second order in Table 4.1:
Table 4.1: First- versus second-order observation

Observation of the Observation of the
first order second order
Observation Observations of something Observation of observations as
in the world through indication observations, that is, observations of the
within the framework of a boundaries of indication in other
distinction observations within the framework of
the same distinction
Reference External reference Self-reference
Distinction Object: this/something else Second-order concepts: concept
Concepts: concept/counter- that can appear as part of their own
concept whole, that is, concepts capable of re-entry
Outlook Mono-contextual Poly-contextual
71
Conclusion
Having used a considerable amount of space on Luhmann’s concept of

observation, I will continue with the rest of this chapter. First, some of the
central conditionings of the system/environment distinction that follow the
concepts of meaning, consciousness and communication are discussed. Then,
some of the alternative guiding distinctions in Luhmann’s systems theory are
introduced, which exist in addition to the system/environment distinction.
The concept of meaning

Luhmann’s theory of observation avoids any kind of anthropocentrism. Not
only people observe: the observing system could be anything from a thermostat
in a refrigerator, or an organic cell to a national bureaucracy. With the theory
of observation, therefore, we have only just begun. Thus, we must move on to
specify which type of system we are fundamentally interested in. What is our
systems reference? Although an organic cell is as capable of observation as a
bureaucracy, their systemic conditions of observation are obviously very
different. Luhmann’s general systems theory assumes that all systems are
autopoietic, meaning that they themselves create the elements they consist of,
including the constitutive boundary between system and environment. On
this general level, systems can be compared to the basic element of autopoiesis.
Luhmann distinguishes between organic systems, psychic systems and social
systems – organic systems create themselves through life, whereas psychic and
social systems create themselves through meaning.
Meaning
When focusing on social systems, meaning thus becomes the first inevitable
concept. In part, the concept of meaning plays the same part in Luhmann’s
systems theory as the concept of discoursivity plays in Laclau’s discourse theory:
• While all social identities in Laclau are embedded in a discoursivity and are
unable to go beyond this discoursivity, psychic and social systems in
Luhmann are similarly unable to operate outside of meaning.
• In the same way that discoursivity in Laclau is constituted by differences
and relations, meaning in Luhmann is also a concept of difference.
• Finally, while discoursivity in Laclau is characterised by floating and unfixable
relationships between discursive elements, meaning in Luhmann is similarly
unfixable, and is always unstable and indefinable.
However, having defined these similarities, the differences become clear. In

fact, the lines of reasoning behind the indicated similarities are very different.
Although both theorists relate meaning to difference, this conclusion is
based on different approaches. In Laclau, meaning is not founded on an
external referential relationship, either in capacity of an external reality or as
72
sign structure. Meaning is neither a structuralist nor a poststructuralist concept.

The main inspiration behind Luhmann’s concept of meaning stems from
Husserl’s phenomenology:“The best way to approach the meaning of meaning
might well be the phenomenological method. This is by no means equivalent
to taking a subjective or even psychological stance. On the contrary,
phenomenology means: taking the world as it appears without asking
ontological or metaphysical questions” (Luhmann, 1985, p 101).
Luhmann defines meaning simply as the unity of the distinction actuality/
potentiality (Luhmann, 1995c, p 65). At the particular moment that something
appears central to the thought or to the communication, something is actualised,
but this always happens in relation to a horizon of possible actualisations (that
is, potentiality). There is always a predefined core surrounded by references to
other possibilities that cannot be used simultaneously. It is important not to
understand potentiality or possibility as structures that precede actualisation,
but as a horizon that lines up with actualisation – something appears and
thereby excludes other possibilities. Thus, meaning is the simultaneous
presentation of actuality and potentiality. Meaning is the actual surrounded
by possibilities; any actualisation of the moment potentialises new possibilities.
For example...
I decide to write a chapter on Luhmann and, simultaneously, a horizon arises of

alternative constructivists who could have been the object of my interest: Bourdieu,
Latour, Castoriades and so on. I choose to begin with the theory of observation
and immediately a horizon of other points of departures emerges: for example,
the concept of autopoeisis or the concept of society. As soon as something is
actualised, meaning can be extended by connecting, in a new selection of meaning,
to the actualised side or by crossing across the difference and actualising something
that was formerly potential. For example, I can decide that it was the wrong
decision to begin with the concept of observation and that the concept of society
would be the better choice. Consequently, the actual and the potential cannot be
separated but can only exist in simultaneous relationships with each other. In
Luhmann’s words: “Meaning is the link between the actual and the possible: it is
not one or the other” (Luhmann, 1985, p 102).
Meaning and psychic and social systems
Psychic and social systems are tied to meaning. They possess no relationship
with the environment, let alone themselves, except through meaning. Meaning
is the shape of the world and overrules the distinction between system and
environment. On the other hand, there are no limits to meaning as such – in
principle, anything can be understood by psychic and social systems but only
within the shape of meaning. Meaning can never extend itself into something
else (Luhmann, 1995c, pp 59-63). Hence, meaning is open-ended in the sense
73
that everything can be understood through meaning, but it is closed and self-
referential in the sense that meaning refers only to meaning.
Like discursive identity in Laclau, meaning can never be fixed. However,
Luhmann’s argument for this is not that a structure is not complete but that
the core of the actualised disintegrates from the moment something has been
indicated. Meaning can never be fixed or maintained; it is fundamentally
unstable. This is in part due to the fact that meaning is always shaped by a
thought or by communication, which disappears the very moment it occurs.
Meaning is always reproduced (or changed) recursively, like decisions within
an organisation.
For example...
It is always decided at the following meeting, when the minutes are being approved,
which decisions are not to be considered decisions after all but simply talk, and
which decisions are to be included in the minutes and should therefore be followed
up and put on the agenda in terms of new decisions. However, the decisions
included in the minutes can never be identical to the decisions as they emerged at
the previous meeting. They now appear in a new situation, in a new specification
in relation to a new horizon of possibilities: for example, the possibility of the
minutes not being approved, or of the decisions not really being decided upon.
The core of actuality thus disintegrates from the moment it appears and,
consequently, meaning provokes change.
Meaning, therefore, is also a continuing rearranging of the difference between

actuality and possibility (Luhmann, 1995c, pp 63-6).
This implies, moreover, that social and psychic systems are dynamic systems
that never stay identical to themselves; they always exist in a movement of
becoming. In Luhmann’s words, the actual is secure but unstable. The potential,
in turn, is stable but insecure. The option of stable security does not exist
(Luhmann, 1985, p 102).
The concept of communication

Meaning is constituted by psychic as well as social systems, and is a collective
medium for the two systems. Psychic systems operate in terms of meaning in
the shape of a closed connection of consciousness. Social systems operate in
terms of meaning in the shape of a closed connection of communication.
The distinction between psychic and social systems is crucial to Luhmann’s
theory of society as communication. Any common-sense perspective would
maintain that communication must necessarily be between people, but also
that only people are able to communicate. Contrary to this view, Luhmann
argues:
74
• First, within a systems perspective, people cannot even be indicated as a

unit, they are, in the words of Kneer and Nassehi,“a plurality of independent
systems – eg, the organic system, the immune system, the neurophysiological
system, and the psychic system all of which operate without any overlap
between them whatsoever” (Kneer and Nassehi, 1993, p 70).
• Second, Luhmann argues that, fundamentally speaking, people cannot
communicate at all, not even in their capacity of psychic systems.
Communication alone is able to communicate (Luhmann, 1996, p 261).
Autopoietic systems
The background for this view is Luhmann’s understanding of systems as

autopoietic systems. The fact that a system is autopoietic means that it invents
itself and everything that it consists of:
Everything that is used as a unit by the system is produced 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,
then, are sovereign with respect to the constitution of identities and differences.
They, of course, do not create a material world of their own. They presuppose
other levels of reality, as for example human life presupposes the small span
of temperature in which water is liquid. But whatever they use as identity
and differences is of their own making. (Luhmann, 1990b, p 3)
The strength of the concept of autopoiesis lies in the possibility of an

unambiguous perspective on society and its social systems as an independent
thing that cannot be reduced to something other than itself, for example to
consciousness or a sum of actions. Again, we see the phenomenological
insistence on observing society as it appears without reference to conditions
external to society.
Psychic systems belong to the environment of social systems. They are
autopoietic systems that produce meaning in a closed connection of
consciousness by the process of thoughts connecting with other thoughts. A
psychic system can never extend itself beyond itself. It is unable to connect
with the thoughts of a different psychic system. Hence, psychic systems are
unable to ever obtain a mutual understanding, let alone a fusion of their
meaning horizons, partial or complete. In this sense, it is not possible for
psychic systems to communicate with each other.
Conversely, social systems also belong to the environment of psychic systems.
Social systems are autopoietic and produce meaning in a closed connection of
communication by connecting communication with other communication.
Likewise, social systems are unable to extend themselves beyond themselves.
Communication cannot connect with thoughts but only with communication.
No consciousness, therefore, can be completely identical to communication,
and no communication can be completely identical to consciousness.
75
Nevertheless, social and psychic systems are structurally linked: first, they
both shape meaning as a medium (but in different forms); second, they make
their own complexity available to each other. Although psychic systems are
unable to communicate with each other, communication is also unable to
communicate unless at least two psychic systems partake in the communication.
Social systems, in turn, make their complexity available to psychic systems,
primarily through language and through disappointment/fulfilment of the
expectations of psychic systems.
Information, form and understanding

To Luhmann, communication is understood not as acts of communication
(since he would then be reducing social systems to human acts) nor as the
transfer of messages between sender and receiver (not even in the more
reception-oriented version). Rather, the notion of communication in Luhmann
is to be seen as a kind of coordinated selection. Communication is a selection
process, consisting of a synthesis of three selections:
1. Selection of information, that is, what is to be communicated.

2. Selection of form of message, that is, how the information is to be
communicated.
3. Selection of understanding, that is, what should be understood about the
message (Luhmann, 1995c, pp 137-176).
Communication does not become communication until all three selection

processes have occurred. Monologues, for example, are not communication –
the message needs to be understood before it can be regarded as communication.
In this context, it is important to stress that the concept of understanding does
not refer to the reception of the message by a psychic system but instead to
the linking up to the message by subsequent communication.
Communication
Any message establishes a horizon of possible links to new communication.

Subsequent communication links up to the message by actualising one of these
possibilities and by leaving the others as a horizon of potentiality. Thus,
understanding is nothing other than the retrospective choice of connection by
subsequent communication.
Therefore, it requires at least two communications in order for communication

to exist.
It should be noted that the three selections noted above all shape meaning as
a medium. The selection of information shapes the distinction of actualised
information/possible information; the selection of message shapes the
76
distinction of actualised message/possible messages; and the selection of

understanding shapes the distinction of actualised connection/possible
connections.
Also, it should be noted that communication is self-referential.
Communication connects with previous communication and, in this sense,
social systems arise in the recursivity of communication.
Preliminary summary
As we have seen, Luhmann’s systems theory is a theory about society as
communication. It assumes that society consists of autopoietic systems of
communication, which produce and reproduce themselves through
communication alone. The different systems of communication are closed
around themselves; they are self-referring. And this closure constitutes the
condition of their openness, that is, of their sensitivity to their environment.
All communication is observation, either in the shape of self-description or
external descriptions. The closure of systems of communication both in regard
to meaning and to communication implies that systems of communication
always observe from a blind spot.
Systems theory attempts to establish itself as observation of the second order
– as an observation of how systems of communication observe. It is about the
description of the limits and blind spots (in the broadest sense) of
communication. It is not only a question of describing how a system of
communication sees something and not something else. It is also not only a
question of localising the moment when silence takes over. But it is concerned
with the relationship between the different systems of communication, the
way they are connected, the way they interrupt each other productively and
destructively, including the existing limits, not only of the individual system of
communication, but also of the overall communication structure of society.
As a second-order observer, systems theory must recognise that it is itself
communication within society. In that sense, systems theory is about
contributing to the self-description of society but, as part of society, systems
theory is faced with a classical sociological problem: self-reference. Within
systems theory, autology provides the solution to this problem of self-reference.
Through the concept of autology, systems theory can describe itself as a form
of self-description describing itself. An autological sociological description of
society in society thus describes the description of society by adapting it to
itself. This means, for one thing, that systems theory explores its own origins,
but it also means that it has to define explicitly those guiding differences on
which it bases its second-order observations, and account for the way they
construct the object of systems theory.
The next section presents some of the central guiding distinctions to
observation of the second order in Luhmann’s systems theory.
77
Form analysis
When taking the concept of observation as the point of departure, it is important
to grant priority to Luhmann’s form analysis. It appears that all choices and
accounts of choices in Luhmann begin with a form analysis of the guiding
distinction itself. The role of form analysis in relation to systems analysis in a
broader sense is highly comparable to the role of deconstruction in the analysis
of hegemony in Laclau.
While deconstruction in Laclau focuses on specific central dualities around
which a discursive game is played out, form analysis similarly focuses on specific
distinctions in connection to which communication plays itself out. The
deconstruction of central dualities produces particular logicities, which
necessarily emerge in certain discourses. Likewise, form analysis analyses the
boundaries of communication and the paradoxes that communication unfolds
when it connects with one particular distinction.
Luhmann defines form as the unity of a difference. This brings us back to
the theory of observation but now we have specified the systems reference to
social, meaning-shaping systems of communication.
For example...
We observe the observations of an observer and note that the observer focuses
on, for example, the risks of nuclear power. We might also note that other observers
focus on different risks – the risk of prenatal diagnosis, of share investments, of
work and so on. We notice a certain recursivity in the communication, when
questions of risk appear. We now need to ask whether a particular form of
communication appears to define itself within societal communication, which
we can call ‘risk communication’.
• If this communication is characterised by its ability to indicate risks, then

what are the conditions of its existence?
If we pose this question in terms of form analysis, the objective is at first to localise
the difference, which defines the scope of the indication of risks.
• Which difference allows an observer to see the environment in terms of

risks?
The question becomes more specific:
• Which is the unmarked side of the difference when communication indicates

risk?
Consequently, form analysis is the analysis of the conditions of communication

given a specific difference, the guiding distinction of form analysis being unity/
difference.
78
Luhmann himself uses the example of risk (Luhmann, 1990d) and asks if the
other side of the difference might be security. Security/risk is a distinction
that is often related to risk communication; however, Luhmann rejects this
proposal, since security cannot be conditioned and cannot therefore operate
as counter-concept to risk. In other words, it is not able to describe the blind
spot of risk communication. Instead, Luhmann suggests ‘danger’ is the other
side of risk. Risk communication is subsequently defined as communication
that connects with the distinction risk/danger. If communication connecting
with risk shapes the distinction risk/danger, then risk is always only risk in
relation to a danger. Risk and danger cannot be separated but only exist in
relation to each other. Now the distinction has been identified. The subsequent
form analysis asks which kind of communication can develop within the
scope of this distinction.
• What must necessarily follow the shaping of this very distinction?

• What are the restrictions on communication due to this distinction?
It is precisely the question of the unity of the distinction. Luhmann’s answer

to this question in relation to risk communication is too far-reaching to be
included here, but he demonstrates, among other things, how communication,
which produces decisions by ascribing risk to the decision itself, simultaneously
assigns an unspecified danger to others.
For example...
One driver’s decision to assign himself the risk of passing another car on a curve
at high speed on the assumption that he can just make it simultaneously implies
the assignment of danger to others. Danger exists in being exposed to the risk
decisions of others (Luhmann, 1990d).
In accordance with Derrida and deconstruction, one might say that form
analysis points to a particular ‘logicity’ given a specific guiding difference in
the communication. Where deconstruction tries to convert the hierarchy of
differences, form analysis begins by inquiring about the other side of the side
indicated. Form analysis tries to locate the distinction that constitutes the
framework of a particular observation. Subsequently, form analysis inquires
about the unity of the distinction and hence about the communication that
the distinction both enables and excludes. Luhmann has conducted a great
number of form analyses, for example on the form of writing, the form of the
distinction system/environment, the form of knowledge and the form of
causality.
It is important to note that form analysis rarely represents the end of the
involvement with a particular problem but often represents the beginning of
the formulation of a problem, which is then followed up in other analytics, for
example in a systems and differentiation analysis. Luhmann himself defines
79
the difference between form analysis and deconstruction by saying that, where
deconstruction often contents itself with uncovering the logicity and
paradoxical basis of a duality, form analysis simply constitutes the incipient
inquiry about the unfolding of a paradox within communication. Given the
fact that, for example, the concept of risk is paradoxical by nature (since
communication on risk is itself risky, whoever runs a risk and inflicts danger
on others runs the risk of becoming visible), how and within which systems
of communication is this paradox handled? In this respect, we can trace a
certain parallel to Laclau to whom deconstruction is merely the beginning of
an analysis of hegemony.
Systems analysis
The most central guiding distinction in Luhmann is the distinction system/
environment. It is not only the difference most often used by Luhmann in his
second-order observations, it is also the difference he uses to regulate the use
of other guiding distinctions and, more generally, to steer his development of
theories.
In systems analysis, the basic notion is that all communication takes place
within a social system, and that all social systems are constituted by a boundary
of system and environment. As indicated in relation to the theory of observation,
it is the very observation that divides the world into system and environment.
The same pertains to social systems in which observations consist in
communicative descriptions. A social system can observe itself or the
environment through descriptions of self-reference or external reference. One
example of self-referential communication in an organisation is financial
accounting in the same way that surveys of waste sites are examples of
communication of external reference in environmental administrations.
The very construction of the distinction system/environment as a guiding
distinction for second-order observation begins with a form analysis of the
distinction, that is, with an analysis of the capacity of communication to shape
the distinction system/environment. For one thing, form analysis demonstrates
that the environment always works as an environment of a particular system
and, conversely, that a system is always a system only in relation to a specific
environment. A social system is simply the unity of the distinction system/environment.
When communication recursively connects with communication, social systems
emerge because of the distinction by the communication between self-reference
and external reference – between that which constitutes the system itself and
that which makes up the environment of the system.
Both system and environment are internal structures of communication.
Environment is not ‘reality’ as such: environment consists of that which is
defined by the communication as its relevant surroundings. The system, in
turn, is only a system in relation to this internal construction of the environment.
System is that which environment is not. Any system therefore is identical
with itself in its difference (and only in the difference) from the internal
environment construction.
80
A second-order observation, employing as its guiding difference the

distinction system/environment, is thus an observation of the way in which a
social system creates itself in the construction of its environment through communicative
descriptions. It is an observation of how a system distinguishes between system
and environment when it observes.
The difference between system and environment is the boundary of meaning
of the social system. The boundary system/environment is not spatial or
material: it pertains to the space of possibility for the creation of meaning in
the system. Any social system constitutes an autopoietic system of meaning of
its own, which produces meaning according to individual rules of selection.
The boundary of meaning indicates that there are different conditions of the
creation of meaning inside and outside the system. An observation of a system’s
use in its observations of a certain distinction between system and environment
is thus simultaneously a second-order observation of the limits to the creation
of meaning within a specific system, that is, the way in which a system ascribes
meaning to itself and its environment.
In this way, each social system is autopoietic both in respect to communication
and meaning. Accordingly, communication can only take place within a social
system and never between social systems. Social systems are unable to
communicate with each other but are able to communicate about each other.
Differentiation analysis
As we have already seen, the distinction system/environment is not merely a
difference, but one that has the capacity for re-entry into itself. A social
system comes into being through a communicative installation of a boundary
between system and environment. However, this is a formation within society
as a social system of communication. Any formation of a social system consists,
therefore, of a transcription of the difference system/environment and its re-
entry into the system itself. Hence, social systems come into being as
differentiations of communication (see Figure 4.6).
Figure 4.6: Re-entry of the distinction system/environment
System Environment
System Environment
System Environment
81
The differentiation of social systems implies the creation of distinct

communication forms of different perspectives of observation. The social
systems observe their difference from within their scope and each in its own
way splits the world in two through the distinction system/environment. Thus,
the world becomes poly-contextual.
Although any social system exists as a sub-system in society, it is not a part
of a whole. Society can never be represented as a whole; there is no total sum
of the perspectives of observation of the different social systems. On the
contrary, society becomes environment to the sub-systems and does so
differently in relation to each of the sub-systems. Any differentiation defines
the system from which it is differentiated as its environment.
Another reason for the impossibility of describing society as a unified whole
is the fact that a description of society is based on one specific system and has
one systems-relative way of creating meaning. Luhmann’s assertion is that we
are able to observe the similarities in the way in which the social systems’
perspectives of observation are differentiated. This is termed the form of
differentiation, that is, the unity of the difference between the systems.
Above the guiding distinction system/environment, we see, therefore, the
guiding difference similarity/difference. The second-order observation in the
differentiation analysis is the observation of the similarity of the difference
between system and environment in social systems. The question is how are
systems differentiated by virtue of the re-entry into themselves of the system/
environment distinction? That is, in which way do the social systems become
parts of their own whole? This pertains to the similarity in the mutual
dissimilarity of the systems’ construction of themselves and their environment.
The analysis of the form of differentiation does not merely suggest the
constitution of the possibilities for communication in one particular system,
but also the general possibilities and limitations for the formation of perspectives
of observation and for social systems of communication given a specific form
of differentiation.
Luhmann traces a shift in the overall form of differentiation in society in
the direction of functional differentiation, which reveals new forms of
communication while simultaneously excluding other previous forms. Within
the segmented form of differentiation, society is differentiated into identical
sub-systems – there are numerous social systems but they all constitute
themselves in the same way with the same perspective of observation. Within
the stratified form of differentiation, the social systems are segregated by layers,
such as, for example, the caste system of old Indian societies, or citizens versus
slaves in Ancient Greece, or the kings, nobility, peasantry and tenants of
European feudal society. Within the functional form of differentiation, social
systems are differentiated in respect of their functions, for example the financial
system, which is centred on the function of price fixing, or the political system,
which is centred on collective decisions, or the educational system, which is
centred on separation. This does not indicate whether or not the functionally
differentiated society is actually functional. Indeed, there is considerable
evidence to the contrary.
82
For example, Luhmann points out that the closure brought about by the
communicative focus on functions blinds society to the destruction of the
ecological precondition of society by the function systems. Each individual
form of differentiation only allows for particular perspectives of observation,
which is the same as saying that each form of differentiation installs its blind
spot within the communication of society. One example of a perspective of
observation that can exist in the stratified society but not in functional society
is the male-dominated perspective of observation. Formerly, the distinction
between man and woman made up a guiding difference in respect of social
stratification. Within the functionally differentiated society, the asymmetric
indication of the male side and the use of the distinction man/woman as a
principle of differentiation for social systems are no longer useful in the same
way.
Table 4.2: The differentiation of society
Segmental differentiation Similar sub-systems, such as tribes, villages and families

Stratified differentiation Differentiation in uneven layers based on the difference top/
bottom
Functional differentiation Differentiation in dissimilar sub-systems that differ from each
other in respect to their function in society
Points of observation
Although Luhmann has primarily studied the form of differentiation of society,

there is no reason why one should not choose other points of observation. The
point of observation could be a specific organisational system for an examination
of how sub-systems are differentiated; it could be the public administration to
consider changes in differentiation form from a formal administration to an
administration divided into different sectors and, hence, changes in the
autopoiesis of the administration; or it could be changes in the differentiation
form of news media from party press via regional press to broadsheets versus
tabloid newpapers. In other words, the object of the differentiation analysis
depends entirely on the choice of systems reference.
Media analysis
A fourth analytics of Luhmann draws on the guiding difference form/medium.
Any observation is an observation by a system in which observation exists in
an indication within the scope of a distinction. Subsequently, any observation
forms differences or, rather, any observation establishes a relationship between
observation as form and its difference as medium. In order for an observation
to be brought about in the first place, difference as difference has to offer itself
up as medium for a form that can condense differences into specific forms
such as horse/not horse, right/wrong, car/not car. Differences as loose elements,
83
without preferences one way or the other and open to any consolidation,
make up the medium of observation, which can precisely only take place by
isolating a single form in the medium of difference. If we look at Spencer-
Brown’s form calculus reproduced in Figure 4.7, we can see that the very line
or gibbet is the medium of form as the unity of the specific separation of m
from not m.
Figure 4.7: The calculus of form
By media, Luhmann understands loosely coupled elements. Media are

characterised by a high resolution and by being accessible to Gestalt fixations
(Luhmann, 1986b, p 101). Conversely, form to Luhmann implies a fixed
connection of elements: “Forms emerge [...] through a condensation of the
mutual dependency between elements, that is, through the selection of the
possibilities offered up by a medium” (Luhmann, 1986b, p 102). Moreover,
media always consist of numerous elements:
Forms, on the contrary, reduce size to that which they can order. No
medium creates only one form since it would then be absorbed and
disappear. The combinatory possibilities of a medium can never be exhausted
and the only reason for restrictions to evolve is the fact that the creations of
forms mutually disrupt each other. (Luhmann, 1986b, p 101)
The relationship between form and medium is itself a form, meaning that any
speech on form is only form in relation to a medium and vice versa. Forms
are only shaped when a medium makes itself available but, on the other hand,
form prevails in relation to the medium, without any resistance on the part of
the medium in regard to the rigidity of form. The difference between form
and medium, however, is relative in the sense that form can be more or less
rigid.
For example...
One example of a medium is money, which is a medium precisely because payments

can:
be offered as random notes, because one payment does not depend on the
significance and purpose of another payment, because the medium is
incredibly forgetful (since it does not have to remember in order to maintain
the paid amount), and because the solvency determines whether payment
is possible. (Luhmann, 1986b, p 101)
84
Decisions are an example of form; they constitute a form because they impress
themselves in a medium and condense its elements into one decision, which is
only a decision in relation to previous decisions and decisions not taken. The
decision to print the company’s logo on paper forms the medium of money by
requiring expenditure. This decision is not interchangeable with other decisions
– its meaning is tied to time and space, and can only be understood in relation to
the company’s other decisions. Moreover, the decision has been made possible
by previous decisions, for example the decision to have a company logo.
A form is not necessarily tied to one medium. It is possible to imagine a

situation in which several media make themselves available to the same form.
An organisation can form numerous different media (for example, money, law
and power), although it can never do so simultaneously.
Which forms and media are available to communication is an historical
question, that is to say, the difference between form and medium varies
historically. Furthermore, the distinction form/medium is capable of re-entry
so that a specific form can later work as medium for a new form. This extends
to an evolutionary question about the way media arise and produce possibilities
for new forms of communication. In the words of Luhmann:
How language, how writing, how alphabetical writing, and how symbolically
generalised media appear. They provide a potential for the creation of forms
which would not exist without them, and we can take advantage of this
potential as soon as the social conditions permit it. (Luhmann, 1986b,
p 104)
Historically, we can trace a development in which forms have, over time,

become media of new forms. In theory, the only limit is the fact that media
cannot be developed beyond the communication of form. Thus, the number
and ways of forms and media offering themselves to communication is
fundamentally an empirical question. As a result of the historical evolution of
new media, one could possibly speak of a media/form staircase on which the
next step is open (see Figure 4.8). Distinction is the medium for the form of
meaning, which further constitutes the medium for the form of language,
which in turn works as the medium for media of distribution and so on. The
staircase is by no means complete.
The implications of this extend to how we observe observations whose
guiding distinction is form/medium. The capacity of distinctions for re-entry
entails numerous potential points of observation. As with the guiding distinction
system/environment, the point of observation can always be moved forwards
or backwards. With the guiding distinction system/environment, one is always
able to choose a new systems reference, and with the guiding distinction
form/medium one can always choose a new form reference. Depending on
what we define as form in our second-order observation, different relationships
appear as medium (although no medium will appear, of course, if there are no
85
Figure 4.8: Media/form staircase
Symbolically generalised media
Media of distribution
Language
Meaning
Distinction
media). Luhmann provides the following example: “A public organisation

can be considered a form but also a medium in which interests clash and
impress themselves” (Luhmann, 1986b, p 104).
As we have already suggested above, Luhmann’s theory of society ascribes a
more prominent position to some media than to others. As a principle rule,
he distinguishes between three types of communication media:
• language;
• media of distribution, such as writing and television;
• general symbolic media, such as, for example, money.
He only ascribes binary codes to the general communication media.
Semantic analysis
Finally, Luhmann employs semantic analysis, which forms the guiding
distinction condensation/meaning, to examine how meaning is condensed in
semantic forms that produce a conceptual pool for communication.
Luhmann makes a distinction between system and semantics in which
semantics are defined as particular structures linking communication to
communication by offer ing up for ms of meaning that systems of
communication treat as worth preserving (Luhmann, 1995c, p 282).
If we were to invoke the concept of discourse anywhere in Luhmann’s
theoretical architecture, this would be the place. In the same way that we can
talk about different discourses in Foucault that enable the enunciation of
particular discursive objects, Luhmann speaks of different semantics (the
semantics of love, of organisational theory, of money and so on) that invoke
specific communication. However, the distinction between system and
semantics is considerably different to Foucault’s concept of discourse. In
Foucault, a discourse can contain characteristics of system; a discourse can
constitute a regime and a discourse can conflict with other discourses; discourses
possess qualities of reality. Semantics do not in the same way posseses these
86
qualities. They do not have a self; they cannot make up a regime; they cannot
conflict with each other. Only systems hold the characteristics of system;
systems decide whether or not they wish to employ specific semantics.
Consequently, semantics do not exist by themselves outside the systems of
communication. An observer can choose to make a semantic distinction
between system and structure but they cannot actually be separated. The
semantic structures are used and reproduced in the selection of communication
that is linked to communication, and can only exist in the constriction of
choices.
The concept of semantics is based on a distinction between meaning and
condensed meaning. As noted above, meaning comprises a constant rearranging
of the distinction actuality/potentiality. Meaning disintegrates immediately
on its actualisation. Thus, meaning is tied to the momentary condition of
actualisation. Communication, on the other hand, is capable of developing
structures that condense meaning into forms, which are set free from the
momentary condition of actualisation. Condensation means that a multitude
of meaning is captured in a single form, which subsequently makes itself
available to an undefined communication. Consequently, semantics are
characterised as the accumulated amount of generalised forms of differences (for example,
concepts, ideas, images, and symbols) available for the selection of meaning within the
systems of communication. In other words, semantics are condensed and repeatable
forms of meaning, which are at our disposal for communication. These
generalised forms are relatively independent of situations and obtain their
specific content from the communication by which they are selected (Luhmann,
1993c, pp 9-72). This definition of semantics largely derives from Koselleck’s
history of concepts.
Objects and concepts are among the forms that semantics can take. As
noted previously, objects constitute a form that has an undefined outer side
(for example, horse/not horse). The concept, in turn, is a form of meaning in
which the indicated inner side delineates restrictions for the outer side (for
example, man/woman), and hence concepts are inextricably bound up with
counter-concepts. There exist no concepts without counter-concepts and
thus no unambiguous concepts (Luhmann, 1988b, pp 47-117).
Luhmann distinguishes between three dimensions of meaning (Luhmann,
1995c, pp 74-82), which enables him to distinguish accordingly between three
forms of semantics. We will not go into this distinction in depth but merely
give a brief précis:
• The fact dimension pertains to the selection of themes and objects for
communication and consciousness. Themes and objects are all designed
according to the distinction this/something else in the same way as ‘object’
as a form of meaning. Similarly, we can speak of semantics of facts as
generalised forms of ‘being-one-thing-and-not-another’.
• The social dimension is based on non-identity in the relationship between
communicators and constitutes the horizon of possibility in the tension
between ‘alter’ and ‘ego’. It thus concerns that which cannot be regarded
87
as oneself. Semantically, it relates to generalised forms of differences between

‘us’ and ‘them’. There can be no ‘us’ without being in comparison with a
‘them’.
• Finally, the temporal dimension articulates the tension between past and future.
The temporal dimension is “constituted by the fact that the difference
between before and after, which can be immediately experienced in all
events, is referred to specific horizons, namely extended into past and future”
(Luhmann, 1995c, p 78). The semantics of time concern the ways in which
we observe and conceptualise past and future. Much like in Koselleck, past,
present and future cannot be regarded as given entities. On the contrary,
they are construed in and by communication and in every communication.
In the words of Luhmann: “What moves in time is past/present/future
together, in other words, the present along with its past and future horizons”
(Luhmann, 1982, p 307).
Connections between the different analytics

Once we have understood the guiding differences presented above, the emerging
question concerns the connections between them. Here, I must admit to
finding myself on thin ice. This chapter on Luhmann has introduced a total of
five central analytics (as compared to, for example, the chapter on Laclau,
which covered only two), and I do not believe that Luhmann has ever fully
elaborated on the overall connections between these. There are three reasons
for this. First of all, the rudimentary attempts that have been made to describe
the connections between the guiding principles are far from having been
revised in relation to the Spencer-Brownian turn in Luhmann’s writings.
Second, the descriptions of connections are not coherent across his different
books (possibly for the same reason). Third, it is possible that Luhmann does
not fundamentally believe in the possibility of one joint analytical strategy
and maintains that the different analytics combine to constitute a flexible
structure whose composition depends on the problem.
The following conclusions are based on the third of the above reasons and
will consider the numerous guiding principles as a pool, rather than as distinct
elements of one analytical strategy. One possibility is to conduct a systematic
reading of the connections between the guiding differences in Luhmann’s
different analyses. That would be an interesting study in itself, but beyond the
scope of this book. Instead, I will provide some examples of how the guiding
differences are brought together in different analytical strategies.
Form analysis and semantic analysis

The relationship between these analyses has a strong resemblance to the
relationship between deconstruction and the analysis of hegemony in Laclau.
Form analysis analyses the unity of a concept and illustrates the concept’s
foundation on a paradoxical distinction. The semantic analysis is then able to
employ the paradox as a guiding principle in tracing the history of specific
88
semantics: the semantics of love, the semantics of law and so on. Subsequently,
semantic history acts as a history of the de-paradoxification of a particular
paradox, for example the history of how the paradox of the justice of justice
has historically sought de-paradoxification through different judicial semantics:
natural law, judicial positivism, Scandinavian judicial realism and so on.
Historical semantic analysis can, in turn, point out new concepts for form
analysis. The relationship between form analysis and historical semantic analysis
is illustrated in Figure 4.9.
Figure 4.9: The relationship between form analysis and semantic

analysis
paradoxes
Points to the
blind spot of the Analyses the semantic
concept by conditions of the formation
illustrating how form analysis semantic analysis
and de-paradoxification of
it is based on a the paradox
paradox
concepts
Semantics and differentiation

Luhmann provides an explicit description of this relationship in his first book
on social structure and semantics (1993c). He understands social structure to
mean the relationship between complexity and the primary differentiation
form of society. This is based on the notion that the differentiation form of
society determines the level of complexity in society (I have chosen to leave
out the concept of complexity in my reading of Luhmann). Translated into
those concepts that I have chosen to actualise here, the concept means that the
differentiation form of society determines the character of the meaning
boundary of the social systems, and thus their openness and sensitivity to their
environment. In other words, the form of differentiation governs how social
systems can be disturbed by their internally constructed environments, and
how they react to these disturbances. Accordingly, Luhmann’s hypothesis is
that semantic development always follows the differentiation form of society,
that is, that compression of meaning into concepts in individual social systems
follow each other in parallel movements. Luhmann calls this a collective
transformation of the semantic apparatuses of society. Accordingly, semantic
analysis meets the complexity–differentiation relationship horizontally (see
Figure 4.10).
This structure invokes an analytical strategy that explores various semantics
related to different social systems (for example, law, politics, art and religion),
according to the thesis that simultaneous transformational ruptures in the
history of semantics indicate a transformation of the social structure. The
89
Figure 4.10: The relationship between differentiation and semantics
Complexity
Semantics
Differentiation
Source: Luhmann, 1993c, p 34
relationship between differentiation analysis and semantic analysis is thus defined

as a relationship between synchronic and diachronic analysis.
This, in turn, does not mean that Luhmann assumes a figure of continuity
in regard to the development of semantics or the existence of a one-to-one
relationship between differentiation form and semantics. Semantics change
continuously as well as discontinuously in relation to the differentiation form,
in the same way that semantics can follow a temporal rhythm different to that
of the differentiation form. Concepts and ideas from one form of differentiation
can live on in the next one. Concepts can change with the development of
new counter-concepts but maintain their designation. Finally, former
distinctions can be put together and condensed in the same concept (Luhmann,
1993c, pp 7-8). However, Luhmann maintains the thesis that the form of
differentiation fundamentally defines the conditions of development for
semantics and not the reverse. The reasons for this appear logical: semantics
are shaped and transformed by communication, which is always communication
within a social system, and the form of differentiation, as it is, is nothing more
than the similarity of the dissimilarity of the social systems1.
Systems analysis and media analysis

In my opinion, the relationship between systems analysis and media analysis
can be presented in different ways. Here, the relationship between the two
analyses is defined as an analytical strategy, whose purpose it is to capture the
formation of a social system.
The basic notion is that systems of communication always emerge in relation
to a medium. Systems of communication are always shaped within a particular
medium; there are no social systems except through the formation of a medium,
at least the medium of meaning. Luhmann’s thesis is that the historical evolution
of new media makes the formation of new social systems possible. The study
of how a new social system arises, therefore, should always begin with an
analysis of the media-related conditions of the emergence of that system. Thus,
the evolution of media of diffusion (such as printed media, radio and television)
can be viewed as an evolutionary justification of the origin of the mass media
system. Correspondingly, the formation of the economic function system
90
presupposes the evolution of money as a general symbolic medium of

communication.
Within this analytical strategy, media analysis thus constitutes a study of the
historical evolution of media and their justification of new forms of
communication. In relation to this, systems analysis becomes a study of how
a particular system comes into being through the connection to new
evolutionary potentials of the media, and how it establishes its own autopoiesis.
This analytical strategy is illustrated in Figure 4.11.
Figure 4.11: The relationship between systems analysis and media analysis
Evolution of media
Emergence of new system

impressed in new medium
Luhmann’s guiding distinctions could create several other analytical strategies,

but the three introduced above can serve as an indication of others that are
possible.
Conclusion
This chapter has examined the most central analytical strategies in Luhmann’s
systems theory. I have attempted to summarise these strategies in Table 4.3,
and have supplemented them with a specific example. The left-hand column
indicates the analytical strategy; the centre column the general inquiry of the
analytical strategies; and the right-hand column an example of a potential
systems theoretical analysis. The example used in this table concerns the
politicisation of businesses. Today, private businesses have become central
political players in a number of fields and are expected to take on social
responsibility that exceeds purely economic concerns. Today, corporations
are expected to consider issues beyond those that are economic; for example,
many businesses have developed extensively into areas including ethics, the
environment and human rights. Many businesses also take an active role in
political decision-making processes; for example, thousands of businesses take
part in standardisation committees and commissions, and are involved in
decision making about the rules for the European market. These decisions
concern consumer interests, the environment, working conditions, industrial
policies and much more.
91
Table 4.3: Luhmann’s analytical strategies

Analytical General
strategy question Example
Form analysis What is the unity of the In what way are organisations
distinction? And which systems that communicate
paradox does it establish? through the form ‘decision’? And
which paradoxes does this form
establish?
Systems analysis How does a system of In what way does the politicisation
communication come into being of the organisations become
in a distinction between system apparent in the internal
and environment? How is the construction by the organisations
system’s boundary of meaning of their environment so that
and autopoiesis defined? they not only construct the
environment as market but also as
political public?
Differentiation How are systems differentiated? In what way does the politicisation
analysis What is the similarity in the of the organisations challenge
dissimilarities of the systems? their internal form of
What are the conditions, differentiation and force them to
therefore, of the formation of institutionalise internal
new systems of communication? reflections of themselves as closed
communication (for example,
through the establishment of so-
called ‘ethics officers’)?
Semantic analysis How is meaning condensed? And How is meaning condensed with
how does it create a pool of respect to environment, human
forms, that is, stable and partially rights, ethics, animal welfare,
general distinctions available to health and prevention, into the
the systems of communication? concept of ‘the socially responsible
corporation’, and bring about new
conditions for corporate
communication?
Media analysis How are media shaped? How do In what way does the politicisation
they suggest a specific potential of the organisation mean that the
for formation? organisation is no longer only
supposed to form the medium of
money, but is also expected to
form a number of other
communicative media such as
power, information and morals? In
what way does this change the
conditions of the organisation from
homophony to polyphony?
Note
1
However, it is debatable whether Luhmann consistently maintains this strategy,
since he, on several occasions, ascribes to semantics a more constitutive role in
relation to systems formation (see Stahäli, 1998).
92
5
A hall of mirrors or a pool of
analytical strategies
n this concluding chapter, I will seek to let the different analytics and
I analytical strategies in Foucault, Koselleck, Laclau and Luhmann reflect

each other and, in so doing, attempt to compare them directly. This
requires some form of standard of comparison – a uniform concept that will
allow the distinctions to stand out. However, that in itself is a questionable
operation, since the principal distinctions that define their constituent diversity
are lost in such an endeavour. Nevertheless, this will not restrain me from
proceeding to define a kind of ‘comprehensive view’, the advantage of which
is that it paves the way for many fruitful discussions about how it is different
to establish, for example, a link to the discourse analysis of Foucault or Laclau
respectively.
As already noted, there is no natural universal standard of comparison. The
four gentlemen do not ask to be compared and neither do they require a
particular standard of reference. However, using Luhmann’s ideas, I have chosen
to construct a concept of analytical strategy on which I will base my
comparisons of the four writers. This will, of course, have the effect of making
the others appear in a Luhmannian light, but there are reasons for my choice
of perspective. The two principal reasons are width and level of abstraction.
First, in my opinion, Luhmann possesses the greatest potential of inclusion
due to his rather general but also highly potent theory of second-order
observations. Second, Luhmann’s concepts appear to permit the development
of a productive concept of analytical strategy. The first part of the conclusion
will therefore offer this as a standard of comparison.
Conditions of comparisons of discourse theories and

systems theories
The basis of comparison is a specific notion of analytical strategy, which is
founded on the theory of second-order observations. Analytical strategy can
be viewed as a second-order strategy for the observation of how the social
emerges in observations (or enunciations and articulations). The elaboration
of an analytical strategy involves the shaping of a specific gaze that allows for
the environment to appear as consisting of the observations of other people.
93
Precondition 1
The first precondition of this idea of analytical strategy is that an observation

is an indication within the scope of a distinction, meaning that it is the distinction
that divides the world into observer and observed, thus making the environment
appear in a specific way.
Precondition 2
Systems theory, conceptual history and discourse analyses are all, in some
sense, analyses of the second order: observations of observation, descriptions of
description, conceptualisations of concepts or designations of signs. Only
while this condition is maintained is the comparison productive.
Precondition 3
The third precondition is the realisation that, on this second-order level, it
does not make much sense to let one’s research be guided by rules of method.
This is simply because a definition of methodological rules leads to an
ontologisation of the social phenomena, when the aim is precisely de-
ontologisation. The equivalent of methodological rules in second-order
observation is analytical strategy. Second-order observation is aware that the
world is not asking to be observed in any particular way. Second-order observers
perceive of the world as poly-contextual, as dependent on the distinction
shaped by observation. Consequently, second-order observers can choose
how the world should appear by deciding on a particular way of seeing, by
connecting with a particular distinction. This is what analytical strategy is
about: choosing a way of seeing and accounting for its implications regarding
the way the world appears and does not appear. Observations of observations
as observations are contingent in relation to the chosen way of seeing. It is
always possible to observe the second order in a different way. A decision has
to be made, therefore, and the decision calls for an explanation.
Precondition 4
The elaboration of an analytical strategy consists of four choices:
1. Choice of guiding distinction.

2. Conditioning of guiding distinction.
3. Choice of point of observation.
4. Choice of potential combination of analytical strategies.
94
A hall of mirrors or a pool of analytical strategies
Precondition 5
One also needs a guiding distinction, by which I mean the distinction that
can define the frame for observations of the second order. The guiding
distinction divides the world to the second-order observer and dictates how
the world can be observed. The guiding distinction creates the necessary
distance between the object and the observer, but also defines which questions
can be posed in relation to the object. The guiding distinction is not predefined
but must be decided. Whether the distinction system/environment, regularity/
dispersion of statements, or discourse/discoursivity is defined as the guiding
distinction is an essential analytical-strategic choice. The guiding distinction
steers the observation and frames the choice of different supporting distinctions.
For example, if the choice is system/environment, the world is divided into
systems and their environment; if one decides on regularity/dispersion, the
world is divided into the dispersion of statements and the regularity of this
dispersion. And one sees only that, which is precisely the purpose of the
guiding distinction: to discipline our way of seeing so that we do not
involuntarily fall back on first-order observation. This is why Foucault expends
so much energy on sharpening his discourse-analytical focus. He knows that
he is only able to observe the discourse as it appears if his way of seeing is
refined enough to keep him from reducing his discourse analysis to discourse
commentary.
Precondition 6
The conditioning of the guiding distinction is the specification of those
conditions by which one and not the other side of a distinction can be indicated
in a second-order observation. It does not suffice to distinguish between
system and environment. One must also define when a system can be perceived
as a system, when a system can historically be said still to be the same system,
when the system has evolved into a different system and when the system has
ceased to be a system. Otherwise the guiding distinction can only obtain the
status of a loose metaphor in relation to which an observer’s indication of
system and environment respectively becomes completely arbitrary. This, of
course, applies to all guiding distinctions. We could thus ask:
• When is a discourse a discursive formation? That is, when can we reasonably

indicate a discursive formation?
• When does a concept of meaning reach into the future through its
condensation?
• When is a nodal point or an anchor point firmly anchored?
Whereas the guiding distinction constitutes the fundamental openness and

closure of second-order observation (that is, the conditions of observation
emerging as object for observation), the conditioning defines the empirical
sensitivity of second-order observation, including the openness of the analysis
95
to criticism. The clearer and more unambiguous the conditioning, the greater
the sensitivity of the analytical strategy to the empirical, which amounts to
saying that the clearer the conditioning of the guiding distinction, the more
evident the falsification criteria of the second-order observation. A discourse
analysis without well-defined criteria for the indication of discursive formations
immunises itself to tangible empirical criticism, thereby dooming any critical
discussion of the analysis to a meta-discussion. (Unfortunately, there are many
of these types of discourse analyses.)
Precondition 7
I understand point of observation to be maintained through external reference
in a second-order observation. It is not enough to decide on a guiding
distinction and to condition it in order to adhere to the analytical eye. As
Foucault points out, different statements can be read very differently depending
on the discursive regularity one employs. In relation to discourse analysis, the
point of observation implies a choice of discursive reference, that is, a definition
of the question whose genealogy one wishes to unravel and the implications
this has for fixing the point of observation. It is my view that this is, as already
mentioned, what Foucault discusses when he, for example, dwells on the criteria
for periodisation. Similarly, in systems theory the analysis depends entirely on
the systems reference of the second-order observer. Do we perceive of the
communication as a system of interaction, an as organisational system or as a
social system? Each communication can appear as an event in several systems
simultaneously. Thus, the appearance of the communication to the second-
order observer relies on the observer’s choice of systems reference (that is,
which system he decides to observe). Again, communication does not itself
call for a particular point of observation; a choice is required, and the choice
must be followed by an explanation.
Precondition 8
Finally, analytical strategies can be combined, as, for example, when Laclau
combines deconstruction and hegemonic analysis. Here, a thorough account
of the complementation of the analytical strategies is critical.
• What are the rules for joining the different guiding distinctions?
Various discourse and systems theories are comparable on the level of their
analytical-strategic potential, that is, on the level of their pool of guiding
distinctions, conditionings and points of observation.
96
Table 5.1: Analytical strategies compared

Analytical strategy Guiding distinction Question
Foucault
Archaeological discourse Regularity/dispersion Why did this and no other statement occur in
analysis of statements this place?
Genealogy Continuity/discontinuity How are different discursive formations and
discursive strategies shaped and transformed?
Self-technology analysis Subjectivation/subjecting How have self-technologies been created and in what
way do they prescribe the way an individual can give
itself to itself?
Dispositive analysis Apparatus/strategic logic How are forms linked together as functional
elements in an apparatus?
Strategic logic/apparatus How are discursive or technical elements generalised in
a schematic, which creates a strategic logic?
Koselleck
History of concepts Meaning/meaning condensed How is meaning condensed into concepts, which
into concepts constitute the space of possibility of the semantic
conflicts?
Semantic field analysis Concept/counter-concept How do concepts appear in relation to their
(generality/singularity) counter-concepts?
How are temporal (past/future), spatial (outside/inside),
and social (us/them) relations produced?
How are general positions established and how is the
general singularised?
Laclau
Hegemonic analysis Discourse/discoursivity How are discourses established in never-concluded
battles about fixing floating elements of signification?
Deconstructivist analysis Signifier/signified Which infinite logic is installed by the duality in
question?
Luhmann
Form analysis Unity/difference What is the unity of the distinction? And which paradox
does it establish?
Systems analysis System/environment How does a system of communication come into
being in a distinction between system and environment?
How is the system’s boundary of meaning and
autopoiesis defined?
Differentiation analysis Similarity/dissimilarity How are systems differentiated?
What is the similarity in the dissimilarities of the
systems?
What are the conditions, therefore, of the formation of
new systems of communication?
Semantic analysis Condensation/meaning How is meaning condensed?
How does it create a pool of forms, that is, stable and
partially general distinctions available to the systems
of communication?
Media analysis Media/form How are media shaped and how do they suggest
a specific potential for formation?
Form/medium How are specific media imprinted in concrete forms,
thus colouring communication in a particular way?
97
System and discourse aligned: guiding distinction

Table 5.1 outlines a comparison of Foucault, Koselleck, Laclau and Luhmann
in regard to their analytical strategies. The table concerns the analytics drawn
out by the different theories, the guiding lines that constitute the analytics and
the questions that these guiding lines offer the present research agenda.
Foucault
In Foucault I have isolated four analytics:
• archaeological discourse analysis;

• genealogy;
• self-technology analysis;
• dispositive analysis.

The guiding distinction in archaeological discourse analysis is regularity/
dispersion of statements and it poses the question of why this and no other
statement appears in a particular place. Archaeological analysis divides the
world into regularity and dispersion, where no regularity can exist without
dispersion, since regularity consists of nothing other than the regularity of the
dispersion. With this guiding distinction, Foucault allows for analyses of the
order of statements without turning the analytical operation into a reductive
operation (seeking to reduce, for example, many discursive events to one or
few causes and many discursive manifestations to one latent structure), but
rather possessing the ability to sustain the statements in their appearance.
Geneaology
The guiding distinction in the genealogical analytical strategy is continuity/

discontinuity, where continuity (that is, the similar or identical) can be
continuous only in relation to the discontinuous (that is, the dissimilar or
different) and vice versa. This guiding principle allows for questions about
the way discursive formations and practices are shaped and transformed. With
this guiding distinction, Foucault lets his analysis be sensitive to any statement
and any practice that presents itself as new (that is, as freed of prior
misconceptions) or as unique (that is, as identical to its origins). The guiding
distinction continuity/discontinuity provides Foucault with an analytical
strategy that is not easily captured by the stories told by the present of the
present but which, in turn, can function as counter-memory.
98
The guiding difference of self-technology analysis is subjectivation/subjection,

which opens up an inquiry into the way the subject is proclaimed as self-
proclaimer, the way self-technologies are created and how they prescribe self-
activity through which the individual can give itself to itself.
Finally, the guiding difference of dispositive analysis is strategic logic/apparatus

but also apparatus/strategic logic. This allows for inquiries into the way
discursive or technical elements are generalised within a schematic that brings
about a strategic logic and inquiries into how forms are linked together as
functional elements in an apparatus through the unfolding of the strategic
logic. Dispositive analysis should be viewed as an extension of Foucault’s
other analytical strategies, since the dispositive analysis inquires about the
mutual ordering of those orders addressed by the other analysis.
Koselleck
In the works of Koselleck we seem to find a distinction between two analytical
strategies:
• history of concepts, which is diachronic;

• semantic field analysis, which is synchronous.
History of concepts
The concept-historical analytical strategy is, like Foucault’s archaeology, of
the second order. It is a strategy for the study of the origins of the linguistic
space of possibilities. But the guiding distinction and hence the opening
question is different. Therefore, although they both work with linguistic units,
they are objectified in different ways. Koselleck’s diachronic guiding distinction
is meaning/meaning condensed into concepts. Koselleck studies the way the
creation of meaning is concentrated and contained in concepts that then
become carriers of a multitude of meaning, which can organise the shaping of
identities and fields of conflict, and thereby extend into the future. The focus
is on concepts, which creates different analytical material to that of Foucault’s.
Whereas Foucault ideally reads everything, Koselleck focuses on what he calls
nurtured semantics, that is, writings that have been worked through linguistically
and thus employ concepts prudently.
Sematic field analysis
Koselleck’s synchronous guiding distinction is concept/counter-concept. This

guiding distinction allows for questions about how concepts always emerge
99
within a semantic field in relation to their counter-concept. For example,

Koselleck points to the fact that ‘us’ is always constructed to the exclusion of
‘them’; that the present is always established in the tension between space of
experience and horizon of meaning; and that any community organises itself
internally through the conceptual opposition of up/down.
Laclau
In Laclau we are able to distinguish:
• hegemonic analysis;
• deconstructivist analysis.
Hegemonic analysis
The former analysis corresponds to the history of concepts and archaeological

discourse analysis insofar as it aims to study the origins of discourses, but,
again, its guiding distinction is different, which means that the question of
discourse presents itself differently. The guiding distinction of the analysis of
hegemony is discoursivity/discourse. The shaping of discourses is therefore
always studied in its capacity to fix the floating elements of discoursivity.
Unlike in Foucault’s work, the object is not statements in their appearance,
but discursive elements that appear relationally and include practice as well as
utterances. In a sense, Laclau’s hegemonic analysis is closer to Koselleck’s
history of concepts than it is to Foucault’s discourse analysis, because they
both emphasise ambiguity and the incessant floating of meaning as that which
establishes the political.
Deconstructivist analysis
Conversely, the deconstructive analytical strategy finds no equivalence in

Foucault and Koselleck. The deconstructive eye makes an object of the
individual differentiation in order to examine the logic that the distinction or
duality might install in the discourse. Through deconstruction, we are able to
unveil particular mechanisms of the creation of meaning, which operate as a
game of dominance between the two sides of the distinction. The combination
in Laclau of the two analytical strategies, by defining deconstructed dualities
as guiding distinctions in the hegemonic analysis, gives a very different result
to a concept-historical or an archaeological discourse analysis.
Luhmann
For Luhmann it becomes difficult to maintain a comprehensive view of the
pool of guiding distinctions and the potential combinations. I believe I can
define five different analytical strategies in Luhmann:
100
• form analysis;
• systems analysis;
• differentiation analysis;
• semantic analysis;
• media analysis.
Form analysis
Form analysis bears close resemblance to Laclau’s deconstructive analytical

strategy, as it concerns the logic of the creation of meaning (identical to paradox
in Luhmann), which is installed with specific distinctions. The guiding
distinction of form analysis is unity/difference and identifies forms of
observation as object by means of this distinction. Form analysis inquires
about the distinction that controls observation and about the paradoxes
established by the distinction. Form analysis asks about the unity of the
separated. In Luhmann, form analysis is never a goal in itself but rather a
strategy for the formulation of further questions – primarily the question of
how social systems de-paradoxify the very paradoxes on which they are built
through communicative operations.
Systems analysis
The guiding distinction for systems analysis is system/environment. With this

guiding distinction, the world is divided into system and environment, with
system the operative side. It is therefore from the side of the system that the
boundary system/environment is defined. In this perspective, communication
is always communication within a system, and the boundary system/
environment is a boundary of meaning that determines that the creation of
meaning in terms of communication happening on different terms outside
the system than inside it. The guiding distinction system/environment poses
the question of how a system of communication comes into being in the
communicative distinction between system and environment.
• How is the boundary of meaning defined in the system and what are the
implications for the continuance of the communication in the system?
• In what way is it closed around itself?
• How does it develop a sensitivity to the environment, that is, how does the
system define the relevance of the environment?
There is no equivalent to the guiding distinction system/environment in the

other programmes for observation. The closest comparison is Foucault’s
occasional descriptions of the individual discursive formations as regimes of
knowledge and truth ascribing a self to discursive formations. In fact, these
descriptions exceed Foucault’s analytical strategy and its intention to describe
discursive formations as the regularity of the irregular. In Koselleck, the closest
notion is in the counter-concept us/them, but here we operate exclusively on
101
a semantic level and not on an operative level as with Luhmann’s concept of

systems.
Differentation analysis
Differentiation analysis employs the guiding distinction similarity/dissimilarity.

It can be argued that this is not an independent analytical strategy but rather
a supplement to systems analysis, since the guiding distinction similarity/
dissimilarity presupposes the distinction system/environment. However, the
guiding distinction similarity/dissimilarity poses questions that are not directly
accessible with the guiding distinction system/environment:
• How are systems differentiated?

• What is the similarity in the dissimilarities of the systems?
• What, consequently, are the conditions of the formation of new systems of
communication?
With the guiding distinction similarity/dissimilarity, we can observe the

similarities defined by the systems of communication in the differences between
themselves and their environment. Thus, we can inquire about the similarities
in how systems distinguish themselves from each other. In a certain sense,
therefore, differentiation analysis is also form analysis because it concerns forms
of differentiation, that is, the unity of the dissimilarity of the systems. The
closest analogy in the other programmes for observation is probably the strategic
level of Foucault’s archaeological discourse analysis. Here, he strives to formulate
the question of the mutual exclusion of the discursive formations. Nevertheless,
the question is not greatly elaborated on the theoretical level by Foucault, but
is unfolded in a number of empirical analyses, for example of the epistemic
basis of discursive formations (for example Foucault, 1974, 1986a).
Sematic analysis
Semantic analysis uses the guiding distinction meaning/condensed meaning.

Semantic analytical strategy asks the questions:
• How is meaning condensed?

• How does meaning establish a pool of forms, that is, stable and partially
generalised distinctions available to systems of communication?
Luhmann’s semantic analytical strategy is largely inspired by Koselleck’s

conceptual history, but differs fundamentally through its origins in the concept
of system. This is critical to the analytical strategy because it can then be
linked with a social theory.
102
Media analysis
Luhmann’s fifth and final analytical strategy uses the guiding distinction form/
medium, where form can only be regarded as such in relation to a medium
and vice versa, with form the operative side. Media analysis only finds its
equivalent in Foucault’s dispositive analysis, in which apparatus corresponds
to form and strategic logic corresponds to medium. The strategic logics that
Foucault pursued largely resemble Luhmann’s symbolically generalised media
in their characterisation. Foucault speaks about legal/illegal, Luhmann about
right/wrong; Foucault speaks about security preparedness/insecurity and
Luhmann speaks about risk/danger. Both medium and strategic logic can
only be brought about through something else; that something else is apparatus
in Foucault and form in Luhmann. However, Luhmann never ascribes a
strategic function to media. The concepts of form and apparatus both constitute
the context in which logic and media are inscribed. However, having
highlighted the similarities, the big difference is that Foucault’s apparatus is a
system of relations between forms, whereas Luhmann appears not to have
unfolded the potential in the concept of form with respect to the relations of
the forms to other forms. (Although he does point to a guiding difference,
which he terms element/relation, and one could argue its potential for
perspectives similar to those that Foucault opens with his concept of apparatus.)
Finally, like dispositive analysis, media analysis constitutes a double movement.
It can be approached from the side of the form as well as the side of the
medium. We can inquire into the generalisation of a form into a medium,
which I call media analysis. But we can also ask which medium is formed by
a specific form and with what effects, which I call formation analysis.
Media analysis asks the questions:
• How are media shaped and how do they suggest a particular potential for
formation?
• How do media render certain forms of communication probable?
Media analysis allows for analyses of the history of forms of communication:
• How are, for example, language – spoken and written – and media of
distribution such as the Internet shaped?
• How do they substantiate entirely new forms of communication?
Moreover, media analysis as a formation analysis creates the possibility of

inquiring about existing media available to particular systems of communication,
and the potential and limitation of these media in this regard. This includes
the way that the formation of specific media in communication (for example,
money) implies a communicative closure around a binary value code of having/
not having.
The different guiding distinctions establish strategies for very different
observations of the second order. They each pose their set of questions and
103
establish observation as an object in their own way. The guiding distinction

not only constructs the object and question, it also implicitly constructs the
criteria for shaping concepts and for valid argumentation. A concept’s meaning
depends on the choice of guiding distinction. It is not possible for a concept
to be given the same meaning based on two different guiding distinctions.
This applies to a writer’s works as well. The concept of ‘discourse’, for example,
is not identical in Laclau’s deconstructive strategy and in his analysis of
hegemony. In hegemonic analysis, discourse is a structural unity of differences,
but such a notion of discourse is inconceivable in deconstructive analytical
strategy, in which discourse is the designation of a specific logic linked to a
specific duality. I have used names (Foucault, Kossellek, Laclau, Luhmann) as
pegs on which to hang analytical strategies, but, in the end, it is not Luhmann
who observes but observation itself, framed by a particular distinction, which
determines the construction of the object and the criteria for the construction
of good arguments and concepts. Depending on the choice of guiding
distinction, the concepts are defined in a particular way. This is an object of
systematic reflection in Luhmann’s conceptual universe.
If one enters via the guiding distinction of semantics (that is, meaning/
condensed meaning), the concepts ‘system’, ‘organisation’, ‘society’, ‘politics’
and ‘complexity’ appear as various semantics. Thus, we can study the semantic
history of, for example, the concept of system, the concept of politics or the
concept of complexity. If we enter via the guiding distinction of systems
analysis (system/environment), ‘system’ is not semantics but an autopoietical
system of communication, and ‘organisation’, ‘society’ and ‘politics’ are not
semantics but different systems of communication. In the same way, the concept
of complexity is granted a central theoretical function to the description of
the complexity-reducing relation of systems to the environment. If we enter
via the guiding distinction of form analysis (indicated/unindicated), the
concepts appear as one side of a distinction, for example system contra
environment and politics contra non-politics.
In this way, we can continue with all guiding distinctions – Luhmann’s as
well as the others’ – and conduct a systematic examination of what happens to
the concepts when a specific distinction is defined as the guiding distinction.
All concepts are fundamentally coloured by the choice of guiding distinction,
which means that there is no ‘true’ definition of ‘discourse’, ‘system’, ‘meaning’
and so on, since the criteria for ‘true’ and ‘real’ are simply construed through
the choice of guiding distinction.
Conditioning and point of observation

Guiding distinctions not only construct questions, object and concepts, they
also, as we will now address, establish their own set of analytical-strategic
problems and their own questions with a choice of observation point and
conditioning guiding distinction. These are questions that can only be solved
in relation to the empiricism or reality to which the analytical strategies seek
to become sensitive. However, there are certain general aspects regarding the
104
character of these questions within the different analytical strategies. I will

attempt to formulate and compare these questions below.
Foucault
Let us, once again, begin with Foucault’s archaeological discourse analysis.
Here, the fundamental question is the conditioning of the discursive formation.
When are we able to identify a discursive formation? The guiding distinction
regularity/dispersion of statements is not sensitive to empiricism until it has
been established when a regularity in the dispersion of statements constitutes
a discursive formation. Without exact criteria for when we can identify
regularity and hence discursive formation, our designation of discourses is
based solely on intuition (Hebrew for revelation).
In fact, Foucault answers this question thoroughly in his work on knowledge
archaeology, but the entire answer depends on which particular form of discourse
we are dealing with. There is no guarantee that regularity can signify the same
thing at all times and in all discursive contexts. Foucault knows this. As it is,
knowledge archaeology precisely designates knowledge as a point of reference.
It extends itself into very different areas of knowledge but always in the shape
of knowledge. This is evident even down to the level of the specifications by
the analytical strategy of the levels of formation, including the emphasis on the
object level and the questions of how objects are shaped, placed in hierarchies
and classified. We can conceive of numerous other forms of discourses that
would not have in common the form of knowledge and would also not, therefore,
share the same form of regularity. Hence, knowledge-archaeological analytical
strategy possesses an inherent notion of the characteristics of knowledge
discourses, but these characteristics themselves become the object of the analysis
and discussion of their origins. Accordingly, The birth of the clinic (Foucault,
1986b) does not merely concern the history of medical discourse but also the
origins of the scientific way of seeing. The conditioning of the guiding
distinction is not therefore a question of deciding on a set of premises once and
for all. In a discourse analysis, conditioning the concept of discourse is itself an
object of analysis. The way a discursive formation is a discursive formation is
an intricate part of the discourse analysis. The discursive formation does not,
therefore, precede the study of it, since the discursive formation as the subject
of discourse analysis can only be demarcated after the completion of the analysis.
In practice, this means that the discourse analyst constantly questions the
discourse’s own criteria for discourse.
Genealogy
In genealogy, the question of conditioning is defined as a question of the

measure for the assertion of continuity and discontinuity respectively. However,
conditioning in genealogy implies an almost reversed conditioning, since the
105
aim is to criticise predefined notions of continuity and rupture, that is, to

criticise the conditioning of continuity and discontinuity through discourses.
The aim is an uninterrupted continuation of contingency in relation to
continuity and discontinuity. In this respect, genealogy is in strict opposition
to archaeology, since genealogy is almost anti-analytical in its rejection of any
naturalisation of change. Genealogy is given a Dionysian life-giving function
by leaving it open, contrary to the tendency to closure in archaeology due to
its propensity for systematism. Incidentally, we can see a parallel here to the
function of deconstruction in Laclau in relation to hegemonic analysis and to
the function of form analysis in Luhmann in relation to systems analysis.
In self-technology analysis, the central problem of conditioning is the ability

to define clearly which form of subjecting applies when. When is a subject
expected to give itself to itself? In other words, when is it a matter of
subjectivation? The second problem of conditioning concerns when one can
speak of technology and self-technology respectively. This problem of
conditioning concerns both the question of when to define something as self-
technology and the question of which analytical variables we make ourselves
sensitive to as analysts. Here, as in knowledge archaeology, Foucault faces the
problem of the purist. He would like to let the way self-technology is self-
technology remain an empirical, and thus historical, question. However, he
needs certain measures that allow him to recognise a self-technology in the
first place. The challenge is to apply a concept of self-technology that is as
empty as possible and then to imply the conditioning as part of the analytics
itself as that which the analysis is fundamentally about. There are, however,
four minimum criteria:
1. The transformative form of subjecting.

2. The objectification of the self.
3. Self-directed activity.
4. Telos.
Foucault requires the objectification of the self, requires self-directed activity,

and requires that the self-directed activity have a form of telos, but in none of
these cases does he specify how this should be done.
Finally, dispositive analysis contains two problems of conditioning. First is the

question of when elements are linked in such a way that it is possible to see
the emergence of an apparatus. The analysis can leave open which elements
have to be present, but one has to be able to define the synchronous cohesion
of the elements. As far as I can see, Foucault provides no unambiguous criteria
but maintains the argument based on a sense of systematism. Second is the
106
question of when a scheme has obtained the characteristics of a general and

strategic logic. Foucault lets the second problem of conditioning depend on
the first, since the criterion is that there is not a strategic imperative until it is
brought about through an apparatus.
Point of observation
A point of observation must also be decided upon. In Foucault’s four analytical

strategies, this decision depends on the question one wishes to pose and pursue
in the analysis, and choice of question always implies a choice of discursive
reference.
For example...
The question might be the birth of the complete employee, but the specification
of this question implies a definition of whether the point of observation is a scientific
discourse and, if so, which one (for example, psychology or organisational theory).
It might also be a political discourse, for example that of administrational politics.
Depending on the discourse one chooses as point of observation, different criteria
for discourse exist and different articulations and notions of the ‘complete
employee’ appear.
In discourses of organisational studies, the ‘complete employee’ operates within

a specific theoretical programme, which obtains its identity through its
differentiation from other theoretical programmes. In discourses of administrative
politics, the ‘complete employee’ does not gain acceptance because of the
concept’s capacity for organisational description but because of its ability to
function as a strategic tool for the commitment of the individual employee to
the changes in the public sector. The choice of observation point directs discourse
analysis and genealogy in different directions. Moreover, well-defined points of
observation allow for studies of interdiscursive relationships, for example in
regard to the ‘complete employee’.
Koselleck
Concept-historical analysis
In Koselleck’s concept-historical analysis the most essential problem of

conditioning is defining the criteria for when a concept can be regarded as a
concept, which means, in Koselleck, a definition of when meaning has been
condensed into the form of the concept. The problem of conditioning is
rather unproblematic in Koselleck, since all his concepts are developed with
the problem of conditioning in mind. The measures for when a concept is a
concept are fairly unambiguous but the entire analysis, in turn, depends on
107
the fact that the creation of meaning possesses exactly these conceptual qualities.
This includes the existence of a nurtured semantics in which the concepts can
appear as object to the history of concepts. Conceptual history is not able to
capture creation of meaning that does not obtain the form of concept (for
example, symbols or images).
Semantic field analysis
In Koselleck’s semantic field analysis the problem of conditioning is a definition

of when a semantic field possesses the characters of a field. The problem of
conditioning in regard to semantic fields has not been unfolded thoroughly in
Koselleck. However, his notion of the fundamental general distinctions (us/
them, inside/outside, past/future and up/down) provides some traceable leads.
At the very least, a semantic field must comprise stable (that is, repeated)
distinctions of the type mentioned above. Moreover, studies of the specific
historical tension between the two sides of the distinctions contribute to the
definition of conditions of particular semantic fields.
In Koselleck, defining an observation point concerns the designation of the

semantic field to which the history of concepts and the field analysis are
linked. As in Foucault’s discourse analysis, the semantics to which a concept
belongs is not a given: one word can indicate various concepts depending on
the reference to a semantic field. A conceptual history of the concept of
‘child’ extends in many different directions and has many different sources,
depending on whether the semantics of education, law or love function as the
point of observation.
Laclau
In Laclau we see the emergence of a number of problems of conditioning,
which, in some cases, take on more serious dimensions than in the analyses of
Foucault and Koselleck. Generally, we might regret that Laclau has abstained
from empirical work, since this kind of experience might have influenced his
theoretical architecture. However, no efforts have been made towards the
conditioning of the apparatus for observation, which means that what appears
coherent and consistent within the theory is transformed into rather feeble
metaphors on confrontation with an actual object. This applies to the concepts
of discourse, and floating and empty signifier. The concept of empty signifier
is particularly anti-empirical.
Hegemonic analysis
The first problem of conditioning directly concerns the guiding distinction

discoursivity/discourse. On the one hand, Laclau’s formulation of a general
108
discourse theory as opposed to Foucault’s specific knowledge archaeology

implies that it is considerably more open concerning the question of which
discourses can be made the object of a discourse analysis. This is a definite
strength. On the other hand, however, it offers no guidelines for when we can
speak of discourse. The only general measure for such a definition is that a
partial fixation of the discourse must occur, but nothing general can be said
about when this happens. Any hegemonic analysis thus faces the problem of
conditioning the measures for partial fixation.
This is not to be read as a fundamental criticism of Laclau’s hegemoic
analysis but as a demonstration of the fact that the analyst will not find much
guidance for such an operation in Laclau. I have no reservations as to its
potency and ability for conditioning the guiding distinction discoursivity/
discourse, although I wish to point out that discourse theory as such does not
invite such an endeavour. Laclau himself does not indicate this as a problem.
Moreover, one might consider whether the assertion of discourse in general
does not oppose any attempt to historise (and hence condition) the way that
discourse can appear as discourse in different contexts and at different times.
Nevertheless, I believe there is one possible extension of the conditioning of
the guiding distinction discoursivity/discourse in the combination of
hegemonic analysis with deconstruction, in which the logicity of dualitites
can perhaps provide conditioning measures for discourse by constituting the
guiding principles for hegemonic analysis. A discourse of power is thus
conditioned by occurring across the distinction power/freedom, and a political
discourse by occurring across the distinction particularity/universality and so
on.
Whereas it is thus possible, in my opinion, to define measures for partial
fixation, I believe that one comes up against fundamental difficulties in relation
to a concept such as the empty signifier. The theoretical function of the
empty signifier is an indication of the boundaries of meaning (not of one
meaning but of ‘meaning’ as such), and the boundaries of meaning cannot be
signified – not even by the discourse analyst. Whereas the empty signifier (of
which I am very fond as a theoretical concept) sustains a significant theoretical
function, it has no empirical analytical potency. It evades any attempt at
conditioning. It is, of course, possible to look for concepts that are ‘reasonably’
empty (such as ecological sustainability or local community) but then the
theoretical status of the concept is lost. It is either empty or not empty. It
cannot be half empty, since it does not then indicate the boundaries of meaning
but at most the floating character of meaning. This means that the empty
signifier, as a result, becomes synonymous with the floating signifier, which is
also an empirically potent concept.
Why is it that the empty signifier does not allow for conditioning? In fact,
this is not strange at all. The empty signifier does not allow for conditioning
because, by defining the boundary of meaning, it simultaneously constitutes
the boundary of the discourse analyst insofar as he works within meaning.
The empty signifier designates the space from which the discourse is analysed
and is therefore unable to be indicated by the discourse analysis itself.
109
Deconstructivist analysis
The question of conditioning obtains a peculiar role in relation to

deconstruction. Deconstruction is precisely not an analysis in the sense of a
reductive operation, which means that there is no analysis that needs
conditioning. Of course, one can maintain that the choice of duality represents
a significant choice. It seems reasonable to ask: Why deconstruction of this
particular distinction? Why, for one thing, is freedom defined in opposition to
power? Why tolerance in opposition to intolerance? Which other oppositions
could be deconstructed? But the answer to the question of why it is this
particular distinction can only be given in hegemonic analysis. In the analysis
of hegemony where the duality is transformed into a guiding principle, the
way, scope and effects by which a specific distinction manifests itself become
evident. It is not until this point that the question of the duality and the
generality, and the historical-empirical status of its associated logic, is
conditioned.
In Laclau the choice of observation point concerns, as in Foucault, the choice

of discursive reference. Moreover, the choice of a specific duality and logic
affects the point of observation. The ‘same’ discursive formation appears
significantly different depending on the deconstructed duality one defines as
guiding principle, that is, which logic the hegemonies are seen as constellations
of. The movement discoursivity discourse happens differently in,
for example, the logic of signification, the logic of representation and the
logic of universalisation. If the chosen guiding principle is the logic of
representation, the movement appears as a description of how different
relationships between representative and represented are articulated and fixed.
If, on the other hand, the guiding principle is the logic of universalisation,
hegemonic analysis becomes a description of how the relationship between
the universal and the particular is articulated and fixed, including how particular
positions are sought, universalised and granted the authority to define the
boundaries for the appearance of particularity.
Luhmann
Form analysis
The problem of conditioning in Luhmann’s form analysis has the same qualities
as the problem of conditioning in relation to deconstruction in Laclau. As we
have seen, form analysis concerns the unity of the distinction, hence, the
primary task is to recognise which distinction is the object of the form analysis.
The first difficulty is therefore in a motivation of the other side of the distinction
in relation to the indication one seeks to observe.
110
For example...
If one focuses on risk, the question is:
• What does the other side of risk consist of?
In his form analysis of risk communication, Luhmann chooses to replace the

opposition risk/security by the opposition risk/danger. The question is therefore:
• Why is danger the other side of the distinction to risk?
As was the case with deconstruction, the question cannot be answered until
form analysis functions as input for another analysis, for example a semantic
analysis or a systems analysis, in which the unfolding of the paradoxical character
of form can be studied. Only in studies of the de-paradoxification of paradoxes
will it become evident to what degree and with what implications a form
manifests itself. Only through de-paradoxification can we follow form in its
chains of distinctions and only then can we determine to what extent
communication follows a specific form.
Systems analysis
I have already touched upon the problem of conditioning in systems analysis

where the guiding distinction is system/environment. The questions are:
• When can we say that communication forms an autopoietical system?

• When is the system identical with itself and when has it developed into a
new system?
Differentiation analysis
In Luhmann’s differentiation analysis, the problem of conditioning pertains to

the measures of similarity in a distinction to be pinpointed as a form of
differentiation.
• When are systems equal in their way of defining themselves in relation to

the environment?
• Moreover, what allows for the description of a specific form of differentiation
to draw conclusions in respect of the possibilities of continued systems
formation?
Evidently, all social systems do not adhere to the same form of differentiation.
Luhmann claims that functional differentiation is presently the primary form
of differentiation, but this does not mean that systems adhering to a different
form of differentiation are not continually formed (for example, biker gangs
111
that define themselves in accordance with codes of honour, or systems that

regulate membership based on ethnicity, or family systems that still present
themselves as patriarchal dynasties). The enunciative power depends entirely
on how demanding one conditions the analysis.
Media analysis
Finally, the problem of conditioning in Luhmann’s media analysis concerns

the measures for the designation of a new medium. As previously noted, a
medium is always a medium in relation to a form, but the tension between
form and medium varies historically. Consequently, we cannot presuppose
deductive measures for media, since this would exclude from the analysis the
historical character of the media. For example, the ways in which language,
radio, television and money create media are very different. Gradually, Luhmann
has constructed a distinction between types of media:
• language;
• media of distribution;
• general symbolic media.
Evidently, each of the types holds definitional measures but, at the same time,
each medium typically exceeds the general definition. The media never
completely function as media in exactly the same way, and it is therefore
critical to the empirical sensitivity of the analysis that the conditioning possesses
an inductive rather than a deductive character. Parallel to Foucault’s discourse
analyses, we cannot predefine a final conditioning of the point at which a
medium can be regarded as a medium. The conditioning happens as part of
the analysis in which the demands on the chosen medium are gradually
intensified, and in which this evolution makes the object appear increasingly
clear and causes the analysis to heighten its sensitivity to empirical details.
The analysis can be seen as completed once one has reached a delimitation of
the characteristics of the medium as medium. As with archaeological discourse
analysis, the object (that is, the specific discursive formation) cannot therefore
be presupposed. On the contrary, the object is the result of the analysis,
including the measures for the assertion of the object.
Semantic analysis
The problem of conditioning in Luhmann’s semantic analysis has the same

definitions as in Koselleck’s history of concepts in as much as Luhmann has
found considerable inspiration here. I will therefore not elaborate any further
on this point, except to say that the fundamental question is:
• When is meaning condensed and when does a semantic rupture exist?
112
The question of choice of observation point in systems theory always involves,

as a minimum, a choice of systems reference. Any communication can exist as
an event in numerous systems of communication. At the very least, one must
indicate whether the point of reference of the analysis is a system of interaction,
an organisational system or a social system. However, as we have seen in
Chapter Four on Luhmann, systems have the capacity to differentiate themselves
into sub-systems where the sub-systems define their own environment and
where the ‘master’-system is observed as the environment of the sub-system.
It is always possible to shift the point of observation and, in so doing, the
observed communications emerge in new ways. It is always possible to move
back in the systems differentiation so that what used to be system now appears
as sub-system, or forward in the differentiation so that what used to be system
now appears as environment. Regardless of the analytical strategy one connects
with, one cannot evade, in Luhmann’s systems theory, the indication of a
systems reference. In this way, the guiding distinction system/environment
obtains a special status in relation to the other guiding distinctions, as it
somehow arranges the relationships between the analyses.
In addition to the choice of systems reference, other questions regarding the
definition of observation point can be formulated depending on the analytical
strategy.
Combining analytical strategies

I have tried to describe Foucault, Koselleck, Laclau and Luhmann as analytical
strategists. I have grouped their writings under the heading of one umbrella by
defining their work as different theoretical programmes for the observation of
the second order. I have suggested and prescribed to them a particular notion
of analytical strategy based on a distinction between guiding distinction,
conditioning and point of observation. Now the question is this: Are we free
to cross-combine these analytical strategies? It is one thing to compare and to
allow for the programmes to reflect each other; it is something else to observe
the programmes as a pool of analytical strategies between which one is reasonably
free to choose and across which elements can be transferred and combined.
I do not have the ultimate answer to this question. Fundamentally, however,
we should remind ourselves that the four programmes have themselves been
pieced together and are made up of fragments from very contradictory theories.
Considering the fact that Luhmann is able to fuse elements from theories as
distinct as Bateson’s theory of information, Husserl’s phenomenology, Parson’s
systems theory and Spencer-Brown’s form calculus, why should we not feel
free to import elements from Laclau’s hegemonic analysis into Luhmann’s analysis
of semantics? When Laclau takes the liberty of bringing together elements
from Foucault’s discourse analysis with elements from Lacan’s psychoanalysis,
despite Foucault’s dislike for Lacan’s work, why should we not consider the
combinatory constellation of systems analysis and hegemonic analysis?
113
Table 5.2: Problems of conditioning related to the analytical strategies

Analytical strategy Problems of conditioning Fixation of point of observation
Foucault
Archaeological discourse When can we speak of ‘regularity’? Which discursive problem is pursued?
analysis
When can a statement be What is the discursive reference?
regarded as a statement?
Genealogy What are the measures for the What is the discursive reference?
identification of continuity and
discontinuity respectively?
Self-technology analysis When can we speak of What is the discursive reference?
subjectivation?
When can something be regarded What is the discursive reference?
as technology?
Dispositive analysis When are the elements sufficiently What is the discursive reference?
connected for them to be called
an apparatus?
When is a schematic a general and What is the discursive reference?
strategic logic?
Koselleck
History of concepts When is meaning condensed? Which semantic field is referred to?
Semantic field analysis When does a semantic field possess What is the discursive reference?
the qualities of a field?
Laclau
Hegemonic analysis Which conditions must be met for What is the discursive reference?
someone to speak of discourse?
Which logic is defined as guiding principle?
Deconstructivist analysis Why this particular distinction? Which discursive problem is pursued?
Luhmann
Form analysis Why this particular distinction? What is the systems reference?
Why this particular ‘other side’ Which form is defined as the guiding principle?
of the distinction?
Differentiation analysis What are the measures for a Which form is defined as the guiding principle?
similarity in a distinction to be
designated as a form of differentiation?
Semantic analysis When is meaning condensed? Which form is defined as the guiding principle?
When does a semantic rupture exist?
Media analysis What are the measures for the Which form is defined as the guiding principle?
recognition of a medium?
114
Luhmann and Foucault have both posited the non-existence of a ‘better’

strategy. Foucault has consciously striven not to form a school – from Foucault’s
perspective, ‘Foucault’ is not a valid argument for a particular analytical strategy.
In the words of Luhmann:‘A choice of guiding distinction is precisely a choice!’
There is no transcendental space from which a specific choice of guiding
distinction can be substantiated in opposition to a different choice. Moreover,
the empirical also does not ask to be observed in any particular way.
In the end, I believe that one should not let the works impede the
development of new analytical strategies. Basically, the guiding distinction is
what guides the observation, not Foucault, Laclau, Koselleck or Luhmann.
What is important, therefore, are the existing and potential relationships between
the analytical strategies one might choose to bring together and the thorough
reflection of how the choice of guiding distinction defines the measures for
other concepts and for their application. This includes the introduction of
any ‘foreign’ concepts.
Evidently, if one does decide to combine (which I have done on several
occasions, for example in Andersen, 1995), certain problems occur.
First, it is not possible to simultaneously employ more than one guiding
distinction without causing the perspective to become unstable and to oscillate
uncontrollably between the different guiding distinctions. One analysis cannot
be linked to both the guiding distinction discoursivity/discourse as well as
system/environment. One might use the two guiding distinctions to define a
combinatory analytical strategy, in which one contemplates two separate
analyses that somehow supplement each other. This requires an account of
the complementarity of the two strategies and the ability to tailor the concepts
in such a way that they are precisely complementary and not competing and
conflicting.
Second, if one imports elements from one analytical strategy into another,
these elements need to be redefined in respect to the guiding distinction. If,
for example, Laclau’s ‘nodal point’ is brought into Luhmann’s semantic analysis,
it must be decided which specific analytical strategic status the concept might
obtain there (if any!). One must answer the question: Which function does
the concept serve in the overall analytical strategy?
I personally have found it a natural progression to introduce elements of
Foucault’s archaeology into a Luhmannian historical-semantic analysis. In
fact, I have simply used a number of Foucault’s discourse-analytical variables
in order to operationalise the fact dimension in Luhmann’s semantics (Andersen
and Born, 2000).
Third, the combination of analytical strategies affects the separate analytical
strategies, for example, by demands on their conditioning. When semantic
analysis is combined with differentiation analysis in Luhmann, it is required
that the conditioning of the two analytical strategies be congenial, so that the
correlation between a rupture in the semantic evolution can be related to
displacements in the form of differentiation. Evidently, the difficulties of a
complementary linking of different analytical strategies do not diminish when
working across the different programmes of observation.
115
For example...
The combination of Foucault’s archaeological discourse analysis with Luhmann’s

systems analysis implicates the problem that Foucault does not employ a distinction
between system and semantics. On the contrary, Foucault’s analyses sometimes
operate almost as systems possessing a self, for example in descriptions of
discursive conflicts. At the very least, a combination of these two analytical
strategies must involve a solution to this problem, for example by weeding out
the systemic elements of the concept of discourse in order to avoid the
simultaneous application of two different concepts of system. Of course, one
might object that it is then no longer a Foucauldian knowledge archaeology but,
the question is whether that difference makes a difference.
Future questions
A second-order perspective does not form a contrast to empirical imminence.

Stepping back and inquiring about the categories and not from the categories,
about the semantics and discourses, and not from the semantics and discourses,
about the systems and not from the systems does not mean a decreased level of
concretion or a greater distance to empirical matters. On the contrary, the
indiscriminate use of preconceived categories and methods of first-order
observations creates immunity from the empirical. Only through the discussion
of the way we discuss are we able to regain sensitivity to the empirical and
avoid automatic social descriptions such as, for example, the aforementioned
discourse on the restriction of the freedom of nations in relation to European
integration (in the Introduction). However, as demonstrated above, a wealth
of analytical-strategic difficulties follow from the shift from a first-order
perspective to a second-order perspective where observation is observed as
observations. Only through a dedicated approach to the analytical-strategic
problems is it possible for a second-order perspective to make the empirical
pose a challenge to the observation.
In this book, I have tried to encourage an analytical-strategic discussion
across a range of different epistemologies based on the idea that they might at
least inform each other in respect to the elaboration of analytical-strategic
problems.
As regards Laclau and Luhmann at least, there exists no tradition for analytical-
strategic discussions. Laclau has never undertaken any empirical work and,
although Luhmann has conducted plenty of empirical work, his students appear
to be primarily interested in his theoretical endeavours. It is my opinion,
however, that there is no point in developing discourse-analytical manifestos
or systems-theoretical concepts without directing these to an object. A second-
order observation is not worth attention unless it is an observation of specific
empirical observations. Within the environments that refer to Foucault,
Luhmann, Laclau and others, we find an anti-methodological sentiment. This
116
Figure 5.1: Analytical strategy
Choice of guiding distinction
Construction of Construction of measures

object for valid arguments
Observations The second-order

observed observer
Specification of point Specification of rules

of observation of observation
Choice of conditioning
is completely understandable. What I am trying to point out is that method is

not the only way in which one can approach the relationship between theory
and empiricism. It is possible to install analytical strategy in the place of
method in an epistemologically founded second-order observation, and hence
discuss problems of operationalisation, without finding oneself with
ontologising rules of method.
In Figure 5.1 I have tried to illustrate my suggestions for a concept of
analytical strategy.
The choice of guiding distinction divides the world into the second-order
observer and his object, the observed observation. Neither observer nor object
is predefined. They are both construed through the choice of guiding
distinction. They are internal analytical-strategic constructions. The guiding
distinction defines the measures for anything to be observable as object while
simultaneously constructing the criteria, which the observer must employ in
argumentation and formation of concepts. The definition of a good and valid
argument, thus, is an internal analytical-strategic construction.
Subsequently, the choice of conditioning is the specification of object and
observation respectively. Throughout this book, the former has been designated
‘the specification of point of observation’, that is, the question of which specific
discourse, system or observation the second-order observer lays claim to in his
observation. The latter is that which I have generally termed ‘conditioning of
guiding distinction’, which concerns the specification of rules for observation
as regards the indication of one or the other side of the guiding distinction
(for example, the specification of when a discourse is a discourse).
Consequently, one cannot refer to any external authority when dealing
with analytical-strategic questions. In this book, I have posed what I consider
some of the key analytical-strategic questions. However, other questions can
be asked, concerning for example sources, text and text analysis, and the
integration of traditional methods in epistemological analytical strategies. I
suppose my claim is that, no matter which question one asks, it must possess
the possibility of reformulation within the concept of analytical strategy stated
above.
117
In conclusion, I will provide, as an example, the question of sources.
For example...
• What is a ‘source’ in an epistemological observation and hence what are

the conditions of criticism of the sources?
This question entails two analytical-strategic questions. First, the construction of

sources is a question regarding the choice of guiding distinction and, second, criticism
of sources becomes a question of conditioning. What is regarded as a source in an
epistemological observation is naturally an internal construction in the analytical
strategy in the same way that the object is a construction. There are no predefined
sources ‘out there’. Sources are constructed as a tool in an analytical strategy to
obtain empirical sensitivity. Thus, the construction of something as a source is an
analytical-strategic choice concerning the empirical sensitivity one wishes to attain
in relation to the internally constructed object.
Subsequently, criticism of sources becomes a question of the specification of the

rules for observation. One example could be Foucault’s knowledge-archaeological
strategy with the guiding distinction regularity/dispersion of statements. Based on
this guiding distinction, Foucault constructs sources as monuments in a discourse.
A discourse-analytical monument is a pivot in discourse, something that stands
out and indicates a difference in the discursive regularity. As monument, the source
is therefore anything but a document. It documents nothing. In Foucault, the
monument is not a source of information. The source constructed as monument
is simply a discursive injection. Foucault has not explicitly developed a ‘criticism of
monuments’, but such a criticism would consist in the specification of rules for
when a monument can be regarded as monument, that is, when we can say that a
specific text or image embodies the injection of a specific discourse (Andersen,
1994).
118
References
Althusser, L. and Balibar, E. (1970) Reading capital, London: NLB.

Andersen, N.Å. (1994) ‘Institutionel historie’, COS-rapport, Copenhagen:
Copenhagen Business School.
Andersen, N.Å. (1995) Selvskabt forvaltning, Copenhagen: Nyt fra
Samfundsvidenskaberne.
Andersen, N.Å. (1996) Udlicitering – når det private bliver politisk, Copenhagen:
Nyt fra Samfundsvidenskaberne.
Andersen, N.Å. (1997) Udlicitering – strategi og historie, Copenhagen: Nyt fra
Samfundsvidenskaberne.
Andersen, N.Å. and Born, A. (2000) ‘Complexity and change: two “semantic
tricks” in the triumphant oscillating organization’, System Practice and Action
Research, vol 13, no 3.
Andersen, N.Å.and Born, A. (2001) Kærlighed og omstilling – italesættelsen af den
offentligt ansatte, Copenhagen: Nyt fra Samfundsvidenskaberne.
Balibar, E. (1994) ‘Subjection and subjectivation’, in J. Copjec (ed) Supposing
the subject, London: Verso.
Baraldi, C., Corsi G. and Esposito, E. (1997) GLU, Glossar zu Niklas Luhmanns
Theorie sozialer Systeme, Frankfurt am Main: Suhrkamp.
Barthes, R. (1990) S/Z, Oxford: Blackwell Publishers.
Beronius, M. (1991) Genealogi och Sociologi, Kunghult: Brutus Östlings Bokförlag
Symposion AB.
Bornmann, L. (1998) ‘Glossar zu Niklas Luhmanns Theorie sozialer Systeme’,
at http://www.uni-kassel.de/~bornmann/luhmann.htm.
Bourdieu, P. (1992) Language and symbolic power, Cambridge: Polity Press.
Brenner, N. (1994) ‘Foucault’s new functionalism’, Theory and Society, no 23,
pp 679-709.
Brunner, O., Conze, W. and Koselleck, R. (1972) Geschichtliche Grundbegriffe,
historisches Lexikon zur politisch-sozialen Sprache in Deutschland, vol 1, Stuttgart:
Klett-Gotta.
Chouliaraki, L. and Fairclough, N. (1999) Discourse in late modernity: Rethinking
critical discourse analysis, Edinburgh: Edinburgh University Press.
Dahlager, L. (2001) ‘I forebyggelsen magt’, Distinktion, no 3, pp 91-102.
Davidson,A.I. (1986) ‘Archaeology, genealogy, ethics’, in D.C. Hoy (ed) Foucault:
A critical reader, Oxford: Basil Blackwell.
de Certeau, M. (1975) Spor af historien, Copenhagen: Vinten.
de Saussure, F. (1990) Course in general linguistics, London: Duckworth.
Dean, M. (1994) ‘A social structure of many souls: moral regulation, government
and self-formation’, The Canadian Journal of Sociology, vol 19, no 2, pp 145-
68.
119
Dean, M. (1998) ‘Administrativ Ascetism’, in M. Dean and B. Hindess, Governing

Australia, Cambridge: Cambridge University Press.
Dean, M. (1999) Governmentality, London: Sage Publications.
Deleuze, G. (1992) ‘What is a dispositif?’, in T.J. Armstrong (ed) Michel Foucault
philosopher, London: Harvester Wheatsheaf.
Derrida, J. (1976) ‘Structure, sign and play in the discourse of the human
sciences’, in J. Derrida, Writing and difference, Chicago, IL: University of
Chicago Press.
Derrida, J. (1980) Of grammatology, Baltimore, NJ: Johns Hopkins University
Press.
Derrida, J. (1988) ‘Letter to a Japanese friend’, in D. Wood and R. Bernasconi
(eds) Derrida and différance, Chicago, IL: Northwestern University Press.
Descombes, V. (1980) Modern French philosophy, Cambridge: Cambridge
University Press.
Esposito, E. (1996) ‘From self-reference to autology: how to operationalize a
circular approach’, Social Science Information, pp 269-81.
Fairclough, N. (1992) Discourse and social change, Cambridge: Polity Press.
Fairclough, N. (1995) Critical discourse analysis, London: Longman.
Foucault, M. (1970) ‘Diskurs og Diskontinuitet’ in P. Madsen (ed) Strukturalisme,
Copenhagen: Rhodos, p 152 (the article is a translation into Danish of
sections 1 and 2 of Foucault, M. [1968] ‘Réponse au cercle d’épistémologie’,
Cahiers pour l’analyse, vol 9).
Foucault, M. (1971) Madness and civilization, London: Routledge.
Foucault, M. (1972) ‘The discourse on language’, in M. Foucault, The archaeology
of knowledge, New York, NY: Pantheon Books.
Foucault, M. (1974) The order of things, London: Routledge.
Foucault, M. (1976) Mental illness and psychology, Berkeley and Los Angeles,
CA: University of California Press.
Foucault, M. (1977) Discipline and punish, London: Penguin.
Foucault, M. (1978) The history of sexuality, Volume 1, London: Penguin.
Foucault, M. (1980) ‘The confession of the flesh: a conversation’, in C. Gordon
(ed) Power/knowledge, selected interviews and other writings, Michel Foucault, New
York, NY: Pantheon Books.
Foucault, M. (1986a) The archaeology of knowledge, London: Tavistock.
Foucault, M. (1986b) The birth of the clinic, London: Routledge.
Foucault, M. (1991) ‘Nietzsche, genealogy, history’, in P. Rabinow (ed) The
Foucault reader, London: Penguin.
Foucault, M. (1997) Essential works of Foucault, 1959-1989, Volume 1: Ethics,
subjectivity of truth, New York, NY: New York Press.
Foucault, M. (1998a) Essential works of Foucault 1954-1984,Volume 2: Aesthetics,
method, and epistemology, New York, NY: New York Press.
Foucault, M. (1998b) ‘The political technology of individuals’, in L.H. Martin,
H. Gutman and P.H. Hutton (eds) Technologies of the self: A seminar with
Michel Foucault, Amhurst, MA: University of Massachusetts Press.
Günther, G. (1976) Beiträge zur Grundlegung einer operationsfähigen Dialectic,
Hamburg: Felix Meiner Verlag.
120
References
Habermas, J. (1972) Knowledge and human interests, London: Polity Press.

Habermas, J. (1991) Communication and the evolution of society, London: Polity
Press.
Habermas, J. (1992) Moral consciousness and communicative action, London: MIT
Press.
Habermas, J. (1996) Between facts and norms: Contribution to a discourse theory of
law and democracy, Cambridge: Polity Press.
Hacking, I. (1986) ‘Self-improvement’, in D.C. Hoy (eds) Foucault: A critical
reader, Oxford: Basil Blackwell.
Heidegger, M, (1978) Being and time, London: Blackwell Publishers.
Ifversen, J. (1997) ‘Om magt, demokrati og diskurs’, Begrebshistoriske studier No
2, Århus: Center for Kulturforskning, Århus Universitet.
Ifversen, J. and Østergård, U. (1996) ‘Europæisk Civilisation, begrebshistorie
og diskursanalyse’, in Østergård og Ifversen (ed) Begreb og historie, Århus:
Århus Universitet.
Kaufmann, L.H. (1987) ‘Self-reference and recursive forms’, Journal of Social
and Biological Structures, vol 10, pp 53-72.
Kent, C.A. (1986) ‘Michel Foucault: doing history, or undoing it?’, Canadian
Journal of History, pp 371-95.
Kneer, G. and Nassehi, A. (1993) Niklas Luhmanns Theorie sozialer Systeme,
München: Wilhem Fink Verlag.
Koselleck, R. (1972) ‘Einleitung’, in Brunner, Conze and R. Koselleck (eds)
Geschichtliche Grundbegriffe, Historisches Lexikon zur politisch-sozialen Sprache
in Deutschland, vol 1, Stuttgart: Klett-Gotta.
Koselleck, R. (1982) ‘Begriffsgeschichte and social history’, Economy and Society,
vol 11, no 4, pp 409-27.
Koselleck, R. (1985) Futures past: On the semantics of historical time, Cambridge,
MA: MIT Press.
Koselleck, R. (1987) ‘Historik und Hermeneutik’, in R. Koselleck and H.-G.
Gadamer (eds) Historik und Hemeneutik, Heidelberg: Carl Winter
Universitätsverlag.
Kosselleck, R. (1988) Critique and crisis, Oxford: Berg.
Koselleck, R. (1989) ‘Linguistic change and the history of events’, Journal of
Modern History, no 61, pp 649-66.
Krause, D. (1996) Luhmann-Lexikon, Stuttgart: Ferdinand Enke Verlag.
Krause-Jensen, E. (1978) Viden og magt, Copenhagen: Rhodos.
Kristeva, J. (1989) Language, the unknown, London: Prentice Hall/Harvester
Wheatsheaf.
Lacan, J. (1982) ‘God and the jouissance of the woman: a love letter’, in
J. Mitchell and J. Rose (eds) Feminine sexuality, London: Macmillan.
Lacan, J. (2001) Ecrits, London: Routledge.
Laclau, E. (1983) ‘Transformations of advanced industrial societies and the
theory of the subject’, in S. Haenninen and L. Palden (eds) Rethinking ideology,
Berlin: Argument Verlag, International General.
Laclau, E. (1985) ‘Hegemoni – en ny politisk logik’, interview in A. Andresen
(ed) Politisk strategi i firserne, Copenhagen: Aurora.
121
Laclau, E. (1992) ‘Universalism, particularism, and question of identity’, October,

no 61, pp 83-90.
Laclau, E. (1993a) ‘Power and representation’, in M. Poster (ed) Politics, theory
and contemporary culture, New York, NY: Columbia University Press.
Laclau, E. (1993b) ‘Discourse’, in R.E. Goodin and P. Pettit (eds) A companion
to contemporary political philosophy, Oxford: Blackwell.
Laclau, E. (1996a) Emancipation(s), London: Verso.
Laclau, E. (1996b) ‘Deconstruction, pragmatism, hegemony’, in C. Mouffe
(ed) Deconstruction and pragmatism, London: Routledge.
Laclau, E. and Mouffe, C. (1985) Hegemony and socialist strategy, London:Verso.
Lévi-Strauss, C. (1996) The savage mind, Oxford: Oxford University Press.
Luhmann, N. (1982) ‘World–time and system history’, in N. Luhmann, The
differentiation of society, New York, NY: Columbia University Press.
Luhmann, N. (1985) ‘Complexity and meaning’, in I. Prigogine, M. Zeleny
and E. Morin (eds) The science and praxis of complexity: Contributions to the
symposium held at Montpellier, France, May 9-11, 1984,Tokyo: United Nations
University.
Luhmann, N. (1986a) ‘The theory of social systems and its epistemology:
reply to Danilo Zolo’s critical comments’, Philosophy and Social Science, vol
16, pp 129-34.
Luhmann, N. (1986b) ‘Kunstens medie’, in N. Luhmann, Iagttagelse og paradoks,
Copenhagen: Gyldendal.
Luhmann, N. (1988a) Erkenntnis als Konstruktion, Bern: Benteli Verlag.
Luhmann, N. (1988b) ‘Frauen, Männer und George Spencer-Brown’, Zeitschrift
für Soziologie, vol 17, no 1, pp 47-117.
Luhmann, N. (1990a) ‘The paradox of system differentiation and the evolution
of society’ in J. Alexander and P. Colomy (eds) Differentiation theory and social
change, New York, NY: Columbia University Press.
Luhmann, N. (1990b) Essays on self-reference, New York, NY: Columbia
University Press.
Luhmann, N. (1990c) ‘The cognitive program of constructivism and a reality
that remains unknown’, in W. Krohn, G. Küppers and H. Nowotny (eds)
Selforganization: Portrait of a scientific revolution, New York, NY: Kluwer
Academic Publishers.
Luhmann, N. (1990d) ‘Risiko und Gefahr’, in N. Luhmann, Soziologische
Aufklarung 5, Wiesbaden: Westdeutsche Verlag.
Luhmann, N. (1991) ‘Am Ende der kritischen Soziologie’, Zeitschrift für
Soziologie, vol 20, pp 147-52.
Luhmann, N. (1993a) Risk:A sociological theory, NewYork, NY:Walter de Gruyter.
Luhmann, N. (1993b) ‘Deconstruction as second-order observing’, New Literary
History, vol 24, pp 763-82.
Luhmann, N. (1993c) Gesellschaftstruktur und Semantik Band – Studien zur
Wissenssoziologie der moderne Gesellschaft 3, Frankfurt am Main: Suhrkamp.
Luhmann, N. (1994a) ‘“What is the case?” and “What lies behind it”: the two
sociologies and the theory of society’, Sociological Theory, vol 12, no 2, pp
126-39.
122
References
Luhmann, N. (1994b) ‘Speaking and silence’, New German Critique, vol 61,
pp 25-37.
Luhmann, N. (1995a) ‘The paradoxy of observing system’, Cultural Critique,
no 31, pp 37-55.
Luhmann, N. (1995b) ‘Subjektets nykker og spørgsmålet om mennesket’, in
J.C. Jacobsen (ed) Autopoiesis II, Copenhagen: Politisk Revy, p 46.
Luhmann, N. (1995c) Social systems, Stanford, CA: Stanford University Press.
Luhmann, N. (1996) ‘On the scientific context of the concept of
communication’, Social Science Information.
Macey, D. (1988) Lacan in context, London: Verso.
Mahon, M. (1992) Foucault’s Nietzschean genealogy, Albany, NY: State University
of New York Press.
Megill, A. (1979) ‘Foucault, structuralism, and the end of history’, Journal of
Modern History, vol 51, no 3, pp 451-503.
Nietzsche, F. (1988) On the advantage and disadvantage of history for life, Indianapolis,
IL: Hackett Publishing.
Nietzsche, F. (1998) On the genealogy of morality, Indianapolis, IL: Hackett
Publishing.
Noujain, E.G. (1987) ‘History as genealogy: an exploration of Foucault’s
approach to history’, in A.P. Griffiths (ed) Contemporary French philosophy,
New York, NY: Cambridge University Press, pp 157-74.
Pedersen, O.K. (1983) Videnskabsproblemet, Copenhagen: Aurora.
Pocock, J.G.A. (1987) ‘The concepts of a language and the métier d’historien:
some considerations on practice’, in A. Pagden (ed) The languages of political
theory in early-modern Europe, Cambridge: Cambridge University Press.
Raffnsøe, S. (1999) ‘Historie- eller diskursanalyse? En introduction til Foucault
Les mots et les choses og L’archéologie du savoir’, COS-report 4/99, Copenhagen:
Copenhagen Business School.
Raffnsøe, S. (2000) ‘Michel Foucault’s dispositionelle magtanalytik’, Grus.
Regnault, F. (1983) ‘Hvad er et epistemologisk brud?’, in S.G. Olesen (ed)
Epistemologi, Copenhagen: Rhodos.
Richter, M. (1990) ‘Reconstructing the history of political languages: Pocock,
Skinner, and the geschichtliche Grundbegriffe’, History and Theory, vol 29,
pp 38-70.
Richter, M. (1995) The history of political and social concepts, New York, NY:
Oxford University Press.
Rose, N. (1996) Intenting ourselves: Psychology, power and personhood, Cambridge:
Cambridge University Press.
Roth, M.S. (1981) ‘Foucault’s “History of the present”’, History and Theory, vol
20, pp 32-46.
Röttgers, K. (1982) ‘Kritik’, in O. Brunner, W. Conze and R. Kosseleck (eds)
Geschichtiche Grundbegriffe, Historisches Lexikon zur politisch-sozialen Sprache in
Deutschland, Vol III, Stuttgart: Klett-Cotta.
Scharff, R. (1974) ‘Nietzsche and the “use” of history’, Man and World, An
International Philosophical Review, vol 7, p 74.
Schmidt, L.H. (1990) Det sociale selv, Århus: Århus Universitetsforlag.
123
Schmidt, L.H. and Kristensen J.E. (eds) (1985) Foucaults Blik, Århus: Modtryk.
Shiner, L. (1982) ‘Reading Foucault: anti-method and the genealogy of power–
knowledge’, History and Theory, vol 21, pp 382-98.
Skinner, Q. (1987) Machiavelli, Oxford: Oxford University Press.
Spencer-Brown, G. (1969) Laws of gorm, London: George Allen & Unwin.
Stahäli, U. (1998) ‘Die Nachträglichkeit der Semantik zum Verhältnis von
Sozialstruktur und Semantik’, Soziale Systeme, Zeitschrift für soziologische Theorie,
no 2, pp 315-40.
Stråht, B. (ed) (1990) Language and the construction of class identity, Göteborg:
Department of History, Göteborg University.
Tully, J. (1988) Meaning and context: Quentin Skinner and his critics, London:
Polity Press.
von Foerster, J. (1981) Observing systems, Seaside, CA: Intersystems Publication.
124
Appendix A: Examples of other
analytical strategies
Analytical strategy Guiding distinction Question Goal of the analysis
Pierre Bourdieu
Field analysis Community/disagreement How are fields created as To demonstrate the
communities of disagreement correlation between
and how do they distribute a history, practice and
network of relative subject power
positions?
Analysis of Authorising/speech How are individuals authorised
communicative position by institutions so that they not
competence only speak from a position of
meaning but also with weight?
How can individuals collect social
and cultural capital that they can
actualise communicatively in
specific fields?
Jürgen Habermas
Discourse analysis Discursive practice/ethics In what way does practice Emancipation and
of discourse deviate from the ethics discourse-ethical
of discourse? communication
Bruno Latour
Translation analysis Translation/association How are ideas, technologies, To demonstrate the
and practices propagated by growth of power
being associated with other ideas, through distribution
technologies, and practices etc, and practice in a
so that they form a network? network
In what way does propagation
through association simultaneously
mean a translation of the
propagated phenomena so that
phenomena constantly change
character and value through
their propagation?
Louis Althusser
Symptomatic reading The invisible in the visible Which questions are posed To analyse the imaginary
by the text? order which regulates
people’s relationship to
Which answers does the text their condition of
offer to the questions it poses? existence with a view
to politicisation
Which questions does the text
pose without answering them?
Which answers does the text offer
to questions it does not pose?
In sum, what is the prohibited and
displaced question that guides and
regulates the text and its visibility?
125
Appendix B: Further reading
Examples of empirical analyses in the spirit of

Foucault, Koselleck, Laclau and Luhmann
Foucault inspired analysis
Amariglio, J. (1990) ‘Economics as postmodern discourse’, in W.J. Samuels

(ed) Economics as discourse, London: Kluwer.
Barnes, T.J. and Duncan, J.S. (1992) Writing worlds: Discourse, text and metaphor
in the representation of landscape, London: Routledge.
Chiapello, E. and Fairclough, N. (2002) ‘Understanding the new management
ideology: a transdisciplinary contribution from critical discourse analysis
and new sociology of capitalism’, Discourse & Society, vol 13, no 2, pp 185-
208.
Cruikshank, B. (1999) The will to empower, Ithaca, NJ: Cornell University Press.
Dean, M. (1992) ‘A genealogy of the government of poverty’, Economy &
Society, vol 21, no 3, pp 215-51.
Dean, M. (1994) ‘“A social structure of many souls”: moral regulation,
government and self-formation’, Canadian Journal of Sociology, vol 10, no 2,
pp 145-68.
Dean, M. (1995) ‘Governing the unemployed self in an active society’, Economy
& Society, vol 24, no 4, pp 559-83.
Dean, M. (1998) ‘Administrating asceticism, reworking the ethical life of the
unemployed citizen’, in M. Dean and B. Hindess (ed) Governing Australia,
Cambridge: Cambridge University Press.
Donzelot, J. (1988) ‘The promotion of the social’, History of the Present, vol 3,
pp 5-15.
Donzelot, J. (1997) The policing of families, Baltimore, MD: Johns Hopkins
University Press.
Eder, K. (1996) The social construction of nature, London: Sage Publications.
Farmer, J. (1995) The language of public administration, University of Alabama
Press.
Foucault, M. (ed) (1982) Pierre Riviere, having slaughtered my mother, my sister, and
my brother, London: University of Nebraska Press.
Foucault, M. (1974) The order of things: An archaeology of the human sciences,
London: Routledge.
Foucault, M. (1976): The birth of the clinic: An archaeology of medical perception,
London: Routledge.
127
Foucault, M. (1987) Mental illness and psychology, Berkeley, CA: University of

California Press.
Foucault, M. (1988) The history of sexuality Vol 2: The use of pleasure,
Harmondsworth: Penguin.
Foucault, M. (1990) The history of sexuality Vol 3: The care of the self,
Foucault, M. (1991) Discipline and punish, Harmondsworth: Penguin.
Foucault, M. (1998) The history of sexuality Vol 1: The will to knowledge,
Foucault, M. (2001) Madness and civilization, London: Routledge.
Hacking, I. (1986) ‘Making up people’, in T.C. Heller, M. Sosna and D.E.
Wellbery (eds) Reconstructing individualism, Stanford, CA: Stanford University
Press.
Hoskin, K. (1992) ‘Control, organization, and accounting: a genealogy of
modern knowledge-power’, System Practice, vol 5, no 4, pp 425-39.
Laqueur, T. (1994) Making sex, Harvard, MA: Harvard University Press.
Lemke, L.J. (1995) Textual politics: Discourse and social dynamics, London: Taylor
& Francis.
Lupton, D. (1995) The imperative of health-public health and the regulated body,
London: Sage Publications.
Miller, P. and Napier, C. (1993) ‘Genealogies of calculation’, Accounting,
Organizations and Society, vol 18, nos 7/8, pp 631-47.
Miller, P. and O’Leary, T. (1987) ‘Accounting and the construction of the
governable person’, Accounting, Organizations and Society, vol 12, no 3, pp
235-65.
Miller, P. and Rose, N. (1990) ‘Governing economic life’, Economy & Society,
vol 19, no 1, pp 1-31.
Miller, P. and Rose, N. (1995) ‘Production, identity, and democracy’, Theory
and Society, vol 24, no 4, pp 427-67.
Mitchell, T. (1988) Colonising Egypt, Cambridge: Cambridge University Press.
Novas, C. and Rose, N. (2000) ‘Genetic risk and the birth of the somatic
individual’, Economy & Society, vol 29, no 4, pp 485-513.
Rose, N. (1985) The psychological complex: Psychology, politics and society in England
1869-1939, London: Routledge and Kegan Paul.
Wodak, R. (1996) Disorders of discourse, London: Longman.
Laclau inspired analysis
Cornell, D., Rosenfeld, M. and Carlson, D.G. (eds) (1992) Deconstruction and
the possibility of justice, New York, NY: Routledge.
Dalton, C. (1985) ‘An essay in the deconstruction of contract doctrine’, The
Yale Law Journal, vol 94, no 5, pp 999-1113.
Daly, G. (1991) ‘The discursive construction of economic space: logics of
organization and disorganization’, Economy & Society, vol 20, no 1, pp 79-
102.
128
Appendix B
Howarth, D. and Norval, A. (eds) (1998) South Africa in transition, London:

Macmillan.
Howarth, D., Norval, A.J. and Stavrakakis,Y. (2000) Discourse theory and political
analysis: Identities, hegemonies and social change, Manchester: Manchester
University Press.
Norval, A.J. (1996) Deconstructing apartheid discourse, London: Verso.
Smith, A.M. (1994) New right discourse on race and sexuality: Britain, 1968-1990,
Cambridge: Cambridge University Press.
Smith, A.M. (1998) Laclau and Mouffe: The radical democratic imaginary, New
York, NY: Routledge.
Luhmann inspired analysis
Andersen, N.Å. (2000) ‘Public market – political firms’, Acta Sociologica, no 1.

Andersen, N.Å. and Born, A. (2000) ‘Complexity and change: two “semantic
tricks” in the triumphant oscillating organization’, System Practice and Action
Research, vol 13, no 3.
Baecker, D. (ed) (1999) Problems of form, Stanford, CA: Stanford University
Press.
Baecker, D. (1992) ‘The writing of accounting’, Stanford Literature Review, vol
9, no 2, pp 157-78.
Graham, P. (1999) ‘Critical system theory’, Communications Research, vol 26,
no 4, pp 482-507.
Hernes,T. and Bakken,T. (ed) (2000) Niklas Luhmann’s autopoiesis and organization
theory, Copenhagen: Copenhagen Business School Press.
Jessop, B., Brenner, N., Jones, M. and Macleod, G. (eds) (2002) State/space,
Oxford: Blackwell.
Luhmann, N. (1989) Ecological communication, Oxford: Polity Press.
Luhmann, N. (1990) Political theory in the welfare state, Berlin:Walter de Gruyter.
Luhmann, N. (1993) Risk: A sociological theory, Berlin: Walter de Gruyter.
Luhmann, N. (1998) Love as passion: The codification of intimacy, Stanford, CA:
Stanford University Press.
Luhmann, N. (2000) Art as a social system (Meridian: crossing aesthetics), Stanford,
CA: Stanford University Press.
Luhmann, N. (2000) The reality of the mass media, Oxford: Polity Press.
Paterson, J. and Teubner, G. (1998) ‘Changing maps: empirical legal autopoiesis’,
Social and Legal Studies, vol 7, no 4, pp 451-86.
Stichweh, R. (1990) ‘Self-organization and autopoiesis in the development of
modern science’, in W. Krohn, G. Küppers and H. Nowotny (eds)
Selforganization: Portrait of a scientific revolution, London: Kluwer.
Stichweh, R. (1996) ‘Science in the system of world society’, Social Science
Information, vol 35, no 2, pp 327-40.
Stichweh, R. (1997) ‘The stranger – on the sociology of indifference’, Thesis
Eleven, no 51, pp 1-16.
Teubner, G. (ed) (1997) Global law without a state, Aldershot: Dartmouth.
129
Teubner, G. (1986) ‘After legal instrumentalism? Strategic models of post-

regulatory law’, in G. Teubner (ed) Dilemmas of law in the welfare state, New
York, NY: Walter de Gruyter.
Teubner, G. (1993) ‘Piercing the contractual veil? The social responsibility of
contractual networks’, in T. Wilhelmsson (ed) Perspectives of critical contract
law, Aldershot: Dartmouth.
Teubner, G. (1993) Law as an autopoietic system, Oxford: Blackwell.
Teubner, G. (1998) ‘After privatisation’, Current Legal Problems, vol 51, pp 393-
424.
Teubner, G., Farmer, L. and Murphy, D. (eds) (1994) Environmental law and
ecological responsibility:The concept of practice of ecological self-organization, London:
John Wiley & Sons.
Weiss, G. (2000) ‘A difference that makes no difference? Decision-making on
employment in European Parliament’, in P. Muntigl, G.Weiss and R. Wodak
(eds) European Union discourses on un/employment, Amsterdam: John Benjamins
Publishing.
Wodak, R. (2000) ‘From conflict to consensus?’, in P. Muntigl, G. Weiss and
R.Wodak (eds) European Union discourses on un/employment, Amsterdam: John
Benjamins Publishing.
Koselleck inspired analysis
Ball, T. and Pocock, J.G.A. (eds) (1988) Conceptual change and the constitution,
Lawrence, KA: University Press of Kansas.
Ball, T. (1988) Transforming political discourse: Political theory and critical conceptual
history, Oxford: Basil Blackwell.
Knemeyer, F.-L. (1980) ‘Polizei’, Economy & Society, vol 9, no 2, pp 173-96.
Koselleck, R. and Presner, T.S. (2002) Practice of conceptual history:Timing history
spacint concepts, Cultural Memory in the Present, Stanford, CA: Stanford
University Press.
Koselleck, R. (1985) Futures past, Cambridge, MA: MIT Press.
Koselleck, R. (1988) Critique and crisis, Oxford: Berg.
Walther, R. (1989) ‘Economic liberalism’, Economy & Society, vol 13, pp 178-
207.
130
Index
INDEX
A ‘concepts and counter-concepts’
‘actuality and potentiality’ (Koselleck) VI, 33-48, 47fig,
(Luhmann) 12, 73 48tab, 99-100
Althusser, L. 2, 3, 49, 52, 125 ‘condensed meaning’ (Luhmann) see
analytical strategies ‘meaning and condensed
Anderson’s views 93-118, 117fig meaning’
definition XIII, 93-6 counter-concepts (Koselleck) see
methodological issues XIII-XV ‘concepts and counter concepts’
ee also strategy types under Foucault;
Koselleck; Laclau; Luhmann D
Anderson, N. A. XIX-XXII, 93-118 Davidson, A.I. 25
apparatus (dispositif) (Foucault) de Saussure, F. 3, 6, 52
27-30, 28fig, 29fig, 31tab, 97tab, Dean, M. 27
99, 106-7, 114tab deconstructivist analysis (Laclau)
see also archaeology; genealogy; 56-61, 62tab, 97tab, 100, 110,
technologies of the self 114tab
archaeology (Foucault) 8-16, 31tab, relationship with discourse analysis
97tab, 98, 105 58-9, 58fig
ee also apparatus; genealogy; see also hegemonic analysis
technologies of the self Deleuze, G. 27
Archaeology of knowledge (Foucault) 8, Derrida, J. 49-50, 51
17 Of grammatology 57
‘articulation’ (Laclau) 50, 51 Descombes, V. 4-5
diachronic and synchronous analysis
B (Koselleck) 35, 43, 47fig
Balibar, E. 2, 24 ‘difference’ (Laclau) see ‘equivalence
Barthes, R. 6 and difference’
‘before and after’ (Koselleck) 44 ‘difference’ (Luhmann) see ‘form and
Beronius, M. 20 difference’, ‘unity and difference’
Between facts and norms (Habermas) differentiation analysis (Luhmann)
XVIII 81-3, 83tab, 89-90, 90fig, 92tab,
Birth of the clinic (Foucault) 105 97tab, 102, 111, 114tab
Born, A. XXI, 115 see also form analysis; formation
Bourdieu, P. XVIII, 125 analysis; media analysis; semantic
Brenner, N. 27, 30 analysis; systems analysis
Brunner, J. 33 Discipline and punish (Foucault) 6-7,
19, 27
C discourse and discoursivity
Chouliaraki, L. XIX Foucault VI
‘communication’ (Luhmann) 74-7 Laclau VI, 50-4, 72, 100
discourse analysis
(Foucault) 1-32, 114tab
131
relationship with deconstructive formation analysis (Luhmann) 97tab,

analysis (Laclau) 58-9 114tab
discourse theory see also differentiation analysis;
and Laclau 12, 49-62 form analysis; media analysis;
comparison with systems theory semantic analysis; systems analysis
93-7 Foucault, M. 1-32, 36, 49, 63, 86
discoursivity see discourse and analysis of statements 10-16
discoursivity analysis of structuralism 6
‘discursive points’ (Laclau) see ‘nodal apparatus (dispositif) 27-30, 28fig,
points’ 29fig, 31tab, 97tab, 99, 106-7,
dispositive (Foucault) see apparatus 114tab
‘dissimilarity’ (Luhmann) see archaeology 8-16, 31tab, 97tab, 98,
‘similarity and dissimilarity’ 105
‘distinction’ (Luhmann) see Archaeology of knowledge 8, 17
‘indication and distinction’ as discourse analyst 2-3
as structuralist 2-3
E Birth of the clinic 105
‘empty-signifier’ (Laclau) 54, 109 Discipline and punish 6-7, 19, 27
see also ‘floating signifier’, ‘signs discourse analysis 1-32, 114tab
and signifiers’ discourse and discoursivity VI
‘environment’ (Luhmann) see genealogy 17-23, 21fig, 31tab,
‘system and environment’ 97tab, 98, 105-6, 114tab
‘equivalence and difference’ (Laclau) History of sexuality 27
55fig Madness and civilization 3, 4, 19
Esposito, E. 69 ‘madness and reason’ 3-6
Mental illness and psychology 5-6
F on prison as punishment/cure 6-7
Fairclough, N. XVIII, XIX on punishment 6-7, 19
first order observation (Luhmann) Order of things 6
see observation Panopticon 7
‘floating signifier’ (Lacan) 53, 109 technologies of the self 24-6, 26fig,
Laclau’s views 53-4 31tab, 97tab, 99, 106, 114tab
see also ‘empty signifier’; ‘signs and
signifiers’ G
form analysis (Luhmann) 64, 78-80, genealogy (Foucault) 17-23, 21fig,
88-9, 89fig, 92tab, 97tab, 101, 31tab, 97tab, 99, 105-6, 114tab
110-111, 114tab see also apparatus; archaeology;
see also differentiation analysis; technologies of the self
formation analysis; media ‘generality and singularity’
analysis; semantic analysis; (Koselleck) 39-41, 47fig
systems analysis Gramsci, A. 49
‘form and difference’ (Luhmann) VI, Günther, G. 64
64, 65fig
‘form and medium’ (Luhmann) H
97tab, 114tab Habermas, J. XII, 125
132
Index
Between facts and norms XVIII semantic field analysis 38-47,

Knowledge and human interest XVIII 48tab, 97tab, 99-100, 108, 114tab
Hacking, I. 25 ‘up and down’ 45, 47fig
hegemonic analysis (Laclau) 49, 55- Krause, D. 64
6, 59, 62, 62tab, 97tab, 100, 108- Krause-Jensen, E. 7
9, 114tab Kristensen XVII, 2
see also deconstructivist analysis Kristeva, J. 6
Hegemony and socialist strategy
(Laclau) 49 L
Heidegger, M. 43, 45 Lacan, J. 49, 53, 54
history see genealogy ‘floating signifier’ 53, 109
history of concepts (Koselleck) Laclau, E. XVII, 1, 36, 37, 39,
33-48, 48tab, 97tab, 99, 107-8, 49-62, 63, 72, 74, 78
114tab ‘articulation’ 50, 51
see also semantic field analysis deconstructivist analysis
History of sexuality (Foucault) 27 56-61, 62tab, 97tab, 100, 110,
114tab
I relationship with discourse
Ifversen, J. 13, 16, 33, 34, 35, 38 analysis 58-9
‘indication and distinction’ discourse and discoursivity 50-4,
(Luhmann) 65-7 72, 100
insanity see madness and reason discourse theory 12, 49-62
‘inside and outside’ (Koselleck) ‘empty signifier’ 54, 109
44-5, 47fig ‘equivalence and difference’ 55fig
hegemonic analysis 49, 55-6, 59,
K 62, 62tab, 97tab, 100, 108-9,
Kaufmann, L.H. 68 114tab
Kent, C.A. 32 Hegemony and socialist strategy 49
Kneer, G. 75 ‘logics’ 59-61
Knowledge and human interest ‘nodal points’ VII, 51
(Habermas) XVIII views on Lacan’s ‘floating signifier’
knowledge archaeology (Foucault) 53-4
see archaeology Latour, B. 125
Koselleck, R. XVII, 33-48, 87, 88 Levi-Strauss, C. 3, 6, 53
‘before and after’ 44 ‘logics’ (Laclau) 59-61
‘concepts and counter-concepts’ Luhmann, N. XV, XVII, 1, 50,
VI, 33-48, 47fig, 48tab, 99-100 63-93, 93
diachronic and synchronous ‘actuality and potentiality’ 12, 73
analysis 35, 43, 47fig ‘communication’ 74-7
‘generality and singularity’ 39-41, differentiation analysis 81-3, 83tab,
47fig 89-90, 90fig, 92tab, 97tab, 102,
history of concepts 33-48, 48tab, 111, 114tab
97tab, 99, 107-8, 114tab form analysis 64, 78-80, 88-9,
‘inside and outside’ 44-5, 47fig 89fig, 92tab, 97tab, 101, 110-111,
114tab
133
‘form and difference’ VI, 64, 65fig mental illness see ‘madness and
‘form and medium’ 97tab, 114tab reason’
formation analysis 97tab, 114tab Mental illness and psychology
‘indication and distinction’ 65-7 (Foucault) 5-6
‘meaning and condensed meaning’ Mouffe, C. 49, 50, 51, 56, 61
12, 72-4, 86, 87, 102
media analysis 83-6, 84fig, 86fig, N
90-1, 91fig, 92tab, 97tab, 103, 112 Nassehi, A. 75
‘media and form’ 83-6, 84fig, 86fig, Nietzsche, F. 17-18
97tab, 103, 112 ‘nodal points’ (Laclau) VII, 51
observation VII, 64-71, 71tab,
77-8, 93-5 O
semantic analysis 86-90, 89fig, observation (Luhmann) VII, 64-71,
90fig, 92tab, 97tab, 102, 112-3, 71tab, 77-8, 93-5
114tab Of grammatology (Derrida) 57
‘similarity and dissimilarity’ 82, On the genealogy of morality
92tab, 97tab, 102 (Nietzsche) 17
‘system and environment’ 66, 67fig, Order of things (Foucault) 6
69, 80-1, 81fig, 101 Østergård, U. 34
systems analysis 82-3, 90-1, 91fig,
92tab, 97tab, 101, 111
P
systems theory 12, 63-92
Panopticon (Foucault) 7
‘unity and difference’ 78-9, 101
Parsons, T. 63
Pedersen, O.K. XI, XII
M Pocock, J.G.A. 33
Macey, D. 53 poststructuralism see structuralism
Madness and civilization (Foucault) 3, ‘potentiality’ (Luhmann) see
4, 19 ‘actuality and potentiality’
‘madness and reason’ (Foucault) 3-6 prison as punishment/cure
Mahon, M. 32 (Foucault) 6-7
‘meaning and condensed punishment (Foucault) 6-7, 19
meaning’(Luhmann) 12, 72-4,
86, 87, 102
R
media analysis (Luhmann) 83-6,
Raffnsøe, S. 28, 32
84fig, 86fig, 90-1, 91fig, 92tab,
‘reason’ (Foucault) see ‘madness and
97tab, 103, 112
reason’
see also differentiation analysis;
Regnault, F. XV
form analysis; formation analysis;
Richter, M. 33
semantic analysis; systems analysis
Roth, M.S. 32
‘media and form’ (Luhmann) 83-6,
84fig, 86fig, 97tab, 103, 112
‘medium’ (Luhmann) see ‘form and S
medium’ Scharff, R. 18
Megill, A. 19, 32 Schmidt, L.H. XVII, 2, 24
second order observation
(Luhmann) see observation
134
Index
self-technology (Foucault) see T

technologies of the self technologies of the self (Foucault)
semantic analysis (Luhmann) 86-90, 24-6, 26fig, 31tab, 97tab, 99, 106,
89fig, 90fig, 92tab, 97tab, 102, 114tab
112-3, 114tab see also apparatus; archaeology;
see also differentiation analysis; genealogy
form analysis; formation analysis; Time and being (Heidegger) 43
media analysis; systems analysis Tully, J. 33
semantic field analysis (Koselleck)
38-47, 48tab, 97tab, 99-100, 108, U
114tab ‘unity and difference’ (Luhmann)
see also history of concepts 78-9, 101
Shiner, L. 32 ‘up and down’ (Koselleck) 45, 47fig
Simmel, G. 24
‘signs and signifiers’ (de Saussure)
V
52-3
von Foerster, J. 64
see also ‘empty signifier’; ‘floating
signifier’
‘similarity and dissimilarity’ 82, Z
92tab, 97tab, 102 Zizek, S. 49
‘singularity’ (Koselleck) see
‘generality and singularity’
Skinner, Q. 33-4
Spencer-Brown, C. 63, 64-5, 84
statements VII
Foucault’s analysis 10-16
Stråht, B. 33
structuralism 2-3, 6
synchronous analysis (Koselleck) see
diachronic and synchronous
analysis
‘system and environment’
(Luhmann) 66, 67fig, 69, 80-1,
81fig, 101
systems analysis (Luhmann) 82-3,
90-1, 91fig, 93fig, 92tab, 97tab,
101, 111
see also differentiation analysis;
form analysis; formation analysis;
media analysis; semantic analysis
systems theory (Luhmann) 12,
63-92
comparison with discourse theory
93-7
135

Discursive Analytical Strategies: Understanding Foucault, Koselleck, Laclau, Luhmann

Uploaded by

Copyright:

Available Formats

Discursive Analytical Strategies: Understanding Foucault, Koselleck, Laclau, Luhmann

Uploaded by

Document Information

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

Discursive Analytical Strategies: Understanding Foucault, Koselleck, Laclau, Luhmann

Uploaded by

Copyright:

Available Formats

Discursive analytical PB.

A-W (reprint) 6/10/08 10:11 Page 1

DISCURSIVE ANALYTICAL STRATEGIES

• focuses on the political and social aspects of their writing;

Niels Åkerstrøm Andersen is Professor of Political Management at the Department

Koselleck, Laclau, Luhmann

Niels Åkerstrøm Andersen

BT001_cover_Bibliovault.indd 1 9/5/2013 11:41:10 AM

Understanding Foucault, Koselleck,

Niels Åkerstrøm Andersen

© The Policy Press 2003

British Library Cataloguing in Publication Data

ISBN 978 1 4473 4220 5 EPDF

Cover design by Qube Design Associates, Bristol.

Front cover: photograph supplied by kind permission of Mary Shaw.

1 The discourse analysis of Michel Foucault 1

2 Reinhart Koselleck’s history of concepts 33

3 The discourse theory of Ernesto Laclau 49

4 Niklas Luhmann’s systems theory 63

5 A hall of mirrors or a pool of analytical strategies 93

Appendix A: Examples of other analytical strategies 125

Appendix B: Further reading 127

List of tables and figures

Communication is a selection process, consisting of a synthesis of three selections:

Concept (Koselleck) is a word condensing a wide range of social and political

Deconstruction is showing how differences are contingent, that is,

Discourse (Foucault), or more precisely a discursive formation, is a system of

Discourse (Laclau) is a structural totality of differences that is a result of an

Form (Luhmann) is the unity of a difference.

Hegemony is only possible when something exists that can be hegemonised.

Information is a difference that makes a difference to a system. It is the system

Method is the rules and procedures required to produce scientific knowledge.

Re-entry is the operation by which a difference is copied into itself, thus

Second-order observation is an observation of an observation as an observation.

Semantics (Luhmann) are characterised as the accumulated amount of

Social system is an autopoietic system of communication, which defines itself

Statement is a function of existence that enables groups of signs to exist. The

Subjecting indicates that an individual or a collective is proclaimed to be a

From method to analytical strategy

• How have they come into being?

Are expressions such as ‘the complete colleague’,‘lifelong learning’ or ‘dedicated

Ontology and epistemology

• What does it mean that something exists?

It is characteristic of ontologically founded thinking that it moves rather swiftly

• Which procedures and rules are necessary in order to obtain theoretical

When a survey of the public administration is conducted by sending

Within ontologically over-determined science, methodology holds ultimate

Whereas ontologically over-determined thinking ontologises the object,

Table 0.1: Method versus analytical strategy1

Observation of an object Observation of observations as observations

The first example looks at analyses directed at the construction of social

Without an accurate identification of the conditions necessary for a social

The other example looks at studies of change. A methodological approach to

The problem holds different definitions in various epistemological programmes.

About the book

• What characterises the way they ask?

Rather, I wish to create an outline for the conception and discussion of

The four writers

1. Their theories are all programmes for second-order observations.