This article was downloaded by: [ULB Bibliotheque des Sciences]

On: 20 February 2012, At: 07:51

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number:
1072954 Registered office: Mortimer House, 37-41 Mortimer Street,
London W1T 3JH, UK

Journal of European
Public Policy
Publication details, including instructions
for authors and subscription information:
http://www.tandfonline.com/loi/rjpp20

Speaking 'Europe': the

politics of integration
discourse
Thomas Diez
Available online: 04 Feb 2011

To cite this article: Thomas Diez (1999): Speaking 'Europe': the politics of
integration discourse, Journal of European Public Policy, 6:4, 598-613
To link to this article: http://dx.doi.org/10.1080/135017699343496

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/
terms-and-conditions
This article may be used for research, teaching, and private study
purposes. Any substantial or systematic reproduction, redistribution,
reselling, loan, sub-licensing, systematic supply, or distribution in
any form to anyone is expressly forbidden.
The publisher does not give any warranty express or implied or
make any representation that the contents will be complete or
accurate or up to date. The accuracy of any instructions, formulae,
and drug doses should be independently verified with primary
sources. The publisher shall not be liable for any loss, actions,
claims, proceedings, demand, or costs or damages whatsoever or

Downloaded by [ULB Bibliotheque des Sciences] at 07:51 20 February 2012

howsoever caused arising directly or indirectly in connection with or

arising out of the use of this material.

Journal of European Public Policy 6:4 Special Issue 1999: 598 613

Speaking
Europe: the politics of
integration discourse

Downloaded by [ULB Bibliotheque des Sciences] at 07:51 20 February 2012

Thomas Diez

ABSTRACT The role of language has as yet been largely neglected in studies of European integration. This article provides the theoretical groundwork to remedy this gap
through discourse analysis. Its main argument is that attempts to capture the EUs nature
both in the political and the academic debate themselves take part in the construction of
the Euro-polity. The article proceeds in three moves: an Austinian move introducing
the notion of a performative language, a Foucauldian move clarifying the political
implications of such a language, and a Derridarean move discussing the possibilities of
change and opening up space for alternative constructions of European governance. Each
move is illustrated by examples from the history of European integration and European
integration studies. The conclusion presents research questions emanating from such a
discursive perspective and discusses some caveats.
KEY WORDS Discourse analysis; European governance; integration theory;
language; polity formation.

NAMING THE BEAST

Suppose a zoologist reveals the existence of an animal so far unknown to mankind.
In an article, she describes its features and gives the beast a name. It is classified and
categorized, put into the framework of zoological knowledge. In recent years, there
have been many attempts at exploring the nature of the beast (Risse-Kappen
1996b) in European integration studies. In many of them, the European Union
(EU) is dealt with as if it were our zoologists unknown animal. It is compared to
other polities and international organizations, its organizational mechanisms are
described and categorized. And there is much effort to name this unknown beast.
Debates abound as to whether it is a postmodern or regulatory state (Caporaso
1996), a confederatio, consortio or condominio (Schmitter 1996), a system
of multi-level governance (Marks 1993) or a multiperspectival polity (Ruggie
1993).
But as long as there is such a proliferation of names, and conceptualizations of
what the name EU means, the EU remains beyond the framework of our political
knowledge. While the efforts of categorization and naming are most often presented as pure descriptions, i.e. as mirrors of reality, the discrepancy between the
existence of the beast and our knowledge of it suggests that reality is not so readily
Journal of European Public Policy
ISSN 1350-1763 print/ISSN 1466-4429 online 1999 Taylor & Francis Ltd

Downloaded by [ULB Bibliotheque des Sciences] at 07:51 20 February 2012

T. Diez: The politics of integration discourse

599

observable as it may seem. Instead, even the zoologist needs a given system of
language, constituting the body of zoological knowledge, for her categorizations.
Language is thus central to our knowledge of reality. It does not only serve as a
mirror of nature (Rorty 1979). Rather, it is possible to know of reality through
linguistic construction only.
This article explores the role of language in the construction of the EU. Its main
argument is that the various attempts to capture the Unions nature are not mere
descriptions of an unknown polity, but take part in the construction of the polity
itself. To that extent, they are not politically innocent, and may themselves become
the subject of analysis, along with articulations from other actors. My plea is
therefore to include discourse analysis in the canon of approaches in European
studies. With a few exceptions, and in contrast to the field of international realtions,
such work is currently missing. Closing that gap would both enlarge our understanding of the integration process, and insert a reflective moment in our analyses.
First, it adds an important dimension to the predominant focus on ideas and
institutions within social constructivist studies of European integration, arguing
that they cannot exist apart from discourse. Second, it introduces a new face of
power. Analyses of European integration have so far by and large focused on
(absolute or relative) material capabilities as power, and on the interests behind the
application of such power. Against such an understanding, Steven Lukes once put
his radical view of power that works through preventing individuals or classes
from realizing their real interests in the first place (Lukes 1974). The notion of
power employed in this article follows the line of Lukes but doubts that there is
such a thing as a real interest independent from the discursive context in which
interests emerge. The power of discourse then becomes crucial.1 Third, it allows for
an analysis of the contestedness of certain concepts, and thus points towards
possible integration alternatives. Finally, it brings with it a reflective dimension to
the research processes, particularly necessary in a field in which many researchers
have traditionally been directly entangled with the political process of integration.
Throughout the article, I will restrict myself to providing some illustrations of
the argument and not conduct a discourse analysis as such. Instead, my aim is to
lay down the theoretical groundwork that relates a constructivism focusing on
language (variously called radical or epistemological constructivism, among
other labels) to European studies.2 The argument proceeds in three moves, each of
which I associate with the name of a certain philosopher or social/political theorist
whose writings have contributed to the elaboration of these moves. The first move
is labelled Austinian and introduces the notion of a performative language. The
second move is called Foucauldian and points to the political implications of the
performativity of language through the definition of meaning. The third move takes
up Derridarean themes and discusses the possibilities of change, opening up space
for the articulation of alternative constructions of European governance.
I introduce these moves as a way of developing and introducing a certain
approach. There are various problems attached to such a procedure. Most
importantly, it is not at all clear whether the work of the respective theorists is
compatible. On the contrary, it has been claimed that lumping together Foucault

Downloaded by [ULB Bibliotheque des Sciences] at 07:51 20 February 2012

600

Journal of European Public Policy

and Derrida, for instance, is to ignore the disagreements both of them expressed
vis--vis each other (see Marti 1988: 167, fn. 2). The exchange between Derrida and
Searle (who uses an Austinian understanding of language) has become a linguistic
classic (Derrida 1977; Searle 1977). It is, however, also the case that the works
associated with each of the three moves are, at least in part, shaped by the others. The
order in which they are presented here roughly follows the historical chronology of
their development, in particular in relation to when each move was taken up by the
social sciences in general, and international relations in particular. Thereby, it will
become clear how the debate proceeded from insisting on the relevance of language
per se to clarifying its power and potentials to change. Each move will therefore
refine, transform and thus move somewhat away from the insights gained from
earlier steps. All of them push the argument in a certain direction, with other paths
available. Readers may thus want to leave the proposed tour of inquiry at a certain
point, and prefer other possibilities opened up by then. None the less, I propose that
the approach I will have elaborated in the end is valuable in that it provides a new
perspective on the development of European governance.
THE AUSTINIAN MOVE: THE PERFORMATIVE LANGUAGE
The common sense of language is that it describes or takes note of a reality outside
language. It is, in other words, constative (Austin 1975: 3). The search for the
nature of the beast EU is in this tradition: European governance is something out
there, the nature of which needs to be captured by language, i.e. by the definitions
and observations entailed in our analysis. But there are several cases in which
language, even to the casual observer, seems to go beyond its constative function.
Examples are the declaration of a childs name at her baptism, the issuing of an order,
or the formulation of a treaty through which a new political organization comes
into existence. In his lectures at Harvard in 1955, J.L. Austin thus introduced
the notion of performative sentences (Austin 1975: 6). In the examples above,
language is performative in that it does not only take note of, say, the founding of the
European Economic Community (EEC). Instead, it is through language that this
founding is performed. Apart from the act of speaking itself (which Austin labelled
a locutionary act), in these cases it is in saying something [that] we do something
(Austin 1975: 94). There is an illocutionary force to language. Furthermore, what
we say may have an effect on other people; by saying something, we may not only
act ourselves, but also force others to do so.
Austin and his student John Searle contributed significantly to the development
of a theory of speech acts acts performed through speech. On the basis of this
theory, Jrgen Habermas was later to develop his theory of communicative action
(Habermas 1984), the influence of which one may trace to his current concerns for
a European citizenship linked to a European politico-communicative space
(Habermas 1992a, 1992b). It is, however, important to note the through in the
above definition of speech acts. In contrast to the following moves, the Austinian
move does not locate action on the level of language as such. Instead, language
serves as an instrument of will and intention: the question posed by Austin is, how

Downloaded by [ULB Bibliotheque des Sciences] at 07:51 20 February 2012

T. Diez: The politics of integration discourse

601

to do things with words, and not, how are things done by words. To the
extent that this presupposes language as a carrier of meaning, the principle of
expressibility, formulated by Searle (1969: 1921), is of crucial importance: It is in
principle possible to say what one means. Habermass discursive ethics, after all,
relies upon exactly this possibility of expression in a discursive space ideally situated
outside coercive power relations (Habermas 1990).
Although speech acts are never purely particularistic but rule-governed and thus
performed within a certain social context, they none the less flow, seen from this
perspective, from the individual. But to the extent that they are conceptualized as
rule-governed, meaning in Searles work is already at least sometimes a function
[and not the origin] of what we are saying (Searle 1969: 45). Speech act theorists are
concerned with politics through, not politics of discourse. But they recognize that
language is not always a neutral and purely descriptive device. Instead, it may
contain evaluations and serve political purposes (Searle 1969: 1326).
When it comes to politics, it is probably uncontested that most articulations, in
the form of negotiation statements, laws, treaties or the like, do or at least intend to
do something. Introducing speech act theory to international law, Nicholas Onuf
cites the statement of rules as an example of typical illocutionary acts (Onuf 1989:
834). The signing of the treaty on the European Coal and Steel Community, for
instance, founded the first European institution on the way to what is now the EU,
and served Frances interest of controlling an important base of German industry,
while it helped Germany to return to the international scene. The system of
governance established since then can be presented as a remarkable collection
of speech acts and their effects, be it in the form of declarations, further treaties,
decisions by the European Court of Justice, or Community legal acts.
In contrast to other attempts to analyse European governance, an approach
informed by speech act theory would pay more attention to language. In looking
for the nature of the beast, Thomas Risse-Kappen, for instance, is mostly concerned
with the domestic structure of certain policy fields and their degree of Europeanization (Risse-Kappen 1996b). The role of language in governance seems to be
as much underplayed as it is in social constructivist scholarship in international
relations more generally speaking, starting with Alexander Wendts focus on state
interaction through gestures, not speech (Wendt 1992: 404; see Zehfu 1998:
1258).3
A most interesting story in this respect is how citizenship developed from concerns about Europes political future and role in the world, via the necessity to
regulate membership of a single market, to being a response to questions about
legitimacy and democracy within the EU. During this process, speech acts
performed by a variety of actors, often with different intentions, not only led to the
establishment of EU citizenship, but also to the reformulation of the concept of
citizenship, with consequences for the shape of the Euro-polity.4 More generally
speaking, the whole history of European integration can be understood as a history
of speech acts (following Onuf: rules) establishing a system of governance (which,
after all, is about rules that are binding for the members of the system; see
Kohler-Koch 1993).

Downloaded by [ULB Bibliotheque des Sciences] at 07:51 20 February 2012

602

Journal of European Public Policy

We should not, however, overstate the distinction between locutionary and

illocutionary acts. In fact, one of Austins central propositions concerned the
practical difficulties in distinguishing between constative and performative sentences (Austin 1975: 94). First, even locutionary acts are performative to the extent
that to state something is to do something: it is to locate something in a specific
context, following certain rules and depending on the given circumstances (Austin
1975: 1467; Searle 1969: 22). Second, the notion of locutionary and illocutionary
acts is an abstraction. Speaking more generally includes both acts (Austin 1975:
147). In the same vein, Searle insisted that the idea that descriptive statements could
never entail evaluative ones amounted to what he called the naturalistic fallacy
fallacy (Searle 1969: 132).
Consider that it was common in the British debate of the 1960s to refer to the
EEC as the Common Market, whereas in Germany the term most often used was
Gemeinschaft (Community).5 One can reasonably assume that, to most people,
the utterance of these words seemed innocent and descriptive, but they were not.
First, in locating the EEC in different contexts according to the rules and circumstances of their respective national debates, they established a specific reading
of the Treaties of Rome. Second, in the case of Britain, this partial fixation of
meaning, together with a referendum as a means of legitimization, served
to structure the evolving debate about possible European Community (EC)
membership, dividing the broad spectrum of opinions into two simple camps:
pro- and anti-marketeers.6
Even if their illocutionary force is not as readily visible as in the case of
rules, such speech acts have important social and political consequences. Whereas
the Austinian move helped us to understand that speaking Europe is to do something, the Foucauldian move will help us to understand better the political force of
such performative language.
THE FOUCAULDIAN MOVE: DISCOURSE, POWER AND REALITY
The British example is, of course, well known and not very original. But it
seems that its implications are rarely understood. More often than not, the British
are taken to be natural Eurosceptics, owing to their history or geographical status.7
But on closer inspection, the problem is less to do with different attitudes towards
Europe, but with the concept of Europe itself. It has to be stressed that neither
the Common Market nor the Gemeinschaft conception was correct or false.
Rather, they were possible readings of the system of Western European governance.
In other words, Europe is not a neutral reality but a contested concept, the
meaning of which is not (yet) fixed (Connolly 1983; see Schffner et al. 1996: 4).
Even assuming (as I will do in the following) that it is somehow related to a system
of governance does not help that much: there are still numerous ways to construct
such a system, in content, nature and scope. It is such constructions that the speech
acts discussed at the end of the last section were about.
Europe might be one of the most typical examples of contested concepts, but
the argument can be made on a more general level. The central proposition is that

T. Diez: The politics of integration discourse

603

reality cannot be known outside discourse, for the moment broadly defined as a
set of articulations. In the words of Michel Foucault:

Downloaded by [ULB Bibliotheque des Sciences] at 07:51 20 February 2012

We must not imagine that the world turns towards us a legible face which we
would have only to decipher; the world is not the accomplice of our knowledge;
there is no prediscursive providence which disposes the world in our favor. We
must conceive discourse as a violence which we do to things, or in any case as a
practice which we impose on them.
(Foucault 1984: 127)
In many ways, this is merely a more radical reformulation of Austins
observation that to state something is to do something. But to phrase it in such
radical terms brings to the fore the political relevance of language beyond the
concept of rhetoric as a means to political ends, and towards a power that rests in
discourse itself. This power makes us understand certain problems in certain ways,
and pose questions accordingly. It thereby limits the range of alternative policy
options, and enables us to take on others. The contest about concepts is thus a
central political struggle (Connolly 1983: 30), not only between individuals and
groups defending one meaning against another, but also between different ways
of constructing the world through different sets of languages. These different
languages are not employed by actors in a sovereign way. It is the discursive web
surrounding each articulation that makes the latter possible, on the one hand
(otherwise, it would be meaningless), while the web itself, on the other hand, relies
on its reproduction through these articulations.
Discourse in this Foucauldian reformulation is thus more radical than the speech
act tradition in that more emphasis is put on the context in its relation to the
individual actor. Although it is we who impose meaning, we do not act as
autonomous subjects but from a subject position made available by the discursive
context in which we are situated (Foucault 1991: 58). The speech act tradition
emphasized the rules and contexts of speaking; the discursive tradition furthermore
emphasizes the constitutive role of discourse in the production of subject identities.
Discourse then takes up a life of its own. It is not a pure means of politics instead,
politics is an essential part of discourse. The struggle to impose meaning on such
terms as Europe is not only a struggle between politicians but also between the
different discourses that enable actors to articulate their positions (Larsen 1997a:
1212).
In a way, this notion amounts to what one may call a linguistic structurationism, adding to Giddens theory the crucial importance of language (see
Giddens 1984). Giddens central aim, shared by Foucault, was to move beyond
structuralism and to reconceptualize the duality of structures and agency. His
theory of structurationism, imported into international relations by Alexander
Wendt (1992), argues that both, structure and agency, were mutually dependent on
each other. Whether Giddens was successful in this endeavour is contested. It has
been argued, for instance, that structurationists eventually privilege structure by
making it their ontological starting point whereas, in a Foucauldian perspective,

Downloaded by [ULB Bibliotheque des Sciences] at 07:51 20 February 2012

604

Journal of European Public Policy

more emphasis is put on practice in that structures are always reinterpreted and
thereby transformed (Ashley 1989: 2767). The major point in the present context,
though, is that Giddens does not take language seriously enough (Zehfu 1998),
whereas a focus on discourse attributes a central importance both to the practice of
speaking and the linguistic context in which articulations emerge and are read.
Before I move on to show the relevance of this to European integration studies
by way of some more examples, I need to clarify that to say that any talk about
reality will always be a specific construction of the latter is not to deny the existence
of reality itself (Laclau and Mouffe 1985: 108; Potter 1996: 7). When entering a
different country, confronted with very real physical barriers, one has to present a
passport. While the Schengen agreement has eliminated borders between some of
its signatory states, it has led to the intensification of such controls at the outside
borders of Schengenland. But there is no neutral language to convey the meaning
of these real borders. Their construction as guarantees of welfare provisions or
illegitimate walls depriving people of their right to move are both speech acts
within a specific discursive context. Furthermore, discourse itself is part of reality.
In that sense, discursive approaches do not fit into the old dichotomy of idealists
versus realists. In fact, the example of Schengenland nicely illustrates this: it
emanates from and reifies a specific discursive construction of European
governance.
EURO-SPEAK
After the Foucauldian move, any description of European governance participates
in the struggle to fix the latters meaning, and thus is a political act. This is hardly
ever recognized. Philippe Schmitter, for one, explicitly acknowledges the role of
language in European integration. He identifies the development of a Euro-speak
defining the space for political action within the EU, while often being hardly
comprehensible to an outsider (Schmitter 1996: 1227; see also Schffer et al. 1996:
8). Elements of this Euro-speak range from acquis communautaire to codecision, from subsidiarity to supranationalit (Schmitter 1996: 137). At the
same time, however, Schmitter sees a need for labels to identify the general
configuration of authority that is emerging in the case of the EU, and doubts that
this can be done by a mere aggregation of currently existing Euro-speak (Schmitter
1996: 137).
But following the Austinian and Foucauldian moves, the new vocabulary that
Schmitter is looking for cannot be used simply to pick up such developments as the
emergence of a new form of multi-layered governance, and to describe the process
of integration (Schmitter 1996: 1323). Instead, such developments are only
knowledgeable to us within specific discursive contexts, and to label them from our
various subject positions is to engage in the struggle for Europe (Wver 1997b).
This struggle is not restricted to the realm of political practitioners as academics
dealing with matters of European integration, we are also part of it.8
Consider the conceptualization of the EU as a system of multi-level governance
(e.g. Marks 1993; Christiansen 1997). The image created by this account is one of a

Downloaded by [ULB Bibliotheque des Sciences] at 07:51 20 February 2012

T. Diez: The politics of integration discourse

605

set of various separated levels of governance (local, regional, national, European)

that interact with each other in some issue areas and follow their own course in
others. This has by now become something of a textbook image of the EU. It
would be nave to assume that this image directly becomes the ground on which
politicians in the EU base their decisions. This is not what is claimed here. Rather,
the point is that such conceptualizations are part of a wider discursive context and
do not stand aside from their object of analysis. They take up the claims made
by German Lnder about their role in the overall system, or by various national
governments leading to the specific construction of subsidiarity in Art. 3b TEC.9 It
is these multi-level representations taken together that reify a notion of politics
working on separate planes. The development of the EU towards such a system that
way becomes a self-fulfilling hypothesis.10
The power of discourse is that it structures our conceptualizations of European
governance to some extent, rather than us simply employing a certain language to
further our cause. The multi-level language gives preference to actors on various
state levels and is linked to an extension of the classical federalist practice of
territorial representation on the highest organizational level, now with three
representational bodies instead of two. What happens if for a moment we employ a
different language and speak of a network polity instead? Our conception of
the EU changes, and instead of levels, we find a more open political space, both
geographically and functionally diversified, undermining the territorial notion of
politics that is still upheld by the multiple levels concept (Kohler-Koch 1999).11
Which of the two languages should be preferred is contestable, and need not
be discussed at this point. Both have their own political consequences in that they
enable different kinds of political actors to claim legitimate existence in different
kinds of decision-making processes. A functional body such as the Economic
and Social Committee does not, of course, simply disappear once the multi-level
language is employed. But it does not figure too prominently in our representations
of the EU, and this quasi non-existence is being reified.
The language of neo-functionalism provides a second illustration. One of the
distinctive features of neo-functionalism was its proposal to bridge the gap between
functional and political association in classic functionalism by transforming the
concept of spill-over (i.e. the notion that integration processes, once started in a
field of low politics, will create a dynamic of their own and sooner or later affect
other policy fields) by adding to it an explicit political content and agent, working
towards the eventual establishment of an overall federal, or at least supranational,
system (Caporaso and Keeler 1995: 334; Kelstrup 1998: 29; Zellentin 1992: 701).
Again, the question is not whether those expectations were right or misplaced.
Instead, my proposition is that while neo-functionalism might thereby have closed
one gap, it opened up another one, and that this is because of the language
employed.
On the one hand, the reformulated spill-over concept had to include democratic
processes at one point or another. Economic policy might well be legitimized by
references to economic output the guarantee of welfare. But this leads to the
construction of Europe as an Economic Community (Jachtenfuchs 1997b; Diez

Downloaded by [ULB Bibliotheque des Sciences] at 07:51 20 February 2012

606

Journal of European Public Policy

1999; Jachtenfuchs et al. 1998). While legitimation through output is already a

position hardly accepted universally in relation to economic policy, things become
even more problematic if one moves into other policy fields. Thus, the inclusion of
participatory elements was unavoidable if spill-over was to be sustained. But, on the
other hand, the language of neo-functionalism was all very technical, the name of
the approach itself being no exception. Accordingly, the central institution in the
emerging polity was given the name of a commission, and the means of governance
were called directives and regulations (Art. 189 TEC). Such terms are hardly
reconcilable with the current language of democracy without a redefinition of
democracy itself. That, however, was not what was proposed in fact, classic
functionalism might have been more apt to such a redefinition by changing
the territorial organization of societies into a functional one, whereas neofunctionalism proposed using the latter to achieve the former.
The democratic deficit charge that has haunted the EU ever since its inception at
Maastricht seems to be directly connected to this problematic. Its citizens claim that
the EU is far too bureaucratic, technical, distant, and its decision-making procedures too intransparent (see Weiler 1998b: 78). This might be the case or not it
seems at least questionable whether politics in any of the national capitals is more
transparent. But the institutional language of neo-functionalism has prevailed until
today, and provides the ground to continuously reconstruct the EU as a monster
bureaucracy concerned with technical matters that increasingly affect the everyday
life of its citizens without their formal consent, while the nation state carries with it
the ideals of self-determination and democracy.
In such a setting, it is hard to make the case that the initiative for a substantial
number of directives can be traced back to member state governments, or that the
size of the EU administration is smaller than that of a single member state such as
Germanys federal bureaucracy (Wessels 1996: 1824), or that non-governmental
organizations are heavily involved in the making of EC policies (Jachtenfuchs and
Kohler-Koch 1996: 24; Kohler-Koch 1998). Surely, none of this makes the EU a
heaven of democracy not on the basis of the predominant current understanding
of democracy, in any case. Instead, the point of this discussion is that the language
of neo-functionalism enables one reading of the EU rather than another. And
furthermore, this language seemed right and innocent (in the sense of being the
objectively best available way) at one point much in the same way as the language
of multi-levellism today.12 In each case, the Foucauldian move points to the politics
involved in discourse, a politics that we are often unaware of and that does not come
to our attention as long as we equate politics with interests and intentions.
THE DERRIDAREAN MOVE: CONCEPTUALIZING CHANGE AND
OPTIONS FOR ALTERNATIVES
Within a universe of discourses, change is only possible if meaning is not eternally
fixed and if the lines of contestation between various discourses are allowed to
shift. Only if this is the case will there be a chance for the development of a new
Euro-speak, and thus for the development of alternative constructions of

Downloaded by [ULB Bibliotheque des Sciences] at 07:51 20 February 2012

T. Diez: The politics of integration discourse

607

European integration. On the other hand, the meaning of words needs to be

relatively stable in a given context for communication to be possible. In his
structural theory of language, Frdric de Saussure argued that national languages
work because they represent crystal grids in which each word has its proper place.
It takes on meaning through the firm opposition in which it stands towards another
word in this grid (Frank 1983: 324). In such a crystal grid model, change is hard
to conceive of. But we all know that meaning is not eternally fixed: dictionaries
provide us with contested meanings of a single word, and, once in a while, such
entries have to be changed because the word is now used in a different or additional
sense. Furthermore, we do experience breakdowns of communication.
This is the reason for a third and final move, which I will call Derridarean.
Change was not absent from the Austinian and Foucauldian moves. They
emphasized the role of action in a continuous reconstruction of and struggle for
meaning. But in order to conceptualize the interplay of structure and agency in
linguistic terms, the Derridarean move will be more helpful. In contrast to Saussure,
French philosopher Jacques Derrida conceptualized language not as a closed and
more or less rigid grid, but as a series of open-ended chains (Derrida 1977). With
each articulation, there is at least a potential of adding new oppositions to the
already existing chain, and thereby of altering it (see Potter 1996: 84). This does not
necessarily result in a breakdown of communication. In fact, communication does
not have to rest on a concept of understanding, assuming the correspondence of
what is said and received in the speakers and receivers minds. Instead, it can be
conceived of as operating on the level of language, where the decisive factor is
the affinity of discourses and thus their mutual translatability (see below).
Furthermore, change and continuity always go hand in hand with each other.
Although the overall discursive space is not as volatile as Derridareans sometimes
suggest, and each addition to a linguistic chain seems to be minor at first, it may
indeed be part of a major transformation, the importance of which becomes clear
only in the long run.
An example of such a change is the development of the construction of
European governance as an economic community in the form of a common
market in the British case. There, the predominant concept of European
integration in the 1950s was indeed a classic Eurosceptic one of pure intergovernmental co-operation. But at the same time, economic considerations played
an increasing role in the overall political debate. This led to the reformulation of
co-operation as a free trade area. The language in which this area was constructed
centred around economic output. Its basic mechanism was still intergovernmental,
but this economic focus laid down a trace that soon made it possible to articulate
supranational governance in the economic realm. And indeed, this is how
Macmillan presented his bid for membership in August 1961 (Hansard 1961: 1481,
1490; see Diez 1999: ch. 3).
Put in a simple way, we all enter into a conversation with a set of preconceptions
from which we set out to reconstruct other articulations. Thus, we not only receive
them passively, but regularly add to the linguistic chain unless our set of preconceptions (or at least those relevant for the given conversation) are exactly the

Downloaded by [ULB Bibliotheque des Sciences] at 07:51 20 February 2012

608

Journal of European Public Policy

same as the ones of the speaker. Borrowing a conceptualization from the radical
constructivist branch of systems theory (Hejl 1987), we may think of ourselves as
being situated in, and our preconceptions resulting from, a node of discourses
providing the basis for our interaction in communication. In other words, our
preconceptions are nothing other than objects of particular discourses, which in
turn are linked to a number of other discourses in what I call a discursive nodal
point (Diez 1998a, 1998b, 1999: ch. 2). There is a simple reason for such linkages
between discourses: the conceptualization of objects in one discourse follows a set
of rules, which, in turn, result from metanarratives providing meaning to the latter,
etc. This creates a web in which discourses are bound up with each other, and which
is held together by nodal points.
The latter, given the Derridarean move, are potentially unstable, but will usually
not change in a radical way. Shifts seem most likely if there is a considerable overlap
between the rules (and therefore the metanarratives) of the two discourses in
question, both in terms of content (that is, concerning the objects of the
metanarratives) and in terms of structure (that is, some overall principles to which
the rules adhere). This overlap makes articulations translatable. On the basis of such
similar languages, it is possible from one nodal point to make sense of articulations
resulting from another one, so that the latter are not rejected right away, opening up
the possibility of (ex)change. Seen from such an angle, the language of a free trade
area in the British case facilitated the move towards the articulation of an economic
community that would otherwise have been much harder, if not impossible.
Finally, the Derridarean move also allows us to address possible alternatives to
the federal state and economic community conceptions that currently dominate
the debate (see Jachtenfuchs et al. 1998). Recent years have witnessed an emerging
Euro-speak that focused on subsidiarity and flexibility.13 Most well known are
the introduction of the principle of subsidiarity into the Treaty of Maastricht,
accompanied by the establishment of the Council of the Regions, and suggestions
ranging from the Kerneuropa and concentric circle visions of German Christian
Democrats Karl Lamers and Wolfgang Schuble and former French Prime Minister
Eduard Balladur to the demands for more flexibility by former British Prime
Minister John Major. All of them, in one way or another, are set in opposition to
centralization and a further unitary development of the EU, either because the
latter are linked to hindrances for further deepening and widening of integration,
or because they are associated with a neglect of nation state identities. While
potentially undermining the acquis communautaire, the emergence of this new
Euro-speak in parts also serves to reify the nation state as a central concept in
politics. Nowhere is this clearer than in the way subsidiarity is invested in legal
discourse through Art. 3b and its sole stress on member states competences.
In terms of the centrality of territorial statehood in political discourses, the
change brought about by these terms thus seems to be of a rather marginal kind. It is
easier to see the problems they pose to the construction of European governance as
a federal state in the making than to the territorial organization of politics as
such. Rather, their usage seems to follow rules similar to those of multi-level
constructions. But seen from the perspective outlined above after the Derridarean

Downloaded by [ULB Bibliotheque des Sciences] at 07:51 20 February 2012

T. Diez: The politics of integration discourse

609

move, these seemingly marginal changes might bring with them more fundamental
transformations in that they lay out a linguistic trace that can be seized upon by
alternative constructions.
Consider the rules of the network discourse. It, too, is set against centralization,
but also against purely territorial politics, and includes both territorial and
functional divisions. Network-like constructions of European governance have
traditionally been marginalized in the overall integration debate. Members of the
Integral Federalists, for instance, argued at the Congress of The Hague in 1948 for
the encouragement, regardless of frontiers, [of] the spontaneous articulation of
interests, energies and hopes (Lipgens and Loth 1985: 49), and stated their wish to
be as far as possible decentralized, both regionally and functionally; not a superstate
but a real democracy, built up of self-governing communities (Lipgens and Loth
1985: 45). But their influence within the federalist movement was never strong, and
if anything became weaker over time.14 Their construction of federalism was too
far apart from that of the dominant discourse, the discursive nodal point from
which they argued too different and outlandish for those used to talking in terms
of modern territorial statehood. The language of the latter is clear, orderly and
relatively parsimonious the waters of the network discourse are much more
muddied. They do not provide a clear outlook and focus on terms such as
spontaneity or living, supple complexity (Lipgens and Loth 1985: 50). From the
discursive nodal point of a federal state conception, it is hard not to see this as a
deficiency. To put it simply, the language of vagueness did and does not translate
well into a language of clear borders, hierarchy and uniformity. The language of
neo-functionalism, in contrast, was in a much better position, having a clear overall
programme. In the same vein, multi-level governance is still a pretty much ordered
one in that it implies, for instance, the clear separation of a minimum number of
levels.
But remember that the exact meaning of a term is context-bound, while at the
same time it can be transformed through the reinvestment of the terms in question
from different discursive positions. Hence, it may turn out to be of some significance that the terms subsidiarity and flexibility are contested concepts that are
not alien to the network language. Instead, they are much closer to it than, for
instance, neo-functionalist language. This increases the translatability of network
articulations into dominant Euro-speak. Much like the movement from free trade
area to economic community in the British case, there is a trace that can be seized
upon by actors working from the networks discursive nodal point.
This is, of course, not to say that, in due course, the debate will have changed so
much that it becomes common to construct European governance in such terms.
The notion of linguistic structurationism reminds us of the need for these terms
to be reinvested by actors from such a discursive position. What is important,
however, is that the current transformations in Euro-speak allow for such a
reinvestment. Thus, the language of day-to-day politics may well be ahead of our
minds in trying to figure out the nature of the beast.

610

Journal of European Public Policy

Downloaded by [ULB Bibliotheque des Sciences] at 07:51 20 February 2012

THE IMPORTANCE OF LANGUAGE

My attempt in this article was to make a case for the importance of language in the
process of European integration. By way of three moves (Austinian, Foucauldian
and Derridarean), I argued that language does more than describe; that all our
accounts of the world (and thus of European governance) are embedded in certain
discourses; that the meaning of words is dependent on their discursive context;
that this context is not rigid but in constant, if only slow, flux; and that recent
transformations of the discursive context enable the construction of Europe as a
network. I have illustrated this string of arguments with a number of examples, but
there is no doubt that there needs to be more research into the workings of each of
the moves in the context of European integration. Among the research questions
that emanate from the above line of argument are the following:

What are the terms with which we speak about European integration? How
did Euro-speak evolve?
What are the political pre-decisions implied in those terms?
What are the alternative meanings of these terms in various contexts?
How are these terms invested? Which rules do they follow? From which contexts do they emanate?

Substantiated by such research, there are at least two further practical implications,
besides the enablement of the network alternative.
First, the future development of the EU will not depend solely on member states
interests, but also on the translatability of the discourses on European governance
that the relevant political actors are embedded in. It seems that the EU is a multiperspectival polity not only because of its lack of a single centre of decisionmaking, but also because it allows for conceptualizations from various angles. The
issue for institutional development is not whether these conceptualizations are
identical, but whether they can make sense of the Treaties and other basic texts at the
core of integration (Wver 1990).
Second, there might be too much focus these days on the change of institutions in
the narrow, organizational sense of the term. The change of institutions, from the
perspective developed above, is not interesting as a fact in and of itself, but as part of
a broader set of practices in which language plays a crucial role. Institutions cannot
be separated from the discourses they are embedded in, and rather than a formal
change of institutions, what seems necessary is a change in the discursive construction of these institutions, of which the former would only be one particular
component. Such change is obviously problematic, for no one can control language,
but everyone contributes to it in each new articulation.
The academic attempts to categorize the EU and give it a place in our order of
political systems are nothing but such contributions. They are attempts to fix the
meaning of European governance, so that we know what the latter is, but they are
not just objective analyses of a pre-given political system. This does not make them
worthless; on the contrary, they are as essential for our knowledge as the zoologists
classification of her beast is, and they are probably more relevant to our daily lives.

Downloaded by [ULB Bibliotheque des Sciences] at 07:51 20 February 2012

T. Diez: The politics of integration discourse

611

Eventually, a further difference to what the zoologist does with her words is
that while it may be relatively easy for her to take the lead in constituting the first
dominant discourse on the newly discovered animal, the many voices involved in
the construction of European governance will ensure that the fixation of meaning in
this case is much harder.
What is the politics involved here? On one level, the answer that this article has
given is that it is a politics of discourse, that within the language in which we operate
lies a set of choices about the political decisions of our day. Since I started out
from the observation that this discursive dimension is largely neglected, it was my
attempt to bring the latter to our attention by focusing on these pre-decisions. But
are we then, according to the above line of argument, dependent on the discourses of
the nodal points in which we are situated? Addressing these questions is a thorny
undertaking, and I can only sketch my (preliminary) answer. But however thorny,
they are of the utmost theoretical and practical relevance. After all, the poststructuralist work in the theory of international relations, from which my argument
is largely derived, set out as a critique both of individualized conceptions of political
agency and of the structuralism of neo-realism, which seemed to undermine any
attempts to change the anarchical international system (Ashley 1989: 2734).
My sketch draws on two distinctive features of discourse at it was set out
above. First, I pointed out that discourses do not cause but enable. They do have a
structural quality in that they are more than the sum of individual acts, but they are
at the same time dependent on the latter. They set limits to what is possible to
be articulated (Wver 1998: 108), but do also provide agents with a multitude of
identities in various subject positions, and are continuously transformed through
the addition and combination of new articulations. In spite of all the epistemological
and ontological differences, their work is thus none the less similar, for instance, to
the structures in Robert Coxs work on international relations, in which structures
predispose, but do not determine (Mittelman 1998: 76). There is room for creativity
on behalf of political actors in the model of discursive nodal points. In Ernesto
Laclau and Chantal Mouffes conceptualization, stressing the practice of articulations, the latter are even the means to link various metanarratives in order to fix
meaning (Laclau and Mouffe 1985: 113). But this creativity is not unlimited, and it
does not originate within the individual because the latter operates from a subject
position that is in itself discursively produced (Laclau and Mouffe 1985: 109, 115),
and so each articulation will already flow from a discursive nodal point. Neither
needs articulations that lead to a reformulation to be consciously conceived of as
such. Their meaning cannot be fixed, and thus they might induce changes beyond
original intentions actors, as Foucault once remarked, may well know what
they do, but what they dont know is what they do does (quoted in Dreyfus and
Rabinow 1982: 187).
Second, it needs to be recalled that, following the Derridarean move, discourses
are different from traditionally conceptualized structures in that they are not
rigid. Their contents can thus only be approximated, and not be once and for ever
determined. The concept of discourse itself might help us to think in novel ways of
structure and agency, since each articulation (a political act) is in itself a constitutive

Downloaded by [ULB Bibliotheque des Sciences] at 07:51 20 February 2012

612

Journal of European Public Policy

part of discourse. It is essential to note the extent to which articulations combine

linguistic elements in novel ways, or whether they largely reproduce the prevailing
rationalities. In that respect, the social constructivisms of Alexander Wendt (1992),
or Jeffrey Checkel (1998; this volume), stressing the co-constitution of structure
and agency and asking for greater attention to be paid to the processes of this coconstitution, are closer to the discursive constructivism espoused in this article than
is often assumed, again despite their differences. Surely, I cannot claim to have
finally solved the general puzzle of transcending the duality of structure and agency.
But the purpose of this article was a more limited one. It was to foster in European
studies, on the ground of theoretical reflections largely taken from the current
debate in international relations, the awareness of the power of language, and of the
discursive situatedness of our articulations and their readings. Speaking Europe, I
hope I have shown, is always to participate in a struggle, as much as is practised
from within a discursive context. The politics of integration discourse should not be
underestimated.
Address for correspondence: Thomas Diez, Research Fellow, Copenhagen
Peace Research Institute (COPRI), Fredericiagade 18, DK-1310 Copenhagen K,
Denmark. Tel. ++45-33455081. email [email protected]
ACKNOWLEDGEMENTS
Previous versions of this article appeared as a paper for the Social Constructivism
in European Studies workshop in Ebeltoft, Denmark, July 1998, and as COPRI
Working Paper 26-1998. For their helpful comments and suggestions, I am indebted
to Thomas Christiansen, Lykke Friis, Knud Erik Jrgensen, Markus Jachtenfuchs,
Marlene Wind, Maja Zehfu, two anonymous referees and the workshop
participants.
NOTES
1

2
3

4
5

For a discussion of these various kinds of power, see Hindess 1996. The latter are not
necessarily mutually exclusive, but my point in this context is to introduce the notion of
discursive power. A discussion of how the different faces of power are related is an
interesting task beyond the scope of this article.
It should be noted that discourse is not reducible to language. But since the latter is a
crucial element of the former, I will restrict myself in this article to the role of language.
Generally, though, the use of speech act theory is more widespread in international
relations than in European integration studies. Examples are the already quoted work
by Nicholas Onuf on international norms, Friedrich Kratochwils study on international law (Kratochwil 1989), or more recently the conceptualization of security as a
speech act called securitization by Ole Wver and the so-called Copenhagen School
(Wver 1995; Buzan et al. 1998; see Huysmans 1998). Wver and his colleagues have
also been among the so far few to analyse the role of language in constructing European
governance (Holm 1997; Larsen 1997a, 1997b; Wver 1990, 1997b, 1998).
This builds upon Antje Wieners work (Wiener 1997, 1998a), although she does not use
speech act theory explicitly in this context.
Until today, the major British-based political science journal dealing with European

T. Diez: The politics of integration discourse

6
7
8
9

Downloaded by [ULB Bibliotheque des Sciences] at 07:51 20 February 2012

10

11

12
13
14

613

integration is called the Journal of Common Market Studies, and the major law journal
Common Market Law Review, whereas the major German journal dealing exclusively
with European integration simply bears the name of Integration.
The pro-/anti-dichotomy may be seen as an effect of having a referendum as such. On
the other hand, referenda themselves depend on the prior formulation of alternatives.
This is analysed in greater detail in Diez 1999: ch. 1.
On the question of the problematic division between a realm of practitioners and a
purely theoretical academic realm, see Ashley 1989: 280.
TEC: Treaty establishing the European Community, as amended by the Treaty on
European Union.
A more general example is even more intuitive: the notion that we live in an age
of globalization has become one of the most important justifications for economic
policies, never mind the question of whether the phenomenon itself is real or not; see
Hirst 1997: 20614.
Schmitter suggests the network-like condominio as one possible future development
path of European governance. However, in the line of the above argument, he does
not treat it as a different reading of what the EU is (Schmitter 1996: 136). For other
conceptualizations of the network, see Diez 1996, 1997; Jachtenfuchs et al. 1998:
4212.
It should be recalled that, in the heyday of neo-functionalism, trust in technology
and science reached a peak in Western development, for instance in relation to nuclear
energy.
Out of a rich literature on this issue, see Adonis 1991; Endo 1994; Hglin 1994; Stubb
1996; van Kersbergen and Verbeek 1994; Wilke and Wallace 1990; Wind 1998.
Integral federalism experienced a small renaissance, though, in the European
governance models of Green Parties in the 1980s, see Diez 1996, 1997.