60 A Discursive Institutional Approach and its Analytical Implications
of rational choice institutionalism, as well as some sociological and, to a lesser degree, historical institutionalists - nor is it clear-cut inductive. Rather, the starting point may be described as analytical inductive. On the one hand, the approach is analytical since the point of departure is taken in general reflections on the concepts of, and the logical sequence between, ideas, discourse and institutions as well as in a series of conceptualisations about the conditions for, and dynamics of, institutional change. On the other hand, the approach is inductive in the sense that the analytical frameworks drawn upon in concrete empirical studies are 'drained' of their causalities, concepts are refined and interrelations established through historical descriptive empirical analysis.
Essentially, institutional change should be understood through the study of the interrelationship between the discursive and the institutional and this interrelationship is historically contingent.
It is thus important to note that present approach to institutional research has a knowledge ambition that differs from the rational choice and sociological and, to a lesser degree historical institutional approaches, which - to a varying degree - strive to establish causal theories. Against this, rather than striving to achieve the formulation of theories that reflect, and are testable against, an objective, stable and observable reality, the aim is to develop analytical strategies, which in turn enable the study of concrete historical institutional developments (Andersen et al. 1996, pp.171-172). In this sense a discursive institutional analytical strategy, on the one hand, takes its lead from theories and concepts of ideas, discourse and institutions while, on the other hand, an analytical strategy obtains its value when it inspires empirical studies, which generate not already available 'insights into aspects of the institutional organization of society' (Andersen et al. 1996, p. 172).
The Dynamics of Institutional Change: A Discursive
The Common Agricultural Policy and Organic Farming: An Institutional 61 Perspective on Continuity and Change
give momentum to processes of articulation and institutionalisation, the outcome of which is institutional change. The suggested conceptualisations of the conditions for, and dynamics of, institutional change will be fleshed out below and situated in the context of comparable concepts within the rational choice, historical and sociological institutional approaches.
Alternative Discourse and Conflicts Over Meaning
The discursive institutional approach outlined here holds that the existence of at least two to each other alternative discourses is a necessary condition for institutional change. The reason for this is that it is only in such a situation that a particular institutional context may be contested through disputes over the articulation of the ideas embedded in this institutional context (Campbell and Pedersen 2001a, p.1 I; Wittrock and Wagner 1996). Apart from constituting a necessary condition for institutional change, conflicts over meaning or conflicts over the articulation of ideas may, however, also be viewed as containing a dynamic, which may give momentum to institutional change.
The rational choice institutional optic on institutional change neither claim nor have a pronounced concern with the formation of preferences outside of what may be deduced from the material conditions of the involved agents, however, both the historical and the sociological institutionalisms operate with conceptualisations of learning processes, which are comparable to the present concern with conflicts over meaning as a dynamic of institutional change. From a historical institutional perspective, the dynamic of change contained in policy- oriented learning processes is often seen as depending on the appearance of technical information, on the rules governing the exchange of viewpoints in a given policy field, and also on conflicts between groupings with different beliefs as to political objectives and causal assumptions (cf. Jenkins-Smiths and Sabatier 1993). It may, however, be argued that this approach assumes too much coherence both of the belief systems between which learning may take place and of the agents ascribing to such belief systems. Hence, this approach leads to analytical blind spots, which renders impossible, for instance, the identification of incoherence and inconsistencies within belief systems and among the agents referring to such. In fact, an important dynamic of institutional change is missed here (I will return to this point below). The sociological institutional conceptualisation of successful learning processes suggests that such processes may be conceptualised as processes of diffusion or isomorphism (cf. DiMaggio and Powell 1991; Strang and Meyer 1994). Although this line of thinking introduces the concept of meaning, the emphasis is still on coherence rather than conflict. Consequently it is unclear whether learning processes are in fact a dynamic of institutional change or, perhaps, instead can be viewed as processes, which increases homogeneity and reinforces already present practises and norms.
62 A Discursive Institutional Approach and its Analytical Implications
Closer to the present concern with meaning conflicts as a dynamic of institutional change is Hajer's (1995) argumentative approach. Hajer suggests that the prime dynamic of change goes through conflicts 'over meaning of physical and social phenomena' (Hajer 1995, p.72) and argues that 'discursive interaction . . . can create new meanings and new identities' (Hajer 1995, p.59).
In order to address the issue of interacting discourses Hajer (1995) makes use of the story-lines concept, which is described as 'narratives on social reality through which elements from many different domains are combined and that provide agents with a set of symbolic references that suggest a common understanding' (Hajer 1995, p.62). Story-lines are thus discursive entities, yet not synonymous with the discursive. Rather, story-lines are seen as metaphors cutting across dissimilar discursive fields, which, in one way or another, all have a share in the construction of a common concern, and contribute to the reduction of the complexity (Hajer 1995, pp.62-68).
The formation of story-lines could then be seen as contributing to change by mediating a shared meaning about a common concern across otherwise dissimilar discursive fields. The main point made by Hajer (1995) is, however, that story-lines may contribute to change as they catch on among a greater number of agents and this may happen to the extent that story-lines appear to be persuasive. A story-line is, in turn, persuasive if it is argued in a credible way, the agents advocating a particular story-line are considered trustworthy and if the practical consequences are acceptable to those agents subject to persuasion (Hajer 1995, p.63). In general, a story-line may gain support to the extent that agents perceive this story to 'sound right' (Hajer 1995, p.61). The notion that ideas must 'sound right' or 'fit' into a particular meaning system in order to be adopted here is, in fact, a recurrent notion in literature addressing the role of ideas in politics (e.g. Hall 1993; Kingdon 1995). Unlike the historical institutional approach to the study of ideas, however, the discursively derived concept of story-lines does not claim that coherency necessarily enforces the spreading of storylines (Hajer 1995, p.61). Likewise, opposed to the historical institutional approach to the study of ideas, the conceptualisation of story-lines does not operate with an analytically predefined notion of the particularity of story-lines which, rather, needs to be identified within a particular discursive and institutional context (Hajer 1995, p.44).
With the objective of capturing the dynamic of institutional change there is, however, also a series of limitations that relate in particular to the concept of story-lines. First, it seems that the introduction of the concept of story-lines overrides the original concern with conflict and instead moves attention towards how story-lines may gain support among an increasing number of agents. In other words, the conceptualisation of story-lines seems to entail a noticeably greater interest with the 'fit' between ideas than with conflicts between dissimilar articulations of ideas. Second, similarly to historical institutional concerns with ideas, those story-lines, which are credible, acceptable and advocated by trustworthy agents, are claimed to have enhanced powers of
The Common Agricultural Policy and Orgatric Farming: An Institutional 63 Perspective on Continuity and Change
persuasion. Yet, even though Hajer (1 995) seeks to address the question of what 'persuasiveness' may originate from, it seems that the explanatory blind spot is merely passed on and still leaves the question open of the origin of credibility, acceptability and trustworthiness in a particular story-line.
To be sure, the discursive institutional approach advocated in the current context is not in radical opposition to Hajer (1995). In fact, in the name of simplicity the wish is merely to push towards an explication of the link between the institutional and the discursive. A discursive institutional response to the analytical problems outlined is thus that the 'fit' between dissimilar articulations of ideas and the 'powers of persuasion' are to be found in the rules that govern discourse (intimations on this point are also found in Hajer's notions of 'discursive affinities' and 'discursive contamination' (Hajer 1995, pp.66-67). In addition, the present discursive institutional optic wishes to level out the bias towards the emphasis on 'fit' introduced by the concept of story-lines by keeping in mind the issue of conflicts over meaning.
In that sense, the concept of conflict over meaning also aims to level out the bias of previous discursive institutional research towards a focus on 'the history of the winners' or 'histories of hegemonies' (Andersen and Kjaer 1996). Even if it is agreed that strategic choices are made during the course of a discursive formation which involves, for instance, certain problems being articulated and institutionalised rather than others, it seems that this concern is rarely turned into an object for discursive institutional empirical research. The concept of conflict of meaning, hence, prepares the field for the study of conflicts over the articulations of ideas and suggests that conflicts over meaning contain a dynamic of institutional change.
Along these lines, the conceptualisation of conflicts over meaning implies, first, that the 'fit' or 'persuasiveness' of ideas as articulated, for instance, in problems and solutions in a given political field, are related to the nature of the rules or institutions governing alternative discourses. To the extent, for instance, an idea finds several non-uniform expressions, yet its articulations are still governed by a set of discursive institutions that have similarities, the way should be paved for such articulations to mutate and possibly for discursive and institutional change. More specifically, for instance, dissimilar problems may notwithstanding be formulated as having similar causes that, in turn, may pave the way for the transfer of solutions.
Second, the discursive institutional perspective suggested, wishes to uphold the conception that disputes over the articulation of ideas contain an important aspect of institutional change and, like above, such conflicts appear around the rules governing discourses. For instance, even though a particular solution may have obtained an institutionalised position in a particular context, conflicts may still evolve around, for instance, the nature of the problem that the solution may resolve. The conflicts, through which a problem is refined, may in turn lead to readjustments in its solution and, hence, the way is paved for discursive and institutional change. Together, it is the 'fit' between and the conflicts over the
64 A Discursive Institutional Approach and its Analytical Implications
rules governing discourses that contain the dynamic of institutional change, which is conceptualised as 'conflicts over meaning'.
The meeting between alternative discourses and conflicts over meanings may ofien leave the institutional context unchanged, yet sometimes this interaction may also lead to mutated meanings, which may give rise to institutional changes over time. That is, first, conflicts over meaning may leave the institutional context unchanged to the extent that the already institutionalised discourse remains intact and its alternatives are rejected. This outcome is referred to as a strategic choice and is the most common outcome since institutionalised discourse, by definition, has the upper hand by already being authorised and linked to some type of sanctions. Second, the outcome of conflicts over meaning may bring about institutional change to the extent that alternative discourse is institutionalised alongside already institutionalised concerns. When conflict over meaning gives momentum to the institutionalisation of alternative discourse alongside already institutionalised concerns, future conflicts over meaning must be expected to appear since conflicts are, in this situation, essentially institutionalised. Finally, the outcome of conflict over meaning may give momentum to institutional change to the extent that a mutation appears out of the meeting of alternative discourses. A mutation is thus the term used to describe the outcome of an interaction between dissimilar discursive rules that have, for instance, produced a combination of problems, their sources and solutions, which differs from the discursive rules as they appeared prior to the interaction and, which through a process of institutionalisation, come to constitute an institutional change. Opposed to the institutionalisation of alternative discourse alongside already institutionalised concerns, when the outcome of conflict over meaning is a mutation of dissimilar discursive rules, then the level of conflict must be expected - at least provisionally - to decrease.
The concept of conflicts over meaning as a dynamic of institutional change has some similarities but also differs from comparable thoughts within the rational choice, historical, and sociological institutional approaches. First, although current study does not aim to specify and conceptualise the processes through which institutional change and change in preferences or interests among agents involved in a given political field are interrelated, it is clear that the two are related. Opposed to, most notably, the rational choice institutional approach, the discursive institutional optic suggested does not, hence, separate preferences or interests from institutions, but rather holds that agent preferences or interests are interpreted through specific historical discursive and institutional contexts.
Second, whereas, for instance, Jenkins-Smiths and Sabatier (1993) assume internal coherence of belief systems and advocacy coalitions, the present approach suggests that discursive incoherence and inconsistency is, in fact, the condition that enables institutional change through conflicts over meaning. That is, mutations of dissimilar discourses appear only when both conflicts exist around the rules governing these discourses - for instance, on the formulation of problems - and where some degree of similarity exists - for instance, on
The Common Agricultural Policy and Organic Farming: An Institzrtional 65 Perspective on Continuity and Change
authorised solutions. Moreover, individuals may not be referring to only one particular discourse but are instead more likely to contribute to the production and reproduction of several alternative discourses even where contributions arise out of critical annotations about such alternatives (see also Hajer 1995, pp.69- 70). Third, emphasising the duality between 'fit' and conflict as a dynamic of institutional change, prepares the field for the study of mutations of alternative articulations (for instance, of problems) rather than the adoption of cut-and-dried problems and solutions, which is, arguably, the study of continuity as opposed to that of change. Finally, it should be noted that the concept of conflicts over meaning prepares the field for the study of the discursive multiplicity within a particular context and aims to capture a dynamic of institutional change related to the interaction between alternative articulations of ideas within, rather than across, social contexts. This is a central characteristic that distinguishes meaning conflicts as a dynamic of institutional change, as opposed to processes of translation, to which we now turn.
Transla tion
The second concept proposed to capture a dynamic of institutional change is that of translation. Translation may be described as the 'process whereby concepts and conceptions from different social contexts come into contact with each other and trigger a shift in the existing order of interpretation and action in a particular context' (Kjaer and Pedersen 2001, p.2 19). The concept of translation proposes that agents operating in one social context may select from concepts and conceptions made available to them through contacts with other social contexts.
The concepts and conceptions selected may, in turn, be connected to concepts and conceptions already embedded in the context into which they are introduced and, essentially, trigger displacements or change in the existing discursive and institutional order (Kjaer and Pedersen 2001, p.219).
Both the historical and the sociological institutionalisms operate with explanations of institutional change, which are comparable to the conceptualisation of processes of translation proposed from a discursive institutional perspective. Among historical institutional frameworks it is common to refer to external factors or shocks such as broader socio-economic developments, change in government, crisis, or even acts of war as one possible source of institutional change (cf. Hall 1989; Baumgartner and Jones 1993;
Sabatier and Jenkins-Smith 1993). From within the sociological institutional approach it has been suggested that a central dynamic of institutional change is to be found in processes of isomorphism or diffusion involving the spreading of standardised practices and models and welldefined and coherent norms within or across organisational fields (cf. DiMaggio and Powell 1991; Strang and Meyer 1994). The concept of translation both has similarities with, and differs from, these historical and sociological conceptualisations of institutional change.
66 A Discursive Institutional Approach and its Analytical Implications
Just as historical institutional research may refer to external factors or shocks as having made an impact on a given policy or polity, or across a number of policies or polities, translation is also a concept that depends, ultimately, on contacts between social contexts. However, opposed to historical institutional approaches that refer to the abrupt impact of external factors or shocks when explaining institutional change, the discursive institutionalism holds that translation - albeit depending on contacts with other social contexts - is an ongoing process among agents within a particular social context (Kjaer and Pedersen 2001, p.219). The process of translation has similarities with the sociological institutional conceptualisations of diffusion and isomorphism insofar as these concepts are concerned with the carrying of concepts and conceptions from one context to another. However, whereas diffusion and isomorphism tend to give attention to the spreading of standardised practices and models and welldefined and coherent norms, the conceptualisation of translation focuses on the selectiveness by which certain concepts and conceptions may be elevated from one social context to another. That is, translation is somehow a more complex process than those of diffusion and isomorphism and implies that a selection of certain concepts and conceptions in one social context may - as translation proceeds - displace or mutate with existing interpretations and, thus, trigger a shift in the articulation of ideas embedded in the field in which a translation is taking place. Translation then has to do with the spread of ideas, but ideas are often translated in a selective way and may displace or mutate with already existing articulations of ideas:
essentially, the extent to which such displaced or mutated ideas are institutionalised, institutional change may said to have taken place.
It is important to note the difference between the dynamic of institutional change related to the concept of translation, and the dynamic of institutional change contained in the concept of conflicts over meaning, as proposed above.
Although both conflicts over meaning and processes of translation are conditioned by the existence of at least two dissimilar articulations of an idea, processes of translation involve - as opposed to conflicts over meaning - contacts with and references to the articulation of ideas within other social contexts. Importantly, this does not make translation an exogenous dynamic of institutional change since the actual processes of change appear as agents within a particular context translate, connect, and displace concepts and conceptions within that context with reference to the articulations of ideas in another context.
The concept of translation thus seeks to deal with external factors or shocks as explanations of institutional change, which may degenerate into ad hoc, residual explanations. Finally, whereas the sociological institutional concepts of diffusion and isomorphism are arguably capturing processes of homogenisation rather than change, the concept of translation prepares the field for the study on how concepts and conceptions may be selected in one context, connected to concepts and conceptions in another context, and how this may trigger