New times. Even a superficial reading of daily newspapers and magazines
is enough to convince anybody that we are living in a period of fast and
profound institutional change!

This is also true in political theory. Here it is broadly accepted that state forms are being transformed (Jessop 1987, 1989), that market institutions are being combined with semi-public and public institutions (Nielsen & Pedersen 1991), that party systems are decomposing or being restructured (Haeusler & Hirsch 1987), that corporatist arrangements are being dissolved modified (Streeck 1984; Teubner 1984; Brulin 1989). that legal structures are being deinstitutionalized or reformed (Harden & Lewis 1986; Graham & Prosser 1988; Teubner 1983, 1985, 1988). and that the role of legislatures is being reformulated (Eriksen 1989).

So, walls are cracking, and the question of how to study changes and processes of change is once again raised. The purpose of this article is to point to a number of methodological and theoretical problems which have to be discussed in connection with a theory of institutional change. The aim is strictly to discuss methodological and theoretical conditions pertaining to the study of institutional change and institutions. No analytical approach for the study of institutions will be presented. No description of actual institutions or institutional changes will be given. Rather, nine philosophical-methodological rarely raised and never answered in institutional theory will be asked.

Resumé

The purpose of this article is to point to a number of methodological and theoretical problems connected with a theory of institutional change. Nine philosophical-methodological questions are asked dealing with a wide range of problems rarely raised in neo-institutional political theory. The questions concern the concept of the state, of institutions, and of language and discourse, and deal with the dialectic between individuals and institutions, choice and rationality, formal organizations and institutions as well as prescriptive versus normative analysis. A critical theory of meaning is sought in order to develop neo-institutionalism from a research program into an actual set of descriptive and interpretative theories.

---

* A special thank you to Bob Jessop and Klaus Nielsen for valuable comments.

Changes and Processes of Change

It is common knowledge that theories of change are studded with methodological How can we evaluate processes without anticipating what we already know? How can we catch trends in the making without describing what has already taken place? It is no exaggeration to state that the history of social sciences is filled with both comic and tragic examples of predictions that never came through, with prognostications that turned out to be wrong, and with hopes and fears that later proved to be either too optimistic or too pessimistic. Still, even with this background, social scientists are once again raising the old question: how to study changes and processes of change?

Since the 19505, mainstream political science has generally taken it for granted that 'the state' can be looked upon as a given and static object, with constant external boundaries and a stable internal organization. Importance has been attached to studying individual and collective action within given institutional frameworks, to describing institutional change in retrospect, and to analyzing political structures as static patterns in terms of the conduct of individuals or organized groups. What might be called 'the myth of the state' i.e. a belief in the state as a sovereign, society-forming, goal-directed, rational, and unitary actor, has prevailed (Olsen 1988a, 1988b; Almond 1988; Fabbrini 1988; Lowi 1988).

This neglect of the state as an actor of its own was criticized in the 1970s and throughout the 1980s. Within this period at least three - on many accounts different but also converging - alternatives were formulated: (1) a state-centered theory (Skocpol 1979; Evans et al. 1985; Nordlinger 1981, 1987, 1988); (2) a strategic relational theory (Jessop 1990); and (3) an institutional theory (March & Olsen 1984, 1989; Olsen 1985).

It is possible to identify at least five theoretical points where these
theories tend to converge:

(1) They tend to view politics more in terms of rule, control and order
than in terms of allocation.

(2) They tend to view the state as an actor in its own right with its own
principles of organization.

(3) They tend to view political actions as shaped by institutional environments
norms and preferences) and not merely by the values,
norms and preferences of political actors themselves.

(4) They tend to view political phenomena in an historical perspective
rather than as actions occurring at the present.

(5) They tend to view institutional alternation and innovation both as
results of open conflict and of incremental processes of learning and
adaptation.

In short, at least two major points of convergence can be identified among

the three alternatives: (1) They do not take 'the state' as an a priori object of analysis. Instead, they present it in both diachronic terms (i.e. as a historical construct), and in synchronic terms (i.e. as an arena for simultaneousrule-oriented (2) They do not regard 'the state' as a static pattern of conduct. Instead, they present it both as a structured ensemble of values, norms, rules, and institutions, and as an arena of competing interests.

However, the three contributions are also in many respects very different. Skocpol, Nordlinger and others translate their awareness of the state into a state-centered theory endowing the state with real and important autonomy vis-å-vis pressures and processes located in civil society. They argue that the state is 'a structured field of action' with a unique centrality and autonomy both in national and international formations.

Jessop, by comparison (Jessop 1990, 379-393), presents a strategic relational theory and refuses to accept the state as a self-determining object in distinct opposition to civil society. He describes the state in terms of multiple centers of legal, paralegal and informal authority, with its own (relatively) autonomous position in relation to other social subsystems. He looks at the dichotomy between state and society as false and tries to analyze how structure (extra-institutional and institutional conditions) and strategy (meaningful, willed and striven for projects for focusing upon the state's role and functions) are related over time.

March and Olsen, from their point of view, introduce an institutional theory. They, too, look at the state as multiple centers of authority, but also present the state as a distinct political order with its own organizing principles. In contrast to Jessop, however, they draw an analytical distinction institutional arrangements and their environments. Relations between the two are only dealt with as questions of internal mechanisms for how organizations can adapt to. imitate and learn from external processes. Extra-institutional environments are described in abstracto.

So, the three contributions converge in some respects, but are also very different in others. They are particularly different in the way in which they present 'the state' as an independent actor and as an independent variable. While the statist theory looks at the state as an object in singular, both Jessop and March and Olsen refuse to deal with the state as such. Neither Jessop nor March and Olsen assign the state any necessary analytical priority. Instead they look at 'the state'

• as a complex set of concrete institutions;

• as an emerging, tendential phenomenon with no pregiven external
boundaries;

* as a symbolic or fictional unit with constant problems of how to secure
its unity and its ability to function as 'a multicentered whole'; and

• in terms of its environmental features and their impact on the dynamics
of the boundaries and organization of the political order.

By doing so, the authors go beyond the existing paradigm in political science and demand an eclectic reinterpretation of political science, sociology, communication theory, economics, history and anthropology. Their merit is to direct attention to at least four features neglected in mainstream political science during recent decades: (1) that institutions and not the state are the prime object of analysis; (2) that institutional history more than patterns of conduct are to be explained; (3) that meaning and action and not behavior is the phenomenon to be understood; and (4) that learning and adaptation, and not blind responses to direct stimuli, are the driving forces behind change. As a result, a major effort to change the theoretical focus not only of political science but also of organization theory and other social sciences is implied.

Below, these four features will be discussed in terms of nine questions to which, in my opinion, any theory of institutional history must specially devote attention if it is to develop from a research program into an actual set of descriptive and interpretive theories. Three books published recently will be used as points of reference - Etzioni (1988), March & Olsen (1989), and Jessop (1990). All three present critiques of mainstream paradigms in political science as well as in economics; all three try to develop analytical strategies for theories of an institutional history; and finally, all three can be said to take part in a neo- (or new) institutionalist attempt to reformulate the purpose of and the boundaries between the social sciences.

Question One

The first question is meta-theoretical by nature and by all accounts the most abstract of the nine. How this question is answered is none the less of vital importance for the other eight. Actually the question consists of three very closely related problems: an ontological, an epistemological and a methodological problem.

In logical terms, the ontological problem comes first, dealing as it does with the question of what reality 'is'. Ontological questions have not commonly been discussed in political science for decades. In the behavioral paradigm it has been taken for granted that the question was solved once and for all by the way logical empiricism within the philosophy of science answered the question with a methodological reference to observable (i.e. positively given) individual or collective phenomena (Lecourt 1981). Reality was identified with concrete, observable phenomena, (eventually)

aggregated to the level of clusters of conduct, and again (eventually)
understood as expressions of underlying values or norms.

When meaning and action (and not behavior) is chosen as the phenomenon be understood, the ontological question becomes more complicated. Jessop appears fully to take this into consideration. Etzioni accepts that every observation is systematically biased. His starting point is an observer; not a 'natural' observer, but an observer supported, affected, formed or influenced by social contexts (Etzioni 1988, xi f., 5, 179. 186). The way he then treats the ontological question is twofold: he affirms the given existence of an / (ibid., ixf., xii, 5, 9, 13 f.) and claims that every observation is systematically affected by communicative considerations, i.e. normative and affective WE considerations (ibid., 95). He claims that the observer cannot reach beyond the mere appearances of reality (ibid.. 123). All reality is intracultural, all observations are distorted by cognitive limitations (ibid., 115, 122 f., 189, 202 f.). He is a subjectivist and a relativist.

March and Olsen, on their part, are more difficult to label. They acknowledge possibility of an objective reality, but treat 'reality' as a term in abstracto (March & Olsen 1989, 41). On the other hand, they acknowledge that individuals see what they can see and see what they want to see, and that reality is formed by interpersonal and affective beliefs (ibid., 45). In contrast to Etzioni, March and Olsen acknowledge the existence of an extra institutional reality but, just like Etzioni. they approve of a given observer (a subject in abstracto). They are relativists and realists.

Jessop, in contrast, is a realist and relativist. He chooses the position of realistic ontology. He assumes the existence of a real world but claims that this justifies only a realist ontology in general. It cannot provide guarantees for any specific variant of the real world, i.e. for a realist ontology in particular. We are all participants in the real world in general. Our position as participants is socially constructed (as subjects in concreto) and can be analyzed as such. As subjects we are also, of course, observers in the real world in particular. As observers we have no direct access to reality. Our observations are both constrained and facilitated by normative properties.

The main difference between the contributions, in short, is the position of the subject. Two different possibilities are chosen. Etzioni and March and Olsen choose to accept the subject as given in methodological terms. The individual is a priori accepted as an observer. The world is out there to be observed and interpreted, understood and explained. Jessop. by comparison, chooses to see the subject as given in theoretical terms. The subject is a participant before he is an observer. The subject is a social construct; he can only observe because he is 'socially constituted' to be a participant. That he is able to observe and to interpret, to understand and to explain, is a function of the world and not of the subject.

Even if the differences between the authors at first glance can seem rather superficial, they do point to the faet that neo-institutional theory cannot avoid making - implicitly or explicitly - some fundamental claims about ontological questions.

The epistemological problem, in turn, is very closely related to the ontological one. So closely that only March and Olsen and Jessop as realists can formulate it. The epistemological problem is twofold: First, it concerns the question of how the world generally is constituted for knowledge to be possible, i.e. by which cultural conditions individuals are 'constituted as subjects' with possibilities to observe and to interpret, to understand and to explain (Douglas 1986, 100 ff.; Edelman 1988, 9, 34, 36, 45). Secondly, it concerns the question of how knowledge can be accepted as scientific knowledge, i.e. under what cultural conditions one discourse (between many alternative ones) is accepted as scientific (Foucault 1969; Lecourt 1972; Bhaskar 1975).

Only Etzioni looks at the problem in subjectivist terms. The epistemological is seen on the one hand as a 'natural' claim laid upon the individual scientist to mirror reality as correctly as possible, and on the other hand laid upon every individual as an ethical claim to take a moral stand in life as such (Etzioni 1988, 2-3, 12, 1990). For Etzioni, in other words, the epistemological problem is turned into a normative claim and not raised to the level of a theoretical problem.

March and Olsen, by comparison, according to their ontological viewpoint, raise the question. They ask why a growing number of scientists - including themselves - are today concerned with institutions and institutional (March & Olsen 1984, 734, 1989, 2). Their answer is only descriptive, however. Scientists today are forced to do so, because institutions over the years have become larger, more complex and more important in collective life (ibid., 734).

Jessop, in contrast, opts for a theoretical answer to the epistemological question. For him the question can only be formulated and answered in theoretical terms (Foucault 1969, 232-255). Looking, as he is, at subjects (e.g. political scientists) as socially constituted participants in structured contexts, the question cannot be answered by a normative claim or dealt with only in descriptive terms.

So, whereas Etzioni erases the epistemological problem by a methodological and March and Olsen deal with it in descriptive terms, Jessop looks for a theoretical formulation and answer. The differences, however, highlight the fact that neo-institutional theory simply cannot avoid a profound discussion of epistemological questions.

The methodological problem, finally, is closely related to both the ontologicaland
epistemological one. It concerns what research techniques
(procedures and remedies) should be adopted in order to ensure the validity

of knowledge in terms of the institutionally given rules and norms for scientific discourse. The problem is not just technical. It is rather a question of how, in methodological terms, to solve problems raised by answers given to ontological and epistemological questions.

One of the major problems in this connection is that the very object of inquiry is 'reality', i.e. 'reality' interpreted and experienced and not reality in its crude extra-discursive form. At our disposal for inquiry we have a large number of sources - statistics, blueprints, reports, laws, accounts, interviews, observations, descriptions, theories, and so on. None of these are unaffected by normative and institutional contexts, and none of them are immediately trustworthy. On the contrary, every single utterance has to be seen as an element in a wider system of meaning and be seen as deliberately used in ongoing power games played between a few or several actors.

What methods can be used to analyze and evaluate such sources, when we have to assume that every utterance is encoded in a wider system of meaning and that every utterance is used deliberately in power games to distort or omit, to emphasize or extol? How can we break the code, and how can we expose the game? This is the essence of the methodological problem.

Question Two

The second question deals with the concept of institutions. Especially from reading March and Olsen one could get the impression that the concept of institution is a term of art, perhaps even another worthy addition to classic residual categories like 'the state', 'the political system", 'the polity", and so on. In Etzioni the term only plays the role of a metaphor. In Jessop the terms are there, but the theory is missing. Institutions and organizations are taken for granted.

March and Olsen, however, offer a theory of institutions and institutional change. They define institutions as arenas for contending social forces, as collections of standard operating procedures, and as structures that define and defend values, norms, interests, identities, and beliefs (March & Olsen 1989, 17 f., 24). Two questions are raised here. First, what is actually meant by institution? March and Olsen point to values and norms as the dependent variable and institutions as the independent variable. Institutions are explained by values and norms and not by any feature of their own. What, then, defines institution? Second, how is one to study the historical constitution of institutions. March and Olsen discuss - in great detail - how institutions are changing (ibid., chs 4. 5, 6. 7). They never discuss how institutions are constituted, only processes of change, and processes of

institutionalization are not discussed. The term institutionalizatiori is only
used twice, and then without any substance (ibid., 38, 53).

Treating institution as the independent variable makes it a non-historical term, and raises the following question: In what way do institutions constitute specific variable compared to other variables such as values, norms, rules, procedures, rituals, and symbols that are normally (March & Olsen 1989, 22) used to describe traits of institutions? Is it possible to define institutions, simultaneously, by including values or norms as dependent variables, and to regard institutions as something more and probably different? Is it possible to distinguish between unstable, short-term or even idiosyncratic preferences and stable, long-term or commonly accepted values or norms, and - on an analytical continuum - to establish at what point a 'break' between the two occurs?

These questions, anyhow, indicate that neo-institutional theory, in whatever cannot avoid making basic assumptions about how to define institution as a dependent variable. It is worth asking whether institutions can be defined both as already existing entities (an independent variable), and as entities in the making (a dependent variable)! It must also be considered, how institutionalization can be defined - on a continuum - as comprising both stable or unstable, short-term or long-term, formal or informal, idiosyncratic or commonly shared values or norms! Likewise it must be considered how institutions can be distinguished from values, and values from norms, norms from rules, rules from procedures and procedures from roles! It is necessary, in short, to identify when an institution becomes an institution and not something else - e.g. a value, a norm, a rule, a role, and so on.

Question Three

The third question links up with the second and concerns the concept of formal organization. From studies in the theory of organization, sociology of law and political economy, a systematic differentiation between institution formal organization is clearly becoming a major analytical problem. The question is not immediately one of definition. It touches rather on how a long list of new types of institutions and organizations since the 1950s and onward have been established, and how this multicentered has been coordinated by the use of both formal authoritative informal normative means (Pedersen 1988, 1990).

Whereas behavioral theories in the 1950s and in the 1960s took it for granted that political decisions could be studied under the assumption that political institutions were relatively few and stable, hierarchically controlled and easily identified neo-institutionalism takes the opposite stand. Both

March and Olsen and Jessop claim that the political order must be perceived as an increasingly complex mixture of competing, complementary and even contradictory institutional arrangements. At the same time, our understandingof orders and composite institutional settings has become inadequate (Olsen 1988c). Still, neither Jessop nor March and Olsen reflect upon the fact that concurrently, it has become more difficult to distinguish between institutions and organizations, and more difficult to apply an unambiguous concept of institutions and formal organizations (even so, see Brunsson & Olsen 1990).

In sociology the concept of institution refers to commonly shared values repeated in rather stable patterns of conduct. The concept of formal organization, on the other hand, refers to (partly) formalized norms, either expressed in positively formulated rules or commonly accepted as such. Formal organizations are defined as one of many possible forms of institutionalization. only used the term institution in a very restricted sense. He developed, in contrast, other terms (tradition, ethics, etc.) covering phenomena later described as institutions, and concentrated on denning organization in detail (including formal organization) (Weber 1978, 48-56).

Modern social theory, on the other hand, attempts to create a more precise delimitation of the two. Habermas, for example, applies the concept of institution in a broad sense. He regards institutions as being established through communication, and understands language as the only medium for the creation of normative agreements (i.e. institutions), able to encompass and integrate systems of actions (Habermas 1984, vol. 1, 85-101). Habermas than Weber has a delimited conception of formal organization. He regards formal organizations as generated by positive law and organizational as premised by legal regulations (ibid., 342 ff.).

What Habermas does not seem to consider, but March and Olsen on their part seem to emphasize (see also Eriksen 1987; Tuori 1988), is the growing interdependence between communicative actions oriented towards mutual understanding and purposive-rational actions oriented to formallegal Today politics are just as concerned with the generation of commonly possessed political values as the generation of organizations by positive law (Nielsen & Pedersen 1989; Berrefjord et al. 1989).

Inter-organizational studies show how formal organizations have become parts of institutionalized environments and how they have become institutionalized themselves (Brunsson & Olsen 1990). Interorganizational too. describe how institutional networks become established over time through communicative actions managed by certain professions and specialized institutions, with the deliberate intent of producing shared worldviews and of transforming these into normatively discourses (Pedersen 1988).

Boundaries between formal organizations and institutionalized environments therefore growing increasingly blurred. Any organization participating institutional networks must be able to formulate and propagate worldviews and self-identities and be able to mobilize other organizations to make decisions in the image of these. To secure commonly shared worldviews has become an important precondition for the coordination and control of institutionalized environments. Organizations have to look upon other organizations as institutions, and have to identify themselves by institutional traits.

This situation, in short, raises at least three problems: (1) How in mixed orders and composite institutional settings can one delimit formal organizations from institutionalized environments? (2) How are interorganizational developed into institutional networks, and how, accordingly, are formal organizations transformed into institutionalized organizations? (3) Finally, is it possible, in more analytical terms, to distinguish between institution and organization and, on a continuum, to recognize when a social process becomes a formal organization and not something else - an institution, for example?

Question Four

The concept in question is 'the state. March and Olsen, but also Jessop, have created grave problems for themselves by proclaiming that the state can be looked upon both as a multiple center of legal, paralegal and informal authority as well as an order in its own right with its own integrative properties.

One major problem can be formulated from the stance taken by March and Olsen and Jessop respectively. It concerns the distinction between political and non-political institutions. March and Olsen for their part claim that the political order (polity) is partitioned into relatively independent domains (March & Olsen 1989, 26, 27 ff., 167). Furthermore, they suggest that it has become increasingly fragmented into small and autonomous constituencies weakening the integrative aspects of polity (ibid., 97 f., 112 f., 135 f., 165 f.). The unity of a polity, therefore, cannot be taken for granted and cannot be understood as exogenously given to any study of political action. The political order, in essence, defies simple description. The delimitation of the political order has become an important analytical problem.

In his terminology, Jessop looks at 'the state' as an emergent, tendential phenomenon. Far from being given, 'the state' must constantly be constitutedpolitically. suggests that the unity of the state only can be defined through state projects, i.e. discourses infusing meaning into what

can be understood as 'the state', but also assigning roles and functions to
the state as such. The state emerges because there is a state project (Jessop
1990, 11 ff., 357, 366, 463, 474, 495).

All three authors, then, regard the polity as a social construct and its unity as a theoretical problem. Neither 'the state' nor its unity can be taken for granted. Yet recognizing that the state is socially defined leads to a series of basic questions about the social context of the polity. Logically, it also leads to questions about how 'the economy' and other social subsystems defined (Stern 1990, 39 ff.). And, finally, it leads to questions about where the boundaries between political science and its scientific neighbors can be drawn.

None of these questions can be solved by referring to the state as the legitimate executor of enforceable sanctions. Neither is any answer to be found in the proclamation of 'natural' (a priori) boundaries between state and market. No boundary can be taken as an a priori given, and no proclamation for indisputable. Answers are only to be found in studying the constant political struggle around the setting and changing of boundaries between political and non-political institutions and in denning where, on an analytical continuum, to place the 'break' between political and nonpolitical

The same is true for the boundaries between social sciences. If mixed orders and composite institutional settings, today more than ever, characterize political order, we also have to change and set new boundaries between social sciences. However, this can only be done by analytical means. The future role and function of different social sciences should be clarified through factual descriptions and deliberate definitions and not be based on myths or traditions upheld by institutions in the realm of sciences.

The merit of March and Olsen. Jessop and Etzioni, by their example, is to draw our attention to the fact that if new times are in the making, then it will also affect the role and function of social sciences. Old boundaries have to crack; old paradigms have to be challenged. Only an eclectic reinterpretation of perspectives in political science, sociology, economics, communication theory, history and anthropology can grasp how new times are creating new trials and new possibilities.

Question Five

The fifth question concerns the concept of language and the construction of discourses. Neither Etzioni, nor March and Olsen or Jessop substantially call into question language and communication. In Etzioni meaning is not discussed, but only looked upon as consisting of commonly shared beliefs, values and norms (Etzioni 1988, 5). March and Olsen, by contrast, place

great emphasis on meaning. They see meaning as a socially constituted interpretative order (March & Olsen 1989, 3, 48) and the development of meaning as the creation of instruments for individuals to make sense of the world they are living in (ibid., 40f., 51 f., 91 f., 94).

Jessop, just like March and Olsen, emphasizes that social conflicts evolve around the development of meaning (Jessop 1990, 473, 478 f., 481 ff., 494 f., 500). He also argues that meaning has to be expressed through discourses and therefore cannot be understood as a 'free-floating' medium for values, norms and beliefs (ibid., 393, 405; March & Olsen 1989, 25). Social relations are not endowed with meaning per se, but only through discourses.

Besides this important addition, Jessop, in opposition to March and Olsen, distinguishes between the extra-discursive and discursive moments of social order. In his ontology, social phenomena can have definite extradiscursive and liabilities (Jessop 1990, 401^-11). And in Jessop's (just as in March and Olsen's) epistemology, knowledge about the world can be experienced through practice and learning and not only through a discursive formation (ibid., 403 ff.; March & Olsen 1989, 66f.).

Although Jessop treats discourses as endowing social relations with meaning, he - in common with March and Olsen - never regards language as such as an independent object for analysis. All three treat language as exogenously given, independent of any theory of how meaning can be shared or commonly possessed. No theory of language is developed to explain how language can endow social relations with meaning or to explain how communication in and through language sets possibilities for and limitations how individuals can participate in and observe the world (Habermas 1984, vol. 1, 37-42). Once again the question appears to be twofold. (1) First there is a question of the systematic place of language and communication in any theory trying to deal with the connection between actors participating in and communicating about the world. It is a question of how a common political culture can be 'materialized' in language and of how this language in itself involves specific validity claims for any communication about political values and institutions. Meaning is not a free-floating medium, but a collective understanding in language, setting concrete claims on any process of communication. (2) In addition, there is a question of how communication is institutionalized in public spheres, setting limits for who can communicate about what with what kinds of arguments. Language is not to be dealt with in abstracto but to be understood as discourses scattered in competing and distinct public spheres with an independent set of participants and validity claims (March & Olsen 1989, 40f., 51 f., 91 f., 94, 129 ff., 155 f.; Jessop 1990, 408 f.; Pedersen 1990; Nielsen & Pedersen 1990).

Question Six

This question concerns the individual and the institution. The most important issue currently debated in political theory is how to deal with meaningful actions and institutional structures in a comprehensive way.

Neither Etzioni nor March and Olsen refute analyses rooted in atomistic individual actors; neither do they want to carry too far analyses that stress structures as determining or shaping individual actions (Etzioni 1988, 9f., 14 f.; March & Olsen 1989, 4f., Bf., 23). They wish to avoid structural determinism as well as atomistic reductionism. The path they stress is one of reaction against treating institutions simply as epiphenomena of selfinterested and groups (Etzioni 1988, xif., 4f., 181-198, 244f.; March & Olsen 1989, 4 f., 40-46, 162-166). It is also one of reaction against treating behavior as epiphenomena of the functional needs of social systems (Etzioni 1988, 9 ff.; March & Olsen 1989, 66).

Jessop, in line with realist ontology, is more eager to argue against structural determinism than he is to criticize others for methodological individualism. He argues against what he calls 'the customary but unhelpful distinction between "structuralist" and "instrumentalist" approaches' (Jessop 1990, 342), and emphasizes that 'structural constraints can only be meaningfully defined in relation to specific agents pursuing particular strategies." (ibid., 342, see also, 408f.)

Such statements, however, contain a number of important problems for all authors. The first problem is that they all seem to take for granted (1) that individuals and institutions can be separated a priori as well as be treated as autonomous; and (2) that this autonomy is generally valid and can be applied exogenously to all societies. The second problem is a particular aspect of the first. None of the authors really make it possible to distinguish analytically between physical individuals on the one hand and role-carrying actors on the other. None of the authors consider the 'constitution of roles' and the institutional conditions for turning physical individuals into 'actors in structure' (Pedersen 1989) as a theoretical question. The whole question of how individuals are 'made up" as individuals not dealt with in any substantial way (Douglas 1986. 100ff.: Edelman 1988, 9, 34, 36, 45).

Jessop is the one who comes closest to raising the question in theoretical terms. He too, though, generally seems to take for granted that individuals and institutions can be separated as autonomous. He also pays lip service to the question of how individuals are "made up' (or interpellated) as individuals (Jessop 1990, 405, 408f.).

Etzioni, by comparison, takes for granted that individuals a priori can
be separated from institutions and also that individuals a priori are individuals.He

viduals.Heproclaims that individuals 'are placed' by structures (Etzioni 1988, 5); but in substance he argues that social contexts support individuals in being free (ibid., xif.), providing them with possibilities to make individualchoices , 186); as well as affecting, forming and influencing them (ibid., 5,179 f.). He does not deal with questions relating to interpellation or analogous mechanisms.

The same thing can be said about March and Olsen. They proclaim that institutions define individual identities (March & Olsen 1989, 17, 161, 164), but in substance they argue that rules constrain and dictate behavior (ibid., 22), define appropriate actions (ibid., 23), specify who has access to what institutions (ibid., 22), and place obligations on a certain role (ibid., 161). The question of how individuals are constituted is avoided.

So, none of the authors actually enquire as to how institutions serve to define individuals, or of how in various ways institutions ensure that individuals will conform to their freedom (Douglas 1986, 100; Pedersen 1989). The question can be formulated in a more paradoxical way: how do institutions define individuals as autonomous from institutions? Or, on what conditions and with what effects will individuals in 'flesh and blood' be interpellated as actors and placed in a certain institutional structure?

All this seems to boil down to the conclusion that Etzioni and Jessop, just like March and Olsen, assume that individuals a priori are endowed with autonomy, and that individuals a priori have roles to play. The lack of any analytical arguments seems to justify the interpretation that they are avoiding the issue of a theory of institutional role-making. This is probably also why they make themselves vulnerable to the criticism traditionally by behavioral theory against classical institutionalism - i.e., properties of individuals should be researched empirically and not taken for granted, and that the performance of individuals can be quite different from the customary values and norms denning their conditions for action.

Question Seven

Question number seven concerns choice and rationality. This question is not usually dealt with in mainstream political science. By contrast, the most decisive question discussed in economics and organization theory for years has been the relation between collective rationality and individual choice.

Two principal lines of argument have been followed in these discussions: one guided by rational choice theory and one oriented towards the theory of bounded rationality. None of these can be described in singular or unilateral terms. In a crude fashion, however, it is possible to claim that rational choice theory (with an influence from methodological individualism

in neo-classical economic theory) is basically deductive. The smallest unit of analysis is claimed to be 'the natural human being', equipped with a set of a priori conditions for action (will, needs, information and preferences, etc.). The theory of bounded rationality, in contrast, is basically inductive. Influenced by social psychology, computer science and cultural theory, it claims that the ability to make fully rational choices is bounded by constraintsinherent the technical, individual and institutional conditions for choice.

In Jessop's writings problems connected to collective rationality and individual choice are - surprisingly enough - not discussed. Jessop, probably more than anyone, has in two ways introduced the concept of strategy into state theory. First, he has done so in describing the state as the site and the object of strategic elaboration by classes and other forces (Jessop 1990. 13). Second, he has done so in treating the state (its structures and functions) as having a differential impact (its strategic selectivity) on how classes and other forces can pursue their own strategies (ibid., 13).

If asked, Jessop would probably argue that it is no accident that the question of rationality is marginalized in his theory. Taking into account his ontology, rationality has to be understood as socially defined and not a free-floating medium for anybody to use for any purpose. Even so. serious questions remain to be answered. What, for example, in Jessop"s terms is to be understood by strategy? Is strategic calculation an expression of rational behavior? If so, is it an expression of a certain type of rational behavior? Should the answer be yes on both accounts, then what is to be understood by rational behavior (as one of several types of behavior) and by strategic rationality (as one of several types of rational behavior)?

Questions like these are probably not solved by the mere proclamation of rationality as socially defined. For one must also explain what can be understood as rational in opposition to non-rational behavior and strategic in opposition to non-strategic behavior.

It is by no means an easy task to disentangle whether and how Etzioni or March and Olsen manage, if at all, to respond to questions like these. In both cases they appear to use the term rationality in a non-defined way. A possible distinction between rational and non-rational behavior is not discussed. It is merely posited that choices can be neither completely rational nor non-rational, only rational and non-rational in varying degrees (see Etzioni 1988, 93, 144 f.)'

In contrast to Jessop, both Etzioni and March and Olsen place great emphasis on discussing institutional conditions for individual choice. They mainly do so in opposition to rational choice theory (Etzioni 1988, 93 ft'.. 115 ft., 144 ff.; March & Olsen 1989, 4-16, 22. 59. 66. 118. 128. 132). but how they do so is quite distinct. Etzioni differentiates normative-affective (N/A) contexts from logical-empirical (L/E) considerations (Etzioni 1988.

93 ff.). He claims that the majority of choices are largely based on N/A considerations and that L/E choices, when they occur, themselves are defined by N/A factors (ibid., 93 ff.). He even brings examples of institutionalzones which L/E considerations are banished (ibid., 154f.).

For their part, March and Olsen distinguish between a logic of appropriateness I) and a logic of consequences (logic II) (March & Olsen 1988, 23 f., 160 ff.). First, they arrange the two logics in a hierarchy according to empirical studies which have revealed that 'even in extreme situations like war, . . . ? individuals seem to act on the basis of rules of appropriateness rather than rational consequential calculation' (ibid., 22). Second, they try to combine the two logics according to a general postulate that 'Although the (decision) process is certainly affected by considerations of the consequences of action, it is organized by different principles of action, a logic of appropriateness and a comparison of cases in terms of similarities and differences' (ibid., 26).

Such statements, however, raise important problems. First, while Etzioni takes it for granted that there can be two different zones for choice-making (the N/A and the L/E zones), March and Olsen claim that there can be two forms of logic (logics I and II). Second, it is empirically possible to distinguish between the two conditions.

Both Etzioni and March and Olsen prioritize the background structures over the intelligible free choices of identifiable political actors, and see the free flow of rational choices as bounded by values and beliefs. But while the purpose of this position is clear, the consequences are disputable. The purpose, of course, is to claim that individual choices are neither without preconditions or limits, nor mere mirrors of institutional forces. The consequences, are twofold: whereas it is discussed in great detail how N/A considerations and the logic of appropriateness (logic I) inextricably bound up with values, norms, rules, procedures, etc., it is not discussed if L/E considerations or the logic of consequentiality (logic II) are bound up in the same way, too.

An important problem is created here. It appears as if the authors regard logic I and N/A considerations as endogenous and logic II and L/E considerations as exogenous to the institutional analysis, and therefore assume that logic II and L/E considerations are given (natural?) features. This is perhaps a misinterpretation of what the authors would argue if asked. But the lack of any analytical differentiation between rational and non-rational behavior seems to justify the interpretation that they believe logic II and L/E considerations to be given (exogenous) features to any institutional analysis.

A solution to this problem might involve viewing both logic I and N/A
considerations and logic II and L/E considerations as endogenous to the
institutional analysis and thus to claim that institutions not only stake out

norms and rules for appropriate and non-appropriate choices, but also constitute norms and rules for rational and non-rational behavior. This would imply that we have to look for how norms and rules enable actors to act (1) as ifthey are determined by their own will (express independent attitudes) and as ifthey are not; (2) as ifthey are goal-oriented (participate in social relations in order to realize self-interests) and as ifthey are not; and (3) as if they are provident (able to calculate consequences of their own behavior) and as ifthey are not.

Such a solution would further imply that the distinction between the two zones and the two logics is analytically abandoned. Logic II and L/E considerations would become one of several aspects of logic I and the N/A considerations. Rational and non-rational behavior would become examples of appropriate and non-appropriate behavior. On the other hand, if both these implications are accepted, then the prime question will no longer be how technical, individual and institutional contexts set boundaries for fully rational choices, but how discourses and institutions are contexts for what is appropriate and non-appropriate (e.g. rational and non-rational) behavior. This is how Habermas, too, understands the problem, when he distinguishes between communicative, strategic, and instrumental rationality, claims that they are all connected to validity claims (Geltungsansprachen) to be met in and through communication (Habermas 1984, vol. 1).

So, a neo-institutional theory (1) has to explain how appropriate and non-appropriate behavior is made possible by discourse and institutions: (2) how rational and non-rational behavior are examples of appropriate and non-appropriate considerations: and (3) how individuals and collectivities act as if they are both rational and non-rational. Only by doing so will we be able to regard logic II and L/E considerations as endogenous to an institutional analysis. And only by doing so, will we be able to make our terminology one of both obligations and codex, of artistry and creativity, criminality and anomie. and even of constructive and unconstructive destruction.

Question Eight

This question concerns processes of change and the design of change.
For several centuries now. the belief in science as a distinct remedv for

knowing the true nature of the physical world has been closely related to macrophysics - the science of sciences in the realm of positivistic epistemology. Since at least the 1840s (Auguste Comte and the publication of the Cows de la Philosophic Positive) social sciences have submitted to the paradigm of macrophysics, asserting that the universe consists of small

elements whose motions are highly predictable, determined as they are by
universal laws of motion. One of the latest examples is rational choice
theory within political economy.

In accordance with this ideal of scientific endeavor, rational choice theories are often unconcerned about whether their models really describe the richness, complexity and confusion of decision processes. Usually they are more interested in whether the models prove able to predict future conduct (Posner 1977, 12ff.).

In this connection Etzioni and Jessop, like March and Olsen, occupy a special position. They are just as interested in analyzing the empirical conditions for particular choices as in deducing consequences from axiomatic of motions (e.g. pregiven needs and preferences). They do not exclude the possibility that concrete political choices may prove to have important, intended or unintended, consequences. Neither do they exclude the possibility that a correct link can be established between past choices and subsequent effects (March & Olsen 1989, 41; Jessop 1990, 409). Thus, they maintain that just as the organization of polity makes a difference, so do individual or collective choices. And just as institutions are expected to shape the conduct of political actors, so are actors expected to reshape institutions.

In Jessop's discussion a whole series of mechanisms for change are identified - for example, struggles for hegemony, conflicts over state projects, strategic selectivity, experiences, learning processes, adaptation and autopoietic self-description, and so on (Jessop 1990, 405-411, 437-446).

In Etzioni's discussion change happens through adaptation, experimentation, and experience. But it also happens through fundamental innovative decisions setting the pace and direction for further decisions (Etzioni 1988, 130-135). A theory of mixed scanning is developed in opposition to both rational and incremental descriptions of change (ibid., 122-135, 157, 159 f.).

March and Olsen operate with several mutually incommensurable theories change. They distinguish between (1) long-term incremental adaptation changing problems employing available solutions within gradually evolving structures of meaning (March & Olsen 1989, 94); (2) short-term ad hoc reorganization, where reorganization depends less on properties of projects and efforts than on random political attention (ibid., 81); (3) intentional transformation, where change cannot be controlled with any great precision (ibid., 56-67); and (4) the design of institutions, which is seen as a question of how to shape citizen preferences and construct democratic political institutions (ibid., ch. 7).

A major difference between the authors can therefore be identified in
the way they deal with questions of change. Etzioni, just like March

and Olsen, maintains that institutional contexts can shape the conduct of political actors, but also that actors are able to reshape institutions through innovative decisions or design. Actors are not only 'blind' participants in processes of change (through experimentation, learning, adaptation and confliction), but are also capable of designing institutions in the image of their proper ideals and interests. Jessop only hints at, but never fully affirms (in the concept of strategic calculation), the possibility for actors to design institutions.

Thus, the authors all emphasize the complexity of processes of change and the combination of different motives (strategies), thereby raising the question of determinism, incrementalism, and blind chaos. The result is a number of problems and paradoxes. First, both Etzioni and March and Olsen assume that actors can deliberately design or invent conditions for future developments. In the reciprocity between structure and actors, they assume that the basis for a general theory of architectural design or innovative construction is already there.

Second, all of the authors postulate that a theory of change must operate with a fundamental unpredictability of systems that otherwise behave partly deterministically. They claim that a system, which otherwise is composed of single and partly predictable elements, may experience a complex order (chaos). So, they postulate (1) that it is possible to apprehend partial causal relations between structure and actions; and (2) that it likewise is possible to apprehend chaotic relations, although such relations cannot causally be referred to distinct actions.

Problems and paradoxes like these immediately call forth very fundamental (1) How can we identify the institutional possibilities for specific actors to design the future or innovate fundamental decisions? and (2) How can we describe the coherence in societies which are supposed to reproduce and change themselves aimlessly through deterministic, incremental and even chaotic processes? Neither March and Olsen nor Etzioni or Jessop try to answer questions like these. They take for granted that actors, through their creative skills, judgment, and innovative faculty, are capable of making designs and taking fundamental decisions. An institutional theory of design is left open.

Question Nine

The ninth question concerns the relationship between descriptive studies
and normative or prescriptive motives.

In the last ten to fifteen years it has once more become appropriate to
take a normative and/or prescriptive stand in political science (Lundquist
1988; Smith 1988). This development is in great contrast to the ideal of

scientific endeavor dominant in political science since the 19505. Etzioni, just as March and Olsen, is among the first to demand either a normative stand and/or a prescriptive element in political theory. Jessop, by comparison, does not deal with the question at all.

The whole purpose of Etzioni's book is to remind scholars that they are responsible for the ethical consequences of their ontological, epistemological methodological choices. His purpose is deliberately normative. and Olsen do not explicitly take a normative stand. They, in contrast, emphasize the possibility for scientists to make prescriptive recommendations on the basis of their analytical results. They see scientific knowledge as an important element in designing democratic institutions. They want to formulate principles for a reasoned debate between alternative

Programmatically, then, March and Olsen espouse a position which has not been appropriate in political science since classical institutionalism: to unify positive analysis with prescriptive concerns. Even so, their attitude seems to be more prescriptive than normative and more based on analytical insights than scientific ideals. They urge us to begin with the values we hold and the political world we inhabit. To them, the primary normative problem is not analytical choices, but the disintegration of polity into multiple centers for segmented and professionalized handling of interests. Likewise, their primary normative objective is not to discuss ethical consequences of analytical choices, but to further a distribution of personal rights and institutional autonomies such that a reasoned public debate can be created (or re-recreated) and a unified democratic culture can be established (or re-established).

The way in which March and Olsen pursue this objective, however, causes problems. First, March and Olsen distinguish between aggregative and integrative political processes (March & Olsen 1989, 117 ff.). Second, they present criteria for how to evaluate such processes (ibid., 119-129). Finally, they claim that such an evaluation can be implemented in two ways: (1) to compare '. . . alternative institutional Utopias - models of hypothetical institutions in hypothetical worlds' (ibid., 129); and (2) to compare '. . . the realized consequences of pursuing alternative Utopian visions in a real political system' (ibid., 129). Both comparisons are to be based on two questions: a question of efficacy, i.e. \ . . whether an institution in an imperfect world what it promises in an imagery one. . . .' (ibid., 129) and a question of virtue, i.e. whether '. . . pursuing a specific virtuous vision brings good citizens closer to, or more distant from, the good life - justice, a moral society, beauty, harmony' (ibid., 129).

Thus, March and Olsen occupy two positions. They want to understand
existing institutional processes, but also to evaluate alternative visions or

utopias. They want to interpret inherited values as a manifestation of different types of processes, but also to assert that it is possible to evaluate these in the light of a model for 'a moral society'. This causes at least two problems. The first is related to one of the issues identified in Question Eight. It is: how can we identify the institutional conditions for actors being endowed with the possibility of formulating visions or utopias? The second concerns the selection of criteria of evaluation and the idea about the good citizen living a good life in a moral society. It is: How can we select criteria for evaluating the 'good' citizen, the 'good' life and the 'moral' society?

March and Olsen identify neither institutional possibilities for how visions can be formulated, nor assert on which premises a choice of criteria should be based. Perhaps once again the authors would argue otherwise if asked. But it seems that they begin, just as Maclntyre does, by assuming that '. . . morality is always to some degree tied to the socially local and particular' (Maclntyre 1981, 119), but then end otherwise, by claiming that their normative ideals of an open and democratic society are exactly so essential that they can be taken for granted as both possible and necessary.

The dilemma, of course, is not new. On the contrary, the dilemma is classic: can visions or Utopias be scientifically argued for? The answer is probably both yes and no and will, in the last instance, also have to be 'solved' by taking a normative stance. But because the problem can only be partly dealt with in scientific terms and in the last instance will also have to lean heavily on normative doctrines, it can in any event be raised as a fundamental question: How. through interpretation and reflexivity, is it possible to criticize the array of existing and inherited values?

What is lacking in neo-institutional theory, then, is a critical dimension. How can we interpret and criticize existing values and norms? How - by our interpretations - can we expose (or discover) possibilities for visions and Utopias? Questions like these have been central in western philosophy at least since Immanuel Kant. And once again we are brought back to Question One. The methodological question can now be turned into a genuine theoretical one. We can ask:

• How can we evaluate existing values or norms without prescribing what
we already know, and without inscribing what we just imagine to be
possible? and

• How can we evaluate dominating values and norms without overlooking
or underestimating alternative visions and objectives pushed aside or
censored by these?

Conclusion

These were the nine questions. Ten or eleven - or many more, just as

relevant - could have been asked. New times are in the making, and
probably one of the most important signs of this is that political theory
once again is forced to raise basic problems.

I have tried to show that the first and foremost question for neoinstitutionalism to be: how can we 'liberate' ourselves from the array of dominating worldviews through a critical interpretation of existing values and norms? The problem is ontological, epistemological and methodological. is a question of how to address basic problems. In concrete (or analytical) terms, the question has already been posed: how can we criticize encoded meaning, and how can we interpret discourses deliberately used to distort or to extol in games of power? How, in other words, can we break the code and expose the game?

Problems of this kind must be paid undivided attention in neo-institutionalism. through a critical theory of meaning can we break the code and expose the game, but also - and probably more importantly - point at alternative visions and possibilities in the values we hold and the political world we inhabit.

REFERENCES

Almond, G. A. 1988 'The Return to the State', American Political Science Review 82, 853-874.

Berrefjord, 0., Nielsen, K. & Pedersen, O. K. 1989. 'Forhandlingsøkonomi i Norden - En
indledning', in Nielsen, K. & Pedersen, O. K., eds., Forhandlingsøkonomi i Norden.
Copenhagen: Jurist- og Økonomforbundets Forlag; Oslo: TANO.

Bhaskar, R. 1975. 'Feyerabend and Bachelard - two Philosophies of Science', New Left
Review, no. 94.

Brulin, G. 1989. Från den 'svenska modelien' till företagskorporatism? Lund: Arkiv avhandlingsserie

Brunsson, N. & Oisen. J. P. 1990. 'Kan organisasjonsformer velges?'. Bergen: LOS-senteret.
LOS-senter notat 90/6.

Douglas, M. 1986. How Institutions Think. New York: Syracuse University Press. Edelman, M. 1988. Constructing the Political Spectacle. Chicago: University of Chicago Press. Eriksen, E. O. 1987. 'Symbols, Stratagems, and Legitimacy in Political Analysis', Scandinavian Studies 10, 259-278.

Eriksen, E. O. 1989. 'State Autonomy in the Age of Corporatism: Towards a Theory of the
Post-Corporate State'. Paper, Seminar on A Crisis of the Welfare State Model? Tromsø.

Etzioni, A. 1988. The Moral Dimension. Toward a New Economics. New York: The Free
Press.

Etzioni, A. 1990. 'A New Kind of Socioeconomics', Challenge 33, 31-32.

Evans, P., Rueschmeyer, D. & Skocpol, T., eds. 1985. Bringing the State Back In. Cambridge:
Cambridge University Press.

Fabbrini, S. 1988. 'The Return to the State: Critiques'. American Political Science Review 82,
891-901.

Foucault, M. 1969. L'archéologie du savoir. Paris: Gallimard.

Graham, C. & Prosser, T., eds., 1988. Waiving the Rules. The Constitution Under Thatcherism.
Philadelphia: Open University Press.

Habermas, J. 1984. The Theory of Communicative Action. Reason and the Rationalization of
Society, Vols. 1-2. London: Heinemann.

Haeusler. J. & Hirsch, J. 1987. 'Political Regulation. The Crisis of Fordism and the Trans-

Harden, I. & Lewis, N. 1986. The Noble Lie. The British Constitution and the Rule of Law.
London: Hutchinson.

Jessop, B. 1987. 'Der Wohlfahrtsstaat um Obergang von Fordismus zum Postfordismus".
Prokla 1.

Jessop, B. 1989. Thatcherism: The British Road to Post-Fordism". Bielefeld: Center for
Interdisciplinary Research. Memo.

Jessop, B. 1990. State Theory: Putting the Capitalist State in its Place. A proof-reading
manuscript is used here for reference. Book forthcoming 1991. London: Polity Press.

Lecourt. D. 1972. Pour une critique de l'epistémologie. Paris: Francois Maspero.

Lundquist, L. 1988. Byråkratisk etik. Lund: Studentlitteratur.

Maclntyre. A. 1981. Åfter Virtue. Notre Dame: University of Notre Dame Press.

March, J. G. & Olsen. J. P. 1989. Rediscovering Institutions. The Organizational Basis of
Politics. New York: The Free Press.

Nielsen. K. & Pedersen. O. K. 1989. "Fra blandingsøkonomi til forhandlingsøkonomi. Mod et nyt paradigme?", in Klausen, K. K. & Nielsen, H. T., eds.. Stat og Marked. Fra Leviathan og usynlig hånd til forhandlingsøkonomi. Copenhagen: Jurist- og Økonomforbundets Forlag.

Nielsen, K. & Pedersen. O. K. 1991. 'From the Mixed to the Negotiated Economy: The
Scandinavian Countries". in Coughlin. R. M.. ed.. Morality. Rationality, and Efficiency:
Perspectives on Socio-economics. New York: M. E. Sharpe.

Nordlinger, E. A. 1981. On the Autonomy of the Democratic State. Cambridge: Harvard
University Press.

Nordlinger, E. A. 1987. 'Taking the State Seriously'. in Weiner. M. & Huntington. S., eds..
Understanding Political Development. Boston: Little, Brown.

Nordlinger, E. A. 1988. 'The Return to the State: Critiques'. American Political Science
Review 82. 875-885.

Olsen, J. P. 1985. •Nyinstitusjonalismen og statsvilenskapen', Statsveienskaplig Tidskrift SS.
1-14.

Olsen, J. P. 1988a. "Administrative Reform and Theories of Organization". in Campbell. C.
& Peters, B. G., eds.. Organizing Governance. Governing Organizations. Pittsburgh:
University of Pittsburgh Press.

Olsen. J. P. 1988b. 'The Modernization of Public Administration in the Nordic Countries:
Some Research Questions". Hallinnon Tutkimus 7. 2-17.

Olsen, J. P. 1988c. 'Political Science and Organization Theory. Parallel Agendas But Mutual
Disregard", Bergen: Norwegian Research Center in Organization and Management and
University of Bergen. Memo.

Pedersen. 6. K. 1988. 'Den sønderdelte stat. Om interesseorganisationernes integration i den
offentlige forvaltning og effekterne heraf. Statsvetenskaplig Tidskrift 91. 259-278.

Pedersen. O. K. 1989. 'Fra individ til aktør i struktur. Tilblivelse af en juridisk rolle",
Statsvetenskaplig Tidskrift 92, 173-192.

Pedersen. O. K. 1990. 'Learning Processes and the Game of Negotiation. Wage Formation
and Negotiated Economy in Denmark". Paper. Seventh International Conference of Europeanists,
DC.

Posner. R. A. 1977. Economic Analysis of Law. 2nd edn. Boston: Little. Brown.

Smith, R. M. 1988. 'The Political Jurisprudence. the "New Institutionalism". and the Fulure
of Public Law". American Political Science Review 82. 89-108.

Stem. P. C. 199(1 'The Social Construct of the Economy', Challenge 33. 38-45.

Teubner, G. 1983. 'Substantive and Reflexive Elements in Modern Law', Law & Society
Review 17, 239-283.

Teubner, G. 1984. 'Enterprise Corporatism: New Industrial Policy and the "Essence" of the
Legal Person', American Journal of Comparative Law 36, 130-155.

Teubner, G. 1985. 'Social Order From Legislative Noise? Autopoietic Closure as a Problem
for Legal Regulation', Florence: EUI Colloquium Papers, doc. lUE 322/85.

Teubner, G. 1988. 'Corporate Fiduciary Duties and their Beneficiaries', in Hopt, K. &
Teubner, G., eds., Corporate Governance and Directors' Liabilities. New York: Walter de
Gruyter.

Tuori, K. 1988. 'Legitimacy: The Problem of Modern Law', Paper, Symposium on the
Legitimacy of Law in Modern Society. Murikka.

Weber, M. 1978. Economy and Society. Berkeley: University of California Press.