Collective Identities and Citizenship

In the following I tI)' to show that there is an attractive notion of citizenship which is independent of any kind of collective identity. Citizenship is understood as one fonn of social interaction. It is argued that in order to understand social interaction adequate1y, it is necessary to introduce a structuraJ. theory ofpractical rationality, to the extent this conception of citizenship is based on the theory of structuraJ. rationality. 1. Collective Identities I will not try to define what a collective identity iso For our purposes it should suffice to name some necessary ingredients of collective identities, some essential characteristics, or to put it in a methodologically less misleading way it should suffice to name conditions which must be fulfilled by a set of persons in order to make it possibleto attribute a collective identity to this set. There are four minimal requirements for a set of persons to form a collective identity: (1) We-intentions form their behaviour to a certain degree.1 (2) The persons out of this set share a common conception 0/ the good. There are common ideas about a good life. Abstractly speaking, the individual value functions share some common traits (value functions in the sense decision theory allows to attribute if certain conditions of coherence of preference are fulfilled). (3) The persons out of this set perceive the collective they form with each other as a quasi-person: they perceive the collective as having interests, heing able to act, competing with other collectives, having ahistory to tell etc.2 (4) There is common knowledge concerning (1), (2), (3). I 'We-Intentions' in the sense R. Tuome1a has introduced and discussed them in detail during the last decade, cf. Tuome1a 1984, chap. 2 and chap. 5; 1989,471-496; TuomelaJ Miller 1985,26-43; Leist 1985,180-205, andNida-RQmelin 1986, 96-108. 2 Whether it is constitutive for a collective identity that other persons outside the collective perceive it as a quasi-person, too, is an interesting question. I personally think that internalism is adequate regarding individual persons (only I myse1fknow for sure whether I fee1 pain or not etc.). In analogy there might be collective identities which nobody outside perceives as such. One should not tI)' to defme collective identities by behaviourist criteria. Analyse & Kritik 15 (1993), S. 120-128 © Westdeutscher Verlag, Opladen Collective Identities and Citizenship 121 The nineteenth century notion of citizenship requiring anation state has all these ingredients of a collective identity. The Hobbesian ideal of an (almost) unlimited sovereignty of state power and its implications such as the principle of non-intervention in international affairs, have become an essential part of this notion which in the realm of political practice is still quite influential .if one regards recent history. The Hobbesian ideal of the state as a quasi-person can easily be transformed into a democratic one if one requires additionally certain rules of aggregating individual preferences to state preferences. 3 Multiculturalism in some versions at least tends to prolong this identity notion 0/ citizenship in segregating the unitarian ideal into many different collective identities loosely bound together by some structural framework of institutions. Citizenship means belonging to one of these collective identities. Individual identity is to a certain degree constituted by the respective collective identity.4 In the following I will concentrate on the relation between the theory ofpractical rationality and the notion of citizenship in order to discuss the role and status of collective identities within such a notion. 2. Structural Rationality5 Traditional rational choice theory is based on a consequentialist notion of practical rationality which is still the predominant view if not in philosophy, then at least in social sciences in the English speaking scientific community. The consequentialist notion of practical rationality has a certain intuitive attraction because at least in modem society humans understand themselves as actively changing the conditions in favour oftheir own well-being. An act is chosen for its consequences and persons are responsible concerning actions in the sense of being responsible for its consequences.6 Modem decision theory seemed to back up the consequentialist notion of practical rationality in showing that coherence of preferences make it possible to attribute a utility function which is maximized by acting in 3 The fact that there are severe limitations to any kind of democratic aggregation as collective choice theory cf. the Arrow-, Senand the GibbardlSatterthwaite-results has shown, till now has not successfully eliminated this notion of citizenship, cf. my 1991a, 184-203; Arrow 1963, chap. Y.; Sen 1970, chap. 6 and chap. 6*; Gibbard 1973, 587-601; Schmeidler/Sonnenschein 1978,227-234; KernlNida-Rümelin 1994, chap. 3, chap. 5 and chap. 11. 4 Interestingly enough there is a common element in libertarianism and these versions of multiculturaiism, cf. Nozick's sketch of a libertarian utopia in. the 3rd part of Anarchy, State and Utopia, New York 1976. 5 The background of this conception is the critique of consequentialist notions of rationality I developed in 1993a. 6 Certainly there are more or less sophisticated versions of this idea of responsibility. Even the early legal systems leamed soon that an actor can not be held responsible for every consequence of one ofhis actions but only for the foreseeable ones. So there emerged more subjectivist notions of responsibility without changing the basic consequentialist approach. 122 Ju/ian Nida-Rflmelin accordance with these preferences. That coherence does not imply consequentialist rationality, as the traditional interpretation of the utility theorem states (cf. Luce/Raiffa 1957, chap. 2 and appendPt 1), has no immediate evidence in its favour. Consequentialism understands acting as a way of changing the world. To choose an action means to change the world in acertain way. Since our knowledge about consequences is limited, choosing actions means to accept a certain lottery of worlds. Maxjmizing expected value of consequences, is the Bayesian answer to limited knowledge. Bayesian consequentialism is the traditional interpretation of the rational choice approach. The behaviour of persons and groups in a society exposes different rank orders of possible worlds as the ultimate consequences of action. This multitude of interpersonally different rank orders causes problems of coordination and especially cooperation in a society. From a consequentialist point ofview it is the task of an institutional framework to provide some basic restraints in order to make social cooperation possible. The traditional view of rational choice takes the system of sanctions based on an institutional framework as the only way to change the outcomes of individual and group behaviour such that social cooperation is guaranteed. But if one takes this claim seriously and thioks of the characteristics such . an institutional framework should have to fulfiU this task, i. e. to force consequentialist optimizers to cooperate because cooperation now has become the strategy which maximizes the individual (or group-)preferences over outcomes (resp. the rank order over possible worlds), it becomes clear that such an institutional framework would be quite horrible. The structures of everyday interaction show that sanctions play a minor role for providing cooperative behaviour. If somebody begs me to elose the door, I elose the door usually because she begged me to do it and not because I am afraid I will be 'punished' in one way or the other if I don't. In general I tell what I think to be true not because I fear the disadvantages I might face if it turns out that I lied. A careful analysis of everyday discourse would certainly show that there is a complex system of rules of interaction whiCh are followed for theil' own sake and not because not following them is sanctioned in one way or the other. It might be that we learned to follow these rules as kids in being sanctioned if we didn't, but this does not mean that we follow the rules because the outcomes of actions are determined by sanctions. The Gricean idea to differentiate between logical inference rules and implicatures7 and likewise the approach of discourse ethic to base moral validity on rules which we implicitly follow (or accept) in communicating (Habermas 1983), allow for some insight into the complexity of the system of constitutive rules of interaction and especially cooperation. Grice and Habermas share the common conviction that social interaction is linguistically constituted. Linguistic analysis can in fact show convincingly that social interaction in general and cooperation specifically is constituted by rule conformity for its own sake. Communication would immediately break down if a miniihal degree of rule conformity were no longer be realized (see Lewis 1975, 3-35). 7 Cf. Grice 1975,41-58, and the paper ofStrawsonIWilsonlSperberlHintikka 1986. Collective Identities and Citizenship 123 To a high degree we act in conformity to rules. Rules are necessary not only for social interaction but also for coordinating one's own life with its different social roles, values, competing obligations and projects. Even more a certain rule conformity is constitutive for action itself in the sense that the identification of action is constituted by mIes. There is a possible 'intellectualist' misunderstanding: these mIes govern behaviour, but the acting persons do not intentionally follow these mIes in general. The plurality of reasons persons have for their actions displays the system of mIes their behaviour is based on, but it is not identical with some adopted normative framework. Following mIes does not imply knowing the mIes one conforms to. Only some parts of this system of rules have found its way into our everyday language. Rules like truthfulness, promising, warning and all the other objects of speach act analysis might be named here. But there are hints, there are gestures, there is a wide range of nonverbal behaviour which follows mIes, too, and which in general has not found its way into our everyday language. In philosophical discourse there is a usage of the term "institution" which covers all these elements ofthe system ofmIes governing social interaction. If one does not associate explicitness and temporal invariance with this term, it can help for our task to introduce a notion of citizenship insofar as the term already indicates the close connection between social interaction in general and politically explicitly institutionalized interaction as it is constitutive for citizenship. Intentional behaviour is governed to a high degree by the motivation to be in conformity with certain structures of acting. Only if Iassurne that in general one should be truthful, do I have good reason to tell the truth even in cases where I may have a personal advantage not to tell the truth. I favour a certain rule if I take the conformity to this rule as a good reason.On the other hand it can be shown that even a person who is interested only in her own well-being, fares better if she acts in conformity to certain rules. Optimizing one's own (longterm) well-being by every singular act would neither maximize well-being collectively nor even individually (see my 1993a, part 4). For this reason, even a merely consequentialist notion of practlcal rationality has to be enriched by structural elements in order not to be individually or collectively self-defeating (cf. Parfit 1986). I personally think that the first step towards a structural version of optimization requires a 'second step' towards a deontological theory of the good. If something is done not because it has best consequences but because it fits into a structure which has best consequences, being in conformity to a certain structure gains intrinsic value. The theory of the good can not be confined to values independent from the structural traits of life-forms, modes of interaction and social cooperation in general. Even if one begins in assuming some theory ofthe (extramoral) good as simply given, the next step in the argument is to accept structural traits as instruments of value maximization. Since a rational person under normal circumstances does not choose structures in general but only singular acts, these acts are not chosen as mere instruments for maximizing the good but as parts of a structure which itself is maximizing the good. The good cannot be understood as a given entity outside 124 Julian Nida-Rumelin any stmcture of individual and collective action and therefore the primacy of the good before the right cannot be defended coherently. Persons do many things for their own sake, and among these things there are some which cannot be understood as optimizing subjective value (or preference fu1fillment). 1 may be truthful for its own sake and not because 1 expect certain consequences by being truthful (which 1 favour). The full blown theory of the good is constituted by both teleological and deontological traits. Rational action is not instrumental for maximizing extemal values, and actions which are not done for the sake of maximizing values are not necessarily communicative actions. It is an illusion to think that there are two ideal types (or paradigms) of practical rationality: strategic rationality and communicative rationality. Instead a mere optimizing (strategic) notion of rationality is self-defeating and communicative action is but one example of a widespread variety of structurally rational types of action. Structural traits are constitutive for practical rationality, and not even in economics can one dismiss them since economic optimization, too, depends on more basic structures of interaction (among others also on communication), and therefore economically rational action can only be optimizing within certain structural traits economically rational action is not optimizing straightforward but is constrained maximizing, too.8 3. Citizenship Based on Structural Rationality Without some elements constraining individual optimizing, many projects essential for a good life cannot be realized and the advantage of cooperation in general is reduced. Structural rationality can at least partly be based on a system of constraining conditions, and this can be provided by a stmctural concept of collective rationality. To the extent that such a stmctural concept of collective rationality is applied in the theory of democracy, it can be combined with a concept well-known from the legal theory of H. L. A. Hart, the concept of secondary rules (cf. Hart 1961, chap. V). Primary mies prohibit and command; they grant members of a legal community rights and impose duties on them. Secondary mies, on the other hand, determine which procedure is to be implemented in a legitimate generation of primary mies. The line demarcating primary mies from secondary rules is not hard and rigid. Because the law of contract determines in which way new legal obligations are generated, Hart subsumes it under the category of secondary rules. Primary and secondary rules cannot be understood as two separate, disjunct sets of mies, because the generation of rules that direct the correct completion and fu1fillment of contracts is itself again bound to secondary rules. The difference between primary and secondary mies can better be formulated in the following way: One specific mle (or one specific set of rules) is secondary in relation to 8 I use here the tenn Gauthier introduced ("constrained maximizing"), even ifI think that his theory of uniting the optimizing model of rational choice and the rule following model does not work, as I tried to show in my 1993b, 53-74. Collective Identities and Citizenship 125 another specific rule (or another specific set of rules). Accordingly, the rules emerging from a legal contract are secondary to a legal obligation that implies concrete contractual conditions, certain rules given by a political institution are secondary to the system of rules by the legal contract, etc. Democratic legitimacy is a feature of collective decisions that accord with certain secondary rules. It is difficult to say in abstract terms which secondary rules are constitutive for a democratic system. As the theory of comparative government shows, the empirically realized variety of such secondary rule systems in parliamentary democracies is very wide. Tbe farnily resemblance between all different kinds of democratic systems involve, on the one hand, a minimal measure of formalized control of the political decision-makers and, on the other hand, a constitutive role for an informal normative consensus. While the first of these elements manifests itself in rules of common, fair and free citizens' elections guaranteed by constitutions and internal democratic structures within the respective system of political institutions, the second element of a democratic system is essential for the specific normative character of its political decision-procedures: The decisions made in these procedures depend on the presupposition of being at least fundamentally capable of achieving common consensual agreement. Citizens' consensus even if never in fact realized is indispensably intentio recta for primary rules issued by the legislator. Legislative projects are universalistically justified }Vith normative arguments like the common interest, political justice, economic efficiency, etc. The universalistic justification of primary rules is essential in a democratic polity. The claim of universal justifiability of primary rules a claim towards which every kind of communicative political acting in a democratic order must necessarily be orientated does not mean a claim for a de facta common agreement on them. A dissent over primary rules, typically appearing in different cognitivistically formulated opinions about the adequacy of specific (universalistically intended) normative criteria, cannot, as a rule, be eliminated by a democratic decision-procedure. This is simply due to the fact that a democratic system is based on a common consensus of a higher order, and this common consensus itself refers to the acceptance of secondary rules. At the level of primary rules common agreement is intentio recta. At this level, however, de-facto common agreement cannot be realized. Citizenship in democratic systems is constituted by a basic social consensus concerning the structural framework of secondary rules which determine under which constraints and with which procedures individuals' first order preferences should be realized. Structural features of interaction and collective decision constrain individual optimization. 0nly within the structural framework which constitutes democratic citizenship is there room left for not necessarily selfor group-oriented optimization. Optimization is constrained by structural features of a democratic order even if optimization is itself oriented towards universalistic normative goals, e. g. political justice. 0nly where the democratic character of a social structure as such becomes dubious, will the obliging character of its constitutive structural traits (secondary rules) also be questioned. In this case the crite126 Julian Nida-Rflmelin rion of optimization becomes dominant, and a Hobbesian helium omnium contra omnes (in terms of either conflicting interests or conflicting normative orientations) will only be prevented by achieving a new basic consensus on the fundamental features of the social structure as a whole. Democratic citizenship should not be based (qua aggregation) on the reductionist constitution of a collective actor, but on structura1 features of interaction constituted above all by secondary rules including the features of democratic decision-procedures. But for the establishment of structural rationality a system of formal rules of interaction and collective decision secured by means of penalties can never be sufficient: Structura1 rationality as a personal guideline of praxis is indispensable as a normative and motivationally effective element in order to constitute citizenship as the corpus-politicum of a democratic order. 4. Structural Rationality and the Common Good One essential feature of the theory of structural rationality is that, according to it, social cooperation is not based on a common good which is jointly maximized. The divergence of individual optimization on the one band and collective rationality on the other in problematic social situations (the Prisoners' Dilemma as ~the paradigm for this type of situations) can be solved and is in fact solved not in changing the valuations of possible outcomes, but in conforming to certain rules of cooperation.9 But if one dogmatically adheres to a consequentialist theory of practical rationality and a consequentialist ethic, then social cooperation and stable institutional frameworks can be understood only as the result of a prevalence of the common good, i. e. of common values maximized individually. Although this might be realized sometimes and in these cases seeming cooperation has in fact been transformed into individual optimization, there is. the alternative of genuine cooperation, which is a miracle only if you adhere dogmatically to consequentialism. Those who think dfAxelrod (Axelrod 1981; 1984; Hegselmann 1993) here, who is widely assumed to have shown that in general there is a convergence of individual optimization and cooperation, should recall that cooperation came about only under some extreme assumptions. One is that the probability of this interaction being the last one between the two players is below 0,3 %! So 333 interactions can be expected. This is quite an extreme assumption except if you are a married or an unmarrled couple. This notion of citizenship, based on the idea of structural rationality, is perfectly independent of any kind of collective identity. There is no need for collective quasi-persons (not even for rules of aggregation even if I personally think that one component of a normatively convincing notion of citizenship is the welfare state, and this certainly implies some weak form of aggregation), no need even for state sovereignity and a fortiori no need for competing collective actors as the 19th century idea of the nation-state assumed and strongly contextualist versions of multi-culturalism assume presently. The reality of a multi-cultural 9 I argue for this in 1991b. Collective Identities and Citizenship 127 society in my view is a good reason for a paradigm-change: from identity-versions of citizenship to structural ones. A structural understanding of citizenship will weaken collective identities and to me this· seems to be an attractive feature of citizenship based on structural rationality.