ARENA Working Papers
WP 00/25



Complex Democracy and the Obligation to Obey the Law

Agust�n Jos� Men�ndez


This paper aims at building the foundations of a deliberative or discursive theory of the obligation to obey the law. On the basis a critical approach to the most frequent arguments on the matter, namely philosophical anarchism and communitarian or associative theories of political obligation, an attempt is made to reconstruct the critical points of reference for determining whether we have an obligation to obey the law. It is argued that we must distinguish the general question from the practical one of deciding in concrete cases whether to abide by the law. This stresses the role of the obligation as an auxiliary of practical reason. It is further claimed that we need to distinguish between the legally imposed constrains derived from the social functions attached to law and the positive requirements for legitimacy of legal norms. The resulting theory of law's legitimacy is said to be complex because it relies on three complementary sources: participation, substantive correctness and guaranteed implementation.


This paper can be seen as preliminary work towards a deliberative conception of political obligation. The basic methodological assumption is that legal and political theory cannot avoid setting as the first issue in the agenda the legitimacy of political institutions. The main institutional assumption is that any relevant theory must be able to cope with the emergence of supranational legal systems. Sovereignty is no longer (was it ever?) a black box in which we can drop all theoretical and normative problems. In this sense, this paper might fit into a more general theory of the European Constitution. The underlying insight is that we can need to take on board the functional needs that the legal system as such is supposed to perform, while given good reasons for it to perform them. It is only from such standpoint that we can start building a positive argument for the general obligation to obey the law.

I. The general obligation to obey the law

This section poses in a more precise and explicit way the question whether there is an obligation to obey the law. We need to reflect on this because a positive or a negative answer is obviously conditioned by what is implicit in it, or what is the same, by how we characterise the obligation [1].

I make four basic claims. First, that the obligation to obey the law is a moral a question which should be posed as a general practical question and not as a narrowly legal one. Second, the answer to it should be seen as a basic auxiliary tool to our daily exercise of practical reason. Third, that the obligation is to be characterised as general, but marginally defeasible. Fourth, that the problem is roughly the same as that of the authority of law or the legal system.

As we will see in section II of this chapter, the present analysis allows us to focus exclusively on relevant issues and dissolves many of the apparently very strong disagreements, that are to a large extent due to the fact that authors are dealing with very different questions

1. A Moral, not a Legal Obligation

�1 The question whether there is an obligation to obey the law must be seen as a moral one, posed within general practical discourse [2]. A corollary of this claim is that we are not interested in considering the question as a legal one [3]. This is so mainly for two reasons.

First, the law cannot provide ultimate or categorical reasons for action. We just need to consider the individual who challenges her legal obligations and asks for a justification of them. When she faces a concrete norm, she asks why she should comply with it. The legal system can provide intermediate reasons up to a point, that correspond to the constitutional norm on which the validity of the whole system rests. But there is no reason why the individual should be satisfied at that point. When she asks why she should comply with the Constitution, no reference to the legal system can constitute a relevant argument. Kelsen's reference to the necessary endorsement of the hypothetical grundnorm or Hart's appeal to the sociological fact of endorsement of the primary norm of recognition by judges are simply not enough. Only moral reasons can provide ultimate reasons for action [4]. This implies that only by reference to moral reasons we can redeem in full the claim to legitimacy of law [5] (though we will see that the evaluation is a complex one that should take into account the peculiar tasks assigned to the legal system) [6].

Second, the existence of a legal obligation to obey the law is at the very least implicit in all legal systems. If a system does not claim that its addressees should comply with it, then it should not be considered as a legal one [7].

2. A Basic Assistant to Practical Reason

�2 The general obligation to obey the law is to be understood as a basic tool of our practical reason, as a valuable assistant to our moral powers.

The combination of limited moral powers and the unstoppable need for complex decisions which are morally relevant implies that our capacity to engage in practical reasoning is limited. Just think about the following facts, familiar to all human beings.

First, we are active beings. Most of the time, we engage into courses of action without much further ado. Even when we stop to evaluate the options open to us in normative and prudential terms, we do so quickly. Only in exceptional cases we stop to reflect at length. This implies that we exercise our moral powers in a selective way. Imagine if that was not the case. Just consider that we had to consider the merits of each single action, a bit like in the artificial reconstruction of utilitarianism as act-utilitarianism. The outcome will be that we will not be able to manage reality. We will end up paralysed. This just indicates that the sheer feasibility of practical reasoning requires that we are selective in its exercise. To do so, we need to map the sphere of action with the help of reflection on standard cases. This explains the differentiation of different layers of reasons for action, something further complicated by the needs of conflict-solving and coordination with others. At any rate, these different layers include reasons of different types, of variant strength and resistentace to be overriden by later reflection.

Second, we need law, at least to the extent that we interact with others [8]. However, we have also seen that law is authoritative and partially heteronomous. This implies an ambivalence. On the one hand, we observe that the functional tasks that law performs as a complement of morality require its autonomy vis a vis morality and general practical reasoning [9]. On the other hand, this implies a potential divergence between law and morality. Sometimes this might be justified by reference to the constrains which law imposes on general practical discourse. But other times this might be due to the fact that law is instrumentalised by interests which cannot be justified. Laws can be unjust. That is the basic insight of the moralistic or normative case for positivism. Such argument goes that we should stricly differentiate law as it is and law as it should be; if we do not do so, we run the risk of lacking a vantage point from which to criticise positive law. We could rehearse the arguments in order to qualify such conclusion, but for our present purposes, the gist of the claim is correct. Because it is precisely the realisation of the potential unjust character of law that makes it so important to exercise practical reason when we come across legal norms. And as we have just mentioned in the previous paragraph, we will be overloaded if we proceeded to pass judgment in a case by case basis. For such reason, our assesment of legal norms must be based on a default or general position, to be qualified or reviewed in concrete circumstance. This is rendered clear by the key pragmatic question which the general obligation to obey the law is expected to answer: whether we should comply with an unjust norm belonging to a system which is considered, as a whole, to be just [10].

This explains why we need to structure the exercise of our practical reason in several layers. This corresponds to Rawls' insight that we need to distinguish justifying a practice from justifying a particular action falling under it [11], Raz's claim that we need to differentiate first and second order reasons [12] or Hare's plea for “some, simple, general and more or less unbreakable principles (...) to keep us from special pleadings and other errors when in situations of ignorance or stress” [13]. It also makes clear that the question whether there is an obligation to obey the law must be seen as a clearly practical, and even pragmatic, one.

3. A general but defeasible obligation.

A) The general character of the obligation to obey the law

�3. When we act, we come across concrete legal norms and generally with no “obligation to obey the law” as here defined (that is, as a practical, not legal question). For example, imagine that a bunch of friends decides to open a bookshop, and to do so under the specific regime of a co-operative. Not knowing how to make heads or tails of legal arguments, they decided to ask to a legal attorney how to proceed. The latter will probably refer to a set of legal norms which they have to observe, from city council regulations concerning health and security to national labour laws on co-operatives. But it will be rather strange that he will say that “you have to obey the law” in general terms, and even less in moral ones (unless this is a patronising sort of attorney or the place where the book shop is open is a quite peculiar one). But now imagine that the friends want to open the book shop in a country where importation of books is burdened with a 100% customs duty. When told about this norm by the attorney, friends could doubt whether the norm is just or unjust. Even if they might end up complying with the law, they might question it. The attorney will be of no help. The measure might be reasonable (for example, it might be designed to protect the emerging publishing industry of the country from dumping from other states) or it might be not (it could be due to a willingness to enrich one of the ministers, who happens to be in the publishing industry). At any rate, the friends will face the practical question whether to comply or not with the law (they could travel abroad and sneak books in, and later on sell them “under the counter”). What is clear is that this question requires a larger framework, which is characteristic of the general obligation to obey the law. But how large should it be? Should they balance the unfairness of the concrete norm with the fairness of the customs duties code? or with the rough fairness of civil laws? or penal laws? or the constitution?

Notwithstanding the relevance of intermediate layers (as the general obligation to pay taxes itself), the general obligation to obey the law must be related to the legal system as a whole. This is so because law makes sense as a system and not as an heterogeneous congeries of norms [14]. Only as a system it discharges certain tasks and furthers certain values. The systemic character of law allow us to realise that the general obligation to obey the law is one of the basic secondary reasons which assist our practical reason. Complying with a legal norm is not an isolated affair, but one which has an impact on the viability of law as a social practice. Let us see this in more detail.

What we value most is not this or that norm, but law as a social order, or what amounts to the same, its capacity to discharge the basic functions of conflict-solving and social co-ordination. This does not require us adopting the extreme view of law as a seamless web (which tips the balance towards an unconditional acceptance of law's legitimacy) but to take into account the systemic dimensions of the problem, and to frame each specific question within the general obligation to obey the law. If the concrete legal system is worthy of our appreciation, we should take into account the fact that its capacity to discharge such functions is endangered beyond a certain threshold of disobedience.

�4. If the conditions for affirming the existence of a general obligation to obey the law are met, we can say that all legal subjects have an obligation to obey the law in all instances to which the law is applied. However, this does not necessarily imply that such obligation must be exception-less. That is so if only because the tasks assigned to law are not impossible to fulfil even if some people do not abide by the law. Moreover, weakening the force of the general obligation to obey the law is the only way to reconcile law as an authoritative order with individual moral autonomy (which is the reason to consider the general obligation to obey the law, to start with [15]). These two considerations move us to consider the general obligation to obey the law as marginally defeasible.

B) The Defeasible Character of the Obligation

a) Characterising the General Obligation as Defeasible

�5. The most controversial step in the characterisation of the general obligation to obey the law is the determination of what kind of reason the obligation is. To put it in different terms, the main source of disagreement concerns the way in which its relationship with other obligations is described. Let us consider once again the central case in which the general obligation to obey the law becomes relevant, namely when we consider that the legal system as a whole is fairly reasonable (and thus that there is a general obligation to obey the law), but we face a concrete case in which the applicable norm is unjust. What role does the general obligation to obey the law play in such cases? Does it give rise to an obligation to obey the law even in such cases?

To exhaust the different ways of considering the same idea, the question in technical terms is what kind obligation the general obligation to obey the law is. Is it to be characterised as a prima facie one? (a sort of rule of thumb that compiles and balances all relevant reasons in abstract terms, but which needs to be balanced against concrete reasons that might be relevant in the context of application). Or as an absolute obligation (that requires us to exclude any reason that might be seen as relevant in the context of application)? Or as something in between, like Raz's exclusionary reason? (which implies that the obligation is seen as a secondary reason which excludes the possibility that its addressees can offer most reasons in contrast to it)? Some remaining arguments can exceptionally be invoked against the general obligation to obey, in which case we can balance them against it and eventually not obey the law.

�6. We have good reasons to disregard the characterisation of the general obligation to obey the law as a prima facie or as an absolute reason to obey the law.

On the one hand, if the obligation to obey the law was seen just as a prima facie one, it would not provide us with anything other than a sort of point of departure in our analysis of whether we should obey or not the law. The law must be something more than just a gatekeeper of the moral agenda if we want it to perform the roles of conflict-solving and social co-ordination successfully. This is something (but not everything) that is implied in Raz's argument that legal norms are reasons for acting, and not merely statements to the effect that there are reasons for acting [16].

But if we consider it as providing an absolute reason, then we cannot any longer see it as part and parcel of the exercise of our practical reason, but as an alternative to it. We will see that it is precisely because philosophical anarchists characterise the obligation in these terms that they come to the conclusion that it is impossible to reconcile the two.

�7. We could turn to an alternative characterisation of the obligation as an exclusionary reason, on the basis of the definition of the latter provided by Raz [17]. Let us consider what Raz understands by it.

First, he distinguishes between first- and second-order reasons. The latter are defined as reasons to act or refrain from acting for a reason [18]. Second, second-order reasons and first-order reasons do not intermingle in our exercise of practical reason. This implies that when they enter into conflict, secondary reasons simply prevail [19]. This means that the relationship between second- and first-order reasons is not a matter of weight, but of simple priority. Third, that does not mean that the dimension of weight is unknown to second-order reasons. However, weighing and balancing can only take place among second-order reasons themselves [20]. Fourth, second-order reasons might be cancelled by “cancelling reasons”, and its scope (or jurisdiction, to say it in more legalistic terms) might be affected by what Raz names as “scope-affecting reasons” [21]. Fifth, exclusionary reasons are a special type of secondary reasons. They are reasons for not acting for a reason [22]. They can be represented as a sort of brackets that exclude consideration of first order reasons altogether.

This characterisation is fruitful when applied to the general obligation to obey the law. First, it brackets most reasons falling outside the legal domain that are contrary to what is prescribed by the law itself. If we state that somebody is under an obligation to abide by the law, we stress the fact that she should no longer weight and balance all relevant reasons. This insight is clearly connected with the value we ascribe to law as a complement of morality. Second, the idea of an exclusionary reason fits better than that of an absolute reason in the pragmatic conception of the obligation to obey the law. It precludes direct weighting and balancing, but it does not rule out a margin of exercise of practical reason. This is enough to render compatible the obligation with individual autonomy.

The application of the concept of exclusionary reason to the general obligation to obey the law puts us on the right track. This could be rendered even more adequate if we take on board some of the criticisms addressed by Cha�m Gans to Raz.

The core of Gans' criticism is oriented to the complete isolation of second- and first-order reasons, which prevents weighing and balancing across classes. In the case that some reasons not excluded by the general obligation to obey the law are relevant in the concrete context of application, he argues that we have no other choice but to balance them against the reasons to abide by the law [23]. This is so because we cannot have a list of reasons which may be violated or which may not be violated by it in advance. It is not even likely that we can foresee all reasons of this sort in advance [24]. On this basis, it seems that the general obligation to obey the law does not exclude weighing and balancing, but it is a reason that refers to the value of having rules independently from the content of such norms [25].

�8. If that is so, the general obligation to obey the law is to be characterised as a general but defeasible one. This is the main idea behind the way in which authors such as Nino structure the way in which we put the question whether we should obey the law or not. This implies that we should distinguish at the very least two different steps in the judgment whether there is or there is not an obligation to obey the law. Or what is the same, when we face a concrete or specific legal norm, we should not decide the question whether we are under an obligation to obey the law in a direct way, or by means of a direct moral assessment of the norm in question. First, we should consider whether there is a general obligation to obey the law, or what amounts to the same, we should take into account the functions performed by the legal system and the values furthered by it. Only then should we go back to the question whether there is an obligation to obey the specific legal norm relevant in an specific context. The answer will be framed by the existence (or non-existence) of the general obligation. Notwithstanding the extent to which the affirmation of the latter tips the question in a favourable direction, the final answer remains open. It is a matter to be decided by the individual with the help of her practical reason.

b) Objections

�9. We now turn to its basic practical relevance. As it has already been said, the general obligation makes a real difference in those cases in which we face an unjust law which belongs to a system that can be said to be just in general or rough terms. In brief, it saves us from the need to go through the same steps of reasoning each and every time that we come across what looks to us to be an unjust norm belonging to a constitutional framework judged to be just as a whole; that is, one within which it is possible to solve conflicts and co-ordinate action in a roughly fair way.

�10. Moreover, there is no incompatibility between the adequate discharge of the basic roles assigned to the legal system and the characterisation of the general obligation to obey the law as defeasible. Conflict-solving and social co-ordination do not need to be matched by an exception-less obligation to obey the law, but only by a massive compliance with the law. This means that marginal non-compliance with the law does not pose a threat to the legal system (even if those who infringe the norms can get away with it). Some extremely prudent legal theorists notwithstanding, we can doubt that the fact that some car drivers cross against a red light poses a serious threat of anarchy on the road, or that the fact that some people devote their lives to burglary places pensioners under the risk of a sudden lack of respect for their private property. It is only when lots of people cross against red lights or crime rates are rampant that the collapse of the legal system becomes a real possibility. Of course there is a threshold at which non-compliance undermines individual motivation to comply and we risk a systemic breakdown. Bu that is not to be blamed on the characterisation of the obligation to obey the law as a defeasible one.

�11. A familiar argument is that law makes a claim to supremacy in the form of the claim to correctness. For that reason, we have to characterise the general obligation to obey the law accordingly [26].

Before tackling this objection, let me specify how the claim to correctness tends to be interpreted. It is said to include three basic small or sub-claims to normativeness, to reasonableness and to peremptoriness. First, the law claims that it should be taken as a system for the guiding of action and not only as a system of knowledge (normativeness). Second, it pretends to take into account all relevant reasons applicable to the each case and, after balancing them properly, to prescribe the correct course of action. This amounts to asserting that a moral judgment which reproduced the content of the law would be a correct one (reasonableness). Thirdly and finally, it requires addressees of the law to take into account only what is prescribed as a reason for action (peremptoriness). We will have a chance to see in a moment that this has been expressed through the idea that law requires compliance because it says so.

Assuming for the sake of the argument that this correctly summarises law's claim to supremacy, it is far from clear why we should characterise the general obligation to obey the law by reference to what the legal system itself claims it to be. Not only has it already been remarked that the question makes sense only from a moral and not from a legal point of view, but the question whether there is a moral obligation to obey the law in the terms established by the legal system itself has little to do with the practical concern that moves us to formulate the question to start with.

The apparent assertion of individual autonomy by means of denying the existence of a general obligation to obey the law comes at the price of leaving us ill-equipped to deal with the related question of whether to comply with a norm, thus risking the overload of practical reason.

4. Is there really a difference between the question of the obligation to obey the law and the issue of legitimate authority?

�12. We could opt for a different approach to the question of legitimacy of the legal system. Instead of focusing our attention on the question whether there is a general obligation to obey the law from the perspective of the legal subject, we could wonder whether the claim to authority on the part of the legal system is legitimate.

The two questions are not exactly the same. We might find instances in which those in charge of public offices could legitimately request compliance with a general rule, while its addressees (or at least some) would be equally justified in not complying with it. If the two questions admit different answers in some instances, we cannot have better proof that they cannot be exactly the same question.

However, in general and substantive terms, they can be regarded as different perspectives from which to tackle the same substantive problem [27]. Normative preference shall be given to the justification of norms to each and every individual requested to comply with them; hence, legitimate authority should be mainly defined by reference to the individual obligation to obey the law (and not the reverse). The marginal cases in which the two questions may receive a different answer would be those in which we are concerned with the moral questions faced by those individuals invested with legitimate authority.

�13. It is quite likely that special reasons to obey the law are applicable to public officials. Their position (which has normally been voluntarily undertaken) implies an explicit endorsement of the legal system. Moreover, they are normally in change, in some way or another, of monitoring individual compliance with the legal system, and, finally, their non-compliance would undermine severely individual motivation to comply. This is specially true with regard to obligations related to co-ordination functions. When individuals see public officials acting in violation of the law, their sense of hopelessness and their distrust in the ability of law to discharge such tasks effectively increases considerably. However, I will not consider in detail the special reasons which might justify the obligation to obey the law of the individuals invested with legitimate authority.

II. For and against the general obligation to obey the law

In this section, I will review two of the main lines of reasoning concerning the general obligation to obey the law, namely philosophical anarchism and the conceptualisation of political obligations as associative.

Before dealing in detail with each of them, the reader should notice that both theories deny that the question whether there is an obligation to obey the law is a really substantive one. Both of them claim that there is no meat in this theoretical pie (and who cares about a meat pie that is meat-less?). Philosophical anarchism considers that the question has always had a negative answer, because it is simply impossible to reconcile individual autonomy (implicit by the phrasing of the question with the help of concepts like “obligation”) and compliance with the law. Hence, the question is a oxymoron or impossibility, of the kind of boiling ice-cream or Aztec equestrian arts. Those who conceptualise the obligation as associative are of the contrary opinion. The question must be answered in the positive. Their line of reasoning tries to show that a positive answer is already implicit in either our political vocabulary (e.g. not acknowledging that a citizen has an obligation to obey the law implies not fully understanding what being a citizen means) or in our constitution as moral beings. Hence, the question is redundant or superfluous, of the like of whether trees are wooden or books have pages.

The general framework outlined in the previous section allows for austerity in the development of arguments, to focus our attention on the weak and strong points of each argument.

1. Philosophical Anarchism: Morality and the Subject

A) The tension between autonomy and law

�14. Any critical morality worth its name assumes a strong conception of individual autonomy, and defines it in some way or another as a capacity to give oneself rules in practical matters (that is, in those cases in which she has to decide what to do, how to behave, etc....) [28]. Whether this process takes place in a monological or dialogical form (a main difference between the Rawlsian and the Habermasian models, that make reference to the concepts of public reason and of communicative reason) does not alter the necessary reference to a strong conception of individual autonomy at the core of it.

�15. The dilemma raised by the philosophical anarchist is this: Once we take individual autonomy seriously, an individual's capacity for discriminating right from wrong in practical matters, law as a social order is under suspicion, because it seems to rule out individual autonomy. If law is to perform its task as a functional complement of morality, if it is to do its work at conflict-solving and social co-ordination, it should not only provide us with rules of action, but it should prevent individuals acting on the basis of reasons or norms outside its domain. To put this in different and more sophisticated terms, law must pre-empt action based on general practical reasons, the reasons that might derive from the monological or dialogical exercise of autonomy by legal subjects. If that is so, what is left of our individual autonomy?

The philosophical anarchist, with an “I-told-you-so” written between the lines, will just reply “None!”, and claim that (a) there is no way of solving the dilemma; (b) therefore, we cannot ground a general obligation to obey the law.

�16. Just consider one of the most famous formulations [29] of the claim, the one made by Robert P. Wolff in his In defence of Anarchism:

“there can be no resolution of the conflict between autonomy of the individual and the putative authority of the state. Insofar as a man fulfils his obligation to make himself the author of his decisions, he will resist the state's claims to have authority over him. That is to say, he will deny that he has a duty to obey the laws of the state simply because they are the laws[30].

Wolff considers that we are faced with a tragic choice. Either we are autonomous or we obey the law. It is not possible to reconcile the two. Accepting the general obligation to obey the law is tantamount to throwing away our moral powers. Once we are prevented from criticising a norm without changing its status by such operation, we renounce our autonomy. From that he smoothly comes to the conclusion that “anarchism is the only political doctrine consistent with the virtue of autonomy” [31].

There are other versions of the argument which can be said to be less ambitious. John A. Simmons, in his Moral Principles and Political Obligations, denies the existence of an obligation to obey the law on the basis of the rejection of all main positive grounds on which such obligation has been said to be founded (such as consent, tacit consent, acts implying consent, communal obligations, the natural duty of justice and so on). Although his argument does not rule out the possibility of finding an alternative foundation, Simmons says that those arguments are not capable of founding the obligation while at the same time preserving the autonomy of the individual. Moreover, his denial is in part due to the emphasis he puts on the obligation to obey the law as a political obligation, as part of the requirements “to support and comply with the political institutions of one's country of residence” [32]. In fact, he is interested in the special bond which connects the individual and the political community (or communities) of which she is a permanent member, in some way or another (either citizenship or permanent residence).

In other instances, philosophical anarchism is associated with the Marxist claim that law is just an instrument of class oppression, that is, it merely reflects the bargaining power of those who have an influence in its design. “The law cannot offer an ethical justification of its demands neither on the basis of its formal structure nor on the basis of its substantive content (...) the law is a coercive order of a historical nature, which reflects the conflict between concrete interests and the dominance of some over others(...) it is an instrument for the domination of one class and its interests over other classes and their interests” [33].

B) Some counter-arguments to philosophical anarchism

�17. Philosophical anarchism poses a major challenge to any attempt to found a general obligation to obey the law. It constitutes itself as a champion of the fundamental value of individual autonomy, and it denounces the fact that law forces us to surrender the exercise of our practical reason, that it attempts to provide us with categorical reasons and that if nothing works, it will be ready to coerce us to comply without further ado. However, this is an overrated argument. It seems to me that philosophical anarchism relies on an unconvincing conception of autonomy when applied to political matters, that it exaggerates what the law expects from us, that it provides an inadequate description of the general obligation to obey the law and, finally, that it models the relationship between law and coercion in a plainly wrong way.

�18. Firstly, it can be argued that philosophical anarchists assume a flawed concept of autonomy [34]. They seem to argue that autonomy means the same in all contexts. It does not matter whether we are dealing with the conception of our good life or trying to find reasonable norms to deal with basic social conflicts and co-ordinate action within society. In the next section, it is argued that the nature of political issues renders this unified conception of autonomy implausible. This is so because the standard of political legitimacy requires the respect for the autonomy of all affected by the political decision, and this is incompatible with understanding law-making as a matter of unrestrained self-legislation. On this basis, it will be argued that public autonomy must be interpreted as sharing the authorship of such rules, and this means participation in terms of equality and symmetry in the process of deliberation and decision-making that lead to the enactment of the rules, and furthermore, that these procedures take seriously the character of the issues at stake (which is turned into a preference for procedures with an epistemic privilege, or in other words, which have a propensity to select the right solution). If we understand public autonomy in this way, a further obstacle to reconciling the general obligation to obey the law and individual autonomy disappears.

That there is something inadequate in the way in which philosophical anarchists conceptualise individual autonomy can be illustrated by showing that if we take their arguments seriously, we should not only condemn law as unavoidably heteronomous, but also other alternative forms of determining common action norms, such as the social practice of contracting. The argument is specially relevant given the fact that we could think about such practices as an alternative to law. Leslie Green [35] has stated quite persuasively that this avenue is closed to philosophical anarchists. On the one hand, any social practice needs to be constituted and defined by certain norms. Saying that they have been decided by a previous contract has the air of a paradox. Given that law is excluded, then what? On the other hand, and more to the point, contracts are binding agreements on how to conduct our affairs in the future. But binding my future autonomy on the basis of an exercise of my own autonomy seems also to be against a strong conception of autonomy defended by the philosophical anarchist. The underlying problem seems to be once again the conception of autonomy.

�19. Secondly, we could say that philosophical anarchists formulate the requirements of the obligation to obey the law too extremely. This allows them to gain some dramatic momentum, but it does not contribute much to their case in substantive terms. They do not distinguish properly between the requirement to relinquish our moral judgment to the law (which is what they say the obligation entails) and the weaker claim to adapt our external behaviour to the action-rules established by the law (which is what it really is about). But why should the law ask us more than to act according to it? [36] By recognising this, we start to undermine the case for a too rigorous understanding of the obligation.

�20. Thirdly, philosophical anarchism makes a further and fatal mistake. Philosophical anarchists characterise the general obligation as exception-less. They consider that we are engaged in determining whether we have to obey the law understood as a source of categorical reasons for action. We have already argued that this constitutes a poor understanding of the obligation. If we construct it in such way, it is clear that there could be no general obligation to obey the law. What is not so clear is why we should conceptualise it in such way. It has been argued that this is not required by the functions performed by law, or by the practical role that the obligation plays in our practical reason, or for this purpose by a correct understanding of law's claim to correctness. On the contrary, understanding it as a defeasible obligation has the advantage that renders it potentially compatible with individual autonomy. We have also indicated why the availability of a general obligation to obey the law plays an important r�le in the structuring of our practical reason.

�21. Finally, it can be argued that a further mistake is associated with the way in which the role of coercion is depicted. Law tends to be mainly associated with the gruesome role of coercing individuals into compliance. For this reason, coercion is reputed to be at the core of law. However, this claim can be challenged with a different conceptualisation of the relation between law and coercion. Though it might be too cumbersome to spell out the argument in full detail, at least two remarks must be made. One is that (at the very least) some of the functions played by law would be necessary even in a brave new world in which people will find motivation to comply with any legal rule in the absence of sanctions for recalcitrant subjects. These functions are mainly associated with its role in co-ordinating action in order to achieve complex social goals [37], but a similar argument could be made concerning its role in conflict-solving. The other is that we could (and should) consider sanctions and punishments as mainly addressed not to the bad man but to the co-operative citizen, ensuring her that her readiness to comply with the law will meet with similar compliance by most citizens. In this view, the main problem tackled by coercion is not that of free-riding, but that of hopelessness. This allows us to realise that coercion does not only pose threats to autonomy, but at the same time can play an essential role in ensuring the conditions for its exercise, stabilising the expectations about the behaviour of other individuals with which we interact some way or another. This shows that we need law in order to constitute our autonomy, even if it coerces us, and even if we have to fight this same coercion which helps us to constitute our autonomy [38].

C) Points to be retained from philosophical anarchism

�22. These are arguments enough as to leave aside the claims made by philosophical anarchists � la Wolff. The previous argument shows that the question whether there is an obligation to obey the law is a valid one. Even if this is what we should come to think at the end of the day, it is important to take stock of some basic insights provided by philosophical anarchism. All these insights are associated with the strong conception of autonomy asserted by philosophical anarchism. However, these insights may stand even if this conception of autonomy falls. Let me say some words about three of them. First, their argument stresses the irreducibly individual character of morality. Even if the conception of autonomy is inadequate, philosophical anarchists are right to claim that at the end of the day, when all is said and done, moral judgment is an individual affair. Not only must the individual take full responsibility for her acts, but any claim to legitimacy on the part of a social institution must be redeemed for each and every individual. Second, this argument might derive wrong consequences from an essentially correct claim. This is that law and morality cannot be reduced to each other. Although they are related in different ways, they are two different social orders. Third, claims made by these insights can be reinterpreted as a refutation of a general obligation to obey the law if understood as without exception. In this sense, it constitutes a sort of appendix to any attempt to found a general obligation to obey the law.

D) Joseph Raz's paradigm of legitimate authority and his denial of the existence of a general obligation to obey the law

�23. The work of Joseph Raz is a standard reference in the literature on the authority of law and the obligation to obey the law. In several of his books, he offers a thorough descriptive and normative conception of authority and also puts forward a model of what constitutes a legitimate authority. With regard to the general obligation to obey the law, Raz argues against its existence in general terms, although he proposes an alternative foundation to it, related to a certain understanding of membership of a certain political community and to the ways in which we can show gratitude to it [39].

There is something puzzling about the whole construction. The Oxonian professor proceeds to characterise the obligation to obey the law as a general and indefeasible one. This implies that it pre-empts that we act in ways contrary to what is prescribed by law on the basis of reasons lying outside the legal domain. That this is the right way of understanding the obligation he justifies by reference to a descriptive statement, namely, by describing what the claim to correctness made by the law itself entails [40]. Then he argues that there cannot be a general obligation to obey the law because the authority which law claims does not fit into the model of legitimate authority. The latter is associated with the understanding of authority as service. This implies that authority is limited in scope and subsidiary to the capacity of the individual for correct moral judgment. This is made explicit in the basic rule Raz argues for in this issue, the so-called normal justification thesis [41]. All this boils down to a sort of authority fragmented not only in objective terms(or what amounts to the same, the claim that authority regards some matters but not others) but also in subjective terms (if an individual is likely to be competent on a matter and get at right solutions, then we cannot justify any authority upon him on the matter). The authority that law claims is not a fragmentary but an absolute one. For this reason, we could never find a justification for the obligation to obey the law.

The move seems to be based on some normative reasons. We have already considered the positivist argument that the strict separation between law and morality is necessary for the sake of critical morality itself. As indicated earlier and now repeated, Raz is one of its main advocates. Moreover, he argues that the general obligation to obey the law gives rise to the danger of an uncritical assumption of authority. Not only does Raz invoke the potential perverse consequences of acknowledging a general and unconditional obligation to obey the law [42], but he insists that the idea that consent is adequate to bind ourselves to obey the law might mislead people into obeying the law when the right moral attitude would be to disobey it. The idea is that people might exaggerate the value of their own consent if that serves their interests [43].

However, Raz's argument does not rule out the possibility of justifying or arguing for the general obligation to obey the law, but only for this obligation understood in these terms [44]. It just denies that it is possible to found an obligation which rules out exceptions, that is, an obligation to which we could not find exceptions in particular cases. Most of the time, Raz's strategy is limited to showing that an exception-less obligation cannot be justified. We can then understand his claim that “To show that there is no general obligation to obey the law, we need only show that there are some central cases in which there is no such obligation” [45]. And his fears regarding the weakening of critical attitudes towards the law do not apply to a defeasible characterisation of the obligation, that entails a call to individuals to check whether law redeems its claim to legitimacy in particular instances.

In this sense, we can see that it is possible to reason with Raz against Raz, in the sense that we could use his arguments and contribution in order to defend the existence of a general obligation to obey the law understood in different terms.

2. Avoiding the problem: Communitarian theories of the Obligation to obey the law.

A) Identity, Gratitude and the Obligation to Obey the Law

�24. The second means of avoiding the question whether there is an obligation to obey the law consists in arguing that the very conceptions and phenomena of politics, political community, and membership of the latter necessarily imply a general obligation to obey the law. The whole issue that underpins the argument of philosophical anarchists (the tension between autonomy and the obligation to obey the law) is seen as out of place. Those who put the problem in such terms simply do not understand what being a citizen entails.

�25. Leaving aside some analytical theories (which claim that the linguistic usage points to the existence of political obligations, but provide no further justification why this must imply that government has authority or that we are to subject to it), the argument constitutes an application of the communitarian understanding of public reason to the problem at stake.

Communitarian critics of liberalism hold that our identity as moral beings is constituted by our membership of a given political community. In other words, it is a matter of assuming a ready-made role or set of roles within the community. This is coupled to the further claim that what shapes and constitutes me is not only the web of social relationships in which I am engaged, but also the duties that define my role. Both membership and role-taking are not voluntary. Communitarian theories of political obligation stress that in this respect, political membership and role-playing are more akin to intimate forms of relationship, like family or other sorts of kinship relationships. They further argue that each role has its point and purpose, defined by local practice and consequently independent of our will. The point is that because membership and role-taking are essential in the definition of our identity, the obligations deriving from them should also be seen as configuring or partially determining who we are. For this reason, reconciling the obligation to obey the law and autonomy is, as it was advanced, a non-issue for the communitarian. We can criticise the definition of the role, but only immanently, or what amounts to the same, on the basis of the same culture that defines it.

Each social role has its point and purpose [46]. Being a citizen entails showing gratitude to our community, something that requires our obedience to its laws. Not accepting this is to reject our identity, and whoever does so is at a loss.

B) Counter-arguments

�26. It seems to me that the basic counter-argument consists in the denial of the communitarian conception of public reason endorsed by these authors. But there are some other questions or arguments specific to their argument for the obligation to obey the law that can be subject to critical review.

�27. I will refer to two of these more concrete criticisms. First, it is not clear why communitarians conceptualise the obligation to obey the law and other political obligations according to the model of kinship relationships. The argument seems to be based on the unavoidable character of politics, but it can be sustained only if we argue that social relationships are either fully voluntary or involuntary. This is not correct. While we cannot decide whether or not we are going to deal politically with a certain issue (this is something which is based on whether certain actions affect others than the actor herself), we are free to decide the way in which we are going to deal with the issue. We cannot get rid of the form of law, but we are free to determine its contents. For this reason, we can subject to critical review the legitimacy of the contents of political obligation. Second, it is far from clear why the definition of political interaction on the basis of role-taking (which is quite plausible as an isolated thesis) must necessarily correspond to the definition of content of such roles by local practice. Even if we accept that membership of a political community is related to what we could call the role of citizen, we could wonder why we should accept the definition of citizenship provided by local practice and why we cannot enter into critical discourses about the proper rights and duties of citizens. For example, to the communitarian argument that being a good citizen necessarily entails obeying the law, one could counter-argue on the basis of Raz's argument that there are other plausible ways of showing a commitment towards our political community.

�28. But the case for a communitarian understanding of the general obligation to obey the law rests on its companion conception of public reason. The reader may allow me to refer to the main content of the critique. First, a critical understanding of practical reasoning can be perfectly sensitive to the fundamental role played by education at large in forging our critical abilities in moral terms and also in the acquisition of basic moral contents. Second, acknowledging the role played by socialisation and education does not imply that we cannot use moral judgment in a constructive way in order to revise the moral content that we have acquired. This capacity is tested by the assimilation of moral contents proceeding from different cultural traditions, with different conceptions of the good, to interact on the basis of an appeal to moral conceptions.

III. A Positive argument

In the previous section, we have made some arguments that seem enough to disqualify the two main “strategies of avoidance” of the question whether there is a general obligation to obey the law.

Philosophical anarchism can only deny that there could be a positive answer to the question by conceptualising the authority of law in an extreme way, not sufficiently attentive to the practical functions envisaged for the obligation itself. However, it was also argued that we would do better to take on board its main message: the irreducible individual character of morality. This must not be interpreted as denying the need of a communicative reinterpretation of practical reason, and more specifically with regard to political issues. However, the fact that we must differentiate two paradigms of autonomy (private and public one) does not mean that individual judgment, and individual judgment alone, is what matters when testing the legitimacy of any scheme.

Associative political obligation theories are too ready to give an unconditionally positive answer to the problem. On the basis of a dubious genealogy of practical reason, they end up always justifying the obligation to obey the law in all instances. This does not take on board the tension experienced by addressees of law themselves, not only with respect to thoroughly wicked legal systems, but also in respect of fair ones, when a given legal norm is perceived as unjust. Some insights, as is the case with philosophical anarchism, are precious. These theories insist on the non-voluntary component of the obligation, and the need to take it into account in any solution to the problem of the obligation to obey the law.

In the coming paragraphs, I consider the conditions under which it is possible to argue that we have a general obligation to obey the law. The structure of the section reflects the concern not to blur the differences between a number of problems related to the obligation. It seems to me that this requires distinguishing clearly between the general problem of justification (or better, explanation) of politics and the form of law and what can be seen as the argument proper for the conditions under which there is an obligation to obey the law.

Thus, in the first sub-section, reference is made to the question whether and when we have an obligation to deal with issues politically and not on the basis of individual judgment. This problem is associated with the delimitation of the spheres of private and public autonomy. In addition, it is considered whether we should make use of the linguistic and practical code of law in such interaction. Both questions reflect the insights provided by non-consent theories of political obligation, and especially those coming from the Hobbesian argument (which stresses law's contribution to the stabilisation of behavioural expectations). However, they are not considered as providing a full justification of the obligation to obey the law.

In the second sub-section, a three-fold argument is made concerning the characterisation of legitimate law. First, and at the core of the model, we have the idea that actual participation in the practice of deliberative democracy contributes most to legitimacy. Secondly, reference is made to fairness, or substantive rightness, based both on the thin conception of the good that underlies deliberative democracy (explaining why we prefer it to other constitutional alternatives) and on the effective distribution of burdens and profits among the members of the political community. Thirdly and finally, a replay of the previous two arguments is necessary at the level of implementation of norms, once we realise the unavoidable interpretative or argumentative character of the application of law.

1. Two Preliminary Questions

In this section, I shall deal with two questions which are considered as preliminary to the normative justification of the obligation to obey the law. Both of them complete the functionalist exploration of the emergence of political issues and the choice of the form of law as a medium to deal with them. In terms of our previous discussion, it is necessary to say that these two elements account for the non-voluntary component in the answer to the question whether there is an obligation to obey the law. We are involved in political relations and communities without our consent, and we enter into patterns of legal regulation of these issues not out of our voluntary option for such a form, but because only law is an adequate candidate to operate as a functional complement of morality under modern conditions. It is necessary to stress that the present arguments do not legitimise law by themselves, because it is not the form of law but the bond with legitimately enacted power that legitimises governmental power.

A) Politics as non-voluntary and the concept of political autonomy

�29. We are not free to decide whether we do or do not want to have action rules in common with others. Once we share a web of common interests with others, we cannot retreat from politics. The choice, which is the same as the classic one between the state of nature and the civil state, would only make sense if we could retreat to our own Walden and leave others unaffected (at least, in a relevant way) by our actions [47]. Political bonds are related to the fact that human action tends to have consequences affecting individuals (even if without intention) other than the actor. From the functional point of view, mention should be made of the argument that repeated interaction leads to the insight that this is a potentially conflictive situation unless we adopt common action norms. From a normative point of view, what matters is that the adoption of such norms is done in ways that respect the individual autonomy of all those affected.

�30. But what is the autonomy that must be respected? The fact that the decision to deal with certain matters in a political way and that the set of matters which should be dealt in this way are not open to our choice implies that it would be self-contradictory to conceptualise autonomy as a matter of unrestrained self-legislation. If we accept as the standard of legitimacy the respect of individual autonomy, we are required to acknowledge others' right to orient their behaviour only according to principles they judge to be correct after sufficient deliberation. Not to recognise this, and to insist on a conceptualisation of autonomy as unlimited self-legislation, would lead to the infringement of the autonomy of some individuals or to tip the balance in favour of things as they are (if we pretend to reconcile the autonomy of all individuals through unanimous decision-making). This requires us to distinguish between public and private autonomy, and to conceptualise the former as participation in the collective processes of deliberation and decision-making. Moreover, this points in the direction of democracy, but it is in need of some further refinement, given the vagueness of the concept [48]. This is what will be done in the following paragraphs.

�31. This argument corresponds quite directly to Habermas' reference to the necessarily split character of autonomy, or what amounts to the same, to his distinction between public and private autonomy [49]. It corresponds to some of the intuitions underlying Rawls' natural duty of justice. The idea that we cannot withdraw from politics is expressed by Rawls when he relates the duty to “how we become bound to one another” and to the ensuing requirements “associated to our institutional ties” [50]. It is because we are under a duty to deal politically with certain issues that it makes sense to claim that “we are to assist in the establishment of just arrangements when they do not exist, at least when this can be done with little cost to ourselves” [51] [52]. It also corresponds to the basic insights that underlie Honor�'s case for the obligation to obey the law (the idea that membership of political communities is not always a matter of choice, and that this is no obstacle to affirm that rights and duties derive from the web of interaction with other members) [53]. Although it does not go as far as to consider that it is possible to base a full obligation to obey the law only upon it.

�32. Some further remarks are to the point concerning the implications of this argument. It is necessary to make it clear that this does not mean either that we are under a duty to engage actively in political affairs or that we must understand political choice as a matter of high virtue. First, the non-voluntary character of politics does not constitute a plea for the virtuous citizen, but just a justification for the inclusion of certain issues and matters on the agenda of politics, something that implies withdrawing them from the sphere of individual autonomous judgment. This is fully compatible with the individual decision not to participate in the actual processes of deliberation and decision-making. Second, the functional overtones of the argument do not imply that politics is just a matter of applying means to ends. Of course, politics can be seen as a tool for solving conflicts of interests and establishing patterns of social co-ordination in order to achieve complex forms of social action. However, it can also be seen as the activity in which, paraphrasing Arendt, we transcend necessity and engage in action. In the latter conceptualisation, politics is not instrumental, but reflects the capacity to act in concert with others, on the basis of an agreement free of coercion, by means of free linguistic exchange. It is important to realise that these are two sides of the same coin. A purely action-based conception of politics sterilises power. Without the realisation that politics constitutes a way of achieving certain purposes in a legitimate way, we might not realise what we want power for [54]. We do not want to rule out the emergence of strategic action in the public sphere, because this might end up leading to behaviour that is exclusively self-interested [55]. Similarly, a purely interests-based politics runs the risk of privatising politics and making it equally sterile.

�33. As already argued, the non-voluntary character of politics is not a sufficient ground on which to rest the case for a general obligation to obey the law. That we have to deal with matters of common interest through a political procedure does not mean that we are forced to accept any common action norm. As we will see, their legitimacy depends on the procedural and substantive characteristics of the norms themselves. The fact that we are not free to enter into political relationships and that our membership of a given political community (or a plurality of them) is usually not a matter of choice constitutes no obstacle to the derivation of burdens and profits from such social relationships. But by means of a direct reference to the procedural and substantive characteristics of the common action-norms, there is no guarantee that the legal system will distribute rights and duties in a legitimate way.

B) The Need for Law and the imperatives of constrained decision-making

�34. Political interaction exclusively based on the practice of “those who talk to one another in order to act in a political autonomous manner” poses several problems [56]. First, rights and duties are exclusively based on their mutual recognition by the members of the political community. This makes it difficult to determine the concrete consequences they have in a specific case, and also leads to a certain instability. Second, there are difficulties related to the organisation of public deliberation and decision-making, some of them related to cognitive problems (mere spontaneous interaction relies on personal memory in the absence of institutional mechanisms for keeping record of past deliberations and decisions). Third, it implies a degree of commitment to public matters that is difficult to reconcile with individual autonomy as a whole (which must include the possibility of not participating actively in them) [57]. These problems are solved by means of structuring political interaction with the help of political institutions, creating stable frameworks for deliberation and decision-making [58].

�35. Law is a basic institution in such an endeavour. We cannot expect morality or general practical discourse to perform on its own the vital functions of conflict-solving and social co-ordination [59]. Not only might it be the case that moral argumentation cannot discriminate right from wrong when confronted with concrete and specific cases, but it is also the case that sometimes morality presupposes the existence of positive systems of common-rule action (that is, instances in which those systems break ground for morality and in which moral judgment is at the very least conditioned by the solution provided by such systems), but also that it cannot on its own avoid conflicts and co-ordinate action within society because of the cognitive, motivational and organisational demands it makes.

�36. The unavoidable character of the form of law implies that the question of whether there is an obligation to obey the law must be attentive to the conditions limiting general practical reasoning that law imposes in order to fulfil the basic tasks assigned to it and to the way in which the legal system evolves.

First, the need to complement morality with law, as repeated several times, is based on the need to reduce the cognitive, organisational and motivational demands made by morality upon individuals. The resort to law implies delimiting a special domain within general practical reason in which several limiting conditions are applied to it [60]. This is done in order to render certain the norms belonging to such domain, and in this way, apt for discharging the basic roles of conflict-solving and social co-ordination. In this way, law can be seen as a special case of general practical reasoning in which decisions are reached with a limited investment of resources in terms of time and effort. If this is something characteristic of law as a social order, the case for the general obligation to obey the law must be sensitive to the fact that law raises a claim to correctness that takes into account these limiting conditions.

Second, law is a complex social order whose norms have been enacted and refined through discourses of application by many different authors and at very different moments in time. To Dworkin's conceptualisation of judicial reasoning as a sort of chain novel [61], we can add Nino's image of law-making as the building of a cathedral [62]. In both cases, the intuition is the same. The ways in which the legal system evolves are partially constrained by the historical constitution, or the legal system as it stands at a certain moment in time. The legitimacy of the present legal system must consider to what extent it has provided a means for increasing the normative legitimacy of pre-existing legal systems and continues to be a vehicle for the improvement of itself.

�37. However, and as it was the case with the unavoidable character of politics, the fact that we have no other option but to use the grammar of law in our political interaction is not enough to claim that there is an obligation to obey the law. This is so because the form of law can be instrumentalised in such a way as to render it illegitimate [63]. That is, no intrinsic legitimacy is associated with the form of law. The legal system can be clearly unfair and even wicked, and in such circumstances we have no obligation to obey its norms. In such a context, the second element of the Rawls' natural duty of justice (namely the duty to help establish fair schemes of social co-operation where they do not exist) requires something very different from mere passive obedience. We cannot pretend either to answer the question on the basis of the formal properties of law as a social order, or to deny from the very beginning the possibility of making a case for it.

2. The positive grounds

Common action norms written in the legal code can be presumed to be heteronomous. That is so because law itself might be perceived as an authoritative, institutional and coercive social order. That is, the characteristics of the legal code lead to a potential legitimacy deficit of its norms. This can only be redressed if the addressees of law are offered reasons that allow them to see themselves as not only as subjects but also as authors of the law. The appropriation of the norms, the transformation of their heteronomous status into an autonomous one can only take place through the medium of reasons acceptable to the individual herself.

This we must do taking into account the two sides of autonomy. On the one hand, it is necessary that autonomy is based on the idea of giving a fair chance to individuals to present their own arguments and account of their interests to others within the process of common will-formation, and to have a chance to influence (even if marginally) the process itself. This idea is associated with deliberative or discursive democracy, that is, with a form of democracy which is divided into deliberation and decision-making arranged according to the principles of equality and symmetry. Only a procedural theory of legitimacy can satisfy the importance of active autonomy, which in its political form must be translated into a fair chance to intervene within deliberation and to affect the outcome at the stage of decision-making. On the other hand, autonomy needs to be supported by arguments that prove that this form of will-formation has some propensity to achieve correct outcomes, according to procedure-independent standards. Legitimacy can only be related to will-formation structures that have some epistemic value, because lots of the questions that we want to solve through the form of law are ones which are not substantively irrelevant (in the same sense as the rules of the road are). Even if people disagree about what is the right solution, they agree that there must be a right solution, so that even if they opt for a procedural form of common will-formation, they could only be satisfied if there is some argument that proves that it tends to produce correct results in such respect. This implies that we have to opt for the democratic arrangement that is best in epistemic terms, at least to take seriously the kind of private convictions of individuals on a good deal of matters concerning common will-formation.

A) Participation in Discursive or Deliberative Democracy

�38. Any conception of political autonomy must have the active exercise of our practical reason at its core. This is a basic insight of liberal theories of legitimacy, all of which revolve around the idea of consent as the embodiment of reason in political matters. This is so because consent provides a bridge between the heteronomy of externally-generated norms and the principle of autonomy as a regulative ideal. It is the means through which we internalise norms and make them our own.

Consent is usually taken to imply that the fact that an individual participates (or at least has the chance to do so) constitutes an argument to claim that she is obliged to comply with the norms adopted through such participation, and in general, with the whole legal system. This idea is reworked in at least three different ways. First, consent is cashed into participation in both deliberative and decision-making procedures. Second, it is claimed that the case for participation as a basis for legitimacy rests on three notions. Firstly, the individual is given an equal right to participate in deliberation and decision-making. Secondly, there is an in-built tendency for the procedure to select the right solution (this is known as the epistemic privilege of democracy). Thirdly, there is no claim to exclusiveness. The case for a general obligation to obey the law put forward here is associated with a complex, not with a simple conception of legitimacy.

The idea in a nutshell is the following. Active participation in a deliberative democratic procedure of law-making binds us to act according to the normative outcomes of such a procedure to the extent that it ensures us an equal right to participate in deliberation and decision-making, and that the procedure is structured in such a way as to have an epistemic privilege, or what amounts to the same, to have a propensity to select the right norms. The argument is procedural to the extent that what triggers the obligation is only our participation in the procedure, but it is also substantive in the limited sense that it refers to average correctness as a supplementary reason to act according to its outcomes (but this is done without any direct reference to procedure-independent standards of correctness).

�39. We have to keep in mind that the main addressees of the argument are those who participate in the procedures of deliberation and decision-making but end up on the losing side. That some individuals will be in such a position is unavoidable once we realise that democratic deliberation has no in-built tendency towards unanimity. Even if we were to enjoy the best possible political conditions, it seems we could not avoid disagreement. Democratic deliberation filters out the most outrageous mistakes, but it also tends to increase (and not to reduce) the alternatives at stake, if only because participants increase their knowledge. Just think about what to say to a citizen on the losing side. Let us call her Loser. Before voting, she thought that a given option was the right one, and voted for it. We know by now that to say that she has an obligation to obey the new rule does not require Loser to review her judgment and convince herself that she was wrong and the majority was right. From the point of view of the law, it does not matter what she thinks if she acts according to the rule supported by the majority. But she will not be happy even with that limited claim. If she is said to be under an obligation to obey the law, and the law is determined by majority-voting, she will have to comply with what she thought was at least not fully right, and maybe wrong. Because even if she does not have to change her mind, Loser cannot forget that what was previously an alternative and is now the majority-supported and institutionally-adopted option was not fully right or even wrong according to her view. So we are requesting Loser to comply with a norm for reasons that she found not very convincing before voting. So how can she see herself as an author of the law, and not only as an addressee that complies for mere prudential reasons? To explain this is the main purpose of the following paragraphs.

a) The structure of opinion- and will- formation

�40. Deliberative democracy structures opinion- and will- formation in three different steps. First, it is necessary to revise and transform individual preferences with the help of a process of public deliberation. If we want to take seriously our autonomy and that of the other participants in the procedure of will-formation we cannot simply rely on our raw preferences [64]. Political issues are sufficiently complex as to make it difficult for us to know enough in advance and moreover, for us to be sure about what we know. We need to confront the interests of others, to cross-examine the reasons that support our claims, to confront those that underlie alternative arguments. This allows us to detect factual and logical errors, to look for points of agreement, and with some luck, to forge a basic empathy derived from the engagement in the common enterprise of deliberation. Second, all this can not guarantee that unanimous agreement will exist at the end of the day, especially if we have limited resources to invest in deliberation, and we have to meet a deadline for adopting a common action norm. This is the same as saying that sadly enough, we have no other option other than to decide heads or tails or something in between. We cannot avoid deciding. The same reasons on which the case for political decision-making rests require that we adopt some form of majoritarian decision-making. There is not much of an alternative. We have seen why political decision-making cannot be based on unanimity. This would lead to a deadlock that privileges the status quo (in case we opt for unanimous decision-making). Third, the fact that both deliberation and decision-making take place under non-ideal conditions, and that we opt for majority voting, makes it implausible that the outcomes are always correct. For these reasons, it is necessary to keep the process open to reforms, to a new examination of the question that could lead to reflexive self-amendment [65].

b) Equal concern requires equal participation in deliberation and decision-making

�41. A second basic requirement for the legitimacy of any procedure of common-will formation is that participants treat each other with equal concern. Because we agree that the procedure has two dimensions, deliberation and decision-making, this entails the application of such principle in both the deliberative and the adversarial dimensions of democracy.

Why equality? The argument is this. The respect for the public autonomy of all individuals must be seen as a matter of distribution of a scarce good. This is so because the chances of understanding, elaborating and pursuing our interests in respect of common action-norms, or what amounts to the same, the collective properties of our societies are mutually limited [66]. If we design deliberation and decision-making in ways that give more chances to a given individual, this necessarily results in reducing the chances of others. The only fair solution is to give all equal chances.

That we need to consciously design deliberation in order to achieve such equality is an insight that is shared by several authors dealing with different related issues. Owen Fiss has explored the question from the slightly different perspective of the relationship between freedom of speech and state regulation. He argues that the freedom can only be attained if speech is regulated in order to treat all equally, and this implies silencing some in order to allow others to speak [67]. The same kind of concern, specifically applied to the political process, underlies the argument for setting ceilings for expenditure on political campaigns, eloquently tackled by Dworkin, Ackerman and the same Fiss [68].

That equal concern implies giving all individuals an equal influence on the decision-making process is an old insight of liberal theory, reflected in the motto of one man, one vote (which has fortunately being modified to read one adult citizen, one vote).

c) Deliberation and the epistemic privilege

�42. But if we accept that aggregating preferences is not enough, and any legitimate process of will-formation requires previous deliberation, this implies that by deliberation we are trying to achieve something, to get at a certain result beyond mere transformation of preferences into a common metric and their aggregation.

�43. The claim to legitimacy of a legal norm would not be redeemed by mere reference to the equal right to participate in deliberation and decision-making processes. This becomes even clearer once we argue that deliberation should be taken seriously. Remember we found it reasonable to reject the claim that the adoption of common action norms is a simple affair of aggregating private preferences, if only because taking seriously our autonomy and that of others requires checking our preferences in the deliberative process.

But if at the end of the deliberative process Loser is even more confident that her option was right and the majoritarian wrong, to justify the obligation to obey to her seems a harder task. Quite paradoxically, the answer is implicit in the reference to deliberation and the transformation of preferences. It implies an interest beyond the mere reduction of all preferences to a common metric and their ulterior aggregation. In brief, it betrays an interest in correctness. Answering to Loser requires showing her that the procedure has an in-built tendency to produce correct results, that takes seriously what motivates individuals to deliberate. From this, it is possible to derive an argument that the individual has a reason to act accordingly to the common action norm and not to her judgment.

Deliberative democracy takes seriously the fact that legal discourses of justification concern not only questions of bargaining on individual interests, but a lot of questions on which what is relevant is to get the correct solution in a political sense. It is common that most people will disagree on what is the right outcome (in terms of what the balance of relevant public reasons requires the outcome to be), but they will assume that there must be a right solution to the question. This makes it clear why taking public autonomy seriously requires something more than giving each citizen an equal right to participation. It requires that the procedures of deliberation and decision-making are designed in such a way that they can satisfy an individual's interest in the correctness of the solution, and produce an outcome that tends to be correct, and she has reasons to act upon even if the solution is different from hers [69].

Why does deliberative democracy tend to produce correct results?

�44. The epistemic privilege of deliberative democracy is based in its tendency to pick out correct norms. The outcomes of deliberation and decision-making in a deliberative democracy are closer to correctness than those of random decision or those of any other political procedure [70].

It has been argued that the privilege can be justified by reference to the arguments laid out in Condorcet's theorem of majoritarian decision-making. It says that the chances that a decision taken by majority rule is correct increase with the size of the constituency if we assume that each individual has a tendency to adopt the right solution (this means that she can spot the right solution more frequently if she resorts to her judgment than to sheer chance). Carlos Santiago Nino has argued on such basis, adding that deliberation based on an equal right of participation of individuals ensures a tendency to be right among those who participate in it [71].

It has been counter-argued that we cannot be so sure about the actual competence of voters, that we do not know what are the consequences of mutual influence (think about those cases in which those with a tendency to get outcomes wrong lower than random play a leadership role within the group) and that the theorem assumes a binary structure of choices (though most political decisions are finally voted in such a format that they presuppose a process of selection that might have selected two wrong alternatives).

All that notwithstanding, we do not need to have resort to a strict version of Condorcet's argument. A somewhat watered-down version might do the trick [72]. Nino's argument on the tendency of deliberation to increase individual competence is essential. The basic intuition is that the case for deliberative democracy can be based on the capacity of such institutional arrangement to collect decentralised information and to incorporate it to the decision-making process. The case for deliberative democracy is, after all, epistemic. One could say that this watered down version of Condorcet resembles Hayek's argument for the epistemological superiority of market over planned economies, only that in this case it is applied to political arrangements.

Moreover, the case is conditional upon institutional arrangements complying with the basic requirements for qualifying them as a deliberative democracy. The basic rights and institutional conditions implicit in the definition of deliberative democracy should be respected [73]. If that was not enough, the epistemic privilege can only be granted to those institutional arrangements which have a capacity for systemic self-improvement. There must be evidence that they are able of reflexively amend themselves towards their normative ideal [74].

Correctness as Public Correctness

�45. A second question concerning the epistemological privilege of deliberative democracy relates to the question of what is meant by a tendency to correctness, and more specifically, by correctness.

It cannot mean moral correctness at large. We assume that reasonable individuals might disagree on moral matters, and we rely on a brief explanation of such disagreement by reference to the many burdens involved in moral judgment. Moreover, moral disagreement renders moral norms uncertain, and that this is the main obstacle to its satisfactory discharge of the tasks of conflict-solving and social co-ordination; it explains why morality needs to be complemented by law. If all that is so, when citizens deliberate about what laws to adopt they cannot be seen as engaged in an unconstrained moral debate. They are not discussing moral truth in a direct way, but they are considering what would be reasonable norms to deal with basic social conflicts and to co-ordinate action within society. That is, they are concerned with the basic problems of living together, and they accept that they have to provide reasons that can be accepted by all participants in the will-formation process [75].

At the same time, correctness cannot mean just aggregation of preferences. Citizens have good reasons for setting aside deep moral convictions for which they cannot find arguments that could be recognised as public reasons. But that does not mean that they have to renounce any currency of correctness. After all, their deliberation is supported and framed by public reasons on the basis of which it is possible to opt for certain common rules and to oppose the adoption of others. The dialogue is not only about the implications of public reasons, but of course about which reasons are in fact public. The concrete instances to be dealt with might lead to some individuals trying to convince all others that their arguments are to be considered as providing public reasons, and others might reject or endorse such a claim. The dialogue might change our views on whether certain arguments are or are not constitutive of public reasons.

On this basis, we can conclude that correctness is to be understood in a political sense, by reference to public reasons. Citizens agree that there must be a right solution to the question on the basis of such reasons, but disagree as to which is the right solution, either because they disagree on which are public reasons or because they derive different consequences from them.

Correctness as mediated correctness

�46. As already argued, the claim to correctness is not to be taken as a direct claim to correctness, but it is to be judged in indirect terms, taking into account the split character of practical reasoning on the question whether there is an obligation to obey the law, which reflects the value of having a tool for co-ordinating action like modern law [76]. The claim is not to be understood as based on the primary correctness of each norm, but as a claim as to the limited soundness of each norm within the framework and taking into account the value to be derived from the legal system to which the norm belongs as a system of conflict-solving and social co-ordination [77].

Why the epistemic element does not render democracy superfluous (why active individual participation matters)

�47. If law not only borrows its legitimacy from morality, but also the question of whether we should obey the law is in itself dependent on the tendency of the law-making procedure towards substantive correctness, then we might come to think that we should opt for the law-making procedure that has a higher propensity to get the correct result, even if it does not include participation of legal subjects. To put it bluntly, if substantive correctness is important, why not recruit some expert Solon and forget about democracy? Note that the suspicion that a theory of political legitimacy trusts too much to substantive correctness triggers severe criticisms even between liberal theorists themselves [78].

�48. Such an argument fails to take into account two factors. First, that the dimension of correctness based exclusively on a direct appeal to correctness does not respect an active conception of individual autonomy. It is true that the theory of legitimacy here advocated is a complex one, and that such complexity derives from the fact that it relies on substantive correctness. But within it, participation is not only the other main source of legitimacy, but it is given preference on two grounds. The first is that the ultimate individual character of moral action. The second, that even substantive standards are determined by reference to the procedure itself. Deliberative democracy allows us to combine a full respect for this active element of autonomy with an in-built tendency to select correct norms. This is so because the tendency is not based on the fact that the outcomes of the procedure are corrected by reference to any substantive standard of correctness, but on the in-built properties of the procedures. This allows deliberative democracy to avoid any direct constraining of the value of individual autonomy. The obligation to obey is triggered by the mere fact of having the right to participate in the procedure of law-making once it fits into the model of deliberative democracy. In this way, we avoid any need to make a direct reference to standards of correctness, thus avoiding invidious comparisons. This is something that has been stressed by authors such as Habermas, who locates sovereignty not in individuals but in the procedure of deliberation or discourse itself [79]. Second, the nature of political issues themselves renders it highly improbable that our Solon deciding solo can get results right. This is not only because we could doubt that his motivation will be impartial, but also due to the colossal task it would be to collect all information and process it. This allows us to find a further and more relevant use for the well-known Hayekian argument against any attempt to impinge upon the self-regulation of the market. We could use it against any temptation to opt for some form or another of epistocracy.

Taking Stock: Who has an obligation to obey on the basis of participation

�49. Why should Loser acknowledge an obligation to obey the law, and why she should consider that this extends to those cases in which she ends up in the losing side?

First, we have explained to her why she should decide not on a case-by-case basis, but why she should consider first whether she has a general obligation to obey the law, and only in some cases test if the breadth of such obligation covered specific instances of her duty. Second, we have argued that respecting her public autonomy cannot imply giving her the right to consider on her own behalf what is the right solution to the case, but must consist in giving her a right to participate in the process of political deliberation and decision-making. This is so because the involuntary character of politics makes it an enterprise in reconciling the autonomy of all participants, and that can be done only in this way. We have further said that common action norms must be written with the help of the grammar of law, because there is no feasible alternative to it. This brings into the picture the basic limiting conditions to general practical reasoning characteristic of the legal form (the imperative of reaching a solution within a limited time horizon, the very limited resources we have at our disposal, the fact that the form of law is characterised by progressive cumulative reforms and not by overall change). It also implies that the answer to the question whether there is a general obligation to obey the law must depend on the substantive values embedded in the norms and not on a generic reference to the form of law.

The next step of the argument consists in defining this general reference to a right to participation as a right to participate in deliberation and decision-making on equal terms. First, we have to show Loser that she cannot ask for more. The right to participate in the political procedure is a scarce good. Only a mutual recognition of the same right to all allows to treat them with equal respect, or in other words, to take seriously their individual autonomy. Second, the structure of the process of deliberation and decision-making ensures that deliberative democracy has an in-built tendency to select the correct result. Though she has all the reasons to maintain her opinion on the matter, Loser also has good reasons to act according to what the law prescribes (at least to adopt it as a general standard for the orientation of her conduct) because it is more likely to be right than herself, especially if we take into account the nature of the questions at stake. It is important to notice that this argument does not interfere at all with the autonomy of the individual, to the extent that no direct reference is made to substantive or procedure-independent standards of correctness is made, but to in-built tendencies of deliberative democracy. Third, the argument is complemented by a reference to the open-ended character of the process. Democratic law-making allows the periodic reconsideration of the question and to the eventual passing of a new judgment upon it [80]. It strengthens the argument for complying with present positive law on the basis that it has a potential to be reformed, something absent in other social orders, and for such reason the practice of law can be the carrier of the correct result [81].

B) Fairness and the Obligation to obey

�50. The right to equal participation in the processes of deliberation and decision-making constitutes the core of the case of the legitimacy of law here advocated. The argument goes that a) it constitutes the most adequate way of showing equal respect for the public autonomy of each individual in way compatible with the equal respect for all others; b) it is the procedure that shows a higher tendency to provide a correct (in political terms) answer to the question at stake. However, it might not purchase sufficient legitimacy for the legal system. Consider the following instances.

First, the right to participate might be distributed far from equally or the conditions for asserting the epistemic privilege of democracy might not be fully met. We stressed in the previous section that the degree of legitimacy contributed by the right to participate in a deliberative democracy depends on the extent to which the right of participation is equally assigned to all members and on the measure to which the conditions on which the epistemic privilege of such political procedure are met. Really existing democracies tend not to fully comply such requirements.

Second, some individuals might be prevented from participating in the political process even if they are willing to do so. Four examples should suffice. Firstly, some human beings simply do not have the capacity to participate meaningfully in the political process. While this condition is in most cases temporary (minors), there are some instances in which this incapacity is permanent (e.g. mentally handicapped people or others suffering severe mental illness). Secondly, not all citizens have access to sufficient material resources as to be able to make a meaningful contribution to the political process, and although there is a good case for socially ensuring this through redistribution, institutional structures are not always responsive to such need. Thirdly, certain minorities might be socially discriminated against, rendering their access to the political process problematic [82]. Fourthly, and by way of conclusion, even in the most liberal states the acquisition of political rights through naturalisation takes place only after the completion of a certain number of years of residence [83]. In the meantime, permanent residents cannot participate in the deliberative and decision-making processes.

Third, there may be cases in which the resort to participation might be not possible for good reasons (the central case is constituted by authoritative legal adjudication, as we will see in the next section).

�51. All these instances point to the need to complete the legitimacy purchased through participation by reference to some other source. This necessarily implies making an appeal to substantive correctness [84].

In a nutshell, the idea is that a legal system can be seen as legitimate and requiring our obedience where its norms comply with standards of substantive correctness. The main issue is to determine on which basis it is possible to assert such standards. The argument that is made in the following paragraphs claims that they must be seen as deriving from the same thin substantive conception that justifies the role of participation in the legitimation of the obligation, as we have described it. This approach is fully compatible with the constructivist approach to ethics, and avoids the confrontation between procedural and substantive strategies for legitimacy. Instead of seeing them as two fully different theories of legitimacy, they are depicted as different ways of elaborating the same values.

I distinguish two different types of standards of correctness. First, I deal with the ways in which we can make a bridge between the rules of discourse and basic rights. This leads to the formulation of some standards of correctness that are applicable with independence of the concrete contents of the legal system. This allows us to see such standards as alternatives to legitimacy through participation when this avenue is closed. Second, I consider the principle of fair play, which is based on the concrete pattern of distribution of burdens and profits within the legal system. This argument must be interpreted as a complement to legitimacy through participation when its case is not fully argued for (for example, when not all conditions for establishing the epistemological privilege of democracy are met).

a) Basic Rights as a Standard of Legitimacy

�52. Deliberative democracy has been said to constitute an approximate institutionalisation of discourse rules. This is an essential part of the argument that leads to the argument that we are under an obligation to obey the laws enacted through such procedure. But why should we opt for discourse rules, and for the conception of public autonomy and of politics as rational conflict-solving and co-operation? This is a question that we have face when arguing against the claims of the communitarian conception of public reason. Against it, it can be said that the option for deliberative democracy is based on a thin substantive theory, whose peculiarity and universal validity allow us to characterise it not just as one among others conceptions of the good, but properly as a conception of the right. There are good reasons for defending such a conception without considering it on a par with traditional or utilitarian conceptions of public reason. So far, so good. How can this be of any avail to us in spelling out substantive criteria that could sustain a claim to legitimacy on the part of the legal system? Keep in mind that the thin character of the conception of the right precludes us deriving from it bold substantive principles. The answer is that though direct derivation is out of the question, it might be possible to find arguments bridging the gap between discourse rules and the affirmation of such principles, in the form of certain basic rights. If that is possible, we would be able to add to the legitimacy derived from participation that anchored in the substantive correctness of norms. And that only by means of elaborating the same values that move us to endorse participation to start with.

If this is so, then the claim to the legitimacy of a certain legal system can be based on the argument that it roughly complies with such standards. It constitutes an independent justification for those for whom there is no possibility of making reference to the procedural argument (like minors, retarded people or those who only occasional fall within the circle of addressees of the law, like tourists) or in those instances in which such an avenue is not available (we will see that the appeal to correctness plays an essential role within the procedure of application of the law).

�53. This bridge has been projected by several authors within the paradigm of constructive ethics. In the proceeding paragraphs I follow the argument made by Robert Alexy [85], on the basis of his own previous work and also of that of J�rgen Habermas and Carlos Santiago Nino.

Alexy grounds his choice for practical reason on a transcendental argument combined with a prudential argument (acting as a sort of closing proviso). The former explores the implications of making an assertion within a discourse and allows the author to complete a chain of reasons with the conclusion that it implies, as far as the discourse is concerned, a claim to equality, freedom from force and universality. The latter ensures that such conclusion is valid, because it is applicable not only to those who participate in discourses motivated by an interest in correctness, but also to all others, who have at least a prudential interest in maximising their utility, something that makes it necessary to pretend to have an interest in correction. Even if considered as a domination technology, discourse is superior to naked force. The latter is simply too expensive given the high number of people who are interested in correctness. However, discourse rules in themselves do not provide any substantive standards of correctness for legal norms. This is because the fact of entering into a discourse implies that we make some pragmatic assumptions from which we can derive some discourse rules. But without further arguments, it is not necessarily the case that such rules should also be binding upon us when we act and not just deliberate We need further a link in the chain of reasoning.

There are two main arguments to ground such a link. Firstly, we can try to establish a bridge between discourse rules and the recognition and endorsement of the value of individual autonomy in full, that is in relation to not only the context of discourse, but also that of action. From this we can obtain a basic catalogue of rights. The argument goes that anyone who seriously participates in discourse is not only interested in grasping correctness for the sake of it, but she does so in order to solve social conflicts and structure co-ordination on the basis of the norms that are agreed upon in such discourse. (Remember that if the participant is not really interested in correctness, the closing proviso based on the prudential argument of maximisation of utility applies). This implies that we assume not only an interest in the autonomy of other participants as participants, but also in a full sense, as individuals who act in the world. This endorsement of autonomy entails the recognition of a general right to freedom as a basic legal norm. On this basis, we can elaborate a catalogue of basic public, private and social rights [86]. Secondly, we could argue that law-making in a deliberative democracy necessarily leads to the recognition of a certain set of rights. This is the track on which Habermas' train of reasoning moves. The idea is that the liberal principle of legitimacy (what he labels as the discourse principle), when tailored to the kind of needs tackled by law as a social order, gives rise not only to deliberative democracy, but also to a system of rights, including private, public and social rights [87].

All these arguments detail the ways in which we can move from the endorsement of discourse rules to the affirmation of substantive principles, articulated into basic rights.

�54. But if that is so, then consent must be shaped and defined by the moral value of autonomy, and the actual value of giving consent might be constrained by reference to the value it furthers. Lawyers are familiar with this move, as private law deals with the value of consent associated with private autonomy, and grounds validity to it within certain limits, beyond which it is considered invalid. The same kind of move can be traced back in some rules of electoral law, this time associated with public autonomy.

Our present task is to show that the validity of consent, the pre-conditions within which it must be given full force because it will protect this kind of value, are those of deliberative democracy. At the same time, it must be argued that deliberative democracy is appropriate to discharge the task of enacting sound legal norms, because its own structure is designed in such a way as to increase the chances of getting the right answer. This epistemic value is the one that answers the challenge posed by the irrelevance of democratically enacted legal norms. The best functional complement of morality is a deliberately and epistemological privileged legal system.

�55. In practical terms, a positive answer to the question whether there is a general obligation to obey the law is something that depends on whether it complies with minimal substantive requirements. Additionally, it is necessary that there are conditions for systemic self-improvement within the system, in order to move it towards the normative regulative ideal.

b) Fairness as a standard of legitimacy

�56. The requirement to comply with the obligation to obey the law can be based not only on the compliance of the legal system with abstract standards of substantive correctness, but also on the concrete way in which the benefits and burdens derived from its existence are distributed among legal subjects.

This foundation corresponds to the argument of fair play or fairness. Its most well-know formulations have been made by Hart and Rawls.

The former argued that:

“when a number of persons conduct any joint enterprise according to rules that restrict their liberty, those who have submitted to those restrictions when required have a right to a similar submission from those who have benefited by their submission” [88].

The latter wrote in the Theory of Justice that:

“The main idea is that when a number of persons engage in a mutually advantageous co-operative venture according to rules, and thus restrict their liberty in ways necessary to yield advantages for all, those who have submitted to these restrictions have a right to a similar acquiescence on the part of those who have benefited from their submission. We are not to gain from the co-operative labours of others without doing our fair share. The two principles of justice define what is a fair share in the case of institutions belonging to the basic structure. So if these arrangements are just, each person receives a fair share when all (himself included) do their part (...) It is important, however, to notice that the principle of fairness has two parts, the first which states that the institutions or practices in question must be just, the second which characterises the requisite voluntary acts” [89].

�57. This standard argument of liberal theories of the obligation to obey the law is here reworked and adapted to the role of complementing the legitimacy provided by participation in the political process, especially on those cases in which the conditions to acknowledge an epistemological privilege to democratic deliberation are not met.

�58. The standard argument of fair play departs from the necessary character of political relationships and from the unavoidable resort to law in order to deal with conflict-solving and co-ordination of action oriented to the achievement of complex social goals [90]. It stresses that these operations in themselves imply the allocation of burdens and benefits derived from the subjection of action to common norms.

The argument of fair play says that the obligation to obey the law can be founded on the duties derived from the reception of the benefits ensuing from the existence of legal norms if three basic conditions are met. First, that the legal system allows for a minimal degree of participation. This precludes that us having an obligation to obey the law on the basis of fair play in the context of authoritarian legal systems [91]. However, it does not constitute a replay of the case for legitimacy through participation. It just points to the complementary character of the argument. Second, that the individual has received a roughly fair share of the benefits derived from enacting common action-norms in the form of law. The assessment of such benefits proceeds according to principles of distributive justice. Third, that the individual has somehow accepted such benefits. Notice that the expression of the individual will is not addressed to the characteristics of the scheme for the distribution of burdens and profits themselves. Furthermore, and because the role of consent is not associated here with the appropriation or internalisation of norms, but to tracing some relevant act of acceptance of the benefits, it is enough that the individual acts in a way that can be objectively said to imply [92] an acceptance of the benefits [93].

�59. The argument of fair play is especially relevant when we deal with the justification for the obligations associated with norms that organise the co-ordination of the action necessary for the achievement of complex social goals, such as the obligation to pay taxes. However, the fact that the general principles of tax law have been mainly developed through judicial reasoning, and that it is very difficult to operationalise the argument into directly applicable legal standards means that the general principles of tax law invoked before courts tend to be associated with abstract standards of correctness specified with reference to the peculiar tasks discharged by the obligation to pay taxes.

C) Guaranteed Implementation

The two previous sections constitute a case for the general obligation to obey the law. This is justified by a complex theory of legitimacy, which combines procedural and substantive arguments. On this basis, legal subjects are provided with sufficient reasons to see the outcomes of deliberative law-making as their law. However, the circle is not yet completed, because law is not only made, but also applied. This leads to further questions that have to be tackled by a complex theory of legitimacy. The argument runs as follows. We have seen that law is a much needed complement of morality. The latter demands too much from individuals in cognitive, organisational and motivational aspects. Moreover, there are certain issues on which what is morally correct depends on what emerges de facto as the socially preferred choice. This is a kind of situation characteristic of scenarios of co-ordination of action through law. However, general norms cannot fully spell out the requested pattern of conduct in all instances of application of the norm. This is so for at least two reasons. First, because legal norms are conveyed in general terms, and pretend to be applicable in a potentially unlimited number of cases. This is something that requires intentionally drafting them without any desire to exhaust the circumstances that might be relevant in the concrete context of application. This means that the actual implementation of the norm might require paying attention to such circumstances and determining how we should construct the norm in such case. Second, the law is conveyed in a human language and produced by humans. As such it can be prey to the shortcomings of language and to the errors and mistakes proper of fallible beings. When the time comes to derive concrete consequences from the norm in question, it is clear that the legal operator cannot avoid solving such questions in one way or another. These two phenomena point to the same conclusion: We cannot avoid interpretation, and interpretation implies a creative jump that necessarily reopens the legitimacy question [94]. This is even clearer once one realises that the paradigmatic discourse of application to which the last legal word is assigned is that of judicial reasoning, characterised by its authoritative and, consequently, heteronomous character. But it also extends to the cases in which laws are applied by administrative bodies, something which is extremely frequent in tax matters.

�60. The complex theory of democracy acknowledges the existence of such gap and the need for specific arguments addressed to fill it or at the very least to reduce its width. However, one should start by referring the exact dimensions of the legitimacy gap. It is strongly constrained by the fitness dimension of legal adjudication, or what is the same, the requirement that they fit it into the legal system as a whole, and by the binding character of precedents, that is, of previous similar decisions. To the extent that we can make a case for the legitimacy of all legal system as such, the legitimacy gap related to legal adjudication is much narrower, because it is clearly constrained by the former [95].

This is illustrated by accounts of judicial reasoning like the one provided by Ronald Dworkin. Leaving aside his basic methodological assumptions, he presents the judge as trying to produce the best possible account of the law [96]. In doing this, the judge is severely constrained by different paradigmatic norms. The impact which the judge can make while interpreting the law is severely constrained [97]. This might not be sufficient to accept his one- right-answer thesis, but it constitutes a good argument against the idea that judges' discretion is complete in hard cases. The same point is stressed by the institutional theory of law [98], and lies behind the development of theories of legal argumentation [99].

If that is so, we have good reasons to add a third pillar to our theory of complex legitimacy, namely, that of guaranteed implementation.

Which are the relevant standards of guaranteed implementation? For sure, we can say that they constitute a replay of substantive and procedural arguments taking into account the specific characteristics of the context of the application of law. However, we must distinguish the problem of the legitimacy of judicial adjudication and administrative decision-making.

�61. On what regards judicial adjudication, two remarks should be made. First, the main source of legitimacy of the judicial decision flows from the reference to substantive standards of correctness on the side of judges. This claim is of course a limited one, that takes into account the different limits to general practical reasoning operative in the legal domain and also the positive content of legal norms as they stood at the time of the decision. The judge needs to provide public reasons to the parties to the case and to all those whom the norm might affect. To the extent that the case might be seen as a precedent for any other judicial decision, the audience extends to all legal subjects. Arguments that the content of the law is this or that are insufficient, because they do not settle what the norm means in the concrete case. The only way to complete them is by reference to substantive reasons obtained by reference to general practical reasoning (replayed within the framework provided by the legal system). This implies that law cannot be seen as a complete or closed system of reasons. It also implies that in order to supplement it, it is not necessary to make a direct appeal to general practical reasoning, but it is possible to have resort to it taking into account the highly institutionalised character of law [100].

Second, the burden of justification placed on an appeal to correctness can only be partially lightened by a resort to participation as a source of legitimacy. No matter how many schemes we conceive in order to increase the number of interests and arguments to which judges are exposed, the fact that it is they and not the parties who have the last word precludes any direct reference to participation as a source of legitimacy. However, three remarks are to the point. Firstly, the structure of judicial procedures forces parties to them to argue by reference to legal norms and occasionally to incorporate considerations concerning the consequences of the decision for the general interest. This counterbalances the tendency of the parties to see the procedure as a matter of strategic action (it is clear that they aim at getting a favourable decision from the judge). Secondly, the fact that the court hears arguments from the two parties ensures her exposure to a certain set of the relevant arguments in the case at hand. This might not be sufficient to ensure that the judge will tend to produce a correct solution, but at least increases such chances. Thirdly, the range of arguments available to the judge can be increased by means of institutionalising procedures to allow a minimal degree of representation of the interests of third parties. This is something already foreseen in some judicial procedures. It is not infrequent to find figures like that of the amicus curiae, a third party (generally, a lawyer) who does not represent a party in a case, but who is called to address the court to help clarify a difficult legal point or to explain something which is in the public interest. This kind of role can be further developed by reference to secondary associations, and thus increasing the breadth of those interests represented [101].

�62. Moving to administrative implementation of laws. The activity of administrative bodies tends to escape the attention of legal theories. It tends to be assumed that all norms give rise to legal relationships which fit either the so-called civil (horizontal) or penal (vertical) paradigms of legal relationships. At any rate, conflicts concerning the concrete way in which norms should be implemented are deemed to be solved either contractually or by resort to judicial adjudication. The reality of the implementation of and even adjudication upon legal norms conducted by the administration is simply bracketed. At most, it is accommodated by allocating to the Administration the role of a private agent (as it was the case in the paradigmatic English solution) or of a quasi-judicial body (as it was the case with the peculiar conceptualisation of the principle of separation of powers within the Continental interpretation of the formal paradigm of law). A democratic theory of tax law needs to offer a different characterisation of the r�le played by administrative bodies, one which takes seriously their activity and the risks that they pose to democracy and individual rights. We must consequently come to grips with the legitimacy problems of the administrative implementation of law. This is a task that should be especially dear to a democratic theory of public law, to the extent that the latter is a field in which the prominence of administrative authorities is more salient and evident to citizens.

First, administrative bodies should act according to the program laid down for them in legitimate law (democratically enacted statutes) and their action should be subject to review by legitimate judicial authorities. This principle simply summarises the standard democratic principles of legality and judicial review. However, a democratic theory of tax law should stress that the principle of legality should not be interpreted as a mere formal constraint upon the action of the administration, but as an enabling condition. The key idea is that democratic law programs the activity of the administration [102]. Moreover, the principle of legality implies that administrative law is not only a source of duties, but also a source of rights of individuals against the administration.

Second, the emergence of administrative bodies must be explained by reference to the principle of division of labour and the need to avoid the overload of participatory institutions. If a democratic theory of law does not want to rely on an appeal to thick political virtues, if it does not want to make democracy depend on the permanent political mobilisation of the citizenry, it needs to recognise that democracy itself requires the emergence of a set of administrative bodies with the specific mission of taking care that legal programs are implemented [103]. Of course, we have just seen that this is no excuse for leaving discretion to tax institutions. Tax law provides a very good example. A democratic theory of tax law denies that taxes are just a matter of coercion. However, it must also acknowledge that tax norms are not self-policing ones. The tax administration is needed in order to monitor compliance with tax norms. The functioning of the tax system is trusted to a certain extent to the activity of taxpayers themselves. However, the administration needs to play a basic surveillance role in order to ensure abiding taxpayers that they will not be abused by free riders. In addition to that, the tax administration has access to basic economic information, and elaborates them not only in order to monitor tax compliance, but also in order to facilitate the data with which to conduct macroeconomic policy

Third, and quite related to what has just been said, the democratic theory of law must be critical with the traditional privileges granted to the administration. In most cases they were based on its characterisation as a quasi-judicial authority, which was consequently independent from the authority of the legislative. To do so, it must reconceptualise the administration as an agent or proxy of citizens. Its justification must be grounded on the need of giving an institutional format to the diffuse active positions which derive from horizontal obligations among the members of a political community. Tax law is once again a good example. The obligation to pay taxes was said to be grounded on the duty to share the burdens derived from the existence of the political community, and in more general terms, it was regarded as based on the duty of solidarity. The active positions of citizens vis a vis other citizens are quite diffuse and imprecise if we only consider the horizontal dimension of this relationship. It is by means of institutionalising them with the help of the obligation to pay taxes that they acquire concrete shape. But in doing so, we concentrate such active positions in the hands of the tax administration, relieving individuals from the burden of policing compliance with the obligation to pay taxes. Moving back to the idea of the administration as a proxy of citizens, this constitutes a basic normative standpoint from which to consider critically the traditional privileges granted to the administration and to proceed to their filtering.

Conclusion: A Complex theory of the legitimacy of law as the basis of the obligation to obey the law

By now, the circle is closed. We can restructure the previous arguments into a complex theory of the legitimacy of law, and relate it to a complex understanding of democracy.

First, the theory is complex because there are two main, differing sources of legitimacy: participation and substantive correctness. To this we have to add a third, guaranteed implementation that derives from the replay of the previous two in a concrete and constrained way within discourses of application of law.

Second, participation contributes to legitimacy in a double sense. On the one hand, individuals are granted an equal right to participate in the process of deliberation and decision-making that leads to the enactment of legal norms. This implies treating them with equal concern and distributing such rights in the only possible way that is equally respectful of the autonomy of all. The right to participate in political terms is a scarce good whose allocation must be mutually limited by reference to the entitlement of others. On the other hand, participation takes seriously the way in which individuals perceive the questions subject to legal regulation, and the interest in correctness relating some of them. This is so because deliberative democracy has an in-built tendency to select correct normative solutions, a tendency which is not based on any direct appeal to procedure-independent standards of correctness, but on the positive influence deliberation has on the individual ability to select right choices, increased by the collective tendency incorporated into the legal system.

This argument justifies the claim to obedience by those who are given a right to participate in equal terms in both deliberation and decision-making procedures (Its main addressees are those that end up on the losing side after voting). However, its strength is based on an effective granting of the right and on the concurrence of the necessary conditions for affirming democracy's epistemological privilege. If the individual has no access to such right (think about minors, retarded people or permanent residents who have not been granted political rights) or if the case for the epistemological privilege is not sufficiently strong (imagine that the way in which public deliberation is structured makes it difficult to affirm that it increases the tendency of the procedure to select correct results) then we need to look for an alternative or supplementary source of legitimacy.

Third, such an alternative or complementary source of legitimacy is provided by substantive correctness. The argument runs as follows. The legal system can legitimately make a claim to obedience to the extent that its norms comply with standards of substantive correctness, Those are to be derived from the same thin substantive conception which upholds that participation constitutes the main source of legitimacy. There are two main kinds of criteria of substantive correctness. On the one hand, those that can be affirmed in abstract terms, to the extent that they are valid independently of the concrete contents of the legal system. Such standards constitute a catalogue of basic rights, obtained with the help of arguments that bridges the gap between justified rules of discourse and basic rights. Three of such arguments are enumerated, following Alexy: the necessary endorsement of the value of individual autonomy, the necessary results of deliberation within a democratic system and the implications of conceptualising democracy as the institutionalisation of discourse. On the other hand, there are standards related to the concrete ways in which legal systems distribute the burdens and profits resulting from channelling conflict-solving and social co-ordination through law. This basically corresponds to the principle of fairness or fair-play.

Fourth, the last element of the theory of complex legitimacy corresponds to the guarantees provided at the time of the application of law, and especially when this is done in an institutional and authoritative way. The general formulation of legal norms and the indeterminacy associated with human language imply that there is a new potential legitimacy deficit in discourses of application of law. This is especially so once we take into account that the last legal word is in the hands of judges. The legitimacy gap is partially filled in if judicial reasoning is based on an appeal to substantive correctness. The judge needs to provide public reasons for her decision, but cannott find enough of them by referring to reasons which conclude that the content of the law is this or that. This requires her to spell out what is the correct legal solution by means of incorporating general practical arguments within legal discourse. The resort to participation as a supplementary source of justification is quite limited, given the authoritative character of the judicial procedure. But the burden of the appeal to substantive correctness can be relieved to the extent that the structuring of the procedure allows for an exposure of the judge to different arguments and perceptions of the interests at stake, something that can be increased with the resort to figures representing general or public interests within the procedure.


Ackerman, Bruce. 1989. Why Dialogue? Journal of Philosophy, 86:5-22.

———. 1991. We The People. Vol. I: Foundations. New Haven and London: Yale University Press.

Alexy, Robert. 1989. A Theory of Legal Argumentation. Oxford: Oxford University Press.

——— 1992. Teor�a de los Derechos Fundamentales. Madrid: Centro de Estudios Constitucionales.

——— 1994a. El Concepto y la Validez del Derecho. Barcelona: Gedisa.

——— 1994b. Basic rights and democracy in J�rgen Habermas' procedural paradigm of the law. Ratio Juris, 7:227-38

——— 1996. Discourse Theory and Human Rights. Ratio Juris, 9:209-35.

——— 1999. The Special Case Thesis. Ratio Juris 12:374-84.

Bengoetxea, Joxerram�n. 1994. Legal System as a Regulative Ideal. ARSP, 53:65-79.

Berlin, Isaiah. 1969. Four concepts of liberty. Oxford: Oxford University Press.

Cohen, Joshua and Rogers, Joel. 1995. Associative Democracy. London: Verso.

Copp, David. 1999. The Idea of a Legitimate State. Philosophy and Public Affairs, 28:3-45.

Dworkin, Ronald. 1982. Natural Law' Revisited. Florida Law Review, 34:165-88.

———. 1986. Law's Empire. Oxford: Oxford University Press.

———. 1987b. What is Equality? Part IV: Political Equality. University of San Francisco Law Review, 22: 1-30.

———. 1996a. Objectivity and Truth: You'd better believe it. Philosophy and Public Affairs, 25:87-139.


. 1996b. The Curse of American Politics. The New York Review of Books. October 17, pp. 19-24

Ely, John H. 1980. Democracy and Distrust. A Theory of Judicial Review. Cambridge: Harvard University Press.

Estlund, David. 1997. Beyond fairness and deliberation: The epistemic dimension of democratic authority. In James Bohman and William Rehg (eds.) Deliberative Democracy: essays on reason and politics. Cambridge (MA): MIT Press, pp. 173-204.

Finnis, John. 1980. Natural law and natural rights. Oxford: Oxford University Press.

Fiss, Owen. 1996. The Irony of Free Speech. New Haven and London: Yale University Press.

Gans, Cha�m. 1986. Mandatory Rules and Exclusionary Reasons. Philosophia: Philosophical Quarterly of Israel, 15:373-94.

———. 1992. Philosophical Anarchism and Political Disobedience. Cambridge: Cambridge University Press.

Gerstenberg, Oliver. 1998. Private Ordering, Public Intervention and Social Pluralism. In Joerges, Christian and Gerstenberger, Oliver (eds.). Private Governance, Democratic Constitutionalism and Supra-nationalism. Luxembourg: Office for the Official Publications of the European Communities, pp. 205-18.

Godwin, William. 1985. Enquiry Concerning Political Justice. Harmondworth: Penguin.

Gonz�lez Garc�a, Eusebio. 1987. La Obediencia al Derecho. Madrid: Civitas.

Gonz�lez Vic�n, Felipe. 1979. La obediencia al derecho. In Estudios de Filosof�a del Derecho, La Laguna: Universidad de la Laguna, pp. 365-98.

Habermas, J�rgen. 1988. Law and Morality. Tanner Lectures on Human Values. Salt Lake City: Standford University Press, vol. 8, pp. 217-79

. 1995. On the Internal Relation Between The Rule of Law and Democracy. European Journal of Philosophy, 3:12-20.

———. 1996a. Between Facts and Norms. Cambridge & London: MIT Press.

Hare, Richard. 1989. Essays on Political Morality. Oxford: Oxford University Press.

Hart, H.L.A. 1955. Are There Natural Rights. Reprinted in Lyons, David (ed.). 1979. Rights, Belmont, California: Wadsworth Publishing, pp. 14-25.

Honor�, Tony. 1987. Making Law Bind. Oxford: Oxford University Press.

Kant, Inmanuel. 1996. Practical philosophy. Cambridge: Cambridge University Press.

Kelsen, Hans. 1945. General Theory of Law and State. Cambridge: Harvard University Press.

Klosko, George. 1989. Presumptive Benefit, Fairness and Political Obligation. Philosophy and Public Affairs, 18:241-59.

———. 1998. Fixed Content of Political Obligations. Political Studies, 46:53-67.

MacCormick, Neil D. 1978. Legal Reasoning and Legal Theory. Oxford: Oxford University Press.

. 1981. H.L.A. Hart. London: Edward Arnold

———. 1982. Natural Law Reconsidered. Oxford Journal of Legal Studies, 1:99-109.

———. 1992. Natural Law and the Separation of Law and Morals. In Robert P. George (de.) Natural Law Theory. Oxford: Oxford University Press, pp. 105-33.

———. 1997. Institutional Normative Order: A Conception of Law. Cornell Law Review, 82:1051-70.

MacCormick, Neil D. and Weinberger, Ota. 1986. An Institutional Theory of Law. Dordrecht: Reidel.

Manin, Bernard. 1996. The Principles of Representative Government. Cambridge: Cambridge University Press.

Mansbridge, Jane. 1996. Using Power/Fighting Power: The Polity. In Benhabib, Seyla (ed.). Democracy and Difference. Princeton: Princeton University Press, pp. 43-65

Nino, Carlos Santiago. 1985. La Validez del Derecho. Buenos Aires: Editorial Astrea.

———. 1989a. �tica y Derechos Humanos. Barcelona: Ariel.

———. 1991a. The Ethics of Human Rights. Oxford: Oxford University Press.

———. 1991b. The Epistemological Moral Relevance of Democracy. Ratio Juris, 4: 36-51.

———. 1993a. A Philosophical Reconstruction of Judicial Review. Cardozo Law Review 14:799-846.

———. 1996a. The Constitution of Deliberative Democracy. New Haven and London: Yale University Press.

Puente, Isaac. 1985. Libertarian Communism. Sidney: Monty Miller Press, available at

Rawls, John. 1964. Legal Obligation and the Duty of Fair Play. In Hook, Sydney (ed.); Law and Philosophy. New York: New York University Press, pp. 3-18.

———. 1969. The Justification of Civil Disobedience. In Hugo A. Bedau (ed.). Civil Disobedience: Theory and Practice. New York: Pegasus Books, pp. 240-255.

———. 1971. A Theory of Justice. Cambridge: Harvard University Press.

———. 1999a. Collected Papers. Cambridge (MA) and London: Harvard University Press.

Raz, Joseph. 1975. Practical Reason and Norms. London: Hutchinson of London.

———. 1979. The Authority of Law. Oxford: Oxford University Press.

———. 1984. Hart on Moral Rights and Legal Duties. Oxford Journal of Legal Studies, 4:123-131.

———. 1985. Authority, Law and Morality. The Monist, 68:295-324.

———. 1986. The Morality of Freedom. Oxford: Oxford University Press.

———. 1987. Government by Consent. In Chapman, John M. and Pennock, J. Roland (eds.). Authority Revisited. Nomos Series, vol. 29. New York and London: New York University Press, pp. 76- 95 (also in Raz, 1994, pp. 339-53).

———. 1994. Ethics in the Public Domain. Oxford: Oxford University Press.

R�denas Calatayud, �ngeles. 1996. Sobre la justificaci�n de la autoridad. Madrid: Centro de Estudios Constitucionales.

Rubio Mar�n, Rut. 1998. National limits to democratic citizenship. Ratio Juris, 11:51-66.

. 2000. Citizenship as a democratic challenge. Cambridge: Cambridge University Press.

Simmons, John. 1979. Moral Principles and Political Obligations. Princeton: Princeton University Press.

Singer, Peter. 1974. Democracy and Disobedience. Oxford: Oxford University Press.

Sunstein, Cass. 1991. Politics and Preferences. Philosophy and Public Affairs 20

Wolff, Paul. 1976. In Defence of Anarchism. New York: Harper and Row.


[1] Most negative answers are conditioned by the way in which the obligation is characterised. If we proceed to conceptualise obligations as general (in the sense of applying o all laws on all occasions) and indefeasible (or what is the same, as excluding action on the basis of whatever reasons outside the legal domain and contrary to the legal prescription), we will unavoidably come to the conclusion that there is no possible foundation for a general obligation to obey the law. This illustrates quite clearly how much is at stake in this section.

[2] Cf. Singer (1974), Gans (1992, 5).

[3] The argument that there are exemptions from the obligation as a legal one is based on an incomplete reconstruction of legal norms. If the rules for the use of a park says that it is forbidden to walk on the grass, and another one says that it is allowed to play football from 12 a.m. to 5 p.m., what at first glance looks as an external exception turns out to be part of the rule when we proceed with its systematic formulation.

[4] Cf. Kant (1996:44): “Everyone must grant that a law, if it is to hold morally, that is, as a ground of an obligation, must carry with it absolute necessity; that, for example, the command `thou shall not lie' does not hold only for human beings, as if other rational beings did not have to heed it, and so with all other moral laws properly so called; that, therefore, the ground of obligation must here not be sought in the nature of the human being or in the circumstances of the world in which he is placed, but a priori simply in concepts of pure reason; and that any other precept, which is based on principles of mere experience- even if it is universal in a certain respect- insofar as it rests in the least part on empirical grounds, perhaps only in terms of a motive, can indeed be called a practical rule but never a moral law”.

[5] This is formulated by Nino as the “fundamental theorem of legal theory”. It was already formulated in Nino (1985, Chapter IX). He put it in the following way in Nino (1993a, 811): “Legal norms do not by themselves constitutive reasons for justifying actions and decisions (like those of judges), unless they are conceived as deriving from moral judgments; normative propositions that exhibits the distinctive traits of autonomy, justificatory finality, universalisability, generality, supervinience and finality”.

[6] Cf. ��34-7.

[7] Alexy (1989); Alexy (1994a, 40): “The relevant question is that in the praxis of any system of dominion there is an implicit claim to correctness, which must be redeemed to anybody. A normative system which does not raise a claim to correctness explicitly or implicitly is not a legal system”. Cf. also Habermas (1988, 243): “Legality can produce legitimacy only to the extent that the legal order reflexively responds to the need for justification that originates from the positivisation of law and responds in such a manner that legal discourses are institutionalised in ways made pervious to moral argumentation” and Habermas (1996a, 106): “In virtue of the legitimacy components of legal validity, positive law has a reference to morality inscribed within it”. MacCormick (1992, 112): “Judges, if anyone, take the law as it claims that it should be taken. They, more than anyone, acknowledge the law at its own estimation. To understand legal statements we should interpret them as meant by those who take them and accept them at face value, those who acknowledge the law in the way it claims a right to be acknowledged. The decisive argument concerning the meaning of statements of legal duties is that the law claims for itself moral force. No system is a system of law unless it includes a claim of legitimacy, of moral authority. That means that it claims that legal requirements are morally binding, that is that legal obligations are real (moral) obligations arising out of the law”. Also Raz (1985, 300-1) who considers that the claim to correctness might also be an intrinsic character of law.

[8] This is further analysed in �39-43.

[9] If law was not kept somehow separate and distinct from morality, we will have to face the cognitive, motivational and organisational problems that require complementing morality with law, so we will be just replaying the previous stage of the game.

[10] Rawls (1964, 7)

[11] Rawls(1999a, 20ff)

[12] Raz (1975, 35).

[13] Hare, 1989, 11.

[14] Kelsen (1945, 3), Bengoetxea (1994).

[15] See �2.

[16] As a side remark, this argument by Raz seems to me difficult to reconcile with his tendency not to distinguish too neatly between theoretical and practical authorities.

[17] As it is indicated later on, Raz himself does not characterise the general obligation to obey the law in such terms. He considers that it must be seen as an absolute reason, according to the claim to correctness made by the legal system itself.

[18] Raz (1975, 39).

[19] Raz (1975, 40).

[20] Raz (1975, 44).

[21] Raz (1975, 44).

[22] Raz (1975, 183).

[23] Gans (1986, 385).

[24] Gans (1986, 385).

[25] Gans (1986, 391): “The fact that a rule was issued by an authority for co-ordination is no reason for not considering, checking and weighting further reasons liable to be relevant to the situation to which the rule applies. However, the fact that co-ordination might be jeopardised if the rule is not obeyed, may be a strong enough reason not to act on any such further reasons that are discovered,. These subtleties are buried in Raz's analysis”.

[26] Cf. Raz (1979, 234): “The obligation to obey [the law] (...) is a general obligation applying to all the law's subjects and to all the laws on all occasions to which they apply”. In the following page, he insists that it should not be as a prima facie obligation, but as a peremptory one”

[27] Against, cf. Copp (1999, 10ff).

[28] Of course, the classic formulation is to be found in Kant. See also Berlin (1969, 131): “I wish my life and decisions to depend on myself, not on the external forces of whatever kind. I wish to be the instrument of my own, not of other men's act of will. I wish to be a subject, not an object; to be moved by reasons, by conscious purposes, which are my own, not by causes which affect me, as it were, from outside (...) I wish, above all, to be conscious of myself as thinking, willing, active being, bearing responsibility for my own choices and able to explain them by reference to my own ideas and purposes”. Cf. MacCormick (1997, 1057): “The telos of moral development is the fully responsible moral agent who takes responsibility for his or her judgments at all levels, and whose volitional commitment to some ideal of order is categorical, not conditional. Only a being that can act in a self-regulating way, judging between possible courses of action through voluntary commitment to some rationally willed order, and seeking to realise the willed order in action, can fully grasp the concept of `wrong' action, and therefore, the concept of right-as-not-wrong action. Only such a being can make full sense of auxiliary verbs such as `ought' or `should'”. A classic statement in terms of political anarchism is to be found in Puente (1985): "[W]e do not mind a restriction if we believe it to be just, and provided that is left up to us to be the judge of that. We do reject it, however, with all the force we can muster, if it something imposed upon us without our having a say on the matter".

[29] It seems to me that there is a philosophical thread moving from liberalism and classical utilitarianism (Bentham) towards philosophical anarchism. In a way, the mapping of moral autonomy is related to the act utilitarian pretence to recalculate, so to say, the pros and cons of social norms in each specific instance of their application. Versions of utilitarianism which come in the rule-utilitarianism format are closer to the discursive argument.

[30] Wolff (1976).

[31] Wolff (1976, 18).

[32] Simmons (1979, 5).

[33] Gonz�lez Vicen (1979), Gonz�lez Garc�a (1987).

[34] Nino (1989a).

[35] Cf. also Rawls (1971, 347); “The practice of promising exists for precisely for this purpose; and so while we normally think of moral requirements as bonds laid upon us, they are sometimes deliberately self-imposed for our own advantage

[36] In his doctoral thesis [A Study in the Grounds of Ethical Knowledge], quoted by Estlund (1997, 184), Rawls already argued that: “[H]ow do we know that the entity in question will always behave in accordance with what is right [?] This is a question (...) which we can always ask, and which we always do ask, and it shows that we do not, in actual practice, hand over the determination of right and wrong to any other agency whatsoever”. Cf. also Rawls (1964, 9), Rawls (1971, 357, 360).

[37] See e.g. Finnis (1980, X.3).

[38] See Mansbridge (1996).

[39] Raz (1986, 92): “Undertaking an obligation to obey the law is an appropriate means of expressing identification with society, because it is a form of supporting social institutions, because it conveys a willingness to share in the common ways established in that society as expressed by its institutions, and because it expresses confidence in the reasonableness and good judgement of the government through one's willingness to take it on trust, as it were, that the law is just and that it should be complied with” He explicitly accepts that this idea is related to an organic conception of the relationship between the individual and the political community. The attitude of trust is not normally formed through deliberate decisions, but through the normal habit-forming process of education and habituation. It includes as a constituent element the obligation to obey the law. Cf. Raz (1987, 93).

[40] Raz (1986, 70): “ States claims to have more authority than the one which could be justified within the framework of the normal justification thesis”Raz (1986, 75): “The law is the only human institution claiming unlimited authority”

Raz (1986, 77): “A qualified endorsement of the authority of the law falls short of acknowledging the authority the law claims for itself [i] it only admits a prima facie obligation to obey the law; [ii] it denies the right on the side of the law to impose certain obligations”

Raz (1987, 82): “All governments claim the right to rule us by the right reason, i.e., to take over from us the job of deciding what we should do, on certain matters”

Raz (1985, 300): “It is exemplified in the claim that even a bad law should be obeyed as long as it is a law in force, while lawful action is taken to try and bring about its amendment or repeal”.

[41] Cf. Raz (1986:53)“The normal way to establish that a person has authority over another person involves showing that the alleged subject is likely better to comply with reasons which apply ton him (other than the alleged authoritative directives) if he accepts the directives of the alleged authority as authoritatively binding and tries to follow them, rather than by trying to follow the reasons which apply to him directly”

[42] Raz (1994, 335): “In some areas and regarding some people, caution requires submission to authority. In others, it leads to the denial of authority. There are risks, moral and other, in uncritical acceptance of authority” (my italics).

[43] Raz (1986, 89). See also MacCormick (1981, 159), MacCormick (1982, 108) and Weinberger in MacCormick and Weinberger (1986, 121). This fear was already quite paradoxically shared by Godwin (see 1985, 241): “[T]he greatest mistake that can arise in the progress of obedience is where it shall lead us, in any degree, to depart from the independence of our understanding, departure which general and unlimited confidence necessarily includes”

[44] Raz (1986, 100, 101): “The two questions [that of legitimate political authority and the obligation to obey the law] are one if we understand the obligation to obey the law as an obligation to obey the law as it requires to be obeyed”; which is in contrast to the understanding of the obligation in contemporary political philosophy: “[A]n obligation to obey the law as it is understood in political writings today is a mere prima facie obligation. Such an obligation, usually thought of as nothing more than a reason to obey, may be based on reasons other than the authority of the law”.

[45] Raz (1994, 327).

[46] MacIntyre offers us a quite revealing list of them: “member of a family, citizen, solider, philosopher, servant of God”.

[47] Cf. Kant (1996, 455): Public law is a “system of laws for a people, that is, a multitude of human beings, (...) which, because they affect each another, need a rightful condition under a will uniting them”. See also Habermas (1976, 179): “the pretence to clear cut politics of the handling of social affairs, to cleanse social questions, a radically democratic formation of consensus that puts a stop to social repression- that is not a conceivable path for any modern society".

[48] Dworkin (1987b, 2): “[T]hough it is clear enough how a democracy differs from these other structures of government in general, democracy is itself an idea of great abstraction if not ambiguity. Democracy requires that officials should be elected by the people rather than chosen through inheritance or by a small group of prominent families or electors. But that abstract statement does not decide: which officials, if any, should be chosen not by the community as a whole but by sections or groups within it, how power should be distributed among officials chose in these different ways, how far elected officials should be permitted or required to appoint other officials to exercise some of their powers, which responsibilities should be held by elected and which by appointed officials, how long officials of either sort should serve, whether their terms of office should be fixed or subject to early termination by those who elected them, how far elected or other officials should themselves be free to change the constitutional arrangements under which officials are elected, whether a constitution should set limits to the powers of officials, so that the officials cannot themselves these limits and so forth. Though we are all democrats, these are lively questions among us, and some are matters of heated controversy

[49] This idea, clearly implicit in Between Facts and Norms, has been elaborated with greater precision in some latter articles. Cf. Habermas (1996a, 121): “The idea of self-legislation by citizens, then, should not be reduced to moral self-legislation of individual persons. Autonomy must be conceived more abstractly, and in a strictly neutral way” with Habermas (1995, 15): “Moral self-determination in Kant's sense is a unified concept insofar as it demands of each person, in propia personam that she obey just those norms that she herself posits according to her own impartial judgment, or according to a judgment reached in common with all other persons. However, the binding quality of legal norms does not stem solely from process of opinion- and will-formation, but arises also from the collectively binding decisions of authorities who make and apply law. This circumstance makes it conceptually necessary to distinguish the role of authors who make (and adjudicate) law from that of addressees who are subject to established law. The autonomy that in the moral domain is all of a piece, so to speak, appears in the legal domain only in the dual form of private and public autonomy” (my italics).

[50] Rawls (1971, 334).

[51] Rawls (1971, 333).

[52] There are two distinct versions of Rawls' argument concerning the existence of an obligation to obey the law. In the first, he grounds it exclusively on the idea of fairness (this corresponds to his 1964 article on civil disobedience). In the second, he adds the argument from the natural duty of justice. The duty of fairness has been converted into an additional ground for individuals that accept specific benefits offered by the social scheme of co-operation. In the Theory of Justice, (Rawls 1971, 335-6) he offers a rationale for the change: “There is still the question whether the parties in the original position would not do better if they made the requirement to comply with just institutions conditional upon certain voluntary acts on their part, or example, upon their having accepted the benefits of these arrangements, or upon their having promised or otherwise undertaken to abide by them. Offhand, a principle with this kind of condition seems more in accordance with the contract idea with its emphasis upon free consent and the protection of liberty. But, in fact, nothing would be gained by this proviso. In view of the lexical ordering of the two principles, the full complement of the equal liberties is already guaranteed. No further assurance on this score are necessary. Moreover, there is every reason for the parties to secure the stability of just institutions, and the easiest and most direct way to do this is to accept the requirement to support and to comply with them irrespective of one's voluntary acts”.

[53] Honor� (1987, 119): “Individual or social needs justify the imposition of special duties on people despite their unwillingness to take on the burdens that are involved. Membership of the communities to which we belong is not cost-free”.

[54] Habermas criticises the fact that Arendt relies on too facile (and historically distorted) distinction between politics and economics, which ends up by sterilising politics. She claimed that institutionalisation of the public sphere is needed in order to avoid the private sphere contaminating the public one. Habermas considers that there are other reasons, like the difficulties related to communicative action.

[55] Some authors have argued that one of the factors that explains the differences between ancient and modern forms of democracy is the realisation of the importance of filtering interests through the political process. In systems in which politics was a matter of action, it was seen as proper that election took place through ballots. Only in such a way it was possible to distribute the chances of getting involved in effective action. But with the emergence of the idea of legitimate pursuit of interests in the public sphere, ballots were substituted by electoral representation of interests. Cf. Manin (1996)

[56] Habermas (1996a, 150).

[57] Ackerman (1991).

[58] Habermas (1996a, 150) argues that an adequate mapping of politics must include not only spontaneous or non-regulated political interaction, but also “the use of administrative power within the political system, as well as the competition for access to that system”.

[59] Habermas (1996a, 460): "(...) [T]hen one can leave open the further question of whether there are moral grounds for entering a legal order in the first place- the problem that rational natural law posed as the transition from the state of nature to civil society. The positive law that we find in modernity as the outcome of a societal learning process has formal properties that recommend it as a suitable instrument for stabilising behavioural expectations; there does not seem to be any functional equivalent for this in complex societies (...) The philosopher should be satisfied with the insight that in complex societies, law is the only medium in which it is possible reliably to establish morally obligated relationships of mutual respect even among strangers"

Rawls (1964, 9): “The acceptance of a constitutional procedure is, then, a necessary political device to decide between conflicting legislative proposals”

[60] Alexy (1989, 16) refers to the following: “[I]ts statute-bound character; its necessary regard for precedents; its involvement with doctrinal studies as developed through an institutionally organised profession of academic lawyers, as well as -and this of course is not true for academic legal discourse- its subjection to the requirements of procedural ordinances and regulations”. See also Alexy (1989, 214ff).

[61] Dworkin (1982) and Dworkin (1986).

[62] The idea was already summarised by Proudhon in his motto Des r�formes tout jours, de utopies jamais.

[63] Cf. Habermas (1996a, 145): “In this situation [Europe of the XVIIth and XVIIIth century], the idea of government by law had the critical sense of uncovering the contradiction built into the established legal orders of the time: a normatively unjustified privileging of the most powerful interests certainly can be concealed in forms of legal authority”.

[64] On the transformative character of preferences, see Sunstein (1991), Gerstenberg (1998). Christiano (1996, 48) applies this argument to the conception of democracy “Democratic decision-making is not merely a matter of each person voting his or her preference. Individual citizens' preferences are formed in society as a result of social interaction they have with others and the institutions that structure social interaction. It is important for them to reflect critically on them and improve their preferences so as to have a sophisticated appreciation of their interests and ideals. A democratic theory ought to have something to say about what constitutes a reasonable and just context for the formation of these preferences”.

[65] Habermas (1996a, 179); “Majority rule retains an internal relation to the search for truth inasmuch as the decision reached by the majority only represents a caesura in an ongoing discussion; the decision records, so to speak, the interim result of a discursive opinion-forming process (...) Doubts about the legitimacy of majority decisions on matters with irreversible consequences are revealing in this regard. Such doubts are based on the view that the outnumbered minority give their consent to the empowerment of the majority only with the proviso that they themselves retain in the opportunity in the future of winning over the majority with better arguments and thus of revisiting the previous decision”. Cf. also Alexy (1992, 237ff), Alexy (1996, 224) and Nino (1996a, 37).

[66] Christiano (1996, 69).

[67] Fiss (1996) argues that equality is distributional. Free speech is not only a matter of expressing our interests, but it is a something connected to the essential pre-conditions for collective self-governance. In this sense, public institutions should be worried about avoiding unnecessary limits on speech, but they should also be concerned with ensuring that all sides are present to the public. This means silencing some to give voice to others. This explains the title of the book and the following remark by its end (at p. 83), that we need to embrace the “ironic truth, that the state can be a friend and an enemy of speech; that it can do terrible things to undermine democracy but some wonderful things to enhance it as well”.

[68] Dworkin (1996b), Fiss (1997).

[69] This is a major insight rendered explicit by Estlund (1997, 195): “I propose, as the counterpart of the idea of procedural fairness in cases where there is an independent moral standard for the outcome, the idea of Epistemic proceduralism: procedural impartiality among individuals' opinions, but with a tendency to be correct; the impartial application of intelligence to the cognitive moral question at hand”.

[70] Cf. Nino (1996a, 129), Estlund (1997, 196). As it is argued latter on, this is not so different from Raz's normal justification thesis.

[71] Nino (1991b, 46), Nino (1996a, 127).

[72] Estlund (1997, 188) for a watering down of Condorcet's theorem.

[73] See Nino (1996a) and Dworkin (1996a, 24-6).

[74] Rawls (1964, 15): “[J]ustice of the Constitution is a judgment influenced by the chances of correcting its injustices”.

[75] Ackerman (1989, 8)

[76] R�denas Calatayud (1996, 73).

[77] Raz (1984, 131): “Judges, if anyone, take the law as it claims that it should be taken. They, more than anyone, acknowledge the law at its own estimation. To understand legal statements we should interpret them as meant by those who take them and accept them at face value, those who acknowledge the law in the way it claims a right to be acknowledged. The decisive argument concerning the meaning of statements of legal duties is that the law claims for itself moral force. No system is a system of law unless it includes a claim of legitimacy, of moral authority. That means that it claims that legal requirements are morally binding, that is that legal obligations are real (moral) obligations arising out of the law”

[78] Habermas (1995, 116) considers that Rawls fills the design of the original position with a considerable deal of substantive content because of the way in which it operationalises the moral point of view (this is related to the pristine goal of the construction, to solve the Hobbesian problem) and offers a counter-model in which “he kept the procedural conception of practical reason free of substantive connotations by developing it in a strictly procedural manner”. Habermas (1995, 117): “Rawls imposes a common perspective on the parties in the original position through informational constrains and thereby neutralises the multiplicity of particular interpretative perspectives from the outset. Discourse ethics, by contrast, views the moral point of view as embodied in an inter-subjective practice of argumentation which enjoins those involved to an idealising enlargement of their interpretative perspectives” Habermas (1995, 131) compares discourse ethics with Rawls' theory of justice: “It can leave more questions open because it entrusts more to the process of rational opinion and will formation (...) I propose that philosophy limit itself to the clarification of the moral point of view and the procedure of democratic legitimation, to the analysis of the conditions of rational discourses and negotiations. In this more modest role, philosophy need not proceed in a constructive, but in a reconstructive fashion. It leaves substantial questions that must be answered here and now to the more or less enlightened engagement of participants, which does not mean that philosophers may not also participate in the public debate, though in the role of intellectuals, not experts”.

[79] Habermas (1996a, 301): “The self of the self-organising community disappears in the subjectless forms of communication that regulate the flow of discursive opinion- and will- formation, in such a way that their fallible results enjoy the presumption of being reasonable”.

[80] Habermas (1996a, 179): “Majority rule retains an internal relation to the search for truth inasmuch as the decision reached by the majority only represents a caesura in an ongoing discussion; the decision records, so to speak, the interim result of a discursive opinion-forming process (...) Hence, the dissenting opinion attached to the justification of a Supreme Court ruling, for example, is meant to record arguments that in a similar case might convince the majority of a future panel of judges”

[81] Nino (1996a, 37).

[82] Ely (1980).

[83] Rubio Mar�n (1998) and Rubio Mar�n (2000).

[84] Nino (1991a, 115): “[L]iberal principles of justice take consent as the primary element for justifying a distribution of rights and obligations, benefits and burdens, and only when there is no such consent do they take into account other factors (...) If that consent is not present, because of deception, ignorance, or the mental state of the person in question, the other criteria of justice come into play, depending on how much love or fraternity there is among the parties (...) So, principles of justice are permanently in the background, not only legitimising the relationships based on consent but also providing substantive criteria which allow us to interpret just how friendly, disinterested, and loving those relationships are”.

[85] See Alexy (1996).

[86] See Alexy (1994, pp. 131-57) and Alexy (1996).

[87] See Habermas (1996a, chapter 3, especially at pp. 122-3).

[88] Hart (1955, 61).

[89] Rawls (1971, 112). Previous formulations can be read in Rawls (1964, 9-10): “If one thinks of the Constitution as a fundamental part of the scheme of social co-operation, then one can say that if the constitution is just and if one has accepted the benefits of its working and intends to continue on doing so, ad if the rule enacted is within certain limits, then one has an obligation based on the principle of fair play, to obey when it comes one's turn. In accepting the benefits of a just constitution one becomes bound to it, and in particular one becomes bound to one of its fundamental rules: given a majority rule in behalf of a statute, it is to be enacted and properly implemented (...) The principle of fair play may be defined as follows. Suppose there is a mutually beneficial and just scheme of social co-operation, and that the advantages it yields can only be obtained if everyone, or nearly everyone, co-operates. Suppose further that co-operation requires a certain sacrifice from each person, or at least involves a certain restriction of his liberty. Suppose finally that the benefits produced by co-operation are, up to a certain point, free; that is, the scheme of co-operation is unstable in the sense that if any one person knows that all (or nearly all) of the others will continue to do their part, he will still to be able to share a gain from the scheme even if he does not do his parts. Under these conditions a person who has accepted the benefits of the scheme is bound by a duty of fair play to do his part and not to take advantage of the free benefit by not co-operating. The reason one must abstain from this attempt is that the existence of the benefit is the result of everyone's effort, and prior to some understanding as to how it is to be shared at all, it belongs in fairness to no one” and Rawls (1969, 241): “A person is required to do his part as defined by the rules of an institution when two conditions are met: the institution is just (or fair), that is, it satisfies the two principles of justice; and, second, one has voluntarily accepted the benefits of the arrangement, or taken advantage of the opportunities it offers to further one's interests”.

[90] This interpretation of the principle of fair play can be traced back to Honor� (1987, 119ff, especially 122): “A citizen need not have chosen to become a citizen nor have made any promise, express or implied, to do so. Nor need he have voluntarily accepted any benefit from the state whose laws are in question. The duty applies to a reluctant and uncooperative as it does to the most active and prosperous citizen”

[91] Rawls (1971, 112): “By the principle of fairness it is not possible to be bound to unjust institutions which exceed the limits of intolerable injustice (...) In particular, it is not possible to have an obligation to autocratic and arbitrary forms of government”. A similar argumentation in Rawls (1969, 240), where he argues that the question of the existence of a general obligation to obey the law based on fairness only makes sense in “a legally established and democratic authority”. In Rawls (1964, 5) this is associated with a legal system based on the rule of law (publicity of rules, similar treatment of similar cases and so on) and that is part of constitutional democracy (that implies equal citizenship, recognition of the freedom of the person, freedom of thought, liberty of conscience and political participation in the political process”

[92] Cf. Rawls (1999a, 60): “The rights and duties so arising are special rights and duties in that they depend upon previous actions voluntary undertaken, in this case on the parties having engaged in a common practice and knowingly accepted its benefits. It is not, however, an obligation which presupposes a deliberative performative act in the sense of a promise or contact or alike”

[93] If we assign to fair play a complementary role, the criticism addressed by Klosko to authors like Rawls and Simmons is correct. The latter consider that we can also be obliged on the basis of fairness in case that we fully consent to the benefits derived from social order. Under the present understanding of the role played by fair play, what is more relevant is the magnitude of the benefits provided. Moreover, this view is no longer open to the kind of criticisms made by Nozick (though it is far from clear whether his arguments bite even to Rawls's understanding of fair play. Having said that, I doubt whether Klosko sees the role of fair play as I do. If he does not, then his argument (that is right under the present understanding of fair play) is not so convincing. For the sake of thoroughness, the reader should know that he supports the suppression of a direct reference to consent within the argument for fair play. However, he does so by means of limiting the breadth of the general obligation to obey the law to those legal systems entrusted with the provision of a very limited number of public goods (those characteristic of a minimal state a la Nozick) and conditioning it to the fact that such goods are worth their cost. Cf. Klosko (1998, 62). He argues that because the benefits provided by such schemes are non-excludable, individuals are no longer free to decide whether or not to receive them. Cf. Klosko (1989, 243). His strategy becomes clearly untenable when he pretends to extend the reach of the obligation. He argues that if the provision of very basic public goods is fair, the legitimacy of the provision of further goods is to be regarded as legitimate, unless there is proof to the contrary. This is quite ingenious, but a trifle too mechanical.

[94] See, for example, MacCormick (1978, 74): “Adjudication implies running beyond the specific and determinate guidance that rules can give). See also Alexy (1989, 1), and in general, all treatises on legal argumentation.

[95] That corresponds to the authoritative dimension of legal reasoning. Cf. Alexy (1999, 375)

[96] Cf. Dworkin (1986, 227): “According to law as integrity, propositions of law are true if they figure on or follow from the principles of justice, fairness, and procedural due process that provide the best constructive interpretation of the community's legal practice”

[97] See, for example, Dworkin (1986, 72 and 251-2).

[98] Cf. MacCormick and Weinberger (1986:chapter 1).

[99] Cf. Alexy (1989, 1).

[100] Alexy (1999).

[101] Cohen and Rogers (1995).

[102] Habermas (1996a, 169, 173, 300). It is worth quoting pp. 173: “The priority of laws legitimated in democratic procedures has the cognitive meaning that the administration does not have its won access to the normative premises underlying its decisions”

[103] Habermas (1996a, for example 39) takes from Parsons the idea that administrative bodies constitute a source of integration in modern societies.

[Date of publication in the ARENA Working Paper series: 15.10.2000]