RITORNA AL SOMMARIO
Analysis of the evolution of affirmative action in the Community system and in the United States reveals different phases: in the first period (which dates to the late 1960s-70s for the United States, and to the mid-1970s in Europe) affirmative action was considered a necessary means to achieve equality for all citizens, an instrument that would go beyond the formal legal dimension and ensure a factual equality of opportunity. Subsequently, the concrete use of this mechanism has forced legal scholars to develop a better theoretical definition of its characteristics and limitation, which proved essential after the first wave of affirmative action plans, when it became clear that the use of this instrument might cause hardship for innocent people, and that it was urgently necessary to devise criteria with which to balance the interests in conflict. The solution was anything but straightforward, partly because the characteristics of affirmative action plans varied greatly according to the concrete circumstances to which they were applied, but most of all because there was more profound disagreement on the objectives to be pursued, and even more on the means to that end (1).
Political ideologies have undoubtedly played a major role in shaping public opinion on affirmative actions: under pressure by powerful movements like the civil rights movement in the United States in the 1960s, or the feminist movement in both the U.S.A. and Europe in the 1970s, it became widely accepted that formal equality was not enough, and that what was required were equal chances for all. But when the tide changed, and this political ideal gave way to more moderate positions, and sometimes even to political agnosticism, it became clear that the legal system moved in rather different directions. It provided mechanisms to satisfy these exigencies, but it failed to provide a strong theoretical basis for them. Therefore, once the political impetus had petered out, the legal framework was unable to cope with attacks directed against affirmative action as a workable and admissible legal tool. The line of cases recently decided, independently of the legal system concerned, shows that the dispute ultimately turns on questions of distributive justice, and that it is at this level that the "match" is decided (2) . The legal form of the solution is a consequence of it, not its basis. This may be an inevitable outcome for all problems with a strong "equity" connotation, and which therefore involve meta-legal questions of resources distribution. Nevertheless, one cannot fail to conclude that the legal framework is extremely weak, and that its contribution to solution of this problem seems almost entirely limited to giving a (not always) suitable form to a primarily political decision.
The history of the Western legal tradition shows that law possesses an array of concepts and solutions that, although they cannot solve distributive problems on their own, may help to rationalize their handling and render it more efficient (3) . Moreover, it also shows that every time that law has sought to avoid social problems that were too controversial, this choice has led to disastrous consequences (4) . The specificity of law consists of its strive for neutrality, and it is therefore poorly equipped to tackle this kind of problems. The sphere of substantial decisions properly belongs to the political arena, where all social actors confront their positions. Nevertheless, the study of the legal process shows that this division of roles implies that every actor has a specific and important task to perform (5) . Law has to make sure that these choices are made and implemented correctly, and that all individuals have at their disposal efficient legal remedies to protect their rights against any kind of breach. This is the sense of one of the most fundamental legal principle, the due process of law, enshrined in the fourteenth amendment of the U.S. constitution (6) and in many similar provisions of the basic laws of several legal systems. This process must be guaranteed by various legal actors, having different characteristics: legislators have to pass meaningful legislation, which has to be duly applied by bureaucracies; courts have to control that the enactment and implementation of legal rules does not infringe on individual rights contained in superior rules or general legal principles. The smooth working of law in guaranteed only if all legal actors perform their duties efficiently and correctly.
During the last part of the nineteenth century, and even more in the present one, the tenet of neutrality of law has come under severe attack. Many scholars with different backgrounds and believes have denounced neutrality as being a myth created by lawyers in order to legitimate their power and their autonomy from the political sphere, a myth that conceals in reality bias and discrimination (7) . There is surely much truth in this belief: law has always gained neutrality by leaving out important sectors of society, as has been the case with slaves in Roman law or in the first period of American law. Nevertheless, neutrality was and remains the fundamental character of law: all legal evolution points to the enlargement to categories of persons previously discriminated against, like slaves, workers, or women (8) . The fact that we are far from attaining this aim does not prevent it from being the very reason for the existence of law in Western societies. We are in a world of second-best, and the existence of affirmative action demonstrates the fact that there are fields where law as a neutral tool shows its intrinsic limits as a means of resolution of social disputes. It demonstrates as well that the abandonment of neutrality is a dangerous path that may lead to the loss of legitimacy of law. Striking a balance between these two contrasting aims is a difficult enterprise, but there is no other choice for lawyers but try it. After all, there may be something more to say about law and affirmative actions.