For the last 30 years there have been debates in philosophy about the concept of rationality. In anglo-american circles they have been primary characterized as discussions about “justified beliefs.” By contrast, the debate in Germany among discourse-theorists (Habermas, Apel) has been linked to the problem of justification of communicative speech-acts (within the concept of the entire communicative situation). Herbert Schnädelbach has modified the discours-theoretical account. His concept can be regarded along a number of dimensions. He has developed a linguistic analysis of the validity claims made in various speech-act situations (following Paul Taylor). In Schnädelbach’s view, normative characterizations of speech acts can be used descriptively, too. Furthermore, the hierarchy of justifications ends on the level of rational choice.My starting point here is: there is a fundamental distinction between justifications in everyday-life and a theoretical level of justification (discoursive justifications). Thus, I discuss the hierarchy of justifications developed by Schnädelbach.Furthermore, I evaluate Schnädelbach’s concept in a semantic perspective. I will show that Schnädelbach’s pragmatical account requires a semantic analysis. Without recourse to such analysis, we cannot understand the universe of normative language.

The ongoing discussion about a notion of pragmatic rationality has evolved in a variety of different approaches, mainly because every author tries to combine his genuine philosophical point of view with the interpretation of “rationality”. Nevertheless there is an agreement of sorts that rationality cannot proceed mereley descriptively but has also normative implications.The paper investigates the proposals of Nicholas Rescher and Herbert Schnädelbach concerning the question of a normativity of rationality. Moreover it deals with the problem of “unity of reason” and its interconnectedness with the inherent normativity of rationality.

Rational explanation, as I understand it here, is the sort of explanation we practise when we try to make intentional sense of a person’s attitudes and actions. We may postulate various obstacles to rationality in the course of offering such explanations but the point of the exercise is generally to present the individual as a more or less rational subject: as a subject who, within the constraints of the obstacles postulated - and they can be quite severe - displays a rational pattern of attitude - formation and decision-making.In this paper I want to draw attention to three distinct, and progressively more specific, aspects of such rational explanation. I do so, because I believe that they are not always prised apart sufficiently. The first aspect of rational explanation is that it is a programming variety of explanation, in a phrase that Frank Jackson and I introduced some years ago (Jackson and Pettit 1988). The second is, in another neologism (Pettit 1986), that it is a normalising kind of explanation. And the third is that it is a variety of interpretation: if you like, it is a hermeneutic form of explanation.

Our rationality depends on the reasons we have for accepting and preferring what we do. But where do reasons come from? What makes what I accept a reason for a conclusion or what I prefer a reason for action? We can explain where reasons come from without postulation or regress. The explanation rests on our trustworthiness combined with our acceptance of it and our preference for it. The explanation reveals that theoretical and practical reason are intertwined in a loop of trustworthiness in what we accept, what we prefer and how we reason. The loop is the keystone of our rationality.

Even though the concept of right is not empirical, Kant does not deduce right in a transcendental manner. If in conformity with the rational principles of transcendental philosophy, we try to understand why this is so, the answer may be found in an analogy with aesthetic reflection. Indeed, aesthetic reflection might contain the transcendental ground of violence in civil society.

The "Why be moral?" problem has been one of the more persistent problems of ethics. The problem is typically posed as a conflict between what is straightforwardly maximal for a person to do in specific circumstances and what is recommended by the principles or rules of ethics, usually what is communally optimal, in those circumstances. Typically ethicists try to convince us that both collectively and individually we will be better off in the long run if we each adopt cooperative strategies despite the temptations of immediate profit offered by straightforward maximization policies. After reworking the notion of "straightforward maximizer" such that it makes sense to say that I may sometimes have rationally good reasons to perform actions that do not in the circumstances, taken individually, maximize my utility just so long as the best-for-me accessible-to-me possible world is realized, l am able to show why it is the case that in social interactions that mirror iterated Prisoner's Dilemmas the constraints of ethics on straighforward maximization are redundant. The policy of straighforward maximization that I defend is more flexible that one of cooperation. It reaps the benefits of cooperation when they are to be had and avoids the disasters of cooperation that lurk in every meeting one has with potentially treacherous strangers. Where the policy of straightforward maximization departs from the ethical choice, I argue, it does so because making that choice would be scrificial, supererogatory, even from the moral point of view. Acting ethically by constraining one's straighforward maximinzation therefore cannot be rationally justified.

It is argued that full-blown cooperation needs collective goals in a strong sense satisfying the "Collectivity Condition". According to this condition, a collective goal ist of the kind that necessarily, due of the goal-holders acceptance of the goal as their collective goal, if it is satisfied for one of the goal-holders it is satisfied for all the others. Not only collective goals but also other group-factors (such as possibly institutionalngroupmoden preferences and utilities) are argued to be relevant to rational cooperative solutions of collective action dilemmas.

A recurrent topic among philosophers as well as social scientists since Novalis, Comte, Weber, modem existentialists, and post-modern sociologists, etc. is that in the absence of what Tocqueville called "dogmatic beliefs” values cannot be grounded : you prefer liberty, I prefer equality; none of us would be neither right nor wrong. Contemporary writers as Rawls and Habermas defend, against this current view, the idea that value statements can be grounded rationally. Habermas' theory of communicational rationality remains procedural, formal and on the whole mysterious, however: how can this peculiar type of rationality be definied and made analytical? A cognitive theory of axiological rationality is developed here starting from the basic point that normative statements and axiological beliefs should be analyzed as grounded on reasons with a transsubjective validity, as positive statements are. This theory is checked in a tentative fashion against some examples of axiological belieft from ordinary experience and against a few pieces ofdata drawn from experimental social psychology.

What type of reason will work under the present conditions? To answer this question a meaningful conception of reason (as distinct from rationality) has to be developed, and is contemporary conditions (due to change in the field of rationality) have to be specified.In part I. of the paper, the radically altered structure of rationality is analysed; it turns out to be characterized by rational disorder. Part II. offers a redefinition of reason; guided by the idea of justice reason operates in transition from one rationality to another. This new kind of reason - "transversal reason " - is further elaborated in part III. It is regarded as a key-element of any type of reason.