Abstract
The work of Rainer Forst constitutes the third generation of the Habermasian School. In Das Recht auf Rechtfertigung [The right to justification] (2007) Forst develops a constructivist approach to justice in a serious effort to find a systematic basis for ‘critical theory’. In this article the relevant arguments of this approach are critically analysed. The position developed in the work of Forst appears to be characterized by a fundamental ambiguity because it oscillates between two irreconcilable points. On the one hand Forst seems to admit the necessity of something like the ‘unconditionality of moral law’ presupposing a kind of Kantian transcendentalism beyond constructivism. But on the other hand, he turns back to a more Habermasian constructivist or proceduralist theory. The ambiguity implicit in Forst’s philosophy is that it does not find a compromise between these two approaches to discursive theory so it fluctuates between them, needing the logical support of both poles.
I Introduction
The work of Jürgen Habermas constitutes a turning point in the history of the Frankfurt School that was very much inspired by an interpretation of American pragmatism and by the work of his colleague Karl-Otto Apel. The Habermasian School, which could be called the second Frankfurt School, distinguishes itself from the first by a multidimensional view of reason. It is a critical theory based on a theory of discourse and it abandons the Marxist paradigm. This Habermasian School is now turning into what could be called its own ‘third generation’. 1 Taking as a point of reference the discursive theory of Habermas, the second generation, led by Axel Honneth, concentrates on the dimension of recognition [Anerkennung] as the foundation of discursive practices and intersubjective relations. The third generation is constituted by the work of Rainer Forst and uses a more abstract theory of justification. In the work of both Honneth and Forst there is a shift from the Habermasian position to a point of interest that claims to be more basic and fundamental for the constitution of a genuine normative dimension inherent to discourse. Both Honneth and Forst turn back to German traditions, the former to Hegel and the latter to Kant, in order to make their claims understandable. The work of Forst is developing rapidly since the publication of Toleranz im Konflikt [Tolerance in conflict] in 2003, with the appearance of Das Recht auf Rechtfertigung [The right to justification] in 2007 and with his Kritik der Rechtfertigungsverhältnisse [Critique of the conditions of justification] in 2011. 2 But in fact the main ideas of all these works were conceived in his earlier book Kontexte der Gerechtigkeit [Contexts of justice] of 1994, published in English in 2002. In Das Recht auf Rechtfertigung [The right to justification], however, Forst undertakes a serious effort to find a systematic basis for his earlier ideas, which gives this work a pivotal position in the development of his philosophy. This book develops a very complex argument which is made difficult to understand by the rather contingent sequence of its chapters (Das Recht auf Rechtfertigung [The right to justification] is a juxtaposition of articles) and by the extremely detailed exposition of, and discussion with, several contemporary thinkers. The importance of this work for the understanding of Forst’s thinking, and for the understanding of the development of Frankfurt philosophy, requires a critical analysis of its main arguments leaving out many of the detailed discussions which complicate the picture. But before doing that it seems appropriate to give a short impression of Forst’s early work in order to understand what the starting point of his philosophy is, what he is aiming at and what his relationship is to his predecessors.
The early work
Kontexte der Gerechtigkeit [Contexts of justice] is an immersion into the debate between communitarians and liberals which took place in the 1980s and early 1990s of the last century. It is an attempt to design a separate position, based on Habermas, that defends the idea of a moral universalism that is also contextual (1994: 239–346). This idea is at the core of Forst’s interest which is to develop a theory of justice synthesizing the ideas of both communitarians and liberals. How can we conceive a theory of morals and justice that is at the same time universal by taking up the claims of liberal philosophy, and contextual in that it does justice to the claims of communitarians? Of course, this is not possible if we consider the contextual approach to be equivalent to relativism. Forst distances himself from contextualism in the sense of a limitation of the validity of moral claims. His contextualism presupposes that discourses claiming normative validity have many domains, the main one being the context of morality which is always directed towards the constitution of universal claims. The ethical, juridical and political domains are relatively independent from morality. They develop their own modes of justification and argumentation and their own criteria of evaluation, which can, but should not, contradict the principles of morality developed in the moral context. A theory of justice that wants to be universal must be based on a consideration of the domain of morality. The ethical, juridical and political are variable in time, culture and community. The philosophical foundation of morality makes an evaluation of ethical, juridical and political issues possible, but even with a positive evaluation these contexts can have different cultural shapes all compatible with the principles of morality. The main issue motivating Forst’s philosophy therefore seems to be the question of how to combine universalism and pluralism. Next to the variance of contextual languages there is a moral Esperanto guaranteeing a neutral and universal position. Together with Apel and Habermas, Forst defines reason as the search for a universal justification that claims general validity and that must be acceptable for all concerned. Discursive theory is an attempt to reconstruct the ‘recursive’, that is, the necessary formal and pragmatic [formalpragmatische] presuppositions of discourse which regulate not only reason but also a morality inscribed into reason, and this includes the ideas of reciprocity, generality, solidarity and respectfulness (1994: 301). From these principles the context of morality derives its right to modulate the contexts of ethical, juridical and political argument.
In Toleranz im Konflikt [Tolerance in conflict] Forst pursues two aims. The first is a historical reconstruction of the idea of tolerance as it appears from the beginning of modern philosophy. The second aim is to integrate the idea of tolerance with his theory of different normative contexts which he developed in his earlier work. For our purpose the theoretical part is more relevant. Forst develops a theory of tolerance starting from a theory of discourse. To be tolerant means to be capable of imagining a community of disagreement (2003: 12). In this new setting, the problem of unifying universalism and pluralism reappears. There must be a domain universally guaranteeing the necessity of tolerance (the moral context) and several domains where specific disagreements should be possible (the ethical, juridical and political contexts). Only by separating a moral context that is universal from other contexts does it become possible to solve the so-called ‘paradox of tolerance’ which is the idea that there are limits to tolerance, that intolerance should not be tolerated (ibid.: 37). Tolerance guarantees the existence of a plurality of opinions on the level of the ethical, juridical and political discourse, but on the level of morality philosophy must develop norms and principles that are capable of being accepted reciprocally and generally. Forst takes up an idea that he introduced rapidly in Kontexte der Gerechtigkeit [Contexts of justice] (1994: 133) and that will be pivotal in his next book: the basic principle making such a development of general norms possible is the principle of justification which denotes an expectation that all participants of discourse necessarily make when they engage in dialogue (2003: 19, 590). There may be no intolerance without a justification capable of being accepted by all concerned. So, in Toleranz im Konflikt [Tolerance in conflict] Forst already conceives all the theoretical terms that will be extensively elaborated in Das Recht auf Rechtfertigung [The right to justification] including the criteria of reciprocity and generality (ibid.: 596, 630).
Searching for a systematic approach
This short incursion into the early work of Forst shows that his main concern is to establish the normative authority of a universal idea of justice. This aim culminates in the systematic approach of Das Recht auf Rechtfertigung [The right to justification] which can be seen as the theoretical outcome of the preceding work. Forst becomes more and more dependent on a constructivist interpretation of Kant, who seems to be the philosopher accompanying all his thoughts. He moves towards a general discursive theory of reason incorporating not only the discursive theory of Habermas, but also the theory of recognition developed by Honneth. In this sense the so-called third generation of the Habermasian School seems to be working on a synthesis of the achievements of its predecessors.
In the following discussion I would like to show that the position developed in Das Recht auf Rechtfertigung [The right to justification] and in several passages of his earlier work is characterized by a fundamental ambiguity because it oscillates between two irreconcilable moments. On the one hand Forst seems to admit the necessity of having something like ‘moral insight’ claiming a kind of ‘unconditionality’, which takes his theory in the direction of transcendentalism showing many resemblances to the work of Karl-Otto Apel and influenced by Kant’s transcendental philosophy. On the other hand, Forst distances himself from this position and turns back to a more constructivist or proceduralist theory following the example of Habermas and the constructivist Kantianism of Onora O’Neill (1994: 291). 3 The ambiguity implicit in Forst’s philosophy is that it does not find a midway nor a synthesis between these two approaches to discursive theory so it moves back and forth between them, needing both as fluctuating poles. The ideas of an ultimate foundation [Letztbegründung] and of an ideal speech community elaborated by Apel, and also the idea of self-evidence that Dieter Henrich developed from Kant’s representation of a fact of reason [Faktum der Vernunft] seem to impose themselves as necessary outcomes of Forst’s theory of discourse. However, he does not think this through, possibly because he wants to keep in line with the constructivism of Habermas and his consensus theory of truth. The overall ambiguity of Forst’s position is that in its interpretation of morality it implicitly moves towards ontological realism, but explicitly presents the contrary, a constructivist interpretation denying this possibility. The deepest reason for this ambiguous position is Forst’s desire to avoid metaphysical claims and to present a totally secularized theory of justice.
The ambiguity appears when Forst relates justice to a theory of reason because he veers towards anthropology, speaking not of reason in general but of human reason, of intersubjective communication. This keeps him far from the metaphysical idealism that inspired Kant and his Romantic followers. But this does not prevent him from trying to give a sketch of a general theory of practical reason, aiming at a transcendental foundation of the criteria of justice which as we will see has clear metaphysical consequences. Three moments seem to be fundamental in this more general theory of reason: first, the idea of the inevitability or inescapability of reason, which is an answer to the question ‘Why reason?’ (§ 2); second, the very basic answer to the question ‘What is reason?’ in the sense of ‘How is reason structured?’ which implies a theory of justification that Forst develops as a theory of practical reason (§ 3); and third, an explanation of the most fundamental motive for the realization of reason, which would be an answer to the question ‘Why follow reason?’ or ‘Why be reasonable?’ (§ 4). In all three moments Forst’s design suffers from an ambivalent approach that takes transcendental insights to be merely constructions.
II The inescapability of reason
The first question to ask when elaborating a rational theory of justice is ‘Why should a theory of justice be rational?’ Of course all theory strives to be rational, so we could say that Forst is considering here a fundamental philosophical problem of epistemology. Reason, according to Forst, is inescapable and therefore it is nonsensical to ask ‘Why be rational?’ (2007: 55). Asking or answering this question already implies a commitment to rationality. And this, according to Forst, also applies to the question ‘Why be moral?’ Once we have understood that there is, in the centre of rationality itself, a normative core (because reason involves a duty to justification), such a question becomes superfluous. Reason holds in itself an imperative directed to the instance using this reason and has in itself a duty to justification.
This argument making reason an inescapable and therefore irrefutable and necessary first insight is also the basic point of the claim that reason is autonomous (2007: 56–60). According to Forst, the autonomy of reason means that there is nothing that can justify reason but reason itself. Reason cannot be deduced from any higher authority. Reason is autonomous, because reason ‘quasi justifies itself’ (ibid.: 56). It justifies itself because reason, in the recursive process of self-justification or self-reconstruction, is searching to understand its own principles and rules (1994: 359; 2003: 592).
This argument of the autonomy of reason is extended to that of the autonomy of morality (2007: 74–99). The autonomy of morals is based on the autonomy of reason, because practical reason is part of reason. Autonomie der Moral [The autonomy of morality] means that there is no higher authority justifying morality than reason itself. Forst is trying to say that morality is not autonomous from reason, but that it is autonomous thanks to reason. He also relates the autonomy of morals to the idea of personal autonomy, but this is of course just a variation of the idea that morality depends on reason because it is rationality that makes a person autonomous. The reference to reason also means that morality cannot depend upon arbitrary cultural presuppositions or prejudices like religious laws or other random customs or manners (ibid.: 10, 77). If we want to answer the question ‘Why be moral?’ then we cannot escape discursive argumentation (ibid.: 57). We are already involved in the normative dimension of reason, because reason implies an obligation: we commit ourselves to the production of reasons [Gründe]. Once we feel this duty we know that we have crossed the bridge from cognitive to practical reason.
So, just as a theory of reason is autonomous when it is based on unconditional and irrefutable principles, a theory of morals can claim autonomy when it develops its principles in a similar unconditional way. Forst in fact follows Kant, who was concerned with the unconditional duty [die unbedingte Pflicht] related to the ‘imperatives of morality’ [das Sollen], but following O’Neill he translates Kant into a discursive theory and situates the development from cognitive to practical or normative reason in the moment in which every cognitive sentence implies an ‘unconditional duty of justification’ (2007: 61). It is this relation to a discursive situation which makes him say that his position is a constructivist and contextualist one. Although he clearly states that the principle of justification is not a product but a kind of evidence, he considers this evidence to be the construction of a ‘recursive analysis’, a self-reconstruction of reason searching to understand its own principles and rules, that has to start in a contextual situation (1994: 359; 2003: 592; 2007: 56). The recursive method starts from an immanent analysis of contextual discourses (he speaks of a ‘kontextimmanente Analyse’) reconstructing several layers of higher foundations which reveal the right to justification inherent in reason itself (ibid.: 62).
Here there is, I think, a problem in Forst’s argumentation. Although he never identifies his own argumentation as being transcendental and he even avoids the term ‘quasi-transcendental’ still applied by Habermas, he in fact has been deploying a transcendental argumentation which has at its core a strong similarity to arguments of ultimate foundation. This amounts to the discovery of a principle of justification and its corresponding right to justification which are immanent to reason and not in any way constructions of man. But he nevertheless declares his position to be a constructivist one because it starts in a contextual discursive situation. Even if Forst was conscious of this ambiguity, the problem remains that although he discovers principles beyond human construction, he nevertheless identifies them as constructions of a recursive contextual analysis.
To clarify this ambivalence let us have a look at Forst’s so-called recursive method. This method starts from a discursive contextual situation in which we need to move to higher and more abstract dimensions of justification. Forst seems not to be very conscious about the fact that in order to understand these evidences of reason (like the first principle of justification and the idea of autonomy) we need a kind of Husserlian bracketing of the lifeworld and its specific contexts of discourse. Indeed, it is only then that we can understand that rationality is inescapable. Forst should have made clear that we need some methodological steps to understand this. One main step needed is an epoché or bracketing of the lifeworld. If we connect the question ‘Why be rational?’ to actions taking place in a specific context we will have to acknowledge the possibility of non-rational actions, which would not allow this question to be seen as being nonsensical. We can only understand that reason is inescapable if we separate reason from a specific context of action. As long as reason stays connected to a specific contextual action, there will be nothing nonsensical in the question ‘Why be rational?’ because it will always be possible to act in non-compliance with reason. But the question becomes nonsensical from the moment that we generalize it and separate it from any particular context. So, we have to abstract this question from the lifeworld and go through a process of bracketing that moves away from specific contexts of argumentation (including the ethical, political and juridical contexts of justification specified by Forst) sticking to a particular highly abstract sentence (‘Reason is inescapable’) that cannot be negated on its logical level without contradicting oneself. This is what it means to come to higher and more abstract dimensions of justification. It is on this level that it becomes evident that we cannot escape reason.
This amounts to a position that is not far from the accentuation of reason which we find in the transcendental tradition from Kant and German Idealism onwards. The point that reason is inescapable is therefore recognized as an irrefutable and necessary first insight, escaping all arbitrariness. In fact Forst works with insights of a theory of ultimate foundation [Letztbegründung], although he never uses these terms to denote his own position. These insights also constitute the basic point of his claim that reason is autonomous, because ‘autonomy’ means that everything could be negated and questioned [Alles einzelne … läßt sich hinterfragen und kritisieren, (2007: 56)] except, of course, reason itself. Reason cannot be deduced from any higher authority. The claim that there is nothing prior to reason justifying reason can be equated to pronouncing sentences like ‘reason is inescapable’, ‘reason is inevitable’, or even ‘reason is absolute’ which are ultimately grounded sentences, because they cannot be denied without contradicting oneself. 4 Instead of naming things for what they are, acknowledging his debt to the transcendental theory of ultimate foundation, Forst stresses the fallibility and fragility of human reason which, according to him, make ultimate foundations impossible (ibid.: 68). But in fact Forst has shown that being finite or fallible does not make it impossible to describe irrefutable evidence. Admitting irrevocable insights does not mean that a theory is complete and that it accepts no further revisions. So, Forst first presents the autonomy of reason and the principle of justification as irrevocable certainties, but then he seems to retract saying that irrevocable foundations are not possible for mankind.
Summarizing this section we may say that although Forst’s theory takes him to a realm of pure reason he never sticks to the point to which his reasoning brings him. He always turns back to contextual situations, forgetting the bracketing that was necessary to make his point. He emphasizes that his reasoning takes place as an immanent analysis of contextual situations (his ‘kontextimmanente Analyse’), but he seems unaware of the fact that his own recursive method needs a bracketing and a passage to transcendental and higher levels of abstraction involving a detachment from all specific contexts in order to arrive at the basic unconditional principles of reason. By emphasizing contextualism he tries to justify his constructivist position, but in fact he is presenting us with a transcendental analysis that takes essential insights from ultimate foundation theory. He admits that although all principles of morals are ‘constructions of man’, the first principle of justification is not (2007: 38). The duty to justification must not be taken like a normal obligation ‘as something produced by us, but must be taken as something that we have in an absolute mode’ [‘die man schlechterdings hat’ (ibid.: 61)]. Forst describes this moment of unconditional evidence using an expression of Henrich who speaks of ‘moral insight’ [sittliche Einsicht (ibid.)] and who used this expression in order to find an equivalent for Kant’s notion of the ‘fact of reason’ [Faktum der Vernunft]. And Forst admits that this ‘fact’ denotes unconditional evidence (ibid.: 62). All claims of constructivism seem to disappear here, because our duty to justify ourselves is unconditional and not a construction of man, but a principle of inescapable reason itself. But despite this admission of a fundamental non-constructivist part of reason Forst sticks to constructivism, because according to him the evidence of the unconditionality of reason can be substituted by what he calls a ‘context analysis of validity claims’. 5 The evidence that Kant, Henrich and he himself are speaking of, is according to Forst not ‘immediate evidence’ because we need a ‘recursive analysis’ to come to it. This means that we have to start in a contextual situation reconstructing several layers of higher foundations bringing us to the revealed first moral principle inherent in reason itself (ibid.). This reconstruction belonging to the method of ‘recursive analysis’ is a construction too and makes it possible to maintain the claims of constructivism. But Forst seems here to be confusing the level of the explanans (the methodological level) with the level of the object of inquiry, the explanandum. Of course, a method is always a construction, but this does not mean that the discovered object is a construction too. The ambiguity of Forst’s situation is therefore that he oscillates between two irreconcilable positions. His argument takes him to the point of claiming a kind of ultimate foundation that is an unconditional and irrefutable evidence (like Henrich’s idea of ‘moral insight’ or Kant’s ‘Faktum der Vernunft’), but then he crawls back clinging to the reality of contextual discourses where his argument and method began. The ambiguity of Forst’s position is that, after leaving the valley of lifeworld existence to climb the mountain of reason and morality, he declares the first principles found on the top to be just constructions made in the valley.
III The fundamental structure of reason
Resolving the question ‘Why reason?’ we have already seen that the right to justification plays a fundamental role in Forst’s theory of justice because this right is a unique root to which all other rights are connected. The insight that the right of justification is inherent in reason is partly an answer to the second main question driving Forst’s work which is the question ‘What is reason?’ in the sense of ‘How is reason structured?’ Forst is mainly interested in practical reason which means that the elements of reason that he emphasizes are all relevant for a theory of justice. I will try to summarize Forst’s position addressing three points that in my view are fundamental for understanding the structure of practical reason. (1) The first point concerns the ultimate principle of practical reason which according to Forst seems to be the right to justification. This principle becomes the starting point of a systematic theory of justice, but it is possible to doubt about the primacy of this right. (2) The second point concerns the main criteria of practical reason which according to Forst are generality and reciprocity. These criteria can be used in order to determine what justice is in a general way (independently of contextual issues) and remain therefore to a large extent formal. I will try to show that Forst is wrong in considering them to be ‘criteria’ and that they need to be complemented by at least one other important formal principle. (3) The third point is about a claim of substantialism that Forst’s theory of justice makes, about how this claim conflicts with his declared constructivism and about what the real presupposition is that confers on Forst’s theory of justice its critical power. I will show that this implicit element, which Forst does not mention, is in fact the transcendental concept of the ideal community of communication. Forst therefore remains indebted to early Habermas and even more to Apel, which also means that he does not escape, as he claims, the transcendental and metaphysical presuppositions related to the ideal speech situation.
1
The principle of justification constitutes, according to Forst, the core of reason and a normative translation of this principle is the right to justification. Like the criteria of generality and reciprocity, which I will consider later, to Forst this principle is not a construction of man (2007: 11, 84). All other rights and moral norms, despite the universality of their claims, are constructions of man, but they depend on the first principle of practical reason and its criteria (ibid.: 14). We have already seen that this comes to admitting that the status of this principle is transcendental and that it is inappropriate to speak of constructivism here. I think this can only be done if we mean that these principles (the first right and its criteria) are constructions made by reason itself. But this of course only means that they are ‘an integral part of’ or ‘inherent in’ reason. If we take constructivism in this large sense the whole tradition from Plato to Hegel would be constructivist. So, to use the term ‘construction’ is not really satisfactory to denote the first level of morality. It is even questionable to use this term for the level of rights that depend on or can be derived from the first right to justification because, although they come from the right to justification, their validity cannot be seen as a construction. To relate rights to a first right that has an evidence status cannot be considered to be a construction when the evidence of its validity is transmitted.
If reason is bound to justification I wonder whether it would not be more consequent to equate reason to dialogue. When we say that the first principle of reason is justification this seems to presuppose that reason is inherently intersubjective and dialogical. Forst shows that reason is always linked to a discursive context. Reason seems to be a kind of epiphenomenon of this context, in the sense that there is no reason beyond the actual discursive situations that take place in the lifeworld. But Forst discovered that the principles of reason are transcendental (although he does not use this word) and independent of these situations, so that he should have said not that reason is linked to discursive situations, but that reason in itself, taken in its transcendental meaning, is dialogical. If reason in this transcendental sense is inherently dialogical then the first right related to it would be the right to participate in dialogue. This right to be part of dialogue makes it understandable why there is also a subsequent right called the right to justification. Justification presupposes the possibility of involvement in dialogue, which means that in fact there is a more fundamental right and a deeper root of justice than the right to justification. Dialogue does not take place because there is a need of justification, but the other way round, there is a need of justification because reason is, in its own dynamic and inner life, dialogical.
This primacy of dialogue would fit better with Forst’s aim to give a rational basis of pluralism, because if dialogue is the core of reason then pluralism would be an a priori condition of reason. To safeguard pluralism it would then not be necessary, as seems to be the case now, to insist on the neutral or agnostic dimension of the procedural frame of moral discourse (2007: 18). Forst has to insist on this procedural aspect because in itself the right to justification does not imply any pluralism. Justification can be taken as a system of foundations constituting a whole with parts and subparts and with the rigidity of logic. Because we could imagine the existence of a closed state justifying its own decisions by decree, and because justification alone is no guarantee of openness, we should be more precise and acknowledge that the ultimate right of justice is real participation in discourse. This ultimate right is of course, like all other rights, a specific good, determining ‘good life’, so it would be erroneous to designate a theory of primordial rights, as Forst does, as being an agnostic theory involving neither goods nor values (ibid.). Dialogue, pluralism, justification, etc., in my view are primary goods and values. They are extremely valuable because they are universal.
The right to participate in community life through reason (dialogue) also implies that there is a higher identity belonging to us all and constituted by reason that does not demand assimilation but participation without any loss of individuality or particularity. Seen from this perspective reason is inextricably bound up with a transcendental community of dialogue striving for a common good and oriented towards the same centre. 6 Diversity and plurality are therefore as irrevocable and indissoluble as reason itself. In fact Forst’s critical theory of power, which he develops in the third part of his book on justification (2007: 291–381) and in his Kritik der Rechtfertigungsverhältnisse [Critique of the Conditions of Justification], supports this idea that it is this right to participate in dialogue that really constitutes the ultimate right.
2
Trying to answer the question of what reason – and especially practical reason – is, Forst refers not only to the right to justification but also to the main criteria involved in justification which according to him are generality and reciprocity. He introduced these criteria in discussions with Thomas Nagel and John Rawls in Contexts of Justice (1994: 65, 75). In order to claim objectivity or validity a justification must meet some requirements. What is valid must be in the first place generally acceptable. This applies to all claims of objectivity, whether descriptive or normative. 7 The criterion of reciprocity relates more specifically to normative claims. What is morally valid must not only be generally acceptable but also reciprocally imputable (ibid.: 66, 205; 2007: 34, 76). 8 Moral law not only demands from each rational being general acceptance, it also demands the acknowledgement that when this valid law applies to me it must also apply to you and others.
It is, I think, very questionable to treat the principle or postulate of ‘generality’ as a criterion for truth or validity. And the same applies for ‘reciprocity’. It is of course true that what is objectively valid must be valid for all rational beings who are also reasonable. Rational beings are those capable of understanding validity-claims; reasonable beings are those of good will disposed to accept a valid insight and act accordingly. In an ideal situation in which all rational beings are also reasonable an objectively valid content must be generally accepted. But it is accepted because it is valid and not the other way round: it does not become valid because it is generally accepted. This means that even in an ideal community of rational and reasonable beings, the authority determining the validity of a claim is not acceptance or agreement but the inner rational power of that claim. Agreement is a (possible) result of argumentation but not its authority-conferring element. It can be an explanandum, but it cannot be the explanans of validity.
Agreement can, however, be taken as an indication (which as such is always highly questionable) but not as a criterion for truth or rightness. Forst is not the first to identify generality with a general agreement that functions like a criterion. Habermas also interprets the Kantian postulate as a criterion. 9 While Habermas uses the word ‘criterion’ sporadically, Forst uses it very frequently. This suggests that in his theory the postulate of generality has become a cornerstone of reason. Being interpreted as a criterion, it is used as the central pragmatic tool to construct, to corroborate and also to dismantle specific contents of morality and ‘good life’ (2007: 14–15). Generality becomes a pivotal element of Forst’s theory but this element is corrupted because general agreement is taken as the authority-conferring element, confusing consensus with truth.
The criterion of reciprocity fully depends on that of generality because only when something is valid and generally acceptable or capable of being generally accepted, is it imputable to others. This means that reciprocity is just a specification of the pivotal criterion of generality. We have seen that a moral claim is not valid because it is generally acceptable, but that it deserves acceptance because it is valid. This general validity creates expectations among rational and reasonable beings towards each other. We expect the other to recognize a right that we consider to be valid and we also accord the other this right. So, reciprocity is an effect of the principle of generality. Reciprocity is based on a belief in the universality of a claim. Forst is right that only these general claims and not arbitrary ones can be expected to be reciprocal. It is indeed unjustified to project arbitrary personal expectations upon others (2007: 126). But this again shows that reciprocity just follows from the principle of generality. We may therefore conclude that if the principle of generality cannot be taken as a criterion of justice then neither can the principle of reciprocity. The authority determining the normative validity of a claim or right cannot be the fact of its being reciprocally imputable. What is valid in this sense must be generally acceptable and reciprocally imputable, but it is not because it is generally acceptable and reciprocally imputable that it is valid. So, generality and reciprocity should not be taken as criteria and this means that they cannot be used to deploy a critical theory of society. If we want to do justice to the principle of justification and to our right to justification we cannot use them as criteria.
Even if we were to accept these principles as criteria they would be too formal to be useful. Fascist ideas like the opinion that each individual should stick to his or her nation and that there should be no mingling between races, could be generally acceptable and reciprocally ascribable in a fascist world. And in this same world the humanist idea that a mingling of races should be possible would be neither generally acceptable nor reciprocally assignable. Of course, this would be different in a humanist world. 10 So, these principles are useful as criteria only if we already have an idea of what world or which moral system is valid. But in that case to use them as criteria would be begging the question.
Forst could counter this argument by saying that it presupposes group-specific indices, because fascists and humanists are groups, which means that the formulated argument is not a moral but an ethical argument (1994: 121; 2003: 683). However, the claims made in this reasoning are not in themselves group-specific, which of course does not constrain them to be acceptable only to certain groups. We will always find groups in favour of or against some moral principles. This is even the case for the universal right to justification; we can find groups (in past and present), or imagine some in the future, denying or accepting its validity, and this of course does not at all diminish its rightness.
If we take generality and reciprocity not as criteria but as formal principles of reason and, being principles of practical reason, as formal principles of morality and justice, then we cannot, I think, deny them a central importance for a theory of distribution. If a right is established to be valid then all rational beings of good will sharing this insight must activate this knowledge by transforming insight into action. That is why these principles should be central to a theory of distribution. A right that is considered to be valid must be distributed following the formal principles of justice, which indeed are generality and reciprocity. A fundamental right that is considered as such deserves to be generally accessible and reciprocally demandable. But these two principles should be joined by an indispensable third principle of just distribution. This is the formal principle of priority. Fundamental rights are different from specific rights because they are inherent in the nature of the instance holding these rights. A police officer, for instance, has specific rights due to his or her social function. However, the fundamental rights of this police officer are due to the fact that he or she is a human being. Fundamental rights are therefore generally and reciprocally assignable and exigible, and for the distribution of these rights the sequence of this assignation and exigency is very important: the assistance of people lacking basic rights, for example, must have priority against the assistance of people lacking other rights. This of course means that rights are hierarchically structured and follow a formal principle of priority. It would harm justice to overlook or deny this structure of priorities. Although Forst alludes several times to a hierarchy of rights (2007: 280–1; 2011: 134–55), the principle of priority plays no explicit role in his work.
We can say that Forst, although confusing postulates with criteria, has correctly identified two major formal principles of justice that need, however, to be complemented by the principle of priority. The triad ‘generality, priority and reciprocity’ could be considered to be a fundamental formal structure of a theory of justice.
3
The principle of justification presupposes the existence of a communicative situation in which the agents are interested in rational outcomes and also have good will that disposes them on a personal level to be reasonable. Because they are rational and reasonable they are autonomous; and because they are reasonable and open to each other’s arguments they are tolerant. Because they are all autonomous and tolerant they are all equal. And, having a sense of responsibility, they are all disposed to explain their points and to help others understand them.
Forst does not use the Habermasian ‘ideal speech situation’, nor the original designation ‘ideal community of communication’ used by Apel, because these terms are too metaphysical and too reminiscent of systems of idealism (2003: 645). But although he avoids these terms he does not avoid the concepts behind them. Albrecht Wellmer has argued that the ideal speech community is in fact a metaphysical concept presupposing not a real but a general (he even says ‘divine’) perspective. 11 The consensus reached in such a community is the agreement of all imaginable beings endowed with reason, existing everywhere and in all times. This is indeed quite a metaphysical concept, but of course this cannot be an argument against it because we should first make a distinction between acceptable (or necessary) and unacceptable (or arbitrary) metaphysical ideas. 12
The ideal speech situation is a necessary idea functioning even at the basis of the epistemological concept of consensus used in modern science. It is counterfactual because it can never be fully implemented in our finite societies, and at the same time it is metaphysical in the sense of presupposing a community of rational and reasonable beings transcending a specific space and time. The principle of justification presupposes the ideal of a community of rational and moral subjects who are always ideal citizens guided by the highest principles of citizenship and considering each other to have symmetrical, equal rights, to be free and autonomous and to be responsible for each other. Although this ideal community is counterfactual, it is nevertheless a necessary idea linked to the principle of justification.
The specific principles inherent in this ideal, such as the principles of freedom, equality, tolerance, solidarity, altruism, respect, participation, recognition, responsibility, etc. are in fact the substantial criteria that make a critical theory of power possible. So, it is not constructivism, but this transcendentalism that constitutes the basis of critical theory. The principles of the transcendental idea of the ideal speech community are implicitly used by Forst to compensate for the vacuity of his formal ‘criteria’ of generality and reciprocity, but he never explicitly refers to them as part of an ideal situation (1994: 194–5; 2007: 106; 2011: 33).
This becomes obvious when Forst states that his supposed ‘criteria’ (of generality and reciprocity) can be used for determining the morality of claims on which there is factual disagreement. In cases of disagreement it is possible, he says, to use the criterion of generality by placing it on a level in which an agreement ‘might have been possible’ (2007: 36). Generality now comes to mean a general agreement as far as it is thinkable for all concerned. Despite factual disagreements, a general agreement might be thinkable. I would say that this means that we have to be capable of abstracting from the real contexts of disagreement in order to situate ourselves in a virtual space in which a transcendental ideal community would accept a certain claim, even when this is not the case in the real world. Here it becomes clear that although Forst explicitly invokes the criterion of generality, he is in fact implicitly supporting this criterion with the transcendental idea of the ideal speech situation and the substantial principles related to it (equality, freedom, tolerance, solidarity, etc.). We may therefore say that for a critical theory this ideal world is in some ways more real than our actual world because we find here principles of justice, which, unlike the principles of generality and reciprocity, are not formal but substantial.
Although Forst distances himself from metaphysics and from the idea of an ideal speech community, this transcendental ‘ideal world’ is in reality the unavowed cornerstone of his critical theory. His aversion to metaphysics is understandable if it means distancing philosophy from undeliberated religious presuppositions, but it becomes irrational when it impedes reflection on our own system of ideas. But metaphysics reappears where it is displaced and so Forst winds up using ontological expressions like ‘the kingdom of ends’ or ‘the kingdom of grounds’ (2007: 38). Critical theory seems to presuppose a ‘kingdom that is not of this world’. This is of course the ideal world situation with the substantial principles related to it. But this is a critical point that could also be made against other representatives of Frankfurt philosophy. The whole dynamics of recognition developed by Axel Honneth in my view depends on this same transcendental idea of an ideal community. Using a term of Dietrich von Hildebrand we might say that Frankfurt philosophy presupposes a ‘metaphysics of community’. 13
We may conclude that Forst’s philosophy not only uses without acknowledgement the idea of an ultimate foundation elaborated by Apel, but also the concept of an ideal community of communication. It is not possible to analyse here all the differences between the ideal community of Apel and that of Habermas. However, one important distinction is that in Habermas’ work the transcendental argument leads to a formal or procedural demarcation of morality. The ideal speech situation offers no material insights into the contents of morality but provides a framework circumscribing moral discussions. 14 Conversely, for Apel the ‘ideal community of communication’ is a directive (regulative and constitutive) idea that can be used to structure the ‘real community of communication’. 15 In fact, Forst wants to provide a substantial theory of justice in this Apelian way (2007: 14). He tries to come to a fundamentum inconcussum (an unshakeable foundation) but he cannot decide to choose for the transcendental and metaphysical part related to ultimate foundation and the ideal speech community. Forst seems partially to move back to Apel without committing himself to his ideas, which, as soon as they are touched upon, provoke an approximation to Habermas again. His theory resembles very much a pendulum moving back and forth between these two philosophers.
This also explains why Forst’s language is often equivocal. Trying to present a substantial theory of justice reconstructing real and existing rights, he speaks of moral constructivism which denotes a human invention rather than a discovery of rights (2007: 14–15). This impression cannot be softened by declaring an invention to be more than an inventio ex nihilo. If there is a moral system behind the words and interpretations of those trying to discover it, then it is inadequate to speak of constructions. Like Habermas, Forst wants to avoid normative realism, because this would lead to a theory of moral, ethical and political blueprints. The necessary presuppositions of discourse offer a procedural framework for real discursive practices in which specific rights, values and political ideals would be produced or constructed, without offering a blueprint of how people must behave or how institutions must be structured. This formal position defended by Habermas is inherited by Forst, who calls this ‘discursive constructivism’ (ibid.: 17). But to Forst the procedural frame is constituted by principles (such as the principle of justification and the criteria of generality and reciprocity) that he does not want to reduce to formal structures, because, as he says, they have ‘substantial implications’ (ibid.: 107, 129, 184, 265). Forst clearly tries to overcome formalism, but without really succeeding in his attempt because the criteria of generality and reciprocity remain, as we have seen, formal.
IV Good will
The third question related to a theory of discursive reason that Forst considers is ‘Why follow reason?’ or ‘Why be reasonable?’ This is a fundamental question for practical reason because once we have acknowledged what is rational we need to know why this cognitive aspect should trigger an action in accordance with our insight. Translating this in terms of practical reason one might say: It is not enough to know what a moral principle is, we also need to want to act in conformity with it (2007: 31, 78). This is of course the point of Kant’s respect or reverence for the law [Achtung fürs Gesetz]. 16 It concerns the passage from cognition to volition.
Forst makes a distinction between normative grounds and motivational grounds, the former being intersubjective and capable of being shared while the latter are just personal (2007: 40). Like Kant, he defends the internalist position that a moral act is really or veritably moral when it is not motivated by external things like sanctions or rewards (ibid.: 42, 78). I think the passage from cognition to volition is also crucial to the understanding of Habermas’ idea of the ‘free power of the best argument’. This idea presupposes that cognitive insights by themselves enact motivational forces. Against the Humean internalist approach of Bernard Williams, which takes moral grounds to depend on personal motives, Forst in line with Habermas defends the idea that a real insight in the universality of (moral) grounds automatically brings about the will to act accordingly (ibid.: 45–9). 17 A necessary condition is that the insight is real and free. Cognition is not really true and moral acts are not really moral when they are not based on free will. If insights and moral acts are imposed then they are not veridical and therefore not personally true or moral. This shows that the discursive aspects of truth, rightness and veracity, as Habermas claimed, are in a certain way interdependent. 18
Although for Forst volition is a direct result of real and free insight into the rationality of a claim, there is still something that must be added to this insight. This something is ‘good will’. Forst starts from Apel who developed this point in his discursive theory following Kant (2007: 54). 19 Indeed, a reasoning can be very strong but in order to accept it and act accordingly there must be a willingness to do so. Only when there is good will are grounds really accepted and interiorized, and become motives. This is why Kant in his Groundwork of the Metaphysic of Morals underlines the concept of free will. Good willingness makes it possible for the imperatives of practical reason not to be perceived as external commandments. Kant puts this in religious terms saying that for the ‘holy and divine subject’ the moral imperative is necessarily identical to its own will [notwendig einstimmig]. 20
Does this mean that reason involves a psychological aspect that is external to reason itself? This has been, I think, the central point in modern philosophy from Arthur Schopenhauer through Friedrich Nietzsche to postmodernism. The will must precede rational argument which means that its first ground cannot be reason. Although the claim that the will is prior to rational argument is in itself rational (showing that reason is inescapable) there seems to be something non-rational determining whether our insights or moral acts are real and veridical. In a footnote to the work of Stefan Gosepath (2007: 56) and in some comments on the work of Ernst Tugendhat (1994: 377; 2007: 57), Forst tries to defend the priority of reason by showing that if we admit the priority of the will, which is something subjective, we would not be able to understand the universal and objective claim of the imperatives of morality (ibid.). 21 So, we can say that Forst defends an internalist position although not in the Humean sense in which the will is prior to reason but in a sense in which the motivation to act in correspondence to a cognitive content is inherent in the insight in that content.
This has some important consequences for Forst’s theory of reason. First of all, if we argue against Humeans, Schopenhauerians and Nietzscheans that the will is not prior to reason because the will is subjective and reason is general or objective, there is a danger of creating a gap between objective contents and subjective motivations. There would be a cleavage between the acknowledgement of a moral law and the motivation to act accordingly. This would end up in relativism because there would be nothing objective in a content to determine my motivation. But objective contents (insights) according to Forst must be able to determine my motivations. There must therefore be an original connection between epistemology (reasons) and psychology (motivations). To close the gap, a theory of reason must be conceived as belonging to something more general that also includes motivations. In a discussion with Henrich, Forst moves towards an extension of the term ‘reason’ to the larger concept of ‘consciousness’ [Bewusstsein]. 22 According to Henrich, Kant’s idea of a ‘fact of reason’ in morality means that an insight into moral law must be complemented by the will to act accordingly. This is the ‘Achtung fürs Gesetz’. But this means that practical reason is part of a larger concept which Forst, following Henrich, calls ‘moral consciousness’ [Bewusstsein des Sollens] and which also includes personal will (2007: 87). So, we may say that the first consequence of Forst’s anti-Humean position is that, motivated by reflections on practical reason, he tends to conceive a theory of reason to be part of a larger theory of ‘consciousness’. Re Contexts of Justice, in a discussion with Michael Theunissen, Forst had already proposed that the connection between reason and will creates ‘Geist’ (1996: 163). 23 But it is clear that Forst does not want to return to a Hegelean metaphysics; there can only be ‘Geist’ in particular individuals. But, as we said, reason is not particular which means that coupling ‘will’ and ‘reason’ on a particular basis might be insufficient to understand that moral motivation is in general inherent in reason.
A second consequence of Forst’s internalist position is that a real and free insight into moral law implies that there is something ‘out there’ which is the content of moral law and which is not a product or a construction of the subject. The cognitive insight is of course something happening in the mind but the content of that insight is not a product of my will. The content is internal but it is independent from the will and in that sense it is external. Forst distances himself from Platonic interpretations. Moral law cannot be ontologically interpreted as a reality. But we can ‘discern and recognize’ [erkennen und anerkennen] moral law as something objective (2007: 87). It is clear that only if there is real and free insight into the moral content can there be a will to act accordingly. The will can be bent towards the moral content, but this already shows that the content cannot be a product or a construct of the will. This has consequences for Forst’s own constructivism because moral law can never be a construct. It makes no sense to say that moral law is partly a human product, that its main principles (of justification, generality and reciprocity) are specifically not constructs whereas the remaining part of morality is. Forst’s internalism shows that he defends a kind of non-ontological moral realism. Although he wants to see this realism embedded into discursive constructivism or proceduralism, we have just seen that moral law cannot be defined in such terms.
A third consequence corresponds to the second one. Forst tries to separate moral realism from an ontological interpretation of the Good, because this would bring the metaphysical idea of divinity too near. Moral realism is a dimension in our minds. But moral law is not subjective. Of course, if my mind gets damaged or confused this does not affect the validity of moral law. But it then becomes almost untenable to say that this realism has no ontological consequences. Forst tries to interpret this reality or objectivity of morality as something inherent only in the discursive situation of humanity. Humanity, being a discursive community, creates moral law. But here we meet the problem we considered before. If we consider morality to depend on our decision to take part in dialogue, it would then be a product of the will, in this case a common will. But neither reason nor moral law can be the product of the will or any higher authority, as Forst himself admits, because this would ruin the idea of the autonomy of reason (2007: 56). This validity belongs to the autonomous realm of reason which has an objectivity of its own that as I said can be discovered but not constructed. Here also it becomes untenable to say that moral realism has no ontological consequences.
According to Forst, the fact that humanity is a discursive community makes it possible to say also that humanity is the last ontological instance to which the moral agent owes reverence. 24 I would concede that human discourse paves a way to reason, to an insight in the validity of moral law, but it is of course not humanity that confers this validity, so that the possibility of taking humanity as the ontological instance of moral law becomes questionable. Forst can only claim that humanity is the last instance to which we owe reverence, because he already uses terms like ‘humanity’ and ‘human reality’ in a normative sense. He certainly does not mean the bad side of humanity, but only humanity as far as it is a discursive community of peaceful understanding and tolerance. It is clear that he implicitly relates this human community to the principles characterizing the ideal speech situation.
So, in answer to the questions ‘What makes us reasonable?’, ‘What makes us respect moral law?’, Forst has humanity as a factual community: ‘The “last foundation” to which we can return, is … our practical world as a human reality … We ourselves are in the end the authority that demands this [justification] and there seems to be no other world of pure values existing backstage’ (2007: 98). 25 The authority of moral law comes from this (life)worldly human reality. Forst solemnly waves away any metaphysical or ontological interpretation of validity with the words: ‘The kingdom of ends belongs to this world’ (ibid.: 99). And in line with Christine Kosgaard he takes humanity to be the ‘cause of morality’ [Grund aller Normativität (ibid.: 90)]. But in fact it is not real humanity, but an idealized one, the virtual image of humanity implicit in the idea of an ideal speech community that determines the basis of moral law. We may therefore say that there is a kind of repressed idealism in Forst’s constructivism. Humanity as a normative term is what is to be explained (an explanandum) and therefore it cannot be used (as explanans) to explain the authority of moral law.
We can show with Apel that discourse is inescapable and that discourse implies certain principles related to an ideal speech situation. We could even call the community involved in such a situation an ‘ideal humanity’. Forst should perhaps have been more precise and should not have spoken of ‘worldly human reality’, but of an idealized humanity. But even then we would say that it is not this ‘ideal humanity’ that confers the authority on moral law, it is not this ‘ideal community’ that determines the validity of morality; this ideal human community is ideal precisely because it per definitionem always respects moral law. It does not create or construct anything. But in fact when it comes to understanding the authority of moral law Forst does not consider ‘ideal humanity’ at all. With Emmanuel Levinas (of whom he presents a very secularized version) he believes that moral law derives its authority from ‘the face’ or presence of the other (2007: 63, 95). 26 Forst thinks that he can use Levinas to understand where the authority of moral law comes from. But he takes Levinas’ ‘face of the other’ in an ametaphysical way so adherence to moral law cannot be understood. A confrontation with the face of the other alone, without any metaphysical depth, cannot be the cause of our willingness to accept moral law because it would not be possible to understand all the atrocities humans carry out on each other. We could combine discursive theory with phenomenology by saying that the face of the other is a kind of language making a certain claim. But then nothing has been won in order to understand where the authority of moral law comes from.
In the end it seems better to acknowledge that moral law takes its authority from itself, which means that it is a reality that cannot be reduced to humanity. Thus we end up with an ontological interpretation of Good. Insight into this reality is indeed not enough for us to act in compliance with it. There must be good willingness before the face of the other can affect us. Our willingness must supersede our insight and this seems only to be possible once we recognize ourselves in the moral norm that belongs to an independent realm. Religions have called such recognition ‘conversion’. In fact such recognition is an identification with something as belonging to the same source. To be convinced by a moral insight means to recognize reasons ‘we identify with, we consider to be ours or to belong to us’.
V Conclusion
To summarize, we can say that the work of Rainer Forst inconsequently moves back and forth between transcendentalism and constructivism. This ambiguity appears in all the main argumentative lines developed in his work. When trying to answer the question ‘Why reason?’ (§ 2) he tends towards a transcendental theory of ultimate foundation and towards the idea of an ultimate fact of reason. Instead of admitting that this theory amounts to transcendentalism, Forst declares that his method relates only to constructions embedded in lifeworld contexts. Answering the question ‘What is reason?’ (§ 3) Forst again clearly tends towards a transcendental theory of derivation of norms from first irreducible principles. But then he again turns back to constructivism or proceduralism denying the possibility of strong derivations because these are always embedded in particular contexts. He develops his right to justification as a fundamental right capable of giving a rational foundation to more specific rights, but then again he turns to constructivism and declares this right to be a mere procedural tool of lifeworld argumentation. And finally in his reflections on the good will necessary for rational action (§ 4) he first tends to determine the authority of moral law in a transcendental way, claiming a separate reality for morality, but then he again hangs this authority on the lifeworld existence of humanity. I have tried to show that all these ambivalences and this continuous back and forth between transcendentalism and proceduralism, substantialism and constructivism, constitute the major difficulty of Forst’s theory of justice. The origin of these ambivalences is Forst’s ambition to create a non-metaphysical, wholly secular theory of justice, which amounts to a position denying its own transcendentalism and metaphysical presuppositions.