In face of the rapid development of neurosciences in the last couple of decades, the controversy over the freedom of human action (or “free will”) becomes more and more one of the most important philosophical problems of the present day, that is approached from a variety of perspectives and in context of various philosophical traditions. Here I would like to examine one project of superseding the opposition between freedom and nature, that was presented by Jürgen Habermas in several essays** over the last decade and that is strongly connected with the notion of detranscendentalization of Kant’s account of free moral action. As I will argue later, this project is interesting not only for its own sake, but also because by an examination of it we can make explicit some historical determinants of the controversy, without a critique of which, in my opinion, no plausible solution of the problem is possible at all.
In the aforementioned essays, Habermas employs the term “detranscendentalization” in at least two distinct, however related, senses. What is common to both of these usages is the object of this procedure, which is Kant’s “realm of intelligibility” or, to be more precise, the conditions of possibility of human cognition and action as presented by the author of the Critiques.
As is well known, those conditions are explicated by Kant in a transcendental framework that involves positing a conceptual space distinct form the realm of empirical nature. Only in this space the notions of objectivity and normativity, and others that depend on them, foreign to the causally determined and natural-law-bound world of phenomena, can have any meaning and any validity. In context of the notion of freedom (I put my stress on that topic throughout the paper, leaving the problem of cognition aside – however, both issues are tightly related) it means that free human action can be conceptualized only as determined by rational reasons and not by physical causes.
Habermas analyzes two ways in which the determination-by-reasons character of human action can be retained while at the same time the “realm of intelligibility” gets detranscendentalized away. One is Hegel’s social-constructivist account of reasons in the Objective Spirit part of his system and the other one is Adorno’s internalization of nature into the subject and construal of the “realm of necessity” as constituted by instrumental reason for the purposes of technical manipulability .
Without going into the details of those subtle conceptual operations, for which I have no space here, I would like to point to the most important features of the detranscendentalization of reason(s) as Habermas seems to conceive it.
(a) Freedom and normativity are extended from moral action to every action. It is essentially what makes actions what they are, i.e. something different than mere natural happenings, that they are determined by reasons, and therefore a subject can deliberate on and be held responsible for them.
(b) Reasons as such are dependent on social practices. It is the practice (or the “language game”) of giving and asking for reasons, that constitutes the proper context of the notions of freedom and responsibility.
(c) It is the natural being, a human, that is the subject of action, deliberation and responsibility, and not some noumenal self subsisting outside the empirical world. Therefore, the “dualism of language games” (or “epistemic dualism”), i.e. the dualism of the normative and the natural-scientific perspectives, dose not imply an ontological dualism – or a transcendental one.
(I omit here the question of the dualism of first-person and third-person perspectives, to which Habermas pays a lot of attention, since I find that the important weakness of his project lies elsewhere.)
The opposition between Kantian “two realms” being thus detranscendentalized, the problem is not yet solved, since we still face a dualism, even if not an ontological or transcendental one, and it can still be – convincingly – argued that in the natural world as natural sciences describe it there is no place for freedom construed along these lines. Thus, Habermas sets out to outline a conceptual framework in which both the notions of natural world and of freedom would be transformed in such a way that any tension between them would disappear.
This project can be conceived of as a further step along the way of detranscendtalizing reason(s). In “The Language Game of Responsible Agency and the Problem of the Free Will”, where it is most explicitly presented, Habermas construes the “space of reasons” (John McDowell) in a Wittgensteinian manner, focusing on the notions of rule-following and learning, i.e. rule-acquisition. The latter plays a crucial role here, as Habermas accepts “the meta-theoretical assumption: ‘our’ learning processes, which depend on the framework of sociocultural forms of life, are in a sense the continuation of prior ‘evolutionary learning processes’ that gave rise to ‘our’ forms of life.” The “continuation” of natural learning processes at the sociocultural level is to be understood in a non-reductionist, but nevertheless naturalistic way. That is, the dualism of language games is to be understood “as an emergent feature of cultural forms of life” , while the cultural forms of life themselves should be interpreted as essentially no different than any other – i.e. natural – forms.
Habermas admits that the notion of learning, if it is to play such a role, must remain a “top-down”, non-physicalistic concept, but he points out that the theory of evolution already has to employ concepts of that kind, such as “self-preservation”, “fitness” or “adaptation” . Therefore, the “radical detranscendentalization” of reason amounts to presenting its natural genealogy.
What Habermas more or less explicitly assumes in his project is that the self-knowledge of Reason as essentially engaged in and constituted by social practices that possess a fundamental normative dimension (as well as the notion of freedom as the capability of following reasons that are constituted by and in the same time constitutive of those practices) – this self-knowledge being a detranscendentalized legacy of Kant – is a product of natural evolution in the same way as language itself is, and that means in the same way as the crow’s use of tools or wolf’s hunting strategies are.
That very assumption I am going to contest, but first I want to stress the importance of Habermas’ project. Not only is it one of the most subtle and convincing attempts at solving the philosophical problem of naturalism. It is also symptomatic of a way of thinking that becomes increasingly popular within diverse philosophical traditions – another important example of it being John McDowell’s work on the concept of empirical experience in Mind and World.
This way of thinking, in all its various forms, stems from a meta-philosophical assumption that the (detranscendentalized) dualism of reason and nature that we inherited from Kant is untenable, but nevertheless it is based on a fundamentally correct insight into the essentially normative character of reason. Thus, we should rethink our notions of both nature and reason in such a way as to save the normativity, but explain away the dualism.
That, as I claim, is impossible. There is no way to escape falling into one of the two pitfalls on the way to realize this project. On the one hand, we risk to end up with an idealistic conception of nature as permeated by normativity itself, which both McDowell and Habermas explicitly want to avoid. On the other hand, if after all the necessary transformations our conception of nature is still naturalistic in the minimal sense required, there is no room in it for normativity – in the minimal sense required. Habermas is right that the synthetic theory of evolution already has to employ top-down, non-physicalistic concepts such as “self-preservation”, that do have an undeniably teleological and normative character. But there is still a long way from that kind of normativity – which is perfectly suitable to a third-person perspective of a natural scientist – to the normativity of concepts such as “responsibility”, “justice”, “freedom” or “reason” – which unavoidably involves a first-person perspective of a participant in social practices. His proposed account of the “sociocultural forms of life” is as naturalistic as many others that he explicitly rejects (and with good arguments). The only difference is that it is biologistic, and not physicalistic, but it does not make it any “weaker” or “softer”.
However, this way of thinking can be tempting because we tend to think that we face an alternative: either we retain normativity of reason(s) in some way or other, or else we have to acquiesce with a bald naturalism that will strip us of everything that we conceive of as essentially human. This alternative itself is the core of Kant’s legacy, detranscendentalized or not. Now, the proper question is whether we must face it, i.e. whether we have to conceive of what is essentially human in a way that involves a non-naturalizable notion of normativity. In other words, is the dualism of language games a necessary part of our self-knowledge?
Habermas clearly thinks it is. For him this dualism is a product of natural evolution of sociocultural forms of life, and therefore it is a necessary trait of human beings in the same sense as language is, or, for that matter, as bipedity is. Thus, all we need is a natural genealogy of this dualism of the same sort as a natural genealogy of language or bipedity.
It is essentially a metaphysical assumption, notwithstanding its evolutionary packaging – to claim that Kant has discovered a necessary (even if this is only an evolutionary necessity) truth about human reason that only has to be detranscendentalized.
Contrary to this, the dualism of language games and the notion of normativity can be construed as results of a historical development. Which does not mean that the “realm of history” is somehow independent of the space of natural evolution, but only that the biological nature of human beings as engaged in sociocultural forms of life can be described in a variety of historically contingent ways, Kant’s and Habermas’s being only one of them. The “dualism of language games” is not an evolutionary trait of human species but only a theoretical description of some evolutionary trait. (Taking a theoretical description for the thing itself is what I called “metaphysical” a few sentences before.)
If that is true, it means that the alternative of “retaining normativity or naturalizing humanity away” is perhaps not unavoidable. It means that we might be able to attain some other form of self-knowledge that would enable us to escape “the problem of naturalism” itself. But the way to that is not through meta-theoretical transformations of our notions of nature, reason and others of that sort, but through a historical critique of the intellectual development in modern philosophy that led Kant to conceptualize our place in nature in the way he did. We would need to examine what determined this development, what theoretical choices were made and why – and what alternative choices could have been made (Spinoza for instance was clearly aiming at a conception of reason that, we can now say, would not employ any notion of normativity). Instead of detranscendentalizing freedom, and reason, we should ask why it has been transcendentalized in the first place.
-----
* This was a paper proposal for some conference. It was rejected, though. But it is also a programmatic text, as the kind of historical critique that I call for at the end is just what I envisage to do in my dissertation (albeit with a broader justification than just an examination of Habermas's proposal) [?]. Hence, all critical comments are all the more welcome.
**My discussion concerns basically four essays: “Fredom and Determinism”, “’I myself am part of nature’ – Adorno on the Intrication of Reason in Nature”, both in Between Naturalism and Religion, Cambridge 2008; “The Language Game of Responsible Agency and the Problem of Free Will”, Philosophical Explorations, Vol. 10, No. 1, 2007; and “From Kant to Hegel and Back Again”, European Journal of Philosophy, Vol. 7, Issue 2, 1999.
