Paul Franks at the Fichte workshop

Paul Franks presented a fascinating paper at the Fichte workshop entitled ‘Intellectual Intuition, Geometry, and Fichte’s response to Kant’s Naturalistic Challenge’. In the paper Franks set up Fichte’s particular brand of transcendental philosophy in terms of his response to what he called Maimon’s naturalistic challenge. The challenge demands of the transcendental philosopher a reason for preferring his account of the mechanism of causal reasoning over a naturalistic one, if both accounts can adequately explain such reasoning. For a Humean style naturalist will assert that causal inferences from observed to unobserved cases can be explained by mechanisms such as habituation and projection. Furthermore, such a naturalist will agree with Kant that we can’t but help make such inferences. What this naturalist will deny however, is that we have any rational justification for making such inferences. The practice of making such inferences merely reflects our (natural) subjective propensities.

According to Franks, Kant sees this lack of justification as intolerable because it leads to a revision in our epistemic practices – that is, it leads to the rejection on philosophical grounds of the validity of causal reasoning. Given that transcendental philosophy can justify causal reasoning by recourse to synthetic a priori principles, according to the principle of conservatism, we should prefer the transcendental story to the naturalistic one. It is important to point up however, what Franks called the ‘reciprocity claim of transcendental philosophy’. This is the claim that the constitutive principles that make possible experience are justified partly because they are actual in experience – the principles are upheld by the facts, so to speak. So the principle of causality underwrites inferences from observed to unobserved facts, but it is also justified by the actuality of what it makes possible.

Franks pointed out that Maimon rejects Kant’s assertion that naturalistic explanation leads to a revision of our epistemic practices. This is because Maimon sees that the naturalist can accept the first half of the reciprocity claim of transcendental philosophy concerning causal reasoning – the constitutive principle – but deny the actuality of the experience of the principle in the world. So the naturalist can hold what we would call today an ‘error theory’ of justification: we cannot help making causal inferences and believing that they are justified by the principle of causality. But although we cannot revise this belief, it is nonetheless false – there is no experience of the causal principle and thus no knowledge, according to the naturalist, of it.

All this means that transcendental philosophy has now lost the philosophical advantage it had over naturalism. For the naturalist need not think that any revision of our epistemic practices is necessitated by the lack of an actual experience of the principle of causality as long as she holds an error theory of its justification. This leaves the naturalist and transcendental philosopher in a stand-off. According to Franks, Fichte saw a way of gaining the upper hand for transcendental philosophy by proving that experience of constitutive principles is actual rather than mere projection. It is in order to make such a proof that Fichte developed his method of construction in intellectual intuition.

Fichte’s idea is that if an analogue of geometrical construction could be developed for the (categorial) principles of transcendental philosophy, then this would demonstrate that it was not possible to take an anti-realist stance towards experience of those principles. This is because we are given the constitutive principles of geometry via the actual iterable constructions in space of particular shapes and figures. Thus necessarily there is no gap between the principles and their application in experience. So although one might still be a semantic anti-realist about the principles of geometry in that one might think they are only made actual as relationally determined constructions in the medium of pure intuition (i.e. one need not think anything beyond those constructions make sentences stating the principles true), one need not be an epistemic anti-realist (error theorist) about our justification of the principles in question.

However, as Franks pointed out, Kant had ruled out any actual a priori intuition of an object as determined by the categories, because the categories only give the rule for the synthesis of what perception may yield a posteriori. So for Kant, there is no real object intuited by applying the categories a priori as there is in geometrical construction: the only object we have knowledge of a priori via the categories is an indeterminate ‘concept of an object in general’. But Fichte’s conception of a priori intellectual intuition does not concern empirical objects, but, as Franks put it: “…an intuition of myself as the subject of purely formal acts.” However, for this conception to work, Fichte will have to articulate a well-defined iterable intuitive procedure which would mirror the basic operations of Euclidean geometry. This Fichte attempts with his ‘principle of Determinability’ which he applies to our (apparent) non-inferential awareness of our conscious acts as our own. I will not go into the details of this procedure, but the end-product is supposed to be a proof that the transcendental principles that govern possible experience can be actualised in experience. Such an actualisation underwrites the transcendental philosopher’s rejection of naturalistically motivated epistemic anti-realism about such principles (although the transcendental philosopher himself accepts semantic anti-realism concerning these principles).

Franks argued that even if this kind of iterable intellectual intuition were clearly defined and demonstrated to be possible, it would not answer Maimon’s challenge. This is because the challenge is not based on some worry about the efficacy of transcendental method to attain truth by being actualized. Rather, the challenge concerns the issue of whether the reasons we give ourselves in philosophy or everyday life actually determine our epistemic practices. For the naturalist, by definition, holds that only the methods of the natural sciences are proven to determine such practices. In other words, the naturalist can accept Fichte’s intellectual intuition as an iterable method of construction, whilst at the same time denying that such a method determines in any way our epistemic practices. So the stand-off simply resumes. And the naturalist may even have the upper hand, because she can argue that her methods are proven to determine many aspects of our epistemic practices, thus she can argue that at least she is a methodological conservative.

In general, I thought that Franks’ paper showed us how transcendental philosophy and naturalism are locked into a dispute that Fichte’s development of transcendental method could not hope to settle. This is because both approaches to philosophical knowledge rest on mutually exclusive assumptions: that our epistemic practices require rational justification in the one case, and that the only acceptable methods of accruing knowledge are the methods of natural science in the other. Franks’ analysis of Fichte’s conception of intellectual intuition brought out specifically how the reciprocity claim of transcendental philosophy gains it no advantage. For example, Hume would argue that all we are given in experience are sequences of individual sensory impressions that we learn to associate as instantiating causal laws though constant conjunction and a habitual inference to unobserved cases. So the principle of causality is actualised in experience, it is just not actualised by means of synthetic a priori rules of synthesis that might also act as rational justification. On his account of how the principle is actualised then, Hume is perfectly entitled to his ‘error theory’ concerning such justification. And, given Hume’s viable account of how the principle is actualised, the burden of proof is on the transcendental philosopher to demonstrate why it can only be actualised according to her story. Appeal at this point to the fact that the latter secures rational justification will simply be question-begging against the naturalist, who denies we need such justification to validate the epistemic practices in question. In short then, Franks’ paper brought us to see how the deep stand-off between transcendental philosophy and naturalism was played out in the short period after Kant initiated the transcendental turn, and how Fichte’s attempt to tip the scales the way of transcendental philosophy failed.