Bristow, W. F. (2007): Hegel and the transformation of philosophical critique

Ch. 1 – Kant’s Subjective Idealism

The main point of the chapter is to show that Hegel’s critique of Kant’s philosophy as ‘subjective’ is well-founded. Specifically, Bristow wants to show that:

First, that the subjectivism of which Kant is guilty does not depend upon any of Kant’s arguments from the nature of Sensibility (e.g. that space and time are ideal) to the subjective ideality of the knowable – i.e. appearances.
Second, that Kant’s subjectivism is not of a Berkeleyan (i.e. Cartesian) variety
Third, that Kant’s subjectivism can be derived simply from his theory of apperception and its role in the transcendental deduction of the categories (19).
Fourth, that recent attempts by Guyer and Ameriks to defend Kant against Hegel’s criticism are unsuccessful.

Premise 1

The first claim is largely taken for granted. This seems ok, since if Bristow’s third point is successful it would end up that Kant’s subjectivism is overdetermined. It’s actually an interesting question though, whether or not Kant’s theory of Sensibility entails any sort of subjectivism in the sense which Hegel worries about. This is all the more important if Bristow is unsuccessful in his argument for his third point (as I think he is).

Premise 2

What of the second claim? Bristow is concerned to distinguish Hegel’s attribution of subjectivism from a reductive phenomenalist subjectivism where,

We can know only representations, which are, in turn, taken to be mental items or ‘ideas’ in Berkeley’s sense. The ‘objects’ of our knowledge, on this reading of Kant’s view, reduce to these mental items, to ideas in Berkeley’s sense (20).

It would be nice to know how exactly we should understand what Berkeleyan ideas are, but the presumption is that they are the kind of things whose esse is percipi. Presumably then we go with either ‘collection’ style idealism or idealism understood as a version of analytic/ontological phenomenalism.

It is notable that Bristow appeals to Allison’s two-standpoints view in order to avoid the worry about phenomenalism (21-2). So Kant’s claim that we know only appearances, or that objects are merely representations should be understood as a claim that comes from the transcendental standpoint rather than the empirical. Since Kant would only be guilty of espousing reductive phenomenalism if he was arguing from the empirical standpoint, the charge fails to meet its mark. As usual I find this move unconvincing, at least without a great deal more discussion of what such standpoint-dependent claims come to. Bristow sides with Allison in thinking that the transcendental standpoint merely makes a claim about the epistemic conditions on our cognition of objects. Interestingly, he also seems to want to deny or avoid the change in Allison’s view from the first edition of his book to the second (see n. 5, p. 22).

Suffice it to say that Bristow’s description of this alternative reading is woefully underdescribed. There is the gesture to the empirical/transcendental distinction, as well as other authors, such as Graham Bird, who have argued against a phenomenalist reading of Kant. But the details remain obscure. But we can assume that neither Bristow nor Hegel understand Kant as committed to phenomenalism. That assumption seems sufficient to move on to an examination of Bristow’s argument for his third claim.

Premise 3

Bristow argues in his third claim for the subjectivity derived from Kant’s theory of apperception and the role this notion plays in his transcendental deduction of the categories. The basic worry, as Hegel puts it in his Encyclopedia Logik is that,

the Kantian objectivity of thinking is itself again subjective to the extent that the thoughts [i.e. the categories], although universal and necessary determinations, are merely our thoughts, and separated from the thing as it is in itself by an insurmountable gulf (EL §41Z2).

Bristow takes this point to be central to Hegel’s criticism of Kant. The issue is not whether the categories are ones which are held by all thinking beings. It is not that sense of ‘merely our thoughts’ that Hegel is getting at, according to Bristow. Instead, the idea is that Kant makes a mistake in how he conceives of us a thinking beings and it is this mistake that entails the ‘unbridgeable gulf’ (23).

The worry then, is that since Kant’s ‘Copernican Revolution’ requires that we define the objects of knowledge in terms of the subjective conditions of our knowing any object, the ‘object’ that we thereby come to know in cognition is one which is merely ‘for’ the subject (27). This still leaves it unclear as to why subjective conditions on knowledge make Kant a ‘subjectivist’ in Hegel’s pejorative sense. The answer, as Bristow sees it, is that Hegel reads Kant’s transcendental deduction of the categories as entailing that,

the objectivity in our knowledge has its source ultimately in the synthetic unity of apperception, thus, in subjectivity (28).

What’s the argument?

Bristow focuses primarily on a few points in the early sections of the deduction, specifically §§15-17. He starts from Kant’s claim that the combination of representations that results from synthesis can result only from a subject’s “act of self-activity” (B130). The notion of self-activity is one Bristow finds extremely important. Bristow spends some time discussing the formalism or representational emptiness of Kant’s conception of the ‘I’ of apperception. I’m not sure why. It doesn’t seem directly relevant for his argument.

Bristow, from what I can understand, seems to be arguing for the following:

The unity of an object depends on the unity of apperception (28-31).
The unity of apperception is self-activity, i.e. the spontaneous combination of representations according a rule.
Hence the unity of an object depends on the legislation of a rule (33).
Since the ‘I’ of apperception is empty of representational content, and the representation of anything else depends upon the ‘I’ of apperception, it must be the case that the rules for such unity originate from the self’s own activity (34).

I have tried to incorporate in (4) Bristow’s discussion of the emptiness of the ‘I’. I’m not sure if this is right since I make it sound like there is an argument from elimination going on here but it’s less clear in Bristow’s discussion that that is what he’s doing. In any case the main idea is one which Bristow borrows from Kant’s practical writings. It is that the kind of normativity required for the unity of apperception, and thus for objectivity, must come from the subject himself rather than any external source (34). Hence the source of normativity is the autonomous subject.

Now, on the one hand this sounds like it has to be true. On the other, it seems deeply mysterious. Of course, Kant was never one to shy away from positing mysteries so it’s not clear that that is a reason to dismiss it. However, I take it that this interpretation of the transcendental deduction posits a much deeper primacy for the notion of autonomy (and thus a privileging of the practical) than one might otherwise be inclined to give. Hence, I wish that Bristow had unpacked his argument a bit more.

So, in short, the worry is that since objectivity depends essentially on the subject, all objects must be only objects ‘for-oneself’ and hence we never get out of the narrow circle of the subject to reality itself.

One might make two replies to this. First, so what? As long as we can interpret Kant as non-Cartesian/Berkeleyan, and thus as not committed to reductive phenomenalism, it is not clear that we lose all that much by being limited to the ‘subjective’ knowledge of objects-for-us. After all, why should it be the case that we can know reality in any other way?

But second, one might worry that Kant’s position is not dialectically sound (i.e. it ‘begs the question’) for the reason that it does seem to make it impossible that we could ever have fundamental metaphysical knowledge of reality. That is, the negative conclusion simply falls out of Kant’s cognitive theory rather than being argued for in an independent way. Since, for example, Leibniz’s cognitive theory entails the opposite conclusion concerning metaphysical knowledge it does not seem that Kant has really advanced the debate concerning the foundations of metaphysics. He has not argued for a position that ends the debate so much as simply provided an alternative position within that debate.

This kind of dialectical objection seems to me the most effective. However, it is not clear that it is as strong as is needed by Hegel. After all, even if Kant doesn’t give us an independent argument for accepting his picture of cognition it still might turn out to be the case that this picture is more compelling than Leibniz’s or Hume’s pictures of cognition (which I think it is). Once we admit this then the best Hegel can do is try and present an alternative theory of cognition that would allow knowledge of things in themselves, but without the bad cognitive theories of the rationalists or the empiricists. In this case the proof will be in the pudding.

Premise 4

In defense of Kant Guyer/Ameriks have argued that the categories are not in fact ‘subjective’ in this way (38). Only the schematized categories are subjective. The unschematized categories may be applied to things in themselves qua noumena. Hence, pace Hegel, there is no ‘insurmountable gulf’ between the subject and things in themselves.

Bristow’s first reply to this is to argue that Hegel’s worry does not rest upon any concern about how widely shared the categories might be, i.e. that they are only ‘for us’ rather than some other beings (39). Bristow makes some points about the dependent or mediated nature of conceptualization (40-1) which do not seem to me as to the point, at least not without an alternative picture of what conceptualization must be like. So the main question is whether the possible reach of the unschematized categories defeats Hegel’s worry that Kant’s cognitive picture mires us in subjectivity.

Bristow emphasizes those aspects of the Kantian texts that claim the categories are devoid of sense and meaning [Sinn und Bedeutung] without any intuition to provide cognitive content (44-6). Bristow then argues that since the categories are empty they cannot be used to appeal to the way things are in themselves. Hence, we cannot appeal to the unschematized categories to avoid Hegel’s objection.

Certainly Bristow is correct to say that application of the unschematized categories cannot give us knowledge of things in themselves, it cannot even give us knowledge that there is successful reference. However, in order for Bristow’s argument to be successful he needs a stronger claim – that we know that the categories do not apply to things in themselves. But how could we know that? Bristow wants to show that Kant’s cognitive theory forces us to the conclusion that the categories cannot apply to things in themselves, but this seems altogether too strong. Kant need only show that it is problematic whether the categories apply to things in themselves, which is equivalent to saying that we cannot know whether a logically possible thought picks out any real entity (either a substance or a property).

So Kant’s reply to the ‘insurmountable gulf’ objection should be that for all we know the unschematized categories do pick out features of things in themselves. However, since this bare epistemic possibility is not enough to ground that highest form of assent – Wissen – we must restrict our confidence to aspects of theory building that do admit of knowledge of reference, or equivalently, the demonstration of real possibility. Isn’t this enough to avoid the subjectivist worry?

This is all to say that the problem of intentionality is not a general problem for Kant. The categories have their content a priori, and we can think of what is logically possible via their deployment. Clearly noumena are logically possible. Hence we can include noumena amongst the ‘things’ we think about. The only aspect of intentionality which Kant was concerned with was knowledge of reference (or successful predication). In his terms this meant knowledge of objective validity. But we cannot know that the categories aren’t objectively valid for things in themselves, just as we cannot know that they are. Hence the best Hegel can do is say that on Kant’s theory it is epistemically possible that there is an ‘unbridgible gulf’. But this is a much weaker form of subjectivity than the one Bristow and Hegel attempt to pin on Kant.

So if Hegel is going to deny even this weak form subjectivity without simply reverting to dogmatic metaphysics he must give some sort of argument for how we can know that the categories aren’t objectively valid for things in themselves that doesn’t beg the question against Kant by simply replacing his cognitive theory with an alternative (as Kant did with the theories of his predecessors).

The initial sections of ch. 2 largely recapitulate the argument from the previous chapter. Bristow now makes explicit the idea that Kant’s critical project begs the question against metaphysics (50/ff/).

The basic idea is that the very notion of a critique presupposes “that the norms of judgment are already self-legislated, in contrast to being determined by the object(s) of pure reason’s inquiry” (52).

Hegel’s ‘suspicion’ is thus that,

…this prior, self-reflective inquiry already effectively establishes in its very stance of inquiry, the formal self-reflection as the highest condition on our knowledge of objects, thus precluding the possibility of metaphysical knowledge, understood as knowledge of the unconditioned (52).

Part of what Bristow then proceeds to do is find a reading of Kant’s texts that supports this. Effectively, this chapter attempts to provide at least some substantiation of the argument in the last chapter that the assumption, in the transcendental deduction of the categories, of epistemological norms originating in a subject’s self-activity (whatever exactly that is), entails subjectivism, and that this assumption is itself present in Kant’s very notion of a critique.