The Trendelenburg/Fischer dispute in the 1860s has a historical interest but its development into an acrimonious polemic considerably diminished its philosophical value.Footnote 1 In my accounts (Bird Reference Bird2006: ch. 9, 173–8; 2010: 486–99) I set aside those aspects of the disagreement to focus on one feature where it seemed to me that Fischer, in criticising Trendelenburg, made a valuable contribution to the philosophical understanding of Kant. His point was essentially to disambiguate, as Kant plainly wished to do, the deplorably unclear general philosophical distinction between what is ‘subjective’ (‘ideal’) and what is ‘objective’ (‘real’). Others of Fischer's views, as both Kanterian and I note,Footnote 2 were more problematic, and I shall not pursue them. Nor is it my wish to continue the polemical tone in the debate, or to pursue the wider literature cited in Kanterian's notes (especially nn. 12, 18), but only to indicate what seem to me to be serious errors in his objections to my account.
Kanterian's primary objections are that my account misrepresented Trendelenburg, that Kant had failed to consider the neglected alternative in the Aesthetic, and had not remedied that neglect later at B148–9 as I claimed. Two crucial grounds for the objections are first that my account commits an alleged logical error (Kanterian, p. 273), and that in any case a passage from Trendelenburg demonstrates that he distinguished types of ‘subjectivity’ and ‘objectivity’ just as Fischer (and I) did (p. 274–5). In the concluding pages Kanterian rehearses a traditionalistFootnote 3 account of Kant's transcendental idealism whose central feature is a commitment to the real existence of things in themselves as an antidote to a traditional and Cartesian philosophical scepticism about knowledge of reality (p. 279).
In section 1 I examine the central objection that I failed to recognize the character of Trendelenburg's neglected alternative; I consider first the Aesthetic (section 1.1) and then B 148-149 (section 1.2). In section 2 I examine A's two central objections first that the criticisms of ambiguity in Trendelenburg commit a logical fallacy (section 2.1) and second that in any case Trendelenburg acknowledged the ambiguity in a quoted passage (section 2.2). Finally in section 3 I say something of Kanterian's traditionalist account of transcendental idealism in opposition to my view that Kant's project is properly understood as a ‘descriptive metaphysics’. It is evidently not possible here to address that issue in the detail given throughout Bird Reference Bird2006, but I note how much more extensive that detail is than Kanterian recognizes, and reject his reading of two brief passages in the Preface and Amphiboly which I used (in Bird Reference Bird2010) to indicate that project of a descriptive metaphysics.
1. The Character of Trendelenburg's Neglected Alternative
Kanterian (and Trendelenburg) represent Kant's view in the Aesthetic as: since space and time are a priori, necessary and intuitive, they are ‘subjective’ and do not therefore belong to ‘objective’ things in themselves (p. 276–7). They belong exclusively, that is only, to appearances. But the suggestion is that this does not take into consideration a further option, which is that space and time belong both to (subjective) appearances and to (objective) things in themselves. Two questions plainly arise here. First is whether Trendelenburg's ‘subjective/objective’ distinction accurately represents Kant's position, and second whether Kant's inference from ‘belong to appearances’ to ‘do not belong to things in themselves’ is warranted. Trendelenburg believed that he represented Kant's position accurately but that Kant's inference was unwarranted. He further claimed that in drawing the unwarranted conclusion Kant had opened the way to a traditional scepticism in which we lack ‘objective’ knowledge of things in themselves (Abstract).
My counter claim is that Trendelenburg's unqualified distinction between what is subjective and what is objective (3A) does not accurately represent Kant's more complex division between what is transcendentally and what is empirically ‘ideal’ (subjective1, 2), and what is transcendentally and what is empirically ‘real’ (objective1, 2). I also hold that Kant's inference in the Aesthetic from the character of our experience of space and time to the conclusion that they do not belong to things in themselves is correct; and that our ignorance of things in themselves is not an invitation to scepticism about the (phenomenal) knowledge we do have. That knowledge is provided in what Kant calls an ‘empirical reality’ compatible with his ‘transcendental ideality’.Footnote 4
1.1. The Aesthetic
I outline relevant features of the Aesthetic not merely to repeat what I have previously said but to note what Kanterian seems not to have understood or acknowledged. In line with Kant's aim in the Aesthetic to ‘isolate’ a priori elements in our experience (B5–6), the detailed arguments demonstrate that space and time in our experience are both a priori and intuitive. Neither Trendelenburg nor Kanterian question those demonstrations. What they question is the conclusion from those demonstrations that space and time belong to appearances and not to things in themselves. Or rather they do not question that space and time in our experience belong to appearances but only that they do not (cannot) also belong to things in themselves. Yet both the a priori and the intuitive character of space and time in our experience conflict with their belonging to things in themselves.Footnote 5 For things in themselves are items wholly independent of our cognitive powers and with no necessary link to our intuition and sensibility. If space and time, as we understand and deploy them in perception, in science and in mathematics, are independent of our cognitive powers then any properties we ascribe to them will be a posteriori not a priori. By the same token whatever the properties of things in themselves may be they have no necessary link with our intuition; and yet it is our intuition which (according to the undisputed arguments in the Metaphysical Exposition) constitutes the spatio-temporal form of our experience. These features combine to show the dependence on our cognitive (sensory) powers of that spatio-temporal experience and which mark their characterization as (transcendentally but not empirically) ideal (or subjective). If, therefore, we ask the question: given the demonstrated character of space and time as we experience them can they (as such) belong to things in themselves?, the answer is patently ‘No’.
1.2. The Passage at B148–9
It might be said that there would be no need to cite another passage against Trendelenburg if it is evident that the disputed inference in the Aesthetic is correct. But the appeal to B148–9 is important because it reveals another, related, unclarity in Trendelenburg's position. I distinguish two often confused questions, one from the Aesthetic considered above and a different question from B148–9. It is in relation to these different questions that the later passage is significant.
In the Aesthetic, as I have represented it, Kant asks the question:
(1) If our experience of space and time is as the Metaphysical Expositions say, could that experience be ascribed to things in themselves?
The answer outlined above is: No; that experience has to be characterized as ‘appearance’ with the weak ‘mind-dependence’ associated with transcendental idealism and in contrast to the strong mind-dependence of empirical idealism.7
At B148–9 Kant puts a different question:
(2) Suppose that there are things in themselves beyond our cognitive experience (and therefore beyond our experience of space and time); could they have (some form of) spatio-temporality?
(1) and (2) are different questions, even though commentators (including Kanterian) do not always distinguish them. In particular Kant's negative answer to (1) does not provide any answer to (2). An unqualified negative answer to (2) might be given so long as there is a strict prohibition on any meaningful description of things in themselves. Kant comes close to such a position but does not hold that references to things in themselves are meaningless.Footnote 6 As I put it in Bird Reference Bird2006, for Kant reference to things in themselves has in our case a cognitive but not a semantic deficit, so that an affirmative answer to (2) would be feasible so long as the cognitive deficit attached to reference to things in themselves is retracted. But by this time in the first Critique Kant has made explicitly clear that we can have no knowledge without some reference to sensibility (B75–6). This, I believe, is why the issue emerges at this point in the Transcendental Deduction; it does not merely repeat the Aesthetic position in answer to (1), but raises and answers the quite different question (2). Even in the light of that general restriction of our knowledge Kant's position allows him to offer a qualified negative answer to (2).
The answer to (2) given at B148–9 is more complex and less decisive than that provided for (1).Footnote 7 This is bound to be so since Kant has now affirmed that we can have no knowledge of things in themselves, that is, knowledge of independent objects involving no appeal to sensibility. Kant has the same kind of difficulty here that he has in characterizing an ‘intellectual intuition’ (B68, 72, 158–9, etc.) or those ‘analogical’ references beyond our experience which he accepts (and thinks important) as on the borderlines of our knowledge and understanding (B310–11; Prol. §§57–9; 4: 350–62). Essentially the point is that, though we lack positive knowledge of purported things in themselves, nevertheless we can characterize them as at least not having the spatio-temporal features which we recognize in experience (B149). Such an alienFootnote 8 experience (as Kanterian rightly calls it, p. 270) cannot be quite like ours because it would have to be mediated by some (intellectual) form of intuition of which we have no more positive knowledge. Kanterian accepts this in allowing on Kant's behalf some ‘analytic’ features of things in themselves but not any more positive ‘synthetic’ features, but he has failed to see how such an account differs from the conclusion in the Aesthetic and does not, as he thinks, merely repeat it. The account, moreover, meets Trendelenburg's requirement that Kant should consider that space and time might belong both to appearances (in our experience) and to things in themselves (outside our experience).
Unsurprisingly, however, Kanterian does allow himself a formal criticism of Kant's position at B148–9 in a familiar dilemma. When Kant supports his qualified admission of those negative, analytic, claims about things in themselves by appealing to an inevitable difference in the required intuition (intellectual as opposed to sensory) Kanterian objects that this either retracts the claim that we know nothing of things in themselves or else disallows even the negative, analytic, claims. That strict formal view is, however, not consistent with his apparent willingness to admit some analytic claims about things in themselves (p. 270). If this is a genuine formal objection to Kant then it will apply also to Kanterian; but evidently there is some room for at least an analytic understanding of a (transcendental) ‘thing in itself’. Such things in themselves have to be understood, defined, in some way, and those definitions will ground just the analytic claims Kanterian (and Kant) are prepared to accept. That definition includes their independence from our cognitive powers and their inaccessibility to our forms of intuition. Those features are for Kant demonstrably so unlike our necessary cognitive powers as to warrant the qualified negative answer which Kant gives to question (2). Kanterian's dilemma is an empty formal gesture.
2. The Central Ambiguity in the ‘Subjective (Ideal)/Objective (Real)’ distinction
Kanterian has two central objections against the claim that Trendelenburg misrepresents Kant's complex ‘subjective/objective’ distinctions. First is the suggestion that to mark these ambiguities in Trendelenburg's account commits some logical error (section 2.1); second is the claim, supported by a quoted passage, that Trendelenburg recognized the same complex distinctions as Fischer and I (and Kant) (section 2.2). My response to these criticisms is first that the logical point is spurious, and second that the interpretation of the quotation is unclear and contentious. The passage shows only that Trendelenburg recognized some of the complex terminology from Kant. It does not show that his distinctions are exactly those advocated by Fischer, myself and Kant, and is even interpretable in a way which conflicts with Kanterian's account.
2.1. An Alleged Logical Error in the Charge of Ambiguity
The claimed point of logic arises from Fischer's primary complaint that Trendelenburg's arguments contain a significant ambiguity in his use of ‘subjective (ideal)’ and ‘objective (real)’. It is easy to see how this might affect the validity of an associated argument, for with the relevant ambiguities in premises and conclusion it may be interpretable as having true premises and a false conclusion and so judged invalid. Kanterian, however, understands such cases differently; he claims that such an argument with these significant ambiguities ‘is logically impeccable, valid if not sound as long as [its terms] are used unequivocally’ (p. 273). But it seems to me absurd to defend any such argument as ‘logically impeccable’ by saying that if its terms were not ambiguous it might be represented as a generally valid form. This is an appeal not to logic but to pseudo-logic. I am confident that Kanterian and I can agree on some appropriate descriptions for such faulty arguments containing significant ambiguities, but ‘logically impeccable’ will not be among them.
2.2. Trendelenburg's Claimed Acknowledgement of the Ambiguities
Kanterian moves towards a more promising point in claiming that Trendelenburg recognized the disputed ambiguity in Kant's four-fold account of empirical/transcendental ideality/reality, and expressed the distinctions in a quoted passage. That the passage is inadequate for this purpose can be shown summarily since the passage does not even refer to all four items in Kant's catalogue. The only two Kantian terms used are ‘empirical objectivity’ and ‘transcendental reality’. Kanterian suggests that we reserve ‘empirical subjectivity’ for Kant's ‘empirical idealism’, but neither of these terms, nor the term ‘transcendental ideality’, are mentioned in the passage.
The issue in any case is not simply whether Trendelenburg recognized that Kant has a four-fold division, but rather how he understands each term in Kant's classification. The question (made clear throughout Bird Reference Bird2010, especially 488–9, 491–3) is particularly how Trendelenburg understands the distinction between empirical and transcendental idealism and how it relates to their (weak or strong) forms of mind-dependence. But here the noted gaps, and other obscurities, in the passage prevent it from providing a clear answer. Consider, for example, two issues of interpretation. Kanterian suggests that we associate ‘exclusive subjectivity’ with transcendental idealism as a rejection of transcendental realism. But an empiricist might reject transcendental realism without subscribing to any form of idealism. Kant's rejection of transcendental realism clearly leads him to accept some form of idealism, and mind-dependence, but the passage offers no account of that commitment. Again, consider the suggestion that what Kant calls ‘empirical objectivity’ is necessarily ‘conditioned’ by the character of the space and time in which that empirical objectivity obtains. But that conditioning character in what James called ‘the great envelopes’ of space and time (James Reference James1912: 94; Reference James1918: 2, 135–6) is their ‘exclusive subjectivity’ which falls short of Trendelenburg's required anti-sceptical objectivity. It is in this way that Trendelenburg convinces himself that Kant cannot support the required objectivity and must give way to a Cartesian scepticism (p. 279). Construed in that way the passage might be said to reflect Trendelenburg's empirically idealist scepticism about our objective (even empirical) knowledge.Footnote 9
3. The Wider Context of Transcendental Idealism
That reference to what, according to Trendelenburg, Kant would need to establish against an idealist scepticism (knowledge of real things in themselves) points immediately to the considerable question of our, and Kant's, understanding of transcendental idealism. Because it is so wide-ranging neither Kanterian nor I can seriously deal with it in this context. He replaces a fuller discussion with two limited objections to my treatment of Kant's project as a ‘descriptive metaphysics’. For in Bird Reference Bird2010 I had pointed to two passages in support of that view, namely the Preface ‘inventory’ metaphor and the account of transcendental reflection and its outcome in a transcendental topic in the Amphiboly. Because Kanterian inevitably says little of the wider evidence cited in favour of Kant's descriptive metaphysics in Bird Reference Bird2006,Footnote 10 his general (traditionalist) account of transcendental idealism functions in this context only as an expression of an approach in conflict with mine rather than as any evidence in its favour. Nevertheless Kanterian provides vigorous objections to my account of those two passages from the Preface and the Amphiboly and here it is feasible to show that his objections are misconceived.
3.1. The Preface ‘Inventory’ Passage
Kant, like Leibniz, calls for a proper ‘inventory’ of our knowledge/experience (Bird Reference Bird2006: ch. 2, p. 28) and in a Preface (Axx–xxii) proposes that his new ‘metaphysics’ should follow such a plan. Kanterian's objection to this is that in the passage Kant is connecting that project not to the propaedeutic Critique but only to the later applied metaphysics in, for example, the Metaphysical Foundations of Natural Science. To read the passage in that way seems to me quite unjustified; it is to strain too hard to press a text into a preconceived interpretation. The metaphysics which embodies the inventory is specifically associated in the passage with the Critique, for the ‘complete’ system which Kant envisages comprises both the complete principles of the system (outlined in the Critique) and what he calls the ‘derivative, richer’ detail of the later work. There is nothing in that contrast, or in the associated difference between approaching the work in its two parts as a ‘judge’ or as a ‘fellow worker’, which restricts the inventory metaphor to the later, derivative project. The natural message is that the metaphor applies to both projects; or, more accurately, to the two parts (general and particular) of the one central, revolutionary, project.
Kanterian adds the surprising objection that the project of a descriptive metaphysics ‘does not accord with the method of the Critique’ because the latter involves ‘direct and indirect proofs, refutations, deductions, etc.’ (p. 280). Why it should be assumed that a descriptive metaphysics cannot contain such material is not explained. Should we expect Strawson's descriptive metaphysics in Individuals to contain no proofs, refutations or deductions?Footnote 11 Should we suppose that geographers, cartographers and navigators do not carry out calculations, engage in argument and deduce or refute associated views? Of course Kant's (and Leibniz's) metaphysical inventory is not exactly like a simple list of items on the supermarket shelves. Kant has to argue, as he evidently does, for the proposed a priori structure of experience and the general principles which embody it.
The essential point in the project of a descriptive metaphysics, which Kanterian misses, is that Kant's revolutionary aim is to begin with an accurate account of our experience before engaging in the standard philosophical arguments about it; we are to see how things in experience are (what experience contains, how it is articulated) before we plunge directly into the received (and endless, undecidable) philosophical issues over the justification for any such items in experienceFootnote 12 (Prol. 4: 255; Critique, Preface Aviii–x). A good example of the benefits of such a revolutionary approach is the demonstration in the Refutation of Idealism that a traditional idealist scepticism was based erroneously on a mistaken order of priorities between inner and outer experience. In that case, as in others, the upshot is to question and perhaps reject a traditional, often idle, scepticism associated with an unthinking ‘school metaphysics’. But the refutation of a sceptical position here rests on a careful examination of the actual relationships between items in a given experience.
3.2. The Amphiboly's Transcendental Topic
In the Note to the Amphiboly (B324) Kant offers a general characterization of ‘a transcendental topic’ as one in which we decide ‘the place which belongs to every concept according to the difference in the use to which it is put and the directions for determining this place’. On the following page he fits this project more specifically into the (preceding and succeeding) criticism of Leibniz and appeals again more generally to ‘a transcendental reflection’ as a procedure for determining the ‘location of the representations of the things compared, … whether thought by pure understanding or given in appearance by sensibility’. Two objections are made by Kanterian to the suggestion that Kant's project in the Critique can be understood as providing a transcendental topic based on a transcendental reflection, of which the criticism of Leibniz is a specific example. First these general notions are ‘not congruent with the Critical project or even the Doctrine of Elements’; second that the Amphiboly procedure remains open to the traditionalist criticism that Kant regards ‘thinking a thing in itself’ as necessary despite the fact that categories are inapplicable to it. I say nothing here of the second complaint which is a familiar but more extensive puzzle in Kant and which is answered in Bird Reference Bird2006: chs 22–23.
As in 3.1 it seems to me short-sighted and unjustified to impose Kanterian's restrictions on the scope of Kant's terminology; in this case to restrict a transcendental topic and transcendental reflection merely to the specific criticisms of Leibniz in the Amphiboly. Kant makes very clear how this procedure prepares the way for all the diagnostic, ‘ground-clearing’ activities in the following Dialectic. The procedure incorporates a diagnosis of philosophical error which goes beyond the case of Leibniz, and is, apparently in conflict with what Kanterian says, quite evidently asssociated with the Doctrine of Elements in its essential appeal to what belongs strictly to the senses and to the understanding. Even if the transcendental topic were confined to the four heads of comparison they are in any case so wide-ranging (covering the ‘inner/outer’, the ‘matter/form’, and ‘appearance/thing in itself’ distinctions) as to involve a great deal of what Kant outlines in the Aesthetic, Analytic and Dialectic.
Kanterian endorses Trendelenburg's belief that Kant has failed to resolve a scepticism about knowledge of things in themselves, even though it is evident that Kant regards such a philosophical aim as hopelessly misguided. This marks a fundamental disagreement between Kanterian and myself over the importance and value of philosophical scepticism which cannot be resolved here. Kanterian asks: ‘What sceptic will be silenced by the assertion that experience is not offered a justification but taken as a datum?’ But this disregards the possibility that Kant's project does not even aim to silence such a sceptic. It might also be said to be precisely because nothing can silence the sceptic that it is better simply to disregard that fruitless argument and turn to something more rewarding. It is difficult not to see that attitude echoed in Kant's remarks about the endless and futile oscillation from scepticism to dogmatism and back again (Aviii–x). Kant's aim is not to join in that philosophical roundabout but to find a more rewarding method of bypassing it. In assuming, with Trendelenburg, that the primary goal must be to rebut, or at least address, such philosophical scepticism Kanterian begs the question against a Kant who endorses the project of a descriptive metaphysics. He has apparently not understood that, so construed, Kant repudiates that assumption.
