Alternative to Ontology or Alternative Ontology? Allison on the ‘Proud Name’ of Ontology (A247/B303)

Abstract

This paper argues that, far from supporting, an oft-cited passage of the Phenomena and Noumena chapter (A247/B303) instead belies, Allison’s influential thesis that Kant’s transcendental idealism is not an ‘alternative ontology’ but a methodological or meta-epistemological ‘alternative to ontology’ that is devoid of specific metaphysical content. Following a programmatic sketch of Kant’s system of principles as a regional ontology of nature, it is argued that the precise wording and original punctuation of that passage suggest that the transcendentally realist ontologies of the past are to ‘give way’ to just such an immanent ontology of the world of outer experience.

‘Metaphysics invariably buries its undertakers’. ¹

1. Introduction

Drawing on the concept of a regional ontology of nature developed in section 2, section 3 of this paper proposes a detailed, phrase-by-phrase re-examination of a brief but important passage repeatedly cited by Henry Allison (cf. 1973: 50, 2004: 120, and 2006b: 7–8) as collateral evidence for his emblematic thesis that Kant’s ‘transcendental idealism is best viewed as an alternative to ontology rather than, as it usually is, as an alternative ontology’ (2006a: 123). ² Guyer and Wood quote the same passage from the Phenomena and Noumena chapter in their co-authored Introduction to the Cambridge Edition translation of the first Critique as proof of ‘Kant’s characteristic tendency to convert ontological … into epistemological questions – that is, … questions about what sorts of thing there must be into questions about the conditions under which it is possible for us to make claims to knowledge about things’. ³ Yet far from dismissing ontology outright, that short passage in fact only rejects ‘the proud name’ (e.a.) ‘ontology’ together with the dogmatic sub-discipline of traditional, transcendent metaphysics to which it had become indissolubly linked. But since it concludes by proposing ‘a mere analytic of the pure understanding’ as the new name for that transcendental discipline which is to supplant the dogmatic transcendental realist ontologies of the past, and since the new name naturally evokes the titles of leading anti-metaphysical works of seventeenth- and eighteenth-century epistemology, notably Locke’s Essay and Hume’s Enquiry (both concerning Human Understanding), it is little wonder that so many commentators regard Kant as having replaced ontological with epistemological enquiries. Nor is it surprising that Allison should invoke the ‘proud name’ passage in support of the thesis mentioned above; for ‘an alternative to ontology’ aptly summarises his radical re-interpretation of Kant’s transcendental idealism as a methodological or meta-epistemological posture devoid of any specific metaphysical or ontological doctrine. Section 4 states (provocatively, yet without, I hope, overstating) the diametrically opposed conclusion reached in section 3. ⁴

Despite the similarity of their views on Kant’s substitution of epistemological for ontological questions, Allison parts company with Guyer when it comes to the ‘transcendental distinction’ (A45/B62) between appearances (phenomena) and things in themselves (noumena). On what has been (rather unhappily) labelled his ‘two-aspect’ view, that distinction embodies a pair of mutually exclusive and jointly exhaustive ‘methodological’ or ‘meta-philosophical’ alternatives facing the epistemologist: either transcendental realism, according to which spatio-temporal predicates apply to all ‘things in general and in themselves’ (B410; cf. also A238/B298), or transcendental idealism, which restricts the legitimate use of all such predicates to appearances alone (cf. e.g. Allison 2006b: 12). This non-metaphysical (which is not to be confused with the more recent ontological) two-aspect approach to what may be termed Kant’s noumenalism is no doubt a salutary corrective to the disastrous two-object or two-world revisionary metaphysics long foisted on Kant by unsympathetic critics; and the same holds for the no less unsympathetic one-world metaphysical interpretations of Guyer and Strawson. ⁵

Apart from the desire to rescue Kant’s noumenalism from patent absurdity, Allison’s non- (indeed anti-) ontological interpretation has a further source as well: those long dominant and even today still not entirely abandoned efforts to cast Kant’s transcendental idealism as an unacknowledged Berkeleyan or other (see Van Cleve 1999) form of empirical idealism or phenomenalism (cf. Allison 1983: 30–1). The prevalence of these ontological approaches, the former dating at least from Jacobi’s famous cri du coeur, the latter from the notorious Garve-Feder review, fully warrants Allison’s claim that transcendental idealism has ‘usually’ been interpreted as an ‘alternative ontology’. Indeed, what he dubs ‘the standard picture’ (1983: 3) combines some form of Berkeleyan (or other phenomenalistic) theory regarding the objects of experience with a very un-Berkeleyan doctrine regarding things in themselves. ⁶

Unfortunately, the glaring inadequacies of ‘the standard picture’ blind Allison to that immanent ontology which Kant set out to establish as the apodictic foundation of a future complete metaphysics of nature consisting of analytic as well as synthetic a priori cognitions and of derivative as well as primitive synthetic a priori principles. ⁷ By dint of insisting, rightly, that these ontological interpretations diminish Kant’s extraordinary achievement, he manoeuvres himself into the position of denying that the Critique contains any positive metaphysical or ontological doctrine at all, even just regarding phenomena or appearances; its positive teaching is rather ‘a transcendental analysis of objectivity’ (Allison 1983: 227), that is, a meta-epistemological theory of what it is for our representations of appearances to be objectively valid cognitions – and in that precise sense an ‘alternative to ontology’.

The next section contains a programmatic sketch of Kant’s ‘System of All Pure Principles of the Understanding’ (A148/B187–A235/B294) as a regional ontology of nature. While mine differs markedly from the ontological interpretations that are ‘usually’ given, Allison would doubtless reject it too; for his are principled objections to any ontological interpretation (see nn. 4 and 16). Alternatively, he might grant mine a certain legitimacy as being in no more than verbal disagreement with his ‘alternative to ontology’ thesis; for others (he might say) are perfectly entitled to call a certain set of principles a regional ontology in something like Husserl’s or Heidegger’s sense (see n. 15), even if, historically speaking, they are in fact principles of knowing expressly intended to supplant those principles of being which they reproduce, at times, almost verbatim. ⁸ To the retort that Kant himself applies the term ‘ontology’ to his transcendental philosophy, not just in the Critique (cf. A845–6/B873–4), but especially in the metaphysics lectures of the 1770s (see n. 19), Allison would no doubt reply that it is an ontology in name only, since it in fact redeploys the concepts and principles of traditional ontology in ‘a whole new game’ (see n. 2). While it would be tempting to turn this hypothetical rejoinder around and insist that Kant’s transcendental idealism is an ‘analytic of the pure understanding’ in name only, and a regional ontology of nature in fact, that would be going too far. Instead, section 3 will argue that Kant’s transcendental idealism is indeed a meta-epistemological (transcendental) ‘analytic of the pure understanding’ as regards its revolutionary method of demonstrating a certain set of principles; with respect to what it seeks to demonstrate, however, it is a doctrine of being, that is, a system of long-familiar, but never before demonstrated first truths about the existence and essence of all actual and merely possible empirically real beings. That this way of defining ‘ontology’ is (despite a certain rupture) still largely continuous with the older and oldest ontological tradition will appear from section 2.

2. Kant’s system of principles as a regional ontology of nature

To see that non-ontological interpretations of transcendental idealism are not just not supported, but are in fact belied, by the ‘proud name’ passage, it is necessary to consider briefly what ontology was for Kant’s eighteenth-century predecessors and how Kant transformed it in passing from his early, dogmatic to his mature, Critical standpoint. The transformation concerns neither (a) the foundational role of ontological first principles in relation to the other branches of metaphysics and to those empirical sciences whose certainty depends on them; nor yet (b) the specific content of Kant’s ontological first principles, most of which are similar, if not identical, to principles his predecessors already regarded as apodictically certain (see n. 25); the rupture concerns rather (c) the domain of application of ontological first principles. ⁹

Among the ontological first principles of Kant’s Leibniz–Wolffian predecessors were many non-trivial – in the later terms of the Critique, ‘amplifying (erweiternde)’ (A765/B793; cf. B11) – necessary and universal judgements (truths) about the possible, actual, or necessary existence and about the essence of all ‘things in general and in themselves’ (B410). ¹⁰ ‘Existence’ and ‘essence’ being, since mediaeval times, the two principal senses of the word ‘being’, such unrestrictedly universal, non-formal principles regarding all logically possible things are eo ipso ontological principles in the literal sense: a doctrine of being. That this scholastic existence–essence distinction is fully preserved in eighteenth-century German metaphysics is clear from the distinction Baumgarten draws (cf. his Metaphysica §311) between principia fiendi and principia essendi (compositionis). The latter are principia possibilitas, that is, the first principles of what a thing or res is (its essence or nature, regardless of whether it exists), whereas the ‘caussa’ [sic] of a logically possible thing’s actual existence (generationis) is its principium fiendi. ¹¹

In accordance with this inherited idea of a general metaphysics or ontology, the relational and modal (the so-called ‘dynamical’) principles set out in Kant’s ‘System of all principles of pure understanding’ govern the existence, while the predicates of the principles ranged under the headings ‘quantity’ and ‘quality’ (Kant’s ‘mathematical’ principles) concern those properties that pertain to the essence of all things. Yet for the mature Kant, unlike his Leibniz–Wolffian predecessors, the things in question are only empirically real or really possible objects of human experience, that is, such things as are or can be given to the senses and subsumed under its categories by the pure understanding aided by the transcendental imagination. ¹² Kant’s dynamical and mathematical principles may therefore be called (adopting Baumgarten’s idiom) rationes or principia fiendi (principles of being understood as becoming or coming-into-existence) and principia essendi (compositionis) (principles of being as nature or essence), respectively. But given their restricted domain of application, the complete ‘system’ of all such principles constitutes a this-worldly or, in Kant’s technical vocabulary, an ‘immanent’, though still general, metaphysics. ¹³

To the extent that the individual principles are principles of being (in one or the other of its two traditional senses), Kant’s ‘Systematic representation of all synthetic principles of pure understanding’ (A158/B197) may be considered a regional ontology of nature, where ‘nature’ designates the infinite class of all actual and possible objects of the empirical natural sciences as well as of ordinary pre- and extra-scientific sensible cognition or experience. Admittedly, ‘regional ontology’ would have seemed an oxymoron in Kant’s day, ontology being a perfectly general branch of metaphysics in contradistinction to those special branches (rational psychology, cosmology, and theology) that contain the most general principles governing some particular object-domain or region. ¹⁴ But what might have seemed a contradictio in adiecto to Kant need not be so regarded today, for just as ontology was then called general metaphysics, so the general first principles of the branches of special metaphysics can be considered regional ontologies. For the interpretation of the passage at A247/B303, it hardly matters whether Kant himself would have countenanced the expression ‘regional ontology’ for what he calls a ‘Rational Physiology’ of corporeal nature or physica rationalis (A846–7/B874–5); the only question is whether what he there describes, now as ‘principles of the exposition of appearances’, now as ‘an analytic of the pure understanding’, declining to apply the name ‘ontology’ (burdened, as it is, with transcendental realist connotations), is in fact a doctrine of the being (existence and essence) of empirically real beings or objects in general (überhaupt), and hence ‘ontological’ in the literal sense of the term. ¹⁵

The restricted domain of application of that ‘System of the principles of pure understanding’ in which the Transcendental Analytic culminates points to the deeper reason why Allison cannot recognise those principles as ontological; for his ‘signature’ juxtaposition of ‘epistemic’ and ‘ontological conditions’ takes the latter to be transcendentally real conditions of things as they are in themselves, his paradigm of ‘ontological’ (or, as he also says, ‘ontic’) conditions being the way noumenal entities like Newtonian absolute space and time make possible other noumenal entities, namely, those finite things (in themselves) that exist in space and time (cf. 1983: 11–12; also 2004: 11, where the same Newtonian paradigm recurs). Allison, in short, understands ‘ontology’ as a transcendental realist project concerned with things in general and in themselves; an immanent ontology being very nearly a contradiction in terms, his insistence that Kant’s positive teaching regarding appearances is ‘non-ontological’ is almost a truism. ¹⁶

To settle (without begging, as does Allison) the question of whether Kant’s transcendental idealism is or contains an ontology, a concept of ontology is required that is neither arbitrary nor anachronistic and above all broad enough to accommodate synthetic a priori principles whose domain of application is restricted to appearances. That the concept of a regional ontology of nature satisfies the third requirement is obvious; that it also satisfies the first two can be made clear by considering briefly a couple of traditional connotations of ‘ontology’.

In the first place, ontology, as traditionally understood, is (1) a general theory of the ens qua ens, which, in the context of Kant’s transcendental idealism, can only mean the ens qua empirically real ens or phaenomenon. This is all that remains of the old Aristotelian science of to on hē on and of the mediaeval discipline devoted to the ens commune once the strictures of the Critique are applied to the unrestrictedly universal ancient, mediaeval, and early modern sciences of the being of ‘all things in general’ (Bxxvii). Nevertheless, Kant’s restrictedly general enquiry regarding the existence and essence of empirically real entities as such has the same right to the title ontologia as the discipline for which that Latin word was coined in the seventeenth century in order to distinguish – as Aristotle had not – the universal science of the ens qua ens or ens commune (ontologia) from the special science of the summum ens (theologia). ¹⁷ Kant’s doctrine of the ens qua phaenomenon is still general by comparison with those special metaphysical disciplines that apply its synthetic a priori principles to some specific domain of objects by introducing empirical concepts. Ontology, so understood, is a pure general metaphysics (metaphysica generalis) by comparison with rational physics as a branch of metaphysica specialis that is restricted to outer phenomena through the introduction of the empirical concepts of matter and motion; and it is likewise general with respect to a putative rational psychology that is confined to inner, mental phenomena, psychologia rationalis being another branch of metaphysica specialis. ¹⁸

A second connotation of ‘ontology’ has strong historical claims to be considered its principal meaning for Kant and the eighteenth century; for apart from (1) the doctrine of the ens qua ens, ontologia is also and above all a (2) doctrine of the praedicata entis generaliora. This is Baumgarten’s formal definition of ontologia in his Metaphysica (§4) and the authoritative one for Kant, even if for him, unlike Baumgarten, the only ens whose praedicata generaliora are in question is the ens qua empirically real spatio-temporal phaenomenon. This sense of ‘ontology’ narrows the focus vis-à-vis the first; whereas the qua of (1) ens qua ens straddles both mediaeval senses of ‘being’, ‘existence’ and ‘essence’, (2) shifts the focus from being as existence and coming-into-existence (ratio fiendi) to being as essence or nature (ratio essendi). The definition of ontologia found in Wolff’s 1728 Logica, namely, the science that treats de ente in genere et generalibus entium affectionibus (§73), places the emphasis squarely on this latter sense of ‘being’, though in Wolff’s, as in Baumgarten’s, definition, there is no thought of restricting ens to ens phaenomenale.

Armed with this concept of a regional ontology of nature, I turn now to the question of whether the passage at A247/B303 supports Allison’s thesis that ‘transcendental idealism is best viewed as an alternative to ontology’, or whether, on the contrary, it highlights the need to replace the old with a new ontology of the sort just described. ¹⁹

3. A new reading of A247/B303

Even those not especially averse to the idea of a universal Kantian science that deals with the empirically real existence of the objects of connected outer experience and with the essence of anything that can be considered an actual or even a merely possible object of empirical cognition may well baulk at the name ‘ontology’. For in what appears to be a definitive ex cathedra pronouncement, and the only reference to ontology in the entire Transcendental Analytic, Kant rejects that title in favour of another. Noting that the principles of pure understanding rigorously demonstrated for the first time in the culminating chapter of the Transcendental Analytic are ‘merely [1] principles of the exposition of appearances (Prinzipien der Exposition der Erscheinungen)’, he goes on to affirm that

… the [2] proud name of an ontology (einer Ontologie), which claims [3] presumptuously (sich anmaßt) to offer [4] synthetic a priori cognitions of [5] things in general in a systematic doctrine (e.g. the principle of causality), must give way to the [6] modest (bescheiden) one [i.e. to the modest name] of [7] a mere (einer bloßen) [8] analytic of the pure understanding. (A247/B303) ²⁰

Does this passage lend colour to Allison’s ‘alternative to ontology’ thesis? ²¹ On the face of it, it does seem to proclaim ‘the end of ontology’ and its replacement by a new discipline whose object is not being, but ‘human understanding’ in something akin to Locke’s or Hume’s sense. However, to read it thus, one must take the clause immediately after the word ‘ontology’ as a non-restrictive relative clause referring to ontologies of every ilk rather than just a particular, presumptuous kind of ontology. Now such a reading is made obligatory by the Guyer–Wood translation, which is taken over (with an ellipsis of no consequence) by Allison; for their translation inserts commas around the entire which-clause. Of course, in German, as these learned men well know, every relative (indeed, every subordinate) clause is preceded by a comma; but in English, the insertion of commas serves to distinguish non-restrictive from restrictive relative clauses. Thus, for example, the sentence ‘Ontology, which is general, cannot be tasked with this specific problem’ means that no ontology whatever can be so tasked given the inherently general nature of all ontologies; in this example, the commas make the relative clause non-restrictive or universal in scope. But if the commas are removed and, in addition, an indefinite article is inserted at the beginning, the resulting sentence, ‘An ontology which is general cannot be tasked with this specific problem’, has a very different meaning: a certain kind of ontology, say, a ‘formal’ ontology (see n. 15), cannot be expected to address the problem, the implication being that there may be a ‘regional’ ontology (ibid.) that is suited to the task. By disregarding the cue provided by Kant’s choice of the indefinite article (einer Ontologie), and by inserting commas (as they do quite deliberately, since the initial comma is not found in the earlier Kemp Smith translation, which they will have consulted), Guyer and Wood make the sense of the proposition expressed in the first sentence: ‘everything to which the name “ontology” is or can be applied “presumptuously claims…”’ etc. ²² However, if Kant’s relative clause is restrictive, as his deliberate choice of the indefinite over the definite article strongly suggests, then he is here juxtaposing an ontology of the presumptuous sort to another ontology – arguably, his own regional ontology of nature – which likewise furnishes ‘[4] synthetic a priori cognitions’ or ontological first principles, but not by any means ‘[3] presumptuously’, and that for two reasons. First, it does not apply such synthetic principles to ‘[5] things in general’ and in themselves, regarding them rather merely as ‘[1] principles of the exposition of appearances’ (e.a.). And, second, it provides a rigorous demonstration of each and every principle by means of ‘[7] a mere [8] analytic of the pure understanding’, that is, in the only way in which such principles can be rigorously demonstrated. ²³

None of this is intended to deny that Kant is here straightforwardly proposing to abandon the historically charged name ‘ontology’ for another, more suitable title that is free of all transcendental realist baggage: ‘[8] analytic of the pure understanding’. The only question is whether, beyond replacing the name, he is proposing to replace what Guyer and Wood call ‘ontological questions’ regarding the being of entities other than ourselves with something like a reflective epistemological enquiry concerned only with the understanding’s capacity to know such objects; or whether, on the contrary, he means by ‘[8] analytic of the pure understanding’ a new and more ‘[6] modest’ method of proving the very same first principles of things found in traditional ontologies, though only for outer things or appearances in space and time, not for ‘things in general’. It seems clear from this passage that Kant was convinced that the name ‘ontology’ could not fail to conjure up the idea of a metaphysical discipline that claims to demonstrate the truth of synthetic a priori principles de ente in genere – contrary to the core teaching of the whole Critique that first truths or principles that are both synthetic and a priori can never be demonstrated for ‘all things in general’ (Bxxvii; cf. A334/B391, A582/B660, A694/B722) or ‘objects in general’ (A56/B1; cf. A63/B88, A130, A235/B294). Yet since the objective validity of those metaphysical first principles whose predicates are the categories can be rigorously demonstrated for all ‘appearances in general (überhaupt)’ (A31/B46; cf. A34/B50–1, A138/B177, A156/B195, and A494/B522), it is not the connotation of generality as such that Kant is rejecting, but only that of absolutely unrestricted generality. Of course, the new name proposed as a substitute for ‘ontology’, ‘[8] analytic of the pure understanding’, does nothing to delimit the legitimate domain of application of a new general metaphysics or ontology, since the understanding, as Kant frequently remarks, can think a great deal that cannot in principle be given to the senses. It was for this reason that Kant began the whole passage by stating unequivocally that his principles are merely ‘[1] principles of the exposition of appearances’. But there may be another reason as well. The expression ‘exposition of appearances’ (e.a.) makes it plain that the new name, ‘[8] analytic of the pure understanding’, does not refer to a merely reflective or epistemological analysis whose sole object is the human understanding; it refers, rather, to a discipline whose object is also all those ‘appearances’ thought by the understanding as something actually or possibly given in outer sensory intuition. In other words, it is indeed a second-order, reflective, or ‘transcendental’, discipline, but one whose object is not just, or even primarily, the understanding itself, but also, and above all, such outer appearances or empirically real objects in space and time as are or can be thought by an understanding to which they are given by the sensory faculty. If so, it is neither an epistemology nor a meta-epistemology nor any other ‘alternative to ontology’ that Kant intends by ‘[8] analytic of the pure understanding’; it is rather an ‘alternative ontology’, though one of a new, ‘[6] modest’ kind, since it consists exclusively of first-order, synthetic a priori ‘[1] principles of the exposition of appearances’ demonstrated through second-order or transcendental reflection on the understanding itself. ²⁴

This reading of the above passage is confirmed by a consideration of the content of the principles themselves. For those principles that Kant believed himself to have demonstrated for the first time by adopting a new, transcendental method are the same as (or at least very similar to) the universal truths to which dogmatic metaphysics had ‘[3] presumptuously’ laid claim. Allison’s rejection of the idea of ‘an alternative ontology’ would certainly be correct had he meant by this that what has here been called Kant’s regional ontology of nature is not a revisionary ontology in virtue of the first principles it contains. ²⁵ But what Allison means by ‘not “an alternative ontology”’ is, of course, not this, but rather that Kant’s system of principles is not an ontology at all (except perhaps in some Pickwickian sense), but rather a meta-epistemological or ‘transcendental analysis of objectivity’ (1983: 227).

How plausible is it, then, that Kant, in this passage, has in mind a further kind of ontology whose demonstrations rest on a transcendental-logical ‘[8] analytic of the pure understanding’, that is, on a reflective analysis of the understanding’s own pure concepts with respect to the conditions of the possibility of their objective validity for outer appearances? Answer: plausible enough, since the mind or understanding is ‘better known’ than is any external object in space (mens notior corpore, in Descartes’s famous phrase) – witness the extent and certainty of its self-knowledge in the parallel discipline of formal logic. The fact that knowledge of the existence and nature of what is other than the understanding (of the rationes or principia fiendi et essendi rerum) can be obtained by an ‘[8] analytic of the pure understanding’, that is, by reflecting on the mind and the necessary conditions of its empirical cognition (on its rationes or principia cognoscendi, to use Baumgarten’s phrase) – this is precisely the thought expressed in another much-cited sentence of the Critique: ‘The conditions of the possibility of experience are, at the same time, the conditions of the possibility of the objects of experience, and on this account have objective validity a priori’ (A158/B197). ²⁶ From this ‘supreme principle of all synthetic judgements’ (Kant means: ‘synthetic a priori judgements of the pure understanding’), it is readily apparent that Kant was not bent on converting or transforming ontological questions about the being of empirically real objects into epistemological or meta-epistemological questions about the conditions of the possibility of their empirical cognition. The principle asserts that the most basic principia cognoscendi of the human mind just are (‘at the same time’) the ultimate metaphysical or ontological principia fiendi et essendi of empirically real objects in space and time; there is no question of epistemological or meta-epistemological reflection on principia cognoscendi being substituted for ontological cognition of the principia fiendi et essendi of outer objects, which is, however, what the talk of converting or transforming ontological into epistemological questions intends. The originally Augustinian, later Cartesian, and finally Husserlian idea with which Kant’s ‘supreme principle’ confronts his reader is that the path to absolutely certain ontological cognition of the existence and nature of outer appearances lies through inner cognition, in this case an ‘[8] analytic of the pure understanding’ in so far as objects are given to it in outer sensible intuition and are objectively known with respect to what they must be like (essence) and whether they can, do, or must exist. But, to repeat, the principles so demonstrated as universally and necessarily objectively valid are themselves none the less first-order principia de ente in genere qua outer phenomenon and de generalibus entium affectionibus for being demonstrated through second-order transcendental-logical reflection on the most general principia cognoscendi of the human mind. This is hardly gainsaid by the little word ‘mere (bloß)’ in the expression ‘[7] a mere [8] analytic of the pure understanding’. On the contrary, that word is a token of Kant’s profound and enduring astonishment at the fact that rigorously demonstrable metaphysical or ontological truths about things (as empirically real outer phenomena) are obtainable independently of any direct access to them, merely by turning inwards and examining one’s own intuitive and discursive faculties and the complex conditions or principia cognoscendi under which representations that are subjective in origin can possess necessary objective validity. ²⁷ It is indeed somewhat odd that, in the culminating chapter of the Transcendental Analytic, Kant chose to call these principles ‘principles of the pure understanding’ (e.a.), after the object of their reflective, transcendental-logical method of demonstration, rather than, as here, ‘[1] principles of … appearances’, in view of their restricted domain of application; but the name does not alter what these principles are or contain, and they are both subjective first principles of thinking and objective principles de ente in genere qua outer phaenomenon – in short, a regional ontology of nature established by a new reflexive method. ²⁸

4. Conclusion

If this is what that brief passage from the Phenomena and Noumena chapter in fact means, then far from telling against, it rather supports, the view that Kant set out to demonstrate for the first time a system of well-known and widely accepted first principles whose ontological import he unfortunately obscured by the titles ‘mathematical’ and ‘dynamical’. He obscured it even further by giving it ‘the modest [name] of a mere analytic of the pure understanding’, by which he hoped to forestall any confusion of his ‘principles of the exposition of appearances’ (e.a.) with the unrestricted universal ontologies of the recent and remoter past. Although the name ‘ontology’ is judiciously withheld in the ‘proud name’ passage itself (but see nn. 18 and 24), those first principles that Kant set out to establish as the unshakeable Cartesian foundation of all actual and possible theoretical cognition in both the pure and empirical sciences of nature represent a revolutionary bid to rehabilitate the once venerable science of being qua existence and essence by confining it within the strict limits dictated by a new method of demonstration. And while it is the reflective method, and not the principles themselves, that is revolutionary, the reorientation of metaphysical enquiries towards the human subject in second-order reflection upon the first principles of knowing must not be allowed to disguise the fact that transcendental idealism is not just a meta-epistemology, but also and above all a first-order science that demonstrates the first principles of being. If so, then Allison’s influential (see n. 3) thesis has done the historical Kant a disservice by removing him from the ranks of those ancient, mediaeval, and modern European thinkers whose philosophical energies were devoted to questions regarding the ‘many ways in which “being” is said’ (Aristotle). Kant no more abandoned ontology in favour of epistemology than did Descartes before him; on the contrary, had he, like Descartes, embodied his revolutionary philosophy in a conspectus intended for use in the universities and entitled The Principles of Philosophy, the frontispiece might fittingly have borne the opening sentence of Book XII of Aristotle’s Metaphysics: perí tēs ousías hē theōría (‘the subject of the enquiry is being’).