2. The fourfold meaning of ‘noumenon’

First and foremost, Kant’s expression ‘things in general and in themselves’ includes in its hyperextension, along with appearances, (a) all those ‘non-sensible (nichtsinnlich)’ (A494/B522) objects that Kant invokes as the ‘intelligible ground’ (A545/B573; A563/B591) or ‘transcendental substratum of outer appearances’ (A383). In view of its complete indeterminateness, Kant calls this unknown ‘cause of appearance (thus not itself appearance)’ (A288/B444) a mere ‘something (Etwas)’ (three times at A358, once at A372) and a ‘something in general (Etwas überhaupt)’ (B307). He also refers to it as a ‘transcendental object’ (using either Objekt or Gegenstand) in many places (cf. A46/B63, A191/B236, A277/B333, A288/B344, A358, A366, A372, A379-80, A393-4, A494-5/B522-3, A557/B585, and A613/B641), though this must be distinguished from another use of the same expression for something which, as he states explicitly, ‘cannot be called noumenon’ (A253).Footnote ⁴ The ‘intelligible ground’ of appearances is, of course, also called a noumenon (in the negative sense specified at B307) as well as a thing in itself (see the quotation from the Groundwork below).Footnote ⁵

This noumenon in what will be called sense (a) is undoubtedly Kant’s main use of ‘noumenon’ in the Critique and the sole use of ‘thing in itself’ up to the end of the first-edition Transcendental Analytic proper: the ‘Something (Etwas)’ about which there has been most ado in the critical reception of Kant’s transcendental distinction. And rightly so. For one thing, Kant’s doctrine of affection appears to be in blatant violation of his own ‘restriction thesis’ (Allison Reference Allison2015: 291), which prohibits application of the categories (here both ‘existence’ and ‘causality’) to any but the objects of possible experience; and, for another, it seems to be a gross ‘violation of noumenal ignorance’ (Robinson Reference Robinson1994: 422), to say nothing of the so-called affection problem: noumena in this sense can no more cause changes in our external sense receptors than can appearances; not appearances, because, as objects first constituted on the basis of sensory affection, they cannot (without circularity) be the causes of sensory affection; and not noumena, for the very reasons noted by Allison and Robinson (among others). And yet without the distinction of appearances from noumena in this first sense, Kant’s Copernican reversal of the epistemological model of metaphysical realism (in both its empiricist and rationalist forms) simply cannot ‘get off the ground’. For only the ‘objects of the senses and the understanding for experience’ (B xviii n.) can plausibly be said to ‘conform to our cognition’ (Bxvi), that is, to the subjective conditions under which alone they can be experienced; to say as much of those same or other things considered as existing in themselves without relation to our minds would be prima facie absurd.

With the Copernican reversal of the Transcendental Analytic complete, however, Kant introduces three further senses of ‘noumenon’ and ‘thing in itself’, each required for the achievement of some specific aim of the Critique. After the Phenomena and Noumena chapter introduces the noumenon ‘in a positive sense’ for the object of the divine nous, the Amphiboly engages in a little noticed use of ‘thing in itself’ and ‘noumenon’ for the object of the pure (in the sense of ‘unschematized’) categories. And in the Transcendental Dialectic, the terms ‘transcendental object’ and ‘thing in itself’ that had been used throughout the Analytic to designate the non-sensible object behind the appearances are applied chiefly to the ‘supersensible (übersinnlich)’ (Bxxi) objects of pure reason, (i) God, (ii) the rational soul, and (iii) the world as that unconditioned totality of all things which no human experience can ever completely encompass. The expression ‘things in general and in themselves’ thus comes to include in its hyperextension this (b) trio of super-sensible objects beyond the world of appearances.Footnote ⁶ While the latter are not called noumena anywhere in the Critique, but rather ‘problems of pure reason’ (A3/B7), ‘being[s] of reason’ (twice at A681/B709), and the like, in the Prolegomena Kant remarks that ‘we should then think for ourselves an immaterial being [the rational soul], and intelligible world [as a complete totality] and the highest of all beings [God] (all noumena), because only in these things, as things in themselves, does reason find completion and satisfaction’ (P, 4: 354 e.a.). Even if these super-sensible objects were not explicitly so called in the Prolegomena, it would be none the less certain that they are things in themselves, since the ideas of pure reason are required in order to complete a series of conditioned things in themselves rather than appearances, and only what belongs to such a series can complete it.Footnote ⁷ The parenthetical ‘all noumena’, on the other hand, seems almost redundant, for apart from enabling the Copernican reversal, Kant’s distinction of phenomena from noumena is equally intended to forestall all pretensions to metaphysical knowledge of super-sensible objects beyond the realm of appearances. Given this its role in relation to the three branches of Wolffian metaphysica specialis, it is obvious that this is a different sense of ‘noumenon’ from that considered first. For one thing, while the ‘transcendental object’ in the first, negative sense of ‘noumenon’ is in most cases (see section 4) the very same object that appears, though considered as existing in itself, there is no question but that the three super-sensible objects of rational psychology, cosmology, and theology are numerically distinct entities from all sensible appearances. They do not, nor can they, appear, while noumena in the first sense do or at least can.Footnote ⁸

It is true, of course, that neither in the Critique nor in the Prolegomena does Kant make any such terminological distinction as has just been drawn between (a) noumena ‘behind’ and (b) noumena ‘beyond’ the sensible world. But in the Groundwork he remarks of the former that ‘we must admit and assume behind appearances something else that is not appearance, namely things in themselves’ (G, 4: 451 e.a.); and in The Progress of Metaphysics he speaks of ‘super-sensible objects, of God, of our own capacity for freedom, and of our soul (in separation from the body)’ (PM, 20: 296), whereby ‘super’ (über) plainly has the sense of ‘beyond’ (Latin: trans, Greek: metá). Unfortunately, this further use of ‘noumenon’ and ‘thing in itself’ in what has just been labelled sense (b) only seems to aggravate Kant’s problems. Martin (Reference Martin1955: 148, 158, 189), who distinguishes just two senses of ‘thing in itself’, one corresponding to part of (b), namely God and finite rational souls, the other corresponding to (a), that is, the intelligible causes of outer appearances, taxes Kant with illegitimately applying, not just the categories of causality and existence, but also ‘unity’, ‘multiplicity’ (both categories of quantity), ‘reality’ (quality), ‘substance’ (a category of relation), as well as the other two modal categories, ‘possibility’ and ‘necessity’ – in short, ‘nearly all’ (p. 198) the categories to ‘noumena’ in one or the other of his two senses.

Before turning next to the Phenomena and Noumena chapter and the noumenon ‘in a positive sense’, it is worth noting that Kant also calls the super-sensible objects of reason ‘beings of thought’ in the Prolegomena (P, 4: 349), and even, surprisingly, ‘being[s] of the understanding’ (P, 4: 352 e.a.) rather than reason. Still, Verstandeswesen may just be a literal rendering of the Greek nooúmenon (cf. Noumena oder Verstandeswesen in Refl CXXXI in Kant’s copy of the first Critique, 23: 36), and a use therefore only apparently at odds with Kant’s many statements to the effect that God, the world-totality, and the rational soul are the objects of the pure concepts of reason rather than the understanding. In the Critique, ‘being of the understanding’ occurs only once (A562/B591), but again, disconcertingly, in reference to the necessary being, God, the object of an idea of reason. Had Kant always been as attentive as he sometimes is to his own distinction between the two higher cognitive faculties in man, he would have reserved ‘being of the understanding’ for two other classes of objects besides noumena in sense (a). One of these he calls ‘extra-sensible (außersinnlich)’ (A287/B343) as distinct from both (a) non- (nicht-) and (b) super- (über-) sensible objects. This extra-sensible object of God’s intuitive understanding is ‘the noumenon in a positive sense’ (B307) or in what will be called sense (c). The other perfectly correct use of ‘being of the understanding’ is explained most fully in a passage of the Prolegomena (P, 4: 332) to be cited below, where it is applied, along with the term ‘NOUMENA’, to objects of the pure or unschematized categories. The object of this fourth use of ‘noumenon’ and ‘thing in itself’ can be called insensible and will be referred to as the noumenon in sense (d). It has been amply documented by Allison (Reference Allison1996, note 13 provides many examples). Whether or not, as Allison maintains, the use of ‘noumenon’ in sense (d) is merely a modification of its use in sense (a) (see n. 14 below), there can be no question but that the Dialectic inaugurates an altogether different use, or that the same is true of the Phenomena and Noumena chapter, to be considered next.

This new use of ‘noumenon’ is found first in the following passage (replaced in the second edition) of the Phenomena and Noumena chapter:

Attempting to disambiguate the latinized Greek term noumena in the second edition, Kant contrasts the noumenon in sense (a) with that in sense (c) as follows:

The description of the first class of ‘beings of the understanding’ or noumena is an unmistakable reference to non-sensible things in themselves or noumena in sense (a). For they are, to the extent that they appear, ‘the [numerically] same objects’ as their appearances, but considered differently, or better, considered also as existing in themselves. What, however, does Kant mean by ‘other possible things which are not objects of our senses at all’? The juxtaposition of ‘the same objects’ with ‘other objects’ suggests that the latter, unlike the former, differ numerically from those objects that appear to us.Footnote ⁹ Now the super-sensible objects, God, the rational soul, and the world-totality, fit this description; moreover, they are designated ‘beings of the understanding’ and noumena, if not in the Critique, so at least in the Prolegomena. But the sequel makes it clear that Kant understands something else by ‘other possible things’. Disambiguating ‘noumenon’ still further in the next paragraph, he juxtaposes noumena in sense (a), which he now designates ‘noumena in the negative sense’, with (c) the objects of a purely intellectual intuition such as only the divine understanding is capable of, calling the latter the noumena ‘in a positive sense’:

The contrast here is between (a) not being an object of sensible intuition, which is merely a negative description, and (c) being the object of a non-sensible or intellectual intuition, which is positive. Leaving further comment on the question of numerical distinctness for later (see section 4), three observations are in order regarding noumena in this positive sense.

The first is merely terminological. While (a) the noumenon in the negative sense is consistently described as non-sensible (nichtsinnlich) by Kant, (c) the noumenon in the positive sense is called ‘extra-sensible (außersinnlich)’ (A287/B343). Even if not something on which Kant lays any stress, the distinction among (a) non-, (b) super-, and (c) extra-sensible objects or noumena is both present in the text of the Critique (though, admittedly, with only a single occurrence of außersinnlich) and useful for purposes of disengaging three quite distinct senses of ‘noumenon’ or ‘thing in itself’. At the beginning of the next section, some Latin equivalents, based on Kant’s own usage, will be introduced for the same purpose.Footnote ¹⁰

Secondly, Kant at times refers to intellectual intuition (intuitus intellectualis) instead of ‘non-sensible intuition’ (A286/B442) when speaking of the extra-sensible object or objects of the divine understanding (e.g. B72; cf. A249, B307, B308, A279/B335, Refl CXXXI, 23: 36, and CLX, 23: 39). For, according to the deleted passage from the A edition quoted above, noumena in sense (c) – literally, ‘things thought by the divine nous’ – are ‘merely objects of an understanding … which nevertheless can be given as such to an intuition, although not to a sensible intuition (thus coram intuitu intellectuali)’ (A249). As Kant explains elsewhere (cf. ID, 2: 396, CPJ, 5: 407-8, RPT, 8: 391, Corr, 10: 130, and Refl CXIII, 23: 35), the intuitus intellectualis of the divine ens originarium differs from the intuitus derivativus (B72) of created minds in that it brings its objects into existence spontaneously through the very act of intuiting them.Footnote ¹¹

The third and final observation concerns what was earlier referred to as the merely problematic or thought-experimental status of Kant’s transcendental distinction. It is important to note that the use of ‘negative’ in the phrase ‘noumenon in the negative sense’ (B307) is not his only use of that term with reference to noumena. While the above is the locus classicus for the strict use of ‘negative’, Kant also applies it in a loose sense to all those ‘boundary’ (cf. A254/B310) concepts whose objects are only ‘problematically’ (ibid.) asserted to exist, that is, to concepts for the existence of whose objects there is only ‘room’ or ‘space’ (Raum or Platz), without its being possible to assert positively (within theoretical philosophy at least) that they possess ‘objective reality’ (ibid.) or are non-empty. Thus, at the end of the Amphiboly, Kant startlingly refers to ‘objects of a non-sensible intuition’ (A286/B342), that is, to (c) noumena in the positive sense, as ‘noumena in this negative sense’ (ibid.). He is not, of course, confounding noumena in sense (c) with noumena in sense (a), having only just introduced a sharp distinction between the two in the preceding Phenomena and Noumena chapter; rather, ‘negative’ is being used here in a different, loose sense to signify that the problematic concept of the noumenon in sense (c) ‘serves for nothing but to designate the boundaries of our sensible cognition and leave open a space (Raum) that we can fill up neither through possible experience nor through the understanding’ (A288-89/B345). And if even (c) the noumenon ‘in a positive sense’ is negative in the sense of ‘merely problematic’, then (a) the noumenon in the negative sense is doubly so: on the one hand, in the strict sense of ‘not an object of our sensible intuition’ (B307), and on the other, in the loose sense of serving ‘for nothing but to designate the boundaries of our sensible cognition’ (A288/B345). As for those (b) noumena which are the super-sensible objects of reason, they too are problematic or negative in the loose sense, and are so described (that is, as problematic, not as negative) in the famous passage of the second Preface: ‘I had to deny knowledge in order to make room (um Platz zu machen) for faith’ (Bxxx). And the same can be said of noumena in sense (d), the objects of the unschematized categories alone: a space must be reserved for objects such as the categories represent on their own, if only to show why it too can never be ‘filled up’ without the contribution of the sensible faculty. The third point, then, can be put this way: ‘problematic’ and ‘negative’ in the loose sense being interchangeable concepts, all four concepts of the noumenon are merely problematic or negative, while the first is negative in a further, strict sense of the term. This is vital to the argument (see section 5) that ‘noumenon’ designates an object whose existence is not assumed as a matter of metaphysical doctrine, but only hypothesized as part of an elaborate meta-metaphysical thought experiment.Footnote ¹²

While the enumeration carried out so far covers three different sorts of noumena that fall within the extension of ‘thing in general and in itself’, there remains a further sense of ‘being of the understanding’ and ‘beings of thought’ still to be considered. It has already been designated the insensible (as opposed to the non-, super-, and extra-sensible) object and the noumenon in sense (d). It is the thing in itself in this sense that Kant has in mind where, in the Amphiboly chapter, he speaks of ‘merely intelligible objects’ that are ‘determinable in accordance with mere concepts’ (A285/B341) and of ‘things that are thought through pure categories without any schema of sensibility’ (A286/B342); and it is again noumena in this fourth sense that are intended where, in the Antinomies, he speaks of ‘objects given to the mere understanding’ (A500/B528) and the ‘thing in itself thought through pure [i.e. unschematized] concepts of the understanding’ (A525/B553). But the clearest reference is perhaps that found in the Prolegomena passage mentioned above:

It is not surprising that in the Amphiboly and Antinomies Kant should have found it opportune to introduce a new idea of the noumenal that brings the specifically Leibnizian conception of ‘things in themselves and as objects of mere understanding’ (A206/B251-2) within the ambit of his general critique of all dogmatic (as distinct from Kant’s own, merely negative or ‘problematic’) distinctions. In the Transcendental Dialectic, that initial transcendental distinction which, in the Analytic, had been contrasted with both the ‘empirical’ distinction of Locke (A45-6/B62-3) and the ‘merely logical’ distinction of ‘Leibniz-Wolffian philosophy’ (A44/B61-2) was expanded to address the claims of contemporary rational psychology, cosmology, and theology regarding knowledge of super-sensible objects; and, in the Phenomena and Noumena chapter, it underwent a further amplification to ward off any confusion of Kant’s problematic transcendental with the dogmatic Augustinian distinction between our sensible and a divine, intellectual intuition that knows extra-sensible objects as they are in themselves by bringing them into existence exactly as it conceives them. Kant, in short, is not equivocating, but consciously articulating the original distinction of the Transcendental Aesthetic and Analytic the better to address metaphysical theories that dogmatically ascribe other modes of existence to the same objects that appear or else posit numerically different objects having other modes of existence than sensible appearances. Since, near the beginning of the Amphiboly chapter, he draws an implicit contrast between sensibilia and these ‘intelligibilia, objects of the pure understanding’ (A264/B320), the designation ‘insensible object’ fits this final sense of ‘noumenon’ well enough.Footnote ¹⁴

3. The narrower and wider senses of ‘transcendental distinction’

In turning next to the expression ‘transcendental distinction’, it may not be amiss (in a Kantian context at least) to resort to a convenient Latin shorthand for the four senses of ‘thing in itself’ or ‘noumenon’ just distinguished. Picking up on Kant’s use of ‘without’ (Latin: sine or absque) at B307 (‘things that the understanding must think without [ohne] this relation to our kind of intuition’), and expanding his use of coram at A249 (coram intuitu intellectuali), the expressions absque intuitu sensibili and coram intuitu sensibili can be employed to capture the main transcendental distinction drawn in the B-Preface and put to work immediately in the Transcendental Aesthetic and throughout most of the Analytic.Footnote ¹⁵ The same Latin terms can serve for another contrast as well. Kant’s distinction of the negative from the positive sense of ‘noumenon’ can be captured by contrasting the object as it exists absque intuitu sensibili with the object considered as existing coram intuitu intellectuali. Furthermore, the object existing coram intuitu intellectuali may be contrasted with the object considered as existing coram intellectu humano puro to mark the change the term ‘thing in itself’ undergoes as Kant passes from the Phenomena and Noumena chapter to the Amphiboly. Adding to the foregoing, as a convenient shorthand for the trio of noumena of the Transcendental Dialectic, objects existing coram ratione humana, Kant’s transcendental distinction may be said to be one between objects considered as existing coram intuitu sensibili (appearances), on the one hand, and, on the other, either the same objects considered as existing in three other ways (absque intuitu sensibili, coram intuitu intellectuali, and coram intellectu humano puro), or a class of numerically different objects existing coram ratione humana. All this maps neatly onto the earlier distinction of sensible from (a) non-sensible (absque intuitu sensibili), (b) super-sensible (coram ratione humana), (c) extra-sensible (coram intuitu intellectuali), and (d) insensible (coram intellectu humano puro) objects; but it takes for granted both a very broad sense of ‘transcendental distinction’ involving all four contrasts and assumes that it is solely in the case of super-sensible objects that Kant’s distinction is a two-object theory. The first assumption will be defended in the remainder of the present, the second qualified in the next section.

It is true that, where first introduced in the B-Preface (Bxviii, Bxxi), the transcendental distinction concerns only (a), absolutely existing non-sensible things or noumena in the negative sense. In the Phenomena and Noumena chapter, where the expression (c) ‘noumenon in a positive sense’ is first introduced, Kant even issues a stern warning that his ‘doctrine of sensibility is at the same time a doctrine of the noumenon’ solely in the negative sense (B307). But even granting that only the thing in itself or noumenon in sense (a) figures in what will henceforth be called the transcendental distinction proper, it will be hard to deny that the ‘doctrine of sensibility’ is enriched by the hypothetical contrast, first mentioned in the Phenomena and Noumena chapter and only subsequently inserted into the second-edition Transcendental Aesthetic (cf. B72), between a putative divine faculty of intuitus originarius and the human faculty described as intuitus derivativus. After all, Kant himself notes (cf. A252/B308-9 and A254/B310) that the noumenon ‘in a positive sense’ serves to underscore the fact that there is no compelling reason to believe that sensible is the only possible kind of intuition, even if it is the only kind known to us and the only kind of which the human mind can fully comprehend the real possibility; and this is as much as to say that there is no reason not to entertain the hypothesis of other modes of existence of these same objects coram other possible kinds of intuition, notably coram intuitu intellectuali, should such an hypothesis prove helpful in determining the limits of our human cognition. Thus, while in the B-Preface Kant speaks only of ‘a twofold standpoint’ (Bxix n. e.a.) and of considering ‘the same objects’ from ‘two different sides’ (Bxviii n. e.a.), to consider appearances from these two sides while also considering existence coram intuitu intellectuali as a further possible mode of existence of those same things that appear may be regarded as Kant’s transcendental distinction in a wider and improper sense. That sensible is derivative and not ‘originary’ intuition is surely part of Kant’s ‘doctrine of sensibility’.

It was argued earlier (see n. 14) that, despite claiming only to ‘modify’ his initial account of Kant’s transcendental distinction, Allison in fact passes unawares from (what may now be termed) his earlier coram-absque distinction to his later coram-coram contrast, juxtaposing the object as it appears to our sensible intuition with the same object considered as existing coram intellectu humano puro. To the extent that the non-spatiotemporal existence of things coram intellectu humano serves as a new foil to their spatiotemporal existence coram intuitu sensibili, it too forms part of Kant’s ‘doctrine of sensibility’ and hence of the transcendental distinction in a still wider sense. At the very least, its introduction in the Amphiboly clarifies the way in which Kant understands the ‘merely logical’ (A44/B61) distinction of the Wolffians between the senses, which represent things in themselves obscurely and confusedly or as they appear, and the understanding that represents those same things clearly and distinctly as they are in themselves through the pure (unschematized) categories alone.Footnote ¹⁶

So much for noumena in senses (c) and (d). What, finally, of the remaining use of ‘noumenon’ in sense (b) for the super-sensible objects ‘God’, ‘the rational soul’,’ and ‘the world totality’? Do they too form an integral part of Kant’s transcendental distinction in what may be called the fourth and widest sense of the expression? It is difficult not to accord the trio a special place in that distinction, given Kant’s famous remark in the second Preface: ‘I had to deny knowledge in order to make room for faith’ (Bxxx). By ‘faith’ Kant means what he elsewhere calls a Vernunftglaube (WOT 8: 140–44) in ‘God, freedom, and immortality for the sake of the necessary practical use’ of reason (Bxxx; cf. A3/B7). Since the transcendental distinction between appearances and these super-sensible objects of special metaphysics dominates the Transcendental Dialectic, and since it is crucial to what Kant himself considered the whole purpose of the first in relation to the second Critique (see section 6), it seems that the distinction between appearances and these super-sensible objects is a key part of Kant’s transcendental distinction in the most encompassing sense of the expression.

Before turning, in the next section, to the earlier suggestion that Kant’s transcendental is a two-object distinction only in the case of these three super-sensible noumena, it is worth considering briefly at least one other interpretation of what has been called the reductive type. It has already been noted that Kant’s expansion of the transcendental distinction proper to include the object existing coram intuitu intellectuali seems to have occasioned Robinson’s attempt to shore up Allison’s ‘one-object’ or ‘one-world’ view by sinking the object existing absque intuitu sensibili into the object existing coram intuitu intellectuali (see n. 10 above). Of course, Allison himself performs a different reduction: from (a) absque mente humana he slides to (d) coram intellectu humano puro, two distinct uses of ‘noumenon’ that were conflated already by Adickes (cf. n. 16). While these conflations were fostered by the Phenomena and Noumena chapter and the Amphiboly, respectively, the (b) super-sensible objects existing coram ratione humana in the Transcendental Dialectic have given rise to another reductive interpretation. Schaper (Reference Schaper1966: 236, 240) interprets Kant’s description of noumena in sense (a), the ‘things that the understanding must think without [ohne] this relation to our kind of intuition’ (B307), as though ‘must’ expressed a ‘demand of reason’ (A305/B362; cf. A332/B389) for the ‘postulation of a noumenal world’. This reduces the noumenon in sense (a) to that in sense (b), ‘the unconditioned, which reason necessarily and with every right demands in things in themselves for everything that is conditioned’ (Bxx). And while Schaper’s reading of Kant’s transcendental distinction is only an isolated attempt to revive the notorious ‘as-if’ interpretation of Vaihinger (Reference Vaihinger1922), Adickes ascribes to ‘Cohen and his [i.e. the whole Marburg] school’ the view that ‘the thing in itself is … a mere Idea’ of reason (Reference Adickes1924: 43; cf. also 121). If this blanket ascription is even partially correct, the reduction of which Schaper is guilty has long been a potent source of confusion.

A coherent response to all the stock objections repeatedly brought against Kant’s transcendental distinction can hardly be expected as long as these different senses of ‘noumenon’ continue to be reduced to a single meaning. On the other hand, it may be possible to dismiss the objections en bloc if none of Kant’s various transcendental distinctions represents an ontological doctrine, being rather parts of a meta-metaphysical thought experiment that Kant regarded as indispensable for the establishment of that new, scientific metaphysics whose foundations are laid in the chapter of the Critique entitled ‘Principles of Pure Understanding’. That is the subject of section 5. First, however, a word is in order on the implications of the foregoing for one- and two-object interpretations.

4. One- and two-object interpretations

That it is not indeed only in the case of (b) the super-sensible objects of pure reason that one may speak of two ontologically distinct objects is most easily seen from Kant’s reference to ‘other possible things’ (B306) in the already-quoted passage from the Phenomena and Noumena chapter that introduces (c) extra-sensible noumena ‘in a positive sense’. As noted in interpreting that passage earlier, by referring to them as ‘other possible things’ (e.a.) Kant seems to imply that all noumena in sense (c) are – like the trio of super-sensible noumena in sense (b), but unlike noumena in sense (a) – numerically distinct objects from appearances. It seems unlikely, however, that Kant wished to deny the theological truism that the same things known through the senses as they appear to us hinc et nunc are known from eternity as they exist in themselves by that ‘special kind of intuition’ (B307) that is God’s alone. And indeed, in a marginal note in his desk copy, Kant expressly disavows any such view: ‘Objects of a non–sensible [divine] intuition are either given [also to us] in a sensible intuition or not’ (Refl CXXXI, 23: 36 e.a.). The note goes on to make the familiar point that we cannot comprehend the real possibility of God’s non-sensible intuition, from which it follows that we cannot grasp the real possibility of such noumena either. And yet if (as will be argued presently) some (but not all) of the same objects that exist as noumena in sense (a) also appear; and if, as theological orthodoxy demands, all things that exist as noumena in sense (a), whether ‘given’ in sensible intuition or not, are present from eternity to God’s intellectual intuition, then it will not be amiss to speak (taking the two-modes-of-existence approach to Kant’s transcendental distinction one step further) of three distinct ways in which one and the same object can exist: first, as appearance before the human mind; second, as (a) noumenon in the negative sense ‘without (ohne)’ relation to the human mind; and, third, as an (c) extra-sensible noumenon intuited by the divine intellect. What both B306 and Kant’s note make clear, however, is that the concept of the (c) ‘noumenon in a positive sense’ also includes in its extension some ‘other possible things’ (B306) which do not appear and are thus numerically distinct from all appearances. Having established that some (c) extra-sensible, like all (b) super-sensible noumena, are numerically distinct from all objects that appear, it only remains to show that the same is true of some noumena in senses (a) and (d).

The case of noumena in sense (c) suggests the following general solution to the long-running one- or two-object dispute: insofar as certain noumena (in any of the disputed senses) also appear to us, phenomena and noumena are the numerically same object existing (or better, considered as existing) in different ways; whereas those noumena that either do not or cannot appear to us are to be considered as numerically distinct objects from all phenomena. It was shown in the preceding paragraph that some (c) extra-sensible objects are numerically distinct from appearances, while others are not. But it only stands to reason that the same holds true for some noumena in sense (a) as well; for Kant surely meant it to be understood (as ordinary common sense dictates) that not everything considered as existing in itself in this, his main sense of ‘noumenon’ must at some time or other actually appear to the senses of some human being (consider: ‘many a flow’r is born to blush unseen’). This means that some noumena in sense (a) are numerically distinct objects from any that appear, though, admittedly, Kant’s transcendental distinction is principally concerned with those noumena in sense (a) that do also appear. Finally, what holds for noumena in sense (a) would seem to be true also of (d) the insensible objects of the pure unschematized categories. Recall the ‘merely logical’ Leibniz-Wolffian distinction between appearances and things in themselves: it is the very same objects that appear obscurely and confusedly to the senses which are also clearly and distinctly cognized as they are in themselves through the pure concepts of the understanding. As noted in the Introduction, Leibnizian pure principles of the understanding were regarded as universally valid for ‘all things in general and in themselves’, including those (d) objects of the pure unschematized categories which do not or cannot appear to the sense at all. In the Prolegomena passage cited earlier, the first example of such an object is a ‘substance, but which is thought without persistence in time’ (P, 4: 332 e.a.). When the unschematized category ‘substance’ is applied to material substances that appear in outer sensible intuition, the same object is being considered as existing in two ways, as an appearance and as a noumenon in sense (d); but there are also substances of which, ‘just because they are immaterial, there is no sensitive intuition’ (ID, 2: 414), and the latter are ontologically distinct objects of the categories alone.

In sum, then, one-object interpretations are correct for some noumena in senses (a), (c), and (d), namely those that also appear to us, though other noumena in all three senses either do not or cannot appear and are thus, like (b) the trio of super-sensible noumena, numerically distinct objects. Both one- and two-object approaches thus have their proper place in the interpretation of Kant’s transcendental distinction in the widest sense. Most of the controversy, however, has revolved around the distinction of appearances from noumena in sense (a), the ‘Something (Etwas)’ of the title. And here the debate between ‘one-’ and ‘two-worlders’ (Van Cleve), or as Aquila puts it, between ‘one-objectism’ and ‘two-objectism’, has been bedevilled by (1) the ambiguities of ‘noumenon’, ‘thing in itself’, and ‘transcendental distinction’, but also by (2) the failure to distinguish between those (numerically identical) noumena that do, and those (numerically distinct) noumena that do not, appear.Footnote ¹⁷

The nowadays most-discussed one-object theory – unhappily baptized the ‘two-aspect’ or ‘dual aspect theory’ – is militantly non- (even anti-) metaphysical, having been variously labelled ‘epistemological’, ‘methodological’, ‘meta-epistemological’, ‘meta-philosophical’, and ‘nonontological’ by Allison, its principal proponent in recent decades. By contrast, two-object theories are typically metaphysical and consequently dismissive of Kant’s transcendental distinction between appearances and noumena in sense (a).Footnote ¹⁸ Precisely because they too are metaphysical in a doctrinal sense, two-modes-of-existence interpretations (one thinks not just of Aquila Reference Aquila1979, but of Sellars Reference Sellars1968) are not much better placed than other metaphysical theories when it comes to meeting those objections that (meta-)epistemological one-object/two-aspect interpretations dispose of with relative ease.Footnote ¹⁹

The multiple-modes-of-existence interpretation to be defended in the next section builds on both the metaphysical two-modes-of-existence theory and meta-epistemological two-modes-of-consideration interpretation. It argues that the various noumenal modes of existence distinguished in the foregoing are not something that either appearances or other things are asserted to have, but something they may and should be considered as having for purposes of a meta-metaphysical thought experiment. What, after all, does ‘problematic’ mean if not that the existence of noumenal entities is not maintained assertorically, as part of a first-order or object-level metaphysical doctrine, but only entertained problematically, for purposes of a meta-level or second-order reflection on the possibility of first-order metaphysics? If this is what Kant means, then all the standard objections to his supposed doctrine of the noumenal simply fall away, as on the two-aspect reading, yet without Kant’s transcendental distinction having been rendered non-metaphysical.

5. A meta-metaphysical thought experiment

It remains to this day a, if not the, scandal of Kant scholarship that there is still no generally accepted reply to the stock objections that have beset Kant’s transcendental distinction almost from its inception.Footnote ²⁰ It has been suggested that the chief obstacle to a satisfactory resolution of these difficulties is the assumption that the transcendental distinction (when there are four such) is put forward as a metaphysical doctrine (rather than a ‘problematic’ hypothesis). By ‘problematic’ (cf. A38/B55, A254-6/B310-1, A259, A286-7/B343, A290/B346, A339/B397) Kant means that nothing prevents us from hypothetically ascribing various noumenal modes of existence (or of transcendental reality) to those appearances that we deem empirically real, nor yet to other things that either do not or cannot appear; and that, since we can, we should do so, not only to avoid the absurdity of appearances ‘without anything that appears’ (see n. 5 above), but also in recognition of the fact that our sensible is not the only possible kind of intuition, nor the objects that appear to us the only things capable of being represented by some (including our own) pure understanding, even though they are the only things that we can be said to know. To quote a well-known passage from the Phenomena and Noumena chapter:

Despite the reference to its merely ‘negative use’, the subject of this passage is clearly not the ‘noumenon in the negative sense’, but precisely the ‘noumenon in a positive sense’ or those objects described in the first edition as ‘noumena (intelligibilia) [existing] coram intuitu intellectuali’ (A249). As noted already (see section 2 in fine), all Kant’s uses of ‘noumenon’ are ‘negative’ in this loose sense of ‘problematic’, ‘empty (for us)’, etc., the noumenon ‘in the negative sense’ being therefore doubly negative. Still, it is precisely the noumenon ‘in a positive sense’ that adds a theological dimension to Kant’s transcendental distinction broadly understood, while at the same time reinforcing the ontological character of his thought experiment.

Unlike the step from (i) ‘noumenon’ as a merely ‘problematic’ or ‘boundary’ concept to (ii) two ways of considering the same object for meta-philosophical purposes; or, alternatively, from (i) to (iii) two ways in which the same object exists, metaphysically speaking; the transition from (i) to (iv) a thought experiment involving multiple modes of existence of the same or numerically distinct objects may well seem a step beyond anything Kant would have recognized as his own transcendental distinction. After all, is it plausible that a pillar of the critical philosophy as central as the transcendental distinction between appearances and things in themselves is neither a metaphysical nor a (meta-) epistemological doctrine, but a thought experiment, and an onto-theological one at that?

Kant’s most famous thought experiment, and perhaps the only one generally recognized as such, is found in his 1768 treatment of incongruent counterparts. Imagining God to have commenced his creation of the universe with a single human hand, Kant remarks that the hand must surely be either a right or left hand; that, in other words, it cannot, without absurdity, ‘be completely indeterminate with respect to such a property’ (GDS, 2: 383) as spatial orientation or directionality. This purely experimental fragment of an onto-theological creation story is deemed fatal to the Leibnizian view ‘that space [and therefore spatial orientation] simply consists in the external relation of the parts of matter which exist alongside one another’ (ibid., e.a.); for, ex hypothesi, there is as yet nothing else in the universe apart from the hand. Much the same technique was already used even earlier, in the Nova Dilucidatio of 1755, where ‘a being considered absolutely’ (NE, 1: 413) is an entity thought of as existing alone in the universe.

As mentioned in the Introduction, the expression ‘things in general and in themselves’ includes in its hyperextension, along with appearances, not just the objects of a hypothetical more-than-human intellectual intuition, but even those of a putative non-human faculty of intuition which, like ours, is finite or sensible, but whose ordering forms are, unlike ours, neither spatial nor temporal. Now if imagining the created universe to have once contained less than it now contains (for example, a single hand) is a thought experiment, why then should it be otherwise when one imagines another type of sensible intuition as existing in the world or – more to the present point – a non-sensible intuition existing outside it? After all, in the Preface to the second edition Kant himself addresses the transcendental distinction proper as an ‘experiment’. The word (now the German Experiment, now Versuch) and its cognates occur over a dozen times. It is used, to be sure, to describe the Copernican ‘transformation of our way of thinking’ (Bxxii n.) in the first instance (cf. Bxvi, Bxviii and n.), but also with specific reference to the transcendental distinction proper (cf. Bxviii and n., Bxix n., Bxx and n., Bxxi). Thanks to a new ‘hypothesis’ (Bxxii n.), which Kant describes as ‘analogous’ (ibid.) to ‘the first thoughts of Copernicus’ (Bxvi), metaphysics is to enter upon the ‘secure path of a science’ (Bix et passim), the natural sciences having done so several generations earlier thanks to the pioneers of ‘the experimental method’ (Bxiii n.) of early modern science (Kant mentions Bacon, Galileo, Torricelli, and Stahl). From these passages, it appears that the transcendental distinction of the B-Preface is intended as an auxiliary meta-metaphysical hypothesis or thought experiment in support of the main epistemological or Copernican hypothesis that objects ‘must conform to our cognition’ (Bxvi). Its purpose is (see section 2) to restrict the scope of that primary hypothesis to objects as appearances by excluding noumena in sense (a), in relation to which it would be patently absurd (ibid.). Still, it by no means asserts the real existence or causality of any entities over and above the objects of possible experience, nor yet the non-spatiotemporality of such entities, much less their one-to-one correspondence with appearances. True, according to the experimental method, hypotheses are framed for purposes of testing. Yet Kant’s intent was not to prove the existence and/or causality etc. of what he was later to call noumena ‘in the negative sense’ by the evident success of his Copernican hypothesis in elucidating the possibility of pure mathematics (cf. Bxxii n.), or by the success of the transcendental distinction in dispelling the Antinomies (cf. Bxix n.). These successes indeed prove the Copernican hypothesis that ‘we can cognize of things a priori only what we ourselves have put into them’ (Bxviii); and they prove it, so Kant maintains, ‘not hypothetically but rather apodictically’ (Bxxii), albeit only as restricted in scope by the auxiliary hypothesis. Yet proof of the transcendental distinction itself (as opposed to showing its usefulness and even indispensability as a secondary hypothesis) is not Kant’s objective. It enables the performance of something like an experimentum crucis that requires at least ‘a space’ for other-than-sensible or noumenal objects, without in any way committing the experimenter to their actual non-spatiotemporal existence or causality, let alone to their possession of still other categorial properties, their non-spatiotemporality, and their one-to-one relationship with appearances.Footnote ²¹

Admittedly, only what has been called the transcendental distinction proper is explicitly described as an ‘experiment’ in the B-Preface. As for the other senses of ‘thing in itself’ and ‘noumenon’ in the Dialectic, the Phenomena and Noumena chapter, and the Amphiboly, Kant could, of course, have come to understand these further juxtapositions of sensible with super-sensible, extra-sensible, and insensible objects in the same thought-experimental vein in which he had come – belatedly, in the B-Preface – to understand his central contrast of objects existing coram and the same objects existing absque intuitu sensibili. But even if only this transcendental distinction proper were a thought experiment, those metaphysical two-modes-of-existence interpretations focused on the Scholastic real distinction between a thing’s realitas obiectiva or being ‘for us’ and its realitas formalis or being ‘simpliciter (i.e. in itself)’, as Sellars (Reference Sellars1968: 40) puts it, would still have to be recast in merely problematic or hypothetical terms as a conceptual distinction. However, there is reason to believe that the entirety of the fourfold transcendental distinction is a meta-metaphysical thought experiment and was so regarded by Kant himself. For while the role assigned super- and extra-sensible noumena may well suggest a merely problematic embrace of something resembling the two-world metaphysics of Christian Platonism, it is far likelier that these two, along with both other senses of ‘noumenon’, all figure in a thought-experimental adaptation of a different, medieval Christian onto-theological creation story, versions or vestiges of which are found in St. Augustine (see n. 11), St. Thomas, Descartes, Leibniz, and Malebranche.Footnote ²²

For present purposes, a schematic and approximative account of the relevant aspects of this medieval Christian creation story will suffice. It provides for as many as three modes of existence of the same natures or things (res). The first is (i) eternal existence coram God’s infinite intellect, which belongs to all possible finite natures or essences. Kant’s concept of God’s intellectus archetypus (A695/B723; cf. ID, 2: 397) attests to his familiarity with this medieval notion of God’s exemplar causality: as ens realissimum, God intuits from eternity the natures of all possible finite things by reflecting on his own infinite nature in which they are all contained to the highest degree of perfection. Second, there is (ii) realitas formalis sive actualis or temporal existence in rerum natura. Kant’s concept of God’s intuitus originarius derives from this old Augustinian notion of the efficient causality by which the divine intellect gives some of those things existing in it from eternity temporal existence outside it in space as well. This further mode of existence obviously belongs only to those finite natures that God makes actual, and only for as long as they are maintained in existence by the efficient causality of God’s creative intellect.Footnote ²³ Third and finally, there is what Descartes and the Scholastics called (iii) realitas obiectiva (in a sense just the opposite of Kant’s ‘objective reality’), that is, intentional or cognitive or ideal existence of some of those same natures in the human mind, whenever, and for as long as, represented there. Of course, many natures that are (i) eternally present to the divine mind, like many that are also (ii) present in the spatiotemporal world of sense, never come to be (iii) re-presented in any finite human mind at all; but while such natures are numerically distinct from any that appear to us, others have temporal existence in the human mind as well as outside it, that is, all three modes of existence.Footnote ²⁴

In adapting the philosophical creation story of dogmatic rational theology, Kant reduces its real to what the Scholastics would have regarded as conceptual distinctions (distinctiones rationis sine fundamento in re). Instead of three ways in which certain natures or essences actually exist, Kant entertains the hypothesis of three non-spatiotemporal ways in which they can be considered as existing, be it in addition to their spatial and/or temporal existence in the human mind as inner or outer appearances, or be it without appearing at all. To begin with, all possible finite natures may be considered as (i) extra-sensible noumena, that is, natures belonging to a mundus intelligibilis existing eternally coram intuitu intellectuali Dei, God himself being (not an extra-, but rather) a super-sensible noumenon or ens realissimum existing outside space and time.Footnote ²⁵ Now some of those natures existing eternally in the divine mind may be considered as also having (ii) contingent existence outside it in the created world order. But here Kant’s thought experiment deviates from all traditional philosophical creation stories. Given his doctrine of space and time as pure forms of our sensible intuition, the contingent existence of created natures outside the divine (but, ex hypothesi, not in the human) mind cannot be existence in rerum natura in the customary sense of belonging to the spatiotemporal mundus sensibilis; the realitas formalis of those things that God makes actual must rather be noumenal or transcendentally real existence, that is, existence as part of the created mundus intelligibilis or non-spatiotemporal world behind the sensible appearances. And just as not everything that exists from eternity in God’s mind (as an extra-sensible noumenon in the positive sense) ever becomes actual outside it (as a noumenon in the negative sense), so not everything that comes to exist in the latter way ever acquires (iii) realitas obiectiva or existence in the human mind. Of those natures that do, however, some acquire existence in space and/or time only as objects existing (iiia) coram the human faculty of sensory intuition (appearances as the object of mere sensation), while others acquire spatial and/or temporal existence as objects existing (iiib) coram both the sensory and intellectual faculties of the human mind (appearances as the objects of isolated perceptions, where not all of the categories are applied), or (iiic) coram both the sensory and intellectual faculties, but where all the categories are applied, that is, appearances in the full sense of empirically real objects of connected human experience or empirical cognition. Now this last mode of existence of the objects of outer intuition in the human mind is (so Kant) tantamount to existence outside it in the empirical sense of ‘outside’ (cf. A373), that is, existence in rerum natura as Kant understands it. Finally, for purposes of engaging with Leibnizian metaphysics, Kant entertains a further mode of existence in the human mind: existence (iiid) coram the pure unschematized categories alone. Like noumenal existence in the negative sense, this third mode of non-spatiotemporal existence of the very things that appear (along with others, like immaterial substances, that do not or cannot appear) represents a clear departure from anything in the traditional creation story. To say that in Kant’s transcendental distinction, broadly understood, certain formerly dogmatic or real become hypothetical or conceptual distinctions is just to say that they no longer form part of a metaphysical doctrine or philosophical creation story, but figure in a meta-metaphysical or onto-theological thought experiment.

If this medieval Christian philosophical creation story is indeed the historical backcloth before which the thought-experimental character and detail of Kant’s fourfold transcendental distinction become comprehensible, then all the standard objections repeatedly brought against the latter as an incoherent metaphysical doctrine are entirely misplaced. Certainly, the metaphysical density of the original use of ‘onto-theological’ to characterize the Continental-European mainstream in philosophy (cf. Heidegger Reference Heidegger and Heidegger2006: 51–79) is attenuated by the merely hypothetical and meta-metaphysical character of Kant’s fourfold distinction; and yet the latter is sufficiently metaphysically oriented to render the present interpretation proof against dismissal as ‘innocuous’ (Strawson Reference Strawson1966), ‘anodyne’ (Guyer Reference Guyer1987), or ‘deflationary’ (Langton Reference Langton1999), on the grounds that the historical Kant was above all else a reforming (and himself a reformed) metaphysician rather than epistemologist. Kant remained scrupulously aware of the limits the Critique had placed on first-order metaphysical cognition as he set about making ‘room’ for other modes of existence besides the phenomenal, the latter being the only mode of existence actually asserted. It is not surprising that he found himself constrained to employ much of the conceptual apparatus of the very (not just Leibniz-Wolffian, but Augustinian and Thomistic) metaphysics he was bent on ousting in his effort to usher in a new, critical metaphysics both of experience (nature) and of morals. To conclude, a rapid glance at the metaphysics of morals, but first at the textual grounds for assigning Kant the ‘disastrous model’ of human cognition with which Strawson saddles him.