DENNIS SCHULTING | Kant’s Radical Subjectivism. Perspectives on the Transcendental Deduction | Palgrave Macmillan 2017


By Dennis Schulting

Robert Watt has provided an excellent précis of the main theme of my book, namely Kant’s radical subjectivism, for which I am extremely thankful. I could not have written a more succinct summary that captures the essence, give or take a few details, of what I take to be Kant’s metaphilosophical stance in the Transcendental Deduction (henceforth simply ‘the Deduction’). So I’m not going to repeat here in my own words what Watt wrote. Rather, what I am going to do is respond to the lingering questions that Watt has, in particular, as to (1) how I see the issue of “epistemic humility” and how this ties in with my thesis of radical subjectivism, and (2) why I think objective validity is not a feature of intuitions, making my position on nonconceptualism vulnerable, in Watt’s view, to being nothing more than a closet conceptualism. Another issue that Watt raises concerns what is probably the most controversial aspect of my reading of the Deduction: namely (3) the contentious claim, which I defended at length in my previous book (Schulting 2012, henceforth KDA), that the categories are all a priori derived from the principle of apperception. Despite Watt’s serious reservations and Corey Dyck’s and Andrew Stephenson’s earlier misgivings, as well as Thomas Land’s doubts about this claim (Dyck 2014; Stephenson 2014; Land 2018; cf. by contrast Quarfood 2014), I remain firmly committed to it as what is in my view the only way to explain the systematicity claim that Kant makes with respect to the two tables, as well as the only way to understand what I call the reciprocity claim that is central to the Deduction (following Henry Allison). I realise I’m perhaps a lone voice in this in current Kant scholarship, but I take comfort in the knowledge of being in the august company of Klaus Reich and Michael Wolff, who both defend the idea.[1]

There is another question that Watt poses in the context of the humility issue (point 1). He asks whether my thesis of radical subjectivism in Kant is itself perhaps not so radical, since P.F. Strawson already proposes something like my thesis of radical subjectivism. Though there is a certain likeness between my reading of Kant and Strawson’s, there is also a clear difference, which in fact concerns an elemental aspect of the thesis of radical subjectivism as I construe it. I shall tackle this first in Section I. I then address, in Section II, the issue of “epistemic humility” (1). In Section III, I discuss the question about objective validity and nonconceptual content (2). Lastly, in Section IV, I confront Watt’s alternative reading of the passages that seemed, to him, not to support my reading of the derivation claim (3).


P.F. Strawson’s The Bounds of Sense: An Essay on Kant’s ‘Critique of Pure Reason’ is undoubtedly one of the most philosophically insightful works on Kant’s First Critique, and no Kant scholar can afford to ignore it, even if it is common knowledge that Strawson’s book is not a straightforward scholarly work in the secondary literature on Kant, and may be flawed as not living up to that standard (cf. Allais 2016). When I started my Ph.D. back in 1999 at the University of Warwick, England, Henry Allison’s Kant’s Transcendental Idealism, Lorne Falkenstein’s Kant’s Intuitionism and Strawson’s classic alongside a copy of the then new Cambridge translation of the Critique of Pure Reason were the first books I purchased, if my memory serves me right, in the university bookshop on campus. Not much later, I was very pleased to find a fine hardback copy of The Bounds of Sense in its second printing from 1968 in a dank cellar of a second-hand bookshop in Charing Cross; the still palpable musty smell that the book gives off makes the book seem even more important. The book has intrigued me ever since, no doubt like many others; my first serious, though needlessly difficult, published article was on Strawson, which was part of the proceedings of the ‘In Memory of Strawson’ conference that Lucy Allais organised in Johannesburg a year after Strawson had died (Schulting 2008).[2] Unfortunately, the one lecture—in my time in England at least—that Strawson was scheduled to give at Senate House, London, in 2004 was cancelled and so I missed the opportunity to see Strawson talk in person.

In The Bounds of Sense, Strawson aims to show that there is a “direct analytical connexion” (1968:96) between experience and objectivity, thus revealing the “self-reflexiveness” of experience. Self-consciousness is possible only if we can differentiate our experiences from the object of our experience—this is an intrinsic element of the possibility of experience that provides room for the thought of experience, i.e. the very idea of self-consciousness as a reflection on one’s experience. Objects in the world, on the other hand, which have an existence independently of experience, must be able to be recognised as existing within a unified framework of reference in order to count as objects of experience. Strawson’s main point is that we can only be self-conscious, be aware of our experiences because we are conscious of objects outside us, while we can only make sense of objects in virtue of our capacity to conceptualise them; the self of experience and the object of experience are reciprocal elements of the same “limiting framework of all our thought about the world and experience of the world”, which dictates what we can conceivably know about objective reality. There is a close resemblance to Kant’s argument in the Refutation of Idealism. But Strawson takes the self-reflexiveness argument to be the core argument in the Deduction as well, however without the alleged excess of transcendental idealism.

And here’s the rub: Strawson’s construal of the analytical connexion between experience and the object of experience is based on an analysis that famously does without a priori synthesis and is disentangled from what he regards as the problematic theory of phenomenalistic idealism (cf. Allais 2016). Most importantly, Strawson denies any role for a priori synthesis, which I take as absolutely integral to Kant’s radical subjectivism. I take Kant to argue that objects are first constituted by the unity of apperception in virtue of a priori synthesis. In other words, it is not the experience of an objective world that determines the course of the experience itself, but rather the other way around: we are able to experience an objective world because we apperceive, in virtue of an act of a priori synthesis, our representations as being constitutive of an objective world. The determinate course of our experience is not grounded on a conception of the objective, but our conception of the objective is first established by an act of synthesis among our representations; and it is this that first grounds both experience and the object of our experience, i.e. the framework of intelligibility that Kant calls possible experience (cf. A110). My thesis of radical subjectivism is therefore not just the analytical claim about the fact that “a certain minimum structure is essential to any conception of experience that we can make truly intelligible to ourselves” (Strawson 1968:11). This intelligibility structure must be seen as grounded in the a priori act of synthesis of a subject’s representations, which Strawson famously rejects as belonging to the “imaginary subject of transcendental psychology” (1968:97); reading Kant in this way, as I do, makes him more radically subjectivist than on Strawson’s analysis of a necessary conceptual structure outside of which we cannot intelligibly conceive of objects.

Strawson’s interpretation of Kant’s phenomenalism is not easy to gloss. But in his interpretation of Kant’s idealism, Strawson is indeed close to mine—this is a closeness that I don’t think Watt had in mind in suggesting that my interpretation of radical subjectivism is very similar to Strawson’s. Strawson interprets Kant’s idealism, rightly I think, as a form of phenomenalism, but in the wrong way, namely as saying that the real, mind-independent objects of experience are mere sensations and thus do not exist outside us (empirically). In other words, Strawson reads Kant as an ordinary Berkeleian. It is not surprising that Strawson has no truck with idealism (he talks about the “doctrinal fantasies of transcendental idealism”; 1968:51), since he construes it, at first blush at least, in the most uncharitable way possible. Take these passages for example:

Spatially and temporally ordered items are not the affecting objects in question, but only, at most, their effects, appearances they present to beings equipped, as we are, with spatial and temporal modes of sensible intuition. Here is the full subjective force of the doctrine that space and time are forms of such intuition. (1968:53)

Space and time are both alike assigned to the subjective constitution of our minds, and all spatially or temporally ordered items occurring in our experience are declared to be merely the outcome of that constitution’s being affected by objects as they unknowably are in themselves. (1968:54)

The doctrine is not that, by affecting our cognitive constitution, things (including ourselves) as they are in themselves produce two distinct kinds of existences, viz. bodies in space (and time) on the one hand and temporally ordered states of consciousness, including perceptions of those bodies in space, on the other. Rather, all the actual effects of these transactions between things in themselves are temporally ordered states of consciousness; but these include […] states of consciousness that we rate as perceptions of bodies in space. (1968:56)[3]

But bodies in space are appearances in a much stronger sense. They are not even effects of things as they are in themselves. It is simply that among the effects of things as they are in themselves are some states of consciousness which we are constrained to regard as perceptions of bodies in space; and apart from these perceptions bodies are nothing at all. (1968:57)

To be sure, Strawson doubts the philosophical validity of these claims, and he therefore dismisses transcendental idealism. But what is striking is that interpretatively Strawson is, in my view at least (and I’m probably in the minority, see e.g. Lucy Allais [2015, 2016], a student of Strawson’s, who vehemently opposes Strawson’s reading[4]), absolutely correct, though he doesn’t provide textual evidence (close to a sin, if he were to have published today, as Allais [2016] points out) and he mistakes Kant’s statements for statements about a particular subject’s states of consciousness, rather than a transcendental subject’s states of consciousness. The mistaken assumption behind Strawson’s dismissal is that things in themselves are the very objects that we experience as ordered in space, and since on Kant’s transcendental idealism this is denied, we must reject transcendental idealism, according to Strawson. This brings me to the question of “epistemic humility” that Watt poses.


The correctness of Watt’s portrayal of my thesis of radical subjectivism notwithstanding, there is one worry that I have with it, and this relates directly to his question about what he refers to as “epistemic humility”. He says that according to radical subjectivism, “Kant thinks that there is only one conceivable reality, i.e. reality according to our framework, and this is a reality of which we can have knowledge”. The problem here is the meaning of “conceivable”. The conceivability or intelligibility condition is of course a condition on what we human beings, as spatiotemporally located, and equipped with the ability to discursively run through manifolds of representations, are capable of knowing. We cannot conceive of a reality that we can know other than the “reality according to our framework”, but we can perfectly conceive, in the strictly logical sense of the term, of a reality outside that framework, which however is to no avail since such conceiving is theoretically at most purely speculative, i.e. not objectively valid. So the second part of Watt’s sentence is an important condition of radical subjectivism. This is close to Watt’s third option: we cannot have a ‘positive’ notion of a reality outside our framework—I would prefer not to call it ‘categorial’, since the framework includes the constraints of spatiotemporality, while our categorial framework strictly speaking goes beyond what is spatiotemporally knowable (see A96; B148)—but we can have a negative conception of such a possible reality, in line with Kant’s distinction between positive and negative noumena.

I should like to emphasise, as this was not entirely clear from what Watt takes me to argue, that when I talk about objective reality this is the reality of the spatiotemporal, empirical objects of law-governed nature (see in particular Chapter 7 in Kant’s Radical Subjectivism, KRS for short), not the reality of things in themselves. This objective reality is wholly constituted by our categorial framework, given sensory input; even matter is part of this categorial framework cum sensory input, i.e. our sensations (in this sense I’m close to Strawson). Both categorial form and the concrete matter of objectively real objects form part of possible experience. Possible experience, on Kant’s view, is not just how we experience objects, but also how objects are constituted such that we can experience them (A111; A158/B197). We do not have to be “epistemically humble” vis-à-vis our knowledge of objects and how “they really are”, for we are intimately connected with them. This is part of what I meant by the phrase “radical subjectivism”, and what Watt correctly characterised as saying “that the objective facts are constituted by the facts according to our framework”, and “cannot outstrip the subjective facts”. Objective reality is everything that we are able to know, given the properly functioning sensory and discursive capacities that we have. Our claims about the world for which we have empirical evidence, match the world. Of course we could be mistaken about this or that fact, but this is an empirical issue that doesn’t impact on the transcendental truth as correspondence between mind and world.

But this truth about radical subjectivism as constitutive of objective reality has a flip side. Though our categorial framework is not ipso facto bound by the constraints of sensibility—as said, Kant allows the conception of objects that go beyond experience, whether they be imaginary or impossible objects—it is of course restricted by the laws of human discursivity, the fact namely that we have to run through manifolds of representations and take these together as falling under a higher concept (A68/B93); as per the laws of discursive, basically Aristotelian, logic we cannot just apply a concept to an object, since a concept is not immediately related to objects; we need intuitions consisting of manifolds of representations, regardless of the fact that these intuitions are spatiotemporal in nature. This was the argument in Chapter 9 of the book, where I argued in some detail that our categorial framework cannot properly think things in themselves defined as thoroughly determined individuals, whereby I made the distinction between, on the one hand, entertaining the notion of <thing in itself> and knowing how to use it in such true analytical judgements as “every thing in itself is non-spatiotemporal” (analytic on account of the thesis of transcendental idealism), which is perfectly possible, and, on the other, to have a genuine intelligible grasp of a thing in itself as a thoroughly determined individual, which is impossible. The logic of discursive judgement is such that the objectively valid determination of any object=x underlying the subject concept of an objective judgement—a judgement whereby reference to the x is not otiose—is limited to the (potentially infinite) limitation of the material of possible predicates that can be affirmed (or negated) of x. The determination of a thing in itself is prohibited as a result of the discursive nature of our judgement, for:

A thing in itself is a thoroughly determined individual, namely an individual that has been determined to have all its properties;

A thoroughly determined individual ex hypothesi cannot be further determined as to which properties it has or has not;

A judgement is by definition a determination of predicates (or properties);[5]

Therefore, a judgement cannot determine a thing in itself.

This, in short, is the argument for idealism from discursivity, i.e. the “epistemic humility” claim that directly follows from the mode of our (human) capacity to judge. It is independent of the claims about space or time, of which Kant says in the Aesthetic that they are not properties of things in themselves.[6]

Does this mean that we cannot know objects as “they really are” after all?[7] Well, this question is confused. It is ambiguous about the difference in meaning between the terms ‘object’ and ‘thing’. It assumes a conflation of things in themselves and objectively real objects as they are supposedly apart from our experience or judgement of them. As I said earlier, objectively real objects are not apart from our experience or judgement of them; there is nothing about objectively real objects that we cannot, in principle at least, capture in judgement. There is no discrepancy between the objective and the subjective facts. But things in themselves are not the constituents of objectivity; they fall outside its scope for the reasons given above.

Of course, there is a relation (or relations) between things in themselves and the objects of our judgement, i.e. the thing’s appearances, but that is a different, complex story concerning the modality in which objects exist as things that I can’t repeat here. One thing that I have tried to point out in KRS is that the relation between things in themselves and the objects of our judgements, i.e. appearances, is asymmetrical, given the definition of a thing in itself and the nature of judgement. No arbitrary object of a determining judgement or even series of objects of determining judgements can ever correspond one-to-one to a thoroughly determined thing. Another thing, which I hinted at in my account of Strawson’s phenomenalist interpretation of Kant’s idealism above, is that the objectively real objects are wholly on this side of experience; there is nothing about them that is outside our experience, outside representation—of course, this is seen from the transcendental perspective or from within possible experience, but seems obvious as the distinction between things in themselves and appearances is a transcendental distinction anyway. And a third aspect is the fact that in judgement we do posit a thing as existing (in itself)—this concerns modality, which as Kant makes it clear, concerns the relation of a thing (that exists in itself) to our cognitive capacity (B99–100), but it says nothing determinate about that thing as such (qua existing in itself). But, insofar as I haven’t touched upon it in KRS (see esp. Chapters 1, 4.10, 7.8 and 9), I have to leave this discussion for another time.


Watt says that my account of empirical intuition “needs … more work”. Let me first state that in his portrayal of my position, Watt says that empirical intuition is not “a conscious representation”. I never say this. In fact, I believe that any representation must be minimally conscious; not in the sense of a “consciousness of” or a “consciousness that” (A103), but in the sense of a representation having an intensity. I have argued this before in KDA and extensively in a more recent article on the roots of Kant’s notion of apperception and consciousness in the Lectures on Metaphysics. I believe that Kant, like Leibniz, held the view that on account of the Law of Continuity no representation can be literally unconscious, i.e. have an intensive magnitude=0 (see Schulting 2015). So a mere empirical intuition, on my reading, is not unconscious in the above sense, unless it is inexistent.

Secondly, reference is often made to the Stufenleiter by those who claim that intuition has a separable status that legitimates seeing an intuition as a kind of objective cognition that is not a cognition of the categorial sort. Basically, Watt thinks that intuition is a “thicker” sort of representation than I take it to be. And here he thinks I need to do more work because I do not make a distinction between manifolds of representations and intuitions (as such). I don’t see why I should because I do not subscribe to the view that intuition is a “thicker” sort of representation. I referred to the Stufenleiter in a note (pp. 240–1n.3), and said that mere intuition is not really a cognition (Erkenntniß), as Kant explicitly says in Metaphysik Volckmann: “the mere intuition is not knowledge [Erkennntniß]” (AA 28:404); usually Erkenntniß is translated as ‘cognition’, so a mere intuition is not cognition. When Kant says in the Stufenleiter that both concept and intuition are “objective perceptions”, this is meant so as to differentiate them from subjective perceptions, namely perceptions as mere modifications of the self (mere sensations); only perceptions as directed at objects are called “cognition”. But this classification should not be taken as if they corresponded to neatly categorisable real separate entities or faculties, which is what nonconceptualists often do; there is no in-between entity or faculty called ‘mere intuition’ that is more than a mere representation or series of such representations and less than a fully-fledged conceptualised objective perception. I think such a categorisation, which nonconceptualists often appeal to, is wrongheaded (see also the rightful critique of such a categorising strategy by Conant 2016).

Intuition is of course cognition only in conjunction with concepts (A51/B75). Now I readily grant that an intuition in conjunction with a concept is a thicker notion than a mere representation or, more precisely, an aggregate of such representations, but there is nothing that differentiates a mere intuition from an aggregate of mere representations (sensations); note that though Kant says in the Stufenleiter that sensations are modifications of the mind, this does not mean that in an empirical intuition of an object no sensations are involved: for, first, all outer appearances are contained in inner sense, and, secondly, sensations, which presuppose the presence of an object (which could be one’s body), are the material content of any empirical intuition (A50/B74). Does this mean that a mere intuition, i.e. an intuition that is not categorially determined or conceptually laden, does not represent the present object? No, but here, as to what this representation of the present object is supposed to amount to, I differ fundamentally from Watt (and most, if not all, nonconceptualists), who to me seems to espouse too strong a conception of objectivity for mere intuition. Watt says he prefers to believe that Kant has “two conceptions of objectivity, and two conceptions of consciousness” (the first option he mentions). I don’t believe he has, and I don’t see any textual evidence for it. My take on it is closer to Watt’s third option: “intuition has different properties depending on whether it is considered in isolation from the faculty of understanding”, although I wouldn’t formulate it thus. The second option he mentions can be dismissed, I think.

Watt argues that also intuitions have object-intentionality. I should first like to mention that I may have been somewhat sloppy in using the terminology of ‘intentionality’ in Kant’s context, a term that Kant never uses (I briefly note this at p. 189n35), without making explicit what I mean by it. Still, I think the terminology of object-intentionality can be used in the Kantian context as implying, on account of the transcendental turn, some sort of subjective directedness to the object, and I deny that in a mere intuition (without any conceptuality involved) there is such a subjective directedness to the object, whereby ‘subjective’ should be read sensu stricto: from the perspective of a subject aware of the intention towards the object. There is no such agency involved in mere intuition; the subject is purely passive in mere intuition, at least from the Kantian transcendental perspective (on a more phenomenological account one might want to argue differently). And there is no determinate object in intuition, to which an agent could be subjectively directed.

Objectivity is strictly a function of transcendental apperception, given that Kant defines an object as “that in the concept of which the manifold of a given intuition is united” (B137) in virtue of an act of apperception, and he considers the objective unity of apperception as defining judgement (B141).[8] The relation to the object in an intuition is, as Kant often says, immediate, and there is therefore no mediation by the apperceiving subject who is aware of the intentional relation between himself and the object. So I don’t deny that Kant “think[s] that [intuitions] put us into some sort of basic connection with objects”, in fact I pointed out that the three criteria for intuition are singularity, immediacy and dependency (p. 247n40), and make a point of the fact that nonconceptualists are right to claim these criteria uniquely differentiate intuitions as independent from the understanding. The immediacy of intuition is an important differentiating characteristic of intuition: it implies that the representational apprehension of the object and the object of representational apprehension collapse, or, to put it differently, the reference to the object is ‘immersed’ in the manifold, as Hegel would say, and has not yet explicitly come to the fore—this happens only when the subject takes a conceptual distance with respect to the object. The explicit subject-object distinction only appears in a conceptual, mediate relation to the object; before that, there is no explicit subject nor an explicit object, in the transcendental sense of those terms, and so also no object-intentionality.

Watt refers to the well-known example of the ‘savage’, mentioned in the Jäsche Logic. I myself referred to an animal’s intuitive sense of its spatial environment (p. 308ff.), which is a less complex example, since animals of course don’t use concepts, while they can still perfectly engage with spatially located objects in their surroundings, as I explained (see also Schulting 2015). The case of the savage is somewhat more complex, as Kant does acknowledge some sort of potential, on the savage’s part, for the use of concepts. I discuss the savage example in a critique of Pippin in Schulting (2016:244n20). Without going into detail here, I certainly don’t deny that the savage has or can have an occurrent representation of a house, and is thereby aware of the object that is a house, without needing concepts (e.g. of a house) to do so (whether he needs to use categories is another matter, perhaps, though he might certainly see the house without them).[9] In my previous book (KDA, chapter 8), I argue extensively, against a whole tradition of Kant interpretation, that the categories and transcendental apperception are not conditions on mere consciousness, and so neither on mere intuition (see also Schulting 2015).

Watt turns to my account of objective validity, and asks if my conception of it as not to be conflated with the truth value of a judgement holds water. If I were to drop this reading, I could then, he says, endorse the reading that in one sense a (non-conceptually laden) empirical intuition is not objectively valid, namely as having a truth value, and in another sense it is objectively valid, namely as being an objective perception as Kant defines it in the Stufenleiter—just as Kantian nonconceptualists argue. But I don’t think stretching the meaning of Kant’s terminology of ‘objective validity’ to cover mere empirical intuition as the representation of an object is warranted. As I argued in KRS, objective validity is a function purely and only of judgement. I cannot here rehearse all of the detail of my account in KRS, so let me just respond to Watt’s objections.

In his criticism of my view that the objective validity of a judgement is not the same as a judgement having a truth value, because not all judgements are objectively valid—i.e. analytic judgements have a truth value, but are not objectively valid (see the discussion in KRS, chs 3 and 4)—Watt cites a Prolegomena passage, where Kant identifies objective validity with universal or general validity (Allgemeingültigkeit). Watt reasons that because analytic judgements are generally valid (i.e. for everyone), they are therefore objectively valid. But this is a bit too quick. He seems to read Allgemeingültigkeit as logical validity, an assumption that he cannot back up with textual evidence.

First, the discussion in the section in the Prolegomena as well as in §19 in the B-Deduction that Watt cites, concerns the problematic of the possibility of having a cognition of an object—note that this comes after Kant has given the definition for object in §17; the discussion in the Prolegomena, and the parallel one in §19 in the B-Deduction, concerns the contrast between judgements of perception and judgements of experience (which was abandoned in §19 after criticism from Schulz). In the section in the Prolegomena that Watt cites (§19; at the outset), right after he says that “objective validity and necessary universal validity (for everybody) are equivalent concepts”, Kant says that

when we consider a judgment as universally valid, and hence necessary, we understand it thereby to have objective validity. By this judgment we cognize the object […] by the universally valid and necessary connection of the given perceptions;

and he makes it clear in the following that it concerns here empirical judgements about objects of sense:

[W]hen by the concept of the understanding the connection of the representations of the object, which are given by the object to our sensibility, is determined as universally valid, the object is determined by this relation, and the judgment is objective. (Prol, AA 4:299 [Kant 1977:42]; emphasis added)

This, then, is the context of Kant saying that objective validity and necessary universal validity are equivalent. Analytic judgements don’t have a role to play here, as analytic judgements do not concern the cognition of an object, least of all a sensible object. For analytic judgements any reference to an object, an underlying x, is otiose. Moreover, if indeed Kant’s Allgemeingültigkeit, or Gültigkeit für jedermann, were logical validity in the sense of having a truth value, then the whole transcendental project of differentiating synthetic a priori judgements from analytic judgements so as to be able to prove the validity of the pure concepts used in the former kind of judgement, would appear to come to nought; for if objective validity held for both kinds of judgement, which criterion would procure us a proof of the validity of employing those pure concepts used in synthetic a priori judgements? Allgemeingültigkeit is not used by Kant to indicate universal logical validity (which obviously holds for everyone as well), but to contrast it with the subjective validity of associative relations between representations that have no basis in an object outside these representations, and therefore do not hold for everyone.[10] This contrast is irrelevant to analytic judgements.[11]

Second, though Kant sometimes identifies objective validity with truth, in the cases where he does so it concerns empirical cognition of an object, not just a statement possibly being true or false, in other words, correspondence to an object and not logical truth (e.g. B816/A788[12]; A125). Van Cleve understandably wondered whether Kant meant by the objective validity of a judgement that the judgement would perforce be true, which is what seems implied by his statements in §19 (given what he says about the copula in judgement), but of course, as Van Cleve also rightly asserts, this can’t be the case (see my account in Chapter 4; cf. Chapter 3).

But if objective validity is not concerned with a judgement being either true or false, there is no problem with Kant stating that a judgement’s objective validity lies in the copula of a judgement “This body is heavy”, say, for objective validity concerns here, as I delineated in KRS, the original relation or intentionality towards the object of my judgement, regardless of the question whether the judgement is empirically true or false (that is, whether this particular body has a particular weight). The objective validity concerns the necessary unity among the representations in a judgement, i.e. its form, whereas the content of a judgement is empirical and contingent;[13] this includes the empirical predicates that are attributed, truly or falsely, to an object of judgement. I can falsely attribute a certain predicate to an object in any arbitrary judgement P, but the necessary unity among or objective validity of the representational constituents P is invariant, for they are necessarily “combined in the object” in order to be about the object. Identifying objective validity with truth value, as does Watt like most commentators, makes it hard to understand how a judgement, on Kant’s definition in §19, can be false, that is, not correspond to, or for its representations not to be “combined in”, the object, precisely why Van Cleve was puzzled about Kant’s notion of objective validity (I address Van Cleve’s legitimate worry, which Watt doesn’t consider, in detail in Chapters 3 and 4).

When Watt refers to §19 in the B-Deduction he uses the analytic judgement “A body is extended” (which is of course to make his point about objective validity also being a characteristic of analytic judgements), but this is inapposite in the context of Kant’s discussion there; it is clear that for Kant the objective validity of a judgement has to do with how representations are necessarily “combined in the object” (B142). See also B137, where Kant clearly associates objective validity with the relation to the object:

Consequently the unity of consciousness is that which alone constitutes the relation of representations to an object, thus their objective validity, and consequently is that which makes them into cognitions…(emphasis added)

Clearly, in the analytic judgement “Bodies are extended” the concepts <bodies> and <extended> are not combined in a particular object, as reference to any object=x is otiose; the relation between those two concepts can be determined purely logically, by means of an analysis of the concept of <body>.[14]

Third, Kant sometimes talks about objective reality (B194/A155; A109; cf. FM, AA 20:279–80), where it is clear from the context he is talking about that which he elsewhere calls objective validity; at B44 and B52, in the Aesthetic, he in fact identifies (empirical) reality with objective validity. This makes it clear that objective validity doesn’t concern the truth value of a judgement. Objective validity concerns reale Möglichkeit, not just logical possibility (Bxxvi n; cf. A89–90/B122; A788/B816). Clearly, analytic judgements have little to do with objective reality, so on Watt’s standard reading of objective validity, analytic judgements would also presumably be objectively real, which is absurd, or else he would have to explain why Kant sometimes identifies objective validity with objective reality (see also FM, AA 20:279).

Fourth, if objective validity were a characteristic of analytic judgements as well as synthetic judgments, the sense in which the synthetic a priori enables us to establish the objective validity of the categories, namely that categories objectively refer to real objects, would be lost. Part of the claim of my thesis of radical subjectivism concerns the fact that transcendental apperception cum sensory content first constitutes the objective reality or validity of objects of experience, i.e. the spatiotemporal objects themselves; on this reading, the objective validity of my judgement is not just a logically valid judgement, necessarily acceptable to all, but is a judgement about objects or about the fact that I have knowledge of these spatiotemporal objects. If analytic judgements were objectively valid in the sense Watt thinks, it would jeopardise the radical subjectivism thesis that Watt endorses, for that thesis relies on a reading of objective validity as saying something about the real object outside my judgement (empirically speaking).

Fifth, if objective validity were a characteristic of analytic judgements, we would also lose the sense in which the synthetic a priori is supposed to solve the conundrum of the definition of truth as a general, universal criterion which should at the same time say something about particular objects, i.e. a criterion of truth that is both sufficient and universal (A58–9/B83). If objective validity were a characteristic of analytic judgements, for which the reference to an object is otiose, objective validity would not provide us the correspondence to real particular objects. And objective validity is precisely what is at the centre of Kant’s endeavour to prove the rightfulness of the employment of pure concepts in judgements about real particular objects. Watt’s standard reading of objective validity would appear to conflict with this endeavour.

I’m not the only nor the first one to note that analytic judgements are not objectively valid or that there are true judgements which are not objectively valid. The best (recent) account is by Alberto Vanzo (2012), who concludes that Kant “did not equate objective validity with truth-aptness”. Analytic judgements are governed by transcendental apperception and thus, in a roundabout way, by transcendental truth, for we can only identify judgements as being strictly analytical because we know how to differentiate them from synthetic judgements. Transcendental truth is thus the condition of all truth[15]. But logical truth as well as empirical truth must be differentiated from transcendental truth (which is the necessary condition of the former), and logical truth, or more precisely truth aptness, has nothing to do with objective validity per se, for, as was said before, reference to an underlying object (or connectedness in the object, as Kant says) is otiose for the analysis of logically valid relations (true or false statements or propositions).

Analytic judgements are a special form of judgement, which abstract from the reference to an underlying object. I do not need to refer to an underlying x as the connecting factor between two related concepts. I can look purely at the analytical relation between two concepts in an analytic judgement in order to see if the judgement is true. All judgements are governed by the principle of logical validity (they have a truth value), but not all judgements are objectively valid. Analytic judgements aren’t; if they were, Kant’s distinction between analytic and synthetic a priori judgements, necessary to establish the truth of claims that aren’t analytic claims, would be completely useless. Thus, not only objectively valid judgements are truth-apt, and truth-aptness does not eo ipso include objective validity. Let me conclude this section by quoting Vanzo:

If only objectively valid judgements were truth-apt, Kant should make statements along the following lines:

an analytic judgement is true if and only if its negation violates the law of contradiction and it is objectively valid;

an affirmative analytic judgement of the subject-predicate form is true if and only if its predicate term expresses a constituent concept of the concept expressed by the subject and it is objectively valid.

However, Kant never includes objective validity among the conditions of truth for analytic judgements. […] Kant claims that conformity to the law of contradiction is sufficient to guarantee the truth of every analytic judgement (A151/B191), thus including the analytic judgements that are not objectively valid. Moreover, God is an eternal, atemporal being is a true analytic judgement (A641/B669; KpV, 5:123), yet it lacks objective validity. (Vanzo 2012:115–16)


The third worry that Watt expresses is not so surprising, as it concerns my most controversial claim first presented in KDA and briefly revisited in Chapter 2 of the book here under discussion, namely the claim about the derivation of the categories from apperception. The chapter that Watt discusses was my response to earlier critics of the derivation thesis. Watt addresses mostly only what I say in that chapter, but the full account is of course to be found in my previous book (KDA).[16]

Kant accused Aristotle of “round[ing] up [the categories] as he stumbled on them”, because “he had no principle” (A81/B107). Hegel accuses Kant of the same. Now, as I pointed out in KDA, either

(1) Hegel was right about accusing Kant of equally failing to “systematically generate” the system of the categories “from a common principle, namely the faculty for judging” (A80–1/B106), because Kant did indeed not seek to systematically deduce the categories from a principle, and rested content with the classification as he enumerated them in accordance with the way this had been done in classical logic since the days of Aristotle, but should have sought to systematically deduce the categories from a principle


(2) he was wrong to accuse Kant of this because Kant was entirely right not to seek to systematically deduce the categories from a principle, and to rest content with the classification by enumerating them in accordance with the way this had been done in classical logic since the days of Aristotle, for the ways of judging and categorial determining are just our ways of so doing and there is no deeper ground that explains why this is the case.

(1) is what Hegelians such as Stephen Houlgate (2006:15–16) say is the case. But if (2) is true, then Kant’s accusation of Aristotle of “haphazardly search[ing] for pure concepts, of the completeness of which one could never be certain, since one would only infer it through induction, without reflecting that in this way one would never see why just these and not other concepts should inhabit the pure understanding” (A80–1/B106–7) is odd to say the least, for he would blame Aristotle for the failure that he is blameworthy for too. There is a third possibility, the one I argued for:

(3) Hegel is wrong because Kant did not fail to systematically generate the categories from a common principle, namely the faculty for judging.

Watt appears to think that we should read the derivation (or deduction) of the categories in line with the reason given for rejecting Hegel’s accusation (2), but we are then owed an account of how to explain Kant’s criticism of Aristotle.

Let me first note that Watt speaks frequently of “forms of judgement”, while Kant actually speaks of finding the “functions of unity in judgement” or “logical functions of […] possible judgments” (A69/B94; A79/B105), so functions instead of forms. This is important, as Kant’s point is to show that the categories are nothing but these functions of unity in judgement under a specific point of view, namely with a view to their objective validity. Kant is not interested in types or forms of judgements per se. The function of unity, of which there are twelve basic kinds, has to do with the function of the understanding of uniting representations under one another (A68/B93). And, as Kant says, the faculty of understanding is the faculty to judge (A69/B94). If the capacity to judge, i.e. to operate a function of unifying representations under one another, is the understanding itself and the faculty of understanding is the principle of transcendental apperception (B136, 137, 138–9; cf. heading §17), then there is a relationship between the functions of unity in judgement and the unity of apperception, and indeed this is suggested by the passage at A79, and also rightly taken to be suggested here by the likes of Henry Allison, Robert Pippin, John McDowell, and James Conant; and also such famous readers of the Metaphysical Deduction as Klaus Reich and Michael Wolff take the unity of apperception to be directly involved in the derivation of the functions of judgement (see Reich 2001; Wolff 1995:177, 180–1). Watt seems to want to disentangle the understanding from transcendental apperception. I come back to this further below.

In my view, transcendental apperception should be invoked, because if it weren’t one could not make sense of the relation between the categories and the functions of judgements that they are said to have in the first part of the B-Deduction (see B143). In whatever way you construe the application of the categories to manifolds of intuition, the principle of transcendental apperception must in some way be involved. If the functions concern the unity of concepts and the categories are the functions of unity of our concepts insofar as they are instantiated in intuitions of objects, and the unity of apperception regulates both unities (by means of the analytic unity of consciousness and the synthetic unity of consciousness respectively; A79/B105), how can the unity of apperception as the principle of understanding (§17) not be centrally involved in the functions of unity in judgement? Lest Kant’s definition of the understanding at A68–9/B93–4 be read as leaving room for disentangling apperception from the faculty of the understanding (for, to be sure, no mention is made of consciousness or apperception in the Metaphysical Deduction, prior to the Leitfaden passage at A79), we can see in a parallel passage in the Prolegomena that the understanding is explicitly connected to consciousness; Kant says that “the business […] of the understanding is to think”, and immediately thereafter continues: “But thinking is uniting representations in a consciousness” (Prol, AA 4:304 [Kant 1977:48]). This implies that consciousness, or apperception, is intimately involved in the acts of the understanding. And if it is indeed thus involved one should be able to tell how it is involved in the account in the B-Deduction, and as far as I can see Watt is not able to tell the story of how it is involved. It may be suggested that surely there can be other ways of reading the passages I quote in direct support of my reading, but as long as one isn’t told what these different accounts are in sufficient detail, such a critique is without cause.

Watt says that Kant only clarifies the “forms” of judgement and that this gives the corresponding categories. But if this is supposed to amount to merely stating the correspondence, it leaves Kant’s claim that we must “systematically generate” the system of the categories “from a common principle, namely the faculty for judging (which is the same as the faculty for thinking) [aus einem gemeinschaftlichen Prinzip, nämlich dem Vermögen zu urteilen (welches eben so viel ist, als das Vermögen zu denken)” (A80–1/B106) wholly unexplained. It seems to me that just summarising without any method is as arbitrary and haphazard as what Kant accuses Aristotle of doing. On Watt’s construal of the story that Kant tells in the passages I and he quote (the Prolegomena passage and the passage in the introductory section of the Deduction) the classification would appear to be arbitrary, just as Hegel complained, and certainly not a priori (notice that in the passage that I and Watt quote at A67/B92 Kant says that the completeness of the categories should be able to be “determined a priori”, which is contrasted with the dependence “upon whim or chance”, but Watt doesn’t notice this). For Watt, the principle is simply “that the categories correspond to the forms of judgement”, and that this “is doing the work in this derivation”. Watt believes that “[t]here is no reason to think that the principle to which Kant refers in each of these passages is the principle of apperception”.

I find Watt’s proposal somewhat puzzling: how can the fact “that the categories correspond to the forms of judgement” be the principle in accordance with which it can be determined that the categories “correspond to the forms of judgement”? Isn’t this viciously circular? Secondly, how does this square with Kant’s emphatic assertion that the division of the categories is “systematically generated from a common principle, namely the faculty for judging”? At least, Watt should acknowledge that the derivation is from the principle of judging, just as Kant says, which is the principle of understanding (given what Kant says at A69/B94), and then explain how that derivation is supposed to go. One can of course then quibble about the fact whether apperception is involved in this (though I firmly believe it does; see further below).

Further, Watt says that there is nothing odd about claiming that the “forms of judgement” are “explanatorily basic”. Watt points to the analogy Kant makes at B145–6 between the categories and the forms of intuition, an analogy he rightly says I didn’t comment on. Just as, as Kant says, no further ground can be given for why we have the particular forms of intuition that we have and no others, no further ground can be given for why we have precisely the kind and number of categories that we do. Watt claims that this says that we can “know, but not explain” that space and time are the pure forms of intuition, and likewise for the categories.

I had emphasised Kant’s use of the adjective “further”, which in Watt’s account doesn’t appear to have a function: a ground can be, and is, given for why we have these and no other categories, and this is the explanation “from a common principle, namely the faculty for judging”, but no further ground can be given. What I, and I took Kant to, mean by this is that we can systematically generate the categories in accordance with this principle and thus explain how they necessarily hang together, in their coherent inner division of four titles of each three moments. Contrary to what Watt suggests, this is the explanation that can be given, as much as the explanation in the Aesthetic concerned the account of why space and time are the forms necessary for the intuition of outer objects and having manifolds of representations at all, respectively. The explanation that the derivation of the functions of thought/categories from a principle is looking for is nothing but this systematic generation of all the moments of the functions of thought (i.e. judging). It seems as if Watt (and others) think that my claim (and hence Hegel’s and Kant’s on my reading) is supposed to be a claim about finding the golden principle that is the underivable, self-standing principle that mysteriously produces the categories (and the forms of intuition). If that were the case, the question would arise as to how that supposedly self-standing properly basic principle relates to what it produces, namely the categories/forms of thought and/or the forms of intuition. But the principle of apperception, and the derivation of the forms of thought/categories from it, is not such a self-standing properly basic principle; rather, transcendental apperception is the methodical principle of thought itself which enables us to see the systematic coherence among thought’s functions and the categories, just by looking at the analytical implications of the ‘I think’. It is this systematic coherence, then, that is basic, which can be made visible by looking at the way the principle of apperception operates in every act of thought, for each function of thought or judging.

Further, Watt finds it difficult to find any location in the text for the derivation of each of the specific categories. This is indeed the case, as I said in KDA. But that doesn’t mean the derivation of the categories doesn’t take place there. In the Deduction, Kant doesn’t proceed discursively, whereby he would be expected to define the categories in the way that his rationalist predecessors might do or did. The Deduction is a piece of prima philosophia, where the basic functions of thought are deduced from scratch, as it were, without presupposing definitional characteristics—this is precisely what I took the derivation to be in its core form from §16 onwards, which starts with the simple analytic proposition of the ‘I think’, and so proceeds systematically to provide the grounds for this analytic proposition, by simply teasing out the analytical implications from this proposition. I argued that these grounds are nothing but the categories, what Henrich (1976; 1988) has helpfully called ‘ways of transition’ between the representations of the ‘I’, which reveal its identity and unity.

Let me put this in a different way. The question in the Deduction uncontroversially concerns the issue of how categories are justifiably applied to objects. To answer this question in a systematic fashion (to give a dogmatic proof of sorts, unlike in the Prolegomena), Kant launches the progressive argument in the B-Deduction starting with the ‘I think’ proposition as the most basic proposition, namely simply the principle of thought. He cannot prove the objective validity of the categories by first defining them and then describing how they apply to objects. If he were to do so he would first have to define what he means by ‘object’ (which he in fact does, but only later on, towards the end of the actual derivation) and then to account for the relation between the categories as so defined and the object as so defined. But an account of such a relation would of course ex hypothesi require that we get to know about how that relation is grounded in both the categories and the object. First defining object and only then accounting for the relation between it and the previously defined categories would land us in an infinite explanatory regress, for how does the defined object relate to the relation?

Often commentators speak of categories ‘governing’ manifolds of representations in intuition or appearances, or of appearances or intuitions ‘standing under’ the categories (unfortunately Kant himself often uses the ‘standing under’ language himself). But such language is in fact utterly vague. How should we qualify such terminology in terms of what it expresses as a form of relation? It leaves the exact grounding of the relation of ‘governing’ or ‘standing under’ unaccounted for, namely how the one relatum is governed by or stands under the other relatum. In some way the categories and the cases of their application must be internally related such that we can justifiably say that, in the occurrent case of an objectively valid judgement about some object, the categories are applied to the intuition of that object. Now how does such an account work? Well, clearly not by providing definitions and standard premise-conclusion style arguments based on such definitions; doing so would be to completely misapprehend the a priori and transcendental nature of what a transcendental deduction is supposed to achieve.

Kant has to carefully ‘move’, to use a Hegelian notion, from the analysis of the mere capacity of thought or the understanding itself to the categories, and then to the object to which the categories are applied, or to put it more accurately, in which the categories are seen to be instantiated or exemplified. This happens in §§16 to 20, and then from §§22 to 26, for the so-called ‘second step’ of the argument, in which the exemplification of the categories in general in spatiotemporal objects is delineated. In §§16 to 17, Kant explains step-by-step how from the principle of apperception, i.e. the ‘I think’ of self-consciousness, evolves the objective unity of apperception that is definitive of an object in general. The very same conditions that ground the possibility of accompanying my representations with an ‘I think’, i.e. to self-apperceive my own representations, are shown, almost in a phenomenological Wesensschau, to be the conditions that ground the concept of an object in general. They are not similar or parallel conditions, they are the same conditions. These conditions are the functions of unity of thought, and the categories are these functions of unity thought insofar as an intuition of an object is determined (B143). Categories are then nothing but the various kinds of unity of consciousness of a manifold of representations in an intuition that make up the transcendental unity of apperception. The categories and the transcendental or original-synthetic unity of apperception come down to the same thing. Consider these passages:

Understanding manifests its power solely in judgments, which are nothing but the unity of consciousness in the relation of concepts in general, regardless of whether this unity is analytic or synthetic. […]

[I]n understanding there will be as many a priori concepts, under which the objects given by the senses must be subsumed, as there are modes of conscious composition (Synthesis), that is, as there are modes of synthetic unity of apperception of the manifold given in intuition. These concepts are the pure concepts of the understanding [….] categories. (FM, AA 20:271 [Kant 1983:75]; emphasis added)

Consequently there must be synthetic unity of apperception in the connection of this manifold; this unity of consciousness requires different functions to connect elements in the manifold in accordance with the differences in the intuitive representations of the objects in space and time. These are called categories and are a priori concepts of the understanding. (FM, 20:276 [Kant 1983:85]; emphasis added)

The unity of apperception in relation to the synthesis of the imagination is the understanding, and this very same unity, in relation to the transcendental synthesis of the imagination, is the pure understanding. In the understanding there are therefore pure a priori cognitions that contain the necessary unity of the pure synthesis of the imagination in regard to all possible appearances. These, however, are the categories, i.e., pure concepts of the understanding. (A119; emphasis added)

The logical moments of all judgments are so many possible ways of uniting representations in consciousness. But if they serve as concepts, they are concepts of the necessary unification of representations in a consciousness and so are principle of objectively valid judgments. (Prol, AA 4:305 [Kant 1977:48])

Apperception is itself the ground of the possibility of the categories, which for their part represent nothing other than the synthesis of the manifold of intuition, insofar as that manifold has unity in apperception. (A401)[17]

If these passages do not provide direct textual evidence for the view that categories are nothing but the various kinds of synthetic unity of consciousness, i.e. apperception—reinforcing my claim of the analytic derivation of the categories from apperception—I don’t know what will. The a priori deduction or derivation of the categories is then nothing but Kant showing, in the Deduction, that the categories, as various kinds of unity in a manifold, directly flow from the various kinds of unity of consciousness in thought itself, i.e. from the self-conscious ‘I’ that thinks.

The various kinds of unity in a manifold of representations in judgement are twelve to be precise. This is not an arbitrary amount, for they exhaust the four properly basic kinds of unity of thought, whereby these kinds have each 3 possible applications (this all gets explained in detail in KDA; which is not to say that the derivation is without its problems, e.g. the internal connection between the four titles is rather tenuous; Hegel has a better account here, in his Science of Logic). In KDA I explained how each of the categories is to be regarded as derived from the ‘I think’ proposition in that each function of unity is a constituent of the synthetic unity that grounds the principle of apperception, which is itself analytical (B135). Kant could have inserted at each stage of this deduction some sort of declaration “….and this is/constitutes the category of…”, but it would have interrupted the flow of the reasoning that leads from the unity of self-consciousness to the definition of object, and then to the definition of a judgement.

Further, what does it mean when Kant says in the conclusion to the ‘first step’ (B143) that the categories “are” the functions of judgement and when in the transitional §21 he says that “with the above proposition” (B144), which is a summary of what was done in the first step, “a beginning of a deduction of the pure concepts of the understanding” has been made? On Watt’s (and others’) account there is no credible story about why all of a sudden Kant here, in §20, mentions the categories (in general), when the very term ‘category’ hasn’t been used in all of §§16–20, i.e. in the actual deduction—in §20 Kant clearly suggests a link to what has come before in terms of a deduction of the validity of the categories. My story provides an account of that link, even though one could quibble about the details of my interpretation. There is an explanatory task to explain the link that Kant makes between the deduction of the categories and the story about apperception and apperception being the definition of judgement (objective unity of apperception). It is hard to see how Watt could tell a plausible story if he denies the link between apperception and the categories. He says that the transcendental apperception is not identical to the understanding, but that conflicts with his earlier statement, where he says that, in accord with B136, the principle of apperception is the principle of understanding, as well as with a passage such as the earlier quoted one from the A-Deduction:

The unity of apperception in relation to the synthesis of the imagination is the understanding, and this very same unity, in relation to the transcendental synthesis of the imagination, is the pure understanding. (A119; emphasis added)

And this one:

By its means [i.e. by means of pure imagination] we bring into combination the manifold of intuition on the one side and the condition of the necessary unity of apperception on the other. Both extremes, namely sensibility and understanding, must necessarily be connected by means of this transcendental function of the imagination. (A124; emphasis added)[18]

So I think Watt is mistaken to think that it is not apperception that is both subject and object of the investigation, but rather the faculty of understanding. The faculty of the understanding is the faculty to think, whereby thinking is the capacity for self-consciousness, aka transcendental apperception as the ability to self-ascribe representations to oneself, as well as the capacity to think about objects.

Acknowledgements: I thank my co-editors Christian Onof and Jacco Verburgt for their helpful comments on an earlier draft of this essay.


[1] In a forthcoming commentary on my book Robert Howell (2018) tersely mentions in a footnote that he sides with Dyck and Stephenson in this matter, without himself having investigated it—that is, he doesn’t give any evidence of it in the commentary. (Notice that Howell doesn’t mention Marcel Quarfood [2014], who in the same symposium in which Dyck and Stephenson took part, is much more positive about the prospects of my derivation claim.)

[2] A reworked, and hopefully improved, version of that article will appear as a chapter in my next monograph Reflexivity and Representation. I also recommend the recent papers by Allais (2016) and Gomes (2016) for very insightful accounts of Strawson’s take on idealism and the transcendental argument of the Deduction. See also Allison (2016) for a critical account of Strawson as well as Guyer (2017).

[3] Strawson is making an important point here that is mostly overlooked in the literature on Kant’s idealism. That is to say, if the relation between things in themselves and appearances is based on affection and the affection relation is to our sensibility, then the affection relation is not twofold, whereby two distinct “existences” are produced, i.e. spatiotemporal objects and sensation/mental representations (of those objects) respectively; rather, if we ignore here the so-called double affection theory of transcendental and empirical affection, there is one affection relation to our sensibility, from things in themselves to appearances/states of consciousness/representations. But if this is true, then the relation between things in themselves and appearances can’t be one of affection on a metaphysical interpretation of Kant’s idealism, which holds that appearances are aspects of the same things that have an itself side to them. So this view has to deny that there is an affection relation between things and themselves and appearances, or it must hold that the affection relation is only an empirical one between the appearance “aspects” of things in themselves and our sensibility. But in any case, there can’t be an “existence” in-between things in themselves and our representations that we would call spatiotemporal objects. There are only two “existences”: things in themselves and states of consciousness/mental representations, and appearances are either aspects of things in themselves or representations, depending on whether one reads Kant’s idealism metaphysical-realistically or phenomenalistically.

[4] I criticise Allais’s refutation of the phenomenalist reading in Schulting (2017b).

[5] Kant doesn’t rigorously distinguish between Prädikat and Eigenschaft.

[6] Contrary to what Howell (2018) alleges. See my reply to him (Schulting 2018).

[7] This is claimed by Alexandra Newton in a critique of my book (see Newton 2018).

[8] As Reich says in his famous book Die Vollständigkeit der kantischen Urteilstafel, “Das Urteil aber seiner Form nach ist gar nichts anderes als eben diese objektive Einheit der Apperzeption” (2001:69).

[9] In the book, I make a distinction between seeing and perceiving, following Quassim Cassam.

[10] Cf. Refl. 3055: “Urteil: die Vorstellung der Art, wie verschiedene Begriffe objektiv, d.i. um ein Erkenntnis des Objektes auszumachen (für jedermann), zu einem Bewußtsein gehören” (AA 16:634).

[11] Watt’s view also conflicts with the definition of judgement given in MAdN: in MAdN Kant defines judgement as an act by means of which given representations first become cognition of an object; analytic judgements are by definition excluded from this narrow definition of judgement, which guided Kant’s rewriting of the Deduction. Cf. also the passage at KpV, AA 5:12–13 where Kant appears in fact to differentiate universal validity (allgemeinen Gültigkeit/allgemeinen Fürwahrhalten) from objective validity (objektiven Gültigkeit).

[12] “[O]nly through the fact that an object is determined for the concept by means of the law of causality does the represented occurrence have objective validity, i.e., truth.” Truth here is true correspondence to the object, not truth value.

[13] Cf. At B247/A202 where the subjective Spiel der Einbildungen is contrasted with the necessary relation of appearances, which alone is objectively valid and concerns the empirical truth of my judgements.

[14] See A6/B10; A151/B190; ÜE, AA 8:228–30; Prol, AA 4:267. The truth of an analytic judgement can be sufficiently determined by means of PNC.

[15] See Schulting (2017a).

[16] The book is currently out of print and will appear in a revised edition with de Gruyter in the Kant-Studien Ergänzungshefte.

[17] See further the references at p. 186n22.

[18] See also B130, B138, B144–5, B151–3.



Allais, L. (2015), Manifest Reality. Kant’s Idealism & his Realism (Oxford: Oxford University Press).

——— (2016), ‘Strawson and Transcendental Idealism’, European Journal of Philosophy 24(4): 892–906.

Allison, H. (2016), ‘Transcendental Deduction and Transcendental Idealism’, European Journal of Philosophy 24(4): 920–33.

Conant, J. (2016), ‘Why Kant Is Not a Kantian’, Philosophical Topics 44(1): 75–125.

Dyck, C. (2014), ‘The Function of Derivation and the Derivation of Functions: A Review of Schulting’s Kant’s Deduction and ApperceptionStudi kantiani xxvii: 69–75.

Gomes, A. (2016), ‘Unity, Objectivity, and the Passivity of Experience’, European Journal of Philosophy 24(4): 946–69.

Guyer, P. (2017), ‘The Bounds of Sense and the Limits of Analysis’, Journal of the History of Philosophy 55(3): 365–82.

Henrich, D. (1976), Identität und Objektivität. Eine Untersuchung über Kants transzendentale Deduktion (Heidelberg: Carl Winter Universitätsverlag).

——— (1988), ‘Die Identität des Subjekts in der transzendentalen Deduktion’, in H. Oberer & G. Seel (eds), Kant. Analysen – Probleme – Kritik (Würzburg: Königshausen & Neumann), pp. 39–70.

Houlgate, S. (2006), The Opening of Hegel’s Logic (West Lafayette, IN: Purdue University Press.

Howell, R. (2018), ‘Deduction Difficulties’, Kantian Review 23(1).

Kant, I. (1977), Prolegomena to Any Future Metaphysics, revised Carus edition by J. Ellington (Indianapolis: Hackett).

——— (1983), What Real Progress Has Metaphysics Made in Germany since the Time of Leibniz and Wolff?, ed. and trans. T. Humphrey (New York: Abaris).

Land, T. (2018), Review of D. Schulting Kant’s Deduction and Apperception. Explaining the CategoriesKantian Review 23(1).

Newton, A. (2018), ‘Review of Dennis Schulting Kant’s Radical Subjectivism‘, Kantian Review 23(1).

Quarfood, M. (2014), ‘A Note on Schulting’s Deduction of Contingency’, Studi kantiani xxvii: 87–93.

Reich, K. (2001), Gesammelte Schriften, ed. M. Baum (Hamburg: Meiner).

Schulting, D. (2008), On Strawson on Kantian Apperception, South African Journal of Philosophy 27,3 (2008): 257–71. [revised version]

——— (2012), Kant’s Deduction and Apperception. Explaining the Categories (Basingstoke/New York: Palgrave Macmillan).

——— (2015), Transcendental Apperception and Consciousness in Kant’s Lectures on Metaphysics, in R. Clewis (ed.), Reading Kant’s Lectures (Berlin/New York: de Gruyter), pp. 89–113.

——— (2016), On An Older Dispute: Hegel, Pippin, and the Separability of Concept and Intuition in Kant‘, in D. Schulting (ed.), Kantian Nonconceptualism (London/New York: Palgrave Macmillan), pp. 227–55.

——— (2017a), Analyticity, Analytic Philosophy and Kant’s Synthetic A Priori: Comments on Robert Hanna’s Cognition, Content, and the A Priori‘, Critique (May issue).

——— (2017b), ‘Kant’s Idealism and Phenomenalism—Critical Notice of Lucy Allais’s Manifest Reality. Kant’s Idealism & His Realism‘, Studi kantiani XXX: 191–202.

——— (2018), ‘Gaps, Chasms and Things in Themselves: Reply to My Critics’, Kantian Review 23(1).

Stephenson, A. (2014), ‘A Deduction From Apperception?’, Studi kantiani xxvii: 77–85.

Strawson, P.F. (1968), The Bounds of Sense: An Essay on Kant’s Critique of Pure Reason, 2nd printing (London: Methuen).

Vanzo, A. (2012), ‘Kant on Truth-Aptness’, History and Philosophy of Logic 33(2): 109–26.

Wolff, M. (1995), Die Vollständigkeit der Kantischen Urteilstafel (Frankfurt a/M: Klostermann).

© Dennis Schulting, 2017.

Dennis Schulting is a former Assistant Professor of Philosophy at the University of Amsterdam. His work concentrates on Kant and German idealism. His most recent journal publications were published in KANT YEARBOOK, KANT-STUDIEN, the PHILOSOPHICAL REVIEW and KANTIAN REVIEW. He is the editor of Kantian Nonconceptualism (Palgrave Macmillan 2016). Two new books are forthcoming: Kant’s Deduction from Apperception. An Essay on the Transcendental Deduction of the Categories (Walter de Gruyter) and Reflexivity and Representation. Essays on Kant and German Idealism (Walter de Gruyter). A book on Hegel’s Science of Logic is in the works.