Subjective Validity, Self-Consciousness and Inner Experience: Comments on Kraus

Abstract I raise three related objections to aspects of Katharina Kraus’s interpretation in Kant on Self-Knowledge and Self-Formation. First, I reject her claim that representations count as merely subjectively valid for Kant if they represent objects from the contingent perspective of a particular subject. I argue that Kant in fact describes consciousness of subjectively valid representations as consciousness of one’s own perceptions rather than of the objects perceived, and therefore that it plays a bigger role in his account of self-consciousness than Kraus allows. Second, whereas Kraus argues that the transcendental unity of apperception structures the content of any consciousness that is possible for a subject, I note that Kant also allows for a merely empirical unity of apperception, which he describes as in principle different from transcendental unity. Finally, I raise some worries for Kraus’s suggestion that we can be aware of the activity of thinking through inner sense.

Second, whereas she argues that transcendental self-consciousness or apperception and its attendant type of unitythe transcendental unity of apperceptionstructure the content of any consciousness that is possible for a subject, I will note that Kant also allows for a merely empirical unity of apperception, which he ties to inner sense and describes as in principle distinct from transcendental unity. Third, and relatedly, I will question Kraus's claim that we can be aware of the activity (or output) of thinking through inner sense.

Mere subjective validity
Kraus argues that the contrast between objective and merely subjective validity has to do with the 'mode under which a subject represents an object ' (p. 21). A representation counts as objectively valid on her view, if its 'representational content is : : : exclusively due to the features of the object, rather than specific contingent conditions of a particular subject ' (p. 22). In contrast, if the representational content is at least partly determined by a subject's particular 'visual' (p. 22) or 'geometrical perspective' (p. 23) on the object, then the resulting representation is merely subjectively valid. As examples of merely subjectively valid representations, she discusses a subject's representation of a chair without a back side (because she perceives it only from the front) or her representation of the chair as 'displaying a red shimmer' when it is in fact brown (because she is wearing red glasses) (p. 22). Kraus's descriptions and examples suggest that, on her view, merely subjectively valid representations represent the objects distinct from those representations, which those representations are taken to be of, albeit only partially (e.g. only from the front) or in a manner that otherwise depends on the subject's perceptual circumstances (e.g. the fact that she is wearing red glasses).
On my view, this construal of the distinction between objective and merely subjective validity conflicts with several features of Kant's discussion. To begin with, we should note that he always applies the distinction to a particular combination (Verbindung), connection (Verknüpfung) or relation (Beziehung or Verhältnis) of representations (e.g. Prolegomena (henceforth, P), 4: 300; B141-2). Furthermore, he makes clear that consciousness of such merely subjectively valid relations between representations amounts merely to consciousness of 'a connection of perceptions within my mental state, without relation to the object' (P, 4: 300, my emphasis). As I understand it, this means that such consciousness is consciousness merely of a relation between my perceptions, and does not yet amount to thinking these perceptions as representing features (even partial or contingent ones) of the object(s) distinct from these perceptions, which the perceptions are taken to be of. 1 The same claim is emphasized in Kant's discussion of the example he uses to illustrate the distinction between objective and merely subjective validity in §19 of the B-Deduction. There, he says that (1) 'When I carry a body, I feel a pressure of weight' expresses a merely subjectively valid relation between representations: it expresses that the representation of a body and the representation of weight are merely 'found together in perception'. In other words, (1) is about the subject and a relation between her perceptions (i.e. it claims that when she represents the body, she represents weight), rather than about the object(s) distinct from her perceptions (i.e. about the body itself). In contrast, Kant says, the judgement (2) 'It, the body, is heavy' expresses an objectively valid relation, insofar as it makes a claim about the object perceived (i.e. the body), rather than merely about the subject and her perceptions (B142). This indicates that Kant's distinction is not intended as a contrast between two modes of representing objects, as Kraus claims, but rather, between a subject's awareness of her own representations, on the one hand, and her judgements about the object(s) distinct from her representations, on the other.
Second, Kant's discussion of mere subjective validity never mentions the perspectives that particular subjects have on objects, which Kraus's understanding of subjective validity turns on. Instead, he says his goal is to distinguish between two possible types of relation between representations: namely, a relation that 'belong[s] to the understanding' in accordance with its principles, which he characterizes as objectively valid, and a 'relation in accordance with laws of the reproductive imagination' -'laws of association'which he says 'has only subjective validity' (B141-2, my emphasis). He correspondingly emphasizes this contrast between objective and merely associative relations between representations in his discussion of the example in §19. Yet the contrast between combination by the understanding and the reproductive imagination does not play any obvious role in Kraus's account of the distinction between objective and subjective validity.
Third, Kraus's reading makes it difficult to understand why the above distinction plays such a central role in Kant's deduction of the categories, both in the Prolegomena and the Critique of Pure Reason. Kant's argument there appears to rely crucially on the claim that full determination by the categories is required for representations that are objectively rather than merely subjectively valid (B142-3; P, 4: 300-1). It is unclear, however, how merely applying the categories could ensure that objects are represented in a manner that does not depend on a subject's particular perspective; conversely, even representations that do depend on a subject's perspectivee.g. the representation of a chair from the front or as shimmering redare fully determined by the categories. In contrast, a merely associative relation between representations is not fully determined by the categories: in particular, as Kant's example emphasizes, such a relation remains undetermined by the relational categories. Thus, he argues that it is by subsuming a merely associative relation between the representation of a body and the representation of heaviness under the relational category of substanceinherence that the representations come to have 'relation to the object'. In other words, the subject moves from consciousness that the representation of a body is followed by a representation of heaviness for her, to thinking of the body itself as a substance in which heaviness inheres.
I emphasize this interpretative disagreement between Kraus and myself because, on my reading, merely subjectively valid representation plays an essential role in explaining the possibility of empirical self-consciousness through inner sense, which both Kraus and I seek to give an account of. I have argued here that consciousness of merely subjectively valid representations amounts to a subject's consciousness of herself and her perceptions; as such, I take it to be essential to Kant's positive account of how subjects come to be empirically conscious of and cognize themselves, rather than the objects distinct from their perceptions. 2

The empirical unity of apperception
My second set of concerns are related to the first. Kraus argues that transcendental apperception and more specifically, its formthe transcendental unity of apperceptionshould be understood as the 'mere form of consciousness' (p. 106)in other words, that any representations of which we are conscious, whether objective or subjective, must have the transcendental unity of apperception (p. 110). In keeping with this, and along the lines of her more general claim that empirical apperception is the 'concrete realization' of transcendental apperception (p. 124), she argues that what Kant calls the 'empirical unity of apperception' is 'the unity of representations of which I am de facto empirically conscious of as my own' (p. 117).
On my view, this conflicts with Kant's text. §18 of the Transcendental Deduction begins: 'The transcendental unity of apperception is that unity through which all of the manifold given in an intuition is united in a concept of the object. It is called objective on that account (darum), and must be distinguished from the subjective unity of consciousness, which is a determination of inner sense...' (B139). 3 Kant goes on to identify this subjective unity with the 'empirical unity of apperception' and once again characterizes it as obtaining 'through association of the representations', giving the following example: 'One person combines the representation of a certain word with one thing, another with something else...' (B140). Now, Kant's claim that the empirical unity of apperception obtains through association is hard to square with Kraus's suggestion that it is the type of unity had by all the representations we are conscious of as our own. The latter also conflicts with the in-principle distinction Kant indicates he means to draw between the transcendental/objective unity that is brought about by the understanding, on the one hand, and the empirical/subjective unity due to the associations of the reproductive imagination, 4 on the other, a distinction that is also repeated elsewhere. 5 This is significant for the topic of empirical self-consciousness through inner sense, because Kant calls the subjective unity of representations the empirical unity of apperception or self-consciousness. On my view, it is the distinctive type of unity that accompanies empirical apperception. Although I do not have space to defend this here, I disagree with Kraus that this consists in consciousness of all of one's mental contents, including one's thoughts and judgements. Rather, it is consciousness in particular of representations as they are apprehended and associatively reproduced, before the synthesis of the understanding through which we cognize the objects distinct from our representations which our representations are of. 6 On my view, this fits well with passages in which Kant presents the contrast between transcendental self-consciousness and empirical self-consciousness through inner sense as a contrast between consciousness of oneself as active and passive, or as thinking and apprehending, respectively: [Apperception through] inner sense is not pure apperception, a consciousness of what the human being does (thut), since this belongs to the faculty of thinking. Rather, it is a consciousness of what he undergoes (leidet), insofar as he is affected by the play of his own thoughts : : : (Anth, 7: 162) The 'I' of reflection : : : is merely the formal element of consciousness. On the other hand, inner experience contains the material of consciousness and a manifold of empirical inner intuition, the 'I' of apprehension (consequently an empirical apperception). (Anth, 7: 141-2, my emphases) 3. Is it an experience that we think?
Both the above points are related to a third disagreement, concerning whether either the activity or the output of thinking are given to inner sense. Kraus argues that the material of inner sense consists in 'all representations "passing before the mind" and, hence, all mental activities from which these representations originate' (p. 32), including 'thinking', 'judging' (p. 17) and perceiving (pp. 76-7). Thus, she claims that we can obtain an 'inner perception of our own activity' by 'turn[ing] our gaze inward' from, say, an object that is seen to the seeing itself (pp. 76-7). In a similar vein, she argues that we can have inner experience of all our 'conscious mental states', where this includes our thoughts (p. 1, see also pp. 4, 24). I will object to this here by focusing on the Reflexion titled 'Answer to the question: Is it an experience that we think? ' (Refl. 5661, 1788-90;18: 318-19). As Kraus understands this Reflexion, it describes the conditions under which we have 'inner experience of one's thinking', that is, become aware of the 'inner appearance of [a] thought' through inner sense (p. 165). I disagree.
On the contrary, Kant argues in this Reflexion that since the acts of thinking that constitute experience involve determining the objective position of events in time, if they were themselves in time, it would have to be in a different, higher-order time in which the acts of first-order time-determination occurred. He concludes that this is absurd, and so that thinking itself is not experienced in time: [If] the consciousness when I institute an experience : : : were itself in turn empirical, then this temporal determination : : : would in turn have to be represented. Yet another time would therefore have to be given, under which : : : the time that constitutes the formal condition of my inner experience would be contained. Thus there would be a time in which and with which at the same time a given time flows, which is absurd. However, the consciousness of instituting an experience or also of thinking in general is a transcendental consciousness, not experience. (18: 318-19, my emphasis) Kraus's reading may seem to be suggested by Kant's claim in the Reflexion that the a priori thought of a square 'brings forth an object of experience or a determination of the mind that can be observed' (18: 319). This could be read as indicating that, on his view, we can be conscious of the thought of the square qua thought through inner sense. However, I think Kant's point is rather that this thought brings forth and gets attached to a sensible intuition of a square, 7 and it is the latter that is 'observed' through inner sense. This is indicated by his subsequent claim that the observation in question puts us in a position to 'demonstrate [the] properties of the square' (18: 319), since geometrical demonstration is never possible through thought alone and always requires construction through intuition on his view. In sum, while thoughts are surely representations that 'pas[s] before the mind', to use Kraus's locution, they do not, on my reading, appear qua thoughts to inner sense. We cannot intuit our thoughts, we can only think thembut this means that we do not have inner experience of thinking or of our thoughts qua thoughts through inner sense.
The three sets of objections I have discussed above each raise a similar point of disagreement: where I take Kant to posit a difference in kind, Kraus sees only a difference in degree, or a difference between a general form and its concrete realization. I have argued that merely subjectively valid representations are not simply less objective than objectively valid ones: they are about the subject's perceptions rather than about the objects perceived. Similarly, empirical self-consciousness is not simply a concrete realization of transcendental self-consciousness; rather, Kant describes it as consciousness of the 'I' that apprehends rather than the 'I' that thinks. The empirical unity of apperception is not an instantiation of the transcendental unity; rather, it is identified as a unity that is due to the associations of the imagination rather than the understanding. And finally, not all of our mental goings-on appear before inner sense; as Kant puts it, 'the human being : : : knows himself also through pure apperception, and indeed in actions and inner determinations which cannot be accounted at all among impressions of sense...' (A546/B574).
These disagreements notwithstanding, Kraus is to be commended for her rich and comprehensive book, which I am sure will set the agenda for future discussions of Kant's account of the self.
Notes 1 I take this to be of a piece with Kant's claim that consciousness of perceptions alone does not amount to cognition of objects. As he puts it, the former 'relates all representation only to our self as modifications of our condition; they are in this case : : : not cognitions of any things and are related to no object" (Refl. 5923, 18: 386, my emphasis). For further discussion of the difference between perception and cognition, see Tolley 2020. 2 For a full defence of these claims, see Sethi 2020 and Sethi forthcoming. 3 Kraus denies that §18 draws an in-principle distinction between the transcendental and empirical unity of apperception, claiming that the former 'is called "objective" only "on that account" that considers specifically cognition' (p. 117). But the phrase 'on that account' in the quoted passage translates the German darum, which could also be simply translated as 'therefore' or 'for that reason'. As I read it, then, it does not restrict the context in which the transcendental unity of apperception is considered objective, as Kraus suggests; rather, it says that the former always unifies the manifold under the concept of an object and is therefore objective. 4 Kraus argues that Kant's claim that the empirical unity of apperception is 'derived' from the transcendental unity of apperception supports her interpretation (p. 117). However, we should attend to how Kant explains his derivation claim: he immediately ties this derivation to the dependence of the empirical unity of apperception on the unity of time (B140). As I understand it, his claim is that since the empirical unity of apperception is represented as a temporal unity (because consciousness of associated representations is consciousness of one representation following another in time), it depends on representing the unity of time, where the latter is an objective unity. This dependence should not be taken to entail that the empirical unity of apperception is itself an instance of objective unity, a claim that Kant expressly denies. 5 B141-2; Refl. 5931-3. 6 For a full defence of this reading of empirical self-consciousness, see Sethi forthcoming. 7 In other words, what is brought forth is simply an intuition of a square, not an intuition of the thought of a square, as Kraus seems to suggest. See also B154.