1. Introduction
Can self-knowledge be objective? There are, undoubtedly, crucial differences between self-knowledge and knowledge of external objects. The former concerns oneself, in particular one’s mental states and empirical-psychological features, whereas the latter concerns spatiotemporal material objects, which are considered in some important respect to be independent of the knowing subject.Footnote 1 Since self-knowledge seems more intimately related to the knowing subject, it is often characterized as subjective and private, as opposed to knowledge of external objects, which is considered to be objective and publicly accessible. This raises the fundamental question of whether (some kinds of) self-knowledge can be understood as knowledge of oneself as object or as objective reality, rather than as the consciousness of oneself as thinking subject or as a merely subjective awareness of one’s mental states.
Immanuel Kant’s transcendental philosophy offers a sophisticated and subtle view that tackles this question, or so I shall argue. In the Critique of Pure Reason (henceforth, Critique), Kant famously enquires into the necessary conditions of the possibility of cognition. On the one hand, he assigns a central role to the thinking subject in the constitution of experience (Erfahrung). On the other hand, experience, for Kant, is more than merely subjective perception and amounts to the empirical cognition of objects (empirische Erkenntnis von Gegenständen). This technical notion of experience as empirical cognition is paradigmatically defined for material objects in space and time. Yet Kant also acknowledges that we can have inner experience (innere Erfahrung) based on some sort of inner perception (innere Wahrnehmung) (e.g. B278; Anth, 7: 142).Footnote 2 That is, experience of myself qua my particular empirical-psychological features in time, primarily in terms of my mental states such as my memories, beliefs, feelings, desires, intentions, fears and hopes. But does inner experience really amount to empirical cognition of myself or does Kant here use the notion of experience in a less technical sense? Is there anything cognized in inner experience in such a way as to fulfil the standards of objectivity set by the Critique? Far from concerning merely a technicality of Kant’s terminology, these questions raise an intricate puzzle regarding the status and nature of the knowledge we have of ourselves as psychological persons. Nonetheless, Kant’s notion of ‘inner experience’ has been relatively neglected in Kant scholarship and beyond. The few interpretations available in the literature differ hugely over the question whether inner experience should be viewed as empirical cognition and, if so, to what extent it is on a par with outer experience, i.e. the cognition of mind-external objects.
The question whether inner experience is cognition depends on whether it has the right kind of content that can underwrite cognition or, more precisely, whether there is an object determined in inner experience according to the forms of empirical cognition, i.e. the transcendental conditions of experience. Kant does not unambiguously answer these questions. Rather, we find two contrasting tendencies in his Critical writings. On the one hand, he explicitly speaks of the ‘object of inner sense’ (as opposed to that of outer sense) and considers the possibility of ‘cognizing myself as object’ in parallel with the cognition of outer objects (B155).Footnote 3 On the other hand, he denies that the self may be given to inner sense as an object referred to in experience (e.g. A22/B37). He argues that the self-consciousness of the thinking subject, i.e. the consciousness of oneself as unified subject of mental activity, is a transcendental condition of cognition, which itself cannot amount to the cognition of an object (e.g. A107, A120, B131, also B275, A350, A381, B412). Corresponding to Kant’s own apparently opposed declarations about inner experience, we find two interpretative approaches in the literature – one positive and the other negative with respect to the cognitive status of inner experience. I call the two approaches the parity view and the disparity view, respectively.
According to the parity view, inner experience should be construed – like outer experience – as the empirical cognition of a mental object.Footnote 4 This view appeals to an alleged structural parallel between inner and outer sense and holds that both – inner and outer experience – are determined through the same set of transcendental conditions, despite some specific differences, such as the non-spatiality of inner states and the representational powers of the mind (in contrast to the physical forces of matter).Footnote 5 These transcendental conditions include the forms of sensibility, time and space, and the pure concepts of the understanding, the categories. On this view, inner experience is therefore taken to be on a par with outer experience.
This parity view contrasts with a set of interpretations that recognize insurmountable disparities between inner and outer experience and therefore reject the identification of inner experience with empirical cognition in the proper sense. Although these interpretations come in different flavours and differ in many details, I subsume them under the common label disparity view. A major widely shared concern is that there may not be a proper mental object to be referred to and determined in inner experience. While the reasons that are assumed for this lack of an object may differ, it is generally argued that whatever is given in inner sense (e.g. ever-fleeting appearances) cannot be determined in a way significantly analogous to an outer material object and in particular cannot instantiate the category of substance.Footnote 6 Some argue, more specifically, that inner experience is more fundamentally dependent on the consciousness of oneself qua thinking subject (transcendental self-consciousness) and thus amounts, not to cognition of a mental object, but to some sort of empirical self-consciousness that is parasitic on transcendental self-consciousness.Footnote 7 Others think that Kant’s notion of inner experience can only be meaningfully discussed in the context of agency and with respect to the whole embodied person. For them, inner experience does not yield theoretical knowledge about one’s mind, but helps in orienting one’s actions in pragmatic and moral respects.Footnote 8
This article critically analyses this bifurcation between the parity and the disparity view with the aim of offering a more compelling account of inner experience – an account that is not only a more adequate interpretation of Kant’s conception, but also systematically more promising in its own terms. After introducing Kant’s notion of inner experience based on textual evidence, I elaborate on the parity view (section 2). Then in section 3 I discuss three concerns commonly raised against this parity (and used in support of the disparity view) and show that they are not fatal for an account of inner experience as empirical cognition. In the final section, I identify a major disparity between inner and outer experience – the lack of an adequate sensible schematization of a mental substance – and then indicate an alternative view that is able to accommodate this disparity (section 4). According to this alternative view, inner experience is empirical cognition of mental states (contra the disparity view), though not cognition of a mental substance (contra the parity view).
2. Inner Experience and the Parity View
Kant never provides a comprehensive account of inner experience, but only scattered remarks, in the Critique and other Critical texts, on inner experience and related notions, such as inner intuition (innere Anschauung), inner perception (innere Wahrnehmung) and self-observation (Selbstbeobachtung). These notions cover a variety of phenomena, ranging from consciousness of one’s sensory impressions and feelings to consciousness of one’s beliefs and desires, and more broadly to consciousness of the temporal sequence of one’s mental states. Kant mainly conceives of inner experience in the narrow sense as experience of oneself ‘from within’ without drawing on self-directed external sources of information, such as behavioural observation or testimony of others.Footnote 9 This means that inner experience is based on, though not identical with, inner perception (or self-observation).Footnote 10 Inner perception involves merely the reception and presentation of appearances in inner sense, yielding inner intuitions of mental states, whereas inner experience seems to require something more.
In the Anthropology and related notes, Kant suggests that inner experience is made possible as empirical cognition by some kind of ‘reflection’ on inner perception. He is less clear about what exactly this ‘reflection’ consists in: (a) reflection under the concepts of the understanding in general, or (b) reflection under the concept of oneself (the concept ‘I’), or (c) subsumption under (psychological) laws of nature:
(a) ‘an empirical intuition which, through reflection and the concept of understanding arising from it, becomes inner experience’ (Anth, 7: 142)
(b) ‘The inner sense is not yet cognition of myself, but we must first have appearances through it, then we first make a concept of ourselves through reflexion on [these appearances], which eventually results in empirical cognition, i.e. inner experience.’ (Refl 6354, 18: 680)
(c) ‘[Inner sense] belongs to psychology (a sum of all inner perceptions under laws of nature) and establishes inner experience’ (Anth, 7: 141)
In the Critique, which presents Kant’s official theory of experience and empirical cognition more systematically, it is unclear whether both – experience and cognition – come together in the case of inner experience as they do in the case of outer objects. There, the notion of inner experience occurs mainly in the context of the Refutation of Idealism (B275–9), which aims to show that ‘inner experience in general is possible only through outer experience in general’ (B278–9). Inner experience is thereby identified with the ‘determination of my existence in time’ (B275; also Bxl, 157n, 430–1), which I take to involve the temporal determination of one’s mental states. This and other passages in the Critique leave it open whether inner experience on its own counts as empirical cognition or whether it remains incomplete if not bound up with outer experience and thus viewed as part of outer cognition.
When talking about matters of inner experience, Kant also uses locutions such as ‘determining my own existence in time’ (Bxl, 157n, 430–1), ‘cognizing myself as I appear to myself’ (B68, 158) and ‘self-cognition’ (B421).Footnote 11 While his writings on anthropology strongly suggest that inner experience is some kind of empirical cognition, the Critique mainly emphasizes its dependence on outer experience and thereby raises doubts regarding its cognitive status. With respect to Kant’s Critical theory of experience, we may now ask: does inner experience satisfy the transcendental conception of empirical cognition of objects, determined through the formal conditions of experience, rather than being some sort of merely subjective awareness a subject has of her own mental states in some indeterminate manner?
The parity view claims that inner experience is empirical cognition of the same kind as outer experience. Jonathan Vogel, who supports this view, suggests that ‘the empirical self’, whose states are given in inner sense, should be viewed as ‘a fully-fledged object’ that can be cognized through inner experience as a ‘genuine, persisting object’ (Vogel Reference Vogel1993: 881). For him, it seems unproblematic to assume an enduring empirical self with changing mental states, which can be observed in inner sense and then cognized through psychological concepts.Footnote 12 Patrick Frierson also argues for the ‘legitimacy of using a substantial soul for empirical purposes’, though more cautiously concedes that Kant is ‘non-committal with regard to the strict metaphysical issue of whether the soul is a substance’ (Frierson Reference Frierson2014: 24–6).
The argument offered in support of the parity view mostly starts from the apparent parity between inner and outer sense. Inner sense is construed as one of the two faculties of sensibility, each of which yields intuitions from some sensation. Explicit support for this parity can be found in the parallel definitions of the senses in the Transcendental Aesthetic. There, Kant defines inner sense as that ‘by means of which the mind intuits itself, or its inner state … as an object’, after having specified outer sense as that by which ‘we represent to ourselves objects as outside us’ (A22/B37).
The argument continues by showing that both outer and inner experience are constituted by the same two kinds of representations – empirical intuitions given through the senses, and empirical concepts through which an object is thought in accordance with the categories – and that these intuitions and concepts are related such that they are sufficient to constitute experience as empirical cognition of an object. In outer experience, outer sense affords spatiotemporal intuitions, which represent empirical properties of an outer object. These properties are represented as states of one and the same object on grounds of the unifying function that the understanding exerts with regard to sensible intuition in accordance with the categories, including that of substance. By way of parity, it is argued that in inner experience, inner sense affords temporal intuitions, which represent empirical properties of the mind. These properties are, again, represented as states of one and the same mind on grounds of the unifying function that the understanding exerts according the categories, including that of substance. On this view, the unifying function of each category is applied in the same way with regard to both inner and outer intuition and, therefore, there is no qualitative difference in the way that the categories determine the object of inner or, respectively, outer experience.
This parity of category application seems supported by the fact that most arguments in the Critique that establish Kant’s theory of cognition, especially the Transcendental Deduction and the Principles of the Understanding, make no distinction between inner and outer intuition and thus seem equally applicable. Thus, it is concluded that, even though Kant’s default case may be outer experience, his arguments hold – by way of parity – for inner experience too. A passage from the A-Paralogisms explicitly suggests such parity, stating that the objects of both outer and inner experience are ‘substances in appearance’ without any qualification:
[I]n the connection of experience, matter as substance in appearance is really given to outer sense, just as the thinking I is given to inner sense, likewise as substance in appearance, and in connection of our outer as well as our inner perceptions, appearances on both sides must be connected amongst themselves in one experience according to rules that the category of substance brings in. (A379)Footnote 13
In sum, the parity view holds that inner experience is empirical cognition on a par with outer experience, since it is determined through the unifying function of the categories in a way significantly analogous to that of outer experience. More specifically, the inner object to which mental states are attached is determined as an empirical substance persisting in time.
3. Disparities between Inner and Outer Experience
The parity view faces several objections. Three main concerns have been put forward in the literature, all culminating in the claim that there is no proper mental object to be cognized in inner experience. First, it has been argued that inner sense does not yield proper material, i.e. a manifold of intuition of its own, and therefore there can be no distinctively inner intuitions that concern mind-internal matters only; rather, what we find in inner sense is merely a ‘reappropriation’ or reorganization of the data received through outer sense.Footnote 14 I call this the no-mental-matter objection. Secondly, it has been argued that, even if there are distinctively inner non-spatial intuitions, they cannot be cognized in an objective way unless they are ‘bound up’ with outer spatial intuition; the principles of the understanding (see A148–235/B187–294), which prescribe how the categories are applied to intuition in order to yield cognition, necessarily presuppose spatial dimensions and are properly applicable only to outer intuitions. I call this the spatiality objection. Thirdly, it has been claimed that, since inner states – unlike physical states – are perceived as highly ephemeral, it seems implausible that there could be anything persistent given in inner sense that would be appropriate to instantiate the category of substance. I call this the no-persistent-substance objection.
These objections are often put forward in support of the disparity view, according to which inner experience cannot be empirical cognition. The common conclusion of all these objections is that there is in fact no mental object that appears in inner sense and that is properly cognized according to all the categories. In turn, only spatiotemporal material objects can be adequate objects of cognition. In what follows, I consider these objections in turn and argue that none of them successfully disables the parity view. Then I raise another disparity, which I take to threaten the parity view most severely and which then leads me to develop an alternative view.
The No-Mental-Matter Objection
Advocates of the no-mental-matter objection claim that inner sense does not yield a manifold of sensation of its own, but rather relies entirely on being provided with sensory material through outer sense. The commentators endorsing this position mainly draw on those passages added in the Critique’s B-edition that highlight the importance of space and outer intuitions for Kant’s theory of cognition in general. For example, Allison develops a strong reading of the claim that ‘outer sense provides the proper material for cognition’ (B67) and argues that inner sense yields only a ‘reflective reappropriation of the contents of outer sense’ (Allison Reference Allison2004: 282).Footnote 15 While outer sense supplies the ‘raw materials for cognition’, the function of inner sense is ‘to combine these materials in accordance with the conditions of time’ (Allison Reference Allison2004: 282–3).Footnote 16 Inner experience is a necessary part of outer experience, as one aspect of every experiential act, but not a self-standing cognition. Moreover, what count as genuinely inner sensation for Allison are primarily feelings. Feelings, however, Allison argues, have no representative function and cannot contribute to any objective content of experience; nor do they themselves represent mental states in such a way that they can be grasped in cognition.Footnote 17
This reading is problematic for two reasons. The first concerns the phenomenal richness of human mental life, which could not be accounted for if the matter of inner sense is fully reducible to that of outer sense. If all that appears in inner sense needs to be related to perceptual representations of outer objects as part of one and the same experiential act, Kant would lack an explanation for how other kinds of mental states that do not directly represent external objects, such as feelings, desires and intentions, could enter our consciousness at all. He presents a diversity of mental phenomena in the Anthropology, in particular in his classification of mental deficiencies, which include not only perceptual illusions and hallucinations, but also various phenomena concerning a person’s mental state in light of her social, medical or psychological conditions.Footnote 18 Moreover, Kant’s theory of ethics appeals to mind-internal causes: the feeling of respect, various sensible inclinations that motivate our actions, and the feelings of pleasure or pain, which enforce or impede an action, can be consciously represented only through inner sense, without relation to outer objects and outer sense (e.g. G, 4: 451; CPrR, 5: 23, 58, 96–7).
Hence, I argue, Kant conceives of inner sense not only as a faculty for perception that receives outer appearances in the a priori form of time, but more broadly as the faculty through which ‘we are internally affected’, i.e. a posteriori, yielding inner sensations (B153).Footnote 19 There is textual evidence that Kant allows for the notion of inner sensation as a sensation that arises from inner sense being ‘affected from within’ by ‘inner causes’ – causes that are internal to the subject, namely in terms of mental faculties such as the cognitive faculties, the will, desires and feelings:
If the faculty for becoming conscious of oneself is to seek out (apprehend) that which lies in the mind, it must affect the latter … there it then intuits itself … in accordance with the way in which it is affected from within. (B68–9, see also A98–9, B68–9, A357–8; Refl 5661 and 6311, 18: 319 and 619)
The Anthropology appeals to ‘a manifold of empirical inner intuition’ and to the ‘matter of inner sense’ as the determinable in inner experience (Anth, 7: 141-2). All these passages suggest that what is able to affect the mind ‘from within’ is not necessarily tied to perceptual states concerning outer objects, but includes any kind of inner appearance. Prima facie, there seems no fundamental disparity between inner empirical affection through mind-internal causes and outer empirical affection through mind-external causes.
A further reason why this objection fails is that there is some systematic pressure on Kant to accept the possibility of a distinctively inner manifold: his theory of a priori self-affection, which explains how the categories relate to the forms of sensible intuition, in particular to time.Footnote 20 A priori self-affection is primarily discussed in the B-Deduction as a crucial element of empirical cognition, since it finally closes the gap between the understanding and sensibility. It does so by showing that inner sense is a priori determined through the ‘synthetic influence’ of the understanding such that the sensible intuitions yielded by inner sense are suitable for determination according to the categories:
The understanding therefore does not find some sort of combination of the manifold already in inner sense, but produces it, by affecting inner sense. (B155)Footnote 21
Time, as the formal condition of the manifold of inner sense, thus of the connection of all representations, contains an a priori manifold in pure intuition. (B177/A138)Footnote 22
It seems to be an indispensable premise for Kant’s argument to work that there is a manifold of inner sense, i.e. the manifold of pure intuition of time, that is determinable in accordance with the categories.Footnote 23 Only if such a manifold is presupposed does it make sense to assume that the understanding shapes a priori what is given to inner sense. Kant illustrates his argument by the example of drawing a spatial figure, e.g. a line, on paper or in thought. Yet this does not mean that his theory is confined to perceptual or imaginative processes that involve spatial figures (see B154). Rather, Kant’s point is that every mental process – perceptual or non-perceptual – induces a manifold of mental states that is intuited in inner sense such that it accords with the a priori ‘synthetic influence’ of the categories on the pure manifold.Footnote 24 Hence, it seems a necessary premise of Kant’s theory that inner sense is capable of having a manifold of its own that is determinable through the understanding and that is in the required sense independent of outer sensation.
The Spatiality Objection
It is frequently argued that the categories – as conditions of possibility of experience – are properly applicable only to outer intuitions and that, therefore, only outer experience amounts to proper cognition of objects. If this argument were correct, it would directly undermine the parity of category application that is central for the parity view. Some interpreters assert that the Critique’s arguments that show the categories’ objective reality must appeal to space and outer intuition. Eckhart Förster even claims that the Principles of the Understanding, which specify the categories with respect to the object’s temporal determinations, remain incomplete unless supplemented by some principles for spatial determinations, such as those provided in the Metaphysical Foundations of Natural Science (henceforth, Foundations).Footnote 25 These Principles are not fully justified by appealing to time and inner intuition alone; rather, in order for the categories to have ‘sense and meaning’, i.e. sensible content, they need to be applied to spatial intuition.Footnote 26 Various other commentators more or less explicitly subscribe to the claim that for an object to instantiate the categories it must have spatial dimensions and outstrip consciousness.Footnote 27
Although Kant nowhere directly argues for restricting the categories’ applicability to outer intuition only, there are a few passages, mainly added in the Critique’s B-edition, that highlight the importance of outer intuition for the categories’ objective reality, such as the General Note to the System of Principles:
In order to understand the possibility of things in accordance with the categories, and thus to establish the objective reality of the latter, we do not merely need intuitions, but always outer intuitions. (B291)
Specifically with respect to the category of substance, it is argued: since nothing ‘lasting’ or ‘persisting’ can be found in inner sense that corresponds to the ‘concept of substance’, we need an ‘intuition in space (of matter)’ to ‘establish the objective reality of this concept’ (B291).
The Refutation of Idealism (in the B-Postulates) aims to refute Cartesian problematic idealism, according to which the existence of external objects cannot be proven, whilst the existence of oneself can be known through inner experience. There Kant maintains that ‘inner experience itself is … possible only through outer experience’ (B277). In particular, he starts from the premise that we are conscious of ourselves as determined in time, which he identifies with having inner experience for the purpose of his argument. Yet for such consciousness of oneself in time to be possible, something persistent must be presupposed that is independent of the mind’s changing representations and that thus is in fact external to the mind itself.
The Refutation of Mendelssohn’s Proof of the Permanence of the Soul (in the B-Paralogism) is supposed to show that the persistence of the soul as the object of inner sense cannot be proven but can be assumed in virtue of being tied to the persistence of the human body:
the persistence of the soul, merely as an object of inner sense, remains unproved and unprovable, although its persistence in life, where the thinking being (as a human being) is at the same time an object of outer sense, is clear of itself. (B415)
These passages indicate doubts that the categories are applicable in the same way to inner and outer experience, though none of them explicitly excludes their applicability to inner, non-spatial intuition. Even if Förster and others were right in arguing that a proof of the objective reality of the categories necessarily relies on the possibility of applying the categories to outer, spatial intuition, it would not follow that, once the categories’ objective reality is established, the categories are not applicable to what is given exclusively in inner senses. Hence, prima facie, these readings do not directly show that the categories cannot also hold for specific instances of inner experience.
These passages leave some ambiguity regarding Kant’s actual view. Systematically we can distinguish three options as to how the applicability of the categories depends on outer spatial intuition. It may depend on (i) the possession of an outer sense as the faculty for spatial representation by means of which one is able to represent distinct parts as external to one another; or (ii) the possession of the empirical concept of (physical) matter in general as the concept of something that is mind-external; or (iii) the possession of outer intuitions of actual outer objects that persist in time.Footnote 28 In what follows, I briefly show that the first two kinds of dependence are unproblematic for inner experience and do not exclude the applicability of the categories to inner non-spatial intuition. The last kind is more complex and requires a closer discussion of the category of substance, which I provide in the next subsection (3.3).
Dependence on outer sense as the faculty for spatial representation
In several passages of the Critique, Kant argues that the capacity for spatial representation is indispensable for the representation of time. Time ‘cannot be made representable to us except under the image of a line’ (B156, cf. A33/B50). Succession is represented ‘in drawing a straight line (which is to be the external figurative representation of time)’ and successively attending to this act of the imagination (B154). This suggests that any representation of time necessarily involves the faculty for spatial representation. The representation of time duration requires the representation of several ‘time parts’ as distinct from one another and spatially distributed on a ‘time line’. The representation of succession requires the apprehension of some spatial change (i.e. motion), such as the drawing of a line. In this sense, any temporal determination – and in particular any application of the categories of relation, i.e. substance, causality and interrelation – must appeal to the faculty for spatial representation. So it is unproblematic that inner experience too relies on this faculty in order to determine the temporal order of mental states.
Dependence on the empirical concept of (physical) matter
In some passages, Kant makes the stronger claim that the application of the categories, in particular those of relation, requires the ‘intuition in space (of matter)’ (B291). In the Refutation (and the corresponding footnote in the B-Preface), he argues that for the experience of any change something ‘persistent in space’ is necessary, which is discerned as matter. Yet he acknowledges that the persistence of matter ‘is not drawn from outer experience, but rather presupposed a priori as the necessary condition of all time-determination’ (B278). This explanation I take to imply that what is required for time-determination, and a fortiori for the determination of mental states in time, is the presupposition of the empirical concept of matter (and its a priori determinations), rather than the actual givenness of some particular physical-material object. Hence, it does not necessarily mean that for inner experience to be possible there must be a physical substratum in which mental states inhere; nor that there must be an underlying physical-material framework that serves as a ‘backdrop’ against which the changing inner states can be measured through direct correlation (i.e. each mental state directly correlates with a physical state of the framework).Footnote 29 Rather, what is important is that we have at our disposal a concept of what it means to be a persistent ‘backdrop’ against which we can understand change and that we have an empirical proof that such a backdrop is really possible; that is, we have a general concept of persistence that is not empty, but objectively real. With the empirical concept of physical matter, we possess precisely such a concept. While this concept itself is acquired through outer experience, Kant argues in the Foundations that its basic determinations can be derived a priori. Yet an analogous concept for the case of inner experience, i.e. an empirical concept of mind, does not seem available for some reason to be specified later (section 4.1). On this reading, the Refutation’s claim ‘that inner experience in general is possible only through outer experience in general’ (278–9) comes down to the claim that in order to determine oneself in time we need to borrow from outer experience the empirical concept of matter that instantiates the concept of something persistent in spatial intuition.Footnote 30
Neither the dependence on the faculty for spatial representation nor the dependence on the empirical concept of physical matter are problematic for inner experience and they do not exclude the applicability of the categories to specific instances of inner non-spatial intuition.
The No-Persistent-Substance Objection
Let us now focus on the third kind of dependence, namely in terms of the actual possession of outer intuitions of outer objects that actually exist and endure in time. Commentators who endorse this kind of dependence for inner experience often base their argument on the claim that the category of substance cannot be properly applied to what is given merely in inner sense. Therefore, inner experience, if possible at all, must be tied to outer experience of a specific physical-material substance. In virtue of the relentless passing of time, my inner states are evanescent and ever fleeting and hence too my inner perceptions of them. Since inner sense receives only an ever-changing flux of inner states, it cannot provide the material appropriate to instantiate the category of substance as something persistent in time (e.g. A22–3/B37, A107, A350, B412–13). Many commentators conclude that inner experience lacks a reidentifiable entity that can be cognized as an empirical substance and that must therefore be substituted by an outer material substance such as the human body.Footnote 31 A person’s inner state, it seems, changes continuously in any experiential episode in as much as the intuitions constituting this episode succeed continuously one another in inner sense: ‘Our apprehension of the manifold of appearance is always successive, and is therefore always changing’ (A182/B225).
This concern might be indicated in Kant’s own claim that in inner sense we lack a ‘standing or abiding self’ (A107, also A350, A381, B275, B412). Furthermore, he claims that the concept of substance can be proven objectively valid only for persistent objects in space, which requires the ‘intuition in space (of matter)’ (B291, also B278).Footnote 32 By contrast, Kant claims in other passages that ‘I cognize myself’ and that there is a ‘substance in appearance’ in inner perception (A379, also MFNS, 4: 542).
In defence, adherents of the parity view offer a deflationary account of outer substance that does not qualitatively differ from an inner object. It may be argued that we intuit a series of varying states, though without having an intuition specifically of the persistent substance that is taken to bear all these states. Even in outer experience, it suffices to perceive a bundle of changing states that are taken to inhere in a substantial outer object by way of the unifying function of the category of substance.Footnote 33 So the evanescence of inner and outer states may differ in degree, but certainly not in kind: both are alterable and can change rapidly, and in both cases the assumption of a reidentifiable object is entirely based on the category of substance and not on what is given in intuition.
While I do not think that Kant holds such a deflationary account (for reasons I spell out in 4.1), I agree with the parity view that some notion of temporal endurance, though perhaps not that of material persistence, is required for inner experience to be possible. As shown in section 2, for Kant inner experience concerns not only the perception of occurrent mental states, but also requires reflection ‘under (psychological) laws of nature’ (Anth, 7: 141) and may also concern temporally more stable phenomena such as dispositions for certain beliefs, emotional response patterns and maxims for action, character traits and other facts of one’s mental biography. Representing such phenomena certainly requires some notion of an enduring self that underlies all these phenomena.
Hence, if the no-persistent-substance objection is taken to deny any involvement of the concept of substance in inner experience, then it is difficult to account for those mental phenomena that seem to require some notion of an enduring self. So I suggest that this objection hints at a crucial disparity that concerns not whether, but how the concept of substance must be construed in the case of inner experience. In what follows, I argue that Kant does not outright reject the applicability of the category of substance to inner experience, but detects a difficulty in finding an adequate sensible explication of what this category amounts to in this case.
4. An Alternative View
No Empirical Schematization of Mental Substance
The previous section has shown that there is no disparity between inner and outer experience at the level of sensation, nor is there a principled exclusion of the applicability of the categories, including that of substance, to inner intuition. Nonetheless, the no-persistent-substance objection indicates a major disparity between outer and inner experience, which finally leads to my conclusion that inner experience is cognition of mental states, rather than of a mental substance.
The major disparity concerns how the concept of substance must be construed for inner experience. Even though we may have to assume an enduring self (as a mental substance) in order to represent mental states, we lack an appropriate schema that explicates the concept of substance in inner intuition. A schema gives a procedural rule for subsuming appearances under the categories (esp. A139–40/B178–9). For inner experience, we do not have available an empirical concept of the mind that could be explicated in accordance with such a schema in the same way as the empirical concept of physical matter.Footnote 34 The reason is that inner sense does not supply the appropriate formal structure to schematize the category of substance. It does not supply ‘a distribution of reality’ that can function as the ‘substratum of all time-determination’, as required by the First Analogy, the Principle of the Persistence of Substance (A183/B226).Footnote 35
In the Foundations, Kant presents the most explicit argument for this disparity between inner and outer experience.Footnote 36 In general, the ‘schema of substance is the persistence of the real in time, i.e., the representation of the real as a substratum of empirical time-determination’ (A144/B183). The real is received in sensation and apprehended as the quality of some property, e.g. heat or colour. As such, it has only intensive (or gradual) magnitude, as the Anticipations of Perception show. Now, in outer intuition, there is ‘something standing and abiding … which supplies a substratum grounding transitory determinations’ (A381). Outer intuition can offer such a substratum because it represents an aggregate of moveable parts, i.e. ‘parts external to one another’ (MFNS, 4: 541). A possible object of outer intuition is not merely an intensive magnitude (regarding its sensory qualities), but at the same time must be represented as spatially extended, i.e. as extensive magnitude. Only an extensive distribution of reality can function as a substratum that can be reapprehended at different times and that thus fulfils the definition of ‘the persistent of the real in time’. So what is given in outer intuition is a spatially extended distribution of sensory qualities and only as such can it be cognized as an empirical substance.Footnote 37 The foundational concept of physical matter represents precisely this spatially distributed substratum. This concept, though itself empirically given, can be explicated a priori according to the principles of the understanding, including that of substance: ‘the concept of a matter as substance is the concept of the movable in space’ (MFNS, 4: 543). The total quantity of matter is preserved, as the Second Proposition of Mechanics states.
By contrast, an analogous sensible explication of substance cannot be found for that which is exclusively given to inner sense, as Kant shows in detail. He argues:
By contrast, that which is considered as object of inner sense can have a magnitude, as substance, which does not consist of parts external to one another; and its parts, therefore, are not substances. (MFNS, 4: 542)
Here ‘parts’ mean the ‘representations in my soul’, i.e. my mental states (MFNS, 4: 542). These parts cannot be set apart and individualized as separable extensive magnitudes; rather, as parts of one and the same consciousness, they have only intensive magnitudes and augment or diminish by degrees. A distribution of a purely intensive reality (that is independent of spatial extension), however, cannot satisfy the condition of a substratum that persists in time. From this, Kant concludes that the ‘very substance of the soul’ and its ‘parts’ cannot be shown to be persistent in time, because they could gradually perish and therefore violate the conservation of quantity of substance (MFNS, 4: 542).Footnote 38 One may object to this conclusion that mental states are temporally extended and therefore supply an extensive distribution of reality. However, mental states, if viewed as temporally extensive distribution of reality, precisely cannot function as a substratum of time-determination, because such temporal distribution already presupposes a substratum according to which they first are represented as temporally extended.Footnote 39 Hence, Kant offers a strong argument that the schematized category of substance cannot be applied to inner intuition because nothing can be intuited as persistent in inner sense.
Yet this argument does not entail that no use at all can be made of the concept of substance for inner experience, as interpreters have often concluded (e.g. Friedman Reference Friedman2013: 322). In this and other passages, Kant clearly grants the applicability of the pure, unschematized category of substance, i.e. as ‘the subject of all predicates’, to the subject of apperception (which is expressed by ‘I’ of ‘I think’):
The I, the general correlate of apperception, and itself merely a thought, designates … the subject of all predicates … a substance, therefore, of which, by this term, one has no concept of what it may be. (MFNS, 4: 452)
In the A-Paralogisms, Kant states that the ‘I’ of apperception ‘signifies a substance only in the idea’ (A350), ‘a substance in concept’ (A400), i.e. as stated in the Second Premise of the First A-Paralogism, a premise that Kant himself endorses: ‘I, as a thinking subject, am the absolute subject of all my possible judgments and this representation of Myself cannot be used as a predicate of any other thing’ (A348).
Yet the applicability of the pure category is not based on intuition, but derived through rational syllogisms from the ‘I think’ of transcendental apperception – a general representation ‘which must be able to accompany all others and which in all consciousness is one and the same’ (B132). The ‘I think’ expresses a transcendental condition of cognition, namely the condition that all representations pertaining to an object must be unified in one and the same consciousness in order to form objective cognition. It constitutes only the generic form of self-referential thought.
Despite the fact that I have to think of myself as a ‘pure substance’ of all predicates applicable to me, I have no sensible concept of myself. In the Paralogisms, Kant argues that the ‘I’ of apperception supplies a representation of ‘the constant logical subject of thinking’, but no ‘cognition of a real subject of inherence’ (A350). On the basis of apperception alone, we cannot take ‘I’ to refer to a persistent substance in which real mental states inhere, without such substance having been sensibly given in inner sense. Nor, moreover, does ‘I’ represent the temporal unity of consciousness that grounds the subject’s personal identity, i.e. ‘the identity of its own substance as a thinking being in all changes of state’ (B408). The ‘I’ of ‘I think’, expressing a transcendental condition, defies any cognitive grasp. The apperceptive ‘I’ may refer to something real, but if it does so, then certainly without conveying any inner experience of the persistence of that something.Footnote 40 What else could guide an adequate sensible explication of substance for inner experience and supply a sensible concept of the mind, even though no substratum is given in inner sense?
The Idea of the Soul as the Analogue of a Mental Whole
In the remainder of this article, I indicate an alternative view of inner experience that preserves the possibility of cognizing mental states, whilst acknowledging that an empirical schematization of a mental substance is unavailable. I suggest that, if neither inner sense nor transcendental apperception can supply a sensible representation of the mind for inner experience, reason as a ‘genuine source of concepts’, i.e. ideas (A311/B386), qualifies as a source (or rather guideline) of such a representation.Footnote 41 I argue that reason’s idea of the soul supplies an ‘analogue of a schema’ that allows for a regulative, rather than constitutive, use of the category of substance with respect to inner intuition (A655/B693). Such regulative use allows us to cognize mental states in inner experience in relation to a projected mental whole, without cognizing a mental substance as persistent in time.Footnote 42
If inner experience is to be cognition of temporally ordered mental states, then it requires the representation of a mental whole to which these states relate as ‘parts’. Since such a mental whole is not given in inner intuition such that it can be subsumed under the category of substance, we have to approximate such a whole – namely as the ‘sum total’ (Inbegriff) of all temporally distributed mental states taken to belong to one and the same subject. This approximation is based on a regulative employment of reason.
Kant defines reason as the faculty for inferring that naturally seeks to ‘find the unconditioned for conditioned cognitions’ (A307/B364) and to complete a unity that must remain incomplete from the perspective of the understanding. Through a regressive use of its syllogistic forms, it gains so-called transcendental ideas, ‘of which no congruent object can be given in the senses’ (A327/B383), but which have ‘an indispensably necessary regulative use, namely that of directing the understanding’ (A644/B672). They give ‘completeness to the empirical synthesis through its progress towards the unconditioned’ (A409/B436). With respect to the categorical syllogism, reason gains the idea of the soul as ‘the concept of the absolute (unconditioned) unity of the thinking subject’ (A334/B391, also B348, A353, A406/B432). This unconditioned unity is taken to be the ultimate condition with respect to all one’s conditioned representations, i.e. inner appearances. Since the categorical syllogism corresponds to the conditioning relation of subsistence, Kant explicates this idea as the concept of a ‘simple [thinking] substance’ (A672/B700).
In the Appendix to the Transcendental Dialectic, Kant identifies this idea with the ‘guiding thread of inner experience’ that directs the understanding in determining mental states, without asserting the existence of a persistent mental substance:
Following [the idea of the soul], we will … connect all appearances, actions, and receptivity of our mind (Gemüth) to the guiding thread of inner experience as if the mind (Gemüth) were a simple substance that (at least in life) persists in existence with personal identity, while its states … are continuously changing. (A672/B700)
This passage is central to understanding how we cognize mental states without cognizing a given mental substance. It suggests that, for inner experience to be possible as the cognition of the relations between inner appearances, we must assume problematically (in the passage above indicated by ‘as if’) that these appearances all belong to one and the same simple substance as its mental states. Unlike the schematized category of substance, which describes (and therefore constitutes) the unity of a sensibly given substance, the idea of the soul prescribes a regulative unity that we ought to approximate by connecting inner appearances in accordance with it: ‘It is not from a simple thinking substance that we derive the inner appearances of our soul, but from one another in accordance with the idea of a simple being’ (A673/B701).
There is no mental unity (or substratum) given prior to any mental activity such that it can be subsumed under the schema of persistence. Rather, there are only given inner appearances that occur in the process of one’s mental activity, such as perceptions and wishes, joys and sorrows, hopes and fears. In order to make sense of those inner appearances as ‘parts’ of one’s mental life, i.e. mental states, we have to presuppose the idea of the soul as the form of a systematic mental whole. The idea denotes the ‘sum total’ of all inner appearances – past, present and future – which one aims to completely cognize.
The idea of the soul serves, more precisely, as the ‘analogue of a sensible schema’ (A655/B693). By employing the idea regulatively, we hypothetically presuppose a mental whole in analogy with the category of substance for the sake of unifying inner appearances, without asserting either a persistent substance on the basis of sensation or a noumenal substance on the basis of pure reasoning. The analogy consists in ‘think[ing] a relation to the sum total of all appearances, which is analogous to the relation that appearances have to one another’ (A674/B702), or more precisely, which is analogous to the relation of a persistent substance to its accidents.Footnote 43
Hence, I propose that inner experience does not lack a sensible explication of substance, but requires a different – analogical – kind of explication than that of outer substance. Since the empirical synthesis according to the schema of persistence cannot be completed for inner appearances, one requires an idea of reason as the guideline to approximate a mental unity throughout time and to form one’s empirical self out of given inner appearances. As the ‘analogue of a schema’, this idea is indispensable for inner experience to be cognition of mental states and their relations, without cognizing a persistent mental object.
5. Conclusion
In this article, I have examined two opposed interpretations of Kant’s theory of inner experience – the parity and the disparity interpretation – and developed an alternative view to overcome this dispute. I have argued that inner experience is empirical cognition of mental states (contra the disparity view), though not cognition of a mental substance (contra the parity view). My analysis has shown that, although inner experience shares the same representational forms as outer experience, it lacks an adequate sensible explication of a persistent mental substance in distinctively inner intuition. My view accommodates this disparity by appealing to the idea of the soul. If employed regulatively, this idea substitutes the empirical schematization of substance and provides the representation of a mental whole in relation to which mental states can first be cognized. The empirical self to be determined in inner experience thus emerges as a systematic whole of mental states that unfolds in time. This view of inner experience preserves the continuity with outer experience and explains its distinctive self-representational character. A full articulation of this view has to wait for a future occasion.Footnote 44
