HEIDEMANN, Dietmar H. (ed.). Kant and Non-Conceptual Content. [?]: Routledge, 2013. 227p. Resenha de: FAGGION, Andrea. Manuscrito, Campinas, v.36 n.2 July/Dec. 2013.
Kant and Non-Conceptual Content is based on a special issue of the International Journal of Philosophical Studies. With the exception of the last of the eight chapters, by Hannah Ginsborg, all the articles were initially presented at a workshop on Kant and non-conceptual content in May 2009 at the Department of Philosophy of the University of Luxembourg. The first chapter is an introduction by Dietmar Heidemann (‘Kant and Non-Conceptual Content: The Origin of the Problem’, pp. 1-10). In the following two chapters, almost half of the book, Robert Hanna presents his arguments in favour of a strong version of Non-Conceptualism that he considers as Kantian Non-Conceptualism (‘Beyond the Myth of the Myth: A Kantian Theory of Non-Conceptual Content’, pp. 11-86, and ‘Kant’s Non-Conceptualism, Rogue Objects, and the Gap in the B Deduction’, pp. 87-103). In their articles, Brady Bowman (‘A Conceptualist Reply to Hanna’s Kantian Non-Conceptualism’, pp. 104-133), Terry Godlove (‘Hanna, Kantian Non- Conceptualism, and Benacerraf’s Dilemma’, pp. 134-151), Stefanie Grüne (‘Is there a Gap in Kant’s B Deduction?’, pp. 152-177), Tobias Schlicht (‘Non-Conceptual Content and the Subjectivity of Consciousness’, pp. 178-207), and Hannah Ginsborg (‘Was Kant a Nonconceptualist?’, pp. 208-221) critically discuss Hanna’s claims.
Kant and Non-Conceptual Content certainly brings a contribution to the Kant scholarship regarding a crucial issue in the first Critique: the relation between concepts and perceptual experience. But its appeal is not merely historical. Since both conceptualists and nonconceptualists have claimed a Kantian root of their arguments, the answer to the question of whether Kant himself was a conceptualist or a non-conceptualist may clarify the framework of this contemporary controversy in philosophy of mind. Furthermore, the question of the role played by concepts in perceptions, if any at all, can only emerge once Kant has drawn a distinction between understanding and sensibility as two qualitatively different sources of representations. In other words, a polemic regarding non-conceptual content in mental representations cannot arise while the distinction between sensible and intellectual representations is drawn as a distinction of degrees of clarity and distinctness. This being so, to sum up, Kant has settled the philosophical paradigm inside which it makes sense to discuss nonconceptual content in mental representations (Heidemann 2013, pp. 2-4).
In order to make their point about Kant being a non-conceptualist, non-conceptualists strongly rely on the distinction itself between understanding and sensibility as independent and irreducible mental faculties (Heidemann 2013, p. 8). On the other hand, conceptualists claim that: ‘In his slogan, “Thoughts without content are empty, intuitions without concepts are blind,” Kant sums up the doctrine of conceptualism’ (Gunther 2003, p. 1). In brief, conceptualists interpret this famous slogan as a Kantian statement of the requirement of concepts for the intentionality or object-directedness of intuitions in such a way that sensible representations would lack representational content without the guidance of understanding (Heidemann 2013, pp. 1-2; Hanna 2013, p. 90). Such a conceptualist thesis would be further developed in the Transcendental Deduction of the Pure Concepts of the Understanding. In their turn, nonconceptualists claim that, in Kant’s view, concepts are required only ‘for the specific purpose of constituting objectively valid judgments‘ (Hanna 2013, p. 93) thus that the blindness of intuitions without concepts should be thought of as less literal than conceptualists suggest.
Following these remarks, it is easy to note that this discussion is about the unity of representations. The revolutionary Kantian distinction between sensible and intellectual representations is a distinction between different kinds of unity in mental representations. According to Kant, the unity of concept is the unity of ‘a representation that is contained in an infinite set of different possible representations (as their common mark), which thus contains these under itself‘ (CPR, A 25/B 40). On the other hand, the unity of intuition is the ‘unity of a set of representations within itself‘ (CPR, A 25/B 40). For this reason, the whole of intuition is a whole whose parts cannot be conceived of as independent representations, but only as components or limitations of the whole, while the conceptual unity is the unity of independent representations sharing a common mark (Heidemann 2013, p. 7 and Bowman 2013, p. 107). Hence, the following question is at issue in Kant and Non-Conceptual Content: once we have agreed that the unity of intuition is intrinsically different from the unity of concept, must we assert that the unity of intuition is also independent of the unity of concept?
Hanna’s answer for the question above is undoubtably positive. In order to make justice to his Non-Conceptualism, it is important to note that he is not saying that the manifold of sensible intuition could bear intentionality by itself. From his point of view, such a claim would
amount to a ‘”sensationalist” conception of non-conceptual content’ susceptible to the objection of adherence to the Myth of the Given (Hanna 2013, p. 14 and p. 75). Rather, on his account, Non-Conceptualism is a theory about ‘representational contents whose semantic structure and psychological function are necessarily distinct from the structure and function of conceptual content, and are not strictly determined by the conceptual capacities of […] minded animals’ (Hanna 2013, p. 20). This is his essentialist content Non-Conceptualism, also considered by Hanna as a Kantian Non-Conceptualism exactly due to the thesis regarding a qualitative difference in the semantic structure and the psychological function of concepts and intuitions. By emphasizing this difference, Hanna supports Russell’s classical distinction between knowledge by acquaintance and knowledge by description, even though he holds that ‘the primary objects of cognitive acquaintance are just individual macroscopic material beings’ (Hanna 2013, p. 40). While knowledge by description is always either ‘knowing X as F‘ (conceptual content) or ‘knowing that X is F‘ (propositional knowledge) (Hanna 2013, p. 41), knowledge by acquaintance (non-conceptual content) is always a context situated, egocentric perspectival, and intrinsically spatiotemporally structured knowledge-how (Hanna 2013, p. 18 and p. 60). This non-conceptual content is:
not ineffable, but instead shareable or communicable only to the extent that another ego or first-person is in a cognitive position to be actually directly perceptually confronted by the selfsame individual macroscopic material being in a spacetime possessing the same basic orientable and thermodynamically irreversible structure. (Hanna 2013, p. 41)
In the second part of his chapter ‘A Conceptualist Reply to Hanna’s Kantian Non-Conceptualism’, Bowman criticizes such a criterion of distinction between conceptual and nonconceptual content that is based on context independence versus the lack of such an independence (Bowman 2013, pp. 120-122). He also defends that ‘[k]nowing-how must […] be analysable in terms of knowing-that’ (Bowman 2013, p. 126). Rather than Non-Conceptualism, Bowman proposes a ‘conceptualist active externalism’ according to which our encounter with the world is already conceptually shaped and involves an at least quasi-conceptual activity on the part of the perceiver (Bowman 2013, pp. 120-121).
Whether deemed acceptable or not, the strong and essentialist version of Non- Conceptualism offered by Hanna is to be contrasted with state Non-Conceptualism, a version of Non-Conceptualism that Hanna believes to be unacceptable (Hanna 2013, p. 26 and p. 32). State non-conceptualists define Non-Conceptualism in terms of failure of concept-possession. Roughly speaking, the state non-conceputalist claim is that the savage who sees ‘a house from a distance, for example, with whose use he is not acquainted, […] admittedly has before him in his representation the very same object as someone else who is acquainted with it determinately as a dwelling established for men’ (Log, AA 09: 33). In other words, the savage from Kant’s example does not need to possess a corresponding concept to specify what he sees as the sight of a house in order to see the very same house as he would see whether he possessed the concept of ‘house’. The same claim is made about little infants and non-human animals regarding their encounters with the world, since they do not possess linguistic tools to judge or describe what they do encounter.
In accordance with Hanna, the problem with state Non-Conceptualism is that a Highly Refined Conceptualism entails that even if it can be shown that some human or non-human cognizers do actually achieve perceptual representations with intentionality and objectdirectedness without actually possessing or even being capable of possessing a corresponding concept for the identification of the perceived object, then Conceptualism is still not undermined (Hanna 2013, p. 32). This is because, as Hanna admits, ‘it is possible to have the ability to deploy and use a concept without also having possession of that concept. In other words, concept-possession requires more and richer abilities than the basic, minimal set of abilities required for concept-deployment and concept-use alone‘ (Hanna 2013, p. 24, see also p. 38 and p. 75).
Thus, that Kantian savage mentioned above could still be deploying a conceptual content, even though he did not possess the corresponding concept. Concept-possession, for Hanna, requires the capability of becoming self-consciously aware of the descriptive or intensional elements of the concept and carrying out analytic a priori inferences involving the concept (Hanna 2013, p. 24). On the other hand, concept-deployment and concept-use only require the ability to recognize an object when one perceives it and to distinguish the object from other sorts of things (Hanna 2013, p. 24). Hence, the truth of a Highly Refined Version of Conceptualism would require only that some possible non-contemporary or non-conspecific cognizer dispositionally possesses the concepts being used and deployed by the cognizer who does not herself possess or is capable of possessing those concepts (Hanna 2013, p. 33; see also Bowman 2013, p. 119).
This is why passages as that from Logic Jäsche quoted above are not enough to make Kant a non-conceptualist. After all, state Non-Conceptualism is compatible with Highly Refined Conceptualism. Nevertheless, as said above, Hanna considers that his content Non- Conceptualism has a Kantian provenance. In fact, the core of his arguments in favour of content Non-Conceptualism relies on Kant’s theory of incongruent counterparts (for instance, a hand and its mirrored image), that is supposed to show that ‘incongruent counterparts are qualitatively identical‘, thus, that ‘there is no descriptive difference between incongruent counterparts’, what amounts to say that there is no conceptual difference between any object and its incongruent counterpart, and, therefore, that if one can perceive the exact and real difference between incongruent counterparts, then ‘essentially non-conceptual content exists’ (Hanna 2013, p. 47).
Although, Hanna insists that he is ‘NOT denying that essentially non-conceptual mental contents can be conceptualized in some other non-essential, non-strictly determining sense’ (Hanna 2013, p. 20; see also p. 31), it is important to note that, according to his Non- Conceptualism, such a conceptualization may also be impossible. In short, essentialist content Non-Conceptualism leaves room for ‘rogue or elusive objects’:
there might then still be some spatiotemporal objects of conscious perception to which the categories either do not necessarily apply or necessarily do not apply: that is, there might be some ‘rogue objects’ of human intuitional experience that are not or cannot also be objects of human conceptual and judgmental experience… (Hanna 2013, p. 95)
Since the Transcendental Deduction of the Pure Concepts of the Understanding maintains that pure concepts ‘are necessary a priori conditions of the possibility of all objects of experience’ (Hanna 2013, p. 89), it precludes the possibility of rogue or ill-behaved objects of experience and, therefore, presupposes Conceptualism. Certainly, only on a conceptualist account of perceptual experience it is possible to guarantee that the unity of conscious perceptions of objects in space and time is determined by (and, therefore, always compatible with) the unity of concepts. Now, if the Transcendental Deduction presupposes Conceptualism, while Kant is actually a non-conceptualist, then there is a Gap in the Deduction (Hanna 2013, pp. 95-97). This is the claim in Hanna’s second chapter in Kant and Non-Conceptual Content.
In her reply to this chapter, Stefanie Grüne notes that there is a Gap in the Deduction if and only ‘if Kant is a strong content non-conceptualist, who believes that there are at least some perceptual states which contain nothing but essentially non-conceptual content’. (Grüne 2013, p. 159). Hanna must accept this claim, since, as we saw above, he believes that state Non-Conceptualism is reducible to Highly Refined Conceptualism all things considered. Furthermore, Hanna does attribute to Kant the strong content non-conceptualist view, as we also saw above. However, Hanna’s arguments for the last claim are only provided in his first chapter, whereas Grüne analyses only his second chapter. This is why she can conclude her own chapter by asserting ‘that characterizing Kant as the founder of Non-Conceptualism is not incompatible with believing in the success of the Transcendental Deduction’ (Grüne 2013, p. 171). She is referring to state Non-Conceptualism, while Hanna must be referring to content Non-Conceptualism, as we realize by combining the claims of his first and second chapter.
After making his point about the claimed Gap in the argument, Hanna goes even further and claims that the Deduction ‘had to fail, given Kant’s other deeper and larger cognitive and metaphysical commitments’ (Hanna 2013, p. 102). For instance, Kant had to make room for moral philosophy. Thus, in accordance with Hanna’s reading, the class of necessarily rogue objects is the same as the class of persons endowed with transcendental freedom (Hanna 2013, p. 99).
Brady Bowman, in his already mentioned chapter, provides good reasons for a Kantian philosopher being cautious about Hanna’s claims regarding rogue objects. As Bowman points out, if we accept that the general idea of rogue objects of experience is compatible with, even necessary for, the overall Kantian project, then ‘our actual experience could be thoroughly Humean and its seeming intelligibility merely contingent appearance’ (Bowman 2013, p. 110). If I understand properly Bowman’s objection, the issue here is that the acceptance of the possibility of rogue objects would imply the acceptance of the possibility that all objects of human experience could be rogue objects, hence, that any regularity observed so far could have been merely accidental, as is the constant conjunction of objects for Hume. In fact, a rogue object would be that cinnabar ‘now red, now black, now light, now heavy’ that Kant mentions in the A Deduction (CPR, A 100). As a result of such a behavior in the objects of perceptions, ‘even though we had the faculty for associating perceptions, it would still remain in itself entirely undetermined and contingent whether they were also associable’ (CPR, A 121-122). On Kant’s view, if they were not, there would be a definitive threat even to the identity of consciousness (CPR, A 122 and B 133), as it is explained by Tobias Schlicht in his chapter ‘Non-Conceptual Content and the Subjectivity of Consciousness’: ‘this consciousness of being the identical single subject can only arise in the light of a regular combination of representations’ (Schlicht 2013, p. 164).
Hannah Ginsborg, in the last chapter of Kant and Non-Conceptual Content, makes exactly the same point as Bownman regarding ‘the anti-Humean aspect of Kant’s view in the Critique‘ (Ginsborg 2013, p. 212) that would be lost if we accepted Hanna’s claims regarding rogue objects. On the other hand, Ginsborg believes to be necessary to deliver a conceptualist reading of the role of the understanding in the constitution of perceptual experience that, like the non-conceptualist view, respects ‘the primitive character of perception relative to thought and judgment’ (Ginsborg 2013, p. 210). In other words, according to her, the role of understanding in perceptual synthesis cannot consist ‘in the application of antecedently possessed concepts to whatever preconceptual material is presented to us by sensibility’ (Ginsborg 2013, p. 214). Rather, on her account, ‘to say that synthesis involves understanding is simply to say that it involves a consciousness of normativity’ (Ginsborg 2013, p. 214) that amounts to the subject taking ‘herself to be synthesizing as she ought’, without having antecedently grasped any concept, pure or empirical (Ginsborg 2013, p. 214).
To be certain, Hanna also claims that perception involves a consciousness of normativity. Nevertheless, according to him, ‘essentially non-conceptual content is inherently normative’ (Hanna 2013, p. 62). This being so, while Ginsborg considers the normativity in our perceptual experience as the distinctive mark of the understanding, Hanna sustains that the ‘essentially non-conceptual content has its own “lower-level spontaneity”‘ or ‘normativity’ (Hanna 2013, p. 74), the ‘body’s own reasons’ (Hanna 2013, p. 75). That amounts to say that Hanna disconnects the imagination from the understanding when it is merely a matter of explaining the constitution of perceptual experience. Hence, in this perceptual level, it does not seem to me that Hanna admits something like what Godlove, in his chapter, describes as judgments ‘about spatiotemporal somethings cognized independently of the application of concepts’ delivered by sensibility (Godlove 2013, p. 148). According to Hanna, the understanding is required only for the constitution of objectively valid judgments, while the nonconceptual content of perception is pre-discursive and pre-reflective (Hanna 2013, pp. 14-15, 41, 60, 67-78), even though it is still normative and spontaneous.
In favour of Ginsborg’s reading, there are the textual evidence of the Transcendental Deduction. After all, even Hanna is claiming that the Transcendental Deduction depends on a conceptualist view of perception. In favour of Hanna’s reconstruction of the synthesis of the intuition, there may be the possibility of avoiding objections of over-intelectuallization of the mind. Although, Ginsborg intends to preserve the primitive character of perception relative to thought, on her account, the subject cannot perceive something as an apple ‘without conceiving it to be an apple, and hence judging that is is an apple’ (Ginsborg 2013, p. 217). This reading could be vulnerable to a familiar kind of criticism regarding conceptualist accounts of perception: if mere intentionality or object-directedness requires conceptual normativity and if conceptual normativity requires an act of judgment, then animals and infants cannot perceive objects since they lack language tools to judge or to possess concepts. Ginsborg herself calls attention to the fact that, on her account, the association involved in our perceptions differs ‘fundamentally from those of animals’ in that our perceptions carry the consciousness of normativity (Ginsborg 2013, p. 217). Indeed, on Ginsborg’s view, animals and toddlers seem to be incapable of perceptions as mental contents with intentionality and object-directedness, for, after acknowledging that we share with animals ‘natural dispositions to associate representations in one set of ways rather than another’, Ginsborg claims that the fact that ‘our perceptual experience has representational content in the first place is not due to the particular ways that we associate our representations, but rather to the consciousness of normativity in those associations’ thus that the understanding is ‘responsible for these perceptions’ having representational content überhaupt‘ (Ginsborg 2013, p. 218).
Regarding Ginsborg’s reading of Kant, we can point out that Schlicht criticizes Kant for thinking that ‘this unification [of a phenomenal manifold of sensory or representational content] amounts to a conceptual synthesis of the non-conceptual content of intuition’ (Schlicht 2013, p. 197). According to him:
If a mental representation is only something for me if and only if intuitional content is brought under categories via spontaneous synthesis, then we are left with the problem that only adult human beings can have phenomenally conscious states. Non-human animals and human infants are excluded from the range of creatures for whom there is something it is like to experience their mental states since they plausibly lack these conceptual capacities. (Schlicht 2013, pp. 197-198)
Although Schlicht sounds reasonable when he adds that: ‘We would prefer an account according to which phenomenal consciousness is more widespread among the animal kingdom. That is the main reason why Kant’s solution seems unsatisfying’ (Schlich 2013, p. 198), we could ask if there would be no alternative between Hanna’s and Ginsborg’s reading such that the synthesis of intuition would be subjected to the understanding and at the same time would not involve the possession of concepts or the over-intellectualization of the mental content. Grüne seems to be offering this alternative. In a way that reminds us of Hanna’s distinction between concept-deployment and concept-possession, Grüne states that ‘the fact that one can have an intuition without possessing concepts does not have any implications for the question what kind of content the intuition has’ (Grüne 2013, p. 164). Her claim is that categories function as rules for synthesis of the sensible manifold into intuitions, whereas ‘synthesizing does not imply judging’ (Grüne 2013, p. 167). Following Longuenesse (1998), Grüne believes that, according to Kant’s Transcendental Deduction, one needs the possession of ‘clear’ concepts in order to be capable of judgment, but only the deployment of ‘dark’ concepts as rules for synthesis in order to be capable of perceptual experience (Grüne 2013, p. 176, n. 39). On this account, perhaps Kant could avoid both the Humean acceptance of rogue objects and the over-intellectualization of the mind.
In any case, as we saw above, Hanna claims that Kant does not only leave room for rogue objects, but also identifies persons endowed with transcendental freedom and necessarily rogue objects. Regarding Hanna’s conception of persons as rogue objects of experience, Bowman also offers a more orthodox Kantian point of view. According to Bowman, Kant is not looking for a way of qualifying persons as ill-behaved or as rogue objects of experience: ‘Instead, he looks for non-contradictory ways of attributing both natural causal determinism and freedom […] to the same objects’ (Bowman 2013, p. 111). The same line of objection is followed by Stefanie Grüne: ‘we can think of ourselves as free beings only if we regard ourselves as noumena, that is as objects, insofar as they are not objects of sensible intuition’ (Grüne 2013, pp. 165-166). To be fair, one needs to admit that this more orthodox reading presupposes the Kantian commitment to a strong version of the Transcendental Idealism, a commitment that Hanna is not willing to accept (Hanna 2013, p. 90). However, it is hard to see how Hanna could bring transcendental freedom to the empirical realm without destroying the natural determinism that Kant intends to preserve as well.
All things considered, perhaps Robert Hanna’s reading of Kant is a misunderstanding of his major philosophical project. But then, as Bowman has said, it is an ‘extraordinarily productive misunderstanding’ (Bowman 2013, p. 115). Kant and Non-Conceptual Content proves that.
GUNTHER, York H. Essays on Nonconceptual Content. Cambridge, MA: MIT Press, 2003. [ Links ]
KANT, Immanuel. Critique of Pure Reason (CPR). Translated and edited by Paul Guyer and Allen W. Wood. Cambridge: Cambridge University Press, 1998. [ Links ]
_____. Lectures on Logic (Log). Translated and edited by J. Michael Young. Cambridge: Cambridge University Press, 1992. [ Links ]
LONGUENESSE, Béatrice. Kant and the Capacity to Judge. Princeton: Princeton University Press, 1998. [ Links ]
Andrea Faggion – Departamento de Filosofia. Universidade Estadual de Londrina. Rodovia Celso Garcia Cid, km 380. 86057-970. Londrina, PR. BRASIL; Programa de Mestrado em Filosofia, Universidade Estadual de Maringá; [email protected]