ROBERT HANNA | Cognition, Content, and the A Priori | Oxford University Press 2015


By David Landy

My plan for my space here is to take Critique’s slogan—discussing new books on Kant and German Idealism—perhaps more seriously than its operators intended by bringing the reader up to speed on a discussion that Robert Hanna and I have been having for a few years now, and then attempting to move that discussion forward at least one more step. The discussion is one about the role of non-conceptual representation in Kant’s account of cognition, and so concerns one of the most fundamental issues in understanding the Critique of Pure Reason, and one that has been the subject of an enormous amount of recent scholarly work.

To put the disagreement between Hanna and me as succinctly as possible, whereas we both agree that Kant holds that we have non-conceptual representations, and that these are necessary for fully-fledged conceptual representation, we disagree about what such non-conceptual representations represent. I hold that all representation of complex states of affairs as complex requires concepts; Hanna holds that non-conceptual representations are also capable of representing complexes as complex. So, for Hanna, we non-conceptually represent objects as spatio-temporally related to one another and to ourselves, as causally efficacious alterations of an omnipresent æther, and as the targets of our desires and actions. For me, non-conceptual representations do not represent anything as anything. They are merely the mental proxies for worldly objects, and so stand in for objects in our mental lives, but do not stand for these objects. For me, to represent anything as anything, and especially to represent two items as related in any way, we must deploy concepts-qua-inferential-rules as the structure of a picture of those items as so related.

That, then, is the issue between us. We both agree that non-conceptual representations exist and are necessary for conceptual representation; we disagree about what and in what way such representations represent. Now for the arguments. The specific concrete discussion that Hanna and I have been having begins with Hanna’s 2008 paper, ‘Kantian Nonconceptualism’. That paper presents a version of Hanna’s Two Hands argument in support of the conclusion that non-conceptual content exists. Here is that argument:

  1. Incongruent counterparts are logically and metaphysically possible. [Premise, supported by Kant’s theory of incongruent counterparts and human geometrical intuition]

  2. Incongruent counterparts, by definition, are enantiomorphs. This entails that they are perceivable mirror-reflected property-for-property spatial duplicates that have exactly the same shape and size, and correspond point-for-point. In short, incongruent counterparts are qualitatively identical. [From 1]

  3. So by definition, there is no descriptive difference between incongruent counterparts. [From 2]

  4. Either of my hands and its corresponding mirror-image are actual examples of incongruent counterparts, and my own actual right and left hands are approximate incongruent counterparts. [Premise, supported by Kant’s theory of incongruent counterparts and human geometrical intuition]

  5. Therefore there is no descriptive difference between either one of my hands and its incongruent counterpart. [From 3 and 4]

  6. Therefore there is no conceptual difference between either one of my hands and its incongruent counterpart. In particular, the difference between either one of my hands and its incongruent counterpart could never be conveyed to someone else who was not directly confronted with these objects—e.g., it is impossible to convey the precise difference between one of my hands and its incongruent counterpart to someone else by means of language over the telephone. [From 5 and the Minimal Constraint]

  7. But I can directly perceive the difference between either of my hands and its incongruent counterpart, and can also directly perceive the difference between my right and left hands. [Premise, supported by Kant’s theory of incongruent counterparts and phenomenological introspection]

  8. Therefore essentially non-conceptual content exists. [From 6, 7, and the notion of essentially non-conceptual content]. (Hanna 2008:55–6)

Hanna’s version of the argument is roughly this. There is no way to distinguish one’s right hand from one’s left hand using only concepts. We can, however, so distinguish our hands. Therefore, we must do so via non-conceptual representations of them. Thus, Hanna sees himself as having refuted the conceptualist by proving that, as he puts it, “essentially non-conceptual content exists”. That is, since the conceptualist claims that all representations are conceptually structured, Hanna takes himself to have refuted the conceptualist by establishing that there are some representations that are entirely non-conceptual, and further, that such non-conceptual representations represent pairs of incongruent counterparts as distinct.

In my opinion, what Kant’s famous slogan about blind intuitions and empty thoughts actually means is that intuitions and concepts must always be combined together for the special purpose of making objectively valid judgements. But outside that context it is also perfectly possible for there to be directly referential intuitions without concepts (‘blind intuitions’, e.g. someone’s first cognitive encounter with a tree) (Hanna 2008:45).

This is Hanna’s full-dress conclusion. When he claims that the Incongruent Counterparts argument shows that essentially non-conceptual content exists, what he means is that it shows that there can be representations of our hands—’directly referential intuitions’—without concepts, and that it is by employing such representations, and no others, that we represent our hands as different. For Hanna, representing a pair of incongruent counterparts as distinct from one another is an entirely non-conceptual affair. We do this using only non-conceptual ‘directly referential intuitions’.

This brings us to the next step in the dialectic wherein I publish, in 2013, ‘What Incongruent Counterparts Show’, showing that the argument from Hanna’s paper was strictly speaking invalid, and attempting to further show that the premise that he would need to add to make it valid is one that Kant would reject. Specifically, I asserted (but did not have the space to defend) the exegetical thesis that Kant holds that all representations of complexes as complex are conceptually structured. My objection to Hanna’s argument was that whereas he was correct to see that incongruent counterparts can be used to show that it cannot be that all of our representations are entirely conceptual, he was incorrect to infer from this that some of our representations must be entirely non-conceptual. Specifically, I defended the position that what incongruent counterparts demonstrate is that our conceptual representations must have non-conceptual components. That is, what I argued is that while concepts alone are not sufficient for representing incongruent counterparts, they are still necessary for doing so. It is the structure that concepts impose on our non-conceptual representations that allows us to picture incongruent counterparts as spatially related to one another.

After 2013, Hanna and I both went off to write books. I wrote Kant’s Inferentialism: The Case Against Hume, which is more or less a book-length study defending and exploring the consequences of the thesis that for Kant all representation of a complex as complex requires concepts (understood as rules of inference) [ED: Landy’s own book is the subject of a forthcoming AmC discussion on this site]. Hanna on the other hand, wrote the book that we are now discussing, which is a little more ambitious, to say the least. Early on in that book, Hanna again presents the Two Hands Argument, this time adding in the premise that I pointed out was missing the first time around, and giving a brief defence of it in a footnote. My plan for the remainder of this essay is to discuss that premise and its attendant footnote. Here is the premise:

In order to represent a complex state of affairs as complex, concepts are not generally required. For example, the egocentrically centered primitive spatial difference between right and left, up and down, front and back, and so on, and also the egocentrically centered primitive temporal difference between earlier and later, now and then, and so on, are immediately given as structurally unified representations in pre-reflectively and non-self-consciously conscious experience. Hence these representations really can be given altogether without concepts. (p. 79)

So, I had suggested that Hanna’s original version of The Two Hands Argument was invalid because it failed to rule out a relevant alternative to his conclusion: that it might be the case that what is needed to represent two hands as occupying different spatial positions is not an entirely non-conceptual representation, but instead a conceptual representation with non-conceptual components. Hanna here explicitly addresses that lacuna by adding the premise to his argument that concepts are not required for representing two hands as occupying distinct spatial positions, and so can now validly conclude that non-conceptual representations can do so on their own.

Now, one thing to note here is that what Hanna actually concludes is stronger than what I just noted was authorised by his new premise. Hanna’s conclusion is not just that it is possible to represent incongruent counterparts via entirely non-conceptual representations. It is that “essentially non-conceptual content really exists” (p. 79), i.e. that we actually represent incongruent counterparts via entirely non-conceptual representations. In fact, later in the book, Hanna goes on to make the even stronger claim that we necessarily represent incongruent counterparts via entirely non-conceptual representations,[1] and as far as I can tell, nothing in the Two Hands Argument warrants that conclusion. Again, I agree that the Two Hands Argument supports the conclusion that our representation of two hands as occupying different spatial locations requires that our representations have a non-conceptual component (which I think is necessarily structured by concepts); what I deny is that these components themselves, and entirely independently of concepts altogether, represent such a complex. Hanna’s conclusion is that the only way to represent hands as occupying different spatial positions is to do so (at least at first) non-conceptually. As noted, though, the premise that he adds to this version of the argument merely asserts the possibility for his own claim—that some representations of complexes as complex can be represented without concepts—not its actuality or necessity.

That, however, is not the main thrust of my criticism of the new argument. Hanna has other resources available for supporting his conclusion (in the event that I turn out to be wrong about the necessity of concepts for representing complexes, that is). To proceed to the main thrust of my critique, we must turn to the footnote in which Hanna defends his new premise, which in itself is just a straightforward negation of the thesis that I have advanced. So, here is that footnote:

My earlier published formulation of The Two Hands Argument assumed this premise, but did not make it explicit. In ‘What Incongruent Counterparts Show’, however, David Landy correctly points out that without this premise, The Two Hands Argument is invalid. For if, necessarily, every representation of a complex state of affairs as complex requires concepts, then even if representing the difference between my right and left hands requires a non-conceptual component, it does not follow that it is essentially non-conceptual. So it fills a logical (or at least explanatory) gap for me to include this premise and its justification explicitly as the new step (8). Correspondingly, Landy’s thesis that necessarily, every representation of a complex state of affairs as complex requires concepts, makes the false assumption that all representation-as involves reflective self-conscious consciousness, which in turn involves concepts. But on the contrary, imaginational representations of complexes can pre-reflectively and non-self-consciously present the complexity of those complexes, via the figurative synthesis of the imagination, without reflectively or self-consciously (even in a dispositional sense) predicating it of those complexes via judgments, concepts, or inferences. This is what Kant calls the “aesthetic comprehension” or comprehensio aesthetica of the imagination, as opposed to the “apperceptive comprehension” or apperceptio comprehensiva of the imagination (R5661, 18:320) (CPJ 20:220). […] Correspondingly, there can be aesthetic or non-discursive clarity and distinctness in the cognitive phenomenology of intuitional representations that is not also logical or discursive clarity and distinctness in their cognitive phenomenology (JL 9:33-39). So, Landy’s thesis is false, the new step (8) is true, and The Two Hands Argument is/remains sound. (p. 79, n.84)

Ok. There is a lot going on in that footnote, and I correspondingly have a lot to say about it. To begin, there are two questions that need at least nominal distinguishing right at the outset. The first is the philosophical question of whether representing a complex as complex requires concepts. The second is the exegetical question of whether Kant held that representing a complex as complex requires concepts. Unfortunately, there is not space enough here to answer the former question, and it is not really the focus of Hanna’s footnote anyway.[2] There isn’t really enough space here to answer the exegetical question either, but we can at least make some headway in that respect by examining more closely some of the claims and texts that Hanna makes and cites here. Hanna’s main claim here is this one:

[I]maginational representations of complexes can pre-reflectively and non-self-consciously present the complexity of those complexes, via the figurative synthesis of the imagination, without reflectively or self-consciously (even in a dispositional sense) predicating it of those complexes via judgments, concepts, or inferences.

The idea here is that concepts are not required for representing the complexity of complex states of affairs per se, but only in making such representations available to a reflective and self-conscious thinking self. Even before we get to the issue of representing complexes as complex, we can notice that there is something at least a little strange about this suggestion insofar as it posits a kind of representation that is not available to self-consciousness, and that of course runs contrary to Kant’s famous dictum that

[t]he I think must be able to accompany all my representations; for otherwise something would be represented in me that could not be thought at all, which is as much as to say that the representation would either be impossible or else at least would be nothing for me. (B132)

That said, I think that Hanna is largely right on this point. Kant clearly does go on to include in his account of mental representation a kind of non-conceptual representation that is not per se available to introspection. Kant sometimes refers to these representations as ‘intuitions’, but when he is being more careful distinguishes them from intuitions proper and calls them ‘sensations’. The role that such representations play in his system is as theoretical-explanatory transcendental posits necessary for our having the kinds of conceptual representations that we do.[3] So, my beef with Hanna is not with his claim that sensations exist, or that they are necessary conditions for conceptual representation, or even in thinking of sensations as a species of representation.[4]

However, Hanna ends up having to answer a sticky question here that I do not. Since, for Hanna, non-conceptual, non-self-conscious representations represent complex states of affairs as complex, he must be able to say who the single subject of such a representation is. It is a familiar and fundamental point in Kant’s account of cognition that the representation of two items as related requires that this relation be represented in a single consciousness. If you have an idea of one of the relata, and I have the idea of the other, neither of us have an idea of these relata as being related. So who or what is it that is the single subject of Hanna’s non-conceptual representations? The natural Kantian answer to this question—the transcendental unity of apperception—is ex hypothesi unavailable to him. As having a non-conceptual representation is, for Hanna, being in what he calls The Grip of the Given—our bodies’ tracking of objects in the causal structure of the world—and, as Hanna repeatedly emphasises, the necessity of our being embodied, my guess here is that he wants to say that the single (non self-conscious) subject of a non-conceptual representation is the human body itself. While Hanna is certainly right that ‘we’ are essentially embodied, the proper Kantian take on that fact is surely that it is ‘we’ transcendental unities of apperception that are embodied, not that our bodies themselves are some sort of proto-unified subjects.[5] That is doubly so since Hanna’s take on physical objects implies that our bodies are themselves merely nominal limits on the material æther (Hanna 2006). So, in the end, I think that this line heads down a distinctly non-Kantian path.

One more philosophical hit-and-run before moving on to the texts. For Hanna, a non-conceptual representation represents its object as having those features that cause it to have the content that it does. To use an example of Hanna’s, it is the features of the martini sitting over there on the table that cause someone looking at it to have a representation that represents that martini as a martini over there on the table. Again, I agree with Hanna that the features of the martini cause me to form mental proxies for those features, but I take these to stand in for those features rather than stand for them. I do not hold that my having such mental states represents those features as features of the martini. Here is one reason for that: if such representations represent the martini as a martini, they do so in a way that does not allow, even in principle, for misrepresentation. Non-conceptual representations represent whatsoever features of the objects that de facto cause them to have the content that they do. Thus, they will necessarily (by definition) represent those objects as having the features that they in fact have. Of course, many philosophers have thought that the notion of a representation that cannot possibly misrepresent is a confused one, and I think this should give us pause about the precise role of non-conceptual representations in Hanna’s system.[6]

That objection, however, is also not my main focus. Again, what I want to focus on is Hanna’s claim that for Kant we can (and do and must) represent complex states of affairs as complex without concepts. Before I turn to Hanna’s defence of that claim, though, it will be worth looking at a few of the many places where Kant explicitly denies it. We can narrow our parameters a bit using the clues that Hanna offers of focusing on the so-called figurative synthesis of the imagination and aesthetic comprehension, i.e. by focusing on specifically aesthetic representations, i.e. representations of space and time.[7] So, I shall now present some passages of my own on those topics, and then with that context provided, return to the ones that Hanna presents for a closer examination.

We can begin with a passage from the Metaphysical Deduction in which Kant contrasts what is given by sensibility, a manifold of representations, with what is produced by the understanding, a representation of this manifold as a manifold.

Now space and time contain a manifold of pure a priori intuition, but belong nevertheless among the conditions of the receptivity of our mind, under which alone it can receive representations of objects, and thus they must always also affect the concept of these objects. Only the spontaneity of our thought requires that this manifold first be gone through, taken up, and combined in a certain way in order for a cognition to be made out of it. I call this action synthesis. / By synthesis in the most general sense, however, I understand the action of putting different representations together with each other and comprehending their manifoldness in one cognition. (A77/B102–3)

Receptivity—our capacity for being affected in such a way that we find ourselves with a manifold of representations—produces only such a manifold. In itself, such a manifold is not yet a representation of a complex state of affairs as complex. While receptivity merely contains a manifold of representations, spontaneity (or the understanding, or our faculty for using concepts) goes through and combines these representations to form a cognition. The process wherein this combination is carried out is called ‘synthesis’ and synthesis, finally, is the action of putting different representations together and comprehending their manifoldness. Synthesis, that is, is the formation of a representation of a complex as complex. This is what Kant means by “comprehending their manifoldness”. A cognition is a representation of a manifold (complex) as a manifold (complex). The synthesis that produces these cognitions is distinctly a function of the spontaneous understanding and therefore a product of conceptual activity. Thus, we can conclude that for Kant any representation of a complex state of affairs as complex will be a conceptual representation.

Now compare the last sentence of the above quotation with another from the Metaphysical Deduction, describing the nature of concepts:

All intuitions, as sensible rest on affections, concepts therefore on functions. By a function, however, I understand the unity of the action of ordering different representations under a common one. (A68/B93)

Synthesis is the “action of putting different representations together with each other and comprehending their manifoldness in one cognition”. A concept rests on “the unity of the action of ordering different representations under a common one”. The two processes are essentially the same: a manifold of representations is, through some action of the subject, combined and ordered to form a single cognition that represents that manifold as a manifold via the concept common to all its elements. So Kant’s thesis in the first passage is that all representations of a manifold as a manifold (a complex as a complex) are produced by the imposition of a conceptual structure on that manifold.

So much for synthesis in general. Now for space and time.

But space and time are represented a priori not merely as forms of sensible intuition, but also as intuitions themselves (which contain a manifold), and thus with the determination of the unity of this manifold in them (see the Transcendental Aesthetic). Thus even unity of the synthesis of the manifold, outside or within us […] can be none other than that of the combination of the manifold of a given intuition in general in an original consciousness, in agreement with the categories, only applied to our sensible intuition. Consequently all synthesis, through which even perception itself becomes possible, stands under the categories. (B160–1)

All synthesis, including the synthesis that is required for representing space and time as complex, stands under the categories.[8] And, of course, what it means for a representation to fall under the categories is precisely that it has the conceptual form that would allow it to appear in a judgement.

Here now is a passage from Kant’s unpublished prize-essay that makes exactly this point again, and does so in the context of discussing the representation of space in particular.

For we can represent a determinate space to ourselves no otherwise than by drawing it, i.e., by adding one space to the other, and so also with time. / Now the representation of a composite, as such, is not a mere intuition, but requires the concept of a compounding, so far as it is applied to the intuition in space and time. So this concept (along with that of its opposite, the simple) is one that is not abstracted from intuitions, as a part-representation contained in them, but is a basic concept, and a priori at that—in the end the sole basic concept a priori, which is the original foundation in the understanding for all concepts of sensible objects. (FM, AA 20:271)

Kant is explicit here that to represent any determinate space at all, one must draw it, and that drawing a space requires conceptual structuring. Here is yet another place where Kant states the thesis that what sensibility provides is merely the matter of cognition, but that it takes structuring by concepts to represent a complex state of affairs as complex, this time from his unpublished notes written during the time when he was composing the Critique:

We know any object only through predicates that we can say or think of it. Prior to that, whatever representations are found in us are to be counted only as materials for cognition but not as cognition. Hence an object is only a something in general that we think through certain predicates that constitute its concept. In every judgment, accordingly, there are two predicates that we compare with one another, of which one, which comprises the given cognition of the object, is the logical subject, and the other, which is to be compared with the first, is called the logical predicate. (Refl 4634, AA 17:616)

Again, Kant articulates his thesis that sensibility provides only manifolds, whereas it is the role of the understanding, of concepts, to use such manifolds to represent complex states of affairs as complex, including representations of objects as existing in space.

We have seen Kant say as much in the Metaphysical Deduction, in the prize essay, and in his notes. Here is yet another passage, this time a rather famous one from the B-Deduction, which presents almost the exact same dialectic:

The manifold of representations can be given in an intuition that is merely sensible, i.e., nothing but receptivity, and the form of this intuition can lie a priori in our faculty of representation without being anything other than the way in which the subject is affected. Yet the combination (conjunctio) of the manifold in general can never come to us through the senses, and therefore cannot already be contained in the pure form of sensible intuition; for it is an act of the spontaneity of the power of representation, and, since one must call the latter understanding, in distinction of sensibility, all combination […] is an action of the understanding, which we would designate with the general title synthesis in order at the same time to draw attention to the fact that we can represent nothing as combined in the object without having previously combined it ourselves, and that among all representations combination is the only one that is not given through objects but can be executed only by the subject itself, since it is an act of its self-activity. (B12930)

All combination is an action of the understanding, which is a capacity for using concepts, and the activity of which is called synthesis. Receptivity, or sensibility, produces only a manifold of representations: it is the way in which the subject is affected. To represent anything as combined, this representation must be produced by an activity of the subject itself, an activity of understanding, a conceptual activity. In the idiom that we have been using, all bringing together of a manifold of representations into a single representation of a complex state of affairs as complex is conceptual. We can represent nothing as combined in the object without this act of the subject itself. No complex object can be represented as complex without the conceptual activity of the representing subject. These passages are merely a few of those in which Kant declares his commitment to this thesis explicitly and emphatically.[9]

That is some of the evidence in favour of my thesis. Hanna likewise presents evidence supporting his denial of that thesis, and it is to that evidence that I shall now turn. We can begin by recalling that Hanna cites the so-called figurative synthesis as the source of the non-conceptual representations that represent items in space and time, and offers the following explication:

This is what Kant calls the “aesthetic comprehension” or comprehensio aesthetica of the imagination, as opposed to the “apperceptive comprehension” or apperceptio comprehensiva of the imagination (R 5661, 18:320) (CPJ 20:220). […] Correspondingly, there can be aesthetic or non-discursive clarity and distinctness in the cognitive phenomenology of intuitional representations that is not also logical or discursive clarity and distinctness in their cognitive phenomenology (JL 9:33–39).

As Hanna does, we can begin with Reflexion 5661:

The action of the imagination in giving an intuition for a concept is exhibitio. The action of the imagination in making a concept out of an empirical intuition is comprehensio. / The apprehension of the imagination, apprehensio aesthetica. The composition of it, comprehensio aesthetica (aesthetic comprehension): I grasp the manifold together in a whole representation and thus it acquires a certain form. (Refl, AA 18:320)

Hanna presents this note as the place at which Kant coins the terms ‘comprehensio aesthetica’ for the non-conceptual representation of a complex as complex. There is no indication whatsoever, though, that this is what Kant intends here. In fact, Kant distinguishes the comprehensio aesthetica from the apprehensio aesthetica and one way of understanding that distinction is precisely that the latter corresponds to having a manifold of representations whereas the former corresponds to uniting a manifold using concepts. Kant explicitly links concepts and comprehension in the second sentence here, and in the last clause of the final sentence attributes a unity to the representations produced by the comprehensio aesthetica, and notes that the comprehensio aesthetica gives such representations forms. Recall that our earlier examination of texts showed that unity is always the result of combination, which is the provenance of the understanding and therefore of concepts. Forms are also repeatedly tied to concepts throughout Kant’s corpus.[10] So, far from supporting Hanna’s thesis, this text can easily be read, especially in concert with the other texts that we have examined (and the many that we have not!), as supporting the exactly opposite view. Comprehensio aesthetica is the uniting of the deliverances of sensibility via concepts.

Here then is the next passage that Hanna cites, this time from the first introduction to the Critique of the Power of Judgement:

To every empirical concept, namely, there belong three actions of the self-active faculty of cognition: 1. the apprehension (apprehensio) of the manifold of intuition; 2. the comprehension, i.e., the synthetic unity of the consciousness of this manifold in the concept of an object (apperceptio comprehensiva); 3. the presentation (exhibitio) of the object corresponding to this concept in intuition. (EEKU, AA 20:220)

Hanna’s citation of this passage is even stranger than his citation of the previous one, as here the comprehension is explicitly identified with “the synthetic unity of the consciousness of this manifold in the concept of an object” and there is no mention whatsoever of this being a specifically aesthetic faculty. Comprehension is self-conscious, conceptual, and only thereby object-directed.

Now, the sentence that follows the above links apprehension, not comprehension, to the imagination, so perhaps Hanna is thinking that it is via apprehension that we non-conceptually represent a complex as complex.

For the first action imagination is required, for the second understanding, for the third the power of judgment, which, if it is an empirical concept that is at issue, would be the determining power of judgment. (EEKU, AA 20:220)

Notice, though, that Kant describes all three of these aspects of empirical concepts as activities and as belonging to “the self-active faculty of cognition”, which is pretty clearly the understanding (broadly construed) rather than sensibility. The key here is to recognise that, as Kant tells us, the imagination, the understanding (narrowly construed), and the power of judgement are all really just different roles that the understanding (broadly construed), qua our inferential faculty, plays in cognition.

The same function that gives unity to the different representations in a judgment also gives unity to the mere synthesis of different representations in an intuition, which expressed generally, is called the pure concept of understanding. (A79/B104–5)

Now since all of our intuition is sensible, the imagination, on account of the subjective condition under which alone it can give a corresponding intuition to the concepts of understanding, belongs to sensibility; but insofar as its synthesis is still an exercise of spontaneity, which is determining and not, like sense, merely determinable, and can thus determine the form of sense a priori in accordance with the unity of apperception, the imagination is to this extent a faculty for determining the sensibility a priori, and its synthesis of intuitions, in accordance with the categories, must be the transcendental synthesis of the imagination, which is an effect of the understanding on sensibility and its first application (and at the same time the ground of all others) to objects of the intuition that is possible for us. (B151–2)

So, insofar as what Kant is calling apprehension here amounts to more than merely having a manifold of representations, it is precisely because the imagination, the understanding, and concepts are deployed.

Two citations down, one to go. The final passage that Hanna cites is the section of the Jäsche Logic in which Kant gives his (in)famous example of the difference between the representation of a house had by someone who is not acquainted with the use of the concept of a house and that had by someone who is so acquainted (Log, AA 9:33–9). I have addressed this example elsewhere (Landy 2015:152–4), so I shall focus here on the remainder of the section. Hanna takes this section to support the thesis that “there can be aesthetic or non-discursive clarity and distinctness in the cognitive phenomenology of intuitional representations that is not also logical or discursive clarity and distinctness in their cognitive phenomenology”. Right off the bat, Hanna’s citing the clarity and distinctness of non-conceptual representations is odd since Kant’s first claim after the aforementioned example is that

[i]f I am conscious of the representation, it is clear; if I am not conscious of it, obscure. (Log, AA 9:33; my underlining)

In the first paragraph of this section Kant tells us of consciousness that

[c]onsciousness is really a representation that another representation is in me. (Log, AA 9:33)

So much for the clarity being a feature of non-reflective, or non-self-conscious, non-conceptual representation. Kant also makes it clear that not only is the subject matter of logic such self-conscious representations, but also that these go hand-in-hand with conceptual as opposed to non-conceptual representations.

Since consciousness is the essential condition of all logical form of cognitions, logic can and may occupy itself only with clear but not with obscure representations. In logic we do not see how representations arise, but merely how they agree with logical form. In general[,] logic cannot deal at all with mere representations and their possibility either. (Log, AA 9:33)

Logic concerns only clear, i.e. conscious, representations, and explicitly “cannot deal at all with mere representations” (emphasis added). Mere representations are sensations, representations that stand in for (are caused by the features of) their objects without standing for them (representing their objects as anything). So much for there being a clarity particular to non-conceptual representation.

What about distinctness? Well, Kant notes that it is only clear representations, and so conscious conceptual ones, that can be distinct or indistinct, so the difference between these will be of no help here either. Hanna appears to be finding succour in Kant’s differentiation of sensible distinctness and intellectual distinctness (Log, AA 9:25), but that is a non-starter as well. Sensible distinctness is not distinctness sans concepts, but is rather distinctness where a concept is deployed, and the objects pictured by that concept are also immediately perceptible as such. Kant’s examples of sensibly indistinct representations are representing something as a house without seeing its windows, and representing the Milky Way as a whitish streak in the sky without perceiving the individual stars that constitute it. The contrasting sensible distinct representations are representing a house using the concept house to unite the manifold of sensations that now include those standing in for the windows of the house, and representing the Milky Way using the concept galaxy to unite the manifold of sensations standing in for the stars seen through a telescope. Nowhere does Kant so much as mention non-conceptual representations themselves representing a complex as complex here.

Intellectual distinctness, in contrast to sensible distinctness, is just the distinctness that accrues to concepts in virtue of their being fully analysed, e.g. knowing that ‘unmarried’ is part of the concept ‘bachelor’. Nowhere in the articulation of the difference between these two kinds of distinctness does Kant commit himself to the thesis that “there can be aesthetic or non-discursive clarity and distinctness in the cognitive phenomenology of intuitional representations that is not also logical or discursive clarity and distinctness in their cognitive phenomenology”. In fact, in explicitly tying both clarity and distinctness to logical form and consciousness, Kant does the exact opposite! He consistently maintains that clarity and distinctness are features exclusive to self-conscious conceptual representations.

Thus, I conclude that the texts that Hanna cites in support of his contention that Kant denies that all representations of complex states of affairs as complex require concepts are at best ambiguous, but are more plausibly read as supporting exactly the opposite exegetical thesis. As we have seen throughout our consideration of as wide an array of texts as one could hope for, Kant holds that all representation of a manifold as a manifold, all combination, all synthesis, requires concepts.[11]

Invited: 16 October 2015; received: 4 January 2016.


[1] “More precisely, autonomous essentially non-conceptual content is presupposed by all rational conceptual/propositional content whatsoever” (Hanna 2015:108).

[2] I have attempted to provide at least the beginning of an answer to this philosophical question in the second chapter of Landy (2015) by considering the grounds on which Kant answers it in the affirmative in replacing Hume’s theory of mental representation with his own brand of inferentialism.

[3] Oddly enough, it is McDowell, the great foe of everything non-conceptual, who notices that while such representations are not available to introspection insofar as they are playing this transcendental role, they can come to be observed by the properly trained empirical consciousness, just as a properly-trained scientist can observe not just a streak in a cloudy box, but also a subatomic particle passing through a gas chamber. Of course, in that case, there will be further sensations-qua-transcendental-posits that are necessary for observing these. See McDowell (2013:447) and Rosenberg (2007:277).

[4] See e.g. A320/B377. I discuss this sense of ‘representation’ in some detail in Landy (2009) and Landy (2015:154–63).

[5] Longuenesse (2007) and Ginsborg (2015) are recent attempts to forge this connection.

[6] I would here refer the reader to Sellars’ (1991:132) discussion of “the classic concept of a sense datum [as] a mongrel resulting from a crossbreeding of […] the attempt to explain the facts of sense perception in scientific style” and the attempt to explain their cognitive, semantic, or epistemic roles. But, of course, Hanna rejects the Myth of the Given as a myth in its own right.

[7] I have provided what I hope is a detailed account of what conceptual synthesis is in Landy (2015). I there note that I have yet to see Hanna, or any of the other proponents of a non-conceptual figurative synthesis, provide any detailed account of what it is or how it functions. Until such an account is presented and defended, the figurative synthesis remains a mere via negativa, a synthesis that we can understand only by contrast with conceptual synthesis. I count that as a strong indicator that Kant would countenance no such thing.

[8] While this passage seems to be a univocal endorsement of the complex representation thesis, it is also accompanied by an infamous footnote, which seems to support the thesis that Kant endorses a kind of synthesis that is pre-conceptual. Limits of space and human decency prevent entering into the fray regarding that footnote here, but I can refer the curious or masochistic reader to my previous discussion of it in Landy (2015).

[9] Here is a brief list of further examples for the as yet unsatisfied reader: V-Lo/Wiener, AA 24:856; Refl 4681, AA 17:668; A97; A120; B134–5; B143; B162n; FM, AA 20:271; Br, AA 11:376.

[10] See Pippin (1982), which is a neglected, but excellent study.

[11] It is worth noting that Hanna also cites the support of various empirical investigations, namely, phenomenological introspection and empirical data in cognitive psychology. I confess to being unclear what the role of such evidence is here since Kant’s is not an empirical question about what happens in brains, etc., but rather a philosophical analysis of the notion of a representation, or an object-concept, itself. That is, Kant is after a transcendental account of our cognitive faculties. So, for example, that an infant or non-human animal has reliable differential response dispositions to spatially and temporally situated objects does not yet show that such a creature represents such objects as spatial or temporal.


Hanna, R. (2006), Kant, Science, and Human Nature (Oxford: Oxford University Press).

——— (2008), ‘Kantian Nonconceptualism’, Philosophical Studies 137: 41–64.

Landy, D. (2009), ‘Sellars on Hume and Kant on Representing Complexes’, European Journal of Philosophy 17(2): 224–46.

——— (2013), ‘What Incongruent Counterparts Show’, European Journal of Philosophy 21(4): 507–24.

——— (2015), Kant’s Inferentialism: The Case Against Hume. New York: Routledge.

Longuenesse, B. (2007), ‘Kant on the Identity of Persons’, Proceedings of the Aristotelian Society 107: 149–67.

Ginsborg, H. (2015), The Normativity of Nature: Essays on Kant’s Critique of Judgement (Oxford: Oxford University Press).

McDowell, J. (2013), Having the World in View: Essays on Kant, Hegel, and Sellars (Cambridge, MA: Harvard University Press).

Pippin, R. (1982), Kant’s Theory of Form (New Haven: Yale University Press).

Rosenberg, J. (2007), Wilfrid Sellars: Fusing the Images (Oxford: Oxford University Press).

Sellars, W. (1991), ‘Empiricism and the Philosophy of Mind, in Science, Perception and Reality (Atascadero, CA: Ridgeview).

© David Landy, 2017.

David Landy is an Associate Professor of Philosophy at San Francisco State University. He works primarily on the history of modern philosophy, especially Hume and Kant, and also has interests in German Idealism and the work of Wilfrid Sellars. He has published in, among other places, the European Journal of Philosophy, History of Philosophy Quarterly, InquiryJournal of the History of Philosophy and Kantian Review. He is also the author of Kant’s Inferentialism. The Case Against Hume (Routledge 2015), shortly the subject of an Author-Meets-Critics discussion on this site, with Steven M. Bayne, Anil Gomes and Tim Jankowiak as discussants.