STEFANIE GRÜNE | Blinde Anschauung. Die Rolle von Begriffen in Kants Theorie sinnlicher Synthesis | Klostermann 2009
By Colin McLear
Stefanie Grüne’s book Blinde Anschauung is a rich, deep, and wonderfully concise examination of issues pertaining to Kant’s theory of cognition in general, and his view on perception in particular. In the course of articulating her argument, she does a great deal to advance our understanding of the structure of Kant’s views on cognition, the nuance and variety of positions he expresses (concerning which she provides a heroic amount of clarification and regimentation), and their ultimate unity and coherence. In what follows I focus on only a limited part of the area that she covers, and primarily only on aspects of her positive view. First, however, I provide some summary of relevant portions of Grüne’s argument.
1. The Inconsistent Triad
Grüne argues (p. 17) that there are sound textual grounds for reading Kant as endorsing an inconsistent triad of positions:
(1) The Independence Thesis concerning the independence of sensibility from understanding
— Intuitions and concepts are representations, for which two distinct cognitive capacities are responsible. Sensibility provides intuitions. The understanding provides concepts.
(2) The Genetic Primacy of Intuition Thesis
— The understanding can only construct concepts under the condition that sensibility has previously provided intuitions.
(3) The Genetic Primacy of Concepts Thesis
— Sensibility provides intuition only in cooperation with the understanding. Intuitions arise only when one already possesses concepts, which function as rules for the sensible synthesis of sensory representations into intuitions.
Grüne sees non-conceptualist readings as rejecting (3), and conceptualist readings as rejecting (2), or (in the case of Ginsborg 2006) arguing that there is no primacy claim in Kant at all. Grüne’s solution is to say that we need to distinguish different ways in which a concept or intuition might have primacy. Her proposal is that Kant posits a distinction between obscure (dunkel) concepts, clear (klar) concepts, and distinct (deutlich) concepts. Obscure concepts are genetically prior to intuition, but clear and distinct concepts are not. The possession conditions for obscure concepts are extremely liberal. This allows Grüne to articulate an intermediary position between traditional forms of conceptualism and non-conceptualism. On the one side, she can recognize the (in one sense) primitive nature of intuition. On the other side, she can show that conceptual capacities are necessary for having any intuitions at all. This position thus encompasses a wide variety of texts and seemingly harmonizes the two strongest forces in the opposing interpretive stances towards Kant’s cognitive theory, providing a version of conceptualism (‘obscure’ conceptualism) that has much in common with non-conceptualism, seemingly without its supposed vices.
2. Grüne on Intuition
Grüne thinks that all empirical intuitions are generated by means of a synthesis performed on purely subjective sensory states (equivalent to ‘raw feels’ such as pains or pleasures) (pp. 40ff., 63n.62, 153n.8). Intuitions are conscious (p. 39) objective representations (pp. 40–1) that are inherently complex (pp. 61, 70) and may be had independently of the existence of any object they might represent (pp. 42–3).
The conscious character of intuition is understood specifically in terms of clarity. An intuition is clear when it allows the subject to distinguish the object occupying one spatial or temporal region from another, be it another object, or simply another part of the subject’s environment (pp. 74, 81).
The objective character of an intuition is understood in terms of its being a kind of intentional state (p. 40). Grüne bases this claim on the canonical Stufenleiter passage in which Kant distinguishes different types of representation (A320/B376–7). However, she also cites (p. 41) an important note from Kant’s Nachlass where he says:
What is an object? That whose representation is a sum of several predicates belonging to it. The plate is round, warm, made of tin, etc. Warm, round, being made of tin, etc., are not objects, although the warmth, the tin, etc., indeed [are]. An object is that in the representation of which various others can be thought as synthetically combined […]. (R6350, AA 18:676)
According to Grüne, relation to an object thus consists in the representation of particular features as unified in one subject. Neither this subject (nor its features) need exist (pp. 42–3).
If intuitions may be had independently of the existence of their objects, we need some characterization of how we determine what an intuition is an intuition of. This, Grüne argues, is done by means of the content of the intuition. Intuitions represent their objects via ‘marks’ understood in terms of the intentional content of a mental state (p. 53). The content of intuition consists of ‘intuitive marks’ glossed in terms of the representation of tropes (pp. 65–71), singular features which only one object may have. Thus, the central difference between the content of an intuition and that of a concept is that intuitions represent objects by means of their tropes, while concepts represent objects by means of general characteristics which one or more objects may share. One thus intuits the particular greenness of a tree, in contrast to thinking of this particular bit of greenness in terms of the predicate green, which is instanced in grass and mint ice-cream, among other things.
This account of intuition plays two central roles in Grüne’s overall interpretation. First, against ‘judgement-theoretic’ conceptualist accounts, it presses the claim that the representation content of intuition is importantly different from that of conceptual thought or judgement. For this reason intuition (and thus perceptual experience) cannot and should not be assimilated to any form of conceptual thought. This is an important step towards vindicating the genetic primacy of intuition (p. 54).
Second, against the non-conceptualist, Grüne’s proposed account presses the importance of a conceptually-guided synthesis in the generation of intuition out of the purely subjective and non-intentional mental states which arise in the subject as it is affected by external objects. In this way, Grüne’s interpretation puts pressure on both sides of the traditional debate concerning the (non)conceptual structure of intuition and the sense in which, in Kant’s famous phrase, ‘[t]houghts without content are empty, [and] intuitions without concepts are blind’ (A51/B75).
In what follows, I present some issues that arise concerning various aspects of this account. First, I ask why we should think that intuition only arises as a product of synthesis (Section 3). Second, I argue that Grüne’s interpretation fails to account for one of the most pressing considerations motivating the non-conceptualist interpretation—viz., the possibility of non-rational animal perception (Section 4). Third, I provide some reasons to question a presupposition of Grüne’s account, namely, that intuitions should be understood as intentional states which relate to their object by means of correctness conditions set by their content (Section 5). Grüne is by no means alone in accepting this presupposition, but I believe there are both textual and philosophical reasons for denying that Kant held such a view. Finally, I argue that Grüne’s interpretation of (empirical) intuition as independent of the existence of its object is unconvincing (Section 6).
3. The Generation of Intuition
In setting up her triad of inconsistent claims, one central assumption is that intuition is the product of a sensory synthesis. Certainly, if intuition depends on such a synthesis we have a clear question of the genetic priority of concepts over intuitions. But if we do not accept the antecedent then the issue of priority is no longer very pressing, at least for the case of the existence of intuitions in general. So why should we accept the antecedent?
Grüne motivates her position textually and via broader philosophical argument. Textually, she notes that numerous passages in the A-edition of the Transcendental Deduction seem to indicate that intuitions are complex representations consisting of a multiplicity of ‘partial’ representations [Teilvorstellungen] (pp. 61–2; 157n.23). Kant says that ‘every intuition contains a manifold in itself’ (A99), and that the drawing of a line in thought, or the representation of number requires that one ‘first grasp one of these manifold representations after another in [one’s] thoughts’ (A102).
I do not find these textual claims convincing. Against texts such as A102, as Grüne herself notes (p. 62), there is no mention of intuition but rather only thought and thinking. If Kant is making a phenomenological point here about the conditions necessary for the experience of a line, he is making an implausible point. As Van Cleve trenchantly puts it,
[w]hen I imagine a line, I am aware of at least some of its parts together. In Kant’s view, this can come about only through synthesizing those parts (or one’s representations of them). ‘We cannot think a line without drawing it in thought, or a circle without describing it’ (B154). But when I imagine a line, I am aware of no such successive generation; I simply plop the whole line down at once.
Now, Kant could indeed be making the implausible claim to which Van Cleve objects, or he could be making a less implausible claim concerning the conditions for representing a spatially unified aggregate (e.g., perceived spatially adjacent points) as a single determinate object (e.g., a line), which could then be utilized in geometrical reasoning. Kant could concede that the experience of a simultaneity of spatially adjacent points is possible without a prior synthesis while denying that this is sufficient for the perception of anything over and above this aggregate of points—viz., the line. The experience of an entity which has unity over and above the phenomenological unity of spatial closeness requires a further synthesis. Grüne might object to this proposal because it allows that there could be synthesis-independent objective representation, and thus intuition, but it seems to me to make better sense of the various texts where Kant is concerned, above all, with our representation of geometric and arithmetical unities in thought, without thereby making what seem to me phenomenologically implausible demands on perception.
As for Kant’s point at A99, it is much less obvious that Kant is making any claim about the generation of intuition from a synthesis of apprehension once the line is put in its full context:
Every intuition contains a manifold in itself, which however would not be represented as such if the mind did not distinguish [unterschiede] the time in the succession of impressions on one another; for as contained in one moment, no representation can ever by anything other than absolute unity. (A99)
As Clinton Tolley has recently noted (Tolley 2013), Kant does not so much seem to be making a point here about the generation of intuition but rather to be signalling the difference between having an intuition and representing it to oneself in thought as a unity. The synthesis of apprehension which aims at intuition is thus not a necessary condition of the intuition’s existence, but rather only a condition of its being represented as having a particular determinate content. As Tolley puts it,
Kant’s point here is thus that the synthesis of apprehension is required only if we wish to ‘apprehend’ an intuition as containing a particular, determinate manifold—that is, only if we wish to have consciousness of a particular manifold as ‘contained in one representation’. In other words, while the synthesis of apprehension ‘in intuition’ is surely a synthesis that is ‘aimed at intuition’, it is not at all one that makes up or puts together an intuition, or puts something ‘in’ intuition, in the first place.
As Tolley further points out, Kant argues only that a threefold synthesis is necessary for the consciousness of the manifold in an intuition—what Kant calls ‘perception’ [Wahrnehmung]. Synthesis is that ‘through which perception, i.e., the empirical consciousness of [the manifold] (as appearance), becomes possible’ (B160). So all of the relevant points Kant makes in the Transcendental Analytic concern the representation, in thought, of intuition, not the generation of intuition itself. Grüne mentions this passage (p. 161) but does not discuss the possibility of such an alternate reading. She instead takes the text to suggest that intuition—an intentional state—is generated via synthesis from a complex of non-intentional states.
I think that, in order to get such a reading from the text, we have to bring Grüne’s broader philosophical apparatus to bear, and show that the application of this apparatus is by no means obligatory. For Grüne, intuition is always the product of the unification of quantitatively (and perhaps qualitatively) distinct sensory states (pp. 64, 70, 155–8). These sensory states are, prior to synthesis, non-intentional (pp. 40, 63, 153) purely ‘raw’ or qualitative feels akin to pain or pleasure. Grüne motivates this reading in part by making use of Kant’s distinction between ‘objective’ sensation and ‘subjective’ feeling [Gefühl] in the third Critique (AA 5:206). While I think she is right that Kant makes a distinction between states that relate to objects (e.g., colour, texture) and those which can only ever relate to the subject having them (e.g., pain, pleasure), I don’t see that this licenses the further claim that objective states are constructed from operations on purely subjective ones. Certainly, such readings of Kant are not uncommon, and substantive argument against such a reading would require more space than I can give it here, but I don’t see how we can get the reading out of Kant’s distinction between subjective and objective states.
As a final point for this section I want to suggest that Grüne’s interpretation sits poorly with Kant’s characterization, in the Transcendental Aesthetic, of our pure intuitions of space and time. I hope that this is not unfair to Grüne’s overall position, since she is primarily concerned with empirical intuition. But I take Kant’s argument concerning empirical intuition to be importantly unified with his account of pure intuition, and (as others have argued) his argument in the B-deduction seems to depend on such a unified account.
As is well known, there are a variety of passages in both editions of the Transcendental Deduction (A99, A102, B136 note, and B160–1) that have been taken by many to show that the representations of space and time themselves depend on a synthesis by the understanding, or the understanding operating in the guise of the imagination.
According to Grüne, the content of every intuition is quantitatively complex because all of space and time is infinitely divisible, and the representations of space and time must have a correlative structure (pp. 65, 70). As far as I can tell, Grüne thus thinks that the pure intuitions of space and time are infinitely quantitatively (though not necessarily qualitatively) complex.
Grüne denies (or is tempted to deny) the thesis that quantitative complexity is itself the result of synthesis (pp. 155–6, and note 18), and correctly points out that the extremely controversial footnote at B160–1 at best shows that a synthesis is responsible for the status of intuitive representation as conscious and objective and not for its character as quantitatively complex (p. 156).
Now, Grüne is mostly concerned with the apparent problem that the non-intentional status of unsynthesized sensory representation would seem to entail that there is no quantitative complexity prior to synthesis (157). However, a broader problem looms.
On the one hand, Grüne’s interpretation allows us to makes sense of part of the infamous §26 footnote, where Kant says:
Space, presented as object (as we are actually required to represent it in geometry), contains more than [the] mere form of intuition—viz. it contains also the grasping-together [Zusammenfassung] of the manifold given according to the form of sensibility, in an intuitive representation—so that the form of intuition gives us merely a manifold, but formal intuition gives us unity of representation. (B160–1, note)
Now, since Grüne holds that all relation to an object depends on synthesis, she seems well-situated to argue that the conditions under which one may represent space (and time) as an object, as is done in geometry, require synthesis. But I think her account faces two problems. First, her account seems unable to construe the pure form of intuition as itself an intuition other than as the outcome of some synthesis. In order to avoid this awkwardness, Grüne would seem to need to collapse Kant’s distinction between the form of intuition (understood as, in the human case, the pure intuitions of space and time) and the ‘formal’ intuition spoken of in the footnote.
The second, and I think more pressing issue, is that Grüne’s position would seem to require that the representation of space (and time), either as form or as formal intuition, is a synthetic construction from quantitatively complex sensory representations. But this goes against Kant’s position, in the Transcendental Aesthetic, that space and time cannot be constructed from any antecedent part. Instead, spaces and times are one and all limitations of the one space and one time:
Space is not a discursive or, as we say, universal concept of things as such; rather, it is a pure intuition. For, first, we can represent only one space; and when we speak of many spaces, we mean by that only parts of one and the same unique space. Nor, second, can these parts precede the one all-encompassing space, as its constituents, as it were (from which it can be assembled); rather, they can be thought only as in it. Space is essentially one; the manifold in it, and hence also the universal concept of spaces as such, rests solely on [our bringing in] limitations. It follows from this that, as far as space is concerned, an a priori intuition of it (i.e., one that is not empirical) underlies all concepts of space. (A24–5/B39; my underlining)
The representation of space (and time) is, according to Kant, both metaphysically and epistemically prior to its represented parts, which are dependent upon this whole. It is partially because the representation of space as a whole cannot be built up out of a previous representation of its parts (particular spaces) that Kant considers it as having the status of an a priori representation. Kant characterizes the activity of the understanding, in contrast, as building representations out of their parts via the process of running through and gathering together a representational manifold (A99). So the activity of the understanding is always from part to whole (CJ, AA 5:407; cf. B72; CJ, AA 5:251–2, 253–4). Hence, any spontaneous activity on the part of the intellect, whether it be conceptual application in a judgement, or implicit rule-governed construction (such as in the careful imagination of complex shape), the nature of the activity must proceed via a movement from part to part, out of which a representational whole may be fashioned.
If this is correct, then Kant’s arguments in the Expositions of space and time require that the fundamental basis of our representation of space and time does not proceed from a grasp of the multiplicity of features of an intuited particular to the whole that has those features. Instead the form of pure intuition constitutes a representational whole that is prior to that of its component parts (cf. CJ, AA 5:407–8, 409).
Grüne’s interpretation thus faces a difficulty had by all interpretations which construe the unity of intuition as resulting from an act of synthesis—viz., how we account for the unity of space and time themselves (or the intuitions thereof). I am admittedly extrapolating from Grüne’s explicit remarks on this subject, and there may well be some way in which she avoids the problem, but on its face her position seems to entail a fundamental tension between the argument of the Aesthetic and the argument of the Analytic. Perhaps Kant really did have such an inconsistent view at the heart of his critical system. Or perhaps we should reject the assumption Grüne (and many others) starts with—viz., that intuition depends for its existence on acts of combination carried out by the understanding (or the understanding in the guise of the imagination).
4. The Problem with Animals
Grüne identifies Kant as distinguishing between three distinct kinds of concept and their corresponding possession conditions. They are: ‘obscure’ or ‘dark’ [dunkel] concepts; ‘clear’ [klar] concepts; and ‘distinct’ [deutlich] concepts.
A concept expressed by the predicate ‘F’ is obscure just in case it functions for the subject solely as a rule for sensory synthesis. This means that the subject, by means of F, can grasp the content of a collection of sensory representations as a unity (p. 197) without thereby coming to make any judgement (or otherwise ‘implicitly’ judge) or infer any conclusion (p. 202), and also without coming to have a general representation (in the sense of a reflected conceptual representation) of any of the features of the object so represented.
In contrast, a clear concept is a concept by means of which the subject is able to classify something perceptually presented to her. Classification is understood in terms of the correct application of the concept to the object in a perceptual judgement (p. 26). Importantly, one can possess a concept both obscurely and clearly (p. 204). The ability to grasp the content of a series of sensory representations as a unity is independent of, and more basic than, the ability to apply a concept in the making of a perceptual judgement. The latter is in fact dependent on the former (p. 204).
Finally, a distinct concept is one where the subject is in the position to infer from the applicability of a concept F to an object, to the object’s possessing a series of other characteristics, all the predicates of which are contained as marks in F (p. 83). All distinct concepts are also clear concepts (p. 98).
As noted in the introduction, Grüne’s signal contribution to resolving the inconsistent triad of claims is to argue that while obscure concepts are necessary for the generation of intuitions, intuitions are necessary for the generation of clear (and distinct) concepts.
Obscure concepts generate intuitions by means of a rule-governed sensory synthesis performed on purely subjective sensory states that lack any representational or ‘intentional’ status prior to their synthesis. However, this means of resolving the apparent inconsistency between the three claims Grüne discusses fails to account for one of the central issues motivating non-conceptualist interpretations of Kant—viz., the possibility of empirical intuition (i.e., perceptual experience) in non-rational animals.
As multiple interpreters have argued, myself included, Kant characterizes the representational capacities characteristic of sensibility as more primitive than those characteristic of the understanding (or reason), and as plausibly part of what humans share with the rest of the animal kingdom. For example, Robert Hanna construes Kant’s distinction between the faculties of sensibility and understanding as capturing the difference between the ‘sub-rational’ powers of the mind that we share with non-human animals, and the ‘rational or higher-level cognitive powers’ that are special to human beings.
Grüne mentions, in a footnote (pp. 202–3n.17), that it is an ‘interesting question’ whether the capacity for obscure concept possession might be had independently of the capacity for judgement and inference. She suggests that an affirmative answer would leave open the possibility that non-rational animals are capable of intuition, while a negative answer (of which Longuenesse’s interpretation is mentioned as an example) would deny such intuitions to animals. As I and others have noted, however, there is significant textual evidence in favour of reading Kant as endorsing the possibility (and perhaps even the actuality) of non-rational animal intuition. Kant is on record in various places as saying that animals have sensory representations of their environment (CJ, AA 5:464; LM, AA 28:449; cf. Anth, AA 7:212), that they have intuitions (LL, AA 24:702), and that they are acquainted with objects though they do not cognize them (JL, AA 9:64–5). So it would be a significant benefit of Grüne’s interpretation if it not only resolved the inconsistent triad she started with, but also in such a way that allowed for the accommodation of non-rational perception in animals, and thus accommodation of a major motivation of the non-conceptualist position.
Unfortunately, however, I don’t see how her interpretation is any better placed than that of other conceptualist readings in allowing the existence of genuine perceptual intuition, as opposed to mere sensory registration, in non-rational animals. Grüne writes:
It seems possible to me that there could be a creature which, just as we do, possessed the inborn capacity to combine representations according to determinate laws, but which didn’t possess the capacity for comparison, abstraction, and reflection. Such a creature could indeed combine sensory representations into intuitions, but would not be in the position to make judgements. As such, they would possess the categories as obscure concepts, without thereby having the capacity for judgement [or inference]. (p. 203n.17)
Whether or not such a creature is possible (though I am inclined to agree with Grüne concerning the conceptual point here), I do not see that it would capture non-rational animal life as Kant took it to exist. According to Kant, non-rational animals have merely the capacity for sensory representation and volition (AA 5:464; 7:196; 16:7; 28:117). Only the human being possesses the capacity for rational thought and self-consciousness via a possession of a faculty of understanding (AA 7:127, 172, 196; B409). Lacking a faculty of understanding, non-rational animals lack the capacity to combine representations in the sense which Kant is concerned with in the first Critique where he says that combination
is an act [Actus] of the spontaneity of the power of representation, and since one must, in order to distinguish this from sensibility, call this the understanding, all combination […] is an action of the understanding, for which we would reserve the general term synthesis […]. (B130)
I take it that Kant’s point here is compatible with allowing that non-rational animals have the capacity for association via exercise of their reproductive imagination (B152; AA 28:277, 689–90). But this is not the same as grasping a complex of sensory representations as a genuine unity, something that seems reserved for the understanding alone. And since Grüne takes such representational unity as necessary for intuition, and thus objective awareness, it seems that her account cannot accommodate the possibility of non-rational animal perception, even with its rather liberal conception of obscure concept possession. Of course, this might be an interpretive nettle that Grüne is willing to grasp. Either way, it would be helpful to hear more about whether or how non-rational animal perception (and the various Kantian texts which appear to endorse its possibility) might be accommodated by her interpretation.
5. The Content of Intuition
I want to now turn to some more general issues concerning the assumption that intuitions have intentional content. Grüne argues that an intuition represents its object via marks, and the marks determine which object the intuition represents:
Intuitions likewise represent their objects by means of marks—albeit by so called intuitive marks, that may only be met with in particular objects. For this reason a particular intuition would then represent a particular object because of the specific [bestimmte] intuitive marks the intuition contains as its content. (p. 53)
Though I do not believe that Grüne ever explicitly states her position in this manner, it would seem that she endorses a version of what, in the philosophy of perception, has come to be known as the ‘Content View’, according to which sensory states count as perceptual experiences just in case they possess correctness conditions. Thus, part of what it is for a mental state to count as an experience is that it can be evaluated for its correctness, and the conditions of its correctness are determined by its content. According to Grüne’s interpretation, such content should be understood in terms of representation of tropes—the particular instantiations of properties possessed by spatio-temporal objects (pp. 66, 70–1).
However, Kant seems to clearly deny that intuitions may be assessed for their correctness. Correct representation is a product rather of judgement:
[T]ruth and illusion are not in the object insofar as it is intuited [Denn Wahrheit oder Schein sind nicht im Gegestande, so fern er angeschaut wird], but are in the judgment made about the object insofar as it is thought. Hence although it is correct to say that the senses do not err, this is so not because they always judge correctly but because they do not judge at all. Thus both truth and error, and hence also illusion as the process of mistakenly leading to error, are to be found only in the judgment, i.e., only in the relation of the object to our understanding. (A293–4/B350)
Here Kant denies that what is delivered by sensibility—namely ‘the object insofar as it is intuited [angeschaut]’—consists in something assessable for truth or error. Error is a product of the relation of the object to the understanding—i.e., in the object as it is judged. To be sure, the senses may provide, in Cartesian phrase, the ‘material’ for error. Kant is clear, for example, that errors in perceptual judgement are due, at least in part, to the causal influence that sensibility and its representations can have on the understanding and its judgmental activity (cf. A294–5/B350–1). But this point doesn’t entail that sensibility yields representational states (intuitions) that relate to an object by means of possessing correctness conditions.
Kant repeatedly expresses the position that the senses do not deliver anything capable of error in his anthropology and logic lectures (cf. Anth §11, AA 7:146; LL, AA 24:83ff., 103, 720ff., 825ff.). Grüne notes the significance of the above passage for conceptualist views that would construe intuition as a form of, or otherwise entailing, perceptual judgement (p. 117). But she does not consider whether such texts raise problems for her own position. It seems to me that they do. If an empirical intuition contains marks which determine the object which it represents, then the intuition is assessable for correctness by virtue of this content. Certainly, the way in which an intuition is correct or incorrect would differ from that of a judgement, since intuitions and judgements contain different kinds of marks. But this shouldn’t keep intuition from being assessable for correctness. Grüne’s interpretation of the content of intuition thus seems to stand in serious tension with Kant’s many remarks suggestive of a doxastic theory of perceptual error.
What’s more, Grüne’s view appears to stand in deep tension with one of Kant’s centrally held commitments—viz., that in intuition we prove the possibility of the subject matter of our thoughts. Kant makes this point explicitly in the preface to the B-edition of the Critique:
To cognize an object, it is required that I be able to prove its possibility (whether by the testimony of experience from its actuality or a priori through reason). But I can think whatever I like, as long as I do not contradict myself, i.e., as long as my concept is a possible thought, even if I cannot give any assurance whether or not there is a corresponding object somewhere within the sum total of all possibilities. But in order to ascribe objective validity to such a concept (real possibility, for the first sort of possibility was merely logical) something more is required. (Bxxvi)
Here Kant contrasts merely logically possible thought, in which one brings together logically compossible concepts, with the kind of real empirical possibility necessary for cognition. Kant distinguishes between the two sorts of possibility in terms of the notion of cancellation [Aufhebung]. The subject matter of a thought is logically possible if the thought’s constituent concepts may be combined in judgement without contradiction, and thus without being logically cancelled out (A151/B190; NM, AA 2:171–2). The subject matter of a thought is really possible, in contrast, if it can be shown that the subject matter to which the thought corresponds consists of properties which are mutually empirically compossible and not, in Kant’s terms, ‘really repugnant’. Kant often illustrates this idea with examples involving physical forces (e.g., opposite motions, opposing attractive and repulsive forces; cf. A264–5/B320–1). Moreover, Kant considered a further kind of repugnance, wherein the subject itself is ‘cancelled out’. In other words, it would be impossible for any being to exist that would instantiate such repugnant properties. In order for knowledge to be possible, the demonstration of the real possibility of the object of knowledge must be secured.
Call this condition on proof of possibility the ‘Modal Condition’. If experience is to provide proof of real possibility, and thus satisfy the modal condition on thought then it must have features which thought alone does not. If experience is conceived along the lines suggested by Grüne, what would those features be? This is a particularly pressing question for her because she considers empirical intuition and thought to be similar in a variety of ways. For example, they are both attitudes to content (though admittedly the contents differ in nature), and the content in both cases sets correctness conditions on mental states that possess them. It seems clear then that the content of an experience cannot be simply the same as the content of a thought, otherwise experience would be no better situated to satisfy the Modal Condition than thought.
The problem is made even more acute by virtue of the fact that Grüne denies that intuition is, with respect to its content, dependent on the existence of the object it represents. Since the very same intuition can be had whether or not its object exists, it is altogether unclear how the having of an intuition alone could ever stand as proof of the real possibility of the subject matter of a thought. Some further condition, from the subject’s perspective, would need to be satisfied, for such proof to be in the offing. But what would that be?
6. Intuition and Presence
Even if we ultimately want to agree with Grüne’s position on the content of intuition, there is a further claim motivating her ‘intentionalist’ reading—viz., that intuition as a mental state is intelligible independently of the existence of its object. This ‘existence-independent’ reading of intuition is argued for on the basis of some of Kant’s remarks concerning intuition in cases of dreams or hallucinations (pp. 42–3; cf. p. 50n.32). She puts heavy weight, in particular, on Kant’s discussion of intuition in the Refutation of Idealism:
From the fact that the existence of outer objects is required for the possibility of a determinate consciousness of our self it does not follow that every intuitive representation of outer things [jede anschauliche Vorstellung äusserer Dinge] includes at the same time their existence, for that may well be the mere effect of the imagination (in dreams as well as in delusions) […]. (B278–9; cf. A376; Prol, AA 4:290)
I agree with Grüne that the apparent problem presented by this passage cannot be avoided by pointing out that he uses the phrase ‘intuitive representation’ [anschauliche Vorstellung] rather than ‘intuition’ (pp. 42–3). But it is not at all obvious to me that Grüne’s point about existence-independence holds.
I take Kant’s general position to be that intuition always involves a presence to the mind of something. In this I believe Grüne and I are in agreement. The question is whether, in the case of empirical intuitions, the occurrence of outer empirical intuition always makes present to mind some actual (empirically) mind-independent thing. If that particular thing did not exist then the intuition would not have occurred.
Grüne’s position is closer to an ‘intentional object’ reading, wherein the object one intuits is determined by the intentional content of the intuition, characterized in terms of the representation of tropes. Thus, in the case of hallucination, according to her reading of Kant, one grasps as a unity a series of sensory representations (sensations) which represent an object which is not in fact present in one’s environment. The object is only ‘present’ in the sense in which it is the unity of the represented tropes (which themselves are only represented as existing). Grüne thus explains the sense in which intuitions are dependent on the ‘presence’ of an object in terms of the ‘presence’ of an intentional object determined by the content of the subject’s intuition(s).
While I take this to be a perfectly coherent position in itself, it doesn’t seem to sit well with Kant’s introduction, in the Prolegomena, of the notion of intuition in terms of the presence of an object which affects the subject. There we have Kant’s explicit claim that intuition depends on the presence of its object, in a sense which strongly implies that the object is or did exist:
How is it possible to intuit something a priori? An intuition is a representation of the sort which would depend immediately on the presence of an object. It therefore seems impossible originally to intuit a priori, since then the intuition would have to occur without an object being present, either previously or now, to which it could relate, and so it could not be an intuition. (AA 4:281–2)
Kant’s framework here depends on the interpretation of ‘presence’ [Gegenwart] as ‘really existent’. This is why a priori intuitions are so problematic. How could there be a priori intuition which nevertheless depends on something’s presence? Kant’s solution is that the pure intuitions of space and time present only mind-dependent objects, which are themselves the formal conditions of anything else’s (i.e., any empirical object) appearing.
If Grüne is wrong to dispute Kant’s apparent construal of empirical intuition as presence-dependent, what other alternative might we have? I suggest that there is a textually plausible alternative way of reading Kant, that allows him to explain what is going on in cases of dreams and hallucination (and thus in the kinds of cases mentioned in the B278–9 passage above) without thereby denying that intuitions generally depend on the presence of their object (in a sense more robust than mere ‘intentional’ presence), such that intuitions, both inner and outer, always have an existing object.
First, note that Kant distinguishes between intuition as the presence to sensibility of some object, and imagination, as the production of intuition independent of any such present object (cf. Anth, AA 7:167). From this distinction one might reasonably conclude that perhaps intuitions, and correspondingly, sensory experiences, are not after all dependent on the presence of the object they bring before the mind. Perhaps they are merely self-standing representations which may or may not correspond to some external object. This would seem to work in Grüne’s favour.
But, as Kant goes on to make clear, a distinction must be made between an ‘original’ (ursprünglich) or ‘productive’ (productiv, dichtend) faculty and one which is merely ‘reproductive’ [reproductiv, zurückrufend] (AA 7:167–8). The productive imagination is partially responsible for the a priori representation of space and time, while reproductive imagination allows a subject to sensorily recall past presentations. So in the case of perceiving a spatial object in vision or touch, the empirical intuition does indeed depend on the presence of that object, as an object of outer sense. What, at this point, is unclear, is whether we are to construe the imagination of that object as the very same kind of representation as the outer intuition of the object. It may be that, in the case of imagination, only inner sense is involved, and thus one has only an inner intuition. But if this is the case then the general claim of presence-dependence could still hold, for in the case of outer intuition what is present is some spatial object, while in the case of inner intuition what is present is an image [Bild]—a reproduction in some sense—of the originally perceived object. In order that Kant’s remarks concerning imagination count against his general claim concerning the presence-dependence of intuition, he would have to be conceiving of all empirical intuition along the lines of inner intuition, and nowhere in the passage from the Anthropology does he do this. So it remains an open possibility that an outer empirical intuition does indeed depend on the presence to mind of the relevant spatially extended object, despite the fact that in the object’s absence, one may have an inner imagined recollection of it.
In a series of notes from Kant’s Nachlass he clearly distinguishes between imagination and intuition. For example, he says:
We have two sorts of intuition: sensible intuition, for which the object must be represented as present, and an imagining as intuition without the presence of the object. The imagining, if one is conscious of it as such, can also be considered as inner sensible intuition. (AA 18:619)
Kant further suggests that the problem of hallucination is primarily epistemic and doxastic—we cannot always tell when we are having a genuinely outer intuition or merely the appearance of one:
in this condition of imagining it [i.e., the real from the imaginary] certainly cannot be distinguished, for this is a deception of the power of judgment; but the question is properly whether it cannot be distinguished in general, i.e., whether one cannot be conscious that the one is an intuition of the senses, the other to be sure a sensible intuition, but only in an imagining, for which no object outside the representation is present. The answer is that consciousness can accompany all representations, hence even that of an imagining, which, together with its play, is itself an object of inner sense, and of which it must be possible to become conscious as such, since we really distinguish such things as inner representations, hence existing in time, from the intuition of the senses. (AA 18:621)
Kant seems here both to concede that imagining cannot be distinguished from outer intuition while, in the latter part of the passage quoted, claiming that our consciousness of imagined objects is really only consciousness of an inner representation—i.e., an object of inner intuition.
A further point which seems to me to favour the interpretation of imagination as offering only inner intuitions, and thus as differing in form from intuitions of outer sense, is Kant’s claim, in the first sentence, that hallucination and dream deceive in virtue of a problem with judgement, rather than the fact that hallucination and dream are simply the same type of representation as one of outer sense. The appeal to judgement suggests that the problem of hallucination is not a metaphysical one concerning the identity conditions of an outer intuition vs. those of a hallucination or dream, but rather the epistemic problem of telling the two kinds of representation apart.
A related text, which might support this idea, occurs in the B-preface where Kant presents a modification of the argument of the Refutation of Idealism. The issue he there raises also suggests that the challenge outer intuition poses is that it cannot always be distinguished from inner intuition, not that there is only one type of representation, which may or may not occur in connection with an outer object. He says:
Now which given intuitions actually correspond to outer objects, which therefore belong to outer sense, to which they are to be ascribed rather than to the imagination—that must be decided in each particular case. (Bxli note)
This suggests that from the first-person point of view one is not always in a position to distinguish outer from inner intuition. One cannot always tell whether what one is experiencing is a genuine spatially extended object, or merely an inner ersatz object generated by imagination.
The admission of the possibility of confusing inner with outer intuition allows Kant to individuate intuition via the particulars to which it relates; in the case of outer intuitions via spatial particulars, in the case of inner intuitions via temporally ordered subjective states. There are several points in the first Critique where Kant appears to suggest that we individuate intuitions in terms of such relations. For example:
This consciousness of my existence in time is thus bound up identically with the consciousness of a relation to something outside me, and so it is experience and not fiction, sense and not imagination, that inseparably connects the outer with my inner sense; for outer sense is already in itself a relation of intuition to something actual outside me; and its reality, as distinct from imagination, rests only on the fact that it is inseparably bound up with inner experience itself, as the condition of its possibility. (Bxl note)
Here Kant seems to be tying outer intuitions to actually existing spatio-temporal particulars that are (at least empirically) distinct from the cognizing subject. This is what distinguishes an intuition of outer sense from a merely inner intuition that might be generated by the imagination, as in a case of hallucination. The contrast Kant makes between intuitions of spatio-temporal particulars (i.e., outer intuitions) and inner intuitions generated by imagination suggests that intuitions of actual spatio-temporal particulars are genuinely different from intuitions generated by the power of imagination, even when one cannot subjectively discriminate between them. Kant distinguishes between genuine experiences and mere imagination of spatial particulars in the Refutation as well:
The proof that is demanded must therefore establish that we have experience and not merely imagination of outer things. (B275; original emphasis)
Finally, in his Anthropology of 1798 Kant explicitly ties perceptual illusion to a confusion of inner sense with representations given via outer sense:
It is said that inner sense is subject to illusions, which consist either in taking the appearances of inner sense for external appearances, that is, taking imaginings for sensations, or in regarding them as inspirations [Eingebungen] caused by another being that is not an object of external sense. So the illusion here is either enthusiasm or spiritualism, and both are deceptions of inner sense. (Anth, AA 7:161; original emphasis; cf. LA, AA 25:61, 281, 1456) 
This passage, and the many passages like it from previous anthropology lectures, gives support to my preferred interpretation, for it takes hallucination and illusion to be the product of a kind of confusion concerning what is present or represented, rather than a problem concerning the representation itself and its connection to some outer thing.
I thus take it that Kant’s considered position is that perceptual representations of external objects are not merely to be causally distinguished from representations generated in inner sense, but are genuinely different representational types than those which are the product of hallucination or dream. Otherwise, he would not need to repeatedly distinguish sensation from imagination, or outer intuition from inner. He could instead have simply held that there is one type of representation which may have different causes, either external or internal. But this seems to be exactly the position he denies. Kant does not locate the difficulty of hallucination in the difficulty of ascribing a single representation to some inner vs. outer cause. Instead, the difficulty is that we cannot always discern when we are having a genuine perceptual experience of a spatial particular. Instead, in any particular case, we must rely on the coherence of our experience of the object, particularly its causal coherence, with the other objects we experience. So the position I have suggested makes good sense of a variety of texts, relies primarily on clear statements in published works such as the Prolegomena and Anthropology rather than statements in his notes, and evinces the proper charity in rendering Kant’s stated positions concerning intuition and imagination coherent.
This position also partially undermines Grüne’s claim that it is the content of an intuition (understood in terms of representation of tropes) which determines which object it is ‘about’. If an intuition could occur in the absence of its object then we need an account of how it is that particular intuitions ‘attach’ to particular objects. The discussion of the intentional content of intuition that Grüne provides is one such account. But if intuitions may be individuated in terms of their objects, as on the account just offered, then there is no issue of ‘attachment’ and the need to ascribe to Kant a theory of intentional perceptual content is less pressing.
 We might thus distinguish the ‘synopsis of sense’ that Kant mentions (A94, 97) from the synthetic apprehension which ‘corresponds’ to what is offered in sense. The synthesis of apprehension runs through and gathers together this manifold but needn’t be construed as generating the intuition itself. We can also see the distinction between the intuition and the representation of its manifold at work in the Axioms of Intuition. There Kant distinguishes an intuited appearance from what must be true if that representation is ‘apprehended, i.e., taken up into empirical consciousness’, while clearly associating the ‘synthesis’ and ‘composition’ (Zusammensetzung) of the manifold in an appearance with the apprehensive act and not the existence of the intuition itself (B202–3) (cf. Tolley 2013:133n.36). Further, Kant distinguishes between the appearance, which ‘as intuition is an extensive magnitude’, such that ‘all appearances are already intuited as aggregates (multitudes of antecedently given parts)’ from the mental act required for an appearance to be ‘cognized’ or ‘represented and apprehended by us as extensive’ (B204). Kant makes a similar distinction in the Second Analogy (B236).
 Grüne’s stance on the non-intentional status of sensory representation is, perhaps, not entirely stable, as she herself notes that the tendency to speak of the parts of intuition as sensations and themselves representational, but only retrospectively, as parts of the intuition, is a vexed issue (cf. p. 158; p. 200n.11).
 I argue for this and related points more extensively in McLear (forthcoming a); cf. Allais (2009:403–4). Tolley (2013:133n.38) makes a related point that Kant’s discussion in the Deduction seems primarily concerned with the conditions under which the conceptual representation of space and time are possible, not with the conditions under which space and time themselves, as pure intuitions, are possible.
 One central notion in Grüne’s argument which is reasonable, though not explicitly defended, is the notion that concept possession and conceptual representation, for Kant, should be understood in terms of a capacity (or capacities) of the subject to do or perform certain actions. The tie between concept possession and the activity of the subject is a reasonable and widespread one. However, I think it might raise trouble for Grüne’s view. I discuss this below.
 An alternate formulation offered by Grüne states that a subject possesses a distinct concept when she is in the position to make a determinate number of general analytic judgements which contain F as subject.
 Grüne’s reasoning here is that in order to explain how the schematized categories legitimately apply to the objects of intuition, it is necessary that the intuitions themselves be so constituted as to allow for such an application. Intuitions do this in virtue of the sensations that constitute them being synthesized in accordance with the category (p. 215ff.). This act of synthesis is not, nor need it issue in, a judgement, but is rather the ‘primitive capacity’ (p. 213) to unify one’s sensations according to rules.
 For further discussion see Naragon (1990); Allais (2009); McLear (2011). For some defense of conceptualist positions that deny this possibility see McDowell (1996), Chs 3 & 6; Ginsborg (2006), Ginsborg (2008); Gomes (2014).
 ‘Nun scheint es meiner Meinung nach möglich zu sein, dass es Wesen gibt, die, genau wie wir, die angeborene Fähigkeit besitzen, Vorstellungen nach bestimmten Gesetzen zu verbinden, die aber nicht die Fähigkeit der Komparation, Abstraktion und Reflexion besitzen. Solche Wesen könnten zwar sinnliche Vorstellungen zu Anschauungen verbinden, wären aber nicht in der Lage, Urteile zu fällen. Insofern würden sie über die Kategorien als dunkle Begriffe verfügen, ohne die Fähigkeit des Urteilens zu haben.’
 I say liberal here because Grüne seems willing to characterize possession of the primitive capacity for grasping a series of marks as a unity in ‘wholly non-intellectual terms’ (p. 202n.16). She suggests that the possession of such a capacity should be understood as the cognitive analogue of the possession of the capacity to digest meat (ibid.). But if this is correct, then it would seem that we’re no longer talking about spontaneity, and thus no longer talking about the capacity that, according to Kant, makes the grasp of higher cognitive unities possible.
 ‘[…] Anschauungen ihre Gegenstände ebenfalls vermittels von Merkmalen—allerdings von so genannten intuitiven Merkmalen, die nur einem einzigen Gegenstand zukommen können—repräsentieren. Eine bestimmte Anschauung würde dann einen bestimmten Gegenstand deswegen repräsentieren, weil sie bestimmte intuitive Merkmale zum Inhalt hat.’
 Lucy Allais (2010) argues that the ‘presence-dependence’ of intuition is part of Kant’s argument for transcendental idealism. She argues that Kant’s notion of intuition needs to be understood as a kind of representation which involves the presence to consciousness of the object it represents, and that this means that a priori intuition cannot present us with a mind-independent feature of reality.
 Grüne never quite talks in this way, but she does explicitly articulate her reading in terms of the conditions under which purely subjective states become ‘intentional states’ (pp. 40, 104, 153n.8, 185–6) which have a relation to an object [Gegenstandsbezug].
 Things are actually more complicated than this, as Kant’s notion of a faculty of imagination covers various subfaculties, which have different cognitive roles, and which are controlled by a greater or lesser degree by the will. See Satura (1971:113–41) for discussion.
 For discussion of the notes and their English translation see Guyer’s discussion in the Cambridge edition of the Reflexionen in Notes and Fragments, Cambridge: Cambridge University Press (2005), note 81, pp. 592–3.
Allais, L. (2009) ‘Kant, Non-Conceptual Content and the Representation of Space’, Journal of the History of Philosophy 47 (3): 383–413. doi:10.1353/hph.0.0134.
— (2010) ‘Kant’s Argument for Transcendental Idealism in the Transcendental Aesthetic’, Proceedings of the Aristotelian Society 110 (1/1): 47–75. doi:10.1111/j.1467-9264.2010.00279.x.
— (2011) ‘Idealism Enough: Response to Roche’, Kantian Review 16 (3): 375–98. doi:10.1017/S1369415411000197.
Bennett, J. (1966) Kant’s Analytic, Cambridge: Cambridge University Press.
Chignell, A. (2011) ‘Real Repugnance and Our Ignorance of Things-in-Themselves: A Lockean Problem in Kant and Hegel’, Internationales Jahrbuch des Deutschen Idealismus 2009, ed. F. Rush et al., pp. 135–59, Berlin/New York: de Gruyter. http://philpapers.org/rec/CHIRRA-2.
— (2014) ‘Kant and the “Monstrous” Ground of Possibility’, Kantian Review 19 (1): 53–69. doi:10.1017/S1369415413000290.
Frierson, P. (2009a) ‘Kant on Mental Disorder. Part 1: An Overview’, History of Psychiatry 20 (3): 267–89. doi:10.1177/0957154X08337642.
— (2009b) ‘Kant on Mental Disorder. Part 2: Philosophical Implications of Kant’s Account’, History of Psychiatry 20 (3) (September): 290–310. doi:10.1177/0957154X08337644.
George, R. (1981) ‘Kant’s Sensationism’, Synthese 47 (2): 229–55. doi:0039-7857/81/0472-0229.
Ginsborg, H. (2006) ‘Kant and the Problem of Experience’, Philosophical Topics 34 (1&2): 59–106. doi:10.5840/philtopics2006341/24.
— (2008) ‘Was Kant a Nonconceptualist?’, Philosophical Studies 137 (1): 65–77. doi:10.1007/s11098-007-9163-3.
Gomes, A. (2014) ‘Kant on Perception: Naïve Realism, Non-Conceptualism and the B-Deduction’, Philosophical Quarterly 64 (254): 1–19. doi:10.1093/pq/pqt019.
Griffith, A. (2012) ‘Perception and the Categories: A Conceptualist Reading of Kant’s Critique of Pure Reason’, European Journal of Philosophy 20 (2): 193–222. doi:10.1111/j.1468-0378.2010.00404.x.
Hanna, R. (2005) ‘Kant and Nonconceptual Content’, European Journal of Philosophy 13 (2): 247–90. doi:10.1111/j.0966-8373.2005.00229.x.
Longuenesse, B. (1998) Kant and the Capacity to Judge, Princeton: Princeton University Press.
McDowell, J. (1996) Mind and World, second edition, with a new introduction, Cambridge MA: Harvard University Press.
McLear, C. (2011) ‘Kant on Animal Consciousness’, Philosophers’ Imprint 11 (15) (November): 1–16. http://quod.lib.umich.edu/cgi/p/pod/dod-idx/kant-on-animal-consciousness.pdf?c=phimp;idno=3521354.0011.015.
— (forthcoming a) ‘Two Kinds of Unity in the Critique of Pure Reason’, Journal of the History of Philosophy. http://philpapers.org/rec/MCLTKO.
— (forthcoming b) ‘Kant on Perceptual Content’, Mind. http://philpapers.org/rec/MCLKOP.
Messina, J. (2014) ‘Kant on the Unity of Space and the Synthetic Unity of Apperception’, Kant-Studien 105 (1): 5–40. doi:http://0-dx.doi.org.library.unl.edu/10.1515/kant-2014-0002.
Naragon, S. (1990) ‘Kant on Descartes and the Brutes’, Kant-Studien 81 (1): 1–23. http://www.degruyter.com/view/j/kant.1990.81.issue-1/kant.19188.8.131.52/kant.19184.108.40.206.xml.
Onof, C. and D. Schulting (forthcoming) ‘Space as Form of Intuition and as Formal Intuition. On the Note to B160 in Kant’s Critique of Pure Reason’, The Philosophical Review 124 (1) (2015).
Pereboom, D. (1988) ‘Kant on Intentionality’, Synthese 77 (3): 321–52. doi:10.1007/BF00869290.
Roche, A. (2011) ‘Allais on Transcendental Idealism’, Kantian Review 16 (03): 351–74. http://journals.cambridge.org/abstract_S1369415410000026.
Satura, V. (1971) Kants Erkenntnispsychologie in den Nachschriften seiner Vorlesungen über Empirische Psychologie, Bonn: Bouvier.
Sutherland, D. (2004) ‘Kant’s Philosophy of Mathematics and the Greek Mathematical Tradition’, The Philosophical Review 113 (2) (April): 157–201. http://www.jstor.org/stable/4147961.
— (2006) ‘Kant on Arithmetic, Algebra, and the Theory of Proportions’, Journal of the History of Philosophy 44 (4): 533–58. doi:10.1353/hph.2006.0072.
Tolley, C. (2013) ‘The Non-Conceptuality of the Content of Intuitions: A New Approach’, Kantian Review 18 (01): 107–36. doi:10.1017/S1369415412000313.
Van Cleve, J. (1999) Problems from Kant, Oxford: Oxford University Press.
© 2014, Colin McLear.
Colin McLear is an Assistant Professor of Philosophy at the University of Nebraska-Lincoln. He graduated with a Ph.D. in philosophy from Cornell University in 2013. His work focusses on the history of modern philosophy, especially Kant, and the philosophy of mind. He published in Philosophers’ Imprint and has publications forthcoming in Journal of the History of Philosophy, Mind, and Philosophy Compass.