1. Introduction
The question of how we experience beauty remains relevant in the contemporary philosophy of art (Plantinga Reference Plantinga2009: 256; Carroll Reference Carroll2012). It is odd that Kant is not part of this discussion, especially since, as Henrich (Reference Henrich1992: 55) points out, the ‘connection [that Kant makes] between aesthetic experience and the fundamental structures of cognition’ should be useful for present-day research in cognitive science. Further, concepts like ‘play’ and ‘harmony’, which Kant employs to describe our cognitive state in aesthetic experience, ‘resonate phenomenologically with aesthetic experience, and consequently were and continue to be influential conceptions of aesthetic experience’ (Zuckert Reference Zuckert2007: 281). One reason for this neglect of Kant may be that he provides a sketchy account of the harmony of the mental faculties in aesthetic experience. In addition, Kant interpreters are divided on how to interpret what he does say about this topic.
Kant says that the imagination harmonizes with the understanding in judgements of beauty. Interpreters have puzzled over whether determinate conceptualization is involved in this harmonization. In this article, I argue that determinate cognition does not form a constitutive part of judgements of natural beauty. Yet the understanding does constrain the imagination: indirectly by restricting the imagination to the broad contours of the image of the beautiful object being judged, and more directly via obscure concepts/principles of ordinary common sense.
Kant’s theory of the harmony of the faculties is embedded in his account of aesthetic judgement in the Critique of the Power of Judgement. For Kant, the mind is an ‘aggregate’ (Aggregat) of the faculties of sensibility, imagination, understanding, judgement and reason (FI 20: 206).Footnote 1 In the Critique of Pure Reason, Kant articulates a theory of cognition in terms of these faculties. The imagination organizes spatiotemporal data into an image. Cognizing this image as, say, the image of a table as opposed to the image of any other object constitutes the task of the understanding. In contrast, aesthetic judgement, for Kant, is not aimed at cognition. Yet such judgements – in which the imagination ‘harmonizes’ with the understanding – are located within the ambit of ‘cognition in general’ (5: 217–18). They are reflective rather than determinate judgements. In determinate judgement, one subsumes a given particular under a cognitively available universal, say, ‘human’. Conversely, in reflective judgement the attempt is to find a universal for a given particular (5: 179). An aesthetic judgement is reflective because it is not ‘grounded on any available concept of the object and does not furnish one’ (5: 190; CPJ 76).
For Kant, judgements of natural beauty (‘the tulip is beautiful’) rather than judgements of art (‘the painting is beautiful’) form the paradigm case of ‘pure’ aesthetic judgement (5: 299). These judgements are ‘pure’ because they are unsullied by theoretical, technical or moral interests. Restricting ourselves to ‘pure’ aesthetic judgements, as I do in this essay, gives us the following overview of Kant’s characterization of aesthetic judgement.Footnote 2 First, aesthetic judgement involves the ‘form’ of the object of intuition and not its ‘matter’. For instance, a tulip is judged beautiful on the basis of its form, not its colour (5: 225). Second, sensory pleasure is interested because it serves one of my ends, e.g. desire for cake. By contrast, aesthetic pleasure is disinterested: I take pleasure in the tulip although it does not fulfil any of my ends. Famously, Kant characterizes (aesthetic) pleasure in the tulip as purposiveness without purpose (5: 221). Third, aesthetic judgements are intersubjectively valid. All human beings share the normative constraints of ‘cognition in general’. Disagreement between them arises if they privilege their own private interests. In judging the tulip as beautiful, my private interests are not involved. Therefore, I expect everyone to judge the tulip as beautiful (5: 211). Fourth, in cognizing an object, the imagination must produce an image suitable for correct cognition, e.g. it should not provide a rope-image if the subject encounters snake-data. In aesthetic judgement, the imagination is free and ‘schematizes without concepts’ (5: 287). Fifth, the harmony between imagination and understanding in aesthetic judgement generates aesthetic pleasure for the judging subject (5: 287).
Even a cursory look at Kant’s doctrine of the harmony of the faculties in aesthetic judgement raises difficult questions regarding the role of the imagination and the part concepts play, given that determinate cognition is not the goal of aesthetic judgement. Paul Guyer frames the contemporary Anglo-American response to the puzzle of the ‘free yet harmonious play of imagination and understanding’ in a widely accepted manner (Guyer Reference Guyer2006b: 165, Reference Guyer2006a: 315).Footnote 3 Guyer identifies three approaches to this puzzle: ‘pre-cognitive’, ‘multi-cognitive’ and ‘meta-cognitive’ (2006b: 165). The pre-cognitivists interpret Kant as ‘claiming that in experiencing an object as beautiful we go through all the steps of normal cognition but are so struck by the unity of our experience of the object we simply stop short of applying any determinate concept to it’ (Guyer Reference Guyer2006a: 315). Guyer regards Dieter Henrich, Donald Crawford, Ralph Meerbote and his own early work (Guyer Reference Guyer1997) as taking this approach, and Hannah Ginsborg and Rudolf Makkreel as variants (Guyer Reference Guyer2006b: 165–9). On the other hand, the multi-cognitivists, Guyer says, ‘have taken Kant to mean that in experiencing an object as beautiful the mind plays back and forth between a number of different conceptualizations for it without being forced to settle on any one of them’ (Guyer Reference Guyer2006a: 315). Gerhard Seel, Fred Rush, Henry Allison (who at least ‘seems to be attracted primarily to the multi-cognitive interpretation’) and Malcolm Budd are multi-cognitivists (Guyer Reference Guyer2006b: 169–71). The multi-cognitivism of Crowther (Reference Crowther2010: 60–1n.) also belongs here. Guyer criticizes both these approaches for denying that determinate concepts are applied in aesthetic judgement (Guyer Reference Guyer2006a: 315–16, Reference Guyer2006b: 178). In opposition to them, Guyer outlines the meta-cognitive approach, in which cognition does take place, i.e. determinate concepts are applied, but the imagination harmonizes with the understanding in aesthetic judgement in a way that ‘goes beyond’ everyday cognition (Guyer Reference Guyer2006b: 183).Footnote 4
In this essay, I will outline a reconstruction of Kant’s theory of the harmony of the faculties consistent with the claim that determinate cognition plays no part in pure aesthetic judgement (henceforth ‘new harmony theory’ [NHT]). I argue for this interpretation as follows. (a) Beauty, for Kant, is relational in that some natural objects and not others are ‘suitable’ for human cognitive faculties, i.e. conducive to aesthetic harmonization. Faced with such a ‘suitable’ form (tulip), the understanding attempts to cognize the tulip as it would any other object. This attempt yields a tulip-image-for-cognition. However, the determinate cognition of this tulip-image as ‘tulip’ does not occur, because the tulip-image-for-cognition triggers the imagination to create a free imaginative representation of its own in a way that neither violates the tulip-image-for-cognition nor ordinary common sense. Judgement compares the tulip-image-for-cognition and the free imaginative representation of the tulip. Based on this comparison, it allows the human subject to apprehend the encountered tulip-data in the form of the free imaginative representation of the tulip instead of the form of the tulip-image-for-cognition, thus giving the imagination primacy over the cognitive interests of the understanding in unifying the object-data at hand. Apprehending the tulip-data in this free imaginative form ipso facto generates aesthetic pleasure, because it brings us relief from the perpetual conflict between the imagination and the understanding in determinate cognition. (b) If we accept (a), then aesthetic harmony need not include determinate cognition. The understanding constrains the imagination indirectly, because its attempt at cognition occasions an image-for-cognition that then both makes possible and constrains the free imaginative representation. In addition, the free imagination cannot violate the obscure concepts/principles of ordinary common sense.
In section 2 I enumerate the Kantian claims that must be accommodated in interpreting Kant’s characterization of the harmony of the faculties. In section 3 I outline NHT and argue that it is compatible with these claims. Finally in the conclusion, section 4, I draw out the implications of NHT for the question of how concepts are involved in aesthetic judgement, and identify my view as ‘non-cognitive’.
Before I begin, some methodological considerations. First, since Kant’s theory of the harmony of the faculties is insufficiently developed, my interpretation, like Zuckert’s (Reference Zuckert2007: 17), is reconstructive, and based on the principle of charity. I offer a theory that coheres with Kant’s statements on pure aesthetic harmony, and to some extent discuss its phenomenological viability.Footnote 5 Second, my discussion restricts itself to CPJ, and does not refer to Kant’s Nachlass. Third, commentators have noted that Kant grounds his doctrine of aesthetic harmony in his outdated faculty psychology. On this count, I provisionally accept Zuckert’s characterization of Kant’s ‘transcendental psychological language’ as ‘describ[ing] types of cognitive activity necessary for producing particular types of representation or experience’ (2007: 283). Fourth, Ginsborg (Reference Ginsborg1990: 46) points to the task of reconciling Kant’s statements regarding harmonization of the faculties in aesthetic judgement with Kant’s theory of cognition. The present essay is a preliminary step in addressing this larger question.
2. Harmony of the Faculties in CPJ
The harmony of the faculties in aesthetic judgement involves ‘conscious[ness] of a reciprocal (wechselseitig) subjective agreement (Übereinstimmung) of powers of cognition with each other’ (5: 218). Imagination and understanding are the ‘powers of cognition’ here, and their harmony is achieved via the mediation of the power of judgement.
2.1
Any interpretation of Kant’s theory of aesthetic harmony must include the following loosely connected propositions.
(a)
The notion of ‘free play’ marks the difference in the roles played by the faculties of imagination and understanding in aesthetic judgement, as compared to the roles they play in determinate cognition (5: 242). In the latter, the imagination synthesizes sensible spatiotemporal data-bits for the correct predication of objects. If it provides a rope-image for snake-data, a snake cannot be cognized as ‘snake’, and so knowledge of objects cannot be gained. Conversely, in aesthetic judgement, the imagination is freed from the understanding’s diktat to cognize (5: 241–2). In this case, we are in a ‘state of mind’ in which imagination and understanding are in free play ‘to the extent they harmonize (zusammenstimmen) with each other as required for cognition in general’ (5: 217–18). ‘Free play’ implies that no ‘determinate concept limits them to a specific rule of cognition’ (5: 217–18), which clarifies Kant’s (5: 190) claim that aesthetic judgement is not ‘grounded’ in concepts, i.e. classification into a genus-species framework is not a precondition for judging beauty. Aesthetic judgement does not aim at concept formation either (5: 217).
(b)
Aesthetic harmony entails that ‘the faculty of intuitions or presentations (i.e., of the imagination) [is subsumed] under the faculty of concepts (i.e., the understanding), insofar as the former in its freedom is in harmony with the latter in its lawfulness’ (5: 287; CPJ 168). This differs from the subsumption of particular intuitions under particular concepts in cognition.
(c)
The imagination ‘schematize[s] without a concept’ in aesthetic judgement (5: 287). It is ‘productive and self-active’ (Selbsttäthig); and, in its role as free inventor, the ‘originator’ (Urheberin) of freely produced (willkürlich) forms of possible intuition (5: 241). Further, an object is judged beautiful if it could ‘provide’ (an die Hand geben) the imagination with a ‘form that contains just such a composition of the manifold as the imagination would design (entwerfen) in harmony with the lawfulness of the understanding in general if it were left free’ (5: 241).
(d)
The imagination–understanding relationship can be considered ‘merely subjectively, insofar as one helps or hinders the other in the very same representation and thereby affects the state of mind’ (FI 20: 223; CPJ 25). In aesthetic judgement, both imagination and understanding are ‘enlivened’ (Belebung) (5: 219), and occur in a ‘well-proportioned disposition (proportionirte Stimmung) that we require for all cognition’ (5: 219). Aesthetic pleasure involves the ‘reciprocal enlivening of the imagination in its freedom and the understanding in its lawfulness’ (5: 287, 291), and arises when the imagination ‘arouses’ (erweckt) the understanding, and the understanding without employing concepts ‘sets (versetzt) the imagination into rule-bound play’ (5: 296).
2.2
Judgement plays a mediating role in harmonizing the imagination and the understanding in aesthetic judgement for the following reasons.
(a)
The imagination cannot ‘apprehend’ the form of the object of intuition if reflective judgement, ‘even if unintentionally’, fails to compare this form to its own ability of relating intuitions to concepts (5: 190; CPJ 76), which suggests that this imaginative apprehension can occur only if reflective judgement is in play. Since reflective judgement involves reflection, this statement is consistent with the claim that satisfaction in the beautiful depends on ‘reflection’ on the form of the object of intuition (5: 192; also 5: 204, 207, 209). Similarly, Kant says that if the imagination (=faculty of intuitions) is ‘brought into accord’ with the understanding (= faculty of concepts) via a comparison undertaken by reflective judgement, then aesthetic pleasure is produced (5: 190; CPJ 76).
(b)
Kant says: ‘Now this merely subjective (aesthetic) judging of the object, or of the representation through which the object is given, precedes the pleasure in it, and is the ground of this pleasure in the harmony of the faculties of cognition’ (5: 218; CPJ 103). In this passage, irrespective of whether and how judgement precedes pleasure,Footnote 6 judgement lies at the base of (‘precedes’) and enables (‘grounds’) pleasure in aesthetic harmony, which suggests that judgement plays a mediating role in producing aesthetic harmony.
(c)
Kant asserts that judgement ‘contain[s] the principle of subsumption’ of the faculty of intuitions (= imagination) under the faculty of concepts (= understanding) (5: 287). Similarly, the ‘procedure of judgement’ enables the ‘apprehension of an object by the imagination … in relation to the understanding’ (5: 292), whatever this ‘procedure’ might entail. Also, pleasure results from the ‘harmonious play of the two cognitive powers [imagination and understanding] involved in judgement’ (FI 20: 224).
3. The Constitution of Aesthetic Harmony: An Interpretative Outline
In section 2 I enumerated the propositions constituting Kant’s description of the harmony of the faculties in aesthetic judgement. I now offer an interpretation that can coherently accommodate these propositions. I will argue that when a human subject encounters certain forms of objects, the understanding initiates but does not complete the process of cognition. The intent to cognize, however, enables the formation of the image required for the process of cognition. This image triggers, at least under ideal conditions, the production of what I call an ‘aesthetic image’; in turn, (aesthetic) reflective judgement permits the aesthetic image to give form to the encountered object-data, but without violating the obscure concepts/principles of the common understanding.
3.1
The beauty of an object is relational for Kant, i.e. it concerns the way in which the object relates to the cognitive powers of the human subject. More specifically, an object can be represented as beautiful if it is somehow ‘suitable’ for human cognition in general. Some objects are ‘suitable’ in a way that engenders the representation of beauty, while others are not. Therefore, beauty emerges in the relationship between the subject and the object. This claim is supported by a passage in which Kant says:
[T]his disposition of the cognitive powers has a different proportion depending on the difference of the objects that are given. Nevertheless, there must be one in which this inner relationship is optimal [zuträglichste] for the animation [Belebung] of both powers of the mind (the one through the other) with respect to cognition (of given objects) in general; and this disposition cannot be determined except through the feeling (not by concepts). (5: 238–9; Kant Reference Kant2000: 123)
Here Kant seems to suggest that different objects bring about different configurations of ‘cognition in general’, or different relationships between the imagination and the understanding. One can interpret this passage as follows. All images can potentially be cognized, i.e. located within a genus-species framework, and so are suited for cognition in general (= cognitively suitable).Footnote 7 Only some of these cognitively suitable images are also, in my terminology, ‘aesthetically suitable’, or suited to ‘cognition in general’ such that it is possible to perceive them as beautiful. Consequently, beauty is not merely subjective, or merely objective. It lies in the relationship between human subjectivity and a set of objects aesthetically suited to it, whatever such suitability or optimal animation might mean. As Fiona Hughes puts it, the ‘harmony within the mind could not arise were there not at the same time a harmony between thing and mind’ (Hughes Reference Hughes2010: 13–14, 2007: 267–8).Footnote 8
3.2
If the form of the object, here the tulip, is aesthetically suitable, it leads the imagination to produce freely an image of its own (henceforth ‘aesthetic image’). Kant does not explicitly speak of an aesthetic image in CPJ. However, as I will show, positing an aesthetic image accounts for his statements regarding aesthetic harmony in CPJ. Kant says:
[I]f in the judgment of taste the imagination must be considered in its freedom, then it is in the first instance taken not as reproductive, as subjected to the laws of association, but as productive and self-active (as the authoress of freely produced [willkürlich] forms of possible intuitions); and although in the apprehension of a given object of the senses it is of course bound to a determinate form of this object and to this extent has no free play (as in invention), nevertheless it is still quite conceivable that the object can provide it with a form that contains [enthält] precisely such a composition [Zusammensetzung] of the manifold as the imagination would design in harmony with the lawfulness of the understanding in general if it were left free by itself.Footnote 9 (‘General Remark’, 5: 240–1; Kant Reference Kant2000: 124–5)
Two parts of this passage are ambiguous. First, what sort of form does the object ‘provide’ the imagination? Is it the image of the object when I encounter it (= tulip-image), or is the tulip-image an occasion for the imagination to act freely? Second, the object is supposed to provide the imagination with a form which ‘contains’ the form the imagination would create if it were free, but within the limits of the understanding. Here, does the imagination actually apprehend a form in nature (= tulip-image) that it would have created itself were it free? Or should we read ‘contain’ to mean that the tulip-image is, in some sense, the outer limit, but includes within these limits (‘contains’) a form that the imagination would create if it could create freely? I argue, from (a) to (c), that the tulip-image forms the occasion for the imagination to produce an aesthetic image of the tulip, and that this aesthetic image is best interpreted as ‘contained’ (=limited) by the tulip-image.
(a)
Two reasons can be cited for positing an aesthetic image. First, Kant distinguishes between the self-active productive imagination that generates ‘freely produced forms of intuition’; and the reproductive synthesis of the imagination which is ‘subject to the laws of association’, and therefore not free to produce just any image for given data (tree shape for tulip-data, etc.) if cognition is to occur (2.1(a)). If the productive imagination in aesthetic judgement must freely compose (zusammensetzen) an image, then it could not be restricted merely to forming the tulip-image via the reproductive imagination. For, in reproductive image formation, the imagination is not free. Consequently, in aesthetic judgement, the productive imagination must create an image of its own (=aesthetic image) distinct from the image generated by the reproductive imagination. Further, both the productive and the reproductive imaginations are involved in producing an image for the sake of cognition (henceforth ‘cognitive image’) (A118). The role of the productive imagination in aesthetic judgement must be distinct from its role in cognition. In the latter case, it must produce a tulip-image in the face of tulip-data to enable its correct conceptualization as a tulip, and therefore cannot act freely. However, if, as Kant says, it acts freely in aesthetic judgement, then it cannot merely create the tulip-image, for it could have produced such an image even if it were not free. Thus, in aesthetic judgement, the productive imagination must create an image of its own, i.e. the aesthetic image, different from the cognitive image.
Second, in CPJ, Kant asserts at 5: 287 that the imagination ‘schematizes without a concept’ in aesthetic judgement. In contrast, in determinate cognition, the imagination schematizes with a concept: it produces a tulip-image for tulip-data, and thus enables cognition of the tulip. If the imagination schematizes, or images, without a concept in aesthetic judging, then the image it produces must be different from the image it would create if it were schematizing with a concept. Therefore, we must posit a distinct image to accommodate Kant’s claim that the imagination schematizes without a concept in pure aesthetic judgement.
(b)
What is an aesthetic image, and how does it differ from the cognitive image? I argue that the occurrence of an aesthetically suitable cognitive image is a necessary condition for the formation of an aesthetic image. The aesthetic image must have a different genetic structure as compared to the cognitive image. Further, it must transform the elements of the cognitive image even if, as already suggested, it cannot violate its essential outline.
I have shown that, for Kant, beauty is relational (3.1), and the reproductively synthesized cognitive image is distinct from the aesthetic image (3.2(a)). In addition, the cognitive image (schematized with concepts) must somehow ‘contain’ the aesthetic image (schematized without concepts) (5: 240–1). How do these claims cohere with each other? It seems to me that the most natural course would be to say that the presence of certain aesthetically suitable forms, like the reproductively synthesized cognitive image of the tulip, trigger the imagination into freely producing the aesthetic image, but this aesthetic image must remain within, or limited to, the contours of the cognitive image of the tulip.
To begin with, how does the tulip-image ‘trigger’ the aesthetic image? One response might be that when I encounter an aesthetically suitable cognitive image (tulip-image), I instantaneously and subliminally give the tulip-image a ‘second look’, as it were, which allows the imagination freely to form an aesthetic image different from the cognitive image of the tulip.Footnote 10 Conceiving of the relationship between aesthetic image and cognitive image in this way also helps give some general content to the notion of aesthetic suitability. Kant says that we must ‘try [each object] out’ to ascertain which objects are aesthetically suitable, because this cannot be said a priori (5: 191). If we accept that the cognitive image of an object occasions an aesthetic image only if its form is aesthetically suitable, then the criterion for such suitability is as follows. Those objects are aesthetically suitable whose cognitive image must occasion an aesthetic image under ideal conditions, that is, abstracted from all individual interests and emotions. This also implies that the formation of the cognitive image is necessary for the emergence of an aesthetic image in pure aesthetic judgement.
Further, an aesthetic image arises when the free (productive) imagination reschematizes – forms a new image out of – an aesthetically suitable cognitive image (already a schema) in accordance with its own heart, as it were. Consistent with the Kantian framework, the aesthetic image can be distinguished from the cognitive image in two ways.
First, the aesthetic image and the cognitive image must have different genetic structures. An aesthetic image necessarily requires for its genesis that the human subject perceive, however obscurely, an aesthetically suitable cognitive image. However, the aesthetic image plays no part in generating the cognitive image. To this extent, the aesthetic image and the cognitive image must be viewed as different sorts of images, which also indicates that the cognitive image can never become an aesthetic image if ‘becoming’ means intra-specific alteration (e.g. child to adult).
Second, if the free imagination cannot violate the contours of the cognitive image, then one can reasonably speculate that the activity of the imagination in reschematizing a cognitive image into an aesthetic image would involve, at least in part, the accentuation and/or obscuration of cognitive image features – e.g. the aesthetic image may be an accentuation and/or obscuration of a subset of the combined features constituting the cognitive image of the rose.Footnote 11
The following example might help make the process of formation of an aesthetic image intuitively more accessible. Suppose I see a grassless and craggy landscape which I do not find beautiful. Now, what would it entail for me to conjure up a mental image of this landscape in which I could take aesthetic pleasure? I venture that it would be the image equivalent of the statement: ‘This landscape would be beautiful if it had more/less grass here and there, and if it had more/fewer crags here and there.’ I suggest that something similar occurs in the formation of the aesthetic image in all pure aesthetic judgements, but with a difference. Namely, in this landscape example, one moves with at least some mental effort from an aesthetically ‘unsuitable’, not-beautiful object, via the formation of an aesthetic image, to the formation of an aesthetically pleasing landscape-image. In contrast, in pure aesthetic judgement, the aesthetic image is, at least under ideal conditions, formed immediately upon encounter with an aesthetically suitable cognitive image.Footnote 12
Finally, recall that Kant claims at 5: 240–1 that the object provides the imagination with a form ‘containing’ the form which the imagination would create if it were free. I take ‘contain’ to mean that the aesthetic image is limited by the cognitive image for two reasons. First, if the cognitive image were to itself provide the form which the imagination would create if it were free, then an aesthetic image need not be posited. Second, Kant says that the free imagination remains ‘bound to a determinate form of [the] object and to this extent has no free play (as in invention)’ (5: 240–1). This suggests that the aesthetic image of a rose cannot violate, or must be limited by, the cognitive image of the rose. But what does this limitation or ‘containment’ entail?
I suggest that the aesthetic image cannot violate the minimal outline that differentiates a cognitive image from every other image. This minimal outline can be compared, mutatis mutandis, to a cartoonist’s outline of a face. The cartoonist may variously alter particular features of the face, but only such that it still remains the same face-image as opposed to all other images. Similarly, an aesthetic image cannot violate the essential contours of a cognitive image that differentiate the latter from all other images. So the aesthetic image of a rose must still remain a rose-image, in that it cannot violate the essential contours of the cognitive image of the rose. Further, for Kant, the differentiation of one cognitive image from all others need not be represented ‘distinctly’, i.e. need not relate to determinate cognition (Jäsche Logik, 9: 34). Instead, the difference may be perceived clearly in that one can be ‘conscious of [how a cognitive image] differ[s] from other representations’ (B414–15n.).
(c)
From (a)–(b), positing an aesthetic image is one plausible way of making sense of Kant’s characterization of aesthetic harmony. Under ideal conditions, the free imagination creates an aesthetic image whenever it encounters an aesthetically suitable cognitive image. The aesthetic image is different from the cognitive image, but cannot violate the latter beyond a point. It must also remain within the more general constraints of the understanding, as I now discuss.Footnote 13
3.3
Neither the application of determinate concepts nor concept acquisition is the goal of aesthetic judgement. Yet Kant says that the aesthetic representation remains within the constraints of the understanding. In what way? I argue that the understanding restricts the aesthetic image both indirectly via the formation of the cognitive image, and directly by ensuring that it does not violate ordinary common sense.
(a)
If we accept that beauty is relational (3.1), and pure aesthetic judgement involves an aesthetic image triggered by the cognitive image, then the understanding in aesthetic judgement must make an initial attempt to cognize the object-data it encounters. This is because the cognitive image is produced for the sake of cognition (3.2). If the understanding makes no effort to cognize, then the cognitive image would not arise. But the cognitive image must emerge for two reasons. First, if we are unable to distinguish between different cognitive images at any given moment (e.g. tulip-image from tumbleweed-image), then we cannot distinguish aesthetically suitable from aesthetically non-suitable object-forms, which means that either all or no objects are even potentially beautiful for us. This contradicts Kant’s claim that only some objects are aesthetically suitable in the sense that they can be deemed beautiful if all the relevant conditions are fulfilled (3.1). Thus the cognitive image must be posited in aesthetic judgement. Second, I argued in 3.2 that, at least under ideal conditions, an aesthetically suitable cognitive image triggers the aesthetic image; so there could be no aesthetic image without a cognitive image. I also argued that the aesthetic image must form part of the overall structure of aesthetic judgement. Therefore, a cognitive image must also be posited.
If a cognitive image is formed, then the understanding has either cognized or begun the process of cognizing an object. Even if the understanding does not actually cognize the object in aesthetic judgement – Kant does assert at 5: 217 that determinate cognition plays no role in aesthetic judgement – one can contend that the understanding attempts to cognize, i.e. begins the process of ‘cognition in general’. One can justify this assumption on both textual and systematic grounds. Kant relates aesthetic judgement to ‘cognition in general’ (5: 217–18; section 2 above). Since ‘cognition in general’ is not determinate cognition, but aesthetic judgement relates to the general process of cognition, and if, as I have argued, the cognitive image forms a constitutive part of aesthetic judgement, then the most natural way to make these propositions cohere is to say that the understanding must attempt to cognize in aesthetic judgement. This interpretation is, of course, consistent with Kant’s claim that the understanding ‘sets’ the imagination into ‘rule-bound play’ (section 2).
(b)
Kant characterizes the ‘free lawfulness of the understanding’ as follows:
[O]nly a lawfulness without a law and subjective correspondence of the imagination to the understanding without an objective one – where the representation is related to a determinate concept of an object – are consistent with the free lawfulness of the understanding (which is also called purposiveness without an end) and with the peculiarity of a judgement of taste. (5: 240–1; CPJ 125)
In this passage, Kant distinguishes the ‘subjective correspondence’ of the imagination and the understanding from the ‘objective correspondence’ between these faculties. Objective correspondence is the correspondence of these faculties in cognition, but Kant leaves subjective correspondence unexplained.
In my view, a passage from the Deduction discussing the subjective universality of aesthetic judgement can help clarify the notion of subjective correspondence. Kant says here that ‘the proportion of these cognitive faculties that is required for taste is also requisite for the common and healthy understanding that one may presuppose in everyone’ (5: 292–3; CPJ 173). This is an odd claim. For, in the Analytic, Kant says that we judge by means of feeling in aesthetic judgement, while common understanding (gemeinen Verstande) – which Kant equates with everyday common sense (5: 293) – ‘judges not by feeling but always by concepts, although commonly only in accordance with (als nach [literally, “as according to”])Footnote 14 obscurely represented principles’ (5: 238). Given this divergence, it is unclear why the ‘proportion’ constituting aesthetic harmony should be the same as that required for common understanding. Yet this passage does provide a clue to the role of the understanding in aesthetic judgement. If common understanding judges by concepts mainly ‘in the form of obscurely represented principles’, and it has the same proportion of the cognitive faculties as exists in aesthetic harmony, then the understanding must also operate with obscurely represented concepts/principles in aesthetic harmony. Since, for Kant, concepts are rules that constrain, and the understanding restricts the imagination in aesthetic harmonization, the understanding can naturally be taken as providing lawfulness to aesthetic harmony through obscurely represented concepts/principles.Footnote 15 Moreover, since the understanding here performs a role not restricted to determinate cognition, one can see why Kant speaks of ‘understanding [or cognition] in general’ (5: 241).
(c)
From (a)–(b), subjective correspondence between the imagination and the understanding means that the understanding constrains the imagination in two ways: indirectly, since the understanding’s attempt at cognition keeps the aesthetic image within the limits of the cognitive image; and more directly, because the free imagination cannot violate the obscure concepts/principles of ordinary common sense. This constitutes what seems to me a reasonable interpretation of Kant’s claim that the faculty of imagination is subsumed under the faculty of understanding in aesthetic judgement (2.2).
3.4
Kant says that reflection and judgement form part of aesthetic judgement. So what is the relationship between reflection, judgement, cognitive image and aesthetic image? As I propose, in aesthetic judgement, judgement compares (or reflects upon) the cognitive image and the aesthetic image. Through this process of reflection, judgement allows an object to be unified in the form of the aesthetic image rather than in the form of the cognitive image. At the same time, it directs the understanding to suspend momentarily its preoccupation with cognition.
In the Jäsche-transcribed logic lectures, Kant characterizes judgement as the ‘representation of the unity of consciousness of various representations, or the representation of their relation insofar as they constitute a concept’ (9: 101). The ‘matter’ of judgement, he continues, consists of representations given to the subject, while the ‘form’ is the way in which the ‘various representations belong, as such, to one consciousness’ (9: 101). Although Kant seems to have determinate judgement in mind here, this passage can illuminate the nature of reflective judgement if we focus on the claim that judgement is a unity of consciousness. If judgement is a unity of consciousness, reflective judgement must provide some ‘reflective’ form to a manifold, and accomplish this in ‘one consciousness’. Kant says as much in CPJ. The harmony of the faculties, he says, consists in the ‘reciprocal relation of the understanding and the imagination to each other in the given representation’ (5: 286; CPJ 166; emphasis mine). Elsewhere, he says that the key characteristic of aesthetic judgement is the extent to which imagination and understanding ‘further or hinder each other in the very same representation’ (FI 20: 223; emphasis mine). It is perhaps this feature of representation in one consciousness that leads Kant to describe aesthetic judgement as ‘reflected perception’ (reflectirte Wahrnehmung), given that perception is instantaneous (5: 191).Footnote 16
Further, use of the term ‘reflected perception’ also shows that aesthetic judgement includes a reflective dimension that demarcates it from mere perception. How does Kant employ the term ‘reflection’? In the first Critique, reflection is the correct determination of representations (A260/B316). In the first introduction to CPJ, reflection is the act of ‘comparing and holding together [zuvergleichen und zusammenzuhalten]’ (20: 211). In CPJ, satisfaction in the beautiful depends more specifically on reflection on the form of the object (5: 192; section 2 above). Reflection also means comparison: if the imagination is ‘brought into accord’ with the understanding in ‘reflective comparison’, then aesthetic pleasure arises (5: 190). I now provide an account of the role of reflection in aesthetic judgement.
If judgement gives form to a manifold in one conscious representation, and given that reflection is both comparison and the correct ordering of representations, then reflective judgement must involve some act of comparison to order correctly a manifold of representations vis-à-vis each other. In fact, Kant says that the imagination cannot apprehend forms without the ‘reflecting power of judgement, even if unintentionally, at least comparing them to its faculty for relating intuitions to concepts’ (5: 190; CPJ 76). Irrespective of how reflective judgement enables imaginative apprehension (3.5 below), Kant here indicates that reflective judgement must compare (reflect on) the form of the given object with its own faculty of relating intuitions to concepts, and that aesthetic pleasure emerges if ‘in this comparison the imagination … is unintentionally brought into accord with the understanding’ (5: 190; CPJ 76). In what follows, I suggest an interpretation of this notion of comparison.
If reflective judgement compares a manifold and orders it in one conscious representation (3.4), and if aesthetic judgement includes both the aesthetic image (schematized without concept) and the cognitive image (schematized with concept) (3.1–3), then one may reasonably assume that aesthetic reflective judgement compares and orders the aesthetic image in relation to the cognitive image. This ordering would presumably consist in judgement permitting the aesthetic image rather than the cognitive image to determine the unity of the aesthetically pleasing representation (tulip) in one consciousness.Footnote 17 Thus, in aesthetic judgement, correct ordering of representations entails that judgement privileges the free activity of the imagination (aesthetic image) over the cognizing activity of the understanding (which includes cognitive image).Footnote 18 Such a judgement is concomitantly ‘reflected perception’, because the aesthetic image provides the form in which the object-data are perceived in one consciousness.Footnote 19
Two questions arise. First, how can a reflected perception also be a judgement?Footnote 20 My interpretation helps answer this question. Judgement in aesthetic judgement compares the aesthetic image and the cognitive image, and privileges the former over the latter. It is ‘reflected’ perception, because this comparing/privileging lies at the base of this perception. Yet, like empirical perception, the beautiful object is perceived in one conscious representation. Kant explains how this is possible: the power of judgement can be called a sense when ‘what is noticed is not so much its [judgement’s] reflection as merely the result of that’ (5: 293; CPJ 173). This passage arguably not only elucidates how an aesthetic judgement can simultaneously be a perception, but also the temporality of aesthetic judgement. Suppose I encounter tulip-data, and form a cognitive image of the tulip at t0. This triggers the formation of an aesthetic image at t1. Judgement compares the cognitive image and the aesthetic image at t2. Finally, at t3, judgement allows the tulip-data to be organized and objectually apprehended in the configuration of the aesthetic image rather than in the form of the cognitive image. We notice this moment at t3 when the tulip-data have been ordered into an object in accordance with the aesthetic image, and we take pleasure in the object. In contrast, t1 and t2 remain unnoticed despite being necessary components of aesthetic judgement.
Second, the claim that judgement compares/privileges the aesthetic image over the cognitive image may seem incompatible with Kant’s characterization of judgement as subsumption. But this is not the case. Unlike determinate judgement that subsumes a particular under an already given universal, aesthetic judgement subsumes the faculty of imagination under the faculty of understanding (5: 287). I argued that this subsumption entails that the aesthetic image, as product of the faculty of the imagination, can violate neither the cognitive image nor ordinary common sense, both of which relate to the faculty of the understanding (3.3). This account of subsumption in aesthetic judgement is compatible with the following characterization of aesthetic judgement as comparison. Aesthetic judgement compares the aesthetic image and the cognitive image, and permits the former to determine the form of the given object-data. But it must accomplish this in such a way that the faculty of imagination is subsumed under the faculty of understanding, i.e. keeping the aesthetic image within the essential contours of the cognitive image and the constraints of ordinary common sense.Footnote 21
3.5
Having offered a reconstructive interpretation of aesthetic harmony (3.1–4), I consider some questions and objections with regard to it.
First, Kant claims that imaginative apprehension of forms requires reflective judgement to compare these forms to its own faculty of ‘relating intuitions to concepts’ (5: 190). This could be read as saying that judgement must somehow precede the formation of the aesthetic image, which is at odds with my view that the aesthetic image emerges first, and is then permitted by judgement to unify the object. However, this statement at 5: 190 can be reconciled with my view if we distinguish between two moments: (a) the formation of the aesthetic image; and (b) judgement permitting the imagination to unify the object-data in accordance with the aesthetic image. If imaginative apprehension at 5: 190 is taken to mean (b) and not (a), then 5: 190 can be interpreted as saying that the imagination gains predominance over the understanding only if reflective judgement is involved, which is consistent with my interpretation.
Second, imagination and understanding are reciprocally animated in aesthetic harmony (section 2). In my interpretation, the imagination creates an aesthetic image freed from the tyranny of cognition; so freedom may account for its animation. But how does the understanding get animated? Following Kant, some commentators have argued that imagination and understanding are like two friends in aesthetic judgement (Henrich Reference Henrich1992: 53; Allison Reference Allison2001: 49, 154; quotation from Kant below). If the understanding and the imagination are like two friends, then the understanding should be positively disposed towards the imagination in aesthetic harmony, even though judgement allows the cognizing function of the understanding to be suspended in favour of the free imagination (3.4). Consequently, the understanding’s animation is best explained by the fact that the imagination is animated. The notion of friendship is helpful in this context. After all, we are happy when our friends get their way for the right reasons, even if it is at our expense. Hence, my interpretation can broadly account for the reciprocal animation of the faculties in aesthetic judgement.
Further, Kant’s description of faculty harmony in aesthetic judgement on the model of friendship can also help explain the generation of aesthetic pleasure. In cognition, Kant says, the understanding and the imagination are ‘like two friends who dislike but can’t relinquish each other, for they live in a continuous fight and yet can’t do without each other’ (24[1]: 707, cited in Henrich Reference Henrich1992: 53), since cognition restricts the understanding to the given in intuition, and subordinates the imagination to the categories of the understanding. This conflict however ceases in aesthetic experience in which an ‘unconstrained accordance prevails’ (24[1]: 707). On my view, at least under ideal conditions, the very perception of an object in the form of an aesthetic image ipso facto entails aesthetic pleasure, because it involves relief accruing from the release of tension between imagination and understanding via the mediation of judgement.Footnote 22
Third, there is the question of whether my interpretation is consistent with Kant’s assertion of the subjective universality, or intersubjective validity, of aesthetic judgement. Without delving deeply into this issue, I think we can see how intersubjective validity can be explained within the ambit of the interpretation proposed here. Kant says,
[aesthetic] pleasure must necessarily rest on the same conditions in everyone, since they are subjective conditions of the possibility of cognition in general, and the proportion of these cognitive faculties that is required for taste is also requisite for the common and healthy understanding that one may presuppose in everyone. [So the person judging with taste] may also demand (ansinnen) [that] his feeling be universally communicable. (5: 292–3)
This passage can be interpreted as follows. Since human beings share the capacity for cognition, the cognitive image must be the same for everyone, and the same forms must be aesthetically suitable for all. Upon encountering an aesthetically suitable cognitive image, the aesthetic image must be formed, at least under ideal conditions. The configuration of this aesthetic image must fall within a range across human beings, i.e. minor variations in the aesthetic image may occur – I may favour the whorls of the rose more than you, etc. However, under ideal conditions, i.e. if private interests are not in play, the aesthetic images formed in two people looking at the same rose cannot be diametrically opposed, because the aesthetic image cannot violate the limits of the cognitive image and ordinary common sense, both of which are intersubjectively valid. Lastly, since the modes of reflection and judgement are also commonly shared, aesthetic judgement in all human beings ought to privilege the aesthetic image over the cognitive image.Footnote 23
In essence, human beings share all the structural features constituting an aesthetic judgement. Therefore, if private interests have been set aside, all humans are entitled to expect that other humans will concur with them about the beauty of natural objects. Kant also says that aesthetic judgement and common understanding are intersubjectively valid in the same way (5: 292–3). This seems right. I can merely demand that others judge the tulip as beautiful, but I cannot expect that they necessarily will. Similarly, I expect others to act in accordance with healthy common sense, but without assuming that they necessarily will.
4. Conclusion: A Non-Cognitive Interpretation
In section 1, I noted Guyer’s distinction between pre-cognitive, multi-cognitive and meta-cognitive interpretations of the harmony of the faculties in aesthetic judgement for Kant. In section 3, I outlined an interpretation consistent with Kant’s statements on this topic, which I called the ‘new harmony theory’ (NHT). I now show that if we accept NHT, and bracket Kant’s discussion of aesthetic ideas and teleological judgement, we can deny that determinate cognition plays a role in aesthetic judgement (4.1). In 4.2, I sketch the advantages of this position, which I call the ‘non-cognitive’ interpretation of Kant’s view of the harmony of the faculties in aesthetic judgement.
4.1
According to NHT, determinate cognition need not play any role in aesthetic judgement. All that is required is that the understanding attempt to cognize (3.3). This claim makes textual and phenomenological sense. For Kant, cognitive knowledge is the primary task of the discursive intellect. Therefore, the understanding must approach all objects, including aesthetically suitable ones like tulips, with the intent to cognize. However, in the tulip case, once the cognitive image has been formed, the imagination is constituted such that it is prompted to create freely an aesthetic image without violating the cognitive image and the obscure concepts/principles of ordinary understanding. This leads judgement in aesthetic judgement to subordinate the cognitive interests of the understanding to the free play of the active imagination. Consequently, NHT gives determinate cognition no role to play in the makeup of aesthetic judgement. Concepts are relevant only to the extent that the aesthetic image cannot violate the obscure concepts/principles of ordinary common sense.Footnote 24
4.2
Pre-cognitivists argue that, in judging an object as beautiful, the initial steps of cognition result in a representation, but we are so ‘struck’ by this representation that we do not apply a determinate concept to it. Viewed from the non-cognitivist perspective, pre-cognitivists claim that we take pleasure in the cognitive image and not a distinct aesthetic image. A general problem with this view is that if we must go through all the steps of cognition but stop short of applying a determinate concept, why do we not consider all objects beautiful? Pre-cognitivism can entail that all objects must be beautiful, but only if the pre-cognitivist rejects the claim that beauty is relational (3.1). For if this claim is accepted, i.e. that only aesthetically suitable forms of objects are beautiful, then the pre-cognitivist could argue that we are ‘struck’ by the cognitive image of some objects but not by all. Therefore, pre-cognitivism need not necessarily imply that all objects must be beautiful.
Non-cognitivism avoids the ‘all must be beautiful’ objection by positing the notion of the aesthetic image. This is the advantage that the non-cognitivist has over the pre-cognitivist. The former can accommodate Kant’s 5: 240–1 claim that the imagination is freely active in aesthetic judgement and results in the aesthetic image. In contrast, even if she accepts the relationality claim, the pre-cognitivist reduces the aesthetic image to the cognitive image, and therefore cannot account for the free activity of the imagination in aesthetic judgement.Footnote 25
The non-cognitive view is of course also incompatible with the multi-cognitive approach. In the latter, an object is perceived as beautiful if the mind takes disinterested pleasure in oscillating between many conceptualizations for it without settling on a determinate one. The non-cognitive view has two advantages over this view. First, the multi-cognitive interpretation of the free play of the faculties in aesthetic judgement as an oscillation of conceptualizations has little textual justification. In addition, the non-cognitivist can offer a textually more grounded interpretation of the notion of free play. The imagination forms the aesthetic image in free play, i.e. through a free creation unhindered by the goal of cognition. Further, the ‘free play of the imagination and the understanding’ (5: 218) can be seen in terms that are to some extent deflationary. The understanding sets the imagination into ‘rule-bound’ play (3.3). Imagination creates the aesthetic image within the confines of the cognitive image and ordinary common sense. This is all there is to free play: imagination and understanding perform roles that they are not free to undertake in the perpetual grind of cognizing. Second, the multi-cognitive interpretation goes against a key Kantian proposition which the non-cognitive interpretation can accommodate. If the representation of the tulip as beautiful must occur in a single representation (3.4), then aesthetic pleasure must be felt in a single moment of pleasure. Any lingering upon this pleasure (as between oscillations) must succeed this moment. Multi-cognitivists seem to confuse this lingering with aesthetic pleasure itself. However, there is no textual evidence for this claim.Footnote 26 Moreover, the non-cognitivist can explain aesthetic pleasure without the oscillation between conceptualizations. Aesthetic pleasure arises when judgement allows the imagination to unify the object-data by way of an aesthetic image. The pleasure is simply the perception of the resulting object; any lingering comes later.Footnote 27 Thus, the non-cognitive interpretation has advantages over the multi-cognitive interpretation of the harmony of the faculties.Footnote 28
Although the non-cognitive interpretation has advantages over the pre-cognitive and multi-cognitive interpretations, it shares with these approaches the claim that determinate cognition is not required for aesthetic judgement. In contrast, Guyer proposes a meta-cognitive interpretation of the harmony of the faculties.Footnote 29 Guyer holds that determinate concepts are indeed applied in aesthetic judgement ‘for the simple reason that in both common sense and Kant’s epistemology … any judgment about an object already applies some determinate concept to it’ (2006a: 315–16). According to Guyer, we say ‘That rose is beautiful’ not ‘That is beautiful’, which means that we do apply concepts (here, ‘rose’) in aesthetic judgements. On the basis of this example, Guyer concludes that ‘the free play of our imagination and understanding … must be an experience of unity that seems to us to go beyond whatever sort of unity or organization is entailed by the concept or concepts that we have to apply to the object in order to think or talk about it at all’ (2006a: 315–16, 2006b: 178). He then articulates the ‘meta-cognitive’ interpretation as follows:
[w]e can, indeed we must be able to have ordinary cognition of the object, but we experience it as beautiful precisely because we experience it as inducing a degree or type of harmony between imagination and understanding – between the manifold it presents and our desire for unity – that goes beyond whatever is necessary for ordinary cognition. (Guyer Reference Guyer2006b: 183)
Here ‘ordinary cognition’, or determinate cognition, is involved in the experience of beauty. An object is beautiful if the imagination harmonizes with the understanding in a way that ‘goes beyond’ everyday cognition.
With regard to Guyer’s controversial meta-cognitive approach, here I simply express agreement with a comment of Malcolm Budd. Guyer argues for the involvement of determinate concepts in aesthetic judgement, because these judgements have the form ‘This rose is beautiful’ rather than ‘This is beautiful’. Budd rightly contends that we can find an object beautiful without knowing what sort of object it is, and we can individuate objects by employing ‘just concepts of colour and shape (“That red, yellow and black round thing over there [is beautiful]”)’ (2008: 113n.).
Further, Guyer’s argument can be challenged from the non-cognitivist perspective. If we accept NHT, the verbal expression of beauty could be taken to occur after the actual experience of beauty. Words need not be involved when in an aesthetic judgement the faculty of judgement allows the aesthetic image to give form to the encountered object-data, thus enabling the judging subject to take pleasure in it. Since words are unnecessary in this moment, determinate cognition need not occur. Kant certainly never claims that the verbalization of beauty necessarily accompanies the experience of beauty. Therefore, neither the aesthetic image nor the cognitive image of a rose needs to be cognized/verbalized in the actual making of an aesthetic judgement. However, our pleasure in the aesthetic image can also be cognized/verbalized subsequently as an ‘S is P’ judgement (‘This rose is beautiful’), since the aesthetic image cannot violate the cognitive image, and therefore, like the cognitive image, can be conceptualized as ‘rose’.
Finally, the non-cognitive interpretation is textually better grounded than the meta-cognitive interpretation. Guyer (Reference Guyer2006b: 183) admits that there is no direct textual evidence for the meta-cognitive interpretation. Indeed, he says that the best evidence for the meta-cognitive approach is not any particular passage, but that it is ‘the only way to make sense of all of Kant’s assumptions’ (186). The same could be asserted about the non-cognitive interpretation. But the non-cognitive interpretation is arguably closer to the text, because it accounts for Kant’s assertion that, in pure aesthetic judgement, the mental faculties are in free play because no ‘determinate concept limits them to a specific rule of cognition’ (5: 217).
Still, the question of whether the non-cognitive or the meta-cognitive interpretation fits better with Kant’s overall system admittedly requires further investigation. In terms of the NHT vocabulary, the choice between them revolves around two textual/philosophical questions. (a) Does the mere intent to cognize suffice to engender the cognitive image, or is actual cognition required? (b) On what basis does aesthetic reflective judgement privilege the free imagination over the concerns of the understanding? Addressing these questions requires exploring the relationship between determinate cognition, aesthetic judgement, aesthetic ideas and teleological judgement,Footnote 30 which I am in the process of doing elsewhere.Footnote 31
