1. Introduction
In the Prolegomena to Any Future Metaphysics, written between the publication of the first and second editions of the Critique of Pure Reason (CPR), Kant draws a distinction between two types of empirical judgements: judgements of perception and judgements of experience. The former are only subjectively valid: ‘they hold only for us, i.e., for our subject’, while the latter have objective validity: we ‘intend that they should be valid at all times for us and for everyone else’. Judgements of perception ‘do not require the pure concepts of the understanding, but only the logical connection of perceptions’, while judgements of experience require ‘in addition special concepts originally generated in the understanding’ (§18, 4: 298).Footnote 1
The Prolegomena distinction is not mentioned in either edition of the CPR. In the B Deduction, Kant draws a distinction between the subjective associations of imagination on the one hand, and the objective unity of apperception, which attaches to all judgements, on the other (§19, B142). This has led some commentators to conclude that the B Deduction abandons the distinction drawn in the Prolegomena in favour of a more unified account of judgements, in which they are all objectively valid. Paul Guyer argues that the distinction between judgements of perception and judgements of experience is inherently at odds with the CPR assumption that any form of apperception presupposes the categories (Guyer Reference Guyer1987: 101). In the first edition of his Kant’s Transcendental Idealism, Henry Allison argues that the Prolegomena distinction is replaced in the CPR by the distinction between subjective and objective unities of consciousness (Allison Reference Allison1983: 148–53). Altering his position in the book’s revised and enlarged edition, Allison aligns himself with Ernst Cassirer, who argued that Kant’s notion of a judgement of perception ‘is only meant to be a methodologically constructed limiting-case’, to provide a contrast to the notion of scientific objectivity (Allison Reference Allison2004: 179–82; Cassirer Reference Cassirer1953: 245–6). Norman Kemp Smith also argued that the distinction is ‘thoroughly misleading’, and fundamentally at odds with Kant’s fully developed critical system (Smith Reference Smith1918: 288–9).
In this paper I aim to salvage the much-maligned Prolegomena distinction. I will argue that, with help from the System of Principles, the Prolegomena distinction can, indeed, be reconciled with the B Deduction account of judgement. To that end, I will begin with an overview of Kant’s treatment of the notion of objectivity, which is central to the distinction between perception and experience in his system (section 2). I will go on to look at the System of Principles in the CPR, arguing that the mathematical principles (Axioms of Intuition and Anticipations of Perception) involve the generation of perception from empirical intuition, while the Analogies of Experience involve the generation of experience from perception (section 3). I will then look at Kant’s account of the distinction between judgements of perception and judgements of experience in the Prolegomena, and argue that his account of the distinction strongly suggests that the ‘special concepts’ required for making judgements of experience are those belonging to the relational categories (section 4). Finally, I will show that the framework I have developed allows us to circumvent the most common problem faced in reconciling the Prolegomena distinction with the account of judgement in the B Deduction (section 5).
2. Overview: Objectivity and Determinacy
In Kant’s critical system, ‘representation’ is the umbrella term used for all mental states. Under ‘representation’ fall sensations, intuitions, concepts, cognitions and even judgements. Human intuitions are either pure or empirical. Pure intuition is the mere form of space and time, and empirical intuition is the spatiotemporal form ‘filled in’ with sensory matter. Concepts, which belong to the faculty of understanding, can also be pure or empirical. If pure, concepts are called categories.Footnote 2 The categories are characterized as general rules of synthesis (B105), where synthesis is to be understood, in the most basic terms, as the process of combining and ordering complex representations into a unified whole. Together, space, time and the categories are the pure forms under which every object that can come before us must necessarily stand.
Intuitions, concepts and judgements are each characterized as ‘objectively valid’ at some point in the CPR.Footnote 3 If one is to make sense of this seemingly indiscriminate usage, it has to be allowed that Kantian objectivity admits of gradations. In the lectures on logic transcribed by Jäsche, Kant characterizes the ‘objective content’ of cognitions in precisely these terms: degrees. Here, he distinguishes between various levels of cognition, allowing that it is only by means of concepts that we can ‘understand’ (verstehen) something (9: 64–5).Footnote 4
In delineating these gradations of objectivity, I will draw on Kant’s notion of determination (Bestimmung) in the transcendental sense (i.e. in the sense having to do with the mind’s a priori contribution to cognition).Footnote 5 Kant talks about transcendental determination in the Amphiboly section of the CPR, where he characterizes form as ‘determination’ and matter as ‘the determinable in general’ (A266/B322).Footnote 6 In rough terms, then, to determine something in the transcendental sense is to impose a priori form on its matter. Again, according to Kant there are of course a number of a priori forms of experience: space, time and the categories. It must be the case, then, that determinacy (i.e. ‘enformedness’), also admits of gradations.Footnote 7
In the Transcendental Aesthetic, appearance is introduced as ‘the undetermined object of an empirical intuition’ (B34/A20). There is, however, a more general sense in which ‘appearance’ is used in the transcendental idealist framework: even when the (originally indeterminate) object of empirical intuition is determinate, it is still ‘only’ an appearance in contrast to things in themselves. This fully determinate object of representation (i.e. appearance in the second sense) is a ‘phenomenon’.Footnote 8 At the most basic level, the object of a sensible intuition is indeterminate (or ‘determinable’) with regard to pure concepts of the understanding.Footnote 9 As the categories are ‘concepts of an object in general’, subsumption under the categories is necessary for the further determination of objects and their relations (B148, A137/B176). As representations are synthesized, their objects go from being appearances (in the first sense) to being phenomena.Footnote 10 What is a composite of matter and form at one level of representation (intuition) can serve as matter at a higher level of representation (experience).
Now we can return to the discrete gradations of objectivity. As with the formal determinacy of objects, so with the objectivity of representations: the more formally determinate an object, the more objective its representation. The objectivity of a representation is, in other words, an indication of the extent to which its object is determined with regard to the a priori forms of experience.Footnote 11 A fully objective representation would be an experience or empirical cognition. An entirely subjective representation, on the other hand, would be a sensation:Footnote 12 sensations are ‘immediate’ representations, meaning they are unprocessed by our cognitive apparatus (i.e. they are pure matter and formless).Footnote 13 As the forms of experience are the a priori conditions of cognition for all discursive spatiotemporal intellects, objective validity can be understood, in the most general sense, as formal inter-subjective validity.
So far, we have established the following preliminary levels of objectivity and determination shown in the table.
It should be clear that this purely formal notion of objectivity is a necessary but not sufficient condition of the empirical truth of cognitions, if by empirical truth we understand the agreement not only of the form but also of the content of cognition with the object to which it is related (A59–60). Objective validity is thus a negative criterion of empirical truth, and objectively valid empirical judgements are those that are capable of being true for everyone.
3. A Priori Principles
Kant characterizes the a priori principles of the understanding as ‘principles of the objective determination of all representations’ (B142) and ‘rules of the objective use of the categories’ (B201). The categories are divided into two groups: the mathematical categories (quantity and quality) pertain to objects of intuition; the dynamical categories (relation and modality) concern the existence of the objects of intuition, ‘in relation to each other or to the understanding’ (B110). The a priori principles of the understanding, likewise, fall into two groups: the mathematical principles (Axioms of Intuition and Anticipations of Perception) and the dynamical principles (Analogies of Experience and Postulates of Empirical Thinking in General).Footnote 14
In a note added to the second edition of the System of Principles, Kant elaborates on the nature of the syntheses corresponding to these two kinds of principles:
All combination (conjunctio) is either composition (compositio) or connection (nexus). The former is the synthesis of a manifold of what does not necessarily belong to each other, as e.g., the two triangles into which a square is divided by the diagonal do not of themselves necessarily belong to each other, and of such a sort is the synthesis of the homogeneous in everything that can be considered mathematically (which synthesis can be further divided into that of aggregation and of coalition, of which the first is directed to extensive magnitudes and the second to intensive magnitudes). The second combination (nexus) is the synthesis of that which is manifold insofar as they necessarily belong to one another, as e.g. an accident belongs to some substance, or the effect to the cause – thus also as represented as unhomogeneous but yet as combined a priori, which combination, since it is not arbitrary, I call dynamical, since it concerns the combination of the existence of the manifold (which can again be divided into the physical combination of the appearances with one another and the metaphysical, their combination in the a priori faculty of cognition). (B201–2)
The syntheses carried out in accordance with the first two categories thus compose the extensive and intensive magnitude of appearances (hence the mathematical principles are also characterized as ‘constitutive’), while those carried out in accordance with the relational categories connect appearances (the dynamical principles are called ‘regulative’). Note that according to the quoted passage, the connection of appearances in accordance with the dynamical principles contains a different kind of necessity than that which their mere composition does. Kant states that the mathematical principles are ‘unconditionally necessary’ in regard to possible experience, while the dynamical principles are necessary only ‘under the condition of empirical thinking in an experience’ (B199/A16). This is because it is formally necessary that every intuition have an extensive and intensive magnitude, while it is formally contingent whether the object of an intuition exists or not:Footnote 15 according to the second Postulate, for something to exist, it must be ‘connected with the material conditions of experience (of sensation)’ (B266, my emphasis).Footnote 16 Thus, it is only formally necessary that if an object exists, it stands in determinate relations with other existing objects. The dynamical principles concern a different kind of necessity than the mathematical ones because they presuppose the formally contingent existence of objects.
Axioms of Intuition and Anticipations of Perception
The Axioms of Intuition govern the synthesis by which an appearance is taken up into ‘empirical consciousness’, which is alternately referred to as perception (Wahrnehmung) (B207). The synthesis in accordance with the Axioms involves the combination of the homogeneous in the manifold of intuition, where ‘the homogeneous’ refers to the continuum of space and time in intuition. The Axioms tell us that all intuitions, as regards their spatiotemporal form, are extensive magnitudes: they are constituted through the successive addition of one homogeneous unit to another. The synthesis in accordance with the Axioms thus generates the representation of a determinate region of space and interval of time (i.e. an extensive magnitude) in the manifold of intuition.
The Anticipations of Perception are concerned not with the spatiotemporal in perception, but with what Kant calls ‘the real’ in them – the degree of the sensation by which the subject is affected. As I said earlier, sensation is a subjective representation – it is a modification of the state of the subject. Despite the subjective character of sensation, however, the Anticipations tell us that it can be cognized a priori that every sensation has a sensory content of a determinate degree (i.e. an intensive magnitude) which corresponds to the degree of influence of the objects of perception on sense. An absence of ‘the real’ in sensation would, thus, amount to the absence of all perception.
Together, the syntheses carried out in accordance with the first two principles – the synthesis of the form of the intuition and the synthesis of the ‘real’ in the intuition – yield perception: the first takes up empirical intuition into ‘empirical consciousness’ (i.e. into apprehension) by generating the representations of a determinate extensive magnitude, and the second ‘anticipates’ all perception as having an intensive magnitude corresponding to ‘the real’ in the object of perception. I will refer jointly to these two syntheses as the ‘synthesis of apprehension’, and understand perception to be apprehended empirical intuition.
Analogies of Experience
It is only in the Analogies of Experience that Kant starts speaking of experience (Erfahrung) in the System of Principles, and he repeatedly distinguishes it from perception, the subject of the previous two principles: he defines experience as the ‘representation of a necessary connection of perceptions’; as ‘empirical cognition, i.e. a cognition that determines an object through perceptions’ (B218); and, finally, as ‘the synthetic unity of perceptions’ (A183/B226). He goes on to characterize the Analogies as rules ‘in accordance with which unity of experience is to arise from perceptions’ (A180/B223).
The Analogies, thus, represent the conditions that are necessary for perception (apprehended intuition) to yield experience. They are ‘regulative’ of intuition but ‘constitutive’ of experience (A664/B692). As was previously stated, Kant characterizes the dynamical principles as dealing with the existence and relations of objects (A178/B221). Recall from earlier that, according to the second Postulate, an object exists if it ‘is connected with the material conditions of experience (of sensation)’ (B266). Roughly, then, for an object to exist, it has to be connected to a sensory state. Kant makes it clear that this connection does not have to be immediate: an object exists so long as it is connected to the object of some sensory state in accordance with the Analogies (A225). Thus, insofar as their existence is concerned, all objects of experience stand in determinate dynamical relations to one another.
Kant says that the Analogies are ‘rules of general time-determination’ (B220). Accordingly, the Analogies present the rules in accordance with which all existing appearances (‘objects of experience’) stand in determinate time relations. The first Analogy states that substance (the persisting, underlying matter of all appearance) is the precondition of all temporal relations between objects of experience. The second Analogy states that all ‘alteration’ (i.e. the succession of one state by another in the same object) in objects of experience happens in accordance with the law of the connection of cause and effect, where the cause is a sufficient condition for the bringing about of the effect. The third Analogy, lastly, states that all objects of experience, insofar as they coexist in space, are in thoroughgoing causal interaction (immediately or mediately).
Kant says that experience, in contrast to perception, represents the ‘relation in the existence of the manifold … not as it is juxtaposed in time but as it is objectively in time’ (B219). Given that apprehension is always successive while appearances can coexist, there is a difference between the ‘subjective’ time of inner sense in which we apprehend intuitions and the ‘objective’ time in which objects of experience exist and interact with one another (cf. Watkins Reference Watkins2005: 189). The various states of an apprehended object might coexist, as do for example the different parts of a house, but they are apprehended successively in time. In other words, perception only presents the succession of states in the subject, not their priority or simultaneity in the object (B233, B257). Without the a priori synthesis of perceptions in accordance with the relational categories, it would not be possible to distinguish between subjective and objective time.
Thus the temporal relations of perceptions in apprehension do not necessarily correspond to the objective temporal relations that hold between actual appearances. This means that the synthesis of apprehension generates representations that are objective with respect to their extensive and intensive magnitudes, but subjective with regard to their temporal relations. We can thus say that perception is a partially objective representation. For a representation to be fully objective, it needs to represent the necessary connections that hold between objects of experience in accordance with the Analogies.Footnote 17 This representation would be an empirical cognition, or experience. We can now fine-tune the preliminary table presented in section 2 (see p. 355).
* Kant characterizes the relations of inherence, causality and reciprocity holding between the objects of experience as ‘the dynamical relations from which all others arise’ (B262).
4. The Prolegomena Distinction
The Prolegomena was written before the publication of the second edition of the CPR and is meant to offer an overview of that work. Its main purpose, as expounded in the preamble, is to answer the question: how is metaphysics possible? This question is tackled in several steps, each of which roughly corresponds to the Transcendental Aesthetic, Transcendental Analytic, Transcendental Dialectic and Doctrine of Method in the CPR. The second part, whose content roughly overlaps with that of the Transcendental Analytic, is concerned with the question: how is pure natural science possible? Later, Kant further clarifies the problem by reformulating the question as: ‘how is it possible in general to cognize a priori the necessary conformity to law of things as objects of experience?’ (§17, 4: 296).
The laws in question here are what Kant terms the ‘universal laws of nature’ (§15, 4: 295), and they are quite different from empirical laws of nature, in that the latter can never be cognized a priori. The universal laws of nature, which comprise the content of pure natural science proper, ‘exist fully a priori’, are purely ‘formal’ and determine the existence of all objects of experience (§15–16, 4: 295–6). Recall from the previous section that the Analogies determine the existence and relation of appearances a priori. Already, then, there is a strong indication that Kant is primarily concerned with the Analogies in the second part of the Prolegomena. The primacy of the Analogies will become more evident in the ensuing discussion of the Prolegomena account of empirical judgements.
Soon after his reformulation of the question at hand, Kant draws the distinction between judgements of perception and judgements of experience:
[A]lthough all judgements of experience are empirical, i.e., have their basis in the immediate perception of the senses, nonetheless the reverse is not the case, that therefore all empirical judgements are judgements of experience; rather, beyond the empirical and in general beyond what is given is sensory intuition, special concepts must be added, which have their origin completely a priori in the pure understanding, under which every perception can first be subsumed and then, by means of the same concepts transformed into experience.
Empirical judgements, insofar as they have objective validity, are judgements of experience; those, however, that are only subjectively valid I call mere judgements of perception. The latter do not require pure concepts of the understanding, but only the logical connection of perceptions in a thinking subject. But the former always demand, beyond the representations of sensory intuition, in addition special concepts originally generated in the understanding, which are precisely what make the judgements of experience objectively valid. (§18, 4: 297–8)
Judgements of perception are subjectively valid – that is to say, ‘they hold good only for us, i.e., for our subject’ (§18, 4: 298), while judgements of experience have objective validity, which, Kant says, is equivalent to ‘necessary universal validity’, i.e., holding for everyone (§19, 4: 298). The main difference between the two types of judgement is that judgements of perception do not require pure concepts of understanding but ‘only the logical connection of concepts’, whereas judgements of experience require ‘special concepts originally generated in the understanding’. The objective validity of judgements of experience is due to the latter’s use of these special concepts of the understanding. All (empirical) judgements, Kant writes, are ‘at first’ mere judgements of perception, and ‘only afterwards do we give them a new relation, namely to an object, and intend that the judgement should also be valid at all times for us and for everyone else’ (§18, 4: 298).Footnote 18
I hope to show that a close reading of the Prolegomena makes it abundantly clear that the ‘special concepts’ that make objectively valid judgements of experience possible are those belonging to the category of relation (subsistence, causality, community).Footnote 19 This would mean that the sense in which judgements of perception are subjective concerns the relation they posit between appearances. The necessity expressed in judgements of experience would, accordingly, concern the relation that holds between appearances in accordance with the Analogies.
The Prolegomena Examples
The first examples Kant provides for judgements of perception are ‘the room is warm, the sugar sweet, the wormwood repugnant’. These judgements ‘express only a relation of two sensations to the same subject, namely myself, and this only in my present state of perception, and are therefore not expected to be valid for the object’ (§19, 4: 299). The judgements in this set of examples can never become judgements of experience ‘because they refer to mere feeling – which everyone acknowledges to be merely subjective and which must therefore never be attributed to the object’ (§19n., 4: 299).
Kant’s second and third examples of judgements of perception are ones that can become judgements of experience. Curiously, the two types of judgement in the second example are identical in their formulation: Kant says that in making the judgement ‘The air is elastic’, one could either mean it as a judgement of perception, in which case only two perceptions (in the subject) are related to one another; or one could mean it as a judgement of experience, in which case one requires that the relation of the perceptions ‘be subject to a condition that makes it universally valid’ (§19, 4: 299).Footnote 20 In the third example, Kant contrasts the mere judgement of perception ‘If the sun shines on the stone, it becomes warm’ with the judgement of experience ‘The sun warms the stone’. In the latter formulation, ‘beyond the perception is added the understanding’s concept of cause, which connects necessarily the concept of sunshine with that of heat’ (§20n., 4: 301).
Kant characterizes the process through which the judgement of perception ‘The air is elastic’ is converted into a judgement of experience as follows:
Now before a judgement of experience can arise from a judgement of perception, it is first required: that the perception be subsumed under a concept of the understanding of this kind; e.g., the air belongs under the concept of cause, which determines the judgement about the air as hypothetical with respect to expansion. This expansion is represented not as belonging merely to my perception of the air in my state of perception or in several of my states or in the states of others, but as necessarily belonging to it, and the judgement: the air is elastic, becomes universally valid and thereby for the first time a judgement of experience, because certain judgements occur beforehand, which subsume the intuition of the air under the concept of cause and effect, and thereby determine the perceptions not merely with respect to each other in my subject, but with respect to the form of judging in general (here, the hypothetical), and in this way make the empirical judgement universally valid. (§19, 4: 300–1)
Kant writes that before a judgement of experience can arise from a judgement of perception, ‘certain judgements must occur’. These judgements ‘subsume the intuition of the air under the concept of cause and effect’ and thus ‘determine the perceptions … with respect to the form of judging in general’. To subsume representations under the categories is thus to determine them with regard to the logical forms of judgement. This means that for a combination of concepts to be indeterminate with regard to the logical forms of judgement is for it to be indeterminate with regard to one of the categories.Footnote 21 In the judgement of perception ‘The air is elastic’, the representation of air is indeterminate with regard to the relational categories: we know this because Kant says it has to be subsumed under the concept of cause and effect in order to be determined with regard to the hypothetical form of judgement. The following is a reconstruction of the process described in the passage above, through which judgements of perception turn into judgements of experience.
We start with:
(1) Judgement of perception: The air is elastic.
Which, in view of a nominal definition of elasticity as ‘expanding after compression’, can be turned into:
(2) Judgement of perception: After air is compressed, it then expands (i.e. returns to its original volume).
The ‘condition’ of the judgement of perception (here, compression) is subsumed under the concept of cause.Footnote 22 This gives us:
(3) Judgement of experience: If air is compressed (cause), it expands (effect).
Which can be turned into:
(4) Judgement of experience: The air is elastic.
I do not imagine Kant is suggesting that, in making a judgement of experience, we explicitly go through steps (1) to (4). The process outlined above seems, rather, to be what would paradigmatically be involved in making judgements of experience. While the intricacies of just what is involved in ‘subsumption’ under the category of relation are controversial and not the focus of this paper, Kant is in any case explicit here that the process would involve making ‘certain judgements’.Footnote 23
Thus it is subsumption under pure concepts (specifically, under those concepts belonging to the category of relation – more on this in the next section), and not the mere logical form of judgement, that gives us licence to characterize the connection of appearances in necessary and universal terms.Footnote 24 Given that the logical forms of judgement belong to pure general logic and the categories to transcendental logic, this should not come as a surprise. In the CPR, pure general logic is described as ‘abstracting from all empirical conditions under which our understanding is exercised’, and dealing with ‘the mere form of thinking’ (A53/B77–A54/B78). In contrast, transcendental logic is concerned with ‘the origin of our cognitions of objects insofar as that cannot be ascribed to the objects’ (B80/A56), i.e. ‘with the part of our thought that has its origin solely in the understanding’; it supplies ‘the principles without which no object can be thought at all’ (A62/B87). Therefore, the logical forms of judgement, on their own, are not sufficient for ‘introducing’ a transcendental content into representations, that is, for determining them with respect to the categories.Footnote 25
‘Special Concepts’
As we saw, judgements of experience require ‘special concepts’ of the understanding. Now, as I argued in section 3, perceptions (i.e. apprehended intuitions) are indeterminate with respect to the category of relation. Likewise, the judgements of perception Kant lists in the second and third examples of the Prolegomena were shown to be indeterminate with regard to the relational categories. Clearly, then, the ‘special concepts’ required for making judgements of experience are those belonging to the category of relation. As we progress in the second part of the Prolegomena, it only becomes more evident that Kant takes all and only judgements of experience to utilize the concepts belonging to the category of relation.Footnote 26 The Analogies, we saw in the previous section, make it possible for experience to arise from perception. They are, accordingly, what makes it possible for judgements of perception to give rise to judgements of experience. In his account of the Analogies in the Prolegomena Kant writes:
With respect to the relation of appearances, and indeed exclusive with regard to their existence, the determination of this relation is not mathematical but dynamical, and it can never be objectively valid, hence fit for experience, if it is not subject to a priori principles, which first make the cognition of experience possible with respect to that determination. Therefore, appearances must be subsumed under the concept of substance, which, as a concept of the thing itself, underlies all determination of existence; or second, insofar as a temporal sequence, i.e., an event is met with among the appearances, they must be subsumed under the concept of an effect in relation to a cause; or, insofar as coexistence is to be cognized objectively, i.e., through a judgement of experience, they must be subsumed under the concept of community (interaction): and so a priori principles underlie objectively valid, though empirical judgements, i.e., they underlie the possibility of experience insofar as it is supposed to connect objects in nature according to existence. These principles are the actual laws of nature, which can be called dynamical. (§25, 4: 307)
The passage quoted makes it clear that the determination of the existence of appearances is objectively valid only if it is ‘subject to a priori principles’, that is to say, subject to the Analogies (since these are the principles that deal with the existence and relation of appearances). All judgements of experience thus involve the subsumption of perceptions under concepts belonging to the category of relation. The Analogies are described as ‘the actual laws of nature’, and, given the description of nature as ‘the whole object of all possible experience’ (§17, 4: 297), it is evident that the Analogies are the rules to which all judgements of experience conform.Footnote 27 The empirical laws of nature (for example, Newton’s universal law of gravitation), insofar as they carry with them an expression of necessity, are grounded in the Analogies as the universal a priori laws of nature.Footnote 28
Corresponding to this account, we can further fine-tune the table developed in sections 2 and 3 (see p. 362).
* The relation of subject to predicate, ground to consequence, or part to whole, corresponding to the categorical, hypothetical, and disjunctive forms of judgement. Cf. A266/B322: ‘In every judgement one can call the given concepts logical matter (for judgement), their relation (by means of the copula) the form of the judgement.’ While the form of both judgements of perception and judgements of experience is the relation of concepts, only the latter are determinate with respect to the category of relation.
** I am following Longuenesse’s account of concept formation, according to which empirical concepts are generated through analysing synthesized representations through processes of reflection, comparison and abstraction (1998: ch. 6).
Two Kinds of Judgements of Perception
Judgements of perception can legitimately only express the connection of perceptions, not the connection of the objects of perception. However, this does not mean that such judgements only ever express what Prauss (1971) terms ‘subjective-private states’ (as quoted in Longuenesse Reference Longuenesse1998: 188). While the relation they posit between perceptions depends on phenomenal consciousness (and is, thus, only subjectively valid), the perceptions themselves can have objects other than the subject’s own representations. It is difficult to see how judgements regarding subjective-private states can ever be conferred with necessity and universality – i.e. hold for everyone, as judgements of experience do. Nonetheless, Kant makes it clear that at least some judgements of perception can be transformed into judgements of experience, and that, in fact, all judgements of experience are initially mere judgements of perception (§18, 4: 298). I suggest, therefore, that judgements of perception are about partially determinate appearances, in the sense laid out in section 3: appearances determined with respect to their quantity and quality, but not with respect to their relations. Some judgements of perception (e.g. ‘sugar tastes sweet’) can never be turned into objectively valid judgements of experience. Others (e.g. ‘if the sun shines on the stone, it grows warm’) can, if the constituent perceptions are subsumed under the category of relation.
Kant’s first set of examples can be said to be judgements of feeling. While Kant does not elaborate any further on feeling (Gefühl) in the Prolegomena, select passages in the CPR provide helpful remarks. In the Transcendental Aesthetic Kant characterizes feeling as ‘an effect of the sensation’, and ‘pleasant taste’ as grounded in feeling (A29). He further characterizes ‘colors, sounds, and warmth’ as ‘merely sensations and not intuitions’, and thus not ‘allow[ing] any object to be cognized, least of all a priori’ (B44/A28).Footnote 29 These accounts are roughly in keeping with the first set of Prolegomena examples, which describe, respectively, warmth, taste and displeasure. These cannot ever become judgements of experience – i.e. they will only ever express the subject’s present state. As Longuenesse puts it, they ‘attribute[e] to the object a predicate that really expresses a mere feeling’ (1998: 173, note 13).
The second group of judgements of perception – those that can become judgements of experience – are more than mere judgements of feeling. I suggest they can be characterized as either (i) singular judgements;Footnote 30 or (ii) inductive generalizations.Footnote 31
A singular judgement of perception is valid only for the subject and at the time of its utterance, since any lasting and inter-subjective validity would presuppose at least a permanently persisting substratum in which properties can inhere (and potentially the concepts of causality or community, if the judgement posits a necessary temporal sequence or coexistence). As was previously argued, just because a certain representation assumes the position of subject in a judgement does not mean that its object is thought of as determined with respect to the pure concept of substance: I can say ‘The air is elastic’ without having subsumed air under the pure concept of substance. In such a case, the judgement, though categorical, is not determinate with regard to the category of relation, which means the relation it posits is not objectively valid.
An inductive generalization would lay claim to what Kant characterizes in the CPR as ‘assumed or comparative universality’, which only licenses one to say ‘so far as we have observed until now, no exception has been found to this or that rule’. An objectively valid judgement of experience, however, lays claim to what Kant characterizes as ‘strict universality’, such that ‘no exception at all is allowed to be possible’ (B3–5). Strict universality can never be derived empirically, which only ever supplies the grounds for comparative universality. Judgements of perception that express inductive generalizations thus can only lay claim to comparative universality, which lacks the a priori grounding that allows for strict universality. Ultimately, the relation such judgements posit is not objectively valid either.
5. The Problem of the B Deduction
In §19 in the B Deduction, Kant characterizes judgement as ‘nothing other than the way to bring given cognitions to the objective unity of apperception’ (B141). The logical form of judgement is thus identified with the form of the objective unity of apperception (OUA), that is, a consciousness that is able to relate to objects. Longuenesse observes that the form of judgement ‘do[es] not relate my states to myself, but marks to objects’. The form of judgement thus carries with it the ‘norm’ to be objectively valid – that is to say, capable of being true for everyone (1998: 186).
This might seem to imply that a judgement of perception is a contradiction in terms: Kant does, after all, emphasize that such judgements are only subjectively valid. The account of judgement in the B Deduction has led some commentators, foremost among them Allison, to argue that the Prolegomena distinction is abandoned in the second edition of the CPR in favour of the distinction between associations of the imagination, which are subjectively valid, and judgements, which are objectively valid (Allison Reference Allison1983: 148–53). Here is what Kant has to say about the former in the B Deduction:
In accordance with [laws of association] I could only say ‘When I carry a body, I feel a pressure of weight,’ but not ‘It, the body, is heavy,’ which would be to say that these two representations are combined in the object, i.e., regardless of any difference in the condition of the subject, and are not merely found together in perception (however often as that might be repeated). (§19, B142)
The laws of association are a function of the imagination, and they ensure that ‘empirical ideas that have frequently followed one another produce a habit in the mind such that when one idea is produced, the other also comes into being’ (Anth, 7: 176). This is, of course, reminiscent of Hume’s characterization of custom or habit as the basis of our belief in causal connections (1993: 28). In fact, Kant mentions the laws of association in connection with Hume in his preface to the Prolegomena. Here he observes that for Hume the notion of a necessary connection between cause and effect is ‘really nothing but a bastard of the imagination, which, impregnated by experience, and having brought certain representations under the law of association, passes off the resulting subjective necessity (i.e., habit) for an objective necessity (from insight)’ (4: 258).
In the revised and enlarged second edition of Kant’s Transcendental Idealism, Allison alters his position with regard to judgements of perception, arguing that there is no doctrinal change in Kant’s account of judgement between the publications of the Prolegomena and the B edition of the CPR. Following Cassirer, he attributes a didactic role to the notion of judgements of perception, claiming Kant devised it simply in order to expose the shortcomings of the ‘standard empiricist’, that is, Humean, account of empirical judgement (cf. Cassirer Reference Cassirer1953: 245–6). Allison goes on to argue, however, that the role of judgements of perception in the Prolegomena is not solely didactic, as Kant explicitly characterizes such judgements as ‘the first stage in a cognitive process’. The seeming incompatibility between the B Deduction account of judgement and the Prolegomena distinction is finally attributed to the Prolegomena’s ‘particular goal and method’, that is to say, its analytic (as opposed to synthetic) method, and its function as merely ‘preparatory’ for those wishing to become familiar with the critical system (Allison Reference Allison2004: 179–82; cf. 2015: 303–4).Footnote 32
I believe the framework developed in this paper provides an alternative to both Allison’s original and revised positions, and has the advantage of neither (i) attributing doctrinal confusion or a change of heart to Kant (as per Allison’s first position), nor (ii) explaining away the apparent tension between the B Deduction and the Prolegomena by an appeal to pragmatic considerations on Kant’s part (as per Allison’s revised position).
If we allow, as I have argued, that appearances admit of degrees of determination, and that the mere form of judgement is not sufficient for the assertion of a necessary and universal connection between appearances, then we have to allow that there can be judgements that are not fully objective (i.e. judgements of perception). However, because (as §19 in the B Deduction makes clear) the form of all judgements is the form of the OUA, judgements of perception, in cases where they are not judgements of feeling, can become judgements of experience: as judgements, they carry with them the ‘norm’ to achieve full objective validity.
On the other hand, subjective associations of the imagination can never give rise to universal rules. They are, properly speaking, mere expressions of the subject’s representational states, and are thus not truth-valued (cf. Allison Reference Allison2004: 88). Of course, it is consistent to argue that judgements of perception utilize empirical associations, habitually produced by the imagination, in asserting a connection between appearances (cf. Höffe Reference Höffe1994: 74). In such a case, however, they would not be mere associations of the imagination: their form would be the form of the OUA, which carries with it the synthesis of apprehension. Strictly speaking, then, in the absence of the OUA, mere associations of the imagination would only be able to express a relation between unprocessed representations.
Unlike what Cassirer and Allison assert, then, it is not judgements of perception but subjective associations of the imagination that Kant uses didactically as ‘a methodologically constructed limiting case’, meant to ‘throw light upon the newly gained concept of scientific objectivity by force of contrast’ (Cassirer Reference Cassirer1953: 245–6). As Longuenesse puts it, the distinction between subjective associations of imagination and objective judgements is not a distinction between two types of empirical judgement (as the Prolegomena distinction is) but between two proposed origins of empirical judgement (1998: 188). The first would be something akin to the ‘bundle theory’ offered by Hume, wherein there is no unified subject of experience, only associated representations. These associations, even if expressed in the form of judgements, would never have license to posit strict universality. The second would be Kant’s account of the OUA as present in the form of all judgements. The OUA drives and makes possible the progression from judgements of perception into judgements of experience, which alone admit of strict universality.Footnote 33
