Start Submission Become a Reviewer

Reading: ‘For Me, In My Present State’: Kant on Judgments of Perception and Mere Subjective Validity

Download

A- A+
Alt. Display

Research

‘For Me, In My Present State’: Kant on Judgments of Perception and Mere Subjective Validity

Author:

Janum Sethi

University of Michigan, Ann Arbor, US
X close

Abstract

Few of Kant’s distinctions have generated as much puzzlement and criticism as the one he draws in the Prolegomena between judgments of experience, which he describes as objectively and universally valid, and judgments of perception, which he says are merely subjectively valid. Yet the distinction between objective and subjective validity is central to Kant’s account of experience and plays a key role in his Transcendental Deduction of the categories. In this paper, I reject a standard interpretation of the distinction, according to which judgments of perception are merely subjectively valid because they are made without sufficient investigation. In its place, I argue that for Kant, judgments of perception are merely subjectively valid because they merely report sequences of perceptions had by a subject without claiming that what is represented by the perceptions is connected in the objects the perceptions are of. Whereas the interpretation I criticize undercuts Kant’s strategy in the Deduction, I argue, my interpretation illuminates it.

How to Cite: Sethi, J., 2020. ‘For Me, In My Present State’: Kant on Judgments of Perception and Mere Subjective Validity. Journal of Modern Philosophy, 2(1), p.9. DOI: http://doi.org/10.32881/jomp.86
212
Views
44
Downloads
  Published on 21 Aug 2020
 Accepted on 04 Jun 2020            Submitted on 30 Oct 2019

Few of Kant’s distinctions have generated as much puzzlement and criticism1 as the one he draws in the Prolegomena between what he calls judgments of experience (henceforth, JOEs)—which he identifies as ‘objectively and universally valid’—and judgments of perception (henceforth, JOPs)—which he says are ‘merely subjectively valid’ (Pr., 4:299).

According to a prominent interpretive narrative,2 the notion of a JOP is short-lived: although it figures centrally in the 1783 Prolegomena, Kant soon comes to see that the possibility of merely subjectively valid judgments is incompatible with his other commitments, and thus omits any mention of them in the 1787 B-edition of the Critique of Pure Reason. However, a number of factors tell against this narrative. First, the difference between objective and merely subjective validity—which forms the basis of the Prolegomena’s distinction—remains central to Kant’s account of experience and continues to play a key role in the argument that is its centerpiece: namely, the Transcendental Deduction of the categories (e.g., B139–43). Second, Kant continues to refer to JOPs as such even after 1787.3 Third, he explicitly allows for merely subjectively valid judgments in the 1790 Critique of Judgment.4 Finally, he himself addresses the change in his terminology in the B-Deduction and insists that it does not reflect any substantive change in his argument, which he says ‘is already stated correctly’ in the Prolegomena (MF 4:476).

These factors suggest that the distinction between JOPs and JOEs continues to be significant for Kant. Indeed, in this paper, I argue for a novel interpretation of the distinction which reveals that it is coherent, is compatible with Kant’s other commitments, and remains consistent across his works. Even more importantly, I will argue that it is crucial for understanding the role played by the categories in generating objectively valid judgments, and thus, for understanding Kant’s argument in the Deduction itself.

My argument proceeds through the following stages. In §1, I begin by considering and rejecting a standard strategy for making sense of the distinction between JOPs and JOEs. According to this strategy, when Kant claims that JOPs are merely subjectively valid, he has in mind judgments that are about perceived objects but that cannot be said to hold for all subjects: perhaps because they are made without sufficient investigation or because they merely assert how things seem to the subject making the judgment. Any reading of this sort, as I see it, overlooks a criterial feature of JOPs: namely, that such judgments do not have ‘relation to the object’ (4:300). In other words, as I will argue, Kant makes clear that judgments of perception are not about the objects of perception, but rather, about perceptions themselves. In §2, I propose my own interpretation, according to which JOPs should be not be understood as insufficiently justified judgments about perceived objects, but rather, as judgments about perceptions that do not represent what the perceptions are of as an object independent of the perceptions. After developing the details of my account, I show that it can make sense of Kant’s seemingly disparate examples. With the positive account on the table, I am in a position to respond more fully to the objection that Kant no longer allows for JOPs by the time he writes the B-deduction. I do so in §2.1. Next, in §3, I respond to four objections to my proposal, each of which argues on different grounds that the kinds of judgments I identify as JOPs are not in fact merely subjectively valid. Finally, in §4, I discuss the key role Kant’s distinction between JOPs and JOEs plays in the Transcendental Deduction. Kant signals that one of his primary goals in the Deduction is to respond to a Humean skeptic who objects that we are not justified in applying a priori categories in experience. I argue that whereas the standard reading of the distinction between JOPs and JOEs undercuts Kant’s response to Humean skepticism, my account can help illuminate it.

1.

In the Prolegomena, Kant distinguishes between judgments that claim to be ‘only subjectively valid’ [nur subjektiv gültig]—that is, ‘valid’ only for the subject making them—and judgments that claim to be ‘objectively’ and ‘universally valid’ [objektiv und allgemein gültig]—that is, valid ‘at all times … and for everyone’ (4: 298). In the case of the former, he says:

I do not at all require that I should find it so at every time, or that everyone else should find it just as I do; they 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: these I call judgments of perception [Wahrnehmungsurteile] (4:299).

As for ‘judgments of experience [Erfahrungsurteile],’ he continues, ‘the case is completely different’: ‘What experience teaches me under certain circumstances, it must teach me at every time and teach everyone else as well, and its validity is not limited to the subject or its state at that time. Therefore I express all such judgments as objectively valid …’ (4:299).

1.1.

The interpretive strategy I will criticize reads passages like this one as claiming that JOPs are judgments that are about objects, but that, for one reason or another, cannot be said to hold for all subjects.5 I begin by considering Beatrice Longuenesse’s account. On Longuenesse’s reading, all judgments are governed by a norm of universal validity for Kant—all judgments, in other words, aim to be valid at all times and for all subjects (1998: 186). Nevertheless, she says, individual judgments can ‘remain “subjective” to a greater or lesser degree’ (1998: 83) depending on the extent to which they ‘rely uncritically’ on a subject’s particular circumstances.

Longuenesse’s discussion of one of Kant’s examples clarifies what she thinks is required to make fully objective JOEs. In §20 of the Prolegomena, Kant offers as an example of a JOP, ‘When the sun shines on the stone, it becomes warm.’6 He contrasts that judgment with ‘The sun warms the stone,’ which he identifies as a JOE (4:301n*). As Longuenesse reads Kant’s example, the former expresses an observed correlation between the sun shining and the stone growing warm. But merely observing a possibly contingent correlation is not sufficient to warrant the judgment that the observed events are actually connected in a manner that holds at all times and for all subjects: this is why, according to Longuenesse, the judgment in question remains merely subjective.7 Converting such a JOP into a JOE that asserts an objective connection between the observed events, she argues, requires:

confronting the correlations already obtained with many more, while perhaps also using the resources of mathematical construction to anticipate and test further possible empirical correlations. Only after such a method has been systematically applied can a causal connection be asserted: ‘the sun warms the stone’ (Longuenesse 1998: 179).

In Kant’s Transcendental Deduction, Henry Allison presents a similar account of JOPs. According to Allison, JOPs ‘express how things seem to a subject under certain conditions and, as such, make no claim to universalizability’ (2015: 299; see also 2004: 180). On Allison’s view, then, JOPs are a species of what Kant calls ‘provisional judgments’ (see VL 24:861–62): ‘initial assessments’ which ‘precede investigation’ and are ‘subject to revision’ (2015: 305). Since such judgments are only made provisionally, Allison explains, they do not claim to hold at all times and for all perceivers and so are merely subjectively valid. Like Longuenesse, however, Allison believes that an objectively valid JOE only ‘emerges as the result of [an] investigation in which the weight of the reasons for and against the initial assessment has been determined’ (2015: 305).

1.2.

Both Allison and Longuenesse take JOPs to be judgments that are about perceived objects, but do not hold for all subjects, either because they are made uncritically (Longuenesse) or provisionally (Allison). Though an analysis of this sort may indeed seem to be called for by the example discussed above, I will now argue that it conflicts with key features of Kant’s description of JOPs.

For one, Kant says explicitly in the Prolegomena that when a subject makes a JOP, what she expresses is ‘merely a connection of perceptions within [her] mental state, without relation to the object [ohne Beziehung auf den Gegenstand].’8 Now, in his discussions elsewhere of types of representation, as well as of the step-by-step progression [Stufenleiter] between them, Kant defines a perception [Wahrnehmung] as ‘a representation with consciousness.’9 He makes clear that by itself, a perception does not yet amount to cognition of an object. As he puts it in a Reflexion dated to around the same time as the publication of the Prolegomena, consciousness of perceptions alone ‘relates all representation only to our self as modifications of our condition; they are in this case … especially not cognitions of any things and are related to no object’ (R5923, 18:386, 1783–84, my emphasis). As I understand it, Kant’s position is that in being conscious of her perceptions alone, a subject is merely conscious of her own subjective states. Such consciousness lacks ‘relation to an object’ in the sense that the subject does not yet think of what her perceptions are of as an object independent of her perceptions. This is the sense in which consciousness of perceptions does not yet amount to cognition of an object.10 It also explains why Kant says that JOPs do not have ‘relation to the object.’ Since JOPs express what the subject is conscious of merely in being conscious of her perceptions, JOPs are not about the objects represented by her perceptions, but rather, about her perceptions themselves. In contrast, Kant says, a JOE expresses not the ‘relation of a perception to a subject,’ but, rather, ‘a property of an object’ (4:298). In other words, whenever the subject thinks of what her perceptions are of as an object, she makes a JOE.

For ease of expression, I will henceforth refer to a judgment that has ‘relation to an object’ in the above sense as a ‘judgment about objects’, and to a judgment about perceptions that lacks ‘relation to an object’ as a ‘judgment about perceptions’. In short, then, Kant makes clear in the Prolegomena that a JOE is a ‘judgment about objects’, and a JOP is a ‘judgment about perceptions’.11

This key feature disqualifies the kinds of judgments that Allison and Longuenesse count as JOPs. Consider the following judgments, for example:

  • (1) ‘When the sun shines on the stone, it becomes warm,’ understood as expressing an observed correlation between external events.12
  • (2) ‘The sun warms the stone,’ asserted either uncritically or provisionally.

Both (1) and (2) are quite obviously about the objects of a subject’s perceptions, rather than about the perceptions themselves. (1) expresses a perceived correlation between the sun shining and the stone becoming warm; (2), whether it is made uncritically or merely provisionally, is about the perceived effect of the sun on the stone. Both judgments have the ‘relation to objects’ that JOPs are said to lack. As such, they cannot be JOPs by Kant’s lights. To put the point generally: uncritical or provisional judgments about objects are still judgments about objects, and this entails that they are JOEs on Kant’s view, not JOPs.

Second, while Longuenesse and Allison claim that what is required to convert JOPs into JOEs is further empirical investigation and the weighing of reasons, Kant in fact says that what must be added to a JOP is ‘beyond the empirical and in general beyond what is given in sensory intuition.’ The only element missing from a JOP, he is clear, is a priori— ‘special concepts originally generated in the understanding’—in other words, the categories (4:297–98).13 But (1) and (2) above are not missing the application of a category: in particular, the connection that these judgments assert between the sun shining and the stone growing warm has already been subsumed under the category of cause. This is obviously true of (2), which claims (albeit uncritically or provisionally) that the sunshine causes the stone to become warm. But it is also true of (1), since per Kant’s argument in the Second Analogy, even representing an objective temporal succession of events—such as the stone’s growing warm after the sun shines on it—requires application of the category of cause.14 In sum, since both (1) and (2) are fully determined by the categories, it would be incorrect to count them as JOPs, as Allison and Longuenesse do.

Third, it follows from Allison’s and especially from Longuenesse’s view that many of our ordinary judgments should turn out to be JOPs, since it is unlikely that such judgments are made with the systematic scrutiny and investigation that they believe is required for making JOEs. However, in an important Reflexion (which is repeated in the Jäsche Logic), Kant makes clear that, on his view, most of our judgments are, in fact, JOEs: a JOP, he says, ‘is really not possible, except insofar as I express my representation as a perception’ (R3145, 16:678; JL, 9:114, my emphasis). What is not possible, as I understand it, is making a JOP about objects—and, of course, that is what the majority of our judgments are about. JOPs, as Kant says, are only possible insofar as they refer to perceptions.

This lack of fit between Kant’s text and the views under consideration obtains, I believe, because they conflate what may be necessary to fully justify a JOE with what is necessary simply to make such a judgment, whether justifiably or not. While systematic investigation and the weighing of evidence may be necessary to fully justify the claim that one’s judgment holds for all subjects, it is not necessary simply for making such a claim in the first place. For example, a subject quite obviously need not go through the procedure Longuenesse describes—viz., ‘confronting the correlations already obtained with many more, while perhaps also using the resources of mathematical construction’—simply in order to claim that the sun warms a stone.15

For Kant, any judgment that so much as claims universal validity—whether it does so hastily or provisionally—is by that token a JOE. This is reflected in his language throughout the discussion in the Prolegomena: in order to make a JOE, he says, the subject need only ‘regard’ [ansehen], ‘deem’ [halten], ‘express’ [aussprechen], ‘want’ or ‘intend’ [wollen] her judgment to be universally valid (4:298–99). Moreover, he makes clear that a subject expresses the claim that her judgment is universally valid simply in virtue of making a judgment that is about objects rather than about her own perceptions. For, as he says, when we make a judgment about an object, we thereby:

intend [wollen] that the judgment should also be valid at all times for us and for everyone else; for if a judgment agrees with an object, then all judgments of the same object must also agree with one another, and hence the objective validity of a judgment of experience signifies nothing other than its necessary universal validity (4:298).16

When a subject makes a judgment about an object, in other words, she ‘intends’ or claims that the object is the way she represents it to be. This amounts to claiming that her judgment is objectively valid. And in virtue of claiming that the object is the way she represents it to be, she in effect claims that others ought to represent it that way as well—she claims, in other words, that her judgment is universally valid.17 Now, it is perfectly possible—and indeed, common—for subjects to make such claims without sufficient investigation. But insufficiently justified judgments about objects are still judgments about objects and so, are not JOPs by Kant’s lights.

On my view, then, the interpretations I have been considering set too high a bar on JOEs and in so doing, wrongly count too many judgments as JOPs. I have argued that any judgment that makes a claim about objects claims objective validity and is by that token a JOE. Correspondingly, only judgments that are not about the objects of perception but rather about perceptions themselves count as JOPs for Kant. In the next section, I turn my attention to the latter.

2.

I have been arguing that JOPs are correctly understood as judgments about a subject’s perceptions rather than the objects of her perception.18 More specifically, as I will now argue, JOPs report sequences of perceptions that a subject finds herself having as a result of her reproductive imagination.

In the Transcendental Deduction, Kant identifies the imagination as ‘the faculty for representing an object even without its presence in intuition’ (B151). He describes the role played by the reproductive imagination in particular as follows. First, it is only in virtue of the reproductive imagination that a subject can be aware of a ‘series of perceptions,’ for it is what allows her to reproduce ‘a perception, from which the mind has passed on to another’ (A121).19 As she scans the façade of a house, for example, it is the reproductive imagination that allows her to recall and thus to keep in mind the door as she moves on to perceiving the windows. Second, the reproductive imagination is also responsible for calling representations to mind on the basis of psychological associations that a subject has acquired as a result of her past experiences (A121). As she looks at the façade of a house, for example, associations with other houses she has previously experienced might make it the case that she expects this house to also have side and back walls, even though she cannot currently see them (see Strawson 1970: esp. 39–41).

These examples demonstrate why Kant takes the reproductive imagination to ‘be a necessary ingredient of perception itself’ (A120n*). Without it, he says, the subject could only be aware, at any moment, of the individual, isolated perception she was having at that very moment (A120). It is the reproductive imagination that allows her to bring to mind perceptions other than the one she is currently aware of, and thus, to be aware of the kind of sequence of perceptions that go into, for example, the complex representation of a house (A121).

Now, Kant frequently explains the distinction between representations that are combined objectively as opposed to merely subjectively in terms of whether the representations are related by the understanding or merely by the reproductive imagination, respectively. In §18 of the B-deduction, for example, Kant distinguishes between what he calls the ‘objective’ and the ‘subjective’ ‘unity of consciousness.’ Whereas ‘subjective unity’ has ‘merely subjective validity’ and occurs ‘through association of the representations’ (B140), ‘objective unity’ results when representations are ‘united in a concept of the object’ (B139). In §19, Kant again differentiates a relation between representations that counts as an ‘objective unity’ and ‘belong[s] to the understanding’ from a relation that obtains ‘in accordance with the laws of reproductive imagination (which has only subjective validity)’ (B141).

Kant’s point in these passages, as I understand it, is as follows. On his view, a subject’s reproductive imagination can make it the case that a representation RB is called up for her while she has representation RA in mind: either because RB preceded RA and is recalled by the imagination, or because of an association the subject has between As and Bs. When this occurs, the subject passively finds herself thinking the sequence of representations <RA, RB>. Now, such passive awareness by itself does not involve any claim about the correctness or incorrectness of these representations being combined as they are—that is, about whether the objects A and B that the representations are of are actually related. Rather, the subject is simply aware of what Kant calls the ‘subjective unity’ of RA and RB. In contrast, the subject can actively relate RA and RB through her understanding. That is, she can judge that the objects A and B are actually related, and so, in effect, that RA and RB are correctly combined and the sequence <RA, RB> is universally valid. This is what Kant calls awareness of the ‘objective unity’ of RA and RB.

This analysis of §§18 and 19 maps onto my preliminary discussion of Kant’s distinction between JOPs and JOEs in §1 above. In the Prolegomena, as we saw, Kant says that JOPs are merely subjectively valid because they express ‘merely a connection of perceptions within my mental state without relation to the object’ (4:300). What Kant adds in the Critique (and elsewhere20) is that the ‘connection of perceptions’ that are expressed by merely subjectively valid JOPs are due to the reproductive imagination;21 in contrast, the connections expressed by objectively valid JOEs are due to the understanding.22

In what follows, I will argue that we can make good sense of the various examples of JOPs Kant offers in these terms.23 First, however, it will be useful to introduce an example of the kind of circumstance in which it might strike us as more appropriate to make a JOP rather than a JOE.24

Consider the following circumstance. Say that in my childhood, my mother frequently made coffee while the neighbor’s rooster crowed, and so the sound of a rooster crowing has become associated with the smell of coffee for me. Visiting a farm one day, I hear a rooster crow and find myself imagining the smell of coffee. Now let’s assume that I am aware of my association and realize that there is no coffee being made nearby. As a result, I do not judge that the sequence (<RROOSTER, RCOFFEE>) corresponds to an objective connection between the rooster crowing and coffee, or that any other subject hearing the rooster ought to represent coffee as well. Rather, I merely report the subjective order of my representations through the following judgment:

(3’)25 ‘My representation of the rooster crowing is followed by my representation of the smell of coffee.’

Or, equivalently, and far more colloquially:

(3) ‘When I hear(d) the rooster crow, I imagine(d) the smell of coffee.’

Neither (3) nor (3’) assert an objective connection between roosters and coffee; rather, they are about a sequence in my representations that is due to my imaginative associations. As such, I argue, they are genuine JOPs by Kant’s lights.

I believe that we should read Kant’s various examples of JOPs in the same way. In §19, Kant offers the following as an example of a judgment that has ‘only subjective validity’:

(4) ‘When I carry a body, I feel a pressure of weight.’26

He contrasts this with the ‘objectively valid’ judgment:

(5) ‘It, the body, is heavy.’ (B142)

Kant characterizes the sequence of representations expressed by (4) as occurring ‘in accordance with the laws of association’; in contrast, he says, (5) does not claim that the relevant representations are merely ‘found together in perception’ but rather that they are ‘are combined in the object.’ We can flesh out this discussion in a manner similar to my example of the rooster above. Say that because of her previous experiences with bodies, a subject has acquired an association between bodies and weight; this association makes it the case that when she now lifts a body, she finds herself imaginatively anticipating a feeling of weight. As a result, she has the sequence of representations <RBODY, RWEIGHT>, which she reports through (4) ‘When I carry a body, I feel a pressure of weight.’ Here, just as in the example of the rooster above, (4) should be understood as describing a sequence of perceptions had by the subject: as reporting, that is, that her perceptions of carrying a body are followed by her anticipating the feeling of weight. In other words, it should be read as equivalent to:

(4’) ‘My representation of carrying a body is followed my representation of weight.’

Just like (3) and (3’) above, neither (4) nor (4’) claim that the sequence of representations had by the subject corresponds to an objective connection between bodies and weight.27 Rather, they express ‘only a relation of two sensations to the same subject … and this only in [her] present state of perception’ (4: 299). As such, they are merely subjectively valid. In order to make a judgment about the object her perceptions are of, it is not sufficient that the subject merely report a sequence of representations that she finds herself having as a result of her reproductive imagination. Rather, she must actively relate those representations through an act of the understanding in accordance with a category. She must judge, in other words, that RBODY and RWEIGHT are correctly combined because bodies and weight are objectively connected. What is required in order to express such a judgment—as we have seen Kant argue in the Prolegomena—is the addition of a category.28 In this case, the subject subsumes the represented relation under the category of substance-inherence, thereby judging that bodies are related to weight as subject to property. In other words, as Kant says, she judges that (5) ‘Bodies are heavy.’

Of course, it is hard to hear (4’) as a judgment one might naturally make upon carrying a body and feeling its weight. Note, however, that this is in contrast with my example of the rooster above, in which it was the JOP (3) ‘When I hear a rooster crow, I imagine the smell of coffee’ that seemed more natural than any JOE that would assert an objective connection between roosters and coffee (such as the false judgment that a rooster’s crowing causes the smell of coffee). In that case, since there is no objective connection between roosters and the smell of coffee, we know that the connection between the subject’s representations is a merely subjective one. In contrast, since there obviously does exist an objective connection between bodies and weight, the corresponding JOE in (5) strikes us as far more natural. Despite this, however, it is genuinely possible to make the relatively unnatural judgment in each case: that is, to make a judgment about the representations of bodies and weight in the former case, and a judgment that asserts an objective connection between roosters and coffee in the latter. On my view, it is these two possibilities—and how one moves from the former to the latter by the addition of a category—that Kant has in mind in each of his examples.

Correspondingly, I believe we should analyze the example discussed above from the Prolegomena in the same way:

(1) ‘When the sun shines on the stone, it becomes warm.’

Though (1) seems to report a correlation observed between perceived events (indeed, this is precisely why Longuenesse and others29 read it as they do), I think Kant simply fails to make explicit in his examples the very feature he stresses in his description of JOPs in the Prolegomena—namely, that such judgments are about perceptions rather than objects. Interestingly, in the passage from the Jäsche Logic I cited above, Kant gives another example of a JOP that involves warm stones, in which he is clearer. ‘In touching the stone I sense warmth,’ he says, ‘is a judgment of perception but on the other hand, The stone is warm, is a judgment of experience.’30 This indicates, I think, how we are meant to read Kant’s example in the Prolegomena. Once again, it should be read as expressing a sequence of representations (<RSUN SHINING, RSTONE BECOMING WARM>):

(1’) ‘My representation of the sun shining on the stone is followed by my representation of the stone becoming warm.’31

Reading (1’) for (1) differentiates it from a judgment in which the represented relation has already been subsumed under a category. The latter is the case for judgment (2) ‘The sun warms the stone,’ and indeed, as I argued above, for (1) itself if it is understood as referring to an objective succession of external events, since both presuppose the application of the category of cause.

Now, one might worry here that I am not entitled to reinterpret (1) as a judgment about perceptions, since it clearly seems to refer to external events. If Kant really meant to refer to perceptions of the sun and the stone in his example, in other words, why does he not say so? To respond to this concern, I want to briefly discuss another of Kant’s examples from the Prolegomena:

(6) ‘Air is elastic.’

Interestingly, Kant says that (6) can express either a JOP or a JOE, depending on whether it refers merely to a relation between the subject’s perceptions or to objects (4: 299–300). In the former case, as I understand it, (6) should be understood as referring to a subject’s perceptions of what amounts to the air’s elasticity: for example, to the fact that her perception of heating air (say, the air in a closed container) is followed by her perception of it expanding (say, when the container explodes).32 In contrast, for (6) to count as a JOE, Kant says, the perception of air has to have been ‘subsumed under a concept of the understanding.’ ‘For example,’ he continues, one might judge that ‘the air belongs under the concept of cause, which determines the judgment about the air as hypothetical with respect to expansion.’ In other words, one might judge that heating air causes it to expand. Thus, in Kant’s example, the judgment (6) ‘Air is elastic’ could be understood as equivalent to:

(6’) ‘My representation of heating air is followed by my representation of air expanding.’ (JOP)

or

(7) ‘Heating air causes it to expand.’ (JOE)33

This shows that, in the Prolegomena, Kant clearly does allow a judgment like (6) that seemingly refers to objects to be disambiguated such that it is understood to refer instead to perceptions. I argue that this licenses us to apply the same procedure when interpreting his example of the sun shining on the stone, as I do above.34

Of course, Kant does not help his case by making his examples so open to misinterpretation. In the Prolegomena, as I mentioned, I believe he is still struggling to find the right way to articulate examples of the kind of judgment he has in mind so that it is clear that they refer to perceptions rather than objects (as his own definitions there require). As we have seen, he is already (somewhat) clearer in the example he provides in the B-deduction, and declares in so many words in the Jäsche Logic that a JOP is only possible if it refers to perceptions. The example offered in the Jäsche Logic does finally make fully explicit that a JOP is meant to be about the subject’s perceptions: ‘I, who perceive a tower, perceive in it the red color,’ as distinguished from the JOE, ‘It [the tower] is red’ (9: 114).

On my view, then, we can give a consistent account of JOPs that makes sense of Kant’s examples while avoiding the criticisms I raised for the standard interpretive strategy in §1. For as I have argued we should understand them, Kant’s examples of JOPs:

  • (i) Are about perceptions rather than objects, and so, do not violate Kant’s claim that JOPs lack ‘relation to the object.’
  • (ii) Only require the addition of a category to be converted into JOEs.
  • (iii) Are not common, since most of our judgments are not about sequences in our own perceptions, but rather about objects. This is in line with Kant’s prediction in the Jäsche Logic that JOPs ‘are really not possible’ unless they refer to perceptions.

Let me sum up the account I am proposing here by returning to the key terms under discussion. I have been arguing that a sequence of perceptions is objectively valid if and only if what the perceptions represent is thought to be combined in the objects the perceptions are of. Correspondingly, a JOE is a judgment that claims that what is represented by a certain sequence of perceptions is combined in the objects the perceptions are of. In order to do this, the JOE must represent what the perceptions are of as an object, which, as we have seen amounts to subsuming it under the categories.

In contrast, a sequence of perceptions is merely subjectively valid if and only if what the perceptions represent is not thought to be combined in the objects the perceptions are of. Correspondingly, a JOP is a judgment that claims merely that a certain sequence of perceptions is had by a subject and does not claim that what is represented by the perceptions is combined in the objects the perceptions are of.35

2.1.

In §2, I argued that Kant’s distinction between JOPs and JOEs can be fleshed out by appealing to his discussion of the difference between merely subjective and objective validity in the first Critique. Let me pause here to consider the important question of whether Kant continues to allow for the possibility of JOPs by the time he writes the B-deduction. If he does not, then not only am I not entitled to map the Prolegomena’s distinction between JOPs and JOEs onto the B-Deduction’s distinction between subjective and objective validity, but it would also be textually inaccurate and misleading to suggest—as I will go on to do—that the former distinction continues to play an important role in the argument of the B-deduction. The worry is as follows: in §19 of the B-deduction, Kant explicitly defines judgment as a way of combining representations so that they are objectively valid (B141). And this seems to entail that he has changed his mind and no longer allows for the merely subjectively valid JOPs discussed in the Prolegomena.36

I think this is the wrong conclusion to reach, however. Though Kant does indeed present a new definition of judgment in §19, there are a number of reasons not to read this as signaling that there has been a substantive change in his position between the publication of the Prolegomena in 1783 and the B-Deduction in 1787.37 For one, just a few lines after stating his new definition in §19, he offers an example of something a subject ‘could … say’ that, as I have already discussed, shares all the features of the Prolegomena’s JOPs: namely, (4) ‘When I carry a body, I feel a pressure of weight.’ Just like in the Prolegomena, as we have seen, Kant characterizes (4) as having ‘only subjective validity,’ and proceeds to contrast it with the ‘objectively valid’ judgment (5) ‘It, the body is heavy,’ arguing once again that it is the addition of a category that distinguishes the latter judgment from the former (B142). This indicates, I believe, that the Prolegomena’s distinction continues to be central to Kant’s purposes in the B-deduction, even if he has decided there to reserve the term ‘judgment’ for the Prolegomena’s JOEs.

Further evidence that Kant continues to allow that JOPs are possible is also to be found in the fact that he returns to referring to them by that name in Reflexionen dated to the period after the publication of the B-edition.38 Additionally, in the Critique of Judgment—published in 1790—Kant explicitly allows for judgments of the agreeable, which he describes as merely subjectively valid judgments, thus demonstrating that he continues to allow for a more capacious use of the term ‘judgment’ that applies to merely subjectively valid judgments (see, e.g., CJ 5:212; 5:214, 5:217).

Keeping this in mind, I suggest that the apparent conflict between the B-deduction and Kant’s other writings can be resolved as follows. We can grant that within the context of the B-deduction, Kant opts to use the term ‘judgment’ in a narrow sense so that it applies only to cases in which representations are actively combined by the understanding.39 Indeed, Kant signals as much in a note added to the preface of the Metaphysical Foundations of Natural Science (published in 1786, between the Prolegomena and the B-edition). In the note, Kant explains that ‘as [he] now understand[s] it,’ the Deduction’s justification of the categories ‘can almost be accomplished through a single inference from the precisely determined definition of a judgment in general (an action through which given representations first become cognitions of an object)’ (MF 4:475–76, my emphasis). Understood in this way, of course, the term ‘judgment’ becomes synonymous with the Prolegomena’s ‘judgments of experience,’ as I have argued we should understand the latter term.

Using the term in its restricted sense lends rhetorical strength to Kant’s argument in the B-Deduction, since it allows him to claim that the Categories are necessary conditions on judgment in general.40 This change does not mean, however, that Kant no longer allows that representations can be merely passively combined by the reproductive imagination, and indeed, that we can be conscious of and report these passive and merely subjective combinations: as we have seen, he describes the contrast between objective and subjective unity in these very terms in §§18 and 19 of the B-deduction itself. As Kant is quick to point out in the note in the Metaphysical Foundations, his more ‘precisely determined definition’ only improves the ‘manner of presentation,’ not ‘the ground of explanation,’ which, he insists ‘is already stated correctly’ in his previous discussions.41

Now, we could easily abide by Kant’s terminology in the B-deduction by expressing the distinction we have been discussing as holding, for example, between the objectively valid judgments that are a result of the activity of the understanding and merely subjectively valid reports about representations that are passively combined due to the reproductive imagination. This would be a merely semantic choice, however, and would not alter the content or relevance of the distinction at hand.42 Since the Prolegomena’s terminology is both better known and more widely used by Kant, I have chosen to express the distinction as holding between JOEs and JOPs in this paper. And, more substantively, I have argued that it is the very same distinction that continues to play an essential role in both the Prolegomena and the Critique and so, that Kant’s discussion in the former can fruitfully be brought to bear on his discussion in the latter, and vice versa.

3.

In this section, I respond to four objections, each of which argues in a different way that the kinds of judgments about perceptions that I have been treating as JOPs are not in fact merely subjectively valid.

3.1.

Judgments about sequences in a subject’s perceptions have a truth value. They are true just in case the subject really does have the sequence of perceptions they report her to have. According to the first objection I will consider, the fact that judgments about perceptions have a truth value entails that they are not merely subjectively valid. In his well-known paper, “Did the Sage of Königsberg Have No Dreams,” for example, Beck argues for this point as follows. Judgments ‘about associations of ideas,’ he says, ‘if correct … are correct for everyone.’ Such a judgment, he continues, ‘is valid (true) for me in the sense that it is claimed to be true of me. Yet this judgment is objectively valid, for it … is a judgment of experience about which others can have evidence and on which they must agree if it is true.’43 To put the point succinctly: if a judgment about my perceptions is true, then it is true for all subjects. But this just means that it is not merely subjectively valid, but, rather, universally and objectively valid.

Though the objection is an initially compelling one, I think it goes wrong in equating the objective validity of a judgment with its having a truth value.44 Though every judgment that has objective validity has a truth value, not every judgment that has a truth value is objectively valid. I have argued that a judgment claims objective validity if and only if it claims that what is represented by a certain sequence of perceptions is combined in the objects the perceptions are of. But a JOP does not do this. Rather, it merely reports a sequence of perceptions the subject finds herself with, without claiming that what is represented by the perceptions is combined in the objects the perceptions are of. It is for this reason, on my view, that JOPs remain merely subjectively valid. Now, as I said above, a JOP is true if the subject really had the sequence of perceptions it reports her as having. And others can and typically should accept the truth of such a judgment on the basis of the subject’s testimony about her own perceptual states. Thus, a true JOP is indeed true for all subjects.45 But this is compatible with its lacking objective validity, since it makes no claim about the objects the perceptions are of.

In sum, both objectively valid and merely subjectively valid judgments have a truth value. An objectively valid JOE is true if the objects represented by a sequence of perceptions really are the way it claims they are, and false if they are not. A subjectively valid JOP is true if the sequence of perceptions really are the way it claims they are, and false if they are not.46

3.2.

The next objection I will consider points out that all the JOPs I have discussed already involve the application of objective concepts and indeed, the categories, since the representations they refer to are already determined to be representations of extensive magnitudes (such as weight), intensive magnitudes (such as warmth or the smell of coffee), substances (such as the sun or bodies) and events (such as the sun shining on a stone or a rooster crowing). How can I maintain, then, that these judgments are not about objects and remain merely subjectively valid?47

We can begin to respond to the objection by pointing out that Kant does not need his examples to consist in perceptions that have not been subsumed under any categories whatsoever. Rather, it is sufficient for the purposes of his argument to focus in on an element in a sequence of otherwise conceptually—and indeed, categorially—determined representations, where that element has itself not been (fully) determined by the categories. If he can demonstrate that it is only when the element in question is determined by a category that it secures ‘relation to the object,’ then his example will be adequate to its function. In (1), for example, the subject’s representation of the sun shining on the stone is presumably already determined as a representation of an event, and the representation of the stone’s warmth is determined as a representation of an intensive magnitude. What remains undetermined, however, is the relation between these two representations. Before that relation is subsumed under the categories, as we have seen, all the subject can report is that former representation is succeeded by the latter—that is, she can only report the temporal succession of these representations that obtains as a result of her reproductive imagination. Once she actively subsumes this relation under the category of cause-and-effect, however, she can now claim that the event of the sun shining is objectively connected to the stone’s warmth as cause to effect. Kant’s example shows how the addition of a category converts the awareness of a merely temporal relation between two individual (but otherwise conceptually determined) representations into a judgment about an actual connection between the objects those representations are of. This is sufficient for his purposes, since it serves as an instance that illustrates the general role the categories are meant to play in relating representations to objects and thus, in building up a subject’s complex objective representation of the world.

A complete response to the objection under consideration, however, will require a further clarification. Kant indicates in the Critique that in order for a temporal sequence of perceptions to even be consciously apprehended and represented as a temporal sequence, it must already at least be determined by the categories of quantity, since these are required to represent time itself as a unified extensive magnitude (B161). As such, it is not quite accurate to say even that the relation between perceptions reported by a JOP is entirely categorially undetermined. Rather, the relation remains undetermined in particular by what Kant calls the relational categories—substance-inherence, cause-effect, and community—the application of which, as we have seen, is required in order to convert a subject’s awareness of a mere temporal sequence in her representations into a judgment that those representations are correctly combined, or, equivalently, that the objects they represent are actually connected.48

In the “System of Principles”, Kant distinguishes between two kinds of combination. The first, which he calls ‘composition’ [Zusammensetzung or compositio], consists in a synthesis that is governed merely by the mathematical categories of ‘what does not necessarily belong to each other.’ The second, which he calls ‘connection’ [Verknüpfung or 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’ (B201n*)—in other words, a combination in accordance with one of the relational categories. On the interpretation I have defended here, JOPs report merely that one representation follows another in time, without determining whether these representations are correctly combined. Borrowing Kant’s language from the Analytic, then, the sequence of representations reported by such a judgment is a mere ‘composition’ of two representations that are not claimed to ‘belong to each other’. In contrast, JOEs express what Kant calls a ‘connection’ between perceptions: by subsuming the relation between the representations under the relational categories, such judgments involve the claim that the representations ‘belong to one another’—that is, that they are correctly combined because they correspond to a connection in the objects represented.49

It follows, then, that it is specifically the addition of a relational category that allows for a JOP to be ‘converted’ into a JOE.50 This is certainly the case with Kant’s own examples, which, as we have seen, involve adding the category of substance-inherence or the category of cause-and-effect. It also maps nicely onto his discussion in the Analogies of Experience, which specifically treat the relational categories. The general principle of the Analogies, Kant says, is that ‘Experience is possible only through the representation of a necessary connection of perceptions’ (A176/B218, my emphasis). Subsuming perceptions that are merely ‘juxtaposed’ (zusammengestellt) in time under relational categories involves claiming that these perceptions are ‘necessarily’—that is, objectively—combined (A176/B219). This is how, on Kant’s view, perception gives rise to experience. And correspondingly, it is the addition of a relational category that allows the conversion of a judgment of perception into a judgment of experience.

I should clarify here that I am not claiming that a JOP cannot employ any relational categories at all.51 Rather, as I have already discussed, it is sufficient that some relation between the objects represented remains undetermined by a relational category. Thus, although in Kant’s example, the representation of the sun shining is presumably already thought to be a representation of an event and, as such, has already been subsumed under the category of cause, it is its connection with the representation of the stone’s warmth that has not been determined by a relational category.52

3.3.

Another objection that denies that judgments that are about sequences in a subject’s perceptions are merely subjectively valid does so on the grounds of Kant’s argument in the Second Analogy. There, as I have already discussed, Kant argues that the application of the category of cause-and-effect is necessary for representing an objective sequence of events. The objection claims that this point should apply to any sequence of events, including sequences in one’s representations. This would then entail that applying the category of cause-and-effect is a necessary condition even on representing a sequence in one’s perceptions, and it should follow from this that judgments about such sequences are already determined by a relational category and, thus, objectively valid.

What this objection overlooks, however, is that Kant’s concern in the Second Analogy is precisely to distinguish a subjective sequence in one’s representations from an objective sequence in events, and to argue that application of the category of cause-and-effect is a necessary condition only on representing the latter. Indeed, Kant argues that we must apply the category of cause-and-effect in order to judge whether or not a subjective sequence of representations corresponds to an objective sequence of events. As for the subjective sequence, Kant says:

I am therefore only conscious that my imagination places one state before and the other after, not that the one state precedes the other in the object; or, in other words, through the mere perception the objective relation of the appearances that are succeeding one another remains undetermined (B233–34).

In order to judge whether the subjective sequence of perceptions I have due to my reproductive imagination corresponds to an objective sequence of events, Kant continues:

the relation between the two states must be thought in such a way that it is thereby necessarily determined which of them must be placed before and which after rather than vice versa. The concept, however, that carries a necessity … with it can only be a pure concept of understanding, which does not lie in the perception, and that is here the concept of the relation of cause and effect … (B234)

In other words, Kant does not argue in the Second Analogy that application of the category of cause-and-effect is a necessary condition on representing even a subjective sequence in one’s representations. Rather, and in line with my discussion so far, his point is that applying the category is necessary in order to make a judgment that refers such a sequence of representations to objects—that is, that claims of the events the representations are of that one actually succeeds the other in time.

3.4.

The last objection I will consider claims that, on Kant’s view, a subject’s consciousness of her own perceptions in time amounts to inner experience. Given this, judgments that express temporal sequences of perceptions—which I have here identified as JOPs—should actually be thought of as judgments of (inner) experience. As such, they cannot be merely subjectively valid and, like all JOEs, require application of relational categories (in particular, the category of cause).

My response to this objection is to deny that consciousness of the temporal sequence of one’s perceptions alone amounts to inner experience. As we saw in §3.3, Kant indicates in the Second Analogy that a subject can be aware of the order in which her imagination calls up perceptions without yet subsuming them under the category of cause (B233–34). It is this immediate awareness of the subjective temporal order of her perceptions that I take JOPs to express. In order for this awareness to amount to any kind of experience of an objective order of events, application of the category of cause is necessary. The subject has outer experience, as we have seen, when she takes her sequence of perceptions to correspond to a causally determined sequence of outer events. She has inner experience, on my view, when she takes the sequence of her representations to itself be causally determined, that is, when she applies the category of cause to her own representations.53 In doing so, she takes the temporal order of her representations to be determined by (psychological) causal laws and goes beyond the immediate awareness of that order which, I have argued, is expressed by a JOP.54

Kant explicitly distinguishes between the mere awareness of one’s perceptions and inner experience along these very lines in the Anthropology. ‘Perception,’ he says, ‘could be called merely appearance of inner sense. However, in order for it to become inner experience the law must be known which determines the form of this connection …’ (7: 144n28, my emphases). ‘[I]t is … necessary,’ he repeats, ‘to begin with observed appearances in oneself, and then to progress above all to the assertion of certain propositions that concern human nature; that is, to inner experience’ (7: 143, my emphases). In these passages, Kant clearly distinguishes between mere awareness of perceptions on the one hand, and inner experience on the other. The latter, he says, involves asserting ‘certain propositions that concern human nature.’ Though a full discussion of this is beyond the scope of this paper, I take the ‘propositions’ concerning human nature that Kant has in mind here to be the natural, psychological laws that humans are subject to. These laws—which Kant identifies as laws of association (A100; 7: 176)—govern the reproduction of representations through the imagination. Kant’s point in these passages, as I see it, then, is that inner experience requires taking one’s representations to be causally determined by laws that govern human psychology.

Most importantly for my purposes in this paper, Kant makes clear that a subject can have mere awareness of her perceptions in time, where this awareness does not yet amount to causally determined inner experience. Judgments that merely express the former awareness are not JOEs, on my view, and do not presuppose that my representations have been subsumed under the category of cause.55 Rather, as I have argued, they are merely subjectively valid JOPs.

4.

In this final section, I argue that a key advantage of my reading of the distinction between objective and merely subjective validity is that it can explain why this distinction appears so centrally in Kant’s deduction of the categories, both in the Prolegomena and in the B-edition of the Critique of Pure Reason. In contrast, as I will explain, the reading I criticized in §1 obscures Kant’s argumentative strategy in the Deduction.

In the Prolegomena, Kant claims that he takes his arguments to provide a ‘complete solution of the Humean problem’ (4: 313). The ‘Humean problem,’ as Kant understands it, is a version of the ‘skeptical doubts’ concerning the idea of a necessary connection that Hume presents in the Enquiry Concerning Human Understanding, generalized so as to apply to other concepts and principles that Kant believes are not derived from experience.56 Now, there are a number of difficult questions concerning Kant’s reading of Hume, as well as the various dimensions along which he might be thought to have offered a response to him.57 For the purposes of this paper, I hope to sidestep most of these questions. First, I will only be concerned with how Kant himself understands the so-called ‘Humean problem,’ setting aside the question of whether he is right to read Hume as he does.58 Second, I will only focus on one aspect of the problem: the question of how we can be rationally entitled to apply concepts that have not been derived from experience to the objects of experience. Since Kant himself describes the main goal of the Transcendental Deduction as consisting in answering how we are entitled to apply the a priori categories to the objects of experience (A85/B117), I take it to be relatively uncontroversial to assume that he aims to respond to this question. Third, we should note that Kant takes the relational categories to be especially central to the ‘Humean problem.’ As he argues in §27 of the Prolegomena, Hume’s doubts about the concept of necessary connection are automatically applicable not only to the concept of a causal connection but also to concepts of the kinds of connection covered by the other two relational categories: namely, a connection between a substance and its accidents, as well as between mutually dependent objects in a community (4: 310). For the purposes of this paper, then, I will focus in on a ‘Humean skeptic’ who denies in particular that we are rationally entitled to apply the relational categories to the objects of experience, and consider that component of Kant’s argument in the Deduction that could be said to respond to such a skeptic by invoking the distinction between objective and merely subjective validity.

We can express Kant’s argumentative strategy against such a skeptic in the terms of the Prolegomena as follows: we are rationally entitled to apply the relational categories because they are necessary conditions on making JOEs. In other words, the application of a relational category is necessary in order to transform a JOP into a JOE, and so, if we were unable to apply the relational categories, we would be limited to making mere JOPs (See 4: 311). But if the relational categories are necessary conditions on making JOEs, the Humean skeptic cannot deny that we are entitled to use them in such judgments.

Now, a full defense of Kant’s transcendental argument is beyond the scope of this paper. Instead, I want to focus in on the following question: why should Kant think his argument has any force against the Humean skeptic? That is, does Kant have reason to think that such a skeptic will grant that it must be possible to make JOEs, such that his argument, if sound, could move the skeptic to accept the legitimacy of the relational categories? If we understand the distinction between JOPs and JOEs in accordance with the standard reading that I criticized in §1, it is hard to see why this should be the case. For recall that, on that reading, a JOP is a judgment about the objects of perception that cannot be said to hold for all subjects because it is insufficiently justified; a JOE, in contrast, is a judgment that has undergone the systematic investigation necessary to justify the claim that it holds for all subjects. Plugging this reading of the distinction into Kant’s transcendental argument, it amounts to the claim that the relational categories are necessary conditions on making sufficiently justified judgments about perceived objects: if we could not apply these categories, on this reading, we would be limited to making insufficiently justified judgments about perceived objects. Even if this were true, however, why should it move the Humean skeptic to accept the legitimacy of the relational categories? For he is by no means committed to the possibility of our making sufficiently justified judgments about perceived objects: indeed, since he denies that we are entitled to apply relational categories like substance or cause, he has already concluded that any judgments that employ such concepts are insufficiently justified.59

On my interpretation, in contrast, any judgment that so much as makes a claim about the objects of perception claims objective validity, and is, therefore, a JOE; JOPs, in contrast, are judgments about the subject’s perceptions rather than about the objects of perception. This allows us to translate Kant’s argument as follows: the application of the relational categories is a necessary condition on making judgments that so much as make a claim about objects; not applying these categories, then, is tantamount to being limited to making judgments about one’s own perceptions rather than about the objects of perception. On this translation of it, I think we can see why Kant could expect his transcendental argument to have force against the Humean skeptic. For, if sound, his argument establishes that if we could not apply the relational categories, then we could not so much as represent the objects of our perception as being a certain way, whether justifiably or not. And Kant could quite plausibly claim that the Humean skeptic he is responding to does not intend his skepticism to go so far as to deny that we are capable of so much as representing the objects of our perception as being a certain way.60 The ‘Humean doubt’ that Kant takes himself to be responding to has to do in particular with whether we are entitled to apply concepts not derived from experience to the objects of experience. It does not call into question whether we are entitled to apply concepts that are derived from experience to the objects of experience, nor, importantly, whether we can so much as take our experience to represent objects in the first place.61 Kant’s claim against this skeptic is that his discussion of the distinction between JOPs and JOEs has shown that even this minimal achievement—relation to the object—requires the application of the a priori relational categories. In a nutshell, since the categories are a necessary condition of making JOEs that so much as represent objects, Kant concludes, the Humean skeptic cannot deny our entitlement to use them in such judgments.62

The same transcendental argument recurs centrally in §19 of the B-Deduction in the Critique of Pure Reason. Though Kant does not there put the argument in terms of the distinction between JOPs and JOEs, as we have seen, he does argue that the addition of a category is necessary to convert a merely subjectively valid report about one’s representations into an objectively valid judgment about the objects represented. Only by relating our representations ‘in accordance with principles of the objective determination of all representations insofar as cognition can come from them’ (i.e., the categories), Kant argues, ‘does there arise’:

a relation that is objectively valid, and that is sufficiently distinguished from the relation of these same representations in which there would be only subjective validity, e.g., in accordance with laws of association. In accordance with the latter 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 … (B142)

In this section, too, I believe, Kant’s primary target is the kind of Humean skeptic I described above. Kant’s argument against the Humean skeptic, again, is that if we were not able to apply the relational categories, we would only ever be able to describe the ‘condition of the subject’—that is, the combination of representations the subject finds herself with—rather than so much as represent (justifiably or unjustifiably) how things are ‘combined in the object’.

In sum, I believe that it is only when the distinction between JOPs and JOEs is understood as I do here that we can correctly understand Kant’s argument against the Humean skeptic. When interpreters read Kant as claiming in §§18–19 of the B-deduction that imaginative associations are insufficient grounds for judgments about objects, they mistake what is meant to be a response to such a skeptic for a statement of agreement with him.63

Now, of course, Kant does agree with Hume on some crucial points. He agrees, for example, that the concept of cause is not derived from our experience of events. And he agrees that our reproductive imagination is governed by laws of association that make it the case that we habitually connect representations that frequently go together. His point against the Humean skeptic, however, is that with these resources alone, our mental lives would be far more impoverished than he anticipates. For even simply to take our imaginative associations as grounds for judging that objects are a certain way requires that our judgments have the categorial form that is necessary in order for them to be about objects. Since our mental lives are not so impoverished, Kant argues, we are committed to the possibility of doing just that.

Notes

1Norman Kemp Smith, for example, declares that ‘the illegitimacy and the thoroughly misleading character of this distinction hardly require to be pointed out’ (1918: 288). Paul Guyer (1987: 101) insists that the notion of a judgment of perception ‘cannot be reconciled’ with Kant’s argument in the B-deduction of the Critique of Pure Reason, and argues that its seeming appearance in that section betrays Kant’s ‘continuing confusion’ about his own argumentative strategy. Kotzin and Baumgärtner argue that Kant’s remarks about JOPs are ‘un-Critical’ and ‘fail to conform to some central teachings of the Critique of Pure Reason’ (1990: 402). 

2See Kemp Smith (1918: 289), Cassirer (1923: 245–46), Beck (1978: 50), Guyer (1987: 101–2), Kotzin and Baumgärtner (1990: 407), Pollok (2008: 335–36), Sassen (2008: 271–72, 281). 

3R3145–46, 16:678–79, between 1790–1804; DWL 24:767, transcript of lectures given in the 1790s; JL 9:114, published in 1800. 

4Namely, judgments of the agreeable. See CJ 5:212; 5:214, 5:217. 

5In addition to Longuenesse and Allison, whose views I discuss below, versions of this strategy can be found in Friedman (2012), Wolff (2012), and Beizaei (2017). 

6I have modified the translation of ‘wenn’ from ‘if’ to ‘when’ here. Whereas the German ‘wenn’ is ambiguous between ‘if’ and ‘when’ in English, I think the latter better conveys the feature of JOPs that Kant emphasizes: namely, that they merely report what is the case in ‘my present state of perception’ (4:299, my emphasis). Cf. Longuenesse (1998: 178n26). 

7Longuenesse (1998: 177–79). The view that JOPs express empirically observed correlations of events and/or inductive generalizations based on such observations is shared by Friedman (2012: 247, 250), Wolff (2012: 132–35), and Beizaei (2017: 363–64). 

84:300, my emphasis. In the Critique of Pure Reason, Kant says the object of representations should be understood as that which makes ‘one way of combining the manifold necessary’ (A191/B236). What this means, as I understand it, is that representations come to have ‘relation to an object’ insofar as the subject takes a certain combination of them to be necessary (i.e., correct) in virtue of corresponding to the way the object is. See also A104–5. 

9A320/B376–77. See also JL 9:64–65. ‘Perception’ should be understood broadly here so as to include representations recalled by the imagination in the absence of the object. I explain this further in §2. 

10For a detailed and very helpful account of the difference between perception and cognition in Kant that I take my discussion here to align with, see Tolley (2017). See also Watkins and Willaschek (2017: 84–87, 101–2). I should note that I take the claim that perceptions lack ‘relation to an object’ in this sense to be compatible with their being related to objects in less cognitively demanding senses: for example, as being causally related to objects or being that in virtue of which particulars are presented. See Allais (2009: 403–5), and, for a discussion of some other less cognitively demanding senses of objectivity, see McLear (2020). In §3.1, I will in effect argue that JOPs do have what McLear calls ‘representational objectivity,’ since they can be assessed for truth. But this does not mean that they have the kind of ‘relation to an object’ that I take to be at issue here, since they are about perceptions, rather than about the objects represented by those perceptions. In contrast, it is clear that judgments like (1) and (2) above—which Allison and Longuenesse count as JOPs—are ‘related to objects’ in precisely the cognitively demanding sense of being about the objects represented by perceptions. 

11It might be objected here that JOPs must take perceptions themselves—or alternatively, the subject who has them—as objects. As I will discuss, however, to represent something ‘as an object’ in the relevant sense for Kant is to represent it as fully determined in accordance with the categories (especially the relational categories). And, as I will argue below, JOPs do not represent perceptions—or the subject who has them—as an object in this sense. In order to do so, the subject must think of perceptions themselves as governed by causal laws. When she does this, however, she makes a JOE rather than a JOP. The latter type of JOE complicates my shorthand above, since it amounts to a ‘judgment about objects’ that is about perceptions. Since this complication does not bear on the bulk of my discussion, however, I retain the shorthand and clarify when necessary. None of Kant’s examples consist in judgments of this sort, nor do the commentators I criticize discuss them. I say more about such judgments in §3.4 and notes 54 and 55. 

12Kant indeed offers this very judgment as an example of a JOP in the Prolegomena, but I will argue below that, as such, it should be understood as referring to the subject’s perceptions of the sun shining and the stone becoming warm, rather than to the corresponding external events. As I will explain, I believe Kant struggles to identify the correct way to formulate his examples so that they exhibit the features he ascribes to JOPs in the Prolegomena. His later examples go some distance towards rectifying this. 

13I discuss Kant’s claim at length in §3.2 below. There, I argue that converting JOPs to JOEs specifically requires the addition of one of the relational categories. 

14A189/B232–A211/B257. Interpreting Kant as allowing that we can represent temporal successions of events and express inductive generalizations on this basis without applying the category of cause conflicts with the Second Analogy, and—as I will discuss in §4—undercuts Kant’s argument against Hume. This point of criticism also applies to the accounts of JOPs defended in Friedman (2012), Wolff (2012), and Beizaei (2017). 

15Friedman recognizes that this is also a possible problem for his view, according to which only ‘well-established laws of empirical natural science’ count as genuinely objective JOEs (2012: 261). As Friedman worries, this appears to entail that there was ‘no genuinely objective experience before Newton established the law of universal gravitation in the Principia …’ (2012: 262–63). He responds to this worry by clarifying his account, such that objectively valid judgments need not be explicitly incorporated into scientific theory, but need only be ‘capable of such incorporation’ (2012: 263). Of course, the merely inductive generalizations that Friedman takes to be expressed by JOPs are typically capable of incorporation into scientific theory; as a result, Friedman is forced to conclude that JOPs are themselves objectively valid (2012: 264), and thus to deny the very feature that Kant takes to be criterial of them. 

16The ‘necessity’ here should be understood as normative necessity, since Kant’s claim applies equally to contingent judgments as he makes clear at 4:305n*. See Allison (2015: 294–95). 

17This is also true for the provisional judgments about objects described by Allison. When a subject makes a provisional judgment, she claims that it is the correct judgment to make given her evidence, even though she recognizes the insufficiency of her evidence. This amounts to claiming that any subject judging on the basis of the same evidence ought to make the same provisional judgment. 

18On this point, my account is closest to the extended account of judgments of perception developed in Prauss (1971). However, I diverge from Prauss on a number of key points. Prauss claims that Kant fails to identify a formulation of JOPs on which they genuinely refer to perceptions rather than objects (1971: 175). According to Prauss, this is the case for all of Kant’s examples, such as ‘When the sun shines on the stone, it becomes warm,’ or even ‘Sugar is sweet.’ The former, he argues, must be reading as expressing an objective succession of events, and the latter as ascribing a certain (secondary) property to an object (i.e., sugar) (1971: 180, 185–88). In place of Kant’s formulations, Prauss argues that it is ‘It seems …’-judgments (‘Es scheint …’-Urteile) that genuinely satisfy the conditions that must be met by JOPs since, according to him, such judgments express the subject’s perceptions and make no claim about the objects of those perceptions (1971: 224–53). Thus, according to Prauss, the judgment ‘It seems that the sun warms the stone,’ for example, counts as a genuine JOP (1971: 234). Even though that judgment obviously involves the category of cause, Prauss argues that it merely involves a use (Gebrauch) of it, not an application (Anwendung) (1971: 251–53, 272–92). As will become clear, I disagree with Prauss that Kant’s examples must be understood to refer to objects rather than perceptions. In this section, I will argue that they should instead be understood as expressing a subjective temporal sequence of perceptions, rather than a succession of objective events. In §3.2, I will argue that this is the case despite Prauss’s worry that such judgments employ objective concepts, such as ‘sun,’ ‘stone’ or ‘sugar’. Moreover, Prauss’s proposed formulation conflicts with Kant’s claim that it is precisely by adding a missing (relational) category—not by subtracting an ‘It seems’ from a judgment that already contains the category—that we transform a JOP into a JOE. As such, I believe he also fails to correctly capture the role Kant’s distinction plays in his deduction of the categories. I discuss this role in §4. 

19More specifically, the reproductive imagination is necessary for the subject to be aware of a series of perceptions as a series. 

20See also R3051: ‘The representation of the way in which different concepts … belong to one consciousness … is the judgment. They belong to one consciousness partly in accordance with laws of the imagination, thus subjectively, or of the understanding, i.e., objectively valid for every being that has understanding’ (16:633, my emphasis). 

21I should note here that I do not take the imagination’s reproduction of representations to itself be an act of judgment. Neither do I think that being conscious of such reproduced representations requires making a JOP. My claim is only that it is possible for a subject to make a judgment that merely reports a series of representations she finds herself having due to her reproductive imagination. When she does so, on my view, she makes a JOP. 

22Thus, I agree with Brigitte Sassen that the Prolegomena’s JOPs reappear in the first Critique as expressions of ‘unities of consciousness based merely on principles of association’ (2008: 269). Sassen also notes that JOPs are meant to lack the relation to objects that is secured by the categories (2008: 272, 276); however, she does not appear to understand this in the way I do here, since what it turns out she takes these judgments to be lacking (like Longuenesse and Allison) is knowledge (2008: 278) or certainty (2008: 284) that the connections between objects that they express actually obtain. 

23Cf. Bird (1962: 142), Beck (1978: 51–52), Kotzin and Baumgärtner (1990), Sassen (2008: 275, 284), Prauss (1971: 176–93). 

24Kant himself attempts to provide examples of judgments which ‘could [n]ever become judgments of experience’ in the Prolegomena: ‘that the room is warm, the sugar sweet, the wormwood repugnant’ (4:299). Unfortunately, however, these examples have been the cause of more confusion than clarity. One source of confusion is that these examples rely on the view—which Kant holds in the Prolegomena—that sensations of warmth and sweetness make a subject aware of secondary qualities, and secondary qualities cannot (at least correctly) be ascribed to objects. But Kant does not consistently treat sensations corresponding to secondary qualities in this way (for example, he calls such sensations ‘objective’ at CJ 5:206). The example I offer above sidesteps these worries. I discuss Kant’s examples involving secondary qualities further in note 34. Pollok (2008: 338–39) takes Kant’s apparent change of heart about secondary qualities in the CJ to entail that at least by this point, he only counts judgments that articulate feelings as JOPs. As I discuss above, I agree that Kant’s later examples of JOPs better illustrate that he does not intend such judgments to ascribe any qualities to objects. This does not entail, however, that such judgments only express feelings; rather, as I argue, they express mere connections of perceptions (including, but not limited to, feelings) that are not determined to correspond to objects. 

25In my subsequent discussion of Kant’s examples, I will use (n’) to number the JOPs that refer explicitly to representations and (n) for the corresponding colloquial judgments. 

26Once again, I have modified the translation of ‘wenn’ from ‘if’ to ‘when.’ See note 6 above. 

27Another way to put this is that the truth conditions for the judgment are set by my representations, not by the objects represented. See §3.1 below. 

28Specifically, a relational category. See §3.2 below. 

29Longuenesse (1998: 177–79). See also Friedman (2012: 246–47, 261–62), Wolff (2012: 134–45n16). 

30JL 9:114. See also R3145–46, where Kant also provides the example, ‘I sense warmth in touching the oven,’ as contrasted to ‘The oven is warm’ (16: 678–79). 

31Or more colloquially, ‘When I see the sun shining on the stone, I feel it becoming warm.’ 

32For a different analysis of this example, cf. Longuenesse (1998: 176–77), Allison (2015: 301–3), Wolff (2012: 132–35), Beizaei (2017: 359). 

33Or one might just as well judge that (8) ‘When air is heated, it expands.’ As I discussed in §1 above, if (8) is understood to refer to a sequence of objective events, then per Kant’s argument in the Second Analogy, it presupposes that that sequence is taken to be causally determined. As such, it is a JOE that involves the application of the category of cause just as much as (7) does. 

34The same goes for the examples I mentioned in note 24 above: ‘that the room is warm, the sugar sweet, the wormwood repugnant’ (4: 299). On my view, these judgments should also be understood as expressing sequences in a subject’s representations, while leaving undetermined whether or not objects really are the way they are represented as being. Just as above, then, I believe these judgments should be disambiguated as follows: a) ‘My perception of the room is combined with a perception of warmth,’ or more colloquially, ‘When I am in the room, I feel warm’; b) ‘My perception of sugar is combined with a perception of sweetness’ or ‘When I taste sugar, I taste sweetness’; c) ‘My perception of wormwood is combined with a feeling of displeasure’ or ‘When I taste wormwood, I feel displeasure.’ My translation of this set of examples is similar to Prauss’s; however, Prauss (1971: 185–88) argues that stated as such, they fail to adequately exemplify what makes JOPs subjectively valid in general and so, are abandoned by Kant. 

35As we can see from my discussion above, I take JOPs to refer to perceptions by specifying their content, where this content does not include a characterization of what is represented as an object (i.e., a fully categorially determined object). I remain neutral here about whether perceptions count as having content independently of thought about them, and also whether perception in Kant should be understood in relationalist or representationalist terms. See McLear (2016) for a discussion of these issues. 

36See Kemp Smith (1918: 289), Beck (1978: 50), Guyer (1987: 101–2), Kotzin and Baumgärtner (1990: 407), Pollok (2008: 335–36), Sassen (2008: 271–72, 281). 

37See Longuenesse (1998: 82–84, 186), Prauss (1971: 168–72). 

38See R3145–46, dated between 1790–1804 (16: 678–79). 

39As an anonymous referee suggests, another way to put this is that in the B-Deduction, Kant is specifically interested in the class of judgments that constitute cognition of an object. 

40We can find historical motivations for Kant’s rhetorical shift. In the note included in the Metaphysical Foundations, Kant is responding to a review by Johann Schultz of Johann August Heinrich Ulrich’s Institutiones Logicae et Metaphysicae, which was published in 1785. As part of his review of Ulrich’s discussion of Kant’s Critique in that work, Schultz complains that the A-edition of the Critique seems to entail that the categories are conditions even for what the Prolegomena would count as a JOP. With respect to the argument of the Second Analogy, for example, Schultz points out that one would have to ‘know that the sunlight is the cause of the warmth of the stone in order to be able to say: when the sun shines, the stone grows warm’ (translated in Sassen 2000: 213). (Note that Schultz’s complaint is valid if we understand Kant’s example in the Prolegomena in the way that Longuenesse does, but not if we understand it as I suggest we should in (1’) above.) As part of his response, as we have seen, Kant opts to restrict the term ‘judgment’ to acts of the understanding for the sake of clarity, since it allows him to state unequivocally that the categories are conditions only of the latter. As I argue above, however, this does not mean that he no longer believes that combinations of representations that are due to the reproductive imagination are possible, and moreover, that we can be conscious of and report such representations. In fact, as we have seen, he provides an example of just such a combination in §19, which does not suffer from the ambiguity of his examples in the Prolegomena that engenders Schultz’s complaint. For further discussion of the Second Analogy and its relation to Kant’s distinction, see §3.3 below. For an extended discussion of Schultz’s review and Kant’s reception of it, see Allison (2015: 306–15). 

41MF 4:476. Since Pollok takes Kant to have substantively revised his view between the Prolegomena and the B-Deduction, he must treat Kant’s insistence that the change merely concerns the ‘manner of presentation’ as a significant mischaracterization (Pollok 2008, 342). 

42On this point, I agree with Friedman (2012: 262) and Allison (2015: 369). 

43Beck (1978: 53). See also Bird (1962: 145), Kotzin and Baumgärtner (1990: 410). 

44This identification can also be found in Allison (2015: 366–67) and Vanzo (2012: 120). 

45Vanzo argues that JOPs are not ‘truth-apt’ for Kant on the basis of his claim in the Doctrine of Method that truth ‘rests upon agreement with the object’ (A820/B848). Since he agrees that JOPs are not about objects, Vanzo concludes that they cannot be true (2012: 120). What this overlooks, however, is that Kant makes this claim as part of his discussion of the type of holding-for-true [fürwahrhalten] he calls ‘persuasion’: an attitude by which the subject takes something to be true of the objects of her perception, albeit on insufficient grounds (see also 9:73 and Chignell 2007). On my view, Kant is describing the truth-conditions of JOEs here, not denying that JOPs can be true. 

46Thus, I agree with Prauss that JOPs can be true; however, I disagree with his claim that they cannot possibly be false (1971: 238–43). 

47See Prauss (1971: 188), Beck (1978: 51–52), Pollok (2008: 338–39). 

48This is in line with Kant’s claim at A664/B692 that while the mathematical categories of quantity and quality are constitutive of intuition, the dynamical categories of relation and modality are constitutive of experience. The dynamical categories also include the categories of modality. I focus on the relational categories in my discussion since, as Kant notes, the dynamical categories do not add any content to what is thought about the object. Rather, they merely express whether the object is possible, actual or necessary in light of the necessary conditions of experience that are due to the understanding, judgment and reason. Thus, the role of the categories of modality, Kant says, is to ‘restric[t]’ the other categories ‘to merely empirical use’ (A219/B266). 

49The contrast is not referred to using these terms in the Prolegomena: there, Kant says that both JOEs and JOPs express ‘connections’ (e.g., at 4: 300). However, he repeatedly distinguishes between two kinds of connection in the Prolegomena: on the one hand, a ‘necessary’ or ‘synthetic connection’ (4: 305) of perceptions that accords with ‘the rules that determine the connection of representations in the concept of an object’(4: 290), and, on the other hand, a ‘logical connection’ (4: 298, 4: 304) or ‘mere aggregate’ of perceptions (4: 310) that merely reflects how they are ‘given in sensory intuition’ (4: 304). It is the latter, on my view, that comes to be called a mere ‘composition’ in the note added to the B-edition that I cite above. Kant also claims that it is the ‘composition of the manifold in an empirical intuition’ that enables consciousness of mere perceptions at B160 (my emphasis). However, he is not always consistent in his usage, sometimes continuing to refer to combinations due merely to the imagination as ‘connections’ (e.g., at B164; B233). Thanks to an anonymous referee for pressing me on this. 

50This has already been noted by a number of other interpreters. See Prauss (1971: 163–64), Beck (1978: 52), Sassen (2008: 277), Pollok (2008: 337–38), Friedman (2012: 258), Allison (2015: 300–303), Beizaei (2017: 348, 351–61). 

51Compare Sassen (2008: 278), Beizaei (2017: 360). 

52This is important for my response to a possible objection. I have argued that judgments like (3) When I hear a rooster crow, I imagine the smell of coffee, and (4) When I carry a body, I feel a pressure of weight, are JOPs. However, the subject of these judgments is the ‘I’ which, as such, is presumably thought of as a substance (though this will need to be qualified: see note 55 below). Moreover, if this is the standard format of all JOPs, then all JOPs already employ the relational category of substance. On my view, however, this is compatible with their status as JOPs, since I do not claim that JOPs cannot employ any relational categories at all. Rather, as I explain above, their status as JOPs derives from the fact that they report a merely temporal relation between perceptions, where that relation in particular has not been subsumed under a relational category, and as such, is not taken to correspond to a connection between the objects the perceptions are of. It is only by judging that the objects the perceptions are of are connected as substance to accident, cause to effect, or as mutually interacting members of a community that a JOP is converted to a JOE. See note 55 for further discussion. I thank Colin McLear as well as an anonymous referee for raising this worry. 

53Indeed, though I cannot defend this claim in this paper, this is another way in which JOPs turn out to be crucial for Kant’s account of experience, namely, in virtue of expressing the kind of empirical apperception that forms the basis for inner experience and knowledge of ones’ own psychology. I develop an account of this elsewhere. 

54Judgments that subsume representations themselves under the category of cause count as JOEs in a derivative sense. They do not—as per my definition in §2—claim that what is represented by a sequence of perceptions is combined in the independent object the perceptions are of. Rather, they treat the perceptions themselves as states of an inner object, in virtue of claiming that their succession in time is causally determined. 

55We might question why the subject needs to bring her representations under the category of cause in particular in order to count as making an (inner) JOE. Why is it not sufficient that she think of herself as the subject of her perceptions—as she does in the judgments I have identified as JOPs—and so, as a substance? In response, it should be noted that Kant makes clear in the Paralogisms that in thinking of oneself merely as the subject of one’s thoughts, one does not in fact use the schematized category of substance, and so, succeed in thinking of oneself as a substantial object (A348–51; B410–11, see also Kraus 2019). Merely thinking of oneself as the subject of one’s thoughts does not amount to intuiting oneself as persisting in time; the latter is required, however, for application of the schematized category of substance. On my view, it is only by thinking of the sequence of one’s representations in time as causally determined that one succeeds in cognizing oneself as an (inner) object. I develop this claim elsewhere. Thanks to an anonymous referee for pressing this point. 

56Hume, Enquiry, §4. Some of the concepts that Kant extends the ‘Humean problem’ to—such as the concept of substance—were considered by Hume himself in the Treatise of Human Nature, which predates the Enquiry by ten years. However, while a German translation of the Enquiry was published in 1755 and was owned by Kant, the Treatise was not fully translated into German until the 1790s, after the publication of the Prolegomena and both editions of the Critique. On this point, see Kuehn (1983: 179), and for an extended discussion, Guyer (2008: Introduction and Chs. 1–2). 

57Stephen Engstrom argues that the Deduction ‘does not aim at a refutation of skepticism’ (1994: 359), either of the Cartesian or the Humean variety. Considering Humean skepticism in particular, however, Engstrom’s main point is that Kant does not seek to refute Hume’s skepticism, but rather to ‘remove’ its source by showing that the categories do not have an empirical origin (1994: 375). As Engstrom notes, Kant’s full response to Hume involves answering the ‘question of our right to employ’ the categories (1994: 376). It is the latter that I will be concerned with here, and I take my construal of it to be compatible with Engstrom’s characterization of Kant’s task. See also Ameriks (2000: 43; 2003: esp. 51, 65–66), Hatfield (2001), Watkins (2004: 381–90) and Dyck (2011: esp. 495–96) for other versions of the claim that Kant does not seek to refute Hume. For responses to the concerns of many of these interpreters that I am in broad agreement with see Guyer (2008: esp. Introduction, Chs. 1–2); see also Forster (2008: 20–32, 40–43). 

58Kant clearly does read Hume as denying that we are rationally entitled to apply the category of cause to the objects of experience and sees himself as responding to this kind of skepticism. For a particularly vivid statement of this, see CPrR, 5: 51–53. 

59For the Humean skeptic, this is not (merely) because of a lack of empirical evidence, but rather, because such judgments employ concepts that have not been derived from experience. Nevertheless, the point stands that Kant cannot hope to get very far with the Humean skeptic if his transcendental argument requires the skeptic to grant that we can make sufficiently justified judgments about perceived objects. Indeed, some interpreters end up concluding that Kant could not even have intended for this stretch of argument to have any force against Humean skepticism. See, for example, Wolff (2012: 131), Allison (2015: 293). This conclusion, however, is at odds with Kant’s claim in the Prolegomena that through his arguments there, he has ‘thoroughly dispose[d] of the Humean doubt’ (4: 310). 

60Of course, a more radical skeptic might deny this, but I do not think that Kant’s argument is meant to respond to this kind of skeptic. See Guyer (2008: 112–13). 

61Hume himself arguably describes us as representing objects as actually being some way or other. ‘When we look about us towards external objects,’ he says in the Enquiry, for example, we ‘find … that the one [event] does actually, in fact, follow the other.’ When we find that one event ‘actually, in fact’ follows another, we make a claim about objects, rather than about our own perceptions. 

62Thus, Kant argues in the Prolegomena that the solution to the ‘Humean problem’ lies in seeing that the relational categories are not derived from experience, but rather that ‘experience is derived from them,’ which he says is a ‘completely reversed type of connection that never occurred to Hume’ (4: 313). Now, Kant makes clear that by ‘experience’ here, he means ‘a synthetic unification of perceptions’ (4: 312) and one might question why we should think he takes the latter to involve any claim to objective validity—that is, to representing what the perceptions are of as an object. However, this is precisely how Kant understands synthetic unity: as he says in the Deduction, it is their synthetic unity that ‘constitutes the relation of representations to an object, thus their objective validity, and consequently is that which makes them into cognitions …’ (B137–38, see also B218). I thank an anonymous referee for pressing me to further clarify this point. 

63See, for example, Allison (2004: 180), Longuenesse (1998: 187), Wolff (2012: 131), Pollok (2008: 327). 

Acknowledgements

For very helpful comments, discussion or advice, I am grateful to Richard Booth, Sarah Buss, Hannah Ginsborg, Colin McLear, Tyke Nunez, Laura Ruetsche, Tad Schmaltz, Umrao Sethi, Daniel Warren, Jessica Williams, participants at the 2019 Eastern Study Group of the North American Kant Society, the 2019 Kant Congress, the Faculty Working Group at the University of Michigan department of philosophy, audience members at the University of South Florida, as well as an anonymous referee for this journal.

Competing Interests

The author has no competing interests to declare.

References


    Primary Sources

      Unless indicated otherwise, translations of Kant’s text are from the Cambridge Edition of the Works of Immanuel Kant (ed. by Paul Guyer and Allen Wood, Cambridge: Cambridge University Press, 1992–). References to Kant’s works are given using the volume and page number in the Akademie edition, Kants Gesammelte Schriften (ed. by Königlich Preussische Akademie der Wissenschaften, vols. 1–29, Berlin: de Gruyter, 1902–). References to the Critique of Pure Reason are given using the standard A and B pagination corresponding to the first and second editions respectively. Other works are referred to using the following abbreviations:

      CJ: Critique of Judgment
      CPrR: Critique of Practical Reason
      DWL: Dohna-Wundlacken Logic
      JL: Jäsche Logic
      MF: Metaphysical Foundations of Natural Science
      Pr.: Prolegomena
      VL: Vienna Logic

    1. Hume, David. 1975. Enquiries Concerning Human Understanding and Concerning the Principles of Morals. Edited by L. A. Selby-Bigge and P. H. Nidditch. Oxford: Oxford University Press. DOI: https://doi.org/10.1093/actrade/9780198245353.book.1 


    Secondary Sources

    1. Allais, Lucy. 2009. “Kant, Non-Conceptual Content and the Representation of Space.” Journal of the History of Philosophy 47 (3): 383–413. DOI: https://doi.org/10.1353/hph.0.0134 

    2. Allison, Henry. 2004. Kant’s Transcendental Idealism: An Interpretation and Defense. New Haven: Oxford University Press. DOI: https://doi.org/10.2307/j.ctt1cc2kjc 

    3. Allison, Henry. 2015. Kant’s Transcendental Deduction: An Analytical-Historical Commentary. Oxford: Oxford University Press. DOI: https://doi.org/10.1093/acprof:oso/9780198724858.001.0001 

    4. Ameriks, Karl. 2000. Kant and the Fate of Autonomy: Problems in the Appropriation of the Critical Philosophy. Cambridge: Cambridge University Press. DOI: https://doi.org/10.1017/CBO9781139173346 

    5. Ameriks, Karl. 2003. “Kant’s Transcendental Deduction as a Regressive Argument.” In Interpreting Kant’s Critiques, 51–66. Oxford: Oxford University Press. DOI: https://doi.org/10.1093/0199247315.003.0002 

    6. Beck, Lewis White. 1978. “Did the Sage of Königsberg Have No Dreams?” In Essays on Kant and Hume, 38–60. New Haven: Yale University Press. 

    7. Bird, Graham. 1962. Kant’s Theory of Knowledge. London: Routledge. 

    8. Beizaei, Banafsheh. 2017. “Kant on Perception, Experience and Judgements Thereof.” Kantian Review 22 (3): 347–71. DOI: https://doi.org/10.1017/S1369415417000127 

    9. Cassirer, Ernst. 1923, reprinted 1953. Substance and Function. Translated by Marie Collins Swabey and William Curtis Swabey. New York: Dover Publications. 

    10. Chignell, Andrew. 2007. “Belief in Kant.” Philosophical Review 116 (3): 323–60. DOI: https://doi.org/10.1215/00318108-2007-001 

    11. Dyck, Corey W. 2011. “Kant’s Transcendental Deduction and the Ghosts of Descartes and Hume.” British Journal for the History of Philosophy 19 (3): 473–96. DOI: https://doi.org/10.1080/09608788.2011.563522 

    12. Engstrom, Stephen. 1994. “The Transcendental Deduction and Skepticism.” Journal of the History of Philosophy 32 (3): 359–80. DOI: https://doi.org/10.1353/hph.1994.0069 

    13. Forster, Michael N. 2008. Kant and Skepticism. Princeton: Princeton University Press. 

    14. Friedman, Michael. 2012. “The Prolegomena and Natural Science.” In Kants Prolegomena: Ein kooperativer Kommentar, edited by H. Lyre and O. Schliemann, 299–326. Frankfurt am Main: Klostermann. 

    15. Guyer, Paul. 1987. Kant and the Claims of Knowledge. Cambridge: Cambridge University Press. DOI: https://doi.org/10.1017/CBO9780511624766 

    16. Guyer, Paul. 2008. Knowledge, Reason and Taste: Kant’s Response to Hume. Princeton: Princeton University Press. DOI: https://doi.org/10.1515/9781400824472 

    17. Hatfield, Gary. 2001. “The Prolegomena and the Critiques of Pure Reason.” In Kant und die Berliner Aufklärung: Akten des IX. Internationalen Kant-Kongresses, edited by Volker Gerhardt, Rolf-Peter Horstmann, and Ralph Schumacher, 185–208. Berlin: Walter de Gruyter. DOI: https://doi.org/10.1515/9783110874129.185 

    18. Kemp Smith, Norman. 1918. A Commentary to Kant’s Critique of Pure Reason. London: Palgrave Macmillan. 

    19. Kotzin, Rhoda H. and Jörg Baumgärtner. 1990. “Sensations and Judgments of Perceptions: Diagnosis and Rehabilitation of Some of Kant’s Misleading Examples.” Kant-Studien 81 (4): 401–12. DOI: https://doi.org/10.1515/kant.1990.81.4.401 

    20. Kraus, Katharina. 2019. “The Parity and Disparity between Inner and Outer Experience in Kant.” Kantian Review 24 (2): 171–95. DOI: https://doi.org/10.1017/S1369415419000013 

    21. Kuehn, Manfred. 1983. “Kant’s Conception of ‘Hume’s Problem.’” Journal of the History of Philosophy 21 (2): 175–93. DOI: https://doi.org/10.1353/hph.1983.0051 

    22. Longuenesse, Beatrice. 1998. Kant and the Capacity to Judge. Princeton: Princeton University Press. 

    23. McLear, Colin. 2016. “Kant on Perceptual Content.” Mind 125 (497): 95–144. DOI: https://doi.org/10.1093/mind/fzv191 

    24. McLear, Colin. 2020. “Animals and Objectivity.” In Kant on Animals, edited by Lucy Allais and John Callanan, 42–65. Oxford: Oxford University Press. DOI: https://doi.org/10.1093/oso/9780198859918.003.0003 

    25. Pollok, Konstantin. 2008. “‘An Almost Single Inference’—Kant’s Deduction of the Categories Reconsidered”. Archiv für Geschichte der Philosophie 90 (3): 323–45. DOI: https://doi.org/10.1515/AGPH.2008.013 

    26. Prauss, Gerold. 1971. Erscheinung bei Kant. Berlin: De Gruyter. DOI: https://doi.org/10.1515/9783110838824 

    27. Sassen, Brigitte. 2000. Kant’s Early Critics. Cambridge: Cambridge University Press. 

    28. Sassen, Brigitte. 2008. “Varieties of Subjective Judgments: Judgments of Perception”. Kant-Studien 99 (3): 269–84. DOI: https://doi.org/10.1515/KANT.2008.021 

    29. Strawson, Peter. 1970. “Imagination and Perception.” In Experience and Theory, edited by Lawrence Foster and J. W. Swanson, 31–54. Amherst, MA: University of Massachusetts Press, 31–54. 

    30. Tolley, Clinton. 2017. “Kant on the Place of Cognition in the Progression of Our Representations.” Synthese. DOI: https://doi.org/10.1007/s11229-017-1625-3 

    31. Vanzo, Alberto. 2012. “Kant on Truth-Aptness.” History and Philosophy of Logic 33 (2): 109–26. DOI: https://doi.org/10.1080/01445340.2011.606645 

    32. Watkins, Eric. 2004. Kant and the Metaphysics of Causality. Cambridge: Cambridge University Press. DOI: https://doi.org/10.1017/CBO9780511614217 

    33. Watkins, Eric and Marcus Willaschek. 2017. “Kant’s Account of Cognition.” Journal of the History of Philosophy 55 (1): 83–112. DOI: https://doi.org/10.1353/hph.2017.0003 

    34. Wolff, Michael. 2012. “Die Analyse der Erfahrung in Kants Prolegomena.” In Kants Prolegomena: Ein kooperativer Kommentar, edited by H. Lyre and O. Schliemann, 127–67. Frankfurt am Main: Klostermann. 

comments powered by Disqus