Kant’s famous distinction between things in themselves and appearances forms a cornerstone of his transcendental idealism. However, notwithstanding its enormous importance, one can hardly find two Kant scholars agreeing on how to read it. Basically, there are two camps. Defenders of a two-world interpretation regard things in themselves and appearances as numerically distinct objects.Footnote 1 To the contrary, those who hold to a two-aspect view say that it is one and the same object merely regarded from different points of view.Footnote 2 Both camps can adduce a large amount of textual evidence in favour of their interpretations. As in the antinomies, the controversy might thus last forever unless it proves possible to provide an interpretation that can do justice to the central textual motivations of the arguments of both camps.Footnote 3 It should then come as no surprise that nowadays many Kant scholars are apparently sympathetic to attempts to resolve the conflict.
The aim of this paper is to propose a novel account of how to end the battle. The root of the battle, I take it, is the equating of a two-aspect with a one-world view. A one-world view is incompatible with a two-world view, of course. But as I shall argue, a two-aspect account is not only compatible with a two-world view; they even entail each other. In the following, I shall argue that things in themselves can be considered independently of any relation to us and of the relation of appearing, i.e. affection, in particular. This involves a distinction of two aspects. But since affection causes sensations and appearances are ultimately grounded in sensations, the relation of appearing requires appearances numerically distinct from things in themselves. Therefore appearing things are different from appearances. As there can be no appearances without something that appears, on Kant’s account, there is a mutual entailment of a two-world and a two-aspect view.
My plan for this paper is as follows. In section 1 I point out why there is a textual stalemate. In section 2 I consider the notion of a ‘thing in itself’ in depth, likewise the distinction between ‘appearance’ and ‘appearing’ in section 3. I will show how this helps to resolve the conflict in section 4. However, my reconciling suggestion would be anodyne if Kant allowed for a distinction between mind-immanent and mind-transcendent appearances. I will put forth arguments for the claim that Kant does not accept non-mental appearances in section 5. This results in a form of phenomenalism of which I give a brief outline in the concluding section 6.
1. A Textual Stalemate
The more recent debate concerning the distinction between appearances and things in themselves has its origin in Gerold Prauss’s Kant und das Problem der Dinge an sich. Prauss argues that the expression ‘in itself’ in ‘thing in itself’ (Ding an sich or much more often Ding an sich selbst) is an abbreviation of ‘thing, considered in itself’. Thus he thinks Kant does not make a distinction between two kinds of entities, appearances and things-in-themselves, but between two ways to consider a thing, i.e. to consider it in itself and not to consider it in itself.Footnote 4
Prauss’s book is the philological basis for two-aspect interpretations. Indeed, Prauss has succeeded in showing that the expressions ‘as they appear’ and ‘as they are in themselves’ refer to two aspects and not to two kinds of things. I settle for two representative passages to underline his claim:Footnote 5
so that the same objects can be considered from two different sides, on the one side as objects of the senses and the understanding for experience, and on the other side as objects that are merely thought, at most for isolated reason striving beyond the bounds of experience. (Bxviii–xix n., translation modified)
But they did not consider that both [sc. space and time], without their reality as representations being disputed, nevertheless only belong to appearance, which always has two sides, one where the object is considered in itself (without regard to the way in which it is to be intuited, the constitution of which however must for that very reason always remain problematic), the other where the form of the intuition of this object is considered … (A38/B55)
Attempts to downplay these and other passages are doomed to failure. Kant explicitly says that the distinction between appearance and thing in itself concerns ‘precisely the same’ things (Bxxvii). And in his Opus Postumum he writes that the thing in itself is not a ‘different object’, but a ‘different relation of a representation to one and the same object’ (OP 22: 26, my translation). Further evidence is provided by the fact that Kant repeatedly emphasizes that there are not two kinds of selves (e.g. noumenal and phenomenal self, subject and object of thought), but only two aspects of them.Footnote 6
The two-world camp, to the contrary, can refer to the fact that Kant makes a distinction between mundus sensibilis and mundus intelligibilis, a sensible and a noumenal world.Footnote 7 In addition, there are numerous passages in which Kant holds that appearances, as representations, exist ‘merely in us’ and not outside the subject (A129).Footnote 8 Appearances are ‘only representations of things’ (B164), ‘nothing but representations’ (A250), ‘merely representations in us’ (A387), and the like. They are not ‘in the same quality as they are in us as things external to us and subsisting by themselves’ (A386) and ‘cannot exist at all outside our mind’ (A492/B520). Kant flatly says that ‘that which we call outer objects are nothing other than mere representations of our sensibility’ (A30/B45).Footnote 9
Kant sometimes explains what it means that representations exist ‘in us’. According to his criticism of the paralogism of substantiality, we need to think the soul as substance in which ‘thoughts only inhere as determinations’ (A349; see also B164, PM, 20: 270). The determinations of inner sense, we are told, are ‘determinations’ of the soul (A381), everybody necessarily has to ‘regard himself as a substance, but regard his thinking only as accidents of his existence (Dasein) and determinations of his state’ (A349).Footnote 10 However, appearances are not representations in the sense of mental items (or acts),Footnote 11 but rather the content of these items.Footnote 12 For when I conceive of a table, we cannot say that the representation is a table. Representations themselves are never tables or the like. In contrast to their content, representations themselves really exist. Thus we should say that representations themselves belong to the noumenal world, whereas their content makes up the phenomenal world.Footnote 13 The moral is clear: If appearances are only the content of representations, they cannot be numerically identical to things in themselves.Footnote 14
These are powerful arguments in favour of a two-world interpretation. So it seems that we have to accept that things in themselves and appearances form, so to speak, two worlds. The result is a textual stalemate between two-aspect and two-world interpretations since both camps have plenty of textual evidence in support of their views.
2. Things in Themselves
To overcome the textual stalemate between the two-world and the two-aspect camp, I shall begin by investigating the concept of a ‘thing in itself’. It is important to keep in mind that Kant never claims to have invented the concept of a thing in itself. Instead, he accuses Leibniz, Locke and the transcendental realist in general (who is opposed to the transcendental idealist because she thinks that space and time are the forms of things in themselves, see A369, A491/B519) of conflating things in themselves and appearances. One can tentatively conclude from this that Kant thought that his predecessors already had the notions of things in themselves and appearances but misapplied them. Furthermore, Kant never really explains what he means by ‘thing in itself’, nor does he explain the concept of ‘appearance’. This suggests that he not only expected his readers to be familiar with these concepts, but that he could reasonably do so. Since, as far as I know, ‘thing in itself’ did not have a specifically philosophical meaning before Kant, we can assume that it was accessible to every competent German speaker of Kant’s day.
As a first approach to the meaning of ‘thing in itself’, let us have a look into a dictionary of the German language of Kant’s day, edited by J. C. Adelung from 1774 to 1786. It is explained there that ‘in itself’ (an sich) means that a thing is determined ‘without a relation to another [thing]’. Examples given by Adelung are ‘The in itself dead richness’ and ‘Tell me how the thing is in itself’ (see ‘An’ in Adelung Reference Adelung1774–86). I take this to mean that we are abstracting from any relation to other objects and are merely considering the properties a thing has independently of any relation. If the same holds for Kant’s things in themselves, a thing in itself is a thing considered independently of any relation. (I will show later that the relation in question is a relation to a cognizing subject since the relation in question is affection.)
To give textual evidence for my suggestion, I now point out that Kant uses the ‘in itself’ construction in other contexts too. An analogy to the thing in itself is the ‘end in itself’. An end in itself, we are told, is something that has ‘absolute worth’ (G, 4: 428).Footnote 15 In contrast, a mere means can be an end, however, not in itself but only in relation ‘to a specially constituted faculty of desire’ (ibid.). A human being ‘exists as an end in itself, not merely as a means to be used by this or that will at its discretion’ (ibid.). The distinction should be clear by now: a mere means has value only in relation to a certain end that can be achieved by it; an end in itself, to the contrary, has value independently of any relation to other ends. Further examples: a proof of God’s existence is not possible, but the possibility of God’s existence ‘in itself’ cannot be refuted (A792/B820); pure categories do not have any meaning ‘in themselves’, but only together with intuition (A348); contradictory judgements ‘in themselves (even without regard to the object) are nothing’ (A150/B189). In all these cases the considered entity has an aspect that exists without a relation to anything else.Footnote 16
All this shows that the ‘in itself’ construction has the same meaning in different contexts. So there is reason to assume that ‘thing in itself’ is analogous to ‘end in itself’, etc. Indeed, there is textual evidence for it. Things in themselves are things ‘without their relation to our intuition’ (A36/B52), ‘without relation to your subject’ (A48/B65), ‘without relation to our senses and possible experience’ (A493/B521). We can thus conclude that things in themselves are things that have an aspect existing independently of any relation.Footnote 17
This runs counter to Prauss’s methodological interpretation. He takes ‘in itself’ for an abbreviation of ‘considered in itself’ and thus for an adverbial determination of ‘to consider’ and not for an adjectival determination of ‘thing’ (Prauss Reference Prauss1974: 13–43). I do not dispute that the linguistic origin of ‘in itself’ is adverbial,Footnote 18 even though Richard Aquila has shown that it should be more aptly related to ‘to exist’ than to ‘to consider’ (Aquila Reference Aquila1979 and 1983: 88ff.). However, I do dispute that things in themselves are not numerically distinct from appearances. To my mind, we have two contrasting predicates – ‘to exist in itself’ and ‘to appear’ – and correspondingly two contrasting substantives – ‘thing in itself’ and ‘appearance’. Let us have a closer look at this.
3. ‘Appearing’ and ‘Appearance’
In the middle of the twentieth century, philosophers such as Ayer or Chisholm distinguished between a ‘theory of appearing’ and a ‘theory of appearances’ (Ayer Reference Ayer1940; Chisholm Reference Chisholm1963). According to a theory of appearing, perception is a direct relation to outer objects without requiring the existence of mind-immanent appearances. To the contrary, defenders of a theory of appearances assume sense-data, through which outer objects are mediately given to us. Stephen Barker has drawn attention to the fact that Kant sometimes seems to endorse a theory of appearing, sometimes a theory of appearances (Barker Reference Barker1967). Kant indeed uses both the noun Erscheinung (appearance) and the verb erscheinen (to appear).Footnote 19 The following passages count in favour of a theory of appearing:
[A] rose … can appear different to every eye in regard to colour. (A30/B45)
We have therefore wanted to say … that the things that we intuit are not in themselves what we intuit them to be, nor are their relations so constituted in themselves as they appear to us … (A42/B59)
I say: in space and time intuition represents both outer objects as well as the self-intuition of the mind as each affects our senses, i.e., as it appears, that is not to say that these objects would be a mere illusion. (B69)
But these passages suggest a theory of appearances:
The undetermined object of an empirical intuition is called appearance. (A20/B34)
But if we consider this empirical object in general and … ask whether it (not the raindrops, since these, as appearances, are already empirical objects) represents an object in itself, then the question of the relation of the representation to the object is transcendental, and not only these drops are mere appearances, but even their round form, indeed even the space through which they fall are nothing in themselves, but only mere modifications or foundations of our sensible intuition … (A45–6/B63)
In the transcendental aesthetic we have undeniably proved that bodies are mere appearances of our outer sense, and not things in themselves. (A357)
Barker thinks that Kant makes use of both theories without realizing their incompatibility. I admit that Kant uses both kinds of theories. But is it true that he conflates them? To my mind this is not the case. In the following passage Kant says that there can be no appearance ‘without anything that appears’:
Yet the reservation must also be well noted, that even if we cannot cognize these same objects [of experience] as things in themselves, we at least must be able to think them as things in themselves. For otherwise there would follow from this the muddled proposition that appearance could exist without anything that appears. (Bxxvi–xxvii, translation modified)
If appearances as entities require something that appears and thus the relation of appearing, not only is a distinction made between appearances and appearing, but both theories are even united into one coherent theory which is both a theory of appearances and of appearing. Kant thinks that the possibility of appearances rests on ‘the relation of certain things, unknown in themselves, to something else, namely our sensibility’ (Prol., 4: 286). The relation in question is affection. This is the relation through which things bring forth sensations in us (be it a causal or some other kind of grounding relation, see A19/B33). Kant identifies affection with appearing in the quotation above from B69, so the affecting objects are the appearing objects.Footnote 20 Since we are merely concerned with outer affection here, the appearing object must exist outside our mind, whereas appearances in at least one sense of the term exist only as representations.
The ‘things unknown in themselves’, to the contrary, are (unsurprisingly) things in themselves, as turns out by a regress argument:
This was the result of the entire Transcendental Aesthetic, and it also follows naturally from the concept of an appearance in general that something must correspond to it which is not in itself appearance, for appearance can be nothing in itself and outside of our kind of representation; thus, if there is not to be a constant circle, the word ‘appearance’ must already indicate a relation to something the immediate representation of which is, to be sure, sensible, but which in itself, without this constitution of our sensibility (on which the form of our intuition is grounded), must be something, i.e., an object independent of sensibility. (A251)
Kant speaks of a ‘constant circle’, but his description goes better with an infinite regress: if the ground of an appearance were again an appearance, we would still need a further appearance, ad infinitum. So the only objects able to ground appearances, according to Kant, are things in themselves. The reason why Kant finds this argument so compelling is probably that we can reproduce sensations, but not create them. Thus we cannot originally produce the matter of experience but need outer objects to ground it (see A376–7, B278–9). So it is things in themselves that affectFootnote 21 us, and appearing things are (at least conceptually) different from appearances.Footnote 22
4. Reconciling Two Worlds and Two Aspects
We are now in a position to determine the meaning of the substantives ‘thing in itself’ and ‘appearance’ on the one hand and the predicates ‘to exist in itself’ and ‘to appear’ on the other. To begin with the latter, the distinction runs as follows:
T-two-aspect: Things as they are in themselves are things insofar as they have aspects that are independent of any relation to us.
A-two-aspect: Things as they appear are things insofar as they affect us.
This does justice to all the passages where Kant says that things in themselves and the things that appear (once again: not appearances) are numerically identical. Appearing and existing in itself are two (metaphysical) aspects of the same thing. There are aspects in the things that are independent of us, while others are constitutive of the relation of affection.Footnote 23 But at the same time, there is a distinction between two kinds of entities:
T-two-world: Things in themselves are mind-transcendent entities.
A-two-world: Appearances are mind-immanent entities.
Appearances, as the mediate result of affection, are numerically distinct from the objects that ground affection. There is a mutual entailment between both distinctions. Things in themselves and appearances are different objects. However, due to the fact that we cannot create appearances out of nothing but need to be affected by objects external to our mind, there must be objects appearing to us. So there would be no appearances without things that appear.Footnote 24 Conversely, the relation of appearing presupposes that there really are appearances. For if we did not synthesize sensations into an organized whole in space and time to which we give the name of an ‘appearance’, there would admittedly be the relation of affection. But affection would not yield more than the raw material of perception, so we could not truly say that things appear to us. Hence appearing requires appearances.Footnote 25 As a result, a two-world account presupposes a two-aspect one and vice versa.
A metaphor recently used by Ralph Walker is suitable to illustrate this. When I look into a mirror, my mirror image is numerically distinct from me. But at the same time, it is the numerically identical person who exists in herself and is reflected in the mirror. So Walker thinks that the dispute on the number of Kant’s worlds is useless (Walker Reference Walker2010). Although I find Walker’s metaphor very instructive, I disagree with him. He seems to equate a two-aspect with a one-world model. But we can see by now why this is false. If I focus on the identity of the person looking at the mirror and being reflected in it, the mirror image does not vanish. Instead, there is a mutual entailment between my mirror image and being mirrored: without being reflected in the mirror, there would be no mirror image; without the existence of a mirror image, nobody would be mirrored. Thus there are two aspects of me and at the same time two objects, I and my mirror image. So the apparent conflict between the two-world and the two-aspect camp is resolved.Footnote 26
The mutual dependence of a two-world and a two-aspect account becomes clearer when we take a closer view at one proponent of each camp. James Van Cleve (Reference Van Cleve1999) subscribes to a phenomenalist two-world interpretation to the effect that appearances are mere constructions out of sense data. I basically agree with him, but add that his interpretation needs to be supplemented by a two-aspect interpretation. For Van Cleve fails to realize that it is the very same object that exists in itself and appears to us. Conversely, Henry Allison (Reference Allison2004) recognizes this on the one hand, but on the other hand he does not realize that, for this very reason, mental appearances as the correlate of the appearing relation are required. Thus his two-aspect interpretation needs to be supplemented by a two-world one.
My mediating proposal rules out two sorts of theories: on the one hand theories which are inconsistent with a two-aspect distinction between things as they are in themselves and as they appear, on the other hand those theories which are inconsistent with a two-world distinction between things in themselves and appearances. However, surprisingly few interpretations can be ruled out by these criteria. Concerning the two-aspect distinction, only those interpretations can be ruled out that deny affection by things in themselves altogether.Footnote 27 For this would be the only possibility as to how appearing things could be different from things in themselves. But very few interpreters hold to that view. Even two-world interpreters are in a position to say that we are affected by things in themselves, though we might at the same time be affected by appearances numerically distinct from things in themselves.
As for the two-world distinction, the same pattern is seen. Many two-aspect interpreters are willing to admit that appearances as the content of representations are numerically distinct from things in themselves. However, they insist that there is yet another sense of ‘appearance’ according to which appearances are the represented objects distinct from those representations through which they are represented. And this kind of appearance, so they argue, is identical to things in themselves. (I will refer to the former kind of appearance as mind-immanent and to the latter as mind-transcendent appearances.) Thus only those scholars who ascribe a relationist account of perception to Kant, and thus deny a distinct notion of ‘content’, are forced to reject the two-world distinction.Footnote 28
As a result, it might seem that my reconciling suggestion is anodyne:Footnote 29 things in themselves are numerically identical to mind-transcendent, but numerically distinct from mind-immanent appearances. However, it would still have the merit of solving the conundrum of the alleged inconsistency of Kant’s text, for it would show how the passages in favour of a two-world and a two-object view are to be understood. But it would be somewhat disappointing after all, as it would not significantly contribute to a better understanding of Kant’s transcendental idealism. However, it seems to me that there are severe problems for the anodyne version. If I am not mistaken, Kant does not consider appearances as entities existing external to our mind. Instead, appearances are nothing over and above the content of representation and the only really existing things are things in themselves. Kant thus accepts some sort of phenomenalism. Let me explain.
5. Are there Mind-Transcendent Appearances?
According to the anodyne view, in addition to appearances as some sort of mental, mind-immanent items (whether as something literally immanent within the process of sensory representing or as the mere ‘contents’ of sensory representing), there are also mind-transcendent appearances in the sense of really existing objects in space and time outside our representations. Ignoring the possibility that there are no appearances at all, there are two ways to challenge the anodyne view. One of them is to challenge the claim that there are mind-immanent appearances. This is the view of those imputing to Kant a relationist approach to perception. On their account, there is no mental appearance but only the relation of appearing. I think this view is flawed: Kant clearly speaks about appearances as a certain sort of mental content.Footnote 30 But I do not discuss the relationist approach here.
The alternative is to deny that there are mind-transcendent appearances. This is tantamount to some sort of phenomenalism.Footnote 31 The wide majority of Kant scholars takes issue with phenomenalist approaches to Kant. It is often stated that there are mind-transcendent appearances, but this is rarely argued for. Some say that we must ascribe this view to Kant due to the principle of charity. However, it is unclear to me why we should do so. Not every sort of phenomenalism is absurd enough not to ascribe, just for that reason, to Kant. And whether or not one likes phenomenalism, it is at least conceivable that Kant’s transcendental idealism is no less coherent on a phenomenalist reading than on a non-phenomenalist one. Therefore one should not reject a phenomenalist interpretation right from the start. In the following I will show that there are no such reasons; instead, there are good reasons to accept a phenomenalist interpretation, as a reading of ‘appearances’ as non-mental items faces many difficulties. In the course of this, I will first show that even the empirical objects that physics deals with are nothing other than mental content, and then that there is no decisive textual evidence in favour of non-mental appearances. It should be noted, however, that this is no more than a brief outline, and that for brevity’s sake some arguments I consider powerful cannot be adduced here.
5.1. Empirical Objects as Content of Representations
If Kant believed in mind-transcendent appearances, one would (having ruled out relationist approaches) expect him to have argued that some kind of appearance is mind-immanent, while some other kind is not. But he does not do this. Instead, he says that even empirical objects – i.e. phenomenal substances which interact with other substances and the like – are nothing over and above representations (or, better, the content thereof). Kant asserts that ‘that which we call outer objects are nothing other than mere representations of our sensibility’ (A30/B45; emphasis mine);Footnote 32 the understanding prescribes laws to nature because appearances are mere ‘representations of things’ (B164) which constitute an object ‘that is merely in us’ (A129); ‘external things – namely, matter in all its forms and alterations – are nothing but mere representations, i.e., representations in us’ (A371–2). This is a hard bullet to bite for those who think that Kant accepts mind-transcendent appearances.
It is occasionally suggested that the term ‘representation’ is sometimes used to refer not to mental items, but to represented objects (see Collins Reference Collins1999: 72; Longuenesse Reference Longuenesse2008: 27–8). But, to the best of my knowledge, there has been virtually no textual evidence provided for this claim in the literature; and I think there is in fact no evidence.Footnote 33 Kant even says (with his own emphasis) that appearances ‘are not in themselves things, but rather nothing but representations, and they cannot exist at all outside our mind’ (A492/B520; translation modified). Some interpreters might think that Kant’s use of the term ‘in us’ is meant to point out that mind-transcendent appearances are mind-dependent. However, it is one thing to admit that appearances are mind-dependent, but quite another thing to establish that mind-dependent things are at the same time objects outside our mind. Kant makes a distinction between two senses of ‘outside us’:
But since the expression outside us carries with it an unavoidable ambiguity, since it sometimes signifies something that, as a thing in itself, exists distinct from us and sometimes merely something that belongs to outer appearance, then in order to escape uncertainty and use this concept in the latter significance – in which it is taken in the proper psychological question about the reality of our outer intuition – we will distinguish empirically external objects from those that might be called external in the transcendental sense, by directly calling them things that are to be encountered in space. (A373)
What we learn from this is that appearances are empirically outside us and that things in themselves are transcendentally outside us. What we do not learn is whether appearances in space are mental or non-mental. Being in space is compatible with existing in our mind only (if the objects are ‘in space’ merely in the sense of being represented in space), and the empirical distinction between objects in us and outside us merely tells us whether objects are only in time or also in space. It is thus natural to expect that the transcendental distinction between objects in us and outside us is about being in and outside our mind. That would conform nicely to Kant’s concepts of reflection of the inner and outer (A265–6/B321–2). This is itself a transcendental distinction; as Longuenesse has pointed out, it concerns determinations only belonging to a substance and those which are either relational or themselves relations (Longuenesse Reference Longuenesse1998: 140–7). On this account, appearances are in us because they are nothing but the mental content of our representations, and representations in turn are inner determinations of the soul.Footnote 34 The following passage from the Fourth Paralogism suggests that this is indeed the transcendental sense according to which appearances are ‘in us’:
For one cannot have sensation outside oneself, but only in oneself, and the whole of self-consciousness therefore provides nothing other than merely our own determinations. (A378)
Representations are our own determinations, thus they are transcendentally in us. Perhaps further arguments are required to make that identification more plausible. But at least it is clear by now that Kant’s distinction between two senses of ‘outside us’ does not give support to the claim that there are mind-dependent, but mind-transcendent appearances.
5.2. Apparent Evidence to the Contrary
Despite Kant’s claim that empirical objects are representations, one might object that there is clear textual evidence for mind-transcendent appearances. But, even worse for those who think that Kant believes in them, there is no decisive textual evidence for them.Footnote 35 One might perhaps think that the notion of an object in space already establishes that there are mind-transcendent appearances. However, this would only be true if space itself existed outside our mind. But we have seen in the last subsection that this is doubtful, to say the least. It is much the same with ‘empirical realism’. Though many commentators consider Kant’s use of this term as a proof for mind-transcendent appearances,Footnote 36 it only means that there are objects in space and time (see A27–8/B43–4, A35/B52). Whether those objects are in or outside our mind is not entailed by that term. The same holds good as to the ‘scientific realism’ often attributed to Kant. While there is no reason to deny that he is a scientific realist, it can still be a matter of controversy whether his scientific realism concerns mind-external objects or rather appearances (or the way things appear to us).
Perhaps the most straightforward way to argue that Kant distinguishes between mind-immanent and mind-transcendent appearances is to point out that he explicitly distinguishes between appearance as apparentia and phaenomenon.Footnote 37 The force of this distinction may seem even stronger by virtue of the fact that Kant refers to the latter as things in themselves ‘in an empirical sense (A29/B45).Footnote 38 However, Kant does not say that the distinction between two kinds of appearances constitutes a distinction between mind-immanent and mind-transcendent objects. He says instead that apparentia is ‘that which precedes the logical use of the understanding’, whereas phaenomenon is the object of reflected experience (De Mundi 2: 394).Footnote 39 Thus apparentia presents us with things as they appear to our senses, whereas phaenomenon is ‘thought of as like (wie) things in themselves’ (PM, 20: 269; my translation).Footnote 40 A further distinctive aspect is that apparentia contains secondary qualities such as colours, whereas phaenomenon does not (see A45/B62–3). All this is compatible with phaenomenon being a kind of mental rather than mind-transcendent appearance. The distinction between apparentia and phaenomenon thus provides no argument in favour of mind-transcendent appearances.
At the same time, the distinction between two kinds of mind-immanent appearances helps us to overcome some apparent difficulties with a phenomenalist interpretation. It has been objected against such an interpretation that appearances have a past and future (Allais Reference Allais2004: 662; Allison Reference Allison2004: 41–2), that phenomenal substances interact causally (Allais Reference Allais2004: 662–3), that Kant distinguishes between primary and secondary qualities (Langton Reference Langton1998: 146ff.; Allais Reference Allais2004, Reference Allais2007) and that there are physically real but unobservable entities such as ‘magnetic matter’ (A226/B273; see also Langton Reference Langton1998: 143ff.; Allais Reference Allais2004: 662; Allison Reference Allison2004: 40–1). Bearing in mind the distinction between apparentia and phaenomenon, none of those objections is compelling. It is true that the past is not given in intuition; however, the past can be thought in appearance as phaenomenon. It is true that phenomenal substances bear a relation of causal interaction to one another. Here the argument might be that forces and perhaps substances are not given in intuition. But why could not appearances as phaenomena interact with each other? It is also true that secondary qualities are given in intuition, though physical objects lack them. However, I have already indicated that, as phaenomena, physical objects can be the content of thoughts despite being different from mere perceptions. And finally, it is true that Kant allows for unobservable entities or qualities. But he says at the same time that they can be derived by means of the analogies of experience (see A225/B273), so they can be thought even though they are not present in intuition.
It might still be objected that Kant distinguishes between ‘subjective’ and ‘objective’ appearances. For example, he distinguishes between the ‘subjective sequence of apprehension’ and the ‘objective sequence of appearances’ (A193/B238). One might argue that the latter is a sequence of mind-transcendent appearances. But there are various meanings of the subjective–objective distinction. To be sure, one possibility is that ‘subjective’ refers to mental states, whereas ‘objective’ refers to real objects outside our mind. But there are other possible meanings. One of them is between subjective and objective representations. According to the famous ‘step-ladder’ passage, sensations are subjective whereas intuitions and concepts are objective (see A320/376–7). This is an inner mental distinction between representations referring to the subject and those referring to objects. If read this way, the subjective sequence of apprehension could be a sequence of subjective representations, while the objective sequence could be a sequence of objective representations.Footnote 41
Marcus Willaschek (one of the few to try to provide textual evidence for mind-transcendent appearances) suggests that the concept of an ‘object of representation’ already indicates that the object is not a representation, but something outside the representation, i.e. a real object outside the mind (Willaschek Reference Willaschek2001: 682–3). However, he neglects that ‘object of representation’ might be read as referring either to the presented object precisely as content in our representation or to a represented object outside representation.Footnote 42 Hence, without further qualification, an ‘object of experience’ can still be understood as a mind-immanent appearance. To my mind, this approach can be applied to all cases in which Kant explicitly uses ‘object of representation’ or comparable expressions. Thus I see no reason why ‘all appearances in general, i.e., all objects of the senses’ (A34/B51), ‘appearances as objects of perception’ (B207), ‘appearances, as things or objects’ (A189/B232) or ‘appearances as possible objects of experience’ (A206/B252) should be taken as referring to anything other than contents represented in representations, as opposed to having a ‘representational’ meaning as referring to represented objects outside representation.
However, it must be granted that Kant sometimes uses the expression ‘object of representation’ in the ‘representational’ sense, as in ‘[a]ll representations, as representations, have their object’ (A108). But the represented object is not for this reason appearance, as is indicated shortly after:
However, these appearances are not things in themselves, but themselves only representations, which in turn have their object, which therefore cannot be further intuited by us, and that may therefore be called the non-empirical, i.e., transcendental object=X. (A109)
Since the object of representation is non-empirical here, it is mind-transcendent, to be sure, but hardly a mind-transcendent appearance. In other cases, representations can themselves be the represented object (see A108), so the represented object is then mind-immanent. Thus formulations such as ‘objects of the senses’ (Bxvii) or ‘things as objects of experience and the very same things as things in themselves’ (Bxxvii) are consistent with appearances not being mind-transcendent objects of representation.
Still, I owe an explanation as to how things in themselves can be represented objects. One might have the worry here that this is impossible, given the restrictions Kant imposes on human knowledge. It would take us too far afield to get into the details here. My short response is twofold. Firstly, there is textual evidence for the claim that Kant identifies the object of representations with things in themselves, at least in a certain context. He says that object and representation are not ‘similar’ (Prol., 4: 289–90; Refl. 6314, 18: 616), by which he means that the object does not have the same qualities as those through which we represent them.Footnote 43 But above all, things in themselves as the represented objects fit most naturally into his ‘formal idealism’. According to ‘formal idealism’, we do not cognize things as they are constituted in themselves, but in the forms of space and time.Footnote 44 Kant is thereby explicitly denying that the represented objects are in space and time and states that the latter only belong to our cognition of them. Once again, however, I do not assert that all empirical knowledge is knowledge of the appearing things.
If appearing things are objects of representation and identical to things in themselves, it of course raises the question how knowledge of them is possible. My response is thus, secondly, that we do not have determinate but only indeterminate knowledge of them. We know that there is something that appears to us in a certain way, but we only know how it appears to us and not how it is in itself. Readers who have qualms about any knowledge of things in themselves will find this unpersuasive. We have seen, however, that Kant thinks appearances require the existence of things in themselves, and he clearly states their existence elsewhere.Footnote 45 So there is no problem about cognizing things in themselves as long as I only have a weak kind of knowledge of them. To be sure, more needs to be said to make my point fully convincing, which unfortunately cannot be done here.
It must be admitted that there are also passages which are, if taken literally, incompatible with my interpretation:
For in the appearance the objects, indeed even properties that we attribute to them, are always regarded as something really given, only insofar as this property depends only on the kind of intuition of the subject in the relation of the given object to it then this object as appearance is to be distinguished from itself as object in itself. (B69; emphasis mine)
Nevertheless, if we call certain objects, as appearances, beings of sense (phaenomena), because we distinguish the way in which we intuit them from their constitution in itself, then it already follows from our concept that to these we as it were oppose, as objects thought merely through the understanding, either other objects conceived in accordance with the latter constitution, even though we do not intuit it in them, or else other possible things, which are not objects of our senses at all, and call these beings of understanding (noumena). (B306; emphasis mine)
But although this is at odds with my interpretation, it is also at odds with those passages that, as I hope to have shown, speak against an identity of appearances with things in themselves. So there is a textual divide; however, the textual evidence supporting an identity is clearly in the minority. One reason for such passages might be that Kant is sometimes a bit sloppy in his formulations. Both sentences have a complicated structure (perhaps in German even more than in the translation), so it should come as no surprise if Kant is not as precise in these passages as one might wish.
However, Kant’s linguistic sloppiness seems to go hand in hand with systematic sloppiness. Speaking about ‘appearances’ rather than ‘appearing objects’ or ‘objects as they appear’ is perhaps not only easier to write, but also to think. We should note here that this imprecision comes in degrees. The formulation that ‘we can have cognition of no object as a thing in itself, but only insofar as it is an object of sensible intuition, i.e. as an appearance’ is already slightly imprecise (Bxxvi; see also PM, 20: 269). Kant does not say that appearances are identical to things in themselves here, but the expression ‘as appearance’ (instead of ‘as it appears’) could easily be understood as if this precisely were the case. Also, Kant’s claim that space and time ‘apply to objects only so far as they are considered as appearances, but do not present things in themselves’ (A39/B56) does not contain any claim about the identity of appearances with appearing things. In the same vein, when Kant says that we have to consider noumena in the negative sense ‘without this relation to our kind of intuition, thus not merely as appearances but as things in themselves’ (B307), no identity claim is made. To be sure, none of the passages mentioned in this paragraph fits in easily with the denial of mind-transcendent appearances. But they are recognizably the result of both linguistic and systematic imprecision and do not reflect what I take to be Kant’s considered view.
6. Conclusion
I have argued that the seeming conflict between two-world and two-aspect interpretations can be resolved once one distinguishes between appearing objects and appearances. Appearing objects are identical to things in themselves, but appearances are not. The two-aspect distinction refers to the distinction between intrinsic properties of a thing and those properties a thing has in relation to us, the latter being that of appearing, ultimately grounded in the relation of affection. The two-world distinction, on the contrary, refers to the distinction between real, mind-transcendent objects and mind-immanent appearances as the contents of certain of our representations. My reconciling strategy would be anodyne, however, if Kant at the same time allowed for mind-transcendent appearances, that is, appearances as anything other than the contents of representations. But as the preceding section has shown, there is no room for them in Kant’s philosophy. As a result, Kant’s transcendental idealism ends up as a kind of phenomenalism. However, there are notable differences to usual forms of phenomenalism, which await further discussion elsewhere.Footnote 46
