In her book which has already received wide attention, Lucy Allais provides a novel account of Kant’s transcendental idealism. She presents her interpretation as navigating between two extremes. On the one hand, deflationist readings see Kant exclusively as a realist: they either trivialize the distinction between appearances and things in themselves or deny Kant’s commitment to things in themselves altogether. On the other hand, phenomenalist interpretations hold that appearances reduce to the content of representations. If we follow Allais, both are untenable. Deflationists cannot do justice to Kant’s idealism – that is, they cannot explain why he thinks that there is something mind-dependent about appearances. Phenomenalism, on the contrary, is too subjectivist in that it ignores Kant’s realism – that is, appearances are public objects that we can directly perceive in space. Consequently, Allais thinks that any successful interpretation of Kant’s transcendental idealism must also explain his realism. It must be a ‘moderate’ one.

Allais’s moderate interpretation considers the notoriously controversial distinction between things in themselves and appearances as one between two aspects of things: the in-itself-aspect corresponds to properties that things have independent of a relation to cognizing subjects (but are epistemically inaccessible to us), while the appearance-aspect concerns properties that objects can only have in relation to cognizing subjects. Allais calls properties of the latter sort ‘essentially manifest’ – they are ‘relational, mind-dependent qualities of things which can be present in experience’ (p. 124). On her relational account of perception, intuition directly presents us objects without mental intermediaries. Allais reconstructs Kant’s argument that space and time are essentially manifest as follows: ‘Kant thinks objects that exist independently of our representing them get to be present to consciousness by affecting us, so he takes it that something that is present to consciousness without anything affecting us does not exist independently of our representing it. Since space and time are a priori intuitions, they do not exist independently of our representing them’ (p. 201). Allais combines this view with non-conceptualism about intuition: intuition presents objects without the contribution of the understanding, and while synthesis requires concepts, the task of synthesis is not to ‘bind’ intuitions, but rather to relate the understanding to objects. Crucially, Allais holds that essentially manifest properties must be grounded by properties that are not essentially manifest – this is why she thinks that we know, according to Kant, that objects have an unknowable in-itself-aspect.

There is much to like about Allais’s book: it proposes one of the most elaborate interpretations of Kant’s transcendental idealism to date. Her presentation is usually very clear and her arguments are often inspiring even for those who disagree with her. Clearly, Manifest Reality is a must-read for anyone interested in Kant’s transcendental idealism, his (meta-) metaphysics, and the conceptualism/non-conceptualism debate. What tempers my enthusiasm for this book, however, is that Allais’s interpretation faces a number of difficulties which she does not adequately address. While she devotes much attention to relatively uncontroversial points, she has frustratingly little to offer at crucial points of her argumentation.

To begin with, Allais’s framing of non-idealist and phenomenalist interpretations as two extremes that must be avoided may be rhetorically compelling in that it suggests that one has to pursue a moderate middle course. But not only is she mostly silent on competing moderate interpretations (Ameriks Reference Ameriks2011 and Rosefeldt Reference Rosefeldt2007 come to mind), it is also doubtful that her framing is adequate. While her arguments against readings that try to get rid of Kant’s idealism altogether are excellent, it is disputable whether phenomenalism ought to be avoided at all. (One could also wonder whether some version of phenomenalism could avoid extreme subjectivism.) To my mind, Allais’s arguments against phenomenalism (pp. 43–56) do not have much force. For example, she claims that Kant’s rejection of Berkeley’s idealism ‘gives us extremely strong grounds to reject any mentalized reading of appearances’ (p. 52). But it seems that Kant can reject Berkeley’s idealism and still accept another kind of ‘mentalized’ idealism. What Kant rejects is Berkeley’s view that space is a property of things in themselves and Kant in fact praises ‘the good Berkeley’ for inferring that objects in space would be impossible on this condition (B71). In Kant’s view, it is the transcendental ideality of space that allows both for the possibility of things in themselves and objects in space, which is compatible with phenomenalism. Allais’s other arguments face similar problems. She holds that the fact that appearances are in space, are directly perceived, interact causally and so on tells against phenomenalism. However, all this can arguably be compatible with a sophisticated version of phenomenalism. Kant does say that outer objects are ‘immediately perceived’ (A371). But why could phenomenalists not allow for this? Allais has no good response here.

The weakness of her anti-phenomenalist argument that we directly perceive objects in space is all the more problematic as this argument seems to carry the main burden for her relational interpretation of intuition. If this argument fails, there is not much reason to think that Kant accepts a relational account (albeit coupled with idealism). And there are independent problems for her views on intuition. For example, Allais cannot do justice to Kant’s epistemology of perception. She denies that, according to Kant, we have to infer the existence of objects outside the mind from the given in perception; the objects are just directly given in intuition. As appealing as this view may be to those who do not want to ascribe a ‘Cartesian’-style ‘veil of perception’ to Kant, he actually accepts that there is a veil of perception – he says that ‘it always remains doubtful’ whether a given perception is caused by an outer object (A368). This can hardly be reconciled with a relationist reading. Having said this, Allais’s reconstruction of Kant’s argument for the apriority of space and time puts her in a strong position to respond to the so-called ‘neglected alternative’ problem. The putative difficulty for Kant is that he infers from the subject-dependency of the representation of space and time to the subject-dependency of space and time themselves. Allais does not face this problem because, on her relationist reconstruction, Kant’s argument proves that space and time themselves, and not merely mind-internal representations of them, are essentially manifest (p. 197). I do not think that this is the only viable response to the neglected alternative problem, but the ability to avoid it is certainly a strength of Allais’s reading.

Notwithstanding this, I do not find the conception of essentially manifest qualities very intuitive or clear. Initially, Allais sounds as if her account is that essentially manifest qualities are properties that are essentially related to the perceiving subject – for example, appearing curvy (qua seen from a certain angle), or the stick in water appearing bent (p. 115). Notably, this means that we see the stick not as bent, but as appearing bent to us. But later, it rather seems that, on Allais’s reading of Kant, essentially manifest properties are analogous to dispositions in the sense that they are ‘features of how objects would perceptually appear to suitable subjects’ (p. 123). This implies that we do not see essentially manifest qualities – the stick is not bent, not even in relation to us. Rather, seeing the stick as bent is a relation which is (on part of the object) made possible by a corresponding essentially manifest quality. These views seem incompatible, and they both invite obvious objections. Clearly, more discussion would be needed; Allais, however, dispenses with this by noting that ‘how exactly to characterise and precisely formalise this kind of mind-dependence is a question requiring much further philosophical work’ (p. 123). Yet, given the centrality of the account of essentially manifest qualities to the book as a whole, Allais might have spent more time showing how to make the account coherent.

With respect to Allais’s defence of non-conceptualism, it appears that her discussion is meant to safeguard her relational account of intuition (cf. pp. 148–50, 160). As she considers intuition (and the properties represented in it) to be subject-dependent, though, it is not clear that this is in fact necessary. Allais might well be able to argue that we directly intuit properties that are nonetheless dependent on the activity of the understanding. On such a view, the understanding somehow shapes the mind-dependent properties of objects (maybe in reaction to given sensations) which are then directly given to us in intuition. Thus, it appears that Allais could have abstained from the extremely complex conceptualism/non-conceptualism debate. Even so, her innovative interpretation of synthesis, according to which synthesis is needed in order to apply concepts to objects, is one of the most interesting parts of the book. I would like to raise two worries, though. First, the role of synthesis is not sufficiently clear. On the one hand, Allais holds that synthesis does not affect intuition, but only our cognition of objects which we base on intuition – in effect, synthesis is part of the process of thinking rather than of intuiting (pp. 274–5). On the other hand, though, she admits that, although synthesis does not produce intuitions, synthesis ‘is something that is done to intuitions’ (p. 169) – that is, synthesis unites intuitions. Allais does not say what the product of the unifying of intuitions is, but it seems that there is a conceptualist element in her account that needs more elaboration. Second, it might be wondered whether her reading is idealist enough. Kant claims, ‘The conditions of the possibility of experience in general are at the same time conditions of the possibility of the objects of experience’ (A158/B197). If the task of the understanding is only to cognize the objects of experience (and not also to somehow shape these objects), then it is not obvious how the subject-dependency of appearances could explain how we prescribe laws to natures (which is clearly Kant’s view; see A127, B164).

Finally, Allais’s interpretation of the relation between what she sees as the appearance-aspect and the in-itself-aspect of things faces difficulties, too. On her reading, the essentially manifest qualities that constitute the appearance-aspect need to be grounded by unknowable non-manifest properties (the in-itself-aspect). Allais refers to the latter as intrinsic properties and holds that, according to Kant, there are no intrinsic properties in appearance (p. 224). As Allais reads Kant, there must be intrinsic noumenal properties that ground the extrinsic properties in appearance (p. 231) – for Allais, this is why we know that things have an in-itself-aspect (p. 228). However, this is a questionable reading which, in addition, presupposes a one-object view rather than establishing it. On an alternative interpretation, Kant does not say that all things must have intrinsic properties. Rather, he thinks that, while things in themselves must have intrinsic properties, appearances only have relational properties, which shows that they are not things in themselves. If this reading is correct (as I think it is), one cannot explain the grounding of appearances by things in themselves as Allais does.

These points of disagreement should not suggest that Manifest Reality is not an important contribution. Yet, I think that it might have benefited from a somewhat different approach. Allais does consider a wide array of textual evidence, but only in the beginning so as to establish the need for a moderate interpretation. Once having reached this point, her procedure is mostly top-down: Allais develops her account of manifest properties by analogy with some contemporary theory of colours and then argues that her account fits certain passages in Kant’s texts. However, the fact that something is consistent with what Kant says does not show that he actually has this view. In my opinion, a bottom-up approach, which begins by careful evaluation of textual evidence and only then develops an account of transcendental idealism, would be more promising. Even so, Allais’s Manifest Reality should be one of the central books in Kant scholarship for the years to come.