HENRY E. ALLISON | Kant’s Transcendental Deduction. An Analytical-Historical Commentary | Oxford: Oxford University Press, 2015


By Lucy Allais

Henry Allison’s monumental commentary aims to provide a critical analysis and evaluation of the Transcendental Deduction of the Categories, using both analytical and historical approaches. The Transcendental Deduction is the notoriously difficult, complex core of the Critique of Pure Reason, an argument so controversial that there is not even agreement in the literature on what its conclusion and premises are supposed to be, never mind on how the argument is supposed to work. The core of Allison’s book consists of a detailed, sentence by sentence analysis of the text of the Deduction. Before getting into the Deduction itself, Allison carefully discusses the development of Kant’s thought prior to the Critique with respect to issues relevant to the topic of the Deduction—what Allison calls Kant’s analytic metaphysics as well as his model of cognition—with the aim of showing what illumination can be found by tracing the path Kant followed in developing his position.

There are chapters devoted, respectively, to Kant’s work in the 1760s, the Inaugural Dissertation, and the so-called Silent Decade (the latter including a detailed discussion of Kant’s work in relation to Tetens). Prior to the core of the book, which consists of two chapters on each of the A and B Deductions, there is a chapter discussing the so-called Metaphysical Deduction and the introductory section of the Deduction. And between the chapters on the A and B Deductions there is a chapter on relevant parts from works written between the two editions of the Critique. As should be clear from this description, the range of works covered by Allison is extensive; however, the discussion is focused because the many texts he deals with are examined with a specific concentration on the ideas in them that are relevant to Kant’s concerns in the Deduction. Part of Allison’s achievement is to chart the course of the development of Kant’s thought on this specific question. The book is also impressive in the exceptionally careful as well as sympathetic attention Allison gives to the details of the argument; it is clearly a labour of love, and a fitting culmination of decades of Allison’s work on Kant’s Critical philosophy. Evidently, my discussion cannot attempt to cover all the topics dealt with in Allison’s extended and detailed commentary, and I shall here focus on the discussions of the Deduction itself.

Before getting into some of the details of the argument, I note three background assumptions that are important to Allison’s approach to the Deduction, as well as similarities and differences with respect to my own approach in relation to these. The first is that both in his discussion of the development of Kant’s thought and in his discussion of the Deduction itself, Allison emphasises the centrality of Kant’s underlying view of sensibility and understanding as two generically distinct cognitive powers, each with its own set of a priori forms. I am in full agreement with Allison about the importance of this distinction for understanding the Deduction. In my view, a common way in which the Deduction is misread follows from failing to take sufficiently seriously the difference between sensibility and understanding and the corresponding difference between intuitions and concepts in Kant’s system. Commentators who take this approach sometimes see Kant’s concern in the Deduction as being with showing the categories to be conditions of objects being given in intuition. As I see it, this results in running together the concerns of the Deduction with those of the Aesthetic, and failing to see what is most interesting in both.

A second background point of agreement is that we both think that the aim of the Deduction is not to address Cartesian scepticism (which Kant aims to do in the Refutation of Idealism), but rather, as Allison puts it, to respond to ‘concept empiricism’ which holds that all the materials we need for empirical cognition are derived from experience (p. 276), or, as I put it, to respond to the sceptic about a priori concepts, who thinks we can account for empirical cognition without any concepts that contain necessity and universality. Allison sees these two points being linked, with the concern about the application of a priori concepts following from the sharp distinction between sensibility and understanding. As Allison reads Kant, the “cognitive spectre” he is concerned with—the worry that “appearances could […] be so constituted that the understanding would not find them in accord with the conditions of its unity” (A90–1/B122–3)—is a direct result of his separation of the contributions to cognition of sensibility and understanding. Allison refers to this as the spectre of lack of cognitive fit or of cognitive dissonance (p. 9). I agree with Allison that Kant is concerned to respond to a sceptic about a priori concepts and that he wants to show that appearances do fit with the conditions of the unity required by the understanding, though I am less sure that this concern follows directly from the separation of the cognitive roles of sensibility and understanding. Allison points out that it does not follow from the fact that whatever is given in intuition necessarily accords with the a priori forms of intuition that it also conforms to the conditions of the understanding and that a further argument is needed in order to establish this result (p 108); this seems right. What I find less obvious is that the mere distinction between sensibility and understanding leads to the ‘cognitive spectre’ of cognitive dissonance; however, I do not think this disagreement has much further impact on our respective readings, since we are in broad agreement about what it is that Kant is trying to show.

The final crucial part of the background for approaching the Deduction is transcendental idealism. Allison and I have different understandings of this position as well as differing views on the relation between it and the argument of the Deduction. I have defended what I call a moderate metaphysical interpretation of the Deduction (see Allais 2015), while Allison sees transcendental idealism as a primarily methodological or metametaphysical position. A further ostensible difference is that while I have suggested that at least a central part of the argument of the Deduction does not depend on transcendental idealism, Allison thinks transcendental idealism is key to the argument’s success. However, here it seems to me that the disagreements are less than they at first appear, because of our different readings of what the position of transcendental idealism amounts to. In particular, Allison and I agree that the argument of the Deduction does not require transcendental idealism understood as a metaphysical position (the way I read the position), and are in agreement that it does require some of Kant’s methodological concerns (the way Allison reads the position). Further, the part of the argument for which Allison brings in transcendental idealism is the final stage of the argument, for which I agree that transcendental idealism, including on my stronger reading of the doctrine, may be necessary. I return to some of these issues after noting some of the details of Allison’s discussion of the argument.

Because of the exceptionally difficult and unclear nature of the argument of the Deduction, many scholars trying to make sense of it end up imposing onto it an argumentative structure, and possibly also philosophical concerns, in ways which may ignore details of the text that do not fit with the argument they want to find there. Allison’s book does not do this. Rather, he carefully goes through and pays attention to all the details. Further, his approach to Kant is throughout refreshingly sympathetic—unlike many commentators, Allison thinks most of Kant’s arguments work, and that his points are good. I give a few examples of details which other commentators brush over, to which Allison pays careful attention, and which I think are important for understanding the Deduction.

The A90–1/B123 passage quoted briefly above in describing the spectre the Deduction aims to address is an example of a passage with respect to which many commentators treat it as if Kant simply cannot mean exactly what he says here. Kant says that

the objects of sensible intuition must accord with the formal conditions of sensibility […] but that they must also accord with the conditions that the understanding requires for the synthetic unity of thinking is a conclusion that is not so easily seen

and he says that

appearances could after all be so constituted that the understanding would not find them in accord with the conditions of its unity […]. Appearances would nonetheless offer objects to our intuition, for intuition by no means requires the functions of thinking.

Those commentators who have what I call conceptualist readings of intuition—who see Kant as arguing in the Deduction that the application of the categories is necessary for objects to be given to us in intuition—dismiss this passage as something Kant is not actually asserting despite the fact that the way the passage is written is that Kant asserts that being given objects in intuition is independent of the functions of thinking. Allison notes and addresses a concern conceptualists about intuition may have about this passage, which may lead them to think that Kant cannot mean what he says here. The concern is whether the term ‘object’, as used in this passage, is compatible with the “‘critical’ conception of object for which he argues in the body of the Deduction” (p. 189). The possible worry, as he sees it, is that if the “objects of possible experience are to be considered as the intentional correlates of a rule-governed unification of sensible data by the understanding”, this might be thought not to be compatible with objects being given in intuition as appearances without any contribution of the understanding. In other words, since Kant is going to go on to argue in the Deduction that nothing can be an “object of experience” (A93/B126) or “an object for me” (B138) without the application of the categories, can he really mean what he says when he says that objects of sensible intuition can be given independently of the categories? Allison’s preferred solution is to say that the term ‘object’ is ambiguous, and that Kant does not always use it in the full Critical sense but also in a commonsense, non-technical way to refer to what we ordinarily think of as an object. He argues that there is no contradiction in the thought that objects understood in this way might not conform to the functions of the understanding. This seems right to me, though I think the point could be made even more simply. Since, in my view, Kant’s central concern in the Deduction is with cognising objects, I read his talk about something being “an object to me” as being about what is required for me to cognise an object as an object, as opposed to simply being presented with an object as a result of its affecting my senses, and this point does not require seeing the term ‘object’ as ambiguous.

Allison sees the passage as highlighting the idea that Kant aims to demonstrate the necessary applicability of the categories to appearances “in spite of the radical separateness of the sensible and the intellectual conditions of cognition” (p. 191), which seems to me exactly right, and to have extensive implications for the argument of the Deduction. For example, conceptualists about intuition may read the threefold synthesis Kant discusses in the A-Deduction as being about something that produces intuition, and therefore take the section to mandate conceptualism about intuition, but an alternative which fits at least as well with the text is that this three-part synthesis is something that is done to intuition. After all, it is the manifold in an intuition and the manifold of intuition that Kant talks about synthesising. Similarly, Allison says that Kant’s point in this section is that “the imagination has the task of unifying the perceived sensory data in a way that makes possible its subsequent conceptualization without itself being a mode of conceptualization” (p. 264). Conceptualists about intuition may point to Kant’s claim (in discussing the synthesis of recognition in the concept at A103) that without the synthesis of recognition reproduction would be in vain; they take this to mean that we would not be presented with an object without this synthesis. However, Allison points out that Kant here claims that without the above-mentioned form of consciousness reproduction would be in vain not that it would be impossible, and further, that the point is that it would be in vain in the sense of not leading to cognition (p. 217).

In his discussion of the threefold synthesis Kant says that

[e]very intuition contains a manifold in itself, which however would not be represented as such if the mind did not distinguish the time in the succession of impressions on one another; for as contained in one moment no representation can ever be anything other than absolute unity. (A99)

As Allison points out, what is crucial here is to distinguish between, on the one hand, being presented with something that in fact contains a manifold and, on the other, representing it as or taking it as a manifold (p. 207). Allison says that

Kant should not be read as maintaining that in a moment we intuit only a miniscule portion of time that constitutes an “absolute unity”, a temporal analogue of Berkeley’s minimum visibile. His point is rather that, if we (mistakenly) ignore the successive way in which the manifold is given and focus instead on what might conceivably be contained in a single moment, we would never have the manifoldness that requires a synthesis in order to be represented as a manifold. (p. 210)

He argues that by “ein Augenblick” Kant did not understand a minimal extent of time, but rather a limit that is not really a part of time at all. It does not seem to me that we need to deny that a moment is a minimal extent of time to make sense of this passage. There are a number of alternative readings, one of which is that recognising an object as a complex of parts or a complex subject of predicates goes beyond what is given in intuition and requires conceptually governed synthesis, and this involves seeing it as an object existing over time.

Rather than seeing the Deduction as concerned with the conditions of intuition (as conceptualists about intuition see it), Allison emphasises Kant’s concern with the conditions of thought. Kant says that it would constitute a sufficient deduction of the categories if we could prove that by them alone an object can be thought. Allison says this might seem to make the task of the deduction too easy, since an object can be thought through any concept that does not contain a contradiction. He suggests that the way to avoid this worry is to realise that what Kant is concerned with is not simply thought but thought of a really possible object. It is not clear to me how this resolves the worry. Kant holds that objects can be known by us to be really possible only if it is possible for them to be given in intuition, so the suggestion seems to be that Kant is saying that it would be a sufficient deduction of the categories if we can prove that by them alone an object that can be given to us in intuition can be thought. The initial worry raised by Allison seems to recur: objects that can be given in intuition can be thought through all sorts of concepts that do not contain a contradiction. My preferred solution to this worry is that when Kant talks about an object being thought in the Deduction he has something quite specific in mind: bringing an object given in intuition under an empirical concept. In other words, he wants to show that the categories are conditions of empirical concept application, and says that if he can show this, it will be a sufficient deduction of them.

As mentioned above, while Allison and I ostensibly disagree about how dependent the argument of the Deduction is on transcendental idealism, this disagreement may not run deep, given our different readings of transcendental idealism, and the extent to which he sees it as a primarily methodological rather than metaphysical position, which is compatible with my reading of Kant’s central concerns in the Deduction. Allison points out numerous points in the Deduction that support reading the argument as concerned with the conditions of cognition rather than with presenting a metaphysical position. For example, in his transition to the Deduction Kant explicitly states that though it “does not produce its object as far as its existence is concerned, the representation is still determinant of the object a priori if it is possible through it alone to cognize something as an object” (A92/B125). Here Kant seems to assert that the role of the categories is epistemic and not metaphysical: to make sense of Kant’s claim here we need to give an account of his reasons for thinking that only through the categories can we cognise something as an object, and then to explain a sense in which this results in a priori concepts being determinant of objects.

Allison discusses Kant’s claim at the end of the A-Deduction that

[p]ure concepts of the understanding are therefore possible, indeed necessary a priori in relation to experience, only because our cognition has to do with nothing but appearances […]. And from this ground, the only possible one among all, our deduction of the categories has been conducted. (A130)

As he says, this makes it look as if transcendental idealism grounds the Deduction, and that a Deduction of the categories would not be possible for a transcendental realist. Allison explains this by saying that if the objects of our cognition were things in themselves, either the categories would be derived from experience which would mean denying their categorial status as a priori concepts of objects in general, or they would be innate, in which case we would not be able to demonstrate that objects (understood as things in themselves) conform to them (p. 280). Crucially, Kant states that the categories are possible and necessary in relation to experience—in other words, in relation to objects of our cognition, rather than things independently of whether or not we can cognise them, and he states that this requires that the objects of our cognition are appearances. At the very least, ‘appearances’ here must mean things that are presented to us. How much further idealistic implication it has is a matter for further argument, and this is a point at which it seems to me the disagreement between Allison and me over the nature of the idealism shrinks our disagreement about the dependence of the Deduction on the idealism. Allison says that Kant’s empirical realism is incompatible with transcendental realism since the latter implicitly appeals to a God’s eye view of things in themselves as the normative standard governing cognition (p. 286) and that for a transcendental realist what counts as an object or objective is a matter of cognitive psychology and has no normative import, which means that a transcendental deduction could be possible only on the basis of transcendental idealism. I agree that the argument concerns the conditions of cognition of objects that can be presented to us, so does not concern a God’s eye view of things in themselves as the normative standard governing cognition or providing an account of objectivity.

I mention two further parts of the text which seems to me to be better understood in the light of something close to Allison’s methodological approach than some other approaches. Allison points out that reflecting on the need for a transcendental ground, Kant begins by suggesting that

[o]ne soon comes upon this [something] if one recalls that appearances are not things in themselves, but rather the mere play of our representations, which in the end come down to determinations of inner sense. (A101)

This passage seems to make transcendental idealism key to the Deduction, but of course that leaves open the question of how transcendental idealism should be understood. Allison argues that Kant here seems to be suggesting that if we recognise that appearances are not things in themselves, but rather mere modifications of inner sense, i.e. if we accept transcendental idealism, we can account for the necessary synthetic unity in question; whereas if we view the matter from a transcendentally realistic standpoint, according to which the necessary synthetic unity would pertain to things in themselves, we cannot do so. On a methodological or epistemic reading, the point is not to claim that appearances are a different kind of (essentially mental) object, but that appearances are things that are presented to and cognised by us, and that we have no access to things that are not presented to us or things independently of the way in which they are presented to us. Similarly, Kant says of the idea of the transcendental object that

this object must be thought of only as something in general = X, since outside of our cognition we have nothing that we could set over against this cognition as corresponding to it. (A104)

Allison says that this characterisation is not so much a solution to the problem as a reformulation of it and that it amounts to a radical internalisation of the problem of knowledge. He explains this as saying that rather than worrying about whether our cognition corresponds to an inaccessible an sich reality, the issue becomes the grounds on the basis of which the mind may attribute objectivity to its synthesis of representations or, alternatively, distinguish between what pertains to the object and what holds merely for the subject. On an epistemic reading, saying that we have to do only with the manifold of our representations and that the X which corresponds to them is nothing to us is not a phenomenalist claim about objects being merely mental entities that therefore need to be unified, but rather a claim about the nature of cognition and our cognitive access to objects. On this reading, Kant’s concern is not with what it takes for an object to exist as a unified complex or unified subject of properties, but what it takes for us to be in a position to represent an object in this way; this is a cognitive problem and one which would be solved by pointing out that objects do in fact exist as unified complexes. Similarly, the so-called ‘binding problem’ in perceptual psychology—the problem of how the brain manages to organise the disparate sensory data to represent unified particulars—is not solved by pointing out that the objects of perception are in fact unified; the problem is how we manage to represent unified objects.

As I have mentioned, Allison is one of the Kant commentators most sympathetic to Kant—he is convinced that there are arguments and that the arguments do work. Further, his detailed account of the argument sticks closely to Kant’s terminology and phrasing, avoiding worries about anachronistically imposing contemporary concerns on the text in an effort to make sense of it. A possible downside of this is that it seems to me that at some points this results in the book having less to offer those who are not already very sympathetic to the text, those who do not share Allison’s basic orientation to understanding the point of Kant’s ‘transcendental’ turn and those who have not thought themselves as far into the text as Allison has and who are looking for more explanation of key concepts without using Kantian terms, or indeed translation out of Kantian terminology into something that might give persuasive reasons to adopt Kant’s approach to someone who has not already done so. In response to this Allison may point out that there is a certain amount of a philosopher’s approach that one has to accept to follow their arguments, and the extent to which one can convince someone who does not accept your starting approach and assumptions is limited. However, in my view it is desirable, where possible, to try to find ways of reading Kant that provide arguments to those who have not already thought themselves into his framework, and, where possible, to translate Kant’s terms in a way that makes it possible to motivate his concerns to those who are not already immersed in and convinced by them. I shall give a few examples of places where it seems to me that Allison’s account requires further explanation, because these, in my view, pick out precisely the crucial points at which Kant’s argument requires further explanation—suggesting that if Allison’s account does not fully explain the Deduction the problem is not in his interpretation, but in his accepting that Kant has given us sufficient means to reach his conclusion, rather than leaving us needing to reconstruct a Kantian argument that would reach Kant’s conclusion.

Allison says that the a priori concepts of an object in general are needed because

empirical concepts are rules for thought of an object as of a certain kind and, as such, presuppose that what they are applied to in a judgment is already thought as an object, understood as something that stands over and against the representation of it

and that,

[a]ccordingly, in addition to the empirical concepts that determine an object as falling under a certain description, judgment, i.e., discursive cognition, presupposes concepts that function to relate the empirical representations to an object in the first place. These concepts are the categories and what enables them to relate representations to objects is just that they are concepts of an object as such. (p. 194)

This seems to me to be accurate in sticking closely to things Kant says, and I agree that Kant does hold that empirical concept application requires the application of the categories, but it does not seem to me to go very far to explain Kant’s claim and to give a reason for thinking that cognition requires a priori concepts that an empiricist would find compelling. The fact that applying empirical concepts to objects presupposes that what they are applied to is already thought of as an object does not show that the latter requires a priori concepts, and that we could not, rather, have a conception of objects consisting entirely of empirical concepts that enables us to think of things as objects in the way needed for empirical concept application.

Allison points out as central to the Deduction the claim that “[e]very necessity has a transcendental condition as its ground”, as well as Kant’s saying that

[w]e find, however, that our thought of the relation of all cognition to its object carries something of necessity with it, since […] the latter is regarded as that which is opposed to our cognitions being determined at pleasure or arbitrarily rather than being determined a priori, since insofar as they are to relate to an object our cognitions must also necessarily agree with each other in relation to it, i.e., they must have that unity that constitutes the concept of an object. (A104)

I am entirely in agreement with Allison both about the importance of these claims for the argument and that the difficulty in interpreting this single-sentence paragraph largely concerns the sense of ‘necessity’ to which Kant appeals. What makes this hard to understand is that Kant is here talking about empirical judgements which relate predicates to objects contingently, but also says that they have some kind of necessity, and that this is required to regard them as related to objects. As Allison reads this, to say that in the cognition of an object the relation between representations is “determined a priori” can only mean determined by an a priori rule, which, Allison argues, indicates that Kant had in mind a normative necessity, such as is imposed by a rule (p. 222). Allison further explains this by saying that to say that the necessity has a normative sense means that while it is neither causally nor logically necessary to unify one’s representations in such a way that they relate to an object, one must do so insofar as one takes them as so related (p. 249).

This does not seem to me to explain why this unification requires representing necessity, which is the key point to be explained. Why and in what sense does representing predicates as unified with subjects in empirical concept application, or as taking them to be related to an object, require representing them as necessarily so unified? Allison later explains this further in terms of the idea of concepts as rules. He says that

the conception of concepts as rules also underlies Kant’s assertion of a place for necessity in empirical cognition and is the reason why he insists that the possibility of such cognition requires a transcendental, i.e. non-empirical, ground. We have seen that the necessity in question applies specifically to concepts qua rules and that it is normative rather than logical (or causal) in nature. The essential point is that as rules concepts are themselves normative in the sense that they involve a kind of epistemic right to expect that whatever objects fall under the rule contain the properties that are expressed in the concept-rule and to require that other cognizers, who possess the concept, will agree with one’s judgment. (p. 237)

But it is not clear what argument has been provided to an empiricist (the sceptic about a priori concepts who both Allison and I take to be the target of the argument) as to why seeing concepts as rules and as normative and as involving an expectation that others will agree with one’s judgement requires necessity. Allison suggests that conceptual breakdowns are only intelligible against the backdrop of a thoroughgoing rule-governed synthetic unity of appearances, which is reflected in the conception of a single experience of which all particular experiences are regarded as instances (pp. 237–8). The empiricists may respond that conceptual breakdowns would not be intelligible against a background of total chaos, but it is not clear why what is required is not simply a contingent order and uniformity, as opposed to necessary unity. This, I take it, is precisely what we need to make sense of to understand Kant’s point in the Deduction. In fairness to Allison, it is not clear what argument Kant gives for this, as well as other points where one might wish for further argument and explanation.

The last point I discuss concerns the nature of transcendental idealism and its role in the Transcendental Deduction. One of Allison’s most detailed discussions of transcendental idealism in this book comes in in relation to his analysis of the final part of the A-Deduction where Kant talks about the understanding as the law-giver of nature, something Allison calls an idealistic thesis. Allison represents Kant as arguing for idealism at this point, and presents the argument as being that the unity of nature must be necessary and a priori certain, together with the claim that we cannot establish necessary a priori claims unless they are grounded in our minds. It is said to follow from this that we ourselves bring into the appearance the order and regularity that we call nature. Thus, Allison says, Kant derives the mind-dependence of the unity of nature from its a priori status and that this follows from the core Critical principle that “we can cognize of things a priori only what we ourselves have put into them” (Bxviii). One way of reading this ‘principle’ is as an expression of transcendental idealism, but if this is the case it surely cannot be used as a premise in an argument for transcendental idealism. On the other hand, if it is a premise in an argument for transcendental idealism it surely needs extensive independent argument. The way it is presented in the text is in a summary overview of the position in the B-preface, which makes it seem to me that we should expect argument for it, and not simply be able to draw on it as something established at the point it is presented. Further, as Allison recognises, we need an argument for the premise that the unity of nature (its order and regularity) is necessary and therefore cognised a priori (or, since Kant defines nature in terms of lawfulness, why appearances have to be unified in the way that makes them constitute nature). This means that the account of the argument for idealism given at this point seems to leave most of the key premises unargued for. Allison argues that Kant rejects the possibility that appearances do not “fit into a coherent whole of human cognition” on the grounds that such a state of affairs would not be apperceivable (p. 275). But why would it not be apperceivable? This is what Kant needs to argue for, and the argument of which we need to give an account. Allison says that unless “appearances were unifiable in such a way as to constitute a common order of nature, they could not be thought together as constituting a whole by a single cognizer” (p. 275). But this does not give an argument for thinking that a common order of nature requires necessary unity, which is the point the empiricist contests. The empiricist could grant that appearances need to be unifiable in a way that constitutes a common order of nature without agreeing that this requires necessary unity.

Allison explains Kant’s claim that the understanding legislates to nature by appealing to a distinction between levels of laws—a nested hierarchy with at the base particular empirical laws that we learn from experience, and at the top the transcendental laws delineated in the analytic of principles which are derived from the understanding—which, he argues, means that the understanding is indirectly the source of all the laws of nature. But this simply states that the system must allow that empirical laws (which are not up to us) are somehow related to the higher laws of the understanding (which are in some sense up to us) without explaining how this works, or explaining what it means for the understanding to legislate the higher laws to nature. These are of course very difficult issues to explain. Making sense of the argument of the Deduction requires giving reasons for and explaining the claim that appearances could not be unified in the way needed for empirical thought about objects unless they fall under the categories, and explaining the result of the Deduction requires explaining the sense in which Kant holds that the understanding legislates to nature. It is not clear to me that Allison’s careful restatements of Kant’s views and arguments with respect to these points show that Kant has convincing arguments here; my suspicion is that Kant’s difficult, complex yet sometimes frustratingly abbreviated argument requires supplementing with a reconstructive argument of the sort which is not Allison’s aim in this book (and which, in my view, has not been written), so the book cannot be faulted for not achieving something which, if I am right about this, cannot be achieved using the method and aims of this commentary. These worries therefore are not a criticism of the success of the book in its own terms, as an unsurpassed and highly sympathetic commentary on one of the most difficult arguments in Kant.


Allais, L. (2015), Manifest Reality. Kant’s Idealism and his Realism (Oxford: Oxford University Press).


© Lucy Allais, 2019.

Lucy Allais is Henry E. Allison Endowed Chair in the History of Philosophy at University of California, San Diego and Professor of Philosophy at Wits University, Johannesburg. Her work focuses on Kant’s philosophy and topics in moral philosophy. In 2015 she published Manifest Reality. Kant’s Idealism and his Realism (Oxford UP).