Escribir el título del artículo


Two Functions of Perception in Kant


Hemmo Laiho*

University of Turku, Finland





Kant uses terms translatable as ‘synthesis’ and ‘perception’ in different ways in different contexts, which suggests that there are different kinds of synthesis and perception. I propose that there are two main basic functions of perception according to Kant: that of singling out a thing and that of getting perceptually informed about the configuration of the thing’s perceptible features. I argue that the first function is not dependent on the kinds of syntheses Kant analyzes in the Critique of Pure Reason but grounds any such synthesis. I also argue that if singling out a thing is considered to involve synthesis, then the term ‘synthesis’ is identified with a unification of sensory information itself, which is not of much consequence for the transcendental philosopher. The paper also aims to clarify what ‘manifold’ consists in and what the starting point of the ‘synthesis in apprehension’ might be.



Intuition, Perception, Synthesis




Kant käyttää ‘synteesiksi’ ja ‘havainnoksi’ käännettävissä olevia termejä eri tavoin eri konteksteissa, mikä viittaa erityyppisiin synteeseihin ja havaintoihin. Esitän, että Kantin mukaan havainnolla on kaksi perustavaa funktiota: olion yksilöinti ja olion havaittavien ominaisuuksien konfiguraatiosta perille pääseminen. Väitän, ettei ensin mainittu funktio riipu senkaltaisista synteeseistä, joita Kant analysoi Puhtaan järjeen kritiikissä, vaan perustaa tuollaisten synteesien mahdollisuuden. Lisäksi väitän, että jos olion yksilöinnin katsotaan edellyttävän synteesiä, synteesillä tarkoitetaan tällöin aisti-informaation itsensä yhdistämistä, jonka tarkastelun voidaan katsoa olevan jokseenkin epäolennaista transsendentaalifilosofille. Artikkelissani pyrin myös selventämään, mistä ‘moneus’ koostuu ja mikä voisi olla ‘synteesin apprehensiossa’ lähtökohta. 



Havainto, Intuitio, Synteesi


1.    Introduction


I take it that much of what Kant says about sense perception implies that outer perception has two basic cognitive functions according to Kant:


(PERCEPTION 1) To single out a thing, where the thing is understood as a “whole”, e.g. that very thing over there.


(PERCEPTION 2) To get perceptually informed about the configuration of the thing and its perceptible features or “parts”, e.g. this rectangular-shaped heavy thing.  


In addition to capturing something deeply true about the phenomenology of perception, the distinction can help us understand Kant’s theory of cognition in many ways. In particular, the distinction can be used to clarify what ‘manifold’ might consist in and what the starting point of the ‘synthesis in apprehension’ in perception might be—a task Norman Kemp Smith deemed impossible (Kemp Smith 2003, p. 97). More generally, the distinction makes one wonder what exactly is the theoretical role of perceptually locatable and traceable particular things in Kant’s philosophical system and what after all does the supposedly necessary linkage between synthesis and perception mean.[1] To that end, this paper challenges the idea that all kinds of perception, PERCEPTION 1 included, is dependent on the kinds of syntheses Kant analyzes in the Critique of Pure Reason. As I hope to be able to show, Kant the philosopher is not even supposed to explain the kind of unification of sensory information that might be required for PERCEPTION 1.[2] This is to say that while some such processes could just as well be called syntheses, and be of great importance in psychological explanations of perception, such processes are not Kant’s real concern in his critical philosophy.

I begin, in Section 2, with preliminary remarks on perception and synthesis. In Section 3, I claim that synthesis, as used in the context of critical philosophy, refers to a cognitive function that stands on the capacity to perceptually single out particulars, not the other way around. The primacy of such an “intuitive” capacity over synthesis is further illuminated in Section 4 by showing that unlike synthesis, intuition does not proceed from “parts” to the “whole.” Finally, in Section 5, I make a tentative proposal how all of this might relate to the Transcendental Analytic in general and to the transcendental deduction of the categories in particular.


2.    Preliminary remarks on perception and synthesis


Kant’s use of the term ‘perception’ (Wahrnehmung, also perceptio and Perzeption) is ambiguous. For instance, perception can indicate entertaining mere sensation (Empfindung, also sensatio) (e.g. A170/B212; see also Falkenstein 1995, p. 161) or even just an inner feeling (e.g. KU, 5:289; cf. Reid 1969, II.XVI). Such uses of the term differ greatly from the ordinary use—as in the perception of a house, for example. Moreover, in Kant’s texts, we find several other terms that refer to the kind of phenomenon that can be regarded as perception or species of perceptual representation. An obvious example would be observation (Beobachtung). Though not as self-evident, apprehension (Auffassung, Fassung, also Apprehension) would be another. The more general term cognition (Erkenntnis, also cognitio) should be mentioned as well, and not solely because Kant sometimes calls cognitions perceptions, but because his use of this term is not exactly consistent either.[3] Last but not least, there is empirical intuition (empirische Anschauung), which Kant sometimes identifies with and sometimes separates from Wahrnehmung (see e.g. Prol, 4:283, 300; V-Met-L2, 28:557; B162).

This brief terminological reminder already suggests that Kant does not have a single psychological process or phenomenon in mind when he speaks of perception, or what we might want to simply call perception. Thanks to the ambiguity we can also quite safely take some freedom in using the term ‘perception’ in the Kantian context to indicate a wider phenomenon than, say, what he might mean by Anschauung or Wahrnehmung in some specific context. (Even if at times these two terms seem to track the distinction between PERCEPTION 1 and PERCEPTION 2.) In any case, it is clear that Kant recognizes and distinguishes several kinds of perceptual representation according to different cognitive functions served by perception (broadly conceived). As already suggested and formulated in the Introduction, I take the main basic cognitive functions of perception to be that of singling out a thing (PERCEPTION 1) and that of getting perceptually informed about the configuration of the thing’s perceptible features (PERCEPTION 2).

To count as genuine perceptions, all kinds of perception require that the perceiver is affected, and the manner the perceiver is affected correlates both with more or less quantifiable features such as shapes and qualitative features such as colors in given empirical circumstances. To this extent, perceiving is beyond our control, as we cannot typically choose what to perceive or how a certain thing appears to us. In this sense, we are controlled by sensibility (Sinnlichkeit) i.e. the faculty responsible for the receptivity of sensory matter, sensations, intuitions and, indeed, for the fact that objects are given to us (e.g. A19/B33).

It should be equally clear that with the notion of synthesis Kant wants to emphasize that representing also requires conjoining, combining, placing or putting together separate elements that cannot do the job individually, and that the final cognitive product is a unification of those elements. An obvious example of such a connecting activity would be a thought according to the logical form S is P. For example, when we think of a bridge built of stones, or rather that the bridge is built of stones, we combine the representations of bridge and stone in a certain manner, thereby achieving a unified representation of a stone bridge. This is a result of an act of predication, and also a result of subsumption, as long as we at least implicitly represent the stone bridge as belonging to a wider class of objects. In such cases, we genuinely understand how things are, which requires more than just sense perception, namely the capacity of judgment. To put it differently, we then have experience (Erfahrung) in the full sense of the term, which is more than just perception, namely “perception that is understood.” (Kant 2005, p. 172; HN, 17:664) Here, I think, we are already cognitively beyond mere PERCEPTION 1 or PERCEPTION 2.[4]

The idea that perception itself involves synthesis is not so clear. To begin with, Kant distances perception (or at least some kinds of perception) from judgment, as he identifies thinking with judging (e.g. A69/B94) while at the same time clearly regarding intuition as a distinct kind of representation that can take place even prior to thinking (B67; B132; HN, 17:620). What is more, Kant links perceptual synthesis with the sensible faculty of imagination, not with the intellectual faculty of understanding (Verstand) (see e.g. A78/B103; A120; B151; V-Lo/Wiener, 24:701-2). Indeed, Kant thinks that imagination (Einbildungskraft, also imaginatio) is no less than necessary for perception. As he points out in a footnote in the so-called A-deduction:


No psychologist has yet thought that the imagination is a necessary ingredient of perception itself. This is so partly because this faculty has been limited to reproduction, and partly because it has been believed that the senses do not merely afford us impressions but also put them together, and produce images of objects, for which without doubt something more than the receptivity of impressions is required, namely a function of the synthesis of them. (A120n)


Here, Kant seems to imply that the role of imagination in perception is to produce images by placing impressions together. However, this is something he never really explains, which makes it difficult to say exactly what he means by this remark. Moreover, he refers to psychologists, and perhaps much of the rest of what he says in that context is not grounded in a psychological point of view, which is why I tend to think that the footnote is merely an off-topic parenthesis.

Be that as it may, the idea that perception itself involves synthesis can mean quite different things. For example, explaining perceptual synthesis in terms of representing concrete particular elements together (such as stones) to obtain a (more) unified view on things (such as bridges made of stones) differs drastically from explaining perceptual synthesis in terms of connecting—to put it somewhat anachronistically—sense-data or bits of sensory information received by the senses to have them processed into representations of stones and bridges in the first place. More generally put, the latter type of explanation reflects the idea that “the three-dimensional visible world is something the mind synthesizes from scratch” (Waxman 2005, p. 162) whereas the former type of explanation does not start from scratch at all, but from particular things such as stones and bridges.

Interestingly, as we will see in the next section, both ideas can be found in the Kantian corpus. At the same time, we find a tension: Should we understand both PERCEPTION 1 and PERCEPTION 2 in the way that suggests a kind of thoroughgoing synthesis, it would mean that PERCEPTION 2 simply denotes a further synthesis, whereas the kind of explanation, which starts from particulars, not only clearly preserves a distinction between perceiving the bridge as such and perceiving it in some more detailed manner, but makes something distinguishable as the basic unit of perception.

The quoted passage from the A-deduction may seem to better support the idea that even PERCEPTION 1 only becomes possible via a synthetic operation that puts together sensory information itself. While some Kant commentators remain ambiguous on this, some commentators do think, more or less explicitly, that Kant’s views of synthesis are to be understood along such lines (e.g. Falkenstein 1995, p. 123; Strawson 1966, p. 32; Kitcher 2011, p. 107; Westphal 2004, p. 89). However, as already suggested, I do not think that this is Kant’s view. At least it is not Kant’s view in the context of transcendental philosophy, though something like it may be a plausible empirical psychological explanation of sense perception according to Kant as well. Instead, as far as Kant’s philosophical account of perception goes, synthesis builds upon the capacity to establish relation to individual empirical things, which is to say that perceptual synthesis starts from particular things or their individuatable features antecedently given in intuition. In a word, synthesis stands on PERCEPTION 1. My take on the matter can be illustrated in a table form roughly as follows:


Transcendental philosophy / Kant in CPR

Unification of bits of sensory information

. . . . . . . → {O}



Singling out a thing (“that”)




Representing perceptual elements together

{ a, b, c } = O

Apprehensive synthesis


Understanding an object of perception

“O is P”

Judgmental synthesis

‘Perception’ (broadly conceived)


3.    What is apprehension targeted at?


To make better sense of what we just suggested, let us ask: What exactly is synthesized in perception? What does ‘manifold’ stand for? As I see it, the best place to look for answers is Kant’s notion of apprehensive synthesis.[5] As Kant describes this kind of synthesis in the second edition Critique,


by the synthesis of apprehension I understand the composition [Zusammensetzung] of the manifold in an empirical intuition, through which perception, i.e., empirical consciousness of it (as appearance), becomes possible. (B160)


The explication does not reveal what exactly Kant thinks is composed or put together in the synthesis of apprehension which he sometimes also seems to take as necessary for perception. It seems, however, that he must have in mind one of the two alternatives examined in the previous section. That is to say that either a totally unstructured sensory manifold is being apprehended, in which case the synthesis in question starts more or less from scratch, or that the manifold itself is prestructured in some fundamental sense, being more like a set of possibilities for further connecting activities rather than the ground zero of every possible perceptual representation.

The “scratch” reading builds on the idea that we synthetically build the visible world out of the motley material provided through affection. In that case, PERCEPTION 1 too must be regarded as constituted by simpler elements put synthetically together—not totally unlike a set of pixels that constitute a desktop icon on a computer screen. In this kind of model, the explanation of perception begins from the “putting-together” of the sensory stuff itself. Kant himself speaks of such an idea in his lectures:


With every manner in which we are affected there are two parts: matter, i.e., the impression of sensation, and form, i.e., [the] manner in which impressions are unified [vereiniget] in my mind. Otherwise I would have millions of impressions but no intuition of a whole object. (Kant 2001, p. 154; V-Met/Mron, 29:800)


Basically, Kant points out here how the input of impressions or sensations, or the sensory stuff itself, must be unified for PERCEPTION 1 to be possible.

            Such a unification operation can well be dubbed synthesis.[6] Then again, when we perceptually represent an object, we do not represent millions of impressions hanging together, but rather this or that thing. In Kant’s terms, we have an “intuition of a whole object.” This is also the level Kant typically speaks at: namely, at the level of conscious representation of objects. The kind of sensory unification processes implied by the quote, however, are clearly inaccessible to us. It can thus be questioned from the outset that such sensory-level “syntheses” would belong to Kant’s transcendental philosophy, where he typically links synthesis with a set of a priori concepts, also known as the categories, and works under the presupposition that that which contributes to a priori cognition “cannot remain hidden from us” (A13/B26). This is not to deny unconscious representations from Kant’s view of the human mind but to press the point that, for Kant, the philosophically relevant explanation of perception hardly consists in giving an account of subconscious information processing or the like. To put it differently, Kant was probably not very interested in how perception actually develops into the kind of phenomenon it is (cf. Hatfield 1990, p. 107).

There is a further reason to believe that the kind of “psychological” model that deals with the sensory information itself does not draw upon those explanatory resources Kant resorts to as a transcendental philosopher. To cite a passage from the Transcendental Analytic:


Things [Dinge] are simultaneous insofar as they exist at one and the same time. But how does one cognize that they exist at one and the same time? If the order in the synthesis of apprehension of this manifold is indifferent, i.e., if it can proceed from A through B, C, and D to E, but also conversely from E to A. For if they existed in time one after the other (in the order that begins with A and ends at E), then it would be impossible to begin the apprehension at the perception of [Wahrnehmung vor] E and proceed backwards to A, since A would belong to past time, and thus can no longer be an object of apprehension. (A211/B258; my emphasis)


Here, as Kant notes how apprehending can sometimes proceed in two different ways, he speaks not only of things—which suggests that these As and Es are individual objects of perception—but also of perceptions of As and Es, which suggests that in apprehending we go from one perception of a thing to another, not from one sensory input to another. Moreover, the quote indicates how the manifold appears as organized into things we can perceive and be aware of—what we dubbed distinguishables in the previous section. Just as importantly, the passage suggests that perceptual relation to things is not itself constituted by apprehension. Rather, apprehensive synthesis builds upon the capacity to single out things (PERCEPTION 1), which suggests more generally that the cognitive subject must have (at least some kind of) perceptual access to things prior to apprehensive synthesis. Further, the quoted passage indirectly evidences that if some special kind of “synthesis” does precondition PERCEPTION 1, it is not the kind of synthesis that would have a core significance in Kant’s critical philosophy, assuming apprehensive synthesis would be the sole candidate for such a thing.


4.    On parts and wholes in intuitive representing


Why does Kant introduce the notion of intuition in the first place? Contrary to what commentators (e.g. Kemp Smith 2003; Smit 2000) have suggested, I would say that because Kant was driven by the idea that not all representing is tied to advancing from parts to the whole. In the Critique, Kant suggests this much as early as the Transcendental Aesthetic. As he explicates the notion of intuition, he attributes to concepts a cognitive limitation of having to proceed from partial representations, while an entire representation can be given in intuition (A32). What is more, in his lectures on logic, where he calls intuitions “thoroughly determinate cognitions” (Kant 2004, p. 597; Log, 9:99), Kant seems to repeat in different words what he has been trying to say all along: namely, that perceptual reference to an individual thing can be fixed (bestimmt) through, and only through, intuition.

            Later in the Critique, Kant emphasizes that every intuition is an extensive magnitude. This emphasis might give a reason to think that intuitions cannot be immediate in the sense that they would not require synthesis from parts up. As Kemp Smith has put it, basing his reading on B202-3, “extensive magnitude cannot be apprehended save through a ‘synthesis of the manifold,’ a ‘combination of the homogeneous’” (Kemp Smith 2003, p. 94). However, I find it more plausible that Kant should be read as holding that this is so only insofar as a certain kind of representation is concerned, namely, precisely the kind of representation by which the cognitive subject actually apprehends or represents the extensive magnitude of a thing to herself. And, as a matter of fact, in the Critique, Kant does remark in a footnote:


We can intuit an indeterminate quantum as a whole, if it is enclosed within boundaries, without needing to construct its totality through measurement, i.e., through the successive synthesis of its parts. (A426/B454n)[7]


Though seemingly marginal, this small quote from the First Antinomy is particularly revealing for our purposes. As the wider context implies, the real problems of the Critique­ concern synthesis in terms of measurement, understood as “repeated addition of units to each other” (A428/B456). This in turn suggests that it is this kind of operation that requires synthesis that proceeds from parts (plurality) to the whole (totality), not intuition as such. It is also worthwhile to note that such an activity belongs to thinking (A428/B456); and as just suggested, it is thinking, not intuiting, that is tied to the cognitive limitations of concept-use and thus necessitated to a progression from parts.

When applied to sense perception—which should be totally legitimate since perception shares the same a priori structure with every humanly possible sensible representation—the primacy of intuition over synthesis suggests that we can perceptually single out a thing without having to successively go through its composition. It also seems then that the ultimate reason for Kant’s introduction of the notion of sensible intuition is to underscore the human capacity to represent things instantaneously, albeit merely sensibly since we are not capable of intellectual intuition (B308). Consequently, it is plausible to take intuition as such as something that is not constituted by the kind of connecting activity Kant analyses in certain parts of the Critique. Rather, intuition or PERCEPTION 1, and with it the possibility of being presented with particular things, understood as “wholes”, grounds the possibility of further “synthesizing” activities, such as PERCEPTION 2.

            This reading is also strongly supported by the way Kant understands sensible or aesthetic clarity and distinctness in intuition. The basic idea goes roughly as follows.[8] We represent clearly when we are able to distinguish objects from each other (Anth, 7:137-38). A clear representing remains indistinct until we become conscious of the manifold, which can be understood in terms of becoming aware of object’s composition (Log, 9:34; Anth, 7:138). In both cases, the kind of awareness involved can also be merely intuitive or aesthetic (Anth, 7:135; EEKU, 20:227*; ÜE, 8:217*). I take this to mean that we can perceive an individual thing in a detailed manner without a conceptual or “logico-predicative” understanding of what the thing is or how its parts are related to the whole. For example, if you do not recognize the thing before you as a house, even if you can perfectly distinguish it from the garden, you do not recognize the function of the rectangular-shaped reflective surfaces either. Yet, you have an aesthetically or intuitively distinct representation of the house. 

            The following quote from the Jäsche Logik is quite illuminating in this regard: 


We glimpse a country house in a distance. If we are conscious that the intuited object is a house, then we must necessarily have a representation of the various parts of this house, the windows, doors, etc. For if we did not see the parts, we would not see the house itself either. But we are not conscious of this representation of the manifold of its parts, and our representation of the object indicated is thus itself an indistinct representation. (Log, 9:34)


As I read it, the quote suggests that to be conscious of the house as a house, or “that the intuited object is a house”, we must be aware of its parts as parts of the house. This is because houses do have doors and windows, and to be aware that the thing is a house is to understand at least that much of such things, whereas if we just saw a big rectangular block in the distance, we could not be said to be consciously representing a house. In this sense, the representation of the house becomes both aesthetically and logically (conceptually) clear and distinct only if we understand how the kinds of parts or features we have distinguished contribute to house-representations in general. More importantly for our purposes, however, the quote suggests a perceptual situation in which “we are not conscious of this representation of the manifold of its parts”. Yet we are looking at that rectangular block in the distance. In other words, though we do not go through the combined features that make a house, and thus are perhaps unable to see that it is indeed a house we are looking at, we nevertheless single out that very thing in the distance. In all, this suggests that the manifoldness as such (in the sense which would indicate an awareness of the object’s composition) does not need to be represented in order to perceive something in the first place. On the contrary, an aesthetically indistinct representation of a thing seems to suffice for PERCEPTION 1.

Thus, on my reading, composing a manifold equals finding out and being aware of a structure of distinguishable parts in some plurality. This involves PERCEPTION 2. However, that does not mean that the object in question, understood merely as this thing here or that very thing over there, is not perceived if no representation of its composition takes place. On the contrary, supposing we have a sufficiently clear aesthetic representation of the thing, we succeed in singling it out. In a word, we succeed in perceiving it in the sense of PERCEPTION 1, which is at the same time necessary for PERCEPTION 2, but not the other way around.


5.    What about the transcendental deduction?


To reiterate, perception has two fundamental cognitive functions according to Kant. The first—PERCEPTION 1—is to establish relation to individual empirical things also dubbed above as distinguishables. The second function—PERCEPTION 2—is to achieve a distinct view on the object perceived. Accordingly, it is one thing to explicate perception in terms of establishing a relation to an object, and quite another to explicate perception in terms of figuring out how the object is composed. Kant himself explicitly refers to this distinction in the transcendental deduction of the categories, as he differentiates between intuiting a house and making the intuition of the house into a perception of the house (B162).

            In above, we also made the further claim that PERCEPTION 1 does not depend on synthesis, at least not in the sense Kant analyzes this capacity as a transcendental philosopher. Granted, this claim is problematic in light of Kant’s transcendental deduction of the categories. To my mind, that is mostly so because the success of the deduction seems to depend on the idea that “synthesis, through which even perception itself becomes possible, stands under the categories” (B162), which in turn can be taken to imply that all kinds of perception depends on categorial synthesis (see also e.g. Allison 2004; Gomes 2014, pp. 14-15).

            It goes without question that the text suggests a deep interconnection of perception and synthesis. Yet the possibility that Kant’s argument is only meant to hold for perception in a very specific sense of the term cannot be ruled out (see also Tolley 2013, p. 124). As we have seen, there is textual basis for taking apprehensive synthesis or PERCEPTION 2 as drawing on the capacity to single out things, and thus as dependent on PERCEPTION 1 and the possibility of being confronted with particulars, not the other way around. If this is so, then we are in the position to claim that whatever Kant has in mind with the link between synthesis and perception, there are reasons to believe that he is not trying to explain what can be taken as the most basic cognitive task of sense perception: namely, perception in the sense of PERCEPTION 1. We have also already refuted the possibility that the kind of sensory unification processes alluded to above could be a function of the categories.

Read against this, it appears quite plausible that the deduction deals with a kind of perception, or a certain function of perceptual representation, not with perception tout court. Accordingly, the claim that all kinds of perception must be subject to categorial synthesis appears to be too broad to give a sufficiently clear view of Kant’s intentions besides the obvious, namely that there must be some kind of affinity between perceptions, syntheses and categories for us to be able to think and judge about sensible objects. In all, given the context-sensitivity of the terms ‘perception’ and ‘synthesis’, which can be used to refer to various kinds of cognitive functions or capacities, it should not come as a surprise that some of them were not Kant’s concern in the deduction.[9]

What is more, when one reads the B-version of the deduction carefully, one discovers that the very idea that perception must be subject to the categories comes with a qualification:


Now since all possible perception depends on the synthesis of apprehension, but the latter itself, this empirical synthesis, depends on the transcendental one, thus on the categories, all possible perceptions, hence everything that can ever reach empirical consciousness, i.e., all appearances of nature, as far as their combination is concerned, stand under the categories [...]. (B164-65; my emphases)


To begin with, and most importantly for our purposes, the emphasis on the synthesis of apprehension leaves PERCEPTION 1 out of the picture. One should also take special note of Kant’s reference to appearances, that is, to undetermined objects of empirical intuition (A20/B34), as opposed to, say, “data of sense themselves unconnected and separate” (Strawson 1966, p. 32). Accordingly, whatever combination exactly means in this context, it is something that happens in relation to (conceptually undetermined) objects, not in relation to the possibility of the objects (or appearances) as such. Moreover, since Kant now uses the German word Verbindung, which connotes intellectual synthesis (Verstandesverbindung) rather than any of the sensible ones (see B151), he might actually have a very specific kind of synthesis in mind in the quoted passage: namely, presumably, the kind of judgmental synthesis we referred to in Section 2. In addition, the dependence relation in question, or the exact meaning of the verb ‘to stand under’, remains unclear. It could simply mean, for instance, that everything that is given to us must fall under the scope of the categories. Besides, one should not ignore the larger context of the passage, which is not sense perception as such, but the far more abstract issue of how the categories are supposed to “prescribe laws […] to the nature as the sum total of all appearances” (B163).[10]

            The overall goal of the Transcendental Analytic is equally revealing in this respect. As Kant explicitly puts it, he aims to analyze thinking and understanding in its pure use (A64-65/B89-90). This can be read as indicating that he has already analyzed intuiting and sensibility in the Transcendental Aesthetic (see also A62/B87). Above all, one should not forget Kant’s stark intuiting-thinking distinction at this point. In fact, important as it is, Kant reminds the reader of this distinction in the introductions to both the Transcendental Logic and the Analytic (A64/B89). In the former, before embarking on his analysis of pure understanding, Kant even emphasizes that the entire analysis will depend on the condition “that objects are given to us in intuition” (A62/B87). He also says that he will scrutinize “merely the part of our thought that has its origin solely in the understanding” (A62/B87; my emphases). Accordingly, assuming that Kant stays true to his goals, he is not going to present new conditions on sensible representation at this point.[11]

            In other words, as we enter the Analytic, it is quite likely that Kant is building on the premise that empirically given particular things are present to the cognitive subject. That this is the starting point from which Kant’s account of thinking and understanding carries on is also suggested in the third Critique, where Kant explicitly says that the role of sense perception is to offer (darbieten) particulars to the understanding (KU, 5:186). Such a starting point is very different from the one in which the mind receives a stream of sensory information from which it shapes actual representations. This is not to say that Kant thought that such a picture would be blatantly wrong. Rather, the point is that, as a transcendental philosopher, he is not supposed to provide such explanations.[12] In particular, we should keep in mind that transcendental philosophy is about a priori cognition, the possibility of which does not rely on outer affection. By contrast, the investigation of actual perceptual processes would require empirical study of the mind and/or bodily functions. Such studies would also include explanations involving sensory systems that operate both independently of and beyond the reach of our will and consciousness, and thus beyond the scope of a priori cognition as well.

            Put differently, Kant’s account of thinking seems to assume that quite much is provided to the mind independently of understanding. Tellingly, at one point in the first edition Critique Kant goes so far as to claim that not only motion, extension, and impenetrability, but also composition (Zusammenhang) can be provided to us through outer sense (A358). A circle closes. If Kant thinks that composition (presumably in some specific sense of the term) can be given sensibly, then it might just as well be that he did not intend to mean that the perception of ordinary objects, such as colorful billiard balls with numbers on them, is dependent on combination as he used the term in the Analytic. It might also be that Kant really means the representation of combination in the Analytic (see B130). In that case, the deduction would be more about the conditions of some such higher-order representation rather than combination or synthesis as such (see also Allais 2017). However that may be, the Analytic appears to stand “methodologically” on PERCEPTION 1 in any case.

            Even in that case, one tricky problem remains. How can one secure the compatibility between perceptions and the categories, given that we must also be able to think about what is sensibly present to us, if the deduction does not really deal with perception in the pritimive sense elaborated above? A detailed answer being beyond the scope of this paper, here is a short answer. The key Kantian idea is that since every humanly possible sensible representation rests on the a priori intuitions of space and time, everything that takes something spatiotemporal as its object necessarily has an element to it that is cognizable a priori. As most clearly evidenced by geometrical construction, the understanding is able to use the otherwise blind imagination for a priori purposes. Imagination in turn is constrained by the same a priori intuitions of space and time. Consequently, understanding has access not only to imagination, but, via imagination, to a priori sensibility as a whole. Now, if the understanding can access the a priori ground of sensible representation in pure a priori representation, then the understanding must surely be able to think of a thing that accords with that same ground, that is, shares the structural features with a priori sensible representation, even if the understanding had nothing to do with the genesis of the perceptual representation of the thing. This, I think, is how understanding ultimately relates to sense perception according to Kant, which explanation, moreover, is separable from any such explanation that deals with the unification of sensory information itself.[13]



Allais, L. (2009), “Kant, Non-Conceptual Content and the Representation of Space”, Journal of the History of Philosophy, 47, pp. 383-413.

Allais, L. (2017), “Synthesis and Binding” in A. Gomes & A. Stephenson (eds.), Kant and the Philosophy of Mind: Perception, Reason, and the Self, Oxford: Oxford University Press, DOI:10.1093/oso/9780198724957.003.0002

Allison, H. E. (2004), Kant’s Transcendental Idealism. An Interpretation and Defense, New Haven: Yale University Press.

Burge, T. (2009), “Perceptual Objectivity”, The Philosophical Review, 118, pp. 285-324.

Carl, W. (1998), ”Die B-Deduktion” in G. Mohr & M. Willaschek (eds.), Immanuel Kant: Kritik der reinen Vernunft, Berlin: Akademie-Verlag, pp. 189-216.

Falkenstein, L. (1995), Kant’s Intuitionism. A Commentary on the Transcendental Aesthetic, Toronto: University of Toronto Press.

Ginsborg, H. (2008), “Was Kant a Nonconceptualist?”, Philosophical Studies, 137, pp. 65-77.

Golob, S. (2014), “Kant on Intentionality, Magnitude, and the Unity of Perception”, European Journal of Philosophy, 22, pp. 505-528.

Gomes, A. (2014), “Kant on Perception: Naïve Realism, Non-Conceptualism, and the B-Deduction”, The Philosophical Quarterly, 64, pp. 1-19.

Griffith, A. (2012), “Perception and the Categories: A Conceptualist Reading of Kant’s Critique of Pure Reason”, European Journal of Philosophy, 20, pp. 193-222.

Grüne, S. (2009), Blinde Anschauung. Die Rolle von Begriffen in Kants Theorie sinnlicher Synthesis, Frankfurt am Mein: Klostermann.

Hanna, R. (2005), “Kant and Nonconceptual Content”, European Journal of Philosophy, 13, pp. 247-290.

Hanna, R. (2008), “Kantian Non-conceptualism”, Philosophical Studies, 137, pp. 41-64.

Hatfield, G. (1990), The Natural and the Normative. Theories of Spatial Perception from Kant to Helmholtz, Cambridge: The MIT Press.

Kant, I. (1992-), The Cambridge Edition of the Works of Immanuel Kant, edited by P. Guyer and A. Wood, Cambridge: Cambridge University Press.

Kant, I. (2000a), Critique of Pure Reason, edited and translated by P. Guyer & A. Wood, Cambridge: Cambridge University Press. (A/B)

Kant, I. (2000b), Critique of the Power of Judgment, edited by P. Guyer, translated by P. Guyer & E. Matthews, New York: Cambridge University Press.

Kant, I. (2001), Lectures on Metaphysics, edited and translated by K. Ameriks & S. Naragon, Cambridge: Cambridge University Press.

Kant, I. (2004), Lectures on Logic, edited and translated by J. M. Young, Cambridge: Cambridge University Press.

Kant, I. (2005), Notes and Fragments, edited by P. Guyer, translated by C. Bowman, P. Guyer & F. Rauscher, Cambridge: Cambridge University Press.

Kant, I. (2009), Anthropology from a Pragmatic Point of View, translated by R. B. Louden, in Anthropology, History, and Education, edited by G. Zöller & R. B. Louden, Cambridge: Cambridge University Press.

Kemp Smith, N. (2003), A Commentary to Kant’s ‘Critique of Pure Reason’, New York: Palgrave Macmillan.

Kitcher, P. (2011), Kant’s Thinker, Oxford: Oxford University Press.

Laiho, H. (2012), Perception in Kant’s Model of Experience, Turku: University of Turku.

Laiho, H. (2018), “Two Kinds of Distinctness, Two Systems of Representation”, in V. L. Waibel, M. Ruffing & D. Wagner (eds.), Natur und Freiheit: Akten des XII. Internationalen Kant- Kongresses. Berlin: de Gruyter, pp. 2683-2690.


Laiho, H. (2019), “Types of Representational Content in Kant”, Kantian Journal, 38, pp. 31-54.


Land, T. (2006), “Kant’s Spontaneity Thesis”, Philosophical Topics, 34, pp. 189-220.

McLear, C. (2015), “Two Kinds of Unity in the Critique of Pure Reason”, Journal of the History of Philosophy, 53, pp. 79-110.

Reid, T. (1969), Essays on the Intellectual Powers of Man, Cambridge: The M.I.T. Press.

Rohs, P. (2001), Bezieht sich nach Kant die Anschauung unmittelbar auf Gegenstände?” in V. Gerhardt & al. (eds.), Kant und die Berliner Aufklärung, Berlin: De Gruyter, pp. 213-228.

Rosenberg, J. F. (2005), Accessing Kant. A Relaxed Introduction to the Critique of Pure Reason, Oxford: Oxford University Press.

Smit, H. (2000), “Kant on Marks and Intuition”, The Philosophical Review, 109, pp. 235-266.

Strawson, P. F. (1966), The Bounds of Sense. An Essay on Kant’s Critique of Pure Reason, London: Methuen.

Tolley, C. (2013), “The Non-Conceptuality of the Content of Intuitions: A New Approach”, Kantian Review, 18, pp. 107-136.

Watkins, E. & Willaschek, M. (2017), “Kant’s Account of Cognition”, Journal of the History of Philosophy, Volume 55, pp. 83-112.

Waxman, W. (1991), Kant’s Model of the Mind. A New Interpretation of Transcendental Idealism, Oxford: Oxford University Press.

Waxman, W. (2005), Kant and the Empiricists: Understanding Understanding, Oxford: Oxford University Press.

Westphal, K. (2004), Kant’s Transcendental Proof of Realism, Cambridge: Cambridge University Press.






Descripción: kant_ctk_TRANSPARENTE1.png

* Hemmo Laiho received his PhD in philosophy from the University of Turku, Finland, in November 2012, upon completion of his award-winning thesis Perception in Kant’s Model of Experience. Since then, he has held various research and teaching positions in the Department of Philosophy at the University of Turku. In 2016, Laiho was a Fulbright Scholar in the Department of Philosophy at the University of California, San Diego, USA. Currently Laiho holds a research position at the Turku Institute for Advanced Studies. His areas of specialization include history of philosophy, especially Kant, philosophy of mind and perception, and aesthetics. His recent articles include: “Kant on Representing Negative States of Affairs” (Topoi, 2016) and “Two Kinds of Distinctness, Two Systems of Representation”, in V. L. Waibel, M. Ruffing & D. Wagner (eds.), Natur und Freiheit (Berlin: deGruyter, 2018). Laiho is an editorial board member of the Amsterdam University Press series Crossing Boundaries: Turku Medieval and Early Modern Studies and a board member of the Philosophical Society of Finland. E-mail: [email protected]

[1] In the Critique of Pure Reason, Kant at times seems to regard synthesis as absolutely necessary for perception (see esp. A120n). Recent interpretations heavily reliant on this idea include Ginsborg 2008; Gomes 2014; Land 2006. See also e.g. Griffith 2012; Grüne 2009. For contrast, see e.g. Allais 2009; Hanna 2005; McLear 2015. ‘A/B’, followed by page numbers, refers to the two editions of Critique of Pure Reason (1781/1787; Kant 2000a). Otherwise, Kant’s works are cited using the abbreviations and volume and page numbers of the Academy edition of Kant’s works (Kant 1900-). The English translations are from The Cambridge Edition of the Works of Immanuel Kant (Kant 1992-).

[2] Lucy Allais has recently argued along similar lines (albeit from different direction, so to speak) in her article “Synthesis and Binding” (Allais 2017). For earlier contributions that bring up the same idea, see e.g. Allais 2009; Laiho 2012, 135-6 and passim.

[3] Compare, for instance, A320/B376-77 and Log, 9:33 with A106. The former texts imply that intuition can be cognition as well, whereas the latter text implies that cognition always requires a concept. In a similar fashion, whereas B146 implies that cognition necessarily requires a sensible component, at A795-796/B823-24 and A817/B844 Kant refers to cognitions of pure reason, which suggests that not all cognition requires a link to sensibility after all. This trend is reinforced in Kant’s practical philosophy (see e.g. GMS, 4:408-9). For further clarification, see e.g. Watkins & Willaschek 2017, pp. 85-7.

[4] I cannot go into all the details here. Further questions include, for instance, what precisely is required for genuine recognition and object re-identification over time and how that might relate to the distinction between PERCEPTION 1 and PERCEPTION 2. It might also be stressed that the main issue in this paper is the theoretical status of perceptual particulars in Kant’s critical system, not the further issue whether and in what sense Kant should be regarded as non-conceptualist.

[5] Since apprehension is successive (A189/B234), the synthesis of apprehension is intimately linked to the synthesis of reproduction, and hence with time (see also A102). That said, restricting my analysis to the synthesis of apprehension should suffice for the purposes of this paper. 

[6] As we see here, though, Kant himself does not necessarily use the exact same word. This happens elsewhere, too, sometimes overlooked by translators: see e.g. Kant 2009, p. 279, where ‘synthesis’ replaces ‘Zusammensetzung’ (Anth, 7:168).

[7] The ‘aesthetic basic measure’ used in “the estimation of the magnitude of objects” (Kant 2000b, p. 135; KU, 5:251) can also be regarded as such a whole. On this, see esp. Golob 2014.

[8] For a more detailed version, see Laiho 2018.

[9] As I see it, Kant’s real concern in the transcendental deduction is experience proper, which is something that requires the capacity of judgment, and thus differs drastically from perception as such. See e.g. HN, 16:494-95 (R 2743); HN, 17:664 (R 4679); HN, 18:320 (R 5661); V-Lo/Blomberg, 24:236; V-Met/Mron, 29:798-99. See also Burge 2009, pp. 295-6; Carl 1998, pp. 206-07; Tolley 2013, p. 125.

[10] Phrases like this leave open the further possibility that not even PERCEPTION 2, when minimally construed as merely sensible non-conceptual synthesis, is Kant’s target in the deduction. This has been at least indirectly suggested by e.g. Hanna 2008; Rohs 2001; Waxman 1991; see also Laiho 2019, pp. 40-2. However, this issue is beyond the scope of the current paper.

[11] The emphasis on purity also prompts the question how much room can there be within an analysis of the pure use of the understanding for an explication of necessarily empirical representation? On top of all this, see A89/B121-22, which suggests that the Transcendental Aesthetic already includes a transcendental deduction of space and time, securing their objective validity with respect to objects given in intuition.

[12] In the above quoted passage (A211/B258), where As and Es refer to particulars, and which served as central textual evidence earlier in this paper, perhaps one could insist that the way Kant uses apprehension should be read against the context of the Analytic of Principles, whereas the deduction has only the sensory input itself to lean on. If this were the case, apprehensive synthesis as some kind of “impression glue” would also be Kant’s topic in the Critique. However, if I am right about Kant’s methodological standpoint in the Transcendental Logic, including all of the Analytic, the goals of which do not point to this direction at all, then this kind of counterargument is not very strong. Still, it should be granted that some things Kant says in the deduction are difficult, or perhaps impossible, to reconcile.

[13] I would like to thank everyone at Finno-Hungarian Seminar in Early Modern Philosophy at the Central European University in Budapest, Perception, Judgment, and Logic workshop at the University of Luxembourg, and German Philosophy Reading Group at the University of California San Diego, for their helpful comments and suggestions on the earlier versions of this paper. In addition to thanking these audiences, I would like to thank the reviewers, with special thanks to the referee who provided extensive comments on the penultimate version.

Enlaces refback

  • No hay ningún enlace refback.
Utilizamos cookies propias y de terceros para realizar análisis de uso y de medición de nuestra web para mejorar nuestros servicios. Si continua navegando, consideramos que acepta su uso.

Creative Commons by-nc 3.0 Logo

ISSN: 2386-7655


  Scimago Journal & Country Rankscopus logo


DOAJ LogoErih Plus LogoCitefactor logoredib Logo

LatIndex LogoISOC Logo MIAR Logo
MLA LogoZenodo Logo
ESCI LogoEBSCO LOGOWeb of Science