Kant’s very particular designation of the status of space in the transcendental aesthetic serves as a major point in the development of the epistemological [1] thesis of transcendental idealism (Kant, 1929). In “The Transcendental Aesthetic” he contends, in opposition to both the rationalists and empiricists, that space is neither an object nor a relation or relations between objects in the world. Instead, Kant sees space (along with time) as a subjective condition for the very possibility of outer experience; without space and time, human beings would have no cognitive access to the world. Kant’s arguments for the status of space are of particular interest to us as they appear before the explication of time, offer more depth, and lead immediately into Kant’s first declaration of his epistemological doctrine. Kant insists that space and time (the “aesthetic”, to use his term) provide the subjective ground of experience: the necessary conditions that make experience of a world possible at all. In his early work, the Tractatus Logico-Philosophicus, Wittgenstein (1961) presents us with a competing view: that language is the subjective ground to all experience.

This essay will begin with an exposition of Kant’s argument for the designation of space and the role of the aesthetic as the a priori [2] conditions for experience, and will then move on to a discussion of whether language can be thought to supersede the aesthetic as the subjective ground for experience. Michael N. Forster’s essay “Kant’s Philosophy of Language?” is useful in illuminating how Kant’s apparent lack of interest in the philosophy of language makes the definition of his attitude toward language very difficult. Despite this, Forster’s own reading of Kant implies that language could possibly have a greater role in Kant’s epistemology than his arguments suggest in Critique of Pure Reason . In turn, the potency of the question of primacy is reduced somewhat if we agree with Forster, and we may have good reason to do so.

Kant’s arguments for the status of space are broken into two parts – the metaphysical exposition and the transcendental exposition. The main bulk of the argument is found in the metaphysical exposition, and the majority of my analysis will concern the four arguments presented therein. Sections 1) and 2) taken together explain how space cannot be thought of as an empirical object, and instead represents something a priori. In this belief, Kant undermines the position of empirical realism, whereby the objects of our experience correspond precisely with real objects in the world. Kant argues in 1) that if we are to represent to ourselves objects that appear simultaneously as outside of me and outside and next to one another, then the space in which we perceive these objects must be presupposed (Kant, 1929, p. A23/B38) . On these grounds, space cannot be observed as a relation between objects in the world, because space must exist independently of these objects for them to appear as existent within it. In support of this claim, Kant adds in 2) that we cannot represent the absence of space; however we can perfectly well represent it as devoid of objects. For this reason Kant argues that, when properly considered, space is an example of an a priori, given that it provides appearance with its very form (that of objects appearing next to one another, and in a different place to me and other objects) as opposed to being part of the appearance itself (Kant, 1929, pp. A24-25/B38-39) . Furthermore, Kant argues that if space were indeed an item of a posteriori knowledge (knowledge dependent on experience), then certain geometric principles (such as that there may be only one straight line between two points) could not be considered a universal rule. Instead, such principles would merely be the product of induction, insofar as our current experience tells us that “no space has been found which has more than three dimensions” (Kant, 1929, p. A24) .

So far, Kant has offered sufficient reasons for us to believe that firstly, space is not an object itself, as it must be presupposed prior to the objects which appear within it and secondly, that space is apparent to us a priori, in that it does not appear to us through our observations, but is a prior condition for our perception of objects to be as it is. In the third and fourth sections of “The Transcendental Aesthetic”, Kant seeks to refute the claim that space is a concept. If he can successfully refute this claim then it would seem that the only option left for the status of space is his own – that it is an immediately given a priori. Insofar as anything is considered a concept, it is a product of the understanding and a rational object. Therefore just as parts 1) and 2) are to be considered a refutation of empiricist claims regarding the status of space, parts 3) and 4) aim to refute a rationalist description of space, i.e. that it exists as a “general concept of relations of things in general” (Kant, 1929, pp. A24-A25) . As mentioned above, Kant requires that space be known a priori on the grounds of 1) and 2). In parts 3) and 4) we discover that not only must it be known a priori, but that it must also be an example of a pure intuition . What Kant means by a pure intuition is an intuition that contains no sense data (Kant, 1929, p. A20/B34) , and is therefore distinct from both sensations and concepts. The argument in part 3) runs parallel with that of part 1). Recall that in 1) Kant argues that we cannot represent to ourselves any object in space without first presupposing space itself; here Kant makes the similar claim that we cannot represent any single place (as in a single constituent part of space) without first presupposing space itself. Cognising a singular place must presuppose space itself because any cognition of a constituent part of space is explained in terms of applying limitations to our current idea of space as a whole, i.e. breaking it into parts by means of applying measurements (Kant, 1929, p. A25/B39) . Kant’s argument relies on the distinction between space itself and any place that occurs within it. A definitive place is conceptual because in cognising any such place, we distinguish a definitive part that can be thought of as in the whole from the immediately given, however we cannot say the same for space itself because we cannot assume the existence of places without space itself first being apparent to us.

At this point, a rationalist may defend their designation of space as conceptual by replying that in the above argument Kant has only proven that the existence of places presupposes the existence of space itself, much like the existence of humans presupposes the existence of primates. The terms “human” and “primate” however, both represent concepts, and the fact that one falls under the other does not provide a satisfactory reason to believe that the prior concept is for some reason another species of understanding. Kant must hereby do more to explain how space is different from other concepts.

The answer, writes Andrew Janiak, lies in Kant’s understanding of extension and intension. Briefly understood, extension refers to subordinate concepts that come under a concept, whilst intension refers to prior concepts that are present within the named concept. Hence “primate” is an intension of ‘human,’ as to call upon the concept “human” you must implicitly call upon the concept “primate”. “Primate” is within “human” according to this understanding, much like places are within space (Janiak, 2012, 11 13 ). Janiak argues that Kant’s suggestion in 4) is that if space is a concept then we ought to be able to construct it out of its many constituent parts. However, space is represented to us as having an infinite number of parts as we can always divide one or the other part further. If it were a concept, it would have infinite intensions. As Janiak points out, Kant’s argument is that while a concept can (in principle) have infinite extensions under it, it cannot have an infinite number of intensions within it, because in order for us to represent the concept to ourselves we must paradoxically call upon every one of its infinite intensions (Janiak, 2012) . Indeed, Kant argues that space is thought of in this way – as having infinite parts, “for all the parts of space coexist ad infinitum ” and therefore must be considered in terms of a priori intuition and not as a concept (Kant, 1929, p. B40) .

Following on from Kant’s arguments regarding the status of space, we now have good reason to accept Kant’s conclusion, which consists of his first mention of the doctrine of transcendental idealism. Essentially, the importance of space’s status is that in as much as it forms the subject-seated ground for all possible experience of an external world, it also reveals that we have access only to the world of appearances and not to things in themselves. We recognise this in acknowledging that empirical reality is only the reality of our experience of the world, and not of the world beyond that (Kant, 1929, pp. A26-A28/B43-B44) . The importance of space being at the very foundations of our experience in forming the conditions becomes obvious through this point in Kant’s work. The a priori intuition of space is necessary for human experience and knowledge and represents the understanding’s action upon sensation – the empirical object of our experience represents sense data that has been brought under the form of space by the understanding. Hereby if the status or the primacy of space were to be undermined, Kantian epistemology would be challenged to explain how we form an experience of the world (and thus the world as it appears to us, the phenomenal world).

As previously mentioned however, Wittgenstein’s early philosophy in the Tractatus Logico-Philosophicus (hereafter Tractatus ) suggests an alternative necessary condition of experience. According to Wittgenstein, the form of all experience must be the language through which we construct our image of the world. Therefore, whatever is outside of these means (i.e. inexpressible in meaningful language) is cognitively inaccessible to human beings. This is what is to be understood by his claim that there is truth in solipsism: “the world is my world,” the experiential world is the only world we can access, and the way we do this is through the use of language (Wittgenstein, 1961, pp. 56-57). This thesis has been called the picture theory of language, as it claims that all meaningful language functions as a somewhat logical picture of the world. If Wittgenstein is correct in his explanation of human world-forming behaviour, then language appears the most likely candidate for what we have been calling the subjective condition for experience. However, Kant’s arguments serve to show that space is the subjective ground of, and defines the form of, our outer experience. Thus, an urgent question of primacy arises, which I have stated in my title: “Does language supersede the aesthetic [in particular - space] as the subjective ground for experience?”

The response we would expect from a defender of Kant’s epistemology would be to insist upon the dualism of knowledge in Kant’s theory. They would rightly point out that Kant names the pure intuition of space as a necessary ground for experience, but not a sufficient ground. In the first case, the pure intuition of time must also be added to the equation. Together these pure intuitions are called the aesthetic, and form the ground of outer and inner experience respectively. The aesthetic, according to Kant, gives us a priori the form of experience – the conditions of space and time that everything in our representations necessarily conforms to (Kant, 1929, pp. A41-43/B59-60) . Similarly to Wittgenstein’s contention that the limits of thought provide the limits of language, Kant would regard the limits of the aesthetic as the limits to sensible knowledge; in setting its form, it also sets its limits. Therefore, for Kant it must be impossible to perceive something that exists under neither of these conditions. The aesthetic is, however, not in itself sufficient for all knowledge, and there must be other capacities that complete our experience. Indeed, Kant introduces his dualism clearly and concisely in the following passage:

Our knowledge springs from two fundamental sources of the mind: the first is the capacity of receiving representations, the second is the power of knowing an object through these representations. Through the first an object is given to us, through the second the object is thought in relation to that (given) representation (Kant, 1929, p. A50/B74) .

The second capacity Kant refers to here is the understanding, and its characteristic spontaneity. He further writes that neither of these two faculties, sensibility nor understanding, can take precedence over the other: “only through their union can knowledge arise” (Kant, 1929, p. A51/B75) . Due to Kant’s clear indication of a dualism of this kind, we should be inclined to reconsider the question of whether language can be the primary condition for experience. It would seem that language has no role in Kant’s epistemology here; indeed he mentions ‘concepts’ of the understanding and logic as science of the rules of the understanding, but neither of these equate to giving any weight to the claim that language has a primary role for two reasons. The first is that, as stated, neither of these capacities takes primacy over the other, so even if language were given a defined role by Kant (which he explicitly does not do), there is no reason to believe that it would have some fundamental and necessary role to play. The second is that Kant clearly ascribes a significant role, in both the transcendental aesthetic and the transcendental logic, to the psychological makeup of the human mind. His psychologistic explication hereby appears to regard language as incidental; language may represent the communication of concepts from one human to another, but the contents of the understanding are not at all identical with language, as described in the Critique of Pure Reason .

Wittgenstein’s position regarding the fundamental makeup of world experience is interesting to compare with Kant’s; the two epistemologies share some common ground, so much so that Erik Stenius, in his commentary on the Tractatus , goes so far as to call Wittgenstein’s project “transcendental lingualism” (Stenius, 1996). Stenius claims that Wittgenstein is undertaking a Kantian project. This, however, should not be taken to mean that Wittgenstein is directly influenced by Kant’s writing. Instead, it is a claim that points to the indirect influence of Kantian ideas that have pervaded Western philosophy since the Enlightenment, and that “his anti-Kantianism meant only (like other Kantians) that he transformed the system of Kant and thus created a Kantianism of a particular kind” (Stenius, 1996, p. 214) .

In parallel, Wittgenstein’s and Kant’s ideas both rely on the input of the subject-seated understanding in giving experience its form and shaping the world. As we have discussed already, Kant’s theory relies on the aesthetic to give reality its form. According to Stenius though, Wittgenstein is engaging in an essentially similar explanation of human cognition of the world when he highlights the role that language plays in providing the conditions and form of our experience. Precisely how these two theories run parallel requires more clarification. Stenius’ argument relies on the claim that the activities of the understanding described by Kant and the activities of the understanding described by Wittgenstein in terms of their world-forming function both refer to the same thing. He argues that on the one hand, what is imaginable is what is logically possible, logic being “the science of the rules of the understanding”, and that what is logically possible is therefore all that is thinkable (Stenius, 1996, p. 218; Kant, 1929, p. A52/B76), making thought the “logical picture of reality”. On the other hand, for Wittgenstein the contents of our world-forming experience (what can be presented via a logical picture) are identical with the contents of language (Wittgenstein, 1961, p. 57; Stenius, 1996, p. 218). In other words, meaningful language is a logical picture, and insofar as it is a logical picture, it also corresponds to what is theoretically possible: thought. Despite the notable differences between the two theories, it is clear that Stenius is suggesting that Wittgenstein and Kant are largely talking about the same processes of world-forming. The implication to be drawn from Stenius’ suggestion is that there is essentially little or no difference between what we mean by “content of the understanding” and “language”. However, as we have already discovered, it appears that Kant deliberately avoided introducing linguistic terminology to describe the content of the understanding in the Critique of Pure Reason . However, a conflation of these two epistemologies offers certain explanatory advantages, namely that it explains how what we think closely correlates with what can be said and also how we cannot seem to think of a concept that we could not put into words.

In his essay, Kant’s Philosophy of Language? , Michael Forster points out that Kant’s attitude to language is on the whole rather ambiguous (Forster, 2012, pp. 485-511). This is notable, he argues, in his change in attitude between writing the three Critiques and some of his later works. Essentially, Forster’s argument is that Kant is content in certain later writings, namely Vienna Logic and Anthropology , to imply that language is indeed part of the essence of thoughts; the content of the understanding. In Anthropology Kant is quoted as having written that “thinking is talking with oneself” and he also objects in Vienna Logic to defining a proposition as a judgement expressed in words on the grounds that a judgement must already be verbal (Forster, 2012, pp. 507-08) . This marks a stark contrast to his earlier work, in which Kant demonstrates his preference to designate the content of thought in purely psychologistic terms, such as “concept” and “intuition” etc., where he might have admitted a relation between thoughts and language. Forster goes on to argue however, that Kant’s refusal to address the question of thought’s relation to language with his own arguments demonstrates that he does not formulate an explicit philosophy of language, but that he engages with the matter by means of agreeing with others who had written on the topic such as Leibniz, Wolff, and his contemporaries and former students Herder and Hamann (Forster, 2012, p. 509) . Hence it is difficult to ascertain Kant’s reasons for this shift, and to determine his own position on the relation between thoughts and language. Forster outlines several reasons why Kant may have never previously explicated thought in terms of language; such as to avoid confusion with his more language-oriented contemporaries, or because of a personal disagreement with Herder, but these fail to provide a satisfactory philosophical answer to Kant’s apparent shift (Forster, 2012, p. 506) . Forster however entertains another possibility, namely that Kant, by his own admittance, was “altogether inexperienced in scholarly philology [and therefore] completely incompetent to make philosophic use of the facts that are related and verified in that branch of knowledge” (1785 cited in Forster, 2012, p. 506).

Kant’s unwillingness to properly engage in the philosophy of language leaves the initial problem unsolved; i.e. that we are unsure what role to assign language in mental activity, particularly that of world-forming. If we could equate thought and language with some certainty then we could in turn conclude that language does have a primary role in forming experience. As it stands however, we are unable to do so, as Kant has no discernible philosophy of language. The answer to the primacy question posed in this paper therefore rests on just how convincing we find Forster’s argument. If we are convinced by him then there is no reason to exclude a role for language in understanding, although it is worth pointing out that this will not assign language the unparalleled primacy Wittgenstein does, as the understanding takes no prominence over sensibility. Even so, the question of primacy’s potency is reduced considerably in this case. If we are not convinced however, then the question of primacy still stands as urgent, perhaps more so than before.

While we are unable to deduce the precise role language plays in experience formation, I believe that I have established that Kantian epistemology could account for an elemental role of language in experience formation. By this I mean that language plays a necessary and grounding role in making human experience possible, much in the same way that the a priori intuition of space makes the appearance of an external world possible. In this sense language too is an a priori condition of experience. This is evidenced by my discussion of the similarity between the Kantian phrase “content of the understanding” and Wittgenstein’s ideas about language’s role in cognition. I further established that Kant himself was unwilling to define concepts as fundamentally linguistic. Yet Forster’s essay shows that although Kant entertained the idea that mental behaviour is linguistic behaviour later in his career, it may not be necessarily inconsistent with the doctrine of transcendental idealism to suggest this. To this extent I would argue that a Wittgensteinian reinterpretation of Kant’s epistemological position is necessary in order to clarify the contentious issue hitherto considered, namely the exact relation between the understanding and language, which I am currently most suitably equipped to draw attention to, rather than solve in full. It is therefore my hope that such an interpretation would prove conducive to further philosophic enquiry.