20th World Congress of Philosophy Logo

Modern Philosophy

Concerning the Relationship Between Non-Spatiotemporality and Unknowability of Things in Themselves in Kant's Critique of Pure Reason

Juan Adolfo Bonaccini
Federal University of Rio Grande do Norte/National Research and Development Council, Brazil

bluered.gif (1041 bytes)

ABSTRACT: In the present paper is analyzed the relationship between Kant's theses concerning unknowability and non-spatiotemporality of things in themselves. First of all, it is argued that even by taking for granted that the Unknowability Thesis does not contradict the Non-Spatiotemporality Thesis, because the former can be thought as a consequence of the latter, this is not enough to avoid another problem, namely, that the Non-Spatiotemporality Thesis is not sufficient to abolish the possibility of thinking consistently of space and time as empirical or material. It is also remembered that this point has already been partially envisaged for the first time by H.A. Pistorius (and later by A. Trendelenburg) and raised as the objection of the "third possibility" or "neglected alternative." Furthermore, it is maintained that although Kant tries to eliminate this possibility in the Metaphysical Expositions of Space and Time (but not in the Antinomies), by attempting to prove that space and time are only formal necessary conditions of sensibility, he cannot do it successfully. Hereafter it is argued that his circumstance is not due to the above objection itself, but to another difficulty that can only be grasped through the analysis of Kant's main argument in the Metaphysical Expositions of Transcendental Aesthetic. Ultimately, in order to show this difficulty, it is argued first that insofar as the Non-spatiotemporality Thesis supposes the validity of the Singularity Thesis, and this supposes the validity of the Apriority Thesis, the whole force of proof reposes on this latter. Secondly, it is shown that, despite his effort, Kant could not justify satisfactorily his claim to the formal apriority of space and time because of his failure to demonstrate necessarily the Apriority Thesis.

bluered.gif (1041 bytes)


We have already given a detailed account of this question in another place, (1) so that here we will try to explain only one of the main arguments. The reason why we have chosen the following one among the others reposes on the fact that it involves an almost unperceived supposition on Kant's part, whose possible implications we would like to explore.

The first thing to be said is that Kant says we cannot know the things as they are in themselves, because in this case they would be essentially neither spatial nor temporal ones. We could surely ask: how can Kant say that, while maintaining simultaneously the Unknowability Thesis? How can he say that things in themselves are neither spatial nor temporal, if he admits that they are unknowable for us? The difficulty was already grasped in Kant's time. The best answer to the question seems to be that Kant formulates the Unknowability Thesis as a consequence of the Non-spatiotemporality one. (2) To know how things in themselves would be, we would need to know how they would be out of its spatiotemporal condition. We know certainly that things cannot be spatiotemporal in themselves, because things in themselves are the same empirical things viewed or considered by abstracting from the subjective conditions of our sensibility, namely space and time. (3) So, if space and time are the only necessary conditions of giveness of anything whatsoever to be perceived by rational finite beings, then, insofar as we are also rational finite beings, we can never know anything whatsoever out of space and time. In other words: we cannot know the things considered as they would be in themselves, without their spatiotemporal features, because we can only know something on condition of its giveness in space and time. So though we cannot know things in themselves, we can, argues Kant, think them. We can really think how they could be without-"by abstracting from"-the (inter-)subjectively added aspects, i.e. completely out of spatiotemporal coordinates. (4)

These considerations show that it is impossible that the Unknowability Thesis contradicts the Non-spatiotemporality Thesis. (5) When we read Kant's Argument correctly, we perceive the sequence or arrangement of the analysis, in a word, its hierarchy. In fact, Kant defends the Unknowability Thesis after proving the Non-spatiotemporality one. (6) But this circumstance is not enough to prove that things in themselves would be necessarily non-spatiotemporal. And moreover, if it is sound, then we ought to acknowledge that Kant cannot demonstrate the Unknowability Thesis either. So, if we want to know why this argument does not suffice to prove what Kant claims, we have briefly to examine the whole argument. That's the aim of the following parts in the present paper.


At the beginning of the Introduction to the Kritik Kant polemically suggests that experience could be a "composition" (Zusammengesetztes) made, on one side, of the content of "impressions" (Eindrücke), and on the other side, of the possible contribution of our "cognitive powers" (Erkenntnisvermögen). (7) Both rationalists and empiricists had generally admitted there are "sensible" and "non-sensible" components in knowledge, as well as the existence of so-called "analytical" and "synthetical" propositions grounded upon them. But they had disagreed with regard to the weight and function of each one of these elements in the whole process of knowing, and mainly in respect of its origins. Despite his differences with both positions, in order to prove the possibility of non-trivial (synthetic) a priori knowledge Kant accepts in principle general features of both points of view. His procedure, we could say, consists in an apagogical way of arguing against and refuting them. He begins by accepting trivial non-polemical aspects for the purpose of showing their limitations, and to guard thereby his thesis: we can know in fact synthetically a priori because we don't know but about appearances, phaenomena. In other words: we have a priori knowledge only because we know nothing about things in themselves. (8)

Kant concedes it is certainly true that we have experience, and that experience is par excéllence the starting-point of human knowledge. But does this mean we have only empirical knowledge? If so, he thinks we would have no science at all, because all knowledge would always involve only sensible, empirical and therefore inductive propositions. Always empirical cognitions, but not universal and necessary ones. Nonetheless, both the rationalist and empiricist positions admit the existence of scientific knowledge, Kant seems to think, so that to accept that all our knowledge comes only from sensible experience would not follow. So we have in one way or another some knowledge that is universal and necessary. And if so, then, while experience qua experience gives neither necessity nor universality, we must admit the existence of non-empirical cognition or knowledge that cannot have its origin or "fount" (Quelle) in sensible experience.

Thus, by starting from this hypothesis of the composed nature of our knowledge, Kant is allowed to admit: first, as the empiricists do, that we have actually sensible experience qua starting-point of knowledge, and secondly, in spite of this chronological "sense-perceptual" origin of knowledge, that we have "rational" knowledge of some universal principles or notions which do not have their origin in perception and does not come from experience, as the rationalists had thought. That's the meaning of Kant's famous statement: "Although all of our knowledge begins with experience, this does not merely imply that all of it originates from experience. Because it is quite possible that even our empirical knowledge would be a composition of what we received from impressions and what our cognitive powers (occasioned solely by sensible impressions) give by themselves..." (9)

If experience is undoubtedly the temporal "starting-point", but not the "epistemic" principle of knowledge, then we have in principle the possibility of thinking there are some principles working as conditions for knowledge of experience, even by starting from sense-perceptions. It is also possible that there is universal, necessary and non-trivial knowledge based on these principles, and if we really have these as conditions, then we will know a priori something about them. Yet it remains to be proved that these are necessary conditions of sensibility. We must prove that rational finite beings could perceive nothing if not under these conditions. When we solve this problem successfully, it appears to suffice to show that we cannot know things in themselves because of their non-spatiotemporality. Yet to prove that we are only able to have perceptions of things under these conditions, it is not enough to take for granted that things in themselves are neither spatial nor temporal. It is not sufficient, unless we have before proved that space and time are undoubtedly the only subjective conditions under which we can have sense-perceptions of things as appearances. In other words, the discussion with the rationalists and empiricists relative to the origins of human knowledge gives Kant the possibility of demonstrating what he wants, but he has yet to prove it.

It has been argued, since Kant's time, that his theory of human sensibility in the Trancendental Aesthetics shows only that we always know the things of experience in space and time, but not that things cannot be at the same time spatial and temporal in themselves too. (10) Yet since this is not the supposition we stated at the beginning, it would have been generally "unperceived". This is the famous objection of the "neglected alternative", which has been quite early grasped by some contemporaries of Kant. In fact, the supposition about which we would like to speak appears to be similar to that one, but it is not so. (11) If, however, we pay attention to the two most frequent replies to this objection, then we will understand the unperceived supposition with which we want to deal.


There are two possible answers which both originate in one and the same principle. The first one consists in maintaining that Kant has absolutely not neglected whatever alternative. The first possibility of answering would consist in arguing that Kant proves in the Trancedental Dialectic, namely in the Antinomies, that it would be contradictory to suppose or admit that the things in themselves could be spatiotemporal. (12) The second one would consist in replying that Kant had already demonstrated in the Transcendental Aesthetics it is not adequate to maintain that the things in themselves could be spatial and temporal. When we put away the former, by replying that Kant does not prove in the Antinomies that the supposition of knowledge of things in themselves is contradictory, because Kant himself recognizes the contrast of propositions performs only a "dialectical opposition" , not an 'analytical' contradiction, (13) and therefore space and time could be still things in themselves or even their qualities; we can try to answer the question by searching for Kant's reasons for the impossibility that things in themselves are spatial and temporal in the Aesthetic. That appears to be rational and quite easy: it's trivial that Kant takes this point, against Newton and Leibniz, as needing to be answered and refuted.

As soon as the Aesthetic begins, Kant definites basic terms and introduces the distiction between mediate and immediate knowledge. The latter is named by Kant "intuition" (Anschauung). (14) Kant analyses this notion in its elements and explains it in contrast with some others like "to be given", "to be affected by objects", "sensation", etc. We have intuitions, he teaches, when objects have been given to us, to our senses before; and this occurs only when we have sensations qua effect of their action upon us. The sensibility itself is defined as the "ability of being affected by objects". (15) So that there are always sensations for us, when we have been affected by things. Later he draws a distinction between empirical intuition (i.e. that given through sensation), and "pure" intuition, negatively defined as the one without any element or remainder of sensation. Afterwards he introduces the concept of appearance (Erscheinung), as well as those of matter and form of appearance. Appearance is defined as the "indefinite object of the empirical intuition" (der unbestimmte Gegenstand einer empirischen Anschauung), sensation as its matter, and form as its non-sensible possibility of ordering. Form, in opposition to matter, says Kant, ought to be a priori, because the ordering relations of matter cannot have its origin in sensation itself. (16) With this elements, after having put aside all mediate components of knowledge, Kant entertains the task of isolating within immediate knowledge the sensual components from the formal ones. In order to be allowed to say that space and time are neither things in themselves nor qualities of them, Kant first raises the question about the essence of space and time. He found two unsatisfactory and contrasting answers and tried to give another alternative. (17) With this aim he tries to prove against Newton that space and time are not concepts empirically acquired from experience, and against Leibniz, that they are not relational or general concepts either. Only after these arguments, Kant presents the thesis space and time are pure intuitions; and only after that can Kant defend the Ideality Thesis, namely, that space and time are the forms of sensibility and hence the unique conditions under which objects can be given to us. According with this line of argument, the whole question has been decided by Kant principally within the Metaphysical Exposition of space and time. (18) We think it occurs, on the whole, in points 1 and 2 of both expositions. Independently of determining if the first through the forth points of the metaphysical exposition of space, and if the first through the fifth points of the metaphysical exposition of time, are themselves different "arguments" or successive "steps" of one or two arguments, (19) we can say that Kant needs first to prove the "Apriority Thesis" and second the "Singularity Thesis"; so that the "Ideality Thesis" as well as the "Non-spatiotemporality Thesis" follow from those first ones. Finally, the Unknowability one is logical implication of this latter. That's the reason why, if Kant could not prove successfully any of the previous theses he undertakes to prove, he could not prove the Unknowability Thesis either. The most important point to prove is therefore the Apriority Thesis, because Kant must prove it first of all to be allowed to prove that space and time are not empirical. If space and time would be empirical, then they could not be pure forms or intuitions. And if it would be so, then Kant could not prove that we can not know things in themselves. The points to be proved are: 1) that space and time are not empirical concepts; 2) (taken for granted that (1) has been satisfactorily proved) that they are not a priori concepts either; 3) that (taken for granted that (1) and (2) have been proved) they are a priori intuitions.

Taken for granted that (1), (2), and (3) have been proved, Kant can implies that: a) we can know only objects given to us in space and time; b) things in themselves could not be spatial neither temporal; and c) hence we cannot know things in themselves, but only appearances. If so, then we see that the force of the whole argument reposes on (1). Kant must demonstrate first of all that space and time could never be empirical to show they must be viewed as a priori. The Apriority Thesis is then the crucial point to discuss. And we make it in the next part.


The Apriority Thesis was been presented and justified by Kant at points 1 and 2 of both metaphysical expositions. So let's examine it to show the hidden supposition mentioned at the beginning of our analysis.

The first step of the argument-or, if one prefers, the first argument-declares space and time could not be empirical concepts abstracted from empirical things, because we must always suppose both of them before having experience of any object. (20) This amounts to admitting space S and time T are conditions of the experience of the objects A, B, C, and so on. And this is equal to saying that space and time are conditions of experience or sense-perception, and therefore, that they are both a priori. Nonetheless, we can certainly ask: is Kant allowed to say that? May this claim be applied justifiably to this purpose? We think not, because this claim was just the aim of the whole argument: if space and time are not empirical concepts abstracted from sensible impressions, they can only be representations preceding experience, i.e. a priori. From which follows that it is not enough to say that space and time are empirical concepts because they are representations a priori. That is precisely what was to be demonstrated! By itself, this step is a petitio principii. In spite of this problem, we could still think it would be only the statement of the thesis; and then we could consider the other step or argument (21) (2.) as the proper proof. (22)

The second step of argument (23) is well known. Kant explains it by saying that we can perfectly represent an empty space and an empty time, but nothing whatsoever out of space and time. (24) On the basis of the consideration that we cannot abstract space and time notions from experienced (perceived) things, Kant concludes its apriority. If space and time were empirical, it appears that we could represent any object without space and time or any of their properties, and space and time might not exist. But though we can think of space and time without any object whatsoever, we cannot represent objects without having to suppose them to be spatial and temporal. Apparently from this we can derive the Aprioriry Thesis: we need to suppose a priori space and time to have experience on the whole. Yet the apriority that we find in Kant's argument is not sufficient to justify the claim that space and time are "formally" a priori. Let's try to envisage why.

The argument for the apriority consists on the whole in arguing that Space and Time are in themselves neither empirical things nor their properties because the experience or perception of things is only possible on condition that one supposes both of them before experiencing or perceiving anything, and this is so because if Space and Time were not a priori, then we could represent at least one thing out of space or time, or even the non-existence them. But we cannot really represent anything out of them, including the possibility of the non-existence of space and time. Ergo, both of them are a priori representations.

Now, we must still remember that Kant takes part in the discussion between Leibnizians and Newtonians concerning the nature of space and time; and that he wants to prove against them his alternative. So that, in relation to those opponents, Kant needs to prove his point of view in a necessary way, i.e. apodictically. That's why we might always ask if Kant can do this. To answer the question, we must first know how he could do it: he could only do it, if he could prove necessarily that the contrary option is a contradictory one, and therefore no option at all. That's why the question is: does Kant really prove that the contrary alternative is contradictory? Apparently he does; but not actually. Because it is not contradictory to think of these representations as apriori and at the same time abstracted from experiences.

It is quite possible the representations of space and time are acquired by successively and unconsciously contrasting a lot of experiencies, which were used to understand things and relations between each other. I can perfectly think of space and time as "psychological" a priori but not "pure" a priori; so I can think that I must suppose space and time a priori but only because experience has taught me that things are themselves spatiotemporal.

To prove necessarily that space and time are a priori representations in the "strong" sense-as "pure" a priori-Kant needs to demonstrate that we cannot absolutely think of them but as a priori. Given that this task implies the impossibility of a non-contradictory thinking of a contrary alternative, and that we can really think free from contradictions that space and time are empirical, we must infer that Kant could not prove the contrary alternative is contradictory. If Kant could do that, then the argued "necessity" of the representations of space and time would be logical. But it is only "psychological" and "subjective": we must always suppose space and time in order to have experiences, but this does not imply necessarily that it is so because space and time could never be empirical things (and our representations derived from them) or properties of them. It's possible to think the contrary possibility out of any contradiction.

There exist at least two possible responses to the argument presented here. The first is that based on the argument of the Antinomies, which is already mentioned above. The second consists in arguing that this necessity is neither logical nor psychological but transcendental. A full response to this point is beyond our present scope, but we can at least suggest our line of the argument: Kant can only speak transcendentally after proving the Ideality Thesis, yet he can only establish transcendentality (through the Ideality Thesis) after having proven the Apriority Thesis.

bluered.gif (1041 bytes)


(1) Cf. J. A. Bonaccini. O Conceito de Coisa em Si no Idealismo Alemão. Sua atualidade e relevância para a compreensão do problema da Filosofia. (The Concept of 'Thing in itself' in German Idealism: Its Relevance to the Understanding of the Problem of Philosophy). Rio de Janeiro: IFCS/UFRJ. 1997, Second Part, Chapter 2.

(2) See: Immanuel Kant, Kritik der reinen Vernunft, Nach der ersten und zweiten Original-Ausgabe neu herausgegeben von Raymond Schmidt. Hamburg: F. Meiner.1930/ Durchgesehener Nachdruck von 1976 (abreviated: KrV), and compare especially A26-27/B42-3 and A42-43/B59-60.

(3) See about this statement: Henry E. Allison, Kant's Transcendental Idealism. An Interpretation and Defense. New Haven/London: Yale University Press. 1983, pp. 3-13; 237-242, and Gerold Prauss, Kant und das Problem der Dinge an sich. Bonn: Bouvier. 1989, pp. 52-61. Cf. Jacinto Rivera de Rosales, La Realidad en sí en Kant. Madrid: Universidad Complutense. 1988, pp. 356-367; Léo Freuler, Kant et la métaphysique spéculative. Paris: Vrin.1992, pp. 171-182/ 200-207; Marco A. Frangiotti, "Refuting Kant's 'Refutation of Idealism' ", pp. 93-95, in: Idealistic Studies 25, n.1 (winter 1995) : 93-106.

(4) See for example KrV, A38/B55.

(5) In other way, something different of us, Lorne Falkenstein have recently considered this point in his paper "Kant's Argument for the Non-spatiotemporality of Things in Themselves", for example at p. 282, in: Kant-Studien 80 (1989) : 265-83. We can also find versions (or the main idea) of this argument in H. E. Allison, op. cit., p. 241, and "The Non-spatiality of Things in Themselves for Kant", p. 319, in: Journal of the History of Philosophy 14 (1976) : 313-321; Ralf Meerbote, " The Unknowability of Things in Themselves", pp. 169-170, in: Kant's Theory of Knowledge (Ed. by L.W. Beck). Dordrecht: Reidel.1974, pp.166-174; Kenneth Rogerson, "Kantian Ontology", p. 3, in : Kant-Studien 84 (1993) : 3-24; Jill Buroker, Space and Congruence. The Origins of Kant's Idealism . Dordrecht: Reidel. 1981, pp. 99-100; Lauchlan Chipman, "Things-in-Themselves", 265ff. in: Immanuel Kant. Critical Assessments. (Ed. by Ruth Chadwick and Clive Cazeaux). London/New York: Routledge. 1992, Vol. II, pp. 263-275.

(6) See for example A26-28/B42-43. Cf. Falkenstein, op. cit., p. 265.

(7) KrV, B2.

(8) KrV, BXVI-XIX. Cf. B XX-XXVIII; B59-60, B73, etc.

(9) "Wenn aber gleich alle unsere Erkenntnis mit der Erfahrung anhebt, so entspringt sie darum doch nicht eben alle aus der Erfahrung. Denn es könnte wohl sein, dass selbst unsere Erfahrungserkenntnis ein Zusammengesetztes aus dem sei, was wir durch Eindrücke empfangen, und dem, was unser eigenes Erkenntnisvermögen (durch sinnliche Eindrücke bloss veranlasst) aus sich selbst hergibt..." (KrV, B1); author's translation.

(10) This is the famous objection of a third alternative "forgotten" by Kant, the so-called "neglected alternative". See about this point Allison (1983: 111-114). The first to raise it clearly and correctly was Hermann Andreas Pistorius; yet it had been thereafter raised by others leibnizians of Eberhard's Circle. On this point see principally Hans Vaihinger's Commentar zu Kants Kritik der reinen Vernunft (Second Edition, Stuttgart: Union Deutsche Verlagsgesellschaft. 1922) , Vol. II, pp.142-151. Almost one hundred years later Adolph Trendlenburg gives a new version of the same objection, especially in his Logische Untersuchungen. Leipzig: Hinzel.1862. Concerning Trendelenburg's objection see also Vaihinger, op. cit., II, pp. 130-142 /290-326. Falkenstein, Buroker and others also treat of this problem.

(11) See our : O Conceito de Coisa em Si no Idealismo Alemão, op. cit., Second Part, Chapter 2.1.1.

(12) So Roberto Torreti,"On the subjectivity of objective space", p. 570, In: Proceedings of the Third International Kant Congress. (Ed. by L.W. Beck). Dordrecht: Reidel.1972, pp. 578-573; Alfred C. Ewing, A Commentary on Kant's Critique of Pure Reason. Chicago: The University of Chicago Press. 1938, pp. 50ff. According to Vaihinger, Kuno Fischer had already used this argument at the end of XIXth Century against A. Trendelenburg. See op. cit.,Vol. II, 301ff.

(13) KrV, B 531-532.

(14) KrV, A19/B33.

(15) "Die Fähigkeit (Receptivität), Vorstellungen durch die Art, wie wir von Gegenständen affiziert werden, zu bekommen, heisst Sinnlichkeit" (ibidem).

(16) Vaihinger observed with reason that Kant supposed it to be evident and did not prove it. (See Vol. II, pp.69ff.). Many skeptic people could raise the following question: why does order exist necessarily apart from sensations? Why could not have its origin in the matter itself, unless we have already suppose it is impossible? And moreover, why it is not possible, unless we assume that we know only appearances? Yet it is just what was to be proved!

(17) Allison (1983: 81).

(18) Though something different to Buroker, Allison and Falkenstein, Manfred Baum have recently started , in principle, from this same supposition against Paul Guyer's new undertaking of Trendelenburg's objection (see Guyer's Kant and the Claims of Knowledge. Cambridge. 1987, pp. 362-368 - apud Baum). See Manfred Baum's paper, "Dinge an sich und Raum bei Kant", pp. 63ff. (especially p. 70), in: Akten des 7. Internationalen Kant-Kongresses (Kurfürstliches Schloss zu Mainz, 1990, Hrsg. von Gerhard Funke, Bonn: Bouvier, 1991), Sektion B/II.1, pp. 63-72.

(19) There has been a large controversy about this point. See Allison (1983: 82ff.) and Vaihinger (Vol. II, pp.156-202).

(20) See KrV, A23/B38 and A30/B46.

(21) We refer to KrV, A24/B38-39 and A31/B46.

(22) Similarly to Vaihinger (opus cit., vol. II, pp. 186ff.) and Paton (Kant's Metaphysic of Experience. New York: Macmillan. 1936, Vol. I, p. 112, apud Allison (1983: 85), to whom Paton is not wright respect to it)

(23) After having already showed that the first point (1) could not be valid alone as a proof of its purpose, we can maintain that it is one of the steps of the argument for Apriority.

(24) See KrV, A24/B38-39 and A31/B46.

bluered.gif (1041 bytes)


Back to the Top

20th World Congress of Philosophy Logo

Paideia logo design by Janet L. Olson.
All Rights Reserved


Back to the WCP Homepage