of God). The immediacy of intuition is crucial because it is what sets them off from concepts, which are essentially representations of representations, i.e., rules expressing what is common to a set of representations. Kant claims that mathematics, and metaphysical expositions of our notions of space and time, can reveal several evident synthetic a priori propositions, e.g., that there is one infinite space. In asking what could underlie the belief that propositions like this are certain, Kant came to his Copernican revolution. This consists in considering not how our representations may necessarily conform to objects as such, but rather how objects may necessarily conform to our representations. On a ‘pre-Copernican’ view, objects are considered just by themselves, i.e., as ‘things-in-themselves’ (Dinge an sich) totally apart from any intrinsic cognitive relation to our representations, and thus it is mysterious how we could ever determine them a priori. If we begin, however, with our own faculties of representation we might find something in them that determines how objects must be – at least when considered just as phenomena (singular: phenomenon), i.e., as objects of experience rather than as noumena (singular: noumenon), i.e., things-inthemselves specified negatively as unknown and beyond our experience, or positively as knowable in some absolute non-sensible way – which Kant insists is theoretically impossible for sensible beings like us. For example, Kant claims that when we consider our faculty for receiving impressions, or sensibility, we can find not only contingent contents but also two necessary forms or ‘pure forms of intuition’: space, which structures all outer representations given us, and time, which structures all inner representations. These forms can explain how the synthetic a priori propositions of mathematics will apply with certainty to all the objects of our experience. That is, if we suppose that in intuiting these propositions we are gaining a priori insight into the forms of our representation that must govern all that can come to our sensible awareness, it becomes understandable that all objects in our experience will have to conform with these propositions.
Kant presented his transcendental idealism as preferable to all the alternative explanations that he knew for the possibility of mathematical knowledge and the metaphysical status of space and time. Unlike empiricism, it allowed necessary claims in this domain; unlike rationalism, it freed the development of this knowledge from the procedures of mere conceptual analysis; and unlike the Newtonians it did all this without giving space and time a mysterious status as an absolute thing or predicate of God. With proper qualifications, Kant’s doctrine of the transcendental ideality of space and time can be understood as a radicalization of the modern idea of primary and secondary qualities. Just as others had contended that sensible color and sound qualities, e.g., can be intersubjectively valid and even objectively based while existing only as relative to our sensibility and not as ascribable to objects in themselves, so Kant proposed that the same should be said of spatiotemporal predicates. Kant’s doctrine, however, is distinctive in that it is not an empirical hypothesis that leaves accessible to us other theoretical and non-ideal predicates for explaining particular experiences. It is rather a metaphysical thesis that enriches empirical explanations with an a priori framework, but begs off any explanation for that framework itself other than the statement that it lies in the ‘constitution’ of human sensibility as such.
This ‘Copernican’ hypothesis is not a clear proof that spatiotemporal features could not apply to objects apart from our forms of intuition, but more support for this stronger claim is given in Kant’s discussion of the ‘antinomies’ of rational cosmology. An antinomy is a conflict between two a priori arguments arising from reason when, in its distinctive work as a higher logical faculty connecting strings of judgments, it posits a real unconditioned item at the origin of various hypothetical syllogisms. There are antinomies of quantity, quality, relation, and modality, and they each proceed by pairs of dogmatic arguments which suppose that since one kind of unconditioned item cannot be found, e.g., an absolutely first event, another kind must be posited, e.g., a complete infinite series of past events. For most of the other antinomies, Kant indicates that contradiction can be avoided by allowing endless series in experience (e.g., of chains of causality, of series of dependent beings), series that are compatible with – but apparently do not require – unconditioned items (uncaused causes, necessary beings) outside experience. For the antinomy of quantity, however, he argues that the only solution is to drop the common dogmatic assumption that the set of spatiotemporal objects constitutes a determinate whole, either absolutely finite or infinite. He takes this to show that spatiotemporality must be transcendentally ideal, only an indeterminate feature of our experience and not a characteristic of things-in-themselves. Even when structured by the pure forms of space and time, sensible representations do not yield knowledge until they are grasped in concepts and these concepts are combined in a judgment. Otherwise, we are left with mere impressions, scattered in an unintelligible ‘multiplicity’ or manifold; in Kant’s words, ‘thoughts without content are empty, intuitions without concepts are blind.’ Judgment requires both concepts and intuitions; it is not just any relation of concepts, but a bringing together of them in a particular way, an ‘objective’ unity, so that one concept is predicated of another – e.g., ‘all bodies are divisible’ – and the latter ‘applies to certain appearances that present themselves to us,’ i.e., are intuited. Because any judgment involves a unity of thought that can be prefixed by the phrase ‘I think’, Kant speaks of all representations, to the extent that they can be judged by us, as subject to a necessary unity of apperception. This term originally signified self-consciousness in contrast to direct consciousness or perception, but Kant uses it primarily to contrast with ‘inner sense’, the precognitive manifold of temporal representations as they are merely given in the mind. Kant also contrasts the empirical ego, i.e., the self as it is known contingently in experience, with the transcendental ego, i.e., the self thought of as the subject of structures of intuiting and thinking that are necessary throughout experience.
The fundamental need for concepts and judgments suggests that our ‘constitution’ may require not just intuitive but also conceptual forms, i.e., ‘pure concepts of the understanding,’ or ‘categories.’ The proof that our experience does require such forms comes in the ‘deduction of the objective validity of the pure concepts of the understanding,’ also called the transcendental deduction of the categories, or just the deduction. This most notorious of all Kantian arguments appears to be in one way harder and in one way easier than the transcendental argument for pure intuitions. Those intuitions were held to be necessary for our experience because as structures of our sensibility nothing could even be imagined to be given to us without them. Yet, as Kant notes, it might seem that once representations are given in this way we can still imagine that they need not then be combined in terms of such pure concepts as causality. On the other hand, Kant proposed that a list of putative categories could be derived from a list of the necessary forms of the logical table of judgments, and since these forms would be required for any finite understanding, whatever its mode of sensibility is like, it can seem that the validity of pure concepts is even more inescapable than that of pure intuitions. That there is nonetheless a special difficulty in the transcendental argument for the categories becomes evident as soon as one considers the specifics of Kant’s list. The logical table of judgments is an a priori collection of all possible judgment forms organized under four headings, with three subforms each: quantity (universal, particular, singular), quality (affirmative, negative, infinite), relation (categorical, hypothetical, disjunctive), and modality (problematic, assertoric, apodictic). This list does not map exactly onto any one of the logic textbooks of Kant’s day, but it has many similarities with them; thus problematic judgments are simply those that express logical possibility, and apodictic ones are those that express logical necessity.
The table serves Kant as a clue to the ‘metaphysical deduction’ of the categories, which claims to show that there is an origin for these concepts that is genuinely a priori, and, on the premise that the table is proper, that the derived concepts can be claimed to be fundamental and complete. But by itself the list does not show exactly what categories follow from, i.e., are necessarily used with, the various forms of judgment, let alone what their specific meaning is for our mode of experience. Above all, even when it is argued that each experience and every judgment requires at least one of the four general forms, and that the use of any form of judgment does involve a matching pure concept (listed in the table of categories: reality, negation, limitation; unity, plurality, totality; inherence and subsistence, causality and dependence, community; possibility – impossibility, existence – non-existence, and necessity–contingency) applying to the objects judged about, this does not show that the complex relational forms and their corresponding categories of causality and community are necessary unless it is shown that these specific forms of judgment are each necessary for our experience. Precisely because this is initially not evident, it can appear, as Kant himself noted, that the validity of controversial categories such as causality cannot be established as easily as that of the forms of intuition. Moreover, Kant does not even try to prove the objectivity of the traditional modal categories but treats the principles that use them as mere definitions