Immanuel Kant is widely acknowledged by philosophers of all persuasions to have been one of the greatest thinkers of all time. He is also notorious for being one of the most difficult to understand. The complexity of his prose, however, is not due to any wilful obscurantism; in reading Kant, one is aware of a thinker struggling to clothe in language ideas of the very highest level of complexity and profundity.

Born in 1724, Kant lived his entire life in the East Prussian town of Konigsberg. He never married, though he was a popular man who by all accounts led a life of the utmost order and regularity. His unique contribution to Western thought is his 'Critical Philosophy': a devastating and ingenious critique of both speculative rationalistic metaphysics, and unfettered empiricism. And yet this monumental system of thought, as set out in the Critique of Pure Reason, stems from just one seemingly humble question: how are synthetic a priori truths possible?

Kant introduced the distinction between 'analytic' and 'synthetic' judgements, despite its having been implicit in the works of Hume and Aristotle, amongst others. He characterises an analytic judgement as one in which 'the predicate B belongs to the subject A, as something which is (covertly) contained in this concept A'. (Critique of Pure Reason) The favourite example of philosophers is 'All bachelors are unmarried'. Here, the predicate ('are unmarried') simply 'unpacks' the conceptual content of the subject ('bachelors'). A distinguishing feature of such propositions is that they tell us nothing about the way the world is; instead, they simply clarify what is involved in our concepts.

In the case of synthetic judgements, by contrast, Kant tells us that the predicate 'lies outside the subject concept'. (ibid.) An example might be 'All humans are under twenty feet tall'. Whilst this proposition is no doubt true, it is nonetheless certainly not a feature of the concept 'human' that anything falling under it is under twenty feet tall. Thus, 'All humans are under twenty feet tall' gives us a substantial piece of information about the world, rather than about the concepts we apply to that world.

It should be easy to see that analytic truths are a priori: that is, knowable independently of any particular experience. I do not have to carry out a survey of bachelors to find out that they are all unmarried. But how could any synthetic truth - one which gives us real information about the world - be a priori?

Kant was of course aware that the vast majority of synthetic truths are knowable only a posteriori - that they require verification through experience. 'All humans are under twenty feet tall' could never be known a priori. He held, however, that there exists a special class of propositions that are both informative and knowable independently of this or that experience. The truths of mathematics (perhaps most significantly those of geometry), he maintained, fall into this class, as do certain other propositions, such as 'Every event has a cause'.

There is nothing about the concept of 7+5 that dictates that it should be equal to 12, nor about the concept of a straight line that it should be the shortest distance between two points. And yet the propositions '7+5=12' and 'A straight line is the shortest distance between two points' are both knowable a priori. Similarly, it is not part of the concept of an event that it should have a cause, and yet we can know with absolute certainty, thinks Kant, that any event will be caused. But how can we know such truths a priori?

Kant's answer to this question is both radical and astonishing. Let us start with the case of geometry. There can only, thinks Kant, be one explanation of our a priori knowledge of the properties of space: the spatial properties of the world must be contributed by the knowing subject. That is, the world as it is in itself is not made up of objects arranged in space. Only the world as it appears to us is spatial, and this is precisely because space is nothing more than our way of representing the world to ourselves. In Kant's own terminology, space is nothing more than a form of intuition (that is, a form of sensory perception). Kant employs a similar, though much weaker, argument to conclude that time, too, is a mere form of intuition. Space and time are features of the phenomenal world - the world as it appears to us - only. The noumenal world - the world of things as they are in themselves - is aspatial and atemporal.

Similarly, causal relations have a subjective origin, being, as it were, 'projected' into the world by the experiencing consciousness. Consequently, causation too is a feature only of the world of appearances, and not of the world independent of our cognitive faculties. However, whereas the forms of intuition are features of our faculty of sensibility (the passive faculty that receives sense impressions), causation is one of twelve 'categories', or 'a priori concepts' imposed on sense impressions by the understanding (the active faculty of reason).

Kant's epistemology stands as a critique of both empiricism and rationalism. The empiricist view is wrong, since the mind is not a mere tabula rasa which passively receives knowledge of the world through the senses. The rationalist view is just as mistaken, as reason alone can never give rise to knowledge, since knowledge demands both concepts and the raw data supplied by the senses. Thus speculative metaphysics - the attempt to achieve theoretical knowledge of non-empirical subjects such as the existence of God, freedom, and immortality - inevitably falls into illusion. It aims to gain knowledge of the world as it is in itself, but theoretical knowledge can only be of the world as it appears.

However, whilst Kant held that we have no theoretical knowledge of such things, he maintained that we can have a practical knowledge of them. Consider free will. When I think of my actions as constituents of the phenomenal world, I am obliged to regard them as produced by rigid deterministic laws, but when I consider those same actions as they are in the noumenal world, I am not so obliged. I can have practical knowledge of that freedom which I am required to postulate in order to account for my inescapable sense of myself as a responsible moral agent.

It seems to many that a choice has to be made between two apparently incompatible ways of looking at the world: the spiritual and the ethical on the one hand, and the scientific on the other. If Kant is right, the dichotomy between these two ways of looking at the world is purely illusory. There is room in the world for both determinism and freedom, for science and spirituality.