Raymond Tallis, reviewing Meillassoux's "After Finitude" in tpm:
At the heart of Kant’s various critiques is a response to David Hume’s observation that there is nothing in experience to explain why the world is ordered – why it is intelligible or even habitable. The causal glue that seems to hold things together is not a property of things in themselves: it is merely a product of our experience of patterns of events. The sense of necessity is just the result of the habits of expectation. Consequently, there is no reason why there should be such patterns and why they should be maintained: I have no grounds for expecting that the sun will rise tomorrow. The laws of nature seem shockingly contingent.
Kant was electrified by Hume’s argument that causation was not an intrinsic property of the world in itself. He went further and said that space and time were also properties of experience, of the phenomenal world, not the realm of things in themselves. Space, time and causation were the necessary conditions or “forms” of sense experience. Any world that was experienced would be experienced as set out in space and time and causally glued together. We do not therefore experience what there is as it is in itself. Hence metaphysics – an account of being qua being, the nature of the in-itself – was impossible.
[Meillassoux] recasts Hume’s problem as being about the contingency of the laws of nature. If the laws are contingent, does that not mean that it is highly unlikely that the world will be ordered? Won’t the laws, being contingent, change at such a frequency that the world will be chaotic, uninhabitable, unable to support life? No, Meillassoux says: this follows only if we assume that the rate of change of the laws, their instability, would be such that they would have an impact on ordinary life. Such an assumption relies on a judgement of probabilities but this, since the “detotalisation” of number through the discovery of transfinite quantities, is groundless. Hence the title of his book. Having established this, he goes on to argue not only that the laws of nature are contingent but that they are necessarily so. Hence the subtitle of his book.