Open AccessThe article argues that Meillassoux's 'After Finitude' underestimates the nature and profundity of Hume's sceptical challenge; it neglects the fact that Hume's scepticism concerns final causes (and agrees fundamentally with Bacon and Descartes in this respect), and that in Hume even the operations of reason do not furnish entirely a priori knowledge. We contend that Hume himself institutes a form of correlationism (which in part showed Kant the way to counter the sceptical challenge via transcendental idealism), and sought not merely to abolish the 'principle of sufficient reason' but to salvage it in a weak form, in turning his attention to the grounds for our beliefs in necessity. We argue further that the 'mathematizability' of properties is not a sufficient criterion to yield realist, non-correlational knowledge, or to demonstrate the 'irremediable realism' of the 'ancestral' statement. Finally, we contend that Meillassoux himself relies on a certain 'Kantian moment' which exempts the reasoning subject from otherwise 'omnipotent' chaos, and that ultimately the 'speculative materialist' position remains exposed to the original Humean sceptical challenge.