KNOWLEDGE wonder how we are to distinguish the belief that Caesar died in his bed from a memory of Caesar dying in his bed, since vivacity is the mark of each But other diYculties are not merely internal The crucial problem is that belief need not involve imagery at all (when I sit down, I believe the chair will support me; but no image or thought about the matter enters my mind) And when a belief does involve imagery, an obsessive fantasy (of a spouse’s inWdelity, for instance) may be livelier than a genuine belief There is something pitiable about Hume’s delusion that in presenting his few scattered remarks about the association of ideas he was doing for epistemology what Newton had done for physics But it is unfair to blame him because his philosophical psychology is so jejune: he inherited an impoverished philosophy of mind from his seventeenth-century forebears, and he is often more candid than they in admitting the gaps and incoherences in the empiricist tradition The insights that make him great as a philosopher can be disentangled from their psychological wrapping, and continue to provoke reXection His treatment of causation, of the self, of morality, and of religion will be treated in the appropriate chapters His main contribution to epistemology was the presentation of a new form of scepticism This begins from the distinction, which we have met in several philosophers, between propositions expressing relations of ideas, and propositions expressing matters of fact The contrary of every matter of fact is possible, Hume says, because it can never imply a contradiction That the sun will not rise tomorrow is as intelligible and coherent as the aYrmation that it will rise Why then we believe the latter but not the former (E II 25 – 6)? All our reasonings concerning matters of fact, Hume argues, are founded on the relation of cause and eVect But how we arrive at our knowledge of causal relations? The sensible properties of objects not reveal to us either the causes that produced them or the eVects that will rise from them Merely looking at gunpowder would never tell you that it was explosive; it takes experience to learn that Wre burns things up Even the simplest regularities of nature cannot be established a priori, because a cause and an eVect are two totally diVerent events and the one cannot be inferred from another We see a billiard ball moving towards another, and we expect it to communicate its motion to the other But why? ‘May not both these balls remain at absolute rest? May not the Wrst ball return in a 154