This article investigates the role of instinct in Hume's understanding of human reason. It is shown that while in the Treatise Hume makes the strong reductive assertion that reason is ‘nothing but’ an instinct, in the First Enquiry the corresponding statement has been modified in several ways, rendering the relation between instinct and reason more complex. Most importantly, Hume now explicitly recognises that alongside instinctive experimental reasoning, there is a uniquely human intellectual power of intuitive and demonstrative reason that is (...) not itself an instinct. At first sight it may look as if this intellectual reason, that is capable of grasping ‘relations of ideas’, is not even grounded in instinct but is a thoroughly non-natural element in human nature. On closer analysis, however, it is shown that intellectual reason, in its apprehension of ‘abstract’ and general relations, is dependent on language – the use of ‘terms’ – and that language itself is grounded in instinctive associations of ideas. Thus, Hume's overall view is that even the intellect is an outgrowth of instinct and his conception of human nature is, therefore, shown to be fully naturalistic. Yet this naturalism can still make room for the ‘exceptionalism’ of human mathematical thought, which has no counterpart in the animal kingdom where language is lacking. (shrink)