1. Wittgenstein took this – ‘Not empiricism and yet realism’ – to be something to aim at in philosophy. My concern in this essay is to suggest what he might have meant by this. The task is made tricky by the fact that the term ‘realism’ is used in a variety of ways in philosophy. Wittgenstein himself is, presumably, understanding the term differently when he speaks of refusing both ‘the scepticism of the idealist’ and ‘the assurances of the realist’.

To give a rough indication of how I will use the term in this essay: ‘realism’ is the view that things and their properties exist independently of human concepts and practices. The formulation leaves room for the suggestion that we should be realists about some but not other of the things people speak of (perhaps trees but not unicorns), and about some but not other general areas of our thought and speech: perhaps science but not ethics. I will suggest that there is nothing in Wittgenstein to cast a completely general doubt on realism so understood. What, then, does he have in mind when he rejects ‘the assurances of the realist’? One possibility is this: he is insisting that it is not the business of philosophy to offer assurances that, for example, trees and mountains exist independently of us. Another, more relevant to my concerns in this essay, is that he is rejecting the assurances of views that hold that we have the ‘right’ concepts: that our experience of the world dictates that these are the concepts in terms of which it is to be described. Such views are, I believe, well characterised as ‘empiricism’. They are, I will suggest in the next section, particularly clearly exemplified in Locke's thinking and may, I believe, be at least part of what Wittgenstein has in mind when he calls for ‘Not empiricism and yet realism’.

2. John Locke remarks that: