Précis of Projection and Realism in Hume's Philosophy

The title of my book, Projection and Realism in Hume's Philosophy (Oxford: Clarendon Press, 2007), might mislead. One might protest, with some justification, that since neither "projection" nor "realism" is Hume's term and that both carry a severe threat of anachronism, discussing them in connection with Hume is misguided. Why might the readers of this journal wish to read such a work?

Well, the first thing to note is that Hume's name has come to be associated with the metaphor of projection, understood as having some kind of "non-realist" connotations, and, at the same time, he attracted readings that make him a "realist" of some sort or another in different areas.¹ So, there seems to be some tension here. Furthermore, the terms themselves can add unnecessary noise, not least because there is no such thing as "realism" per se, let alone any settled way of understanding the metaphor of projection. Attempts to define "realism" fall foul of different objections, and in any case what is a non-realism for some (behaviorism, say) is a realism for another.² Simon Blackburn's use of the metaphor of projection differs greatly from Freud's.³ And even if there were fixed contemporary points, an exercise in taxonomy is not a thing of great fascination and is, indeed, damaging with respect to a thinker such as Hume whose very brilliant uniqueness is one of the things exerting its pull on us. My question, then, is what is it about Hume's discussions that occasion labels such as "projection" and "realism," labels that appear to connote very different attitudes? The exercise provides a focus for understanding these two apparently different directions in Hume's philosophy. How are we to understand Hume's talk of "gilding and staining" in the context of claims that seem to make straightforward identifications of virtue with the useful and agreeable? My aim, then, is to understand key aspects of Hume, not by bringing to the text preformed notions of what "projection" or "realism" consist in, but by looking at the arguments and texts that occasion those terms.

What emerges is, I think, a nuanced reading of aspects of Hume's philosophy that is illuminating, even if sometimes a little unusual. But I was equally interested in interrogating the projection metaphor on independent philosophical grounds. Hume and the metaphor of projection are constantly joined, and so it seemed that his text afforded the occasion to discuss a term that throws as much shadow as light. (I am pleased that it has made some impact on discussions in areas as diverse as Kant on the experience of time and sexual objectification.)⁴

The book is divided into three parts, one concerning religious belief and belief in the external world, the second on modality and the self, the third on sentiment and values. I begin by considering projection. An initial thought about projection is that it is a matter of representing something "in here" to be "out there." I call this "feature projection." If one turns to the Natural History of Religion, talk of feature projection might be appropriate, because Hume's explanation of the emergence of polytheism appeals to our disposition to ascribe psychological states to natural objects and events. But this initial thought, though not mistaken, misses different and equally interesting ways of understanding projection, and furthermore, it cannot account for other elements of Hume's thought that involve projection. For example, when we say that belief in God is a projection of our fear, we are not representing the world to contain our fear but saying that we hold the belief because we are fearful. A belief or some other way of taking the world might be a projection of a feature of our mental life in the sense that it explains why the subject has that belief. I call such uses instances of explanatory projection. The relevant explanations are projective because they implicitly contrast with alternative "detective" explanations, where the subject holds the world to be thus and so because they are appropriately responsive to...