The above diagram summarizes, as succinctly and clearly as I can manage, the reading of Hume given by Deleuze in his amazing book Empiricism & Subjectivity. Though often ignored and rarely read, I actually think this is one of Deleuze’s most important books, and reading it slowly and carefully was crucial in helping me pull other pieces of Deleuze’s project together and greatly enhanced my understanding of his work.
Not surprisingly, in Deleuze’s reading Hume has more in common with Bergson than with Locke and Berkeley. Deleuze’s typically subtle and effortlessly brilliant reinterpretation of Hume’s “empiricism” allows Hume’s most important work, the Treatise of Human Nature, to be read alongside Spinoza, Nietzsche, and other Deleuzian mainstays. Unlike both Cartesian rationalists and British empiricists, Hume, like Bergson, posits the subject and the material world not as two absolutely divided substances but as associative constellations formed by a kind of “carving up,” what Hume calls ‘extension’ and ‘correction,’ gradations of difference that produce a new phenomenon immanently from within an existing phenomenon. Belief is the affective condition which moderates the subject’s understanding of the given; invention is the affective condition which moderates the subject’s action in the given. The difference between the two is qualitative, but, as in Bergson’s “memory” and Nietzsche’s “relations of force,” the qualitative difference is produced by a process of quantitative distinction.