Empiricism and Experience: Two Problems

Click to increase image sizeClick to decrease image size

Notes

1 I will use the abbreviation E & E (followed by page references) for Gupta, Empiricism and Experience (Oxford: Oxford University Press, 2006).

2 An example will help. Think of three different scenarios involving a subject S (and of the experiences e–e′ she may have in them): a green wall under normal lighting conditions covers S’s visual field (e); a blue wall covers S’s visual field, but she wears yellow sunglasses (e∗); S dreams that she is seeing a green wall (e′). The experiences e, e∗ and e′ issuing from these situations are different. Different things are true of the world and of S if S has any of them. However, they are subjectively identical. In all of them things would look the same to S: S would feel as if she were seeing a green wall. MFE captures the intuition that this fact holds of experience in general, rendering it impossible for S to recover the state of the world w and the state of the self s (i.e., the world–self combination <w, s>) that resulted in her experience e since many different world–self combinations are compatible with the subjective character of e (E & E, pp. 5–9).

3 The problems of the Cartesian view (CV) are discussed in E & E, pp. 40–6; 55–8. One of the main arguments against it is given in E & E, pp. 142–54. Gupta grants the coherence of CV, but questions its truth. The phenomenology of our experience is compatible with different metaphysical possibilities, not just with the one stated by CV. If we had direct awareness of external objects in a public space (rather than of private objects in a subjective world), the phenomenology of our experience would be the same. The status of CV is no better than that of any other view: on the one hand (by IE), we can neither rule out nor show the truth of CV a priori; on the other (by MFE), the phenomenology of experience cannot establish by itself the truth or falsity of CV.

4 There is an initial tension between Gupta’s revolutionary claim that the given in experience is not propositional and the fact that we can describe it as a collection of conditional propositions like ‘If I suffer e and v is correct, then the propositions in Γ _e (v) are true.’ I cannot talk about this issue here, but I think it could be accommodated in Gupta’s theory leaving its main virtues untouched.

5 Gupta makes three crucial claims: (1) the logical form of the given is that of a function or argument form; (2) interdependent notions are fundamental in epistemology; (3) convergence in revision processes introduces rational epistemic obligations.

6 NCS is a minimal realist in Gupta’s sense (E & E, pp. 180–5).

7 This paper was finished during the tenure of a postdoctoral grant at Arche at the University of St Andrews funded by the Spanish MEC and is part of the research project HUM 2006–04907/FISO, also funded by that institution. Thanks to A. Gupta and to those who attended two seminars at the Universities of Valencia and St Andrews for their comments.