1 See Leibniz, New Essays on the Human Understanding IX, 8;Google Scholar Reid, Thomas Inquiry Into the Human Mind (1764) ch.6, XI;Google Scholar and Thompson, Judith Jarvis ‘Molyneux's Question,’ Journal of Philosophy 71 (1974) 637–50.CrossRefGoogle Scholar

2 E.g., Morgan, M.J. Molyneux's Question: Vision, Touch and the Philosophy of Perception (Cambridge: Cambridge University Press 1977).Google Scholar

3 I do not mean to imply that innate mechanisms can reasonably be postulated without regard for wider empirical evidence. See Eilan, Naomi ‘Molyneux's Question and the Idea of an External World’ in Spatial Representation, Eilan, N. McCarthy, R. and Brewer, B. eds. (Oxford: Blackwell 1993).Google Scholar The point is that for any such postulation to have definite Content, it needs tying to one of the issues above.

4 Reid, Inquiry, ch.6 XX;Google Scholar Berkeley, Essay Towards a New Theory of Vision (1709).

5 For the Reidian view, see Hopkins, R. Picture, Image and Experience (Cambridge: Cambridge University Press 1998),Google Scholar ch. 4. For a Berkeleian approach, see Peacocke, C. Sense and Content (Oxford: Oxford University Press 1983),Google Scholar ch. 1. John Mackie in effect reads Locke as adopting the former position, in Mackie, J. Problems from Locke (Oxford: Clarendon Press 1976), 29–32.CrossRefGoogle Scholar Martha Brandt Bolton reads him as taking the latter view, in Bolton, M. ‘The Real Molyneux Question and the Basis of Locke's Answer,’ in Locke's Philosophy: Content and Context, Rogers, G.A.J. ed. (Oxford: Clarendon Press 1994), 75–99.Google Scholar

6 Since the circular appearance correlates with the presence of a particular 3-D shape, there is a sense in which that appearance ‘represents’ sphericality. Perhaps this yields a derivative sense in which the experience itself also does so. The point is that it does not do so in the sense that is nowadays Standard. That is, the experience does not, in itself, make a claim about how the world before the subjeet is, viz. that it contains something spherical. Locke may not have used ‘represents’ in this sense, but he was certainly au fait with the idea it captures. For, as I go on to explain, his thought is precisely that, since the only connection between the circular appearance and the spherical presence is that the latter standardly causes the former, Molyneux's subject will have to learn that the two are so linked. The force of denying any stronger connection lies in rejecting the idea that the experience represents, in the modern sense, 3-D shape.

7 Compare that in Molyneux's first letter to Locke on the subject (quoted in Park, Desiree ‘Locke & Berkeley on the Molyneux Problem,’ Journal of the History of ideas 30 [1969] 253–60CrossRefGoogle Scholar).

8 Lotze, H. Metaphysic II (Oxford: Clarendon Press 1887),Google Scholar on which see Morgan, Molyneux's Question, ch. 6;Google Scholar and Senden, M. von Space and Sight, Heath, P. trans. (London: Methuen 1960).Google Scholar

9 This is how Heil, John reads Locke: ‘The Molyneux Question,’ Journal for the Theory Of Social Behaviour 17 (1987) 227–41.CrossRefGoogle Scholar

10 Condillac, Treatise on the Sensations (1756), in Philosophical Writings of Etienne Bonnot, Abbe de Condillac, Philip, F. and Lane, H. eds. (Hillsdale, NJ: Lawrence Erlbaum 1982).Google Scholar

11 The traditional reading is not compulsory. On one alternative, Berkeley accepts that vision represents certain properties, but denies that these are properties of the environment. All vision presents us with is colours in a quasi-spatial array. On another, he insists, consistently with his later idealism, that such patterns of colour are as independent of us as any properties we experience. However, on any reading, Berkeley holds a general position about vision which entails that the world as presented in visual experience does not share properties with the world as presented in touch. Thus, on any reading, his position — whether or not it amounts to global scepticism about vision's representational powers — addresses issue One.

12 Berkeley, New Theory, 42;Google Scholar Reid Inquiry, ch. VI, III. Reid's position here differs from that he adopts in the text cited in n. 1. For an explanation of the discrepancy, see Hopkins, R. ‘Thomas Reid on Molyneux's Question’ Pacific Philosophical Quarterly 86 (2005) 340–64.CrossRefGoogle Scholar

13 Jackson, F. Perception — A Representative Theory (Cambridge: Cambridge University Press 1977);Google Scholar Foster, J. The Nature of Perception (Oxford: Clarendon 2000)CrossRefGoogle Scholar

14 Perkins, M. Sensing the World (Indianapolis: Hackett 1983)Google Scholar

15 For instance, for the property Reid called ‘visible figure,’ there is a contemporary debate over whether vision represents it, and one over whether touch does. For the former, see Peacocke, Sense and Content, ch. 1 and Tye, Michael ‘Visual Qualia’ in The Contents of Experience, Crane, T. ed. (Cambridge: Cambridge University Press 1992).Google Scholar For the latter, see Lopes, D. ‘Art Media and the Sense Modalities: Tactile Pictures,’ Philosophical Quarterly 47 (1997) 425–40,CrossRefGoogle Scholar and Hopkins, R. ‘Touching Pictures,’ British Journal of Aesthetics 40 (2000) 149–67.Google Scholar

16 For accessible exposition of these ideas, see T. Crane, ‘The Nonconceptual Concept of Experience,’ in Crane, The Contents of Experience

17 The claim is not that concepts are formed by a process of abstraction: it is just that any concept Applying in experience does so in a way allowing for a good deal of Variation in the details of the scene thereby presented.

18 Prima facie this reintroduces the issue of innateness. Innate concepts are, presumably, not perceptual in the sense defined. Thus framing Two in terms of perceptual concepts leaves it turning on the falsehood of nativism, for those concepts. However, I never claimed that innateness was irrelevant to Molyneux's question; just that its role was subsidiary to that of other issues. If innateness does enter here, it does so precisely in such a subordinate role.

19 Note the disjunction in (1*). It is obvious that there is a concept meeting (1*) and (2) if we deploy a conjunction instead. Consider the concept tactual cube or visual cube.

20 Ex hypothesi touch has failed to produce a cross-modal concept F, but that proves nothing more than that some perceptual concepts are mode-specific, something we already knew (think again of tactual cube).

21 One way in which we may outperform the blind person is that we can deploy concepts derived from combining visual experience with tactual. However, this is irrelevant to Two, which concerns what each modality can by itself provide.

22 Campbell, John ‘Molyneux's Question’ in Perception, Philosophical Issues 7, Villanueva, E. ed. (Atascadero, CA: Ridgeview 1996) 301–2.Google Scholar This is not Campbell's only attempt to identify the underlying issue — see below, n. 24. In what follows I refer to the issue just sketched as ‘Campbell's,’ but do not thereby pretend to solve the exegetical difficulties Campbell's paper presents.

23 Does this commit us to non-conceptual experiential content? Usually that is defined as content that could be enjoyed by subjects lacking the particular concepts required to capture how the world is then represented as being (see, e.g., T. Crane, ‘The Nonconceptual Content of Experience’). What is denied here is stronger: that the subject's visual experience involve his tactual concept. That leaves open whether the content of his visual experience is dependent on his possessing particular concepts (including indeed — though this does not seem likely — the tactual concept cube).

24 ‘Molyneux's Question,’ in Evans, G. Collected Papers (Oxford: Clarendon Press 1985).Google Scholar For examples of Evans's influence, see Eilan ‘Molyneux's Question and the Idea of an External World’; and Campbell ‘Molyneux's Question.’ Despite the Claims expounded above (VI), Campbell takes the deepest issue raised by Molyneux's question to be between ‘radical externalism’ and ‘internalism’ about shape perception (302-3). Since the internalist view is, in essence, that of Evans's B, Campbell accepts Evans's view of the lie of the land here, at least in part.

25 In fact, Evans is here discussing the case of tactual and auditory experience. However, when he turns to vision he clearly intends to apply the same argument to the visual-tactual case.

26 For defence, and some examples, see Campell, J. Post, Space and Self (Cambridge, MA: The MIT Press 1995),CrossRefGoogle Scholar ch. 1.

27 Hopkins, R. Picture, Image and Experience (Cambridge: Cambridge University Press 1998), ch. 7Google Scholar

28 See, as well as the last citation, M.G.F. Martin, ‘Sight and Touch’ in Crane, ed., The Contents of Experience.

29 See, for instance, Kennedy, J.M. Drawing and the Blind (London: Yale University Press 1993).Google Scholar For some discussion of the bearing of this work on Molyneux's question see Lopes ‘Art Media and the Sense Modalities.’

30 For detailed comments on written versions of this paper, I thank Laura Berchielli, Darragh Byrne, the late Greg McCulloch, Paul Noordhof, David Owens, and three anonymous referees. Mike Martin was especially generous, offering insightful comments on several versions. Thanks are also due to Edward Craig, Jane Heal, Penelope Mackie, Hugh Mellor, Harold Noonan, Donald Peterson, Joss Walker, and other members of audiences in Cambridge, Birmingham, Sheffield, and Glasgow. Some of the work on this project was made possible by the award of a Philip Leverhulme Prize.