Abstract: Ordinary common sense suggests that we have just one set of shape concepts that we apply indifferently on the bases of sight and touch. Yet we understand the shape concepts, we know what shape properties are, only because we have experience of shapes. And phenomenal experience of shape in vision and phenomenal experience of shape in touch seem to be quite different. So how can the shape concepts we grasp and use on the basis of vision be the same as the shape concepts we grasp and use on the basis of touch? I think this is the intuitive puzzle that underlies the question sent by the Dublin lawyer Molyneux to John Locke. This concerns a man born blind, who learns by the use of his touch to discriminate cubes from spheres. Suppose him now to gain the use of his sight. And suppose him to be presented with a cube and a sphere, of nighly the same bigness. Quaere, will he be able to tell, by the use of his vision alone, which is the sphere, and which the cube? (Locke 1975, II/ix/8.)

Abstract: in Enrique Villanueva (ed.), Perception (Philosophical Issues vol. 7) (Atascadero: Ridgeview 1996), 301-318, with replies by Brian Loar and Kirk Ludwig

Abstract: Molyneux's Question, that is, “Suppose a man born blind, and now adult, and taught by his touch to distinguish between a cube and a sphere... and the blind man made to see: Quaere, whether by his sight, before he touched them, he could now distinguish, and tell, which is the globe, which the cube”, was discussed by many theorists in the 17th and 18th centuries, and has recently been addressed by contemporary philosophers interested in the nature, and identity conditions, of perceptual concepts. My main concern in this paper is to argue – against Evans, Campbell, and a number of other contemporary philosophers – that a test of the sort Molyneux envisioned, at least if carefully designed and administered, can indeed be a crucial experiment for the claim that we deploy the same perceptual concepts when identifying shapes by sight and by touch. I will explore some implications of this argument for a theory of recognitional concepts. And I’ll try to trace out some unhappy consequences of various alternative views