«Elliott Sober has written an interesting paper containing striking generalizations and speculations that relate pictures, language, and perception. His inquiries display an admirable breadth and imagination, and I think that we must immediately grant some of his major points. Sober is right, for example, to reject the traditional objections which he considers to postulating mental pictures. He also seems correct in holding that the currently available evidence-for postulating pictures in the brain is, as he presents it, at best inconclusive.
Nevertheless, not all of Sober’s views are beyond challenge. In the present essay, I will question (i) whether Sober’s functional characterization of pictures captures important aspects of pictorial representation, (ii) how he understands the notion of a picture’s content, (iii) whether, as Sober claims, every picture is representational, (iv) the Sober-Berkeley thesis that pictures have a peculiar specificity, (v) Sober’s idea that concatenation can achieve pictorial conjunction, and (vi) his account of logical form and pictures. I will also state, quite informally, my own position on some of these topics. A general consequence of my criticisms is that Sober’s way of characterizing pictures is less adequate than he supposes. This consequence, to which I will not return below, suggests that Sober’s own reasons for not now postulating mental pictures may need re-examination.
Some of the following discussion is supported by the results in my ‘The Logical Structure of Pictorial Representation.’ But none of my present observations presupposes any familiarity with that paper.»