Shape Discrimination in the Octopus

Abstract

IN a recent communication¹, P. C. Dodwell raises some doubts about the adequacy of a theory of visual shape discrimination in octopus which I outlined in Nature ². Dodwell makes four points: (1) He thinks it would be difficult to devise an analysing mechanism which would analyse relative differences in rates of firing at different points over a row of neurons: such a mechanism is necessary on my theory to explain the discrimination of certain figures (for example, a square and a circle). (2) He maintains that the theory fails to account for discrimination of shape irrespective of the locus of excitation on the retina. (3) He writes: “the reason for postulating the array of cells [in the optic lobes] becomes obscure, because … the analyser … should enable it to deal directly with information from the retina”. (4) He maintains that a theory of shape discrimination which he is putting forward overcomes the difficulties inherent in my own theory by transforming a spatial pattern into a temporal pattern.