SHAPE RECOGNITION IN RATS

Abstract

Reasons are given for doubting whether certain of Lashley's experiments, purporting to be investigations of shape discriminations, really are such. In particular his finding concerning square‐circle discrimination cannot be held to lend support to Deutsch's theory of shape recognition. An experiment is reported which shows that rats can discriminate square from circle under certain conditions of pre‐training, and do so under these conditions by responding to horizontal base of square; response is transferred to other shapes having a horizontal base, but not to shapes lacking it. There is some evidence to show that this happens for small shapes‐where Deutsch's system could be operative‐and this is therefore evidence against his theory. A theory of shape recognition is proposed of same logical type as Deutsch's, which accounts for results here reported, accords well with other known characteristics of shape recognition in primitive visual systems, and presents less difficulties than Deutsch's system as a possible neurological model.