Tuesday, July 21, 2009

Insightful problem solving in birds.

An exchange of letters in PNAS as followup to the subject of the July 3 post on tool use in birds, in which the birds spontaneously use a series of tools to obtain a reward. Lind et al. question whether insight (internal mental modeling of possible problem solutions) or shaping (through successive reinforcement of behaviors) was involved. Authors Bird and Emery respond:
...our study found that rooks not only learned to obtain food by dropping a stone into a tube to collapse a platform, but that they were able to spontaneously choose the correct size and shape tool when presented with subsequent transfer tests and spontaneously solve a completely new task, whereby they had to lift a bucket by using a novel hook tool and manufacture a hook tool from a piece of wire. While we agree that the question of animal “insight” is a complex one in light of studies on chimpanzees and pigeons, we find shaping an unsatisfactory explanation for the spontaneous choice of the correct tool and the bending of the wire. We do not dispute the argument that initial acquisition of stone dropping was brought about through shaping, however, because shaping requires one or more stimulus/behaviour–reward pairings, a novel behaviour such as hook manufacture cannot simply be a case of shaping.

No comments:

Post a Comment