Richard Gregory: “Why is learning slow? Is it set by physiological (hardware) limitations, or is it due to cognitive (software) strategy?
As learning can be single-trial, in dangerous situations, the general slowness may well be a cognitive strategy – following Mill’s Methods for induction. Many instances are needed to establish that A is related to B. The more the noise (and other possibilities etc.), the more instances are needed, so learning should be slower.
The experiment would control environment ‘noise’, to see whether learning is generally faster in a simpler, more easily predictable world.
I tried this 50 years ago on Guppy fish, but they died! (The tank had metal sheets with many regular holes, giving moire patterns, which moved more or less consistently with the movements of the fish.)
If children are brought up in a simpler, more regular environment – do they learn faster, with fewer trials? Would this extend to any learning? How generally is the inductive strategy applied?”
Professor Richard L. Gregory CBE FRS is Emeritus Professor of Neuropsychology at the University of Bristol.
Photo credit: Martin Haswell.