There is not clear-cut boundary between heredity and environment, which leads to Minsky's notion of "predestined learning". "Why make the brain use a tedious learning process when the final outcome seems so clear? Why not build in the answer genetically? One reason could be that learning is more economical." Why not make learning about space predestined -- this will give economy of learning.
Constraints on learning make it predestined. "We acquire our conceptions of space by using agencies that learn in accord with processes determined by inheritance. These agencies proceed to learn from experience -- but the outcomes of their learning processes are virtually predestined by the spatial geometry of our body parts."
(NB: The hypothesis emerging here is that mental universals -- e.g., conservation of quantity -- may result from learing in an enviroment that is structure or constrained in an important way. Two things to keep in mind though -- the constraints could be in the structure of the environment being learned about, or could also be in the structure of the system that is learning! One example of this is Dawson, Kremer and Gannon's paper on finding simple cell receptive fields in connectionist networks, but only when the network is initially constrained.)
Pearl Street | Society Of Mind Home Page | Dawson Home Page |