Home > Messages from the Ether > Mental representations

Mental representations

A recent Cornell study moves further towards a weakening of the computational model of the mind. My most in-depth study of this subject was from May 2003 while reading Stephen W. Horst's Symbols, Computation, and Intentionality. The book was published in 1996 and in it he criticized the computational model. I'm certainly not in the scene, but I really didn't think that the computational model had that strong a following anymore.

Anyway, the Cornell methods reminded me of some recent observations of my own behaviour. In their study, students were given the name of an object and had to point to it from a pair of images. Sometimes the non-correct object would have a name that sounded distinct from the correct object; sometimes the names would sound similar. The study found that students took a measurably longer time to point to the correct object when the names sounded similar ('candle' and 'candy' were more difficult to differentiate than 'candle' and 'jacket'). Over the past several months, I've noticed an odd pattern in how I mistype letters on the keyboard. While my mistypes will often be from swapped letters ('teh'), there are a considerable number of occations when I'll mistype similar looking letters. So, I'd type 'q' for 'g' or 'p' for 'b'. The mistakes involve different fingers and/or different hands and so have nothing to do with the mechanics of typing.

This is all very unscientific (how many times do I mistype unrelated, non-similar keys?), but interesting w/r/t the Cornell study. The arbitrary linguistic label for an object can affect our visual recognition of that object, just as the arbitrary visual representation of a letter/sound can affect our recognition of it. Equally interesting: the Cornell study may support similar linguistic games used in the analysis of dreams.

The author of the study is Michael J. Spivey. I swear I've heard that name before but can't find the reference right now.