What is the nature of neural representations underlying high
level cognitive processing, such as language? I suggest the that
question of how letter position within a word is encoded offers a
circumscribed problem for investigation into this realm. First I
give an overview of the problem, and proposed solutions. Then I
discuss some relevant experimental evidence: positional error
patterns from patients suffering from acquired dyslexia, letter
perceptability patterns from normals, and word-naming priming
results. Following a brief tutorial on methods used in
computational modeling of neural processing, I describe my
proposed model of letter-position coding. I show, via simulations,
how this model accounts for the experimental data.
Last updated 11/12/00
Comments, Questions about this page to Colin
Phillips
