3: Visual Word Recognition: Problems And Issues
Kenneth I. Forster · The Mental Lexicon · 2007
The study of visual word recognition is about about one of the most fundamental properties of the brain-the ability to store and rapidly retrieve information about familiar visual patterns. Visual word recognition is also particularly appropriate as an arena to stage the battle between different models of pattern recognition, and it is no accident that much of the debate about neural network models of cognition centers on this topic. This chapter discusses the central issues in the field. It talks about theories of complex cognitive systems, Chomsky's distinction between descriptive adequacy (of a grammar) and explanatory adequacy, the concept of parallel activation, semantic categorization and masked priming. The development of computational models is a move in the direction of providing detailed descriptions of actual processing mechanisms. Current procedures deal only with the means of distributions, and a significant effect in an item analysis can easily be generated by just a small subset ofthe items. If measures of generality were also included, it might turn out that much of the disagreement about experimental findings would eventually disappear.