ABSTRACT
Models of even small parts of the reading process are becoming more and more complicated. At conferences, people have trouble getting their model on a single slide. Clearly, for lovers of parsimony, this is a regrettable situation. While it may turn out that the complexity of the models is only commensurate with that of the subject process, this need not be conceded given our current knowledge. The thesis of this chapter is that a simple mechanism which has hitherto been overlooked can account for at least as much data as any of the extant, complex models.
