Is there serial processing in the reading system; and are there local representations?

There is wide agreement that the goal of any cognitive theory of lexical processing must be to describe the mental architectures and processing components that enable the transformation and retransformation of information across three broad domains of knowledge: orthography, phonology, and semantics. There is also a degree of consensus that processing within and between these broad domains of knowledge can be understood in terms of a triangle with connections between each domain,1 as indicated in Figure 1.1. This triangle framework is a venerable one, dating back at least to Baron (1977; see Coltheart, 2005).