This article examines the relationship between eye movements and word recognition, moving from what we know about isolated word recognition to the reading of text. There are clear points of contact between behaviour in laboratory tasks with isolated words and word recognition in text; for instance, Schilling et al. (1998) have shown consistent word frequency effects in naming, lexical decision, and fixation times in silent reading. However, there are also major differences between these two domains. Radach and Kennedy (2004) cite “integration with work on single word recognition” as an issue for future reading research. Equally, though, research on isolated visual word recognition can benefit from a consideration of normal reading; when a word is processed in isolation, its normal context has been replaced by a “null” context. The article concentrates on some of the anatomical and computational principles governing access to the mental lexicon. First, it considers eye movements in reading isolated words and then explores how fixation in word recognition is controlled. Finally, the article discusses the issues raised by binocularity.
Neighborhood frequency effects in visual word recognition: A comparison of lexical decision and masked identification latencies
