Most quantitative models of visual word recognition assume that letter information is directly combined with word knowledge to produce recognition of short words (3 - 5 letters). By contrast it has been argued that, for longer words, intermediate units such as syllables and/or morphemes are involved. A lexical decision experiment was carried out to investigate this issue. The stimuli were compound words and nonwards made of a combination of lexical and nonlexieal parts. The fixation point was varied within the stimuli. The results indicate that nonwards may indeed be decomposed into their constituent parts. However, words appear to be recognized as a whole, even while more information may be extracted from their initial than from their final parts.
|Number of pages||7|
|Journal||IPO Annual Progress Report|
|Publication status||Published - 1984|