VISUAL WORD RECOGNITION ACROSS ORTHOGRAPHIES - ON THE INTERACTION BETWEEN CONTEXT AND DEGRADATION

Single-word semantic context and stimulus degradation yield an overadditive interaction during word recognition for readers of alphabetic English (a deep orthography in which lexical access relies primarily upon orthographic information). Lukatela and Turvey reported that this over-additive interaction is absent when subjects read a shallow orthography like Serbo-Croatian. They argued that this arises because subjects rely upon a prelexical phonological code for lexical access. The present experiment on the reading of alphabetic English suggests that Lukatela and Turvey's results are due to their choice of a baseline against which priming is assessed. Some implications for models of word recognition are noted.