abstract Renske Hoedemaker

Renske Hoedemaker (University of North Carolina at Chapel Hill, USA)

Eye movements and recognition of words in lists.

While studies of the recognition of words presented in text and in isolation have yielded highly consistent evidence about many facets of lexical processing, there are important exceptions: Semantic priming is very robust in tasks where words are recognized in isolation but is quite limited during text reading, and orthographic neighborhood density facilitates response times in isolated word recognition tasks but is associated with elevated processing times during the reading of text. Following Brysbaert (1995), we use eye movements to study recognition of words in lists during tasks that combine the explicit control of subjects’ goals found in isolated word-recognition tasks with the fast, well-practiced ocular response mode used in reading text. In the ocular lexical decision task (LDT) eye movement responses replace the typical manual response mode. The distributions of gaze durations in the ocular LDT (analyzed by fitting ex-Gaussian distributions) show effects of lexical characteristics that may be obscured or distorted by the high response-time floor found with manual responses. Semantic priming during the ocular LDT was most pronounced for slower responses, suggesting recruitment of prime information occurs only on trials that require more effortful target resolution. In contrast, high neighborhood density was associated with faster word recognition times across the reaction time distribution, suggesting an encoding-based facilitation effect. When the task was changed from lexical decision to recognition memory, the effects of semantic priming and neighborhood density were weakened substantially. Taken together the results show how task goals and time required to perform a task interact to determine which lexical characteristics most prominently influence word recognition.