Effects of Probabilistic Contingencies on Word Processing in an Artificial Lexicon

Artificial lexicons have been used with eye-tracking to study the integration of contextual (top-down) and bottom-up information in lexical processing. The present study utilized these techniques to study the role of probabilistic information in lexical processing. Participants were trained to associate novel nouns and modifiers, with certain combinations occurring more frequently than others. Participants heard a modifier-noun phrase and were asked to select the words in a display. We predicted that participants would make anticipatory eye movements to nouns based on the probabilities they previously learned. While no anticipatory effects were found, delayed effects consistent with our predictions were found.