Repetition priming of words, pseudowords, and nonwords

In this thesis, I report five experiments designed to assess repetition priming for words, pseudowords, and nonwords. The task used combines an implicit perceptual fluency measure and a recognition assessment for each list item. Words and pseudowords generate a consistently strong repetition effect even when there is a failure to recognize the stimulus when probed immediately following the perceptual fluency measurement of that stimulus. Nonwords show a weaker overall repetition effect which is resolvable from noise for items failing recognition in two experimental conditions. A connectionist model based on the work of McClelland and Rumelhart (1985) is proposed as a way to understand the mechanisms potentially responsible for the pattern of findings. While the error-driven nature of learning in the McClelland and Rumelhart (1985) model results in a poor fit to the nonword priming data, this attribute is not endemic to all connectionist models. Using a model based on Hebbian learning, we can instantiate a property that I believe is characteristic of implicit memory--that learning is primarily based on the strengthening of connections between units that become active during the processing of a stimulus. This model is found to provide a far more satisfactory account of the data.