Gradience, Phonotactics and the Lexicon in English Phonology

Experimental work has established that when subjects judge the phonological wellformedness of nonsense forms, they are strongly affected by the frequency of the phonological elements of the form and by the number of actual words that such a form is similar to. These results challenge phonological theory by suggesting a central role for frequency and the lexicon. In this paper, I review these results and show how they can be easily modelled with Probabilistic Optimality Theory. The payoff is that from very few phonological assumptions we can derive virtually the whole panoply of experimental effects. We can also derive various Local Conjunction effects as well.