A reader's view of listening

There is a view that the fundamental processes involved in word recognition might somehow be different for speech and print. We argue that this view is unjustified, and that the models of lexical access developed for the written form are also appropriate for speech, provided that we allow for obvious differences due to the physical characteristics of speech signals. Particular emphasis is given to the role of word frequency in the recognition process, since this places restrictions on the types of models that can be considered (e.g., the cohort model). We reject the view that there are no frequency effects in spoken word recognition, and we also reject the view that frequency effects in printed word recognition can be relegated to the minor status of a post-access decision effect.