When racquets are baskets but baskets are biscuits, where do the words come from? A single case study of formal paraphasic errors in aphasia

This paper provides evidence for the existence of real-word errors phonologically related to targets (formal paraphasias) in the naming attempts of a person with aphasia, MF. Zn particular, this is the first demonstration that such errors are genuine lexical errors and not jargon homophones (phonological errors that happen to be words by chance; Butterworth, 1979). This finding was replicated on a second naming attempt. MF's naming accuracy was influenced by length and by imageability/concreteness. There was no effect of frequency or age of acquisition. In addition to formal errors, he made semantic errors and nonword errors that were phonologically related to the target. The apparently mixed (semantic and formal) errors are argued not to be truly mixed on the grounds that they do not share phonology with the targets at greater than chance rates. A detailed analysis of MF's errors is followed by discussion of the findings in relation to a variety of models of speech production including those with one (Morton, 1970) and two (Butterworth, 1989, 1992) stages of lexical access and an interactive activation model (Dell, 1989). Whereas all of the models can account for some aspects of his naming, no model, as it stands, can give a complete account of MF's performance.