Background: Previous studies have shown that brain-damaged patients with selective deficits to phonological processes produced frequent phonological errors and similar error patterns in all spoken tasks, exhibited the effects of word frequency, grammatical class, and imageability, and were unable to make rhyming judgements, due to impaired lexical retrieval and/or phonological representations.
Aims: This paper describes a Cantonese-speaking brain-damaged patient, LKK, whose performance patterns in spoken tasks indicate impairment to both the lexically mediated non-semantic and semantic pathways of oral production, as well as the phonological output buffer.
Methods & Procedures: A range of tasks was conducted including repetition, reading aloud, oral naming, written/spoken word-picture matching, non-verbal semantic tests, written lexical decision, and homophone judgements.
Outcomes & Results: LKK performed normally on written lexical decision, word-picture matching, and non-verbal semantic tests, but he was unable to make homophone judgements and showed impaired production in all oral tasks. He was better able to read aloud names of objects than to name them. He also made more semantic errors in naming than reading. His accuracy in reading single words was affected by word frequency and form class. Further observations of his oral production included better (but nevertheless impaired) performance on repetition than reading and naming, a consistent effect of word length across tasks, and a tendency for phonological errors to occur on the coda compared with the onset.
Conclusions: There was sufficient evidence for deficits of the phonological lexicon and/or the access to it along the non-semantic route and the semantic pathway at the post-semantic level in LKK. The effect of word length and comparable patterns of error distribution across spoken tasks suggested additional impairment to the phonological output buffer. The different levels of accuracy in repetition, reading, and naming, as well as the differential rates of semantic errors in these tasks were consistent with predictions of the summation hypothesis.
Open access content
Free trial content