Lexical Competition Effects in Aphasia
Some aphasic patients show single word production deficits in some situations where object naming is required (e.g., they perform well when objects are presented in unrelated groups (e.g., Cat, Fork, Bread...), but deteriorate when the same items are presented in semantically related groups (e.g., Cat, Cow, Dog...)) (see Wilshire & McCarthy, 2002). We investigated whether context-sensitive single-word production impairments reflect an impaired ability to resolve lexical competition. Three groups of participants (non-fluent aphasics, fluent aphasics, and older controls) completed four tasks that manipulated lexical competition: 1) A category exemplar task, where a high competition condition involved generating items from broad categories (e.g., Animals: "Cat. Dog" etc.), and a low competition condition involved generating items from narrow categories (e.g., Pets: Cat. Dog" etc); 2) A verb generation task, where participants were presented with objects and were required to generate related verbs. The high competition objects were related to a range of verbs (e.g., Penny: Spend"/"Pay"/"Buy" etc), and the low competition objects were related to one dominant verb (e.g., Scissors: "Cut"); 3) A name agreement task where a high competition condition involved naming low name agreement objects (e.g., Artist/Painter), and a low competition condition involved naming of high name agreement objects (e.g., Anchor), and; 4) A sentence completion task, where extrinsic competition was introduced via presentation of auditory distracters. The low competition distracters did not make sense (e.g., Barry wisely chose to pay the RANGE: "Bill"/"Cashier" etc), whereas the high competition distracters did (e.g., Barry wisely chose to pay the FINE: "Bill"/Cashier" etc). Our first hypothesis was that all participants would show high competition costs in increased response latencies and/or decreased accuracy. At the group level, this hypothesis was supported in all four tasks. At the individual level, there was mixed support as some participants showed predicted effects on the verb generation, name agreement, and sentence completion tasks. The second hypothesis was that exaggerated competition costs would occur in some or all non-fluent aphasics. At the group level this hypothesis was not clearly supported on any task. At the individual level there was mixed support, with some indications that non-fluents may be more likely to show significant competition effects than fluents. The third hypothesis was that non-fluent aphasics with relatively well preserved single word production but relatively impaired sentence production may be most likely to show exaggerated lexical competition effects. There was little support for this hypothesis. It was concluded that the data do not support the hypothesis that context-sensitive single-word production impairments are symptomatic of an impaired ability to resolve lexical competition. However, we have gained information on how heterogeneous aphasics perform on tasks that manipulate lexical competition, and we have gained some insights that may direct future research down a path towards more informative results, and increased knowledge on the complex process of speech production.