Adverb distribution in Broca's aphasia; A different view on the Tree Pruning Hypothesis
MetadataShow full item record
The goal of this study is to test the plausibility of the Tree Pruning Hypothesis (TPH; Friedmann and Grodzinsky, 1997) for Broca’s aphasia by examining adverb distribution in spontaneous speech. The TPH claims that in Broca’s aphasia higher nodes of the syntactic tree are unavailable during sentence production, whereas lower functional heads are available; the syntactic tree is pruned. The plausibility of the TPH was analyzed using the universal adverb hierarchy by Cinque (1999). Cinque claims that adverbs specify distinct functional heads. In this approach adverbs are located at specific fixed positions in the syntactic tree. Cinque’s claim combined with the TPH leads to the prediction that adverb distribution should reveal the pruning phenomenon during spontaneous speech of Broca’s aphasic subjects. This prediction was tested by analyzing spontaneous speech samples of ten Broca’s aphasics. Ten unimpaired controls and ten Wernicke’s aphasics acted as control groups. The universal adverb hierarchy is divided in six hierarchical groups for the analysis. In addition, a more semantic classification of adverbs in six groups was used which does not follow Cinque’s hierarchy. Broca’s aphasics preferred low adverb groups most. This was not observed in unimpaired controls. This distinct preference for low adverb groups is predicted by the TPH. The preference pattern for the remaining higher adverb groups in Broca’s aphasics was almost equal to unimpaired controls. Preservation of this preference pattern for higher adverb groups is not compatible with the TPH, since higher syntactic nodes are assumed to be inaccessible. The results suggest the existence of a metaphoric ‘on/off switch’ for tree pruning. When the ‘switch’ is on, high syntactic nodes are inaccessible. When the ‘switch’ is off, high syntactic nodes are available. If the ‘switch’ is turned off a preference pattern in Broca’s aphasics is observed which resembles preferences of unimpaired controls. However when the ‘switch’ is turned on only low adverbs can be produced. It is proposed that this ‘on/off switch’ is explained by Avrutin’s Weak Syntax theory (2006). Avrutin claims that in unimpaired controls narrow syntax processes are the most economical solution for information encoding. Information however might be encoded by discourse, provided a special register is present. Avrutin states that in Broca’s aphasia narrow syntax is weakened because of brain damage and loses its status of most economical system. Consequently, narrow syntax and discourse compete with each other, which results in a 50/50 chance for either system to win. Situations in which narrow syntax wins the competition are compatible with situations in which the ‘tree pruning switch’ is off. When discourse wins, the ‘tree pruning switch’ is turned on. The alternative non-syntactic division of adverbs did not reveal a semantic deficit. In fact the alternative division further supports the proposal of an ‘on/off switch’ for tree pruning. The expected differences between Broca’s and Wernicke’s aphasics regarding syntactic processes could not be verified in the hierarchical grouping of adverbs. In fact, Broca’s and Wernicke’s aphasics’ adverb distribution bear considerable similarities. A syntactic deficit appears to be present in Wernicke’s aphasia also. Close analyses of the alternative division however showed that production patterns in Wernicke’s and Broca’s aphasics differ. Regarding the highest and lowest adverb groups Wernicke’s aphasics seem to perform more equal to unimpaired controls than to Broca’s aphasics. Consequently, it is ruled out that Broca’s and Wernicke’s aphasics have a similar underlying deficit. A deficit of both syntactic and lexical processes is proposed for Wernicke’s aphasia.