Abstract
This study was designed to explore the production of word stress following traumatic brain injury TBI. Ten subjects with TBI and ten matched normal controls produced a sentence with stress elicited on different words. The difference between stressed and unstressed productions of the same word was calculated for intensity, fundamental frequency and duration. Subjects intensity range, fundamental frequency range and vital capacity were also obtained. Naive listeners judged which word was stressed within each sentence. Individuals with TBI were significantly less accurate conveying intended stress compared with normal controls. Individuals with TBI produced significantly less difference in duration between stressed and unstressed words. There was no correlation for either group between percentage change in intensity, fundamental frequency, or duration and the related physiological range. Durational control requires subtle physiological adjustments that individuals with TBI may be unable to accomplish. Further, compensatory strategies may place excessive cognitive demands on the speaker. Thus, the production of stress contrasts may not be amenable to therapeutic intervention. Rather, listeners may be required to rely on context to infer intended stress.