Frontiers in Psychology (May 2014)
The ability of left- and right-hemisphere damaged individuals to produce prosodic cues to disambiguate Korean idiomatic sentences
Abstract
Introduction Speech prosody plays a significant role in signaling speaker’s intention to convey literal or non-literal meanings (i.e., idioms). Investigations of acoustic features associated with idiomatic utterances have provided evidence that those utterances can be signaled by specific prosodic contours. Considerable research has shown conflicting results regarding the effects of brain damage on the production of prosody (Baum & Dwivedi, 2003; Cancelliere & Kertesz, 1990; Gandour, Dzemidzic, Wong, Lowe, Tong, & Hsieh, 2003; Van Lancker, 1980; Van Lancker & Sidtis, 1992). Relatively little research has been done on role of prosody in the production of idiomatic expressions by brain-damaged individuals. The primary focus of this study was to examine the contribution of the left and right cerebral hemisphere in producing prosodic features for idioms that are “ditropically ambiguous” (having either a literal or an idiomatic meaning, e.g., David spilled the beans) Korean sentences. Method and Procedures Individuals with left- (LHD) and right-hemisphere damage (RHD) due to cerebrovascular accidents (CVA) and age- and education-matched health controls (HC) produced 6 ditropically ambiguous sentence pairs, one with the idiomatic and one with the literal meaning, in two different production tasks, an elicitation task and a repetition task. Speech samples were analyzed utilizing acoustic measures, native listeners’ identification of sentence type and perceptual ratings, and expert listeners’ perceptual ratings. Results During the elicitation tasks, participants with LHD significantly differed significantly from HC in durational measures used to distinguish between the literal and idiomatic productions. Significant differences between participants with RHD and HC were seen in measures of fundamental frequency. Differences associated with task were seen. For the repetition tasks, acoustic and perceptual analyses showed that the LHD and RHD groups produced utterances comparable to HC performance. Overall, native Korean listeners were successful in discriminating the intended idiomatic and literal meanings of ditropic sentences produced by HC for all tasks. However, they showed decreased performance in discriminating utterances produced by brain-damaged individuals during the elicitation tasks, especially those produced by RHD. Statistical analyses on the relationship between acoustic features and goodness ratings suggested final F0 change and F0 variations were found to be significant predictors for goodness ratings of both elicited and repeated utterances. Three speech language pathologists with training in phonetics participated as raters for vocal qualities. Nasality was significantly salient vocal quality of idiomatic utterances. Conclusion The findings support that (1) LHD negatively affected the production of durational cues and RHD negatively affected the production of fundamental frequency cues in idiomatic-literal contrasts; (2) healthy listeners successfully identified idiomatic and literal versions of ambiguous sentences produced by healthy speakers but not by RHD speakers; (3) Productions in brain-damaged participants approximated HC’s measures in the repetition tasks, but not in the elicitation tasks; (4) Nasal voice quality was judged to be associated with idiomatic utterances in all groups of participants. Findings agree with previous studies indicating HC’s abilities to discriminate literal versus idiomatic meanings in ditropically ambiguous idioms, as well as deficient processing of pitch production and impaired pragmatic ability in RHD.
Keywords
