Music & Language Studies

BUCHANAN, Tony W. Kai Lutz, Shahram Mirzazade, Karsten Specth, N. Jon Shah, Karl Zilles, and Lutz Jäncke. “Recognition of Emotional Prosody and Verbal Components of Spoken Language: An fMRI study.” Cognitive Brain Research 9 (2000): 227-38.

The current study utilized a Functional Magnetic Resonance Imaging (fMRI) technique to identify the neural areas involved in the recognition of emotional prosody and phonemic characteristics of spoken language by normal listeners. Ten right-handed male subjects, lacking any known neurological disorders, were presented with the recorded words: BOWER, POWER, DOWER, TOWER, spoken by a native-English speaker in one of four emotional tones of voice: happy, sad, angry, neutral. The stimuli remained the same (though in random order) for all conditions. What differed were the task requirements. In each case, the subjects were to depress a button upon detection of the target stimulus. There were four conditions: (1) identify POWER in any of its emotional forms; (2) identify BOWER in any of its emotional forms; (3) identify any of the words in a happy tone; and (4) identify any of the words in a sad tone. This is similar in intent to the study by Bartholomeus (1974), but more elegant in experimental design, and avoiding the limitations of the dichotic listening paradigm. In both the present experiment and Bartholomeus’ the same stimuli were used with different tasks in order to test whether requirements on attention impact neural processing, in particular whether they alter laterality effects. The findings in the current case were affirmative, and defined the regions of the brain far more precisely than is possible with dichotic listening. The results demonstrate two major findings: first, bilateral activation during both phonetic and emotional detection tasks, in comparison with baseline (resting state) activity; and second,

significant lateralization of cortical activity during the perception of both emotional prosody and the perception of verbal characteristics of words (236).

In particular, it was found that detection of emotion tasks produced significantly greater activity in right frontal regions than corresponding verbal tasks; and verbal tasks involved significantly greater left frontal activity than emotion detection tasks with the same input. The authors point out that localization of activity in the frontal lobes is counter to the findings of some clinical studies, which would predict lateralized activity in temporoparietal regions. However, they explain:

This discrepancy most likely arises from the inherent differences between experiments involving patients with specific lesions and functional imaging experiments. While studies involving patients with lesions are able to demonstrate those areas critical for a specific function, functional neuroimaging studies are only able to demonstrate those areas that are involved in a specific function (234).