Perception of the Emotional Content of Speech

Abstract

White and Indian adult males, speaking whatever words came to mind in their native languages (English and Cree, respectively), attempted to express vocally the emotions of happiness, sadness, love, and anger. After these simulations of emotional state were recorded, the speech samples were rendered unintelligible by means of an electronic filter (which removed verbal meaning while leaving intact the tonal aspects of speech). The filtered vocal expressions from the white and Indian speakers were then played to other groups of white and Indian adult males (receivers). When both heard the samples from the white speakers, the white receivers were able to identify the emotional content of tone of voice more accurately than the Indian receivers. On the other hand, Indian receivers performed more accurately than whites when judging the samples from Indian speakers. The significant interaction suggested that language and culture are crucial factors in the transmission of emotion-even on the nonverbal vocal level.

Characteristics and Recognizability of Vocal Expressions of Emotion

“…Both techniques yielded results which may be interpreted as indications that for all three subject groups, but most clearly for the Taiwanese and Japanese subjects, confusions among emotions were primarily a function of level of activity. This outcome is consistent with the findings of Davitz (1964c), Davitz and Davitz (1959), Dimitrovsky (1964), and Fairbanks and Prono vost (1939) with respect to American English subjects and with the findings of Albas et al (1976) with respect to Anglo-Canadian white and Cree-Indian subjects: The confusions found in these studies can also be more easily explained in terms of emotions being similar in level of activity than in terms of their being similar in, for example, evaluative meaning or strength. This finding is also in accordance with the results of our own recognition experiments with children, reported on in Chapters 7 and 8.…”

“…Thus, the active emotions of anger and hap piness tended to be confused, as did the passive emotions of sadness and love. A similar finding has been reported with respect to the performance of adults (Albas, McCluskey, & Albas, 1976). It is thus suggested that both children and adults might employ the same dimensions in their categorization of vocal expressions of emotion.…”

“…If the expressions of joy and anger do indeed hardly differ, subjects will have a hard time keeping them apart, which will result in a great number of mutual confusions. And indeed, confusions between joy and anger have occasionally been reported, cross-culturally (Albas, McCluskey, & Albas, 1976) as well as intra-culturally (Dimitrovsky, 1964). However, the confusions are not frequent enough to warrant the conclusion of a complete absence of any differences.…”

“…Thus, Albas et al (1976) found that both AngloCanadian and Cree-Indian subjects often confused the active emotions of happiness and anger on the one hand, and the passive emotions of sadness and love on the other, regardless of the culture of the speakers. This suggests, as has been reported before for American English (Dimitrovsky, 1964;Davitz, 1964c), that in general errors in the identification of vocal expressions of emotion seem to be a function of similarity in level of activity of the emotions concerned rather than, for example, similarity in evaluative meaning (see also Chapters 7 and 8).…”

“…Albas, McCluskey, and Albas (1976) had forty Anglo-Canadians and forty Cree-Indians identify emotional expressions of happiness, sadness, love, and anger, produced by six Anglo-Canadian and six Cree-Indian adult males. The speakers were allowed to speak any two sentences that came to mind (in their native language) in an attempt to express each emotion.…”