Perceptual identification of voices under normal, stress, and disguised speaking conditions
Harry Hollien, Wojciech Majewski, Patricia A. Hollien · The Journal of the Acoustical Society of America · 1974
A number of authors have contended that, at present, the human auditor is the most accurate “system” for correctly identifying speakers from tape-recorded materials; other authors have indicated that individuals do not have to know the speakers in order to make highly accurate judgments of this type. This research attempts to test these two assumptions. Talkers were 10 adult males who read a modernization of R. L. Stevenson's “Apology for Idlers” under three conditions: (a) normal speech, (b) stress (talkers were subjected to randomly distributed electric shocks while speaking), and (c) disguised speech. Three types of listeners were utilized: (a) listeners who knew the talkers, (b) listeners who did not know the talkers but who were native speakers of English, and (c) listeners who knew neither the talkers or the language (i. e., native speakers of Polish who did not know English). The listeners who knew the speakers scored very high for all speaking conditions, nearly 100% for both normal and stress and 79% for disguise. The scores for the English-speaking listeners, who did not know the speakers, were about one-half as high (even after training) as those for the first group for both normal and stress and were poorer by a factor of three for the disguised condition. Scores for the listeners who knew neither the talkers nor the language were lower than for the other groups but patternings among the data were the same. Implications of these results are discussed.