Roles of subjective and objective tone in tone-phoneme identification

Researchers have described tone phonemes by their changing fundamental frequency (Fo) contours, suggesting that they are cued by low-pitch objective tones. Possible roles of higher-pitch objective tones (heard as higher pitch harmonics), or low-pitch subjective tones (evoked by higher-frequency harmonics and heard as the missing Fo) in tone-phoneme identification have not been investigated. The purpose of the present investigation was to study relative roles of subjective tone and objective tone in tone-phoneme identification by Mandarin-Chinese speakers.;During phase I, 20 subjects identified Mandarin-Chinese words at progressively lower intensities under five filter conditions: (1) Fo only, (2) resolved harmonics above the Fo, (3) unresolved harmonics, (4) all harmonics without Fo, and (5) all harmonics. During phase II, subjects identified the missing fundamental of a complex tone at progressively lower intensities under the same filter conditions. The spectrum of the complex tone was similar to the long-term spectrum of the tone phonemes, and the Fo of the complex tone was equal to the tone phonemes' average Fo.;Tone-phoneme MILs were measured at low sensation levels (SLs) under all five filter conditions. This finding suggests that many harmonic combinations are potential tone-phoneme cues when other harmonics are missing. Several other results, however, suggest that roles of subjective and objective tone in tone-phoneme identification depend on which harmonics are heard. Subjects, for example, identified tone-phonemes only at high SPLs when they heard the Fo only or unresolved harmonics, whereas they identified tone phonemes at high and low SPLs when they heared resolved harmonics. Resolved harmonics, therefore, may afford Chinese listeners more practical tone-phoneme cues than the Fo only or unresolved harmonics when listening to everyday speech. Furthermore, when subjects listened to unresolved harmonics they identified tone phonemes at intensities below their subjective-tone MIL. Thus, objective tone alone probably cued tone phonemes at low SLs when subjects listened to unresolved harmonics. Conversely, a significant number of subjects identified tone phonemes near their subjective-tone MIL when they listened to resolved harmonics. Apparently, objective tone, or subjective tone, or both, cued tone phonemes at low SLs when subjects listened to resolved harmonics.;These findings are explained in part by a pattern-recognition theory of pitch perception. These findings, moreover, have implications for understanding listening problems of tone-language speakers with different configurations of hearing loss.