Z. S. Bond

A note on the acoustic-phonetic characteristics of inadvertently clear speech

Abstract Even when speaking conditions and listener responses are very tightly controlled, some talkers are easier to understand than others. In a series of intelligibility tests using both English words without context and English sentences, both native and non-native listeners found one of five talkers difficult to understand. Since all talkers read similar materials and the task was the same for all listeners, the differences in intelligibility must have resulted from particular phonetic characteristics used by the talker. Spectrograms were made of all test words produced by the talkers and compared on selected acoustic-phonetic properties. In comparison with the more intelligible talkers, the least intelligible talkers produced test words at shorter durations; abbreviated vowel durations; used the least differentiated vowel space, as defined by the first two formants; used minimal cues for consonantal contrasts; and had the most varied amplitude of stressed vowels. These characteristics are similar to those distinguishing deliberately clear speech. That non-native and native listeners found the same talker difficult to understand suggests that the effect of clear speech, though different in degree, reflects the use of the same acoustic-phonetic information by both groups of listeners. Further, clear speech has similar acoustic-phonetic characteristics whether deliberately or inadvertently produced.

Acoustic-phonetic characteristics of speech produced in noise and while wearing an oxygen mask

The present study investigated changes in the prosodic and acoustic-phonetic features of isolated words by four male talkers speaking in quite and in pink noise at a level of 95 dB SPL. Speech samples were collected both with and without an oxygen mask. Changes in duration, fundamental frequency, total energy, and formant center frequency were analyzed. In addition to the expected changes of increased pitch and amplitude associated with speaking in noise without an oxygen mask, significant effects were found (particularly in the formant center frequencies) as a result of using the oxygen mask. When the oxygen mask was employed, no further significant changes were caused by adding noise to the speaking situation.

Identification of acoustically modified Mandarin tones by native listeners

This study investigated identification of fragmented Mandarin tones by non-native listeners. Monosyllabic Mandarin words were digitally processed to generate intact, silent-center, center-only, and onset-only syllables. The syllables were recorded with two carrier phrases such that the offset of the carrier tone and the onset of the target tone were either continuous or discontinuous in fundamental frequency (F0). The syllables were presented with an original carrier phrase, excised from the carrier phrase, or excised and cross-spliced with another carrier phrase. Response accuracy and reaction time were measured, and tone confusion patterns were analyzed. Overall, tone identification varied as a function of modification and tone. Intact and center-only syllables were identified more accurately than silent-center and onset-only syllables. Tone 2 was consistently the most challenging tone to identify. Although the performance level of the third-year students approached that of native listeners reported in Lee, Tao, and Bond (2008), the non-native listeners did not show evidence of using coarticulatory information. Nonetheless, the continuity or discontinuity in F0 between the carrier and target tones did affect tone identification, suggesting the influence of context in non-native tone identification.

Identification of Acoustically Modified Mandarin Tones by Non-native Listeners

This study investigated identification of fragmented Mandarin tones by non-native listeners. Monosyllabic Mandarin words were digitally processed to generate intact, silent-center, center-only, and onset-only syllables. The syllables were recorded with two carrier phrases such that the offset of the carrier tone and the onset of the target tone were either continuous or discontinuous in fundamental frequency (F0). The syllables were presented with an original carrier phrase, excised from the carrier phrase, or excised and cross-spliced with another carrier phrase. Response accuracy and reaction time were measured, and tone confusion patterns were analyzed. Overall, tone identification varied as a function of modification and tone. Intact and center-only syllables were identified more accurately than silent-center and onset-only syllables. Tone 2 was consistently the most challenging tone to identify. Although the performance level of the third-year students approached that of native listeners reported in Lee, Tao, and Bond (2008), the non-native listeners did not show evidence of using coarticulatory information. Nonetheless, the continuity or discontinuity in F0 between the carrier and target tones did affect tone identification, suggesting the influence of context in non-native tone identification.

Speaker variability and context in the identification of fragmented Mandarin tones by native and non-native listeners

Abstract This study investigated the individual and joint contribution of speaker variability, context, and type of acoustic input to the identification of Mandarin tones. Mandarin syllables produced by single vs. multiple speakers were digitally processed to generate intact, silent-center, center-only, and onset-only syllables to be presented in isolation or with a precursor carrier phrase. Forty native listeners and 55 non-native listeners were put under time pressure to identify the tones of the syllables. The results showed higher identification accuracy for single-speaker tones and tones presented in context. Tone identification accuracy also decreased as acoustic input was reduced. The speaker variability effect showed comparable magnitude for the native and non-native listeners. In contrast, non-native tone identification, compared to the native performance, was facilitated less by context and compromised more when acoustic input was minimal. Tone confusion analyses showed a Tone 2–Tone 3 confusion and a bias towards Tone 4 responses for both groups of listeners, but the patterns of confusion and bias are more variable in the non-native responses. The non-native listeners also showed inconsistent evidence of using F0 height for tone identification when acoustic input was minimal.

Synthetic Speech Intelligibility: Comparison of Native and Non-native Speakers of English

This study sought to determine the intelligibility of DECTalkTMsynthesized speech to both native and non-native speakers of English. The ability of the groups to understand DECTalk sentences delivered both in quiet and at a signal-to-noise (S/N) ratio of +10 dB was investigated. Results suggested that non-native speakers made significantly more errors transcribing DECTalk sentences than did native speakers. DECTalk was found to be significantly less intelligible in noise than in quiet for both groups, but non-native speakers experienced significantly more difficulty understanding DECTalk in noise than did native speakers. Implications of these findings and the need for further research are discussed.

The Elusive / Illusive Syllable

A common phonological reduction in casual speech is vowel loss in unstressed syllables as in s ‘ pose . How the phonetic detail of this reduction is implemented in

Interaction between Auditory and Oral Sensory Feedback in Speech Regulation

To investigate the interaction between the auditory and oral sensory feedback modalities during speech production lingual vibrotactile thresholds were obtained from subjects in the following conditions: (1) before and after speech production with normal auditory feedback, (2) before and after speech production under exposure to auditory masking, and (3) before and after exposure to auditory masking without performing speech tasks. In addition duration measurements were obtained for selected speech sounds to investigate temporal changes in the articulatory patterns of subjects in the various conditions. Lingual sensory decreases and temporal reorganization were observed only in subjects speaking under auditory masking. These data suggest a balanced interaction between auditory and oral sensory feedback modalities which, when disturbed, results in non-phonemic change in speech production.

Identification of multi-speaker Mandarin tones in noise by native and non-native listeners

The similarities and contrasts between native and non-native identification of multi-speaker Mandarin tones in quiet and in noise were explored in a perception experiment. Mandarin tone materials produced by three male and three female speakers were presented with five levels of signal-to-noise ratios (quiet, 0, -5, -10, and -15dB) in two presentation formats (blocked by speaker and mixed across speakers) to listeners with various Mandarin experience (native, first-year, second-year, third-year, and fourth-year students). Stimuli blocked by speaker yielded higher accuracy and shorter reaction time. The additional demand of processing mixed-speaker stimuli, however, did not compromise non-native performance more than native performance. Noise expectedly compromised identification performance, although it did not compromise non-native identification more than native identification. Native listeners expectedly outperformed non-native listeners, although identification performance did not vary systematically as a function of duration of Mandarin experience. It is speculated that sources of variability in speech would affect non-native more than native tone identification only if syllable-internal, canonical F0 information is removed or altered.

Acquisition of the English Voicing Contrast By Arab Children

Children are typically more proficient than adults in learning the phonetic detail of a second language. The purpose of this study was to investigate the acquisition of the English voicing contrast as cued by voice onset time in syllable-initial position and preconsonantal vowel duration in final position in the speech of children whose native language is Arabic. The children, 24 to 135 months in age, were recorded producing minimal pairs differing in the voicing of consonants. Voice onset time and vowel duration were measured from spectrograms. Children were highly variable in their mastery of the voicing contrast. For this sample, neither age nor experience with English could predict phonetic proficiency.