Julie Mapes Lindholm

Phonetic identification by elderly normal and hearing-impaired listeners

Young normal-hearing listeners, elderly normal-hearing listeners, and elderly hearing-impaired listeners were tested on a variety of phonetic identification tasks. Where identity was cued by stimulus duration, the elderly hearing-impaired listeners evidenced normal identification functions. On a task in which there were multiple cues to vowel identity, performance was also normal. On a/b d g/identification task in which the starting frequency of the second formant was varied, performance was abnormal for both the elderly hearing-impaired listeners and the elderly normal-hearing listeners. We conclude that errors in phonetic identification among elderly hearing-impaired listeners with mild to moderate, sloping hearing impairment do not stem from abnormalities in processing stimulus duration. The results with the /b d g/continuum suggest that one factor underlying errors may be an inability to base identification on dynamic spectral information when relatively static information, which is normally characteristic of a phonetic segment, is unavailable.

Stimulus factors influencing the identification of voiced stop consonants by normal-hearing and hearing-impaired adults

The effects of mild-to-moderate hearing impairment on the perceptual importance of three acoustic correlates of stop consonant place of articulation were examined. Normal-hearing and hearing-impaired adults identified a stimulus set comprising all possible combinations of the levels of three factors: formant transition type (three levels), spectral tilt type (three levels), and abruptness of frequency change (two levels). The levels of these factors correspond to those appropriate for /b/, /d/, and /g/ in the /ae/ environment. Normal-hearing subjects responded primarily in accord with the place of articulation specified by the formant transitions. Hearing-impaired subjects showed less-than-normal reliance on formant transitions and greater-than-normal reliance on spectral tilt and abruptness of frequency change. These results suggest that hearing impairment affects the perceptual importance of cues to stop consonant identity, increasing the importance of information provided by both temporal characteristics and gross spectral shape and decreasing the importance of information provided by the formant transitions.

An interpretation of age-related differences in letter-matching performance

Comparisons were made of the response latencies of old (mean age = 69.2 years) and young (mean age = 26.8 years) subjects on simple and choice reaction time (RT) tasks and “physical identity” (PI) and “name identity” (NI) trials of a letter-matching task. Young subjects were faster than old subjects on all tasks, and the absolute difference between groups increased with processing complexity (simple RT < choice RT< PI <NI). However, in support of the hypothesis that aging is associated with a general reduction in processing speed, the relative difference between groups did not vary with task, except for a subset of the NI trials. Response latencies for the NI trials varied with stimulus letter for both age groups, but the magnitude of the letter effect was greater for the elderly. Their latencies were disproportionately long for the more difficult letters. A second experiment showed that NI latency reflected the visual similarity of the letters with respect to the other letters in the stimulus set. It is suggested, therefore, that the NI letter effect resulted from differences in letter identification time. The disproportionately long latencies of the elderly for the visually similar letters are discussed in terms of the hypothesis that aging is associated with an increase in internal noise.

Vowel intelligibility in the absence of the acoustic reflex: performance-intensity characteristics

Performance-intensity functions were defined for ten synthetic vowels whose durations (50 ms) were shorter than the time necessary for the acoustic reflex to alter the transmission characteristics of the middle ear. Recognition accuracy for ten listeners with normal auditory function was asymptotic from 72 to 90 dB and then fell linearly to 108 dB. Thus, at SPLs greater than approximately 90 dB, the auditory encoding of vowels is altered by the absence of the acoustic reflex.

Nonlinear auditory coding of vowel formants at high sound pressure levels

To study vowel intelligibility in the absence of the acoustic reflex, ten synthetic vowels whose duration was shorter than the effective latency of the acoustic reflex were presented to normal hearing listeners at levels ranging from 72 to 106 dB SPL. Signal intelligibility varied as a function of both SPL and vowel identity. Vowels with widely spaced formants, i.e., /i/ and /ɪ/, were unaffected by presentation level. Vowels with more proximal formants began to show a decrement in intelligibility at 96 dB and suffered as much as a 40% decrement at 106 dB. Confusion errors were not symmetrical in F1‐F2 space suggesting a systematic distortion in the auditory coding of formant location. These data indicate another source of distortion for many severely hearing‐impaired listeners who commonly listen to speech amplified to greater than 100 dB and who do not have measureable acoustic reflexes. [Research supported by NIH.]

Factors underlying errors in the recognition of stop consonants by listeners with cochlear hearing impairment

In a series of experiments we have assessed the contribution of selected stimulus factors to the recognition of stop consonants by listeners with mild to severe hearing impairments. Errors in recognition do not appear to be conditioned in a significant manner by a high amplitude first formant or by brief duration formant transitions. Errors are conditioned by the spectral tilt at signal onset and by the duration of the steady‐state spectrum at signal onset. In severely hearing‐impaired listeners, discrimination among the stops appears dependent on the presence of distinctive temporal properties among the signals. In general, our results point to the critical role played by accurate temporal resolution in maintaining accurate speech intelligibility.

The role of the acoustic reflex in modifying identification accuracy for vowels

“Rollover,” poorer word identification accuracy at high than at low SPLs, is correlated with the absence of the acoustic reflex in individuals with lesions to c.n. VII and c.n. VIII. The increased number of word identification errors at high SPL is due largely to errors in vowel identification. Individuals with these types of lesions, however, are rare. In order to study rollover in a normal population, we constructed 10 vowels of 50‐ms duration. These were presented to normal hearing listeners at moderate and high SPLs. We reasoned that, since the latency to onset of the acoustic reflex, at the test levels, was longer than the total duration of the vowels, we should observe poorer vowel identification at the higher SPL than at the lower. This was, in fact, the outcome. Vowel confusions were usually with vowels with higher F2 and lower F1. This outcome was confirmed by the results of identification tests with stimuli along an /i/‐/e/ continuum.

Spectral tilt and burst duration as determinants of stop consonant identification accuracy for hearing‐impaired listeners

In a series of experiments we have manipulated the tilt of the onset spectrum and duration of the release burst in syllables with initial voiced stop consonants. In identification tests we have found that normal hearing listeners are largely unaffected by abnormal tilt or burst duration. Hearing‐impaired listeners, on the other hand, are significantly affected. We speculate that when moderately and severely hearing‐impaired listeners achieve normal recognition accuracy for stop consonants, they do so by giving greater‐than‐normal perceptual weight to the information provided by burst duration and spectral tilt. The results of identification tests in listeners with congenital severe to profound sensorineural hearing impairments are consistent with our speculation.

Procedures which alter the intelligibility of voiced stop consonants for hearing impaired listeners

To assess the hypothesis that spread of masking from F1 is responsible for poor identification of voiced stop consonants by hearing impaired listeners, /bdg/ were synthesized with bursts and formant transitions in configurations with and without the first formant. Synthesis in the /eI/ environment allowed minimal acoustic differences to cue the contrasts. To maximize the probability of finding evidence of masking spread the stimuli were presented at levels both above and below the point of maximum intelligibility. At no point on the function was performance without the first formant better than performance with the first formant. To investigate other ways in which the intelligibility of /bd/ could be enhanced, we altered, in separate manipulations, the amplitude and duration at signal onset of the cues for place of articulation. Increasing the amplitude of the onset cues improved identification accuracy for /d/ but not for /b/ or /g/. In contrast, increasing the duration of the onset cues improved identif...