Field studies indicate that Japanese macaque (Macaca fuscata) communication signals vary with the social situation in which they occur [S. Green, "Variation of vocal pattern with social situation in the Japanese monkey (Macaca fuscata): A field study," in Primate Behavior, edited by L. A. Rosenblum (Academic, New York, 1975), Vol. 4]. A significant acoustic property of the contact calls produced by these primates is the temporal position of a frequency peak within the vocalization, that is, an inflection from rising to falling frequency [May et al., "Significant features of Japanese macaque communication sounds: A psychophysical study," Anim. Behav. 36, 1432-1444 (1988)]. The experiments reported here are based on the hypothesis that Japanese macaques derive meaning from this temporally graded feature by parceling the acoustic variation inherent in natural contact calls into two functional categories, and thus exhibit behavior that is analogous to the categorical perception of speech sounds by humans. To test this hypothesis, Japanese macaques were trained to classify natural contact calls by performing operant responses that signified either an early or late frequency peak position. Then, the subjects were tested in a series of experiments that required them to generalize this behavior to synthetic calls representing a continuum of peak positions. Demonstration of the classical perceptual effects noted for human listeners suggests that categorical perception reflects a principle of auditory information processing that influences the perception of sounds in the communication systems not only of humans, but of animals as well.

No abstract

No abstract

No abstract

Directional hearing acuity, as measured by the minimum audible angle (MAA), was determined in four Old World monkeys, Macaca radiata. The acoustic stimuli were linear changes in frequency (sweeps) for different frequency ranges and sweep rates. The sweeps ranged between 0.5 and 1.3 kHz, at two durations, 100 and 200 ms. In upsweeps which began at 0.5 kHz and were 200 ms in duration, MAA decreased as sweep rate and frequency range increased. These thresholds were compared to MAAs of sweeps which traversed the same range of frequencies but at a different rate, to MAAs of sweeps with identical rates but over different frequency ranges, and to the MAAs of downsweeps. These comparisons indicated that range, and not sweep rate, exerts the greatest effect on the MAA. Interaural phase differences derived from the upper limits of the frequency range are discussed as potential FM localization cues.

The directional dependence of the transfer function from free-field plane waves to a point near the tympanic membrane was measured in anesthetized domestic cats. A probe tube microphone was placed from a ventral approach, which minimally affects the head and pinae. Acoustic measurements were taken using a quasianechoic method in which an impulsive stimulus was presented and room reflections were rejected using a Kaiser window. The transfer functions are similar to those reported earlier [A.D. Musicant et al., 12th ICA, Toronto (July 1986)]. They can be broken into three distinct frequency ranges: A low-frequency range (< 5 kHz), where interaural level differences vary smoothly with azimuth; a midfrequency range (5–18 kHz), where a prominent spectral notch is observed; and a high-frequency range (> 18 kHz), where the spectrum varies greatly with source location. The frequency of the midfrequency notch is a robust cue for sound source direction; it varies with both azimuth and elevation in a monotonic fashion in the frontal field. Source direction is uniquely determined by combining the midfrequency notch cue from both ears. These results suggest that simple consideration of interaural level differences may not be sufficient for the study of sound localization at high frequencies. [Work supported by Grant No. NS12524 from NINCDS/NIDCD.]

Many species, including humans and cats, have head-related transfer functions (HRTFs) that introduce sharp features to the amplitude spectrum of broadband sounds as auditory stimuli propagate to the eardrum. Our behavioral results indicate that HRTF-based spectral cues, in particular mid-frequency spectral notches, are crucial for accurate spatial hearing in cats. This talk will summarize a series of physiological and behavioral assessments that follow the neural encoding of biologically relevant spectral information from a distributed representation in the auditory nerve to notch-selective neurons in the inferior colliculus. Just as anatomically defined pathways have been described for binaural directional cues, our results support the existence of a functionally segregated auditory pathway that is specialized for the processing of spectral cues for sound localization. [Work supported by NIDCD Grant No. DC00954.]

No abstract