The monaural localization of tonal stimuli

0 "This work was supported by USPHS Clinical Research Center Grant NS 03358-09 and a Research Career Development Award from NINDS 1 ROBERT A. BUTLER Department of Surgery (Otolaryngology), University of Chichgo , Chicago, l11inois 60637 With one ear occluded, 17 listeners were asked to locate tone bursts, .25, A, .6, .9. lA, 2.0, 3.2, 4.8, and 7.2 kHz, generated by a loudspeaker concealed from view. The S's response was to callout that number, from a series of numbers arranged horizontally, behind which he thought the tone bursts originated. The listeners perceived the sounds as emanating from the side of the unoccluded ear, but their judgments bore no consistent relation to the actual location of the sound source. Rather, the listeners showed a strong tendency to locate a tone burst, within the range of .9 through 7.2 kHz, in a fixed spatial relation to the next higher- and lower-pitched tone burst. Distorting the pinna of the unoccluded ear failed to modify the perceptual pattern. It was suggested that the perceived spatial relations among the various frequencies was a by-product of the tonotopic organization of the auditory nervous system. - "The more nearly the sounds approach pure tones, the more inaccurate the localization. This is true regardless of the absolute pitch. Genuinely pure tones are essen tially unlocalizable in monaural hearing [Angell & Fite, 190I, p. 2451" So concluded the authors after testing extensively with a variety of sounds the localizing ability of a unilaterally deaf person. Their statement should have been qualified, since the listener did assign 72 out of 90 sounds presented by tuning forks to a specific position in space. It was only that his choices were usually wrong. Whether he systematically positioned those tones with respect to their pitch cannot be deduced from the description of the data. This does occur, however, when the sounds originate in the vertical plane, directly ahead of the listener. High-pitched tones are perceived to originate above lower-pitched tones (Pratt, 1930). In both the Angell and Fite and the Pratt studies, the cues for binaural time and intensity differences were absent; hence, listeners were deprived of the information necessary for distinguishing between the locations of the sound sources. The fact that Pratt's listeners located higher-pitched sounds in the vertical plane above lower-pitched sounds could have been the result of extrinsic associations. We use the words "high" and "low" when describing the pitch of a sound. In the absence of binaural cues for spatial position, these word associations may cause us to perceive sinusoids with brief periods as high, and those with longer periods as low in space. On the other hand, the correct interpretation of this phenomenon may be just the opposite; i.e., we use the words "high" and "low" to refer to the pitch of the S and a Mine Safty Appliance Muff (Noisefoe Mark 11M) was placed over the same ear. The ear chosen for occlusion was on the side of the nonpreferred hand. After occlusion, thresholds for the tone bursts of .25, A, .6, .9, lA, 2.0, 3.2, 4.8, and 7.2 kHz were obtained. These stimuli, whose rise-fall time was 5 msec and whose plateau was 10 rnsec, were repeated approximately six times per second. Following threshold measurements, Ss were instructed to callout the number of the loudspeaker from which the tone bursts appeared to originate. They were requested to keep their heads firmly in the headrest that was attached to the chair. Each S was oriented so that the array of loudspeakers extended from directly ahead to 80 deg toward the side U i his tones because tonal stimuli of differing unoccluded ear. Only loudspeakers at 10, frequencies do indeed appear spatially 30, 50, and 70 deg from the S's medial: differentiated in this way in the vertical sagittal plane were used, although the plane. Some data support this suggestion. listeners were not aware of this restriction Roffler and Butler (1968), studying Each of the nine frequencies was generated vertical plane localization, reported that three times by each of the four children 4 and 5 years old perceived loudspeakers. Hence, every frequency was high-pitched tones as emanating above presented 12 times and the test session low-pitched tones. During posttest consisted of 108 tonal presentations, or conversations the children gave no trials. Orders of presentation for indication of associating the words "high" loudspeakers and frequencies were and "low" with the stimulus frequency. quasirandom. Tone bursts were delivered at The present paper explores this problem 20 phons and continued until the listener further. An experiment was designed to made a location judgment. Testing was find out whether a systematic spatial conducted in a sound-treated room with relation among various tonal frequencies the sound-generating equipment exists when listeners, deprived of binaural (Krohn-Hite oscillator, Grason-Stadler cues, are asked to locate these stimuli in electronic switch and interval timer) the horizontal plane. Positive results, located in an adjoining room. An intercom should they occur, could not be explained system served to maintain voice contact in terms of word association. The words between Sand E. "high" and "low" obviously are not applicable to this type of listening situation; nor are any other words associated with tonal frequencies likely to influence the perception of their relative position in the horizontal plane. Results Upon inspection of the data, there was no indication that Ss were locating sound sources with an accuracy exceeding chance. However, a definite pattern of positioning those frequencies of .9 kHz and above was PILOT STUDY observed. Specifically, the tone bursts 3.2 Method and 4.8 kHz were perceived as nearer the Five listeners participated. Their Ss' median sagittal plane than were .9 and audiometric thresholds for frequencies 2.0 kHz. The latter frequencies, in turn, ranging from .25 through 8.0 kHz were were perceived to originate nearer the within IS dB of 0 re ISO standards. The median sagittal plane than were 1.4 and listeners were seated comfortably at the 7.2 kHz. Since, under conditions of center of a semicircular reinforced screen binaural listening, some frequencies whose radius was 5 ft. Nine KLH consistently appear further displaced loudspeakers, 4-in. diam, were attached to toward the median sagittal plane than the screen at eye level. Each was housed in others (Butler, Roffler, & Naunton, 1967), a wooden cabinet, 6.25 x 6.25 x 4 in., and the possibility arose that subliminal stimuli was identified by a number, 1 through 9, impinging on the occluded ear may have according to its horizontal position on the influenced the results. Certainly stimuli screen. With reference to the vertex of the whose intensities are below threshold value listener's head, the speakers were placed obtained by standard audiometric 10 deg apart, center to center. techni (...truncated)