A contralateral "cue" tone presented in continuous broadband noise both lowers the threshold of a signal tone by guiding attention to it and raises its threshold by interference. Here, signal tones were fixed in duration (40 ms, 52 ms with ramps), frequency (1500 Hz), timing, and level, so attention did not need guidance. Interference by contralateral cues was studied in relation to cue-signal proximity, cue-signal temporal overlap, and cue-signal order (cue after: backward interference, BI; or cue first: forward interference, FI). Cues, also ramped, were 12 dB above the signal level. Long cues (300 or 600 ms) raised thresholds by 5.3 dB when the signal and cue overlapped and by 5.1 dB in FI and 3.2 dB in BI when cues and signals were separated by 40 ms. Short cues (40 ms) raised thresholds by 4.5 dB in FI and 4.0 dB in BI for separations of 7 to 40 ms, but by ∼13 dB when simultaneous and in phase. FI and BI are comparable in magnitude and hardly increase when the signal is close in time to abrupt cue transients. These results do not support the notion that masking of the signal is due to the contralateral cue onset/offset transient response. Instead, sluggish attention or temporal integration may explain contralateral proximal interference.

S An abstract of no more than 200 words must accompany each paper submitted. Careful preparation is important to make possible its use verbatim in Current Physics Index. It should mention the subjects studied and new methods used and it should set forth quantitatively new observations and conclusions. Brief numerical results and their accuracy may be included. (1) Abstracts must be submitted doubleor triple-spaced as one paragraph only. u 1 (2) (Special care should be given to make equations and formulas clear to the typesetter, and all handwritten letters and symbols that cannot be typed should be identified in the margin and only the first time that such occur. (3) Clearly distinguish capital and lower-case letters where there could be confusion. (4) Use fractional exponents to avoid root signs. (5) Extra symbols should be introduced to. avoid complicated exponents or where it is necessary to repeat a complicated expression a number of times. (6) Use the solidus (/) wherever possible for fractions. (7) Do not repeat mathematical derivations that are easily found elsewhere in the literature; merely cite the references. (8) Mathematical expressions should be consecutively numbered throughout the body of the paper at the right-hand margin in parentheses. Numbering starts anew with each appendix: Appendix A: (AI), (A2), (A3), etc. Appendix B: (Bl), (B2) (9) Equation numbers mentioned in the running text should be enclosed in parentheses, i.e., Eq. (1), Eqs. (la), ( 2 a ) , . . . . (10) If a "List of Symbols" is included, it should be placed between the abstract and the Introduction. Symbols should not be redefined in the text. (11) Metric units should be used. The preferred form of metric unit is the Systeme International (SI). ILLUSTRATIONS (1) Line drawings in India ink on good-quality white drawing paper (not translucent tracing paper) are preferred for the original set, but sharp glossy prints are acceptable. All photographic figures should be submitted as glossy prints. Xerox and other similar reproductions are not acceptable for the original set. The triplicate set of figures may be in any form that is clearly legible. (2) Every illustration must be referred to in the main text in consecutive numerical order. Numbering starts anew in each Appendix and is further designated by affixing the Appendix letter to the figure number. Examples: For Appendix A: Fig. A-l, Fig. A-2, etc. For Appendix B: Fig. B-l, Fig. B-2, etc. (3) A caption (legend) must accompany each illustration. Captions must be typed doubleor triple-spaced on a sheet of paper at the end of the manuscript. (4) It is recommended that all nongraphic material that can be printed in the figure legends without hindering the transmission of information should be placed there rather than on the figures themselves. Contributors who do not wish to follow this course should give careful reasons for their choice and will be held responsible for any expense incurred in special art work needed to produce clearer figures. (5) Drawings should be planned for reduction to, preferably, a single-column width (4% in. on a full-size page.) (This is later reduced to Vh in. when the Journal is printed.) Since figures are reduced according to lettering size, it is advised that lettering be as large as possible consistent with the legibility of the figure. • Principal letters on figures are reduced to approximately 3.1 mm, 1/8 in., or 9*/2 points on a full-size page. (This is later reduced to approximately 2.3 mm, 1/12 in., or 7 points when the Journal is printed.) This reduction is done regardless of the overall size of the drawing. iv J . Acoust. Soc. Am. 68(1), July 1980 • Overlarge lettering with oversmall data points cannot reproduce well. Data points must be the same size as an " O " in the lettering size used, and A's and D'S, etc., must be able to be circumscribed by the "O." For best results, draftsmen should be given a recent copy of the Journal to familiarize themselves with the proportions to be followed. • Handlettering is not acceptable. • Standard AIP abbreviations should be used in lettering: e.g., kHz, not KHz....dB, not db or Db. (6) Coordinate paper is not satisfactory, but if used must be blue-lined with all coordinates to be reproduced ruled in with India ink. (7) Illustrations must be held, with the rarest exceptions, to 8y2Xllin . (8) Photographs of standard apparatus in a laboratory setup are not instructive and should not be included. (9) Electron micrographs, in most instances, should carry an overlay sheet indicating areas of prominence where strongest resolution should be achieved at the engraver's. (10) Photos with a superimposed screen, no matter how fine, are unusable. (11) For figures that have been published elsewhere, permission must be obtained from their author and publisher to reproduce them. Proper acknowledgment at the end of the figure caption is made as follows: [From J. Jones, Sonics (Smith, Chicago, 1964).) where this is the only reference in the paper to J. Jones; [From J. Jones.4 ] -where J. Jones has been cited in the main text.

The present experiments examine the effect of a weak 40-ms tone burst (cue) on the detection of a closely following 40-ms signal at the same frequency. Detection becomes more difficult as the temporal separation (onset to onset) between them shortens from around 300 ms to under 52 ms. The threshold increase or proximal interference is similar whether signal frequency is constant from trial to trial--frequency certainty--or changing--frequency uncertainty. The increase is also similar whether the cue goes to the same ear as the signal or to the opposite ear. This contralateral interference by such weak cues, only 4 dB SL against a continuous broadband noise, appears to exclude a role for forward masking by the cues. When the preceding tone burst differs in frequency from the signal, threshold increases little at any temporal separation. Combined with earlier results on frequency uncertainty (Scharf, B., et al., 2007, J. Acoust. Soc. Am. 121, 2149-2157), the present results show that a listener can shift focusing to an unexpected signal frequency in less than 52 ms. However, the rapidity of focusing is usually obscured by proximal interference, which possibly occurs whenever cue and signal share the same period (approximately 200 ms) of temporal integration.

Overshoot, the elevation in the threshold for a brief signal that comes on close to masker onset, was measured with signal frequency certain (same frequency on every trial) or uncertain (randomized over trials). In broadband noise, thresholds were higher 2 ms after masker onset than 200 ms later, by 9 dB with frequency certainty, by 6-7 dB with uncertainty. In narrowband noise centered on the signal frequency, thresholds at 2 ms were not elevated with certainty, but were elevated 4-5 dB with uncertainty. Thus, frequency uncertainty leads to less overshoot in broadband noise, to more overshoot in narrowband noise. Reduced overshoot in broadband noise may come about because the masker, given its many frequencies, disrupts focusing at onset as much under certainty as uncertainty. Once the initial disruption dissipates, threshold is lower with certainty so overshoot is greater. In contrast, a narrowband noise with frequencies only near the signal does not disrupt focusing when the signal frequency is known beforehand, so overshoot is absent. When frequency is uncertain, the narrowband noise serves to focus attention on the signal frequency; as this requires time, detection near noise onset is poorer than later on, so overshoot is present.

How quickly can a listener focus on a single tonal cue that indicates the frequency of an upcoming signal? Initial measurements were made with frequency uncertainty (signal frequency varies randomly from trial to trial) and with certainty (same frequency on all trials). Measured by a yes-no procedure, thresholds for 40- and 20-ms signals presented in continuous broadband noise at 50 dB SPL were higher in uncertainty than in certainty; the difference decreased monotonically from 5 dB at frequencies below 500 Hz to under 3 dB above about 2500 Hz. This decrease in the detrimental effect from uncertainty, which comes about with increasing signal frequency, may result from preferential attention to higher frequencies. In a second experiment, frequency again varied randomly, but each trial now began with a cue at the signal frequency. The critical variable was the delay from cue onset to signal onset. A delay of 352 ms eliminated the detrimental effect of frequency uncertainty at all frequencies. At the shortest delays of 52 and 82 ms the detrimental effect was reduced primarily at lower frequencies. Our analysis suggests that shifting focus to a cued frequency region, under optimal stimulus conditions, requires less than 52 ms.

The induced reduction in the loudness (ILR) of a weaker tone caused by a preceding stronger tone was measured with both tones in the same ear (ipsilateral ILR) and also in opposite ears (contralateral ILR). The two tones were always equal in duration and were presented repeatedly over several minutes. When the tone duration was 200ms, for 24 listeners the loudness reduction averaged 11dB under ipsilateral ILR and 6dB under contralateral ILR. When the duration was 5ms, ILR was 8dB whether ipsilateral or contralateral. For each duration, ipsilateral and contralateral ILR were strongly correlated (r around 0.80).

Induced loudness reduction (ILR) is the decline in the loudness of a weaker tone induced by a preceding stronger tone. In this study we investigate how ILR depends on exposure time and signal frequency. For 12 listeners, successive magnitude estimation was used to measure the loudness of 70-dB-SPL test tones, presented with and without preceding 80-dB-SPL inducer tones at the same frequency. Experiment 1 measured the evolution of ILR over time at 0.5 kHz. The results suggest that ILR may begin after a single inducer presentation, and increases over at least 2 to 3 min as the inducer and test tones are repeated every few seconds. Following the cessation of the inducer, the recovery of loudness is slow and still incomplete after 1 min. Experiment 2 extended the measurements to additional signal frequencies. The results show that the amount of ILR and its evolution over time are approximately the same at frequencies from 0.5 to 8 kHz. Similarly, loudness matching showed no effect of frequency on ILR, which averaged 8.2 dB. These findings, together with previously noted similarities among ILR, ipsilaterally induced loudness adaptation, and temporary loudness shift, indicate that loudness reduction induced by stronger sounds is a very common phenomenon.

In the matching of loudness by the method of adjustment, one sound—the variable—is varied in intensity and the other—the standard—is fixed. Listeners usually judge a sound as softer when it is the variable than when it is the standard. They set sound A to a higher level to match sound B when A is the variable than when B is the variable. This difference or adjustment error may be as large as 30 dB but is usually under 5 dB. A large part if not all of the error appears to result from induced loudness reduction, or ILR. ILR is the loudness decline imposed by a stronger tone on a weaker one that follows within a few seconds. Its magnitude and temporal characteristics, including formation and disappearance, are compatible with the characteristics of the experimental conditions that give rise to the adjustment error.

As part of Stevens Day, 26 July 2006, some of his former students and collaborators participate in an informal roundtable titled, S. S. Stevens: The man, the scientist. They present a series of vignettes that focus on personal relationships and exchanges, mostly about psychophysics. We hope thereby to gain insight into the way Stevens identified a problem, placed it in its appropriate context, related it to what was already known, developed needed analytic and experimental procedures, provided the theoretical even philosophical underpinnings, and finally came up with clear and impelling presentations in word and figure. A primary goal of the round table is to provide inspiration and advice to our younger colleagues.

Under appropriate stimulus conditions, a tone may decline in loudness the equivalent of 10 dB and more when preceded by a stronger tone. This induced loudness reduction or ILR was uncovered, indirectly, in a large number of studies by L. E. Marks and his associates [e.g. Marks, J. Exp Psychol HPP 20, 382–396 (1994)]. Those studies seemed to suggest that ILR required that tones be presented at two widely separated frequencies over a relatively large range of SPLs. Although part of the measured loudness changes seemed to stem from response biases, recent studies show unequivocally that the reduction in loudness is mostly sensory and that tones need be presented at only a single frequency and at two levels some 10 to 20 dB apart. The present paper puts together what is known about the dependence of ILR on signal frequency, level, duration, temporal relations, and hearing loss. The role of ILR in various other psychoacoustical phenomena such as induced loudness adaptation and loudness enhancement is reviewed. Much of this knowledge was obtained in direct and indirect collaboration with So/ren Buus. [Work supported by NIH/NIDCD Grant No. R01 DC 02241.]

A tone usually declines in loudness when preceded by a more intense inducer tone. This phenomenon is called "loudness recalibration" or "induced loudness reduction" (ILR). The present study investigates how ILR depends on level, loudness, and duration. A 2AFC procedure was used to obtain loudness matches between 2500-Hz comparison tones and 500-Hz test tones at 60 and 70 dB SPL, presented with and without preceding 500-Hz inducer tones. For 200-ms test and comparison tones, the amount of ILR did not depend on inducer level (set at 80 dB SPL and above), but ILR was greater with 200- than with 5-ms inducers, even when both were equally loud. For 5-ms tones, ILR was as great with 5- as with 200-ms inducers and about as great as when test and inducer tones both lasted 200 ms. These results suggest that (1) neither the loudness nor the SPL of the inducer alone governs ILR, and (2) inducer duration must equal or exceed test-tone duration to yield maximal amounts of ILR. Further analysis indicates that the efferent system may be partly responsible for ILR of 200-ms test tones, but is unlikely to account for ILR of 5-ms tones.

Two opposite sequential loudness effects concern the effect of a stronger Tone 1 on the loudness of a subsequent weaker Tone 2, as assessed by loudness matches with Tone 3. Loudness enhancement is reported when Tone 1 precedes Tone 2 by 50 to 100 ms. Loudness recalibration (or induced loudness reduction) is obtained for delays of about 1 s. This letter argues that what appears as an enhancement of Tone 2’s loudness is, in fact, an induced reduction of Tone 3’s loudness, which occurs because Tones 1 and 3 are at the same frequency. Preliminary experiments support this analysis.

The amount of loudness recalibration (i.e., the drop in loudness of a moderate-level tone caused by a preceding intense recalibration tone of the same frequency) was measured as a function of the recalibration tone’s duration and level using a 2AFC procedure. The 500-Hz test tone—presented at 60 and 70 dB SPL—and the 2500-Hz variable-level comparison tone both lasted 200 ms. Results for 10 normal listeners show that 5-ms recalibration tones yielded only 3 (test tone at 60 dB SPL) to 4 dB (at 70 dB SPL) of recalibration, whether their level was 80, 95, or 110 dB SPL. In contrast, 200- and 500-ms recalibration tones at 80 and 95 dB SPL (200 ms only) yielded 6 (at 60 dB SPL) to 10 dB (at 70 dB SPL) of recalibration, again with no apparent effect of level. Note that 5-ms recalibration tones at 95 dB SPL yielded much less recalibration than 200-ms recalibration tones at 80 dB SPL, despite their nearly equal loudness. These data indicate that recalibration is not governed by the loudness of the recalibration tone and that recalibration-tone duration is a crucial parameter for recalibration of loudness. [Supported by NIH/NIDCD R01DC02241.]

This paper reviews the effects of one sound on the loudness of a following sound. The following sound is usually perceived as softer than when presented in isolation. At least five sequential effects can be identified. (1) Simple loudness adaptation: the earlier part of an ongoing sound results in a decline in the loudness of later parts. (2) Ipsilaterally induced adaptation: increments in the level of an ongoing sound induce a decline in the loudness of the ongoing sound. An intermittent louder sound at a nearby frequency also causes a decline in loudness. (3) Loudness recalibration: the stronger and weaker sounds of induced adaptation are separated by a silent interval, but otherwise the decline in the loudness of the weaker sound, called recalibration, seems to follow much the same rules as induced adaptation. (4) Temporary loudness shift: a very intense sound often causes a temporary decline or shift in the loudness of a following weaker sound. This temporary shift is attributed to fatigue of the cochlear hair cells. (5) Loudness enhancement: a brief sound is louder when it follows a stronger sound within 200 ms or so. These various sequential effects are largely perceptual, but their physiological bases can only be guessed at. This paper is about changes in the loudness of one sound caused by exposure to a preceding sound. With no silent interval between the two sounds and with no stimulus change, such sequential effects are generally referred to as loudness adaptation. With a silent interval, among the notable sequential effects are loudness recalibration, temporary loudness shifts, and loudness enhancement. Except for loudness enhancement, the effect of the preceding sound is either to leave unchanged or to diminish the loudness of the following sound. Fechner had little to say about loudness, no doubt because control of sound intensity was so difficult in the 19 century. He did refer to sequential effects in psychophysics with respect to the measurement of difference thresholds but not with respect to sensory magnitudes, such as loudness. Many contemporary psychophysicists (see Baird, 1997) do consider effects of preceding stimuli and responses, even of the whole context, on responses, especially in scaling procedures. I limit myself to what appear to be sequential effects on perceiving rather than uniquely or mainly on responding.

Neuhoff 1 reported that “rising level tones... change (in loudness) more than falling level tones despite having the same actual change in level... indicating that direction of change is an important (and previously unaddressed) factor in the perception of dynamic loudness change”, and speculated that: “In a natural environment this over-estimation could provide a selective advantage, because rising intensity can signal movement of the source towards an organism.” Leaving aside the question of why it may not be as important for survival to detect the movement of a sound source away from an organism, we dispute the assertion that there is no prior evidence about the influence of direction of change on the degree of change in perceived loudness. This evidence does exist and shows, in contrast to the result reported by Neuhoff 1 , that declining signal intensity covers a greater range of loudness than does rising signal intensity.

The present study was inspired by results of Allen and Wightman [J. Speech Hear. Res. 38, 503–511 (1995)]. The goal of the present study was to explore the effects of uncertainty of a weak tonal masker (or distractor) on the detection of a tone in noise. The thresholds of three listeners were measured for 1-kHz tone burst (350-ms duration) in broadband noise (300–1800 Hz, at 60 dB SPL). A 2IFC tracking procedure (3 down, 1 up) with six interleaved tracks was used. On each trial, the distractor frequency was selected randomly from six frequencies (525, 800, 925, 1075, 1280, and 1600 Hz) outside of the critical band surrounding the signal. The distractor came on simultaneously with the signal in the signal-plus-noise interval and also at the corresponding moment in the noise-alone interval. The preliminary results indicated that thresholds for the signal increased in the presence of distractors by 3–10 dB, depending on the frequency of the distractors and type of noise. Thresholds were higher for distractors close to the signal frequency, more so in intermittent noise than in continuous noise.

Simple loudness adaptation for pure tones was measured at frequencies from 0.125 to 16 kHz and at sensation levels from 5 to 60 dB. Sixteen young listeners with normal hearing participated in four experiments. Most of the loudness measurements were obtained by the method of successive magnitude estimation; some were also obtained by loudness matching. The two indices of loudness adaptation gave similar results. At all frequencies, loudness adaptation increased as sensation level decreased. After 6 min, average loudness declined at most frequencies by about 20% at 40-dB sensation level (SL) and by between 70% and almost 100% at 5-dB SL. Adaptation also increased with increasing frequency, and was especially marked at 16 kHz, where loudness declined more than 60% at a sensation level as high as 40 dB. Most of the adaptation occurred usually within the first 3 min of exposure, but loudness continued to diminish at a slower rate up to around 6 min. The dependence of loudness adaptation on frequency and level can be largely accounted for by the restricted-excitation-pattern hypothesis. Adaptation is assumed to take place when excitation is restricted to a narrow region of the cochlea. This hypothesis is supported by a quantitative analysis based on excitation patterns derived from measurements of masking.

Earlier we presented data (Scharf et al. (1994) Hear. Res. 75, 11–26) from a young patient (S.B.) who had undergone a vestibular neurotomy, during which the olivocochlear bundle (OCB) was severed. Those data are complemented by measurements on 15 other patients — some like S.B. with normal audiometric thresholds, none with a loss greater than 35 dB at experimental frequencies. Comparisons of performance for the same ear before and after surgery or between the operated and healthy ears do not provide evidence that the lack of OCB input impairs the following psychoacoustical functions: (1) detection of tonal signals, (2) intensity discrimination, (3) frequency selectivity, (4) loudness adaptation, (5) frequency discrimination within a tonal series, (6) in-head lateralization. Data on single-tone frequency discrimination are equivocal. These mostly negative results apply to listening both in the quiet and, where relevant, in noise. The only clear change in hearing after a vestibular neurotomy is that most patients detect signals at unexpected frequencies better than before. This change suggests an impaired ability to focus attention in the frequency domain. Although limited in scope, our finding that human hearing without OCB input is essentially normal agrees with much of the relevant literature on animal behavior and with the patients' self-reports.