Top-down gain control in the auditory system: Evidence from identification and discrimination experiments

Abstract

The influence of intensity range in auditory identification and intensity discrimination experiments is well documented and is usually attributed to nonsensory factors. Recent studies, however, have suggested that the stimulus range effect might be sensory in origin. To test this notion, in one set of experiments, we had listeners identify the individual tones in a set. One baseline condition consisted of identifying four 1-kHz, low-intensity tones; the other consisted of identifying four 1-kHz, high-intensity tones. In the experimental conditions, these baseline tone sets were augmented by adding a fifth tone at either 1 or 5 kHz. Added 5-kHz tones had little effect on identification accuracy for the four baseline tones. When an added 1-kHz tone differed substantially in intensity from the four baseline tones, it adversely affected performance, with the addition of a high-intensity tone to a set of low-intensity tones having a more deleterious effect than the addition of a low-intensity tone to a set of high-intensity tones. These and further results, obtained in an exploration of this asymmetrical range effect in a third identification experiment and in two intensity-discriminationexperiments, were consistent with the notion of a nonlinear amplifier under top-down control whose functions include protection against sensory overload from loud sounds. The identification data were well described by a signal-detection model using equal-variance Laplace distributions instead of the usual Gaussian distributions.

Knowledge alters visual contrast sensitivity

“…The results support a model in which the function of top-down contrast sensitivity control is to optimize performance while protecting against overload. In this regard, it is similar to the top-down control mechanism proposed by Parker et al (2002) for auditory intensity. We suggest that contrast sensitivity is controlled in a top-down manner in addition to bottom-up control.…”

“…Here it has been suggested that the function of a nonlinear intensity control mechanism is to maximize discriminability and protect the auditory system from an overload. Specifically, Parker, Murphy, and Schneider (2002) showed that one function of auditory gain control, the auditory analogue of visual contrast sensitivity control, is to protect the auditory system from an intensity overload.…”

“…For reasons clarified below, instead of the standard detection threshold paradigm, we used a new paradigm that was employed by Parker et al (2002) to show that auditory-intensity control, the auditory analogue to visual contrast sensitivity control, is under top-down influence. In particular, we combined an absolute identification task with a cuing paradigm.…”

Posttransient shifts in auditory lateralization

“…One discrepancy between the two observations concerns the role of listener expectancies. The reduction in identification accuracy was essentially eliminated when listeners were instructed to ignore a predictable loud tone, implying that the gain-control mechanism can be adjusted according to the experimental contingencies (Parker et al, 2002). In contrast, reduction of loudness usually occurs even when the listeners are instructed to ignore the inducer (Arieh & Marks, 2003b).…”

“…Loud transients have also been shown to affect identification and discrimination (Parker, Murphy, & Schneider, 2002). In one condition, listeners were instructed to identify four weak tones (e.g., 25,30,35,and 40 dB [SPL]), and in a second condition, a fifth tone, either 90 or 45 dB (SPL), was added to the stimulus set.…”