Auditory Induction: Illusory Continuity of the Fainter of Two Alternating Sounds

Abstract

If a louder sound (α) and a fainter sound (β) are presented alternatively (α,β,α,β,α⋯), β will seem to be on continuously (i.e., β will not only be heard alone, but will seem to be present along with α) if the intensity difference between the two sounds exceeds a critical value (varying from less than 3 dB to more than 50 dB depending upon the intensity of α and the relative spectral characteristics of the sounds). This illusory carry-over or “auditory induction” (AI) can persist for durations of α lasting several seconds. AI of the fainter sound is a general phenomenon: α and β may be different noise bands, tones of different frequencies, or one a noise and the other a tone. As a rough approximation for some stimulus pairs, if α is capable of masking β with simultaneous presentation, then AI will usually occur when they alternate. Silent gaps of 50 msec between α and β prevent AI. “Phonemic restorations” [R. M. Warren and C. J. Obusek, “Speech Perception and Phonemic Restorations,” Perception Psychophys. 9, 358–362 (1971)] appear to be a special type of AI with illusory phonemes determined by semantic and grammatical rules. Another form of AI having unique rules is found when α and β are identical except for intensity.