Evidence for Two Mechanisms to Account for the Speech to Song Illusion, the Verbal Transformation Effect, and the Sound to Music Illusion

Abstract

Introduction:Five studies examined the speech to song illusion, the verbal transformation effect, and the sound to music illusion in order to determine if they were distinct phenomena and to assess if they could be accounted for by a single perceptual/cognitive mechanism. Methods:In Study 1, word lists varying in length from 1 word (as often used to study the verbal transformation effect) to 4 words (as often used to study the speech to song illusion) were presented to participants for 4 minutes to investigate the percepts that were elicited. In Study 2 participants were asked to indicate YES/NO if they experienced the speech to song illusion when listening to word-lists modified by a vocoder. In Studies 3-5 participants were asked to click a button as soon as the shift in percept occurred from speech (or sound) to a music-like percept to assess the time-course of the speech to song (or sound to music) illusion.Results: Study 1 shows that the verbal transformation effect and the speech to song illusion elicit similar percepts. In Study 2 participants indicated that the speech-like stimuli elicited the speech to song illusion more than the noise-like stimuli. In Studies 3-5 similar time-courses were observed for the speech to song illusion and the sound to music illusion.Discussion: Previous, single-mechanism accounts of the speech to song illusion are discussed, but none of them adequately account for all of the results presented here. A new model is proposed that appeals to both a perceptual/"lower-level" mechanism and a cognitive/"higherlevel" mechanism.