What Turns Speech Into Song? Investigations of the Speech-to-Song Illusion

Abstract

In the Speech-to-Song Illusion a spoken phrase is presented repeatedly and begins to sound more like it was being sung. The present study used several different types of stimulus phrases to elicit the SpeechtoSong Illusion to determine the underlying cognitive mechanism responsible for this illusion. Previous stimuli used to elicit the SpeechtoSong Illusion have been phrases extracted from real sentences. We examined whether having a meaningful phrase (in Experiment 1), or even real words (in Experiment 2), were necessary for eliciting this illusion. We also examined how the length of the stimulus impacts the occurrence of the SpeechtoSong Illusion by manipulating the number of words in the repeated phrases (in Experiment 3) and by manipulating the number of syllables in each word of the phrases (in Experiment 4). The results of these 4 experiments support an underlying mechanism as described in Node Structure Theory: satiation of lexical nodes allows for the syllables to "pop out" as music-like beats.