Exploring the Node Structure Theory as an Explanation for the Speech-to-Song Illusion

Voiceover

Presenter(s)(s)

Kavery Kallichanda

Abstract or Description

The speech-to-song illusion is a phenomenon in which the repetition of a spoken phrase causes the words to be perceived as sung instead of spoken. The Node Structure Theory (NST) has been proposed as the underlying mechanism for the illusion. NST proposes that increased activity in a node, known as priming, causes the node to be activated, with repeated activations over a small space of time resulting in satiation, or a temporary reduction in the node’s ability to be primed and activated. Thus, this theory suggests that the repetition of a phrase causes the satiation of the lexical nodes associated with the phrase, resulting in the loss of the initial “speech” percept, with the continued priming of the node increasing conscious awareness of the metrical pattern found in the repeated phrase, causing the perception of song. According to this model, the effect of the speech-to-song illusion on a spoken phrase should be temporary. Hence, this study looked at the duration of the illusion to either provide support for or evidence against the NST model. Participants were asked to complete a survey that requested them to rate their perception of a spoken phrase after hearing it once and after hearing it twice on a scale of 1 (“sounds exactly like speech”) to 5 (“sounds exactly like singing”). They then completed the survey once more the following day. The results showed that the effects of the speech-to-song illusion are indeed temporary, which supports NST as the underlying mechanism for the illusion.