Music to Your Ears

Abstract

Listeners usually have no difficulties telling the difference between speech and song. Yet when a spoken phrase is repeated several times, they often report a perceptual transformation that turns speech into song. There is a great deal of variability in the perception of the speech-to-song illusion (STS). It may result partly from linguistic properties of spoken phrases and be partly due to the individual processing difference of listeners exposed to STS. To date, existing evidence is insufficient to predict who is most likely to experience the transformation, and which sentences may be more conducive to the transformation once spoken repeatedly. The present study investigates these questions with French and English listeners, testing the hypothesis that the transformation is achieved by means of functional re-evaluation of phrasal prosody during repetition. Such prosodic re-analysis places demands on the phonological structure of sentences and language proficiency of listeners. Two experiments show that STS is facilitated in high-sonority sentences and in listeners’ non-native languages and support the hypothesis that STS involves a switch between musical and linguistic perception modes.

Validation and applicability of the music ear test on a large Chinese sample

“…Numerous cognitive and developmental connections exist between music and language, including bidirectional transfer [ 13 ], the speech-to-song illusion transfer [ 37 ], shared cognitive neural bases [ 38 ], and the association of musical ability with phonological pitch discrimination [ 39 ]. Notably, speakers of tonal languages, such as Mandarin Chinese, exhibit distinct perceptions of music compared to non-native speakers of tonal languages [ 32 , 40 , 41 ].…”

Rhythm is a timescale

“…It seems more likely that the task of the listener is to infer the timing relations such that a "sing-song" effect can arise (which in turn might make the message more appealing and/or long-lasting in memory, due to its atypical "musicality"). The speech-to-song illusion (Deutsch et al 2011) shows that stretches of spontaneous speech can be perceived as music when a certain portion is repeated in a loop (various factors, like the size of the loop, the number of repetitions and the segmental makeup of the looped speech, can affect the strength of this illusion; see Rathcke et al 2021a). This is of interest in the context of the current synthesis because a looped auditory signal is, in fact, isochronous at the level of the entire loop size, and it can become "musical" as soon as the structure of the repeating pattern is revealed in perception (see tapping to speech in Rathcke et al 2021b).…”

Weighting of cues to categorization of song versus speech in tone language and non-tone-language speakers

“…2). In addition, to test the hypothesis that the strength of the speech-to-song illusion is modulated by second language proficiency (Rathcke et al, 2021), we compared the illusion strength of the UK-resident and China-resident groups with a Bayesian independent samples t-test using JASP, collapsing across tone language (Mandarin and Cantonese). (As reported in the Participants section above, UK-resident participants had significantly higher self-reported English speaking and listening proficiency than China-resident participants.)…”

“…This difference in familiarity could potentially drive any group differences in either overall speech-to-song perception or in musicality cue weighting. Indeed, prior work has shown that the speech-to-song illusion is enhanced in less familiar languages, especially languages that are difficult to pronounce (Margulis et al, 2015), and that the illusion is enhanced when hearing phrases in a non-native language (Rathcke et al, 2021). However, our main finding is that native English speakers and tone-language speakers are strikingly similar in their perception of the illusion: they report the same degree of transformation into song, rank stimuli similarly in musicality after repetition, and largely use the same cues when deciding whether a stimulus sounds more like song or speech.…”

“…These findings lead one to expect that the song illusion would be stronger in the Cantonese and Mandarin participants in our study, given the phonological differences between these languages and English, which the participants did not speak natively. Additionally, there is evidence that song illusion strength is stronger for phrases in a non-native language (Rathcke, Falk, & Dalla Bella, 2021), perhaps because semantic representations are not as strongly activated as when hearing one's native language, allowing more focus on prosodic representations.…”