Slide Slam C9 Sandbox Series
Is neural entrainment language specific at 6 months of age?
Ege Ekin Özer1, Konstantina Zacharaki1, Silvana Silva-Pereira1, Marcela Peña2, Nuria Sebastian-Galles1; 1Universitat Pompeu Fabra, 2Pontificia Universidad Católica de Chile
Neural oscillations at multiple frequencies simultaneously entrain to the rhythms in the continuous speech signal (Giraud & Poeppel, 2012). Through neural entrainment, the human brain is allowed to segment speech into phrases, syllables, and phonemes (Meyer, 2017). Adults show stronger neural entrainment to their native speech than non-native speech (Pérez et al., 2015; Ding et al., 2016; Etard & Reichenbach, 2019), but native speech specialization through neural entrainment in infants is relatively unknown. Infants at birth have a coarse ability to discriminate between languages that differ greatly in rhythm, such as French from Russian (Mehler et al.,1988). From 4.5 months of age, they develop the ability to distinguish their native language even from a prosodically similar one (Bosch & Sebastián-Gallés, 1997; Nazzi et al., 2000). Later at 6 months, infants start showing the first signs of attunement to the native language (Werker & Tees,1984; Kuhl et al., 1992). Studying brain-to-stimulus coupling in 6-month-old infants, Barajas et al (2021) did not find differences in amplitude or phase tracking between native and non-native languages, but in their design, infants repeatedly listened to the same sentence from each language. In our project, we want to measure brain-to-stimulus coupling in 6-month-old infants listening to a variety of sentences. With this goal, we are analyzing the data collected by Peña et al (2010), where twenty-six 6-month-old infants listened to sentences from Spanish (native language), Italian, and Japanese, presented both forwards and backwards. All three languages differ in their segmental and phonotactic repertoires, but Spanish and Italian are prosodically similar as they are both syllable-timed languages, whereas Japanese is a mora-timed language (Ramus et al., 1999). Peña et al (2010) found that 6-month-old infants show stronger spectral power in gamma frequency for their native language than non-native languages regardless of rhythm. Our goal is to explore infants’ neural entrainment to the prosodic contour of languages prosodically similar to and distant from their native language. We measure entrainment as the correlation between speech envelopes reconstructed from their EEG and the original speech envelopes (Vanthornhout et al., 2018), filtered below 8 Hz, to capture the prosodic cues in speech: both intonation and syllabic rhythm (Meyer, 2017). We have two potential outcomes regarding our hypotheses. 1) If the neural entrainment to speech in infants is native language specific, we expect to find higher entrainment to native language (Spanish) than to non-native languages (Italian and Japanese). 2) If it is rhythm specific, we expect to find similar neural entrainment to Spanish and Italian, and we expect neural entrainment to Japanese to be lower. We expect no differences between languages in the backward speech condition, as sentences presented in reverse lack the rhythmic information in speech (Ramus et al., 2000).