Investigadora Responsable: Marcela Peña.
__________________________

In recent decades, many studies have revealed that young children are extraordinary learners. Most studies have explored what and when learning occurs but only recently infant research has focused on studying the mechanisms underpinning such learning. The current proposal attempts to contribute to filling this gap in knowledge by exploring the role of the neural entrainment on the early steps of language acquisition, communication skills and vocalization development, using artificial intelligence methods for analysis and modelling.

Neural entrainment is defined as the ability of the brain signal to synchronize with other signals. An example of neural entrainment is the synchrony observed between the oscillatory brain activity and the oscillatory changes in the speech signal, which may involve synchrony between their amplitudes and/or frequencies. Adults’ and children’s studies support the idea that the neural entrainment is a mechanism that the hearing brain exploits to attune to speech units and enhance the mental representation of the attended sounds. For instance, in adults, the amplitude of the neural entrainment to syllables when listening to normal speech (i.e. 4-7 syllables per second) positively correlate with speech intelligibility, word comprehension, and word segmentation from fluent speech. Moreover, in children, the neural entrainment to auditory beats is atypically out-of-phase in dyslexia, allocating it a crucial role in reading.

In contrast, up to date, little is known about how such a sophisticated process operates during infancy, despite some studies have shown that neural entrainment is functional at early age. Indeed, 8-month-old infants show neural entrainment to syllables and the larger the entrainment the better is the discovery of words from fluent speech.

This proposal aims at exploring whether and how the amplitude and precision of the neural entrainment to speech relate to the performance in word learning, communicative abilities and vocalization complexity in 12-month-old infants. In study 1, we will measure the neural entrainment to speech when infants listen to either fast or slow speech. Here, we expect to unveil some critical aspects of the neural entrainment triggered by incoming stimuli. In study 2, we will expose infants to a cocktail party-like situation to assess the ability of the infant brain to selectively entrain to one voice over other when listening to two voices simultaneously, one of them using infant-directed and the other adult-directed speech. Here, we expect to elucidate some top-down aspects of the preference for speech that human exhibit from birth.

Our main hypothesis predicts that the infants who exhibit greater amplitude and precision of neural entrainment to speech (slow and fast in study 1, and infant- and adult-directed speech in study 2), would perform better in a word learning task, score higher in communication skills and exhibit more complex vocalizations. To assess an eventual predictive value of neural entrainment for communicative capacities and vocalizations we will measure those abilities at 18 and 24 months.

We will submit our brain, word learning and vocalization data to deep learning and machine learning approaches trying to identify predictive patterns for language development. We will then enrich our model by including bio demographic data of the infants and families, and generate new analytic tools for the exploration of the relationship between the brain activity and linguistic and communicative abilities during the early steps of language acquisition.

Together, by applying highly innovative experimental design and methods, our proposal seeks to provide new knowledge on whether the infant brain activity entrain to the speech signal and about the potential implications of such entrainment for early linguistic and communicative development.

Our results may serve as inputs for the current models about the brain mechanisms underpinning language acquisition, may nourish the literature concerning the study of biological markers for language learning and language difficulties during early development and hopefully may provide new innovative analytic tools for the study of the earliest steps of language acquisition.