Alejandrina Cristia, CNRS research director and director of the Laboratoire de Sciences Cognitives et Psycholinguistique (LSCP), received an ERC Consolidator Grant for her project entitled "Experience effects in early language acquisition" (ExELang).
The "Experience effects in early language acquisition" project, a worldwide study
How does experience affect child language acquisition? Answering this is crucial for the understanding of human cognition, but current evidence is biased towards rich English-speaking families and insufficient to establish causality.
The researchers intend to fight bias by using an innovative and ecological technique: Audio-recordings collected with a recorder worn by the child throughout a normal day. They will start with 14 extant large-scale corpora, totaling over 24,000 hours of audio, collected from about 1,000 infants, living in rural and urban sites in every peopled continent, and learning a wide variety of mostly unrelated languages, which they will grow to about 40,000 hours. They will develop cutting-edge machine learning algorithms to automatically analyze these long, messy, and multilingual recordings. This will yield measures representing infants' experiences (the speech addressed to them, and the speech they overhear) as well as how advanced their own vocalizations, and how correlated these measures are across individuals and cultures.
They will try to establish causality by including data from Randomized Control Trials (RCTs) where treatment groups have received behavioral interventions aimed at increasing the quantity of speech infants are addressed.
In parallel, they will use a reverse-engineering approach to derive predicted effects of experience from computer programs that model infant learning. Each program will contain different subsets of the learning mechanisms that infants theoretically have. All programs' input is the speech heard by infants, which the program processes to produce "vocalizations”. They will check which subsets of learning mechanisms lead to experience-outcome curves like those found in the infants’ data. Together, these approaches will provide a unique and universal view of how experience causally affects language acquisition.
Linguistics and cognitive science at the heart of an exceptional career
Alejandrina Cristia received a BA in Letters from the Universidad Nacional de Rosario, Argentina, in 2004, which earned her the Annual Prize of the Academia Argentina de Letras. Joining the interdisciplinary Linguistics program, she obtained a masters (2006) and a PhD (2009) at Purdue University, United States. Postdoctoral funding by Gilles de Gennes and the Fyssen Foundation brought her the LSCP in Paris, France. She then joined the Neurobiology of Language department at the Max Planck Institute in the Netherlands for two years, before returning to the LSCP as Chargé de Recherche CNRS in 2013, and completing her HDR, from Ecole normale supérieure, in 2014. She became Directrice de Recherche CNRS in 2019.
In 2017, she received a John S. McDonnell Scholar Award in Understanding Human Cognition, the same year in which she started a team entitled 'Language acquisition across cultures'. She was awarded a bronze medal in linguistics by CNRS in 2020.
A career dedicated to understanding the mechanisms of language acquisition
Her work focuses on how young children learn their native language(s). More specifically, she aims to shed light on the learning mechanisms that are involved in the acquisition process. She employs an interdisciplinary, multi-method approach: Audio- and video-recordings and controlled experiments are used to document infants’ input and outcomes; meta-analyses, to integrate across multiple (necessarily partial and noisy) results; and computational models, to better describe the input exhaustively, and check whether a hypothetical mechanistic explanation linking experiences to outcomes is actually feasible. In current (and future work), she extends this systematic, large-scale approach to cultures that are seldom studied, particularly those that have been described as having distinct properties for language acquisition. This line of investigation allows to assess how resilient language acquisition processes are, which has both theoretical and practical implications.