Soutenance HDR Isabelle Dautriche

Date: 
Vendredi, 13 Janvier, 2023 - 10:00
Date fin: 
Vendredi, 13 Janvier, 2023 - 18:30
Lieu: 
Campus Saint Charles, salle des Voûtes

La soutenance HDR d'Isabelle Dautriche est programmée à 14 heures

10H Judit Gervain (University of Padoa, Italy & CNRS)

How do infants represent speech: the developmental origins of the embedded neural oscillations model

The proposal (Giraud & Poeppel 2012) that a hierarchy of embedded neural oscillations support speech and language processing in the brain has received ample evidence in adults. However, the developmental orgins of such a neural architecture remain to a large extent unknown. The talk will present EEG data from newborns and older infants showing that neural oscillations are already in place early in development and are shaped by prenatal and postnatal experience.

—————-

11H Coffee break

—————–

11H30 Luca Bonatti (University Pompeu Fabra, Spain)

The role of logic in infant cognition

Do infants have a logic in their mind, and what role could it have in their cognitive system? David Hume argued that no novel content can arise by  simply inspecting ‘relations of ideas’ — what we would now call logic, but only by knowing ‘matters of fact’ –  objects, events, causal relations. This creates a puzzle: how is it possible that a system that adds no novel knowledge be useful to our cognition of the world? I will argue that the puzzle is a bit less puzzling if one thinks that knowledge is not acquired not only by adding novel information, but also by reducing uncertainty. I will present  experiments with adults and infants suggesting that elementary logical processes are indeed spontaneously triggered in different contexts, and precisely serve the role of reducing uncertainty by eliminating possibilities.

—————-

12H30 Lunch ILCB

—————–

14H Soutenance HDR, Isabelle Dautriche (CNRS & Aix-Marseille University)

How children learn the meaning of words 

One central task in most linguistic theory is to provide an account of the acquisition of language: What kind of machine in its initial state, supplied with what kinds of input, could acquire a natural language in the way that infants of our species do? This question is usually confined to the acquisition of phonology and syntax, leaving vocabulary aside. After all, words must be learnt by noticing the real-world contingencies for their use. No other theory could explain why English children learn to associate the sound sequence /ˈbɛtə/ to the meaning “better” and Turkish children to “worse”. I argue that focusing on distributional or statistical abilities provides an incomplete picture of word learning as it fails to account for how learning occurs, what kind of knowledge learners bring into the word learning problem and what properties of language may facilitate it. I will present experimental and computational work that suggest that 1) children are active learners, able to combine and weigh several sources of information while learning the meaning of words; 2) that the vocabulary of natural languages has evolved to be more learnable by infants and 3) that infants come equipped with non-linguistic biases that guide their learning, some of which may be shared with other species.

Committee:

Jennifer Culbertson, University of Edinburgh
Balthasar Bickel, University of Zurich
Luca Bonatti, University Pompeu Fabra
Judit Gervain, University of Padoa & CNRS
Noël Nguyen, Aix-Marseille Université

—————–

17H Apéro