31 mai 2019, 11h30–12h30
Language is such an ubiquitous feature in human societies that one may think it is trivially easy to learn. However, this is clearly not true for anyone who has actually tried to learn a language in adulthood, or to design artificial agents who can process language. There is something in the cognitive system that children bring to the task, which makes all the difference: A child exposed to several languages may actually acquire them all, without any special effort by themselves, or by those around her. Or do they? Many current behavioral intervention programs assume that children's optimal language and brain development is powered by one-on-one interactions with caregiving adults, who should play with the child and adapt both activities and speech to the child's current interests. Such recommendations seem to stand at odds with practices anthropologists have documented in many cultures, where the child is just one more participant in multi-group interactions, and is expected to catch on and even contribute to joint tasks. In this talk, I discuss the insights that cognitive science can provide on this controversy, as well as the evidence bearing both on the importance of adult-child interactions, and its prevalence across cultures.
Alejandrina Cristia (EHESS, CNRS), « Is learning language child's play? Resolving a controversy using cognitive science », IAST General Seminar, Toulouse : IAST, 31 mai 2019, 11h30–12h30.