Proceedings of the Annual Meeting of the Cognitive Science Society

Prenatal auditory learning is well-documented, but not at the phonetic level. It may be that phonetic information is so degraded in the prenatal acoustic environment that Prenatal Phonetic Learning (PPL) becomes impossible. Yet, this has never been clearly established, and some results suggest directly or indirectly that some degree of PPL may be taking place. In this study, we assess the feasibility of PPL from external sources through extensive computational simulations of early phonetic learning in simulated prenatal and postnatal environments. We examined the learnability of phonetic contrasts in utero, the added value of prenatal exposure for subsequent learning and whether it could explain earlier attunement to vowels than to consonants. Our results provide preliminary evidence against the feasibility of PPL. We discuss alternative hypotheses and the remaining work needed to complement our results.