How Well Can We Predict Second Language Learners’ Pronunciation Difficulties?

Abstract

Mid-20th–century scholars argued that second language (L2) instruction should be rooted in a comparison of the structural characteristics of the first language (L1) and L2. Their enthusiasm for a “scientific” approach to errors reflected the view, based on the contrastive analysis hypothesis (CAH), that learners’ difficulties could be predicted through purely linguistic analyses. Pronunciation seemed particularly amenable to this treatment. If teachers knew their learners’ problems in advance, they could presumably design curricula and activities to address their students’ needs. Although it soon became clear that many aspects of CAH were seriously flawed, interest in a linguistic account of L2 pronunciation difficulties has persisted. This synthesis of empirical findings from pronunciation research demonstrates that the enthusiasm for error prediction has been misguided, largely because of 2 erroneous beliefs: the assumption of uniformity and the assumption of equal gravity. The need for an alternative perspective promoting evidence-based teaching practices is demonstrated.