In L2 vocabulary acquisition, psycholinguistic evidence hardly translates into educational programs. Some controversies arise around words whose form is similar across languages: cognates (formally and semantically similar) and false cognates/friends (formally similar). Laboratory experiments demonstrate the cognate facilitation effect: Cognates are responded to faster and learned better than other words. On the other hand, classroom-oriented second language acquisition (SLA) studies emphasize the importance of learner’s "cognate awareness" that is consciously focusing on cross-linguistic similarity between L1 and L2 words. Because such awareness is enhanced in proficient multilinguals, less proficient L2 learners are advised to have their cognate awareness raised in class. For false cognates, the situation is even more complex: Psycholinguistic studies show that their processing may be inhibited or even facilitated compared to non-cognates, depending on task, but when translated, false cognates yield more errors than non-cognates. Here, SLA literature advises that L2 learners are warned about their existence, although there is surprisingly little robust empirical evidence that false cognates are more difficult to learn than other words.

How can we bridge the gap between psycholinguistics and SLA to gather robust but ecologically-valid evidence on word acquisition? What is exactly meant by "cognate awareness" in SLA studies? Are awareness-raising attempts effective in L2 classrooms? This talk will discuss several experiments that have compared the acquisition of L1-L2 cognates, false cognates and L2 non-cognates. It will answer whether raising "cognate awareness" indeed modulates the effectiveness of learning L2 words.