Elisa Fiorenza ¹ Will Travers ^2*

• ¹ University of Messina, Messina, Sicilia, Italy
• ² Davidson College, Davidson, United States

Metalinguistic awareness is considered by many to be a crucial ingredient for successful adult foreignlanguage acquisition. Researchers further suggest that it helps bilinguals learning a third language (L3) even more than it does monolinguals learning their second language (L2). In response to a recent proposal that differences in metalinguistic awareness may be responsible for the variation in grammatical development often witnessed amongst bilingual university students, and that L3 teaching methods should accordingly be modified as a result, research that actually measures these varying levels of metalinguistic awareness has now become urgently needed. Due to a lack of standardized assessments and an inability to converge on how best to measure this set of abilities in adults, few studies have yet attempted to operationalize this variable in the US-based L3 context, and those that did adopted diverse methods, raising issues of comparability. For multilingual language learners, especially those who grew up speaking a heritage language, the challenge that researchers face is whether to measure metalinguistic awareness in a native language, a prior-learned L2 (if one such language exists), the learner's current target language, or a language unknown to the learner entirely. This article highlights these methodological complexities and calls for a principled approach to measuring metalinguistic awareness before implementing any pedagogical changes in terms of how third languages are taught.