In this commentary, I question Westergaard's argument that third language (L3) data can be used to decide between theories such as Full Transfer Full Access, involving wholesale transfer initially, and Full Transfer Potential, involving property-by-property transfer. I suggest that much L3 data will be amenable to explanation under either theory.…

In this commentary, differences between feature re-assembly and feature selection are discussed. Lardiere's proposals are compared to existing approaches to grammatical features in second language (L2) acquisition. Questions are raised about the predictive power of the feature re-assembly approach. (Contains 1 footnote.)

Responds to a reanalysis of study findings that refute the claim that negative evidence can lead to parameter setting in second-language acquisition, presenting empirical evidence from French learners of English, suggesting that positive second-language acquisition data do not guarantee the loss of native language parameter settings. (26…