Turtles all the way down: Micro-cues and piecemeal transfer in L3 phonology and syntax
There are several theories which tackle predicting the source of third language (L3) crosslinguistic influence. The two orthogonal questions that arise are which language is most likely to influence the L3 and whether the influence will be wholesale or piecemeal (property-by-property). To my mind, Westergaard’s Linguistic Proximity Model (LPM) is preferable to other theoretical models (say Rothman’s Typological Primacy Model) insofar as it is consistent with many aspects of L2/L3 phonological learnability that I am familiar with. Westergaard proposes a structure-based piecemeal approach to the explanation of third language acquisition (L3A). The model is driven by parsing and dictates that the first language (L1) or second language (L2) structure which is hypothesized to be most similar to the L3 structure will be the one to transfer.