I found that even brief experience with learning an L2 led to rapid and pervasive changes in production of the L1 (“phonetic drift”). Moreover, this cross-language influence occurred at segmental, subsegmental, and global levels, suggesting that structural links between the L1 and L2 are established at the onset of L2 learning between corresponding phonological units of varying size (Chang, 2010c, 2011, 2012a).
Interestingly, L2 effects on the L1 were less pronounced when learners had previous knowledge of the L2, suggesting that the influence of recent L2 experience is enhanced in novice learners by a novelty effect and diminished by progressive familiarization with the L2 (Chang, 2013b).
In my latest work on this topic (Chang, 2019b), I tracked the progress of phonetic drift over the course of a year abroad in the L2 environment, finding that L1 modifications persisted in some (but not all) aspects of the L1 after L2 instruction and, crucially, in the absence of frequent active L2 use. This phenomenon argues in favor of increased transparency and consistency in reporting language background data in behavioral linguistic research.
Recently, I also wrote a review chapter for the Oxford Handbook of Language Attrition synthesizing the research in this area over the last four decades (Chang, 2019c).