The notion of transfer can be traced to Weinreich's 1953 Languages in Contact, which emphasised the notion of interference, and Lado's 1957 Linguistics across cultures, which examined the phenomenon in the light of the contrastive analysis hypothesis (CAH). Both works highlight the negative aspects of transfer, sometimes also called interference.
More recent research recognises that L1/L2 influence can be positive as well as negative and operates in both directions: as an effect of the L1 on L2 acquisition, and an effect of L2 acquisition on the L1. For these reasons we now prefer the term cross-linguistic influence.
Odlin defines transfer as
the influence resulting from the similarities and differences between the target language and any other language that has been previously (and perhaps imperfectly) acquired.
Kellerman's early study of Dutch learners of English found that learners judged more metaphorical uses of verbs as less likely to be correct (Kellerman, 1978, U-shaped learning). These findings contradict the CAH, which predicts ease of transfer between languages at points of similarity.
Cutler has shown cross-linguistic effects in phonology, when L1 structures handicap later L2 perception:
Non-native listening [is] hindered by the efficiency with which the native procedures operate; even when native procedures are ill-suited to the structure of a non-native language, their application is difficult to inhibit. [...] The baby's hard work in the first year of life pays off handsomely in native listening efficiency; drawbacks only become apparent when adults also engage in non-native listening.