We are very pleased to welcome Jason Rothman of the University of Reading. He earned a Ph.D. in Hispanic Linguistics from UCLA in 2005 and focuses on language acquisition (especially L2 and L3) and multilingualism. His talk - "Constraining transfer in L3 acquisition: Introducing the typological proximity model" - will be taking place on Thursday, May 19 at 3:10 PM in SS 1070.
In recent years, formal linguistic approaches to language acquisition have seen a sharp increase in interest in true multilingualism, that is, investigating what happens when a bilingual acquires an additional language (e.g., see Garcia Mayo and Rothman, 2013; Halloran and Rothman, 2013 for literature review). In light of this interest, several models of cross-linguistic influence have been formalized: the Cumulative Enhancement Model (Flynn et al., 2004), the L2 Status Factor (Bardel and Falk, 2007; Falk and Bardel, 2011), the Typological Primacy Model (Rothman, 2011, 2015) and most recently the Scapel Model (Slabakova, in press) and the Linguistic Proximity Model (Westergaard et al. in press). Because the sources of potential transfer (with two sources to select from) are more dynamic than in L2 acquisition, and given the crucial role that transferred representations play for learnability across development and ultimate attainment, it is not surprising that the majority of work in the nascent field has focus on determining how the mind selects between the L1 and L2 for transfer in L3 interlanguage formation. In this talk, I will introduce the audience to the current state of the science in the field of generative L3 studies. I will argue that determining with accuracy the source and dynamics of transfer at the level of mental representation in the very first stages of L3 interlanguage must be done before any viable developmental theories can be meaningful. I will explain the tenets of the Typological Proximity Model, which sits at the interface between formal linguistic constraints and cognitive economy to explain why initial stages L3 transfer seems to be complete (from either the L1 and L2 in its entirety) and obtains as a result of the parser’s processing of specific linguistic cues in the L3 input stream after initial exposure.