RESEARCH

RELATIONSHIP BETWEEN LEARNER’S COMMUNICATIVE PREFERENCES AND LANGUAGE STRUCTURE

with T.Florian Jaeger and Elissa Newport

Fedzechkina, M., Jaeger T.F., & Newport, E.L. (2013). Communicative biases shape structures of newly acquired languages. Proceedings of CogSci 2013. [pdf]


Fedzechkina, M., Jaeger T.F., & Newport, E.L. (2012). Language learners restructure their input to facilitate efficient communication. Proceedings of the National Academy of Sciences. [pdf]


Fedzechkina, M., Jaeger T.F., & Newport, E.L. (2011). Functional biases in language learning: Evidence from word order and case-marking interaction. Proceedings of CogSci 2011. [pdf]

                                                   

 
 

Why do languages share structural properties? It has been long argued that grammatical structures that reduce processing complexity or increase the communicative efficiency of a language tend to persist diachronically and cross-linguistically. By what means do communicative or processing pressures come to shape grammar over time, however, remains unknown. In a series of artificial language learning experiments I’ve explored the possibility that these pressures operate during language acquisition biasing learners to slightly deviate from the input they receive as they acquire the language.

PROCESSING INFLUENCES ON LANGUAGE ACQUISITION

with John Trueswell, Lucia Pozzan,  and T.Florian Jaeger

It has been long argued that processing pressures, such as a preference to maximize the number of linguistic dependencies processed at each point in time, contribute to the way language structures change over time (cf. ‘Maximize On-line Processing’ principle (Hawkins, 2004)). In collaboration with Lucia Pozzan and John Trueswell (University of Pennsylvania), I am currently conducting miniature artificial language learning experiments involving child and adult learners to address the question of whether the correlation between headedness and marking type in a language (head-final languages typically have dependent marking while head-initial languages tend to have head-marking) can be explained by learners’ preference for early disambiguation.


With Florian Jaeger, I’m also testing dependency-length minimization accounts (e.g., Gibson, 1998) against behavioral data.