Syntactic distance and L2 learnability: the case of English Relative Clauses

Dora Alexopoulou, University of Cambridge

(joint work with Xiaobin Chen and Ianthi Tsimpli)


Similarity between L1 and L2 facilitates learning. The more L1-L2 similarities exist, the more comprehensible the L2 input is for the learner, leading to faster learning (Kellerman 1983, Jarvis and Pavlenko 2007). This long-held view in the SLA literature has found recent confirmation in studies showing that the linguistic distance with the L1 can predict attainment in L3 (Schepens et al 2020).

Linguistic distance is usually measured on the basis of  lexico-statistical metrics used  in phylogenetic typology, and more recently on metrics of morphological complexity and phonological similarity (van der Silk et al 2019, Schepens et al 2020).  Its effect  on learning  is  identified in broad outcomes like test scores in proficiency tests.

Our aim in this talk is to understand the effect of L1-L2 syntactic distance focusing on  the process of acquisition rather than overall outcomes.  Specifically, we focus  on the effect of syntactic distance on the acquisition of the different  relativisors in  L2 English Relative Clauses (RCs) addressing the following research questions.

  1. Is there an effect of global linguistic distance, as captured by broad typological classifications, on the acquisition of RCs, independent of L1-L2 differences around individual features?
  2. Is there an effect of local syntactic distance regarding variation in the domain of the  syntax of RCs?
  3. Do individual features contribute to syntactic distance equally? Or do inherent properties of linguistic features (e.g. their interpretability) make some features more challenging for learners independently of L1-L2 linguistic distance?

To answer these questions we measure syntactic distance on the basis of differences in parameter settings, adopting the methodology of Longobardi and Guardiano 2009.  We use  macro-parameters (Huang and Roberts 2016) to measure global distance and micro-parameters to calculate local distance.  RCs show rich crosslinguistic variation regarding the choice of relativisors which allows us to  investigate the effect of local distance as well as individual features. For our empirical investigation we use the EFCAMDAT corpus, a corpus of writings of students of English as a foreign language (Geertzen et al 2013). The corpus has data from learners around the world with diverse linguistic backgrounds enabling an investigation of the following linguistic backgrounds: Chinese, Japanese, Russian, German, Brazilian Portuguese and Arabic.


Geertzen, J., Alexopoulou, T., & Korhonen, A. (2013). Automatic linguistic annotation of large scale l2 databases: The EF-Cambridge Open Language Database (EFCAMDAT).In in proceedings of the 31st Second Language Research Forum (SLRF), Carnegie Mellon, Cascadilla Proceedings Project.

Huang, C.-T. J., & Roberts, I. (2016). Principles and parameters of universal grammar. In I. Roberts (Ed.), The Oxford handbook of universal grammar. Oxford University Press.

Jarvis, S., & Pavlenko, A. (2007). Crosslinguistic influence in language and cognition. Routledge.

Kellerman, E. (1983). Now you see it now you don’t. In S. Gass & L. Slinker (Eds.), Language transfer in language learning (p. 112-134).

Longobardi, G., & Guardiano, C. (2009). Evidence for syntax as a signal of historical relatedness. Lingua, 119 (11), 1679 – 1706.

Schepens, J., van Hout, R., & Jaeger, T. F. (2020). Big data suggest strong constraints of linguistic similarity on adult language learning. Cognition, 194 , 104056.

van der Slik, F., van Hout, R., & Schepens, J. (2019). The role of morphological complexity in predicting the learnability of an additional language: The case of LA (additional language) Dutch. Second Language Research, 35 (1), 47-70.

Calendar (black)

10 Nov 2020, 2:00 pm

Location (black)

Online – MS Teams