The development of semantic relations in second language speakers: A case for Latent Semantic Analysis


  • Scott A. Crossley
  • Tom Salsbury
  • Danielle S. McNamara


Latent Semantic Analysis, Lexical Networks, Corpus, Linguistics, Lexical Proficieny, Computational Linguistics, MTLD


This study explores how a Latent Semantic Analysis (LSA) index provided by Coh-Metrix can be used as a method to examine the development of lexical networks in second language (L2) speakers. Using the LSA index along with an index of lexical development (lexical diversity), a measure of general linguistic development (TOEFL scores) and indices of lexical overlap in a yearlong longitudinal study, this study demonstrates that semantic similarity along with TOEFL and lexical diversity values significantly increase as learners study a second language. Measures of word overlap did not demonstrate significant growth. The findings demonstrate that the LSA value reported by Coh-Metrix can be used to approximate lexical growth and that L2 learners likely begin to develop tighter semantic relations between utterances and words within a short period of time. This finding implies the growth of lexical networks. This study also has important implications for the development of coherent speech patterns in L2 learners.


Download data is not yet available.