NP Complexity in EFL Argumentative Writing at the CEFR B1, B2 and C1 levels: Corpus-driven Fine-grained Measures

Abstract

Fine-grained measures of Noun Phrase (NP) complexity are being employed to describe linguistic complexity in learner language.
However, the use of pre-established fine-grained NP complexity indices prevents researchers from analysing the actual range of NP
complexity types that learners display in different text types and at different levels. Likewise, the use of learner corpora which are not aligned with the CEFR hinders the identification of NP complexity criterial features, i.e. NP complexity types which show statistically significant differences in their use at different levels, between CEFR levels. This study contributes to the existing literature by exploring NP complexity in L1 Spanish EFL learner writing using corpusdriven fine-grained measures obtained from a CEFR-aligned learner corpus with two main objectives: a) to analyse the variety of NP complexity types employed at the CEFR B1, B2 and C1 levels; and b) to reveal NP complexity criterial features at those levels. After manually parsing 6,182 NPs in 140 argumentative texts (44 texts at B1, 50 at B2 and 46 at C1), 39 NP complexity types were identified and employed as fine-grained NP complexity measures. The findings reveal an increase in the variety of NP complexity types at the CEFR B2 level. The results highlight the importance of multiple premodification and simple or multiple postmodification, especially prepositional phrases, in NP complexity development across B1, B2 and C1. Specific NP complexity criterial features are also found at B2 and C1. This study underscores the use of corpus-driven fine-grained complexity measures to describe NP complexification in different text types and CEFR levels.