An investigation into the lexical and grammatical characteristics of English and French L2 textbook reading materials

Publication type
U
Publication status
In press
Authors
Van Parys, A, De Wilde, V., Macken, L., & Montero Perez, M.
Publisher
Anéla (Dutch Society for Applied Linguistics) (Ghent, Belgium)
Conference
Anéla Conference 2024 (Ghent, Belgium)
Download
(.pdf)
View in Biblio
(externe link)

Abstract

L2 textbooks are a fundamental source of input in the L2 classroom. To optimally develop
reading skills, linguistic characteristics of input should increase as learners grow more proficient
(Berendes et al., 2018), ideally both in terms of lexis and grammar (Jeon & Yamashita, 2014).
However, little is known about the lexical and grammatical characteristics of L2 textbook reading
materials. This study focuses on a corpus of reading materials retrieved from L2 textbooks targeting
English and French (i.e., the main L2s in Flemish education) and aims to assess how grade level is
predicted by a set of lexical and grammatical (syntactic/morphological) features.

The corpus contains the reading materials from three English and three French L2 textbook series
used in Flemish secondary education (grades 1-6; ages 12-18), totalling 18 textbooks per L2
(English: 392,341 tokens; French: 293,855 tokens). Using a custom Python script, a range of
features were calculated measuring dimensions of lexis (e.g., word frequency), syntax (e.g., mean
sentence length, clausal subordination), and morphology (e.g., inflectional diversity).

Regression analysis reveals that none of the lexical features significantly predict grade level in any
of the textbook series. However, various grammatical features do significantly predict grade level,
both in terms of syntax and morphology. When comparing L2s, the French materials yield fewer
grammatical features predictive of grade level and generally smaller effect sizes, though
morphology appears more impactful compared to English. Overall, these findings put forward
grammar as the main driver in predicting grade level.