Previous research on alternation phenomena in English as a Foreign Language has mostly focused on how learners’ mother tongue influences their choice of variant, while ignoring the impact of their proficiency level. Including proficiency level allows one to track the development of probabilistic grammars that guide the choice between variants at different stages of language learning. In this way, Dubois et al.’s (2023) study on the genitive alternation showed that possessor animacy, otherwise the strongest constraint in the genitive alternation, is weaker for low-proficiency learners of English, which they attribute to general learning mechanisms that apply regardless of the learners’ mother tongue. In the present study, we investigate the impact of learners’ proficiency level on the choice between will and be going to (he will read the newspaper vs. he is going to read the newspaper) in spoken language, which differs from the genitive alternation as it does not involve a change of word order. Additionally, investigating how proficiency level affects the choice of future marker forms a desideratum within research on the acquisition of the future markers, where learners’ mother tongue does not appear to be highly influential (Bardovi-Harlig 2000: 411–412). Methodologically, we collected 3616 instances of will and be going to from the Trinity Lancaster Corpus, a three-million-word corpus consisting of transcribed recordings from a spoken language exam between an examiner, who is a native speaker of British English, and low-intermediate to advanced learners of English from a wide variety of mother tongue backgrounds. The future marker observations were annotated for constraints known to probabilistically influence the choice of variant, including structural persistence, the type of sentence, clause, verb and subject, the presence of temporal adverbs, the temporal proximity of the future event and the length of the clause (see Engel & Szmrecsanyi 2023). The choice of variant was then analyzed using mixed-effects logistic regression, where the probabilistic constraints were entered as predictors in interaction with the speakers’ proficiency level. Results show that learners differ from native speakers regarding most relevant constraints at specific stages of language learning, regardless of their mother tongue background. Specifically, low-proficiency learners are sensitive to more constraints than native speakers, which is due to their more restricted usage of be going to for events that are relatively certain to happen in the near future. At the same time, these learners might be influenced by prescriptivist rules from English textbooks, which consistently cover the usage of the future marker variants (Burton 2023). By contrast, native speakers do not distinguish between the variants to the same extent, resulting in their more frequent use of be going to. Bardovi-Harlig, Kathleen. 2000. Chapter Seven: Past, Present, and Future. Language learning 50(s1). 409–437. Burton, Graham. 2023. Grammar in ELT and ELT Materials: Evaluating its History and Current Practice (Second Language Acquisition 164). Bristol, Blue Ridge Summit: Multilingual Matters. Engel, Alexandra & Benedikt Szmrecsanyi. 2023. Variable grammars are variable across registers: future temporal reference in English. Language Variation and Change 34(3). 355–378.
Dubois, T., Paquot, M., Szmrecsanyi, B., & et al. (2024). How do probabilistic grammars develop in spoken EFL? The influence of proficiency level on the choice between will and be going to. ICAME45, Vigo, Spain. https://hdl.handle.net/2078.5/215592