Prediction in language learning: Can we teach it and what sort of knowledge is generated?

Supervisor: Professor Emma Marsden

A) Rationale for the project

When we hear or read language in real time, we constantly, and extremely rapidly, anticipate which sounds, words and grammar might come up next. It is not clear whether this phenomenon is the result of having already learned language - that is, after multiple experiences we become adept at predicting what will come next - or whether, in fact, 'prediction' is a key mechanism by which we actually learn language- that is: if our predictions are met by what we subsequently hear, this establishes and consolidates knowledge of the language; if our predictions are not met, we learn from our error and tally the likelihood of particular combinations in language (not) occurring. To date, there is strong evidence of prediction in native speakers, but evidence is much less clear in second language (L2) learners. Also, L2 research to date has (a) focused on a narrow domain of grammar in the noun phrase (gender, animacy, case) and (b) not yet investigated whether explicitly teaching and practising prediction can help learning. This research project would make a cutting-edge contribution to both learning theory and teaching practice by investigating these issues in a classroom experiment, focusing on hitherto neglected syntax.

B) References that should be read

Andringa, S., and Curcic, M. (2015). How explicit knowledge affects online L2 processing: Evidence from differential object marking acquisition. Studies in Second Language Acquisition, 37(2), 237-268.

Hopp, H. and Lemmerth, N. (2016). Lexical and syntactic congruency in L2 predictive gender processing. Studies in Second Language Acquisition, 1-29.

Foucart, A. (2015) Prediction is a question of experience. Linguistic Approaches to Bilingualism 5(4), 465–469. https://doi.org/10.1075/lab.5.4.04fou

Huettig, F., and Mani, N. (2016). Is prediction necessary to understand language? Probably not. Language, Cognition and Neuroscience, 31(1), 19-31. https://doi.org/10.1080/23273798.2015.1072223

Kamide, Y., Altmann, G.T.M., and Haywood, S.L. (2003). Prediction and thematic information in incremental sentence processing: Evidence from anticipatory eye movements. Journal of Memory and Language, 49 (1), 133–156

Lew-Williams, C, Fernald, A (2010) Real-time processing of gender-marked articles by native and non-native Spanish speakers. Journal of Memory and Language 63: 447–64

McManus, K., and Marsden, E. (2017). L1 explicit instruction can improve L2 online and offline performance. Studies in Second Language Acquisition, 39(3), 459-492. DOI: 10.1017/S027226311600022X

McManus, K., and Marsden, E. J. (2017). Online and offline effects of L1 practice in L2 grammar learning: a partial replication. Studies in Second Language Acquisition.

C) Research aims and questions

Focusing on one syntactic phenomenon in one language (such as one type of relative clause, the passive voice, one type of interrogative), this project would investigate

the extent to which L2 learners and native speakers generate syntactic predictions
the extent to which prediction can be taught, as measured on a battery of outcome tests

D) Methods

The study would probably use two groups of learners: highly advanced and intermediate, with approx. 30 in each group, and a native control (n=20).
Experiment 1 will develop and administer a battery of tests to determine whether anticipation is observable online (via eye-tracking and/or self-paced reading) and to measure oral production and explicit acceptability judgments. Experiment 2, adopting an experimental pre, post, delayed post-test design, will randomly assign learners to one of (1) ‘explicit information plus prediction practice’, (2) explicit information plus exposure’, or (3) a test-only group. The type of knowledge would be measured via the same test battery, immediately and 10 weeks after the intervention.

E) Skills and opportunities you could gain

You would work with various experimental environments such as PsychoPy or EPrime, statistical analysis packages (SPSS, R) and experience pre-registering your studies via the OSF in line with cutting edge open science practices. Research assistant opportunities could also be available on the IRIS project.