Data-Driven Learning: Can and Should Language Learners Become Corpus Linguists?

Friday, May 5, 2017 4:00 pm - 4:00 pm EDT (GMT -04:00)

ALERT: DUE TO FLIGHT CANCELLATION, THIS TALK HAS BEEN CANCELLED. IF YOU WERE HOPING TO COME, YOU CAN READ UP ON NINA VYATKINA'S RESEARCH ONLINE. WE KNOW; IT'S NOT THE SAME THING.

Corpora, or large digital textual databases, have been attracting the attention of language educators since their emergence in the 1960s. Materials developers have used corpus information to create reference grammars, dictionaries, and textbooks that more adequately represent the target language than artificial examples. More recently, applied corpus linguists started using corpora in their language classes in a more direct fashion: while either creating corpus-based worksheets for classroom use or teaching their students to search corpora in order to complete learning tasks. Nevertheless, such more direct corpus-based applications, a.k.a. Data-Driven Learning (DDL), are still far from common pedagogical practice. One of the obstacles for a wider spread of DDL is its discovery-based nature that is akin to hypothesis-building and hypothesis-testing in academic research, which is deemed too challenging for language learners, especially at lower second language (L2) proficiency levels. In this talk, I address this challenge and ask the question: can and should we expect language learners to become researchers, or corpus linguists?

I will answer this question using empirical data from my own classroom-based DDL studies conducted with American university students at different proficiency levels in their L2, German. I will begin by outlining specific theoretical and pedagogical principles undergirding DDL and a brief overview of available DDL research. Then, I will demonstrate DDL activities that I used in my language classes with special attention to how I integrated these activities with broader course goals and other, non-DDL course components. I will then present my analysis of some quantitative and qualitative learner performance and perception data. I will argue that DDL was overall very successful and will discuss how the design of my pedagogical interventions helped overcome known limitations of this teaching method and how DDL effects were modulated by various contextual factors. I will conclude by arguing in favor of a wider application of DDL in language teaching and more empirical research in the area.

About the Speaker

Head shot of guest speaker, Nina Vyatkina of the University of Kansas
Nina Vyatkina is an Associate Professor of German and Applied Linguistics at the University of Kansas. Currently, she is an investigator on a learner corpus research project that involves international collaboration with Humboldt University in Berlin, Germany. She is a co-recipient of the 2009 Paul Pimsleur Award for Research in Foreign Language Education (American Council on the Teaching of Foreign Languages / The Modern Language Journal). Read an interview with Dr. Vyatkina on digital humanities research.