TWIRLL: Targeting Wordforms in Russian Language Learning

TWIRLL: Targeting Wordforms in Russian Language Learning

TWIRLL’s main objective is to correct the non-strategic traditional approach in Russian language learning, which focuses on memorizing paradigms of wordforms. A typical noun has 12 wordforms, and a verb can have over 100 wordforms. However, most forms of most words are rarely or never encountered. A computational learning simulation experiment (Janda & Tyers 2018) shows that mastering Russian wordforms is much more successful when only single high-frequency forms are learned. Learning whole paradigms gives worse results because it overpopulates the search domain and overwhelms students.

TWIRLL’s target groups are learners and instructors of Russian at the A1-B2 CEFR levels (elementary and intermediate).

TWIRLL’s main outcome is the Strategic Mastery of Russian Tool (SMARTool) that puts learners and instructors in contact with all and only the wordforms that they need most, along with the grammatical and lexical environments in which those wordforms are most likely to appear. Other outputs include: extension and new interface for the Russian Learner’s Constructicon, revised courses and curriculum, student and staff mobility, and dissemination of SMARTool through conferences, promotional video, and social media.

The Arctic University of Norway and The Higher School of Economics collaborate on TWIRLL’s objectives through a coordinated program of mobility, student involvement in scientific work, and shared teaching and supervision of students, bracketed by a kick-off workshop and a dissemination closing conference.

The Principal Investigator of TWIRLL is Laura A. Janda, who is a member of the CLEAR (Cognitive Linguistics: Empirical Approaches to Russian) research group at UiT The Arctic University of Norway). TWIRLL is funded by a grant (CPRU-2017/10027) from the Direktoratet for internasjonalisering og kvalitetsutvikling av høyere utdanning (Diku: https://diku.no/). Diku has selected TWIRLL as an exemplary project, see this site.