HLT4LL: Mathias Schulze

Bio
Mathias Schulze is an Associate Professor of German at the University of Waterloo in Canada. He was born in Finsterwalde in Brandenburg (Germany) and did his teacher training for German and Russian in Leipzig (Germany) and Kaluga (Russia). For ten years, Mat worked at universities in Sunderland and Manchester (England) before coming to Waterloo. Currently he is the director of the Waterloo Centre for German Studies, editor of the CALICO Journal (on computer-assisted language learning), and the editor of the H-Net list on Transnational German Studies. His research in computer-assisted language learning focuses on the acquisition of foreign language grammar and the development of proficiency. He has worked on Intelligent CALL—the application of the artificial intelligence techniques natural language processing and user modeling to CALL—and on tutorial CALL as well as on online language learning. Currently he analyzes the complex dynamic system of individual proficiency development in online language courses.

Human Language Technology and Tutorial CALL—The Odd Couple?
Tutorial CALL—the interaction of language learners with the computer—has changed from a mainstay in computer-assisted language learning to a fringe aspect of research and development in the last two decades. This is in stark contrast to a renewed emphasis on instructed language learning with practice and awareness-raising activities in both applied linguistics research and practical language teaching methodology. To improve the quality of the learner computer interaction, its central components—corrective and preemptive feedback, contextualization and tailoring, and (metalinguistic) text augmentation and feature-based retrieval—need to be enhanced. I will argue in this context that human language technologies have the—as yet largely unrealized—potential to help ameliorate these core components of tutorial CALL. In the course of my presentation, this argument will be exemplified in two ways: (1) by discussing broader language-learning activities and tasks which can be facilitated for learners and instructors through the use of these technologies, and (2) by reviewing technologies that are suitably robust for useful employment in language learning contexts. Issues that have all too often hindered the utilization of human language technology in tutorial CALL will also be addressed. These issues will have to be resolved if tutorial CALL and human language technology are to become an odd couple, in which both sides benefit from one another. I will attempt to show that the existing and potential benefits are mutual and many. Throughout this presentation, I will focus on activities and technologies in which written text production is central.