ABSTRACT
Recent developments in speech technology have enabled a human-computer interaction in which the user can converse directly with a machine for the purpose of transferring information. For example, one can make airline reservations and check on weather and traffic conditions. With a leap of the imagination, one can picture an intelligent tutor that can work with a student in a learning situation, overcoming issues of distance as well as affording tantalizing economies of scale. Such a system might seem ideal for language learning, given its ability to recognize and parse the user’s speech. Indeed, the chapters in this volume illustrate the variety of systems, from reading coaches to second language tutors, that have adapted speech technology for language learning. These chapters assume that speech technology to automate selected teaching processes is desirable, and some chapters (e.g., Aist & Mostow; Mostow et al.) put forth evaluation results that suggest the effectiveness of this automation. However, to address comprehensively the implications of speech technology for computer-assisted language learning (CALL), it is useful first to examine the minimum requirements for learning in a computer-assisted context, then to look at what remains to be done to make such a capability an instructional reality. We begin by observing common pedagogical limitations of CALL, then introduce a framework for CALL that reflects contemporary theories of second language acquisition, and finally examine how speech-interactive CALL systems, in particular the ones in this volume, relate to this framework.
