Speech Synthesis for Language Tutoring Systems

In this paper we shall be concerned with the use of Text-to-Speech Synthesis, or TTS for short, as a tool for Language Learning. We shall present a number of applications where TTS plays a fundamental role in helping the student in Second Language learning. TTS can be a fundamental tool in helping to recognize and get aware of the contrastive features that constitute the main learning targets of the student. TTS can also be used simply as a speaking Tutor when help is needed in any self-instructional system or just to provide feedback on some exercise the student is practicing. It can be used as a Reader for Dictation exercises where there is a need to vary voice quality and speaking rate. Eventually, it can be used to help students working on a Listening Comprehension task in giving hints on what the main task to be accomplished consists of, and other similar Oral drills. We shall be presenting all these examples of the use of TTS in a CALL without always assuming that it is the only way to cope with oral linguistic practice. In general, having a human tutor to do the same kind of tutoring activity guarantees a much better result: the question is whether a human tutor may always be available when the student needs one, which is usually not the case. So the possibility to have a substitute, for how much of lesser quality it may be, is worth pursuing. And there is at least one case in which the computer-based speaking tutor constitutes the only viable alternative to the human tutor: when mimicking the levels of speaking proficiency in L2, or levels of interlanguage, as will be explained in detail further on.