If determining a language student's entry level is difficult for traditional campus-based institutions, it is all the more so in the case of distance-based post-secondary operations where students are often older and have interrupted their studies at some point. Not only does one have to address the issue of the faux debutants, that is of the many students who claim to be real beginners when in fact they are not, but, more importantly, one is also obliged to ensure that returning students possess in practice (not just on paper) the prerequisite skills on which a course may be built. While some would argue that the locus of responsibility does not necessarily lie with the institution, it would be foolhardy to suggest that it rests entirely with the student. In campus-based institutions, students have access to their peers and to faculty advisors when faced with decisions of this kind. Some students at a distance may be fortunate enough to have a student advisor or counsellor, but rarely will this individual be in a position to advise students adequately in this regard. Computer-assisted evaluation, on the the other hand, could serve as a primary, if not the sole, medium for determining whether or not a student possesses the required preparation for a specific second language course.

In fact, Wyatt (1984) has argued in favor of the development of computer-adaptive testing for placement tests, in part because these would be more humane and user friendly than non-computerized tests, since the former could branch students according to their level and thereby greatly reduce the number of questions which are beyond their competence. In a distance education setting, one can envisage developing one placement test which, by the appropriate use of branching, could recommend which of several courses a particular student would be adequately prepared for.

Computerizing challenge examinations and summative examinations within a course (normally final examinations) allows not only for the individualization of the testing material through data banking, but also for the testing in a removed location, of audio comprehension and speech generation. If this proves to be a time-consuming task in traditional learning environments, the distant educator is further hampered by the distance between the learner and the teacher. In this instance, different configurations of computer- controlled cassette recorders, providing audio comprehension or incorporating speech production via voice- activated recording capability, would prove to be a viable alternative to current practice.

Callapplications

Discussion of the relative worth of the various software developed during the past 10 to 15 years for second language learners has been heavily influenced by different theories of second language teaching. At the heart of the question is Krashen's (1982) distinction between "learning" and "acquisition" and the resulting emphasis which has been placed on the communicative approach to language learning by both supporters and detractors of Krashen's thesis. Just as the audio- lingual methods of the 1960s and 1970s (symbolized in many ways by the use that language laboratories were put to) have had difficulty weathering the storm, so to have the behaviorist drills that have been synonymous with CALL during the past 15 years. Underwood (1984) is one of several who believe that the value of the computer as a learning aid in language acquisition lies in the use of creative communicative software (games and simulations) rather than in the use of "wrong-try-again" drills. Not only does he consider the latter to be unimaginative and boring, but he is also concerned that they emphasize form rather than content and in so doing replicate many of the shortcomings that one finds in most classrooms: the "classic" form of CALL recreates precisely those features of the classroom we are trying to avoid: it is teacher- (ie. computer-) controlled, evaluative and highly structured. (p. 49)