Die simbiose tussen masjienvertaling en rekenaargesteunde taalonderrig met betrekking tot die leksikon

Abstract

The form and function of vocabulary or dictionaries (lexicons) in the fields of second language learning, translation, computer-assisted language instruction (CALl) and machine translation (MT) are investigated. The nature of the thesis is theoretical. The hypothesis underlying the study is that MT and CALl have both developed ways of representing or utilising lexicons that can be used to mutual advantage. The primary aim is not to search for or describe such cases of mutual interaction. Preliminary research has shown that there is not much interdisciplinary interaction between MT and CALl, so that substantive cases of this kind of interaction would be unlikely. The primary aim is to investigate the possibility of such a symbiosis and to examine areas where useful interaction could occur. The study shows that vocabulary learning is not as well researched as other areas of language learning. In the field of CALl a few interesting specimen programs were found, but the general lack of depth with respect to vocabulary learning programs is perhaps a reflection of the neglect of this aspect of language learning. The field that makes the most extensive use of lexical information is MT. Formalisation of lexical knowledge utilisation is not on the same level in translation theory, and this is a shortcoming of the trade that resulted in MT not having too much collaboration with it. Translation theory does, however, contain a mass of very specific lexical knowledge. This would ultimately have to be incorporated into MT systems - whether there is a model behind it or not. Several similarities are found between second language learning theory and translation, and opportunities for mutually advantageous co-operation exists. Among the ideas of CALl by which MT can benefit, is the incorporation of knowledge about • relationships between translation equivalents and • relationships between word meanings and the syntactic contexts of their forms in the vocabulary learning process. Beneficial to CALl can be the classification of lexical material in MT • on the basis of influence on syntactic environment and • according to how they fit in with the levels of learnability - as postulated by influential language learning theories.