ABSTRACT
Advances in the coding of messages in the Second World War and the application of computers to areas other than mathematics led to the notion that translation could be done ‘automatically’. Early computing research workers who were not linguists tended to think that translation could be effected by substituting a word in one language for its lexical equivalent in another and, while morphology was addressed, syntax and semantics were hardly considered; the frequently hilarious results have been well documented (Knowles 1979, inter alia). Two more or less parallel developments to produce computerized aids for translators began to take shape: the creation of machine translation systems in the late 1940s and the building of computerized dictionaries in the form of multilingual terminology data banks (term banks) in the early 1960s. Initially these developments took place independently but in recent years there has been a merging of the two forms, in which the terminographical information in a term bank may be incorporated into machine translation systems in the form of ‘back-up’ dictionaries. Latterly, the information has been used to provide the data for expert systems which may help, for example, with the semantic problems in machine translation.
