Résumé:
The large number of users of computers and information technology understand the relationship between computer and language as the computer provides great potential to study, analyze and simplify human languages and also to facilitate their education. But the relationship is not so, it involves a lot of complexities and ambiguities. The natural language processing is a difficult and an incomprehensible expression/term because of the challenges that it poses. The automatic processing of language is primarily for the development, the creation and the formulation of machine translation, whether written or spoken, because translation is the only means by which nations can cope with the explosion of knowledge and information; no society or nation can live independently of this knowledge revolution in front of the hegemony of the English language in the world. The purpose of this research is to know the challenges posed by translation in this field and to what extent the computer succeeded in providing a translation equal to the human translation.
In fact, Automatic translation poses the problem of the universals of language and that of the principles and inherent parameters of internal functioning to each natural language. Two essential facts are to be noted here: the dynamics of the natural metalanguage and the dynamics of the enunciation context. the firsts expresses the abstract microphenomena of the natural language. It is the foundation of any use of language in context. The capture of this metalanguage makes it possible to explain the expressions or effects of meaning. The second is about the spatio-temporal parameters in which Any statement is produced. These parameters are modified at each instance, and this underlies the fact that the computation base of the computer model is constantly obsolete.
Therefore, the translation activity is a truly (socio-) cognitive activity that the machine will have a lot of troubles to integrate it totally. it is under the full domination of Man.
