The author of this article sees that, in 1979, more and more monographs about Translation Studies and analytical papers on translation emphasise the definition of the concept ‘translation’. This tendency in modern research finds its origin in the attempts of both translation theory and translation criticism to provide the translation practice with some clear guidelines and to fix the norms of this text-creating activity. Beside an undeniable cognitive value, this normative approach also has a negative side-effect: for the study of the phenomena involved, preference is given to static definitions instead of dynamic models. In this article the author reflects on this situation and provides some thoughts on both static and dynamic approach.