Considered as a 'problem solving activity' (Guilford 1950), creativity is part of the translator's everyday life. Many translators, however, are insecure about their own creativity. They are victims of their erroneous, preconceived ideas about the concept of 'fidelity' in translation. Our experimental groups show two kinds of reactions to that insecurity. The 'semi-professional' learners tend to go back to a 'playing it safe' strategy at the first sign of criticism. Translators with a certain position in society and with a certain 'reputation' as authors have the necessary courage to stick to their creative problem solving, but with the awkward feeling of a betrayal which they heroically feel obliged to lay claim to, though they try with numerous explanations to show that it is not a betrayal. Their 'mechanisms of justification' are partly false and can partly be integrated into a theory of translational activity which is mainly based on the results of cognitive research and which is liable to give the translator more security and courage with regard to his creativity.