Abstract
A CRITICAL EVALUATION OF THE ANALYTICAL RESEARCH OF SPEED Here S.B. attempts the critical analysis of the research which she led two years, ago with a view to comparing the language of different proposals put forward at the Metz congress of the socialist party (1979). It is an opportunity for her to refer back to the two methods that she used: the automatic analysis of language and lexicometry. Taking the way they work step by step she brings out their weakness, detects certain arbitrary aspects and notes the moments when the interference of the research worker deflects a process which is on the surface rigorous. She also shows how the conditions of application and efficiency belonging to each method have led her to modify the corpus submitted for analysis, and the difficulty that that has involved in comparing their results. S.B. tries to extend her conclusions into the discourse analysis general. Her study also deals with crucial questions such as the formulation of hypothesis, the constitution of corpus and the use of several conjoined methods.