The case of astele is used in this article to demonstrate how Old French organized its semantic fields differently from the modern language, with more synonymy and polysemy. Consequently, dictionaries of Old French need to take account of this situation when compiling their entries if they are to reflect accurately the semantic organization of the language rather than that of modern French. This raises the question of the use to be made of the thousands of multilingual glosses which, in line with this medieval linguistic attitude, are liable to be dismissed as vague or inaccurate and so excluded from dictionaries.