When one compares the AR 1912 and the AM 1934, one is struck by the number of si flats that have been changed into si natural. In studying the charts compiled under the direction of D. Gajard, one can conclude that it is he who, at the last moment, i.e., on the galley proofs of the AM 1934, crossed out by hand numerous si flats to change them into si natural. Why did he do that? There are several reasons.

The first stems from an erroneous interpretation of the manuscripts from Italy, especially those of Benevento. These manuscripts do not distinguish between si flat and si natural. They do not recognise the flat sign; they write si with any other indication and entrust the rest to the memory of the singers. As these are also very reliable manuscripts for the teneurs si and mi of the 3rd and 4th modes [J. GAJARD, “Les récitations modales des 3e et 4e modes dans les manuscrits bénéventains et aquitains”, Etudes grégoriennes 1 (1954), p. 9-46.], D. Gajard believed that the absence of the flat sign was for them a kind of evidence in favour of si natural.

The second seems to correspond to a taste of the period: to maintain the original character of medieval chant, certain people tried to give it an exotic, extraordinary flavour. But the comparison of the medieval manuscripts shows that it is often the Vatican edition owed to D. Pothier (1912) that has the right of it on this point, in accord with the monastic tradition (Saint-Denis, Saint-Maur des Fossés, Metz), a part of the Germanic tradition (Saint-Georges de Willingen, Aix-la-Chapelle, Utrecht), and the Solesmes antiphonaries of D. Guéranger until 1934. Thus numerous antiphons of the new antiphonary bring back the traditional si flat, especially in the 1st and 4th modes, but also, sometimes in a little less notable way, in the 8th mode and even by way of exception in some antiphons in the 3rd mode, where it contrasts with the structural si natural.