A couple of years ago Colette Bodelot (2014, Les propositions complétives dans la Chronique originale de Frédégaire (I.4, chap. 1–90). In Piera Molinelli, Pierluigi Cuzzolin & Chiara Fedriani (eds.), Latin vulgaire – Latin tardif, Vol.II , 183–203. Bergamo: Bergamo University Press, Sestante Ed.) discussed in a rich and worthy paper the use of direct and indirect speech by Fredegarius, one of the most representative Merovingian storytellers, and showed that he rather used the subjunctive instead of the Accusativus cum Infinitivo (AcI). Therefore, I decided to take again into account this author and his historical work which on the other hand had been object of a keen inquiry by Lyliane Sznajder (2005, Stratégies de prises en charge énonciatives dans le discours indirect. In G. Calboli (ed.), Papers on grammar IX, 2, Latina Lingua , 749–761. Roma: Herder.), and I started from these two papers. I myself had considered the direct and indirect discourse in Latin and Indo-European languages (Calboli, in print), however in that paper the core of my inquiry was rather the AcI as a peculiar construction of the indirect speech. In order to choose a specific text with all stylistic implications, I concentrate myself on the clash between the king Theuderich and saint Columbanus, where I could compare the Vita Columbani by Jonas and Fredegarius’ Chronicle . I could therefore take into account also a kind of epic style proper of the Histories of Saints, which suggested the use of AcI, a typical construction of the most authoritative Latin. In this case Fredegarius’ text was a reproduction of Jonas’ text, but with some differences in the use of subordinate clause: Jonas employed AcI, Fredegarius the simple subjunctive. I took into account also the use of Gregor of Tour, and pointed out a fluctuation, in Banniard’s (2012, Le latin classique existe-t-it. In Biville Frédérique, Marie-Karine Lhommé & Daniel Vallat (eds.), Latin vulgaire – Latin tardif IX , 57–78. Lyon: Maison de l’Orient et de la Méditerannée) sense, which produced in Merovingian Latin a larger frame of constructions than in classical and imperial Latin. This depended also upon the abandon of some constraints which in classical and postclassical Latin reduced the number of possible clauses. The following passage was the reduction of such a frame in Carolingian Latin. In previous Merovingian language, AcI was consistently challenged by subjunctive, both introduced by conjunctions of subordination ( quod, quia, quoniam , etc.) or without any conjunction in a kind of simple subjunctive. This phenomenon was connected with the expansion of subjunctive, in particular of pluperfect, which was extended in most Romance languages (cf. Stotz 1998: 333, Handbuch zur lateinischen Sprache des Mittelalters . Vierter Band. Formenlehre, Syntax und Stilistik . München: C. H. Beck), and in Merovingian Latin was employed also instead of indicative (cf. Vielliard 1927: 224, Le latin des diplômes royaux et chartes privées de l’époque mérovingienne . Paris: Honoré Champion). Therefore, I highlighted that these uses in Merovingian Latin have to be considered variations and enlargements of the linguistic frame rather than mistakes as they would be considered from the point of view of classical Latin, though they were the product of a decomposition of Latin, in particular of the cases system.