The question of from where the text speaks is related to another: in which language? The language of the Devisement has been problematic in its reception from the outset, though this has been frequently occluded both by early transmitters and by modern scholars. One illustration of uncertainty concerning language comes early in the Z redaction. At the beginning of Chapter 4, we read: ‘In Iorgia est quidam rex qui David Melic totis temporibus nuncupate, quod in lingua Galica dicitur Rex David’ (4, 1: ‘In Georgia their king is always called David Melic, which is to say in French King David’). Here an early Latin translator of a lost version of the Franco-Italian redaction, himself thought to be a Venetian, glosses the Georgian name ‘David Melic’ (derived from the Arabic malik, ‘king’) by telling us that ‘in French’ (even though he is writing in Latin) this means ‘Rex David’. We are almost certainly dealing here with an instance of the Z redactor translating his source unthinkingly, which is characteristic mainly of just the early part of his work: he seems to have quickly become more attentive to such details and to have sought to make his own version increasingly more internally coherent. But the slip is nonetheless symptomatic of the Devisement's early linguistic mobility: it seems to move so easily between languages that, when translated, it invariably retains traces of the language of the source text. Thus, as we will see, the French redaction retains Italianisms from its Franco-Italian source, while the Tuscan redaction retains Gallicisms.