Ceci est une ancienne révision du document !


Early Latin and late Latin

Michèle FRUYT

University of Paris-Sorbonne (Paris 4)



2. Usual diastratic variations in late Latin

2.1. Written texts vs oral speech

The texts never exactly reflect the spoken language, the fundamental means of communication in any given society. They represent only a small sub-group of the speech instances and they neglect many aspects of the linguistic reality. There are different kinds of relationships between the oral instances produced by the speaker and the written versions of them. Some are written texts containing oral features (fr. écrit oralisé), while, conversely, some texts are oral features adapted to a written text (fr. oral écrit). The first category is documented by Cicero’s letters to Atticus, which, although they are supposed to represent a colloquial conversation between two literate, close friends, do not give an exact image of the spoken language, since they contain some literary elements that would not have existed in a uiua uoce colloquial conversation. The second category is represented by the passages of the freedmen’s speeches in Petronius. The author wrote them in order to reproduce the colloquial language of the illiterate people, but they are probably not the exact reproduction of the oral instances since they are an exaggerated caricature of the freedmen’s language. A third category could be represented by the texts that were dictated to a professional scribe by illiterate soldiers (e. g. some Vindolanda letters and some letters from soldiers in Egypt). These contain formulas (e.g. rogo ‘I ask you’) introduced by the scribe, who re-modelled and re-formulated the speech of the sender of the letter. The litterary turns of phrase to be found in the letters are partly the work of the scribe, who was simply doing what he had been taught to do. However, a few late texts do give us some reliable information about the spoken language of their time, e.g. Egeria’s Itinerarium (end of 4th century AD) and Augustine’s Sermones, the latter being considered as the closest text to the language of the non-literate or illiterate population at the beginning of the 5th century AD in the Roman province of Africa. Augustine called his listeners imperītī and indoctī, writing that they spoke uulgī mōre(Aug. Doct. christ. 4,10,24). It is generally thought that Augustine sometimes, in his sermons, spoke extempore and that some secretaries (tachygraphs) would take notes. His speech would then be written up later, by a secretary or by Augustine himself. While some passages in the Sermones do indeed show oral features, even so, these works remain written adaptations and therefore interpretations of their oral sources.

Linguistic innovations already in use in oral speech are reflected in the texts only some time later. Sometimes early changes are documented in a passage of direct speech. For example, in several passages of direct speech in Egeria’s Itinerarium, iste is already used as a deictic for what the speaker and the addressees see in front of them:

  • Egeria, Itinerarium 14, 2:
    Ecce ista fundamenta in giro colliculo isto quae uidetis, hae sunt de palatio regis Melchisedech’.1) \\‘See these foundations around this small hill that you see before you, they come from the palace of the king Melchisedech’.

This is an evolved linguistic feature, since it has developed as part of the grammar in some Romance languages (Old-French cist < ecce istum). But Egeria never uses iste when speaking in her own name in the narrative passages where she is linguistically more conservative (Fruyt 2009, 2011).

Another interesting and rare indication of linguistic innovation in oral speech comes from Suetonius, who mentions the first occurrence of the verb mandūcāre with the generic meaning ‘eat’ in a sentence written by Augustus in a private, familiar letter :

  • Suet. Aug. 76,1-4:
    1. Cibi … minimi erat. 3. Verba ipsius ex epistulis sunt …4. Et rursus: ‘Ne Iudaeus quidem, mi Tiberi, tam diligenter sabbatis ieiunium seruat quam ego hodie seruaui, qui in balineo demum post horam primam noctis duas buccas manducaui.
    ‘1. He was very sober as far as food is concerned. 3. He wrote himself in a letter … 4. And again: ‘A Jew, my dear Tiberius, does not observe the Sabbath as scrupulously as I observed it today, since it was only in my bath, after the first hour of the night, that I ate two mouthfuls’.’

This generic meaning ‘eat’ for mandūcāre was established only much later in the standard language of the texts. But this occurrence in Suetonius shows us that, since the speaker is the emperor himself, this evolved meaning was already in use at the end of the first century BC and the beginning of the first century AD in the spoken language of literate individuals in the political spheres of high Roman society. We see here a semantic change from a specific to a generic meaning, since the specific feature and original meaning of the verb mandūcāre was ‘eat in great quantities’, while the context shows here that Augustus was relating the fact that he ate in a very small quantity (see infra § 3.1.).

2.2. Register variations

In late Antiquity, Latin was submitted to the usual diastratic variations of register. One and the same author may use several registers according to the speech situation and the identity of the addressees. Augustine’s works can be divided into three diastratic levels: the highest level is illustrated in certain treatises e.g. the De ciuitate Dei, the medium level in the Confessiones, and the lowest in the Sermones. There are, indeed, important differences between the language of the De ciuitate Dei and that of the Sermones, as shown by the use and frequency of is, hic, iste, ille, ipse in their various functions as endophors (anaphors and cataphors), correlatives of relative pronouns and deictics. The frequency of is decreases from the De ciuitate to the Sermones where is survived only in the correlative function (André and Fruyt 2012).

Retour au §1 ou Retour au plan ou Aller au §3

1) cf. 14, 3; 15, 1; 13, 4.