Banniard, Michel
▶ Watch on YouTubeCoverage note: full - automatic transcript machine-translated into English from a French lecture. Latin quotations and many proper names are garbled.
Core ideas
Michel Banniard opens with Caesarius of Arles' peasant who has no time to learn his prayers, and calls himself a rusticus who prefers paper handouts to PowerPoint. He then sets out his framework for Poitou, Aunis and Saintonge, a historic border zone between oïl and oc. He rejects the 19th-century model of two Latins, a respectable literary Latin and a "Vulgar Latin" that supposedly corrupted itself into the Romance languages. He also rejects the image of ageing languages and golden, silver and iron ages. Latin, he argues, was a spoken language for everyone, Latinophonie on the model of Francophonie, with all the variations described by Coseriu: regional accents, social levels, registers and individual choices. Classical Latin itself already contained small fluctuations that would later grow into Romance. Latinisation was consolidated by Christianisation. In the 6th-7th centuries the Latin type became the Romance type and split into languages and dialects. He places the shift from a "more Latin" to a "more Romance" spoken language between 650 and 750 and calls the 8th century "proto-Romance", meaning early Romance, not a primitive one. He also dismisses the "apocalyptic" view of the barbarian invasions, which he says French historians projected back from the defeat of 1940.
Until the 8th century, written Latin remained in real contact with speech, and the documents show this. Fortunatus' life of Radegund is in learned Latin, but the nun Baudonivia's version is in sermo rusticus, language for general communication meant to be read aloud on feast days. Its sentences about the aristocratic saint sweeping, cleaning and drawing water are "as much Latin as Romance" once the spelling is set aside. From the 9th century, documents become "Latinised Romance". In the Polyptych of Irminon, abet ibi ecclesiam bene constructam is Old French in Latin spelling. In the dialogues of the Count of Vendôme agreement of around the year 1000, modo mi tolit meum fiscum reflects Romance phrasing, and Banniard reads this as a direct draft rather than a lost chanson de geste. Merovingian inscriptions, including Abbot Mellebaude's in the Hypogeum of the Dunes and epitaphs collected by Cécile Treffort, show spellings to be read for the speech behind them. Inconsistent spelling does not mean inconsistent speech, any more than modern children's spelling mistakes mean they cannot speak French. After the Carolingian period, inscriptions return to a "restored" Latin that says more about culture than about language. Banniard adds that some very old French syntax, as in Raoul de Cambrai, is more elaborate than late Latin (romanitas maior).
On Poitou, he supports Jean Pignon's thesis that the speech of the region up to north of Poitiers was originally Occitan of a Limousin type, later "re-pronounced" with oïl phonetics while keeping a southern lexicon and morphology, as dialect atlases show with words such as fin (hay) and fissou. This would explain why William IX of Aquitaine, a trilingual aristocrat immersed in Limousin speech and in the tropes and sequences of Saint-Martial, wrote in Occitan. In the discussion Banniard insists that regional variation does not stop Latin being Latin, citing Martin of Braga and Martin of Tours, who preached far from their native Pannonia. He prefers a non-linear model of change, inspired by Stephen Jay Gould: small sporadic phenomena that suddenly accelerate, producing a break between 650 and 750. He holds that by the year 1000 ordinary peasants from the Loire and the Garonne could not understand each other, and that only the aristocracy was polyglot. He also points to da in Italian charters as early as the 7th century, well before the usual date of 900.
Excerpts
Michel Banniard on the myth of corruption:
“There were two Latins: the literary Latin that can be taught and respected, and the vulgar Latin that takes every form like a linguistic vampire. And it is from this vampire that the Romance languages were born. [...] It is a traditional 19 th-century teaching that persists, which teaches that the history of languages is formed, in this old 19 th-century view, exclusively by, in fact, corruption.”
Michel Banniard on Latin spelling over Romance speech in the Polyptych of Irminon:
“it has sentences like " Abet ibi ecclesiam bene constructam." That is anything but Latin; it means " he has a well-built church there." You immediately recognize Old French, right ? Well, I recognize it immediately. There you go. Simply put, they know the spelling, so they re-Latinize it. But the phrasing and the morphology are pure Old French.”
Michel Banniard on spelling and speech:
“what you must do instead of letting yourself be frightened by the apparent orthographic inconsistency is to reconstruct the actual orality behind it, and the actual orality is perfectly coherent. [...] just because people don't master spelling well doesn't mean they speak incoherently or speak a primitive language.”