Les traductions de la Bible aux XIIe-XIIIe siècles, videlicet Biblia Francorum (Les psautiers)

Starting with a short analysis of the still unclarified relations between heresy, art and French medieval translations of the Bible, and taking into account that one does not have the means to write about all the biblical translations of the first two centuries of French literature, the present PhD thesis resorted to the only remaining possibility, that of addressing the corpus through a synecdoche. It delves into the first vernacular Romance psalters: the corpus of psalm translations, paraphrases, and commentaries into French. There were countless reasons supporting this choice: the book of psalms is by far the most important of the Middle Ages. It was, on the one hand, the most read and commented part of the Bible at the time. On the other hand, it had certain autonomy. The psalter was a separate book, with its calendar, hymns, prayers and litanies, and it was often a luxurious object, not only because of its richly ornamented initials, but mainly because of the Christological cycle inserted between the calendar and the rest of the book. As for the links with heresy (or heterodoxy), one may bear in mind that the Chanson de Saint Alexis, the text which impressed Valdes, was copied into a psalter, the St Albans Psalter. Valdes followers’ presented a French psalter with learned gloss at Lateran. Those of Metz used another French psalter. And Lambert the Stutterer had apparently used a commentary of the psalter in vulgar French. It is therefore easy to grasp the relevance of a study of the vernacular works inspired by the psalms. Historical sources are the first to highlight it, but there is also the testimony of libraries. The frequent appearance of the Psalteria in Gallico among the books of the English monasteries in the 12th-13th centuries and their presence at the head of the lists compiled in the following century are the best signs of the success that these translations have known. Half of the French manuscripts of the twelfth century are of insular (Anglo-Norman) origin, of which another half - that is to say a quarter of the corpus - are psalters. One has the impression that everything revolves around the psalters, that they constitute a pars pro toto. This is why the thesis starts from the link between psalmic texts and art, and ends with their links to heresy / heterodoxy. The two ends are linked in a loop, in order to delineate the initial stakes of the translatio. However, the research never intended to produce a synthesis. Given that one is dealing with specific cases, the demonstration has been commonly diluted into the analysis of details serving to prove precise points. At the centre of the investigation are the interactions with the Medieval Latin sources, sometimes unidentified. The precise nature of these relations can never be evaluated on the basis of the hypotheses formulated by previous researches. The author avoided basing his research on previous interpretations. He deliberately chose not to analyze each text according to the amount of its bibliographic entries. Some works should or could have occupied a disproportionate place (the Eructavit poem), while others, with little bibliography (psalm commentaries), or none (the first verse translation of the psalms) could have simply been neglected. To give each text its rightful place in the economy of research, such was the choice to be made. The slow movement from one detail to another, from prose translations to verse adaptations and later on to prose commentaries, conjured up a new outlook on these psalmic texts. One does not need to insist upon the division of the corpus into three categories (prose, verse and commentaries). In this case, the three categories provided certain stability. They were structured around three main literary works (the Oxford Psalter, the Eructavit, and the first French commentary of the Psalms), with elements swinging occasionally between two criteria (such as the two verse translations). The slow movement focused upon text and manuscript details provided a series of results. The Oxford Psalter may indeed be an autograph translation, as recently proposed, and its source may very well be the St Albans psalter, but for a couple of contradictory variae lectiones needing further analysis. The double psalters (bilingual, with two columns) are secondary developments of its manuscript tradition, but they also present a long series of peculiarities. The Copenhagen Psalter remains a mystery, because of its palimpsest nature, as well as the Oxford fragment. Nonetheless, all other spin-offs of the Oxford Psalter include various interesting small texts. The French last word of the Latin litany in the Additional 35283 of the British Library makes one think that the main language was French, as well as the titles accompanying the images in the Winchester Psalter. Among the interlinear glosses, the Arundel Psalter shows an oscillation between the double columns disposition and the subsequent passage to the superscript gloss. The mirrored presence of Old English translations in the Eadwine Psalter also points towards possible links with previous insular translations. All these details and many more provided the basis for a comparison with Old English, Medieval Latin, and Oriental psalters. The layout of the French psalters could in fact be the result of many influences. The short analysis of the first verse translation also clarified certain aspects: the base manuscript is the Harley 4070 of the British Library, written by two scribes. Its manuscript tradition was certainly greater, since a poem translating the Athanasian Creed, written in the same rhyme and meter, was copied at the end of the second verse translation of the psalms, the one composed on the Continent. Next, the Miserele and the Eructavit poems drew both of their contents from a couple of learned commentaries, in Latin. The Eructavit also played the role of an entire psalter in certain manuscripts, such as the Additional 15606 of the British Library, where a poem dealing with the siege of Jerusalem by Nebuchadnezzar has been linked to it, forming a single structure, namely De David li prophecie. But it is only in the analysis of the French psalm commentaries that one may find the real links with the Medieval Latin tradition. The first commentary, probably written for Laurette of Alsace, was assembled in three of four stages, by three or four authors. Its source (or sources) was/were a series of learned commentaries developed from the Media Glossatura of Gilbert of Poitiers. The second French commentary constitutes a spinoff of the first one, and the third commentary, completely independent from the first two, is a word for word translation of the Magna Glossatura of Peter Lombard. All of these conclusions needed however further proof. And since the huge manuscript tradition of the learned Latin commentaries on the psalms dating from the 11th-13th centuries is impossible to grasp, one had to make a second compromise by adding a corpus de contrôle: the texts inspired by the Book of Proverbs, less numerous but belonging to the same category of biblical works: the poetic books. The analysis of these other translations, the verse version of Samson de Nanteuil in particular, validated the conclusions of the first three chapters: Samson did indeed use a Latin learned commentary, but not the Bede gloss, as previously stated. He used a spinoff of this gloss, such as the one identified in the Burney 17 manuscript of the British Library. Next, one had to explore the role played by the Bible translations in the creation of the French language. The traces of these pillars of medieval culture would be found in almost all of the Old French literature, and yet the biblical texts have not been studied systematically. The role of these letters in the creation of a new literary language remains little known. Thus the end-chapter of the present PhD dealt with a vast unknown: the contribution of translation in the creation of the language. Results have been deceiving nonetheless, because what one may find is not the influence of translation, but the very preexistence of oral practices before any of the studied texts were even written down. It was because bilingual and diglossic situations created four linguistic parameters instead of two, and the translation from Latin into French always passed through a low prestige variety of Latin. It was a safe zone for the translator