2.5. Alphabetical Latin-French Lexica

"Les dictionnaires monolingues des langues occidentales qui nous sont aujourd'hui si familiers ne sont pas issus par simple mutation, traduction ou adaptation, des dictionnaires latins antérieurs. C'est au contraire par une lente évolution des versions bilingues plus anciennes que se sont dégagées les formes essentielles de nos modernes répertoires."

2.5.1. Abavus series

A sixth member of the group, which was unknown to Roques, is a fragment of a Latin-French glossary found in a Latin-German glossary, ms. Trèves, BM n° 1125 (fourteenth century), described by Holmér (1964). The Latin-French portion of the manuscript, ff.1r-2r, consists of 279 articles beginning with the letter 'A'. Similarities which this fragment shares with the Abavus manuscripts edited by Roques (1936b) led Holmér to conclude that it is also part of the series. Holmér (1964: 86) believes that the Trèves fragment (Ab6) is closely related to Ab1. However, through a comparison of the first thirty lemmata from Ab6 with the same quantity from Ab1 and from Ab3, Lindemann (1994: 148) illustrates convincingly that Ab6 is, in fact, related to Ab3.

Roques (1936b: xxiii f.) also describes another manuscript, Paris, BN lat. 4120, dated 1352, which contains on ff.122-125 a short glossary beginning with the word absconsa. He is uncertain whether this glossary is part of the Abavus group and resolves to publish it separately. Merrilees (1994c: 14) concludes, from a sampling of letters 'A', 'B' and 'C' from this text, that it is probably not a version of Abavus. It contains many lemmata not found in the Abavus family and it appears to have sources in addition to Papias and Balbi. He also notes that alphabetical order is followed only to the first letter and that the whole glossary seems to be composed of blocks of words which may have an internal alphabetical order, yet do not show evident thematic or textual relationships.

Roques (1936b: xxv) observes that Ab1 and Ab2 are obviously linked by their similarity of inventory, but their differences are such that it is unlikely one derived from the other. He also considers Ab4 and Ab5 to be intimately related and probably derived from the same model (1936b: xxiii). He published Ab4 in its entirety, but published only the variants of Ab5, deeming that it would be of little interest to print both of them (1936b: 239).

Roques (1936b: xxxiif.) thinks that the source of the two oldest Abavus manuscripts (Ab1 and Ab2), as well as the Absconsa fragment, is a Latin-French glossary based on a Latin lexicon deriving from Papias; that Ab3 is a revision of this first stage; and that Ab4 and Ab5 represent a later revision of the second stage.

Interpretamenta are mainly one-word equivalents in the two oldest manuscripts, Ab1 and Ab2. Paraphrases or additional synonyms are rare, as are indicators of metalanguage. Lindemann (1994: 137) confirms that an "overwhelming majority" of the lemmata in Ab1 derive from Papias, as the following extract shows:

The two youngest Abavus manuscripts, Ab4 and Ab5, both of which date from the second half of the fourteenth century, are very different from the Douai-Évreux tradition. Their nomenclature is much more extensive and their structure is also more complex: multiple equivalents are common, glosses sometimes include both Latin and French definitions, there is a rudimentary Latin metalanguage, and each contains approximately 175 versus memoriales. Those versus which Roques was able to identify (1936b: xxxii) were drawn from Eberhard of Bethune.

Ab3 is smaller (5856 articles) than Ab4 and Ab5, both of which number more than 9000 articles, but its structure resembles theirs. The most important characteristic shared by the three later manuscripts is their obvious link with the Catholicon of John Balbi (Lindemann 1994: 145). Balbi's influence is evident in this sample:

Some derivations, i.e. abavus, aberrare, abiectio, have been given the status of lemmata and placed in alphabetical order. They do not appear as lemmata in the Catholicon, but occur respectively as sublemmata of Avus, Erro, and Abiectus (id. 1994: 146).

2.5.2. Brito 2: ms. Montpellier, Faculté de Médecine H236

Such appears to have been the case with ms. Montpellier, Faculté de Médecine H236, of which 103 out of 127 folios contain a double glossary, Latin-Picard and Picard-Latin, consisting of 4825 entries (Grondeux 1994: 41).

The anonymous compiler of the alphabetical Latin-Picard glossary was probably a schoolmaster living in northern France, specifically in Artois, in the first half of the fourteenth century. His principal source (id. 1994: 41; Hauréau 1885: 594) was the Expositiones vocabulorum biblie, compiled by William Brito between c.1248 and c.1267 (see 1.3.4.), and his secondary source (Grondeux 1994: 41) was the Graecismus of Eberhard of Bethune, written near the end of the twelfth century.

Grondeux (1994: 42f.) explains that the author used Brito's entries as his basic framework, at the same time shortening them considerably, and then inserted lemmata selected from the Graecismus together with their explanations, commentary, etymology, etc. With this foundation established, the compiler made several modifications to it: he took the verbs out of the general language collection to make a separate section, reworked Brito's alphabetical order, and provided grammatical indicators for some entries. Entries do not follow a fixed structure; material from the Expositiones may be juxtaposed with material from the Graecismus, a versus and grammatical information may be included, and there may or may not be a Picard equivalent. Only about one article in four includes a Picard translation.

From a lexicological standpoint, the interest in this unique manuscript arises from the Picard forms in the second glossary. A second anonymous writer used the alphabetical glossary to compile a 'reverse' glossary, starting with Picard terms and finding their Latin equivalents. This second author provided a translation for some 2900 of the 4825 lemmata in the first glossary. He wrote the Picard forms in the margins of the alphabetical glossary in the order in which they occurred; that is, first those which begin with 'a' in Picard and 'a' in Latin, next those which begin with 'a' in Picard and 'b' in Latin, and so on. Thus the primary ordering is alphabetical according to the vernacular forms, and the secondary ordering is alphabetical according to their Latin equivalents (id. 1994: 45f.).

(An anonymous fifteenth-century scribe adopted a similar method, inverting Latin lemmata and French equivalents, when writing ms. Paris, BN lat. 7684, which is discussed in 2.5.5.)

The interest in the Picard-Latin glossary in ms. Montpellier, Faculté de Médecine H236 from a lexicographical point of view, lies first in the attempt to create a glossary in which Latin is the exit language and a vernacular (in this case, Picard) language is the entry. Its second lexicographical interest stems from a reference system which the compiler devised as an aid for users. He assigned an alphanumeric notation to each Picard word in the 'reverse' glossary. The notation corresponds to the location in the text of its Latin equivalent (Hauréau 1885: 594; Merrilees/Shaw 1994: 8). The scribe's system for referencing the text is described more fully in 3.1.2.2.2.

The script and the page layout of ms. H236 make it a very difficult work to consult, but despite its drawbacks Grondeux (1994: 44) states that it shows signs of considerable use. Anne Grondeux edited this manuscript for her thesis at l'École Nationale des Chartes, and it is currently being prepared for publication in the Nouveau Recueil des Lexiques latin-français du Moyen Age.

[Next] -- [Table of Contents]