The Carolingian scholars had a higher opinion of Martianus Capella’s De Nuptiis than most scholars have had since. At least, they were sure that he was important, partly because his allegories were not immediately penetrable, and several of them devoted considerable energy to trying to understand him. Martianus Capella himself had been one of the scholars working in Roman North Africa in the early fifth century, and his works had been studied in the Roman West during the intervening centuries, but, as Sinéad O’Sullivan’s study shows, the revival of interest in the details of Latin language and culture led to an extraordinary new ninth-century determination to analyse the work; there is no evidence of pre-Carolingian glosses on the text, so the glossing tradition on the De Nuptiis can indeed be said to begin with the ninth-century Carolingian scholars. And many of them were involved, as we can tell from the large number of different glosses and hands which survive on the margins or as interlinear annotations in most of their manuscripts. O’Sullivan quotes with approval a perceptive comment made recently by Malcolm Godden and Rohini Jayatilaka about glosses on Boethius, in a volume she co-edited, that “annotations have been added ‘in successive stages by different hands in ways which would suggest a variety of sources and commentators’” (viii-ix); the conclusion of her introductory remarks is that there is no point in searching for a “putative perfect original” (xxviii). This is more or less the point also recently made by Alice Rio as regards the chimeric “formularies”, which hardly existed as collected units at the time. Under such circumstances we can probably be justified in referring to a collective cultural and educational tradition rather than just the outstanding work of a few individuals.
This volume only considers the first two books of Martianus Capella’s magnum opus, the introductory ones which ostensibly concern the marriage of Mercury and Philology, but the text of these two books, with their glosses, fills 447 pages (3-449). The edition chosen to be the base published version is that made by James Willis (Leipzig, 1983). Copious relevant material from the twenty manuscripts studied is made available to the reader. Ninth-century glosses are the relevant ones here, so later annotations even on these same manuscripts are not normally reproduced or considered; O’Sullivan realizes that this is not always a neat distinction to make (cxxxiii-cxxxiv), not least because “the distinction between text and gloss was often blurred” (xxvii). A typical page of this edition will contain a single sentence, four or five lines of printed text, but some have none of the text at all, as the glosses from the text reproduced on the previous page spread over to the following one. In addition, there is both a separate apparatus at the foot of each page recording manuscript variants in the actual text (although only of words which are glossed), and an indication of where pre-Carolingian scholars such as Fulgentius and Isidore of Seville (the two most often referred to here) considered the word in question, on occasions when they seem to have been consulted by the Carolingians, although there is no indication there of what the intervening scholars actually said. The editorial effort involved has thus been near superhuman. The glosses are presented after each sentence of the text; many are single words, and often these are illuminating for a Latin linguist (or a Romanist, if Romanists see these volumes) interested in vocabulary; for example, such a simple matter as having textual freta glossed as maria by Romance-speaking commentators probably attests the obsolescence of fretum from Romance as well as from their knowledge of Latin (e.g. in Liber I, 37, Willis 15.21; pages 150-51 in this edition). But the glosses often turn on the point of the use of words whose literal meaning is clear enough; thus three words before the freta there appears noctibus, which has given rise to four different lengthy annotations altogether: e.g., in six manuscripts, “Non quia solis noctibus hoc agat, sed noctes nominat ut uigilias ostendat. Multos enim amor sapientiae uigilare facit”. The thoughts and assumptions manifested in these glosses are not analysed or commented on or translated further here, so despite its pantechnicon appearance this volume provides as many questions as answers; that is, there is ample scope for further scholarly analysis by historians, philosophers, experts in pagan cosmology, literary specialists, Latinists, linguists, Greek specialists (since several of the glosses manifest a determination and curiosity to understand Greek) and others. It is particularly intriguing to note that the Carolingian scholars seem to have been attracted rather than put off by the work’s “overt paganism” (xix). Unfortunately for the linguistically inclined reader, however, “[t]he orthography of the manuscripts is normalized except where a gloss is only found in one witness …. Abbreviations are silently expanded …. Modern punctuation is followed rather than that found in the manuscripts ….” (cxl); it seems that even when we are dealing with a linguistically-interesting text the modern philologists and linguists are still going to find their basic data misrepresented in recent editions. I know, there is rarely room to include every orthographical variant in a volume; but it would be nice to think that the details as printed were a reality, and abbreviations could at least be indicated as such in the good old-fashioned way.
Nineteen of the twenty manuscripts are studied individually (xxxv-cix), in alphabetical order of the city where they are now housed (not of the initial used to refer to them here). Leiden has four, Paris three (plus a fourth mentioned elsewhere but not studied in this section, BnF.lat.12960), Cambridge and Rome two, and Besançon, Cologne, London, Orléans, Oxford, St Petersburg, Trier and Wolfenbüttel have one each. Sixteen of them are sorted out into a stemma (cxxx). O’Sullivan spent six years travelling from Ireland to study these in minute detail, and the resulting stemma inspires confidence. And yet, as so often with such carefully researched textual studies, the result leaves questions of many kinds to be considered further by others wishing to use the data in the service of cultural and sociophilogical studies of the culture of the age. One more general question that worries me, if nobody else, is why textual historians such as the undeniably excellent O’Sullivan still prefer not to help the philologists, even now “emending” (that is, rendering unusable) the attested texts without necessarily indicating when they have done so. If this fails to worry you, you will be extremely impressed by the volume.