Two Fragments of the Carmen Saliare
and the Manuscript Tradition of Varro’s De Lingua Latina
Giulia Sarullo (ITALIA) and Daniel J. Taylor (USA)
Marcus Terentius Varro (116 - 27 BC) is ancient Rome’s most prolific scholar, but unfortunately only precious little of his voluminous corpus has survived. The 600 fragments of Menippean Satires, three books of De Re Rustica, and six of the original 25 books of De Lingua Latina all testify to Varro’s antiquarianism and his profound love of Latin – its vocabulary, etymologies, grammatical structures, history, and literature. He is, as Quintilian famously proclaimed, the vir Romanorum eruditissimus and peritissimus linguae Latinae.
De Lingua Latina VII is especially important, because it studies poetic diction and therefore includes numerous quotations of archaic Latin poetry, many unknown from other sources. In this presentation we wish to address two fragments of the ancient Carmen Saliare that Varro quotes in a passage (VII.26 et seqq.) of linguistic and literary significance.These two fragments of the Carmen Saliare – a longer one in chapter 26 and a shorter one in chapter 27 – provide a fascinating glimpse into the details of digitally reading and editing a text of truly archaic Latin.
Scholarly consensus affirms that Laurentianus Pluteus 51.10, known as F, is, the extant archetype of Varro’s LL. F contains six major lacunae, several noteworthy transpositions, and seemingly innumerable errors. Its 11th century Beneventan script is often hard to read.Moreover, in the case of our two fragments, the unfamiliarity of the archaic Latin has resulted in errors that are not attributable to the usual palaeographical difficulties. Such features recur in all the codices descripti, which number around 75.
The earliest apographon of F is Vallicellianus D. 49.3, which was misdated to the late 15th century and therefore never figured in any account of the manuscripts until Taylor’s 1996 edition of book X; it was copied by none other than Coluccio Salutati in the 1390s. The next earliest apographon is Laurentianus 51.5, known as f and executed by Antonio di Mario in 1427. Thanks to modern technology, especially its ‘zoom’ feature, we digitally and painstakingly autopsied F, Vall, and f and have determined slightly new readings for both quotations from the Carmen Saliare in F and several diagnostic errors or corrections in both Vall and f that reoccur in other codices descriptiand in the editio princepsthat we also autopsied electronically.
The distinctive Beneventan i-longa and small case /l/ are virtually indistinguishable, and it is therefore no surprise that F's /l/ in lancusianesis retained by f but read as an /i/ by Coluccio.F's word divisions are often idiosyncratic, and so Antonio breaks up F's Cozeulodorieso into Cozeu lodori eso, thereby providing the starting point for Pomponio Leto's list of putative archaic Latin words in the editio princeps.Perhaps due to his failing eyesight, Coluccio omits a /p/ in supplicante but more importantly splits the word into suplice cante, a procedure followed by at least ten other manuscripts surely copied from Vall.Likewise, Antonio reconfigures F's Ianusve vet as Ianus vevet, which is reproduced in a half dozen manuscripts copied from f.F and f both read duonus ceruses, but Vall has duo nusceruses, a reading which recurs elsewhere and, more importantly, is read by Leto as duo nuseruses, which then appears in most subsequent editions.These are but a few of the interesting and valuable textual issues that we will address.
Easily the most curious and most diagnostic reading occurs in Vall.F reads dun; = dunus, but Coluccio doesn't understand the Beneventan semi-colon that indicates a final -us; he only understands the semi-colon used after a q to indicate -ue, that is, the enclitic-que.So he reads dun; as dunue and inserts the q in order to make dunque, which he emends to dumque, a perfectly good Latin word but a far cry from F's dunus.Vall's newly created locution appears in other manuscripts and enjoys a lengthy existence afterit is printed by Leto and then by almost all subsequent pre-scientific editions.
We conclude that F, f, and Vall are the triple fount of the Renaissance manuscript tradition of Varro's De Lingua Latina.Moreover, it is perfectly obvious that Pomponio Leto incorporated many of Vall’s readings in his editio princeps. We are therefore provisionally prepared to conclude that Vall is the mysterious but so far undiscovered primary manuscript that Leto used for his edition.That manuscript has been described as "a manuscript similar to M," a statement that makes perfect sense once it is realized that M was copied from Vall.The innovations and errors in the two fragments of the Carmen Saliare that we have been able, thanks to digital philology, to examine in such detail provide a microcosmic picture of the overall textual and editorial tradition of Varro's LL and its attempt to produce that ever elusive definitive text.