Latin grammatical manuscripts (including also commentaries, glossaries, and the like) contain surprises. While literary texts are pretty stable, these books are meant to be useful tools rather than cultural monuments, and they are often personal anthologies: a bit of Donatus here, a little Priscian there, some handy word lists, something on metrics, and who knows what else? Given that the components of such manuscripts are largely parts of familiar texts, it is easy to overlook anomalous and unique works that have been inserted without much ceremony among more standard material. Students of the scholarly traditions of late antiquity and the early middle ages recognize the need to expect the unexpected in these manuscripts, and over the past half-century or so some of them have been examined very attentively: new works have been found, and familiar texts have been found to incorporate less familiar ones.
To recognize, edit, and elucidate such material requires great patience and even greater knowledge. That the manuscript now catalogued as Oxford, Bodleian MS Add. C 144 (henceforth: O), written in central Italy in the eleventh century, contains surprises has long been known, and it has been described more than once, notably by Louis Holtz in 1981 and by Mario De Nonno in 2013.[1] Those names also figure in the present story: it was Holtz who long ago suggested to A.C. Dionisotti that there was something of interest in a glossary contained in O, and De Nonno who, in 1990, announced the discovery in O of a lost work on metrics by Martianus Capella.[2] The latter text remains unpublished; publication of the glossary that precedes it in the manuscript and is given there the title Expositio Notarum (henceforth: EN) was announced by Dionisotti in 1996;[3] she has now published it in a magnificent edition with apparatus, commentary, a long introduction, and an even longer, and invaluable, set of appendices and indexes.
The title Expositio Notarum prompts a simple question: what notae are explained in this glossary of some 1800 words and phrases? Dionisotti in 1996 identified extensive overlaps between the lemmata of EN and the words and phrases found in the list of stenographic symbols generally known as notae Tironianae. That system of shorthand was probably created in the late Republic and remained in use for centuries; it fell into desuetude but was revived probably in the eighth century. It was not used (at least not in surviving manuscripts) for transcribing whole texts, but it appears in Carolingian marginalia, and many copies of lists of the notae survive; an edition, with the signs depicted and explained in some 132 tabulae, was published by W. Schmitz in 1893.[4] EN and the Commentarii Notarum Tironianarum (henceforth: CNT) are clearly related: they are similarly organized (the first four commentarii divided into capitula, EN divided into 23 numbered sections); they show similar methods of arrangement, often grouping words by meaning (names of rivers or weapons or articles of clothing), grammatical function, or similarity of sound and spelling rather than using the (often rough) alphabetical order found in most glossaries; they also both contain uoces nihili (as stenographic exercises), which EN endearingly glosses by Latinum non est. By no means all the lemmata in EN and CNT match, but there are enough to be significant; it is all but certain that the collection of notae that EN was originally written to elucidate was another version of CNT. The difference is not surprising: many such sets of notae must have existed, because in antiquity it was a living system of stenography that changed to meet the demands of different environments, and both EN and CNT show signs of change and accretion over time.
Both EN and CNT, however, have entries that might seem odd for stenographic training: along with obvious elements (prepositions, verb and noun endings, legal terminology), we find an entry (A.27=CNT 39.66) for “Quousque tandem abutere Catilina patientia nostra”;[5] all told, there are more than a dozen entries in EN, some of them also in CNT, that seem to come straight from Cicero and not only from schoolbooks like In Catilinam but from letters and from philosophical and rhetorical works. There may be many more lemmata drawn from Cicero, but most of his vocabulary is shared with other authors and thus cannot be securely identified as being taken from his works. Virgil may be even better represented (although again, the conservatism of poetic diction often makes it impossible to tell what is specifically Virgilian), and these lemmata too are partially shared with CNT. Section 17 of EN, for instance, begins with the following sequence (Q.1–7): Pecua, Pecuarius, Tityrus, Elefantus, Caprea, Asinius Pollio, Hic et haec bos. Dionisotti, introducing section 17, describes it as “remarkably coherent: a zoo of animals familiar, exotic, and fantastic” and closely related to CNT tabulae 108–109. EN Q.1, 2, 4, 5, 7 are indeed found in one form or another in CNT, but Tityrus and Asinius Pollio are neither animals nor in CNT. Someone in the world of stenography was clearly interested in the Eclogues, just as someone (perhaps the same person) interrupted a list of birds in section 18 related to a similar list in CNT with R.17-20: Leda, Harpya, Harpalyce, Celaeno before returning at R.21 to Chelido (the bird, not Verres’ mistress). The train of thought is erratic, but birds (even though there is no swan here) lead our stenographer to Leda and to the harpy; the first syllable of Harpya leads to Harpalyce the Amazon; and the idea of harpies leads to dira Celaeno. A final example of the range of interests of our stenographer can be found in the longest entry in EN, M.60 (I give Dionisotti’s slightly emended text of O):
Hic balatro, leno et gulosus tenebr<i>o; de ipso genere sunt et isti: lurco, glucto, bardalio, glutto, boveo†; de ipsis vitiis sunt et isti: malceo, ganeo, quos nos dicimus <…>; popi<n>ones: qui amant ire in popinam.
“A thoroughly tiresome bunch of wastrels, possibly some antiquarian’s haul from comedy or satire” writes Dionisotti, and the text is unfortunately as corrupt as their morals. But whoever these characters are and whatever their origin, they attracted some notarius’ attention; and I would love to have listened to the trial or discussion where the stenographer needed abbreviations for such words in his minutes.
For what it tells us about the educational system that both studied and fostered Latin letters, EN is a genuinely important text. It shows, as do the dialogues of the Hermeneumata Pseudodositheana recently edited by Eleanor Dickey,[6] that the version of Roman education described by Quintilian (and many modern accounts)—with an orderly progression from home study with a paedagogus to the formal study of Latin language and literature with the grammaticus and then to specialization in rhetoric, law, or something else—was relevant to only a small part of the population, the higher reaches of respectable society that produced government officials, advocates, and gentlemen of learning and elegance. That is not news to specialists (particularly students of papyri), but Quintilian’s scenario is a comfortable one—particularly to academics, who tend to be pretty well-educated and who appreciate order and stability. But then as now, this high-flown curriculum was not for everyone, and how sharply the various kinds of education were divided is a real question: texts like EN demonstrate that the lofty study of texts and language under the grammaticus overlapped with what we might think of as secretarial school, the stenographic training of the notarius. And that is something that merits the attention of anyone interested in Roman education and culture. I do not know just what that overlap meant, and Dionisotti does not speculate on the subject. Notarii were probably not making stenographic copies of Cicero and Virgil; did they invent stenographic signs for Ciceronian and Virgilian words because they received the same basic literary education as the upper classes? Was it considered appropriate for those upper classes also to be able to take dictation? The grammarian Valerius Probus also compiled a list of Notae iuris, some of which turn up in EN too; the literate cultures of Rome were varied and complex, and EN offers hard evidence of the links among them, the meaning of which needs much more study.
Dionisotti’s edition of EN is remarkable for much more than the significance of the text it presents: it is a work of profound scholarship. In her 1996 article, Dionisotti explained in more detail than in the introduction to EN the history of glossaries and of the editing of glossaries. The two great collections of such texts, Goetz’s Corpus glossariorum latinorum and Lindsay’s Glossaria latina, are opposites in their approach: Goetz was a splitter, emphasizing the particularity of each glossary and printing them separately; Lindsay was a lumper, attempting without notable success to explain many of the extant glossaries as offshoots of one great, master glossary composed in antiquity. But while there was never any ur-glossary, they are not isolated texts either, and nobody has done more than Dionisotti to clarify the glossographic tradition. In this case, she shows in detail in the introduction (and in the appendices listing all the entries in EN that are related either to CNT or to other glossaries) not only how EN and CNT mesh, but also how a great many glossaries, including those containing Anglo-Saxon glosses and other major collections that are some centuries older than O itself, drew on an ancestor of EN. This information is included in the critical apparatus attached to each entry in EN; this is, to the best of my knowledge, the first truly critical apparatus composed for any glossary, and it is an apparatus of glossographic traditions, not (as sometimes supplied by Goetz) of manuscripts.
Dionisotti goes beyond that, however. While every entry in EN has its list of sources and critical apparatus, entries that need it also receive a commentary, and Dionisotti’s research into the language and possible sources of EN extends far and wide. She points out parallels for EN’s language (both the lemmata and the glosses) ranging from Plautus to the middle ages, from poetry to law to natural history and medicine; from Roman Africa to the Anglo-Saxons. She has read not only the text but the apparatus of works familiar and unfamiliar, for those one can search in dictionaries or online and those for which, as far as I know, no index exists. She has an encyclopedic command of the language and methods of ancient philology and the peculiarities of EN’s diction, all of which are explained in the commentary and distilled in the appendices. I have immense admiration for Dionisotti’s scholarly acumen and her ability to present difficult material clearly and succinctly; I am simply in awe of her knowledge and the sheer amount of toil that went into this volume.
Dionisotti’s work is not easy for someone unfamiliar with ancient scholarship to read; her writing is compressed and her observations are often stated obliquely. But her superb exploration of EN reveals the history of a text and a mode of thinking about language that provide significant contexts for our reading of more familiar and more appealing works; so too, her scholarship has a depth, honesty, and generosity that deserve our attention and respect.
Notes
[1] L. Holtz, Donat et la tradition de l’enseignement grammaticale (Paris, 1981) 409–12; M. De Nonno, “Ancora ‘libro e testo’: nuova descrizione del ms. Oxford, Bodl. Libr. Add. C 144, con osservazioni codicologiche e testuali,” in R. Casavecchia et al. (eds.), Libri e testi: Atti del Seminario internazionale, Cassino, 30-31 gennaio 2012 (Cassino, 2013) 63–109.
[2] M. De Nonno, “Un nuovo testo di Marziano Capella: La metrica,” RFIC 118 (1990) 129–44.
[3] A. C. Dionisotti, “On the nature and transmission of Latin glossaries,” in J. Hamesse (ed.), Les manuscrits des lexiques et glossaires de l’antiquité tardive à la fin du moyen âge (Louvain, 1996) 205–52.
[4] W. Schmitz (ed.), Commentarii Notarum Tironianarum (Leipzig, 1893). Schmitz’s notae are not all found in any one manuscript, and his last dozen plates contain supplements from various sources.
[5] Note that Dionisotti identifies sections by letter, rather than number, for clarity; thus item 27 in section 1 becomes A.27. Entries in CNT are identified by tabula and entry number in Schmitz’s edition.
[6] E. Dickey (ed.), The Colloquia of the Hermeneumata Pseudodositheana (2 vols., Cambridge, 2012 and 2015). Review of vol. 1 at https://bmcr.brynmawr.edu/2013/2013.08.34.