Thursday, May 12, 2016


Maria Chiara Scappaticcio, "Artes Grammaticae" in frammenti: I testi grammaticali latini e bilingui greco-latini su papiro: Edizione commentata. Sammlung griechischer und lateinischer Grammatiker (SGLG), 17. Berlin, Boston: De Gruyter, 2015. Pp. xv, 585. ISBN 9783110445039. $210.00.

Reviewed by Eleanor Dickey, University of Reading (

Version at BMCR home site


What are the Latin grammatical texts, and where does one find them? For well over a century 'Latin grammatical texts' has been more or less synonymous with the lengthy treatises on Latin grammar in Keil's massive Grammatici Latini.1 But over time more and more fragments of further grammatical texts have turned up on papyri and ostraca, and these are important not only for what they tell us directly but also for the context they give to the Grammatici Latini texts. The fragmentary materials have been difficult to find and often overlooked, however, because their publications were scattered through the papyrological literature. Now for the first time the fragmentary texts have a collected edition like that of Keil—and considerably more up-to-date. Moreover, Scappaticcio's edition comes with extensive discussion: it is a detailed study as well as a collection of texts and makes a significant contribution to our knowledge of how Latin grammar was taught and used during the Roman empire. However, it is pitched towards readers with extensive background in the subject and excellent Italian (one often has to actually read the discussion in order to find information that in some other works could have been identified by skimming), and those lacking these qualifications may find it prohibitively difficult.

The work begins with an introduction explaining the different types of ancient grammatical text and how they relate to other types of ancient work on language, such as glossaries and colloquia. The texts themselves are then presented in an order based on their content: first alphabets (seven texts), then 'Schulgrammatik-type' grammars (two continuous prose texts) and 'Regulae-type' texts (four declension tables and four conjugations). These are treated in great detail and are followed by briefer discussions of glossaries with grammatical information (four texts), grammatical scholia (two texts), and two doubtful fragments that may be grammatical. There is also a brief preface by Louis Holtz (in French) and an afterword by Pierre Swiggers and Alfons Wouters (in English). Extensive indices and selected photographs (high quality and in colour, but too small to be easily legible) conclude the volume.

Three of the texts, two (very fragmentary) alphabets and a conjugation table, are completely new. The conjugation table, which is a significant find, is from Oxyrhynchus and will also appear in 2016 as P.Oxy. LXXXII 5302; the co-operation that allowed this text to be published almost simultaneously in two different works is as admirable as it is unusual. The new alphabets are from the Bu Njem ostraca: an extensive archive of notes by Robert Marichal, who did the original Bu Njem publication,2 has recently been rediscovered in Paris and contains meticulous drawings of many unpublished ostraca, including the ones containing these alphabets. As Scappaticcio has access to the entire archive, we can look forward to more new Bu Njem texts in the future.

The different texts are not all treated the same way: sometimes there is a full edition preceded by a heading with basic information on date, location, etc.; sometimes an edition is given without the heading (in these cases much of the basic information can usually be found in the discussion if the reader is persistent, but the search is not an easy one); and sometimes no edition of the text is given. This last is the case with the two Bu Njem alphabets, which are presented via photographs of Marichal's drawings (unproblematically, as the drawings are very clear).

When editions are provided the texts are generally well presented, preserving the original layout, alphabets (some of the Latin is in Greek script), diacritics, and punctuation; Scappaticcio has previously worked on diacritics,3 and this shows in her unusually keen awareness of the importance of the original symbols. Remarkably, she has even managed to preserve the original colour scheme of one papyrus, with red headings actually printed in red ink (pp. 194-9); this is a triumph for De Gruyter as well as for the author and the reader. (Surprisingly, the same was not done for another text where the original used both red and black ink: pp. 70-1.) Line numbers do not always match those of previous editions, and when a papyrus is given both a diplomatic transcript and a restored edition, line numbers in the two versions do not always match each other. Non-grammatical material relevant to a grammatical text and preserved on the same papyrus (for example, one of the alphabets is followed by a line of Virgil, traditionally interpreted as a copying exercise for the person learning the alphabet) is not included in the edition but is mentioned in the discussion. No translations are provided, but the nature of the texts is such that translations would be helpful for only a few of them. In those few Scappaticcio's extensive discussion of each text and its possible meaning(s) will be preferable to a translation for readers working seriously on the text concerned, though not for those looking for a quick key to what the Latin probably means.

The wide-ranging discussion makes valuable new points. For example the words bovabo, bovabis, bovabit on P.Vindob. inv. L 156, dismissed as 'incomprensibile' by the previous editor, are shown to be a conjugation in the future tense of bo(v)o 'bellow'. And Scappaticcio considers not only the individual texts, but also how they fit into the rest of the Latin grammatical tradition —not only the well-known texts in Grammatici Latini, but also the comparatively obscure Hermeneumata Pseudodositheana. Though the Hermeneumata contain no discrete grammars, they incorporate bits of grammatical information, and Scappaticcio shows that they are relevant to the grammatical fragments. As indeed they should be, since the Hermeneumata are bilingual language-learning materials and most of the Latin grammatical papyri were clearly designed to help Greek speakers learn Latin: the papyri in this collection come from the same educational environment as the Hermeneumata and could indeed have been created by some of the same authors. This origin does not separate the papyri from the treatises in Grammatici Latini, many of which also come from the Latin-learning environment of the Greek East: Charisius, Dositheus, and Priscian are only a few of the grammarians whose works, despite being written in Latin, were directed towards non-native-speaker Latin learners.

A section of particular interest is that on alphabets. Some readers may wonder about the classification of alphabets as 'grammatical' texts, since we do not now consider the alphabet part of grammar, but in antiquity grammarians regularly included the alphabet in their discussions, often explaining it in considerable detail. Moreover, learning the Roman alphabet was a common first step for Greek speakers learning Latin—though not an essential first step, as the presence in this book of several Latin texts in Greek script makes clear.

The alphabets included are very diverse: in addition to straightforward Latin alphabets we have a Greek alphabet with Latin transliteration, Latin alphabets with Greek transliteration (in both types the Greek is written above the Latin, but the order of the letters allows one to determine which alphabet is primary and which the transliteration), Greek alphabets in Latin script (i.e. the Latin letters come in the Greek order and include digraphs such as 'ps'), and Latin alphabets with the Latin letter names written above them in Greek script. The juxtaposition of these texts allows Scappaticcio to offer interesting discussion of matters such as diacritics, fluctuating transliteration practices, and the different systems for naming the letters of the Latin alphabet. Many of the texts contain a pair of alphabets (by which I mean not merely a Latin alphabet with Greek transliteration, but two Latin alphabets, perhaps both with Greek transliteration), raising questions about why alphabets should come in pairs. Usually, on the original papyrus the two alphabets appear in different scripts (e.g. one in capitals and the other in minuscule), and many scholars assume that this distinction was a deliberate strategy to help students learn both sets of letter forms. Many editors of these alphabets, therefore, try to reflect the distinction by printing one alphabet in upper case and the other in lower case. In this edition, however, all the alphabets are printed in lower case and the difference in scripts is mentioned only in the narrative introduction.

Another interesting feature that arises from putting the different texts together is the order of the persons in verb conjugations. It has long been observed that one of the ancient conjugation tables (P.Strasb. inv. g. 1175, from the third or fourth century AD) presents verb forms in the distinctive order third person singular, second person singular, first person singular (plurals are not given). This order has usually been explained as an aberration of this particular papyrus, perhaps under Semitic influence, since mainstream grammatical texts are unanimous in using the normal 1-2-3 order. But Scappaticcio's new Oxyrhynchus conjugation table, which dates to the second century AD, also has the peculiar 3-2-1 order (without the wholesale omission of the plural): now that two of the four surviving papyri with conjugations use this order, it cannot be explained as an aberration. In a thought-provoking discussion of the problem (pp. 247-9), Scappaticcio makes the startling claim that a parallel for the papyrus order can be found in the alphabetical glossary of the Hermeneumata Monacensia, which includes some conjugated verb forms. And indeed she is right: in the sample I checked,4 the 3-2-1 order occurs four times—but it is less popular than the usual 1-2-3 order (which occurs 21 times) and also less popular than the bizarre order 1-3-2 (8 occurrences in Latin and 9 in Greek). Yet the presence of the 3-2-1 order is not simply accidental, for other orders (e.g. 2-3-1) never occur. And if the appearance of the 3-2-1 order in the Hermeneumata is not accidental, the parallel is significant here, even if not yet fully understood.

In short, this book is important both for its collection of primary texts and for its discussion, which raises new questions as well as contributing to old ones. It is to be hoped that the work will stimulate further debate about Latin grammatical texts and lead to the publication of more such materials, some of which must be lurking in the world's major papyrological collections.


1.   Heinrich Keil, Grammatici Latini, Leipzig 1855-1880. Many of the works in this collection have, of course, also received more recent editions.
2.   Robert Marichal, Les ostraca de Bu Njem, Suppléments de Libya Antiqua VII, Tripoli 1992.
3.   Maria Chiara Scappaticcio, Accentus, distinctio, apex: l'accentazione grafica tra Grammatici Latini e papiri virgiliani, Turnhout 2012. She has also published on the Virgil papyri: Papyri Vergilianae: l'apporto della papirologia alla storia della tradizione virgiliana (I-VI d.C.), Liège 2013.
4.   Georg Goetz, Hermeneumata Pseudodositheana (volume III of Corpus Glossariorum Latinorum), Leipzig 1892, pp. 142-7. This section of the Monacensia glossary must have a very complex history: it is ostensibly alphabetized by first letter of the Greek (which is transliterated but still uses Greek alphabetical order), but within the section of Greek words beginning with 'h' (i.e. eta), the different verbs are alphabetized by first letter of the Latin. For example, p. 144 starts with forms of neglego, numero, noceo, nos, posco, probo, prandeo, pecco, quiesco, saluto, etc. This alphabetization is not recent; it must have been done on a version of the glossary that contained more Latin misspellings than at present, for on p. 143 Latin verbs beginning with 'a' are followed by venio and vinco, then coquo and other words beginning with 'c'. Evidently the glossary was alphabetized using the common misspellings benio and binco, and the spelling was later corrected.

No comments:

Post a Comment