Donati Graeci

Columbia Studies in theClassical Tradition

Editorial Board

William V. Harris (editor)Eugene F. Rice, jr., Alan Cameron, Suzanne Said

Kathy H. Eden, Gareth D. Williams

VOLUME 32

Donati GraeciLearning Greek in the Renaissance

By

Federica Ciccolella

LEIDEN BOSTON2008

Cover illustration: Milan, Biblioteca Trivulziana, MS. 2167 (Liber Donati, end of the fifteenthcentury), fol. 13v: young Earl Massimiliano Ercole Sforza at school.

This book is printed on acid-free paper.

Library of Congress Cataloging-in-Publication Data

Ciccolella, Federica.Donati Graeci : learning Greek in the Renaissance / by Federica Ciccolella.

p. cm. -- (Columbia studies in the classical tradition ; 32)Includes bibliographical references and index.ISBN 978-90-04-16352-2 (hardback : alk. paper) 1. Greek language--Study and

teaching--History--To 1500. 2. Greek philology--History--To 1500. I. Title. II. Series.

PA57.C53 2009480.71--dc22

2008039458

ISSN: 0166-1302ISBN: 978 90 04 16352 2

Copyright 2008 by The Trustees of Columbia University in the City of New York.

Published by Koninklijke Brill NV, Leiden, The Netherlands.Koninklijke Brill NV incorporates the imprints Brill, Hotei Publishing,IDC Publishers, Martinus Nijhoff Publishers and VSP.

All rights reserved. No part of this publication may be reproduced, translated, stored ina retrieval system, or transmitted in any form or by any means, electronic, mechanical,photocopying, recording or otherwise, without prior written permission from the publisher.

Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill NVprovided that the appropriate fees are paid directly to The Copyright Clearance Center,222 Rosewood Drive, Suite 910, Danvers, MA 01923, USA.Fees are subject to change.

printed in the netherlands

Marco carissimo

CONTENTS

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi

Chapter One. The Latin Donatus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. Aelius Donatus Artes: A Pedagogical Program . . . . . . . . . . . . . . 12. The Shaping of the Medieval Donatus . . . . . . . . . . . . . . . . . . . . . . 83. Donatus(es) as Schoolbook(s) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164. Ianua . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205. Ianua(e): Structure and Variants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296. Vernacular Donatus(es) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447. Donati meliores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478. The Association with Disticha Catonis . . . . . . . . . . . . . . . . . . . . . . . . . 529. Sixteenth- and Seventeenth-Century Printed Editions of

Ianua . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5410. Learning Latin: Repetition, Memorization, and

Translation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5611. The Latin Curriculum. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

Chapter Two. The Greek Curriculum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 751. The Position of Greek in Roman Education . . . . . . . . . . . . . . . . 772. Greek Grammar in the Middle Ages: An Impossible

Dream? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 853. Humanism and the Revival of Greek Studies . . . . . . . . . . . . . . . 974. The Byzantine Tradition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1035. Practicing Greek Grammar: Ertmata, Epimerismoi, and

Schedography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1096. The Making of Humanist Greek Grammar . . . . . . . . . . . . . . . . . 1187. Other Grammars and Course Books . . . . . . . . . . . . . . . . . . . . . . . . 1248. Teaching Greek in Humanist Schools . . . . . . . . . . . . . . . . . . . . . . . 1309. Schools of Greek . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13910. Teaching Greek in Greek: Michael Apostolis and the

Direct Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146

viii contents

Chapter Three. Donati graeci . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1511. In Search of the Greek Donatus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1522. The Four Donati graeci or Pylai . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1543. Pyl a: The Manuscripts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1584. Pyl a: Toward a Stemma Codicum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1725. Pyl a as a Grammar Book: The Variable Parts of Speech . . 1806. Pyl a as a Grammar Book: The Invariable Parts of

Speech and the Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1957. The Other Donati graeci: Pylai as Compilations or Donati

compositi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1988. The Manuscripts of the Donati compositi . . . . . . . . . . . . . . . . . . . . . . 2009. The Donati compositi as Grammar Books . . . . . . . . . . . . . . . . . . . . . 20910. The Language of the Greek Donatus: Between Greek and

Latin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22111. The Greek Cato . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225

Chapter Four. The Greek Donati and Their Context . . . . . . . . . . . . . . 2291. Latin in Byzantium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2292. Maximus Planudes and the Greek Donatus . . . . . . . . . . . . . . . . . 2373. Places of Origin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2444. Using the Donati graeci in Schools: Reutilization and

Superimposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259

TEXTS

Donatus graecus aSiglorum conspectus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263Textus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266Appendix latina . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398

Donatus graecus bSiglorum conspectus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401Textus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403

Donatus graecus cSiglorum conspectus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429Textus. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430

Donatus graecus dSiglorum conspectus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 493Textus. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 495

contents ix

NotesDonatus graecus a . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513Donatus graecus b . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 523Donatus graecus c . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 529Donatus graecus d . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 544

Appendix I. Comparing the Four Donati graeci . . . . . . . . . . . . . . . . . . . . . 555Appendix II. The Manuscripts of Ianua . . . . . . . . . . . . . . . . . . . . . . . . . . . . 583

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587Index of Personal Names . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 623Index of Manuscripts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 635

INTRODUCTION

Hi nunc Constantinopoli capta quis dubitet incendio quaevis scriptorum monumenticoncidentur? Nunc ergo et Homero et Pindaro et Menandro et omnibus illustrioribuspoetis secunda mors erit. Nunc Graecorum philosophorum ultimus patebit interitus.Restabit aliquid lucis apud Latinos, at fateor neque id erit diuturnum.

[Now that [the Turks] have captured Constantinople, who can doubtthat every memorial of the ancient writers will be set on fire? NowHomer, Pindar, Menander, and all the most famous poets will die forthe second time. Now the last destruction of the Greek philosopherswill be at hand. A gleam will survive among the Latins, but, I wouldsay, it will not last for a long time either.]1

Few phenomena shaped Western European culture as significantly asthe rediscovery of ancient studies during the Renaissance. As JacobBurckhardt has pointed out, though the essence of the phenomena[i.e., the cultural aspects of the Renaissance] might still have beenthe same without the classical revival, it is only with and through thisrevival that they are actually manifested to us.2 As the ideal trainingfor the ideal citizen of the new era, the system of humanist educationreplaced the medieval curriculum, which had equipped individualswith complex skills appropriate to specialized tasks, but was based onthe authoritative message of a few selected texts.

The influence of the culture of antiquity had not died out inEurope during the Middle Ages. The revival of classical antiquity pro-moted by Charlemagne in the ninth century was already a form ofRenaissance; even some aspects of monastic scholarship can be under-stood by considering the direct influence of Latin writers, whose workshad continued to be copied, studied, and imitated within the wallsof medieval monasteries. Despite the increasing spread of vernacularlanguages, Latin kept its role as the language of the church, law, andinternational aairs, as well as of science and learning, throughout

1 From a letter by Enea Silvio Piccolomini (Pope Pius II) to Cardinal Nicholas ofCues, 21 July 1453 (Pertusi 19902, 2. 54).

2 Burckhardt 1990 [1860], 120.

xii introduction

the Middle Ages, especially in Italy, where the Roman past had leftthe most outstanding traces.

However, the Renaissances attitude toward classical antiquity wasvery dierent from that of the Middle Ages. The rebirth of city life,which started in Italy in the fourteenth century, favored the rise ofa new culture and, at the same time, the rediscovery of the past:Culture, as soon as it freed itself from the fantastic bond of theMiddle Ages, [] needed a guide, and found one in the ancientcivilization.3 Men of culture continued to study the Latin authorsof the medieval curriculum, but expanded their knowledge by addingother authors and literary works. Recovering ancient texts that, for along time, had lain neglected in monastic libraries of Europe becamethe goal of many humanists. The new manuscripts made up largecollections, and the texts they contained reached a wider audience:the use of a simplified handwriting in manuscripts and, later, theinvention of printing sped the reproduction of books and made theircirculation easier. Translations spread the knowledge of these textsamong a wider public.4

This new culture had enormous eects on education. In the hu-manist system of global education of the perfect citizen, the humani-ties acquired a significant place, along with the seven liberal arts andmore practical disciplines, such as law and medicine.5 At the sametime, ethical and religious values were nurtured: finally, the conflictbetween ancient pagan culture and Christianity found a solution in

3 Burckhardt 1990 [1860], 123. See also Kristeller 1979 [1955], 1920; Witt 2000,173; and Marcucci 2002, 1519. Scholars have often emphasized that, among the Ital-ian city-states, Florence had partially recreated the political and social environmentof fifth century Athens: the rise of a new class of merchants and traders encour-aged the development of human abilities and being open to the world, while wealthand powerand not the nobility of birthdetermined participation in political life.Florentines grounded the pursuit of human happiness in the use of creative intel-ligence and the fulfillment of the values of antiquity: virtue, justice, wisdom, andprudence, i.e., the four Platonic virtues (cf. Symp. 209A). Thus, there was no contra-diction between the individual quest for material wealth and Christian morality. See,e.g., Stinger 1988, 176.

4 On Renaissance translations, see in particular the excellent works by Baldassarri(2003) and Botley (2004).

5 The idea that the humanities are a necessary component of a global educa-tion is expressed in Pier Paolo Vergerios treatise De ingenuis moribus et liberalibus studiisadulescentiae (text and translation in Kallendorf 2000, 291, in particular 2829). SeeBuck 1959, 273 f.; Gundersheimer 1965, 7 and 25 (with a partial translation of thetreatise, 2638); and Garin 1966, 114 f.

introduction xiii

the West. By the first half of the fifteenth century, the studia humani-tatis became a clearly defined circle of scholarly disciplines: grammar,rhetoric, history, poetry, and moral philosophy. In each of these dis-ciplines, the reading and interpretation of Latin and Greek authorsplayed an important role.6 The humanities were not regarded as anencouragement to otium, but as a necessary support of negotium, a stim-ulus to action; the imitation of the style and content of the works ofthe classical writers provided an excellent source of inspiration.7

The picture, however, is not homogeneous. Although the crisisof medieval pedagogy had already emerged during Petrarchs time(13041374), there was no conscious break with the past until the fif-teenth century. For a long time, humanist teachers continued to usemedieval teaching methods, readings, and schoolbooks, and to regardrepetition, memorization, and imitation as the students main tasks.8

Moreover, in spite of the steady rise of the vernacular languages, theprevalence of Latin as the medium of instruction remained unchal-lenged for a long time.9

The manifold aspects of the humanist revival of ancient culturehave been extensively studied. In the last hundred years, the discov-ery and publication of many documents has oered a more preciseand detailed picture of Renaissance education. Evaluating the rela-tionship between the new pedagogy and medieval culture and educa-tion, however, is much more dicult:10 the interpretation of the extent

6 According to Kristeller (1979 [1955], 23.), the humanists main concern wasliterature. In fact, most of them were teachers, professors, or secretaries to princes orcities; most of their works were orations, letters, poems, or historical works. Therefore,Renaissance humanism must be understood as a characteristic phase in what maybe called the rhetorical tradition of Western culture. Kristeller denies a philosophicalorigin of Humanism, even if he acknowledges the impact of the new culture onphilosophy because of the emphasis placed on the individual and the rediscoveryof the Greek philosophers.

7 Witt (2000, 8.), instead, interprets Renaissance education as being based on apolarity between vita contemplativa and vita activa and considers grammar and rhetoricto represent the terms of this opposition. Such a contrast was not felt in antiquity:grammar and rhetoric were complementary in education. For an overview of thehistory of modern studies on the Renaissance after Burckhardt see Rabil 1988; andCelenza 2004, 157.

8 See Padley 1976, 9: Rather than a training in original thought, [humanistpedagogy] proposed a model of elegance. Therefore, everywhere, Renaissancegrammar remains to a large extent bound up with rhetoric.

9 See, e.g., Bolgar 1954, 267.; and Garin 1958, XIII.10 For example, Garin (1958, 91104, al.) emphasized the contrast between medie-

val education and Renaissance pedagogy, which he regarded as a sort of revolution.

xiv introduction

and eects of the changes in humanist education, as well as the evalu-ation of its continuity with the past, are still being discussed. However,it is undeniable that the classical revival was extremely important, atleast in the intentions of humanist educators: the idea of an indissolu-ble bond between past and present inspired the pedagogical theoriesof the Renaissance.11

The re-introduction of Greek studies in the West represented a sig-nificant innovation in Renaissance culture. Most contemporaries wereaware of its importance and described it as a sort of miracle, dueto the initiative of some individuals (Coluccio Salutati and the Flo-rentine humanist circles) and to the ability of a Byzantine teacher(Manuel Chrysoloras) who taught Greek to Westerners. Actually, theGreek revival was the culminating point of a long process; the alter-nately friendly and hostile relationships between Byzantium and theWest had contributed to the mutual knowledge of the two worlds.Interestingly enough, the West and Byzantium followed parallel pathsin their mutual approach: the interest in Greek culture that, duringthe fourteenth century, began to develop in the West, corresponded tothe spread of Latin culture in Byzantium.12 The migration of Byzan-tine scholars to Italy, which had started long before the capture ofConstantinople by the Turks (1453),13 disclosed to Westerners a culturealmost unknown to them: Greek writers, whom Westerners had for-gotten since the end of antiquity, were read, translated, commentedon, and imitated. Gemistos Plethons lectures challenged centuries ofScholasticism and, in this way, the foundations of medieval culture.On the other hand, both Byzantine scholarship and Western culture

Almost three decades later, Anthony Grafton and Lisa Jardine (1986, xiixiv) chal-lenged Garins assumption by pointing out the aspects of continuity with the past dis-played by humanist culture, in general, and with school, in particular. Black (1991a,315.) oers a summary in an analysis of the conclusions reached by Grendler (1989;see, however, Grendler 1991, 335.). See also Witt 1988.

11 See Grafton-Jardine 1982, 55.12 See in particular Kristeller 1966, 21. In any case, Byzantiums interest in Latin

culture was less intense and was basically restricted to the Latinophrones, the supportersof the Union (i.e., of the return of the Orthodox Church under the sovereignty of theChurch of Rome).

13 See Monfasani 2002, 30: We no longer believe in the myth that Greek emigrsfleeing the fall of Constantinople in 1453 caused the Renaissance in Italy. Thisidea, expressed by the humanist Pier Candido Decembrio, has been repeated forcenturies in textbooks as the logical cause of the Italian Renaissance; see Burke2001 [19992], 243, who compares this migration to the flight of central Europeanscholars to England and America after 1933; and Bianca 2006, 4.

introduction xv

were heirs to the homogeneous Greco-Roman of late antiq-uity. Thus, for example, Byzantine and Western scholars practiced thesame methods of teaching and approaching ancient texts; the focus onstudia humanitatis in humanist schools resembled the rhetorical trainingof Byzantine schools in the Palaeologan age.14 For this reason, Byzan-tine curriculum and pedagogy could be easily transferred in the West;Western scholars who wanted to practice their Greek read the sametexts and did the same exercises as Byzantine students.

The re-introduction and the spread of Greek studies in the West,however, required important transformations with respect to the an-cient Greek and Byzantine traditions; for example, it became neces-sary to adapt Byzantine grammar, conceived for native Greek speak-ers, to the needs of students who approached Greek as a foreignlanguage, usually through Latin. Only those changes could grant toGreek studies a permanent place in Western culture. So far, scholar-ship has paid little attention to grammar books, lexica, and dictionar-ies: in other words, to the tools that made this revival possible, as wellas to the methods that teachers followed to impart to their students aknowledge of Greek.15

Even a partial analysis of these tools reveals that the return ofGreek studies to the West was everything but a straightforward pro-cess; its history includes bright triumphs and depressing failures, con-sent and criticism, acceptance and resistance. Byzantine emigrs whotaught Greek in Italy became rich and famous (e.g., Manuel Chrysolo-ras) or were frustrated in their ambitions (e.g., Michael Apostolis);Westerners who spent their energies learning Greek could reach animpressive mastery of the language (e.g., Bruni, Filelfo, and Politian)or remain obscure and mediocre scholars. More importantly, thisminor grammatical material, still largely unexplored, reveals thatChrysoloras, the author of the first Greek grammar for Westerners,

14 See in particular Geanakoplos 1988, 350 f.15 See Cortesi 1986, 164 f.: La dicolt maggiore nasce [] dalla mancanza di

studi sistematici e organici relativi alluso dei manuali di grammatica per lapprendi-mento del greco in occidente, alla loro struttura e alle fonti in essi confluite [] Laconoscenza del fenomeno grammaticale greco imporrebbe [] un preciso lavorodi raccolta dapprima di tutti i testimoni rimasti, fase indubbiamente faticosa perlimmenso materiale che giace inedito nelle biblioteche, per passare poi allanalisie alla verifica dei contenuti. Pontani (1996, 135) also stresses the importance of thestudy of lexica and grammar books: Qualunque discorso sulla presenza della linguagreca in Occidente deve fondarsi sula positiva conoscenza degli strumenti attraverso iquali la lingua poteva essere appresa: cio grammatiche e lessici.

xvi introduction

was not the demiurge or the deus ex machina that his contemporariesdescribed. First of all, Byzantine scholarship of the Palaeologan agealready had promoted a rethinking of Greek grammatical tradition;during the Palaeologan age, Byzantine elementary grammar wasslowly undergoing a process of simplification to meet the demandsof students for whom the Attic Greek still used in literature anddocuments was like a foreign language. Secondly, Chrysoloras maininnovationthe use of the Latin system of declensions based on theending of the genitive singular for Greek nominal inflection, antici-pated by Roger Bacon in the thirteenth centurywas not the onlycase of application of a Latinate scheme to Greek grammar. The fourGreek Donati, the object of this book, represent four attempts to createa Greek grammar modeled on Latin; thus, students could learn Greekusing patterns as familiar to them as those that they had used to learnLatin.

A study of the four Donati graeci must take into account the com-plex balance between continuity and innovation in grammatical stud-ies and pedagogy, as well as the interaction and exchanges betweenEast and West during the Renaissance. The Greek Donati are Greektranslations or adaptations of the so-called Donatus or Ianua, an ele-mentary book used during the Middle Ages and the Renaissance inlearning Latin. By unveiling the origin, function, and fate of the GreekDonati, it will be possible to analyze an almost unknown aspect of therevival of Greek studies in the Renaissance. As Greek grammar books,in fact, the Greek Donati are failed experiments: they were not deemedworthy of a printed edition, nor, apparently, did they circulate outsideof Italy and/or Crete.

One of the four texts, version a, probably originated as a sim-ple word-for-word translation of the Latin textbook for Greeks whowanted to learn Latin. During the fifteenth century, this version,perhaps originally written in the interlinear spaces of a Latin text,became an independent grammar and was used to learn Greek;apparently, Greek Donatus a did not undergo the process of adap-tation of Latin morphology to the target language that led to thecomposition of Donati in modern languages.16 More advanced andimproved grammars replaced the Greek Donatus, but the replacement

16 See below, 4446. Greek Donatus a applies to several Greek versions of Ianua,which are dierent from each other but, apparently, closely related for their approachto the original text and the language used.

introduction xvii

may not have been complete. In fact, when the extant manuscripts ofversion a were produced, the Greek grammars of Manuel Chrysolo-ras, Theodore Gaza, Constantine Lascaris, and some other Greekscholars already existed. Thus, Greek Donatus a, the Donatus transla-tus, may have been transformed into an independent grammar in anarea where such books were not circulating. Many elements point toCrete or some other Venetian colony in Greece as possible places oforigin. It is highly probable that some Venetian ocers and their fam-ilies brought the most widespread elementary Latin schoolbook, Ianua,and that the book was later translated into Greek.

Versions b, c, and d, on the other hand, are clear and consciousattempts to create real Greek elementary grammars using Ianuasstructure. Some parts maintain a tight link with the Latin text, whilesome other parts are taken from authentic Greek grammatical mate-rial, especially from Moschopoulos ertmata or from his source(s).In all three Donati compositi, for example, the paradigms of the sec-tion on verbs are derived from the same source, which was prob-ably an improved Greek translation of Ianua. Most probably, Donaticompositi were just some of the many compilations of grammaticalmaterial available at that time. In any case, compiling seems to bea typical method in the composition of Renaissance Greek gram-mars: for example, some sections of Chrysoloras Erotemata closelyecho Moschopoulos work.

As Greek grammars, in spite of their many dierences, the fourDonati graeci show some common features. All of them consider fivenominal declensions and four regular verbal conjugations (with theaddition of a variable number of irregular verbs, all modeled onLatin). The parts of speech expounded are the same and in the sameorder as in Ianua: this means that in Greek there is no article, but,as in Latin, there is an interjection. Greek nouns lack a dual buthave an ablative (and Greek prepositions can take the ablative, too);Greek verbs lack both the dual and the middle voice but have theimpersonal voice, the future imperative, all tenses of the subjunctive(in a), and even a supine and a gerundive (in a). The Donati compositi,on the other hand, acknowledge the existence of the aorist, but oftenconfuse it with the perfect. These four grammars build up a type ofGreek grammar that I would call Donatus-type or Pyl-type.17 The

17 The use of the term Pyl is explained below, 154155.

xviii introduction

Greek Donati are closely connected with Latin elementary grammar:they represent the introduction of Latin parsing grammar (Poeta quaepars est, etc.) within a dierent system, which privileged definition ( , etc.).

Greek Donatus a, first of all, is a translation from Latin. Therefore,the first chapter of this study deals with its Latin original, the Pseudo-Donatan Ianua, by sketching out the evolution of Aelius Donatus Arsminor in its medieval forms; Ianua was just one of them. ChapterTwo examines the rediscovery of Greek culture and the revival ofGreek studies in the West. After a brief description of the study ofGreek in antiquity and in the Middle Ages, this chapter describesand evaluates the role played by Byzantine emigrs in re-establishinga Greek curriculum in the West. Three aspects receive particularemphasis: the creation of a new Greek grammar for Westerners, thesimilarities between Byzantine and Latin pedagogy, and the positionof Greek in Renaissance schools.

The last two chapters focus on the Greek Donati. Chapter Threeoers an analysis of the texts, describing the manuscript tradition andthe content of the four versions, with particular emphasis on versiona. Since, however, a critical edition of a grammatical text makes senseonly if it takes into account the pedagogical context within which thetext was produced,18 the final chapter deals with the questions raisedby the Greek Donati: chronology, authorship, place of origin, and usein classrooms and/or for the self-study of Greek.

The anonymous translator(s) of Greek Donatus a probably did notintend to write a Greek grammar and Maximus Planudes did notknow that his Greek translation of the Disticha Catonis would becomean elementary reading for Western students of Greek. In the sameway, I did not approach the Greek Donatus originally for its signif-icance in Renaissance Greek studies. This project started as a finalpaper for a graduate course on the reception of antiquity, taught atColumbia University by Professor Suzanne Said in the fall of 1998.While inquiring into the reception of Latin culture in Byzantium, Icame across Maximus Planudes translations from Latin; among theworks of this extraordinary polymath, the (supposed) translation of

18 See Carlotta Dionisottis programmatic assertion (1984, 208): when dealing withgrammatical texts, it is necessary to engage with the text, to ask it questions, in shortto understand both it and why one is working on it.

introduction xix

Aelius Donatus Ars minor still awaited a critical edition. I decided toundertake this task in my dissertation: the limited number of manu-scripts transmitting the textsix, five of which are in Italian librariesand directly accessible to meseemed to oer an ideal condition fortextual criticism. Also, that topic granted my study a fair level of origi-nality; in fact, the only existing monograph on the Greek Donatus wasan unpublished dissertation that Wolfgang Oskar Schmitt defendedin 1966 at the Humboldt Universitt of Berlin. Schmitts outstand-ing workwhich, unfortunately, had almost no circulation outside ofthe former German Democratic Republiccontained a thorough dis-cussion of the problems related to the Greek Donatus (especially ofPlanudes authorship) and a critical edition of the text.19

I began my work by checking the existence of other manuscriptscontaining the Greek Donatus; I was able to use catalogues, electronicresources, and data banks that were not yet available at the time whenSchmitt was doing his research or were not accessible to him as ascholar living beyond the Iron Curtain. The result was the discoveryof five new manuscripts, containing three dierent versions of thesame text. In my dissertation (Columbia University, 2004), I presenteda critical edition of the Planudean Greek Donatus, i.e., version a, aswell as an overview of other two versions, i.e., b and what I believedwas the only extant part of c, the section on verbs.20 In my editionof the text, the new manuscripts allowed me to confirm or to correctmany of Schmitts assumptions.

Further research has led me to discover the missing parts of ver-sion c, as well as a fourth version, d, in the grammar attributed toZacharias Calliergis in one of the manuscripts of version a. Anothersignificant result has been the identification of some elements that

19 Schmitts dissertation was made available to me in microforms, thanks to theeort of the Interlibrary Loan Service of Butler Library at Columbia University.Schmitts dissertation still represents an indispensable starting point for further re-search on the Greek Donatus. Starting from a general outline of Planudes activityas a translator (pp. 136), Schmitt analyzed the text of the Latin Ianua (37103), itsGreek translation, and its manuscript tradition (104211). He equipped his editionof the Greek text (1+92+) with a commentary centered mainly on language andstyle (212265). The hypothesis of Planudes authorship was examined and rejected inthe final part of the work. My attempts to contact the author in Berlin through hisformer advisor, Johannes Irmscher, have yielded no result. This book is also dedicatedto Wolfgang Schmitt, an extraordinary scholar, and to the memory of JohannesIrmscher.

20 See Ciccolella 2004, 260.

xx introduction

point to Crete or the Northeastern part of Italy as the place of ori-gin of the four Greek Donati.21 These new discoveries have immenselywidened the field of inquiry. One Greek Donatusthe vulgate, ver-sion amight still be viewed as the result of the initiative of an indi-vidual or as the authentic or spurious work of a Byzantine scholarand, as such, as a contribution to literary history only. But four ver-sions of the same text, which probably originated within the sameenvironment, indicate that they were created to respond to precisecultural demands; thus, we must consider the four extant Donati graeciin general, within the context of the revival of Greek studies in theWest, and in particular, as products of the search for adequate toolsfor the teaching of Greek, which distinguished the first stages of thatrevival.

Working on Renaissance Greek grammar represents a challengeand requires, so to speak, the spirit of a pioneer. First of all, no mod-ern critical editions are available for the most important Byzantine-humanist grammars: Manuel Chrysoloras Erotemata, Theodore Ga-zas Introduction to Grammar ( ), Constantine Las-caris Summary of the Eight Parts of Speech ( ), and, of course, minor works such as Chalcondyles and Cale-cas grammars. Secondly, there is no general survey of the study ofGreek in the Renaissance comparable, for example, to Robert Blacks2001 extensive study on Latin education. A monograph that may con-sider not only the products of high scholarship, such as translationsof classical texts, but also the tools available for elementary instruc-tion in Greek is still a desideratum for the study of Greek culture inthe Renaissance. Also, we need systematic studies on the structure,sources, and use of Greek grammar manuals in the West, but notwithout the preliminary, careful work of collecting and cataloguing allthe immense amount of extant material scattered throughout Euro-pean and American libraries.

This book is intended as a first attempt to fill this gap by describ-ing a tradition of Greek studies somehow connected with Venice andCrete; this tradition was certainly inferior in quality and circulationwhen compared with the Florentine scholarship, but it is still impor-tant in the cultural history of the Renaissance.22

21 See Ciccolella 2005, 1520.22 On the dierences between Venetian and Florentine humanism, see Witt 2000,

8587, 454458.

introduction xxi

The edition of grammatical texts raises particular problems. It istrue that Renaissance texts present the advantage of being closer intime to our age than classical texts.23 However, texts created for schooluse were subject to continuous modifications. As a matter of fact, agrammar was not merely read but used: like all secular books, gram-mar books were not made to last.24 In the past as well as in thepresent, a grammar was continuously open to corrections. Unusualforms were likely to be corrected and replaced with more normalforms or perhaps deleted entirely. Teachers were free to modify thetext by excluding superfluous material or including what they con-sidered appropriate to their own pedagogical methods and to thedemands of their classes. Typically, later editions of schoolbooks arebetter than earlier ones: this reverses the idea of an archetype, inLachmanns sense, as a possible goal of a critical edition. We shouldalso consider that elementary school texts were usually much moretolerant of forms of the spoken language than literary texts. Theoreti-cally, then, an excess of normalization of language and style wouldbe inappropriate. Practically, however, we should not exclude thatsome modern or extravagant forms were introduced in the copy-ing of manuscripts unintentionally by copyists rather than intention-ally by grammarians. Applying common sense or skeptical suspensionof judgment is often the best solution: in some cases, abandoning theidea of correcting the text or even using cruces desperationis may be notsigns of defeat, but ways to respect what is no longer accessible tominds educated in analogical Greek grammar.25

23 See Celenzas considerations about the editing of Renaissance texts (2004, 136).24 See Cavallo 1980, 158.25 For a discussion of Lachmanns method and of its use in modern textual criti-

cism, see Reynolds-Wilson 19913, 214; Irigoin 2003 [1977]; and Montanari 2003, 3340. Polara (1991) has eectively described the diculty of editing grammatical works;his remarks deserve to be quoted extensively: (102) Tra gli obblighi pi evidenti perun editore di testi grammaticali seri c quello di correggere, o almeno tentare dicorreggere errori visibilmente non dautore, che rendano lesposizione incomprensi-bile, incoerente o contraria alla dottrina consolidata. (103) Di fronte ad evidentiinesattezze, che non si possono neppure attribuire ragionevolmente ad un momentodi distrazione del grammatico, lintervento editoriale si impone ed il rispetto dellatradizione manoscritta significa solo incomprensione della dottrina. On the otherhand: (107) [] nel tentativo di formire un testo ragionevole, non solo si pu incor-rere in madornali errori, quando ci si lasci prendere la mano dal gusto dellinterventoe si sostituiscano il proprio pensiero e la propria dottrina a quelli dellautore, ma canche il pericolo sottile di eliminare, come sviste involontarie e perci emendabili,veri e propri usi stilistici, o addirittura consuetudini di lingua.

xxii introduction

I have corrected or modified the texts only when they appearedclearly altered or corrupted and when their alterations could be ex-plained by the manuscript tradition.26 In general, however, I havetried to preserve elements that I consider as constituent features ofthe Greek Donati and closely bound to their Latin model, such asthe confusion between declensions and verbal tenses or the Latinatelexicon and syntax of some passages. Similarly, I have preserved formsthat reveal the influence of demotic Greek in the study of Greekclassical grammar. In fact, I have not aimed to make the Greek Donatisimilar to grammars in the modern sense, but to oer to readers andscholars four texts that, with their imperfections, might document aparticular stage of the revival and spread of Greek studies in theWest.

In laying out the texts, I have tried to oer a complete picture of thevariae lectiones found in the manuscripts of versions a and b by present-ing them in parallel columns when they represent dierent branchesof the manuscript tradition.27 Often, I have given my preference to

26 Of course, fixed rules do not exist: choices may vary according to specificsituations. The fact that most of the scribes of the Greek Donati so far identifiedwere professional copyists, priests or notaries and, as such, presumably in controlof the texts they were copying would discourage any textual intervention; an excessof conservatism, however, may harm the clarity and readability of texts, which isthe purpose of any modern critical edition. Thus, on the one hand, assuming thatthe Greek Donati were tools to teach literary Greek, I have corrected nominal andverbal forms influenced by the spoken language (). Most probably, in fact,these anonymous grammarians relied on the grammar taught by late antique andByzantine grammatical texts, which were based on Attic Greek; the influence ofthe current language may have easily aected the writing and the transmission ofancient grammatical forms. On the other hand, I have preserved some non-classicalforms and constructions attested in Hellenistic and Byzantine literary texts, espe-cially in the parts of the Greek Donati intended for describing conceptse.g., defi-nitions and exampleswhere grammarians may have chosen language and style inorder to communicate more easily with their students. I have preserved inconsisten-cies and omissions, as well as some forms attested by all manuscripts and reflect-ing the common usage of the spoken language, because correcting them would havemeant introducing an abstract idea of Greek into a text that originated as a responseto concrete demands. Conversely, I have corrected forms and passages that mayhave appeared obscure to modern readers. Finally, I have avoided textual interven-tions in the many cases of non-attested grammatical forms or irreparably corruptedpassages.

27 In the Greek Donatus a, I have italicized the passages that belong to MS. R only.This manuscript is closely related to the vulgate text (x), but is often independentfrom it; see below, 172. In the Latin texts facing the Greek Donati a and c, I haveindicated in italics the passages where the Greek and Latin texts do not coincide. Also,in the transcription of cs Latin version (Pl), I have italicized words and forms that do

introduction xxiii

readings contained in the manuscripts bearing the most recent ver-sion of the text, which happen to be more complete and correct thanthe earlier versions. On the other hand, I adapted punctuation andorthography to modern usage. In the negative critical apparatus, Idid not mention the mistakes of iotacism, psilosis, and orthographyin each manuscript, unless they represent significant (possible) textualvariants. Versions a and c have been edited with a Latin translationfacing the Greek text. In cs case, I have oered a transcription of theLatin text as it appears in the manuscript, with all its orthographicpeculiarities, its gaps, and its mistakes. In as case, I have not oeredan existing text, but a sort of ideal Ur-Ianua a that may or may nothave been the original of Greek Donatus a, but that functions as apoint of reference and a support for better understanding the Greektext. For this reason, unlike the c-text, for a I have maintained thestandard classical writing (e.g., ae and oe instead of e).28

I wish to express my gratitude to my advisors, Alan Cameron andCarmela Vircillo Franklin, who have supervised my dissertation withpatience and care and provided invaluable intellectual and moral sup-port during the various stages of the process. I am also grateful to theother members of the defense committee, Consuelo Dutschke, KathyEden, and James Zetzel, for their constructive criticism and sugges-tions. Alexandros Alexakis and Roger Bagnall were very helpful at thefirst stage of my research, while Maria Luisa Angrisani and MariaGrazia Jodice gave me the opportunity to present the first results in alecture at the University of Rome La Sapienza, in April 2001. Thequestions raised during that lecture and the other presentations that

not correspond with regular grammar and vocabulary. The notes on pages 513553were not intended as an extensive commentary on the four texts, which would deservea separate volume. With these notes, which supplement the description of the texts inChapter 3 and the critical apparatus, I have tried to explain some diculties in thelanguage of the Latin and Greek Donati, their relationship with other grammaticaltexts, and my choices in establishing the texts. I have devoted more attention to theDonati compositi, which, unlike version a, are not treated individually in Chapter 3 andhave never been objects of specific studies until now.

28 I have used the Latinized form of Greek names that are familiar to modernscholars and readers (e.g., Marcus Musurus, instead of Markos Mousouros).Conversely, I have transliterated less familiar Greek names (e.g., Moschopoulos orMargounios). Greek words have been usually transliterated (e.g., ertmata instead of, but Erotemata as the title of Chrysoloras grammar). The bibliography isupdated to March 2007.

xxiv introduction

I gave on Renaissance Greek grammar have stimulated me to recon-sider many of my assumptions and widen the field of my research.

Two grants from Texas A&M University have allowed me to ex-plore some Italian libraries in order to find the Latin originals of theGreek grammars edited in this book; I wish to thank in particularIsabella Fiorentini of the Biblioteca Trivulziana in Milan for herassistance.

I completed the revision of the manuscript as a fellow of Villa ITatti, the Harvard University Center for Italian Renaissance Studies.Like all those who have had the privilege of working at I Tatti, Ibenefited from a stimulating intellectual environment and invaluableresources.

I am extremely grateful to my colleague Craig Kallendorf and toChristopher Celenza for their continuous encouragement and theirpatience in reading the manuscript and pointing out mistakes andomissions. Giuseppina Magnaldi helped me solve many doubts con-cerning the constitutio textus, while Wolfgang Haase, Steve Oberhel-man, and three anonymous referees were lavish with advice when anarticle based on my dissertation appeared in the International Journal ofthe Classical Tradition (2005). Their contributions have been so numer-ous that it would be impossible to acknowledge all of them as theyoccur. I thank Concetta Bianca, Msgr. Paul Canart, Guglielmo Cav-allo, Mario De Nonno, Elizabeth Fisher, James Hankins, UmbertoLa Torraca, Athanasios Markopoulos, Antonio Martina, and DavidSperanzi for many useful conversations, for a wealth of bibliography,and for often making available to me their works before publication.

I am grateful to William Harris and the editorial board of theColumbia Studies in the Classical Tradition for accepting this work intheir prestigious series, to Krista May for revising my English carefullyand patiently, and to the sta of Brill for solving all kinds of editorialproblems.

My professors and fellow graduate students at Columbia, myfriends in Italy and in the U.S., my colleagues at Texas A&M, andthe other fellows at I Tatti have shared in one way or another the longprocess that has led to this book. Particular thanks go to my family,especially to my mother Paola and my husband Marco, for toleratingmy long absences and making me feel their warm support in all wayspossible.

I consider this book as a point of departure rather than a point ofarrival. Much research still needs to be done on Renaissance Greek

introduction xxv

studies, and I hope that, in the future, new discoveries will add newinformation, widen the perspective, and even challenge the conclu-sions reached so far.

Florence, May 2007

chapter one

THE LATIN DONATUS

The use of Aelius Donatus Ars minor and of other grammars derivedfrom it during the Middle Ages and the Renaissance bears witness tothe eectiveness of Donatus method for teaching elementary Latin.At the same time, the many modifications that the text underwentthroughout the course of the centuriesin particular, its contamina-tion with Priscians Institutiones and other medieval works on gram-mar, as well as the massive insertion of paradigmscorrespond toimportant changes in the teaching methodology used for Latin. Thischapter analyzes the causes and eects of the evolution of the Ars minorinto one of its many new forms, Ianua, which became the most com-mon Latin elementary grammar in the Italian schools of the MiddleAges and the Renaissance. An early version of Ianua constituted theoriginal of Greek Donatus a.

1. Aelius Donatus Artes: A Pedagogical Program

During the Middle Ages, the Latin grammarian Aelius Donatus, al-though pagan, was held in the greatest esteem. He owed much of hisfame to Saint Jerome, who proudly referred to him as his teacher(praeceptor meus Donatus: Contra Rufinum 1. 16, PL 23, 429 A; etc.).1

Donatus, grammaticus urbis Romae, was active between 354 and 363C.E.The name Donatus is especially attested to in Africa, which wasalso the place of origin of other grammarians of that age, includingProbus, Nonius Marcellus, and perhaps Charisius.2

Aelius Donatus grammatical works constitute a corpus (Ars gramma-tica Donati) of four books. The first book, known as Ars minor, containsa synthetic treatment of elementary morphologythe eight parts of

1 Humanists explicitly attributed to Donatus the merit of Jeromes refined literaryeducation. See Brugnoli 1965; Rice 1985, 85; and the passages quoted ibid., 231 n. 5.

2 On Donatus life, see in particular Holtz 1981, 1520; and Kaster 1988, 275278.For a survey of Donatus life, work, and fate, see Holtz 2005.

2 chapter one

speechin a catechistic format. The last three books make up theArs maior, where grammar is treated more extensively. The secondbook (Ars maior 1) deals with the constituent elements of words andphrases: letters, syllables, accents, and punctuation. The third book(Ars maior 2) analyzes the eight parts of speech: noun, pronoun, verb,adverb, participle, conjunction, preposition, and interjection, payingparticular attention to the properties (accidentia), i.e., the changes towhich words are subject when related to other words. The fourthbook (Ars maior 3) focuses on style and contains a description of itsqualities (figures of speech) and defects (barbarism, soloecism, etc.).

The genetic structure of the Ars maior and its transition from sim-ple to complex, from essential to ancillary elements, correspond tothe method followed in classrooms. At the same time, however, theArs maior has the framework of the rhetorical-philosophical treatisesof antiquity: there is, in fact, a clear attempt to codify previous knowl-edge into an all-inclusive, self-contained system.3 The pyramidal struc-ture of each chapter, the division of the matter into classes and sub-classes, the exposition by antitheses and complements, and especiallythe dogmatic tone, make the Ars maior a complex work. Adding tothis complexity, Donatus does not ever declare his aim or his generalplan.4 At any rate, a comparison between the Ars minor and the Arsmaior reveals that Donatus conceived the two works as two stages inthe study of grammar. The Ars minor is a compendium, a textbook foran introductory course; it focuses on the rudiments of the languageand presents them in a form that is easy to memorize. The Ars maiormeets the demands of more advanced students, who require a real ref-erence book in order to accomplish the stylistic ideal of artistic prose.

Although the treatise in three parts was by far the most commonmethod in expounding grammatical material, short grammars like

3 See the observations by Holtz, 1981, 61. Lomanto (1987, 1113.) has focused onthe uniformity in structure of Roman Artes grammaticae, of which DonatusArs maiorrepresents a clear example. Based on the Stoic distinction between (dictio, theword considered per se) and (oratio, the word within a context), most Artes dealfirst with the elements of (de voce, de littera, de syllaba, de dictione, de oratione; deaccentibus, de tonis or de distinctione or de posituris; de rhythmo, de metro, de pedibus), thenwith what concerns (the eight parts of speech), and finally with the elements ofstyle (Latinitas). This tripartite structure seems to have been elaborated by RemmiusPalaemon in the first century C.E.

4 As Holtz remarks (1981, 54): Il ne sagit pas ici dun corps de doctrines qui secre sous nos yeux dans une libre mditation, [] mais dune srie de trs pesantesarmations qui ont pour eet de constituer une sorte dinventaire.

the latin donatus 3

the Ars minor were not unknown in antiquity. The most successfulhandbook for the study of Greek, the " by Diony-sius Thrax, of the second century B.C.E., shows the same character-istics found in Donatus Ars minor: it consists of a set of definitions,sometimes explained with examples. Dionysius " allows us tolocate the origin of the short grammar within the Alexandrian cul-tural milieu.5

In the initial chapters of his Institutio oratoria, Quintilian criticizedthe use of manuals (commentariola) for the teaching of elementary gram-mar (1. 5. 7):

Ex quibus ( scil. grammaticis) si quis erit plane impolitus et vestibulum modo artishuius ingressus, intra haec, quae profitentium commentariolis vulgata sunt, consistet;doctiores multa adicient.

[If a teacher is quite uneducated, and has barely crossed the thresholdof his profession, he will confine himself to the rules commonly knownfrom teachers manuals; a more learned man will be able to add manymore.]6

Quintilian was perhaps aware of the incongruities and obscuritiesthat such elementary books could contain: they synthesizedoftenarbitrarilycomplex issues, summarized longer works, or reportedteachers dictations in classrooms, but they lacked thoroughness andconsistency.

The innovative character of the Ars Donati lies in its combinationof an elementary textbook with a grammatical treatise, thus oeringa complete course of Latin. Moreover, Donatus devotes much morespace to practical issues, such as inflection, than to theory, defini-tions, and concepts. We cannot say to what extent Donatus inno-vation reflects an actual change in the teaching of Latin in classrooms.However, it is interesting to note that other grammarians of the latefourth centuryfor example, Charisius and Diomedesalso reducedthe extent of the traditional theoretical parts in their Artes: a morepragmatic view of grammar seems to have prevailed during this time,and it was perhaps a general custom not restricted to individual schol-ars.

A comparison of the treatment of a topic common to both the Arsminor and the Ars maior, the eight parts of speech, will clarify thispoint. The Ars minor is conceived in question-and-answer format, as

5 On Dionysius ", see below, 106.6 Translation by Russell 2001, 125.

4 chapter one

a dialogue between a teacher and a pupil who repeats his assignmentin the classroom. The colloquial style of the Ars minor is evident also inintroductory expressions typical of the language of classrooms, such asDa declinationem verbi activi (4, p. 593 Holtz); Da adverbia loci; Da temporis,etc. (5, p. 596 Holtz). As noted by Holtz, this pattern, which is asold as the school itself, reflects not only the philosophical (Platonic)dialogue, but also the Roman tradition of a father educating his son.7

The dialogical form was not uncommon in grammatical works, butin the Ars minor it is used extensively and systematically for the firsttime. The Ars minor became a model for later works on grammar: thedialogical form, in fact, was more suitable than the enunciative formfor recognizing and memorizing rules or definitions at the elementarylevel.

The following passages on the noun oer an interesting example ofthe similarities and the dierences between Donatus two Artes:

Ars minor, p. 585 Holtz: (2) Nomen quid est? Pars orationis cum casu corpus autrem proprie communiterve significans. Nomini quot accidunt? Sex. Quae? Qualitas,comparatio, genus, numerus, figura, casus. Qualitas nominum in quo est? Bipertitaest: aut enim unius nomen est et proprium dicitur, aut multorum et appellativum.

[What is the noun? A part of speech with cases, which signifies a personor an object either specifically or generally. How many properties applyto the noun? Six. Which ones? Quality, comparison, gender, number,form, and case. What does the quality of nouns consist in? It is twofold:either the noun belongs to one, and it is called proper; or it belongsto many, and it is called common.]

Ars maior 2, pp. 614 f. Holtz: (2) Nomen est pars orationis cum casu corpusaut rem proprie communiterve significans, proprie ut Roma Tiberis, communiterut urbs flumen. Nomini accidunt sex, qualitas, conparatio, genus, numerus, figura,casus. Nomen unius hominis, appellatio multorum, vocabulum rerum est. Sed modonomina generaliter dicimus. (3) Qualitas nominum bipertita est. Aut enim propriasunt nomina, aut appellativa. Propriorum nominum secundum Latinos quattuorsunt species, praenomen, nomen, cognomen, agnomen, ut Publius Cornelius ScipioAfricanus [] Appellativorum nominum species multae sunt. Alia enim suntcorporalia, ut homo, terra, mare; alia incorporalia, ut pietas, iustitia, dignitas. etc.

[The noun is a part of speech with cases, which signifies a person oran object either specifically or generally: specifically like Rome orTiber, generally like city or river. Six properties apply to thenoun: quality, comparison, gender, number, form, and case. The namerefers to one man, the noun to many, the term to objects. However,

7 Holtz 1981, 100.

the latin donatus 5

we call them nouns only, in a general sense. The quality of nounsis twofold: nouns are either proper or common. In Latin there arefour kinds of proper names: the personal name, the gentile name,the surname, and the nickname, like Publius Cornelius Scipio theAfrican [] There are many kinds of common nouns. Some of themare corporeal (= concrete), like man, land or sea; others areincorporeal (= abstract), like devotion, justice or dignity; etc.]

Apart from some slight dierences, the definitions of the Ars minorbuild up, so to speak, the framework of those of Ars maior 2. Byfirst learning, most likely by heart, the basic rules in the Ars minor,the pupil would acquire the background necessary to understand themore complex classifications of the Ars maior. Moreover, the Ars minorcontains charts of declensions and conjugations, but very few techni-cal details and examples that may distract pupils from absorbing theessential concepts. On the other hand, in book 2 of the Ars maior theauthor has generously supplied examples and details but has omittedcharts. The student who tackled the Ars maior, in fact, was supposed tohave already mastered the Latin language well enough to concentrateon word analysis (); therefore, examples were an indispensabletool for remembering rules and grammatical categories.8

The Ars minor and the Ars maior, the handbook of elementary mor-phology and the encyclopedic treatise on grammar, were not ends inthemselves, but were conceived as preparatory to the study of liter-ature, the final purpose of the teaching of the grammaticus. Donatuscommentaries on Terence and Virgil represent the fulfillment of hiseort to equip late-antique teachers with pedagogical tools useful fortheir task. The first has been handed down with gaps and interpola-tions. Of the second commentary, which is probably the earlier of thetwo, only the biography of Virgil taken from Suetonius De poetis, theletter of dedication to Munatius,9 and the introduction to the Bucolics

8 See Holtz (1981, 110117): isolated words are taken from everyday language andfrom the school environment or refer to the Roman tradition, whereas sentences areusually quotations from Virgil, the auctor mainly studied in schools.

9 This letter (published by C. Hardie in Vitae Virgilianae antiquae, Oxford 1960)is particularly interesting in identifying Donatus aims and methodology. Donatusintended to write a book that might be useful to his colleagues, especially thosewho were just beginning their career; at the same time, he wanted to improve theancient commentaries by taking into account the dierent pedagogical demands ofhis age. For this purpose, as Donatus claims, he had collected a considerable amountof documentation, had carefully selected the material, and had decided to reproducehis sources to the letter. Donatus shows the tendency, typical of Roman scholarlyprose, to juxtapose heterogeneous elements, as well as a remarkable taste for inserting

6 chapter one

are still extant. The rest can be reconstructed through Servius, espe-cially the so-called Servius Danielis or Servius auctus.

Literature had become a science during the Hellenistic age thanksto the accurate studies of the Alexandrian scholars, and had receiveda place of distinction in the Hellenistic school. In its three-level cur-riculum studiorum, , and #the secondstage, that of the grammarian, was in fact devoted to the study ofpoetry. When literary studies were introduced in Rome in the late sec-ond century B.C.E., the grammaticus was assigned the same role in theRoman school system: for Cicero, Horace, Juvenal, and other Latinauthors, the grammarian was first of all a teacher of literature whosemain task was the interpretation of poetry. Donatus himself was bet-ter known for his commentaries than for his grammatical treatises.Throughout the imperial age and up until the final decline of theancient educational system, Virgil and Terence (with Cicero and Sal-lust for prose and, from the late fourth century, also Lucan, Statius,and Juvenal) held a steady place in the Latin curriculum. The fourstages of the teaching of auctores in classrooms remained unchanged:lectio (reading a text, usually aloud), enarratio (exposition and inter-pretation of the content and the language), emendatio (correction andimprovement of the transmitted text), and iudicium (evaluation of theauthor and his work).

From the fourth century B.C.E., with Plato, Aristotle, and the Sto-ics, the origin, development, and functions of language had becomeobjects of study as well.10 Consequently, grammatical studies followedtwo parallel lines, codified in Quintilians distinction between theinterpretation of the authors (historice, exegetice, enarratio auctorum) andthe science of speaking and writing (methodice, horistice, scientia loquendi etscribendi).11

quotations in the text. However, he rarely mentions his own interpretation. See Holtz1981, 29.

10 See in particular Irvine 1994, 2539.11 Cicero (De oratore 1. 187) indicated the main tasks of grammar in the study of

poets (poetarum pertractatio), the knowledge of the contents of literary works (historiarumcognitio), and the explanation of words and their correct pronunciation (verborum intepre-tatio and pronuntiandi sonus). Quintilians division of grammar into the study of correctspeech (recte loquendi scientia) and the interpretation of poets (poetarum enarratio: Inst. or. 1.4. 2; 1. 9. 1, etc.) substantially followed Ciceros assumption. In the fifteenth century,Niccol Perotti still hinted at Quintilians definition in his Rudimenta grammatices: Gram-matica est ars recte loquendi recteque scribendi, scriptorum et poetarum lectionibus observata (quotedby Percival 1981, 237).

the latin donatus 7

The interpretation of literary texts (scientia interpretandi) and the rulesfor the correct use of language in speaking and writing (ratio scribendi etloquendi) were both still relevant for late-antique grammarians. Audax,who lived some time between the fourth and the sixth centuries,specifies this in the question that introduces his grammar book (GL7, 321):

Grammatica quid est? Scientia interpretandi poetas atque historicos et recte scribendiloquendique ratio.

[What is grammar? The science of interpreting the poets and thehistorians and the method of writing and speaking correctly.]

The same concept is expressed in the pompous words that Diomedesaddressed to Athanasius when introducing the three books of his Arsgrammatica (GL 1, 299):

Artem merae Latinitatis puraeque eloquentiae magistram sub incude litteraria doci-liter procudendo formatam humanae sollertiae claritas expolivit.

[The brightness of human ingenuity refined the art of the true Latinstyle (Latinitas) and the master of pure eloquence [i.e., grammar],molded by hammering (it) on the anvil of literature in a teachable way].

Literary sources demonstrate that Latin grammarians taught exactlyas their Greek colleagues: grammars distinction between historice andmethodice applied also to Roman schools. As for methodice, Varro andRemmius Palaemon had adapted to Latin the Greek grammar codi-fied by Dionysius Thrax. The adaptation of Latin to Greek was car-ried to such a point that Roman grammarians felt obliged to finda Latin equivalent for the article in the demonstrative pronoun hic,haec, hoc; to distinguish between subjunctive and optative, even if inLatin the same forms function for both meanings; and to introducean eighth part of speech, the interjection, in order to make the num-ber equivalent to that of their Greek models.12

However, the use of correct language in writing and speaking,which in late antiquity had already gained importance in the commonconcept of grammar,13 gradually came to prevail over the study of

12 See Bonner 1977, 193.13 See, for example, Dositheus definition (1, p. 2 Bonnet): "

$ % & ' & () ( ' *-+ , / Ars grammatica est scientia emendati sermonis in loquendo etscribendo poematumque ac lectionis prudens praeceptum. On this subject, see Sandys 19583, 1.8 f.; Percival 1988, 83 n. 22; and Irvine 1994, 5355.

8 chapter one

literature. The gradual but inexorable changes in the linguistic faciesof the Roman world required that teachers should pay more attentionto the correctness of language than to the reading of the auctores.Thus, the study of literature became subsidiary to that of language.At the same time, instead of being a preliminary stage for the study ofthe auctores, grammar became an end in itself. The attempt to saveLatinitas from inevitable decline made elementary grammar booksincreasingly necessary to meet the demands of students who, becauseof the diusion of the vernaculars, more often than not had to learnLatin as a foreign language. Over the course of time, Donatus Arsminor became the preferred handbook for Latin grammar in schools.

Donatus twofold commitment to language and literature fully con-forms to the Greco-Roman grammatical tradition. However, he canalso be considered as a point of departure: his Ars minor, which wasused to learn the Latin language for about twelve centuries, built upthe foundations of medieval and Renaissance elementary grammar.

2. The Shaping of the Medieval Donatus

The Ars Donati was just one of many works on grammar handeddown from late antiquity to the Middle Ages. Both abbreviated schooltexts, like the Ars minor, and encyclopedic treatises, like the Ars maior,belonged to the so-called Schulgrammatik-type, which was based on themethodical analysis of the morphological elements of the language,the parts of speech. For each part, a definition, a list of properties(accidentia), and a fairly detailed discussion of its functions is given;definitions usually prevail over examples and paradigms.

With more than one thousand pages, Priscians Institutiones grammat-icae is the most monumental work of this kind. A medieval studentof Latin could benefit also from Artes dealing with specific aspects,such as metrics, orthography, and figures of speech: for example, Deorthographia by Caper, Agroecius, and Bede, or De arte metrica by Bede.Less systematic guides to nominal and verbal inflection, known as Reg-ulae and containing lists of paradigms, were available also: for exam-ple, Pseudo-Palaemons Regulae, Probus Catholica, and Priscians Insti-tutio de nomine pronomine et verbo.14 These categories, however, should be

14 See Irvine 1994, 57; and Law 1997, 54.

the latin donatus 9

considered in a very broad sense because few works on grammar fallwithin only one of them: every kind of variation and cross-influencewas possible, depending on the goals of the grammarian and thedemands of his audience.

This point explains why Donatus Ars, the most successful textbookever written in the history of Western education,15 underwent anendless process of transformations and adaptations from late antiquityto the sixteenth century. Like many other ancient grammatical texts,Donatus work survived a transformation of the culture that hadproduced it. Christian communities, in fact, realized very quickly thatLatinitas and emendatiowriting correctly and establishing a correcttextwere indispensable to the preservation of a written traditionand the continuation of their textual culture. In order to achievetheir goals, Christian scholars had to rely on a well-established set ofrules and on the authority of accepted literary models.16 The demandfor a normative grammar became even more compelling after thesixth century, when vernaculars were gradually replacing Latin ineveryday usage: therefore, access to the Latin texts of ecclesiasticaland monastic culture had to be assured through the systematic studyof Latin grammar. Donatus pedagogy, based on a course of Latin ontwo levels, was seen as particularly suitable in satisfying this urgent,practical need.

It was Servius, grammaticus urbis Romae at the beginning of the fifthcentury, who mainly contributed to making the Ars Donati an authori-tative text. His commentaries on Donatus Ars minor and Ars maior werewidespread. As in his commentary on Virgil, Servius treated Donatusas an auctor, whose statements were subject to comments by teachersand were memorized by students. Servius often used the formula utDonatus dicit to reinforce his assertions.17

15 Irvine 1994, 58. See Murphy 1974, 139.16 Irvine 1994, 74. A strong and lasting opposition to the use of pagan grammari-

ans, however, arose among the most conservative Christians. Law (1994a, 100) quotesa short text, handed down by two ninth-century manuscripts, which bears witness tothe debate between those who asserted the usefulness of ancient grammar and thosewho considered it as a vessel of the wine of error which lying teachers poured out(vinum erroris quod [] propinaverunt magistri mendaces: a quotation attributed to Gregorythe Great).

17 On Servius and Sergius, see Holtz 1981, 223., 428429; and Kaster 1988,169196, 356359, 429430. Two versions of Servius commentary (GL 4, 405428and 428448) have been handed down by an eighth-century manuscript (Paris, Bib-liothque Nationale, MS. lat. 7530), severely corrupted in spite of its antiquity. The

10 chapter one

From the fifth to the seventh centuries, Donatus work had severalimitators and commentators: for example, Cledonius, a Latin gram-marian in Constantinople, who wrote the first commentary on Dona-tus equipped with lemmata (thus bearing witness to the conditionof Donatus text that was circulating at Constantinople at the time);Servius and the mysterious Sergius or Pseudo-Cassiodorus; Pompeius,who taught in Africa; and Julian of Toledo, a bishop of VisigothicSpain.18 Writers and grammarians from the sixth century onwardsdemonstrate that Donatus had risen to the level of a symbol.19 Pri-scian, who taught Latin in Constantinople during the sixth century,praised Donatus and made extensive use of his work, even if hisattitude toward the Ars Donati was cautious and realistic rather thanblindly laudatory.20 Cassiodorus, Gregory the Great, Isidore of Seville,and Bede held Donatus work in great esteem. In most cases, however,they relied not on the original text but on commentaries that ampli-fied the authentic Donatan material through more or less arbitraryadditions.21

Between the fifth and the sixth centuries, ancient grammar wasChristianized. Donatus work also underwent the same treatment:words taken from Christian texts or lexica and passages from theScriptures replaced, or were added to, the examples taken from Vir-

first editor, Jenson (Venice, ca. 1476), attributed the first to Marius Servius Honora-tus and the second to Sergius grammaticus. A Servius plenior, a more extensiveversion of a commentary on Donatus Ars with a massive use of Virgil to explainDonatus rules, is also attributed to a Sergius, which is most likely just a wrong read-ing of the name Servius; see Heinrich Keils introduction to his edition in GL 4,LII. This Servius plenior was used by Cledonius, Pompeius, and the anonymousauthor of the two books of Explanationes in Donatum (GL 4, 486565). Holtz (1981, 429),in turn, attributed the two books to two dierent authors. On the vexatae quaestionesof the origin and authorship of the Explanationes, see De Paolis 2000. A fragmentarytreatise (Sergii fragmenta Bobiensia: GL 7, 537541) and a commentary on the Ars minor(GL 8, 143148) have also been attributed to Sergius: see Holtz 1981, 429.

18 On Cledonius (GL 5, 779), see Holtz 1981, 235, 429431. On Pompeius (GL5, 95132), ibid., 236 f., 431, and the comprehensive study by Kaster 1988, 139168.Excerpts from Julian of Toledos work have been edited in GL 5, 317328, and 8,CCIVCCXXXIX.

19 For example, Boethius (Categ., PL 64, 257 D, 260 A) considered Donatus andAristarchus as the undisputed authorities on Latin and Greek grammar respectively.See Holtz 1981, 238.

20 Priscians works are edited in volumes 2 and 3 of GL. On Priscians use ofDonatus Ars, see Holtz 1981, 239244, 425.

21 Holtz 1981, 245259, 318 f.

the latin donatus 11

gil and from the Roman pagan tradition.22 This Donatus Christianusbecame the basic textbook in the cloisters of eighth-century Ireland,where learning Latin was a compelling necessity for religious prac-tice.

Donatus, however, had conceived his Ars for fourth-century nativespeakers of Latin. With his workespecially the Ars maiorhe hadaimed to impart not knowledge of the Latin language, which wastaken for granted in native speakers, but a taxonomy and classificationof words necessary to define the style of the auctores. Therefore, inthe Ars minor Donatus had oered the conjugation of only one verb,lego, and the declension of five nouns or adjectives, one for eachgender traditionally recognized: hic magister (genus masculinum), haec musa(genus femininum), hoc scamnum (genus neutrum), hic et haec sacerdos (genuscommune), and hic et haec et hoc felix (genus omne). This certainly was notenough to acquaint a non-native speaker with the complex verbal andnominal inflections of Latin. Donatus grammar, in fact, does not dealwith Latin morphology, but restricts the treatment of the language tobasic definitions and to very few examples of the main grammaticalcategories.23

In order to adapt this text to the needs of Christian students forwhom Latin was a foreign language, Irish grammarians resorted toseveral strategies: from the contamination of the Ars minor with BookTwo of the Ars maior to the improvement of Donatus text by theworks of other grammarians, such as Charisius, Diomedes, Probus,Consentius, and Priscian, probably made accessible through an influxof manuscripts from the European continent. This is the case, forexample, in Virgilius Maro and the source of the seventh-century ArsMalsachani. Commentators also continued their activity, focusing onthe Ars maior: the anonymous Ars Laureshamensis, the Ars Ambrosiana,and Quae sunt quae of the seventh-eighth centuries.24 The Ars minor

22 For example, in Donatus Christianus the nouns ecclesia and templum and the adjec-tive fidelis replaced Donatus musa, scamnum, and sacerdos as examples of nominal gen-ders and as paradigms of inflection; fructus and species were added as paradigms offourth- and fifth-declension nouns; the question (part of speech) quid est? becameQuid est (part of speech)?. See Law 1994, 7374.

23 See Percival 1988, 72.24 Vivien Law has devoted several important studies to the problems connected

with the preservation, transmission, and use of Latin grammars in medieval Irelandand England (1994; 1997, 2849, 91123; etc.). On the reception of Donatus in Ire-land, see Holtz 1981, 272320. On the treatise Quae sunt quae, see Munzis observationsin his recent edition (2004, 915; text and commentary on pp. 1766).

12 chapter one

gradually prevailed over other short grammar books, such as the man-uals by Asper, Dositheus, and Scaurus, or the Ars breviata attributed toAugustine. Even more important, however, was the fact that the Arsminor was integrated with Regulae-type grammar books. Irish gram-marians of the seventh and eighth centuries (e.g., Tatwine and Boni-face) concentrated their eorts on improving the content and struc-ture of the Ars minor by adapting it to the needs of their pupils: theyabandoned the semantic and derivational criteria of ancient grammarand reduced the space traditionally devoted to definitions in favor ofa more thorough description of accidence and morphological phe-nomena.25 The merging of Schulgrammatik with Regulae created a modelof elementary foreign-language grammar unprecedented in antiquity.The creation of a grammar for foreigners seems to be peculiar tothe insular grammarians of the seventh and eighth centuries. Worksconceived by teachers who taught Latin to Greek students, such asEutyches, Phocas, and Priscian, although based on a comparativeapproach to the Latin language, in fact were written according to theusual formal and descriptive criteria.26

The earliest surviving example is the Ars Asporii. Although stilldefective in its disposition and organization of contents, we can con-sider it as a first experimental grammar for non-Latins: written per-haps in Gaul at the end of the sixth century, the Ars Asporii consists ofa contamination between the Ars minor and the second book of the Arsmaior. Donatus Ars is reproduced in its Christianized form and witha significant increase in the number of examples quoted.27 TypicalChristian contents and methods are displayed by another anonymous

25 Law 1997, 104.; see also Contreni 1986 (1992). Word separation, graphicallyvisual forms, and modification of the word order were some of the devices throughwhich Insular grammarians made Latin grammar more easily accessible to pupils; seeSaenger 1997, 8390.

26 See Law 1986; 1997, 58, 73.27 A version of the Ars Asporii, or Ars Donati exposita ab Aspero, was published in GL 8,

3961 from MS. lat. 207 of the Brgerbibliothek in Bern. On the transmission of thetext and its versions, see Holtz 1981, 432. The author keeps and expands Donatusdistributional classifications of nouns by gender (GL 8, 40: Genera nominum sunt quattuor,masculinum, femininum, neutrum, commune. Masculinum ut hic iustus, femininum ut haec ecclesia,neutrum ut hoc ieiunium, commune duobus generibus ut hic et haec finis vel hic et haec sacerdos vel hicet haec dies. Est praeterea trium generum, quod omne dicitur, ut hic et haec et hoc ingens felix prudens.Est epicoenon nomen, id est promiscuum, ut vultur ardea accipiter vel aquila, etc.). Christianparadigms are quoted for the verbal system (4951): ieiunio, oro, vigilo, praedico, supplico,commendo, etc. See Law 1994a, 91 and 1997, 102 f.

the latin donatus 13

elementary grammar in question-and-answer format, the fragmen-tary Ars Bernensis, which was composed in Ireland during the eighthcentury. As in the other Artes of that age, Donatus is considered anabsolute authority: his words are constantly quoted in the form oflemmata, paraphrased, and explained, like the Scriptures, in a literalsense. The anonymous author of the Ars Bernensis contaminated Dona-tus Ars minor and the second book of the Ars maior with Pompeius,Priscian, Sacerdos, Charisius, and Isidore, following a procedure thatreminds us of the catenae of the early commentators of the Bible.28

The adoption of the formal classifications of nouns and verbs intofive declensions and four conjugations, used by Priscian in his Institutiode nomine pronomine et verbo, represented a significant improvement: itprovided the framework for a description of Latin morphology thatcomplemented Donatus definitions.29 In turn, this new PriscianicDonatus was supplemented by commentaries, which often precededthe text in manuscripts, and by tools aimed at making the gram-mar book suitable for beginning students. They included, for instance,exercises and lists of paradigms taken from the stock material avail-able to teachers and often circulating separately in schools, such asDeclinationes nominum, i.e., lists of nouns declined without a connectingtext, composed in England perhaps during the seventh century.30 ThisInsular elementary grammar was, therefore, an elementary descrip-tive grammar, which Vivien Law has defined as a succinct systematicexposition of Latin grammar in which morphology takes first place.31

It originated in England during the seventh century and built up thefoundations of later elaborations of the Ars minor.

28 Text in GL 8, 64142. Another catena grammaticalis is the so-called Donatus orti-graphus (edited by J. Chittenden, CCCM 40 D, Turnhout 1982), a dialogue between ateacher and a pupil, in which Priscian and other grammarians are also used exten-sively. See Holtz 1977, 70; 1981, 434436.

29 Cf. 1. 1, GL 3, 443: Omnia nomina, quibus Latina utitur eloquentia, quinque declinationibusflectuntur; 3. 38, 450: Omnia verba [] habent coniugationes quattuor. On the (probable)origin of the traditional classification of Latin morphology from Varro and RemmiusPalaemon, see Barwick 1922, 236.

30 See Holtz 1981, 341, 344348; Law 1994, 7475 and 1997, 75., 104; and Colom-bat 1999, 69 (on Vestibulum, a list of three thousand Latin words in 427 sentences,which usually supplemented Ianua). Declinationes nominum were copied, for example,after Donatus Ars maior 2 in the third part of Paris, Bibliothque Nationale, MS. lat.5570, dating from the beginning of the twelfth century. According to Holtz (1981, 419),they are a free exploitation, in catechistic form (at least at the beginning of the text)of Priscians Institutio de nomine pronomine et verbo (my translation).

31 Law 1984 and 1997, 78.

14 chapter one

The British-Irish case was not isolated: in the rest of Europe, Dona-tus Ars also underwent transformations and adaptations. Among themany variants of the Ars Donati handed down to us, Louis Holtz hasrecognized another main tradition: the Visigothic, which is of Span-ish origin. In Italy, the Visigothic and the Irish traditions merged inthe middle of the eighth century, perhaps in the Irish monastic foun-dation of Bobbio.32

Whereas the transmission of the Ars minor, constantly used inschools, was continuous and uniform, the three books of the Ars maiorhad a more varied fate. From the time of Servius, because of twobooks devoted to the morphology of the parts of speech (the Arsminor and the second book of the Ars maior), Donatus corpus was oftenhanded down in two independent parts: Ars minor and Ars maior 1,and Ars maior 2 and 3. Book Two of the Ars maior competed with theArs minor, which was considered too elementary, especially during theso-called Carolingian Renaissance, when an elevation in culture alsocaused a change in pedagogy.

An increase in the number of manuscripts of classical authors (e.g.,Virgil, Terence, and Horace) copied in the Carolingian age attests to arenewed interest in ancient literature and, most probably, to a revivalof the practice of reading poetry in schools. The teaching of grammar,however, remained generally restricted to the doctrine of the partsof speech.33 Elementary morphology, based on Donatus, built up thecore of the most widespread grammatical works of that age: the tractsby Paul the Deacon,34 Peter of Pisa,35 and Smaragdus.36

Because of its simplicity, the Ars minor became in later centuries themost common elementary schoolbook for Latin. From the thirteenth

32 See Holtz 1981, 446499.33 On the conservatism of Carolingian education, see Munzi 2000. For example,

the parts of speech constitute the main theme of Alcuins De grammatica (PL 101, 848902; see below, n. 39), a dialogue between a teacher and two students, one Frankishand one Saxon. Alcuin, although trying to insert the grammatical doctrine within awider context, did not go very far beyond elementary grammar. See Holtz 1989, 155.

34 In his Ars Donati quam Paulus Diaconus exposuit (re-edited by M.F. Bua Giolito,Genova 1997), Paul contaminated an interpolated version of Donatus Ars minorwith a complete Declinationes nominum-type treatise and other shorter supplements(on formae casuales, monosyllabic nouns, compound pronouns, impersonal verbs). SeeHoltz 1989, 155; Giolitos introduction, 2331; and Law 1994, 73. and 1997, 134.

35 A partial edition in GL 8, 159171.36 Liber in partibus Donati, ed. by B. Lfstedt, L. Holtz, and A. Kibre, CCCM 68,

Turnhout 1986.

the latin donatus 15

century on, an increasing number of manuscripts handed down theArs minor with other more advanced grammars, such as the Doctrinaleby Alexander of Villedieu, a verse grammar through which studentswould complete their study of Latin grammar. Commentaries byIrish and Carolingian grammarians (Murethach, Sedulius Scottus,and Remigius) replaced those transmitted from late antiquity.37

Carolingian teachers also introduced a significant innovation inforeign-language grammar by supplementing the Ars minor with a newtool, the parsing grammar, in which the descriptions of the character-istics and properties of each morphological element were introducedby the analysis of a headword. The closest model was Priscians Par-titiones duodecim versuum Aeneidos principalium, a detailed analysis of everyword of the first line of each of the twelve books of the Aeneid. Priscian,who lived and worked in Constantinople, may have been influenced,in turn, by the methods used in Greek and Byzantine schools.38

The return to the traditional practice of parsing (i.e., identifyingand labeling grammatical forms) was a consequence of the rediscov-ery of Priscian promoted by the Carolingian grammarians and influ-enced the later development of medieval grammar and teaching.39

The most important elementary grammar books used in the Mid-dle Ages, Remigius and Ianua, were parsing grammars resulting from acontamination between Donatus Ars minor and Priscians Institutiones.Parsing grammar was an open form that easily permitted varia-tions and insertions of new material: this flexibility was its key to suc-cess.40 On the other hand, the imitation of Priscians Partitiones causeda revival of the dialogical form, which had been used already in theArs minor and perhaps corresponded to the actual practice that wasbeing followed in schools. In his dialogue De grammatica, Alcuin had

37 See Holtz 1981, 505 and n. 23; Holtz 19891990; and Law 1997, 60 f., 144146.38 On Byzantine schools, see below, 106.39 Quintilian (Inst. or. 1. 8. 13) recommended the practice of parsing in the teaching

of grammar. The rediscovery of Priscian, which corresponded to the higher stan-dard of Carolingian scholarship, was promoted by Alcuin of York, abbot of Tours,who lived between the eighth and ninth centuries. By means of his Dialogus Franconiset Saxonis de octo partibus orationis and his abstracts of the final two books of Priscian,Alcuin attracted the attention of his contemporaries to Priscians main work, theeighteen books of Institutiones grammaticae. The transmission and use of Priscians Insti-tutio de nomine pronomine et verbo had continued without interruption from late antiquityonwards, and in Mediterranean Europe the knowledge of Priscians whole corpus hadnever ceased. See Law 1997, 61 f., 83; and Holtz 2000.

40 Law 1997, 85.

16 chapter one

tried to set the content of Priscians Institutiones grammaticae in question-and-answer format. However, the earliest proper example was oeredby Peter of Pisa, who taught grammar to Charlemagne: in Peters Ars,questions such as What is pater? and To what word class does itbelong? introduced an analysis of each part of speech. This formatwas later maintained and improved to such an extent that it becamea distinguishing feature of these handbooks, for the most part anony-mous, and usually known by their initial question: Codex quae pars, Doc-tus quae pars, Quid est doctus,41 Magnus quae vox, or, in the more distinctlyChristianized versions, Anima quae pars and Propheta quae pars. At theend of the ninth century, Usuard of St. Germain introduced in hisgrammar the first declension with the parsing of poeta. Poeta quae parsest? was also the introductory question of the twelfth-century Ianua,which, together with the contemporary Dominus quae pars and Magi-ster quae pars, was one of the most widespread grammar books of theMiddle Ages.42

Therefore, together with Donatus, Priscian gradually became anauthority and an object of commentaries.43 In particular, the Scottiperegrini Sedulius and John Eriugena commented on Donatus A