Early Latin: Constructs, Diversity, Reception 9781108476584, 9781108671132

Citation preview

EARLY LATIN

This is the most detailed and comprehensive study to date of early Latin language, literary and non-literary, featuring twenty-nine chapters by an international team of scholars. ‘Early Latin’ is interpreted liberally as extending from the period of early inscriptions through to the first quarter of the first century bc. Classical Latin features significantly in the volume, although in a restricted sense. In the classical period there were writers who imitated the Latin of an earlier age, and there were also interpreters of ‘early Latin’. Later authors and views on early Latin language are also examined as some of these are relevant to the establishment of the text of earlier writers. A major aim of the book is to define linguistic features of different literary genres, and to address problems such as the limits of periodisation and the definition of the very concept of ‘early Latin’. J. N. A da ms CBE, FBA was the author of many books and articles on the Latin language, including The Latin Sexual Vocabulary (1982) and Bilingualism and the Latin Language (2003). He was awarded the Kenyon Medal of the British Academy in 2009. A n na Ch a houd, F TCD  is Professor of Latin at Trinity College Dublin. She is the author of C. Lucili Reliquiarum Concordantiae (1998) and various articles on Republican Latin and on the grammatical tradition, and has co-edited, with E. Dickey, Colloquial and Literary Latin (Cambridge, 2010). Giuseppe Pezzini is Fellow and Tutor in Latin at Corpus Christi College, Oxford. He has published extensively on ‘early Latin’ language and literature, Roman comedy, ancient philosophy of language and fiction theory, including Terence and the Verb to-be in Latin (2015) and a forthcoming commentary on Terence’s Heautontimorumenos.

Published online by Cambridge University Press

E A R LY L AT I N Constructs, Diversity, Reception Edited by J. N. A DA MS All Souls College, Oxford

A N NA CH A HOU D Trinity College, Dublin

GI USEPPE PE Z ZI N I Corpus Christi College, Oxford With the assistance of

CH A R L I E K ER R IG A N Trinity College, Dublin

Published online by Cambridge University Press

University Printing House, Cambridge CB2 8BS, United Kingdom One Liberty Plaza, 20th Floor, New York, N Y 10006, US A 477 Williamstown Road, Port Melbourne, V IC 3207, Australia 314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre, New Delhi – 110025, India 103 Penang Road, #05–06/07, Visioncrest Commercial, Singapore 238467 Cambridge University Press is part of the University of Cambridge. It furthers the University’s mission by disseminating knowledge in the pursuit of education, learning, and research at the highest international levels of excellence. www.cambridge.org Information on this title: www.cambridge.org/9781108476584 DOI: 10.1017/9781108671132 © Cambridge University Press 2023 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2023 Printed in the United Kingdom by TJ Books Limited, Padstow Cornwall A catalogue record for this publication is available from the British Library. A Cataloging-in-Publication data record for this book is available from the Library of Congress ISBN 978-1-108-47658-4 Hardback Cambridge University Press has no responsibility for the persistence or accuracy of URLs for external or third-party internet websites referred to in this publication and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

Published online by Cambridge University Press

D. M. J. N. A.

Published online by Cambridge University Press

Contents

List of Illustrationspage x List of Tablesxi List of Contributorsxiii Acknowledgementsxviii List of Abbreviationsxix 1 Introduction: What Is ‘Early Latin’? Giuseppe Pezzini and Anna Chahoud

1

Pa r t I   T h e E pig r a p h ic M at e r i a l  2 Alphabet, Epigraphy and Literacy in Central Italy in the 7th to 5th Centuries bc

15

3 Identifying Latin in Early Inscriptions

41

4 The Egadi Rostra: A Linguistic Analysis

63

Rex Wallace

Simona Marchesini Wolfgang de Melo

Pa r t I I   Dr a m a  5

Metre

Wolfgang de Melo and Giuseppe Pezzini

79

6 Morphology and Syntax

100

7 Support Verb Constructions in Plautus and Terence

118

Wolfgang de Melo José Miguel Baños

vii

Published online by Cambridge University Press

viii

Contents

8 Ecquis in ‘Early Latin’: Aspects of Questions

138

9 Indirect Questions in ‘Early Latin’

157

Colette Bodelot

Peter Barrios-Lech

10 Latin edepol ‘by Pollux!’: Background of a Latin Aduerbium Iuratiuum206 Brent Vine

11 Early Latin Lexicon in Terence (and Plautus)

221

12 A Comparison of the Language of Tragedy and Comedy in Early Latin Drama

251

Giuseppe Pezzini

Robert Maltby

P a r t I I I  O t h e r G e n r e s a n d F r a g m e n t a r y Author s 13 The Language of Early Latin Epic

275

14 Early Latin Prayers and Aspects of Coordination

292

Sander M. Goldberg

J. N. Adams and Veronika Nikitina

15 Some Syntactic Features of Latin Legal Texts

311

16 Repetition in the Fragmentary Orators: From Cato to C. Gracchus

327

17 How ‘Early Latin’ Is Lucilius?

351

18 Greek Influences on Cato’s Latin

373

19 Greek Loanwords in ‘Early Latin’

386

Olga Spevak

Christa Gray

Anna Chahoud

Neil O’Sullivan James Clackson

Pa r t I V   R e c e p t ion  20 ‘Early Latin’ and the Fragments of Atellane Comedy Costas Panayotakis

Published online by Cambridge University Press

411

Contents

ix

21 Lucretius and ‘Early Latin’

434

22 Cicero and Early Dramatic Latin

453

23 Early Latin Texts in Livy

466

24 Pliny Rewrites Cato

485

25 Gellius’ Appreciation and Understanding of ‘Early Latin’

511

26 Views on ‘Early Latin’ in Grammatical Texts

527

27 Nonius Marcellus and the Shape of ‘Early Latin’

549

28 ‘Early Latin’ to Neo-Latin: Festus and Scaliger

563

29 Conclusions: ‘Early Latin’ as a Concept

582

Barnaby Taylor

Gesine Manuwald John Briscoe

Cynthia Damon

Leofranc Holford-Strevens Alessandro Garcea Jarrett Welsh

Anna Chahoud J. N. Adams

Bibliography597 Index Verborum638 Index of Non-Latin Words643 Index Locorum Potiorum644 Subject Index650

Published online by Cambridge University Press

Illustrations

Figures 2.1 The Fibula Praenestina (CIL 12.3) page 20 2.2  The Forum inscription, side 2 (CIL 12.1)21 2.3  The Forum inscription, sides 1–4 (CIL 12.1)22 2.4  The Duenos inscription, top (CIL 12.4)22 2.5  The Duenos inscription, drawing, top (CIL 12.4)23 2.6a  Abecedarium of Lanuvium 31 2.6b  Abecedarium of Lanuvium, drawing 31 2.7  The Madonetta inscription (CIL 12.2833)35 2.8 The Lapis Satricanus (CIL 12.2832a)35 2.9  The Tibur inscription, drawing (CIL 12.2658)35 2.10  The Tibur inscription (CIL 12.2658)36 3.1  Drawing of the scene of the cista from Praeneste CIL 12.56057 3.2  Etruscan mirror CIL 12.54759 3.3  Detail of the inscription CIL 12.54760 3.4  Detail of the inscription CIL 12.54760 7.1 Frequency of support verb construction from Plautus to the Historia Augusta122 7.2  Percentual frequency of the four most generic support verbs  126 7.3 Percentual frequency of esse over the totality of support verb constructions128 7.4 Percentual frequency of facere over the totality of support verb constructions130

Map 2.1 The languages of pre-Roman Italy

x

Published online by Cambridge University Press

16

Tables

2.1 Early Latin inscriptions page 19 2.2 Spelling of velars in early Latin inscriptions 24 2.3 Letterforms in early Latin inscriptions 32 6.1 Present subjunctive siem/sim in Plautus and Terence 102 6.2 Medio-passive infinitive doublets 104 6.3 Percentage of medio-passive infinitive doublets  104 6.4 Imperfect and simple future forms in Plautus 106 6.5 sciebam/scibam in Plautus and Terence 107 6.6 sciam/scibo in Plautus and Terence 107 6.7 Absence of subject accusatives 114 6.8 Absence of subordinator ut115 7.1 Frequency of support verb constructions from Plautus to the Historia Augusta121 7.2 Total number of support verb constructions in Plautus (Aul. Mil. Poen.) and Terence (Ad. And. Ph.)128 7.3 Frequency of facio and do in support verb constructions 132 8.1 ecqu- in Roman comedy 141 9.1 Corpus of early Latin indirect questions: number of indirect questions by genre and register 161 9.2 Indirect questions in Plautus, by type (+indicative or +subjunctive)165 9.3 Indirect questions in Terence, by type (+indicative, +subjunctive, +infinitive)165 9.4 Indicatives in indirect questions, by text-type 166 9.5 Anticipation and prolepsis in the comedies of Plautus 171 9.6 Anticipation and prolepsis in the comedies of Terence 171 9.7 Placement of matrix-clause verb and mood of subordinate clause in Plautus 172

xi

Published online by Cambridge University Press

xii

List of Tables

9.8 Placement of matrix-clause verb and mood of subordinate clause in Terence 172 9.9 Matrix verbs in the imperative: fragments of early Latin tragedy 179 9.10  uide with indirect questions in Roman comedy 186 9.11  scin used in pre-announcements and pre-threats in Plautus and Terence 194 11.1 Typology and distribution of potentially ‘early’ words in Terence 224 11.2  Patterns of word-formation in Terence 225 11.3  Typology of Plautine ‘neologisms’ 241 12.1  Compound nouns and adjectives in early Latin authors 259 14.1  Forms of coordination: statistics 294 16.1  The basic forms of repetition according to Wills 330 17.1 Frequency and distribution of sigmatic ecthlipsis in Lucilius’ hexameters369 19.1 Distribution of Greek loanwords in ‘early Latin’ across grammatical classes 398 19.2  Semantic fields of Greek loanwords in ‘early Latin’ 400 20.1  Distribution of fragments of Atellana in the source-authors 417 20.2  Number of citations of Pomponius and Novius in Nonius 419 20.3  Typology of Nonius’ quotations of Pomponius and Novius 420 21.1  Passive infinitive in -ier in Lucretius’ and Cicero’s hexameters 436 21.2  Disyllabic genitive -ai in Lucretius’ and Cicero’s hexameters 437 26.1  Gender variation in Caper/Priscian with quotations 537 26.2 Active/passive variation in Caper/Priscian (Flobert 1975, with additions) 541

Published online by Cambridge University Press

Contributors

J. N. A da ms (†2021) was Research Fellow of All Souls College, Oxford, and previously held positions at the University of Manchester and Reading. He wrote numerous works on aspects of the history of the Latin language, such as The Latin Sexual Vocabulary (1982), Bilingualism and the Latin Language (2003), Social Variation and the Latin Language (2013), Asyndeton and its Interpretation in Latin Literature (2021). José Miguel Ba ños is Professor of Latin Philology at the Complutense University of Madrid. Among his research interests are Latin syntax and semantics, the figure and works of Cicero (particularly speeches and letters), hispanic humanism and missionary linguistics. He has coordinated the Sintaxis del latín clásico (2009) and the Sintaxis Latina (2021). Peter Ba r r ios-L ech is Associate Professor of Classics at the University of Massachusetts Boston. His published work has taken a pragmatic perspective on the Latin and Greek languages. He has also written articles on puns as well as allusions to Greek literature in Latin literature. Col et te Bodelot, a member of the Academia Europaea, is Professor Emeritus of Latin Language and Literature at the University of Clermont Auvergne. Her work is mainly concerned with Latin ­syntax: indirect questions, substantive clauses, correlative structure, macro-syntax, grammaticalisation of subordinators and valency ­ of adjectives. The corpora she explores range from early Latin to late Latin. She is interested diachronically in the evolution of the Latin language and synchronically in the stylistics and pragmatics of ­multi-genre literary writings. John Br iscoe  is an Honorary Research Fellow at the University of Manchester and a Member of the Academia Europaea. He has ­published four volumes of commentary on Livy books 31–45 (1973– 2012), as well as Teubner editions of those books (1986–1991), an Oxford xiii

Published online by Cambridge University Press

xiv

List of Contributors

Classical Text edition of books 21–25 (2016) and a Teubner edition of Valerius Maximus (1998); his most recent book is a commentary on Valerius Maximus book 8 (2019). A nna Ch a houd  is Professor of Latin at Trinity College Dublin. She has published on Republican Latin language and literature, and the grammatical tradition. Major ongoing projects include the first English-language commentary on Lucilius (Cambridge University ­ Press) and an edition of fragmentary satire and popular verse for the Fragmentary Republican Latin series in the Loeb Classical Library. Ja mes Cl ack son is Professor of Comparative Philology at the University of Cambridge. He is a Fellow and Director of Studies in Classics at Jesus College, Cambridge. He has published widely on the history of the Greek and Latin languages, the comparative philology of IndoEuropean languages and the reconstruction of Proto-Indo-European. C y nt hi a Da mon  is a Professor of Classical Studies, emerita, at the University of Pennsylvania. Her books include The Mask of the Parasite (1997), commentaries on and translations of Tacitus (Histories 1 (2003), Agricola (2017); Annals (2013)), texts of Caesar’s Civil War (2015, 2016) and, with Will Batstone, Caesar’s Civil War (2006). Current projects include a critical edition of the Bellum Alexandrinum for the Library of Digital Latin Texts, a new Loeb edition of Caesar’s Gallic War, and studies of Pliny’s Natural History and its reception. Wolfg a ng de Melo is Professor of Classical Philology at Oxford. His first book, The Early Latin Verb System (2007), was followed by the Loeb edition of Plautus in five volumes (2011–13). His most recent work is a critical edition of Varro’s De lingua Latina, accompanied by an introduction, a translation, and a commentary (2019). A l essa ndro G a rce a  is Full Professor of Latin Language and Literature at Sorbonne University and honorary member of the Institut Universitaire de France. His main works include Gellio e la dialettica (2000), Cicerone in esilio: L’epistolario e le passioni (2005); Caesar’s De analogia (2012); Tout César (2020). He coordinates the ‘CGL – Corpus Grammaticorum Latinorum’ and ‘GRADIS – Grammatici disiecti’ website projects. Sa nder M. Goldberg,  Distinguished Research Professor at the University of California, Los Angeles, is co-editor with University College London’s Gesine Manuwald of a new edition of Ennius for

Published online by Cambridge University Press

List of Contributors

xv

the Loeb Classical Library (2018). Other publications include Epic in Republican Rome (1995), Constructing Literature in the Roman Republic (2005), and an edition of Terence’s Hecyra in the Cambridge Greek and Latin Classics series (2013). Chr ista Gr ay  is Assistant Professor in Classics at the University of Reading. She has co-edited two volumes on Roman oratory for the Fragments of the Republican Roman Orators project at the University of Glasgow. She is the author of Jerome, Vita Malchi: Introduction, Text, Translation, and Commentary (2015) and co-editor of The Hagiographical Experiment: Developing Discourses of Sainthood (2020). L eofr a nc Holfor d-Str ev ens,  before his retirement in 2011, was Consultant Scholar–Editor. He has published on classical, musicological, and computistic topics, but especially on Aulus Gellius, the subject of his monograph Aulus Gellius: An Antonine Scholar and his Achievement (2003). His most recent publications are an Oxford Classical Text of Gellius and a companion volume of Gelliana (2019). Ch a r lie K er r ig a n is a Research Fellow in the Department of Classics at Trinity College, Dublin, where he teaches Latin language and literature. His study of Virgil’s Georgics – Virgil’s Map – was published in 2020, and his Latin blog Confabulations can be found online as part of Trinity College Dublin’s Living Latin project. He has been responsible for the production of the three indices in the volume. Robert M a ltby is Emeritus Professor of Latin Philology at the University of Leeds. His main research interests are Latin drama and elegy, Roman commentators and grammarians, and etymological play in Latin. His main publications include A Selection of Latin Love Elegy (1980), A Lexicon of Ancient Latin Etymologies (1991), Tibullus Elegies: Text, Introduction and Commentary (2002), Terence: Phormio (2012) and Corpus Tibullianum Book III: Introduction, Text, Translation and Commentary (2021). Gesine M a nu wa ld is Professor of Latin at University College London. Her research interests cover early Roman drama, Roman oratory, Roman epic and the reception of antiquity, especially in neo-Latin literature. She has published widely on Ennius, Pacuvius, fabula praetexta, Roman comedy and Republican theatre more generally. Simona M a rchesini  is a historical linguist with expertise in the languages of ancient Italy and anthropology of writing. Following her doctorate in Tübingen (Germany), she was research fellow at the Scuola

Published online by Cambridge University Press

xvi

List of Contributors

Normale Superiore (Pisa) and at the University of Verona. She lectured at Tübingen and Verona and in 2009 she founded Alteritas (Verona, Italy), a multidisciplinary research institution devoted to the study of interactions among peoples in time and space. She has authored eight monographs on the languages of pre-Roman Italy and more than sixty papers, and edited six miscellanies. Among her books are Prosopographia Etrusca, Gentium Mobilitas (2007), Le lingue frammentarie dell’Italia Preromana (2009) and Monumenta Linguae Raeticae (2015). V eronik a Nik it ina  holds a doctorate in Latin Linguistics at the University of Oxford, gained under the supervision of Professor J. N. Adams. Her doctoral thesis dealt with the question of standardisation and variation in Classical Latin. Her research interests include Latin epigraphy and early Latin literature and she is currently working on the topic of euphemisms in Latin. Neil O’Sulli va n  is a Senior Lecturer in Classics and Ancient History at the University of Western Australia. His recent publications focus on Cicero’s use of Greek and its transmission in Latin manuscripts. Costa s Pa nayota k is is Professor of Latin at the University of Glasgow. He researches fragmentary Roman comic drama (mime and Atellane comedy) and Latin fiction, and is author of Theatrum Arbitri: Theatrical Elements in the Satyrica of Petronius (1995) and Decimus Laberius: The Fragments (2010). He is currently preparing critical editions (with ­facing translation and commentary) of the fragments of Atellane comedy, the sayings (sententiae) associated with the mimographer Publilius, and Petronius’ ‘Dinner at Trimalchio’s’. Giuseppe Pezzini  is Fellow and Tutor in Latin at Corpus Christi College, Oxford. He has published especially on early Latin language and ­literature, Roman comedy, ancient philosophy of language and ­fiction theory. He is the author of Terence and the Verb to-be in Latin (2015), and of a forthcoming commentary on Terence’s Heautontimorumenos. Olg a Speva k,  a member of the Academia Europaea, is Assistant Professor of Latin and Greek Linguistics at the University of Toulouse Jean Jaurès. Her interests are primarily in the areas of Latin syntax, semantics and pragmatics. Her publications include Constituent Order in Classical Latin Prose (2010) and The Noun Phrase in Classical Latin Prose (2014). She is currently working on nominalisation and features of Legal Latin.

Published online by Cambridge University Press

List of Contributors

xvii

Ba r na by Tay lor is Fellow in Classics at Exeter College in the University of Oxford. He is the author of Lucretius and the Language of Nature (2020) and co-editor, with Giuseppe Pezzini, of Language and Nature in the Classical Roman World (2019). Br ent V ine  is Distinguished Professor of Classics and A. Richard Diebold, Jr. Professor of Indo-European Studies at the University of California, Los Angeles. His research is mainly devoted to phonological, morphological and etymological problems in Classical and IndoEuropean linguistics, with attention also to metrics, stylistics and Latin literature. R e x Wa ll ace  is Professor Emeritus in the Department of Classics at the University of Massachusetts Amherst. He is author of Zikh Rasna. A Manual of the Etruscan Language and Inscriptions (2008), co-author of The Archaeology of Language at Poggio Civitate (Murlo) (2018), and coeditor of Lucilius and Satire in Second-Century Rome (2018). His research interests include the languages, inscriptions and alphabets of ancient Italy, writing systems, and comparative/historical linguistics. Ja r r et t W el sh  is Associate Professor of Classics at the University of Toronto. His research focuses on the transmission and editing of fragmentary Republican Latin drama literature and on the texts, especially lexicographical, that preserve the fragments.

Published online by Cambridge University Press

Acknowledgements

Most chapters in this volume derive from a conference entitled ‘What is Early Latin?’, held in the Department of Classics, University of St Andrews, on 10–11 July 2018. The event would not have been possible without the financial support of the University of St Andrews and of Trinity College Dublin, as well as the help and support of the colleagues in the School of Classics in St Andrews, especially Jason König and Sam Dixon. Giuseppe Pezzini conducted the bulk of the work on the volume during a period of research leave, generously funded by the Loeb Classical Library Foundation and the British Academy, to whom he is also very grateful. A main aim of the volume has been to investigate the concept of ‘early Latin’ and whether it is meaningful to speak of the periodisation of the Latin language as it has survived in written sources. New discoveries come up, such as the inscriptions on warship rostra dating from the first Punic War: these have been duly discussed. Contributors have engaged deeply with the many facets of the question and provided new insights on a wide range of texts, from the Latin extant in texts and inscriptions of the early period to the reception of ‘early Latin’ language in the late Republic, the imperial period, late Antiquity and beyond. We are very grateful to the contributors for their enthusiastic answer to our call, for their responses to our comments and suggestions, and for their commitment and patience during the long production of the volume. Special thanks go to Charlie Kerrigan for all the work he has done on indexing and proofreading. Finally, we are very grateful to Michael Sharp of Cambridge University Press for his support and tolerance, and to Juliet Wilberforce for her ­excellent copy editing. J. N. A., A. C., G. P.

xviii

Published online by Cambridge University Press

Abbreviations

Abbreviations of ancient Latin works are given as per OLD where possible, with no indication of the edition used if this is the same used in the OLD (e.g. Acc. trag. = Acc. trag. Ribbeck 1896), except in a few cases for the sake of clarity (e.g. ORF). Additional references are at times given in brackets where individual contributors have used a different edition. References from ­editions other than those used in the OLD are always indicated (e.g. Enn. Ann. Sk., Trag. J.). Abbreviations of late antique Latin works and ancient Greek works are given as per TLL and LSJ respectively.

General Bennett BTL

Bennett 1910–1914 Bibliotheca Teubneriana Latina. 4th ed. Munich and Leipzig, Saur/Turnhout, Brepols, 2006 CE Bücheler, F. (1895–1897), Carmina Latina Epigraphica. Leipzig Ch. Charpin 1978–1991 CGL Corpus Glossariorum Latinorum, vol. 5: Placidus, Liber glossarum. Glossaria reliqua. Edidit G. Goetz (Leipzig, 1894) CIL Corpus Inscriptionum Latinarum. Berlin, 1862– CILA Corpus de inscripciones latinas de Andalucía. Seville, 1989 Clauss-Slaby  Epigrafik-Datenbank Clauss-Slaby (www.manfredclauss .de/) CSbC Cugusi and Sblendorio Cugusi 2001 CT Daviault 1981 D. Dangel 1995 DELL Ernout and Meillet 1985 EDLIL De Vaan 2008 EF Bakkum 2009 xix

Published online by Cambridge University Press

xx ELM ET F. FPL4 FGrHist FrHist GL G–M GRFF GRFM2 H–S ILLRP J. K–H K–St L. LEW LIMC LIPP LIV LLT LSJ LTUR M. MRR OCD3 OCD4 OLD ORF P.

List of Abbreviations Cugusi 1970–1979 Meiser 2014 Frassinetti 1967 Blänsdorf 2011 Die Fragmente der griechischen Historiker, ed. F. Jacoby (Berlin, 1923–) Cornell 2013 Keil, H. (1855–1880), Grammatici Latini, 8 vols. Leipzig Goldberg and Manuwald 2017, 2018 Funaioli 190 Mazzarino 1955 Hofmann and Szantyr 1965 Degrassi, A. (1957–1963), Inscriptiones Latinae Liberae Rei Publicae, 2 vols. Firenze Jocelyn 1967 Kühner and Holzweissig 1912 Kühner and Stegmann 1955 Lindsay 1903 (Nonius), 1913 (Festus) Walde and Hofmann 1965 Lexicon iconographicum mythologiae classicae. Zürich, 1981– Dunkel, G. E. (2014), Lexikon der indogermanischen Partikeln und Pronominalstämme, 2 vols. Heidelberg Rix, H. (ed.) (2001), Lexikon der indogermanischen Verben: Die Wurzeln und ihre Primärstammbildungen, 2nd ed. Wiesbaden. Brepols Library of Latin Texts online (https://about.brepolis .net/library-of-latin-texts/) Liddell, H. G. and Scott, R. (1940), A Greek-English Lexicon, 9th ed. Oxford Steinby, M. et al. (eds.) (1993–2000), Lexicon topographicum urbis Romae, 6 vols. Rome Marx 1904–1905 Broughton 1950–1968 Hornblower, S. and Spawforth, A. J. S. (eds.) (1996), The Oxford Classical Dictionary, 3rd ed. Oxford Hornblower, S., Spangenberg Yanes, E., and Eidenow, E. (eds.) (2012), The Oxford Classical Dictionary, 4th ed. Oxford Glare, P. G. W. (ed.) (2012), The Oxford Latin Dictionary, 2nd ed. Oxford Malcovati 1976 Panayotakis 2010

Published online by Cambridge University Press

List of Abbreviations RE RhL RRC RS Sch. Sk. ST SVF TLL TrRF V.

xxi

Wissowa G., et al. (eds.) (1893–1980), Paulys Real-Encyclopädie der classischen Altertumswissenschaft. Stuttgart and Munich Halm 1863 Crawford 1974 Crawford 1996 Schierl 2006 Skutsch 1985 Rix 2002 von Arnim 1903–1905 Thesaurus linguae Latinae. Leipzig, 1900– (vol. 1) Schauer 2012, (vol. 2) Manuwald 2012 Vahlen 1903

Inscriptions and Texts not Recorded in OLD Agr. Bant. Corn. de XX quaest. Decret. Paulli Fragm. Tarent. Gall. Narb. Irnit. Iul. munic. Malac. Par. fac. Puteol. Prov. praet. Repet. Rubr. SC de Bacch. SCPP Sent. Minuc. Spolet. Tarent. Vrson.

Published online by Cambridge University Press

Lex agraria (CIL 12.585) Lex Latina tabulae Bantinae (CIL 12.582) Lex Cornelia de XX quaestoribus (CIL 12.587) Decretum L. Aemilii L. f. Paulli procos. Hispaniae (CIL 12.614) Lex incerta tabulae Tarentinae, fragmentum (CIL 12.2924) Lex Galliae Narbonensis de flamonio provinciae (CIL 12.6038) Lex Flavia Irnitana (CILA 2.4, n. 1201, Clauss-Slaby) Lex tabulae Heracleensis quae Iulia municipalis dicitur (CIL 12.593) Lex municipii Malacitani (CIL 2.1964) Lex parieti faciendo Puteolana (CIL 12.698) Lex de prouinciis praetoriis, uulgo lex de piratis persequendis (Crawford 1996: 1.238–252) Lex repetundarum (CIL 12.586) Lex Rubria de Gallia cisalpina (CIL 12.592) Senatus consultum de Bacchanalibus (CIL 12.581) Senatus consultum de Cn. Pisone patre (copy A) Sententia Minuciorum (CIL 12.584) Lex luci Spoletini (CIL 12.336a) Lex municipii Tarentini (CIL 12.590) Lex coloniae Genetivae Iuliae sive Vrsonensis (CIL 12.594)

chapter 1

Introduction: What Is ‘Early Latin’? Giuseppe Pezzini and Anna Chahoud

What is ‘early Latin’? The main contention of the present volume is that this question does not have a single answer. Rather, ‘early Latin’ is one of those ubiquitous labels (like ‘old’ or ‘archaic’ Latin) which have been used by classical scholars to denote different linguistic entities, and above all to describe a variety of linguistic features, in an often confusing and potentially contentious way. ‘Early Latin’ is above all a linguistic construct, which evokes frameworks of periodisation (often diverging), and posits a distinction between a supposedly discrete and cohesive linguistic variety (‘classical Latin’) and another one, equally discrete and cohesive, belonging to an earlier time period (‘pre-classical Latin’, a notion which has often carried negative value judgments since antiquity). Far from aiming to replace one theoretical framework with another, the studies presented here contribute, through a fresh analysis of specific linguistic phenomena and stylistic trends, to challenge the myths of periodisation and standardisation, and to expose the limited usefulness of evolutionary models to explain language change.

1.1  The ‘Early Period’ and the Problems of Transmission A first related problem concerns the chronological boundaries of the period, and especially its end point. Standard discussions point to the period 450–100 bc (Courtney 1999: vii) or 400–150 bc (Clackson and Horrocks 2007: 90–129), with various references to discontinued ‘archaic’ or ‘old’ ­orthography, morphology, syntax and stylisation traits. In this ­volume (which intentionally avoids a single, superimposed framework) ­different authors have set it to 80 bc (Cicero’s early works), 100 bc or 150 bc (­approximately coinciding with the death of Lucilius and Cato respectively), but the list of dates could be easily expanded, especially if one includes the ­perspectives of ancient linguists, from Varro to Pliny, from Valerius Probus to the grammarians of late antiquity, each with a different notion of the 1

https://doi.org/10.1017/9781108671132.001 Published online by Cambridge University Press

2

Giuseppe Pezzini and Anna Chahoud

ueteres, their authority and usefulness (see in particular Garcea’s discussion in Chapter 26). Similar divergences concern the starting point of the ‘early period’, which is often (but not always) made to coincide with the date of production of the first literary texts (mid–late 3rd century bc). The scope of this volume is purposefully broad, and for this reason it also includes an opening chapter on pre-literary Latin and its sources (see Chapter 2 by Wallace). At the other end, the last ‘early author’ discussed in the volume is Lucilius, but this should not be taken as a neat demarcation point: in fact, another important contention of the book is that there is no linguistic watershed between what are normally described as ‘early’ and ‘classical’ Latin, and that continuity appears to be more prominent than change, as evident in the case of Terence (cf. e.g. Chapter 11 by Pezzini). Even more problematic is the criterion applied, whether consciously or not, to identify supposedly ‘early features’. Scholars still tend to define as ‘early’ any linguistic feature attested in ‘early texts’ that does not correspond to the linguistic standards codified by modern grammars, which are mainly based on (selected) texts of Caesar and Cicero (‘classical Latin’). In fact, as shown in several chapters, many of such supposedly ‘early features’ can be easily paralleled in late republican or classical sources, without any particular archaising stylisation (this is the case for instance of phrases with ecquis and derivatives, discussed by Bodelot in Chapter 8). Moreover, any comparison between ‘early’ and ‘classical’ texts is severely vitiated by the problems of transmission. De Melo, in Chapter 6, notes, for example, that archaic morphology is often modernised in the manuscripts of Cato (p. 100). More generally, Damon reveals in Chapter 24 the amount and range of alterations which Pliny introduced in his quotations of Cato, ­including for instance the use of abstract nouns, and the avoidance of old-fashioned adjectives. It is not only the text of early authors that was changed. As mentioned in Chapter 9 by Barrios-Lech (p. 164), the manuscripts of Cicero were heavily standardised by ancient editors, often under the bias of periodisation. Also more careful analyses, focusing in particular on features stylised as archaisms in classical Latin, often fail to take into account the diachronic and synchronic variety within ‘early Latin’ itself. The contrast between the morphosyntactic and prosodic systems of Plautus and Terence (noted here by Bodelot, pp. 155–6, and de Melo and Pezzini, Chapter 5, passim), would suffice to illustrate this; equally revealing would be a comparison between the language of the Egadi inscriptions (de Melo, Chapter 4) and that of Lucilius (Chahoud, Chapter 17). In fact, even within the boundaries of the ‘early period’ there are authors who would consider the language of their

https://doi.org/10.1017/9781108671132.001 Published online by Cambridge University Press

Introduction: What Is ‘Early Latin’?

3

predecessors as ‘old’ or ‘archaic’, and deal with it in different ways, from anchoring their own diction to it, to purposefully rejecting it. Similarly, as discussed in Chapter 3 by Marchesini, the epigraphic corpus shows a great fluctuation between the preservation of fossilised, old fashioned and innovative traits (e.g. phonetic spellings), sometimes occurring together in the same text or type of text. Indeed, as noted by de Melo (pp. 67–8), the first attestation of a modern form in a particular inscription does not mean that the change has been completed and the attestations of the older form are spelling archaisms. For this reason, early Latin inscriptions cannot be dated reliably by purely orthographical features, and this makes extremely difficult any attempt at reconstructing language change in precise diachronic terms. To complicate the picture, one must stress that the linguistic variety of ‘early Latin’ is not only diachronic, but above all synchronic.

1.2  Synchronic Variety: Genre, Register and Style Synchronic variety is one of the topics more often addressed in the volume, especially in relation to the distinction between language change on the one hand, and divergences related to genre, register and style on the other. As neatly stated by Adams and Nikitina in Chapter 14 (p. 307): [‘Early Latin’] was not a monolith whose features were general across all types of texts in particular periods. It was generically varied within periods, and the possibility was there that a genre from, say, the early period (or, if composed later, written in imitation of the practice of an early period) might have a feature not, or hardly, shared by other genres of its own period.

That is to say, there are many features that are often associated with the ‘early period’ as a whole, which are in fact restricted to or distinguished according to different genres. A case in point is asyndeton, which, as noted by Adams and Nikitina in their chapter, is frequent in early legal texts, but rare in prayers. Some of these generic features may disappear from later texts of the same genre (e.g. end-of-list coordination, typical of early prayers, comes out of fashion in the imperial period; Adams and Nikitina p. 146), but in many cases they do persist in classical Latin; in this case periodisation is a meaningless concept. Another example is the plurality of metrical systems attested in ‘early Latin’, distinguished according to genre, which in many respects is preserved also in classical poetry (de Melo and Pezzini). Some generic features reflect divergences of register: an example is the use of the indicative mood in indirect questions, which in his chapter BarriosLech describes as a typical feature of low-register informal speech. Its frequency

https://doi.org/10.1017/9781108671132.001 Published online by Cambridge University Press

4

Giuseppe Pezzini and Anna Chahoud

in (low) comedy, and rarity in (high) tragedy, is not surprising, and reflects a register and generic distinction which, again, is also attested in classical Latin. At the same time, instances of the subjunctive in indirect speech are in fact also found in Plautus and (more rarely) Terence, usually in the speech of high-status characters. A neglect of generic and register factors has often produced very distorted pictures of ‘early Latin’, especially given the particular literary nature of the extant ‘early’ corpus. In fact, the vast majority of early Latin texts (240–80 bc) belong to a specific literary genre, that is comedy (c. 221,000 words or 82% of the total), including Plautus and Terence above all, but also Caecilius Statius, Afranius, Titinius, etc. To these one should also add the c. 8,000 words of the satirist Lucilius, who imitates the language of comedy in many respects (Pezzini 2018; cf. Chahoud, pp. 358–9). In the history of Latin literature, the comic genre is represented almost only by the above ‘early’ authors. The few exceptions of equivalent texts (including the fragments of Atellanae, analysed by Panayotakis in Chapter 20) form a very small corpus (no more than 3000 words combined), and indeed display many inherited generic features. The inherent ‘low-register’ or ‘colloquial’ nature of the early Latin corpus should be taken into consideration when assessing supposedly ‘early’ features, together with the particular linguistic idiosyncrasies of ‘comic’ (and especially Plautine) diction: in many cases features that have been considered ‘early’ are in fact better described as ‘low-register’ and/or as specifically ‘comic’. The picture is even more complicated by the characteristic fondness of both comedy and early satire (Lucilius) for the interplay of different registers and genres, resulting in a considerable variety of stylistic effects. This point introduces another key variable to consider, which is ­stylisation: many forms attested in early texts do not have a neutral ­linguistic value, reflecting standard speech, but are rather artificial and ­linguistically marked, and chosen by writers for particular stylistic effects. Stylised forms can derive from and evoke other genres: this is the case for instance in the official, legal and religious terminology that abounds in early ­tragedy (Chapter 12 by Maltby) and epic (Chapter 13 by Goldberg), and, ­conversely, the amount of tragic diction used by Plautus and Lucilius in mock-tragic passages (Maltby, p. 271, Chahoud, pp. 356–7). Other forms can be modelled on the style of previous literary models, also foreign, such as the Homeric-style neologisms used by Ennius or Naevius (Goldberg, pp. 278–80), which in turn will enter high-style Latin poetry (Lucretius, Virgil) not as ‘archaisms’ but as markers of genre-specific poetic diction (Leumann 1974: 153–7).

https://doi.org/10.1017/9781108671132.001 Published online by Cambridge University Press

Introduction: What Is ‘Early Latin’?

5

The typical result, and indeed goal, of stylisation is the creation of an artificial language to contrast with contemporary, common speech. An extreme case of this contrast is discussed by Clackson in Chapter 19, which presents the results of a systematic collection of Greek loanwords in early Latin sources. Clackson notes in particular that many of these words had a limited occurrence or could be better explained as code-switches, and in fact were probably not in common use and/or were not fully integrated into Latin. Against communis opinio, statistical comparison suggests in fact that the proportion of (Greek) loanwords in ‘early Latin’ is strikingly low compared to that in other languages, including later stages of Latin, but appears to be analogous to that of other languages spoken in Italy in the same period. These results are also confirmed in Chapter 18 by O’Sullivan in his discussion of Cato’s avoidance of so-called ‘Grecisms’, that is, lexical borrowings from Greek. In contrast, the relative preponderance of Grecisms in Plautus is thus revealed as a distortion of actual language practice, and was presumably considered as stylistically marked by ancient audiences. Clackson’s analysis shows the need for extreme caution when using the extant Latin corpus as evidence for the reconstruction of early Latin language. Similar considerations are raised by Gray in Chapter 16, dedicated to patterns of repetition in the fragments of republican orators, and to the elucidation of their variety and development. Arguing against traditional discourses on the ‘linguistic primitivity’ of ‘early Latin’, Gray’s analysis raises awareness ‘of the different stylistic levels in which repetition can participate, and to account for their motivation without giving in to the temptation of ascribing differences merely to the diachronic evolution of language’.

1.3  ‘Early’ Archaisms Artificial Grecisms and stylised repetition are not the only aspects of ­stylisation in early Latin sources; another important one is the fondness for old-fashioned words. We refer here to ‘archaism’, a label that from ancient critics to the present day refers to two quite distinct phenomena: a) the adoption of ‘a word or usage long since recognised as obsolete … as a type of stylistic ornament’, or b) ‘the last throes of an older way of writing’, thus ‘archaism’ in a ‘weaker sense’, in which an author deliberately or not uses words/usages on their way out or refuses to adopt ‘the latest stylistic vogues’ (Adams 2005: 79). Both traits are widespread in extant early texts, and present the greatest obstacle to any attempt at diachronic analysis. In fact, another extensive

https://doi.org/10.1017/9781108671132.001 Published online by Cambridge University Press

6

Giuseppe Pezzini and Anna Chahoud

part of the ‘early’ corpus is made of ‘high-register’ texts (c. 45,000 words), including the works of Cato and Sisenna (notorious for their oldfashioned, archaising style), but also the works of Ennius (Annals and ­tragedies), Naevius, Livius Andronicus, Pacuvius and Accius. All these texts are characterised by a style fond of fossilised linguistic features, which were probably already ­obsolete in contemporary speech. Maltby and Goldberg discuss some of these f­ eatures, including e.g. the use of an artificial syntax, or the e­ xploitation of archaic morphology for elevated stylistic effect. A similar fondness for archaism characterises another important group of ‘early Latin’ sources, namely inscriptions, which typically tend to ­preserve fossilised features, no longer reflecting actual linguistic practice. An example mentioned by Wallace (p. 20) is the spelling q before u, an ­orthographic practice standard in pre-literary Latin, which is still attested in the ‘early period’ (e.g. CIL 12.586 pequnia, + nine other inscriptions in CIL 12), down to the 1st century bc (e.g. CIL 1.593, 45 bc); and yet it is likely that this spelling was already old-fashioned in the early period, as suggested by the higher frequency of the spelling pecunia (eighty-eight inscriptions in CIL 12) and by the concentration of -quC- spelling in sources marked by a general archaising orthography (especially legal texts; cf. e.g. CIL 1.584, 117 bc, which features, among other things, the old-fashioned spelling of geminates as single consonants, e.g. iouserunt, casteli). Also Marchesini and de Melo note in their chapters the fossilising tendency of c. 400–200 bc inscriptions; de Melo for instance comments on the archaising spelling -o for -us or the old-form probauere, both of which could be thus described as ‘early archaisms’. Early archaisms are not only found in high-register texts and inscriptions: Pezzini and de Melo’s chapters on Roman comedy (Chapters 5, 6 and 11) discuss a great number of obsolete or obsolescent linguistic forms, prosodical, lexical or morphological, which are used by Plautus and (more sparsely) Terence for stylistic or other purposes. These include old forms preserved as fossilised metrical variants (e.g. the subjunctive siem, siet, the passive infinitive -ier, the future scibo), stylised morphological variants, often used with bombastic effect (e.g. the genitive -ai, the genitive -um), as well as a great number of lexical archaisms, used to provide a range of different effects, from characterisation to emotion. Again, this kind of archaising stylisation, attested in a large part of the ‘early’ corpus, must be taken into account when discussing the supposedly ‘early’ features of the language of the period.

https://doi.org/10.1017/9781108671132.001 Published online by Cambridge University Press

Introduction: What Is ‘Early Latin’?

7

1.4  Linguistic Layers: Three Case Studies The linguistic complexity of ‘early Latin’ can be epitomised by focusing on three chapters, very different in content and approach, but similar in their illustration of the problems inherent to the notion of ‘early Latin’, construed as a discrete diachronic variety. The first is Vine’s case study in Chapter 10, dealing with the interjection edepol. On the one hand edepol appears as a typical ‘early Latin’ form, very common in and almost completely restricted to early sources, and displaying phonological phenomena normally associated with ‘early Latin’, which Vine reconstructs in his newly proposed etymology: syncope of final vowel (pollux > -pol), monophthongisation ei > ē. (*deiu- > -dē.(u)-), shortening by enclisis (-dē- > de˘ , the asseverative particular *ē> ˘e -). On the other hand, as noted by Vine, the distribution of edepol suggests a strong generic affiliation, with almost all instances found in comedy (including the late republican Atellanae of Novius). Given that the ‘post-early’ poet Novius is virtually the last comic playwright of Latin literature, one wonders whether the absence of edepol in later texts has any diachronic significance. Moreover, the phonological features displayed by edepol, although still attested in early sources, are in fact not predominant: the monophthongisation ē. , in particular, is much rarer than ei or ī, and by the end of the early period was considered rustic (cf. Adams 2007: 52–61). Edepol is thus better described as fossilised comical jargon, with no particular affiliation with ‘early Latin’ per se. Similar problems concern the form pol, abundant in ‘early Latin’ (especially in association with women), but also found in quoted speech of later times: as noted by Vine, ‘[t]his is therefore a word that belongs to spoken language, with a profile that is broader than that of edepol, both in terms of genre and chronology’ (p. 210). The second is Chapter 15, by Spevak, who discusses a number of syntactic features attested in early legal texts (siremps(e) ‘the same’, dominant participle, autonomous relative clauses and their mood, restrictive quod, quo setius), which could be distinguished between four different typologies: (a) Features attested in legal (con)texts of all periods, but also found in non-legal (con)texts, in both early and classical periods (dominant participle) (b) Features attested in legal (con)texts of all periods, and not found in non-legal (con)texts, in both early and classical periods (sirempse, autonomous relative clauses)

https://doi.org/10.1017/9781108671132.001 Published online by Cambridge University Press

8

Giuseppe Pezzini and Anna Chahoud

(c) Features attested in legal (con)texts of all periods, which are also found in non-legal (con)texts in the early period, but not in the c­ lassical ones (quo setius) (d) Features only attested in legal (con)texts of the early period (­restrictive quod) Of these, only category c) and d) can be properly described as ‘early’ features, and with different stylistic values, whereas for category a) and b) periodisation is not a fruitful concept. The problem is of course that ­discrimination between these different categories is often impossible. The third is Chapter 23 by Briscoe, who collects passages in Livy which purport to quote verbatim original early Latin sources, and ­presents a close reading of three of them (22.10.2–6, 25.12.5–10, 40.52.4–6), as sample-cases to illustrate the methodological problems involved. He identifies many features that can be paralleled in ‘early Latin’ sources and/or apparently archaising contexts, and yet many of these cannot be straightforwardly construed as ‘early’: these include for instance ­variants that were obsolete by Plautus’ time and used in stylised contexts already in ‘early Latin’ (e.g. duellum for bellum); formulaic phrases likely to have been tralatician; stylistical features typical of formal Latin, also attested in classical Latin (e.g. repetition of the antecedent). In addition, one finds ‘early forms’ in manifestly forged passages (e.g. duis and ast in 10.19.17), or even ­potentially forged forms, evoking ‘early’ morphology or phraseology (e.g. the unique passive faxitur). In fact, few of the ‘early features’ identified by Briscoe in his sample texts (also) appear in early non-stylised contexts, which is the condition required to use the label ‘early Latin’ in any meaningful way.

1.5  Early Latin Features If one takes into account all the above methodological considerations, one should indeed be led to consider as ‘early’ only features that (1) are only or predominantly attested in ‘early’ sources and/or elsewhere only occur in archaising (con)texts and (2) are evenly distributed in ‘early Latin’ across different authors and/or do not betray obvious signs of particular stylisation and/or generic association (and especially no signs of being marked as old-fashioned). Some examples are discussed in Chapter 6 by de Melo and Chapter 7 by Baños, including for instance the use of esse in support verb constructions with nouns in -tio, the omission of the subject in infinitive clauses, or of ut in subject and object noun-clauses.

https://doi.org/10.1017/9781108671132.001 Published online by Cambridge University Press

Introduction: What Is ‘Early Latin’?

9

In fact, few of the features analysed in the volume could be identified as belonging to this category of specifically ‘early Latin features’. In any case, this kind of identification is not the primary purpose of the volume. Moreover, to strictly apply this criterion would not be helpful in describing the linguistic situation of ‘early Latin’, which is characterised, as seen, by the co-existence of a range of different variants, to be synchronically and diachronically distinguished. Although linguistic variety cannot be associated only with ‘early Latin’, it does appear as a distinctive trait of the language of the period, as noted in many chapters. Other more general features of ‘early Latin’ appear to be ­independence from Greek, described by O’Sullivan as an ‘assertion of identity against a Greek cultural hegemony’ (p. 385), and a general fondness for ­lexical ­coinage (cf. Maltby’s chapter). And yet, even in this respect, ‘early Latin’ should not be treated as a monolithic entity: as shown by Pezzini, the ­language of Terence is free from the linguistic variety and e­xuberance admitted by many of his early colleagues, and it is indeed, in many respects, indistinguishable from that of Cicero, notwithstanding Cicero’s own notions of periodisation. This leads to the last, but certainly not least, key issue discussed in the volume.

1.6 Reception Although the volume does illustrate the features most commonly associated with ‘early Latin’ as a real linguistic entity, its main focus is rather ‘early Latin’ as a construct, or rather a set of different constructs. This explains the variety of approaches, the broad and non-dogmatic perspective, but also the prominence of ‘reception’, unusual in studies of this kind: nine chapters are dedicated to the ancient notions of ‘early Latin’ from Cicero and Lucretius down to Scaliger, with special attention to grammarians (in Chapter 26 by Garcea, and in Chapter 27 by Welsh), aiming to trace the history of the periodisation of Latin, and its limits. A protagonist in this history is of course Cicero (see Chapter 22 by Manuwald), who is the first extant author to quote extensively from early authors (‘Ennius, Accius, Pacuvius, Lucilius, Terence, Caecilius and o­ thers’, Quint. Inst. 1.8.11). Cicero’s attitudes to ‘early Latin’ anticipate trends that will be developed by later authors and grammarians, including above all the contradictory use of ‘linguistic peculiarities of the playwrights as a parallel or model for contemporary orators or as features to avoid in oratory’ (p. 464). The opportunistic idealisation of ‘early Latin’, tempered by an aversion to obscurity, is visible in many other authors, from Gellius to late

https://doi.org/10.1017/9781108671132.001 Published online by Cambridge University Press

10

Giuseppe Pezzini and Anna Chahoud

grammarians (discussed in Chapters 25 and 26 by Holford-Strevens and Garcea respectively), and complicates the tradition of ‘early Latin’. As seen above, archaism is a literary notion, and one that came to the fore especially at the time of the second-century ad revival, in which ­deliberately archaising authors such as Gellius, Fronto and Apuleius were seen to p ­ erform direct imitation of, or allusion to, words and ­phraseology of ­earlier writers (Pezzini 2016: 29–30). This volume does not include a study of Apuleius, whose ‘archaism’ is more precisely a case of Plautine r­eception (cf. e.g. Harrison 2004: 17, Pasetti 2007: 159–63), or of Fronto, whose s­upposed ­archaisms may well be conversational traits, except when he resorts to ­quotations when discussing contemporary topics (e.g. 47.1–6, with van den Hout ad loc.: ‘Fronto’s being an archaist does not mean that he writes archaic Latin’). Gellius’ preference for the early authors signalled a deeper nostalgia for a long gone past. Holford-Strevens notes that ‘his love of the early writers is not a matter only of language or style; they also serve a moral and patriotic discourse’ (p. 520). It is a discourse that started with Cicero and Varro in the late Republic, and continued with the late grammarians in their effort to preserve what they perceived to be the integrity of the Latin language. Welsh’s chapter in particular illustrates these problems by focusing on the most extensive indirect source for republican Latin, the lexicographical collection of Nonius Marcellus. Against traditional views (especially concerning the lack of a discernible rationale behind the De Compendiosa Doctrina), Welsh argues that Nonius’ excerption was biased by his own interests and by his readers’ needs, including in particular the concern to provide a selection of old-fashioned (but not obsolete) words, to be used as linguistic preciosities in public speech. From a linguistic point of view this approach resulted, in practice, in the distorting preference given to morphological variants, and above all in the neutralisation of the stylistic charge of a word (e.g. the original elevated tone of the verb caluor, excerpted from a Plautine song, is obscured by Nonius’ presentation, which gives the false impression that the verb was ordinary in ‘early Latin’ speech). A consequence of the opportunistic reception of ‘early Latin’ is described by the notion of ‘literary renewal’, which is discussed in several chapters, including those by Taylor and Panayotakis. In Chapter 21, Taylor ­analyses a sample of apparently ‘early Latin’ features in Lucretius’ poem, such as infinitive -ier, genitive -ai, escit, cupīret, subjunctive ausim/ausis, ­gerund with direct object, and the triad terra mare caelum. All these supposedly early features are ultimately derived from the model of poetic diction ­stylised by Ennius and imitated by Cicero (Aratus), and are freely ‘renewed’ and expanded by Lucretius into an archaising poetic language; this is an

https://doi.org/10.1017/9781108671132.001 Published online by Cambridge University Press

Introduction: What Is ‘Early Latin’?

11

artificial linguistic entity, designed to fulfil the needs of literary expression, and has little to do with ‘early Latin’ per se. In fact, Taylor uses ‘early Latin’ not as a diachronic concept, but rather as a ‘certain linguistic and cultural construct … associated … with the styles and languages of an earlier period’ (p. 434). The same is true for Catullus, in a completely different genre (or rather a variety of genres) that deliberately distanced itself from the diction of third- and second-century poets; when archaising vocabulary or morphology make an appearance, irony or parody is at work (e.g. poem 44), or an elevation of style in the long poems: in any case, we are dealing with poeticisms and not with ‘archaisms’ in any sense of the word (cf. Chahoud 2021: 127–30). Similar considerations are valid for the language of the fabula Atellana, which renews the diction of earlier comedy; in this case, however, the feature is not archaism, but rather generic affiliation. As noted by Panayotakis, the authors of Atellana re-used (and renewed) ‘early features’ not in order to sound ‘early’ or ‘old-fashioned’ but for reasons associated with the genre in question, namely to ‘acknowledge the genre’ within which they are writing. In fact, the construct of ‘early Latin’ as emerging from the volume ‘was not static as a linguistic concept or as a literary period; it was fluid, continuous and renewable’ (p. 416). A language created from the remains of times past to fulfil the needs of literary expression: this is what ‘early Latin’ became once more, over a thousand years later, when the fragmentary republican authors were rediscovered in the mid-16th century. Chapter 28 (Chahoud) discusses a lesser known aspect of the work of Joseph Scaliger, the classical scholar who more than anyone else promoted the critical study of fragmentary Latin and its sources. While engaged in the study of Festus’ De uerborum significatione, the young Scaliger noted the variety and strangeness of the vocabulary preserved in Festus’ quotations from early authors, and used it in his Neo-Latin translation of Sophocles and Lycophron. The language of Ennius, Pacuvius, Lucilius turns, in this very particular idealisation and reception, into a repository of rare words to be borrowed, or used as the basis for archaising neologism, for the production of an elevated, recherché, or deliberately obscure Latin verse diction.

1.7  Structure of the Volume The volume is structured in four thematic parts. Part I covers the epigraphic material, with chapters by Wallace, Marchesini and de Melo. Wallace’s chapter presents an overview of

https://doi.org/10.1017/9781108671132.001 Published online by Cambridge University Press

12

Giuseppe Pezzini and Anna Chahoud

pre-literary Latin inscriptions, dated up to the 4th century. This variety of Latin (‘pre-historic’ or ‘pre-literary’ Latin) is to ‘early Latin’ what ‘early Latin’ is to imperial Latin, that is, a repertoire of obsolete forms or ‘archaisms’, to be preserved in the later stage in fossilised contexts or used in place of current forms to add an old-fashioned patina. Marchesini ­carries forward the analysis, presenting an overview of epigraphical evidence between the 4th and the end of the 2nd century (c. 480 inscriptions); this corpus is complemented by de Melo’s close analysis of one of the most fascinating sets of republican inscriptions, the Egadi rostra. Part II of the volume is dedicated to the most widely documented genre in the literature of the early period, namely drama (comedy above all, but also tragedy), and covers different levels of language analysis: prosody and metre (Pezzini and de Melo), morphology (de Melo), syntax (Baños, Bodelot, Barrios-Lech) and vocabulary (Vine, Pezzini, Maltby). Part III presents case studies illustrating some (supposedly) ‘early’ features of the main other genres, preserved in fragmentary form: epic (Goldberg), prayers (Adams/Nikitina), legal texts (Spevak), oratory (Gray), Lucilius (Chahoud), Cato (O’Sullivan). This part closes with a general chapter on the use, or rather avoidance, of Greek (Clackson), which is revealed as another distinctive trait of the period. Part IV includes chapters discussing the reception of ‘early Latin’ in later authors, from late republican (Panayotakis, Taylor, Manuwald) through early imperial literary texts (Briscoe, Damon, Holford-Strevens), and then the grammatical tradition (Garcea, Welsh), to the creative reuse of ‘early Latin’ as a stylistic marker in the early modern period (Chahoud). The concluding chapter (Chapter 29), by Adams, offers insights into the findings of the volume and questions the concept of ‘early Latin’ as a whole.

https://doi.org/10.1017/9781108671132.001 Published online by Cambridge University Press

Part I

The Epigraphic Material

https://doi.org/10.1017/9781108671132.002 Published online by Cambridge University Press

chapter 2

Alphabet, Epigraphy and Literacy in Central Italy in the 7th to 5th Centuries bc Rex Wallace

2.1 Background In the period before Roman expansion, central Italy was inhabited by a culturally and linguistically diverse population. The names of the peoples and the territories they inhabited have come down to us, although in later, often much later, Roman and Greek historical sources. The languages they spoke are known primarily from epigraphic texts incised on durable materials such as stone, metal, ceramic, bone and ivory. The earliest inscriptions date to the 7th – 5th centuries bc.1 Matching the language of the inscriptions with the names of peoples mentioned in historical sources is not always clear-cut, especially for this early period, but there is consensus about the languages and the areas where they were spoken. The inhabitants of central Italy occupied areas adjoining the Tiber River (see Map 2.1). The Latini occupied Latium vetus, an area bounded by the Tiber River to the north, the valle Latina to the east and Monte Circeo to the south. The Latini were ringed round by Sabellic-speaking peoples, Hernici to the east, Volsci and Aurunci to the south. Etruscan territory was delimited by the Tiber River and the Tyrrhenian Sea; its northernmost boundary stretched beyond the Arno River. The Falisci and the Capenates inhabited small tracts of land to the north of Etruscan Veii. Umbrian-speaking tribes settled along the east bank of the Tiber, Umbrians to the north, Sabines to the south. South Picene speakers populated communities in the mid-Adriatic area, in southern Picenum. 1

The terminology used to refer to the inscriptions of the 7th – 5th centuries bc varies from language to language. The periodisation of the languages varies too. In order to avoid confusion, I use the phrases ‘early Latin’, ‘early Faliscan’, ‘early Umbrian,’ etc. to refer to the languages of the inscriptions attested during this period. I employ the phrases ‘early Latin alphabet’, ‘early Umbrian alphabet’, etc., to refer to the alphabets in which the inscriptions of this period were written.

15

https://doi.org/10.1017/9781108671132.003 Published online by Cambridge University Press

16

rex wallace

Map 2.1  The languages of pre-Roman Italy. Reproduced by permission of Rex Wallace, University of Massachusetts Amherst.

With the exception of Etruscan, the languages spoken in central Italy belonged to the Italic branch of Indo-European (Fortson 2017). Faliscan and Latin formed a particularly close-knit group within Italic known as Latino-Faliscan; Umbrian, South Picene and other languages of limited

https://doi.org/10.1017/9781108671132.003 Published online by Cambridge University Press

Alphabet, Epigraphy and Literacy in Central Italy

17

attestation were members of a group of related languages referred to as Sabellic. Etruscan, the major non-Indo-European language of ancient Italy, stood apart linguistically. Our map gives the impression that the various peoples inhabiting this part of the peninsula were linguistically homogeneous, but this depiction is deceptive. Speakers of different languages lived in close proximity to one another, so it is plausible to think that competence in multiple languages was not unusual (Clackson and Horrocks 2007: 41–2). The geography of the region – the rivers and their valleys – promoted economic and social interaction between communities of speakers. Material evidence demonstrates that craftsmen and traders moved freely from one community to another; members of the aristocracy intermarried and established networks of common social and economic interest, cemented by locally elaborated versions of gift-exchange. Inscriptions, reinforced by literary evidence, point to the movement of ethnic groups across regional boundaries as well. Livy’s description of the Etruscan kings of Rome (Liv. 1.35–60) is complemented by epigraphic evidence. Etruscan inscriptions recovered at S. Omobono (ET La 2.3) and the Campidoglio (ET La 2.4) bear regional dialect features, suggesting that Etruscans had an enduring presence in the city (de Simone 1968; Cornell 1995: 157; see also the discussion in Adams 2003: 159–63). An early 6th century Umbrian inscription incised on a locally made mixing bowl documents the presence of Umbrians at Tolfa (Caere) in southern Etruria (Benelli 2013).2 The intensity and duration of linguistic contact in central Italy is demonstrated by common developments such as word-initial stress accent and the formation of an onomastic system composed of praenomen and gentilicium (see Clackson and Horrocks 2007: 41–8). Arguably, the most consequential result of contact was the adoption, adaptation and diffusion of the alphabet. Once introduced, the alphabet spread over the area quickly and established itself in communities on both sides of the Tiber well before the end of the 6th century bc.

2.2  Origins and Diffusion Euboean Greeks, who settled in southern Italy at Pithekoussai and Kyme (Cumae) in the 8th century bc, were responsible for the introduction of the alphabet to the inhabitants of central Italy (Cristofani 1972, 1978). Greek traders sailed to ports along the central Tyrrhenian coast and moved 2

In this paper, I assume, as most researchers do, that the inscriptions written in the early Umbrian alphabet, including those recovered from Sabine territory, were Umbrian in language.

https://doi.org/10.1017/9781108671132.003 Published online by Cambridge University Press

18

rex wallace

inland to developing urban centres attracted by commercial opportunities (Camporeale 2015: 21). They trafficked in new ideas and technologies – pottery production, gold-working, metal-working, sculpture and building technology. Some Greek traders and the craftsmen who accompanied them were literate; they introduced the Greek alphabet that had been developed at their home settlements on the island of Euboea (Bartoněk and Buchner 1995: 189–98). The oldest Euboean Greek inscription in the western Mediterranean, older than the inscriptions recovered from Pithekoussai, was discovered in an elderly woman’s grave at the Osteria dell’Osa cemetery at Gabii in Latium, indicating that Greek speakers were members of the community. The inscription, which reads ευλιν(ος) ‘spinning well’, was incised on a locally-produced ceramic flask dated to the second quarter of the 8th century (Bartoněk and Buchner 1995: 204–5; Watkins 1995a: 37–9; see Bellelli and Benelli 2018: 23–7, esp. 27, for a reappraisal). Etruscans were the first to adopt the practice of writing from Euboean Greeks. The oldest Etruscan inscriptions recovered at Tarquinia and Caere date to the beginning of the 7th century bc (Cristofani 1972: 404–6). The inscription on the so-called Jucker vase of Tarquinia (ET Ta 3.1), considered the earliest in Etruscan, exhibits an orthographic feature – the ci, ka, qu convention – that distinguishes it from its Euboean Greek source. By this convention, the voiceless velar stop /k/ was spelled by the letter c before the vowel letters i and e, by the letter k before the vowel letter a, and by the letter q before the vowel letter u. (Etruscan did not have the vowel /o/.) The Etruscan word kacriqu [meaning unknown] (ET Ta 3.1) demonstrates that the convention permitted a letter representing a resonant consonant to stand between the velar letter and vowel letter.3 The spelling of the phoneme /f/, a sound that was not part of the Greek phonological inventory, was another significant Etruscan innovation. In the earliest Etruscan inscriptions, dating to before 680 bc, this sound was spelt by means of a digraph, hv/vh (both orders were common), for example, θavhna /thafna/ ‘chalice’ (ET Cr 2.1), melehvra /melefra/ [a type of ceramic vessel] (ET Cr 2.8). Latin speakers learned to write from the inhabitants of Caere or Veii, their closest Etruscan neighbours across the Tiber (Maras 2009b). The earliest Latin inscriptions have letterforms that are formally identical to those found on contemporaneous Etruscan inscriptions. More substantial evidence of origins is based on spelling conventions borrowed from 3

Words in Etruscan are transcribed in italics; words in Latin are in small capitals; words in Faliscan and the Sabellic languages are in boldface. For the transcription of Etruscan sibilants, I follow the conventions in Wallace 2008: xviii.

https://doi.org/10.1017/9781108671132.003 Published online by Cambridge University Press

Alphabet, Epigraphy and Literacy in Central Italy

19

Etruscan. The spelling of the phoneme /f/ by means of the digraph vh is the most important piece of evidence. This spelling is found on the Fibula Praenestina (fhe⁝fhaked /fefaked/ ‘he made, had made’), one of the earliest inscriptions in the Latin corpus (Table 2.1; Figure 2.1). Another Table 2.1  Early Latin inscriptionsa Find spot 1. 2. 3.

Dateb

Inscription

Praeneste Praeneste Satricum

c. 675–650 c. 675–650 c. 700–650

Vetusia inscription Fibula Praenestina amphora inscription

4. 5. 6. 7. 8. 9. 10.

Acqua Acetosa Gabii Caere [Roma] Roma [Forum] Roma Roma [Forum] Ficana

c. 650–600 c. 620–610 c. 625–600 c. 600–550 c. 600–550 c. 525–500 c. 600–500

11.

Satricum

c. 575–550

]×tartispo[inscription Tita inscription Vendia inscription Forum inscription Duenos inscription Rex inscription Monte Cugno inscription Dolium inscription

12. 13. 14. 15. 16. 17.

Ardea Roma [S. Omobono] Acqua Acetosa Roma [Forum] Roma [Forum] Lanuvium

c. 550 c. 500 c. 550–500 c. 550–500 c. 550–500 c. 525–500

Kauidios inscription ]ọ uouios inscription Karkavaios inscription Eco rai×[inscription Trepios inscription Abecedarium

18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29.

Garigliano Acqua Acetosa Satricum Tibur Corcolle Lavinium Incertae originis Incertae originis Roma [Domus Flavia] Roma [Domus Flavia] Roma [Domus Flavia] Ager Signinus

c. 500 c. 500–450 c. 500 c. 550–500 c. 500–450 c. 550–500 c. 500–400 c. 500–400 c. 500–400 c. 500–400 c. 500–400 c. 400

Garigliano inscription Manias inscription Lapis Satricanus Tibur inscription Corcolle inscription Madonetta inscription Pulpios inscription Pias inscription Pacua inscriptions ]×iciam. [inscription ]bios inscription Morai inscription

Citation ET La 2.1 CIL I2.3 Colonna and Beijer 1992 CIL I2.2902d Colonna 1980b Peruzzi 1963 CIL I2.1 CIL I2.4 CIL I2.2830 CIL I2.2917c Colonna and Gnade 2003 CIL I2.474 CIL I2.2829 CIL I2.2917a Pensabene 2000 Fortini 2005 Attenni and Maras 2004 Vine 1998 CIL I2.2917b CIL I2.2832a CIL I2.2658 CIL I2.2833a CIL I2.2833 CIL I2.479a CIL I2.479b CIL I2.2916c, d Colonna 1980a CIL I2.2916ka Colonna 1995a

Notes a. The inscriptions in the list are a selection. See Colonna 1980a for additional inscriptions incised on fragments of pottery. b. Dates are based on Hartmann 2005 and Maras 2009a: 435–6.

https://doi.org/10.1017/9781108671132.003 Published online by Cambridge University Press

20

rex wallace

Figure 2.1  The Fibula Praenestina (CIL 12.3). Reproduced by permission of the Center for Epigraphical and Paleographical Studies, The Ohio State University.

significant datum is the ci, ka, qu convention, which was expanded in Latin to include the spelling of all velar and labio-velar stops, /k/, /g/ and /kw/. On the Forum inscription, which provides the most extensive example, c was written before e (recei /reːgej/ ‘king’), k before a (kalator|em ‘herald’), and q before v (qvoi ‘who’) (Figures 2.2–2.3). The rule is found also on the Kauidios inscription; the letters k and q spelled the voiced velar /g/ (kavidios /ga:widios/ ‘Gauidios’, eqo /ego:/ ‘I’). The ci, ka, qu rule was not applied consistently in all early Latin inscriptions. Some scribes appear to have been unsure about what letters – c or k – to use to spell /k/ and /g/. This may be the reason why the scribe who wrote the Duenos inscription corrected the letter k to c in both feced ‘made’ and pacari [meaning unclear], and yet wrote q for the labiovelar /kw/ in the relative pronoun qoi /kwoj/ ‘who’ (Figures 2.4–2.5). Ultimately, this rule was responsible for the fact that the spelling of velars fluctuated considerably from one early Latin inscription to another (Table 2.2). The Faliscan alphabet is also Etruscan-based, perhaps borrowed from the inhabitants of Narce in the south-eastern part of the Ager Faliscus. The ci, ka, qu convention was employed to write the earliest Faliscan inscriptions, consistently so in the Ceres inscription (EF 1), for example, ceres ‘Ceres’, f[i]fiqod ‘fashioned’, 3rd pl. perf., eqo ‘I’, karai ‘beloved’, dat. sg., etc., less capably in EF 3 (kalpena ‘Calpena’, kalketia ‘Calcetia’, kaios ‘Gaius’, sociai ‘friends’, nom. pl.) and EF 4 (eco ‘I’, quto ‘pitcher’). By the 6th century,

https://doi.org/10.1017/9781108671132.003 Published online by Cambridge University Press

Alphabet, Epigraphy and Literacy in Central Italy

21

Figure 2.2  The Forum inscription, side 2 (CIL 12.1). Reproduced by permission of the Center for Epigraphical and Paleographical Studies, The Ohio State University.

the rule appears to have been moribund. Scribes wrote k for /k/ and /g/ in EF 6–7, eko /egoː/ ‘I’, kaisiosio /kajsiosjo/ ‘Caesius’, gen. sg., and in EF 8–9, fifiked /fifi(n)ged/ ‘fashioned’, 3rd sg. perf. At its inception, the Faliscan alphabet was distinguished from the Latin by the spelling of the phoneme /f/, which was represented by the ‘arrow’ sign, of uncertain origin.

https://doi.org/10.1017/9781108671132.003 Published online by Cambridge University Press

22

rex wallace

Figure 2.3  The Forum inscription, sides 1–4 (CIL 12.1). Reproduced by permission of the Center for Epigraphical and Paleographical Studies, The Ohio State University.

Figure 2.4  The Duenos inscription, top (CIL 12.4). Reproduced by permission of the Center for Epigraphical and Paleographical Studies, The Ohio State University.

https://doi.org/10.1017/9781108671132.003 Published online by Cambridge University Press

Alphabet, Epigraphy and Literacy in Central Italy

23

Figure 2.5  The Duenos inscription, drawing, top (CIL 12.4). Reproduced by permission of the Center for Epigraphical and Paleographical Studies, The Ohio State University.

Another distinctive feature of early Faliscan writing was the use of z to spell the sibilant /s/, e.g., z[e]xtos ‘Sextus’ (EF 1). The conditions under which this spelling occurred are not clear; it appears sporadically throughout the history of Faliscan writing (Bakkum 2009: 84–6). In contrast, z appears but once in the early Latin corpus, in an inscription of uncertain interpretation (zka ,̣ Colonna 1980a: 63, no. 29). Even so, the letter maintained its position in the alphabetic series in Latin for another 300 years, until it was replaced by the letter g (Wachter 1987: 32–3; Wallace 2011: 15, 17). Although the Faliscan and Latin alphabets had an Etruscan source, they employed several letters – beta, delta, omicron and xi – that were not used at all, or were rarely used, to write Etruscan. These letters appeared in

https://doi.org/10.1017/9781108671132.003 Published online by Cambridge University Press

24

rex wallace Table 2.2  Spelling of velars in early Latin inscriptions /k/

/k w/

/g/

1. Fibula Praenestina 2. Vendia inscription 3. Forum inscription

— eco recei

— — qvoi, qvos

4.

virco

qoi

eco —

— —

5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.

fhe⁝fhaked — kalatorem kapia(d), sakros Duenos inscription pacari feced, cosmis Monte Cugno inscription — Dolium inscription mamarcom loucios, placiom Kauidios inscription — Eco rai×[inscription —

kavidios, eqo — eco —

Garigliano bowl Tibur inscription Karkavaios inscription Madonetta inscription Corcolle inscription Pulpios inscription ]×iciam. inscription Pacua inscriptions

— kavios — — — eqo — —

kom, sokiois qetios (?) karkafaios castorei, qurois [d]ic̣ a se — ]×iciam. [ pacva (?)

— — — podlouqueique — — — —

Etruscan abecedaria until the end of the 7th bc occupying their canonical positions in the series. Presumably, they were pronounced in the recitation of the alphabet and so were available to the bilingual (or multi-lingual) speakers who adapted the Etruscan alphabet to write Latin and Faliscan. Greek influence could have played a secondary role in the elaboration of both writing systems, for example, the value /ks/ for 𐌢, but the fact that gamma could be used as a sign for the voiceless velar stop /k/ establishes Etruscan as the primary source. Early Umbrian inscriptions dating to the last half of the 7th century bc were recovered at sites on the east and west banks of the Tiber River, at Magliano Sabina (ST Um 3), Poggio Sommavilla (ST Um 2), Chiusi (ST Um 40) and Poggio Civitate (Murlo) (Tuck and Wallace 2018a). Although the find spot of the ceramic vessel on which the oldest inscription, ST Um 41, was incised is unknown, it was probably written in the Umbro-Sabine area of the lower Tiber (Colonna 1999b). Most likely, the alphabet of these inscriptions was borrowed and adapted directly from a Euboean Greek source (see Agostiniani, Calderini and Masserelli 2011: 8; Benelli 2013: 316). The following features distinguished early Umbrian writing from Etruscan

https://doi.org/10.1017/9781108671132.003 Published online by Cambridge University Press

Alphabet, Epigraphy and Literacy in Central Italy

25

(see Agostiniani 2013: 47–8): (i) /k/ was spelled by k; (ii) m was written with four, rather than five, bars; (iii) theta and/or samekh was repurposed to represent the Umbrian front vowel /ẹː/ (transcribed as í), and khi was repurposed to represent the Umbrian back vowel /ọ:/ (transcribed as ú); (iv) s was written with four, rather than three, bars; and (v) a new letter, , was invented to spell the sound /f/. By the 6th century bc, the early Umbrian alphabet had spread beyond the lower Tiber River valley and was modified in regionally distinct ways (Benelli 2008: 23; Maggiani 1999: 67–8). South Picene stone-cutters who incised the epitaphs of local chieftains on funerary stelae made striking modifications to the forms of some signs (Marinetti 1985: 47–60). For example, the letter (/f/) was reduced to a di-colon (˙·) and the letter o to a point ( ). The transverse bars of a and t were sometimes replaced by points, and the letters u, ú, and l were inverted (∧, ⩚, ). Salient alterations were made too by the scribe who incised the 5th century Volscian inscription from Satricum (ST VM 1). In this script, /f/ was written by a tri-colon (⁝) and /ọː/ was written by digamma ( ). Punctuation separating words had the form of four vertical dots and the end of the inscription was marked by five vertical dots. The early Umbrian inscription from Tolfa in southern Etruria (ST Um 4) appears to have been written with letterforms drawn from multiple sources. The sign of the ‘arrow’ (𐌣), which represented /f/, and multi-bar sigma ( ), which represented /s/, point to early Faliscan; four-bar m, í in the form of a windowpane ( ), and letters with odd orientation (inverted u (∧), sideways m ( ) and e), point to early Umbrian and South Picene (see Benelli 2008: 25). A 5th century Hernican inscription (ST He 2) also appears to be a hybrid formation (Colonna and Gatti 1992; Maggiani 1999: 68). Most letterforms and spelling conventions follow those of southern Etruscan writing systems (Caere and Veii), for example, the spelling of /f/ by the digraph hv and right-to-left direction of writing, but four-bar m and inverted letters are features shared with early Umbrian and South Picene writing systems. The diffusion of the Euboean Greek alphabet among the peoples of central Italy followed two main tracts, one Etruscan, the other Umbrian. In the 6th century bc, the area of the lower Tiber served as a focal point for the development of regionally distinct writing systems that drew on resources from both traditions and whose features were further embellished by local developments. Recent epigraphic discoveries at Satricum (see Morandi 2009), which have yet to be incorporated into the complex picture of the spread of writing in central Italy, stand to further enhance our understanding of the development of the writing systems in this part of the peninsula. 8

8

⋅

https://doi.org/10.1017/9781108671132.003 Published online by Cambridge University Press

26

rex wallace

2.3  Epigraphic Categories The earliest inscriptions in Italy were incised on instrumenta domestica. They followed a small set of formulaic structures that were transmitted along with the alphabet. Commonly attested epigraphic categories can be attributed ultimately to Greek models, although not without a significant amount of Etruscan intermediation, particularly in the case of Latin and Faliscan (Colonna 1999a: 439–40). Popular categories were proprietary inscriptions, donative inscriptions, votive inscriptions and artisans’ signatures. During the 7th and 6th centuries bc, inscriptions in these categories had the form of ‘speaking inscriptions’; the object on which the inscription was incised ‘spoke’ to the reader. ‘Speaking inscriptions’ drawn from the epigraphic categories listed above are presented in I–IV.4 I Proprietary inscriptions (1) Etruscan mi qutum karkanas (ET Cr 2.18)

I (am) the drinking vessel of Karkana.

II Donative inscriptions (2) Etruscan min. [i velθ]ur paiθinaie [mu]lu[vani]ce (ET AS 3.6)

Velthur Paithinaie presented me as a gift.

(3) Etruscan mini l[a]uχu[ś]hie paiθin[a]σ muluvan[ice] mlak[aσi] (ET Vn 3.2)

Laukhushie Paithinas presented me as a gift to a noble man.

III Votive inscriptions (4) Etruscan mine muluv[an]ece avile vipiie.n.nas (ET Ve 3.11) (5) Etruscan (6) Etruscan 4

Avile Vipiennas presented me as a gift. mi me ṇ ervas (ET Cr 3.39) I belong to Menerva. [mini avi]le zuqume tu.r.ace men.[er]ạvas (ET Vs 3.10) Avile Zuqume offered me to Menerva.

The following epigraphic symbols are employed in this chapter: Angled brackets indicate l­etters restored by an editor in place of those incised by the engraver; curly brackets {a} indicate letters erroneously made by the engraver; round brackets (a) indicate letters omitted by the engraver but supplied by an editor; square brackets [a] indicate letters that can no longer be read but have been supplied by an editor; the underdot ạ indicates letters that are damaged and/or no longer clearly ­legible, but are restored by an editor; the pipe | indicates line breaks. The symbol × indicates a letter that is damaged but cannot be restored by an editor. Syllabic punctuation is marked by placing a period before and after a letter.

https://doi.org/10.1017/9781108671132.003 Published online by Cambridge University Press

Alphabet, Epigraphy and Literacy in Central Italy

27

IV Artisans’ signatures (7) Etruscan mi mamarce zinace (ET Ve 6.2) (8) Latin (9) Umbrian

Mamarcus made me. manios : med : fhe⁝fhaked : numasioi (CIL 12.3) Manios made me for Numerius. setums : míom | face (ST Um 5) Septimus made me.

Most inscriptions recorded relationships among elites, or among elites and the divinities they worshipped. Proprietary inscriptions named the person who gave the object as a gift. This is true also for most artisans’ signatures. The objects on which the signatures appeared are to be interpreted as gifts commissioned by the subject, that is to say, ‘so-and-so had me made’. Donative inscriptions named the donor, rarely the recipient. The structures found in I and II were employed also in votive texts (III), an epigraphic type that first appears in central Italy in the 6th century. Although the syntax of the inscriptions was simple, more complex structures were made by combining the formulaic patterns of I–IV. V Inscriptions with combinations of epigraphic formulae (10) Latin eco vrna tita(s) vendias mamar[cos m]ed fhe[ced] (Peruzzi 1963)

I am the urn of Tita Vendia. Mamarcus made me.

(11) Etruscan [mi a]uve ḷ es feluśkeσ tuσnataị[eσ pa|]panalaσ mini muluvanice hirumi[n]a φerśnal.[n]as (ET Vn 1.1) I belong to Avile Felushke, (son of) Tusnataie and Papanai. Hirumina Phershnalna presented me. (12) Etruscan

mi titelas θị[na] {mla} m[l]ạχ mlakas (ET Cr 2.9)



I am the beautiful water jar of the beautiful Titela.

(13) Faliscan eco qutoe uotenoisio titas duenom duenas salue[to]d uoltene (EF 3) I (am) the drinking vessel of Voltenos, a fine (gift) of the beautiful Tita. Cheers, Voltenos!

Inscription (10) combines a proprietary inscription with an artisan’s signature. The oldest funerary inscription in Etruscan (11), which is dated to c. 600 bc, was written on a stone stele in the form of two speaking inscriptions. The inscription presents the name of the deceased (Category I)

https://doi.org/10.1017/9781108671132.003 Published online by Cambridge University Press

28

rex wallace

and the name of the warrior’s comrade who commissioned the stele (Category II). Inscriptions in which the object and the proprietor were modified by an adjective phrase modelled on Greek kalos-inscriptions appear soon after the introduction of writing. One such Etruscan inscription, (12), was configured in a chiastic ABBA structure whereby the name of the proprietor and its modifier embraced the name of the object and its modifier. Interestingly, the Etruscan mlaχ mlakas phrase finds a comparable phrase in Faliscan (13). The modifiers in the appositional noun phrase, titas duenom duenas, were arranged in a manner similar to the Etruscan. Inscription (13) also includes an epigraphic type (V) as yet unattested in Etruscan, but with models in Greek (χαῖρε, χαίρετε): the so-called convivial salutation (saluetod uoltene ‘Cheers, Voltenos!’). Such salutations were incised on wine vessels earmarked for use at social events such as banquets. Inscription (14) is the only example in the early Latin corpus. V Salutation (14) Latin

salvetod tita (Colonna 1980b) Cheers, Tita!

Etruscan proprietary inscriptions sometimes appeared with a coda in the form of a prohibition: ein minipi capi ‘don’t steal me’. The Etruscan injunction is the counterpart of the Greek; compare the prohibition in the Tataie inscription (Bartoněk and Buchner 1995: 199–200). The final line of the Latin Duenos inscription concludes with a prohibition: ne malos (s)tatod ‘let no evil person steal (me).’ Early Umbrian inscriptions do not lend themselves as easily to epigraphic categorisation. Some are too fragmentary to classify, others are of controversial interpretation, but do not appear to be composed in categories common to Etruscan, Latin and Faliscan. The only proprietary inscription in the South Picene corpus, (15), copies the syntactic structure of the Greek model: name (genitive) + verb ‘be’ (1 sg. pres.). (15) South Picene apies esum (ST Sp Te 4)

I am (the cup) of Appius.

Inscriptions no longer bound to inherited formulae and composed in language characteristic of elevated styles appear soon after the adoption of writing, perhaps inspired by Greek inscriptions found at Pithekoussai such as ‘Nestor’s Cup’ (Colonna 1988: 28–9). At the same time, scribes who composed these inscriptions drew on native poetic resources. Such inscriptions appear in Etruscan in the second half of

https://doi.org/10.1017/9781108671132.003 Published online by Cambridge University Press

Alphabet, Epigraphy and Literacy in Central Italy

29

the 7th century (ET Fa 0.4; Cr 0.1; Vn 3.1); they are, for the most part, uninterpretable, but attempts – some more convincing than others – have been made to enumerate the texts’ stylistically elevated properties (Colonna 1988: 27–9; Maras 2015: 204; Maras 2020). The Faliscan Ceres inscription and the Latin Duenos inscription, while incorporating inherited epigraphic formulae, also attest features characteristic of poetic style (Eichner 1988–1990b; Martzloff 2018; Mercado 2012: 277– 9; and Watkins 1995b: 131–4). South Picene epitaphs commemorating deceased warriors were sometimes composed in syllable-­counting metrical structures; other poetic properties included distraction of constituents, marked word order and interlocking patterns of alliteration (Eichner 1988–1990a; Martzloff 2018; Mercado 2012: 293–312; and Watkins 1995b: 125–31). The earliest inscribed objects were intended for private consumption. Inscriptions earmarked for public spaces appear at the beginning of the 6th century. A notable example of the civic use of writing comes from the Crocifisso del Tufa cemetery at Orvieto. Families whose members were interred at the cemetery incised their names on the lintel of the family tomb edifice by means of a ‘speaking’ inscription of the proprietary type: mi mamarces velθienas ‘I (belong to) Mamarce Velthienas’ (ET Vs 1.4). Thefarie Velianas, the king of Caere, ‘published’ the gold tablets of Pyrgi (ET Cr 4.4), which recorded his votive dedication to the goddess Uni (in Etruscan and Punic) by affixing them to the doors of Temple B in the site’s sanctuary. The Etruscan inscription on the cippus from Tragliatella, according to the analysis of Colonna (2005), marked the territorial boundary between the communities of Veii and Caere. The inscribed Stele of Vicchio, recently recovered at the site of Poggio Colla, located north of Florence, may be the non-secular counterpart. Although difficult to interpret, it appears to identify the cult site of the Etruscan divinities Uni and Tinia (Maggiani 2016; Wallace 2018). The Latin Forum inscription, of which segments of sixteen lines survive, was erected between the Forum Romanum and the Comitium (see Figure 2.3). Because the cippus was broken off near its base, the surviving portion of the inscription defies interpretation. An attractive proposal suggests that it recorded regulations regarding carriage-traffic on the via sacra (Eichner 1995). Interestingly, the Forum inscription and the Etruscan cippus of Tragliatella share formal similarities: a rectangular stone with boustrophedon writing. The monumental style of writing points to the involvement of administrative officials in composition and layout (Colonna 1988: 30–1).

https://doi.org/10.1017/9781108671132.003 Published online by Cambridge University Press

30

rex wallace

2.4  Palaeography and Orthography of Early Latin Inscriptions The earliest Latin inscriptions, of which a partial list is presented in Table 2.1, are a mixed bag. Geographically, the find spots are scattered throughout Latium, from Rome and Praeneste to Satricum and the sanctuary of the goddess Marica on the Garigliano River. One inscription, an inscribed olla presumably made and inscribed at Rome, was recovered outside Latium at Caere. Numerically, the total is somewhat greater than the combined total of early Umbrian and South Picene inscriptions, but nowhere near the size of the early Etruscan corpus, which now stands at roughly 250 for the 7th century bc alone. The largest number of early Latin inscriptions come from Rome, but most are short segments of texts incised on ceramic fragments; it is impossible to determine the epigraphic category and function in all but a few cases (Colonna 1980a). Early Latin inscriptions, considered as a whole, are not uniform in terms of ortho­ graphy or palaeography. As a result, developments can only be described in the broadest terms. The oldest Latin abecedarium, incised in right-to-left direction on a fragment of the foot of a bucchero cup dating to the second half of the 6th century, was recovered from the sanctuary of Juno Sospita at Lanuvium (Figures 2.6a and 2.6b). The abecedarium is incomplete; the first half of the series, up to the letter k, was lost because of damage to the cup. The series of letters reflects several reforms to the alphabet inherited from Etruscans. The number of letters in the series was reduced by elimination of tsade (san) and the supplemental letters phi and khi. Distinctive palaeographic features appear too – m with four bars, modern-looking n, diminutive o, hooked p, r with half-loop and retrograde s with four bars; these features are attested in Latin inscriptions of the second half of the 6th century bc. Letterforms extracted from twelve inscriptions – among the most substantial in the early Latin corpus – are presented in Table 2.3. Although the letterforms are approximations, it is possible to represent areas of ­palaeographic variation and change (highlighted). The letter m in the earliest inscriptions was written with five bars, the vertical bar varying in length. This form gives way to m with four bars in the second half of the 6th century. The earliest forms of p had the shape . Later forms were made with an open loop ( ), becoming similar in shape to r ( ). The form of s varied by the number of strokes – three, four or more – and by its orientation with respect to the direction of writing. Retrograde s is common. s with three-bars eventually supplanted the other variants, but this

https://doi.org/10.1017/9781108671132.003 Published online by Cambridge University Press

Alphabet, Epigraphy and Literacy in Central Italy

31

Figure 2.6a  Abecedarium of Lanuvium. Reproduced by permission of Daniele Maras. Figure 2.6b  Abecedarium of Lanuvium, drawing. Reproduced by permission of Daniele Maras.

form did not become standard until the 5th century. Variation existed too in the form of v; in a few inscriptions, the letterform Y was written instead of, or alongside of, V. The preference for Y in the Forum inscription may be due to the conservative nature of legal texts, and perhaps also to the fact that the inscription was incised under the management of public officials to be erected in a public space. Regardless, the form Y did not long remain in vogue. Considerable ink has been spilled over the shape of r in the rex inscription (for details, see Hartmann 2005: 274–6 and Wachter 1987: 94–6). The lower half of the letter’s vertical bar curves moderately in the direction of writing, a feature considered by some to be indicative of a later, ‘cursive’ form of r and therefore indicative of an inscription to be dated well into the republican era. However, the form of the descending stroke may just as easily be attributed to the difficulty of writing on the sloping surface of the inside of a cup. All early Latin inscriptions, whether from Rome or from ‘non-urban’ areas, were written with the half-loop form. The first ‘classical’-looking, bicaudal r appears on the Morai inscription from the Ager Signinus, which dates to c. 400 bc. In the earliest inscriptions, a was written with an oblique transverse bar, ascending or descending in the direction of writing. A ‘new’ form of a, written with its transverse bar detached from one of the oblique bars (⋀/), is attested in the second half of the 5th century on the Ianaias inscription. The form appears too in two inscriptions on stone recovered from Gabii, both probably dating to this same period (Fortson and Potter 2011; Johnston 2015). By the end of the century, the letter could be written with

https://doi.org/10.1017/9781108671132.003 Published online by Cambridge University Press

32

rex wallace Table 2.3  Letterforms in early Latin inscriptions a b

c

d

e

f

Notes: a. The letterforms are approximations. For detailed drawings of the letters in these inscriptions, see Urbanová 1999. For comparison of Roman and ‘non-urban’ letterforms, see Maras 2009a. For chronological developments, see Maras 2009c. b. The tri-colon separates reduplicated syllable and verb stem (fhe⁝fhaked). c. Add to the forms of punctuation, the ‘snail-like’ sign that separates the inscription into sections. d. The form of p is unique. The sign lacks a vertical bar. According to Maras (2009a), the vertical was omitted in order to distinguish p from r, which were, by c. 500, becoming similar in form, at least in the hands of some scribes. e. Punctuation in line 1 is missing the lower dot. It was obliterated by the nail hole in the middle of the plaque. f. The letter b, if indeed the letter is b and not f, has a larger loop on the bottom than the top. For discussion, see Vine 1998: 261.

∧,(⩚,). Such disarticuthe medial bar descending from the apex of the letter lated forms are a common variant in mid- to late-republican inscriptions. Changes in orthography accompanied changes to the forms of letters. By the middle of the 5th century, several orthographic changes had circulated throughout Latium creating a more uniform system of orthography. An important innovation, perhaps one of the earliest, involved the vowel letters i and v. The convention whereby the letter i stood for the

https://doi.org/10.1017/9781108671132.003 Published online by Cambridge University Press

Alphabet, Epigraphy and Literacy in Central Italy

33

vowels /i/ and /iː/ as well as the approximant /j/ was borrowed from the Etruscan writing system in which the letter i did double duty, representing both /j/ and /i/. In Latin (and Faliscan), the scope of the rule was extended to include the letter v, which represented the velar series of vowels and approximant /u, uː, w/. The first example of /w/ spelled by the letter v is attested in the Tita inscription (salvetod /salweːtoːd/). Praenestine fetusia /wetusia/ (f = ), which dates to the second quarter of the 7th century, adhered closely to Etruscan orthography, maintaining the distinction between f /w/ and v /u/. Interestingly, the letter f may have remained as an option to spell /w/ in Latin writing systems outside Rome, if it is indeed the case that the letter f in the Karkavios inscription from Aqua Acetosa represented /w/. Perhaps the most noteworthy orthographic change was the simplification of the digraph fh – which spelled the fricative /f/ – to f. This change appeared in Rome no later than the beginning of the 6th century. In the Duenos inscription, the letter f spelled the initial fricative of feced ‘made’. This change was made possible by the fact that the approximant /w/ was spelled by v; the final letter of the digraph was redundant. The spelling of the velar consonants /k, g, kw/ was another area in which there was orthographic change (see Table 2.2). The ci, ka, qu convention was employed in the Forum inscription and in the Kauidios inscription, but in other early Latin inscriptions, either c or k were the letters of choice, q appearing sometimes, though primarily as a sign for the labiovelar. The selection of c as the major representative of /k/ and /g/ was not firmly established until the 5th century. The digraph qv, or less commonly q, was used primarily, but not exclusively, to spell the voiceless labiovelar /kw/. q is attested as a spelling for /k/ sporadically throughout the republican period, particularly before v and o (for example, peqvnia, CIL 12.586; for discussion, see Adams, forthcoming). The spelling of the velars, as it appears in later republican inscriptions, did not fall into place until after the invention of the letter g, which took place at the beginning of the 3rd century. The oldest Latin inscriptions – the Fibula Praenestina and the Vetusia inscription – were written in right-to-left direction, following the Etruscan norm. Left-to-right direction is found before the end of the 7th century, perhaps under the influence of Etruscan writing at Caere and Veii. In these settlements, Etruscan scribes adopted left-to-right direction for a short period of time at the end of the 7th century, before returning to right-to-left direction. By the end of the 6th century, left-to-right direction is preferred for writing Latin. The most unusual writing style in the

https://doi.org/10.1017/9781108671132.003 Published online by Cambridge University Press

34

rex wallace

‘early’ Latin corpus is found on the Forum inscription. It was incised in boustrophedon, although some lines were inverted in the manner typical of what is known as ‘false’ boustrophedon. This style stands out as the only example in the early Latin corpus and one of the few in the Latin corpus as a whole. Most early Latin inscriptions lacked punctuation. Indeed, punctuation was rare in eighth- and seventh-century Euboean Greek inscriptions and it was rare in early Etruscan inscriptions too. The Fibula Praenestina, then, is a conspicuous exception, and not just for its word-punctuation. The reduplicated syllable of the verb fhe⁝fhaked ‘made’ was set off from the verb stem by a tri-colon. Inscriptions incised on stone, perhaps of a more official, public character, employed word-punctuation more consistently than inscriptions incised on ceramic. Both the Forum inscription and the altar base inscription from Corcolle have word-punctuation in the form of triand di-cola. Punctuation appeared too on the bronze lamina of Lavinium (Figure 2.7), separating the names of the divine heroes Castor and Pollux (castorei : podlovqveiqve). Particularly noteworthy, because it is so unusual in the earliest inscriptions, is the ‘higher level’ punctuation found in the Forum inscription (see Figure 2.2). A punctuation mark resembling a spiral-like sign separated the text into sections (see Vine 1993: 41–50). Although word-punctuation, especially in the form of a point set at midline level, became the norm in late republican Latin inscriptions, scriptio continua was possible well into the republican era. The layout of inscriptions must be taken into consideration too. Some appear, at least from the modern reader’s vantage point, to have been carelessly placed on the object on which they were incised or painted. Others were more skilfully displayed. The layout of the Lapis Satricanus, for example, was carefully planned (Figure 2.8); the second line of the inscription was centred on the stone with respect to the first. The inscription on the bronze lamina from Lavinium was incised in two lines, both right-­justified. The names of the heroes, Castor and Pollux, were placed on line one (the last two letters of the enclitic conjunction spill over into line two). The appositional noun, qvrois ‘sons (of Zeus)’, was placed at the beginning of line 2. Unfortunately, the nail-holes in the bronze detract from this aesthetically pleasing arrangement. Alphabets and writing systems do not exist in a vacuum. Potters and engravers were mobile, relocating at workshops in different communities. It is not surprising then that external influences in writing styles can sometimes be detected in the early Latin corpus. The most striking example is the inscription on the altar base from Tibur (Figures 2.9–2.10).

https://doi.org/10.1017/9781108671132.003 Published online by Cambridge University Press

Alphabet, Epigraphy and Literacy in Central Italy

35

Figure 2.7  The Madonetta inscription (CIL 12.2833). Reproduced by permission of the Center for Epigraphical and Paleographical Studies, The Ohio State University.

Figure 2.8  The Lapis Satricanus (CIL 12.2832a). Reproduced by permission of the Center for Epigraphical and Paleographical Studies, The Ohio State University.

Figure 2.9  The Tibur inscription, drawing (CIL 12.2658): Reproduced by permission of the Center for Epigraphical and Paleographical Studies, The Ohio State University.

The inscription was laid out on the stone in a ‘spiraliform’ style reminiscent of South Picene funerary stelae; the inverted v and l recall those on South Picene inscriptions as well. For example, the letters v (kavios) and l (fileod) were inverted. Tokens of m and n were written retrograde;

https://doi.org/10.1017/9781108671132.003 Published online by Cambridge University Press

36

rex wallace

Figure 2.10  The Tibur inscription (CIL 12.2658): Reproduced by permission of the Center for Epigraphical and Paleographical Studies, The Ohio State University.

and most tellingly, o had the form of a point (see Table 2.3). These features may be attributed to a stone-cutter who, having worked in the medioAdriatic area, perhaps on South Picene stelae, relocated to Tibur and carved in a workshop there (Wallace 1987: 128).

2.5 Literacy The topic of literacy in ancient Italy is beset with difficulties (for discussion, see Harris 1989). The evidence for the period discussed in this chapter is limited to inscriptions belonging to a few epigraphic categories. The quantity is limited too; apart from Etruscan, the number of inscriptions in other languages is small. The quality of the corpus is also an issue. Some inscriptions are fragmentary; others are difficult to interpret. As a result, it is impossible to reach reliable conclusions about the state of literacy in central Italy for the period in question. For the 7th and 6th centuries, the epigraphic evidence indicates that writing was in the hands of the aristocracy and the scribes and craftsmen who worked in their employ. Roman historical sources, if not entirely unreliable, provide some insight into the world of writing beyond what is attested by inscriptions (Cornell 1991). Laws were copied onto tablets for public display; treaties were drafted; commentaries compiled about

https://doi.org/10.1017/9781108671132.003 Published online by Cambridge University Press

Alphabet, Epigraphy and Literacy in Central Italy

37

religious practices. There is mention too of archives housing both public and private documents. Texts of a temporary nature were written on perishable materials. Replicas of writing tablets and styluses – some on precious material such as ivory – were placed in aristocratic tombs as status symbols as early as 650 bc. Representations of linen books, wooden tablets and papyrus scrolls appear together with images of the deceased on Etruscan sarcophagi. Religious sanctuaries appear to have been the institution par excellence for the cultivation and transmission of writing (Maras 2015: 210–13). Sanctuaries sponsored workshops whose artisans composed votive texts for believers and celebrants. At Veii, craftsmen working in the Portonaccio sanctuary fashioned vases and inscribed them with texts to the goddess Menerva. They employed distinctive spelling conventions in the inscriptions, for example, xi (𐌗) for /s/, and they developed a unique form of punctuation whereby the final letter of a closed syllable was marked by dots or points, ostensibly as a means of facilitating the instruction of writing (Maras 2015: 213–14; Wallace 2008: 26). The scribes were also responsible for adapting the structure of donative inscriptions for votive purposes (see III.3). Towards the end of the 6th century, a scriptorium was active in the sanctuary of the port city of Pyrgi. Scribes incised votive inscriptions to the divinities Cavatha, Shuri and Uni on vases offered by worshippers. A writing school may also have existed in Latium at the sanctuary of Juno Sospita at Lanuvium, where a Latin abecedarium dating to the end of the 6th century was found (Attenni and Maras 2004: 68–78; Maras 2015: 213). Scribal schools also served as an important conduit for the spread of writing. Scribes at the sanctuary of the goddess Reitia at Este in the Veneto borrowed the convention of syllabic punctuation from their Etruscan counterparts at Veii. Workshops producing pottery, sometimes under the direction of elite families, employed artisans who were literate and who were familiar with the epigraphic categories current in the 7th and 6th centuries. These workshops and their craftsmen were another important source of the transmission of writing. A remarkable example from Etruria illustrates this point. Among the oldest Etruscan inscriptions are those on custom-made, relief-ware bucchero drinking cups that were presented as gifts in order to cement social or economic relationships. The inscriptions spiraled up around the conical bases of the cups in a script with several letterforms unique to the artisan and the workshop in which they were produced. Cups manufactured at Caere and commissioned by the members of the Paithinas family were discovered at sites in the mineral rich

https://doi.org/10.1017/9781108671132.003 Published online by Cambridge University Press

38

rex wallace

area of the Colline Metallifere in northern Etruria (see inscription II.2). Cups similar in style to those made at Caere, bearing inscriptions with the same unusual letterforms and bearing the names of members of the same Paithinas family, were also made at Vetulonia (see II.3). Artisans who worked for the Paithinas family at Caere relocated to Vetulonia and set up shop there (Maras 2015: 207–8). In this instance, it is possible to construct the social network of the family who commissioned and distributed the cups, the movement of their artisans and the routes by which writing was transmitted. Objects used in textile production – spools, spindle whorls and loom weights – were frequently marked with letters or with non-alphabetic signs, and occasionally incised with texts (see Tuck and Wallace 2012). In antiquity, the production of fabrics was a task typically associated with women. As a result, some have argued that women may have been involved in the spread of literacy (Bagnasco 1999; Camporeale 2015: 21; Maras 2015: 209–10). The claim cannot be verified, but it is conceivable that female members of elite families learned to read and write. Etruscan, Latin and Faliscan proprietary inscriptions of the 7th century bc were sometimes incised with the names of women. A passage in Livy may also bear, albeit obliquely, on the question of the literacy of women during this period. At 1.34.9, he informs us that Tanaquil, the wife of the Etruscan king Tarquinius Priscus, was well schooled in the interpretation of portents, a responsibility that presumably required some degree of literacy (Maras 2015: 210). Poggio Civitate (Murlo), an Etruscan hilltop community inhabited by an aristocratic family, artisans and craftsmen, provides another perspective from which to assess the question of literacy during the 7th and 6th centuries (Wallace, forthcoming). Locally-produced ceramic objects recovered from the site’s aristocratic residence were sometimes incised with proprietary inscriptions. Ivory plaques carved in the form of the mistress of the hunt or her animal avatars were incised with the names of the aristocratic family’s social and/or commercial associates. Imported relief-ware bucchero cups, several engraved with muluvanice inscriptions, were recovered from the cella of the site’s religious building. An inscribed bucchero cup, perhaps manufactured at Vetulonia, was found in the remains of the site’s workshop where it may have served as a model for the production of reliefware cups and for writing on ceramic materials (Tuck and Wallace 2018c). More intriguing, but more difficult to evaluate, is a terracotta tile fragment on which the initial letters of the word raσ[na- /rasna-/ ‘public’ were

https://doi.org/10.1017/9781108671132.003 Published online by Cambridge University Press

Alphabet, Epigraphy and Literacy in Central Italy

39

written. The tile was unearthed among the remains of the site’s 6th century monumental edifice. It is tempting to think that the word raσ[na- was part of an inscription referring to the zilaθ meχl raσnal, that is to say, to the ‘leader of the community’. The epigraphic evidence recovered from the Poggio Civitate, coming as it does from the archaeological context of a pre-urban settlement, suggests that its inhabitants read and wrote in a wider range of contexts than can be inferred from inscriptions recovered from burials and sanctuaries (Tuck and Wallace 2018b). The observations in this section, though limited in scope, provide a glimpse of the state of literacy during the period in question.

2.6 Conclusion For most of the peoples of central Italy, writing systems based on Etruscan or Umbrian models were in use before the end of the 6th century bc. They rapidly diversified, developing distinctive characteristics in the shapes of letters, writing styles and spelling conventions. Unfortunately, many of these regional varieties did not long survive; some appear to have been lost before the end of the 5th century. Inscriptions in the early Umbrian alphabet are not attested after the middle of the 6th century. Umbrian inscriptions of the Hellenistic period were composed in alphabets borrowed from Etruscan and, later, from Latin. South Picene inscriptions are not found later than the 5th century, although the alphabet appears to have been used to inscribe toponyms on two bronze helmets dating to c. 300 bc (Marinetti 1985: 252–5). A Hernican inscription dating to the 3rd century (ST He 3) was written in a Latin alphabet (Colonna and Gatti 1992). The native alphabet must have passed after Hernican settlements were annexed by Rome (c. 306 bc). The Faliscan alphabet survived into the late republican period, although letterforms, orthography and language were under increasing pressure from Latin. A small cache of inscriptions recovered from Capena, dating to c. 300 bc, were written in a Latin-based alphabet. The inscriptions incorporated two letters of the early Umbrian and South Picene alphabets, ú ( ) and í (⋈), as well as the convention of inverting letters (u and m). The language of the inscriptions is principally Sabellic, but the inflection of onomastic forms reveals Latin influence (Bakkum 1996). The loss of native alphabets and writing systems in cases such as these foreshadow the demise of the languages in which they were written, not only in central Italy, but throughout the Italian peninsula. 𐌣

https://doi.org/10.1017/9781108671132.003 Published online by Cambridge University Press

40

rex wallace Acknowledgements

I thank Wendy Watkins, curator of the Center for Epigraphical and Paleographical Studies at The Ohio State University, for assisting me in assembling the photographs of Latin inscriptions. I also thank my colleague, Daniele Maras, for permitting me to publish his photograph and drawing of the Latin abecedarium from Lanuvium. Finally, I am grateful to the editors of this volume and to my colleague, Bill Regier, for comments on my contribution. They do not share all of the views expressed here; they are not responsible for any errors.

https://doi.org/10.1017/9781108671132.003 Published online by Cambridge University Press

chapter 3

Identifying Latin in Early Inscriptions Simona Marchesini

3.1 Introduction The corpus of Latin inscriptions surviving from the beginning of the 4th to the end of the 2nd century bc, while undoubtedly more consistent than the antiquissimae dated from the 7th to the end of the 5th century (CIL 12 Pars prima) dealt with in the previous chapter by Wallace, nevertheless presents several methodological issues, which have to be discussed before any in-depth analysis of single inscriptions or group of inscriptions may be attempted. These limitations restrict the overall hermeneutic approach to the corpus and reduce its representativity in respect of the study of Latin as it was spoken and written in the 4th to the 2nd centuries bc. During this period, which must be considered in its historical framework, the role played by Rome not only in the Italic context, but also in the Mediterranean, increases dramatically. Originally a small town in Latium, Rome was in this period rising to become the centre of an extended empire (Coarelli 2011: 85–8; Nonnis 2012: 136). Additionally, probably also as consequence of this new role, Rome ‘creates’ its literary tradition and first schools of grammar during the 3rd century bc. It cannot be forgotten that Livius Andronicus and Spurius Carvilius, the initiators of these two traditions, came from the Greek milieu of Southern Italy (272 bc: defeat of Taranto and arrival of Livius Andronicus at Rome; see Wachter 1987: 338). It is therefore unsurprising that the inscriptions of this period present a m ­ ultifaceted and sometimes mixed situation in relation to graphemic and phono­ lo­gical features. Fluctuations between archaic and innovative traits ­characterise a differentiated level of literacy in the documents from Rome and from the neighbouring towns and districts of old Latium, such as Praeneste, Tusculum and Ardea. 41

https://doi.org/10.1017/9781108671132.004 Published online by Cambridge University Press

42

Simona Marchesini 3.1.1  Etymological versus Phonetic Solutions

Sometimes phonetic spellings, which seem to mirror the spoken l­ anguage – for example1 cosoleretur versus consoluerunt in the Senatus consultum de Bacchanalibus (CIL 12 581), or cosentiont in an Elogium Scipionum (CIL 12.9) – alternate, even in the same inscription, with etymological reconstructions (consol, or even cesor versus censur; see Wachter 1987: 304). The level of literacy, considered before the spreading and the strengthening of a literary tradition, is often limited, thus producing varying solutions to the same phonological issues. 3.1.2  Personal Names The majority of inscriptions, as in almost all the languages of ancient Italy, contain exclusively personal names, which represent more than 90% of the total lexicon. Personal names have a special and independent status within the lexicon of a language, and as such they exhibit a grammar of their own and often preserve archaic traits even in an innovative milieu. For this reason, it is difficult to argue for grammar rules starting from the propria. The remaining 10% of lexemes, as we will see, may be considered to be more faithful records of spoken practices and therefore more useful for an inquiry into questions relating to spoken versus standard language. 3.1.3  Lexical Categories Represented in the Inscriptions Another limitation affects the evidence offered by the other lexical categories of the given records: the inscriptions known to date belong mostly to a few textual categories, which generally do not include, for example, references to future actions or facts, and therefore do not generally exhibit expressions of futurity in either the verb or noun system. The short texts, which can be attributed mostly to a religious context or are private dedications, are not elaborated (in the sense, for example, in which philosophical or technical treatises may be viewed as ‘elaborated’ or ‘speculative’), and only in few cases is the text longer than a line or two. Many lexical and grammatical categories are thus missing. The verbs are mostly expressed in the past and in the present tenses, the persons are the third singular or the third plural. The few lemmata belonging to an institutional lexicon (e.g. cosul, praetor, iurato, 1

The conventions I follow for the transcription of pre-Roman languages are italics for Etruscan, Faliscan and Latin, and boldface for the Sabellic languages.

https://doi.org/10.1017/9781108671132.004 Published online by Cambridge University Press

Identifying Latin in Early Inscriptions

43

coirauerunt, iustei, sententiad) are indeed very interesting, as they shed light on the historical and social frameworks of Latin society at the time. 3.1.4  Chronological Aspects The last issue, which is overwhelming for the inscriptions dated between the 4th and the 2nd century, is the chronological uncertainty of their execution. Recurring archaisms, which appear vis-à-vis innovative features in the same document, prevent us from establishing the date of the inscriptions unambiguously. Only twelve out of the very ancient inscriptions (antiquissimae, see Wallace in Chapter 2 of this volume, Table 2.1, nos. 1, 2, 7, 8, 9, 12, 18, 20, 21, 22, 23, 24) have been reassessed by means of modern archaeological and epigraphical considerations (Hartmann 2005) and in many cases with a very large chronological span. Later inscriptions, which are the focus of this chapter, are still difficult to attribute to a specific context, owing to the lack of archaeological evidence and to the insufficient dating clues that the few stylistic elements displayed in the inscribed objects offer. Yet some graphemic solutions of phonetic features, such as the introduction of the letter by Spurius Carvilius (Scaur. GL 7.15–16 = 4.9.5 Biddau = GRF p. 3), or the monophthongisation of oi to o (Wachter 1987: 127, 187) might establish a terminus post quem for the chronology of some texts. As it seems impossible to reconstruct the chronological seriation of all the inscriptions – that is, the relative dating method based on the coherent association of letter types (definition and examples in Marchesini 2004 and 2012; an attempt of seriation for Latin inscriptions from the 7th to the 5th century in Maras 2009c) – and many of them may be ascribed to a large time span, it is difficult to follow the development of the language and to establish even a relative chronology of grammatical features. An enlightening case comes from the inscriptions on the two sarcophagi of the Scipiones, the chronology and historical value of which have been extensively discussed among specialists, historians, linguists and archaeologists (see 3.2.4.2 in section 3.2.4, and section 3.7) without any conclusive agreement. The catalogue of the epigraphic collection of the Museo delle Terme in Rome, published in 2012, includes some of the inscriptions of this period which were considered chronologically uncertain by Wachter (1987): among these are CIL 12.545, dated towards the end of the 3rd century on the basis of epigraphic and linguistic features (Nonnis 2012: 27 = Wachter 1987: 469) and CIL 2.477, dated by Caruso to the middle of the 2nd century also on the basis of linguistic and epigraphic features (Caruso 2012: 29 = Wachter 1987: 347, l).

https://doi.org/10.1017/9781108671132.004 Published online by Cambridge University Press

44

Simona Marchesini 3.1.5  Tendency to Preserve Fossilised Features

In inscriptions of the period between the 4th and 2nd centuries, we observe the tendency to preserve fossilised features of morphology and spelling, both in personal names and in other nouns. Examples of such features are the loss of final -s (see section 3.4.8) and the preservation of pre-monophthongised forms (section 3.4.6). We must emphasise once more that archaic and/or archaising traits, where given, undermine the representativity of inscriptions as reliable linguistic sources (Penney 2011: 221). 3.1.6 Faliscan Another aspect which deserves attention is the relationship between Latin and Faliscan. The c. 350 short inscriptions coming from Falerii Veteres (Civita Castellana), from Falerii Noui – the city of southern Etruria rebuilt by the Romans after the conquest of Falerii Veteres in 241 bc – and its surroundings, dated from the 7th to the 2nd centuries bc, must be considered the expression of a dialect variety close to Latin but divergent in some aspects. There is no agreement as to the status of Faliscan, viewed by some as the language closest to Latin but separate from it (see for example Wallace and Joseph 1991) and by others as a dialect of Latin, albeit divergent in some respects (see Weiss 2020: 16–17). The main distinguishing feature is the treatment of the PIE aspirates, which result in Faliscan word-initial /f/ and /h/ as developments of both *bh/*dh and *gh (Fal. hileo, fe), whereas Latin and Oscan–Umbrian have respectively /f/ from *bh/*dh (filius), and /h/ from *gh (hic). Wallace and Joseph (1991), after reconsidering the evidence in a diachronic perspective, explain this unexpected behaviour as a sound change (f > h) and ‘subsequent hypercorrection as the source of Faliscan f/h variation, [which] happened after the 4th century’ (see also Stuart-Smith 2004: 61–2; for a consideration of Faliscan as a dialect sharing strong similarities with the Latin language due to a common stage of development called ‘Urlatinofaliskisch’ (see Meiser 1998: 9–10). Depending on whether the separation appears justified or, vice versa, the Faliscan evidence is to be included in our consideration of ‘early Latin’, differences arise in the evaluation of some phonological and morphological aspects. In support of the Latin-Faliscan hypothesis one could mention the fact that the first record of (Latin) rhotacism comes from

https://doi.org/10.1017/9781108671132.004 Published online by Cambridge University Press

Identifying Latin in Early Inscriptions

45

Faliscan (Bakkum 2009, II, Nr. 59–60), where the verbal form carefo (< *kasēfō: ‘to be free from’, ‘to lack’, first pers. sing. future indicative) presents the development /s/ > /r/, which Latin completed by the early 4th century (Fortson 2011b: 203–4; for rhotacism see also section 3.4.3). Another relevant feature is the preservation of the archaic form of the genitive singular *-osi̯o, as displayed in the inscription on the impasto pitcher from Civita Castellana (Bakkum 2009, II, Nr. 3, p. 409), where the form euotenosio is recorded. Even if this inscription is dated to the 7th century bc, the development of this genitive is later represented in Faliscan by the form titoio (*