The Roman Clan: The Gens from Ancient Ideology to Modern Anthropology (W.B. Stanford Memorial Lectures) 0521856922, 9780521856928, 9780511161827

Citation preview

This page intentionally left blank

T H E RO M A N C L A N

The gens, a key social formation in archaic Rome, has given rise to much controversy in modern scholarship. In this comprehensive exploration of the subject, Professor Smith examines the mismatch between the ancient evidence and modern interpretative models influenced by social anthropology and political theory. He offers a detailed comparison of the gens with the Attic genos and illustrates, for the first time, how recent changes in the way we understand the genos may impact upon our understanding of Roman history. He develops a concept of the gens within the interlocking communal institutions of early Rome which touches on questions of land-ownership, warfare and the patriciate, before offering an explanation of the role of the gens and the part it might play in modern political theory. This significant work makes an important contribution not only to the study of archaic Rome, but also to the history of ideas. c . j . smi t h is Professor of Ancient History at the University of St Andrews. His previous publications include Early Rome and Latium: Economy and Society c. 1000 to 500 bc (1996), Trade, Traders and the Ancient City (1998), Religion in Archaic and Republican Rome (2000) and Sicily from Aeneas to Augustus (2000). He is the co-editor of Fragmentary Roman Historians (forthcoming).

THE W. B. STANFORD MEMORIAL LECTURES

This lecture series was established by public subscription, to honour the memory of William Bedell Stanford, Regius Professor of Greek in Trinity College, Dublin, from 1940 to 1980, and Chancellor of the University of Dublin from 1982 to 1984.

T H E RO M A N C L A N The gens from ancient ideology to modern anthropology

C. J. SMITH

   Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo Cambridge University Press The Edinburgh Building, Cambridge  , UK Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9780521856928 © C. J. Smith 2006 This publication is in copyright. Subject to statutory exception and to the provision of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published in print format 2006 - -

---- eBook (EBL) --- eBook (EBL)

- -

---- hardback --- hardback

Cambridge University Press has no responsibility for the persistence or accuracy of s 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.

For Susan

Contents

List of illustrations Preface List of abbreviations

page ix xi xiii

General introduction

1

pa rt i the evidence f or the g e n s 1 The ancient evidence 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

12

Introduction Etymology, meaning and definition The gens: A definition The nomen The twelve tables Gentilicial exogamy and the nature of the gens Legendary and historical genealogies Gentes, cults and mores The gens at law The gens in Cicero and Livy Conclusion

2 Modern interpretations 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

12 13 15 17 21 30 32 44 51 55 63

65

Introduction Carlo Sigonio Giambattista Vico and his influence Niebuhr and the beginnings of modern ancient history Theodor Mommsen Lewis Henry Morgan Morgan’s Marxist legacy Fustel de Coulanges Henry Sumner Maine The gens in the early twentieth century Conclusion

vi

65 66 71 81 85 88 99 101 104 108 112

Contents 3 The gens in the mirror: Roman gens and Attic genos 1. 2. 3. 4. 5. 6. 7. 8. 9.

Introduction Early views of the genos Bourriot and Roussel: A French critique The evidence: Philochoros and the Athenaion Politeia The evidence: Demotionidai and Dekeleieis The evidence: Gene¯ and Cults The gene¯ and the Eupatridai Reconstructing the genos The genos and the gens

4 Archaeology and the gens 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

Introduction From Bronze Age to Iron Age The gens at Osteria dell’Osa The gens at Satricum The Auditorio site Summary The Etruscans Onomastic structures The evidence of chamber tombs Crisis Part I: Conclusion

vii 114 114 115 117 121 126 130 133 134 140

144 144 144 147 150 153 155 156 158 160 160 164

part ii toward s an interpretation of the g e n s 5 The Roman community

168

1. Introduction 2. Clientes and plebeians 3. Historical outline

168 168 176

6 The Roman curiae 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

Introduction The origins of the curiae Dionysius of Halicarnassus and the Roman curiae Curiae and Quirites The names and locations of the curiae The religious aspects of the curiae The military functions of the curiae The political functions of the curiae The lex curiata The political assemblies of the curiae Curiae and tribus The archaic curiae

184 184 186 192 198 202 205 208 210 217 223 225 230

Contents

viii 7 The patricians and the land 1. Introduction 2. Gentes and tribus

8 The patriciate 1. 2. 3. 4. 5. 6. 7. 8. 9.

Introduction Patres, patricii, and the senate Patrician privileges Priestly office Auspicia Political office Patrum auctoritas Patricians and plebeians The problem of the patriciate

9 Warfare in the regal and early Republican periods 1. The early Roman army 2. The gens Fabia at Veii 3. Conclusion

10 Explaining the gens

235 235 236

251 251 252 258 260 263 268 273 275 278

281 281 290 295

299

Introduction Patricians and gentes in the regal period Patricians and gentes in the early Republic Patricians, gentes and the community The gens in the Roman historical tradition

299 302 306 315 324

11 Roman history and the modern world

336

1. 2. 3. 4. 5.

Appendix 1: Dionysius of Halicarnassus on the Roman curiae and religion

347

Appendix 2: The missing curiae

356

Select bibliography General index Index of ancient persons Index of passages discussed

363 384 388 390

Illustrations

1 Schematic representation of relationship between gens and agnatic descent groups; after Linke (1995) 25 2 Schematic representation of intermarriage between gentes; after Linke (1995) 35 3 The centuriate organisation

ix

page 16 33 283

Preface

The origins of this book lie in my dissatisfaction with my own account of the gens in a previous account of early Rome. For almost a decade, I have been trying to improve upon a few sentences in an earlier book, and I am slightly horrified that the result is as long as it is. Perhaps that is a testament to my obsession, but I hope it also reflects the importance of the subject. Over such a long period, more people have contributed in direct and indirect ways than any preface can acknowledge. Nonetheless it is a pleasure to bring to mind at least some of my debts. Early versions of parts of the argument were offered as papers to stimulating audiences at Leicester, Cambridge and St Andrews. The real impetus to put my thoughts down as a whole came from the very kind invitation from Trinity College Dublin to present this as the Stanford Lectures in spring 2001. I am immensely indebted to my friends and colleagues there for their hospitality, and the stimulating environment in which I was able to present an early version of these ideas – I hope now to have answered some of the questions so pertinently posed on those occasions. I was then fortunate enough to gain research leave from the AHRB, which gave an invaluable period of time to think, reflect and write. By that stage the revision of H. Peter’s Historicorum Romanorum Reliquiae of which I am an editor was well under way. This has been an immensely formative intellectual journey for me, and much of what follows has been coloured by insights gained from that enterprise. It has deepened relationships already formed and created new ones, and to all involved I am very grateful. Several colleagues kindly commented on all or part of a draft, and I am both lucky in my friends, and indebted to them. Michael Crawford and Jill Harries helped me with Roman law, and Robin Osborne and Robert Parker with the chapter on the genos; in so doing they improved the whole immeasurably. Ed Bispham, Guy Bradley and Fay Glinister provided insights and encouragement just when I needed them, and made me feel that the project xi

xii

Preface

was worthwhile. Sally Humphreys provided immensely valuable comments; Nicola Terrenato was both a generous but also a discerning reader of the text, and I was enormously pleased to be able to talk through some of the issues with him and Laura Motta in Chapel Hill, where by good fortune I was also able to discuss Osteria dell’Osa with Anna Maria Bietti Sestieri. Andrew Lintott and Fergus Millar were immensely supportive, as always. Tim Cornell, Andrew Drummond and John Rich read between them several drafts, commented in great detail, and have improved, quite literally, every page of what follows by their good sense, acumen and scholarship. This was particularly kind, since I have from time to time had the temerity to disagree with them. There is but a pale reflection of the dialogue which I have enjoyed with each of them, and their published works, in this book. Colleagues in St Andrews have been remarkably tolerant of and responsive to my importuning in many ways, and not just in Classics; Nigel Rapport helped me with some of the anthropological issues, and the University Library staff were tremendously helpful; in this context I would also like to thank Sophia Fisher at the Institute of Classical Studies Library. I owe a particular debt to Iveta Adams for her helpful and extremely valuable copy-editing in the final stages. I have learnt much from all those who have commented on the text, or discussed my ideas, and I have been hugely encouraged by their support. They are not responsible for the positions I have taken, or the errors I have made, but they, and many others who I have bothered and badgered, have contributed both to whatever is good in the book, and to my sense of purpose in writing it. In what follows, despite my best endeavours, much is prefaced by ‘perhaps’ and ‘maybe’ so one definitive statement seems appropriate. Susan, who has read this book several times, always with tremendous acuity and faith, and has lived with it for almost as long as I have, gave me the strength to finish what I often thought was beyond me; no dedication could ever say enough.

Abbreviations

Bremer IA CAH 2 CIL FGrH Funaioli GRF GL IG ILLRP ILS Inscr. Ital. XIII LIMC LTUR Malcovati ORF MRR RE Roman Statutes RRC TLL

F. P. Bremer, Iurisprudentiae antehadrianae quae supersunt (Leipzig 1896–1901). The Cambridge Ancient History (2nd edn.) (Cambridge, 1961–). Corpus Inscriptionum Latinarum (Berlin 1862–). F. Jacoby (ed.), Die Fragmente der griechischen Historiker (Leiden 1923–50). H. Funaioli (ed.), Grammaticae Romanae fragmenta (Leipzig 1907). H. Keil (ed.), Grammatici Latini (Leipzig 1857–80). Inscriptiones Graecae (Berlin 1873–). A. Degrassi (ed.), Inscriptiones Latinae liberae rei publicae (Florence 1957–65). H. Dessau (ed.), Inscriptiones Latinae selectae (Berlin 1954). A. Degrassi (ed.), Inscriptiones Italiae XIII (Rome 1937–63). Lexicon iconographicum mythologiae classicae (Zurich and Munich 1981–99). E. M. Steinby (ed.), Lexicon topographicum urbis Romae (Rome 1993–2000). H. Malcovati (ed.), Oratorum Romanorum fragmenta (Turin 1967). T. R. S. Broughton, The Magistrates of the Roman Republic (New York 1951–86). A. Pauly, G. Wissowa and W. Kroll (eds.), Realencyclop¨adie der classischen Altertumswissenschaft (Stuttgart 1894–1980). M. H. Crawford (ed.), Roman Statutes (BICS Suppl. 64) (London 1996). M. H. Crawford, Roman Republican Coinage (Cambridge 1974). Thesaurus linguae Latinae (Leipzig 1900–). xiii

General introduction

The gens, of all Roman institutions, is the one most alluded to and least explained. Only the absolute power of a father over his son has had such influence in subsequent philosophical and political thought.1 Historians have made the gens the key to Roman politics, archaeologists have sought the gens on the ground, and both have described as ‘gentilicial’ a huge array of activities and traces of social behaviour. Early modern thinkers found the justification for their definitions of contemporary nobility in the concept of gentilitas. Social anthropologists have used the gens as a model to help them understand societies as distant as Africa and native America. Engels developed Marx’s belief that the Roman gens helped to explain the origin of private property. One of the most profound divisions in twentieth-century Italian jurisprudence has been between those who thought the gens (embodied by family) predated the state, and those who saw it as the product of the state. This debate is not only still ongoing, but also shadows a much wider, and much deeper, concern in modern thought about the nature of identity, as a real ethnic, biological fact, or a fictitious, political fig leaf concealing darker motives and deeper fears. Yet there has been no substantial treatment of the gens in English for nearly a century, and none that I know of in any language which sets out to establish both the reality of the institution, and the myriad interpretations that have been laid upon it. This book is therefore at its heart the history of a debate which began in antiquity, and which, in unexpected ways and along surprising paths, continues to be relevant to this day. The history of the Roman gens has fascinated scholars for centuries, and the obscurity of the institution has not prevented imaginative if unfounded reconstruction. This book has two themes, an attempt to state as clearly as possible the evidence for the role of the gens in early Roman history, 1

For a powerful attempt to demonstrate that the power of life and death held by a Roman paterfamilias was a myth, but one deeply embedded in the Roman self-image, see Shaw (2001).

1

2

The Roman Clan

and an account of the use which scholars have made of the gens from the early modern period onwards. The reason for this twofold purpose is simple. My argument is that we have largely misunderstood the nature of the Roman gens because we have brought to it preconceptions which derive from the way the gens has been presented in other disciplines such as social anthropology. For the gens, of all Roman institutions, has been unusually significant outside the ancient historical discourse. The word itself is very difficult to translate without importing meanings from other contexts. ‘Clan’ has been popular, though ‘House’ has had its champions, especially in the early translations of Niebuhr and Mommsen. Modern anthropologists distinguish ‘lineage’ from clan through the accuracy of genealogical knowledge; a lineage traces descent from a common ancestor through known links, and a clan is a group where the genealogical links are not all known. This definition of a clan would actually fit the gens well, since, as we shall see, the relationship between members was based on a largely fictive kinship.2 In some parts of the study of kinship, ‘gens’ is used itself as a technical term. On the whole I have tried to avoid translation, but where appropriate I have tended to use ‘clan’, and for one important reason, which is that this most aptly indicates the way that the concept has travelled far beyond its own time and place. The translation is intended more as a signifier of the dangers of the cross-cultural comparison.3 Paradoxically, this book will argue that both in its own time and subsequently, the Roman gens has been more important as part of an argument than as a social institution. Most Romans did not regularly and explicitly refer to their gens as their core self-definition, but it represented aspects of aristocratic behaviour which were important, and were disputed and contested. Much of this work will focus on the debates about the community at Rome in the fifth and fourth centuries bc, a period characterised by the sources as one of violent and bitter antipathy between patricians and plebeians. At the same time, the book will demonstrate that, consciously and unconsciously, scholars from Vico to Morgan and Maine used the perception of the gens to reinforce their vision of the world. The disjunction between the Roman reality and the presentation in later writers is 2

3

In an influential account, Finley (1983) 45 rejected the concept of the clan, and insisted on lineage. He was denying the importance of any form of kinship as a basis of power. Finley stated that the gens unlike the genos was a lineage, and nothing like a clan or tribe, but it is not clear to me that Finley’s usage is exactly the same as modern anthropologists’ usage, and it makes assumptions about the ancestry of these lineages which are difficult. Nevertheless, Finley was correct in his observation that one cannot find either a tribal or a feudal system behind the gens. For definitions of clans and lineages, see Stone (1997) 62–6; Parkin (1997) 17–18.

General introduction

3

sometimes very sharp, and has led to the confusing use of a concept of a ‘gentilicial’ society in archaic central Italy which actually has very little relationship to the social institution from which the adjective is derived. This argument has profound consequences for archaeological research in particular. In seeking a new understanding of the Roman gens, our argument flows naturally into many areas of early Roman history at least as obscure as, and if anything more controversial than, the gens itself. The connection between the gentes and the Roman patriciate (a hereditary nobility) is undeniable, though the extent to which the connection was exclusive is one of the most difficult of issues. It is impossible to understand the gens without looking at Roman society as a whole, and this book will lay unusual emphasis on the importance of the citizen community in the early Republic. To this end, we shall discuss in detail both the patriciate and the assemblies of the people, most particularly the curiate assembly, which has received a great deal of rather eccentric analysis over the past thirty or so years. Much of this has been dismissed, and the institution has consequently been rather underestimated in standard accounts. These are not digressions, however, but form part of a vision of the early Republic which stresses the wider social context, and which finds a role for the gens within that community. On this argument, the gens was not an obscure archaic survival, which was symptomatic of aristocratic disdain for community, but a form of organisation which reveals the tense and difficult negotiation of power between aristocracy and people. This argument is itself not without consequences for the way we see the Struggle of the Orders, a modern phrase describing the rivalry between patricians and plebeians which emerges clearly from the sources, and was an aspect of Roman history of intense interest to the early modern period. Since this book is neither a history of Rome at that time, nor an account of the influence of the Roman narrative on later periods (both of which would be valuable nonetheless), it is at times much more cursory than the subject matter deserves, but it is my hope that it will contribute to the debate on both topics. What underlies this account as a whole is a belief first that Roman history has been and continues to be of central importance in political discourse, and second that Roman politics was neither without ideological argument, nor alien to ideas of democracy. No apologist can make an Athens out of Rome, but at the same time, far too little is made of the important debate, which I believe was perennial at Rome, over the nature of community, the proper roles and duties of its citizens, and the interlocking of the various

4

The Roman Clan

institutions of the city state. There is a temptation to see regal and early Republican Rome merely as a kind of idea-free bloodbath, a militaristic machine run by selfish aristocrats, and tergiversant demagogues only too willing to pull the ladder of political advantage up after themselves. Readers will find a different early Rome here. Every argument about early Rome is also an argument about sources, and indeed has been ever since de Beaufort set out to demonstrate the unreliability of the tradition in the eighteenth century. The problem is easily stated: none of our sources were remotely contemporary with events (the first Roman historian wrote at the end of the third century bc), and they are themselves pessimistic about the quality of the information which they had to go on. Yet write they did, and demonstrably with the concerns and the political language of their own day. What evidence exists for the nature and functions of the gens, how reliable is it, and can we reconstruct a reliable picture of the role of the gens in early Rome? Chapter 1 sets out all the evidence which can be used in the process of defining the gens. It is varied in nature; we have legal definitions, some of considerable antiquity, others much later, some stories in historians or antiquarian writers, etymologies, and snippets of information about customs and practices. None of this can be overlooked, but equally, it may not all be of the same value. We must not underestimate the difficulty late Republican and early imperial writers had in discovering about their more distant past, or the extent to which they could resort to weak arguments and invention to supplement the facts as they saw them. In chapter 2, I will argue that preconceptions and methodologies can influence conclusions; the same was true of ancient writers. The additional important source we have is the names of magistrates from the beginning of the Republic. This is far from uncontroversial, and we shall have to address it in detail, but readers should be aware of my position from the outset. We can produce a list of magistrates from a variety of sources that survive to us: the historians, primarily Livy, Dionysius of Halicarnassus and Diodorus Siculus, and the inscribed list of magistrates on the Capitol, known as the Capitoline Fasti, which is also Augustan in date. There are problems and discrepancies; some versions have additional years, and there are disputes over some names, and more particularly over the inclusion of the cognomina of early magistrates. (These are the names that are added after the nomen, often thought to indicate families within gentes, although some are nicknames or markers of a particular success, and they may have been added later.) That said, there is a striking degree of uniformity which suggests a single source, and we know that the names

General introduction

5

of consuls and magistrates, along with significant events, were inscribed in the pontifical tables, a record kept on an annual basis.4 This is not unimpeachable evidence. Livy himself claims that the records were mostly destroyed in the Gallic sack. There are uncertainties in the earlier parts of the lists, to be sure, and the list creates a picture of a rather smooth transition from kingship to Republic that may conceal a more troubled period. However, complete scepticism is, to my mind, unjustified. The sequence of magistrates is, at a broad level, reliable, both in indicating the existence of constitutional magistracies, and in demonstrating a pattern of office-holding which will become an important element in our understanding of the gens. Some scholars have called for a methodology to sift the good evidence from the bad.5 The extent to which this can be successful depends on the degree of control we have over the accuracy of the sources. It is not possible to prove that Livy was telling an accurate story, and even the recovery of all the fragmentary historians who preceded him would only help us piece together how the story was built up, and not whether it was true, at any philosophical or historiographical level. It would be much more useful if we knew what sources the earliest historians used (and it is important to note in this context that Cato the Elder, who wrote a history of Rome in the second century bc, makes reference to the pontifical tables). When we come to sources which are not telling a narrative, the problems grow. Lawyers’ definitions are not necessarily a guide to past practice so much as a way of tidying up present reality; antiquarian information may look archaic, but that is simply a modern perception. The absence of evidence does not provide an argument in the context of early Rome, since so much evidence is missing. The view taken here is that it was possible for information and knowledge about both events and structures to have been transmitted from early times to the time when history was being constructed, and in a number of different ways, not simply through the lists of magistrates. A whole range of media were available, from inscriptions and monuments to stories, oral tradition, family archives, and the simple continuance of some practices and features of archaic Rome down into the Republic and beyond. If one accepts the possibility that Romans could know about their past, as I do, then the questions become ones of reliability, and the general approach taken here is that whilst the ancient sources are capable of misunderstandings and 4 5

Frier (1979); more generally on the Fasti, see Oakley (1997) 21–109. There are several such statements in Raaflaub (1986b), but little agreement nonetheless.

6

The Roman Clan

mistakes, they are rarely setting out to mislead. Moreover, contradictions and confusions may be a more accurate reflection of a contradictory and confusing reality than a picture produced by a rational preference of one source over another. Given these deep-seated problems, it is tempting to look for assistance from other kinds of evidence. Archaeology is helpful for much Roman history, especially the development of the early city, but it is much less helpful in the context of a political argument. As we shall see, one major problem is the influence of preconceived notions of the gens on archaeological descriptions. It has been common to describe evidence of gentilicial activity in the archaeology of central Italy, but this is not the same as finding the gens on the ground. Comparative history and social anthropology have their part to play, but one has to start from some perception of what Rome was like in order to choose the comparandum, or apply the model. One can use the same sources to describe Rome in the fifth century bc as a society of feuding condottieri with local powerbases, or a society collectively developing complex and sophisticated, and dare one say rational, responses to the problems of maintaining a civic community. We can give as an example an event in fifth-century history, directly relevant to the history of the gens, and to which we will return at length later. Early in the fifth century, Rome was at war on several fronts, and under pressure. One of the best known patrician gentes, the Fabii, offered to undertake the war against Veii themselves, and they marched out as a kind of state-sanctioned private army and formed a garrison on the Cremera river. They were successful for some time, before being ambushed, and killed almost to a man. Did this event really happen? We do not know for sure; presumably the Fabii claimed it did, but the sources give quite different versions, from the numbers involved, to the reasons for the disastrous denouement, and careful source criticism can reveal putative reasons for all the variants. Part of the importance of the episode lies in what it might or might not tell us about early Rome. Was this how all Roman warfare was conducted at that time, or was it the last gasp of an antiquated tribal mode or do we have the embellishment of a local brawl? We do not know, and all three interpretations (and several others too) are, strictly speaking, possible. In order to resolve this one way or the other we have to refer to the nature of the Roman army, the reconstruction of which is, in the first place, controversial in the extreme, in the second place, often predicated on our response to the prior question about the Fabii, and in the third place actually capable of supporting any answer about the Fabii. One might argue that the Roman army had achieved a measure of uniformity which shows

General introduction

7

just how out of date the Fabii were, but it is not incontrovertible, nor is it impossible to believe that Romans sometimes conducted warfare for communal purposes, and sometimes for local purposes. This is not meant to be a criticism of the tremendous efforts of many scholars to establish more clearly the nature of Roman history and the historical record; nor is it a denial of the necessity for careful analysis of every source on every point. It simply reflects a conviction that Roman history is not a problem which can be cracked if only one applies the right method, nor is it a jigsaw which only admits of one solution. Much of the history of early Rome was approached through debate in antiquity, and we see the traces of later arguments. One may think, for instance, of the origins of the Republic itself. There were a number of stories which were told, and a number of different ways of thinking about why and how (and when) the Romans expelled their last king and established a Republic. Which version is ‘true’ is not the only, and not necessarily the most interesting, question.6 This does not mean that we cannot write early Roman history, or that we must disbelieve everything we read in the sources. It does however render it a peculiarly difficult task. The first part of this book, which takes a rather sceptical line towards many current reconstructions of the gens, indicates a number of pitfalls, and makes the second part, where I attempt my own account of what the gens was and how it operated, vulnerable to similar criticism. This is all the more the case because of my belief that an understanding of this social institution can be arrived at only through an effort to understand those institutions which made up the contemporary political and military structures, all of which are themselves the subject of much dispute. Others will judge if the picture which emerges reflects the sources we have, and is coherent in its own terms, but the book will have achieved one of my major aims if it provides a basis for further debate. 6

Wiseman (1998b).

part i

The evidence for the gens

The first part of this book focuses on the range of evidence which can be and has been brought to bear on the problem of the gens. We begin with the ancient evidence itself. What we know about the gens is actually confined to a few areas: inheritance, and various markers of identity such as burial grounds, legendary genealogies and religious rituals. One problem is that it is very difficult to find specific aspects of the gens which are not shared with many other social groups, so that what makes the gens different is elusive. The absence of a political dimension is also striking. The sources do not describe the gens as a political unit, though it is clear that members of gentes participate in political life in various ways; this will be the focus of the second part of this book. What is important here is to note that one source, Livy, indicates that an argument could be made that the gens was an institution which was possessed only by patricians. At the same time, our analysis of the evidence indicates the difficulty of making that argument with any cogent force, and indicates equally the existence of clear counterindications, including evidence for non-patrician gentes, and definitions which are at variance with a patrician monopoly. The evidence does not give a single, straightforward picture, and my argument will be that this reflects ancient realities. The second chapter considers how the concept of the gens has been treated by historians from the early modern period to the present day. It should be noted that many of the early figures are somewhat isolated. Renaissance thought, represented here by Carlo Sigonio, had already developed important arguments about the gens and the patriciate. These had their own context in contemporary debates about the definition of nobility. Although Sigonio was an enormously significant figure in the development of the discipline, his importance was rather overlooked in the nineteenth century, yet his approach was far more analytical than that of our next key thinker,

10

The evidence for the gens

the early eighteenth-century scholar Giambattista Vico. Vico’s approach is confused and inaccurate, and to a large extent fantastic; moreover, it had practically no influence whatsoever on the nineteenth-century German development of the discipline of ancient history. Vico became important once more through quite different channels, particularly in France where he was championed by Michelet, an important influence on Fustel de Coulanges, and in early twentieth-century Italy, where he was taken up by Croce. German scholarship developed in a different direction, but the conclusions of Niebuhr especially were identified as the factual basis from which scholars outside the discipline would work. The gens became a crucial part of the development of the discourse on kinship invented by Lewis Henry Morgan and Henry Sumner Maine amongst others. The conclusions of the former were taken up by Marx and Engels, and have been influential ever since, but our investigation shows how Morgan’s creative misrepresentation of the Roman gens led to the development of a model of gentilicial society that was in fact radically different from the evidence which the sources give us. This chapter concludes with a brief statement of the key modern theories, and an indication of their intellectual inheritance. Study of the Roman gens has not proceeded with the same degree of methodological sophistication as research into the Athenian genos. The genos has always been the obvious parallel institution for the gens, but the radical re-evaluation of the Greek evidence has not been systematically juxtaposed with the Roman evidence. The purpose of the third chapter is to do exactly that, and I have therefore developed in some detail the Greek parallels. Since these owe much, consciously and unconsciously, to the development of social anthropology subsequent to Morgan, this chapter also continues one aspect of the interaction between ancient evidence and modern interpretation. Whilst the gens and the genos appear far more dissimilar now than they used to, the kinds of interpretation which have been applied to the genos will be useful for our own final attempts to explain the gens. One area where modern theory and ancient evidence have been most closely connected with regard to the gens is archaeology. In the fourth chapter, I consider some of the most relevant archaeological evidence, and show the way that preconceptions drawn from the social anthropological models outlined in chapter 2 have directed archaeological identifications of material remains with an ancient social institution. This connection raises serious problems of interpretation, whilst remaining an extremely

The evidence for the gens

11

exciting and productive area. The problems are at their most acute when one extends the field of investigation into central Italy more generally, and Etruria in particular. These considerations bring us directly to the problem of the historical context of the alleged gentilicial structures in the sixth to fourth centuries bc, which is the subject of the second part of the book.

chap t e r 1

The ancient evidence

1. introd uction We have significant evidence for the nature and history of the gens but, as so often for social institutions which are obvious to those who live with them, but obscure to those who come after, our information is scattered in time and disparate in character. Much is missing that might have helped; the wreckage of Varro’s abundant work is particularly unfortunate for our understanding of early Rome. From what we know, Varro was both knowledgeable about the traditions of early Rome, and interested in many of the questions which we will have to address. However, there is no reason to suppose that even if all that the ancients wrote was restored to us, we would have an account that was accurate and cogent. The idea that the gens had a single form which even ideally might be recovered is one which this book explicitly rejects. As we shall see, the evidence suggests that the gens was not an institution which stood still, and that it was part of an argument, an argument which was conducted with some passion on a number of fronts over a long period of time. Our task must be to recover the lineaments of the argument, without the expectation that putting them together will reveal even a half-complete jigsaw of a single picture. If one considers the chronological spread of the references, it is immediately clear that there is a large gap between our first source, the Twelve Tables of the mid-fifth century bc, and the rest of our sources, most of which come from the historians and antiquarians of the late Republic and early Empire. We shall uncover differences in the ways in which writers use the word gens, which may imply that, at least for different writers, the word had different resonances. Gaius in the second century ad informs us that the law of the gens had fallen into desuetude and obscurity by his own time, but he preserves important earlier information. From the start, therefore, we must recognise that each reference must be taken on its own terms. Later sources use earlier sources themselves; we should say for instance that 12

The ancient evidence

13

Aulus Gellius has a second-century ad understanding of earlier material, not that Aulus Gellius tells us about the gens in the second century ad. We begin with the most significant structural facts about the gens, some of which will be discussed further later on. This will establish the basic nature of the group, and identify some of the questions we must address.1 2. et ymology, meaning and definition We may begin with the word gens itself. The word has distant roots connected with family and descent; K¨ubler gives various Sanskrit equivalences, which indicate that it is Indo-European in origin.2 It is cognate with the verb gigno and in Greek

 and the general sense is of a social group based on relationship through common descent. From it we have gentilicius, gentilis as adjective and noun, and gentilitas. This fundamental concept of relationship through descent allowed the group of related words to develop a wider reference. The easiest way to demonstrate this is by reference to the conspectus materiae at the beginning of the article in the Thesaurus linguae Latinae on gens, which commences with the sense of the group of those who draw their descent from a single princeps, and then develops the technical Roman sense of a conjunction of several families, and thereafter the wider sense of a nation or people. This wider meaning is found quite early; Ennius has magnae gentes opulentae in what appears to be a reference to the peoples of Latium, and the word may be found in this sense of other peoples in Accius, Naevius and Plautus, whereas the earliest literary references to the gens in the sense of a familial group are found in Cicero, and the antiquity of the concept is based on the appearance of the gentiles as an inheritance group in the Twelve Tables.3 Eventually, by translation of a Hebrew term, the word develops connotations of opposition to true religion, and is equated with paganism, which brings us to its reincarnation as our familiar word ‘gentiles’.4 All the related words follow the same trajectory, referring both to the narrow Roman group, and to the concept of a wide nationality, often not Roman. 1

2 3 4

I omit here discussion of the relationship between gentes and tribus, e.g. the Claudii and tribus Claudia. No ancient source fully explains the fact that some tribes have the names of patrician clans, and therefore this will be discussed later; see below, chapter 7.2. I also omit here discussion of one of the most famous actions by a gens, the military action conducted by the Fabii in the fifth century bc; this will also be considered in full at a later stage (chapter 9.2). K¨ubler (1910). Ennius Ann. 141 Skutsch; Acc. trag. 523 Dardaniis gentibus; Naev. com. 109 qui apud gentes solus praestat; Plaut. Rud. 1 qui gentis omnis mariaque et terras mouet. See Ernout and Meillet (1959) 250–3 for a clear account.

14

The evidence for the gens

We should certainly note that there is a clear parallel to be drawn between the search for familial ancestry and the process of identifying the founders of nations, or searching for the origines gentium.5 It is from this sense of a ‘nation’ that Roman writers developed the concept of the ius gentium, a law of nations, and later scholars from Grotius onwards developed this into theories of natural justice. This meaning is a long way from, and not directly connected with, gens in the sense of a ‘clan’, but the use of the same word as an early social institution at Rome, and in a phrase loaded with meaning for early moderns, was productive, as we shall see, of some interesting and perhaps deliberate confusions. The early gens does not easily escape the shadows of its later definition. It is worth briefly noting that, as the lexicographers show, gens is not alone in referring in the narrow sense to a group with common descent. TLL is most explicit here, referring to familia, stirps and domus as parts of the whole referred to by gens; and there is also the clearly related word genus in its sense of a birth group. This lexical entry is derived from K¨ubler’s account, and may make too tidy the relationship between these different groups, while some writers use them interchangeably.6 There may be distinctions to be drawn between these words, but as the lexica show, these distinctions are not clear and are often ignored by the time we have substantial amounts of literature to deal with. The word gens and its cognates had already had centuries of use and development by the time Latin literature begins, and that in itself is an important fact. We shall see below a number of definitions of the gens, but at this point we may note a late one, that of Isidore (Orig. 9.2.1), who gives no indication whatsoever that the word has any meaning more narrow than that of a nation or people. When he writes that gens est multitudo ab uno principio orta and gens . . . appellata propter generationes familiarum, id est a gignendo, sicut natio a nascendo,7 he is thinking of the nations of the world in a sense which modern theorists of natural law would have understood more readily than those who drafted the law of the Twelve Tables, which is where the gens in the sense in which this book is interested

5

6 7

Bickerman (1952); see now Fromentin and Gotteland (2001). Rodriguez (1996) largely discusses the use of the word gens for a nation, people or state, and suggests that the word implies an elemental or primary collection of individuals, based around sometimes fictitious concepts of kinship, language, customs and historical past rather than political identity or territory. K¨ubler (1910). ‘The gens is a multitude sprung from the same beginning’; ‘the gens is named because of the generations of families, that is from gignendo, coming into being, just as the nation is from nascendo, being born’. See Rodriguez (1996) 44–5, acknowledging here the importance of the fictive kinship of the Roman gens in the creation of legendary ancestries for the nations.

The ancient evidence

15

first makes its appearance. To indicate what the gens in the sense of a ‘clan’ looked like, a preliminary definition will be given to clarify the argument. 3. the g e n s : a d ef inition The clearest and most comprehensive definition comes from Q. Mucius Scaevola, a notable and highly regarded individual who held the consulship in 95 bc, was a distinguished governor of Asia, and subsequently pontifex maximus, and who wrote an eighteen-book codification of the ius ciuile. The definition, which is quoted by Cicero as a good example of how to make a definition, runs as follows: itemque [ut illud – sc. hereditas]: gentiles sunt inter se qui eodem nomine sunt. non est satis. qui ab ingenuis oriundi sunt. ne id quidem satis est. quorum maiorum nemo seruitutem seruiuit. abest etiam nunc. qui capite non sunt diminuti. hoc fortasse satis est. nihil enim uideo Scaeuolam pontificem ad hanc definitionem addidisse.8

The nomen was inherited through the male line, and so the gens can be seen as a group comprising all agnates, and so, potentially at any rate, a collection of families. Agnates are defined as legitimate descendants of a common male ancestor, through the male line only, but for some purposes at least this stopped at what was known as the sixth grade (i.e. second cousin); there was no such limit on the gens. In the gens of the Fabii then, every free member born into the group took the nomen Fabius and was a gentilis. The basic situation may be illustrated figuratively (Figure 1). As we shall see, the sources refer to collectives such as the Fabii, Claudii, Cornelii and so forth as gentes, although they also describe them in other ways. From the definitions we have seen in the preceding section, we can see that there will have been a tendency to try to identify the first holder of a given nomen, the ancestor of the gens, and this is the subject of a later section, but, to anticipate that discussion, it will be evident that these figures are mythical, and that the relationship to a single ancestor is fictitious, and this raises questions over how genuine was the kinship between members of a gens. Many complications follow of course, and Scaevola’s definition indicates some of them; not everyone with the nomen Fabius is on an equal footing with other Fabii. Freed slaves took the nomen of their master, but 8

Cic. Top. 29: ‘Again: “gentiles in relation to each other are those who share the same nomen.” That is not enough. “Those who are born from freeborn citizens.” That too is not enough. “Of whose ancestors no-one has served in slavery.” There is still something missing. “Who have not suffered capitis diminutio.” Perhaps that is enough. For I see that Scaevola the pontifex has added nothing to this definition.’ On this work, and on Scaevola, see Reinhardt (2003).

The evidence for the gens

16

Figure 1. Schematic representation of relationship between gens and agnatic descent group; after Linke (1995) 25.

were not thereby part of the gens. Another issue is whether every nomen reveals a gens, and hence whether every freeborn Roman citizen who fitted the Scaevolan definition belonged to a gens or not. Finally, some gentes were indisputably patrician. We find the phrase patricia gens in Livy, and we know of a division of patrician gentes into gentes maiores and minores, though this is mysterious.9 A freeborn individual, descended from others who were freeborn and where no instance of capitis diminutio had taken place, would inherit patrician status and share that status with all other members of the gens. It is tempting to see the gens, especially the patrician gens, as a collective body, and to assign it functions, identity and agency, and many modern accounts, as we shall see in the following chapter, have done precisely that. It is important therefore to 9

Below, chapter 8.2.

The ancient evidence

17

state at the outset that this has more to do with modern interpretation than ancient evidence, and we should beware of confusing the action of all or some members of a gens with the action of the gens as a collective group, with or without a leader. Scaevola is, as De Sanctis pointed out, defining gentiles inter se.10 In other words, these are characteristics which relate to whether someone did or did not belong to the same gens as someone else. The definition excludes freedmen, amongst other categories, and precedes the massive influx of citizens after the Social War. It would have been very interesting to know the context in which Scaevola gave his definition. It is possible that he was trying to elucidate previous law – the word gentiles appears in the Twelve Tables, as we shall see shortly. Alternatively, Scaevola may have been defining the situation in his own day. Whether anything had changed between the fifth and the first centuries bc is part of the concern of this book. In the following sections, we shall consider the evidence in more detail, and we shall return to Scaevola’s definition, but nothing that follows seriously affects the validity of the positive statements made in this section. We may move forward then from this basis to consider the rest of the evidence for the nature and function of the gentes, and the activities of their individual members. We begin with the nomen itself. 4 . the n o m e n Scaevola’s definition of a gens begins with the shared name. The nomen is the second name, after the praenomen and before the cognomen (which not all Romans used).11 We appear to have a technical term for the nomen, nomen 10 11

De Sanctis (1960) 226–7. For an admirably clear discussion of Roman practice generally, see Salway (1994). The issue is enormously complex, and the bibliography is considerable, especially since we have to understand the system as a whole to understand the nomen. Many aspects of the problem were first set out by Mommsen (1864–79) I.1–284 (two essays originally published in 1860 and 1861). There is an early and important contribution by Chase (1897), and a major study followed shortly afterwards in Schulze (1904). Two studies by different scholars published in the same year came to completely different conclusions about where the system originated; Bonfante (1948) suggested a Sabine origin, whilst Pulgram (1948) suggested that the nomen began as an Etruscan practice, but was taken on by the Italic peoples. A major overview was provided by Rix (1972), whose preference was for a Faliscan origin. When such varied answers are offered to a question, it may well be suggested that the question is the wrong one, or the evidence insufficient to answer it, or both. Several contributions to Duval (1977), an important collection of papers, are relevant; see in particular Heurgon (1977); Nicolet (1977); cf. M´enager (1980). There is an important and current Italian debate; see Colonna (1977); Marchesini (1997) 154–9; and below, chapter 4.8. Finally, we should note the position of Franciosi, who denies that the nomen was a patronymic and associates the Roman onomastic system with a sort of totemism; Franciosi (1999) 223–60; Franciosi (1984b).

18

The evidence for the gens

gentilicium,12 in a work on praenomina, found in Julius Paris’ epitome of Valerius Maximus. The unknown author quotes Varro, and the whole work is therefore thought to be an essentially Varronian account. This account appears to underlie all later grammatical accounts, which do not differ substantially.13 The author notes that the earliest figures in Roman history, Faustulus, Romulus and Remus, had only one name. With the admixture of Alban and Sabine families, the necessity to distinguish between individuals so that their gens might be known led to the introduction of the second name, the gentilicium. Varro estimated that there were 1,000 gentilicia, and about 30 praenomina. It is hard to see 1,000 gentes in early Rome, so at best Varro must be making an observation about his own time. It is also not clear that there has to be a gens for every nomen gentilicium. The number is a suspiciously round figure, and when combined with the number thirty for praenomina (admittedly transmitted as approximate) raises suspicions that Varro is playing a numbers game.14 Certainly, this is extremely problematic evidence, under any reconstruction, for early Roman society, or for the gens. Modern scholarship has shown that the nomen was in essence an adjectival patronymic.15 The date of the introduction of the two-name system is unclear, as with the three-name system with the addition of the cognomen which distinguished between individual branches of a family, and also between individuals. The debate hinges partly on epigraphic evidence, which, scanty as it is, would appear to allow for the binominal system by the end of the sixth century, at least among the class of people who used writing, but the trinominal system not before the third century. The versions of the list of chief magistrates provide additional information. There are two rather different types of evidence for this list. One comes from historians such as Livy, Dionysius of Halicarnassus and Diodorus Siculus. The other is the inscribed list of magistrates on the Capitol which dates from the Augustan period. The historians and the Fasti Capitolini are in broad agreement, but the latter gives far more cognomina and sometimes multiple cognomina for an individual, for the early period. Mommsen doubted whether this could be genuine, and suspected later 12 13 14

15

Cf. also Varro. Ling. 9.60; Suet. Claud. 25.3; for nomen gentile, see Suet. Nero 41.1.7, Porphy. Ep. 1.13.8, Sil. Pun. 5.646, Stat. Theb. 6.342. See Prisc. in GL II.57.12–17; Diom. in GL I.321.3–11, quoted in Salway (1994) 125 n. 5, 128 n. 29. Salway (1994) 125 n. 8 counts fifty-six praenomina. It is also notable that according to D. Hal. 1.85, Varro claimed there were fifty Troianae gentes, though we can count only fourteen; see Nicolet (1977) 51–2. Rix (1972).

The ancient evidence

19

interference, since we do not see the cognomen epigraphically before the third century.16 The later insertion of the cognomina has been disputed, and the current view appears to be that the evidence does not permit us to deny the use of the tria nomina from an early period.17 This may be optimistic. It is extremely significant that a relatively recently discovered inscription, dated to the late 270s and found at Caere, gives us the name C. Genucio(s) Clousino(s), which must be C. Genucius Clepsina, consul in 276 and 270 bc.18 The cognomen is Etruscan, and the context may explain its appearance. Slightly earlier are the well-known inscriptions with the names of Cornelius Scipio and Cornelius Scapula.19 There is still a long gap from the third century back into the fifth, and a further difficulty which is the date of the transformation of the cognomen from a nickname into an inherited marker of a family, especially a family within a gens. Since most of our argument has to do with the nomen, the cognomen is of lesser significance, but scepticism has important consequences; first it becomes difficult to make arguments from the transmitted cognomina about the number of branches within a gens, and second we cannot be absolutely certain that the trinominal system was used extensively from an early period, and this means that we cannot with complete confidence use onomastic evidence to support a picture of 16

17

18 19

For the binominal system, a date around 650 bc is common; see Heurgon (1977); for the tria nomina, Bonfante (1948), and Kajanto (1977) rely on the epigraphic evidence, first used in this way by Mommsen (1864–79) 42–68. The sceptical approach was most firmly pressed by Cichorius (1886), and see the balanced and far-reaching account of Ridley (1980). See also Alf¨oldi (1966). The counter-argument to Mommsen on the authenticity of the cognomina in the early consular Fasti was put by Beloch (1926) 46–52, accepted by Cornell (1995) 440 n. 14, though in fact Beloch’s arguments are not particularly strong. Mommsen argued that some of the cognomina were patently retrojected (Augurinus for the Genucii for instance); Beloch responded that this simply meant that some, especially plebeian, cognomina were interpolated into a list which had genuine cognomina. Beloch also argued however that no-one could have engaged in the full-scale invention that Mommsen envisaged, though he has to accept that some invention has taken place; and that it did not follow from the absence of cognomina in other contexts that they were also absent from the lists of magistrates, where the cognomen was of greater significance. Beloch’s strongest argument, as he saw it, appears to be that cognomina appear in Licinius Macer’s account of the Libri Lintei. This argument is only strong if one believes that the Linen Books were a genuine source, and not one displaying precisely the sort of invention which Mommsen was suggesting. Both Mommsen and Beloch doubt that Greek influence was sufficient to permit the names Q. Publilius Philo (cos. 339) and P. Sempronius Sophus (cos. 304), but that might underestimate plebeian contacts with the Greek world. Badian (1988) describes the cognomen as being in a ‘no-man’s land between name and description’; see now Bruun (2000) 49–56. Torelli (2000), with earlier bibliography. CIL i2 6 = vi 1285 = ILS 1 = ILLRP 309; ILLRP 1274a. On the problems over dating the first inscription, see LTUR IV.281–5, s.v. Sepulcrum (Corneliorum) Scipionum; there is a longer inscription, which may date to around 200 bc, but there is also a painted name, which could still be early third century.

20

The evidence for the gens

gentes comprising large numbers of families from an early period.20 One other oddity of our evidence is that the tribunes, whose names survive only in the historians who tend to give the tria nomina less conscientiously than the epigraphic Fasti, usually appear with only two names. Does this mean, as some have thought, that the plebeians began to use cognomina later than the patricians? A very few early tribunes do have three names, and although Dionysius of Halicarnassus, who gives consular names in full more often than Livy, not infrequently uses only two for tribunes, there is enough doubt for us to suspect that the dividing line here was not between patricians and plebeians, or was not used rigorously.21 The binominal system arose at much the same time across central Italy. It has been argued that the name indicated membership of a group wider than the nuclear family, i.e. the gens or a similar group, and that this was not confined to any one social group, but was a universal system.22 This has profound implications, and has been particularly useful in the analysis of Etruscan society where the epigraphic record takes on an additional significance in the absence of any substantial historical accounts. It is worth pausing to note a possible objection. The connection between the nomen and the gens might be derived from observation of common behaviour, rather than an original feature of either phenomenon. If the nomen is indeed an adjectival patronymic, one cannot simply derive from it the notion of social institutions larger than the agnatic family, and of course these groups may have varied enormously in terms of size and complexity. The use of the cognomen might have emerged precisely in those groups which had indeed developed this greater level of complexity, and needed additional markers. For now, we must note that, whatever the nature of the relationship, those who had the same nomen were likely to have belonged to the same gens, and were gentiles. We turn now to the earliest evidence for what the gentiles did, Rome’s first lawcode, the Twelve Tables.

20

21

22

It is possible that after the restoration of the consulship in 367, or after the requirement for a plebeian as well as a patrician consul in 341, the lists of magistrates may have been kept more accurately, but that is only a speculation. For early tribunes with tria nomina, see C. Terentilius Harsa (462), and M. Volscius Fictor, whose dates are disputed; it is interesting that both names indicate non-Roman descent. ‘Harsa’, unattested elsewhere, is perhaps Etruscan (Schulze (1904) 357); ‘Volscius’ will have been connected with the Volsci. The distinction is admitted by Beloch (see above), and more generally; see Kajanto (1977). Salway (1994) 127 leaves the matter somewhat open in his formulation ‘the evidence suggests that they were pioneered by the e´lite, perhaps keen to differentiate a noble family ancestry’; there are strong arguments for seeing a plebeian nobility imitating the patricians, or sharing patrician habits, from an early stage. Conveniently and elegantly, Momigliano (1989) 98–9; see also below, chapter 4.8.

The ancient evidence

21

5. the t welve tables The members of a gens, or gentiles, are first referred to in the Twelve Tables, Rome’s earliest surviving lawcode, created in the middle of the fifth century bc. The story of the promulgation of the code in 451 and 450 bc is famous; a Decemvirate was established to set out a new basis for Roman society as a result of the pressure from the plebeians against the dominance of the patricians.23 Ten tabulae were set up for discussion, and, after amendment, were adopted by the comitia centuriata, and as Livy says of these ten laws (3.34.6), qui nunc quoque, in hoc immenso aliarum super alias aceruatarum legum cumulo, fons omnis publici priuatique est iuris.24 The remaining two laws were more controversial, and we need not consider them at present; and the ultimate disgrace of the Decemvirate did not affect the success of their code. Much of the narrative story in which the development of the code is embedded is highly problematic, including the suggestion that the Romans drew inspiration from Athens,25 but the most recent edition of the laws accepts the general reliability of the laws as transmitted to us.26 Arguments against a fifth-century code27 are different from arguments that the code we have has changed since the fifth century, but any reconstruction would then be purely speculative. The creation of such a code was within the capacity of the Romans of the fifth century,28 and the provisions we have were appropriate for the period.29 There are specific historical problems associated with the Decemvirate, but this has little effect on the laws themselves. The evidence we have can still relate to a fifth-century lawcode compiled by a body similar to the Decemvirate.30 23

24

25

26 27 28 29

30

Livy 3.33–55 and D. Hal. 10.56–11.44 give the major narrative accounts; see Cornell (1995) 272–92 for a balanced discussion of the issues surrounding the Decemvirate and its lawcode. A number of articles in Raaflaub (1986b) also address the Decemvirate from a variety of standpoints. ‘[Laws] which even now, with the immense mass of legislation heaped up since, are still the fount of all public and private law.’ This is an exaggeration, but reflects the reverence with which the laws were regarded; cf. Cic. de or. 1.195; Leg. 2.59. See also Eder (1986) 272–3. An account gathering all the sources can be found at Ogilvie (1965) 449–50, concluding ‘the whole episode is a fiction of the early first century bc’. Nevertheless there are some interesting similarities with Greek lawcodes; see Cic. Leg. 2.59–64; Roman Statutes II.560–1; Martini (1999). Wieacker (1971) suggests that the idea of codification may itself have had Greek origins. Roman Statutes II.555–721 gives introduction, reconstruction, commentary and bibliography; see also Wieacker (1988) 287–309. For instance, Nap (1925) 407–29 argued implausibly that the Twelve Tables were invented in 225 bc. See Cornell (1991) for a clear positive statement that written material could have existed in, and survived from, the archaic period. See for instance Ogilvie (1965) 452: ‘the parochial character of some of the provisions of the Tables . . . is only compatible with a fifth century date’. See also Cornell (1995) 273 for further discussion of the Decemvirate. Ungern-Sternberg (1986) accepts the existence of the Decemvirate and the laws, but little else.

22

The evidence for the gens

A different kind of problem relates to the transmission of the laws. It is complex; there is the issue of the transmission from the original laws to our sources (and one might add that the sources are disparate, and whether all of them had exactly the same view of the content of the laws is an insoluble issue), and then from the sources to the surviving manuscripts, and then again to modern editions and their distribution of the fragments to specific Tables.31 Cicero attests that some form of the Twelve Tables existed in the late Republic;32 the provisions remained of interest to lawyers for obvious reasons, and to antiquarians and grammarians because of oddities in language. The careful analyses in Roman Statutes demonstrate that the provisions of the laws were known, if not always clearly understood because of the language and the passage of time, and we can therefore hope to get close to the contents of the code, albeit in a fragmentary and distorted fashion. Our main concern is with testamentary law, and we can tell from Cicero that this was still a controversial area.33 To anticipate though, it is extraordinarily difficult to see why the provisions concerning the gentiles should have been inserted at a date later than the middle of the fifth century. The references to the Twelve Tables are the best and indeed almost the only evidence we have for the operation of the gens at such an early date, and in the field of inheritance; the evidence is therefore crucial. Our knowledge of the Twelve Tables depends on later citations, and the reconstruction of the order of the fragments and their assignment to specific tables is a difficult task, based as it is on the most recalcitrant of information. It is a reasonable inference from evidence in Cicero that the fifth Table dealt in detail with issues of succession, and this is where our reference to the gens lies.34 It is impossible to say with certainty what a complete Table looked like, but what we have of the fifth Table suggests that an attempt was made to set out a coherent position on inheritance as a whole, and for that reason, it would be a mistake to treat V.5 and V.7 in

31 32 33 34

Roman Statutes II.556; compare for instance the scepticism of Levi (1995a). de or. 1.195; Leg. 2.9. de or. 1.176 on the case between the Claudii and Marcelli; see below, 1.9. Two important texts, Cic. Top. 26–7 and de or. 1.173, give what appear to be lists of contents from a narrow band in Tables V, VI and VII (Roman Statutes, II. 567); we also know from Dig. 38.6.1. pr. that the Twelve Tables dealt with testamentary succession and then intestate succession. It is fair to say that given the rarity with which reference is made in the sources to a specific Table, in addition to the difficulty of reconstructing clauses, any attribution of a clause to a specific Table can be contested.

The ancient evidence

23

isolation. I begin by presenting the whole of the Table as reconstructed in Roman Statutes first, and then consider each provision in more detail.35 V.1 uirgo Vestalis V.2 V.3 uti legassit super familia ?pecuniaue? tutelaue sua, ita ius esto. V.4 si intestato moritur, cui suus heres nec essit, agnatus proximus familiam ?pecuniamque? habeto. V.5 si agnatus nec essit, gentiles familiam ?pecuniamque? h[abento]. V.6 V.7 si furiosus ?prodigusue? esst, agnatum gentiliumque in eo ?pecuniaque? eius potestas esto. V.8 si libertus . . . ex ea familia . . . in eam familiam . . . V.9–10 V.1 A Vestal Virgin V.2 V.3 As he has disposed by will concerning his familia ?or goods?, or guardianship, so is there to be source of rights. V.4 If he dies intestate, to whom there be no suus heres, the nearest agnate is to have the familia ?and goods?. V.5 If there be no agnate, the gentiles are to have the familia ?and goods?. V.6 V.7 If there be a madman ?or spendthrift?, power in respect of him ?and goods? is to belong to his agnates and gentiles. V.8 If a freedman . . . from that familia . . . to that familia . . . V.9–10 The first two provisions relate to the guardianship of women. The Vestal Virgins are exceptional in their legal status, as a number of sources indicate.36 The more normal situation is found in V.2, which refers to the guardianship of women and their rights to dispose of and obtain property. The formal position is that a woman of any age is required to have a tutor or 35

36

See Roman Statutes II.581 for the basic text and translation; 634–51 for details; and discussions in Voci (1960) 1–83 and Kaser (1955) 81–98. [. . .] and are used to indicate where the word, not simply the form, is restored or results from correction; reflects a belief that a text can be reconstructed with reasonable plausibility, but no source claims to report it. Gell. NA 1.12, including citation of Antistius Labeo’s commentary on the Twelve Tables; Gai. Inst. 1.144–5; Plut. Num. 10.5 attributes this rule to Numa, so it may be one of the so-called leges regiae, on which see Roman Statutes II.561–3.

24

The evidence for the gens

guardian. A woman who is in the tutela of her agnati requires their authorisation for the usucapion (acquisition) or alienation of res mancipi. The intention appears to be to protect the interests of the prospective intestate heirs by disallowing any action that might materially affect the woman’s wealth. It is important to note that although it is more common to see a single tutor operating, and that is envisaged in the discussion of V.6, here the woman is in the tutela of her agnati, which may imply some element of co-operative action.37 The intention of these provisions however is not simply a reflection of Roman male mistrust of the ability of women to make intelligent decisions, but relates to marriage. In Cicero’s pro Flacco, the defendant Flaccus is accused of having deprived Sextilius Andro, whose wife (Flaccus’ relative) had died, of his inheritance.38 The defence is that the tutores had given their approval neither to the transfer of property consequent on a marriage by usus nor to a marriage by coemptio.39 The tutores, or one of them at any rate, thus prevented the transfer of property to the new husband. What Flaccus did was to ensure that the property went instead to a young relative of his, L. Flaccus. This is a later case of course, but there is no reason to suppose that the Twelve Tables, in discussing arrangements for succession, did not allow for the wider group, either collectively or through a single representative, to regulate what was potentially a loss of property through marriage.40 We can see therefore that this provision allows for the preservation of property within the agnatic family, unless the guardians permit it to be transferred. The main problems over V.3 relate to the reconstruction of the text of the Twelve Tables, and whether the original wording included pecunia, as argued in Roman Statutes, the meaning of legare, which simply must 37

38 39 40

The reconstruction mentions a single tutor, but Gai. Inst. 2.47 allows for multiple tutores. There were two kinds of tutores: those appointed by will, and those to whom the decision fell in cases of intestacy, or as here, the automatic tutores of one not sui iuris, a group later defined by the term tutor legitimus. On tutela see Fayer (1994) 379–611; Kaser (1955) I.76–8; briefly, Crook (1986) 62–3. The subject needs a new treatment which takes account of the social context of tutela, as has been provided for tutela over children by Saller (1994) 181–203. Cic. Flac. 84–9. Cic. Flac. 84: nihil enim potest de tutela legitima nisi omnium tutorum auctoritate deminui (‘for nothing can be taken from an estate in lawful guardianship without the permission of all the guardians’). It should be said that this issue has been the subject of considerable controversy. Watson (1967) 21–3 argues that the original clause expressed ‘the restriction on usucaption with reference to res mancipi only’, and goes on to argue that the extension of this provision to the acquisition of manus through usus must have been a late development, intended to enable most marriages to be sine manu. Drummond (1989a) 151 writes that the agnati ‘had to authorize any disposal of [a woman’s] property, including probably her conclusion of an in manum marriage’ (a better formulation might be disposal of any res mancipi she owned). Whether this extended to testamentary tutores is quite unclear. The nature of the dispute in the pro Flacco is also controversial; for the suggestion that Flaccus claimed the inheritance through his gens, see Watson (1971) 181; Kaser (1955) 74–81.

The ancient evidence

25

be ‘dispose by way of testament’ here, and the kind of will envisaged, on which Roman Statutes states that ‘the bibliography is enormous, and there is no evidence’ (II.639).41 Although the arguments are difficult, one may conclude that the outcome was that the testator had freedom to dispose of (at least some of ) his property and goods42 and to assign tutela.43 The provisions we have considered so far indicate that the fifth Table was concerned with the control of property. V.4–5 refer specifically to the case of intestacy, and those who inherit by default.44 In the case of intestacy and the absence of a suus heres, the estate devolves first to the agnatus proximus and then to the gentiles. Some definitions are required. A suus heres is one who is in the patria potestas or manus of an individual, and who becomes sui iuris, independent, after that person’s death.45 Put another way, sui heredes are all those for whom the deceased was the sole surviving male ascendant. The succession of the sui heredes was automatic in cases of intestacy, but it was entirely possible that there were none (this would always be the case for a woman). The agnati encompass those who, outside the sui heredes, are related to an individual through the male line. Thus all brothers, and their descendants, are agnatically related to our hypothetical individual; as indeed are his sisters, because all of them descend from the same man; but sons of a sister are not encompassed, nor are any relations whatsoever on the wife’s side 41

42

43

44

45

See Watson (1975) 52–60 for a more substantial treatment, arriving at somewhat different conclusions, followed by Drummond (1989a) 148–9. Watson argues from the absence of the word familia in the clause that it ‘was not envisaging the appointment of an heir. The testamentary provisions involved relate only to the making of individual bequests and the appointment of tutors’ (59). Watson (1975) 59 argues that pecunia appears alone here as meaning ‘property’ without the word familia which means ‘a man’s property considered as a unit’ and relates to the assets of the paterfamilias, including people; whereas in the following two clauses, familia appears without pecunia. Roman Statutes prints both words in all three clauses. For tutela in this clause, see Dig. 50.16.120; Tit. Ulp. 11.3.14; Dig. 50.16.53 pr. (Paul). Gai. Inst. 2.224 comments that this freedom could so diminish the value of any inheritance, by spreading legacies too widely outside the family, that heirs refused to take up inheritance, and many died intestate (quae qui scripti heredes erant, ab hereditate se abstinebant et idcirco plerique intestati moriebantur). This led to a rash of laws, beginning probably in the second century bc, which tried to limit the testator’s capacity to give away more than a certain amount of his estate to people other than his family; see Watson (1971) 163–74. This leads us to conclude first that there was substantial disintegration of family property and second that there were efforts to restrict this. It is interesting that laws had to be passed later to protect the continuity of the family (and its sacra, Cic. Leg. 2.21.52, the concern of two pontifices maximi, and see below, 1.8), an intention which the aspects of the Twelve Tables we discuss here also seems to enshrine. Doubts about the text centre on whether the word pecunia appears here as suggested by Ciceronian evidence, as reported in Cic. Inu. 2.148 (and identically in Rhet. Her. 1.13.23) or not, as in the legal texts, Tit. Ulp. 26.1–1a; Coll. 16.4.1–2; Dig. 28.2.9.2 (Paul); Dig. 50.16.195.1 (Ulpian). The whole issue is dealt with in Gai. Inst. 3.1–17. Gai. Inst. 3.1–9; Watson (1971) 68; Johnston (1999) 51.

26

The evidence for the gens

(brother-in-law, etc.). Two brothers born to the same father but different mothers are still related agnatically. Agnati do have a hierarchy, and hence the term agnatus proximus; so a brother is closer than a nephew; again Gaius Inst. 3.10–16 is helpful. It has been argued that the law may have meant that if the agnatus proximus refused the inheritance or had died, the whole estate reverted to the gentiles immediately. This might suggest that the agnati were inserted into the succession at a relatively late stage.46 This fits an evolutionary model where the primitive clan is the earliest form of social organisation, and the family emerges from within it; we shall see in chapter 2 ways in which this argument has been developed. There is however no evidence for this; the nearest agnate may have become agnatus proximus until there were no agnates left, or it may have been too difficult to decide priority between members of a more extended group of relatives. In the working definition at the beginning of this chapter we saw that the agnati were probably defined up to the sixth grade (i.e. second cousin). Scaevola’s definition indicates that at least towards the end of the second century bc, those who shared a nomen and fulfilled certain other requirements formed the gens. It is extremely unfortunate that, as we shall see below, Gaius, whose clear definitions for other groups are helpful, deserts us here, but it is possible to say that the gens encompasses relations in the male line beyond the agnati. This follows from Scaevola’s definition coupled with the fact that one inherited one’s nomen through the male line, and it makes sense of the provision for the gentiles or members of the gens to inherit after the agnatus proximus. What does not follow from the provision in the Twelve Tables is the conclusion that the gens as a corporate body inherited collectively.47 The provision specifies the gentiles, and the natural way to read this is that each member of the gens inherits an equal share. There is no mention in any source of a gentilis proximus, and it follows that the gentiles were not organised hierarchically like the agnati. It has sometimes been claimed that references to a princeps gentis suggest otherwise, but I shall demonstrate below that this always refers to a mythical founder of the gens, and there is absolutely no evidence for a formal ranking system within the gens. This is one area where preconceptions of the gens are extremely important. For those who stress the collective importance and nature of the gens, collective 46 47

Watson (1975) 68–9; (1971) 176–8; Drummond (1989a) 150–1; Guarino (2001) 443–4. The gens may nonetheless be seen in anthropological terms as a corporate group, in something of the way that Maine described it (below, 2.9), defined as ‘a set of individuals who have socially recognised claims – rights – to consume or use a specific resource or set of resources’; Bell (1998) 188.

The ancient evidence

27

inheritance is a crucial structural fact, but it should be stressed that it rests almost entirely on this passage. For those who seek to diminish the role of the gens as a collective, it is important to stress that the gentiles inherited individually. In practice, this system of partible inheritance may have reduced the distinction drawn here; an estate, divided equally into many parts, may only have functioned economically if farmed as a whole, but that aside, we must remember that it is unnecessary to see the gens as a primitive communitarian organisation to acknowledge that at the point that property devolves to the gentiles, it devolves to them all equally. This means that, in the case where the gentiles inherit, property is not concentrated in one person’s hands, but each gentilis has legal ownership of his own share. Thus the gens accommodates but limits private ownership. To return to the Twelve Tables, it will immediately be seen that there is a problem regarding the position of women, because a woman can have no suus heres, since she cannot have the power of patria potestas.48 Therefore in the event that a woman with property and sui iuris dies intestate, her agnatus proximus and then gentiles must inherit.49 Moreover, there were significant obstacles for a woman who wished to make a will, and if the situation is confused for the early Empire it is impenetrable for the preceding periods.50 It is enough to say here that the rules surrounding inheritance, by focusing on the two limiting conditions of patria potestas and agnatic descent, make it by default hard for women to transfer property outwith the agnatic group except through marriage, which was itself subject to the approval of a woman’s tutores, who in first instance are her agnati, as is shown by V.6. This is important because, as Drummond demonstrated, if a woman with property of her own married cum manu, and the husband died, her agnati had lost control of her property (and whatever she may have inherited).51 V.6 refers to the establishment of tutela and is thereby related to V.2, and V.3.52 As discussed by Gaius,53 the situation may have arisen in wills which 48 49 50 51 52

53

Dig. 50.16.195 (Ulpian): mulier autem familiae suae et caput et finis est (‘a woman is both the beginning and end of her family’). Saller (1994) 166. See Volterra (1991) and Fayer (1994) 544–9 for discussion. There is no evidence that a woman could make a will in the period of the Twelve Tables, though the Vestal Virgins may have been exceptional. Drummond (1989a) 151. The difference between the provisions of V.3 and V.6 is neatly demonstrated at Roman Statues II.643: ‘what is willed under Tabula V, 3 is tutela over potential sui heredes (and others); what is established [in Tabula V, 6] is tutela over instituted heirs (and others) in the case of a will which does not assign a tutor and over sui heredes in the case of intestacy’. Gai. Inst. 1.155: quibus testamento quidem tutor datus non sit, iis ex lege xii agnati sunt tutores, qui uocantur legitimi (‘for those to whom no guardian has been given by a will, by a law of the Twelve , the agnates are tutors, and are called statutory’).

28

The evidence for the gens

did not stipulate the assignment of a tutor as well as in the absence of a will; in the Digest quoting Paul, the reference is to the second situation.54 It is further clear from the Digest, citing Gaius again, that there would be a single tutor if a single agnatus proximus survived, but that tutela would be shared between agnati in the same degree.55 We learn from a dispute related in the Laudatio Turiae i.13–26 that ‘Turia’ argued against her reversion to the tutela of her gens after a dispute over the will of her murdered father; she claimed that no familia or gens could be proved to exist.56 This passage implies that tutela, like property, reverted to the gens if there were no agnati, and Roman Statutes (II. 643) states that it is ‘likely that there was another clause relating to gentiles, as with Tabula V, 4–5.’ Given the care shown over preserving property, it is unsurprising that in V.7 we find a clause which appears to allow the agnati and gentiles to act against someone who was either mad or spendthrift. The issue was submitted to a magistrate. The clause seems to have been absolute, but the Tituli Ulpiani (12.1–3) suggest that reference was only to prodigi who inherited through intestacy, to be consistent with the absolute freedom of a testator in V.3, but this is probably a later development.57 V.8 extends the issues of succession to freedmen. If a freedman died intestate without a suus heres, his property reverted to his patron; if his patron was dead, then to the patron’s suus heres and thereafter to the agnati and gentiles; this must be at the heart of the argument related in Cic. de Or. 1.176 between the Claudii and Claudii Marcelli, which we shall discuss below.58 Finally, there is a stipulation in the Twelve Tables regarding the actio familiae erciscundae, the function of which was to divide an inheritance. This is an obscure process, and also requires more discussion later, but the sources explain it as the division of property owned by a consortium. It is clearly relevant in the context of the numerous occasions within the 54 55

56 57

58

Dig. 26.4.6: intestato parente mortuo adgnatis defertur tutela (‘with an intestate parent dead, guardianship reverts to the agnates’). Dig. 26.4.9: si plures sunt adgnati, proximus tutelam nanciscitur et, si eodem gradu plures sint, omnes tutelam nanciscuntur (‘if there are many agnates, the nearest is to take guardianship, and if there are many of the same degree, they all are to take guardianship’). Watson (1967) 121–2. The Laudatio Turiae proves what we would otherwise have to assume, that tutela also existed to control the disposition of property in intestacy. The clause is cited together with V.4–5 at Rhet. Her. 1.13.23 and Cic. Inu. 2.148, in discussing the various clauses applied to the dispute over the will of one Malleolus, who killed his mother and was tried in 101 bc; cf. Cic. Tusc. 3.11 (also quoted at Non. p. 711L); Rep. 3.45; Varro Rust. 1.2.8 (cf. Colum. Rust. 1.3.1); Hor. Sat. 2.3.217–18; and in the legal sources Dig. 26.1.3 pr.1 (Ulpian), 27.10.1. pr. (Ulpian), 27.10.13 (Gaius), 50.16.53 pr. (Paul); Pauli Sent. 3.4a.6–7; Tit. Ulp. 12.1–3; Gai. Inst. 2.64; Just. Inst. 1.23.3. See Continisio (1987); Guarino (1973) 244–53; Fayer (1994) 559–87. Watson (1975) 109; below, 1.9.

The ancient evidence

29

Table that we have come across the actions and rights of a group of individuals as opposed to a single individual.59 The object of this account has been to set out, hopefully without preconceptions, the evidence for the actions of the members of a gens as described in the Twelve Tables. We have here a valuable indication of what the gentiles did. Succession and tutela are defined by a series of concentric circles as it were: sui heredes, agnati (or at least agnatus proximus), gentiles. Almost all the provisions which we have considered are predicated on the desire to maintain property within a group; tutela controls the movement of property, especially in the event of a woman marrying; the provisions of wills are confirmed; in the case of intestacy, the gentiles are the final inheritors if all other groups fail; guardianship of those declared incapable of managing their affairs is asserted, and again, we can infer, I believe legitimately, that in the case of the exhaustion of sui heredes and agnati, the gentiles have a role. The group’s interests defend against the actions of individuals, such as women in a position to marry who could take property away from the group, madmen or spendthrifts, and there is a clear procedure that avoids any doubt if someone dies without a will, or if eventualities have conspired to render the will invalid or irrelevant.60 Notice that only in one instance, as far as we can tell, are the rights of the group overruled, and that is for Vestal Virgins, whose freedom is assured, and whose intestate property reverts in publicum; and notice also that in extreme circumstances, even the absolute freedom of a testator could be curtailed in favour of the group through V.7. At the same time, we have no evidence of inheritance or action by a corporate body, and actually no reference as such to the gens, only to the gentiles, and the gentiles operate only in instances of intestacy. How common intestacy was is an issue which is practically impenetrable for the period in which the laws were written down, but it is not implausible that it was relatively common, not least because the outcome of intestacy need not have been different from the testator’s wishes. It has also been pointed out that the rules on intestacy are unlikely to have been wholly at odds with the expectation of a will, and that wills may have taken the intestate provisions as a default; the ‘equitable division of patrimonies’ may have been commonplace.61 We may conclude with a famous summary and an unfortunate tease. Although we rely heavily on the Corpus Iuris Ciuilis of Justinian for information on individual clauses and on wording, the early nineteenth-century 59 60

See Fayer (1994) 24–8; Bannon (1997); below, chapter 7.2, for detailed discussion. 61 Saller (1994) 164. See Voci (1960) 83, ‘la funzione dell’hereditas e´ di continuazione’.

30

The evidence for the gens

discovery in the library of the cathedral at Verona of a more or less complete text of Gaius’ Institutes gave us a tremendously important additional source.62 Justinian’s Institutes relied heavily upon this lucid introduction to Roman law, dated to the second century ad. We know Gaius also wrote a commentary on the Twelve Tables, which must have drawn on earlier sources, but it only survives in citations within the Digest – a sad loss, but an indication of Gaius’ knowledge and expertise. The beginning of Book 3 of the Institutes is concerned with intestate succession, and draws on the Twelve Tables; at 3.17, Gaius reports: si nullus agnatus sit, eadem lex XII tabularum gentiles ad hereditatem uocat. qui sint autem gentiles primo commentario rettulimus; et cum illic admonuerimus totum gentilicium ius in desuetudinem abisse, superuacuum est hoc quoque loco de eadem re iterum curiosius tractare.63

Frustratingly, we do not have the description in the first commentary; it is either lost from this text, or was in a different work. Nevertheless, the reference is precious; it gives us a secure terminus ante quem for the gens as a significant institution in terms of inheritance law. We can reasonably state that what we learn from Gaius and later lawyers is a reconstruction of an alien system and not one still in the process of development. 6 . g entilicial exogamy and the nature of the g e n s One aspect of the gens, which helps us to define its structure, is the apparent fact that it was an exogamous insitution. This has been maintained in several works by Gennaro Franciosi, but the evidence is not substantial.64 First, there is no known instance of marriage between people with the same nomen. In addition, in a fragment of Livy Book 20 (which covered the period 241 to 219 bc), we find the following: P. Celius patricius primus aduersum ueterem morem intra septimum cognationis gradum duxit uxorem.65 62

63

64 65

In the context of the next chapter it is important to note that Niebuhr was one of the first to be able to use this text, though a considerable amount of evidence was nonetheless available through the Digest to scholars like Sigonio and Vico. ‘If there are no agnates the same law of the Twelve Tables gives the inheritance to the gentiles. Now we explained who the gentiles are in the first commentary; since we pointed out there that the whole law relating to the gens has fallen into disuse, it is a waste of time to go into any details here.’ This is the major theme of Franciosi (1999); cf. Franciosi (1988); Biondo (1995) and Franciosi (1995). See also Linke (1995) 28–31; and Bettini (1990). Livy Cod. Per. lat. 3858C (from Book 20): ‘P. Celius was the first patrician to take a wife against the ancestral custom within the seventh grade of relationship.’ Celius is best emended to the patrician name Cloelius. Bettini (1990) suspects the passage of being a mediaeval forgery.

The ancient evidence

31

This shows that marriage between cognates and agnates to the seventh degree, essentially siblings and cousins, was not customary. How far mos was binding is difficult to say, but we shall meet below some customs which were rigorously observed. Franciosi argues that what he sees as a prohibition against marriage within the seventh grade was secondary to that which forbade marriage within the gens, and notes that the rule uses the same terms as the ius osculi or right of relatives to kiss each other on the mouth (famously thought to allow the paterfamilias to check if there was alcohol on the breath of a female relative).66 The primary prohibition was marriage within the gens. Apart from the argument from the absence of evidence, an important additional indication is provided by Plutarch in two of his Roman Questions. In his explanation of why women kiss the kinsmen on the lips, Plutarch states that men did not marry women related to them by ties of blood (   ), but that at some point, this provision was relaxed, as a response to a particular case where a man of good character married his cousin, an heiress, and was thought to be growing rich, but the people refused to prosecute and allowed marriage as far as the grade of cousins, but no further. In QR 108, Plutarch reiterates the statement that the Romans did not marry close kin ( 


 ) and suggests three reasons: because this enlarged relationships, or because disagreements in marriages of near kin disrupted natural rights, or because women needed many protectors, and if they married outside their household, then their kinsmen were available to the women if their husbands wronged them. The laws of the Twelve Tables which control passage out of the gens would seem to reinforce the general principle that marriage took a woman and (more importantly) her property from one gens to another, and was therefore a matter of the deepest seriousness. Some confusion has been caused because of Livy’s account of the reward to Hispala Faecennia, the freedwoman who exposed the Bacchanalian conspiracy in 186 bc.67 Livy describes her reward as giving her, essentially, the rights of a fully free citizen woman, which include the obscure terms datio and deminutio; and also gentis enuptio (marriage out of the gens) and the choice of her own tutor, just as if she were a free woman who had been given these rights by her husband in his will. It has sometimes been thought that the explicit grant of gentis enuptio implies that this was not the common practice for free women, but the sense of the passage is quite against 66 67

Polyb. ap. Ath. 10.440f; Plut. QR 6 = Mor. 265B-E, Comp. Lyc. et Num. 77b; D. Hal. 2.25.6; Cic. Rep. 4.6; Val. Max. 2.1.5, 6.3.9; Plin. HN 14.89; Gell. NA 10.23.1; Tert. Apol. 6. Livy 39.19.5; discussed at length and with extensive bibliography in Franciosi (1999) 21–55.

32

The evidence for the gens

this. What it implies is that as a freedwoman, there were restrictions on her freedom to marry, but that these were lifted as a reward, and she was thus permitted to marry like a free woman. The passage actually confirms the expectation of exogamy. Although the restrictions seem to have been loosened over time, exogamy remained the standard practice.68 From the evidence we have already discussed, it is possible to come to some kind of picture of what the gens looked like in relation to other gentes. It is perhaps best shown schematically, as Linke has done in his very important account, and the table in Figure 2 illustrates the scheme well.69 Membership of the gens passes through the male line. The gens extends beyond the agnatic family to include all members who have the same nomen70 and claim some kind of kinship; as we shall see, this is largely a useful fiction.71 The gens is exogamous, and so is, in optimum circumstances, refreshed by the admission of women from other gentes as it loses its own women. This basic model lacks only one key piece of information, which is the size of the gens; naturally, this will have varied depending on the success of families in reproducing themselves agnatically. The only evidence we have, for the Claudii and the Fabii, is desperately difficult to use,72 and both of these gentes may have been unusual. Having come thus far, and established a picture of what gentes looked like, and some of the ways in which gentiles acted with regard to each other within and between gentes, we turn in the next two sections to the stories which were told about their origins, and some of the customs they adopted. 7. legendary and historical genealogies In the previous sections we have moved towards a working definition of the gens using lexical and legal evidence. The outline is not greatly in dispute, inasmuch as it represents a theoretical statement of the gens as a collection of families, but it is very basic. Adding complexity brings controversy. We may begin with the issue of the alleged kinship of the members of a gens. In the Digest (50.16.195.2) we find a definition which refers explicitly to this aspect. The definition uses the word familia as is not uncommon in later 68

69 70 72

For later developments, including the permission of marriage of cousins, see Watson (1967) 38–9; for marriage of first cousins, see Treggiari (1991) 108–19; for the rarity of this even in the early Empire, see Saller and Shaw (1984). See Linke (1995) 17–44, and below, 10.2, for an engagement with the major arguments of the book. 71 See below, 1.7. See above, 1.4. See below, 7.2 (the immigration of the Claudii); 9.2 (the Fabii at Cremera); and general comments on the size of the Roman population at 5.2 below.

The ancient evidence

33

Figure 2. Schematic representation of intermarriage between gentes; after Linke (1995) 35.

sources which confuse gens and familia,73 but it is clear that it means a gens because it refers as an example to the Iulii, and it characterises this group as ab eiusdem ultimi genitoris sanguine (‘from the blood of the same ultimate parent’).74 The definition of Isidore which we quoted earlier, though it 73 74

See below, 1.10. item appellatur familia plurium personarum, quae ab eiusdem ultimi genitoris sanguine proficiscuntur (sicut dicimus familiam Iuliam), quasi a fonte quodam memoriae; Franciosi (1989) 8 describes this as a ‘definizione nebulosa’, showing that by the later principate, the gens was not more than a memory.

34

The evidence for the gens

refers to a gens in the sense of a nation, has the same feature.75 Let us consider these genealogies in detail.76 It is clear that one aspect of Roman historical scholarship which fed into both historical narrative and antiquarian scholarship was the development of family history. This latter aspect was recognised by Cicero and others to be problematic, since it led to accounts which glorified individual families and therefore, allegedly, corrupted the early lists of magistrates.77 Moreover, it is plain that this process of self-invention at a family level was a deep structural element of Roman political discourse, reinforced within the context of the funeral, where the masks of ancestors were displayed and the previous glories of families rehearsed.78 Arguably, therefore, Roman history, with its concentration on a relatively narrow group of noble families, itself tells the story of familial self-definition. The purpose of this section, however, is to look at some family stories and see whether they relate specifically to the concept of the gens, and most particularly to ask whether certain kinds of genealogies are in any sense a defining feature of a gens, and when the creation of these kinds of stories may have begun. Two clear conclusions will emerge along the way. First, there is no evidence for the concept of the pater or princeps gentis as a formal office or position. Gentes were acephalous; there was no system of primogeniture, and therefore in a gens which comprised several families, no paterfamilias could formally outrank another. If on occasion a leader of a gens emerged, this must have been by consent or force majeure, and not formal entitlement. Furthermore, princeps gentis refers usually to the ancestor of a gens, a clearly mythical figure as we shall see, or, in one instance, the first of a clan to reach curule office.79 This has long been recognised,80 but frequently forgotten, and it is important that when modern writers, particularly in the context of Etruria, use the term principes, which derives from Livy, we should not confuse this with the concept of the princeps gentis. Second, we 75 76 77

78 79 80

See above, 1.2. It would seem that the traditions of creating legendary genealogies for nations may have partly influenced the Roman need to create them for gentes, as well as vice versa. This whole section is enormously indebted to Wiseman (1974); see also H¨olkeskamp (1999); for a possible Praenestine example, see Peruzzi (1996). Cic. Brut. 62, referring to speeches and funeral orations, which were preserved and reused at successive funerals, and which contained false triumphs, too many consulships (an interesting comment, perhaps suggesting accounts at variance with the Fasti), false relationships and alleged transitiones ad plebem (in other words, claims of original association with a patrician gens); Cicero gives the example of a claim he might make to be descended from the patrician consul of 500 bc, M’. Tullius. Flower (1996). Cornell (1995) 246; mythical ancestors, D. Hal. 6.69.1; first curule magistrate, Cic. Fam. 9.21.2; for Suet. Tib. 1, which Cornell takes as a reference to a mythical ancestor, see below, n. 106. See for instance K¨ubler (1910) 1183.

The ancient evidence

35

can clearly see that, at least by the late Republic, genealogical thinking was not confined to patricians, but was open to plebeians as well. Let us start with the Aemilii, whom Plutarch says most writers agree to have been one of the ancient patrician houses of Rome.81 We learn one story about their beginnings from Paulus’ abbreviation of Festus: quod ab Ascanio descendat [sc. Aemilia gens] qui duos habuerit filios, Iulum et Aemylon.82

Livy does not mention Aemylos, but does acknowledge that the details of the founder of the Iulii are obscure, and mentions the possibility of brothers, though not in a way that would admit Aemylos.83 Silius Italicus, in his description of L. Aemilius Paulus, has a different version: . . . numerare parentem Assaracum retro praestabat Amulius auctor, Assaracusque Iouem.84

Amulius and Numitor are the rival kings of Alba in the story of Romulus and Remus; the family tree goes back from Amulius to Assaracus, father of Tros, who founded the Trojan royal house,85 and was the great-grandfather of Aeneas. So just as the Iulii claim both the Trojan and the Alban (and thereby Roman) royal houses, so do the Aemilii. There is yet a further story that Romulus’ mother was Aemilia, the daughter of Aeneas and Lavinia, who slept with Mars.86 Clearly, the ancestor is the person from whom the Aemilii derived their nomen, the name Aemilius which they all shared. 81

82 83

84 85 86

Plut. Aem. 2.1:       ! "#$%&!
"
 ' $!  $"() )

%*"( +
,) (‘most writers agree that the house of the Aemilii was one of the ancient and patrician houses at Rome’). Festus (Paul.) p. 22L: ‘Because they [the gens Aemilia] descended from Ascanius, who had two sons, Iulus and Aemylos.’ Livy 1.3.2: haud ambigam – quis enim rem tam ueterem pro certo adfirmet? – hicine fuerit Ascanius an maior quam hic, Creusa matre Ilio incolumi natus comesque inde paternae fugae, quem Iulum eundem Iulia gens auctorem nominis sui nuncupat (‘I shall not discuss – for who could state a matter so old as certain – whether this was Ascanius, or an older brother, who was born to Creusa whilst Troy was unharmed, and was from there a companion of his father’s flight, whom the gens Iulia claim as the author of their name.’). Sil. Pun. 8.294–6: ‘Amulius the founder allowed him to count back to Assaracus as an ancestor, and Assaracus to Jove.’ Hom. Il. 20.215–40; Aeneas gives his lineage. Plut. Rom. 2.3: - &  . " ' /0 , 1%" )

" (‘some say that Aemilia, daughter of Aeneas and Lavinia, slept with Ares [Mars]’). This is an interesting story because the chronological closeness of Aeneas and Romulus must indicate that it predates the demonstration that there was a vast chronological gap between the Trojan Wars and the eighth-century foundation of Rome. Ennius may have worked on a short chronology, but Fabius Pictor already knew better; see Cornell (1975); (1986) 247. Although Aemilia cannot have founded the Aemilii, the relationship to Romulus is a strong claim for the antiquity of the lineage.

36

The evidence for the gens

The Aemilii also claim a connection with Numa Pompilius, the second king of Rome. The connection is through Pythagoras: Aemiliam gentem appellatam dicunt a Mamerco, Pythagorae philosophi filio, cui propter unicam humanitatem cognomen fuerit Aemylos,87

but in another version, Numa, the disciple of Pythagoras, named his son Mamercus after Pythagoras, thus founding the Aemilii Mamerci.88 Finally we should note a further version: Gens Aelia appellatur, quae ex multis familiis conficitur.89

So we have several different genealogical versions, all related to the name of the Aemilii; they connect the family to Aemylos (son of Ascanius), or Amulius (rival of Numitor), or Aemilia (daughter of Aeneas and mother of Romulus), and Mamercus nicknamed Aimilios or Aemylos (son of Pythagoras or Numa Pompilius). The etymological explanation is perhaps less flattering, although it indicates the size of the gens and its constitution as a number of linked familiae, but it relates again to the nomen. According to Festus, the Fabii took their descent from Hercules, the great hero, and god, via Fabus or Fovus. Foui, qui nunc Faui appellantur, dicti, quod princeps gentis eius ex ea natus sit, cum qua Hercules in fouea concubuit. alii putant, eum primum ostendisse, quemadmodum ursi et lupi foueis caperentur.90

Plutarch derives the first Fabius from a liaison beside the Tiber between Hercules and either a nymph or a local woman, and then suggests the alternative that the first Fabius invented pits for trapping animals; Silius 87

88

89 90

Festus (Paul.) p. 22L: ‘they say that the gens Aemilia was named after Mamercus the son of the philosopher Pythagoras, who because of his unique humanity was named Aemylos (-2 )’; cf. Plut. Aem. 2.2: 3 & + $%! #! ' !4
" . $ $$5 67 ! 9:
;%? . @A , 0) $&") : (‘that the first of them and the one who gave his name to the family was Mamercus, son of Pythagoras the wise, known as Aimilios because of his grace (haimylia) in words and charm, is said by some who hold that King Numa’s education came from Pythagoras’). Note that Plutarch has a different version at Numa 21.2–3; Numa’s wife is from the Marcii, and hence Ancus Marcius is his grandson; his sons are named Pompon, Pinus, Calpus and Mamercus, the last of whom Plutarch says used the cognomen Rex, and founded the Marcii. The other three sons give rise to the Pinarii, Pomponii and Calpurnii Pisones (the last two plebeian families); see below. The importance of Pythagoras at Rome from an early point is indicated by the fact that the Romans set up statues of Pythagoras and Alcibiades during the Samnite Wars (Plin. HN 34.26). Festus p. 83L s.v. Gens Aemilia: ‘the gens is called Aemilia, because it is brought together from many families’; the word aemulus means striving or rivalling. Festus p. 77L; ‘The Fovi, now called Favi [Fabii], are called this because the princeps gentis was the son of the woman whom Hercules slept with in a pit. Some think that he was the first to show how bears and wolves might be caught by pit traps.’

The ancient evidence

37

Italicus makes the woman Evander’s daughter, which tidies up the story; for the legend of Hercules and Evander was well-established.91 Although largely absent from the regal period, the Fabii do find a place via their ancient connection with the Lupercalia; the two priesthoods are the Fabii (or Fabiani) and the Quintilii (or Quintiliani). Ovid, who was close to Paullus Fabius Maximus, and wrote to him from exile, gives an explanation at Fasti 2.375; the Fabii were on Remus’ side when he chased after some bandits, and caught them before Romulus and his band.92 The associations of the Iulii are exceptionally well known, because so much is made of them in the Augustan period, and because so much of our surviving antiquarian material survives through the exegesis of Virgil. The derivation of Iulii from Iulus son of Aeneas left rich opportunities for story-telling, but the story is older than Virgil. The first reference we have may be in Cato the Elder, deriving the Iulii from Iulus or Ascanius; for various reasons the fragment is controversial, but the genealogy was wellknown.93 By 69 bc, it was being proclaimed by Julius Caesar himself in his funeral oration for his aunt Iulia, widow of Marius: amitae meae Iuliae maternum genus ab regibus ortum, paternum cum diis inmortalibus coniunctum est. nam ab Anco Marcio sunt Marcii Reges, quo nomine fuit mater; a Venere Iulii, cuius gentis familia est nostra. . .94 91 92

93

94

Plut. Fab. 1.2; Sil. Pun. 6.627–36. Cf. Ov. F. 2.237. See Wiseman (1995) 80–1, 126–7 for details and references. Wiseman believes that the distinction between the two groups is connected with the censorship of Q. Fabius Maximus Rullianus (304 bc). Plin. HN 18.10 notes that like the Lentuli and Cicerones the Fabii take their name from a legume (faba). Cato ap. Origo Rom. 15.5; cf. Cic. Rep. 2.20, Leg. 1.3; Livy 1.16.5; D. Hal. 2.63.3–4; Ov. Fast. 2.499 for Proculus Iulius from Alba Longa, who attested to Romulus’ apotheosis. There are two problems; first, the insecurity of the citations in the Origo Rom., on which see Momigliano (1958) and Smith (forthcoming c); second, the claim that in the second century the Iulii were ‘obscure’ (M¨unzer (1999) 27, 413) compares the Iulii with the Servilii, Cluilii, Curiatii, Geganii and Quinctilii in having most of their appearances in the Fasti concentrated before the Licinian–Sextian legislation of 367 bc). Since the specific reference to the derivation of the Iulii from Iulus is found in the Origo Rom. and the citation also includes a reference to (L.?) Caesar, Chassignet ((1986) 59–60) thought it could not really belong to Cato, but neither objection is sustainable. Serv. A. 1.267 clearly cites from Cato Ascanius’ change of name to Iulus after his defeat of Mezentius, so Cato must have dealt with the name. The Origo Rom. reference may derive its knowledge of Cato’s reference to the subject from L. Caesar, who may himself have elaborated it to draw the connection between Iulus and the Iulii, in which case this family history would be surprisingly late. It is equally possible that Cato did make the (etymologically obvious) connection, since the Iulii may well have been vaunting their ancestry even if they were, as far as we know, less prominent in politics at the time. (Even this claim of obscurity is thrown into question by the existence of Iulii holding office in the second century bc.) It is not a scientific certainty anyway that the origin of a genealogy must coincide with a period of political power. Suet. Iul. 6.1: ‘On her mother’s side Iulia, my maternal aunt, was sprung from kings, and on her father’s connected with the immortal gods. For the Marcii Reges (that was her mother’s name) descend from Ancus Marcius, and the Iulii, to whom my family belongs, descend from Venus.’

38

The evidence for the gens

We have given here the three most significant patrician genealogies, but there are many others. For example, the sources are largely consistent in their account of the responsibility of the Potitii and Pinarii for the cult of Hercules at the Ara Maxima, awarded them by Hercules himself. There are two key elements of the tradition; one is the lateness of the Pinarii, who were therefore excluded from the sacrifice, which explains their name by analogy with $"7, to be hungry. The other aspect is the disaster that befell the two gentes in 312 bc, when the censor Appius Claudius Caecus persuaded the Potitii to let the cult be handled by public slaves, and the gens, described as being made up of a dozen families, died out within a generation.95 Clearly there is a strand of tradition which links some gentes back to the time of Evander, and gives not only massive antiquity but also a useful link with Greece. Other houses were descended from the companions of Aeneas, including the Cluentii, Geganii, and Sergii who participate in the famous games in honour of Anchises in the fifth book of Virgil’s Aeneid.96 The Sergii may perhaps be set to one side; there is no doubt that the failure of the angry Sergius in Virgil’s race is meant to indicate the disappointments and disasters in the career of L. Sergius Catilina. The Geganii reappear among the leaders of Alba Longa, the site allegedly founded by Aeneas’ son. Another gens mentioned in Virgil Aeneid 5 and the Servian commentary is that of the Nautii. The Nautii are interesting because they have the responsibility for the cult of Minerva. In one account, Ulysses (Odysseus) and Diomedes stole the Palladium, or sacred image of Minerva (Athena), from Troy. It brought them little luck, and Diomedes, whose connections with Italy, like those of Odysseus, are close and detailed, tried to return the Palladium to Aeneas; he was in the middle of a sacrifice however, and so Nautes took it.97 95

96

97

Serv. A. 8.270; Livy 1.7.12–14; Diod. Sic. 4.21.2; D. Hal. 1.40.3–5; Cic. Dom. 134; Origo Rom. 8; Macrob. Sat. 3.6.12–14 giving the details of the sacrifice, and cf. CIL VI.313; Plut. QR 60 = Mor. 278e and Macrob. Sat. 1.12.28 bringing in the fact that the cult is forbidden to women; Livy 9.29.9, Festus p. 237L; Val. Max. 1.1.17; Vir. ill. 34.2; Macrob. Sat. 3.6.12–14 for the disaster to the Potitii. Plut. Num. 21 derives the Pinarii from Pinos, one of the four sons of Numa; see above on the Aemilii. Serv. A. 5.117: et bene laudat familias nobilium: nam a Sergesto Sergia familia fuit, a Cloantho Cluentia, a Gya Gegania, cuius non facit mentionem. nobiles autem familias a Troianis fuisse Iuuenalis ostendit, ut ‘iubet a praecone uocari ipsos Troiugenas’ (‘and he well praises the families of nobles, for the Sergia family comes from Sergestus, the Cluentia from Cloanthus, and the Gegania from Gyas, whom he does not mention. Juvenal shows also that noble families descended from Trojans, when he says “he ordered those Trojan-born to be summoned by a herald”’ ( Juv. 1.99–100)). Serv. A. 2.166, 3.407 and 5.704 are the key passages, and note that the latter is a quotation from Varro’s work, de familiis Troianis; on this, and the similar work produced shortly afterwards by Hyginus, see Wiseman (1974) 157.

The ancient evidence

39

A few houses claimed descent from the period of the kings themselves; no fewer than four houses claimed descent from Numa: the Calpurnii Pisones, Pomponii (both plebeian), Pinarii and Aemilii. The Marcii Reges, as we have seen, claimed Mamercus son of Ancus Marcius as founder. A peculiar story in Plutarch, but taken from Juba and Varro, is suggestive of the kind of myth-making in the late Republic; it is connected with the foundation of the temple of Diana, founded by Servius Tullius, and has a Sabine duped by a Cornelius over the sacrifice of a heifer; evidence from coins may suggest that the same story was told of a Postumius.98 The Iunii, peculiar in many respects, are at some point bound into the Tarquinii, and the Tarquinii themselves have a brief afterlife in the first year of the Republic. As indicated, the duality between Rome and Alba Longa allowed an important historical strand to emerge. The Horatii and Curiatii were connected with a battle in the reign of Tullus Hostilius, and are part of a narrative about Rome’s final absorption of Alba Longa. The Cloelii are probably descended from a king of Alba; Livy tells us of king Cluilius, after whom the fossa Cluilia was named.99 Later in the same narrative, Livy tells us that the Iulii, Servilii, Quinctii, Geganii, Curiatii and Cloelii are admitted to an enlarged senate as a result of this.100 The Quinctii also provide the unsuccessful Luperci who joined Romulus instead of Remus in chasing after bandits. One could continue, but the major stories are all here, and we are already seeing not just the great patrician houses but also those which at least later were plebeian. The Iunii are a good instance; later they were distinctly 98

99 100

Plut. QR 4 = Mor. 264CD; Juba FGrH 275 F12. Cf. Livy 1.45.3 with Ogilvie (1965) 183–4; Val. Max. 7.3.1; Vir. ill. 7.10–14; Zonar. 7.9. There was a prophecy that the outstanding heifer in Antro Curiatius’ herd would, when sacrificed in the temple of Diana on the Aventine, bestow on the city of the one who sacrificed it dominion over Italy, but Cornelius the priest told Curiatius to bathe before sacricifing, and then Servius sacrificed the animal and hung its horns (cornua, hence Cornelius) in the temple. For a coin with a bust of Diana on the obverse and a scene of sacrifice on the reverse, see RRC 372; the moneyer was A. Postumius Albinus, who was related to the annalist. Livy gives the story without the names, and it seems that only Juba had got the story in full, since Varro’s version is slightly different again. Livy 1.22.4, 23.3. Livy 1.30.2: principes Albanorum in patres ut ea quoque pars rei publicae cresceret legit, Iulios, Seruilios, Quinctios, Geganios, Curiatios, Cloelios (‘in order that that part of the commonwealth should grow, he numbered the Alban chiefs among the patres, the Iulii, Servilii, Quinctii, Geganii, Curiatii, Cloelii’); cf. D. Hal. 3.29.7 '  B C $D: E! " ! $% F( &! )""( " * ' *%7% "%):, 0D &B " 
 S H  &B #! )"T "D): &U $)   ( ( V% . W7) &B &%D): " % % ! *!, 3$
  $7 &&" %, :7$"% - D" "  ;, "(): &B # %, ' *%% H " &B . *%% %7
 &")D):, :7$"% - F% "  D,  &B
 " %7 &%!. See Lambert (1993) 371–80 for this passage, and Rhodes (1993) 65–79; cf. Bourriot (1976) 15–18, 460–91, 595–711. The influence on Dionysius of Halicarnassus of this and similar passages is unmistakeable. 24 Wade-Gery (1931a). L´evˆeque and Vidal-Naquet (1996). Philoch. FGrH 328 F35, with Jacoby’s commentary; Andrewes (1971). The passage is cited in a variety of contexts for the use of genn¯etai for what used to be called homogalaktes, but most fully in Phot. s.v. X%
"!" : . . .  &B %7% $7
" & $%)
%"2. "G & c  d7&
`  X; , ":"%T ";" *. B ' % F %0 , *%7% &B ' ;^" #D` $%: ,  W7)` &B ! %7 *%%! -&%)7" W), * q >: - *%%"( , :$; #! $)"  ! %! F
";  $% o d) >: )7" , k $% ( % )' -"%7, ' :"%$"2" $% !