The Family in Roman Egypt: A Comparative Approach to Intergenerational Solidarity and Conflict 1107011132, 9781107011137

Table of contents :
Contents
Preface
Map
Intergenerational solidarity and family support networks in cross-cultural perspective
Household structures, marriage patterns, and inheritance strategies
Balancing benefits and obligations. Parental love and filial piety over the life course
Widowhood, remarriage, and residence patterns
Growing old in the household
The patriarchal household and the incoming daughter-in-law
Childless old age. The worst of all fates?
Conclusions
Bibliography
Index of sources
Index

Citation preview

THE FAMILY IN ROMAN EGYPT

This study captures the dynamics of the everyday family life of the common people in Roman Egypt, a social stratum that constituted the vast majority of any ancient society but rarely figures in ancient sources or in modern scholarship. The private letters and the census returns that survive among the rich papyrological evidence from Graeco Roman Egypt provide us with a wealth of information on these people not available for any other period or region of the ancient Mediterranean. The book discusses such things as family composi tion; the differences between urban and rural families; obligations of parents and children; choice of spouse; postmarital residence patterns; widowhood, divorce, and remarriage; the role of the daughter in law; adoption and fosterage; and provision for old age. The study analyses what can be ascribed to cultural patterns, economic considerations, and/or individual preferences by setting Roman Egypt into context with other traditional patriarchal societies where families adopted such strategies to deal with similar exigencies of their daily lives. sabine r. huebner (Dr. phil., Jena, 2005; Habilitation, Berlin, 2010) is Privatdozentin of Ancient History at Freie Universität Berlin and a Heisenberg Visiting Fellow at the Institutum Romanum Finlandiae and the British School at Rome (2012/13). Among her publications are a monograph on the clergy in the later Roman Empire (Der Klerus in der Gesellschaft des spätantiken Kleinasiens, Stuttgart, 2005), a co edited volume, Growing Up Fatherless in Antiquity (with David M. Ratzan, Cambridge, 2009), and the co edited thirteen volume Blackwell Encyclopedia of Ancient History (Oxford, 2012). She has also contributed many articles on aspects of the family in Greek and Roman society to international journals and handbooks.

THE FAMILY IN ROMAN EGYPT A Comparative Approach to Intergenerational Solidarity and Conflict

SABINE R. HUEBNER Habilitationsschrift, Freie Universität Berlin 2010

cambridge university press Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo, Delhi, Mexico City Cambridge University Press The Edinburgh Building, Cambridge cb2 8ru, UK Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9781107011137 © Sabine R. Huebner 2013 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2013 Printed and bound by CPI Group (UK) Ltd, Croydon cr0 4 yy A catalogue record for this publication is available from the British Library Library of Congress Cataloguing in Publication data Huebner, Sabine R., 1976– The family in Roman Egypt : a comparative approach to intergenerational solidarity and conflict / Sabine R. Huebner. pages cm Includes bibliographical references and index. isbn 978-1-107-01113-7 1. Families – Rome. 2. Rome – Social conditions. 3. Rome – Social life and customs. 4. Rome – Civilization. I. Title. hq511.h84 2013 306.85094560 32–dc23 2012040129 isbn 978-1-107-01113-7 Hardback Cambridge University Press has no responsibility for the persistence or accuracy of URLs for external or third-party internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

Contents

page ix

Preface Map 1

xi

Intergenerational solidarity and family support networks in cross-cultural perspective 1.1 1.2 1.3 1.4 1.5 1.6

2

Household structures, marriage patterns, and inheritance strategies 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10 2.11

3

The graying of the Western world and intergenerational solidarity Intergenerational solidarity in the ancient world Definitions Sources The Roman Egyptian census Overview of the chapters

Regional diversity of historic household forms Definition of the household Household forms in antiquity Household structures from inscriptions Household structures from census returns Physical layout of the house in Roman Egypt Demographic regime and household formation Household life cycles Marriage patterns in Roman Egypt Landholding and inheritance patterns Reasons for and timing of fission of multiple family households

Balancing benefits and obligations: parental love and filial piety over the life course 3.1 Fertility in Roman Egypt 3.2 Downward or upward wealth flows? 3.3 Parents and children in Roman Egypt

v

1 1 4 17 20 21 28 31 31 33 34 35 38 39 42 47 48 50 53 58 58 59 65

Contents

vi

3.4 Balancing benefits and obligations: parents 3.5 Balancing benefits and obligations: children

4

Widowhood, remarriage, and residence patterns 4.1 4.2 4.3 4.4

5

Growing old in the household 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 5.10 5.11 5.12

6

The household as main provider of care and support in cross cultural perspective Household and old age in Roman Egypt The elderly in simple and extended family households The elderly in multiple family households Solitary residence of the elderly Widowers and widows and marriage propensity of children Contact with children living far away Authority, retirement and household hierarchies Female heads of household Property transmission during lifetime Retirement contracts Conclusions

The patriarchal household and the incoming daughter-in-law 6.1 6.2 6.3 6.4 6.5 6.6

7

The chaste widow in cross cultural perspective Widowhood and remarriage in Graeco Roman society Widowhood and remarriage in Roman Egypt Remarriage in the census returns

The “mother in law belt” The daughter in law in Roman Egypt Sisters in law Tensions between husband and wife The daughter in law and care in old age Consanguineous marriages

Childless old age: the worst of all fates? Growing old without children Reasons for childlessness De facto childlessness The living situation of the childless elderly Emotional consequences of childlessness Alternatives to care by children: the spouse as principal source of assistance 7.7 Support by other kin in the household 7.8 Care by slaves and freedmen 7.1 7.2 7.3 7.4 7.5 7.6

66 80 92 92 93 94 97 107 107 109 111 112 113 114 118 122 130 131 134 139 141 141 144 150 151 154 155 162 162 163 165 167 168 169 171 172

Contents 7.9 7.10 7.11 7.12 7.13

Support networks beyond the household walls Adoption Uxorilocal marriages and the adoption of the son in law Fostering Charity

vii 174 175 187 197 198

Chapter 8 Conclusions

199

Bibliography Index of sources Index

214 250 257

Preface

It is a great pleasure to acknowledge the many individuals and institutions without whose considerable support this book would not have been written. This book was accepted as a Habilitationsschrift at Freie Universität Berlin in November 2010. I would like to express my deep gratitude to the members of my Habilitation committee, Ernst Baltrusch, Roger Bagnall, and Winfried Schmitz, for their endorsements and detailed comments on the manuscript. I am especially grateful to Roger Bagnall for his continuous guidance, unwavering support, and encouragement over the years and for the inspiration that his own work has often provided. I would also like to thank Glen Bowersock, Brent Shaw, and Mireille Corbier for their generosity in carefully reading and meticulously commenting on drafts of chapters. Over the years, Ernst Baltrusch, Walter Scheidel, Raffaella Cribiore, and Walter Ameling have been invaluable sources of general advice, insight, and encouragement and I am sincerely grateful to them. Finally, I should not forget to thank the anonymous readers of Cambridge University Press for their most diligent perusal: their comments were a tremendous source of help and inspiration when redacting the final version of the manuscript. Many thanks also go to my colleagues and friends at Columbia University, the Institute for the Study of the Ancient World, UC Berkeley, and the Institute for Advanced Study in Princeton for their active interest and helpful discussions. Special thanks go to David Ratzan, Anna Boozer, Graham Claytor, Jinyu Liu, and Richard Payne for their unfailing support and constructive criticism. I presented preliminary results and works-in-progress of this study at various institutions and conferences in the USA and Europe, and I am grateful to all (too many to acknowledge) who provided valuable and thoughtful comments. I am grateful to Mireille Corbier, Isabella Andorlini, Jinyu Liu, Joshua Goldstein, Roberta Mazza, Katariina Mustakallio, Hartwin Brandt, and Walter Scheidel for invitations to present aspects of this study in the form of seminars and workshops at the École Normale Supérieure in Paris, the Università di Parma, DePauw ix

x

Preface

University, the Max-Planck Institute for Demographic Research in Rostock, the University of St. Andrews, the Università di Bologna, the Institutum Romanum Finlandiae, Bamberg University, and Stanford University. My deep thanks also go to Elizabeth Bulls, Bärbel Trettler, and Marian Zelazny for handling all of the administrative tasks on my research grants. I also would like to thank Terrie Bramley for proofreading the manuscript, and Karen Downing, Kirstie Venanzi, and Gabriella Hoskin at the Institute for Advanced Studies in Princeton for locating various books for me. I also wish to thank a number of institutions for their generous financial support that enabled me to undertake the present work: the European Commission for a three-year Marie-Curie Research Fellowship at Columbia University and Freie Universität Berlin, the German Research Foundation (DFG) and the German Academic Exchange Service (DAAD) for a one-year research grant at Columbia University, the Institute for the Study of the Ancient World (ISAW) at New York University for a one-year visiting fellowship, and the Institute for Advanced Study (IAS) in Princeton, where I spent the spring semester of 2010 as a Member at the School of Historical Studies generously supported by the Herodotus Fund. Special thanks go to my editor at Cambridge University Press, Michael Sharp, and the members of his team Joanna Breeze and Elizabeth Hanlon, who were extremely helpful, kind, and efficient at all stages of production. I am also grateful for the sharp eye and intelligent advice of my copy-editor, Merle Read. I dedicate this book on the Roman Egyptian family to my own Roman Franco-German one, my two little girls, Nina and Marla, and Stéphane.

Map

xi

M e d i t e r r a n e a n

S e a

Alexandria Pelusium Naukratis Boubastos Heliopolis Memphis Soknopaiou Nesos

Neilopolis

Moe

Karanis

ris

Arsinoe

Philadelphia

Heracleopolis Magna

Fayum

Philoteris

Aphroditopolis Ancyronopolis d R e

e Lak

S

Narmouthis

0

10

Theodosiopolis Hermopolis Magna

Kerkeosiris Tebtunis

20

a

Oxyrynchus

Arsinoe

Magdola Kerkethoeris

e

Theadelphia

Antinoopolis

Heracleopolis Magna

30 km

Aphroditopolis 0

5

10

15 miles

Coptos Dakleh Oasis Thebae vel Diospolis Magna

0 0

50 25

100 50

150 75

100

200

250 km

125

150 miles

Roman Egypt.

chapter 1

Intergenerational solidarity and family support networks in cross-cultural perspective

In a family where the old do not fear death, the young do not fear life. Erik Erikson

1.1 The graying of the Western world and intergenerational solidarity This study deals with one of the most crucial social issues of our times – generational relations, familial organization, and the support that elder generations can expect to receive from their adult children and the wider kin group. This topic has only recently found heightened attention, owing to rapidly aging societies in the Western world over the last few decades, a process that after a time lag will also affect the developing countries. The combined effects of higher life expectancy and a decline in birth rates have been producing fundamental changes. By the middle of the twenty-first century, the old and the young will represent an equal share of the world’s population. Globally, the proportion of those aged 60 years and over is expected to more than double, rising from 10 to 21 percent between 2000 and 2050, while the proportion of children will decline by a third, from 30 to 21 percent. We will soon enter an unprecedented period of history in which persons over 60 years old will outnumber those aged 15 years and younger. The same trend is expected in developing countries, where the proportion of elderly persons is expected to rise from 8 percent at present to 19 percent by 2050, while the proportion of children will fall from 33 percent to 22 percent.1 Just one hundred years ago, the average life expectancy was somewhere in the mid-forties in Western Europe; at the turn of the millennium it had reached already the late seventies and is expected to rise another 10 years by the year 2050.2 In comparison, the average life

1

United Nations 2002: §§3, 4.

2

Maddison 2001: 28–30.

1

2

Intergenerational solidarity, family support networks

expectancy in the ancient world was somewhere between the early twenties and early thirties depending on the local disease environment.3 The graying of many modern societies goes hand in hand with considerable medical, social, and financial long-term consequences, a process that has alarmed social scientists, politicians, and the public alike. To give just one example, while in Germany today roughly 2.5 working adults support one retiree, in 2020 there will be only 2 workers for every dependent pensioner, and in 2050 the projected number is 1.3 working adults for every retiree.4 Increasing life expectancy and decline in fertility will lead inevitably to higher taxation of the working population and/or reductions in the financial support of the elderly, if more drastic measures such as the development of private pension alternatives or the delay of retirement age are not taken very soon. Lagging far behind traditional discourses about equity between classes, milieus, genders, or races, discourse about equity between generations has thus gained momentum and is currently one of the most fiercely debated issues in many developed nations. As the aging society endangers the stability and sustainability of social security schemes, policy makers faced with these national challenges in dealing with an aging society are advocating measures to strengthen intergenerational solidarity. Some scientists argue that the expansion of public welfare has negatively influenced the foundation of family solidarity. For the history of the modern welfare state beginning in the late nineteenth century and continuing through the twentieth, we can see in general an increasing involvement of the state in the life of the individual through the implementation of national programs on health, housing, and income security – not only for the elderly but also for other weaker members of society, such as widows, orphans, the ill, and the unemployed.5 The question of intergenerational support is therefore tightly interwoven with broader issues in the development of modern Western society.6 Industrialization, urbanization, the demographic transition, increased migration, and globalization have also been accused of contributing to a deterioration of traditional intergenerational obligations and the erosion of the extended family network, ostensibly perceived to provide a safe haven for the elderly.7 Increasing divorce rates, single 3

4 7

Scheidel 2012a, in press a. See also Livi-Bacci 1992: 31; Parkin 1992; Frier 2000: 787–816; Scheidel 2001a: 118–62; 2001b: 1–26. For classical Greece, see Hansen 1985, 1988; Sallares 1991: 42–293. For Roman Egypt, see Bagnall and Frier 2006: 87–8, who restore for females an average life expectancy at birth of 22.5 years. Mayer and Hillmert 2003: 74–100. 5 e.g., Mayer and Müller 1986. 6 Kertzer 1995: 365–6. Cowgill and Holmes 1972; Cowgill 1974, 1986; Caldwell 1982; Treas and Logue 1986. More recent studies expose this view as too simplistic: Hermalin 2003; Aboderin 2004.

The graying of the Western world

3

parenthood, delayed childbearing, decreasing family size, and voluntary childlessness are further factors which significantly transformed the dynamics of family ties and solidarity. Some scientists have even warned of an intergenerational warfare over limited resources, a discourse that manifested itself in an astounding media hype in the 1990s.8 These fears have been mitigated, however, by more recent studies. Generational relations and networks prove to be quite resilient to these changes.9 In addition, increasing economic wealth has not only benefitted the working generation but has also caused the standard of living for the elderly to rise constantly.10 The Madrid International Plan of Action on Ageing was produced during the United Nations World Assembly on Ageing in 2002. It recognizes the vital role the family still plays for the well-being of the aged, and further advocates respect for their dignity, equality, and non-discrimination. The Plan promotes intergenerational interdependence, solidarity, and reciprocity in order to ensure support and care, alleviate poverty, especially in developing countries, and integrate the elderly better in society to advance their health and well-being.11 While today functions such as education and old-age support are mediated by the state, this sphere was, in earlier times, almost exclusively organized by the family and household. Thus these institutions will be the main focus of our study. In antiquity in the absence of public provision the household was the most important institution for the health and welfare of its members, and the basis for redistributing resources between generations, which depended on their stage in the life cycle, their special needs, and their capabilities during this phase. The household played a critical role in caring for the vulnerable members of society: children, the ill, the disabled, and the old. Both kinship by blood and bonds established by marriage and joint living were important. Household social networks could balance needs and authority, while organizing housework, childcare, care for the elderly, and financial assistance. To make such an endeavor work, it 8

First proclaimed and widely publicized by the right-wing group Americans for Generational Equity, or AGE, who predicted that the underprivileged young who have to shoulder the growing expenses for the elderly will eventually mobilize against the elderly to reclaim their share of the pie (see, e.g., Wall Street Journal, January 13, 1986; cf. Hess and Warren Markson 1991). 9 Pelting and Smith 1991; Kertzer 1995: 363–83; Troyansky 1996: 233–43. 10 Knodel and Chayovan 2008. 11 United Nations 2002. Governments in many developing states are in fact banking on the preservation of such cohesion and mutual assistance between family members as a cornerstone for their approach to dealing with population aging, which is also on the rise in their countries. In sub-Saharan Africa we are dealing with a serious problem of its own sort, however: owing to the impact of the HIV/AIDS pandemic, many elderly people have been increasingly forced to care for grandchildren orphaned by the disease, even though they themselves are living in poverty and need support.

4

Intergenerational solidarity, family support networks

was necessary to put the interests of the family and household above those of the individual. The fulfillment of these obligations to provide mutual support and assistance was based on traditions, social expectations, affection, and reciprocity. Informal codes of behavior and traditional patterns of family support that formed public opinion were supplemented by formal laws that regulated the measures of intergenerational support. Household formation, economic means, and the availability of kin, seem to have been the most determining factors. Many, if not most, of these strategies involved the interplay of different generations from one family. However, this model mediated by the family, not the state, also had its demographic pitfalls. High mortality left many children orphaned and many elderly parents without children. We can thus observe a multitude of strategies that were used to cope with life’s uncertainties, such as widowhood, childlessness, early orphanhood, and frail old age. The regulations, debates, and struggles concerning these challenges left their traces in our sources, providing us with precious information about intersections of family relations and family economies in the society of Roman Egypt.

1.2 Intergenerational solidarity in the ancient world In this study I want to ask what were the normative ideals of intrafamilial support in Graeco-Roman Egypt, and how did these ideals play out under real-life conditions. By intrafamilial support I mean the duties and the responsibilities of parents towards their children, on the one hand, and of adult children towards their elderly parents, on the other hand. My study pursues these questions on the household level, the basic unit of society, often described as a microcosm of society in which social relationships, economic systems, and cultural norms find their reflection. In a comparative, cross-cultural approach with specific emphasis on other preindustrial patriarchal societies, I want to study the role of intrafamilial support as a source of security and mutual assistance, and shed light on the question of how these patterns of mutual assistance and support were formed over the course of life. In the present study I do not aim to portray old age or the elderly, but rather to provide insights into intergenerational interactions among family members of all ages. I would like to explain how these continuing interactions generated expectations and how they should be seen as strategies to secure support in later life. I will further discuss different forms of property transmission from one generation to the next, and link them to different family types and household patterns. Inheritance and succession patterns of

Intergenerational solidarity in the ancient world

5

a society reveal information about its cultural values, the idea of reciprocity between generations, the relationships between parents and their children, males and females, the relationships between siblings and those between spouses. The study focuses on the urban and rural middle class as documented in our papyrological material from Roman Egypt: individuals who owned some property but were not rich, small and moderate landowners, veterans, small traders, craftsmen, merchants, doctors, and scribes are the focus of this study.12 The elite and upper social classes, which dominate the literary accounts, will take a back seat. Such a study, looking explicitly at intergenerational solidarity on the household level of the common population, is lacking not only for the ancient world, but for many other preindustrial societies as well. The importance paid to interdisciplinary methods from historical demography, anthropology and sociology will become evident over the following pages. Graeco-Roman Egypt, thanks to its dry climate, has left us with rich source material not found anywhere else in the Mediterranean, and thus provides the natural regional focus for such an inquiry. Close to 50,000 published documentary papyri give us rare unmediated access to the humble and middle strata of society and their daily lives. Nonetheless, a consistent problem in the study of the ancient family is the lack of evidence for particular aspects of daily life, something which this study aims to deal with in part by investigating the possibilities of comparative studies.13 Therefore, this study will not only limit itself to the situation in Egypt, but will also raise questions of a broader comparative historical sociological bent, making extensive use of studies on other early modern and modern societies’ domestic groups, household complexities, and intergenerational support networks. These societies which will be addressed and taken for comparison in this study share some features: a common base in agriculture, high fertility and mortality rates, a proportion of the population aged over 60 of about 5 percent (compared with 20–30 percent in today’s Western societies), a high preponderance of the very young, a patriarchal family system, and the lack of a public pension system and regular retirement.14 When studying the ancient family and household, historians usually stay within the realm of the Mediterranean, and thereby tend to localize explanations for phenomena and developments. While a comparative approach is 12 14

Cf. Bagnall and Frier 2006: 72–3. 13 See section 1.4. The transition from high fertility and mortality rates to low fertility and mortality rates is termed the “demographic transition,” a process a country typically goes through when it develops from a preindustrial to an industrialized economic system. This model is based on a theory first proposed by the American demographer Warren Thompson.

6

Intergenerational solidarity, family support networks

still unusual for traditional studies on the ancient family, it is intrinsic to the study of early modern household form and composition.15 I strongly believe, however, that only by placing the Graeco-Roman world in a wider global context is it possible to recognize general aspects, on the one hand, and specific, particular ones, on the other. Comparative history can liberate historiography from these self-imposed restrictions, and make a less local or regional, or even less Eurocentric, historiography of the ancient world possible. Effects and phenomena suddenly seem less self-evident. Looking at a different unknown society can help to distance oneself from the society one knows best. Ideally, by using comparison in historical research, we can study two or more phenomena, regions, or events systematically with respect to their similarities and differences, and thereby can clarify the profiles of the respective cases. However, since I want to focus in my study on Egypt in Roman times, I attempt here what Kocka calls an “asymmetric comparison.”16 If I refer to family structures in the Ottoman Empire or early modern China, I do not intend to study these societies in their own right but rather use them as a background and reference point, to instrumentalize these cultures for a better understanding of the culture under question, Roman Egypt.17 By turning to evidence of early modern Europe and Asia, denser information allows us to reconstruct, by analogy, patterns and facts for Roman Egypt that would otherwise be lost for the ancient world. Because of gaps in our evidence or lack of clear indications, certain patterns and structures that we find elsewhere in early modern societies have escaped the attention of the ancient historian. Such a form of comparison has, of course, its risks, namely because it does not and cannot pay equal attention to the other societies. We simply cannot reach equal proximity to the original sources of these societies owing to the missing command of ancient Chinese, Arabic, or Hindi, and have to rely on secondary literature with the inherent risks of accepting interpretations of these studies without any possibility of rechecking them. And time constraints also forbid studying the historical context and its continuity in detail. So we merely concentrate on certain issues that are of interest, an approach that risks superficiality and distortion. On the other hand, in this case, it is the only way to open the possibility of comparison at all, and widen our horizon in the study of intergenerational relations, the household 15 17

Hajnal 1965, 1982, 1983; Laslett 1972, 1983, 1988. 16 Kocka 2009: 17. Kocka 2009. Cf. also van den Braembussche 1989; Haupt and Kocka 1996: 9–45; Kocka 1999; Lorenz 1999; Haupt 2001.

Intergenerational solidarity in the ancient world

7

and the family. For example, with an understanding of the structures, patterns, and mechanism of household formation in these societies, the case of Roman Egypt can be put in perspective, making common features and respective uniqueness apparent. Comparisons across time and across culture and region will allow the construction of concepts that go beyond one region or period of time, and allow recognition of some “world pattern” of intergenerational support in connection with household composition. To get an idea of daily life as it must have been for people in the ancient world, the best thing would probably be to live in a developing country for some time. Reading studies on daily life in various preindustrial societies, which were also affected by high fertility, high infant mortality, and short life expectancy, but for which we have as well denser sources than for the ancient world, is a step in this direction. Descriptions of the daily routine of the lower social strata, their households, and their families have vastly expanded my horizons, pointing me to discoveries that helped me to understand aspects of Roman Egyptian family life that I would otherwise have missed in my sources. Of course, there is always the inherent danger that whatever distinctive traits there were in Roman Egypt (or any other part of the Roman world) are potentially disregarded in the process of applying a comparative model, and we have to be aware of this. Nonetheless, studies on ancient family life have too often restricted themselves by particularism, ignorant of theories and models developed by social historians of other historical periods for studying the same phenomena. Comparisons between the ancient world and other preindustrial societies do exist, of course, but have taken place mainly in the fields of religion, production, and economic growth.18 Social history, and especially the history of the family and household, is a further very fruitful area that still awaits exploration. Ancient Mediterranean household formation has never been the object of comparative analysis,19 even though we have quality data that hardly ranks behind the data from much later societies inviting comparisons. This study will be a first step in this direction.20 By conducting this study I came to the conclusion, as far as family and household structures are concerned, that early modern and modern north18

19 20

Religion: Borgeaud 2004, esp. 207; Bodel and Olyan 2008; Gwynn et al. 2010. Economy: Meyer 1924; Rostovtzeff 1926; Finley 1973; Frederiksen 1975; Osborne 1991; Rathbone 1991; Parkins 1997; Davies 2007; Saller 2007; Bang 2009. Huebner 2010. I have started a further project comparing family and household structures in the Mediterranean from antiquity up to the nineteenth century. In this project we seek to identify significant variables that might explain shared or contrasting features and characteristics of family and household structures between regions and historical periods.

8

Intergenerational solidarity, family support networks

western Europe seems problematic for a comparative approach owing to quite different family and succession patterns. In Roman Egypt, marriage of women was early and universal. The young couple generally joined existing households, and rarely formed new ones. Brothers inherited equally, and often resided together. In contrast, in north-western Europe, the prevalent albeit not exclusive pattern was late marriage for both men and women. A relatively high proportion of men and women never married at all, and young couples generally formed new households upon marriage. When the young couples co-resided with parents after their wedding, it was often only temporary. Early modern and modern southern Europe, however, exhibited quite similar structures and actually provides good comparative material,21 as do many early modern and modern Near Eastern and Far Eastern societies. Fieldwork studies on households and family life in pre-twentieth century China and India, as well as those on the Arab world, point to remarkable similarities with Roman Egypt, and might give us insights into the potential aspects and realities of ancient family life that cannot be recovered from our Roman Egyptian sources. Furthermore, these studies offer the opportunity to employ methods and theories developed for other periods of history.22 Although the early modern and modern family in north-western and southern Europe has received a tremendous amount of attention from anthropologists during the past forty years, the same cannot be said of the family in the Near and Middle East. The number of scholars who have advanced groundbreaking studies is still in the single digits.23 Of particular interest for the family in Roman Egypt are the studies on earlier and later periods of Egypt, that is Pharaonic and Ptolemaic times and Egypt in the Middle Ages and the nineteenth and early twentieth centuries, that allow us to consider household and family forms in this region as a continuum and not in their usual isolation, thus enabling us to see the developments and broader pictures of Eastern Mediterranean households and family structures.24 Discussions on intergenerational equity focus on the economic, social, and moral obligations of the young and middle-aged toward the older 21 22 23

24

Herlihy and Klapisch-Zuber 1985; Benigno 1989: 165–94. Hajnal 1965, 1982; Skinner 1997; Reher 1998; Goody 2000; Engelen and Wolf 2005. Petersen 1968; Tucker 1985; Ahlawat and Zaghal 1989: 251–73; Al-Haj 1989; Tucker 1993; Cuno 1995, 2005; Okawara 2003: 51–75. Other recent studies on family history in the Middle East include Duben 1985; Tucker 1985, 1988; Gerber 1989; Duben 1990; Duben and Behar 1991. Pharaonic Egypt: Pestman 1961; Eyre 1984, 1992, 1999, 2004, 2007; Wente 1990; Janssen and Janssen 2007. Ptolemaic Egypt: Pomeroy 1984; Lewis 1986; Clarysse and Thompson 2006. Medieval Egypt: Cohen 2005. Nineteenth- and twentieth-century Egypt: Tucker 1985, 1993; Cuno 1995, 2005; Inhorn 1996.

Intergenerational solidarity in the ancient world

9

members of society, and in return the obligations that the adult generations have to the younger ones. The main aim of policies concerning the aging population is to secure the economic, social, and emotional well-being of the weakest generations of society, the old and the young. The heightened attention this subject received in academic literature, owing to developments in contemporary societies, led in the 1970s to the establishment of the new field of social gerontology, which studies intergenerational solidarity multidimensionally. On the macro level, the field takes into consideration the social-structural and institutional contexts, cultural and religious values, legal norms, and political agendas that govern reciprocal relations among kin. On the micro level, the field takes an in-depth look at single families, studying how they interacted with and were influenced and affected by historical processes. The American sociologist Vern Bengtson and his colleagues developed the theoretical paradigms and frameworks for studying intergenerational solidarity, a model based on social exchange theory.25 They took on the definition of solidarity by Durkheim as the structural forms by which individuals are integrated within groups.26 According to their classification, intergenerational solidarity, defined as the bond between parents and children, is based on six variables:27 (1) structural solidarity, which means the geographic proximity between parents and children, with co-residence as the highest level of structural solidarity; (2) associational solidarity, i.e., the frequency of contact between individuals; (3) affective solidarity, such as emotional assistance and the degree and reciprocity of positive sentiments among family members; (4) consensual solidarity in the form of shared opinions/worldviews; (5) functional solidarity, i.e., the degree of support and exchange of resources among family members, which includes practical assistance; and finally (6) normative solidarity, the persisting cultural and religious norms and values pertaining to obligations across generations. In Bengtson’s view intergenerational solidarity is the sum of these six dimensions, which can be substantially divided into two groups: (a) structural-behavioral (associational, functional, and structural) and (b) cognitive-affective (affectual, consensual, and normative) solidarity. More recently, Bengtson and others have turned to the solidarity/conflict model, in part responding to previous criticism that the solidarity model implies an emphasis on harmony and consensus and closes out any negative 25

26 27

Bengtson et al. 1976. Cf. also Mangen et al. 1988; Roberts et al. 1991; Silverstein and Bengtson 1997; Bengtson 2001. Durkheim 1933. For a review of more recent discussions see Roberts et al. 1991. Bengtson and Roberts 1991.

10

Intergenerational solidarity, family support networks

aspects of family relationships.28 Bengtson thus stresses that solidarity and conflict are not mutually exclusive in family relationships and can be present at the same time.29 While Bengtson and other modern sociologists study relations between parents and their adult children by conducting quantitative studies testing the model of intergenerational solidarity/conflict through empirical data from surveys and in-depth interviews, this approach is difficult to pursue for studying intergenerational solidarity in ancient societies for which we often have only anecdotal evidence. Nonetheless, we want to read our sources available for the period and region under study with these definitions in the back of our mind. Even if we lack the quantitative data, we will see that we will find all six dimensions of intergenerational solidarity also reflected in our ancient source material. The study of the history of aging and old age is also a fairly recent field in historical studies, having begun in earnest in the early 1980s with the study of early modern and modern England. At that time Peter Stearns declared the research on the elderly in societies before the industrial revolution “virgin territory.”30 Since then, however, an increasing number of studies have appeared in which scholars approach the subject from a historical, demographic, sociological, and anthropological perspective. This recent interest is not surprising, if we take into account that there are few historical topics that have as much contemporary relevance, involving economic as well as moral issues. These in-depth studies have unmasked the “golden age” theory where the elderly held a high status in societies of the past owing to prevailing moral standards, elderly people’s scarcity, and their control of wisdom and economic resources as a historical myth.31 Laslett has called this phenomenon “the world we have lost” syndrome – people have forever lamented the moral downfall of their own times compared with those of the past.32 We find many examples of this sort in antiquity as well.33 28 29

30 31

32 33

Bengtson et al. 2002. In recent years the solidarity/conflict model has been challenged by the concept of intergenerational ambivalence. The ambivalence model developed by Lüscher and Pillemer stresses the sociological and psychological contradictions or dilemmas, the coexistence of both positive and negative elements in intergenerational relations. See Lüscher and Pillemer 1998; Lüscher 2002; Pillemer and Suitor 2002; Pillemer and Lüscher 2004. Stearns 1982: 1. Quadagno 1982; Stearns 1982; Kertzer and Laslett 1995. Representing rather common opinion: Simmons 1945; Tönnies 1957; Fischer 1977; Achenbaum 1978. Laslett 1965. For the Roman world, see, e.g., Cicero’s Cato Maior de Senectute. For further references: Parkin 2003: 61–7.

Intergenerational solidarity in the ancient world

11

The last few decades have now seen the study of old age blossom into a thriving subfield of social history for the early modern and modern period34 and also the ancient world.35 They show that the perception and experience of old age depended on social class, regional and economic circumstances, specific family situation, the individual’s health and control over property, and last, but not least, the individual’s gender. These studies examine old age through the analysis of various evidence, such as literary, legal, medical, visual, papyrological, and epigraphic sources, which study definitions of and attitudes toward old age, employ life-course approaches, and consider the impact of age on the construction of identity. The authors of these studies agree that the sources reflect a great ambivalence toward old age and the elderly. Depending on the literary genre, the elderly are shown in a more positive or negative light and either the positive or negative characteristics of old age are stressed. The respective, often multilayered picture drawn of old age correlated with the perspective and intention of the author and the respective society, social status, and individual background in question. On the one hand, we have examples of old people being depicted as frail, vulnerable, spiritless, whiny, and lecherous,36 especially old women past childbearing age.37 On the other hand, we have numerous examples in our sources for elderly people who were highly esteemed for their wisdom and balanced judgment.38 While in earlier studies the influence of demographic factors for social history and old age in particular had been widely ignored, the more recent studies on old age and aging benefit greatly from the demographic sophistication we have acquired in recent decades.39 Parkin’s book, a solid and highly original introduction to the experience of old age in the Roman world, is a fine example of the benefits that derive from such an approach.40 34

35

36 37 38

39

40

De Beauvoir 1972; Stearns 1982; Hajnal 1983; Wrigley 1982; Laslett 1983. For an excellent overview, see Brettell 1991: 444. Greek and Roman society: Bertman 1976; Finley 1981; Falkner and de Luce 1989; de Luce 1994; Mattioli 1995; Brandt 2002; Gutsfeld and Schmitz 2003. Ancient Near Eastern societies: Stol and Vleeming 1998. Greek society: Richardson 1933; Garland 1985; David 1991; Falkner 1995. Sparta: Hodkinson 1983; Schmitz 2005: 125; 2007: 134–5. Roman society: Harlow and Laurence 2002; Cokayne 2003. Graeco-Roman Egypt: Rupprecht 1998. For late antiquity: Gnilka 2001; Gilleard 2007; Amerise 2008. Hubbard 1989; Baltrusch 2003. Bremmer 1987; Henderson 1987. For old women in classical art, see Pfisterer-Haas 1989, 1990. Cf. Corvisier 1985; Falkner and de Luce 1989; Herzig 1994; Mattioli 1995; Brandt 2002; Gutsfeld and Schmitz 2003; Schmitz 2007: 105–8. For the field of historical demography of the ancient Mediterranean, see Hopkins 1966; Saller and Shaw 1984; Shaw 1984, 1987, 1996; Hansen 1986; Saller 1987, 1994; Parkin 1992; Bagnall and Frier 2006; Scheidel 1996b, 2001a. Parkin 2003. See also Parkin 1997.

12

Intergenerational solidarity, family support networks

In a society with high fertility and mortality rates household composition in Roman Egypt changed much more rapidly than today. Child mortality was high, about half of all live-born children did not live up to the age of 5, and the risk of dying of infectious diseases, in childbed, or in warfare remained high over the lifespan.41 At birth, an individual had an average life expectancy in the low or mid-twenties. With high fertility and low population growth, the census population of Roman Egypt of classical antiquity ranged on the same level as India and China as late as the early twentieth century.42 While our best sources for establishing average life expectancy are the census returns from Roman Egypt,43 skeletal evidence – considered unhelpful as evidence for life expectancy – can tell us something about diseases and disabilities these ancient people suffered from during their lives, from traces left in their teeth and bones.44 Studies on the skeletal evidence have shown that, because of malnutrition and heavy physical work from a young age on, serious health problems were common. A high percentage of examined skeletons from the Roman world were affected by porotic hyperostosis, a condition that causes bone tissue to appear soft and spongy, which is usually triggered by anemia, probably due to an iron deficiency in the diet. Another pervasive condition found in the examined skeletons was enamel hypoplasia, a condition of the teeth likewise caused by malnutrition, but also by illness, parasites, infections, or fever during tooth formation. Unsanitary conditions due to fecal contamination of drinking water facilitated the spread of tuberculosis, cholera, typhus, and the plague bacillus, which also left skeletal lesions. Many succumbed to endemic diseases such as malaria at certain periods of the year. The victims’ bones show pathological lesions caused by severe anaemia.45 Furthermore, we know that infectious viral diseases, such as meningitis, measles, mumps, scarlet fever, and, most famously, the Antonine plague (believed to have been smallpox), afflicted many individuals and periodically killed a considerable percentage of the population, distorting long-term demographic trends. In general, life in the countryside was healthier than in cities. Densely settled environments were a fertile soil for the spread of infections, 41 42 43

44

45

Parkin 1992; Corvisier and Suder 2000; Bagnall and Frier 2006; Woods 2007. Barclay et al. 1976; McEvedy and Jones 1978; Bhat 1987; Anderson 1988; Wilson and Airey 1999. Bagnall and Frier 2006. The model life-table systems by Coale and Demeny, who used data drawn primarily from early modern European societies, constitute a point of reference and comparison for our data from antiquity (Coale and Demeny 1983). For a more detailed discussion of the Roman Egyptian census and the quality of data obtained, see section 1.5. Garland 1995: 20; Scheidel 2001a, in press a; Hin 2009. For a general introduction into reconstructing history through skeletal analysis: Grauer 1995. See especially Scheidel 1994, 1996b, 2001a, in press a; Shaw 1996, 2006; Lo Cascio 2006; Hin in press.

Intergenerational solidarity in the ancient world

13

which negatively affected life expectancies. Respiratory diseases due to air pollution from indoor heating and cooking were also widespread. Literary accounts tell us of many more diseases and health problems going along with certain life stages. Poor eyesight and loss of hearing were common for elderly people, as were complaints of pain in joints of the knee, neck, foot, and ankle, probably caused by arthritis.46 Elderly people in the papyri from Roman Egypt complain especially about frailty, infirmity, and poor eyesight.47 One of the major issues in the history of old age – and one that ties historical study quite closely to contemporary policy debates – is that of the residence of the elderly as it has a direct effect on availability of support and their quality of life. In order to examine the living conditions of the elderly in a given society and the options that the elder family members had when it came to sustenance and maintenance, it is crucial to study the dominant forms of household composition.48 In contrast to the modern individualized Western world which is characterized by institutionalized old-age support and thus often isolation of elderly people, the elderly in past societies were believed to have regularly lived in multi-generational households, cared for by their children and grandchildren. Yet, as more sophisticated studies have shown over the past decades, there was a wide diversity of living circumstances and household forms. In early modern Europe, for instance, one finds several different forms of living circumstances of aged parents, depending on cultural patterns, an urban or rural context, social class, occupation of the head of household, economic considerations, and respective labor demands. Some parents remained living on their own in the family home after their children married and moved out. Others left the family home to their eldest son on his marriage and moved to a smaller house on the same lot. Still others continued to live in the same household as their married sons, with the father remaining head of the household until his death. Finally, some elderly parents were circulated among their children and lived with each of them in turn for some duration.49 Peter Laslett and his Cambridge Group were the first to conduct systematic research on the 46 47

48 49

Parkin 2003. PSI 5.484 from 257 bce: ἐσμὲν γὰρ πρεσβύτεροι καὶ ἀδύνατοι; P.Enteux. 25 from 222 bce: πρεσβύτερός εἰμι καὶ ἀσθενῶ τοῖς ὀφθαλμοῖς; P.Enteux. 26 from 222 bce: ἀκληρήσαντος δέ μου κατὰ τὸ ἴδιον σῶμ[α] καὶ τοῖς ὀφθαλμοῖς ἀδυνατοῦντος; P.Oslo 3.124 from the late first century ce: [δ]ιὰ̣ τὸ ἀσθενῆ με εἶν̣α̣ι τ̣ῇ̣ ὁράσει καὶ ὑπὸ γ̣ήρους̣ σ̣υ̣ν̣ε̣χόμεν̣[ον]; P.Flor. 3.312 from 92 ce: ἀπολυθῆναι τῶν λειτουργιῶν διὰ γῆρας καὶ ἀσθένειαν; P.Oxy. 5.889 from the fourth century ce: π]ερὶ ἐμὲ γῆρας καὶ τὴν τοῦ σώ[ματος ἀσθένειαν. For an introduction to household studies and further bibliographical references, see Chapter 2. Das Gupta 1997: 46–7. For further bibliographical references see Chapter 2.

14

Intergenerational solidarity, family support networks

form and size of the preindustrial family and household and advanced standardized definitions of household types.50 Methodological considerations about the joint family household were later discussed by Wheaton and Kertzer.51 On the multiple family household, the best overall discussion is to be found in Kertzer, who also elucidates the differences between stem and multiple family systems.52 The clearest and most comprehensive exposition of the geography of historical Italian family life has been advanced by Barbagli and Kertzer.53 I will address this field of research on historical household forms in more detail in Chapter 2. This study has also been tremendously enlightened by the vast scholarship on the family, one of the most prolific areas of historical research over recent decades. Studies on the ancient family have used juridical, literary, medical, visual, archeological, and epigraphic sources, depending on region and period and availability of respective source material. More recently, the study of the family has received additional impetus from other disciplines such as gender studies, anthropology, comparative studies, and historical demography. In the progression and refinement of the field, regional and cultural differentiations within the ancient Mediterranean – development over time, on the one hand, and continuity from pagan to early Christian times, on the other hand – have been stressed.54 Of particular interest for our endeavor are those studies that treat family strategies concerning life course, marriage patterns, child rearing, and old age.55 Most studies have concentrated so far on the classical Greek56 and Roman families,57 with dominance on the Roman side, and here especially on the city of Rome, Italy, North Africa, Gaul, and Spain. Works on family and kinship in the 50

51

52 53

54

55 56

57

Laslett and Wall 1972. For a general introduction to the study of early modern and modern household systems, see Ehmer et al. 2001. Wheaton 1975; Kertzer 1989. Viazzo 2003 provides a comprehensive review of the literature on the “Mediterranean model” of household formation and deals with many of the topics that are not covered here. His article contains useful discussions of the sources, problems, and methodology, as well as annotated and critical bibliographical guidance. Kertzer 1989. Barbagli and Kertzer 1990. See also Herlihy and Klapisch-Zuber 1978; Kertzer and Barbagli 2001; Kertzer and Saller 1991. There is an immense literature on the Roman family: for an overview, see Dixon 1988, 1992; Rawson and Weaver 1997; Gardner 1998; George 2005. For the impetus from other disciplines, see especially Hopkins 1965a, 1965b; Sallares 1991, 2002; Shaw 1996; Scheidel 1996b, 2001a, 2001b. For a detailed overview of the state of research, see Rawson 2003b. Andreau and Bruhns 1990; Bradley 1991a; Corbier 1999a, 1999b, 2001b, 2006; Parkin 2003. Harrison 1968; Lacey 1968; Garland 1990; Golden 1990; Pomeroy 1997; Cox 1998; Patterson 1998; Nevett 1999; Schmitz 2007. Rawson 1986, 1991, 2003a; Dixon 1988, 1992; Bradley 1991a; Corbier 1991c; Kertzer and Saller 1991; Treggiari 1991b; Saller 1994; Rawson and Weaver 1997; Gardner 1998; Evans Grubbs 2002; Severy 2003; George 2005.

Intergenerational solidarity in the ancient world

15

Eastern Mediterranean are still rare.58 Cox has advanced an admirable study on household interests in classical Athens. She discusses marriage patterns, the relations among family members, property transmission, and the structure of the oikos of the Athenian elite.59 Her analysis, however, relies mainly on literary sources (court speeches and politic biographies), while my study is primarily indebted to the papyrological evidence (mainly the census returns, marriage contracts, legal petitions, and private letters) and thus stresses different aspects and covers different strata of society. The same applies to Gallant’s examination of the Greek household.60 The problem with his study is the scanty literary evidence, produced by a literate elite, from which he tries – with the help of modern proxy data – to extrapolate information about average household size and composition of the Greek peasant population. Earlier studies that have addressed sociological aspects of housing in Roman Egypt have been done by Luckhard (1914), Calderini (1924), Montevecchi (1941), Hobson (1985), and Alston (1997, 2005, 2006). In my 2007 article in the Journal of Roman Studies I discussed the peculiar marriage pattern of Roman Egypt, responding to and partly refuting earlier works.61 I have also studied remarriage and stepfathers in Hellenistic and Roman Egypt, household structures and family relations in the inscriptions from Graeco-Roman Egypt, and adoption and fosterage in Egypt and the Eastern Mediterranean more generally.62 Furthermore, we have several studies on the life and status of women in Graeco-Roman Egypt, the most notable by Pomeroy (1984), Rowlandson (1998), Melaerts and Mooren (2002), and Bagnall and Cribiore (2006). Studies on women in later Byzantine periods include Beaucamp (1990–2), Krause (1994–5), and Arjava (1997). Montserrat (1996) advanced a study on sex and society in Graeco-Roman Egypt and Yiftach-Firanko (2003) on marriage documents of Hellenistic and Roman Egypt. Nonetheless, there is not a single study in monograph form so far that explicitly deals with family and household in Roman Egypt, which is surprising in view of the region’s richness of sources which provide us with fascinating information about family life below the social and political elite, and the variations between urban and rural families. 58

59 61

62

Vandorpe 2002 discusses a Ptolemaic family archive; Thompson 2002 and 2006 gives an excellent overview of the Hellenistic family in Ptolemaic Egypt. Cohen 1993 and Perdue et al. 1997 studied the family in ancient Israel. Also very informative is Krause 2003. Cox 1998. 60 Gallant 1991. Huebner 2007. Earlier studies on brother–sister marriage in Roman Egypt include Pestman 1961; Hopkins 1980; Shaw 1992; Parker 1996; Scheidel 1996a, 1997a. For responses to my study, see Remijsen and Clarysse 2008; Rowlandson and Takahashi 2009; and Chapter 7 below. Huebner 2007, 2009a, 2010, in press a, in press b.

16

Intergenerational solidarity, family support networks

Despite the enthusiasm with which findings on early modern household forms which have dramatically altered widespread views about the family in past times were met by historians of later periods, these methods developed to study household composition and household strategies have rarely been exploited by ancient historians for studying the ancient family and household. Household forms in the Graeco-Roman world have never been studied in comparison with other preindustrial societies either, even though communication between these two areas of research could tremendously enhance our understanding of ancient family life, and could help us to recognize common features and identify cultural specifics or unique characteristics. One of the problems for such a study on ancient household structures is the lack of reliable source material. Saller and Shaw produced a study on commemorative patterns and likely household composition among the non-elite strata in the early Roman Empire in the West based on funerary inscriptions; Edmondson and Martin did the same for Roman Lusitania and Asia Minor.63 On the basis of their conclusions drawn from the epigraphic evidence, the assumption that the nuclear family was the typical form of living in the western Roman world has found its way into virtually every publication on the Roman family of the last two decades.64 I have argued elsewhere on the basis of an extensive survey of inscriptions for Roman Egypt that our epigraphic evidence might not be reliable for any assumptions about actual household structures.65 Moreover, studies on later periods of the Mediterranean stress the great diversity of family and household forms even within the same region. The idea of there being just one “Mediterranean family” clearly has to be abandoned.66 These studies on the family in later periods of the Italian peninsula should make us very wary that anything like the “Roman family” existed at all. We should rather expect to find enormous variations between regions just within Italy, not to speak of the other provinces of the Roman Empire. I will return to a discussion of the 63 64 65

66

Saller and Shaw 1984: 136–7, with n. 49; Shaw 1984; Martin 1996; Edmondson 2005. See Huebner 2010, in press b, and Chapter 2. As has been noted before, the Romans had no word for the nuclear family; neither familia nor domus denoted the nuclear family household. Roman law and linguistic evidence reflect extended or joint family structures (Martin 1996: 40; Potter 2006: 56–7; cf. Frier and McGinn 2003, who placed special emphasis on the degree to which Roman private law can in fact be seen as a reflection of actual family structures). The only evidence that points indeed to nuclear structures is the funerary inscriptions and the literary evidence composed by an elite reflecting the living circumstances of only a tiny minority (e.g., Cic. Off. 1.53). For a valuable general introduction to ancient epigraphy, see Bodel 2001; anyone wishing to ponder further the relevance of thousands of tombstones to the Roman social historian can profitably read Valerie Hope’s article from 1997 and Walter Scheidel’s discussion (2009). See also Chapter 2 of this study. Cf. Hajnal 1983; Barbagli 1991; Hareven 1991; Kaser 1996; Mitterauer 1996; Reher 1998.

Definitions

17

epigraphic evidence and the conclusions drawn from it regarding ancient household structures in Chapter 2. I believe that the only available source material for the ancient world that provides us with empirical data on household structures for the common people is in fact the census returns from Roman Egypt. In combination with more qualitative sources such as rental or sales contracts or divisions and registrations of property, we are able to reconstruct everyday life on the household level, a field of study that, despite its interest for the history of the family, the history of childhood and old age, and the history of household economy, has not received its due attention so far. This study is thus indebted to four areas of research: (1) studies on intergenerational solidarity with their current focus on contemporary Western and increasingly modern developing countries’ societies; (2) old age studies, a specialized field within social history that has also found heightened attention by ancient historians over the last two decades; (3) household studies, a flourishing field since the early 1970s that so far has focused mainly on early modern and modern societies; and finally (4) family studies, which have experienced a boom over the last three decades in the field of classical studies.

1.3 Definitions Various terms in this study need definitions. The study employs the concept of the life cycle of the household, which refers to periodic recurring processes during which households and families undergo changes in size and composition. Life-cycle concepts are well known in the discipline of economic history. We have the cycles of growth and recession on the macro level, and individual economic life cycles that exhibit a changing ratio between period of labor earnings to consumption and dependency over the individual’s lifespan. Household life cycles mean the different stages in the development of a household, from the point of its foundation over periods of expansion and contraction to its dissolution. Transition between stages is marked by changes in the individuals’ biographies, such as marriage, the birth of children, the death of a spouse, and so on. Not every household goes through the same cycle of stages, and each stage can vary in its length depending on the respective household members’ needs and options.67 The term “generation” can take on many different meanings varying from popular to scientific use, and varies between the disciplines such as 67

Cf. Le Bras 2008.

18

Intergenerational solidarity, family support networks

sociology, demography, and history. Its many uses are destined to cause a confusion of concepts.68 In this study I understand generation as descent relationship, not as a concept referring to age or historical time. When talking about intergenerational relations, solidarity, and conflicts in Roman Egypt, I am referring to interfamilial conflicts, not larger society-level processes. I understand generation as liberated from age-bound connotations, which does not mean that age does not play a significant role in the interaction of generations. I study parents and children and their relations that change over the family’s life course. Moreover, I wish I could study these changing relationships and contingencies also through historical time and note changes from first- to third-century Egypt, but for this our source material is unfortunately not abundant enough. Likewise, I try to avoid marking out a chronologically defined age group as “the young” or “the elderly” in this study. If I need a cutoff age, as in Chapters 3, 5 and 7, I use 14 as the end of childhood and 60 as the starting point of old age. At the age of 14, boys reached legal majority and started to pay the poll tax. Many girls entered their first marriage at this age. At the age of 62, poll tax liability ended in Roman Egypt, and the age of 68 qualified for membership in the gerousia of Oxyrhynchus, and the age of 70 for excusal from participation in certain liturgies.69 In general, it is, however, more helpful for this study to define the individual by their status in a specific life-cycle position. To give an example, her age was much less decisive for the position of a young woman in the household than whether she was the unmarried daughter of the head of the household or the young bride of one of the sons of the household. For a daughter-in-law, as long as her parents-in-law were alive and her children not yet grown up, whether she was 15 or 35, her situation and status did not really change. Likewise, for a widow with adult sons it did not really have a bearing whether she was 45 or 60 at the time when her husband died. The authority and headship of the household would regularly go over to an adult son. Since it was not likely that the widow remarried, she often took up a subordinate position in the household, dependent on financial support by her children. On the other hand, as long as an elderly woman’s husband was alive, she occupied the role as the female head of the household regardless of whether she was 40 or 70. For an elderly male head of the household it likewise did not matter whether 68

69

Cf. Kertzer 1983. Bertman and Eyben, influenced by the ideology of the student uprisings of the late 1960s and early 1970s, tried to identify certain “generations” or rather cohorts in Greek and Roman societies that were characterized by the same norms and values (Bertman 1976; Eyben 1977; see also Eyben 1993), an attempt that has not proven successful (Kleijwegt 1991). For taxation in Roman Egypt, see Wallace 1938.

Definitions

19

he was 50 or 70, since it was up to his death that he enjoyed control over all household matters. Chronological age becomes almost irrelevant in terms of these considerations. However, what mattered apart from their status in a specific life-cycle position was the old man’s or woman’s physical wellbeing, and their need for financial support and practical care. Usually the term “caregiving” has a wide range of meanings that go beyond distinct medical care. In this study I understand caregiving as a form of “social care” that ranges from medical care and nursing, in the case of an acute or chronic illness or old age frailty, to social care as an everyday activity, which included living together in one household, helping with housework, emotional support, and personal affective attention. Studying the household is an attempt to understand past societies from the bottom up, still a new approach to understanding ancient Greek and Roman societies. The fascinating thing about the household is the fact that it occurs as a concept and structure across cultures and societies. Nonetheless, as we have seen above, the household has become an analytical concept only a few decades ago, and ever since then its definition has been a highly debated issue in modern scholarship. “Household,” “family”/“kinship,” and “houseful” are categorically different concepts.70 While household usually defines the domestic group, including lodgers and servants, living under the same roof and sharing activities and the same table, the family can reach beyond household walls and encompass all those related by blood, marriage, or adoption, regardless of where they live.71 The “houseful,” on the other hand, comprises all the people who live in the same building or the same premises but form independent households. Furthermore, what has been understood in recent years is that the household cannot be tackled as a monolithic entity, but has to be studied as a flexible and permeable concept, as the sum of the interactions of its members, such as production, consumption/distribution, and reproduction.72 Household has to be understood as a process rather than a norm.73 Household boundaries are fluid, defined by the everyday practices and relationships of household members and influenced by inheritance and marriage patterns, the allocation of authority within the household, organization of production, resource allocation, and labor participation.74 70 71

72 74

More on this in Chapter 2. For the problems of defining the meaning of familia and domus in a legal and colloquial sense, see Bradley 1991a: 4; Saller 1994: 75–6, 81–2, 91–4; Gardner 1998. Hammel 1984: 31; Wilk and Netting 1984: 5; Wilk 1991; Burton et al. 2002. 73 Hammel 1972. Segalen 1986; Yanagisako and Collier 1987; Cheal 1989: 11–22; Kabeer 1991; Fraad et al. 1994; McKie et al. 1999; Han 2004: 408–23; Komter 2005; Stone 2005.

20

Intergenerational solidarity, family support networks

I often refer to patriarchal societies and mean by them societies in which the eldest male of the family, usually the father or grandfather, holds authority over all women and children as well as the younger men of the household, with children belonging to the father’s lineage. Women in these patriarchal societies, being assigned a role primarily in reproduction and administration of the household, were usually barred from any political leadership, but were not necessarily excluded from control over property or the right to inherit. The concept of patria potestas in its absolute form as defined in Roman law and tradition, however, was unknown in Egypt, and even after 212 ce, when Roman citizenship was given to basically all free inhabitants of the Roman Empire, was never as rigorously interpreted there as in classical Roman Law.75

1.4 Sources The study of the life of the lower social strata, focusing particularly on the humble peasants and craftsmen, children and the elderly, is enormously constrained by the limitations of our source material. Literary texts were written by and for an elite, and their value is curtailed when one is studying the experiences of the majority of the population. Literary sources become more informative for the middle and lower classes only in the later period of the Roman Empire. Moreover, we have few contemporary accounts about household structures in general. Ancient authors mention the topic only en passant, but never make it their central point, as domestic arrangements were probably too obvious or tedious to talk about. So the exceptional receives attention, not the typical. The description by Plutarch, for instance, of the crammed living conditions of the Licinii Crassi of the first century bce,76 or of the Aelii Tuberones of the second century bce,77 might indeed portray the typical living style of the humble peasant population in ancient Latinum. In this context, these cases serve, however, rather as literal examples rendering an ideologically transfigured view of Rome’s glorious past when senators were still unassuming and hard-working men than as an account of historic realities. Moreover, when assessing the ancient family support system, literary accounts provide us with the major features of its ideal working, but often not with its historical reality. The juristic writings can give us the framework but does not tell us much about lived reality. Also archeology’s contribution to this interdisciplinary problematization has 75 76

Taubenschlag 1916; Lewis 1970; Lippert 2008: 117–18; cf. Arjava 1998. Plut. Crass. 1. Cf. section 2.1. 77 Plut. Aem. 5. Cf. section 2.1.

The Roman Egyptian census

21

been limited, despite the publication of several archeological works on ancient houses and households in recent decades.78 One problem is that the definitions and methodology developed by sociologists for studying the household have not yet been incorporated in the archeological studies of the Greek and Roman household, or in their theoretical and interpretative practice. This might be partly due to the fact that archeological studies cannot recover household dynamics, and theories and models used for household studies have little bearing on the archeological material remains of the house.79 Nevett (2010) has recently advanced an innovative study which tries to push the archeological evidence further and explores new approaches to contribute to the academic mainly text-based discourse on the ancient family and domestic space. Daniel has contributed an analysis on architectural orientation in the papyri (2010). However, recovering the size and composition of a family from archeological remains alone seems to be a difficult endeavor. The papyrological evidence from Graeco-Roman Egypt, however, is an exceptional case for the study of the ancient world for which information about the common population is very scarce. It yields information about the life of the lower social strata, such as farmers, small traders, and craftsmen who constituted the vast majority of the population, and therefore has special importance in our inquiry.

1.5 The Roman Egyptian census The population of Egypt in Roman times consisted of Roman citizens, citizens of the Greek cities, and indigenous Egyptians who mainly lived in the small towns and villages and doubtless constituted the vast majority of the population. The Roman government took every effort to maintain this clear separation by a complex set of rules. They eventually became obsolete in 212 ce with the Constitutio Antoniniana, which granted all free inhabitants of the Roman Empire full Roman citizenship.80 The capital, Alexandria, had in Roman times about 200,000 to 600,000 inhabitants and was, after Rome, the largest city in the Roman Empire.81 The low estimation for population size of a nome capital or 78

79 80

81

Cf. Nevett 1994; Alston 1997; Ault and Nevett 1999, 2005; Nevett 1999. For a new approach of studying the household as a process in a social archeology, see now also Souvatzi 2008. Alexander 1999; Souvatzi 2007. Cf., however, also Segalen 1986: 27; Allison 1999; Nevett 1999: 6–12. Cf. Bagnall 1997b on the complex question of the ethnicity of the people in the Roman Fayum and whether they saw themselves as Greeks or Egyptians. Cf. Diod. 1.31.6–8; Joseph. Bell. Jud. 2.385. For discussion of the population size of Alexandria, see Tacoma 2006: 30–1. Antioch in Syria ranked third, according to Joseph. Bell. Jud. 3.2.4 §29.

22

Intergenerational solidarity, family support networks

metropolis (Arsinoe and Oxyrhynchus being two of the around fifty nome capitals in Roman Egypt) is about 15,000 to 25,000 inhabitants.82 The total number of people living in the metropoleis is thus estimated to be about 0.75 to 1.25 million. The number of villages in Roman Egypt ranged, according to Bagnall and Frier, from 2,000 to 3,000 with an average population of 1,000 to 1,500, thus totaling approximately 3 million.83 While in 1994 Bagnall and Frier set the percentage of the population who lived in the cities to one-third of the total population of Roman Egypt,84 Bagnall, influenced by subsequent studies that argued for smaller urban population sizes,85 later lowered his estimate to about 15–25 percent of the total population, which would still mean a fairly high degree of urbanization due to the high population density in the inhabitable portion of land.86 Among the papyri many private letters, contracts, court proceedings, wills, and, most importantly for our angle, census lists have survived.87 Our modern term “census” comes from the Latin word censere, “to count or estimate.”88 The procedure of counting people is already testified in our earliest historical records. Knowing the number of the (adult male) population was decisive for a government’s ability to assess its labor availability, defense readiness, and potential tax revenues. The first known census was taken by the Babylonians in 3800 bce. Every six or seven years the Babylonians systematically recorded people, livestock, and agricultural products. In Egypt the practice of conducting a periodic census reaches as far back as the late third millennium bce. The principal aim of the pharaohs was to establish the availability of labor. Household listings have also survived from the Egyptian workmen’s village Deir el-Medina from the second half of the second millennium bce. For late eighth- to early seventh-century Assyria, a collection of tablets known as the “Harran census” have come down to us, on which the heads of all households, their family members, landholdings, animals, and all other property were recorded. Most of the individuals recorded were common people, farmers, 82 83 84 85 86 87

88

For the debate on the numbers, see Rathbone 1990: 119–37; Bagnall and Frier 2006: 55; Tacoma 2006: 31. Goldsmith 1984: 271–2, 283; Bagnall and Frier 2006: 55–6. Bagnall and Frier 1994: 56 (page numbers are the same for the 1994 and 2006 edns.). Scheidel 2001a: 183–4, 247–8; Tacoma 2006: 31; cf. Lo Cascio 2009: 97–103. Bagnall 2009: 107–11. These census returns have been analyzed by Hombert and Préaux 1952; Bagnall and Frier 1994 (I generally refer to the 2006 edn.); and Bagnall et al. 1997 (P.Oxy.Census). See also the comments and criticism by Scheidel 1996b and 2001a. For a general overview of the history of the census in the ancient world and further literature, see Huebner 2012b.

The Roman Egyptian census

23

gardeners, and shepherds, and average household size was four individuals. The Bible refers to several censuses for ancient Israel.89 For Attica, the orator Demetrius of Phalerum is said to have held the first census in 317 bce. He counted 21,000 adult male citizens, 10,000 metics, and 400,000 slaves.90 The Ptolemaic census in Egypt, in operation at least from the reign of Ptolemy II, seems to have been based both on occupational groups and on the household as declaration unit, and was used mainly as the basis for raising taxes.91 The declarant gave the names of all adult members of his or her household, their gender, and their relationships to each other. Their livestock and agricultural produce were also listed. In contrast to the later Roman census in Egypt, however, ages were not supplied and minors were not recorded, which makes the earlier census unsuitable for any more detailed demographic analysis. From the few declarations we have so far, we can only conclude that the Ptolemaic census was little organized and standardized, and it did not yet take place periodically. Existing lists were, however, updated on a fairly regular basis.92 The Romans according to tradition conducted a census as far back as the sixth century bce, an introduction of Servius Tullius. It was first organized by the consuls, and the office of the two censors was introduced in the midfifth century to release the consuls from this burden.93 Varro gives an account of the procedure.94 Every five years all adult male citizens had to register themselves with their full names and ages. Each head of household also had to declare his wife, his children, and all other members of the household, including slaves. Finally, he had to list all his property and give an assessment of its total value. Widows and orphans were registered by their legal guardians, but their number was not included in the total count of capita civium.95 First, only citizens of the city of Rome itself were included, until later in the early third century bce citizenship was extended 89

90 92 93 95

In about 1500 bc, at the time of the Exodus, Moses was given the order to count the Israelites (Num. 1:1–4 “The Lord spoke to Moses . . . ‘Take a census of the whole Israelite community by their clans and families, listing every man by name, one by one. You and Aaron are to number by their divisions all the men in Israel 20 years old or more who are able to serve in the army. One man from each tribe, each the head of his family, is to help you.’” Cf. Exod. 30:1). According to tradition, in 1017 bce King David conducted a census for military purposes, which brought divine wrath upon Israel (2 Sam. 24:1–9; 1 Chr. 21:1–6, 23:3, 27:24; cf. for the census of Solomon, 2 Chr. 2:17). Finally, the famous passage in the Gospel of Luke (2:1–3) refers to the Roman census for which Mary and Joseph had to travel to Bethlehem. Athen. 6.103 272C. 91 P.Lille 1.27; P.Frankf. 5; W.Chrest. 198; P.Alex. 553; P.Tebt. 3.814.45–58. Clarisse and Thompson (2006) recently advanced an impressive analysis. Liv. 4.8.2–7, 4.22.7, 4.24.2. 94 De ling. Lat. 6.86–7. Liv. 3.3. The census was held in the Campus Martius in a special building called the Villa Publica, erected in 435 bce for the purpose of the census by the second pair of censors, Gaius Furius Pacilus

24

Intergenerational solidarity, family support networks

to Italy as well. Caesar Augustus finally included the entire Roman Empire. The Roman census was conducted for taxation and military purposes, and played a crucial role in the administration of the empire. Taking part in the census was obligatory, and voluntary absence or refusal was punished severely.96 Censuses were also held by the Roman provincial governments, though they do not seem to have taken place all at the same time or with the same frequency. We are best informed about the census of Roman Egypt, conducted at a fourteen-year interval from at least 11/12 ce to 257/8, for which nearly 400 individual census returns have survived among the papyrological evidence. For most parts of this study we focus on the 304 returns which almost exclusively originate in Middle Egypt and have been treated in Bagnall and Frier’s first edition of The Demography of Roman Egypt (1994). They record 1108 persons. The additional 63 households (of which 36 are complete or nearly complete), probably from Lykopolis in Upper Egypt, published as P.Oxy.Census, which record an additional 256 individuals, are fraught with various problems, partly due to the small size of the sample, and partly due to the bad preservation of the papyrus roll they were written on.97 The Upper Egyptian returns also exhibit a high degree of lost ages and the tendency of age-rounding. After the first edition of The Demography of Roman Egypt was in print, about 33 additional census returns from Middle Egypt became available to the authors, which they added as a supplement to the second edition (2006), but did not take into consideration for the discussions and the tables. These additional census returns record around another 100 individuals. All inhabitants of Roman Egypt, even those exempt from the poll tax, had to take part in the census.98 The principal use of the census was to gain accurate lists of men liable to the poll and other capitation taxes, but this was not its only use as otherwise women, children, and men above the age of 62 (when the capitation tax ceased to be due) would not have been included. Because they are all-encompassing, the census documents allow studies into the domestic group, household complexity, and marriage patterns. In order to facilitate the operation of the census, individuals had to return to their legal domiciles, as difficult as in some cases it must

96

97 98

and Marcus Geganius Macerinus (Liv. 4.22; Varro, Rust. 3.2). The censors drew up registers of the census, called the tabulae censoriae or libri censorii, and once finished, they counted up and announced the number of citizens. We hear about the death sentence or sale by the state into slavery (Cic. Caec. 34). Service in the army was, however, a valid excuse for absence from the census. Bagnall et al. 1997: 89–94. For place of origin, see Montevecchi 2000: 175–84. Bagnall and Frier 2006: 12.

The Roman Egyptian census

25

have been to define this legal or fiscal domicile. We have a fragmentary edict on papyrus of the prefect of Egypt, Gaius Vibius Maximus, from 104 ce: “The census by household having begun, it is essential that all those who are away from their nomes be summoned to return to their own hearths so that they may perform the customary business of registration and apply themselves to the cultivation which concerns them.”99 The text of the edict reinforces of course the famous passage in the Gospel of Luke 2:1–3. The census of 257/8 ce seems to have been the last of this regular fourteen-year cycle to take place in Roman Egypt. The discontinuation might be connected with changes in the system of capitation taxes. Under Diocletian the tax system of the late Roman Empire was restructured, which should probably be seen in connection with his other efforts in reorganizing provincial administration and the coinage.100 Diocletian introduced a five-year census in 287, before the fifteen-year indiction cycle was established in 312 ce. The quality of the Roman Egyptian census, “the best available demographic source for any population prior to the Renaissance,”101 is in general regarded as very high. The quality of reporting ages is, for instance, comparable to that of census data from Western European countries in the nineteenth century, and considerably higher than the quality of reported ages in census data from any country in the Middle East and Africa in the nineteenth and early twentieth century.102 This evidence is, however, biased in several ways. First, those several hundred surviving documents constitute only a tiny fraction of the original number of about 18 million documents that must have been produced over those more than 250 years.103 Furthermore, following the general pattern of our papyrological documentation, most of these documents originate from the second century ce, and most come from Middle Egypt, where papyri had a better chance of surviving in the dry desert sands. That means that we have almost no evidence for the large part of the population that lived in the Nile Delta and in Alexandria. As a result of unequal preservation almost half of the returns come from the big cities, and the metropoleis Arsinoe and Oxyrhynchus are predominant. The countryside is under-represented, with returns from villages being much less well preserved. The returns that we have from a rural environment are also unevenly distributed. 99 101 102

103

P.Lond. 3.904 (p. 124, col. 2). 100 Lact. De mort. pers. 23.1–4. Bagnall and Frier 2006: 50. Crayen and Baten 2010. Cf. Scheidel 1996b: 53–91; Bagnall et al. 1997: 91; Bagnall and Frier 2006: 44–7. For the calculation of an approximate number: Tacoma 2006: 168 n. 19.

26

Intergenerational solidarity, family support networks

Nearly two-thirds of our evidence comes from the Arsinoite nome in the Fayum, a fertile region of agricultural towns and villages where in Roman times most of the inhabitants were independent small-scale farmers.104 We have one separate census register from one city in Upper Egypt, probably Lykopolis, already referred to above.105 When discussing individual census returns in this study, I refer to the numbers in the catalogue that Bagnall and Frier provided as an appendix to their study on the demography of Roman Egypt.106 As in modern censuses, in Roman Egypt the household was the unit for the declaration.107 The usual term given in these documents to define the registration unit is oikia.108 Every fourteen years the head of each household, normally the eldest male, had to file such a census return to facilitate future taxation. He had to declare his property and the persons who lived in it, that is, not only his immediate family, but also lodgers, freedmen, and slaves.109 The returns record the number of members of a household and give their name, age, and relationship to the other members of the household. Status is given only for declarations from the major cities and is missing for villages. The same applies to occupations. Living off land revenues was not considered as an occupation in antiquity, and words such as ἄτεχνος (without a trade, not trained to any craft or profession), ἰδιώτης (unskilled, lay person) and ἀργός (not working the ground, living without labor), which we find in the census declarations for some adult males, seem to have been the terms for it.110 Most often only names, ages, and family relations are given, the most important information for our analysis. Unless the papyri are heavily damaged, “the kinship between household members can virtually always be reconstructed with considerable confidence.”111 Most of the women and men who registered in the census lists for Roman Egypt were of relatively humble origin and belonged to the lower and middle social strata of their community. We have many artisans, small traders, and peasant farmers. The rest were landowners who were able to live 104 106

107

108 109 110

Cf. Gazda 1983; Grenfell et al. 1990; Tacoma 2006: 168. 105 Bagnall et al. 1997. Bagnall and Frier 2006: 179–325. To give an example of how to read those references: 117-Ar-6 tells us that this return belongs to the census of 117 ce and comes from the Arsinoite nome (in this case from the village Philadelphia). The number at the end is an arbitrary serial number assigned by Bagnall and Frier to every census return. 187-Hm-1 belongs to the census of 187 ce and comes from the Hermopolite nome (in this case from its metropolis, Hermopolis). The reference ???-Ar-3 refers to a census return from the Arsinoite nome whose date is, however, lost. Bagnall and Frier 2006: 13. For a discussion of the Roman Egyptian census returns for household studies, see also Alston 2005: 129–31. Cf. Bagnall and Frier 2006: 12–14. Bagnall and Frier 2006: 22–5. For the fourth century, see Bagnall 1993: 181–207. Bagnall and Frier 2006: 24. 111 Bagnall and Frier 2006: 53.

The Roman Egyptian census

27

off the revenue from their land, and only a few professionals. The number of town councilors recorded in these returns is very small, also because town councils were only created under the rule of Septimius Severus at the start of the third century ce.112 Some 11 percent of all persons registered in the census lists were slaves who belonged to less than one-sixth of all households. Slaves were a symbol of status and the fact that five-sixths of all households registered without slaves seems to indicate that they were of modest means.113 As we have seen, villages are under-represented. Returns from the metropoleis account for roughly 50 percent of all our data, though their population constituted probably only one-third to one-half of those of the surrounding villages.114 Moreover, the limitations of the census returns, as has often been noted for other regions and periods, are that they rarely state what criteria the people who drew up these documents used to define a household. The possibilities are sleeping under the same roof, residing on the same farm, sharing meals, or participating in the same economic activities. By studying the household within a specific society we thus need other sources to establish residence and production patterns, and put the information we gain from the census returns into context.115 I will discuss the definition of a household and the history of household studies more in detail in the following chapter.116 I will thus also draw on other documentary evidence, especially private letters which survive among the papyri and bring to light ancient family life. For practical purposes and since our source material is severely limited, I assume when analyzing the census returns from Roman Egypt that each return represents one household, defined as a residential unit, with members sleeping under the same roof, preparing meals together, and sharing common living and production space, even if some of the members of the family might have spent a considerable time of the year away from home traveling for business matters, seasonal work, or visiting relatives.117 We also know little about how the Roman returns were filed. We can probably assume that teams of officials visited the villages for a certain period of time and filled out the form in conjunction with each head of the household, but we do not know if other household

112 113

114 117

Cf. Tacoma 2006: 170. Bagnall and Frier 2006: 48–9. For Ptolemaic Egypt, see Thompson 2002: 152–3; Clarysse and Thompson 2006: ii.262–7. Bagnall and Frier 2006: 49. 115 Cf. Goody 1972. 116 See also Huebner 2010. Cf. for the problems of defining a household: Goody 1971, 1972; Laslett 1972: 23–8; Hammel and Laslett 1974.

28

Intergenerational solidarity, family support networks

members were also required to be present – probably not, because occasionally absent sons, married daughters, or runaway slaves were listed as well. In the towns the heads of each household probably had to address themselves to the local census bureau within a certain period of time after the edict had been issued and enrol their household members for the census. Finally, we cannot stress often enough that family and household are constantly changing units: their respective compositions depend on the individual’s point of progress through the life course, marriage, divorce, and remarriage, all of which could change its composition, as well as births, deaths, and the dissolution of formerly joined households. The census returns catch only one specific moment in a household’s life cycle. They are unable to record a historic progress through time except in those few cases in which we have several consecutive returns for the same household. However, here we are just given snapshots of the households’ composition at a specific point at an interval of fourteen years. This is counterproductive for the study of family dynamics and family as a social organism, for which we again need to fall back on additional sources such as private letters, petitions, and contracts. In conclusion because we do not have the same quality and quantity of evidence for other regions of the ancient world as we have for Roman Egypt, there is simply no way for making any conclusive assumptions about family and household structures in the ancient Mediterranean in general.

1.6 Overview of the chapters The present chapter having laid out the aim of this study, the second chapter delves into detail and gives an overview of the history and methodology of household studies, and tries to locate the Roman Egyptian household pattern within this system. I will discuss the physical structures of a typical household, the average number of residents, household life cycles under the demographic regime of high fertility and mortality, differences between rural and urban regions, and inheritance patterns that affect and were shaped by prevalent household structures. This chapter sets the scene for later chapters which will refer back to it. Chapter 3 looks at intergenerational reciprocity on the household level. While economic support for children and the elderly are two separate systems in most modern societies, and intergenerational transfers from adult to elderly generations occur through taxation and social expenditure, old-age support in antiquity was based on direct reciprocity. By contrast to most parents in the modern

Overview of the chapters

29

West, elderly parents in Roman Egypt expected to be cared for by their children in direct reciprocity for raising them and providing them with an education or dowry in earlier years. This chapter investigates normative obligations parents had toward their young children and children toward their elderly parents, authority struggles between the generations living in one household, gender differences between sons and daughters and mothers and fathers, and strategies such as retirement contracts that elderly parents employed to secure old-age support. Chapter 4 tackles an important subject in the sociology of the Roman Egyptian family, thinking through the consequences of widowhood within concrete household formations. The chaste widow who remained faithful to her late husband was the ideal in Graeco-Roman society – at least among the upper classes. The census returns from the first to third century seem to confirm that also in Roman Egypt remarriage for widows was in fact quite rare: the recorded instances in the census returns indicate that most widows, especially those over 30 years of age, apparently did not remarry but raised their children alone and managed their husbands’ property single-handedly until the children came of age. I argue, however, that we should not take the information in the census returns at face value, since these documents only record a woman’s previous marriage if it met two conditions: first, her previous union resulted in children; and second, these children lived with the mother and her new husband, a very unlikely living situation given the prevalent postmarital residence patterns in Roman Egypt.118 Drawing on cross-cultural studies on various preindustrial patriarchal societies, I suggest that the likelihood of remarriage for widows is directly linked to household composition upon the husband’s death. Living arrangements and household forms were closely connected to the quality and quantity of the care and support for the aged. To examine the living conditions of the elderly in Roman Egypt and the options that existed to provide for the elder family members, an analysis of the dominant forms of succession patterns and household structures is relevant. Urban and rural areas in early modern Italy, for instance, show considerable differences as regards inheritance patterns and the organization of old-age support, as, e.g., Kertzer and Hogan have emphasized.119 In Chapter 5 I want to examine if the same can be said for Graeco-Roman Egypt. Inheritance patterns, often viewed as the “engine that drives household composition itself,”120 were decisive for residence patterns and thus for the decision regarding which 118 120

Cf. Huebner 2009a. 119 Kertzer and Hogan 1995: 120–39; Brandes 1996: 16. Kertzer and Hogan 1995: 120.

30

Intergenerational solidarity, family support networks

child was chosen as a primary caregiver in old age, and both inheritance and residence patterns thus have to be studied in unison. Property has always been the most powerful leverage in achieving support in old age. Chapter 6 addresses for the first time the crucial role of the daughter-in-law in a patriarchal household. Under the system of patri-virilocal marriage patterns, a young daughter-in-law could face a dire fate during her first years of marriage. She had little influence in her husband’s home, and was subdued and potentially mistreated. Classical sources tell us that a young bride in her husband’s household was expected to act in a demure and shy manner, and endure the hardships and constant criticism by her mother-in-law to work harder without complaint. It was the task of the mother-in-law to familiarize her new daughter-in-law with her new duties and responsibilities and supervise her work. Consanguineous unions, the marriage to a cousin or uncle, softened tensions that arose from this constellation as they accorded the daughter-in-law a higher value within her husband’s family and a greater autonomy in household decision-making. And in fact consanguineous marriages were and are very common in societies where the multiple family household is the ideal. In consanguineous marriages the mother-in-law was, instead of a stranger, an aunt or otherwise closely related woman whom the young bride had known since childhood. The retention of family wealth and the dowry within the family and the strengthening of kinship solidarity were further motives that led parents to look for a groom for their daughters among their close relatives. As we will see, also in the Hellenistic world in general and in Roman Egypt more specifically we find a strong tendency toward this form of unions. When assessing the ancient family support system, literary accounts provide us with its major features, but often not with its historical reality. The ancients indubitably relied on their children in old age, but we should not underestimate the impact of the demographic regime. Although coresiding sons and their families were the most important factor in old-age care, this family model constituted only the ideal that could often not be achieved due to high mortality rates and the migration of adult children. Chapter 7 will be especially informed by the life-course perspective that has proven to be successful for studying aging in other periods of history. The life-course approach traces how historical circumstances and individual events challenged, converted, improved, and transformed over the life course an individual’s provision for old age. We will see that the support system of the elderly was generally shaped by relationships formed early in life, and that often from young adulthood strategies were pursued to secure support in one’s later years.

chapter 2

Household structures, marriage patterns, and inheritance strategies

The peasants’ houses were all alike, consisting of only one room that served as a kitchen, bedroom, and usually as quarters for the barnyard animals as well . . . The only light was that from the door. The room was almost entirely filled by an enormous bed, much larger than an ordinary double bed; in it slept the whole family, father, mother, and Carlo Levi, Christ Stopped at Eboli. London, 1953: 120. children.

2.1 Regional diversity of historic household forms If we want to learn more about the inner dynamics of the family, intergenerational relations, and the strategies and decisions family members took to achieve certain goals or adapt to new circumstances, it is indispensable to begin with a study of living arrangements. Decisions about the marriage of a daughter, a new family enterprise, the purchase of more land, sending a son away for an apprenticeship, and provision for old age and death did not affect only one individual, but all the family members who lived together. The timing of those decisions was the response to the opportunities or needs that arose from certain household constellations. On the other hand, certain cultural patterns of predominant living forms influenced decision-making to achieve the household form that was considered the ideal for traditional and economical reasons. Household composition also affected the way an individual or family coped with situations of crisis, the death of a spouse, divorce, orphanhood, or childless old age. Before the beginning of serious studies on historical household composition, intrinsically connected with the names of Peter Laslett, John Hajnal, and Richard Wall, it was a widely held assumption that the simple family household was the result of progress and modernization brought forth by the Industrial Revolution. The French sociologist Frédéric Le Play termed the nuclear family the “unstable family,” since children left the household upon marriage and nothing remained once the elder parent generation 31

32

Household structures, marriage patterns, inheritance

died.1 The nuclear household was interpreted as an evolutionary step away from the multiple family household of previous times that comprised a descent group of several generations living together under one roof. These evolutionary presumptions were connected with the idea of a general decline of modern society away from the golden age of the patriarchal multiple family household and toward a more unstable, individualized elementary family. The smaller modern unit was also perceived as more apt for the requirements of industrial urban society. More recent studies over the last four decades have shown that in societies within north-western Europe (such as England, Scandinavia, the Netherlands, Germany, and northern France) nuclear family structures indeed predominated as far back as the thirteenth century. Marriage was typically late for men and women, and couples set up their independent household upon marriage. In southern and eastern Europe, however, a different family system prevailed over the centuries. Households were larger in size and the multiple family household was not at all uncommon. Sons brought their wives into their parents’ household, and marriage was early, especially for women. In places such as central and northern Italy,2 southern France,3 Russia,4 the Balkans,5 and the Eastern Mediterranean region, such as Turkey,6 Syria,7 and Egypt,8 the multiple family household was a common experience among the peasant population until very recently. Yet, no region was homogeneous. Various ecological, socio-economic, and demographic conditions and factors, such as urban or rural environments, agricultural forms, inheritance patterns, and social and demographic structures, influenced the dominance of specific household compositions within one locality.9 For Italy in particular, regional differences have been well studied. While central and northern Italy was characterized by complex family forms and multi-generational households, the southern regions of Italy exhibit far more nuclear family structures. Moreover, while postmarital residence in the northern and central regions was customarily patrilocal, young couples in the south 1

2

3

4 8 9

Le Play 1871: 11. For further references, see Burgess 1916; Ogburn 1933; Davis 1941; Parsons 1944; Goode 1963; Mitterauer and Sieder 1982: 11–39; Sieder 1987: 12–13. For a detailed discussion of the history of household studies of the Graeco-Roman world, see Huebner 2010. Klapisch 1972; Levi 1973; Barbagli 1984, 1991; Kertzer 1984, 2001: 44–5; Herlihy and Klapisch-Zuber 1985; Kertzer and Hogan 1989; Kertzer and Saller 1991. e.g., for southern France: Le Roy Ladurie 1966; Biraben 1972; for Corsica: Dupâquier and Jadin 1972. Czap 1978, 1982. 5 Kaser 1994a, 1994b. 6 McCarthy 1979. 7 Meriwether 1999. Cuno 2005. Hajnal 1983; Segalen 1986; Yanagisako and Collier 1987; Cheal 1989; Barbagli 1991; Hareven 1991; Kabeer 1991; Fraad et al. 1994; Kaser 1996; Mitterauer 1996; Reher 1998; McKie et al. 1999; Han 2004; Komter 2005; Stone 2005.

Definition of the household

33

usually set up their independent household. These differences between the north and south are usually explained by different forms of landholding. In the south the peasants resided in towns, while in the central and northern parts of Italy they lived rather in hamlets on their land. Everywhere, not just in Italy, urban structures seem to have been more conducive to nuclear structures of families; in rural settings we find a higher degree of multiple and extended families.10

2.2 Definition of the household The standard definition of a household as a residential unit is used here, since studies on the early European family have shown that households were not necessarily a unit of production that shared its means of production such as livestock, farming equipment, and land, and did not inevitably hold all its property jointly.11 Also those family members that spent a considerable time of the year away from the household for work or traveling are considered as members of the household if this seems to have been their main residence. The respective type of household is based on the number of kin-related conjugal family units in one household.12 Conjugal units consisted of the three basic elements: father/husband, mother/wife, and children, or only two of these three.13 There are five main categories: (1) solitary households (a household in which an individual is living alone); (2) households with multiple unmarried persons living together (mostly co-resident siblings); (3) simple family households, also called nuclear households (normally a couple with their children, but also childless married couples and widowed spouses who lived with their unmarried children are included); (4) simple family households extended by co-resident kin (an elderly parent, an unmarried nephew or sister, extending a nuclear household either upwards or laterally); and finally (5) multiple family households in which more than one married couple lived together. I differentiate here between vertically and horizontally extended multiple family households. Vertical households are those in which two or more married couples belong to different generations, 10

11

12 13

See Hajnal 1983; Hareven 1991; Barbagli 1991; Kaser 1996; Mitterauer 1996; Reher 1998. See now also Huebner and Nathan in press. For the problems of defining a household, see Goody 1971, 1972; Laslett 1972: 23–8; Hammel and Laslett 1974. Sometimes also separate simple family households worked on a piece of land jointly (Hajnal 1983: 67; cf. Cuno 1995: 489). Laslett and Wall 1972; Hammel and Laslett 1974. Skinner 1997: 56: “Thus, a childless couple, a mother and her children, and a father and his children all count as CUs [conjugal units].”

34

Household structures, marriage patterns, inheritance

usually parents and their married son(s). Horizontal households were built by married siblings living together with their families. This form is often referred to by its French name, frérèche (fraternity), denoting a household consisting of at least two brothers with their wives and children. Conjugal couples in these multiple family households were usually related through the male line. In this study I do not take into consideration lodgers or slaves but consider exclusively the co-resident kin group related by blood, marriage, or adoption.

2.3 Household forms in antiquity Interestingly, we find Le Play’s nineteenth-century view of the nuclear family household as an evolutionary step away from the multiple family household also in antiquity. Plutarch, the Greek biographer of the first and second centuries ce, laments on several occasions the dwindling of the joint generational household in Roman society. According to Plutarch, joint brother households were more common in the past than during his time.14 Plutarch tells us about the upbringing of Marcus Licinius Crassus (c.115–53 bce), who, despite the family’s wealth, grew up with his parents and brothers under quite humble circumstances. Upon their marriages, his two older brothers did not leave home but brought their wives into their parents’ house. All of them continued to live under the same roof even though the house was small and very modest.15 Plutarch regarded this circumstance as an explanation as to why Crassus, despite his later wealth, was temperate and moderate throughout his life.16 For Plutarch it was a sign of poverty in his day for brothers to live together on an undivided estate. Men of the Roman elite preferred to set up their own household upon marriage – mainly, as Plutarch tells us, to avoid conflict.17 As an example of this archaic form of living, Plutarch presents elsewhere the family of the Aelii Tuberones of the second century bce. Sixteen male members of this family were living with their wives and numerous children under one roof and farmed their land jointly. According to Plutarch, this family’s lifestyle was the result of their impoverished economic situation, which made the sons unable to afford to establish their own households when they married.18 Plutarch sees this movement toward an elementary family of 14 15 16 18

De fraterno amore 478.1. When one of his brothers died, Marcus Crassus married the latter’s widow and had children with her. Plut. Crass. 1. 17 Plut. Aem. P. 5.5; Val. Max 4.4.9. Cf. Garnsey and Saller 1987: 129. “For there were sixteen near relations, all of them of the family of the Aelii, possessed of but one farm, which sufficed them all, whilst one small house, or rather cottage, contained them, their numerous offspring, and their wives; amongst whom was the daughter of our Aemilius, who, although her father

Household structures from inscriptions

35

contemporary Roman society as an evolutionary yet deplorable step. However, how should we understand these anecdotes? Should we attribute the differences between living circumstances of Roman senators of the second century bce and those of the second century ce to a general change in household formation patterns in Roman Italy, or, rather, ascribe it to the fact that while Roman senators of the second century bce had been poor and humble farmers who made Rome strong, they now lived in luxury? And what should we assume about the household patterns among the lower social strata in Plutarch’s time, the peasant farmers in the countryside that formed the vast majority of the population? Plutarch alludes only in passing to the household composition in late republican Rome, from which we can scarcely draw any conclusions about general household patterns in the Italian countryside.

2.4 Household structures from inscriptions While the literary evidence seems unrewarding for any conclusions about republican and imperial Roman household structure outside the small elite, many scholars have questioned the legal evidence as a reliable reflection of everyday life. Roman law indubitably stresses the extended or multiple family, the old patriarch heading a family of several married sons, but, as many have shown, Roman law outlined principles and standards “push(ing) things to the limits of logic,” and should not be seen as reflecting social reality.19 Epigraphic sources, especially the many funerary inscriptions we have from the city of Rome and the Italian peninsula, have enjoyed a higher regard. Saller and Shaw’s large-scale study of tombstones from the Roman West has been praised as a milestone in our understanding of the ancient family. Saller and Shaw questioned whether “the linguistic and legal material alone might lead us to downgrade the significance of the nuclear family.” Their analysis of the epigraphic evidence, in their opinion, corrected this view: “On the basis of the tombstone inscriptions we have come to the conclusion that for the populations putting up tombstones throughout the Western provinces the nuclear family was the primary focus of

19

had been twice consul, and had twice triumphed, was not ashamed of her husband’s poverty, but proud of his virtue that kept him poor. Far otherwise it is with the brothers and relations of this age, who, unless whole tracts of land, or at least walls and rivers, part their inheritances, and keep them at a distance, never cease from mutual quarrels” (Plut. Aem. P. 5, trans. J. Dryden; cf. De fraterno amore 478.1). See also Dixon 1992: 7, 232. See, e.g., Crook 1967: 114 on patria potestas in Roman law; Gardner 1986; Dixon 1992: 3–5.

36

Household structures, marriage patterns, inheritance

certain types of familial obligation.”20 Even though Saller and Shaw were cautious to merge these two angles, they did not refrain from a general hypothesis about the longue durée of Roman household forms: Modern historians have shown that in most areas of Western Europe the nuclear family was the main type of familial organization as far back as dependable records are available. On the basis of our evidence, it seems a reasonable hypothesis that the continuity of the nuclear family goes back much further in time and that it was characteristic of many regions of Western Europe as early as the Roman Empire.21

The view that the Roman West was dominated by nuclear family structures has been accepted by virtually every publication on the western Roman family since Saller and Shaw’s article appeared in the 1980s. Bradley states: “the emphasis historians have placed on the predominance in Roman society of the nuclear family [is such] that it is now almost unorthodox to think of the Roman family in any other way.”22 Parkin and Pomeroy go even further: “It has become increasingly clear over recent decades that the nuclear family structure was the norm among Roman citizens in the classical period, at least in the western half of the empire.”23 I have argued in my contribution to Beryl Rawson’s Companion to Families in the Greek and Roman Worlds that the trust placed upon the epigraphic evidence might not be warranted.24 I was certainly not the first who questioned Saller and Shaw’s conclusions,25 but I was the first to compare the inscriptions from Graeco-Roman Egypt that had never been considered before with our knowledge about household forms from the Roman Egyptian census returns. And in fact results from the census returns and from the epigraphic evidence differ considerably regarding their diagnostic value for any assumptions about family and household composition. While we would conclude from the epigraphic evidence that Roman Egypt displayed family and household structures similar to those in the Roman West, with a strong dominance of the nuclear family, we know from our census returns that the majority of the people actually lived in multiple family households consisting of several married couples and their offspring. If this is true for Egypt, we might wonder whether funerary inscriptions in the Roman West might not be as strongly governed by commemorative patterns and provide little information about actual household composition. It must be stressed again that where Roman Egypt seems to be distinct 20 22 23

Saller and Shaw 1984: 124. 21 Saller and Shaw 1984: 145–6. Bradley 1991a: 6. Cf. Dixon 1992: 3–5; Severy 2003: 8; Hasegawa 2005. Parkin and Pomeroy 2007: 74. 24 See Huebner 2010. 25 Martin 1996; Scheidel 2012b.

Household structures from inscriptions

37

from other regions of the Roman Empire and the ancient Mediterranean is in the kind and degree of its documentation, not necessarily in its cultural and social practice.26 Saller and Shaw were perhaps too quick to draw conclusions about the pervasiveness of the nuclear family, working as they were under the influence of Laslett’s hypothesis, which held that the simple family household was the common experience in Western Europe long before the Industrial Revolution. Therefore they saw their results as corroborating these findings and extending the assumption of the pervasiveness of the nuclear family in European history many centuries back, to ancient Roman times. However, as we have seen above, Saller and Shaw’s underlying hypothesis has in the meantime been proven wrong. Further studies on early modern household forms have brought to our attention considerable regional differences within Europe.27 Rural areas of central and northern Italy, for instance, were characterized by a multiple family system as far back as the late medieval period: for the regions of Lombardy, Piedmont, Liguria, and Emilia-Romagna (c.1550–1750), the combined percentages of extended and multiple family households were between 27 and 42 percent.28 Households of peasant farmers in fifteenth-century Tuscany looked similar to those that Plutarch described for the families of the Aelii Tuberones and the Crassi in the second and first centuries bce.29 To date we do not have a survey of funerary inscriptions from Greece or the Greek islands that could provide us with comparative material. Cox has studied household dynamics for classical Athens mostly on the basis of court speeches. She discusses marriage patterns such as endogamy vs exogamy, the role of slaves in the domestic economy, and inheritance strategies, but does not elaborate on actual household compositions, e.g., by employing the models provided by the Cambridge Group. Her observations seem to imply, however, that the stem family household was the ideal, with one of the sons staying at home. Joint brother households seem to have been rare in classical Athens; brothers seem to have set up their individual households after their father’s death.30 This corresponds with findings of studies on later periods of mainland Greece and on the Greek islands: stem families with just one child, most often a son, remaining at the parents’ home, and being responsible for their maintenance in old age, were the norm.31 26

27 29 31

For this tendency, see, e.g., Lewis 1984: 1077–84; Bagnall 1995: 11–13; Cribiore 2001: 6; Bowman 2005: 313–14; Hanson 2005: 85. Wall et al. 1983; cf. also Hammel and Laslett 1974: 75–8. 28 Viazzo and Albera 1990: 467. Herlihy and Klapisch-Zuber 1978. 30 Cox 1998. Cf., e.g., Vervier 1984; Sant Cassia and Bada 1992; Caftanzoglou 1994, 1997; Hionidou 1995.

38

Household structures, marriage patterns, inheritance

2.5 Household structures from census returns This general deficiency in our knowledge about actual household composition for the ancient world is, however, more than offset by our source material for Ptolemaic and above all Roman Egypt, the exceptional documentation in the form of tax lists and census returns having survived in the dry desert sands of Middle and Upper Egypt. The various Ptolemaic administrative registers in Greek and Demotic from the third and second centuries bce studied by Willy Clarysse and Dorothy Thompson record only Ptolemaic Egypt’s adult inhabitants (not their ages), the poll tax they had to pay, their livestock, and their occupation. For this reason their value for establishing family and household sizes is limited.32 This means that for Ptolemaic Egypt we can ascertain the number of adults per family or household, but we do not know the actual family size. Households with two adults, most often a married couple, were the most common type (39.2%). Another 19.5% of all households contained three adults, 11.7% had four adults, and 17.8% of all households had only one adult. When classifying this data, Clarysse and Thompson tried to distinguish between the ethnic affinities of the family members. They came to the conclusion that among Greeks a household with only one adult was much more common (30.2%) than among Egyptian families (10.5%), while two-adult families, most of them simple family households, were more common among Egyptians (46.5%) than among Greeks (27.2%). From the available data Clarysse and Thompson established that most Egyptian households were simple, based on a conjugal couple, often extended by an old mother or father but no non-kin dependents. Households of the Greek settlers, on the other hand, tended to be larger and to include co-resident siblings and various non-kin dependents. For other family types in Ptolemaic Egypt, figures do not differ much between Greeks and Egyptians. The high number of apparently solitary households among the Greeks consisted mainly of single women, most likely widowed and living together with their minor children, who were not registered because of being underage. Clarysse and Thompson calculated on this basis a total average number of family members of 4.2 (4.0 for Egyptian families and a slightly higher figure of 4.4 for Greek families).33 The census records from Roman Egypt studied by Roger Bagnall and Bruce Frier, however, allow for a more complex analysis of the structure of the household. The average family size in Roman Egypt was about 4.3 32

Clarysse and Thompson 2006.

33

Clarysse and Thompson 2006: ii.237–49.

Physical layout of the house in Roman Egypt

39

people, virtually the same number that Clarysse and Thompson established for Ptolemaic Egypt. Households that included slaves and lodgers had an average of five members.34 While multiple family households had as many as 8 to 9 members, simple family households consisted of 3.3 to 3.7 people, and extended family households of about 4 to 5 people. The extended or multiple families that also accommodated slaves and lodgers had an average size of about 10 people. Bagnall and Frier arrived at an average household size of 5.3 members for metropolitan households and 4.8 members for households in the countryside.35 This difference in size is mainly due to the fact that households in the cities more often owned slaves. In general, household size was surprisingly small in Roman Egypt, but in line with other countries where the mortality and birth rates were both high.36 As a point of comparison, the average US American family size in 2000 was 2.6 members; households in Roman Egypt were almost twice that size.

2.6 Physical layout of the house in Roman Egypt Before we delve deeper into this topic and the data from the census returns, let us summarize what we know about the physical layout of the house in Roman Egypt. Our most fruitful sources here are the archeological remains and the documentary evidence. The documentary evidence, such as papyri concerning property transfers, wills, leasing or sales contracts, declarations of property, and litigation about houses or parts of them, provides us with information about prices and dimensions of houses. Records of burglaries tell us something about the accessibility of the houses. Archeological remains, on the other hand, inform us about the layout and appearance of domestic living space. In systematic examinations of the excavated household contexts in different regions of Egypt, such as the Nile Valley, the Fayum, Alexandria, and the oasis towns and desert settlements, archeologists have noted considerable differences between cities and villages, and the variety of house types in general. Most of the archeological evidence for Roman housing comes from villages in the Fayum, such as Karanis, Narmouthis, Tebtunis, and Soknopaiou Nesos, and, to a lesser degree, from Kellis and Trimithis, villages in the Dakleh Oasis.37 34 36

37

Bagnall and Frier 2006: 67–8. 35 Bagnall and Frier 2006: 67–9. In late medieval Tuscany, for instance, the average household size was 4.4 (Herlihy and KlapischZuber 1985: 283). Husselman 1979 for Karanis; Rathbone 1990: 124–37 and Gallazzi 1994 for Tebtunis; Boak 1935 for Soknopaiou Nesos; Davoli 1998 for the Fayum in general; Boozer 2007 and 2010 for Trimithis; Hope and Whitehouse 2006 for Kellis. Cf. also Gallazzi and Hadju-Minaglou 1989; Alston 2002: 50, 52; Bagnall and Rathbone 2004.

40

Household structures, marriage patterns, inheritance

A typical home in the Fayum, where most of our census material comes from, was a courtyard house, a housing type which has been built for millennia all over the Mediterranean and the Middle East.38 The houses were mainly built of economical mud brick interspersed with wooden beams; stone was used for exterior stairs and sometimes the outer walls.39 Houses consisted usually of two or three floors. In addition, many houses had basements where food was stored, and access to a roof terrace.40 The main entrance was from the street, and the wooden doors could be locked with keys or bolts.41 The price one had to pay for a building plot was about half as much as the price for a house. We find a great variety in house sizes: floor space could range from 30 m2 to 200 m2, the mean size being about 75 m2. A house could cost several hundred to several thousand drachmas, depending on its rural or urban location, size, age, building materials, and state of preservation. In general, houses in villages cost much less than in the metropoleis, on average about half the price, which might be due to the more expensive building materials used in urban architecture (such as stone as opposed to mud brick), higher cost of land, and more multistory constructions.42 Not only were entire houses bequeathed or sold, but quite often shares of the houses were traded. Individuals often owned ½ or ⅔ of a house, one time even 1 ⁄42 of a house, which was rather a theoretical subdivision, a “share of interest”, that did not find its reflection in the actual occupation of the house.43 In general, we do not know, however, how these houses were used by their owners, and whether there were different parts of the house reserved for the men and women of the family. Most of the houses were, in any case, too small, consisting only of one or two rooms, to allow for any differentiation between the sexes and age groups. It becomes clear that privacy was limited if several couples lived together under one roof. Houses in Karanis, for instance, show that rooms in many houses were organized linearly, which meant there was no central room or hallway to reach all of the other rooms, but one had to pass through all the intervening rooms in order to reach the room at the back of the house. Houses in the Dakleh Oasis, on the other hand, were clustered around a central room.44 Comparative material suggests that in the latter case the central room was used by all members of the household for daily activities, and that each 38 42 43

44

Barnow 2001. 39 Husselman 1979: 33–6. 40 Alston 2002: 53, 59. 41 Alston 2002: 53. Maehler 1983: 120. For private houses in Roman Egypt as mentioned in the papyri, see Luckhard 1914; Maehler 1983; Drexhage 1991: 74–89; Daniel 2010. Especially for a discussion of fractions of houses, see Muhs 2008. Boozer 2010: 119.

Physical layout of the house in Roman Egypt

41

couple occupied one of the bedrooms with their children.45 The layout of the house, with an often windowless façade and a locked front door toward the street as the only passage to the outside world, seems to indicate that priority was given to privacy within the household, but certainly there were also practical reasons for it, such as guarding the interior from the wind and sand.46 Excavations in the villages of the Fayum show that individual houses were often modified by their owners over the course of several decades or even centuries, to restore parts that had fallen into decay or to adjust the living space to the inhabitants’ needs.47 The courtyard was the center of activity and socialization for the immediate family and the members of the neighboring houses. Ovens made of plastered mud and built into the wall of the courtyard, grain bins, millstones, and mortars were located there, all used by those making bread for the household. All the other food processing and cooking is believed to have taken place in the open courtyard. In addition to bread baking and cooking, clothes were made there as well. Various instruments for spinning, weaving, and sewing have been found, as well as tools for making wooden domestic utensils and simple furnishings.48 Domestic animals were also kept in the courtyard. Since the courtyard was often shared with other houses of the insula, it was accessible to more than the immediate family. In a petition from second-century Oxyrhynchus we learn that upon leaving the house, inhabitants used to lock the front door as well as the door to the inner terrace as we might today, fearful that even neighbors next door would take the opportunity to pilfer something. In fact, when a certain Diemous from second-century Oxyrhynchus returned home and found her back door lifted and her jewelry stolen, she suspected that her neighbor the weaver Heras, or one of his employees, had taken it, as they were the ones who had easiest access to her courtyard.49 This gives us an interesting insight into the concept of the private sphere, relations with neighbors, and living in an apartment block. In the census returns courtyards are sometimes listed as separate property; they did not belong necessarily to the owner of a house.50 45

46 49 50

See, e.g., Eldem 1968 and 1984 for nineteenth- and twentieth-century Turkish multiple family households and their use of domestic space. Cf. Sartre 2005: 226. 47 Maehler 1983: 128. 48 Bellah 1980: 11–13. P.Oxy. 10.1272 from 144 ce. e.g., a census declaration from Karanis dating from ce 189 (187-Ar-27) states: “There belong to the persons whose interests I represent, to Tasoucharion, whose father is unknown, her mother being Sarapias, an Antinoite, the mother of the persons named below, a house and courtyard and a third share of another house in the village, and to Gaia Apolinaria and Gemellus Horion, her children, Antinoites, in common and equally, a house and two courtyards formerly the property of Valeria

42

Household structures, marriage patterns, inheritance

2.7 Demographic regime and household formation Let us look at the household structures of the census population in Roman Egypt. Bagnall and Frier point out that not an inconsiderable number of people lived alone in (according to Laslett’s scheme so-called) solitary households, but also that their numbers might be inflated simply because short returns are more likely to have come down to us complete. About 6 percent of the population recorded in the census returns lived alone, and the solitary households constituted about 16 percent of all household forms.51 Even if the proportion of solitary households seems to be quite high, solitary households were in general much less common than they are today. In the EU, North America, Japan, and Australia, about 25 percent of all households had only one member in 2000. In Sweden today the tendency to solitary households is even twice as high: almost one-half of all households contain only one member.52 Many of the solitaries in Roman Egypt were already of very advanced age. Unfortunately we cannot say anything about the type of household in which they lived before living alone. We can, however, imagine different scenarios that seem likely, such as a widowed mother whose children had married, died, or migrated for work reasons. The most popular type of household recorded in the Roman census returns was the simple family household, which accounted for 43.1 percent of the total; 35.5 percent of the recorded individuals lived in this form of household. Nuclear family households of a married couple and their unmarried children constituted the major share. Couples living in a simple family household were often of advanced age. They apparently had not established this nuclear household when they married but it was caused by attrition through the death of parents and brothers.53 Simple family households that were extended by co-resident kin constituted 15 percent of all households.54 Nine simple families in the census returns were extended upward by a spouse’s elderly parent. In six cases it was the elderly mother who lived with the couple and their children: in four cases the mother of the husband, in one case the mother of the wife, and in one case both mothers lived with the couple. In three cases the husband’s father lived with the younger couple. The example of a husband’s parents living with the young

51 54

Diodora and a third share of two houses and two courtyards, and elsewhere a half share of a house and courtyard and of another courtyard, and elsewhere a house and courtyard, and two courtyards formerly the property of Gaius Longinus Apolinarius, veteran, and a house and courtyard formerly the property of Ptolemais, and elsewhere a third share of a house and courtyard, which I declare for the house by house registration of the past 28th year.” (Transl. APIS.) Bagnall and Frier 2006: 60. 52 Therborn 2004: 312. 53 Bagnall and Frier 2006: 61. Bagnall and Frier 2006: 60–2.

Demographic regime and household formation

43

couple was much more common, undoubtedly because it was usually their home in which the whole family lived. We have thirteen examples for laterally extended families, mostly by unmarried brothers of the husband, sometimes by his unmarried or divorced sisters or cousins, and in two cases by a brother of the wife. While the simple family household, extended or not, constituted the majority of all households, the multiple family household was also a frequent configuration. Over one-fifth (21%) of all households contained more than one married couple, and over two-fifths (42.8%) of all people lived in multiple family households.55 We differentiate again between laterally extended and vertically extended types. Well over half of all multiple family households (56%) in the Roman census returns are the laterally extended type, usually consisting of an elderly couple and their married son(s), while 44% contain married couples from the same generation, a form also called frérèches, consisting of at least two married brothers and their families.56 The family of Miusis, 59, and Senpachompsais, 53, from the village of Tanyaithis, in the early second-century Apollonopolite nome, is an example of a vertically extended multiple family household.57 The two daughters of the elderly couple, 24 and 18 years of age, were not yet married. The son, 29 years old, however, was married and had brought his wife into his parents’ home. The young couple already had two children, a one-year-old boy and a just-born daughter. The household of Horos, 48, his brother Horion, 35, and their sister Tapekysis, 45, is an example of a laterally extended multiple family household, a frérèche. Horos was married to his sister Tapekysis, and they had two sons. Their brother Horion was married to Thenatymis, 39, an apparently unrelated woman with whom he had a little son.58 The fact that the multiple family household was so relatively common suggests that it constituted the ideal family form, for demographic probability worked strongly against achieving it.59 Roman Egypt’s demographic regime, which was prevalent in many societies before the demographic 55 58

59

Bagnall and Frier 2006: 60, 62–4. 56 Bagnall and Frier 2006: 62–3. 57 117-Ap-6. Another example of a two-generation multiple family household is a household of four brothers, aged 47, 45, 42, and 38. They were all married and all four couples had between one and three children. The eldest brother, a quarry worker, was the head of the family. It was he who declared the property (131Pr-1). We know from a later census declaration that 14 years later the second eldest of the brothers, his wife, and his two children had moved out and had established their own household (145-Pr-1). Cf. Hajnal 1983: 69: “The joint-household system did not normally produce a situation where the majority of households were joint at any one time, though there may have been joint-family household systems which have operated in that way. However, under a joint-household system, the majority of people were members of a joint household at some stage in their lives.”

44

Household structures, marriage patterns, inheritance

transition, was dominated by high mortality, high fertility, and low population growth.60 Given the high mortality rates, the overlap between generations was relatively small compared with that of modern societies. Most fathers did not live long enough to see their offspring marry and have children. In addition, two-fifths of all fathers did not have a son that reached marriage age, or only had daughters.61 A precondition for a joint brother household, a frérèche, however, was that a family had at least two or more sons who reached marriageable age, married, and decided to reside with their wives in their father’s household. Even in a society in which the multiple family household was the ideal, usually the majority of households were not joint, owing to demographic and socio-economic constraints.62 Not only biological fate hindered this family form. Work migration might have driven sons away from their paternal home to the cities. For these reasons we might infer that, in those cases in which the formation of a multiple family household was indeed possible in the development cycle of a household, the family actively pursued this form of residence and saw it as the ideal form of living.63 Roman Egypt exhibited a household pattern that we also find in, for example, ancient China, Ottoman Anatolia, and nineteenth-century Syria and Egypt (places where the proportion of multiple family households ranged between 20 and 40 percent).64 Late medieval central Italy offers slightly higher percentages (39 to 45 percent),65 while on the Mishino estate in early nineteenth-century Russia the percentage of multiple family households was as high as 82 percent.66 Roman Egypt, with its 21 percent of joint family households, thus ranged at the lower end, which could have several explanations. Low life expectancy, lower than in the nineteenth century, might have prevented the formation of too many multiple family households. On the other hand, we also have indications that a multiple family household was not kept up at all cost. As I will discuss in more detail further 60 61 62

63 64 65

McEvedy and Jones 1978; Anderson 1988; Wilson and Airey 1999. Cf. Saller 1994; Huebner and Ratzan 2009b. Cf. Levy 1965; Wheaton 1975: 614–15: “In traditional populations, infant mortality is high and the further life expectancy of those who reach the age of marriage is low. Many married couples, therefore, will produce no children who survive to maturity, one son only, or only daughters; in these cases the family cannot pass through a joint phase. When two or more sons survive to maturity, the life expectancy of the father may limit the length of time during which the family is joint if the household breaks up at the father’s death. In the Tuscan instance this joint phase was further reduced because men often postponed marriage until age twenty-five.” For a qualification of this thesis, see Kertzer 1989: 1–15. Cf. Wheaton 1975: 601–28; Kertzer 1989: 1–15; Douglass 1991: 292–3. McCarthy 1979: 314; Gerber 1980: 244; Duben 1985: 91; Gerber 1988; Cuno 1995; Liao 2001. Laslett 1977: 22–3. 66 Czap 1983: 128–9.

Demographic regime and household formation

45

down in this chapter, only couples from the immediate family, parents and their married sons, or married brothers stayed together in one household. We do not find couples living together under one roof where the males were related more distantly than brothers and fathers and sons, i.e., as uncles, nephews, or cousins.67 On the Mishino estate in nineteenth-century Russia just mentioned, where four-fifths of households were multiple family, we do find, however, many households of co-resident cousins. Early modern and modern societies of north-western Europe, on the other hand, show percentages of multiple family households between 0 and 4 percent, giving credence to the assumption that there was a completely different family and household formation system in play (see Table 2.1).68 When we compare the situation in Egypt in the early centuries under Roman rule with the household structure recorded for Ptolemaic times, we see that the Roman Egyptian census population quite closely resembles the earlier Egyptian rather than the earlier Greek population.69 The percentage of multiple family households is about the same for Egyptian families in Ptolemaic times (22.4%) and in the general population in Roman times (21.0%), but on a different order of magnitude from the Ptolemaic Greek one (12.2%). Among the Egyptian households for which we have data from the Ptolemaic period, 51% should probably be classified as simple, 12% as extended, and 22.4% as multiple family households.70 Among the Greek families in the Ptolemaic period, 39.1% can be classed as simple, 13.5% as extended, and 12.2% as multiple family households. As Clarysse and Thompson remark, the percentage of simple family households most probably would have been considerably higher if the data also recorded minors. Many of the apparent solitaries would then turn into a household consisting of a widow (or a widower) living with minor children, i.e., a simple family household.71 The rather low percentage for multiple family households among the Greek population might be attributed to cultural patterns: sons went off to form a new household upon marriage, brothers did not continue living together after their parents’ death, and the fact that the Greeks resided in the more urbanized areas. This correlates with our records for Roman Egypt where the percentage of multiple family households among metropolitan households was just 15.3%.72 67 69

70 71

Cf. Chapter 5. 68 Flandrin 1979: 71; Laslett 1977: 20–3. We have to stress once more that the Ptolemaic evidence records only adults, not minors, while the Roman and all later records presented in Table 2.1 take into account all family members (cf. Clarysse and Thompson 2006: ii.255). See Clarysse and Thompson 2006: ii.248–60. See Clarysse and Thompson 2006: ii.249–50. 72 See Bagnall and Frier 2006: 67, tab. 3.2.

Table 2.1. Household composition in various preindustrial societies (percent).a

Household type (1) Solitary (2) No family (3) Simple (4) Extended (5) Multiple

Ptolemaic Egypt: Egyptian familiesb 11.2 3.3 51 12 22.4

Ptolemaic Late Northern Egypt: medieval Northern Russia, 19th France, 17th c. Roman Tang England, 16th c. Greek Egypt, China, Tuscany Italy, 18th c. c. (Mishino, (Brueil en Vexin, (Ealing, familiesc 1st 3rd c.d 6th 8th c.e (1427)f (Piedmont)g 1625)i 1822)h Middlesex, 1599)j 31.4 3.8 39.1 13.5 12.2

16.2 4.8 43.1 15.0 21.0

12.4 5.1 46.4 14.5 21.4

13.6 2.3 54.8 10.6 18.7

All non classifiable households were excluded and percentages recalculated. See also Huebner 2010. b Clarysse and Thompson 2006: ii.255. c Clarysse and Thompson 2006: ii.255. d Bagnall and Frier 2006: 60. e Liao 2001: 341. f Herlihy and Klapisch Zuber 1978: 292. g Viazzo and Albera 1990: 467. h Czap 1983: 128 9. i Lions 1967: 521 37. j Laslett 1972: 85. a

0 3.7 57.4 16.7 22.2

0.7 0 6.7 10.7 82.2

7.3 1.5 83.8 7.3 0

12 2 78 6 2

Household life cycles

47

2.8 Household life cycles Here I need to stress once more that these different forms of family composition were not really alternatives but stages in an ideal model of household progression. They all might be experienced by a single family over the course of several decades. Different types of household forms – solitary, nuclear, extended, or joint – found in Roman Egypt should be seen as phases in a cycle. In those cases in which we can follow a household from one census to the next, we can observe major changes in household composition during this fourteen-year cycle, with households becoming either simpler or even more complex. Simple family households turned into fragmentary household forms as the consequence of the parental generation’s death, or into extended households or frérèches when one or more of a couple’s sons married and had children. Let us give an example for a household where we have two consecutive census returns: in the year 117 ce, a widowed 53-year-old woman lived together with her three adult children, one son, 32 years old, and two daughters, 28 and 33 years old. The late husband’s unmarried 70-year-old sister, the paternal aunt of the children, also lived with them.73 The younger of the two daughters, the 28-year-old Aphrodous, was married to her 32-year-old brother, Sokrates (see for this “brother–sister marriage” the discussion in section 7.12). Fourteen years later, in 131 ce, the same household was registered again, and the return has come down to us. The older generation had died, the aunt, the mother, and apparently also the elder of the two sisters, if she had not married and moved out.74 The brother–sister couple, now 46 and 42 years old, lived in the family house with their five children, three sons and two daughters, all of whom had been born since the last census fourteen years earlier. Here we witness a change from an upward extended family household to a nuclear one. If we had another census return for the same family fourteen years later, we most likely would see that the household had become complex again when the three sons of the couple had married and brought their wives into their parents’ house, and daughters had moved out to live with their parents-in-law. We will encounter more often over the following chapters the necessity of a dynamic approach to the structure of the individual family during its life cycle, and the subsequent sequences of types through which the respective household passes.

73

117-Ar-1.

74

131-Ar-3.

48

Household structures, marriage patterns, inheritance

2.9 Marriage patterns in Roman Egypt In the Roman Egyptian documents we discover a pattern of early and pervasive marriage for women, later marriage for men with an age gap between spouses of several years, and patrilocal residence. Males in Roman Egypt ideally spent their entire lives in the household they were born into, whereas daughters routinely left their parents’ house upon marriage and then joined their husbands’ families. Young couples virtually exclusively set up their households with the husband’s parents. Uxorilocal marriage, where the young couple resided with the wife’s family, is a common feature of many other societies that are otherwise dominated by virilocal marriage patterns when the wife’s parents were lacking a son.75 Confronted with the same unfortunate demographic situation, many parents who had only daughters took recourse to arranging a uxorilocal marriage for at least one of them in order to win a male for the household. We have, however, no evidence in our census returns from Roman Egypt that husbands moved into their wives’ paternal home. Cohabitation between married daughters and their elderly parents is documented only for those daughters who had married a “brother.” I have argued elsewhere that it would make more sense to read our Egyptian evidence on “brother–sister” marriage in light of this custom as the arrangement of an uxorilocal marriage, i.e., the adoption of a son-in-law in the absence of a biological son, rather than seeing Roman Egypt as the only instance in human history of widespread institutionalized incest, a point to which I will come back in Chapter 7.76 Marriage in Roman Egypt usually did not result in new household formation. The new couple was integrated into a pre-existing household structure and an ongoing domestic economy.77 While in early modern and modern north-western Europe marriage usually shaped the structure of the household, because the young couple set off to form their own household, in Roman Egypt marriage had hardly any effect on the household as an economic entity and consumption unit. It was the death of the patriarch (sometimes later his widow) which was the crucial factor for any realignment of residence patterns.78 Neolocality upon marriage seems to have been the exception in Roman Egypt. 75

76

77

The in-marrying son-in-law was sometimes explicitly adopted by his parents-in-law, and sometimes took over informally his wife’s family name (see, e.g., Shimizu 1991; Bowen 1996; Hammel and Gullickson 2004: 151). Cf. Huebner 2007. On adoption in the ancient Eastern Mediterranean in general, see Huebner in press a. This hypothesis has stirred some new studies on this topic: see now Remijsen and Clarysse 2008; Rowlandson and Takahashi 2009. I respond to their objections below in Chapter 7. Cf. Skinner 1997: 62. 78 Cf. Chapter 5 on succession of authority in the household.

Marriage patterns in Roman Egypt

49

We find marriage patterns similar to those documented for Roman Egypt in other regions and periods of the Mediterranean,79 in many early modern and modern Asian societies,80 and in the Arab world.81 In all these societies female marriage was early and universal. Young couples generally joined existing households, and rarely formed new ones. Even in nineteenthcentury Egypt more or less the same pattern of patriarchal and patrilocal family structure prevailed as in Roman Egypt.82 I suggest interpreting this as parallels rather than as continuities. Children belonged to the patrilineage; fathers acted as guardians for minor children; in the case of a father’s death a close male relative took over this role. Inheritance was partible and sons as well as daughters inherited. A girl would marry in her mid- to later teens and join her husband’s family unit. The young couple set up household with the groom’s family; a divorce or the husband’s death, however, severed her ties to his family and she regularly returned to her natal family. In early modern and modern north-western Europe, however, marriage occurred relatively late in life for men and women. A relatively high proportion of men and women never married at all, and neolocality upon marriage was the rule for young couples. When the young couple co-resided with their parents after their wedding, this was often temporary and was usually followed by the establishment of a new household.83 These differences and the regional persistence and continuity over a long period of time suggest that marriage patterns are affected above all by cultural and economic determinants, such as traditional residence forms, inheritance patterns, and forms of landholding. Changing demographics and changing modes of production can, however, also have rapid effects on traditional residence and household patterns, as we have seen for instance for the nineteenth century. As Hoch has observed for serfs in imperial Russia, patrilocal rather than neolocal marriage made it possible for men and women to marry earlier, since no additional capital was necessary to establish a new household. In addition, introducing a daughter-in-law and her dowry meant that the availability of labor and total wealth of the household increased.84 For this reason, parents urged their sons to marry relatively young. Compared with the age at the time of first marriage for males in early modern and modern 79

80 81

82

For instance, for late medieval Tuscany, see Herlihy and Klapisch-Zuber 1985, or for the fourteenthcentury village Montaillou in southern France, see Le Roy Ladurie 1975. Hajnal 1965, 1982; Skinner 1997; Reher 1998; Goody 2000; Engelen and Wolf 2005. Petersen 1968; Ahlawat and Zaghal 1989; Al-Haj 1989; Tucker 1993: 195; Cuno 1995; Okawara 2003. Other recent studies of family history in the Middle East include Duben 1985: 75–97; Tucker 1988; Gerber 1989; Duben 1990; Duben and Behar 1991. Tucker 1985. 83 Hajnal 1965: 103; Coale and Treadway 1979. 84 Hoch 1982: 245.

50

Household structures, marriage patterns, inheritance

north-western Europe, an average age at marriage of about 25 for men in Roman Egypt was, in fact, quite low.85 Not only was it not necessary to have the means for founding an independent household when couples resided with the husbands’ families: childbearing and rearing of children were also less costly than in nuclear households. In multiple family households children could be raised together, tended by their mothers in turn (i.e., co-residing sisters-in-law) or the older generation of the household, which allowed the children’s mothers to work or even become pregnant again quite soon. By bringing children into the household the couple gained status and power and established themselves within the household, so that besides being married at a young age, which often equals higher fertility, there were also strong social incentives which benefitted higher birth rates.

2.10 Landholding and inheritance patterns Social anthropologists have linked different kinds of household formations to prevalent inheritance and succession rules within a society or region. Under a stem family system, characteristic for early modern and modern central European societies, only one of the children, usually the eldest son, inherited the paternal property and took over the farm, leaving his siblings to find a living elsewhere. In the multiple family household system usually all the children were perceived to have a claim to their parents’ property. And in fact, in the Roman period of Egypt, all sons and daughters received a share.86 It had been an old Egyptian practice to give the eldest son a double share,87 a tradition that might still be found in the Roman period in the customary transmission of household headship to the eldest son and parents sometimes assigning the eldest son a larger portion of the inheritance.88 In the New Kingdom and the Late Period, siblings inherited together while the eldest son acted as a trustee and administrator of the property for the community of heirs, organizing the farming or leasing of the land, and allocating to the individual siblings their respective revenues.89 Land, 85

86

87 88

89

Bagnall and Frier 2006: 116. For Roman Italy and the argument that impoverishment in late republican time restricted marriage and fertility, see Brunt 1987; for a new view, Hin in press. For partible inheritance in Roman Egypt, see Kreller 1919; Taubenschlag 1955: 239–43; Rowlandson 1996: 152–70. Cf. Bagnall 1992: 138; Tacoma 2006: 209–20. Maehler 2005: 133–4. See Kreller 1919: 149–54; Montevecchi 1935: 101–3; Seidl 1965: 185–92; Pestman 1987: 61–2; Donker van Heel 1990: 93–103; Clarysse 1995: 95–6. For late medieval Tuscany, cf. Klapisch-Zuber 1985: 19. Lippert 2008: 62, 125. Cf. Codex Hermupolis 8.30–1.

Landholding and inheritance patterns

51

houses, and slaves were owned jointly by all heirs. If the patrimony was to be divided among the siblings, the eldest son received a double share and could choose first.90 In Roman times a similar system seems to have been at play in the rural multiple family households consisting of several married brothers with their families. Property was held κοινὸν καὶ ἀδιαίρετον, “jointly and undivided.”91 Here the eldest acted as head of the household and represented the family to the outside world. The inheritance system in Roman Egypt seems to have followed indigenous Egyptian roots rather than Greek practice, where sons usually inherited at a certain point during their father’s lifetime, not just upon his death.92 Nonetheless, all siblings were legally independent: a system of patria potestas under which sons were unable to hold and govern their own business as long as their fathers were alive did not exist for Greeks and Egyptians in Egypt.93 The system of partible inheritance in rural Roman Egypt would match with the observed preponderance for multiple family structures. Under this household form the requirements for labor on the family land were met, and, instead of the size of the agricultural holding being diminished, the farm supported all of its members.94 Rowlandson notes that joint ownership by siblings was most common among smallholders, probably to avoid the fragmentation of already small land parcels.95 Since under a system of partible inheritance in Roman Egypt the number of surviving heirs determined the size of each person’s share, the avoidance of fragmentation of land after a patriarch’s death apparently enabled the household to remain in control of greater wealth and resources. It is quite likely that those members of a family who farmed together also lived together under one roof, even though cooperation and sharing of the means of production did not necessarily involve joint living.96 However, partible inheritance is in general seen as stimulating actual cohabitation between inheriting siblings when they held their land and houses jointly and undivided.97 Multiple family households in which all the sons stayed in their parental home even after marriage seem to have been better off than simple family households across many early modern and modern patriarchal

90 92 93 95 96

97

Lippert 2008: 125. Cf. Codex Hermupolis 9.9–11. 91 Cf., e.g., P.Oxy. 34.2713 from 297 ce. Cf. Schmitz 2007: 35–6; Golden 1990: 119–21. Cf. Chapter 5 of this study. Taubenschlag 1916; Lippert 2008: 117–18. Cf. Arjava 1998. 94 Lemaître 1976; Shaffer 1982. Rowlandson 1996: 144, 173. Cf. Erdkamp 2005: 66–70. Dio Chrysostomus tells us about two brothers, humble peasants, who married two sisters and farmed their property together, but lived in separate households (Dio Chrys. Or. 7.10). Cf. Erdkamp 2005: 66. See also Alston 2002: 69–75 for Egypt. For late medieval Tuscany, see McArdle 1978: 137–8; Ring 1979: 16–19; Brettell 1991: 353.

52

Household structures, marriage patterns, inheritance

societies; at least as far as peasant households are concerned. In families in which resources, such as arable land, were not sufficient to maintain a large household, only one son would remain with his parents; the others would depart shortly after reaching maturity so that they could find employment elsewhere and establish their own households when they married.98 Considerable differences existed between urban and rural settings in Roman Egypt regarding predominant household forms. An urban environment seems to have led to an increase in nuclear families, while in rural environments multiple family households were more common.99 Similar observations that multiple family households were more common in the villages than in urban centers have been made for other agrarian societies, e.g., for medieval Genoa,100 the Middle East,101 and China,102 far into the twentieth century. The differences between urban and rural environments in Roman Egypt were most likely prompted by the same exigencies that we find in those other historic agrarian societies. In rural regions the household functioned as a unit of labor pooling: peasant farmers lived in complex family households in order to maintain a sufficient domestic workforce, with all the implications for headship succession and inheritance patterns at the time of the head’s death. Wage-earning families in the cities were much more likely to live in nuclear households which they set up upon marriage; the marital relationship dominated the small family enterprise.103 Moreover, as Klapisch-Zuber has shown for late medieval Florence, the majority of the urban population was in general poorer than the peasants in the countryside, which went hand in hand with smaller household forms, the dominance of simple family households, and the establishment of new households upon marriage.104 The same has been claimed by Kertzer and Hogan for nineteenth-century northern Italy.105 Also, rural elite families in nineteenth-century Egypt, who formed the largest households, with several conjugal couples living together, keeping the paternal inheritance undivided, and being able to organize labor more efficiently, were distinguished

98

99

100 101 102 104 105

Freedman 1979: 235–9. More recent scholarship has, however, meanwhile argued that joint household formation was also common among the poor, and rather represented the Chinese experience in general (Wolf and Huang 1980: 68–9; Hajnal 1983; Ma 1988). Bagnall and Frier 2006: 67: “Complex households obviously played a much larger role in the Egyptian countryside than in the more urbanized metropoleis.” Hughes 1977. Bates and Rassam 1983: 147, 196; Gerber 1989: 413–19; 1995: 149; Duben and Behar 1991: 75. Lang 1946: 136, 148; Yang 1959: 17–18. 103 Herlihy and Klapisch-Zuber 1985: 298. Klapisch-Zuber 1985: 19. Cf., however, Herlihy and Klapisch-Zuber 1985: 95. Kertzer and Hogan 1989. See also Flandrin 1979: 61; Torti 1981: 190; Barbagli 1984.

Fission of multiple family households

53

by their wealth.106 Bagnall and Frier doubt that rural families in Roman Egypt were on average wealthier than metropolitans:107 the comparative evidence, however, points in the opposite direction. A high mortality regime with an inheritance pattern of partible inheritance could mean serious perils. Too few children could easily lead to none surviving their parents. Too many, however, could mean fragmentation of property and, over the course of just one or two generations, impoverishment.108 In many societies multiple family households thus served according to Goody as a “strategy of heirship,”109 a strategy to ensure the continuation of the family. Since it was only sons who could guarantee that the paternal property remained in the agnatic lineage, a male heir was decisive in patriarchal societies. If a daughter inherited, the paternal property would eventually be lost to her children, i.e., to another lineage. Under a demographic regime as grim as that of antiquity, however, one-fifth of all couples remained childless, while another one-fifth had only daughters, resulting in 40 percent of all couples not having a male heir. In a multiple family household, however, the risk of having no male successors available from the paternal lineage was much lower than in nuclear households. For if one man could not boast of having a son, it was likely that at least his brother had a son, who would be available to perpetuate the household and prevent the loss of the land to the family.110 In multiple family households the children of brothers grew up like siblings. Since the usual term for cousin, ἀνεψιός, is very rarely documented in the private correspondence of Roman Egypt, we might actually assume that many of the “brothers” and “sisters” we find in the papyri were in fact cousins. We find the same habit of expressing the relations between cousins as siblings in many societies up to this day, for example in the Middle East or Asia where multiple family households are common.111

2.11 Reasons for and timing of fission of multiple family households Even multiple family households in Roman Egypt experienced fission at some point in their domestic life cycle. Differences regarding the timing of the household’s break-up have been noted from one society to another 106 109 110 111

Cuno 1995: 486–7. 107 Bagnall and Frier 2006: 69–70. 108 Cf. Tacoma 2006: 211–29. Goody 1973: 16. Cuno has described this strategy for nineteenth-century Egypt: 1995: 492–3. See also Kertzer 1989: 8. In island Southeast Asia, for instance, the relation between cousins is expressed as “we are siblings within the origin-house” (Waterson 1995: 55). For the literal and extended use of kinship terms in documentary papyri, see Dickey 2004.

54

Household structures, marriage patterns, inheritance

where the multiple family household prevailed as an ideal. The factors that affected the decision to divide and set up several new households are various and sometimes difficult to determine. The death of a male head, however, obviously increased the probability of household division, but the time that passed between the patriarch’s death and the break-up of the family varied considerably between individual families, localities, and societies. For seventeenth- to eighteenth-century Aleppo in Syria, for instance, evidence shows that brothers, after their father’s death, did not split up as long as their mother was still alive and living with them.112 And while in Ottoman Anatolia multiple family households usually split up right after the patriarch’s death,113 in nineteenth-century Egypt brothers continued to uphold the multiple family household long after their father and even mother died.114 Cuno has tried to explain why there were these differences between brothers in rural nineteenth-century Anatolia and brothers in nineteenth-century Egypt. For him, one of the most decisive differences between these two regions was the availability of land. While land was plentiful in rural Anatolia, in Egypt, owing to the scarcity of arable land, strategies to keep land together were necessary for the family to survive and prosper.115 Also in Roman Egypt the limited amount of available farmland and overpopulation of fertile regions may have favored the continued co-residence of married brothers. Joint living and joint farming formed a way of avoiding the fragmentation of land through inheritance. Kertzer similarly explains a tendency in the late nineteenth-century Emilia Romagna in northern Italy – married brothers staying together long after the parent generation had died – with the need for labor among the sharecropping families in this region.116 Czap has given a further reason for the preponderance of large multiple family households among the Russian serfs of the eighteenth and early nineteenth centuries: the more available hands on a farm, the higher could be the degree of economic specialization, which would allow some members of the family to engage in handicraft, thereby reducing the risk of economic ruin in case of a series of bad harvests.117 Moreover, in contrast to households that relied only on one wage-earning male, in multiple family households, in the event of illness or death of a male family member, there were still other males present to fill his place and take over his obligations to his land and his family.118 112 115 118

Meriwether 1999: 86, 94. 113 Duben 1985: 83–4. 114 Cuno 1995: 490. Cuno 1995: 486, 491, 495–6. 116 Kertzer 1989; cf. Douglass 1991: 302–3. 117 Czap 1982: 6–7. Cf. Bagnall and Frier 2006: 73. For the pervasive problem of fatherlessness in the ancient world, see Huebner and Ratzan 2009a.

Fission of multiple family households

55

Interestingly, the same pattern as for nineteenth-century Egyptian households can be observed for Roman Egyptian ones as well.119 The census returns record a quite considerable percentage of co-residing married brothers, sharing hearth and home, even if both parents were already dead. A petition from late third-century Oxyrhynchus by a young woman exemplifies such a constellation: the young woman tells us how her widowed mother lived together with her brothers in one household long after their parents had died. The siblings had inherited jointly their father’s property: “For all that was left to us by the foregoing inheritance – being one household and one family – was in that very house in which they (i.e., the petitioner’s mother and her two maternal uncles) lived in partnership, I mean because the slaves and lands and household and moveable goods were all held in common.”120 After her mother died, her two maternal uncles tried to cheat their niece of her rightful third part of the inheritance. However, if available farmland was limited in Roman Egypt and overpopulation of fertile regions of the country, such as the Arsinoite nome, made the upkeep of multiple family households advisable, these exigencies were not pressing enough that all households that would have been able to keep up their joint status in fact stayed together. At some point their members decided to split up. There seems to have been a cross-cultural ranking of household forms regarding the probability of splitting up. In societies where the multiple family household was the ideal, of all forms of multiple family households the vertical households consisting of a married parent generation and their married son(s) were the least likely to split up. Horizontal households of married brothers ranked second, and diagonal households with married cousins and/or uncles living together were deemed the most likely to split up. This tendency can presumably be explained partly by the degree of authority the head of household held over its members. While in a vertical multiple family household the head of household was the father of other adult males, in horizontal and diagonal households the head’s authority was less indisputable, especially if there were other adult males of his own generation or even of the one above him.121 Czap has shown, for instance, that in nineteenth-century Russia household fission most often occurred when a household head was survived by a brother but headship of household went to his son, with the consequence that the uncle became inferior to his nephew.122 119

120 121

Cuno 1995: 490. Cf. also for Ottoman Syria, where we find the same tendency to live in frérèches, Okawara 2003: 63–4. P.Oxy. 34.2713 from 297 ce; trans. Rowlandson 1998: 75; see also Evans Grubbs 2002: 54–5. Cf. Lee and Campbell 1997: 122–3. 122 Czap 1982: 19–20.

56

Household structures, marriage patterns, inheritance

Moreover, studies on nineteenth-century China have shown that an increase in population density went hand in hand with a rise in multiple family households. While such households consisting of a married couple and their married sons constituted less than 20 percent of all household types in early nineteenth-century China, about the same proportion as in Roman Egypt,123 we can observe an increasing tendency for multiple family households to be formed by married brothers and even cousins over the course of the nineteenth century.124 Above all, the percentage of so-called diagonal multiple family households consisting of married uncles and/or nephews rose dramatically, from practically 0 to 25 percent within a couple of decades.125 These numbers translate into a proportion of almost half the people living in a diagonal multiple family household in 1873, while in 1792 not even 1 percent had done so. These changes over the course of the nineteenth century were probably triggered by the fact that it became economically more difficult for family members to set up a new household and support it.126 With limited resources of land, the multiple family household seems to have constituted the most economic solution. By extending the range of suitable co-resident family members, the formation of complex multiple family households was no longer precluded despite high mortality and low fertility. This means for Roman Egypt that, despite high mortality rates, a higher rate of multiple family households would have been possible if there had been a greater tendency among more distant relatives to co-reside. It seems, however, that families in Roman Egypt preferred to discontinue living together at the latest when the multiple family household consisted only of the diagonal stage of married uncles and nephews. We do not find a single example of this household formation type in the Roman census returns. The division of property in Roman Egypt was also more likely when the late father left his property to children from different marriages. When the old Psenamounis from late second-century Oxyrhynchus died, he left an adult son and an adult daughter from his first marriage, and two adult sons and three minor children from his second marriage.127 After his death his children drew up a contract dividing their father’s estate among them. His son and daughter from his first marriage agreed to take a smaller portion than what they were entitled to in their father’s will because the children from their father’s second marriage offered to cover the deceased father’s debts. The division of the property concerned the two groups of siblings 123 126

See Table 2.1 (p. 46). 124 Lee and Campbell 1997: 111. 125 Lee and Campbell 1997: 51. Lee and Campbell 1997: 49. 127 P.Oxy. 14.1638 from 282 ce.

Fission of multiple family households

57

from the father’s two marriages, but did not discuss any divisions within these two groups. The full siblings apparently agreed to hold their inherited property jointly.128 As we have seen above, bequests to children from the same marriage were often made together, and it was left to the children to divide the property at a later point whenever they seemed to see a need for it. In sum, partible inheritance, scarcity of land, and the necessity to maintain a sufficient labor pool were therefore probably the three most important factors that influenced household composition in Roman Egypt. While strong family ties surely were a reason, economic considerations might have outweighed them and convinced several married brothers to continue to live in one household despite tensions that indubitably arose. Over the following chapter we will repeatedly come back to these patterns of household composition influenced by various culturally, socially, and economically determined factors that have an influence also on family life, marriage arrangements, fertility, and intergenerational solidarity. 128

For a father’s or mother’s will benefitting children of different marriages, cf. also P.Oxy. 4.716 from 186 ce; P.Oxy. 9.1208 from 291 ce.

chapter 3

Balancing benefits and obligations Parental love and filial piety over the life course

He who does not support a wife, he who does not support a child, the evil state of things is doubly for him, he grinds flour, he has no rushes, Old Sumerian proverb1 he is not reckoned among people. The father who sired a son will say: “You are my son! I have raised you that you will repay me the favor.” I, the god Erra, however, will let the son die that the father has to bury him. Later I will let the father die, but no one will be there who will bury him. Poem from ancient Mesopotamia2

3.1 Fertility in Roman Egypt Artificial limitation of fertility seems to have been rarely employed in Roman Egypt.3 Virtually universal marriage for men and women, young age at first marriage for women, short spacing between births, and the fact that women had children at quite high rates well into their forties point to an unrestricted natural fertility regime. The average age of mothers at childbirth was 27.1 years, and each woman had to give birth to almost 6 children to prevent the overall population from decreasing.4 Means for restricting fertility seem to have been the spacing of birth intervals achieved by prolonged breastfeeding and the abstention from remarriage by women from a certain age on.5 Furthermore, in the metropoleis, families reported significantly fewer female than male children to the census officials, which could have been a result of under-reporting of young girls, or rather a preference for rearing male offspring. The latter could have found its expression in the neglect of baby girls, leading to higher mortality rates for girls, or, even if rare, infanticide or exposure of female 1 4

Stol 1998: 59. 2 Cf. Westbrook 1998: 13. 3 Frier 1994; Bagnall and Frier 2006: 148–51. Bagnall and Frier 2006: 135–9. 5 Cf. Chapter 4.

58

Downward or upward wealth flows?

59

children.6 Female slaves were much more common than male slaves in the census returns, constituting roughly two-thirds of the slave population,7 a pattern which finds its equivalent, e.g., in the Chinese Tang census lists dating to the sixth to eighth century ce.8 Thus, many exposed females might have been picked up and raised as slaves by other families.9 Infertility might have been another check on the number of children. From developing countries we know that a considerable percentage of women (one in four) are infertile owing to complications sustained in a previous childbirth. Inadequate nutrition which causes anemia and lack of ovulation, too strenuous physical work, and generally poor health are other causes of unwanted childlessness.10 Female genital mutilation, a common practice in parts of Africa and the Near East and potentially also in Roman Egypt,11 leads to high rates of infertility.12 Lower fertility was certainly also the outcome of other indirect circumstances such as seasonal migration, when husbands were away from their wives for many months in a row.13

3.2 Downward or upward wealth flows? According to classic wealth flow theory first proposed by Jack Caldwell, the leading demographer in the field of fertility transitions, intergenerational wealth flows, measured over a lifetime, was primarily upward from children to parents in societies before the demographic transition. In societies after the demographic transition, on the other hand, wealth flows occur mainly downward, from parents to their offspring.14 Children in preindustrial societies, according to Caldwell, were not regarded as an economic burden but rather an asset, as education costs were low and children could 6

7

8 9

10 13 14

For the discussion about the incidence of infanticide and exposure in the Greek and Roman world, see Engels 1980: 112; Golden 1981: 320; Harris 1982, 1994, 1999; Boswell 1984: 10–33; 1988: 53–137; Scheidel 1997b; Corbier 2001a; Evans Grubbs 2010. Bieżuńska-Małowist 1971, 1977; Masciadri and Montevecchi 1984; Straus 1988: 854–6; Bagnall 1993: 202; 1997a; Scheidel 1997b, 2005, 2011: 287–310; Bagnall and Frier 2006: 92–4, 151–3, 158–9; Evans Grubbs 2010; Harper 2011: ch. 10. For wet nurses hired for foundlings, see Masciadri and Montevecchi 1984: 10–20. Liao 2001: 345–6. The raising of foundlings as slaves is a well-documented practice in the papyri from Roman Egypt (cf. Bieżuńska-Małowist 1971, 1977; Straus 1988: 854–6; Bagnall 1993: 202; 1997a; Scheidel 1997b, 2005, 2011: 287–310; Bagnall and Frier 2006: 92–4, 151–3, 158–9; Thompson 2002: 152–3; Evans Grubbs 2010; Harper 2011: ch. 10). For wet-nurses hired for foundlings, see Masciadri and Montevecchi 1984: 10–20. See for adoption and fosterage in the ancient Eastern Mediterranean: Huebner in press a. Agigian 2000: 1138–9; cf. section 7.2. 11 Huebner 2009b. 12 Agigian 2000: 1139. For other societies, see Kaplan and Lancaster 2003; Bavel 2004; Hin in press. Caldwell 1976b, 1982, 2005; Cain 1977; cf. Hin in press. For a definition of the model of demographic transition, see sections 1.1 and 1.2.

60

Balancing benefits and obligations

contribute to the household’s income from a very early age, especially in peasant societies where children provided a cheap source of labor. For this reason, Caldwell argues that individuals in early modern and modern societies had no economic reason to limit fertility. In addition, the fact that children from an early age contributed more to the household than they consumed, and by this net bargains were created for their parents, stimulated procreation. Caldwell’s theory has been an object of contention since he first proposed it in 1976. Opponents in the field of anthropology and demography contend – generally basing their argument on empirical evidence – even in poor peasant societies intergenerational wealth flows were downward from parents to children, if one considers the entire life course.15 For contemporary developing countries, social anthropologists use micro-level data to assess the correlation of intergenerational wealth flows and fertility.16 Thanks to quantifiable data for Roman Egypt we are in the exceptional situation for the ancient Mediterranean to measure approximately downward and upward resource flows as well. We have the necessary data to calculate how expensive it was to raise a child and provide him or her with food and clothing.17 If we offset these expenses against the contribution those children could bring to the household in later years, we are able to calculate the cost–benefit ratio of bringing up a child from a purely economic point of view. Table 3.1 lists the caloric requirements for children and adults at different ages who are moderately active,18 and the prices in drachmas for food and clothing, as far as we know them from the papyri from Roman Egypt. The physical activity level for most people in antiquity was certainly higher than moderate, but these lower rates allow for the smaller size and lighter stature of children and adolescents in Roman Egypt compared with modern Western populations (see Table 3.1). In my calculations I assume that 80 percent of the caloric requirements were covered by bread and 20 percent by more expensive groceries such as oil, eggs, wine, and meat.19 During the late first century ce to around 15 16

17

18

19

Kaplan 1994; Stecklov 1999; Lee 2000; Lee and Kramer 2002. Caldwell 1976a for south-west Nigeria; DeLancey 1990 and Makinwa-Adebusoye 1994 for subSaharan Africa; Chang 1990 for China. For the problems associated with prices, inflation, dating of documents, and comparability, see Drexhage 1991. Estimated energy requirements from the Institute of Medicine, Dietary Reference Intakes for Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids, Washington, DC, 2002. Cf. Drexhage : 442, who argues for oil and wine as daily constituents of the peasant diet. See also Bagnall 2007, who argues on the basis of the fourth-century Theophanes archive that for the average population 36 percent of caloric requirements from sources other than wheat is perhaps too high a figure.

Table 3.1. Costs of bringing up a child in Roman Egypt. Caloric requirements Age Males 2 3 4 8 9 13 14 18 19 30 31 50 51+ Females 2 3 4 8 9 13 a

Per day 1,000 1,400 1,800 2,400 2,600 2,400 2,200

Per year

Costs in drachmas Other Breada provisionsb Clothing etc. Poll tax Total Ac

1,400 365,000 511,000 36 50 1,600 511,000 584,000 50 58 2,200 657,000 803,000 65 80 2,800 876,000 1,022,000 87 102 2,800 949,000 1,022,000 94 102 2,600 876,000 949,000 87 94 2,400 803,000 876,000 80 87

1,000 1,400 1,400 1,600 1,600 2,000

365,000 511,000 511,000 584,000 584,000 730,000

36 50 50 58 58 73

20 20 40 25 30 30 25

10 20 20 20 20 20 20

20 20 30

10 20 20

80% of caloric requirement. Oil, eggs, wine, cheese, meat (20% of caloric requirements). c Costs per year. d Costs up to the respective age. e Total annual income (drachmas). f Annual contribution to the household (drachmas). b

20 20 20 20

66 90 125 152 164 157 145

Total Bd

80 98 140 167 172 164 172

132 160 492 552 992 1,112

66 80 90 98 108 123

132 160 492 552 924 1,044

Incomee Contribution f

333 333 333 333

166 161 166 166

181 169 181 188

62

Balancing benefits and obligations

160 ce, 1 artaba of wheat cost about 8 drachmas (its price increased to 18 drachmas in the period from the 190s to the 270s, but we want to focus here on the early Roman Empire).20 One artaba was 38.8 liters or 30.3 kg of wheat. Scheidel estimates that 1 artaba yielded about 35 kg of bread, taking into account a 0–10 percent weight loss for milling and the higher moisture content of bread compared with wheat.21 One kilogram of whole-wheat flat bread provides about 2,300 calories. A small child of 2–3 years, who consumed about 1,000–1,400 calories per day, thus needed at least 4.5 artabai of wheat per year (36 drachmas), while an older child in his teens consumed up to 10 artabai of wheat per year (80 drachmas) to meet his caloric requirements. For the additional groceries, I add a fixed price of 20–40 drachmas per year (depending on the various age groups) for easier calculation purposes, and for clothing another 20 drachmas per year.22 I assume that children up to age 2 were nursed by their own mothers, and consequently caused little expense. According to my calculations, food and clothing cost for a toddler about 66–80 drachmas per year, an older child a little less than 100 drachmas per year, and boys in their early teens about 125–40 drachmas per year. A document from 150 ce mentions a monthly alimony payment over 6 drachmas for a boy, thus 72 drachmas a year.23 We do not know how old the boy was. If he was still a toddler, 72 drachmas were certainly enough to cover for his food and clothing.24 The fact that contemporary apprenticeship contracts mention annual sums of 48 drachmas for maintaining a teenage apprentice, payable either by the parents or the master depending on where the boy lived, can only be explained by the master and parents splitting the costs for feeding him and half the expenditure in cash going to whomever the boy lived with.25 Children, boys as well as girls, started to contribute from an early age in peasant societies. They tended the flocks, and helped with weeding, babysitting, or light household chores.26 Usually, only from around the age of 14, males worked full-time on the farm and/or as an agricultural wage laborer.27 It has been argued that many of these wage laborers came from families who owned land themselves but had surplus labor available to rent out when the children grew into adolescence.28 By working only temporarily as wage 20 23 25 27 28

Rathbone 1997; Scheidel 2002: 103. 21 Scheidel in press b. 22 Drexhage 1991: 448–9. P.Mu¨nch. 3.101. 24 Cf. Drexhage 1991: 446–8. P.Tebt. 2.385 from 117 ce; SB 6.9374 from 169 ce. 26 Varro, Rust. 1.17.2, 2.10.1. Apprentices in a craft often did not live at home, so we do not want to consider them here. Rathbone 1991: 153; Garnsey 1998: 148; Erdkamp 1999.

Downward or upward wealth flows?

63

laborers on other people’s estates, these local peasants earned a supplement to the income from their own land. Wage laborers were employed for irrigation works, weed clearing, tree or vine pruning, sheep shearing, pack animal loading, earth moving, plowing, sowing, mowing, threshing, and the harvest of olives or grapes.29 Virtually all the wage laborers recorded for Roman Egypt were males, and aged 14 to 40.30 As seen above, my calculations omit contributions that children of a younger age made to the household. I compensate for this by assuming that 14-year-old boys had the same labor productivity as an adult male which, of course, cannot have been the case.31 In the first century ce agricultural wage laborers earned about 3 to 7 obols per day. Their wages increased to 8 to 16 obols from 125 ce on.32 I calculate expenses from around 125 ce with an artaba costing 8 drachmas and a wage laborer earning 8 obols a day. I further assume that an agricultural wage laborer found employment for half of the year. With a daily wage of 8 obols and 180 days of employment, his annual wage amounted to 242 drachmas. In our case I include an additional 70 days, giving 250 days as the total number of working days, as a son of the household would work on the family farm on those days when he was not laboring in someone else’s fields. Using 365 days of employment would certainly be too generous, since during certain periods of the year there was little work to be found and agricultural societies suffered in general from a structural underemployment over the course of a year. In sum, when a son started working and earning wages, or working on the family farm for which the family saved employing a wage laborer, he had already cost his parents the total sum of more than 1,000 drachmas. From the age of 14 he could contribute around 2,000 obols, i.e., 333 drachmas, per year to the household. However, as we can see from Table 3.1, with an average cost of 160 drachmas per year for food, clothing, and taxes, about half of his income would have been eaten up by expenses for his own maintenance,33 meaning he could contribute only 173 drachmas to the household. To repay his parents the debts for raising him, he would have to work for 5.8 years. That means, only from the age of around 20 would a

29 30

31 33

Cf. Drexhage 1991: 412–25. Scheidel 1996c. For independent laborers on the third-century Appianus estate, see Rathbone 1991: 148–74. Cf. Hin in press. 32 Cf. Drexhage 1991: 405. For the third century, see Rathbone 1991: 156–8. Cf. Rathbone 1991: 165, for the difficulties to make ends meet as a wage laborer, as wages were very low and employment seasonal, temporary, and unreliable.

64

Balancing benefits and obligations

son’s benefits start to outweigh the costs of his upbringing. This is also called the cumulative break-even age.34 If one or more sons survived to adulthood, fine. If there were, however, more daughters than sons or no sons at all, the family risked a losing bargain. Daughters only cost but never contributed significantly if they got married at an early age, and then even required an advance part of the inheritance in the form of a dowry. A son brought at least a daughter-in-law to the household who came with a dowry and whose labor could be employed. So far we have not even taken into account the high infant mortality rate, which in over half of all cases caused parents to invest in children who died before they reached adolescence and could pay back the favor of their upbringing. One-third of all children died before reaching their first birthday, and only slightly less than 50 percent survived up to their fifteenth birthday.35 The grim mortality regime in theory forced a surviving son to contribute considerably more to his parents’ household than his own rearing had cost, in order to compensate the parents financially for the early death of his brothers and the upbringing of his sisters. Each son thus had to work not only for himself but on average also for a brother who died in infancy, that is up to another 6 years. At this point the son was already 26. Before a son could repay his parents the debts for raising him, his late brothers who did not survive to adulthood, and in addition his sisters, it was very likely that his parents were no longer alive. Only 63 percent of all males still had a father when they turned 15, only 28 percent still had a father when they reached 30, and at the age of 40 only 9 percent of all men still had a father.36 Having children for older parents was thus even less financially beneficial than for younger ones, as the older might not live long enough to see their sons start working.37 Raising children for economic purposes thus seems to have been a risky endeavor at least. Even the argument that sons provided labor to work the farm becomes obsolete. Hiring wage laborers who earned little and could be employed when needed was more economical and less risky. Interestingly, Mueller, who conducted a study in the 1970s on twentieth-century rural Egypt, also comes to the conclusion that 34

35 36

37

Cain 1977. The point at which a child compensates for the means it consumes has been variously identified for other peasant societies as between 12 and 29 years old. Cf. Kaplan 1994; Stecklov 1999; Lee 2000; Lee and Kramer 2002. Model West, Level 4 males. Cf. Bagnall and Frier 2006: 100, table 5.3. Saller 1994: 52, table 3.1.e, male, “ordinary,” Level 3 West. For the problem of pervasive fatherlessness in the ancient world, see Huebner and Ratzan 2009. Cf. Stecklov 1999: 14.

Parents and children in Roman Egypt

65

children were net costs.38 In sum, the link between the costs of children and fertility remains highly dubitable. Further studies on a variety of societies have, however, added more evidence that perceptions regarding the likelihood of receiving support in old age correlate positively to higher fertility.39 This seems also to have been a factor for parents in Roman Egypt. Children seem to have added considerably to their parents’ emotional well-being, their status in the household and their community, and their perceived security in old age. I will come back to this point in Chapters 5 and 7.

3.3 Parents and children in Roman Egypt In the following we want to study the emotional and social aspects – and less the financial ones – of the relations between parents and young children, on the one hand, and elderly parents and their adult children, on the other. In general, our ancient sources are silent about children’s lives, at least about the children that did not belong to the highest echelons of society: “Of all social groups which formed the societies of the past, children, seldom seen and rarely heard in the documents, remain for historians the most elusive, the most obscure.”40 The papyri are undoubtedly again the best source material when we want to learn more about the children of the humble masses. By retrieving intimate details and personal aspects of family life from the private letters of Roman Egypt that have survived in the desert sands, we try to ground intergenerational solidarity and interdependence in practice, place, class, and time.41 The perils of anecdotal evidence are well known, above all the doubtful relevance of scraps of information for the historical reality of a wider group of society. Can we paint a picture of intergenerational relations among family members in Roman Egypt based on pieces of evidence that have come down to us by chance? But this is all that we have, and if we want to learn anything at all about the relationship between parents and their children from the sub-elite strata, we have to fall back on this kind of evidence in our endeavor to paint a bigger picture. The reader will find episodes recounted from the lives of the lower and middle 38 39

40 41

Mueller 1976. Cf. Kaplan 1994. Leibenstein 1975; Cain 1981a; Caldwell 1982; Willis 1982; Hammer 1986; Schellekens 1993; Hugo 1997. Boswell 1984: 10. There are not many studies that deal with children per se in Graeco-Roman Egypt: Hobson 1989; Bradley 1991a: 103–24; Cribiore 1996, 2001; Montserrat 1996; Phang 2001; Vuolanto 2002, 2003; Mirkovic 2005; Huebner 2009a, 2009b.

66

Balancing benefits and obligations

classes that offer exemplary illustrations of individual relationships. These documents, most of them private letters, certainly reflect, despite formulaic conventions, the individuals’ personalities.

3.4 Balancing benefits and obligations: parents Parental love In general, parents in Roman Egypt seem to have been expected to feed and clothe their children, watch over their moral development, celebrate their coming of age, provide their sons with an education and their daughters with a dowry, and find suitable spouses for them.42 Like today, in Roman Egypt attitudes and behavior of parents toward their children varied considerably from individual to individual and between different social and economic backgrounds. We find examples ranging from serious abuse to deep emotional attachment.43 There is no doubt that the ancients sometimes exposed, killed, sold into slavery, or prostituted their children and routinely expected them to contribute to the household income from a very early age. Yet, children also represented the future of the family. They were deeply loved and their raising and education were taken seriously and watched closely.44 Parents did not spare expense and proudly invited friends and neighbors to celebrate their sons’ and daughters’ births, coming-of-age ceremonies, and weddings.45 A father’s relationship to his children ideally rested less on patriarchal authority, control, and subordination than altruistic love, care, and indulgence. If fathers acted harshly, it was because they believed it to be in their children’s best interest. Patria potestas was not interpreted in its strict Roman sense in Egypt: sons were independent and could own property once they had reached the age of legal majority. The father’s authority represented a form of guardianship for underage children rather than an absolute power.46 42

43

44 46

Focusing on Roman Egypt, we will not recount the legal background and literary references to the duties of parents toward their children in the Roman world. Readers will find this done much better elsewhere: Rawson 1986, 1991, 2005: 1–11; Wiedemann 1989; Bradley 1991a; Dixon 1992; Eyben 1993; Bertier 1996; Rawson and Weaver 1997; Backe-Dahmen 2006; Huebner and Ratzan 2009a: 3–28. Cf. Roman laws from the second century ce: CIust 5.25 De alendis liberis ac parentibus. For the discourse about Ariès’s theory that childhood as a separate life stage was an invention of the eighteenth century, see Wilson 1980; Burton 1989; Dixon 1992; Garnsey 1991; Harlow and Laurence 2002: 35. Cribiore 2001. 45 Cf. Huebner 2009b. Taubenschlag 1916; Lewis 1970; Arjava 1998; Beaucamp 2007: 278. For reciprocal obligations, patria potestas, and pietas in Rome, see Saller 1994: 102–32; for classical Athens, see Golden 1990: 80–114; for the ancient Near East, see Stol 1998.

Balancing benefits and obligations: parents

67

A charming example of how a son could twist his father around his little finger comes from third-century Oxyrhynchus. Shortly after joining the army, a young man realized that the life of a legionnaire was not exactly what he had wished for.47 Rather than serving in a legion he wished to do service in the cavalry. Not only did he appeal to his father, he also sought the support of his sister and mother, who then entreated the father on his behalf. The young man asked his father to come down to Alexandria and exert his influence on the authorities to help him succeed in his wishes. Pausanias, his father, was at first unwilling, as he tells us himself. Only after his wife and his daughter begged him to go and support his son did Pausanias travel to Alexandria. After some difficulties, he succeeded with the authorities. His son was assigned to the cavalry and Pausanias accompanied him to his new point of service in Coptos. Another example is the well-known letter from late second- or early third-century Oxyrhynchus of a little boy trying to blackmail his father to take him with him to the capital, Alexandria. If his father did not consent to send for him, he warned him that he would not greet him or take his hand anymore, or write him a letter. And he even threatened that he would refuse to eat and drink. His father, who had apparently sailed away without telling him about his imminent departure, had sent him presents instead but the spoiled boy was not satisfied with them. If we believe the young boy, he unnerved his mother so much with his complaints that even she wished her husband had taken him with him.48 As in the case of the young cavalryman, we have further evidence that mothers sometimes intervened in conflicts between fathers and sons. A mother put a question to the oracle of Serapis in the first century ce to solve a conflict between her husband and her son.49 Her husband apparently tried to convince their son Phanias, who was already grown up and married, to make a contract – about what we are not told. It seems that Phanias was hesitant whether to comply with his father’s wishes or to oppose his will, and so his mother went to the sanctuary to ask for the god’s opinion. We have more evidence that gods were approached for their intervention when parent–child relationships were strained. Of the twenty-one questions that survive from a list of addresses to an oracle, one reads: “Am I to be reconciled with my offspring?”50 A certain Cornelius and his son Hierax from second-century Oxyrhynchus represent a case of a very affectionate and caring relationship 47 50

P.Oxy. 14.1665. 48 P.Oxy. 1.119. Trans. Winter 1933: 60. W.Chrest. 122, no. 78 from 6 ce.

49

P.Oxy. 8.1148.

68

Balancing benefits and obligations

between father and son.51 Hierax had apparently left home in order to study in a major city and his father advised his son, while using several terms of endearment, to be studious and to attend to his books, which would then bring him profit, and reminded his son not to offend anyone. He confirmed that he was going to visit him as soon as he was done with his business. Cornelius also announced that he would send Hierax money and clothes, and he advised his son how to wear them. A certain Ision in fourth-century Oxyrhynchus received a letter from both of his parents, Psais and Syra, who were looking forward to his arrival back home, telling him about the preparations for his coming. Among other things they had made pickles for him.52 We have another letter from the third-century ce from a young boy who was away from home at school. He wrote to his father, pleading for him to visit. He had already written five letters but his father had only replied to one and had not come to see him thus far. The boy’s teacher had also asked for the boy’s father to travel to see him, probably because he was waiting for his salary. The boy reassured his teacher that his father would come soon, though he did not know when this would be. The boy felt lost and homesick, and missed his old teachers, his brothers, and his pigeons back at home.53 Even if respect for one’s parents was a core value, we are occasionally offered a glimpse into real-life relationships between parents and children that speak rather of equal standing, and quite frank relationships between children and their parents. In a letter from third- or fourth-century Oxyrhynchus a certain Demetrius reproached his father that he had not sent him the hay for the oxen that he had asked for some time ago.54 As a consequence, his oxen grew emaciated without fodder and were not able to work, and Demetrius could not irrigate his fields. “You seem to be mocking my industry,” Demetrius accused his father. His attitude mirrored that of children down the years who have not felt taken seriously by their parents even as adults. Origenes and his son Trophimus from late third- or early fourth-century Alexandria were, however, on very good terms. The son treated his father as a buddy rather than an honored adult.55 Trophimus flippantly quoted remarks from his father’s preceding letter: Origenes had called his son mockingly “gift of Zeus” because of his boastfulness when he had sent 51 53

54

P.Oxy. 3.531. 52 P.Oxy. 10.1299. SB 3.6262; cf. Cribiore 2001: 112. See also Schubart 1923: 69; Metzger 1974: no. 42. For an English translation, see: Winter 1933: 64–5. P.Oxy. 6.938. 55 P.Oxy. 8.1160.

Balancing benefits and obligations: parents

69

money home, and he had added some playful remarks about his son’s lover in Alexandria. Trophimus wrote back in mock innocence asking who his father thought his lover was. At all times, parents have worried about the health and safety of their children. Papyri often speak of love for one’s children, and record phrases such as “for you know the anxiety for a child.”56 Mothers in particular seem to have worried greatly that their sons could be ailing.57 From a secondcentury literary account of the healing power of Imouthes-Asclepius we are given details about how a mother was anxious about her adult son who suffered from periodic attacks of fever over a period of three years, probably malaria. When her son was lying in a fever and pain, she sat at his bedside the whole night without sleep, being grieved by her son’s agonies.58 Another woman, named Isidora, from second- or third-century Oxyrhynchus wrote to her husband to rush home because their son was sick and had not eaten in six days. She was filled with such despair that her son might die that she threatened to hang herself if her husband did not hurry to her side to support her.59 Theonas, a soldier from the early second century ce, assured his mother Tetheus back at home in Oxyrhynchus that he was not seriously ill, but just had not written back because he was in camp.60 He told her not to worry, and he blamed the person who had spread the rumor that he was sick. It becomes apparent that Theonas was very upset that his mother had been worried about him. He told her twice that he was fine and that she should not take any trouble or send him anything. In another case from the late third or early fourth century the writer of a letter, Thonis, reassured his friend or relative that he had received the latter’s son safe and sound in every respect and that he would take care of him as if he were his own.61 The father was no doubt worried about his son’s safety and well-being when the boy journeyed to the city alone, leaving his family behind. Since Thonis added that he would make sure that the boy attended to his work, we might assume that the boy had been sent to an uncle or a father’s friend for an apprenticeship. Even if the following story lies outside our period of analysis, we may nonetheless take a short diversion to hear about the family of Kleon, an 56 57

58 59

60

e.g., BGU 2.380 from the third century ce. As we will see also further below, relationships between mothers and sons seem to have been especially close and emotional (see Chapters 5 and 6). P.Oxy. 11.1381, lines 102–7. PSI 3.177. Cf. Bagnall and Cribiore 2006: 77, 280: “The letter is evidence against the notion that high child mortality led to a lack of ‘emotional investment’ in young children.” P.Oxy. 12.1481. 61 P.Oxy. 12.1493.

70

Balancing benefits and obligations

engineer in third-century Ptolemaic Egypt.62 The surviving correspondence of his archive allows an incomparable insight into ancient family life, the responsibilities of a father, and his relationships with his children. We first meet Kleon when he was a middle-aged married man with two sons in their teens. By royal appointment Kleon held an influential position as a regional architekton. He had a substantial income and was high enough in the hierarchy of officialdom to approach the king in person. He and his wife lived in Crocodilopolis (later Arsinoe) where he was the commissioner in charge of all public works in the area, while their sons pursued their studies in Alexandria, around 300 km downriver. Kleon provided for the expenses of his sons, and when they finished their education he took care of their successful career entry, fulfilling the social role society expected him to undertake. Kleon introduced his older son to the king when he was ready to enter his chosen career. When his younger son, Polykrates, had just completed his training to be a surveyor, the son impatiently wrote to his father from Alexandria, reminding him that it was now his turn to be introduced: Polykrates to his father, greeting! I trust you are well and everything is going satisfactorily. We too are well. I have written to you repeatedly to come here and introduce me [to the king] so that I may be rescued from my present state of unemployment. And so now again I ask, if it is possible and there is nothing in the public works to prevent you, try to come here for the festival of Arsinoe, for I am sure that commending me to the king will be easy if you are here in person . . . Write us in reply, so we may know how you are and not worry [about you]. Take care of yourself to stay well and come to us in good health. Farewell.63

Kleon’s sons behaved well and loved and respected their father. Between the lines of their letters we discover a sense of deep emotional attachment. They were worried when they had not heard from him for some time and were anxious about his well-being. Both sons were poised to enter successful careers, and their parents had obviously every reason to be proud of them. Parents in antiquity, however, also faced the opposite end of the spectrum. What if a son showed a lack of respect and had the wrong friends, or even became a criminal? How was a father expected to react? When one of Kleon’s younger co-workers, a certain Demetrios, was put into prison, he 62 63

Cf. Lewis 1986: 37–45; van Beek 2006. P.Petr. 2.11 from the middle of the third century bce; trans. Lewis 1986: 38. Cf. Van Beek, 2006: 58–9; B. Van Beek, “‘We too are in good health’: the private correspondence from the Kleon archive,” in: P. Van Nuffelen ed. Faces of Hellenism: Studies in the History of the Eastern Mediterranean (4th Century B.C . – 5th Century A.D .), Leuven, Paris, Walpole, 2009: 147–60.

Balancing benefits and obligations: parents

71

called Kleon for help because he obviously had no one else whose assistance he could have sought: To Kleon, greetings, from Demetrios. I have already written you before about the arrest for which I have been arrested now. You also know that we have already been hard pressed during the works, and now I am being utterly pressed, since I have been carried off to prison. Think of me as your own son, and lead me out of prison. You will not suffer any harm. For I am in need of a lot of things in prison. Farewell.64

This hints at Demetrios’ belief that Kleon, if his own son were in such a miserable situation, would have pulled out all the stops and used all of his influence to free his son, regardless of whether he thought he was guilty or not. Thus, Demetrios begged that he might act in the same way in his case. Unfortunately, we do not know if Kleon responded to his pleas. We are able to follow the life of Kleon for a few more years. He was kept busy with the supervision of his ten foremen and a labor force of around one thousand workers. They constructed new irrigation works in Crocodilopolis and carried out the required maintenance and repair for old ones. For one reason or another, however, Kleon failed to complete his tasks as expected and eventually lost the favor of the king that he had enjoyed for so long. We know from a letter of his wife that when the king came to the Arsinoite nome on an inspection tour, Kleon had not managed to complete the irrigation works at the village of Ptolemais in time for the royal visit. In addition, some irregularities were found in his accounts of receipts and expenditures. At this point we might wonder how Kleon’s sons reacted when they heard that their father had made an unfavorable impression on the king and was in danger of losing his position. His money had given them the opportunity to go to Alexandria to conduct their studies, and they had relied heavily on his connections and influence for their own advancement. His fall from royal grace could well mean the end of their own nascent careers. However, Kleon’s sons did not let their father down. When his older son wrote him, true filial love emanated from his words: Absolutely nothing will be more important to me than to care for you for the rest of your life in a manner worthy of you and worthy of me, and to see to it, if something happens to you as it does to mortals, that you obtain all due respect. The most important thing to me will be to care for you, both as long as you live and when you have departed to the gods.65 64 65

P.Petr. 2.4(7) from the middle of the third century bce; trans. Van Beek 2006: 155. P.Petr. 2.13(19) from around 252 bce. Cf. Lewis 1986: 44–5.

72

Balancing benefits and obligations

These two sons were obviously well aware that they owed their father a great deal. They begged him to take his leave and live with them: So please make every effort to be released permanently and if you find that is impossible, then at least for the duration of the river’s ebb, when there is no danger [of flood damage], and besides you can leave Theodoros there to take over for you, so you can spend at least that time with us. Just bear in mind that you will experience no disgrace and that everything will be seen to by me for your being free of pain and sorrow.

We know that Kleon was indeed convinced by his son to take some time off.66 Nevertheless, he seems never to have returned to his office, and nor do we ever hear of him again. We can assume that his sons convinced him to stay with them permanently and enjoy his final years with his family. Household composition and children Household composition is regarded as an important factor for the quality of a child’s life. Studies have shown that differences in household type have an impact on the nutritional status of children, the quality of childcare, and their access to higher education. The household ratio of adults to children is especially decisive when measuring these factors. More adults in a household mean that their economic contribution can afford a better living standard for the children. On the other hand, if these adults are dependent on the household economy, the result might be negative consequences for the nutritional status of children. More children always meant that resources had to be divided into smaller shares. Again, it mattered if the children were siblings or cousins, and which position in the household hierarchy their father held. Children seem to profit more from a nuclear family household in which their father had the headship than from an extended or multiple family household with their grandfather or uncle as the head.67 Studies on multiple family households in Bangladesh, for instance, show that the presence of a father’s married brothers in the same household, and therefore increased competition with cousins for limited resources, posed a challenge and has been shown to decrease the well-being of the children.68 The more children present in the household, the fewer resources were allocated to any one child. Unfortunately, we cannot test the prevalence of malnutrition of children from different social strata and household 66 68

P.Petr. 2.42a from 250–249 bce. Cf. Lewis 1986: 45. Angel and Tienda 1982.

67

Edlund and Rahman 2005.

Balancing benefits and obligations: parents

73

backgrounds for Roman Egypt, as has been done for modern developing countries. It might indeed have been the case for Roman Egypt also that as long as the father was in his best years and able to work and provide for his family, living in a nuclear household was more beneficial for an individual child. As has been shown elsewhere, however, in the case of a father’s death, living in a multiple family household was more advantageous. Adult male relatives, such as an uncle, grandfather, or adult brother, could significantly mitigate the effects of orphanhood and replace the father to some degree: they took over the roles as guardians, managed the orphans’ property, and protected them in their vulnerable social position.69 More than one-fifth (21 percent) of all children and teenagers under the age of 18 recorded in the census returns lived with a single parent without any other relatives. Living with a single mother was only slightly more common (12.2 percent) than living with a single father (8.8 percent). Onethird of all children lived in an intact nuclear family with both father and mother, sometimes alone, but in general with one or more siblings (32.2 percent). On the other hand, two-fifths of all children grew up in large households. They lived in multiple family households with their parents, their parents’ married and unmarried brothers and sisters, and the latter’s children (24.4 percent). Even three-generational households with one or both grandparents still alive were not uncommon (15.3 percent). Much smaller fractions are obtained for other configurations: 1.3 percent of all children lived with an adult sibling without any parents or grandparents; 1.3 percent lived with an adult married cousin; 1.9 percent of all children lived together with minor siblings without any family member above the age of 18 and a not inconsiderable number, 2.8 percent, lived without any close relatives. Adoption contracts as a source for intergenerational obligations We can also learn a great deal about what people considered to be parental obligations from the adoption contracts that have come down to us from later Roman Egypt, since the main aim of adoption was the establishment of a relationship that resembled that of natural parents and children. From the content of these contracts it becomes clear that Roman law was not the underlying basis for adoption in Roman Egypt. Neither was an imperial script sought nor did the prefect intervene, as was obligatory for a valid 69

Huebner and Ratzan 2009b; see especially the contributions by Scheidel, Huebner, and Harders in this volume.

74

Balancing benefits and obligations

adoption according to Roman legislation.70 Our earliest surviving contract of adoption comes from the first half of the fourth century ce,71 and it is from this and later evidence that we are able to gain more information on the procedure of adoption in Egypt. The adopting parent pledged to raise the child, provide him with food and clothing and an education, and make him his lawful heir.72 The contract in addition stipulates that the adoptive father was forbidden to disavow his adoptive son or even reduce him to slavery. In another surviving contract, an adopting father declared his responsibility for “feeding and clothing the child in decent and appropriate fashion as my own legitimate and physical son.”73 Another adopting couple assured the mother who gave away her daughter that they would take over the role and duties of genuine parents and agreed to care for all the girl’s needs, fulfilling the position of parents.74 No obligations of the adoptee, such as support and care in old age for the adopting party, are given in the contract; to secure maintenance in old age or a proper burial does not seem to have been the main aim of the adopters as it usually was in classical Athens.75 The adoption was formally carried out by the adoptive father by registering the child as his own lawful, legitimate son or daughter.76 Adoption was regarded as an entirely private affair based on indigenous Egyptian customs mixed with some elements of Greek law.77 Patria potestas, for instance, seems to have played almost no role in the conception of father–child relations. Adoption procedure in Hellenistic or Roman Egypt has not yet attracted much interest in historical scholarship, largely because it is rarely attested in our sources.78 Demographic analysis of the census population, however, suggests that adoption in Roman Egypt may have been more common than previously assumed.79 We will discuss the institution of adoption in more detail in Chapter 7.

70

71

72 75 76 78

79

Cf. CIust 7.47.2 from 286: “Imperatores Diocletianus, Maximianus. Impuberem, quem ad vicem naturalis subolis adrogare desideras, si hi, qui sanguinis necessitudine iunguntur, id ei expedire apud praesidem provinciae confirmaverint, filium habebis, ita ut bonorum tuorum quarta pars tam in postremo iudicio tuo, quam si a te emancipatus fuerit, ei praebeatur et super patrimonio eius idoneis fideiussoribus datis servo publico caveatur, ne sub copulandae adoptionis obtentu in facultates eius, quae ei diligenti provisione servandae sunt, inruas.” P.Oxy. 9.1206 from 335 ce. Also: P.Lips. 1.28 from 381 ce; P.Oxy. 16.1895 from 554 ce; P.Cair.Masp. 3.7305 from 568 ce. P.Oxy. 9.1206. 73 P.Lips. 1.28. 74 P.Oxy. 16.1895. Rubinstein 1993; for more literature, see Chapter 7 below. P.Oxy. 9.1206: ἀπογράψομαι αὐτὸν εἰς ἐμαυτοῦ γνήσιο[ν υἱόν]. 77 Cf. Chapter 7. Taubenschlag 1959: 261–321; Kurylowicz 1983: 61–75; Beaucamp 1990–2: 48–52; Krause 1994–5: iii.80–1. See now Huebner in press a. Cf. Huebner 2007, 2012c, in press a, and Chapter 7 in this study.

Balancing benefits and obligations: parents

75

Education and training for a craft While concrete information is lacking for other regions and periods of the Greek world, in classical Athens parents were obliged to provide their children, at least their sons, with some form of education.80 In Roman times only upper-class families could afford a higher education for their sons that comprised rhetoric, law, or philosophy and that equipped them with the knowledge for a career in law and politics. Members of the lower classes, that is, the vast majority of people, could not afford the considerable expenses such an education would have cost.81 Many boys from the lower classes attended primary school, but at least from the age of 12 or 13 they were expected to learn a trade or work full-time on the family farm in order to be able to contribute to the family income. Lucian of Samosata in Commagene, who lived in the middle of the second century ce, was only in his early teens when his father considered it best for him to leave school and start to learn a trade.82 Lucian was sent by his father to an uncle to learn stonemasonry, a craft already practiced by two uncles and a grandfather. Lucian tells us about the advantages an apprenticeship had in the eyes of his father and his friends in comparison to a higher education: To most of them, higher education seemed to require much labor, consid erable time, no small expense and an illustrious position, while our family fortunes were small and needed some quick assistance. If I were to learn one of the artisan’s crafts, first and foremost I would immediately earn a living from my craft and would no longer eat at home at my age. Indeed, in a short time, I would even delight my father by regularly bringing him income.83

This passage shows that as early as a child’s teens, roles between parents and children began to be reversed. While the father at the beginning of his son’s apprenticeship had to cover the costs of his education, food, clothing, and last, but not least, taxes, only a few years later, while still in his teens, the son was able to make his own living and even contribute to his parents’ household income. Nonetheless, as we have seen above, it took a son many years to pay back his parents the expense of raising him. In Roman Egypt, if in rural areas no adequate teacher was available, but parents placed emphasis on a formal education and had the money to pay for it, at an early age children were sent away for schooling to the next city. A 80

81 83

Plut. Sol. 22. Cf. Harrison 1968: 77. Jewish fathers were obliged to feed and clothe their children (Epistle of Aristeas 248) and provide them with an education (Philo, De spec. leg. 2.233). Cf. Cribiore 2001. 82 Cf. Bradley 1991a: 112. Cf. Bradley 1984: 40–1. Lucian, Som. 1–4; trans A. M. Harmon (Cambridge and London, 1921).

76

Balancing benefits and obligations

couple from Heracleopolis sent away their 9- and 12-year-old sons for schooling, while their 7-year-old daughter stayed at home.84 If the head of the family was a craftsman, his son often continued in his father’s trade. Not all fathers trained their sons themselves, however.85 We have evidence that a weaver sent out his three sons to other weavers to learn the craft, while he took in someone else’s son as an apprentice.86 The purple-dealer Aurelius Pachymios from early seventh-century Panopolites took in his nephew, the son of his wife’s brother, as an apprentice for two years and later planned on marrying him to his daughter.87 He apparently had no son, and his son-inlaw was probably destined to take over his business.88 Apprenticeship often meant that the child left his parents’ household and moved into his master’s home, a rite of passage to adulthood. Only if the master lived nearby was it possible for the young trainee to continue to live at home. The apprentice had the opportunity of on-the-job training, and the master enjoyed a source of cheap labor. Examples for apprenticeships from all over the ancient Mediterranean world are numerous.89 In the case of a transfer of craft knowledge from father to son or from one relative to another, the details of the apprenticeship were apparently agreed on verbally and informally. In other cases the fine points of the apprenticeships, such as duration, number of holidays, food and clothing, as well as the fee to be paid by the parents, were laid down in a contract that was signed by the master and the apprentice’s parents, as we know from surviving examples from Roman Egypt. Formal contracts were already common in the ancient Near East.90 Xenophon in fifth-century bce Athens recommended that “when a man puts his son out to apprenticeship to be trained, a contract should be made 84

85

86

87 89 90

215-He-1. The poet Horace was deeply grateful for his father’s loving endeavors to support his son’s higher education. Even though he was a former slave, he had been able to have his son schooled in Rome, as was worthy of an equestrian’s or senator’s son. He had managed to pay the teachers’ high fees and to equip his son with clothing and slaves, like his fellow students. Later in life, when he mixed in the best society, Horace did not apologize for his father’s low standing or feel ashamed of him. In this way he repaid his father through his deepest gratitude and praise (Hor. S. 1.6.65–92). For apprenticeship contracts in Roman Egypt, see Bradley 1991a: 103–24; Bergamasco 1995: 95–167; 1997: 7–26; 1998: 193–6. Apprenticeship contracts for his three sons: P.Mich. 3.170 from 49 ce; P.Wisc. 1.4 from 53 ce; P.Mich. 3.172 from 62 ce. For the weaver’s own apprentice, see P.Mich. 3.171 from 58 ce. CPR 4.23 from 608 ce. 88 Cf. P.Par. 21ter; SB 1.4503–5, 1.5285–6. Cf. Bradley 1984: 40–1. §189 of the Babylonian Code of Hammurabi from 2100 bce already committed artisans to teach their handicraft to their apprentices, otherwise they would face the loss of their apprentices to their biological parent (Meek 1969: 163–80). §200B of the Hittite law set the price a father was committed to pay to the instructor of his son who ought to teach him the craft of a carpenter, smith, potter, leatherworker, or fuller. And we have an apprenticeship contract from Nippur drawn up between a free man and his master, a bleacher (Gardiner 1953: 145–9).

Balancing benefits and obligations: parents

77

concerning what the son should know.”91 Children entered their apprenticeship around the age of 12,92 and apprenticeships lasted from a few months up to six years. We find apprentices in crafts such as weaving, forging, flute-playing, hairdressing, embalming, and nail-smithing. Around 60 percent of all apprenticeship contracts were entered between a father and the future employer. In all other cases the mother or a near relative who acted as legal guardian was responsible, presumably because the father had already died.93 Parents were not allowed to take away their child before the date set by both parties and had to pay a fine if they did. If the child was lazy or refused to do the work assigned to him by his master, he had to pay a fine or stay as many days longer as he had been negligent and lazy. The master, however, had likewise to pay a fine if he failed to instruct his apprentice within this agreed period.94 The census returns from Roman Egypt record several teenage boys who by working in a trade contributed their share to their parents’ household finances. All of them were already above the age of 15 and most probably had already completed their apprenticeship.95 We encounter a 16-year-old linenweaver, a 17-year-old doctor, a weaver, a workman, and a scribe of the same age, an 18-year-old wool-carder, a 19-year-old goldsmith, and a 19-year-old workman.96 The children who entered such an apprenticeship in Egypt originated from the lower social strata. Girls were employed under the same conditions, but seem to have been trained mainly in weaving and shorthandwriting, and do not seem to have been freeborn.97 We have one contract in which a father agrees to apprentice his freeborn daughter for four years to a weaver in return for a loan of 400 silver drachmas. Aurelia Aphrodite, the girl in question, was to be fed and clothed during this period by her master. In return she was obliged to work for him without receiving a wage and was not allowed to leave his house, not even for a day or a night. If Polydeukes, the father, was not able to repay his debts, she had to stay with her master (and presumably had to continue working for him without receiving a wage).98 While the arrangement that the parents were not allowed to take 91 94 96 97

98

Xen. Equest. 2.2. 92 Bradley 1991a: 108. 93 Krause 1994–5: iii.7. e.g., P.Oxy. 2.275 from 66 ce. 95 Bagnall and Frier 2006: 72. Bagnall and Frier 2006: 72 n. 78. P.Mich. 5.346a from 13 ce; SPP 12.40 from 150 ce; P.Oxy. 14.1647 from 239 ce; P.Mich.inv. 5191a from 271 ce. Compare also P.Wisc. 1.5. For a discussion, see Bradley 1991a: 107; Rowlandson 1998: 267–8. P.Oxy. 67.4596. Cf. P.Oxy. 31.2586 from 253 ce: the father Hermias also received 400 drachmas from the master Dioscorus the linen-weaver, and promised to return the sum at the end of his son’s apprenticeship. See also BGU 4.1124 from 18 bce: the document speaks about a loan of 100 drachmas.

78

Balancing benefits and obligations

away the apprentice from the master’s house was common in apprenticeship contracts, the fact that Aphrodite was free to return to her parents’ house only if her father was able to repay the 400 drachmas at the end of this four-year period was rather exceptional. In this arrangement the girl seems to have served as a pawn for the debts of her father. Polydeukes must have been in considerable financial difficulties. It is not hard to imagine what status the girl would have held in her master’s household, if her father was not able to repay his debts at the end of the four-year period. This arrangement therefore resembled rather a sale of one’s child than an apprenticeship, the daughter serving as a security for the loan’s repayment.99 Divorce and child maintenance support If we believe the census returns from Roman Egypt, divorce seems to have been quite common even in the lower social strata. When a marriage ended in divorce, the children in the majority of cases remained in their father’s household,100 while the divorced mother returned to her natal family to live with them or to remarry in due course.101 In the few cases in the Roman census returns from which we know details about the living situation of women who kept their children after the divorce, we find that these women were homeowners in their own right, not living with their parents or relatives.102 That means that women who took their children with them most probably did so only if they could offer them a place to live, and that the woman’s natal family or the second husband often were not willing to take in her and her children from a previous union. However, divorce did not absolve the absent parent from his or her obligations to provide for existing children. Whatever the living situation of the children, at least according to later Roman law the absent parent was nonetheless still responsible for their maintenance.103 A certain Dionysos from late second- or early third-century ce Oxyrhynchus, for instance, 99 100 101 102

103

For an alternate interpretation, see Hengstl 1972: 96–7; Bergamasco 1996: 174. Dig. 25.4.1.1 (Ulpian); cf. Bagnall and Frier 2006: 124–5. 145-Ar-4; 145-Ar-12; 159-Hm-3. Cf. for classical Athens: Isager 1981–2: 85; Golden 1990: 142. 145-Ar-1; 145-Ar-2; 187-Ar-29. Cf. for the place of residence for the divorced wife: Barker 1997; Huebner 2009a. We will return to this problem in Chapter 5. Dig. 25.3.5.14 (Ulpian). Cf. CIust 5.24 (from 294 ce): “Divortio facto apud quem liberi morari vel educari debent. 1 Imperatores Diocletianus, Maximianus. Licet neque nostra neque divorum parentium nostrorum ulla constitutione caveatur, ut per sexum liberorum inter parentes divisio celebretur, competens tamen iudex aestimabit, utrum apud patrem an apud matrem matrimonio separato filii morari ac nutriri debent.”

Balancing benefits and obligations: parents

79

pledges in his divorce settlement to provide for his children until they come of age.104 From a papyrus of second-century Neilopolis in the Arsinoite nome, we know of a couple that divorced not long after the birth of their son, Horion. The mother kept the little boy and her ex-husband reimbursed her for the costs of clothing and food.105 We have a similar case from late sixth-century Syene. After the divorce, the mother kept the little daughter, while the father gave his ex-wife four solidi for her maintenance. Later, however, the mother remarried and threw out her daughter, who then returned to her father. The daughter consequently sued her mother for those four solidi her father had once paid for her upbringing.106 In the same vein, if the children in case of their parents’ divorce stayed with their father, he could retain parts of his ex-wife’s dowry and use it for maintenance and education of their children.107 Governmental help for poor parents We have evidence from nearly every century of antiquity that under various circumstances parents were forced to sell their children.108 We do not hear who bought these children and it is in general assumed that slave-dealers did. Private charities occasionally tried to prevent parents from selling or killing their children. They provided poor families with money to encourage them to keep their children. The aim was that these children stay with their families. From an honorary inscription we know, for instance, that the well-known consul and orator Marcus Aurelius Cotta provided his freedman with enough money to be able to raise all the children born to him, feed them, provide his daughters with a dowry, and secure his son the rank of a military tribune in the imperial army.109 In the late second century Publius Licinius Papirianus, a leading man of Cirta in North Africa, donated a considerable sum of money to his home community, laying down that the interest this money would earn each year should be invested in the upbringing of 300 boys and 300 girls of the town. Each month the boys were to receive 2½ denarii and the girls, 2 denarii. To receive support children had to be at least 3 years old, on the one hand certainly because infants who were nursed cost little, and on the other hand to prevent money 104

105 106 107 109

P.Oxy. 6.906: ὁ δὲ Δ[ι]ογένης καὶ ἐπιχορηγή[σ]ε̣ ι ἀπὸ τοῦ νῦν τὰ [δ]έ̣ οντα τοῖς αὐτοῖς υἱοῖς αὐτῶν παρ᾽ αὑτῷ διαιτωμέν[ο]ις ἄχρι ἡλικίας. Cf. Taubenschlag 1936: 508. P.Strasb. 7.666 from 145 ce (cf. Krause 1994–5: i.128 n. 70, 251). P.Lond. 5.1731 from 585 ce; Rowlandson 1998: 79–80, no. 62. P.Oxy. 2.265 from 81–95 ce. 108 Cf. Hezser 2003: 1–26; Evans Grubbs 2010. CIL 14.2298 ILS 1949 (found on the Appian Way near Rome).

80

Balancing benefits and obligations

being squandered on children who were not expected to live beyond the early years of childhood. Boys received support until they were 15 years old, apparently the time when they were old enough and able to earn their own livelihood. Girls were supported up to the age of 13, the age at which many of them usually got married. Publius Licinius Papirianus wished that the duoviri of each year choose the children who should be supported.110 The monthly 2½ denarii corresponded to 10 drachmas per month or 120 drachmas per year, a sum that at least in Roman Egypt would certainly have been high enough to provide a child with the necessities of life.111 In 315 ce the emperor Constantine issued a law to prevent poor families in Italy from exposing or killing their children when they were unable to provide for their upbringing because of poverty. If a parent reported to the praefectus praetorio a child that the family was not able to raise, food and clothing were to be issued to them without any further delay.112 The program was financed by the imperial treasury and the emperor’s private account, and can be seen as a public measure to help poor people plan for their old age by supporting them financially to keep and raise their children. Public support of this sort was, however, no more than a drop in the bucket since it reached only a fragment of the entire population. In Egypt, only in Antinoopolis, the city founded and endowed with numerous privileges by the emperor Hadrian in remembrance of his drowned lover Antinoous, could parents, who had registered their child within thirty days after birth, ask for public sustenance for his or her upbringing.113 The money came from the imperial treasury and was not directed to the poor but came as part of a package of other privileges the citizens of Antinoopolis enjoyed, such as exemption from liturgies and the poll tax. No similar measures of public help for children in any of the other cities in Roman Egypt are known.

3.5 Balancing benefits and obligations: children Filial piety Now in return we want to ask which obligations children had toward their parents. In modern Western society the role of children in caring 110 111 112 113

CIL 8.1641 ILS 6818. Cf. above the calculations of maintenance costs for children in Roman Egypt. CTh 11.27.1 from 315 ce: “cum educatio nascentis infantiae moras ferre non posit.” P.Fam.Tebt. 33 from 151 ce; SB 16.12742 from 157 ce. Cf. Corbier 2001a: 65.

Balancing benefits and obligations: children

81

and materially supporting their aged parents has been minimalized and the latter’s welfare has become largely a community concern. Intergenerational interdependence has passed from the micro to the macro level. We pay into pension funds to finance today’s retirees and expect the next generation to do the same for us: we pay directly to maintain the younger generation and indirectly to maintain the older one. The question of equity between the generations thus comes down to developing fair mechanisms for redistributing waged income via taxation to the elderly in the form of pensions. However, this “contract” fails insofar as there are no written obligations, and no generation can be sure what to expect and if their expectation of reciprocity will not be disappointed in the future even if they fulfill their obligations today. People in antiquity had similar worries. They also feared that their efforts in their younger years, when they were providing for the very young and the old members of their families, would not be rewarded later in life. Today the risk is lower as it has shifted from the individual to society as a whole and the state has to suffer bankruptcy before all social security systems fail. In ancient times, however, it was not in fact unlikely that one would find oneself utterly without means in old age. For the most part, in societies in which poverty is constant and widespread and a general public social security system is not widely available, there are few alternatives to family members as a safety net in case of job loss, death of a spouse, sickness, or old age. For this reason, great importance is assigned to the bonds of responsibility and reciprocity between kin. In virtually all ancient societies children were the primary source of old-age support and were socially and morally expected to support their elderly parents.114 The fifth commandment, for instance, adjured children to honor their parents (Exodus 20:12). In texts from ancient Mesopotamia and New Kingdom Egypt it is made explicit that children had to give respect as well as economic support to their parents in their old age.115 In both Greek and Roman societies, filial piety was a normative factor that regulated intergenerational solidarity. It mattered not only in private circles but also in public. Filial behavior was seen as evidence of a person’s general trustworthiness, reliability, and moral worth. The doctrine of filial piety made lack of submission, respect, and obedience socially unacceptable. Disrespect and disobedience towards one’s parents and the refusal to provide them with food, clothing, and practical care in their old age had detrimental effects on one’s social integrity and

114

Cf. Parkin 2003: 203–35.

115

Cf. Stol and Vleeming 1998.

82

Balancing benefits and obligations

prestige. In classical Athens, for instance, a candidate for public office had to bring proof of fulfilling these obligations towards his parents.116 Aeschines gives us an explanation for this rule: “Because if a man is mean toward those whom he ought to honor as the gods, how . . . will such a man treat the members of another household, and how will he treat the whole city?”117 The story of Aeneas and Anchises, or the founding legend of the temple of Pietas in Rome, in which a mother was imprisoned and kept alive by her own daughter’s breast milk, embodied pietas as unconditional filial love in Roman society. In a letter to Atticus (176.2) Cicero stated that “to refuse to maintain one’s parents is wicked.” Plutarch stressed that nature and law “have assigned to parents, after gods, first and greatest honor (τιμήν)” and “there is nothing which men do that is more acceptable to gods than with goodwill and zeal to repay favors to those who bore them up.”118 While there was a public prosecution foreseen in Attic law for neglecting to care for one’s old parents (γραφή γονέων κακώσεως),119 in the Roman world it was only from 161 ce on that support of one’s old parents was explicitly called a legal duty in an imperial rescript: “The competent judge will order that you be supported by your son, if he has the means to provide you with maintenance (alimenta).”120 Roman law from Ulpian to Justinian emphasized the importance of filial obligations toward elderly parents. Despite their unequal power over their children, father and mother earned the same grade of respect and honor.121 Negligence of this filial duty was to be punished.122 The delinquent’s property could be seized and sold,123 and he himself could be put in prison.124 For illegitimate children the law held that the reciprocal obligation to support existed only between them and their mothers and maternal relatives.125 Obedience and affectionate 116

117 118

119

120 121 122 124

125

Aristot. Ath. Pol. 55.3. Cf. Lumpe 1959; Lacey 1968: 116–18; Finley 1981; Gnilka 1983; Garland 1990: 244; 256–62; Strauss 1993: 43; Baltrusch 2003; Huebner 2005; Schmitz 2007: 35–6, 106. Aeschin. 1.28. Plut. Mor. 479f; trans. W. C. Helmbold, Plutarch’s Moralia, vol. vi, Loeb Classical Library (Cambridge, MA, 1939). For attitudes and expected behavior of children toward parents in Roman elite circles, cf. Eyben 1993. Diog. Laert. 1.55; Isaeus 8.32. Cf. Lacey 1968: 116–18; MacDowell 1978: 92; Strauss 1993: 44; 65. For the evidence for Delphi, see esp. Rupprecht 1998: 228–9 n. 32. CIust 5.25.2 from 161 ce; cf. also CIust 5.25.1. Dig. 27.10.4. Cf. Dig. 37.15.1.1–2; CIust 8.46.4 from 259; NovIust 155.1. See also Saller 1994: 113–14. Dig. 37.15.1.2, 37.15.2.pr., 48.2.11.1, 48.9.1, 48.9.9.1, 1.18.4. 123 Dig. 25.3.5.10. Quint. Inst. Orat. 7.6.4–5: “Let children maintain their parents, or be put in prison; for an infant will surely not be put in prison. From this exception there will be a possibility of proceeding to others, and of making a distinct inquiry whether everyone who does not maintain his parents is to be put in prison.” Trans. J. S. Watson (London, 1856). Kaser 1975: ii.156.

Balancing benefits and obligations: children

83

devotion were favorite qualities parents looked for in their children because this ensured support in old age.126 Care for the elderly appears several times in the papyri from Roman Egypt. The terms used for old-age care are γηροβοσκεῖν,127 γηροβοσκία,128 γηροκομία,129 or γηροτροφία.130 Old-age support in Roman Egypt could take different forms. In a peasant family sons likely assisted their father in farming, and with his continually decreasing health and strength he would need help more often. Craftsmen or small traders would have liked their sons to eventually take over their business after training them and working with them for several years. In environments where we find a high percentage of out-migration for the male working group, practical care and co-residence could have been compensated by sending money or other forms of material support. Those elderly individuals who were wealthy enough not to be in the position where they needed their children’s financial support still most probably expected them to be nearby to help in time of illness and disability. The simple assurance that the children, if away from home, had not forgotten about their old parents was sometimes enough for their contentment. The old Hikane from first-century ce Berenike wrote a letter in a highly vexed tone to her son Isidoros, who was away from home, probably for business reasons. Hikane begged her son to think of her and send her a letter showing her that he had not forgotten about his old mother at home: “Was it for this that I carried you for ten months and nursed you for three years, so that you would be incapable of remembering me by letter?”131 A daughter was also expected to repay her parents the favor of an upbringing by her care and support later on.132 A funerary inscription for a 20-year-old girl called Lysandre erected by her parents in Roman Karanis shows the grief and despair of parents whose child had died too early: What profit is there to labor for children, or why to honor them above all else, if we shall have for our judge not Zeus, but Hades? My father took care of me for twice ten years, but I did not attain to the marriage bed of the wedding chamber . . . The breasts of my mother nourished me with their 126 127 128 129

130 131 132

Sen. Helv. Matr. de consol. 2.4; Sen. Ben. 3.38.2; Sen. De ira 2.21.2–6; Quint. Inst. Orat. 6, preface 6–11. BGU 7.1578 after 212 ce; P.Cair.Masp. 3.67314 from 569/70 ce. P.Oxy. 9.1210 from the first century bce to the first century ce. SPP 1, pp. 6–7, no. 1, ll. 21–2, from 454 ce: καὶ τοῦ φρέατος, δέξασθαι τὴν αὐτήν μου γαμετὴν ἥνπερ ταύτῃ πολλὰ εὐχαριστῶ ἕνεκεν εὐεργεσιῶν τ[ε] καὶ γηροκομιῶν. See also SB 1.4845. P.Flor. 3.383, l. 38, before 223 ce. P.Berenike 2.129; trans. Bagnall and Cribiore, 2006: 169–70. Herodotus contends that – by contrast to Greece – in Egypt daughters were obliged to care for their elderly parents, while Egyptian sons cared for their parents only if they chose to do so (Hdt. 2.35). Cf. Taubenschlag 1936: 509; 1956: 175; Seidl 1966.

84

Balancing benefits and obligations milk to no purpose at all, and to those breasts I cannot repay the favor of nourishment for their old age . . . Weep for Lysandre, whom Philonike and Eudemos bore in vain.133

In early third-century Philadelphia in the Arsinoite nome, an old veteran brought a claim against his negligent daughter, Longinia, since she failed to provide him with the necessities for his old age.134 The elderly father accused his daughter of ingratitude and impiousness.135 Even though he had given her a generous dowry upon her wedding, she neglected to care for him in his old age. Siblings encouraged each other to look after their elderly parents. While a certain Sarapion was dealing with legal proceedings at the prefect’s court in Alexandria, he wrote to his brother Dorion, who apparently resided on the family estate back home in the countryside: “Before all else take care of your health. Look after Demetrous and our father Dorion.”136 Demetrous could have been Sarapion’s wife, or a sister, or even Dorion’s wife. Dorion the elder apparently lived with or close to his adult sons but his sons had taken over his business. Another man, Sempronious, who was also traveling for business, wrote home to his brother: “for we ought to reverence her who bore us as a god, especially when she is so good. This I have written to you brother, since I know the sweetness of dear parents.”137 Providing care for one’s elderly parents seems to have been also the precondition for succession to their estate. A certain Aurelia Techosis from late third-century Oxyrhynchus stresses that she had fulfilled all of her obligations to her old mother. Her mother had apparently been living alone after her husband’s death. When she fell ill, her daughter moved from the Small Oasis to Oxyrhynchus, where her widowed mother lived, to take care of her: “My mother was stricken with illness, and I in the goodness of my heart nursed and tended her and was assiduous in performing what is owing from children to parents.”138 And when her mother eventually died, Techosis provided for her funeral and “did all that was fitting on the 133

134 135 136 138

I.Métr. 83: τί πλέον ἐστ’ εἰς τέκνα πονεῖν, ἢ πρὸς τί προτιμᾶν, εἰ μὴ Ζῆνα κριτὴν ἕξομεν̣, ἀλλ’ Ἀΐδην; δὶς δέκα γάρ μ’ ἐκόμησε πατὴρ ἔτη, οὐδ’ ἐτέλεσσα νυμφιδίων θαλάμων εἰς ὑμέναια λέχη . . . μαστοὶ μητρὸς ἐμῆς κ̣ ε̣ νεὸν γάλα τοί με ἐκόμησαν, οἷς χάριν οὐ δυνάμην γηροτρόφον τελέσαι . . . κλαύσατε Λυσάνδρην, συνομήλικες, ἣν Φιλον[ί]κη καὶ Εὔδημος κούρην τήν ἐτέκοντο μάτην. For a translation, see Rowlandson 1998: 347, no. 273. We find similar examples in the Anthologia Palatina, e.g., 7.261: Τί πλέον εἰς ὠδῖνα πονεῖν, τί δὲ τέκνα τεκέσθαι, ἣ τέκοι εἰ μέλλει παιδὸς ὁρᾶν θάνατον; ἠιθέῳ γὰρ σῆμα Βιάνορι χεύατο μήτηρ· ἔπρεπε δ’ ἐκ παιδὸς μητέρα τοῦδε τυχεῖν. BGU 7.1578. BGU 7.1578, l. 7: ἀχαριστίας; ll. 8–9: κα[τ]ὰ ἀσέβειαν ὑπò τηˆ ς θυγατρóς μου Λογγεινί [ας ?]. P.Oxy. 2.294 from 22 ce. 137 SB 3.6263 from the second century ce. P.Oxy. 8.1121 from 295 ce, ll. 8–12: ἡ προκειμένη μου μήτηρ Τεχῶσις νόσῳ κατα[β]λ̣ [η]θεῖσα κατὰ τὴν ἐμαυτῆς μετριότητα ταύτην ἐνοσοκόμησα καὶ ὑπηρέτησα καὶ οὐκ ἐπαυσάμην τὰ πρέποντα γείνεσθαι ὑπὸ τέκνων γονεῦσι ἀναπληροῦσα.

Balancing benefits and obligations: children

85

occasion of her death.”139 For this reason, Techosis claimed that she had every right to her late mother’s property.140 When the only son of Aurelia Maria died prematurely, a niece took over the late son’s duties for her elderly widowed aunt. She cared for her and promised to arrange for her burial. In return, she received from her aunt a donatio mortis causa comprising the widow’s household goods.141 A certain Aurelia Ammonia from mid-fourth-century Oxyrhynchus left her fuller’s workshop and other shops to her two sons as a donatio mortis causa. She explicitly stressed the devotion that her younger son had shown to her. He had remained with her since the death of his father, who had died a long time ago, worked at the fulling trade, and did not abandon her, but treated his mother kindly. Ammonia probably mentioned these good deeds of her younger son as the explanation why she gave away her property in equal shares to both her sons, the elder son not receiving, as customary, a larger portion.142 Sometimes parents chose one of their children, who then took over the sole responsibility to care for them, as in the case of a firstcentury Oxyrhynchite family with several sons.143 It has been argued that the sons who were chosen to care for their elderly parents were exempt from the poll tax.144 This assumption rests, however, only on this single papyrus, and we do not even know if the tax exemption was granted. Another question was the duration of time the elderly parents required this support. As there was no retirement in antiquity, people continued working as long as they were able to do so. The elderly thus posed less of a burden on the younger generation than they do today in the contemporary West, where widespread and compulsory retirement is usually set at the age of 65 but is scheduled to increase over the coming decades. With an average life expectancy of 80 years and more, it is not uncommon for individuals to live and receive pensions for over twenty years after they retire. In antiquity, on the other hand, less than 5 percent of the population even attained the age of 65. And not only do more people grow old today, many more need a 139

140 142

143

144

ll. 14–15: πάλιν τὰ πρὸς τὴν κηδείαν αὐτῆς παρέσχον καὶ τὰ καθήκοντα ἐπὶ τῷ θ̣[α]ν̣άτῳ ἐξετέλεσα. English translation by Winter 1933: 110–11. 141 P.Mu¨nch. 1.6 from 540 ce. P.Coll.Youtie 2.83 from 353 ce, ll. 6–10: ἐπειδὴ σύ, ὁ Διονύσιος, ἐκ [μ]ακρόθεν ἀπὸ τῆς τελευτ̣ ῆ̣ ς̣ τοῦ σοῦ πατ[ρ]ὸς παρέμεινάς μοι τῇ μητρὶ καὶ ἐργαζόμενος τὴν τῶν κναφικῶν τ̣ έ̣ χνην καὶ μὴ καταλείψας με τὴν μητέρα εὐχαριστῶν, ὁμολογῶ κα̣[τα]γεγραφέναι ὑμῖν ἀμφοτέροι[ς] τ̣ οῖς τέκνοις μου κοινῶς ἐξ ἴσου μέρους πάντα ὅσα κατεσκεύασα. English translation by Rowlandson 1998: 279, no. 217. For the eldest son’s privileges, cf. section 2.10. P.Oxy. 9.1210 from the first century bce/first century ce: ἐπιλελεγμένων ὑπὸ τῶν γονέων εἰς γηροβοσκίαν ἀφ᾽ ὧν ἔχουσι υἱῶν. Wallace 1938: 120; Taubenschlag 1956: 179; El Mosallamy 1994: 495; Rupprecht 1998: 226.

86

Balancing benefits and obligations

high degree of support for a significantly longer period of time. A state of utter helplessness and the need for intensive care was certainly much rarer in antiquity than today given the absence of effective medication or lifesustaining measures. In antiquity severe illness often led to a quick death. In sum, clearly in Roman Egypt to care for one’s aged parents and provide for their proper burial was thought to be the repayment of the favor of nourishment in earlier years.145 Despite the general consensus on filial obligations there was, however, no exact definition of what children were expected to do.146 Were economic sustenance payments regarded as sufficient, or were children also required to live nearby or in the same household and provide practical care if needed? The details of care seem to have been rather a question of negotiation between the parties in each individual case. The precise application of the norms was rooted in specific circumstances. Repayment of education The following case from the late third-century bce Arsinoite nome clearly reflects Greek tradition being upheld in Ptolemaic Egypt. The old Pappos from Ghoran denounced his son Strouthos, who neglected to pay him the monthly subsistence of 1 artaba of wheat and 4 copper drachmas, even though, as Pappos claimed, he had provided him with an education and taught him grammar.147 In addition, Strouthos physically and verbally abused his old father, whose eyesight was already deteriorating, and tried to take away the latter’s furniture, or at least this is what his father reported to the authorities.148 The son eventually pledged himself to support his father by 2 copper drachmas per month, easily not enough to provide for his father’s living.149 Another petition from the same era shows us that sometimes daughters received some form of formal education and were also held liable to return the favor and to support their elderly parents. In this petition from Magdola in the Arsinoite nome, an old father named Ktesikles complained to the king about his ungrateful daughter, Nike.150 Ktesikles declared that he had brought up his daughter and provided her with an 145 147

148

149 150

I.Métr. 83; Rowlandson 1998: 347, no. 273. 146 Cf. Rupprecht 1998: 229–31. P.Enteux. 25 from 225 bce, ll. 1–2: ἀδικοῦμαι ὑ]πὸ Στρουθοῦ [τοῦ υἱοῦ μου. . . ἐμοῦ γὰρ δι]δάξαντος αὐτὸν τὴν .[ . . . . . . . . . . . . κ]αὶ τὴν γραμ[ματικὴν - - ]. P.Enteux. 25, ll. 7–9: ὅταν ποτέ μοι ἀπαντήσῃ λοιδορεῖ με τά αἴσχιστα, καὶ ἀποβιαζόμ[ενος - - ]δ̣ọν εἰς τὴν οἰκίαν μου ἁρπάζει μου ἀεὶ τὸ πρόχειρον τῶν σκευῶν, καταφρονῶν μου ὅτι πρεσβύτερός εἰμι καὶ ἀσθενῶ τοῖς ὀφθαλμοῖς. Cf. Taubenschlag 1956: 510; Rupprecht 1998: 230; Cribiore 2001: 107–8. For further literature about prices in Ptolemaic Egypt, see Rupprecht 1998: 230 n. 36. P.Enteux. 26 from 221 bce; cf. Cribiore 2001: 106; Bagnall and Derow, 2004, 245–6, no. 152.

Balancing benefits and obligations: children

87

education, but now that he was old and his eyesight was deteriorating, she refused to support him.151 Ktesikles had already taken legal action against her eight years earlier, and had obtained her written oath that she would provide him with 20 drachmas each month. Nevertheless, his daughter neglected to fulfill the contract over the following years because a certain dancer, Dionysios, “seduced” her.152 Then Ktesikles petitioned the king that he might order the strategos to start an investigation and, should Ktesikles be telling the truth, compel his daughter to fulfill her obligations. It is interesting in this case that Nike was compelled to support her old father in the form of money. The sum of 20 drachmas per month was quite a considerable amount that certainly would have paid for his entire maintenance. These two old men, Pappos and Ktesikles, clearly based their claim on Greek law. We do not have comparable cases for Roman Egypt in which a mother or father brought the argument that providing the child with an education in younger years entitled him or her to support in old age. Post-mortem obligations At least as important for the old as the provision of practical care, food, and clothing was the assurance that someone would care for their proper burial. The Athenian γραφή γονέων κακώσεως referred not only to proper care for one’s old parents but also to the duty to bury them. Who received the inheritance, not necessarily the closest kinship ties, was decisive for the assignment of this duty.153 According to Roman law, the expenses of a burial had to be met by one’s heirs. If the deceased was too young or a slave, the father or his master had to pay for it. A man was allowed to retain from his wife’s dowry the amount that he had spent on her funeral before he returned it to her family.154 Petersen has convincingly argued that the epithets φιλάδελφος, φιλότεκνος, φίλανδρος, and φιλοπάτωρ bestowed to the deceased person in his or her grave inscription give us an indication of

151

152

153 154

P.Enteux. 26, ll. 2–4: ἐμοῦ γὰρ ἐκθρέψαντος τὴν ἐμαυτοῦ θυγατέ[ρα] καὶ παιδεύσαντος καὶ ε[ἰς ἡ]λικίαν ἀγαγόντος, ἀκληρήσαντος δέ μου κατὰ τὸ ἴδιον σῶμ[α] καὶ τοῖς ὀφθαλμοῖς ἀδυνατοῦντος, οὐχ οἵα μοι ἦν ἐπαρκεῖν τῶν ἀναγκαίων οὐδέν. P.Enteux. 26, ll. 8–10: [νῦν δὲ φθαρεῖσα ὑπὸ Διονυσ]ίου, ὄ̣ ντος κιναί[δο]υ, οὐ π̣[οιεῖ] μοι τῶν κατὰ τὴν χειρογραφ[ί]αν οὐδ[ὲν] καταφρονοῦ[σά μου διὰ τοῦ γ]ήρως κ[αὶ τ]ῆς ὐπ[αρ]χούσης μ̣ οι ἀκληρίας. Cf. Isaeus 2.2–11. See also Kreller 1919: 32–3. Fontes iuris Romani anteiustiniani, ed. S. Riccobono and V. Arangio Ruiz, Florence, 1940–3 (FIRA), 2.334–5; Paulus, Opinions 1.21.2–5, 8–14.

88

Balancing benefits and obligations

who set up the funerary stela and who took care of the grave site. Petersen argues that φιλάδελφος had the meaning “dear to brother (or sister)” rather than “fond of brother (or sister).”155 In one of the earliest surviving Greek papyri from Memphis from the second half of the fourth century bce, a certain Artemisia cursed her husband, who had neglected to care for the burial of their deceased daughter and perform the funeral rites. In retaliation, she wished that her husband would not receive funeral rites either from his children, and would not be able to bury his own parents, indubitably one of the worst curses.156 In a Demotic papyrus from the early third century bce we learn about a mother who left her son a variety of property under the condition that he fulfilled three duties, one being that he took care of her proper embalming and burial.157 We notice little change from Ptolemaic to Roman times in Egypt. In the year 211 ce a certain Isidoros from Hermopolis arranged with his two brothers, Hermeinos and Theognotos, that, if he waived his right to succession to their paternal estate, only his brothers would be liable to pay for their late father’s debts, support their old mother, feed and clothe her, and provide her with the accustomed burial offerings after her death.158 The widow Aurelia Maria from mid-sixth-century Syene bestowed a gift on her daughter in return for her services in her old age and, apparently even more importantly, because she stressed it in the text three times, for proper burial offerings after her death.159 She had formerly entrusted one of her sons with this gift and duty but he had predeceased her. Now the daughter was to receive this special share of Maria’s property. The remaining maternal estate, however, was to be split on an equal basis among all of Maria’s children upon her death. Also, the previously mentioned Techosis from late third-century Oxyrhynchus stressed that she had performed the customary funeral rites for her mother and thus was entitled to receive the inheritance.160 Funeral expenses could be considerable: for the tomb itself and the sacrifices, for the pyre, cakes, and grave clothes, and for a so-called “gift to the corpse.”161 Not only a proper burial but also the performance of the customary rites at the grave were considered filial duties. These customary rites included commemorative visits to the grave, wreathing, ribboning, and 155 158

159 160

Petersen 1964: 159. 156 UPZ 1.1; cf. Rowlandson 1998: 63, no. 37. 157 Glanville 1950. P.Lond. 3.932. The three brothers had a sister, who is not mentioned as heir to her parents. She might have already received her dowry as a form of pre-mortem inheritance (cf. Sijpesteijn 1989: 214). P.Mu¨nch. 1.8; P.Lond. 5.1857; cf. Farber 1996: 455–7 (D23). P.Oxy. 8.1121 from 295 ce, ll. 14–15. 161 Cf. Champlin 1991: 184–6.

Balancing benefits and obligations: children

89

anointing the grave stela, animal sacrifice, and drink and food offerings.162 We know from the papyri that it was customary that relatives and friends send food to the bereaved family for the deceased’s sacrifice.163 In sum, proper filial behavior was understood as kind treatment and economic support of one’s elderly parents as long as they lived, and provision with a proper burial and observation of the customary grave rites after their death. Small communities and tightly woven neighborhoods in which everyone saw and knew everyone made it almost impossible to evade this social pressure. The public demonstration of such filial loyalty enhanced one’s social reputation, fostered one’s career, and secured one’s rightful inheritance. Limits of filial obligations Yet, filial obligation had its limits. Solon according to tradition laid down a law that if parents had not provided their son with an adequate education, had sent their daughter or son into prostitution, or sold them as a slave, filial obligations expired.164 That is, when parental obligations had not been met, filial obligations likewise ended.165 Children who had been disinherited without apparent cause were exhorted by the law to start proceedings for unduteous will (querella inofficiosi testamenti).166 Parents were also morally expected to nominate their children as heirs, and disinheritance was accepted only when children had failed to render the

162 163

164 165

166

Garland 1985: 104–20. BGU 3.801 (second century ce): “sweetmeats, 160 in number, and 10 pine-cones for the sacrifice in her honor”; SB 14.11646 (first/second century ce): “one hundred walnuts.” While pagans continued to celebrate the birthday of their family members even after their death, coming together at their tombs for a ritual banquet, for Christians it was the day of death of their beloved when they met at his grave (Petersen 1964: 167). From Roman Asia Minor, especially from the Lydian-Phrygian borderland, we have evidence that those post-mortem duties could also include the expiation of the deceased’s sins. The so-called “confession inscriptions” from Lydia and Phrygia dating mostly to the second and third centuries ce were erected not only for oneself or living family members: we see that wives, husbands, sons and daughters, and also foster children erected stelas for deceased family members (Petzl 1994). They were erected to reconcile the gods rather than to confess a sin, as Rostad has shown, and should therefore rather be called “reconciliation inscriptions” (Rostad 2002). Plut. Sol. 22. Cf. Taubenschlag 1936: 509; 1956: 175. Alexander Severus issued an edict in 227 that children were not obliged to pay back all the expense their mother had incurred when she had acted out of maternal lavishness (CIust 2.18.11 from 227 ce: “Alimenta quidem, quae filiis tuis praestitisti, reddi tibi non iusta ratione postulas, cum id exigente materna pietate feceris. Si quid autem in rebus eorum utiliter et probabili more impendisti, si non et hoc materna liberalitate, sed recipiendi animo fecisse ostenderis, id negotiorum gestorum actione consequi potes”). Cf. Gardner and Wiedemann 1991: 121–5.

90

Balancing benefits and obligations

appropriate duty.167 On the other hand, children who had not shown the expected obedience and respect, and had behaved with ingratitude towards their elderly parents, were not entitled to inherit anything from them when the latter had explicitly wished this to be so. If their parents had disinherited them, the burden of proof was on the children; they had to bring evidence that they had rendered the expected duty to their parents.168 It was also acknowledged that some parents were led astray and wrongfully disinherited their own children, pressured for instance by a second spouse. In this case the will was to be declared void because it contradicted the claims of filial and parental piety.169 From 213 ce on children could no longer attack their father’s will as unduteous if their father had left them at least the fourth part of his estate.170 Nephews and nieces, uncles and aunts had no right to claim for unduteous wills if they were passed over in a will; only children, parents, and siblings from the same father171 were allowed to start proceedings for unduteous will if they were not left at least this legal portion.172 To bequeath children unequal shares if all the children had behaved dutifully and showed the expected filial piety and devotion to their parents was stigmatized and regarded as unduteous.173 To bequeath equal shares to all the children, especially to all the sons even if they were from different marriages, was praised as righteous parental piety. Impartiality towards one’s children was expected.174 For Roman Egypt we have a case from the second half of the sixth century of a father disinheriting his own children, accusing them of attempted parricide.175 The singularity of this source points, however, to the fact that parents only very rarely disinherited their own children: Proclamation of disownment and rejection, having my mind and under standing unaffected, with true and unerring judgment, without any guile or fear or violence or compulsion or deceit, in a public place of business. And this I transmit to my parricidal children, though children in name only . . . thinking to find you helpful in all things, a comfort to my old age, submissive and obedient, and on the contrary you in your prime have set yourselves against me like rancorous things, as I learned through experience of your 167 168 170 171

172 174 175

CIL 6.10299, ll. 1–2 FIRA 3.48. Cf. Eck 1978: 277–9; Champlin 1986: 253–5. CIust 3.28.28.pr. from 321 ce. 169 Dig. 5.2.2–4. Dig. 35.2.1.pr., a limit already set in the lex Falcidia from 40 bce. Half-siblings of the same mother were barred from bringing proceedings against the will of a brother or sister (CIust 3.28.27 from 319 ce). CIust 3.28.21 from 294 ce. 173 Valerius Maximus 7.8. Cf. P.Oxy. 4.716 from 186 ce, P.Oxy. 8.1208 from 291 ce, and section 2.11. P.Cair.Masp. 3.67353 v Sel.Pap. 1.87 from 569 ce. This disinheritance document is part of the famous Dioscorus archive.

Balancing benefits and obligations: children

91

heartless parricidal conduct and lawless disposition, seeing that I fell griev ously ill through you . . . and it is no longer lawful for you in future to call me father, inasmuch as I reject and abhor you from now to the utter end of all succeeding time as outcasts and bastards and lower than slaves . . . for ravens to devour the flesh and peck out the eyes, in this manner I debar you from receiving or giving anything on my behalf, whether I am alive or dead, because I have rightly and justly thus resolved.176

176

Trans. by Bowman 1986: 132.

chapter 4

Widowhood, remarriage, and residence patterns

You scorn those whom you bore, you marry when you are infertile, you vindictively upset the proper order of a will, nor do you blush to hand over your entire patrimony to a man to whose already corpse like body you submitted your own withered old age. Valerius Maximus 7.7.4

4.1 The chaste widow in cross-cultural perspective To lose one’s spouse to death in the early years of marriage was almost as common in all societies before the demographic transition as it is today to have a divorce. In accordance with the age differences between husbands and wives, many more women lost their spouses to death than vice versa. While in modern Western societies most widows are elderly (over twothirds of widows in Western Europe are older than 65),1 a significant proportion of women in earlier centuries lost their spouses while they were still in their twenties and thirties. The chaste widow who refrains from remarriage after her first husband’s death is a nearly universal paradigm of female virtue across societies. When studying the likelihood of remarriage after the loss of a first spouse, we must take care to distinguish between the ideal and reality. We meet the chaste widow much more often among the elite than among the broad masses of the respective societies; the British anthropologist and historian Jack Goody sees economic factors as the main cause for this discrepancy.2 Goody concludes that poor widows without further support from family members were forced to remarry because they were in dire need of an economic provider for them and their children. Among the elite, however, concerns about property divisions actually kept wealthier widows, at least those who had children already, from remarrying.3 To give an example, in China in the 1

Palmore 1987: 96, table 2.

2

Goody 1990: 202.

92

3

Goody 1990: 205.

Widowhood and remarriage in Graeco-Roman society

93

early modern period the widow was theoretically expected to remain loyal to her late husband, devoting her life to the worship of his spirit. The literary celebrated ideal was the widow who disfigured herself in order to avoid remarriage, or committed suicide as an act embodying the highest Confucian convictions.4 As we should expect, however, these customs reflect only the ideals and conventions of the social elite, a very small proportion of Chinese society. Remarriage rates for the lower social strata were high.5 In India far into the nineteenth century the ideal of the chaste widow escalated in the ideal of self-immolation, the widow following her husband’s dead body on the funeral pyre, an act regarded as the ultimate proof of wifely loyalty and devotion. The widow who could not bring herself to die on her husband’s funeral pyre was destined to lead the life of a social outcast.6 However, this practice was widespread only among upper-caste Hindus, who had banned remarriage for widows entirely. Widows of lower castes in India remarried in large numbers.7 For Ottoman Aleppo in northern Syria the same differences between the social strata have been noted. While relatively few widowed women in secure economic circumstances remarried after their first husband’s death – only about 5 percent of all the recorded cases – remarriage among poor widows was widespread.8

4.2 Widowhood and remarriage in Graeco-Roman society The ideal of the widow remaining faithful to her late husband also reigned in Greek and Roman society. Widows in classical Athens who decided not to remarry, but stayed with their children and raised them on their own, were highly praised as unselfish because they placed the welfare of their children higher than their own happiness.9 And in Rome the univira, the woman who had known only one man, was praised as the paradigm of female virtue and marital faithfulness. However, again, ideal and reality were not the same. Scholarly opinions about the frequency of remarriage in antiquity vary greatly. Thompson, for instance, claims that remarriage for divorced and widowed women was considered normal in classical Athens, and the incidence high.10 Gallant estimates, on the basis of information on remarriage in forensic speeches, that approximately one-third of all widows 4 8 9

10

5 6 7 See Waltner 1981. Johnson 1983: 14. Sangari and Vaid 1996. Nair 1996: 9. Meriwether 1999: 131–2. Compare, e.g., Dem. 27.13–15, 29.26; Lib. Or. 1.26, 58; Jo. Chrys. Sacerd. 1.2; Gr. Nys. Vit. Macr. 6, 11. Cf. Krause 1994–5: iii.130–45; Van Dam 2003: 102–3. Thompson 1972: 211–24. See also Isager 1981–2: 81–96; Just 1989: 66–7; Garland 1990: 339; Pomeroy 1997: 120.

94

Widowhood, remarriage, and residence patterns

remarried, many of whom had children from their first husband.11 We have several examples from fourth-century Athens where husbands even arranged in their wills for their wives’ subsequent remarriages.12 Among Romans the incidence of remarriage after a husband’s death is commonly regarded as a frequent phenomenon that served as a tool for personal and family advancement.13 Augustan marriage law even required that widows up to the age of 50 and widowers up to the age of 60 remarry within a short time span after the death of their spouse.14 Roman law, even in late antiquity under Christian influence, never discouraged remarriage. Its only concerns were preserving the property rights of the children from the first union.15 The question is to what degree did these laws, originally aimed at the elite of the city of Rome, apply to and affect the lower social strata, especially in the provinces?

4.3 Widowhood and remarriage in Roman Egypt The census returns from first- to third-century Roman Egypt are once again promising source material for the provincial lower social strata. As we have seen, these returns record the common people of Middle Egyptian villages and towns, most of them farmers, craftsmen, and small traders. Only very few of the recorded individuals can be regarded as belonging to the social elite.16 From the census returns it appears that women began to marry from age 12 onwards, and by their late twenties virtually all women seem to have married at least once.17 Most of the men, 69 percent of those 78 cases for which the ages of both husband and wife survived, were as old as or older than their wives, by 0–13 years, with the apex at 2–6 years. In 19 percent of cases the husbands were older than their wives by 15 years or more.18 In some cases women were married to men older by 20 years or more. The largest age gap between spouses recorded in the census returns is 31 years.19 In only 12 percent of the cases women were older than their husbands, in one case by as many as 8 years.20 This means that, on average, women were considerably younger than their husbands and had a good chance to outlive their 11

12

13 14 15 18

Gallant 1991: 54. We have fifty-three cases of remarriage in the oratorical sources, almost as many for women as for men. Cf. Hunter 1989; Günther 1993; Pomeroy 1997: 188. e.g., Dem. 30.7, 36.8, 57.41; Isaeus 2.7–9. Guardians appointed for the children were also favored candidates for marrying the widow (Dem. 27.5, 28.15, 29.43, 36.8). Cf. Pomeroy 1997: 169; Cox 1998: 90. Humbert 1972: 76–112; Bradley 1991b; Champlin 1991: 10–15; Krause 1994–5: i.2–4, 89–95. Rules of Ulpian 14; cf. Treggiari 1991b: 37–80; Evans Grubbs 2002: 83–7, 220. Cf. Clark 1994: 18. 16 Cf. Bagnall and Frier 2006: 72–3. 17 Bagnall and Frier 2006: 113. Bagnall and Frier 2006: 118–19. 19 117-Ap-7; cf. Bagnall and Frier 2006: 118–19. 20 187-Ar-4.

Widowhood and remarriage in Roman Egypt

95

husbands. The odds that a 15-year-old girl would become a widow within the first 15 years of marriage were over one in four.21 If this 15-year-old girl married a 35-year-old man, the likelihood to become a widow within the next 15 years was almost two in five (38 percent). The proportion of women who thus faced the choice between remaining a widow or remarrying at least once (and perhaps even more than once) in their lives must have been considerable.22 What happened to these women? Did they remain faithful to their late husbands? Were they financially supported by his family? What effect did their spouses’ death have on their and their children’s living situation? To our surprise, we find only 7 cases of widows or divorced women who had remarried recorded in the census returns – out of about 400 surviving returns which recorded almost 1,500 individuals.23 On the other hand, we have 46 cases recorded in the census returns in which a formerly married woman is declared as living with her children in a simple family household.24 In about one-fourth of these cases at least one of her children was no older than 14 years, i.e., a legal minor.25 At first glance, we would conclude from such numbers that in Roman Egypt many women who lost their husbands did not remarry, even if the widowed or divorced woman was still young and capable of bearing children. Bagnall and Frier, in fact, argue on the basis of the census returns that remarriage for women was exceedingly rare.26 Yet this conclusion (supported also by Hanson27) patently contradicts the patterns of remarriage just discussed for comparable societies in which women from all strata but the social elite remarried in large numbers. The following questions thus arise: (1) Did the remarriage of widows in Roman Egypt really occur less frequently than is generally assumed for common people from other traditional patriarchal societies? Or, if this is not the case, then (2) does Roman Egypt present us with an exception to what is otherwise a widespread pattern of remarriage of young widows? Or rather (3) should we reconsider the nature of our evidence (the census returns) in case these returns may have misled us either by not recording or 21 22

23 24

25

26

Cf. Bagnall and Frier 2006, table 5.3. Saller 1994: 189; Bagnall and Frier 2006: 77, table 4.2, 123; Scheidel 2009. This is indirectly confirmed by our literary sources that point to a high number of children whose father died before they reached maturity (Sen. Ben. 5.5.2; Cic. Ver. 2.1.153; Aug. Serm. 14.1). Cf. Krause 1994–5: iii.4. 117-Ar-7; 131-Me-1; 145-He-2; 173-Ar-11; 187-Ar-22; 187-Ar-29; 187-Ar-32. Only in a few cases is it explicitly specified that a woman was widowed or divorced. Often we can just assume from the fact that she had children that she had been married before (cf. Hanson 2000: 152). 33-Ar-2; 103-Ar-9; 117-Ar-5; 145-Ar-2; 173-Pr-4; 187-Ar-29; 187-Ox-4; 243-Ar-1 (? age of one child lost); 243-Ar-3; 257-Ar-1; P.Oxy.Census 89-Pt-51. Cf. Bagnall and Frier 2006: 126–7. Bagnall and Frier 2006: 115, 126–7. 27 Hanson 2005: 86–7.

96

Widowhood, remarriage, and residence patterns

perhaps somehow suppressing a substantial number of widows who had indeed remarried? We suggest that in this particular case we should attempt to see whether our socio-cultural model can shed some light on the evidence, instead of the other way around.28 I believe that any statement about remarriage rates based solely on the census returns is most likely to be unreliable. Remarriage in Roman Egypt was likely far more common than is indicated by these documents at first sight (in other words, option 3). Bagnall and Frier have already pointed out that the returns do not show us all cases of remarriage: “Since surviving issue of prior marriages provide the invariable occasion for mentioning remarriage, its incidence was certainly higher than the returns indicate.”29 Marital status (single, married, widowed, or divorced) is often not stated as such in the census returns. Thus, married or previously married status can be inferred only if a spouse or legitimate children were present in the same household.30 Since the census returns regularly recorded only those children who were living in the same household,31 we must thus acknowledge that these documents only recorded a woman’s previous marriage if it happened to meet two conditions: first, her previous union had to have resulted in children; and second, these children had to be living with her after her remarriage. Therefore, we are only able to recognize a woman as having been previously married (widowed or divorced) if she happened to have her children living with her in the same household at the time of the census declaration.32 To understand the implications of this point, we need to set the probability of this living arrangement in the context of the traditional virilocal marriage pattern that prevailed in rural Roman Egypt. As we have seen in Chapter 2, the majority of all people recorded in the census documents, nearly 60 percent, lived in extended or multiple family households, with complex household forms being more common in the villages than in the cities.33 The young couple, even if they already had children themselves, thus lived with the husband’s family, at least until the husband’s father died.34 Often, however, they lived there even longer, since many brothers decided to delay the splitting up of the household and to continue living 28 29

30 32

Cf. Scheidel 2001b: 3. Bagnall and Frier 2006: 126. The same seems also to be applicable to divorce. Bagnall and Frier 2006: 124: “Since the returns mention divorce only when there are surviving children, it may have been appreciably more frequent than the attested cases suggest.” For the frequency of divorce among the lower classes, see also Humbert 1972: 343–4; Saller 1991: 38; Treggiari 1991b: 473–82; Treggiari 1991a: 31–46; Bagnall and Frier 2006: 123–4. Bagnall and Frier 2006: 112. 31 For an exception see, e.g., 145-Ar-12. Cf. Krause 1994–5: i.98; Hanson 2000: 152. 33 Bagnall and Frier 2006: 66–8. 34 Cf. Chapter 2.

Remarriage in the census returns

97

together. Couples living in nuclear households were in general of a relatively advanced age, a sign that these households came into being by the attrition of kin rather than by setting up a new household. Young couples usually lived with the husband’s family.35 Under ideal circumstances children grew up in multiple family households together with their cousins, grandparents, uncles, and aunts. But what happened to his family when the father died prematurely? Did the orphans stay in their paternal home where they had lived before, did the family all leave together, or did the widow alone leave, with her children remaining behind with their paternal relatives?

4.4 Remarriage in the census returns Let us look closer at those widows who are identifiable as such in the census returns. These 46 widows had children and at the point when the census was drawn up they had not remarried but continued living with their children. It is, however, hard to draw any conclusion from this number alone about the relative frequency of widowed mothers refraining from remarriage, since we do not know how many women with children had remarried but are not identifiable as such since they had left their children behind with their first husband’s family.36 From the 46 single mothers we can subtract 4 women who were the mothers of apatores and therefore not widows but probably living in an illegitimate union with the father of their children.37 In 22 of the 42 remaining cases of single mothers, the mothers functioned as the heads of the households, mostly the ones whose children 35 36

37

Bagnall and Frier 2006: 60–2. 33-Ar-2 (mother is declarant); 103-Ar-9 (mother is declarant); 117-Ar-1 (lodgers); 117-Ar-5 (mother is declarant); 117-Ar-6 (son is declarant); 117-Ar-7 (son is declarant); 117-Ar-12 (mother is declarant); 131Ar-12 (mother is declarant); 131-Ar-13 (mother is declarant); 131-He-2 (son is declarant); 145-Ar-1 (mother is declarant); 145-Ar-2 (mother is declarant); 145-Ar-3 (mother and daughter; daughter is declarant); 145-Ar-20 (son is declarant); 145-Ar-22 (mother is declarant); 145-Ly-1 (mother is declarant); 159-Ar-10 (son is declarant); 173-Ar-9 (two apparently unrelated widows with children; one widow’s son is declarant); 173-Ar-11 (mother and daughter; daughter is declarant); 173-Pr-4 (son is declarant); 173-Pr-11 (son is declarant); 173-Pr-14 (mother is declarant); 173-Pr-15 (mother is declarant); 173-Pr-17 (son is declarant); 187-Ar-29 (mother is declarant); 187-Ar-34 (mother is declarant); 187-Ar39 (mother is declarant); 187-Me-1 (mother is declarant); 187-Ox-4 (mother is declarant); 201-Ar-1 (mother is declarant); 201-Ar-8 (mother is declarant); 201-Ar-9 (mother is declarant); 215-Ar-1 (son is declarant); 215-Ar-5 (son is declarant); 243-Ar-1 (mother is declarant); 243-Ar-3 (mother is declarant); 243-Ar-4 (mother is declarant); 257-Ar-1 (mother is declarant); ???-Ar-2 (mother is declarant); P.Oxy. Census (Bagnall et al. 1997): 89-Pt-27 (son is declarant); 89-Pt-28? (son is declarant); 89-Pt-36 (son is declarant); 89-Pt-37 (son is declarant); 89-Pt-42 (son is declarant); 89-Pt-51 (son is declarant). Whether a widow was acting as head of household or not obviously did not depend on the age of her children or whether or not she had sons. We have widows in their seventies with sons in their fifties: the former nevertheless filed the census return as head of household (e.g., 201-Ar-9). 145-Ar-1; 173-Ar-9 (2); 187-Me-1.

98

Widowhood, remarriage, and residence patterns

were still minors. We have, however, also mothers in their fifties,38 sixties,39 and even seventies40 who, despite living with adult sons, occupied the role as head of household. A recent study of wills by Yiftach-Firanko partly illuminates the legal grounding of this arrangement.41 A man’s heirs were his children or his natal family, not his widow.42 Widows were, however, often granted by their husbands an extended usufruct of the latter’s estate, on the condition that they took responsibility for the administration of the property for the benefit of their common children until they came of age.43 Yiftach-Firanko sees in this arrangement not a new entity after the husband’s death but a continuation of the wife’s rights as “mistress” (kyrieuousa) of the family estate which she had already enjoyed during her husband’s lifetime.44 In such a situation the widowed mother stepped into her late husband’s role as the head of household and enjoyed a degree of independence unusual for women at any level of society. The same pattern has been described for many other traditional societies.45 If these widows remarried, however, they lost this privilege.46 If a widow in Ottoman Aleppo, for instance, was assigned the management of the husband’s estate until their joint children could take over, she usually refrained from remarrying, since she would have irreversibly lost all the rights to her late husband’s property and the custody of her children.47 When studying the wife’s usufruct and administrative responsibilities, Yiftach-Firanko nevertheless does not take into account that the respective 38 41 42

43

44

45

46

47

145-Ar-22; 201-Ar-1; 201-Ar-8. 39 117-Ar-12; 173-Pr-14. 40 131-Ar-12; 201-Ar-9. Yiftach-Firanko 2006. It should not be forgotten, however, that the fact that widows did not inherit from their husbands did not mean that women could not hold property. Women could inherit from their parents, and some women held considerable property in their own right (Rowlandson 1998). Cf. P.Cair.Masp. 2.67156 from 570 ce. Cf. Yiftach-Firanko 2006. See also Dig. 7.2.8 (Ulpian); 33.2.37 (Scaevola). For mothers in the Greek East serving as guardians of their minor children, see Evans Grubbs 2002: 248, 254–60. Yiftach-Firanko 2003: 122–3, 242–3. Cf. also Häge 1968 for matrimonial property arrangements in the papyri. For late Ottoman Palestine, Agmon 2006: 162; for twentieth-century Iran, Fridell 2003: 156; for ancient China (Qing dynasty), Mann 1987: 44; for late imperial China, Sommer 2000: 191; for nineteenth- and twentieth-century China, Wolf 1981: 140 and Holmgren 1985: 2; for nineteenthcentury Egypt, Tucker 1985: 60. Cf. in general Palmore 1987: 101–2; Goody 1990: 202–3. CPR 6.1 from 125 ce; P.Diog. 9 from 186–224 ce; SB 8.9642 from 117–38 ce; Chr.Mitt. 306 from 155 ce; cf. Dig. 35.1.62.2 (Clementius). Goody 1990: 202–3: “The ideal widow was one who remained chaste after her husband’s death and by so doing retained an interest in the property, in the conjugal fund, established at their marriage.” For early Hindu society in pre-Christian India, see Altekar 1959: 250–1: “A vast majority of widows therefore used to get their husband’s shares, if not directly as their heirs, at least indirectly as the guardians of their minor sons . . . In the absence of the son the property should devolve, not upon the widow, but upon the nearest male sapinda . . . the widow is, however, nowhere mentioned as possessing any rights of inheritance.”

Remarriage in the census returns

99

families in Roman Egypt must have exhibited certain living arrangements as a precondition for such a legal arrangement to make sense. For the widow to become the head of the household and receive a usufruct in return for the management of the estate, she must have been living in a conjugal household with her husband and their children in which her deceased husband had been head, since only in this case would he have been able to grant these rights to his widow. If the family had been part of an extended household, and the husband’s father or brothers were still alive and all property was held in common,48 this arrangement would not have been possible. A young widow who had not yet established her own household with her husband, but still lived with her husband’s family, surely could not expect to be granted the same rights upon her husband’s death as the women mentioned above, i.e., the administration of his property and the household headship. In fact, in those cases in which a deceased husband’s father or elder brother was still alive, and moreover living with them in the same household, he administered all the family’s property, as long as it was not divided.49 As discussed in section 2.11, only when the members of a multiple family household decided to split up were the shares of the inheritance divided among all legal heirs. That means that the orphaned children’s share was often tied up in a multiple family household. They gained independent ownership only at a much later point.50 In other patriarchal societies in which the multiple family household was common this arrangement, the widow functioning as trustee of the conjugal fund and acting as head of household, was thus only possible if her late husband and his brothers had already divided their parental estate and the couple thus had already established their own household. If not, his widow merely took her dowry back and returned to her natal family, often leaving her children behind. In classical Athens, young widows were usually urged to leave their deceased husbands’ families very quickly, regardless of whether or not they had children.51 “One does not find a single example of a widow of childbearing age who chose or was allowed to choose to remain in her husband’s house.”52 Our literary sources refer to a law stipulating that only if a widow was pregnant with 48

49 51

52

Cf. P.Oxy. 34.2713 from 297 ce: “For all that was left to us by the foregoing inheritance – being one household and one family – was in that very house in which they lived in partnership, I mean because the slaves and lands and household and moveable goods were all held in common.” Rowlandson 1998: 184. 50 Codex Hermupolis 8.30–9.4; cf. Lippert 2008: 125. Isaeus 1.15, 8.8, 9.29; Dem. 36.45–6. Cf. Thompson 1972: 211–18; Isager 1981–2: 85; Gallant 1991: 54; Pomeroy 1997: 120, 188; Cox 1998: 89. Hunter 1989: 291.

100

Widowhood, remarriage, and residence patterns

her late husband’s child would she remain in her husband’s oikos, and only until she gave birth.53 Cuno notes for nineteenth-century Egypt that households were headed by a widow only when the couple had been living in a nuclear household before the husband’s death and no other male adult was available within the household to take over the headship.54 Likewise in traditional India, a wife could only be entrusted with managing her late husband’s property if her husband did not own property in common with his family.55 Only if the husband and his brothers had already divided their parental estate, and the couple had thus already established their own household, could the widow succeed her late husband as head.56 Breschi et al. come to the same conclusion for nineteenth-century central Italian sharecropping families, who lived predominantly in multiple family households. If a couple were still living with the husband’s family, the likelihood that his widow would remarry after his death was significantly higher than if the couple had been living on their own.57 Among nineteenth-century Italian wage laborers who mainly lived in simple family households, it was quite common that widows took over headship. However, this almost never happened in multiple family sharecropper households. Since here several adult males were available to take over the late head’s responsibilities, the deceased head’s widow could not succeed to her husband’s position.58 So what happened to those younger widows and their children in Roman Egypt who did not happen to live in their own household? Was a woman who had lived with her husband’s family able to remain there with her children after her husband’s death, if she wished? The census returns from Roman Egypt give us the impression that this living arrangement was in fact regarded as quite unusual. A considerable number of seemingly unattached women late in and past their marriageable age lived with their parents or brothers.59

53

54 58 59

Ps.-Dem. Or. 43.75: Ὁ ἄρχων ἐπιμελείσθω τῶν ὀρφανῶν καὶ τῶν ἐπικλήρων καὶ τῶν οἴκων τῶν ἐξερημουμένων καὶ τῶν γυναικῶν, ὅσαι μένουσιν ἐν τοῖς οἴκοις τῶν ἀνδρῶν τῶν τεθνηκότων φάσκουσαι κυεῖν. Cuno 1995: 491–4. 55 Altekar 1959: 261. 56 Sommer 2000: 191. 57 Breschi et al. 2007. Kertzer and Hogan 1995: 125. 103-Ar-3 (20-year-old daughter); 117-Ap-5 (2[.]-year-old daughter); 117-Ap-6 (24-year-old daughter, coeval sister-in-law married the brother at 16 or younger and has already two children); 117-Ar-1 (33year-old daughter and 70-year-old sister of late head of household); 117-Ar-11 (22-year-old daughter; sisters-in-law married at 19 and 17 or younger); 117-Ar-12 (20-year-old daughter; mother had married at age 15 or younger); 117-Ar-13 (55-year-old woman living in her brothers’ household); 131-Ar-12 (31year-old daughter); 131-Ar-18 (40-year-old daughter with her 4 children in her parents’ household); 131-Ox-1 (20-year-old daughter; sister-in-law married at 19 or younger); 145-Ar-3 (40-year-old

Remarriage in the census returns

101

In the following analysis, I use 20 years as the age where I assume that women who still lived in their natal home might have been married before, especially when their sisters-in-law living in the same household had married the women’s brothers at the age of 15 and younger. Twenty-two women over the age of 20 lived (again) with their parents; another eight women over the age of 20 lived with their brothers when their parents had already died, and another two lived in their uncles’ multiple family household. The average age of women living unattached in their natal home was 31 years. We explicitly know that some had been married but had returned to their natal home upon their divorce or the death of their husbands.60 If remarriage of young widows in Roman Egypt were truly as uncommon as has often been stated, we should find a much higher number of widows living with their deceased husbands’ families. Yet, we have only one case in the census returns in which a widow stayed in her in-laws’ household: a widowed mother with three children who lived with her husband’s married brothers and their families.61 She was, however, already 60 years old.62 Moreover, her 17-year-old son seems to have been the only male offspring of this joint brother household, and thus represented the single male heir on which the continuity of the entire frérèche depended. And we can identify only four cases in the census returns where a widowed or divorced mother returned to her natal home together with her children.63

60

61 62

63

divorced daughter returned to her natal home together with her children); 145-Ar-12 (33-year-old divorced daughter); 145-Ly-1 (21-year-old daughter); 145-Pr-1 (24-year-old daughter); 159-Ar-4 (47year-old divorced or widowed sister with her 19-year-old son living in her brothers’ household); 159Hm-3 (40-year-old daughter); 173-Ar-11 (32-year-old twice widowed or divorced daughter with her two children from two different unions); 173-Pr-3 (48- and 44-year-old sisters living with their brother); 173-Pr-5 (24-year-old daughter; sisters-in-law married at age 16 and 13 or younger); 187Ar-4 (23-year-old daughter; sisters-in-law married at age 15 and 16 or younger); 187-Ar-8 (20-year-old daughter); 187-Ar-22 (22-year-old daughter; half-sister married at age 18 or younger); 187-Ar-39 (29year-old daughter); 201-Ar-9 (56-year-old daughter); 215-Ar-2 (25-year-old woman living with her brother, not married to each other; 50-year-old woman living with her sister); 215-Ar-3 (26-year-old daughter); 215-He-2 (33-year-old woman living in her brother’s household); ???-Ar-3 (35-year-old and 38-year-old daughters living in their uncles’ multiple family household). Moreover, two women formerly married to their brothers but then divorced also stayed in their parental home, obviously because they had nowhere else to go (159-Ar-4; 159-Ar-5). Two apparently still married women nonetheless lived with their parents: 201-Ar-8 (21-year-old) and 201-Ox-1 (20-year-old). 131-Ar-18 (40-year-old daughter with her 4 children in her parents’ household); 159-Ar-4 (47-year-old divorced or widowed sister with her 19-year-old son living in her brothers’ household); 173-Ar-11 (32year-old twice widowed or divorced daughter with her two children from two different unions). ???-Ar-3. Cf. 117-Ar-1 for a 53-year-old widow who lived with her 70-year-old sister-in-law and her three adult children as renters in Arsinoe. Bagnall and Frier 2006: 131-Ar-18; 145-Ar-3; 159-Ar-4; 187-Ar-10. Cf. P.Oxy. 34.2713 from 297 ce for a widowed mother who had returned with her fatherless little daughter to her father’s home.

102

Widowhood, remarriage, and residence patterns

For this reason, the most likely assumption is that most of the younger widows, whose number we must presume to have been sizable, given the high mortality and typical age gap between spouses, left their children behind and returned with their dowry to their natal home to be remarried in due course. We see them in the returns simply as unmarried “daughters” or “sisters,” not as “widows,” or, if they remarried, as “wives,” and not “remarried widows.” As suggested above, such distinctions are invisible in the census returns because they recorded remarriages only for those households that included the children of the previous marriage. We just do not know if, for instance, the 24-year-old Senosiris who lived with her parents and siblings was not widowed or divorced, and had returned to her natal home leaving her children with her in-laws. Her sister-in-law, the wife of Senosiris’ brother living in the same household, had already two children at the age of 18.64 Another case is the 24-year-old apparently not yet married Taaronnesis, who lived with her parents and siblings. Her 13-year-old sister was already married, and her sister-in-law, the wife of Taaronnesis’ brother, was only 16 years old. These are strong arguments for the possibility that also Taaronnesis had been married before.65 The two sisters-in-law of another 23-year-old unmarried woman who lived with her parents were both only 16 when they had their first child.66 These examples show that marriage in these families generally took place much earlier for women. Therefore, it does not go too far to assume that these single women in their mid-twenties were in fact already widowed or divorced. Hanson in her study on ancient widowhood also notes an indubitably clear aversion of widowed women to live in a household whose head was only related to them by marriage, fearing disrespect and mistreatment from their in-law kinsmen.67 A further issue in this regard touches on the community’s concern for the widow’s reputation and respectability. A young widow simply could not be permitted to remain unattached in her late husband’s parental home, living together with his married or unmarried brothers and cousins. Such a woman was a source of constant tension in a household where she no 64 65

66

117-Ap-6. 173-Pr-5. Further potentially young widows and divorcées who had returned to their natal families: 103-Ar-3 (20 years old); 117-Ap-5 (in her twenties); 117-Ar-1 (33 years old); 117-Ar-11 (22 years old); 145Ar-12 (33 years old, divorced, returned to her parents); 145-Ly-1 (21 years old); 145-Pr-1 (24 years old); 159-Hm-3 (40 years old); 173-Pr-14 (20 years old); 187-Ar-8 (20 years old); 187-Ar-21 (22 years old, married sister is 18); 187-Me-1 (20 years old); 201-Ar-1 (54 years old); 201-Ar-6 (57 years old); 201-Ar-9 (56 years old); 215-Ar-2 (25 years old); 215-Ar-3 (26 years old); 215-He-2 (33 years old); ???-Ar-3 (38 and 35 years old); 117-Ar-12 (20 years old); 131-Ar-12 (31 years old); 117-Ar-13 (55 years old); 187-Ar-39 (29 years old); 201-Ar-14 (27 years old). 187-Ar-4. 67 Hanson 2000: 152.

Remarriage in the census returns

103

longer had any sexual or reproductive role to play, and was exposed to sexual abuse. In a joint household of several married brothers, a young widowed woman could thus be a cause for a considerable amount of conflict and anxiety.68 There were two solutions to this dilemma. Either the widow had to leave or she married a younger brother of her late husband – a practice widespread in many Near and Far Eastern societies, including early modern and modern Israel, India, China, Taiwan, and Japan.69 By remarrying the widow within the family, the household continued utilizing the woman’s productive and reproductive labor while keeping land structures and the joint household’s patrimony intact.70 However, leviratic unions were unknown in Greece and Rome, as well as in Roman Egypt. Here it is worth reflecting on a sad fact of life: for the children living in a multiple family household, the untimely death of a father often seems to have led to the loss of the mother as well. Fatherless children in classical Athens or in Rome seemed to have regularly stayed with their paternal relatives rather than to have followed their mothers into a new marriage.71 A widow in late medieval Florence, for instance, who had to return to her natal family on the day of her husband’s funeral, faced leaving her children behind for her husband’s agnates to rear: “When a widow left a house in order to remarry, she left with her dowry but without her children.”72 We have several testimonies from late medieval Tuscany that small children feared when their fathers died that they would also be abandoned by their mothers, who remarried rather than stayed with them.73 The same was true in almost every patriarchal structured society, so for instance for twentiethcentury rural Turkey, where children remained with the husband’s family.74 For nineteenth-century Egypt we hear of many cases where widows or divorcées lost custody of their children to their husband or husband’s families, since children belonged to the paternal line, which furnishes us with a parallel to Egypt many centuries earlier. If the widow remarried, she regularly left her children behind because children were forbidden by law to live as a “foreigner” (i.e., not a close relative) in the stepfather’s household.75 We have also several cases in the papyri from Roman Egypt where a widow

68

69 71 72 74

Cf. Mann 1987: 44 and Wolf 1981: 140 for early modern and modern China, where widows were urged to leave their deceased husbands’ families. For traditional India, see Weinberger-Thomas 2000: 147. Mattson 2005: 1. 70 Drèze 1996: 265. For Athens, see Cox 1998: 75, 119–20. For Rome, see Dixon 1988: 50. Klapisch-Zuber 1985: 125; cf. 123. 73 Cf. Haas 1998: 175; Klapisch-Zuber 1998: 271–3. Delaney 1991: 53. 75 Tucker 1985: 58–9.

104

Widowhood, remarriage, and residence patterns

reclaimed her dowry from her late husband’s family,76 and we find fatherless children in Roman Egypt living with a host of relatives. They lived in the households of their adult brothers,77 less often, cousins78 or paternal or maternal uncles;79 once even a paternal granduncle took his deceased nephew’s offspring into his household, “cherishing the kindness due by nature towards my nephew’s children,” even though the mother of the children was still alive.80 A receipt for the return of a dowry issued by the mother of a young widow from Oxyrhynchus dating to the second half of the second century ce illustrates such a case.81 The young woman had lost her husband and consequently had returned home to her parents to be remarried in due course. The late husband’s family returned the dowry to the widow’s parents, who were already thinking about their daughter’s next marriage and planning to bestow the same dowry on her once more (“all of which shall be bestowed upon the same daughter of mine when she is married to another

76

77

78

79

80

81

e.g., P.Coll.Youtie 2.67; BGU 3.970; SB 6.9065; P.Cair.Isid. 62; M.Chr. 88; probably also BGU 2.388; cf. Yiftach-Firanko 2006. 89-Ar-1 (7-year-old Horion lived with his two adult brothers); 103-Ar-12 (two girls aged 14 and 8 living with their adult brother); 215-Ar-9 (a 10-year-old girl living with her adult brother and sister). 145-Ar-20 (two brothers aged 8 and 1 living with their aunt and 20-year-old married paternal cousin); 159-Ar-10 (a boy or young man lived with his elderly aunt and his paternal cousins’ families); 159-Ar-13 (a 12-year-old girl probably lived with her 25-year-old paternal cousin and his wife); 173-Ar-11 (a fatherless boy or young man whose age is lost lived with his elderly paternal aunt and her children); 187-Ar-7 (a 17-year-old boy and his 14-year-old sister lived alone). To the fifteen children who were explicitly identified as nephews, cousins, or minor siblings, we may add seven apparently fatherless minors recorded in the household lists, whose relationship to the declarant and his family is left unstated: 131-Ar-11 (the 6-year-old Philip and his younger sibling of unknown sex lived in the family of a divorced 30-year-old man and his two children); 145-Ly-1 (an 8-year-old girl, identified as an “orphan”); 187-Ar-8 (an 8-year-old girl living in the household of the declarant and his children from two marriages); 187-Ar-26 (a woman of 25 and two girls, aged 15 and 4, “kin” of the declarant); 229Hm-2 (a 6-year-old boy living with the declarant and his 8-year-old son). For the discussion of whether or not Esther (2:7) had been adopted by her cousin Mordecai, the son of her uncle, as his own daughter after her parents had died, see Wahl 1999: 78–99. 173-Pr-13 (13-year-old girl living in the household of a paternal uncle?); 215-Ar-3 (a 10-year-old and an 8-year-old girl living with their paternal married uncle and his family). Cf. Bagnall and Frier 2006: 126. See also P.Oxy. 34.2713 from 297 ce: after her father’s death, Aurelia Didyme moved with her mother back to her mother’s paternal home, where her grandfather and her two adult uncles lived. After her mother’s and grandfather’s deaths, Aurelia Didyme continued to live with her two uncles, who acted as her guardians. In general, however, our evidence from Roman Egypt does not seem to corroborate Bremmer’s thesis that maternal uncles had a special role in fosterage (Bremmer 1976: 65–78; 1983: 173–86). P.Oxy. 34.2711 from 268–71 ce: the granduncle had taken in the three destitute children of his deceased nephew (two sons and a daughter), while their mother lived elsewhere. The papyrus then breaks off. See also SB 14.11881 from the fourth century for a woman who had problems supporting her deceased brother’s children. P.Coll.Youtie 2.67 from 260/1 ce. Cf. Rowlandson 1998: 191–3, no. 141.

Remarriage in the census returns

105

household”).82 The young widow’s minor son had remained behind with his paternal uncles, one of whom was acting officially as the boy’s guardian.83 Just because the paternal family did not wish to keep a young, marriageable woman did not mean that her children were no longer valued. On the contrary, leaving the children with their late father’s family held advantages for both the children and the paternal family. If they had lived with their parents in an extended household, the children were also after their father’s death able to remain in the only house they had ever known, continuing to live together with their paternal grandparents, uncles, or elder brothers, so their father’s death at least entailed no dislocation.84 This was also the household to which they were legal heirs.85 In return for its maintenance of the children, the family for its part secured the continuation of the paternal family line and property. However, even if the mother moved out to remarry, this did not necessarily mean that the children lost contact with her, since she probably still lived nearby in the same town or village and they remained her legal heirs – together with children she might have in a subsequent union.86 Moreover, through this living arrangement the perils that were associated with living together with a stepfather under the same roof were avoided.87 These important changes that accompanied parental divorce or death of the father emphasize the necessity of a dynamic approach to the structure of the individual family during its life cycle, and the sequences of types through which the respective household might pass.88 Our census returns come to an end in the mid-third century ce, and so it is impossible to make any statements about whether or not the proportion of widows who remarried in Roman Egypt decreased under the influence of Christianity, as has been often suggested.89 However, while the religion of the land may have changed, the underlying economic constraints persisted. In line with this assumption, Arjava has, in fact, found no hint in the 82

83

84 85

86

87

88

P.Coll.Youtie 2.67, ll. 33–4: ἅπερ πάντα τῇ αὐτῇ θυγατρί μου ἑτέρῳ ἀνδρὶ γαμουμένῃ προσφέρεσθαι μέλλει. For other widows reclaiming their dowry from their late husbands’ families, see, e.g., BGU 3.970; SB 6.9065; P.Cair.Isid. 62; M.Chr. 88; probably also BGU 2.388. Cf. Yiftach-Firanko 2006. Cf. Dixon 1988: 50. In addition to this kind of fosterage by relatives, actual adoption of orphaned children seems to have been much more common in Roman Egypt than previously assumed. Relatives, mostly brothers of the dead father, adopted these orphans (e.g., P.Lips. 1.28 [381 ce]; Lib. Ep. 25). These children cannot be distinguished from natural children in the census returns since, according to Greek nomenclature, adopted children took over the patronymic of their adoptive father (cf. Huebner 2007). Bagnall and Frier suggested that “Egyptian females of all ages appear to be considerably less mobile than males” (2006: 165). See Huebner 2009a. For the evil stepmother in Pharaonic Egypt, see Eyre 2007; for the stepmother in the Roman world, see Watson 1995. Cf. section 2.8. 89 For further references, see Evans Grubbs 2002: 208–9.

106

Widowhood, remarriage, and residence patterns

sources that people in late antiquity remarried less frequently than in earlier periods.90 Again, this conclusion coincides with Nathan’s observation that despite the considerable ideological pressure exerted by the Church and Christian thinkers, strong social, financial, and legal reasons prevented social customs from changing significantly from the classical period to late antiquity.91 Finally, other documentary material from later Roman Egypt – such as divorce settlements that conceded to both spouses the option of remarriage, wills, or marriage contracts that mention children from previous unions – also indicates that in late antiquity remarriage among the lower strata remained a common phenomenon.92 90 92

Arjava 1996: 171. Cf. Beaucamp 1990–2: 61–70; Krause 1994–5: i.181–91. 91 Nathan 2000. Divorce settlements from later Roman Egypt: P.Oxy. 43.3139 from the third or fourth century; P.Cair. Masp. 2.67154 R from the sixth century; P.Cair.Masp. 2.67153 from 568 ce; P.Flor. 1.93 from 569 ce; P.Cair.Masp. 3.67311 from 569/70 ce; P.Lond. 5.1712 from 569 ce; P.Cair.Masp. 1.67121 from 573 ce. For remarriage for women in late antiquity, see P.Cair.Isid. 113 from 303 ce; P.Col. 7.181 from 342 ce; SB 16.12265 from the late fifth century; P.Cair.Masp. 1.67005 from the sixth century; P.Cair.Masp. 3.67340 from the sixth century; P.Prag 1.42 from the early sixth century; P.Lond. 6.1717 from 560–73 ce; P.Cair.Masp. 3.67305 from 568 ce; BGU 1.319 from the seventh century. Remarriage for men in late antiquity: SB 16.12226 from the fourth century; P.Sakaon 38 from 312 ce; P.Berl.Zill. 4 from the middle of the fourth century; P.Herm. 31 from the sixth century; P.Cair.Masp. 1.67028 from 551 ce; P.Lond. 5.1709 from 570 ce. Cf. Krause 1994–5: i.244–54.

chapter 5

Growing old in the household

Take to yourself a wife when you are twenty years old, that you may have a son while you are (still) young. Instructions of Ankhsheshonq (BM 10508)1

5.1 The household as main provider of care and support in cross-cultural perspective The so-called “familial support ratio” describes the ratio of provider generation and consumer generation.2 It is established by relating the number of family members aged between 15 and 59 to family members aged 14 and below and 60 and over. Those under 14 years of age and those over 60 are defined as dependents. The familial support ratio is most commonly employed to measure and predict the ratio of the working vs the dependent population in contemporary societies affected by rising life expectancy and declining birth rates. Data taken from the Roman Egyptian census returns can tell us also about the familial support ratio, that is, the theoretical availability of support within the household, and constitutes thereby one of the very cases where this is possible for a society before the Renaissance.3 The total proportion of dependents of the Roman Egyptian census population was roughly 35 to 40 percent, about as much as it is in contemporary Western societies. But the ratio of those under 15 to those over 60, which today is about 1 : 1, was in Roman Egypt about 7 : 1. This means young children dependent on the support of the parent generation outnumbered dependent elderly by a huge margin. However, as far as the adult children/ elderly parents ratio is concerned, the support ratio in Roman Egypt was 1 3

Trans. Glanville 1955: 11.7–8. 2 Siegel and Hoover 1984. The Middle Babylonian population counts (Tenney 2011) and the Tang Chinese census documents of sixth- to eighth-century China (Liao 2001) would allow us to do the same for the respective census populations.

107

108

Growing old in the household

much higher than in our modern societies owing to higher levels of fertility and lower life expectancy. Different societies have developed different strategies and traditions for those perceived as the optimal caregivers for the elderly. Most often children were supposed to care for their parents in old age, but even here we find considerable variations as to whether all or just one child was expected to undertake this role, depending on the gender and the particular position in the order of birth. Among the Sherpa in Nepal, for instance, it was the youngest son who was expected to assume the role as caregiver and inherited in return two-thirds of his parent’s property while his brothers shared the remaining third.4 The unmarried daughter who looked after her aging parents was a common feature in many societies, from Neo-Babylonian times5 until today in contemporary Western societies.6 In !Kung families in the Kalahari desert, it is the grandchildren who act as the caregivers for their old grandparents.7 The availability and quality of care was decided not only by who took over the burden, but also by the living situation of the elderly. For early modern and modern societies in Europe several forms of residence patterns for elderly parents are testified.8 Some parents continued to live on their own in the family home while their children left to set up independent households; others left the home to their eldest son on his marriage and moved to a smaller dwelling on the same property. Some continued to live in the same household with their adult married sons, and the old father remained the head of the family until his death. In other societies the elderly were circulated among their children and lived with each of them in turn for some duration.9 A number of important studies on old age and the elderly have been published in recent years on the Greek and Roman world.10 As already laid out in Chapter 1, the present study does not aim to describe the philosophical or cultural conceptions of old age or the legal and social status of the elderly in Roman Egypt, but rather to provide insights into intergenerational interactions among family members of different generations within the context of the family or household. By this I hope to elucidate

4 7 9 10

5 6 Beall and Goldstein 1982: 141–8. Van Driel 1998: 188. Ikels 1983. 8 Draper and Keith 1992. Cf. Plakans 1989. For more literature, see Das Gupta 1997: 46–7. Ancient Near Eastern societies: Stol and Vleeming 1998. Greek society: Richardson 1933; Garland 1985; David 1991; Falkner 1995. Graeco-Roman world: Bertman 1976; Finley 1981; Falkner and de Luce 1989; de Luce 1994; Mattioli 1995; Brandt 2002; Gutsfeld and Schmitz 2003; Schmitz 2007 (including Sparta); Timmer 2008; Schmitz and von Prittwitz 2009. Roman society: Harlow and Laurence 2002; Cokayne 2003. Graeco-Roman Egypt: Rupprecht 1998. For late antiquity: Gilleard 2007; Amerise 2008.

Household and old age in Roman Egypt

109

how these continuing interactions generated expectations and how they should be seen as strategies to secure support in later life.

5.2 Household and old age in Roman Egypt To determine who took care of the elderly in Roman Egypt and under what circumstances elderly men and women lived, our most fruitful sources are again the census returns. Taken together with more qualitative sources such as wills, petitions, and private letters, they provide us with the most detailed and vivid information about residence patterns of the elderly that we have for antiquity. While for the analysis of private letters and the like, the actual age in years is often impossible to ascertain, and I regard someone as old from his relationship to others, his status, authority, and need of care and support, I need to set a cutoff point for any empirical analysis. As for defining whom I want to consider as “old” in our analysis of the census returns, I thus chose, as set out in the introduction, these individuals who were 60 years of age and older. The number of declarations recording such persons is relatively small. From the 1,108 individuals that Bagnall and Frier took into consideration in their first edition of The Demography of Roman Egypt,11 only 49 (31 men and 18 women) were aged 60 or older. This amounts to roughly 4 percent of all recorded persons, a percentage of the total population that comes close to that of many other societies of the past, but is of a completely different order of magnitude from today. In modern Western Europe more than 20 percent of the population is over 60, and this percentage is expected to rise within the next decade to over 30 percent.12 A high proportion of the old people in the towns and villages of Middle Egypt seem to have co-resided with one or more of their children. Many lived with their sons, the latter’s families, and their not yet married daughters. As regards marital status, there are noticeable differences between men and women. Even though marriage was nearly universal for women, widows, as we saw in the previous chapter, rarely remarried, at least from a certain age when they had already established some residential and economical independence with their husbands upon the latter’s death, while men after their wife’s death or after a divorce usually remarried until far into 11

12

Bagnall and Frier 1994. They included the census returns that became available to them after this date only in the supplement to their 2nd edition published in 2006, but did not use them in their analysis. See section 1.5. See section 1.1.

110

Growing old in the household

their forties, a pattern not dissimilar to that found in most other patriarchal preindustrial societies across the world.13 These differences in remarriage rates had a powerful impact on the living circumstances of the elderly, leading to stark differences between men and women. While the majority of old men are recorded as living with a spouse, often one considerably younger, more than two-thirds of women over the age of 60 did not reside with a spouse; thus they were either divorced or widowed and had not remarried. Of those elderly individuals over the age of 60 recorded in the census returns, only 6.5 percent (n=2/31) of all men,14 but 11.1 percent (n=2/18) of all women15 were living in households classed as solitaries.16 About 30 percent (n=9/31) of all elderly men17 and about 33.3 percent (n=6/18) of all elderly women18 were living in simple family households with unmarried children. About half of those elderly people living in nuclear family households, men as well as women, were still married; the others were living with their unmarried children. Co-residence of an elderly parent with a married child, that is living in an extended conjugal household, seems to have been more common for women than for men: 33.3 percent of women aged over 60 (n=6/18) were widowed or divorced and were living with a married child, forming a so-called upward extended conjugal household,19 but only 13 percent of old men (n=4/31) were widowed and living with a child.20 Co-residence with a son was the rule, undoubtedly because a young couple regularly resided with the husband’s family, not the wife’s. Living with a non-kin son-in-law seems to have been a constellation that one tried to avoid by all means.21 Many more old men than old women were living in multiple family households consisting of several conjugal couples: sixteen men (or about 52 percent)22, but only four women (or about 22.2 percent)23 13 16 17 18 19 20 21

22

23

Cf. Chapter 4. 14 173-Pr-8; 187-Ar-5. 15 187-Ar-30; 215-Hm-3. See below for a discussion. 131-Ar-8; 145-Ar-12; 131-Be-1; 145-Oa-1; 131-Ar-14; 117-Ox-2; 131-He-3; 145-Ox-2; 159-Hm-3. 117-Ar-6; 145-Ar-12; 131-Be-1; 173-Pr-14; 173-Pr-17; 215-Ar-1. 11-Ar-1; 117-Ar-1; 145-Ar-3; 173-Ar-9; 117-Ar-12; 131-Ar-12. 117-Ox-2; 131-He-4; 145-Ar-9; 187-Ar-18. In only one case a mother was living with her daughter and the latter’s husband and children (131-He2). In another case a widow’s son-in-law was simultaneously her nephew (201-Ar-10). In one other case, an elderly mother was living with a daughter and the latter’s children; the daughter was divorced and she was the owner of the house they were living in (145-Ar-2). 75-Ox-1 (two elderly widowed brothers living with their children); ???-Ar-3; 117-Ap-5; 117-Ar-11 (since the wife is 60, we assume her husband whose age is lost was older than her); 131-Ox-1 (again, since the wife is 60, we assume her husband whose age is lost was older than her); 159-Ar-11; 173-Me-3; 173-Pr-5; 187-Ar-8; 187-Ar-22; 201-Ar-9; 215-He-3; 131-Ar-18 (the husband’s age is lost but his wife is already 60); 145-Ar-9; 215-Ar-4. 117-Ar-11; 131-Ox-1; 131-Ar-18; ???-Ar-3.

Elderly in simple and extended family households

111

over the age of 60. Not many elderly men still lived with their brothers. At this advanced age it was much less likely to have living siblings than in younger years.24 Most often the multiple family household was formed by conjugal couples of two different generations, for instance, an elderly couple with their married son(s), or by several conjugal units in the younger generation. Some 38 percent of those men living in multiple family households (n=6/16) were widowed.25 Of the elderly women over the age of 60 living in multiple family households only one was widowed.26 In most cases the elderly parents, or a widowed parent, were living together with one or more married sons.27

5.3 The elderly in simple and extended family households Let us look at some examples of the elderly in simple and extended family households. In the early first-century Arsinoite nome, a certain Esersythis, a 70-year-old woman, lived together with her son Harthotes, who was in his early fifties, and her 9-year-old grandson Hapatothoes.28 Her son, a public farmer and priest of Tothoes, had evidently taken over the headship of the household when his father, Marres, died. Either his wife Taanchoriphis had also died or they were divorced, and the young son was left with him. We know from other documents from the same archive that Harthotes also had a daughter who was 12 years old at the time of the census. We meet her at a later date married to a paternal cousin: despite her young age she thus might have already lived in her uncle’s/father-in-law’s household when this census document was drawn up.29 Also in the Arsinoite nome but a hundred years later, the 75-year-old Sentaos resided with her 56-year-old son, his 53-year-old wife, and their two children.30 After the death of her husband, her son had become the owner of the house.31 24 25 27

28

29 31

75-Ox-1; ???-Ar-3. 75-Ox-1 (two old widowed brothers); ???-Ar-3; 173-Me-3; 201-Ar-9; 215-He-3. 26 ???-Ar-3. 117-Ap-5; 117-Ar-11; 131-Ox-1; 159-Ar-11; 173-Me-3; 173-Pr-5; 187-Ar-8; 187-Ar-22; 201-Ar-9; 215-He-3; 131-Ar-18; 145-Ar-9. 11-Ar-1. Harthotes gives his age in a document from 1 ce as 40 (P.Oslo 2.32) and as 55 in 12 ce (SB 20.14440 11-Ar-1), a clear case of age rounding. P.Mil. 1.1 from 38 ce. 30 117-Ar-6. As we saw in Chapter 4, it was the norm that children inherited before widows. When the husband of Tnepheros, 30, from Philadelphia in the Arsinoite nome, died, it was his eldest son, a 10-year-old boy, who inherited his father’s house, not the widow (117-Ar-5). His widow was, according to custom, provided for by a usufruct, while her two daughters probably received their share in the form of a dowry. The mother acted as guardian for her minor son and filed the census declaration. Since she was illiterate, her kyrios, a kinsman, helped her.

112

Growing old in the household

Even though the 74-year-old Tasoucharion from early third-century Karanis had two adult sons, 56 and 46 years old, and a 56-year-old daughter, she functioned as head of household for the census officials, probably because she was the owner of the house in which her children, grandson, and great-granddaughter lived. The elder son merely acted as her guardian.32 Tasoucharion probably had not inherited the house from her late husband but from her parents; women in Roman Egypt could hold considerable property that they had inherited from their natal family.33 Although she was not yet 60, we do not want to pass over Thephrokos, a widow in her fifties, from the Heracleopolite nome, who apparently had no sons to live with and therefore had moved in with her married daughter, her son-in-law, and the couple’s 3-year-old son. The house in Ankyronpolis in the Heracleopolite nome in which they lived belonged to her son-in-law, who declared the entire household.34 The 70-year-old Petechon, residing in the same nome as Thephrokos, lived with his 27-year-old son Piathres, who already in his father’s lifetime had taken over the ownership of the house.35 In another case, from the late second-century Prosopite nome, the 60-year-old Tatihoes lived together with her 20-year-old daughter Thermouthis and functioned as head of household.36

5.4 The elderly in multiple family households Research on past centuries of southern Europe has shown that in the areas characterized by multiple family systems the great majority of the elderly lived in large complex households, which furnishes us with parallels to Roman Egyptian society. In a sharecropping parish of central Italy, for instance, Kertzer found that in 1880 over 70 percent of those aged over 60 years resided in extended or multiple households. Only very few widows or widowers lived alone.37 In the following I want to describe some residence patterns for the elderly in Roman Egypt who lived in multiple family households, a living constellation, as already mentioned, much more common for elderly men than for elderly women. We have a typical multiple 32 36 37

201-Ar-9. 33 Cf. Rowlandson 1998. 34 131-He-2. 35 131-He-3. Cf. 201-Ar-10. 173-Pr-14. Kertzer 1984: 95–7. Similar patterns and proportions have also been described for periods and regions where we find a high proportion of “patricentric” family arrangements: Bourdelais 1985 for nineteenth-century southern France; Viazzo 1989: 229–44 for nineteenth-century northern Italy; Fauve-Chamoux 1996 for seventeenth- and eighteenth-century southern France; Perez-Fuentes Hernandez and Pareja Alonso 1997 for nineteenth-century northern Spain.

Solitary residence of the elderly

113

family in the case of the household of the 65-year-old Hartbos from early second-century Tanyaithis in the Apollonopolite nome.38 He was married and had three sons and two daughters with his considerably younger wife: her age is uncertain but she could not have been much older than in her early 40s, as one of their sons was only 2 years old. The eldest son, on the other hand, apparently in his early twenties, was already married, and his 20-year-old wife had moved into the household. Even more complex was the household of the 75-year-old Peteamounis from late second-century Moithymis in the Mephite nome.39 He was a widower who lived with his three adult sons, 45, 36, and 30 years old, their wives, and his two 4-year-old granddaughters, and acted as the head of household.40 The 69-year-old Pantbeus from late second-century Thelbonthon Siphtha in the Prosopite nome apparently did not want to spend his old age alone, having lost his first wife, the mother of his eldest son, and had married a woman who was 15 years younger than him.41 He now lived with her, their sons aged 3 and 21 and daughters aged 13 and 24, and his married 35-year-old son from his first union. The younger of the daughters was married to her 21-year-old brother, but the couple did not have any children yet. The 66-year-old Stotoetis from early third-century Soknopaiou Nesos in the Arsinoite nome was in a less fortunate situation than Pantbeus.42 He apparently had no children (anymore) and resided with his wife in the household of his adult married nephews and their families. This household constellation was most probably the outcome of a frérèche where all but one brother of the elder generation had passed away. It was not the 66-year-old Stotoetis but his 33year-old nephew who acted as head of the household, surely a situation that could create tensions between the two generations. We also do not know to what extent the nephews felt obliged to care for their old uncle and his wife.

5.5 Solitary residence of the elderly “Living together,” however, did not automatically mean that all family members had access to the same resources and were equally provided for. We know from studies on gender and household relations in various societies that the consumption of household resources among its members can be highly unequal.43 Residential patterns do not tell us if the elderly living in their sons’ household really received adequate support, or were in fact isolated within the family, abused, and neglected. Neither do we know 38 41

117-Ap-5. 39 173-Me-3. 40 For a similar case, see 215-He-3. 173-Pr-5. For similar cases, see: 187-Ar-8; 187-Ar-22. 42 215-Ar-4.

43

Papanek 1990.

114

Growing old in the household

if an elderly person living alone did not actually live next door to his children and receive adequate support and loving care. Even if the household seems to have been the fundamental unit for organizing intergenerational support, family support and co-residence were not necessarily identical. It is highly debatable if complex household forms really secure the economic and emotional welfare of their elderly members, and whether or not the incidence of older individuals and couples living alone really needs to be taken as a sign of their vulnerability, disregard, and isolation.44 To give a few examples from Roman Egypt for the solitary elderly: in late second-century Arsinoe the 70-year-old Sarapammon lived alone in his house, but his married daughter lived nearby only a few streets away with her husband and their two children, so that in theory she was available to look after him.45 In addition, Sarapammon also owned a male slave who most likely maintained his house, did the cooking and the laundry for him, and kept him company. Thus, Sarapammon’s solitary residence did not mean that he was lacking support. Likewise at the end of the second century ce, the 60-year-old Isidora lived without any family members in her house in Tebtunis in the Arsinoite nome, but she did not lack financial means, nor care or company either.46 She had inherited property from her parents, and was entertained and cared for by her two slave women, 45 years and 20 years old, and the women’s four daughters, who were between the age of 1 and 8 years.47 The 72-year-old Aurelia Demetria, widowed and apparently childless, owned two slaves, a 34-year-old of unknown sex and a 28-year-old male, who were both available to take care of her.48 The 70-year-old freedman Pnepheros seems to have been worse off, since he lived entirely alone in his house in Thelbonthon Siphtha in the Prosopite nome without any family members, lodgers, or slaves, but it is very likely that he had children or close relatives living nearby – we just cannot tell from the census returns alone.49

5.6 Widowers and widows and marriage propensity of children In his study on peasant families in early modern Japan, Skinner observes that age at first marriage differed for men and women whose parents were still alive, who had lost a father or a mother, or whose parents were both 44 45 47

48

Cf. for this problematic among the elderly in traditional Thailand: Knodel and Chayovan 1997. 187-Ar-5. 46 187-Ar-30. Three further slave women, 68, 42, and 38 years old, most probably a mother and her two daughters, were classed as runaways in this declaration. 215-Hm-3. 49 173-Pr-8.

Widowers, widows, marriage propensity of children

115

dead.50 He shows that those sons seem to have delayed marriage who lived alone with a widowed mother. He notes the same for girls who lived alone with their widowed father. Skinner’s explanation of the phenomenon is that marriage of the son or daughter would have disrupted the gender balance in the household formed by the child and his surviving parent. On the other hand, a widow living alone with a daughter seems to have had her married as early as possible in order to win a young man for the household. Likewise, a widowed father strived to marry his son early in order to bring a woman into the household.51 We want to have a look at our census data from Roman Egypt to see whether this model is also applicable for Roman society. We have twenty sons living with their widowed mothers in the same household.52 In ten of these cases the sons were currently married,53 in ten cases not married.54 Two out of these ten were too young to be married. Eight, however, were in their “best years,” from their late twenties to midfifties. Potentially the widowed mother in the household did not in fact encourage their (re-)marriage. We have seven further cases in which a widowed mother lived with one55 or more adult daughters.56 Two of them were married,57 and five not married (anymore).58 The 60-year-old Tatihoes from the late secondcentury Prosopite nome, who lived with her 20-year-old daughter Thermouthis, would remain alone should her daughter marry.59 For her it would thus have made sense to arrange a uxorilocal marriage for her daughter and thereby win a young man for the household who could not boast a house of his own.60 We might wonder how these female households made a living. If the household drew its income from farming, they must have hired wage laborers for working the fields. We know from twentiethcentury rural Ethiopia that women were involved in harvesting and took over virtually all food processing; clearing of land, plowing, and hunting, however, was predominantly done by men.61 We have some evidence from Ptolemaic and Roman Egypt that at least some daughters supported their elderly widowed mothers financially, namely by practicing a trade, an art, or 50 52

53 54

55 57 59

Skinner 1997: 65–6. 51 Skinner 1997: 65. 11-Ar-1; 117-Ar-1; 117-Ar-6; 117-Ar-7; 131-Ar-11; 131-Ar-12; 131-Ar-13; 145-Ar-22; 159-Ar-10; 173-Ar-9; 173Pr-11; 173-Pr-15; 173-Pr-17; 187-Ar-9; 201-Ar-1; 201-Ar-9; 215-Ar-1; 215-Ar-5. 117-Ar-1; 117-Ar-6; 117-Ar-7; 131-Ar-11; 131-Ar-12; 159-Ar-10; 173-Ar-9; 173-Pr-15; 201-Ar-1; 215-Ar-5. 11-Ar-1 (son, 55 years); 131-Ar-13 (son’s age lost, mother is 57); 145-Ar-22 (two sons, 26 and age lost); 173-Pr-11 (son, 15 years); 173-Pr-17 (son, 47 years); 187-Ar-9 (son, 16 years); 201-Ar-9 (two sons, 56 and 46, unmarried); 215-Ar-1 (son, 35). 131-He-2; 145-Ar-3; 145-Ly-1; 173-Pr-14; 201-Ar-8. 56 117-Ar-1; 187-Me-1. 131-He-2; 117-Ar-1 (to a brother). 58 145-Ar-3; 145-Ly-1; 173-Pr-14; 187-Me-1; 201-Ar-8. 173-Pr-14. 60 For uxorilocal marriages, see section 7.11. 61 Waring 2004: 186.

116

Growing old in the household

even prostitution.62 None of them, however, seems to have been married. Two of three cases that came to us because they caused some conflict originate from the Ptolemaic period.63 The first one dating to the Roman period comes from mid-fourth-century Hermopolis, but the respective incident took place in Alexandria. This papyrus has created a lot of scholarly discussion whether we are dealing here with a rhetoric exercise or some sort of fiction or with a real incident.64 Theodora and her daughter lived in extreme poverty. When her daughter was old enough, Theodora sold her into prostitution so that she might earn a living for them. As a prostitute the daughter found the favor of a city councilor of Alexandria, named Diodemos, who visited her regularly in the evenings. But one day, supposedly because he caught her with another man, he killed her. The city councilor was imprisoned and his case went to trial. Surprisingly we meet there as a joint plaintiff the prostitute’s old mother. She asked that Diodemos be compelled to provide her with financial support for her old age because he had killed her daughter who had been her only recourse.65 To cut a long story short: the governor decided in her favor. The city councilor was sentenced to the sword, and the prostitute’s mother inherited the tenth part of his fortune. The governor conceded that the old mother, instead of depriving her daughter of her modesty and sending her into prostitution, could have married her off. However, even the governor was probably well aware that without a dowry this would have been rather 62 63

64

65

P.Cair.Zen. 1.59028 from 258/7 bce; P.Lond. 7.1976 from 253 bce; BGU 4.1024 from 360 ce. A girl, Satyra, from the middle of the third century bce, supported her elderly, apparently single or widowed mother by working as a harp player for the dioiketes Apollonios. Apollonios promised the girl clothing and a living allowance for her services, but we have a letter from Satyra submitted to Zenon, a local notable in third-century Philadelphia, in which she complained that Apollonios neglected to pay her and her mother their clothing allowance and other provisions for already the second year. (P.Cair.Zen. 1.59028 from 258/7 bce; cf. Bagnall and Cribiore 2006: 98). About the same time, another daughter helped her mother, Haynchis, with selling beer and thereby supported her in her old age. When Demetrios, a local vinedresser, abducted the daughter and concealed her from her mother with the intention of setting up house with her without her mother’s consent, Haynchis also submitted a petition to Zenon. Without her daughter, the mother complained, she could not keep her business afloat. Haynchis asked Zenon to help her get her daughter back, contending that Demetrios had deceived her daughter since he was already married to another woman by whom he also had children (P.Lond. 7.1976 from 253 bce; cf. Rowlandson 1998: 209; Bagnall and Cribiore 2006: 102). BGU 4.1024, ll. 6–8. Cf. Wenger 1904; Whitehorn 1979: 245; Poethke 1981: 457–62; Keenan 1989: 15– 23; Bagnall 1993: 196–7. BGU 4.1024.8: ἡ δὲ μήτηρ τηˆς πόρνης Θ̣ ε̣ ο̣ δ̣ ώ̣ ρα τις γραυˆ ς καὶ πένης ἠξίου κατανα[γκ]ασθηˆ ναι Διόδημον πα̣ρ̣ α̣σ̣ χ̣ ειˆ ν α̣ὐ̣τῃˆ εἰς λόγον δ̣ι̣ α̣τροφω ˆ ν ὀ̣ λ̣ ί̣ [γ]ην τινὰ τουˆ βίου παραμυθίαν. ἔλεγε γάρ, ὅτι διὰ τουˆ το δέδωκά μ̣ [ο]υ̣ τὴν θυγατέρα πορνοβοσκῳ ˆ , ἵνα δυνηθω ˆ δια τραφηˆναι. ἐπεὶ οὐˆ ν τηˆς θυγατρός μου [τ]ελευτησάσης ἀπεστερήθην τω ˆν τροφω ˆν, διὰ τουˆτο ἀξιω ˆ δοθηˆναί μοι ὡς μετρίᾳ γυναικὶ ὀλίγα πρὸς διατροφήν.

Widowers, widows, marriage propensity of children

117

difficult. And even if the mother had found a husband for her daughter, it would not have been certain that her son-in-law would be willing to support his mother-in-law – whether by sending her money or by letting her move in with them. Though we do not know about the case’s actual connection to reality, as we have seen, the arrangement that a parent moved in with their married daughter and son-in-law is indeed very rarely documented in the census returns. In only two cases do we find a widowed mother living in her son-in-law’s household. In the first case the relationship between the elderly widow and the rest of the family is not securely documented.66 In the other case the wife’s mother was a sister of the husband’s mother and thus simultaneously the head of the household’s maternal aunt.67 In short, coresidence with daughters seems to have been rare and occurred only in those cases where mothers were widowed and had no sons. The meager evidence is spread out over several centuries. We have just two cases where a widower lived with an unmarried daughter who seemed to be of marriageable age, even though in both cases the ages of the daughters are uncertain.68 In ten cases a widower lived with one or more adult sons. In five cases these sons were married69 and in five cases they were not married (anymore).70 Therefore, we cannot draw any conclusive results as to whether or not Egyptian widows and widowers conformed to Skinner’s model and respectively encouraged or discouraged their children to get married. Widowers living with their unmarried sons in most cases did not have female relatives or slaves living in their household who could take over household chores that were typically assigned to females.71 The 70-year-old Petechon and his 27-year-old son Piathres, in the Heracleopolite nome, had not even female slaves to keep the house for them, but, of course, we do not know whether or not any female

66 68

69

70

71

131-He-2. 67 201-Ar-10. 173-Ar-7 (age lost); 187-Ar-28 (age lost). Cf. the case from third-century bce Magdola in the Arsinoite nome, already discussed in Chapter 3, where the old father, Ktesikles, complained to the king about his ungrateful daughter, who failed to support him (P.Enteux. 26 from 221 bce). Cf. Cribiore 2001: 106; Bagnall and Derow 2004: 245–6, no. 152. 89-Hm-1 (one son age lost, one 20 years old); 131-He-4 (son is 31 years old); 159-Ar-1 (son is 33 years old); 173-Me-3 (sons are 45, 36, and 30 years old); 215-He-3 (sons are 32 and 29 years old). 131-He-3 (son is 27); 145-Oa-1 (sons are 32 and 48); 145-Ox-2 (both sons’ ages are lost); 145-Pr-1 (son is 20, daughter is 24; both unmarried); 187-Ar-18 (son is 43). Exceptions are 173-Pr-13 (a 40-year-old widow with a 16-year-old son and 13-year-old niece or cousin); 145-Pr-1 (a divorcé or widower aged 59 [52 written in the census return] with his 20-year-old son and 24-year-old daughter; cf. 131-Pr-1); 173-Me-1 (a 51-year-old divorcé with a 1-year-old baby son and an unmarried 17-year-old daughter by two different wives).

118

Growing old in the household

relatives lived nearby who (to employ some gender stereotyping) did the cooking and the laundry for them.72 In sum, elderly mothers or fathers living with their daughters were in the minority, a fact that should not surprise us, as we saw above that daughters regularly moved out upon marriage and only sons stayed in their parental household.73 Widows and widowers or formerly married parents mostly lived with their sons, whether married or not.

5.7 Contact with children living far away Private correspondence between family members shows that at least some parents enjoyed an intense and close relationship with non-resident children, even if their daughters had married away or their sons had migrated to find work in another village or town, or to serve in the Roman army. A certain Saturnilus writing in the early second century ce, together with his young pregnant wife and their two little children, had left his home town, Karanis in the Fayum, in order to serve in the Roman frontier camp at Pselkis in Nubia.74 But he stayed in close contact with his widowed mother at Karanis by frequent letter writing. In one of his surviving letters, sent three months after his departure to Pselkis, Saturnilus assured his mother that he had already written three letters that month and that he would seek any opportunity to come home to visit her. Saturnilus hoped to be sent north to Alexandria to deliver official correspondence to the prefect. On his way he hoped to have the chance to make a stop in Karanis, showing an earnest desire for seeing his mother: Everybody who comes will testify to you how I am trying to come every day. If you want to see me a little, I want it a lot, and I pray to the gods every day that they soon give me an easy passage for coming. Everything in the army happens when an opportune moment arrives. If I find an opportune time I am coming to you.75

Saturnilus even tried to bribe an officer to secure the assignment as an official postal courier.76 In his letter he reported that his wife had recently had a son, and he thanked his mother for sending him a monthly allowance and a basket of olives. His mother was not living alone, but apparently living with Saturnilus’ two married brothers and their children, and maybe some younger not yet married siblings. Also, Saturnilus’ married sisters were 72 74 75

131-He-3. Cf. 201-Ar-10. 73 Cf. Chapter 2 on marriage patterns in Roman Egypt. P.Mich. 3.203 ( SB 4.7356) from 114–16 ce. Cf. Winter 1933: 50–2; Rowlandson 1998: 74. Trans. Rowlandson 1998: 74. 76 Cf. P.Mich. 8.468, ll. 38–41, for bribery in the Roman army.

Contact with children living far away

119

apparently living close by with their families. Saturnilus wished to send his wife and his children home, and asked that a friend of the family come down to accompany them back home. He himself was to stay for at least another two years at the Nubian frontier on guard duty. Mothers seem to have cultivated an especially strong emotional relationship with their sons, a phenomenon that we will discuss more in detail and from a different perspective also in Chapter 6.77 When a certain Apollinarius, living around the same time as Saturnilus, entered the military and had to leave Karanis, the same village where Saturnilus’ family lived, he kept in touch with his mother by regular letter writing. On a journey to Rome he sent her a note from Cyrene in present-day Libya and then from Ostia, the harbor of Rome. He asked his mother repeatedly to write him back and tell him about her health and the well-being of his brothers: Apollinarius to Taesion, his mother, many greetings. Before all else I wish you good health and make obeisance on your behalf to all the gods. From Cyrene, where I found a man who was journeying to you, I deemed it necessary to write to you about my welfare. And do you inform me at once about your safety and that of my brothers. And now I am writing you from Portus, for I have not yet gone up to Rome and been assigned. When I have been assigned and know where I am going, I will let you know at once; and for your part, do not delay to write about your health and that of my brothers. If you do not find anybody coming to me, write to Sokrates and he forwards it to me. I salute often my brothers, and Apollinarius and his children, and Kalalas and his children, and all your friends. Asklepiades salutes you. Farewell and good health. I arrived in Portus on Pachon 25. (2nd hand) Know that I have been assigned to Misenum, for I learned it later. (Verso) Deliver to Karanis, to Taesion, from Apollinarius, her son.78

In another case probably also from Karanis from the late second century, a certain Sempronius’ thoughts were also with his widowed old mother, who lived with his brothers and their families, while he himself was hundreds of miles away in Alexandria. He beseeched his mother to write him back that he might be less troubled about her health. The letter was read to her by her younger son, Sempronius’ brother Maximus: 77

78

Cf. the old widow of the second century bce who rejoiced when she heard that her son was learning Demotic and had the chance to obtain the position of a tutor in the house of the local physician, who was apparently looking for someone who could teach his children. The old mother hoped her son would gain provision for old age, not only for his own old age but also for that of his mother: “I rejoiced on both your account and my own,” she wrote to him (UPZ 1.148; for a translation, see Bagnall and Cribiore 2006: 113). P.Mich. 8.490 from the second century ce; trans. APIS (Advanced Papyrological Information System).

120

Growing old in the household Sempronius to Saturnila, his mother and lady, very many greetings. Before all I pray for your health together with that of my brothers may the evil eye not touch them. At the same time I make obeisance for you daily in the presence of the lord Sarapis. I have sent you so many letters and you have written me none in return, although so many people have sailed down. Please, my lady, write me without fail about your well being in order that I may be less troubled. For your well being must always be my prayer. I salute Maximus and his wife, and Saturnilus and Gemellus [probably other, not yet married sons], and Helene and her household [probably a married daughter living close by with her family]. Tell her that I have received a letter from Sempronius from Cappadocia. I salute Julius and his household by name, and Scythicus and Termuthis and her children. Gemellus salutes you. Do keep well for my sake, my lady, at all times.79

After her husband’s death Saturnila apparently lived with her son Maximus, who was already married, and her two not yet married sons Saturnilus and Gemellus. Helene, likely her daughter, was married and probably did not live in the same household but close by. As in the letter of Saturnilus discussed above, in Sempronius’ letter we may also sense a son’s deep emotional attachment to his mother. Sempronius assured her how much he cared for her well-being, and that he wished to receive a letter from her. At the end of the letter he begged her: “Do keep well for my sake, my lady, at all times.” Sempronius addressed his mother as the honored matron of the household. That this standing did not measure up to reality is brought to light, however, by a second letter attached to the first one, written by Sempronius for the eyes of his brother Maximus only. It reveals that Sempronius wrote the first letter mainly as a consolation for his mother, an assurance that someone cared about her and regarded her as important and influential. The second letter, however, shows that the old Saturnila had in fact complained to her son that she was not receiving the proper reverence due to her status and that she felt that she was treated like a slave in her own household. Perhaps we capture in her son’s letter his mother’s own words of complaint: Sempronius to Maximus his brother, very many greetings. Before all I pray for your health. I learned that you are treating our revered mother harshly as if she were a slave. Please, dearest brother, do not distress her in anything. If anyone of the brothers talks back to her, you ought to box their ears. For now you ought to be called father . . . For we ought to worship her who bore us as a god, especially when she is so good. This I have written to you, brother, 79

SB 3.6263, ll. 1–17; trans. Winter 1933: 49.

Contact with children living far away

121

since I know the sweetness of dear parents. Please write to me about your health. Farewell, brother.80

Sempronius reprimanded his brother to abstain from harsh treatment of their mother and to make sure that his younger brothers met her requests without protest. He admonished his brother to act as a father to his younger brothers and to slap them if they behaved disrespectfully. Sempronius even gave a reason as to why he thought that his old mother had earned better treatment or, in his own words, had to be revered as a goddess. She had given birth to them and was extraordinarily good and virtuous. The degree of compassion and sensitivity with which Sempronius dealt with this conflict is touching: everyone’s emotions were taken into account and he acted in a very thoughtful and loving way. If sons were away from home, their old mothers cared a lot about receiving at least a letter from them as a sign that they had thought about them – just as today an elderly mother might expect a weekly phone call from her children if they are unable to visit. Kophaëna, a widowed mother from fourth- or fifth-century Heracleopolis Magna, the metropolis of the Heracleopolite nome, was very troubled and upset that she had not heard back from her son Theodoulos, who was away from home, probably for business reasons: I want you to know what the steward told you, that “Your mother, your mother Kophaëna, is ill.” Look, for thirteen months and you have not even tried to write me a letter, because you know that I have treated you better than (my other son?), and you have not even tried, hearing that I am ill, you have not tried to send anything to me, not even something short.81

We can imagine that an even more marginal position than to a widowed mother was assigned to an elderly woman who did not even belong to the immediate family, such as a widowed or divorced aunt or sister who returned to her natal family.82 Daughters who moved out of their parents’ home upon marriage stayed in close contact with their parents, if possible by regular visits if they had married in the same village, or by exchange of foods and textiles via a courier.83 Letters written by daughters to their natal families were usually 80 82 83

SB 3. 6263, ll. 18–31; trans. Winter 1933: 49. 81 BGU 3.948; cf. Bagnall and Cribiore 2006: 224. 117-Ar-1. P.Oxy. 14.1679 from the third century ce: a daughter sends clothes to her mother (cf. Bagnall and Cribiore 2006: 352); BGU 2.615 from the second century ce Arsinoite nome: a daughter sends a basket to her father (cf. Schubart 1912: 68; Bagnall and Cribiore 2006: 395); P.Fay. 127 from second or third century Bakchias (Arsinoite nome): a daughter sends her mother some pieces of pottery (cf. Bagnall and Cribiore 2006: 329).

122

Growing old in the household

addressed to their mothers, but they included greetings to the rest of the family: the father, if still alive, brothers and their wives, and unmarried sisters. A certain Thaisous for instance wrote to her mother, Syras: “Salute Ammonas my brother and . . . and my sister . . . and my father, Theonas.”84 An early second-century ce letter by a young woman named Thermouthas addressed to her mother, Valerias, who lived in Philadelphia, a village in the Arsinoite nome, also illustrates the emotional closeness and affection a married woman retained to her family of origin, even if she lived far away.85 Thermouthas seems to have had up to five younger sisters who were still living at home with her mother, her sick father, and her grandmother. Her letter gives us a vivid glimpse into a three-generational household, and brings to life the mere statistics we gain from the census returns about such an extended family: Thermouthas to Valerias her mother, very many greetings and always good health. I received from Valerius the basket in which there were twenty pairs of wheat cakes and ten pairs of other cakes. Send me the blankets . . . and fine quality wool . . . Also, I am at the moment seven months pregnant. And greet Artemis and little Nikarous and Valerius . . . and all those in your house. And how is my father doing? Please send me news because he was ill when he left me. I greet grandma . . . Deliver to Philadelphia, to Valerias my mother.86

From the information her letter provides, we can assume that Thermouthas and her parents kept in regular contact, her father had paid her a visit only recently, and another member of the family, a certain Valerius, maybe a brother, acted as frequent courier between the two households. Not only letters but also goods were exchanged between the two families.87 All these letters reflect close emotional relations, respect, and filial love. The children took pains to assure their parents that they themselves were well and stressed that they cared very much about their parents’ well-being.

5.8 Authority, retirement and household hierarchies The life cycle of a household was tightly connected with the ways in which resources were transferred between generations. These property 84

85 86 87

P.Oxy. 2.295. For the use of kinship terms also for individuals that were not relatives, see Bagnall and Cribiore 2006: 85–6. Cf. Chapter 6 on the daughter-in-law. SB 5.7572. Cf. Winter 1933: 90; Rowlandson 1998: 284–5, no. 220; Burnet 2003: no. 160. Trans. Rowlandson 1998: 285. Cf. also P.Wisc.inv. 12, a letter of Termouthas’ husband Antonius to his mother-in-law, asking about some blankets for the winter. Another letter by Antonius to his mother-in-law is P.Mich.inv. 121. He also conveys his wife’s greetings: “Thermouthas sends you many, many greetings, and she reproaches you often because you do not send her a letter in reply.”

Authority, retirement and household hierarchies

123

transmissions afflicted intergenerational, gender, and sibling relationships and are usually associated with negotiations about rights and obligations of kin. Inheritance was only one of these, although the most important. Others were the allocation of dowries upon marriage, gifts inter vivos, or retirement contracts. In Roman Egypt, property transmission occurred usually at either the marriage of the donee for female heirs, or the death of the donor for male heirs.88 Parents provided their daughters with a considerable dowry upon their marriage, usually, however, not the complete share, but only a partial transfer that would be completed at the death of the girl’s parents.89 In Hellenistic and Roman Egypt husbands did not control the dowry of their wives. It served as a security measure in case of divorce or widowhood.90 If a girl’s parents died before she married, her brothers were expected to provide the dowry.91 Sons inherited at their father’s death, usually equally, at least if their father died intestate.92 If a father wrote a will, the eldest son often received a slightly larger share than his younger brothers, but generalizations are not possible, and partible inheritance as in pharaonic Egypt,93 even if not in its most radical form of complete equality of all the children, seems to have been the norm in Roman Egypt.94 The Athenian concept of the ἐπίκληρος did not exist in Graeco-Roman Egypt.95 Unlike the position in Athens, where the heiress was required to marry her father’s nearest male relative since she was not allowed to hold property in her own name, women in Roman Egypt could hold property in their own right and daughters were able to inherit the entire patrimony if there were no sons96 – as they had been able to do in pharaonic times.97 Inheritance customs in Roman Egypt did not follow archaic or classical Greek succession patterns.98 It may surprise us to learn that equal inheritance among all offspring regardless of gender existed parallel to virilocal marriage patterns. While the former speaks usually for late marriage for both sexes and neolocality upon marriage, the latter is a characterization of family systems in which women

88

89 91 93 94 95 96 97

Goody calls the dowry an “inheritance premortem,” which assigns to daughters a share of the paternal property – a counterpart to their brothers’ share, however, often not an equivalent (Goody 2000: 86–7). Tacoma 2006: 70, 208, 218. 90 Cf. Yiftach-Firanko 2003. P.Oxy. 6.907 from 276 ce; cf. Tacoma 2006: 69–70. 92 Tacoma 2006: 161, 207–8. Seidl 1962: 180; Pestman 1969; Lippert 2008: 154–6. Kreller 1919; Tacoma 2006: 208, 229, 250. Harrison 1968: 132–3, 151; Just 1989: 95–8; Ruschenbusch 1990; Karabélias 2002. Cf. Vatin 1970; Pomeroy 1984; van Bremen 1996; Yiftach-Firanko 2003. Seidl 1962: 180; Pestman 1969; Lippert 2008: 154–6. 98 Cf. Huebner in press c.

124

Growing old in the household

by contrast to their brothers received only a dowry, and were married relatively early to husbands who were significantly older.99 Support in old age was based in no small measure on parental control of productive resources such as land. According to the Roman succession and inheritance pattern, an old father was the head of household until his death. A similar succession pattern must have been at play in indigenous pre-Greek Egyptian society. Ptolemaic population counts seem to indicate that Egyptians favored multiple family households100 characteristic of lifelong authority of the head of household and transfer of property only upon his death. This is the system that still prevailed in Egypt in Roman times at least among the rural population.101 Even though about two-thirds of all men would have lost their fathers when they reached the age of full majority at 25, there were indeed adult men who were married, had children of their own, and occupied influential positions in their town, but were still financially dependent on their old fathers. An example is the councilor Ptolemaius from Oxyrhynchus in the late third century ce who tried to evade yet another liturgy by claiming that he did not have the means since he was still supported by his father.102 Patriarchal family systems typically involve a strong differentiation of authority along gender and generational lines,103 and this hierarchical system thus seems to have found its reflection in the census. In the Roman Egyptian census returns household members were usually listed in a specific way, with little variation from household to household. Sex, age, and closeness to the head of household were the determining factors for one’s rank in the (census return’s) hierarchy of household members. The oldest male usually acted as head of household and declarant, and he usually appears at the top of the list. He declared all the other household members by name and filiation. Further, he identified these persons as spouse, parents, children, or consanguineous relatives through their relationship to him, or in the case of the daughters-in-law and grandchildren, through their relationship to their husbands or parents. We do not know if this order represented the view of the head of household as the official declarant, or rather that of the scribe who drew up the census lists. While similarities to

99 101

102

103

Skinner 1997: 61. 100 Clarysse and Thompson 2006: 314. For patria potestas in Roman and later Roman Egypt, see Chapter 3 and Taubenschlag 1916; Lewis 1970; Beaucamp 2007: 278. P.Oxy. 12.1415: δέομαι ὑμῶν, οὑ δυναμαι. μέτριος εἰμι, παρὰ πατρὶ τρέφομαι. Cf. Tacoma 2006: 130. Oppenheim Mason 1992: 13–32.

Authority, retirement and household hierarchies

125

hierarchical orders in Middle Babylonian,104 Tang Chinese,105 or late medieval Florentine population counts are striking,106 sex discrimination was much more pronounced in Roman Egypt than in the population registers of these other societies. As we will see on the following pages, gender was ahead of age the most crucial factor in determining one’s rank in the household lists. It would certainly go too far to assume that the order family members were listed in the census reflected exactly the hierarchy that prevailed in their respective family or household; in practice one should imagine that it was subjected to even daily alterations depending who of the household members was present and which situation or problem had to be dealt with. Nonetheless, the fact that the head of household was listed first, males usually before females regardless of age and within one gender according to seniority, lodgers below family members and slaves always last, does imply that there was a culturally and socially determined reflection of authority and status within the household with a strict sense of male dominance and a hierarchy based on age.107 In Roman Egyptian society there were also separate sets of socially determined behavioral norms and roles within the family attached to each gender that dictated the sharing of work and decision-making in the household.108 Let us first have a look at how the members of a simple family household were listed. The husband and father as the head of household always ranked first, even if his wife was older than him. Children usually ranked below their parents. Sons, however, were always listed after their fathers.109 Sometimes sons were listed before their mothers even if the mother was the head of household and the son a minor.110 Most often, however, sons were registered after their mothers.111 In rare cases, even daughters were 104

Tenney 2011. 105 Cf., e.g., Lee and Campbell 1998: 128; Liao 2001. Klapisch-Zuber 1985: 54–5. 107 In a modern census children would certainly be listed in their order of birth, not first all sons and then daughters, even if they were older than their brothers. 108 Cf. Taubenschlag 1955: 130–49; Lewis 1970; Arjava 1998: 49; Rowlandson 1998: 184. 109 33-Ox-1; 33-Ox-2; 75-Ar-2; 89-Hm-1; 103-Ar-3; 103-Ar-9; 117-Ap-5; 117-Ap-6; 117-Ap-7; 117-Ar-12; 131Ar-12; 131-Ar-13; 117-Ap-10; 117-Ar-13; 131-Ar-20; 117-Ox-2; 131-Ox-1; 159-Ar-11 (a special case because this was a “brother–sister” marriage); 159-Hm-3; 173-Ar-9; 173-Pr-16; 187-Hm-1; 215-Ar-1; 215-Ar-5. Sons ranking above their mothers seem to have been the rule in the Oxyrhynchite nome (P.Oxy. Census 89-Pt-1; 89-Pt-7; 89-Pt-8; 89-Pt-10; 89-Pt-13; 89-Pt-14; 89-Pt-16; 89-Pt-19; 89-Pt-33; 89-Pt-35; 89-Pt-36; 89-Pt-37; 89-Pt-42; 89-Pt-47; 89-Pt-51; 89-Pt-56; 89-Pt-60). 110 33-Ar-2; 103-Ar-9. 111 103-Ar-1; 103-Ar-5; 103-Ar-11; 117-Ar-2; 117-Ar-4; 117-Ar-5; 117-Ar-6; 117-Ar-7; 117-Ar-11; 131-Ar-3; 131Ar-18; 131-Be-1; 131-He-2; 131-He-4; 131-Me-1; 131-Ox-16; 145-Ar-2; 145-Ar-3; 145-Ar-9; 145-Ar-17; 145Ar-22; 145-Ar-23; 145-He-1; 145-He-2; 159-Ar-1; 159-Ar-26; 173-Ar-3; 173-Ar-15; 173-Pr-5; 173-Pr-7; 106

126

Growing old in the household

registered before their mothers.112 Daughters were usually listed below sons, even if they were considerably older,113 and they were even listed below sons of their brothers, i.e., their nephews.114 Siblings of the same sex ranked according to age.115 We note a remarkable exception to pervasive sex discrimination in the records of the household of the unmarried, widowed, or divorced Methe, a 32-year-old declarant who was the head of household even though she had a maternal half-brother and a male cousin living with her.116 She lists in order her 3-year-old daughter and her 1-year-old son before her brother and cousin, and even before her 48-year-old mother. The reason might be that she was the owner of the house. We get a highly interesting view on the inner hierarchies of so-called patchwork families from these nuclear families that were composed of couples who both had been married before and had brought children by former partners into the relationship. A stepson stood in his stepfather’s household at the bottom of the hierarchy,117 listed even after a daughter-inlaw and the patriarch’s nephews.118 We have one case, however, where the wife in such a patchwork family owned the house. Here she occupied the first rank, followed by her husband and then her adult son from her previous relationship.119 If a man had married a second time and had children living with him by both his first and second wife, the children from his first union sometimes ranked below those from his present marriage, even though they were older.120 However, we also have cases where these half-siblings were registered according to the usual order of sex and age.121 In one extreme case a man, formerly married to his full sister, divorced her and married again. His ex-wife and sister, however, still lived with him and his new family in the same household, evidently because she had nowhere else to go. She and her two children, a son and a daughter, 173-Pr-10; 173-Pr-11; 173-Pr-13; 173-Pr-15; 187-Ar-2; 187-Ar-4; 187-Ar-8; 187-Ar-9; 187-Ar-11; 187-Ar-12; 187-Ar-23; 201-Ar-1; 201-Ar-10; 215-Ar-3; 215-He-1; 215-He-2; 243-Ar-1; 243-Ar-3; 243-Ar-4; 257-Ar-1; ???-Ar-3; ???-Ar-5; ???-Me-1; P.Oxy.Census 89-Pt-24. 112 61-Ar-1; 131-Oa-1; 145-Oa-2; 201-Ar-9; P.Oxy.Census 89-Pt-42; 89-Pt-63. But in general they followed their mothers: 103-Ar-1; 103-Ar-3; 117-Ap-5; 117-Ap-6; 117-Ap-7; 117-Ap-8; 117-Ar-2; 131-Ar-3; 131-He2; 131-He-4; 131-Ox-1; 131-Ox-6; 145-Ar-3; 145-Ar-9; 145-Ar-12; 145-Ar-23; 145-Ly-1; 145-Ox-1; 159-Ar7; 159-Ar-10; 159-Ar-20; 159-Ar-21. 113 Exceptions: 131-Pr-1; 173-Ar-11. 114 89-Hm-1; 117-Ap-6. 115 Exceptions: 173-Pr-5 (the 3-year-old Isidoros is listed before his 35-year-old half-brother and his 21year-old full brother); 173-Pr-3 (the 44-year-old Therothbechis is listed above her 48-year-old full sister Therobastis). 116 173-Ar-11. 117 187-Ar-22 (a 5-year-old daughter from the couple’s present marriage ranked above her adult halfbrother on the mother’s side, but below her adult half-brother on her father’s side). 118 145-He-2. 119 131-Me-1. 120 187-Ar-8; 201-Ar-5; 215-Ar-4; P.Oxy.Census 89-Pt-21; exception: P.Oxy.Census 89-Pt-33. 121 131-Ox-1.

Authority, retirement and household hierarchies

127

were listed above her ex-husband’s second wife and their two daughters, perhaps because of their closer consanguinity.122 As a daughter of the house and mother of heirs to the family, she and her offspring remained of high standing and well respected in her father’s and ex-husband’s household. Orphaned cousins and nephews living with their relatives always ranked below the children and siblings of the head of household – regardless of their sex or age.123 However, if they were male, they did not necessarily rank below the declarant’s wife or his elderly mother.124 The same applied to childless unpropertied elderly uncles125 or aunts126 who lived in the household of a nephew or niece. They usually held a very low rank. In households in which we find several conjugal couples of different generations living together under one roof, the elderly father usually occupied the function of the head of household. His wife regularly ranked directly below him, above their common sons and daughters.127 In the only case we have where an elderly father had yielded his authority to one of his adult sons, he moved to the bottom of the list and was listed after all the other family members.128 Loss of his wife did not change the position of the household’s patriarch. He remained in his leading position. Some widowers even married again quite late in life and had more children with their second, considerably younger wives.129 If, however, the patriarch died and one of his sons succeeded his father as the head of the household, his widow’s position, judging from her rank in the census lists, was apparently considerably weakened. After her husband’s death, she usually dropped in the hierarchy of listed household members to the lowest rank.130 While she held a position of authority over her daughters-in-law during her husband’s lifetime, after his death her daughters-in-law,131 her grandchildren,132 and even her grandsons’ wives133 ranked above her in the census returns.134 In just a few cases we see that the widowed former wife of the head of household did not lose ground, but became herself the new head of household, even when she had 122 126 130

131

132 134

159-Ar-5. 123 145-He-2; 173-Ar-11; 187-Ar-4; 187-Ar-26. 124 145-Ar-20. 125 215-Ar-4. 117-Ar-1. 127 117-Ar-11; 187-Ar-4. 128 131-He-3. 129 e.g., 173-Pr-5; 187-Ar-8: 187-Ar-22. 11-Ar-1; 117-Ar-6; 117-Ar-7; 131-He-2; 145-Ar-3; 145-Ar-20; 159-Ar-10; 173-Ar-9; 173-Ar-11; 173-Pr-17; 215-Ar-5. 117-Ar-6; 117-Ar-7; 131-He-2; 145-Ar-20 (“brother–sister” marriage); 159-Ar-11 (a special case because this was a “brother–sister marriage”); 173-Ar-9 (again “brother–sister” marriage); 215-Ar-5; P.Oxy. Census 89-Pt-36; 89-Pt-44. For the same phenomenon in late medieval Tuscany, see Klapisch-Zuber 1985: 54. 11-Ar-1; 117-Ar-6; 117-Ar-7; 131-He-2; 145-Ar-3; 173-Ar-9; 173-Ar-11. 133 145-Ar-3. Cf. Chapter 6 on the daughter-in-law/mother-in-law relationship. For the relations between elderly parents and their adult children in nineteenth-century Italy, see Kertzer and Hogan 1995: 128.

128

Growing old in the household

adult sons living with her. In these cases where the widow took over the formal position as head from her late husband, her sons, daughters, and daughters-in-law remained listed below her.135 Ownership and headship of household did not necessarily coincide; gender was nonetheless more decisive. In one case, two sisters jointly served as declarants of the household and probably owners of the house. However, an older brother ranked first in the list of household members.136 Finally, let us look at those census returns that record several conjugal families of the same generation living together in one household, that is the frérèches. In those cases where several adult brothers with their families lived under one roof, either the conjugal units were listed one after the other according to the birth order of the brothers,137 or first were listed all the brothers according to their age, then all the latter’s sons, and then all their wives and daughters, according to the order that had been chosen for the males to whom they belonged.138 That means that in cases where the head of household was married, his wife either followed immediately after him, even before his brothers, or she was listed among the other women, most often as the first one among them.139 If conjugal units were listed en bloc, they were ranked according to the father’s ranking among his brothers. Unmarried sisters of the head of household ranked below all the brothers, but often, even if they were younger than them, before their brothers’ wives, i.e., their sisters-in-law.140 A woman ranked among her sisters-in-law, the wives of her husband’s brothers, not according to her own age but according to her husband’s position among his brothers. Daughters-in-law usually ranked below their mother-in-law, at least as long as the old patriarch was still alive.141 Either sons of the head of the household were listed immediately after their father and before their uncles (sometimes the head’s daughters were also treated this way),142 or all the children of the household followed their respective mothers further down on the list.143 The ranking 135 136 137

138 139

140

141 143

117-Ar-12; 131-Ar-12; 173-Pr-15; 201-Ar-1; 201-Ar-8; 201-Ar-9; 201-Ar-10; P.Oxy.Census 89-Pt-27. 173-Pr-3. e.g., 117-Ar-4; 117-Ar-11; 131-Pr-1; 173-Me-3; 173-Pr-5; 173-Pr-10; 201-Ar-9; 215-Ar-3; 215-Ar-6; 215-He3; ???-Ar-3; ???-Ar-5. 145-Ar-19; 159-Ar-4; 159-Ar-21. 89-Ar-1; 117-Ar-4; 117-Ar-7; 131-Ox-16; 59-Ar-26; 187-Ar-4. For a man listed below his wife who was the head of household, see 103-Ar-7; 173-Pr-7. 117-Ap-5; 117-Ap-6; 131-Ox-1; 159-Ar-4; 187-Ar-4 (the 29-year-old Thermoutharion ranked even below her 1-year-old nieces). Sometimes sisters-in-law preceded the unmarried sisters of their husbands: 117-Ar-11; 173-Pr-5; 215-He-2; 215-He-3. 117-Ap-5; 117-Ap-6; 117-Ar-11; 131-Ox-1; 173-Pr-5; P.Oxy.Census 89-Pt-47. 142 159-Ar-10. 33-Ar-1; 117-Ar-4; 117-Ar-7; 117-Ox-1; 131-Pr-1; 159-Ar-4; 159-Ar-5. In only one case the father’s halfbrother is listed before his son: 145-Ar-18.

Authority, retirement and household hierarchies

129

of sons and daughters depended on their father’s standing in the extended household. Regardless of which system was used, the male head of the household and his immediate family (that is, the wife and children) thus occupied a special position in the household. His wife ranked first among all the other women, his children first among all the other children. As we have seen, when the old patriarch died, normally his eldest son acquired the headship by inheritance.144 Letters addressed from younger brothers to the eldest support the thesis that the eldest brother acted as the leading figure of the family not only on paper but also in everyday life, where he occupied a position of special reverence among his siblings. The eldest brother was the one in whose name the family’s holdings were registered, even if it was owned jointly by him and his brothers. Together with the headship, the eldest of the brothers also took over the social role of his father for his minor siblings who had not yet reached majority. We have heard above of Sempronius, the eldest of several brothers, that he reminded his younger brother Maximus that in his absence it was now his responsibility to assume authority in the household: “As long as I am gone, you ought to be called father.”145 Sempronius admonished Maximus to reprimand their younger brothers if they behaved disrespectfully towards their widowed mother. Again the sensitivity with which Sempronius reminds his younger brother of his duties and the respect due to their elderly mother is surprising: “I know that without my letter you are able to please her. But do not take amiss the advice of my letter.” At the wedding of a young woman in early fourth-century Oxyrhynchus, it was not the parents but a brother who sent out the wedding invitations for his sister. Their parents had likely already died.146 In yet another case, from early third-century Oxyrhynchus, a certain Apollonios sent out invitations for the coming-of-age party of his younger brothers, representing their late parents.147 Not all men could expect to ever become head of their own household. Possibilities for younger sons to become head in their own right were normally either to outlive their elder brothers, to set up a household on

144 146

147

Cf. section 2.10. 145 SB 3.6263; trans. Winter 1933: 49. P.Oxy. 12.1487: “Theon, son of Origines, invites you to the wedding of his sister tomorrow, which is Tubi 9, at the 8th hour.” P.Oxy. 12.1484: “Apollonios requests you to dine at the table of the lord Sarapis on the occasion of the approaching coming of age of his brothers at the temple of Thoeris . . .” For the μελλοκουρία see Montserrat 1996: 39–41.

130

Growing old in the household

their own, or to marry uxorilocally, which meant marrying a woman without brothers, moving into her parents’ household, and then succeeding to his father-in-law’s position when he died.148

5.9 Female heads of household Among the Roman Egyptian census population the number of female heads of household was high. An analysis of the Roman census returns yields about 75 percent male heads of household and about 25 percent female heads of household. Surprisingly the percentage of female-headed households was considerably lower for Ptolemaic Egypt, despite similar demographic contingencies that must have left many mothers widowed at a young age with young children. Of the total number of households taken into consideration by Clarysse and Thompson, just 11.5 percent were headed by females. Among the Greek population in fact just 9.8 percent of all households had a female head, while among Egyptian households the number was at least 12.9 percent.149 None of these households contained an adult male. Since the Ptolemaic documents do not record minors we cannot say whether these households were headed by single, divorced, or widowed women or women who were still married but whose husbands were away for work or for army service.150 Most of these women who acted as head of household in Roman Egypt were widows who had taken over the headship from their late husbands and were managing the estate, as long as their children were minors. Once their sons reached majority they were supposed to hand over the headship and the patrimony undiminished. We might assume that in those cases where an elderly widowed mother had not ceded the headship of the household to her adult sons she was in fact the owner of the house inherited from her family of origin. The 55-year-old Tetanoupis from second-century Memphis declared the household in which she lived with her 33-year-old son from a previous marriage and her second husband.151 The 64-year148

149

150

We have one exceptional case from second-century Arsinoe in which 31-year-old Herodes was the head of the household even though he was the youngest of three brothers. His brothers were 45 and 33 years old. All lived together under one roof (159-Ar-4). Probably he acted as head of household because he was the guardian of the two minors who owned the building the family was living in. Clarysse and Thompson 2006: 301–2. (There is a calculation mistake on p. 302, n. 274: it should be 46 female household-heads in 403 units in order to calculate the percentage of households headed by women in general! Alternatively, in order to calculate the percentage of households headed by women among the Greek population, the numbers would be 25 female household-heads in 255 units.) Clarysse and Thompson 2006: 302. 151 131-Me-1.

Property transmission during lifetime

131

old Kronous from early second-century Arsinoe, the head of household, declared the household in which she lived with her 46-year-old son, his wife, and their three children.152 By chance also the next census return from this household, drawn up fourteen years later, has come down to us. Here all family members were still alive. Kronous was then 78 years old but still functioned as head of household.153 To look at yet another family from early third-century Karanis, the 74-year-old Tasoucharion lived with her two adult sons, 56 and 46 years old, her 56-year-old daughter, and a 6-year-old great-granddaughter, but she was the one who acted as declarant of the household.154 This distribution of households headed by women as opposed to men is about the same for both the metropoleis and the villages. Female heads in the cities in Roman Egypt were, however, on average much younger than those in the countryside. It is not surprising that women in the metropoleis started at a younger age to take over the headship than did those women in the countryside because, as stated above in Chapter 4, couples in the cities were younger when they established an independent household. For male heads of household we can also see interesting differences between rural and urban households. While in the countryside in Roman Egypt men over the age of 55 accounted for 30 percent of all the male heads of a household, the number of heads of household over the age of 55 dropped in the cities to about 13 percent of all male heads of household whose ages we know. This can also be explained by the fact that young men in the cities were more likely to set up their own household early than to stay for an extended time with their parents or brothers, which increased the proportion of urban nuclear households headed by younger men.

5.10 Property transmission during lifetime Property transmission to the next generation during one’s lifetime took place in Roman Egypt only rarely. It was the exception that elderly fathers gave up control over the management of the household to their sons while they were alive, even if they had grown-up sons who were married and had children themselves. We have only one instance in the Ptolemaic tax lists,155 152 154

155

117-Ar-12. 153 131-Ar-12. 201-Ar-9. Cf. also the 57-year-old Sempronia, who lived with her adult son, whose age is lost (131-Ar13), and the 59-year-old Thenpestsokis, who lived with her 33-year-old son and his wife and baby girl (201-Ar-1). Clarysse and Thompson 2006: 296.

132

Growing old in the household

and only one in the Roman census returns for a transfer of the patrimony during a father’s lifetime.156 In classical Greece, however, it was the rule that fathers turned over the estate to their children during their lifetime, and this custom is also widely documented for the early modern period in many north-western European societies where the stem family and impartible inheritance were common.157 In these north-western European societies, the patriarch stepped down at a certain point from his role as head of household and handed the ownership of his property down to his eldest son in exchange for a contract securing his support in old age. This succession pattern is usually called by anthropologists “filiocentric” (as opposed to “patricentric”). In classical Athens it seems to have been a common practice for men upon their sixtieth birthday or upon the marriage of their son to transfer their property to the younger generation.158 However, this arrangement, i.e., a property transfer inter vivos, is held responsible for an increased potential for intergenerational conflicts and a worsening living situation for the elderly. In societies in which parents retired at a certain age and handed over the management of the farm to one of their children in return for support in old age, considerable tensions between the generations and a decidedly pragmatic and harsh treatment of the elderly once they were deprived of their power have been noted.159 As their status and influence declined, the elderly were in danger of being regarded as a burden, and of being neglected. Nor did many old men feel very comfortable relinquishing all their power and property to their sons, since it left them with little bargaining power and made them dependent on their sons’ goodwill and sense of moral obligation. Our sources from classical Athens provide us with several instances where elderly parents were mistreated, regarded as a burden, neglected, and resented. The old father wished to postpone the relinquishment of his power as far in the future as possible, whereas the son was no longer willing to wait for his turn.160 Contentions and disputes regarding the point of time when this transfer of ownership and managerial 156

157 158 159

160

The 70-year-old Petechon lived with his 27-year-old son Piathres, who already in his father’s lifetime had taken over the ownership of the house (131-He-3). Other old fathers functioned as heads of their households far into their sixties and seventies (75-Ox-1; 117-Ap-5; 117-Ox-2; 131-Ar-8; 131-Ar-14; 131Be-1; 131-He-4; 145-Ar-12;145-Oa-1; 145-Ox-2; 159-Ar-1; 159-Hm-3; 173-Ar-9; 173-Me-3; 173-Pr-5; 187Ar-8; 187-Ar-22; 215-He-3). Berkner 1972; Goody 1982; Gaunt 1983; Moring 1998. MacDowell 1978: 92; Golden 1990: 111–14; Schmitz 2007: 35–6. Gaunt 1983, 1987; Robin 1984; Plakans 1989; Sorensen 1989; Wall 1995. For similar arrangements in eighteenth/nineteenth-century Austria, see Berkner 1972; Sieder and Mitterauer 1983. Cf. for instance Dem. 47.34–5. See Lacey 1968: 119–20; MacDowell 1978: 92; Golden 1990: 111–14; Strauss 1993: 66–72; Schmitz 2004: 94–8; 2007: 106–8 for further references.

Property transmission during lifetime

133

authority from the older to the younger generation should take place were common, even giving rise to rifts between sons and their parents in some cases. We just have to think of the old Philokleon in Aristophanes’ Wasps.161 By contrast Iophon, the son of the tragedian Sophocles, unsuccessfully brought a charge against his father accusing him of mental incapacity, in order to get the administration of his property taken out of his hands.162 A peculiar institution in Roman Egypt constituting rather the exception than the rule was the father handing over his entire estate to his children during his lifetime but retaining the right to enjoy its profits. After the transfer, the father could no longer dispose of the property as he wished or alienate it without his children’s permission, but continued to derive the revenues from the property that was formally assigned to his children.163 This institution finds its equivalent in later Roman law on patria potestas, reflecting potentially customs in the Eastern provinces.164 Legally a son standing under patria potestas could not own anything in his own right. Property that the son acquired by his own labor or through a donation or legacy, however, called peculium adventicium in Justinian times, belonged to the son, but the father enjoyed a lifelong usufruct.165 In the Egyptian cases the property under question seems to have been, however, the paternal estate. The corresponding law in Roman Egypt was the one that gave a father the right to the property of his predeceased son; only after the father’s death would it devolve to the closest next of kin. Both laws, being attributed to “The law of the Egyptians,” a collection of Greek and indigenous laws, thus protected the family estate, which stressed once more the importance paid to the unity of the paternal property.166 Revillout found the earliest evidence for this institution in Demotic deeds of Egypt under Persian rule (525–332 bce).167 According to Kreller, this formal Anwartschaft (prospective entitlement) of the children to the paternal estate survived through Ptolemaic times, as a couple of Demotic papyri show, and it still pops up, but in vanishingly low numbers in our Greek documents from early Roman

161

162

163

164 167

Cf. MacDowell 1971; Gadamer 1976; Martin 1984: 105–6; Hubbard 1989; Minois 1989: 49–65; Garland 1990: 286–7; Baltrusch 2003: 702. Cf. Aristoph. Cl. 844–5; Aristot. Ath. Pol. 56.6; Isaeus 6.9; Aeschin. 3.251; Cic. Senec. 7.22. Cf. Jouanna 2007: 120. Cf. Kreller 1919: 204–7; Pestman 1995: 80–1; Yiftach-Firanko 2009: 550–1. The edict of Mettius Rufus, prefect of Egypt in 89 ce, refers to this custom (Boak 1933: 1.121). Cf. P.Oxy. 42.3015 from 117 ce. CIust 6.61; NovIust 117.1. 165 Kaser 1975: ii. 166 Cf. Yiftach-Firanko 2009: 551. Revillout 1912.

134

Growing old in the household

times.168 Egyptian inheritance practice differed in this regard considerably from Roman traditions, and occupied the middle ground between Greek and Roman rules of succession. While in classical Athens a father stepped down and handed over the family estate to his son at a certain point in his life, the Roman paterfamilias stayed in absolute power until his death. Indigenous Egyptian inheritance practice, as far as we can tell from the Demotic and very few Greek documents, gave the children a say regarding sales and leases of the family estate, and limited the father’s ability to make a will. It seems, however, that in Roman times the children’s say had to be contractually agreed to be valid; Roman law naturally did not recognize this children’s right.169

5.11 Retirement contracts As just mentioned we occasionally come across cases recorded in our papyri from Hellenistic and Roman Egypt where parents drew up an agreement with their children settling contractually their old-age maintenance.170 The custom of drawing up formal “retirement contracts” between parents and children was not limited to Egypt in the ancient Mediterranean. The contracts drawn up by parents and their children in the ancient Near East, for instance, were of a similar scope. They either specifically mention the form and degree of care the two generations agreed upon (e.g., the amount of grain, oil, and clothing and meat on special occasions), or they speak of care and support in general terms.171 In ancient Anatolia, elderly parents initiated a “brotherhood contract” among their grown-up sons which was destined to prevent the dissolution of the household as long as the parents were alive. The sons obliged themselves to continue living together in the parents’ household, not to accumulate private capital, not to ask for their share of the property, and to take care of the (surviving) parent(s).172 Retirement contracts in Roman Egypt could take various forms. They could be part of a marriage contract, a deed of gift, or part of a will. Some left 168

169

170 172

Kreller 1919: 180–1. Cf. P.Oxy. 2.237, col. iv, v, viii after 186 ce; P.Oxy. 4.713 from 97 ce; P.Giss. 1.30 from 140–61 ce; UPZ 1.118 from 83, 136, or 147 bce; BGU 3.998 from 101 bce; CPR 1.4 after 52 ce; CPR 1.188 from 106/7 ce; P.Bas. 7 from 117–38 ce; BGU 1.76 from the second or third century ce. Cf. Kreller 1919: 49: “Nun ist allerdings mit der Tatsache zu rechnen, dass dieser Gedanke der Hausgemeinschaft mit dem Vorwärtsschreiten des modernen Geistes in der griechischen Welt mehr und mehr abgestorben ist. Das ergibt sich für das Recht der Papyri deutlich daraus, dass wir von Wartrechten der gesetzlichen Erben nur noch geringe Reste nachweisen können.” See also Kreller 1919: 178–200. Kreller 1919: 204–7; Pestman 1995: 80–1. 171 Greenfield 1982: 309–16; Westbrook 1998: 10. Veenhof 1998: 145–60.

Retirement contracts

135

only part of a father’s property to a son enabling him to establish his own household,173 while others left a father’s property more or less undivided and complete to one or all of his sons during his lifetime, with the eldest son succeeding to the place of his father and the property remaining intact.174 The respective obligations of children and parents were outlined in detail. The early retirement contracts from Ptolemaic Egypt clearly rested on Greek custom. In the early third century bce, a couple, both originally from Temnos, a small Greek city-state, on the western coast of Anatolia, and now living in Elephantine in Upper Egypt, appointed each other heirs, and their common sons as heirs of the surviving spouse. In the event that their sons got married before both of their parents had died, ownership of their property was destined to go over to the sons in exchange for a lifelong pension payable to the parents and the obligation to take care of a proper funeral for them.175 Let us look in more detail at the “retirement contracts” that we have from Roman Egypt: an example is the contract P.Mich. 5.322 from Tebtunis, dating to 46 ce, which belongs to the archive of Kronion, son of Apion, nomographos of the grapheion of Tebtunis. It was drawn up between an elderly couple, 69-year-old Psyphis and his wife Tetosiris, and their two sons, two daughters, and grandson.176 It was registered through the record office in Tebtunis and Kerkesoucha Orous, two villages in the south-eastern Fayum. From the locations of the property that Psyphis and Tetosiris left behind, we can deduce that Psyphis originated from Tebtunis, while his wife came from “Kerkesis” (Kerkeosiris), a village a short distance to the north-west of Tebtunis. Psyphis dictated this text to a scribe, and the style of the contract is colloquial and in parts ungrammatical. In this contract, the old Psyphis and Tetosiris divided their entire property among their children and one grandson. The term used here is μεμερικέναι “divide among.” The children and grandson were to take immediate possession of the property. The elderly couple declared that from this point they should not be the owners of their property anymore and should have no power to sell or mortgage it, alienate it, leave it to anyone else, or otherwise dispose of it, but that their children should have full control over it. And whatever they wanted to do with it, they did not require the consent of their parents. The elderly couple did not even retain any usufruct rights. Then the 173 174 175 176

See, e.g., P.Petr.2 1.3 from 238/7 bce from Crocodilopolis (later Arsinoite). P.Cair.Goodsp. 6 from 129 bce; cf. Kreller 1919: 208. P.Eleph. 2 from 284 bce; cf. Kreller 1919: 225–6; Bagnall and Derow 2004: 240–1, no. 147. Cf. Rupprecht 1998: 234–5.

136

Growing old in the household

contract comes to speak about what the heirs in return were required to provide for the old couple. Their two sons, Onnophris and Psyphis, and their grandson, Psenkebkis, had to furnish Psyphis and Tetosiris, as long as they lived, with 2 artabai of wheat and 6 kotylai of oil each month, and with 300 drachmas a year for other expenses and clothing. It was not uncommon for a husband to lay down in his will that his heirs had to pay his widow a lifelong pension in kind and money since women usually did not inherit from their husbands.177 But it was rare that he himself entered into such an arrangement whereby his heirs had to pay him a lifelong pension. Furthermore, the heirs in this case obliged themselves to cover all the couple’s debts. Finally, they had to provide Psyphis and Tetosiris with a proper funeral after their deaths according to their rank.178 What makes this case so special is not just the aspect of contractually regulated old age maintenance, but also the fact that it concerned the entire property. Kreller in his Erbrechtliche Untersuchungen (1919) called this form of pre-mortem transfer of the entire property Gutsabtretung, by contrast to Abschichtung, a transfer that affected only a certain share of the paternal estate. Such a Gutsabtretung is very rarely documented in our sources: one of the few parallels for such a retirement contract comes from around the same time from the same archive of Kronion, son of Apion, nomographos of the grapheion of Tebtunis, which makes us think of a local custom. This second contract dates to the year 42 ce, and is thus just four years older than our just discussed case.179 The 65-year-old Orseus left land, building sites, furniture, and household goods to his three sons and one daughter. Just as in the previous case, his eldest son, Nestnephis, received more than his younger brothers.180 Orseus acknowledged that from the present day he gave up all power to sell any of the aforesaid property or to mortgage it or to give any part of it to any children except the aforementioned. And whatever he did contrary to these conditions should be null and void. In return for receiving the major share, the eldest son, Nestnephis, was required to pay all the public taxes in money and in kind. Additionally, he was obliged to provide his father as long as he lived, from the present day annually in the month Drousilleios, with 12 artabai of new, pure, and unadulterated wheat, and with 12 drachmas of silver for clothing, oil, and other expenses. The old couple from the contract discussed above, Psyphis and his wife Tetosiris, 177

178

179

P.Fay. 124; cf. Kreller 1919: 365; BGU 1.86 from 155 ce; cf. Kreller 1919: 365; P.Lond. 3.932 from 211 ce; cf. Kreller 1919: 134, 365. For similar cases, see P.Eleph. 2 from 284 bce; BGU 4.1013 (Claudius or Nero); P.Oxy. 3.472 from c.130 ce; P.Grenf. 2.71 from 244 ce; P.Lond. 3.1044 from sixth-century Hermopolis. P.Mich. 5.321. 180 Codex Hermupolios 9.9–11. See Lippert 2008: 16, 36, 61, 125.

Retirement contracts

137

also asked for 1 artaba per month per person, which thus seems to have been the standard ration. Instead of 12 drachmas for additional expenses, however, their heirs had to come up with 300 drachmas per year. We know of only a handful of comparable cases. In a retirement contract from early imperial times, a woman in her fifties from the Arsinoite nome transferred her property to her three daughters; her husband assisted as her guardian. In exchange for the transition of property rights, she stipulated that her daughters had to provide her with regular provisions, clothing, and a cash allowance as long as she lived.181 In a similar case a certain Hermione from second-century Oxyrhynchus bestowed on her daughter her property in exchange for a regular pension payable until her death.182 In another document from 244–8 ce from Hibis in the Kargha Oasis a father transferred his complete property to his two sons during his lifetime. The lower part of the document is heavily destroyed and not readable, but Kreller assumes that we had here similar provisions for the father’s maintenance during the last years of his life.183 Where a contract bound one child to provide care in reciprocity for inheritance, it seems unlikely that the other children provided assistance. A certain Isidoros from early third-century Hermopolis was willing to relinquish all rights to his paternal inheritance, both from his father’s and his mother’s side, under the condition that his two brothers shouldered the debts of their parents and took care of their widowed mother as long as she was alive, and provided her with a proper burial after her death.184 In a letter from mid-second-century Euhemeria in the Arsinoite nome, a certain Theogiton, probably a relative or friend, severely reprimanded another Apollonios that he did not comply with his duties to provide for his old mother and was defrauding her of her rightful allowance (chorēgia). Lines 13f. seem to allude to a contractual obligation, probably laid down in the latter’s father’s will.185 Theogiton threatened to initiate legal proceedings if Apollonios was not willing to comply with his duties. Here we are not dealing with a retirement contract per se but with provisions payable by the son and heir to his widowed mother who had not inherited from her husband. We encounter a similar case in the will of Stotoetis from midsecond-century Soknopaiou Nesos in the Arsinoite nome. Stotoetis’ widow was destined to receive a lifelong pension from the heirs in kind and cash.186 181 182 183 184 186

BGU 4.1013 (Claudius or Nero); cf. Kreller 1919: 209, 365. P.Oxy. 3.472 from c.130 ce; cf. Kreller 1919: 365. P.Grenf. 2.71 from 244–8 ce; cf. Kreller 1919: 212–13. P.Lond. 3.932 from 211 ce; cf. Kreller 1919: 134, 365. 185 P.Fay. 124; cf. Kreller 1919: 365. BGU 1.86 from 155 ce; cf. Kreller 1919: 365.

138

Growing old in the household

Up to the Byzantine and early Arab period of Egypt we find the occasional “retirement contract” drawn up between parents and children in order to formally organize old-age support. An old man from sixth-century Antinoopolis left his daughter his property as a donatio mortis causa in return for her continuous care in his sickness and old age.187 An old man named Daniel from the eighth-century Coptic town of Jeme drew up an agreement with his daughter Mariam and her husband, Papnute, in which they settled on the terms of care for the old man. Mariam and Papnute agreed to care for the old Daniel in exchange for ownership of a house, which was to be transferred immediately. Should Mariam and her husband fail to care properly for the old father, the house would pass back into Daniel’s ownership.188 We might wonder why some parents chose to draw up such a contract with their children. Did they prefer to have a written deed when they suspected that their children might not honor the moral obligation to provide for them when they were no longer able to care for themselves? Alternatively, one or both parties might have voted for a legally binding contract after a more informal agreement between them had not been met. The contracts just discussed do not mention any potential motives. There does not seem to have been a special occasion, such as the marriage of a son or a fatal illness of one of the parents, which could have induced them to give up all rights to their property during their lifetime. And their property, at least that of Psyphis and Tetosiris, was considerable, consisting of land, vineyards, storehouses, fishing, dykes, wells, a watchtower, several multistory houses in Tebtunis, vacant building sites, a barn, slaves, cattle, and all their household furniture and utensils, while the maintenance they asked for in return was modest by all means. The only other motive that comes to mind is perhaps the avoidance of taxes.189 An inheritance tax in Egypt is believed to go back to pharaonic times, with the state collecting 10 percent of the value of the land that was transferred. To enforce these taxes, a land register was kept by the Egyptian tax collectors.190 In Ptolemaic times paying the inheritance tax was a precondition for succeeding to an estate.191 Everyone seems to have fallen 187

188 189 190

191

P.Cair.Masp. 2.67154 v from 566–73 ce: ἐπειδήπερ πᾶσαν ἀνάπαυσι̣ ν καὶ ἐπιμέλειαν ἐποίησας μοι ἐν τῷ ἐμῷ νόσῳ καὶ γήρᾳ. P.KRU 116; cf. Wilfong 2002: 83. For other societies, see: West 1893; Shultz 1926; Lymer and Hasseldine 2002: 25. Wallace 1938; cf. Jouguet 1911: 356. We are in dire need of a comprehensive update and expansion of Wallace’s studies on the Roman Egyptian tax system. Seidl 1962: 183–4; Lippert 2008: 125.

Conclusions

139

under this tax, and no exemption was granted even for direct lineal descendants. Wallace believes that it is most likely that this inheritance tax continued into Roman times, even if we do not have any proof for this.192 This indigenous inheritance tax needs to be distinguished from the Roman vicesima hereditatium that only Roman citizens were liable to pay.193 By masking the transfer of property from the parents to their children as a maintenance contract, i.e., as an exchange, the family might have aimed at avoiding the inheritance tax that would have been due upon death. Even though they could not avoid all taxes, since a tax had also to be paid for the transfer of property, the tax on gifts of property, the telos ekstaseos, was assessed at a lower rate, though the exact percentage is not known.194

5.12 Conclusions In conclusion, in Roman Egypt growing old usually did not mean a sudden loss of power and status for the old father at some discrete point. Only if the old patriarch wished so, he could give up step-by-step some of his responsibilities to his adult sons. He and his wife were not at the mercy of the younger generation but remained the head of the household. Aging thus seems to have been characterized by a smooth transition of power, which became final at the old father’s death. Old men and women also seem to have continued to be active until physical incapability prevented them from working. This had the advantage that there was no sudden change in the pattern of daily living at a defined point regardless of one’s physical capability. The elderly parent did not feel useless but was still able to contribute to the household’s income and felt socially involved. An old man’s sons helped him by working the land or practicing his business, but he remained the owner of the family estate and head of the family. In societies in which the old patriarch remained the owner of all the family property until his death and only relinquished managerial authority gradually to his sons, relations between the generations living in one household seem to have been more favorable, and the elderly enjoyed a stronger 192 193

194

Wallace 1938: 234. In Rome the inheritance tax was introduced in 6 ce, but only had to be paid if the heirs were not descendants in direct line and the property was worth more than 100,000 sesterces (cf. Wilcken 1899: 345; Kreller 1919: 98–100; Gilliam 1952: 397–405; Duncan-Jones 1990: 195). The Roman inheritance tax, as its name vicesima hereditatium indicated, constituted 5 percent of the value of the property. Allowances were made for funeral expenses. For sales of property, the so-called enkyklion was assessed. In the first centuries ce, it comprised 10 percent of the value of the real property that was sold (Loewenstein 1973: 301). Wallace 1938: 227.

140

Growing old in the household

position than in family systems where the elderly father transferred his property during his lifetime. An argumentum ex silencio, which we should not overrate of course, is the fact that we do not find any authority struggles between fathers and sons over ownership of property in our evidence from Roman Egypt. As already discussed in Chapter 4 and section 5.9, when the old patriarch died, his wife, however, did not possess a similar powerful position. Ownership did not go to her but to the couple’s children. For this reason men often provided their widows in their wills with a lifelong usufruct. The transfer of property from the older to the younger generation during the patriarch’s lifetime seems to have been the exception in Roman Egypt, while it was the rule in classical Athens. Very few actual contracts between parents and children have survived for Roman Egypt that formally regulated pre-mortem property transfer and old-age support. They might be seen as evidence that care for the elderly was not always exclusively defined by the customary obligation of children toward their parents, and that an old patriarch did not always want to rely merely on the parental authority that was traditionally invested in him. Sometimes it seems to have been necessary to lay down contractually the responsibilities, commitments, and liabilities of both parties. Other motives, such as tax evasion, however, might also have played a role. Be this as it may, the contracts discussed above offered the advantage of security for the elderly and justice for children who acted as caregivers.

chapter 6

The patriarchal household and the incoming daughter-in-law

It is a great benefit both for the support and for the advantage of widows, that they so train their daughters in law as to have in them a support in full old age, and, as it were, payment for their teaching and reward for their training. For to her who has well taught and well instructed her daughter in law, a Ruth will never be wanting who will prefer the widowed life of her mother in law to her father’s house, and if her husband also be dead, will not leave her, will support her in need, comfort her in sorrow, and not leave her if sent away; for good Ambrose on Ruth 1:20 11 instruction will never know want.

6.1 The “mother-in-law belt” While the census returns constitute an invaluable source for household composition, they tell us very little about household dynamics, alliances and tensions between household members. For common sources of conflicts between parents and children, husbands and wives, mothers-in-law and their daughters-in-law, we have to look for different sources, private letters or literary accounts, and also comparative studies on peasant families in other patriarchal societies. How was living experienced in a joint family household? We know from other traditional patriarchal societies that life in a multiple family household did not mean only harmony. One of the main issues that often led to severe tensions was the fact that the male head and his wife imposed their views and thoughts on the younger couple(s), constantly interfered in their conjugal matters, and transgressed their personal space. This often led to repression and suppression, giving rise to feelings of discontentment and unhappiness. Many families had a single point of income, such as a family workshop, joint farm work, or revenue 1

Vid. 6.33–4 (P.L. 16.257–8); Ambrose, “Concerning widows”, trans. H. de Romestin, E. de Romestin and H. T. F. Duckworth, Nicene and Post-Nicene Fathers, 2nd ser., vol. x, Buffalo, NY, 1896.

141

142

Patriarchal household and incoming daughter-in-law

from jointly owned property, and the male head was usually the one who controlled the family fortune. For every purchase or investment he had to give his consent. Dishonesty between family members and unequal and inappropriate distribution of funds and rights to individuals could create severe tensions. While socio-anthropological gender studies have mostly focused on women’s power, status, rights, and influence relative to men, more recent studies have added the dimension of age, focusing on exploring women’s status relative to other older and younger women.2 The daughter-in-law/ mother-in-law relationship is thus accorded major importance by anthropologists studying traditional patriarchal societies. Their roles in ancient family systems, however, have not been addressed so far. This omission is all the more surprising as the daughter-in-law assumed a central role intersecting gender and generation divisions in kinship systems of patriarchy, a system that we also find in Roman Egypt. Comparative evidence, in fact, suggests, however, that the most common source of conflict was the relationship between mothers-in-law and incoming daughters-in-law. Depictions in Greek and Roman myth and poetry tell us about common prejudices and conflicts between daughters- and mothers-in-law.3 Their relationship was often dominated by suspicion, jealousy, and authority battles. Plutarch, for instance, knew that tensions between mothers- and daughters-in-law easily arose. He tells us about an old custom that existed in the city of Leptis in Northern Africa: The day after her wedding the bride should ask her mother-in-law for a pot. The mother-in-law, however, was supposed to refuse it to her and deny that she had one. This was to prepare the young woman for the mischievousness and jealousy of her husband’s mother, and if something worse should happen in later years, not to become angry and bear a grudge. The young woman should understand that her husband’s mother was jealous because she loved her son, and after his marriage was insecure about her position. The best thing, according to Plutarch, was for the bride to convince her mother-in-law that she had no intention to detract from or lessen her son’s affection for her.4 Plutarch

2 3

4

See, e.g., Brown et al. 1994. Cf. Wolf 1972; Kandiyoti 1988; Inhorn 1996; Brown 2004. Cf. the legend of Cupid and Psyche, told by Apuleius, where Venus is depicted as the stereotypical mother-in-law envious of her daughter-in-law’s beauty and the affection her son has for her. Learning of Psyche’s disobedience, she has her pregnant daughter-in-law whipped and tortured by Worry and Melancholy, and punishes her further by making her complete four difficult tasks (Apul. Met. 4.28–6.24). Cf. also Demand 1994. Plut. Mor. 138a–146a.

The “mother-in-law belt”

143

compared the relationship between a mother-in-law and her daughter-inlaw to the relationship between daughter and stepmother, and he suggests a strategy of deference and subordination on the part of the younger woman endeavoring to build a relationship of filial affection and trust to cope with the older woman’s jealous rivalry. More than her own parents, a daughterin-law should honor her husband’s parents, in whose household she lived now, and if she had problems or was hurt, she should not run back to her parents but rather conceal it from them. Her parents-in-law should be the first to be asked for advice; by this proof of trust and affection they would be won over. “To be seen to trust makes you trusted, to be seen to love makes you loved,” Plutarch advises young brides.5 As anthropological studies on other patriarchal societies show, under a virilocal marriage pattern the young bride faced a dire fate after marriage: she entered into a life of servitude in her husband’s household, and was supposed to be obedient and submissive in the face of constant mistreatment.6 The traditional gender division of labor in a multiple family household usually kept the young couple apart for most of the day, the men working outside in the fields while the women stayed at home minding the young children, cleaning the house, and preparing the food, under the close supervision and scrutiny of the wife of the head of the household. Since the young couple spent hardly any time together during the day, the most important connection a young bride formed in her new home was thus not to her spouse but to her sisters- and mother-in-law.7 And in accordance with the unwritten right of the oldest woman in the family, the mother-in-law aspired to exercise close supervision over the life of her young daughter-inlaw.8 A mother-in-law even interfered in the intimate life of the couple and the decision about how many children her daughter-in-law had to bear. She also strictly regulated the visits of her daughter-in-law to her own mother and other relatives.9 Moreover, the distinctive feature of the patriarchal family was that the unity between parents and sons usually held primacy over the conjugal bond, at least as long as the elder generation was alive. The son was expected to ally with his mother, stand on her side in case of conflict, and make his wife submit to his mother, if necessary by force. It is 5

6 8 9

Plut. Mor. 143a–c. Interestingly, Plato took a quite modern approach to solve the problem of tensions and quarrels resulting from co-residence of a young couple with their in-laws. He recommended that newlyweds “must leave their own houses to their parents and the bride’s relations, and act themselves as if they had gone off to a colony, visiting and being visited in their home, begetting and rearing children” (Laws 776a–b; trans. R. G. Bury, Cambridge, MA, 1967; cf. Demand 1994: 17). Brown 2004: 168; Wang 2004: 20–1. 7 Mandelbaum 1970; Goody 1983; Wolf 1995. See, for instance, Foner 1989: 231–2; Guidicini and Alvisi 1994; Xiaonan 1999: 87. Kagitcibasi 1982: 12.

144

Patriarchal household and incoming daughter-in-law

therefore no wonder that a daughter-in-law’s sense of emotional isolation in such a household constellation has been described by anthropologists as most painful.10 There is no doubt that the status of daughter-in-law was higher than that of handmaids and paid servants, but it did not change the fact that daughters-in-law were often subjected to abuse, both psychological and physical. The anthropologist Judith Brown has labeled the regions in which we find patriarchy and the virilocal household, such as western Africa, the Mediterranean, south-west Asia, and China, the “mother-inlaw belt.”11 She emphasizes the commonality of these regions in the power of the senior woman in the household and the exploitation of the incoming daughter-in-law.

6.2 The daughter-in-law in Roman Egypt Setting up house for a young couple in Roman Egypt usually followed the virilocal residence pattern; thus it entailed the introduction of a young woman to the family and the creation of a new nuclear unit within the household. All this required changes in roles, as well as redefinitions of relationships. On her wedding day the young bride had to leave her natal home and move in with her in-laws, i.e., her husband’s parents, his brothers, unmarried sisters, and the married brothers’ wives and children. Sometimes uncles and aunts of her husband, along with their children, were also living with them.12 The young woman thus entered a complex household where all male members were closely linked to one another by bonds of blood relation. Her peculiar situation as an outsider was further aggravated by the fact that the newlywed couple was not economically autonomous but imbedded in the economic activities of the household with limited access to independent financial resources. Neolocality upon marriage had the advantage that it favored much closer conjugal bonds. The only authority the young wife had to face was her husband, and she profited from the independent economic basis of her husband. When discussing the daughter-in-law’s relationship to her in-laws in Roman Egypt, however, we should not forget to take into account the demographic probabilities of such a constellation. Judging from the life tables, it cannot have occurred very often that a daughter-in-law lived 10 11

12

Wolf 1972: 124. Brown 1997: 79–99. For cross-cultural comparisons of intergenerational relations between the women of the household, see Brown et al. 1994; Dickerson-Putman and Brown 1998. See sections 2.8 and 2.11.

The daughter-in-law in Roman Egypt

145

together with a father-in-law for an extended period of time. According to model life tables that reflect ancient demographic conditions, the probability that a man had a living father at age 25 – on average around the time when he got married13 – was about 39 percent and decreased over the next fifteen years to virtually zero.14 That means that just around one-quarter of all women had a father-in-law in the early years of marriage. In those cases in which a woman married a man who had been married before and who was therefore considerably older, the likelihood that her husband’s father was still alive tended to be negligible. It was demographically much more likely that a young bride faced a mother-in-law in her new home. Since women regularly married several years earlier than men, they were likely to outlive their husbands.15 Yet, here also only half of all men had according to model life tables a living mother when they got married for the first time, and after ten years of marriage the mother of only one-fourth of all men was still alive.16 The percentage of married women living with a parent-in-law recorded in the Roman Egyptian census returns is indeed not high. Some 82.5 percent (n=165) of all married or previously married women (n=200) did not co-reside with a mother- or father-in-law. The majority lived with their husband and their children, with their brothers- and sisters-in-law, or – if they were divorced or widowed – with their children or their natal family. Only about 17.5 percent (n=35/200) of all women did live with a parent-inlaw. In 45.7 percent (n=16/35) of these cases a woman’s in-laws were both alive.17 As we predicted from the model life tables, it was much more common to co-reside with a mother-in-law than a father-in-law: 14.3 percent (n=5/35) lived with their widowed father-in-law in the same household.18 In about 40 percent (n=14/35) of cases a woman lived with her widowed mother-in-law.19 Married women who lived with a father-in-law or both of her husband’s parents were, as we would expect, on average 13

14 15

16 17

18 19

For a median age at 25 years or slightly higher for men at first marriage in Roman Egypt, see Bagnall and Frier 2006: 116. Saller 1994: 52, table 3.1.e, male, “ordinary,” Level 3 West: proportion having living kin. Bagnall and Frier 2006: 118–21. For the considerable age differences between spouses in late medieval Florence, see Herlihy and Klapisch-Zuber 1985: 205, fig. 7.1; for fifteenth-century Ragusa on Sicily, see Rheubottom 2000: 86–9. See note 14 above. 117-Ap-5 (daughter-in-law aged 20); 117-Ap-6 (18); 117-Ar-11 (30 and 18); 131-Ox-1 (25); 145-Ar-9 (13); 145-He-2 (16 and 18); 159-Ar-11 (40); 173-Pr-5 (16 and 13); 187-Ar-4 (??, 29, 17); 187-Ar-8 (??); 187-Ar-22 (18). 159-Ar-1 (22); 173-Me-3 (4[.]; 31; 19); 215-He-3 (31). 117-Ar-1 (28); 117-Ar-6 (53); 117-Ar-7 (three sisters-in-law of unknown age); 117-Ar-12 (36); 131-Ar-12 (60); 131-He-2 (20); 145-Ar-3 (18); 145-Ar-20 (2[.]); 159-Ar-10 (2[.]); 173-Ar-9 (38); 173-Pr-15 (16); 201Ar-1 (35).

146

Patriarchal household and incoming daughter-in-law

younger than those who had just a widowed mother-in-law. In almost half of the cases where a daughter-in-law lived with her widowed mother-in-law, the old lady, not her son, was functioning as the official head of the household.20 And not all women who lived with a mother-in-law were young. We find a 53-year-old woman living with her 75-year-old mother-inlaw,21 and a 60-year-old woman whose mother-in-law was 78 years old.22 From the private letters preserved on papyrus we can catch a glimpse of these women’s daily lives, their hopes and struggles in their new homes. From second- or third-century Alexandria a certain Serenilla wrote home to her father, in the Arsinoite nome: “Serenilla to Sokrates, her father, many greetings . . . I want you to know that I am lonely. Keep in mind, ‘My daughter is in Alexandria,’ so that I may know that I have a father, so that they may not see me as someone without parents.”23 Serenilla felt lonely and missed her family. We might wonder who “they” were that should be reminded that she was not without any backing from home. It probably does not go too far to assume that they were in fact her in-laws with whom she was quarreling. To give an example for the other side of the coin, we have a letter by an elderly woman (her name is lost) from the second-century Arsinoite nome sent to her son Kopres that illustrates vividly how a woman feared that her daughter-in-law might drive a wedge between her and her son. In a disagreement with her son, the old lady tried to shift all the blame onto her son’s young wife. This should not surprise us since a woman’s most vital old-age security was her sons, and she had every interest in keeping their allegiance and lifelong loyalty alive at the expense of the young couple’s conjugal bond.24 The mother warned her son that his wife was trying to influence him and was striving to alienate him from her: To Kopres, greetings. I know your quick temper but your wife inflames you when she says every hour that I do not give you anything . . . I am keeping nothing back from you because I trust you in everything. Your wife says in fact “She does not trust you . . .” Nobody can love you because she shapes you to her advantage.25

Another elderly woman from late third-century Hermopolis expressively stresses in her will that she does not wish her daughter-in-law, the widow of 20

21 24 25

117-Ar-12 (the mother-in-law is 64, the daughter is 36); 131-Ar-12 (the mother-in-law is 78, the daughterin-law is 60); 145-Ar-3 (the mother-in-law is 45, the daughter-in-law is 18); 173-Pr-15 (the mother-in-law is 42, the daughter-in-law 16); 201-Ar-1 (the mother-in-law is 59, the daughter-in-law is 35). 117-Ar-6. 22 131-Ar-12. 23 W.Chr. 100; cf. Bagnall and Cribiore 2006: 288. Cf. section 5.7. SB 3.6264 from the second-century Arsinoite nome. See Bagnall and Cribiore 2006: 282.

The daughter-in-law in Roman Egypt

147

her late son, to have any share in her inheritance, “for I, too, though independent, have suffered at her hands.”26 A daughter-in-law suffered especially then if her husband was away for work or military service and left her behind in his brothers’ or parents’ household to fight on her own for resources for herself and her children without his backing and support.27 High rates of labor migration recorded for Roman Egypt point to the fact that sons were away for long periods of time at or traveling to or from their migration site, while their wives remained behind in their husbands’ household alone with their in-laws. For societies in East and South Asia where virilocal residence patterns were prevalent far into the last century, high rates of suicide have been reported for young daughters-in-law; not a few saw no other escape from the torments of their in-laws than to hang or drown themselves.28 Some young wives in Roman Egypt also seem to have been so unhappy in their new homes that they were considering extreme measures. We have a letter from a certain Didymarion, an elderly widow from the second-century Arsinoite nome, who wrote to Paniskos, the brother of her son-in-law, with whom she apparently enjoyed a cordial relationship.29 Didymarion confided in the young man her worries about her daughter, who was married to Paniskos’ younger brother. Didymarion feared that Paniskos’ mother made her daughter’s life such a misery that the girl considered suicide by drowning herself in the sea. Didymarion pleaded that Paniskos might act as the father of the family and settle this dispute. I want you to know that my daughter is very much bothered by your mother. For she wrote to me, saying: “If she spends another month with me like this, I will throw myself into the sea.” I am writing to you now since you are like a father to them. Look, what is the truth? So far she did not blame your brother, but I say this so that he may not treat her badly.30

In Chapters 3 and 4 we saw that a young woman would take shelter in her parents’ home in a marital crisis or in case of divorce or the death of her husband. A married woman sometimes also returned home to her natal family if she was sick or about to give birth, as comparative studies on other traditional patriarchal societies show. It was in her parents’ home that she found some rest from the strenuous household chores in her husband’s house.31 Also for Roman Egypt we occasionally find references in the papyri that a pregnant 26 28 31

P.Lips. 1.29 from 295 ce. Cf. Rowlandson 1998: 196–7. 27 Shahd 2003: 27. Skinner 1997: 75. 29 P.Petaus 29. 30 Trans. Bagnall and Cribiore 2006: 276. Janelli 1982: 38; Broude 1995: 111; Bloch 1998: 92.

148

Patriarchal household and incoming daughter-in-law

woman left her marital home shortly before giving birth.32 Where would a pregnant woman go if not home to her natal family where she would find some respite in her final stage of pregnancy, and the love and support of her mother and sisters during this critical period? We have a letter from a soldier named Maximus who wrote to his mother-in-law Zenobia in Oxyrhynchus about his wife Salamaei, who stayed with her natal family. The letter dates to the fourth or fifth century ce. It was apparently not the first time that Maximus asked his mother-inlaw to send her daughter to him, but so far Zenobia had found excuses not to do so, perhaps because she did not want to lose her daughter’s help, or the daughter refused to go back to her in-laws: To my lady, my mother Zenobia! Maximus greets you in the name of the Lord . . . Send to me my wife Salamaei. My mother Rufina sends you her regards; greet all my brothers! Once and twice and thrice have I told you to send my wife and you didn’t want to; but now don’t lose a day or night to send me my wife!33

While the birth of a son in itself raised a young woman’s status in her inlaws’ household, this reached a peak when the son grew up and brought in a bride, and so the cycle repeated itself. Deprivation in early life was thus rewarded eventually with authority later on, and gave younger women something to aspire to with advancing age.34 Thus only through giving sons to the household could a young woman eventually assert her autonomy and increase her rank and influence. The failure to bear children was thus disastrous for a woman’s standing in her husband’s home and for her security and well-being in her later years.35 The case of the 60-year-old widow Taphiomis from the second century Arsinoite nome is interesting in this regard.36 We do not find another widowed woman in the census returns who after her husband’s death continued to live in his paternal home with her in-laws.37 A closer look at the composition of Taphiomis’ household gives us a potential clue as to why she had

32

33 35

36

A. Bülow-Jacobsen, H. Cuvigny, and J.-L. Fournet, “The identification of Myos Hormos: new papyrological evidence,” Bulletin de l’Institut français d’archéologie orientale 94 (1994): 33–4, no. 3 (O.Max.inv. 267 from the second century ce); cf. Bagnall and Cribiore 2006: 166. O.Florida 14 from the second century ce; cf. Bagnall and Cribiore 2006: 167. P.Oxy. 34.2731. Cf. Bagnall 1993: 191 n. 62. 34 Cf. Kandiyoti 1988. This is still the case in traditional families in modern Egypt (Inhorn 1996: 18). Studies on the Indian family have shown that the number of sons a woman had borne in the household has a decisive impact on her position in the household’s hierarchy (Laungani 2007: 143). ???-Ar-3 (Soknopaiou Nesos). 37 Cf. section 4.4.

The daughter-in-law in Roman Egypt

149

been able to stay. Six brothers had at one point lived and raised their children together under one roof, but only two of the brothers were still alive when the census return was drawn up. Of the other four brothers’ families, just their children and grandchildren were still alive, all of whom were female. Only Taphiomis had a 17-year-old son, who was thus the only male heir of the household on whom the future of the family rested. This was no doubt the reason why his mother had been able or willing to stay in her late husband’s household. Evidence that reflects extreme tensions between a man’s mother and his young bride is rare. The silence of the sources alone, however, is not very significant, as interfamilial tensions between the women of the household would rarely have found their expression in written correspondence. Moreover, it may be worth remembering here that many women’s letters would have been dictated, mostly to a male household member; letters of this kind would therefore not have been private communications, but their contents would have been known in the household where the woman lived.38 However, we also have widespread evidence that mothers- and daughters-in-law in Roman Egypt lived together relatively harmoniously. A letter by a certain Eudaimonis to her “daughter” Aline from early secondcentury Hermopolite nome is usually regarded as a letter by a mother-in-law to her daughter-in-law.39 Aline seems to have been commuting between her in-laws’ and her husband’s home. From the letters of the archive to which this letter belongs, it becomes clear that Aline was married to a certain Apollonios, a son of Eudaimonis,40 and that Eudaimonis and Aline had been living and working together in one household, until Aline, heavily pregnant, had sailed away with her servant to give birth in Heptakomia, where her husband was stationed. She had left her little daughter Heraidous behind with the girl’s grandmother Eudaimonis.41 Eudaimonis wrote to Aline that she was praying that she might give birth to a baby boy. She further reported about the difficulties of finding slave-girls to help her with weaving, probably as a temporary replacement for Aline as long as she was away.42

38 39 40 41

42

Cf. in general Bagnall and Cribiore 2006. P.Brem. 63 from 117 ce; cf. Bagnall and Cribiore 2006: 143–4. Bagnall and Cribiore 2006: 139–63. The little Heraidous also stayed with her grandmother in other instances when her parents were away (P.Giss. 21 from 115–17 ce; P.Giss. 24 from 116 ce). This apparently led to an estrangement of Aline and her daughter; we have a letter in which Aline complains that Heraidous forgot to greet her in a letter to her father (P.Giss. 78). Cf. P.Giss. 20 from 117/18 ce in which Aline writes to her husband about her wool working.

150

Patriarchal household and incoming daughter-in-law

6.3 Sisters-in-law Of those women who lived with parents-in-law, 57.1 percent also had siblings-in-law living with them (n=20/35). It was even more common for a woman to live with her husband’s siblings than with his parents – 21 percent (n=42/200) lived with brothers- or sisters-in-law without any parent generation, compared with the 17.5 percent (n=35/200) who lived with parents-in-law. We do not hear anything about disputes between sistersin-law in our papyrological evidence, which should not lead us to conclude, of course, that these conflicts did not exist. As already stressed above, domestic discord rarely manifested itself in our sources.43 Comparative studies show, moreover, that the more sons of the same family married and brought in their wives, the more tensions between mother-in-law and daughter-in-law were replaced by disputes between sisters-in-law competing with each other over family resources. In fact, relations between sistersin-law often turned out to be the most aggressive ones in the multiple family household. Wives of brothers were natural competitors and each tried to advance the interests of her own family unit. Constant rivalry for the loyalties of husbands and sons seems to have been the rule in multiple family households, and wives fostered competition among brothers. They bore grudges against each other when they felt that household resources were not shared equally, or if one of them enjoyed the special favor of the mother-in-law. Antagonism between brothers’ wives has also often been blamed for the break-up of a multiple family household.44 Cross-cultural studies suggest that the relative rank of a woman among her sisters-in-law depended on her husband’s relative rank. The same seems to have been true also for Roman Egypt, if we assume that the rank that was accorded to her in the census returns reflected her actual position in the household hierarchy.45 As we saw in Chapter 5, gender and age were the major determinants of rank in the census lists, with the head of household ranking first followed by his wife, sons, and daughters. Wives of elder sons were listed above wives of younger sons, and the wife of the head of household was listed above all other female household members. Studies on other patriarchal societies with similar household patterns have found 43

44

Conflicts and jealousy between sisters-in-law figure prominently, however, in mythology, as for instance in the myth of Aedon and Niobe (Hom. Od. 19.508–53). They were married to twin brothers, Zethos and Amphion, sons of Zeus. Aedon had only one son, while her sister-in-law Niobe had many. Jealous of her sister-in-law’s many children, Aedon tried to kill Niobe’s eldest son, Amaleus, in his sleep, but by mistake killed her own son, Itylus, who shared the bed with his cousin. Caplan 1970: 35; Wolf 1972: 54. 45 Cf. section 5.8.

Tensions between husband and wife

151

that the rank a woman held in a frérèche was directly related to her risk of death.46 This should not surprise us, since the more sisters-in-law a woman had, and the lower her rank among them, the more work she was probably assigned and, above all, the less care she could expect in sickness and childbed, which increased the hazards of mortality.

6.4 Tensions between husband and wife Common conflicts between spouses, at least those that received official recording, were over sufficient maintenance of the wife, meaning the provision of food and clothing.47 This obligation was laid down in many marriage contracts and its neglect was the most common accusation in a woman’s petition against her husband.48 Receiving sufficient maintenance seems to have become a problem for the wife especially when for work reasons her husband did not live with his family for long stretches of time, perhaps turning to occasional wage labor to supplement their farm earnings during slack agricultural periods because the family land was not large enough to provide for the whole family, or because he served in the army and was on military duty.49 Women and children spent their time in the home village either on their own or with the husband’s extended family, while husbands and fathers were semi-permanent residents of a major city or in an army garrison. Since usually everyone was expected to return to their family for the census registration,50 such short-term migration is, however, difficult to detect on the census returns.51 It is not hard to imagine that some men who spent most of their time away from home neglected to send home the proper amount of money; many might even have kept a concubine at their place of work. Hence Apollonia’s demands of her husband in her marriage contract: “Philiskos, whether he is at home or away from home, shall furnish Apollonia with everything necessary and clothing and whatsoever is proper for a wedded wife, in proportion to their means.”52 Wives and the rest of the family were often left worrying when husbands failed to send letters home about their well-being.53 Some men seem to have been away from their wives for several months or even longer, 46 47 48 50 51 52 53

Cf. Hammel and Gullickson 2004: 10. e.g., BM 10593 from 170 bce. Cf. Yiftach-Firanko 2003: 188. For references, see Yiftach-Firanko 2003: 188 n. 17. 49 Cf. Bagnall and Frier 2006: 165–9. P.Lond. 3.904 from 104 ce. For some indirect evidence of migration in these documents, see Bagnall and Frier, 2006: 167–8. P.Tebt. 1.104 from 92 bce; trans. Winter 1933: 120. e.g., P.Bad. 4.48 from 127 bce; cf. Bagnall and Cribiore 2006: 107.

152

Patriarchal household and incoming daughter-in-law

asking about the sex of their newborn child and instructing their wife by letter whether to raise the newborn or let it die. Hilarion, a soldier stationed in Alexandria, wrote to his wife, Alis, in Oxyrhynchus: “If among the many things that are possible, you do bear a child, if it is male, let it be, but if it is female, cast it out.”54 Hilarion apparently had already been away from home for some time and did not even know whether or not his wife was pregnant. We hear from other wives complaining in letters to their husbands about the difficulties they were facing alone at home, managing the household and raising the children. Alis had expressed her fear of being forgotten, but her husband in Alexandria assured her: “You have told Aphrodisias, ‘Do not forget me’; but how can I forget you?” Hilarion also sent money home regularly: his remark about exposing a potential female baby should thus not be interpreted as a lack of love and care for his wife. A certain Isias in the second century bce wrote to her husband, Hephaistion, who had already been away for a considerable time.55 Isias had heard from a third person that her husband was in katoche (a form of religious detention) in the Serapeum in Memphis, and also her letter was found there. She tells him that she is pleased that he is alive and that she thanked the gods, but is clearly annoyed that her husband did not show any intention of returning home: For the news that you are well I straightaway thanked the gods; but about you not coming home, when all the others who had been detained there have come, I am ill pleased, because after having piloted myself and your child through such bad times and been driven to every extremity owing to the price of wheat, I thought that now at least, once you got home, I would enjoy some rest. But you have not even thought about coming home, nor given any regard to our situation, how I was in want of everything even while you were still here, not to mention this long lapse of time and such crises, during which you have sent us nothing. Moreover, since Horos, who delivered the letter, reported that you have been released from detention, I am thoroughly ill pleased. Nonetheless, since your mother also is distressed, please both for her sake and for ours return to the city, if nothing more pressing holds you back.56

Isias clearly had the backing of her husband’s family, with whom she was living and who were as upset as she was that Hephaistion made no effort to return home and take care of his wife and child and the family business. Hephaistion’s brother dispatched a second letter on the same day urging him

54 55

P.Oxy. 4.744 from 1 bce; cf. trans. Winter 1933: 56; Rowlandson 1998: 230. UPZ 1.59 from 168 bce. 56 Trans. Bagnall and Cribiore 2006: 111.

Tensions between husband and wife

153

to return home as soon as possible.57 A certain Maria living with her daughterin-law in fourth-century Oxyrhynchus complains in a letter to her son, who was away from home because of his work as a business agent for wealthy Oxyrynchites, that he had not sent money, neither to her nor to his wife.58 Another spillover effect of conflict between spouses was certainly domestic violence. We have charges of abuse and beating brought forward against a husband by his wife, but also vice versa.59 In a petition for recovery of her dowry, Syra from early first-century Oxyrhynchus reports the misconduct of Sarapion, her husband: “But Sarapion, after he squandered my dowry to suit himself, continually illtreated, insulted, and attacked me, and deprived me of the necessaries of life: finally he deserted me, leaving me in destitution.”60 Many marriage contracts contained a formula that forbade a husband to insult or injure his wife.61 In their marriage contract the above-mentioned Philiskos promised Apollonia: “It shall not be lawful for Philiskos . . . to eject or insult or illtreat her nor to alienate any of their property with injustice to Apollonia.” While abused or neglected wives often sought sanctuary in their paternal homes, we also find wives preferring loyalty to their husbands who were tangled up in disputes with their fathers-in-law. When the father of a certain Dionysia tried to assert the right to dissolve his daughter’s happy marriage because he was quarreling with his son-in-law about the daughter’s property and dowry, the daughter submitted a petition to the prefect of Egypt asking to be allowed to remain with her husband.62 There existed apparently a law attributed to the “law of the Egyptians” allowing a father to take his daughter (and her dowry) away from her husband, even against the daughter’s will. While according to the strictly legal definition of patria potestas a Roman father also had the right to dissolve the marriage of his daughter who had married sine manu, Roman officials in Egypt regularly rejected this law as too cruel, usually recommending the parties who appealed to them to comply with the daughters’ wishes. Roman law was only mitigated in the second century ce, forbidding a paterfamilias to break up his daughter’s harmonious marriage, apparently in response to individual petitions.63 In our case, neither Dionysia nor her father was, however, a Roman citizen. 57 58 59

60 62 63

UPZ 1.60 from 168 bce. P.Oxy. 48.3403; cf. Rowlandson 1998: 243–4, no. 180 (a); Bagnall and Cribiore 2006: 212. P.Oxy. 2.281 from 20–50 ce; P.Oxy. 6.903 from the fourth or fifth century ce; P.Oxy. 1.3581 from the fourth or fifth century ce. For an aggressive ex-wife, cf. SB 10.10239 from 37 ce. P.Oxy. 281 from 20–50 ce; trans. Winter 1933: 126. 61 e.g., P.Tebt. 1.104 from 92 bce. P.Oxy. 2.237 after 186 ce; cf. Rowlandson 1998: 183–9, no. 138; Yiftach-Firanko 2009: 550–1. For legal references, see Evans Grubbs 2002: 196.

154

Patriarchal household and incoming daughter-in-law

6.5 The daughter-in-law and care in old age While the daughter-in-law initially was situated at the bottom of the gender and generational hierarchy of the household,64 over time, however, the power balance shifted in favor of the younger woman. First, by having children a daughter-in-law asserted her status in her husband’s household, and second, when parents-in-law became older and frail, they also became dependent on the daughter-in-law’s assistance. Caring activities normally fell into the domain of daily housework, the traditional field of responsibility for women.65 Sons in Roman Egypt acted officially as the main caregiver for their elderly parents; the actual practical care, however, was most likely not performed by them, but instead by their wives owing to the typical labor division in the household.66 However, while a son’s care for his elderly parents was motivated by affection, piety, and feelings of reciprocity, in an in-law relationship these feelings were usually absent.67 A study on twentieth-century Chinese families living in joint families has shown that daughters-in-law provided their frail parents-in-law with all the necessities, but did not feel obliged to provide emotional and psychological support, such as listening to them patiently or providing them with companionship.68 A daughter-in-law assumed her responsibilities only because the culturally assigned role was part of her domestic duties, not because she felt any direct emotional affinity or filial obligation towards her parents-in-law. As long as the old patriarch, the husband’s father, lived, the authority rested clearly with his wife, the mother-in-law. But this power constellation changed when the patriarch died and his son assumed the leading role in the family. The authority of a mother-in-law, economically dependent on her son, was considerably weakened – especially when she became older and frailer and had to rely more and more on the care and support of her daughter-in-law.69 Both sides, the daughter-in-law and the mother-in-law, thus had to learn to get along with each other, and as Ambrose 64 66

67 68

69

Cf. section 5.8 for the correlating rank accorded to her in the census lists. 65 Wong 2000: 311. In Asian societies, such as Japan, Korea, and China, where the multiple family household was common until the middle of the twentieth century and daughters usually left the household upon marriage, it was the daughter-in-law who was expected to play the main role of nurturer for the aged (Liu and Kendig 2000; Wong 2000). Wong 2000: 306; cf. Liu and Kendig 2000: 17. Cf. Chapter 3. Wong 2000: 306–7. Such an attitude between caregiver and care receiver, as studies have shown, has major implications for the contentment and emotional stability of the elderly parents and the daughter-in-law alike. For instance, studies on modern East Asian societies have shown that “daughters in law exhibited inordinate amount of subjective burden even though the objective burden remained relatively low” (Liu and Kendig 2000: 14). Cf. also Chapters 2 and 5.

Consanguineous marriages

155

recommends, from the beginning a wise mother-in-law treated her daughter-in-law well and taught her to love her in order to be able to count on her support in later years.70

6.6 Consanguineous marriages If a daughter-in law was the main source of practical care, the choice of a suitable bride for one’s son must have been of major importance for parents in order to provide for a time when they would be entirely dependent on her goodwill. As long as the elderly couple was healthy and active and commanded authority, even a recalcitrant daughter-in-law would have had to pay deference to her parents-in-law. But once they became frail and dependent, they would have had to rely on the care and support of this woman who may not have been bound to them by any feelings of reciprocity or filial affection, and perhaps even harbored vengeful feelings against them for being treated like a maid in earlier years. Societies have found varying means to overcome these difficulties: In China and Taiwan far into the twentieth century it was a widespread custom for couples to adopt little girls, unwanted by their natal families because of poverty, and to raise them side by side with their biological sons in order to marry them eventually to one another. This practice was employed to circumvent the tensions usually resulting from the introduction of an adult stranger, i.e., a daughter-in-law, into existing domestic relationships and power balances. These Chinese marriages, studied extensively by Arthur Wolf, are known as minor marriages (or sim pua marriages).71 In these unions in which the daughter-in-law knew her parents-in-law since childhood, the elderly could expect a higher level of emotional affinity from her than from a daughter-in-law who had come as an outsider to the family at a much later point. Around the middle of the last century this form of marriage accounted for as much as 70 percent of all marriages in some Chinese provinces.72 An alternative to adopting a daughter-in-law at a very early age, and from a cross-cultural perspective a far more common strategy for securing the loyalty of a daughter-in-law, was to look for a suitable bride for one’s son among one’s own relatives. The advantages of close-kin marriages are 70

71

Cf. Ambrose (Vid. 6.33–4. P.L. 16.257–8, cited in the epigraph to this chapter), who referred to the example of Ruth (1:20–1), who left her own people to remain with her mother-in-law, Naomi, in her old age. Wolf 1995. 72 Wolf 1968: 864–5.

156

Patriarchal household and incoming daughter-in-law

usually seen in higher loyalty and compatibility of the daughter-in-law to the husband’s family. A further bonus is the retention of family wealth and property within the family, and the strengthening of kinship solidarity.73 Finally, consanguineous marriages are often thought to soften the consequences of patrilocal residence patterns for the young bride who had to leave her natal home and move into the household of her groom upon marriage. As we have seen above, studies on patriarchal societies with virilocal residence patterns paint a grim picture of the difficult situation an isolated young daughter-in-law faced in her new surroundings: conflict of authority with her mother-in-law, a heavy workload assigned to her, and a lack of recourse to any means of independent support. In consanguineous marriages, however, her mother-in-law is instead of a stranger an aunt or an otherwise closely related woman whom the girl has known since childhood. Marriage to a close relative allows the bride sustained contact with her parents and other social networks developed before marriage, and accords her a higher value within her husband’s family and a greater autonomy in household decision-making.74 In the event that her husband dies prematurely, the risk of getting thrown out or cheated out of her dowry is considerably lower. It should thus come as no surprise that the majority of the world population favors consanguineous marriages, with numbers even on the rise because lower rates of child mortality mean that more marriage partners within the wider family circle are now available. In the year 2000 consanguineous unions were virtually non-existent among about 1 billion people of the global population. Among almost 3 billion people the proportion of consanguineous marriages was 1–10 percent, and among 991 million people there was a prevalence of 20–50 percent for close-kin marriages.75 In the Pakistani Muslim population of modern Britain, for instance, cousin marriage of first or second degree is as common as 69 percent of all marriages, and 46 percent of marriages in the Arab community and 44 percent in the Indian Muslim community are consanguineous.76 Among Bedouin communities in Kuwait and Saudi Arabia, consanguineous unions reach up to 90 percent of all marital unions.77 Far into the twentieth century also the societies around the Mediterranean exhibited a strong tendency toward consanguineous marriages.78 This 73 75

76 77

Barakat 1993. 74 Barakat 1993; Yount 2005: 410. For about 1,064 million people the level of consanguinity was unknown since it was not reported or the quality of data was insufficient (Bittles et al. 2001: 68–78). Schott and Henley 1996: 316; Guilmoto and Rajan 2005: 295. Al-Rashoud and Farid 1991; Panter-Brick 1992: 69–72. 78 Cf. Bonte 1994.

Consanguineous marriages

157

tendency can probably be traced back not only for centuries but even millennia. In biblical Judea, for instance, cousin marriage seems to have been a common practice.79 In classical Athens, first-cousin marriages were “clearly an important factor in the life of the upper classes.”80 Thompson suggests that for Athenians the strengthening of kinship ties and intimate knowledge of the groom’s background and character were reasons for preferring to marry one’s daughter to a cousin: “Whenever an Athenian father sought a worthy husband for his daughter he would often turn to a relative whose qualities he knew firsthand and whose loyalty he could expect.”81 Consolidation of property and consolidation of the kin group under the threat of political disgrace were other motives mentioned by the Attic orators.82 If a father did not look for a bride for his son among his kin this was even seen as a sign of a lack of loyalty and trust among family members.83 Consanguineous marriages also play a role in the Hellenistic novel. In Apuleius’ novel, The Golden Ass, for instance, the young woman Charite is about to marry her cousin Tlepolemos, when robbers interrupt the wedding ceremony, abduct her and demand ransom from her family. This cousin was most probably her father’s brother’s son, since Charite herself tells us that she and the groom, only three years older than her, were brought up in the same household.84 Plutarch noted in the second century ce that, in contrast to tradition in the Greek world, Romans did not marry their daughters to close relatives. Plutarch supposed that Roman parents chose a husband for their daughter from outside the extended family because their daughter’s own family could 79

80

81 83

84

An example is Jacob, who married his two cousins, Leah and Rachel, who were the daughters of his mother’s brother (Gen. 29:10–30; cf. Goody 1983). Thompson 1967: 273–82, esp. 279. Thompson found seven parallel cousin marriages and three crosscousin marriages in the literary sources for classical Athens. Cf. Cox 1998: 4–10, 12, 15, 19, 24–6, 219. Thompson 1967: 280; Cox 1998: 3, 10. 82 Cox 1998: 33–4. “A convincing proof of their enmity is the fact that though Eupolis had two daughters and was descended from the same ancestors and saw that Apollodoros was possessed of money, yet he gave neither of his daughters to him in marriage” (Isaeus 7.11–12; cf. Just 1989: 80–1). An example for a matrilateral parallel cousin marriage was the marriage between the speaker’s grandparents in one of the orations by Isaeus: “My grandfather Kiron married my grandmother, who was his cousin, born of the sister of his mother” (Isaeus 8.7; cf. Thompson 1967: 274). Also, Demosthenes’ sister married a matrilateral parallel cousin, the son of her mother’s sister (Plutarch, Lives of the Ten Orators, 847c); for the identification of Laches as a cousin of Demosthenes and his sister, see Thompson 1967: 275. “There was a young man, handsome to behold, the leading light among his friends and equals. The whole town chose him as First Citizen, and more than that: he was my cousin, and older than I in age by a mere three years. From our earliest years he was brought up in my company and so grew up, inseparable from me in our living arrangements, in our little house; no, in the same room, in the same bed, bound to me, as I was to him, in the mutual affection of a holy love. Through legitimate nuptial vows he had from long ago been pledged to the bonds of marriage; he had even been named my husband in the official wedding agreement, with the full consent of our parents” (Apul. 4.26.3; trans. J. C. Relihan 2007: 85). Cf. Treggiari 1991b: 109–10; Corbier 2005: 272.

158

Patriarchal household and incoming daughter-in-law

then serve as recourse in case of conflict with her husband’s family.85 Despite Plutarch’s testimony for Roman aversion to close-kin marriages, however, the practice of consanguineous unions seems not to have been limited to the eastern part of the Roman Empire.86 We find several firstcousin marriages recorded in various sources from late republican times onwards.87 We have, however, no means to ascertain how high the proportion of close-kin marriages was of all marital unions. The only thing we can firmly say is that Roman law forbade marriages between full or half-siblings, while in the Greek East we have some evidence for half-sibling marriages.88 Roman law conceded, however, unions between first cousins, a form of union that is usually regarded as endogamous by social historians.89 The marriage of uncle and niece, who were even more closely related to each other than first cousins (third vs fourth degree), was declared legitimate by a senatus consultum in 49 ce, and this concession was not abrogated until 342 ce.90 Only at the end of the fourth century did Theodosius I try to banish cousin marriages. His law was later reinforced by Arcadius and Honorius for the Eastern and Western halves of the empire,91 but it was annulled for the East again in 405.92 The practice of marrying within the family apparently went on even in the fifth century. The bishop Theodoret of Cyrrhus in Syria reprimanded the leading men of Zeugma (in modern south-east Turkey) for still allowing their daughters to marry cousins and uncles.93 Cousin marriages were still legal under Justinian I, despite several hundred years of Christian influence, with the Christian literate elite vehemently arguing against them.94 Unfortunately, we find it nearly impossible to detect cousin marriages in the Roman Egyptian census returns. To do so we would need not only the parents’ but also the grandparents’ names for both husband and wife. We can only suspect that at least some of those women who lived with an 85 86 87

88

89

90 91

92 94

Plut. Mor. 289d–e. See Pomeroy 1976: 215–27; Moreau 1978: 41–54; Krause 1994–5: iii.240. See, for instance, Saller and Shaw 1984. Cf. Thomas 1980: 345–52; Treggiari 1991b: 109–19; Corbier 1991a: 140–1; 1991c: 180–2; 2005: 276. The senator Publius Paquius Scaeva from Histonium (modern Vasto in the Abruzzo region) married his cousin Flavia, as they tell us in their grave inscription: “consobrinus idemque vir, consobrina eademque uxor” (CIL 9.2845–6 CIL 5244 ILS 915). Cf. Corbier 1991c: 184. For an overview, see Huebner 2007; 2012a. See Chapter 7 for a discussion of full brother–sister marriage in Roman Egypt. Inst. 1.10.4 (213 ce); Dig. 23.2.3 (Paulus); Dig. 28.7.23 (Marcellus); Dig. 28.7.24 (Papinian); CIust 6.25.2 (213 ce). Cf. Corbier 2005: 275–6 with nn. 60–4. CTh 3.12.3 (396 ce); CTh 3.10.1 (409 ce). Cf. Roda 1979; Puliatti 2001: 176–8; Moreau 2002: 192–8; Corbier 2005: 278. CIust 5.4.19 (405 ce). 93 Thdt. Ep. 8 (Sources chrétiennes 40, ed. Y. Azema, Paris 1955). Cf. Jo. Chrys. Hom. Ep. I Corinth. 34.4. See also O’Roark 1996: 403–4; Moreau 2002: 135.

Consanguineous marriages

159

apparently unrelated mother-in-law were also bound by some kinship ties to their husbands’ families. One of the few certain first-cousin marriages is recognizable to us only because the young couple were living together with their widowed mothers (who were twin sisters) in one household in early third-century Tebtunis.95 The 54-year-old sisters, both divorced or widowed, had married their children, a 33-year-old son and a 35-year-old daughter, to each other, i.e., thus making a matrilateral parallel cousin marriage. They all lived in a house that belonged to one of the two sisters. The young couple had just started a family, and their babies were again twins like their grandmothers. The sisters surely felt secure about their old-age support, since neither of them was a mother-in-law related merely by marriage, but both were at the same time mother and maternal aunt to the younger generation.96 The daughter of the public farmer and priest Hartothes from the early first-century Arsinoite nome had married her paternal cousin, i.e., she had made a patrilateral parallel cousin marriage.97 Another cousin marriage in the census returns from late second-century Antinoopolis was a cross-cousin marriage.98 In many societies cross-cousin marriage, i.e., the marriage of children of a brother and his sister, is usually seen as the ideal arrangement, but parallel cousin marriages, like the one above, also occur quite frequently.99 I have argued elsewhere that the “brother–sister” marriages recorded in the Roman Egyptian census returns, constituting up to one-quarter of all marriages in some regions,100 were probably also cousin marriages, unions arranged by parents between a biological daughter and an adopted son, often the biological son of a brother or sister.101 In an adoption contract from the Hermopolite nome dating to 381 ce we have an example for the formal adoption of an orphaned nephew: a 60-year-old grandmother, Aurelia Teeus, gave her orphaned 10-year-old grandson for adoption to her son, the younger brother of the boy’s late father. The uncle promise to raise the boy as if he were his own and to manage his considerable property, consisting of land and 95 96

97

98 99

100 101

201-Ar-10; cf. P.Tebt. 2.480. In a Coptic papyrus related to Pisentius, bishop of Coptos in the early seventh century, a man is mentioned who lived together with his niece, the daughter of his brother, and was obviously married to her (P.Pisentius 26; cf. Wilfong 2002: 43). See the Harthotes archive: www.trismegistos.org/arch/archives/pdf/99.pdf [accessed July 2012]. Cf. section 5.3. 187-An-2. We find seven examples for parallel cousin marriages in the literary sources for classical Athens (see n. 80 above). Hopkins 1980; Goody 1990; Shaw 1992; Scheidel 1996a; Bussi 2002. Huebner 2007. Cf. Remijsen and Clarysse 2008; Rowlandson and Takahashi 2009.

160

Patriarchal household and incoming daughter-in-law

buildings and movable goods the boy had inherited from his parents, until he came of age. He gave the assurance that the adoptee held all the rights of a freeborn and firstborn son, “as if he were engendered by you from your own blood,” as the grandmother said.102 By contrast to modern Western practices, where the details about a child’s origin usually are not disclosed to the adoptee or adoptive parents, nephews were preferred candidates for adoption in most societies. We should probably assume that the “brother–sister” marriages recorded in the Roman Egyptian census returns constituted only the tip of the iceberg of strong tendencies toward endogamous unions in the Roman Egyptian population, a form of marriage pattern that is widespread in these regions to this day.103 To come back to the daughter-in-law’s position in her husband’s household, however, the strong tendency to endogamous marriages meant that in almost one-third of all cases (n=10/35) where a woman lived with her husband’s parents we thus cannot really speak of an in-law relationship. In seven of these cases the parents-in-law were the natural parents because the woman had married her own “brother” (cousin),104 and in two cases the mother-in-law was an aunt.105 For one woman who had married her halfbrother, her mother-in-law was the new wife of her father, thus also her stepmother.106 Therefore just about 5 percent (n=10/200), if not less, of all married or previously married women co-resided with a mother-in-law who was not in fact her natural mother or at least very closely related to her. On the other side of the coin, 13 percent of all the adult women in the census returns were living with a daughter-in-law (n=26/200). In almost 30 percent of these cases (n=7) it was, however, her own daughter who had married an (adopted) brother; in one case the daughter-in-law was also a stepdaughter, and in two cases the daughter-in-law was a niece. That means that only 8 percent of all the women lived with an apparently unrelated daughter-inlaw (n=16/200). Interestingly, it has been noted that “brother–sister” marriages in Roman Egypt were extraordinarily fertile. In one case a married “incestuous” couple had eight children;107 in another, six or seven children.108 Bagnall and Frier attribute this trend to the fact that this form of marriage was usually entered into by spouses who were younger than the average.109 Regardless of the question whether these couples were natural siblings or not,110 an additional 102 104 106 109

P.Lips. 1.28; trans. Rowlandson 1998: 297. 103 Holy 1989. 117-Ar-1; 145-Ar-9; 145-Ar-20; 159-Ar-11; 173-Ar-9; 187-Ar-4; 187-Ar-8. 105 187-Ar-4; 201-Ar-10. 187-Ar-22. 107 187-Ar-4. 108 103-Ar-3. Bagnall and Frier 2006: 128. Cf. Scheidel 1997a: 368. 110 See Chapter 7 below.

Consanguineous marriages

161

explanation is certainly that women in these uxorilocal marriages did not have to endure conflict with their mothers-in-law. Constant domestic tensions are generally believed to lead to reduced fertility in young women and are commonly found in virilocal marriages.111 Studies on exogamy versus endogamy marriage patterns for twentieth-century north and south India, for instance, have shown the importance of kinship and household structures for fertility and infant mortality.112 Exogamous marriages in which an outsider woman moved into her husband’s household apparently resulted in higher maternal and child mortality, an outcome that is attributed to the vulnerability and low standing of the young woman in her new household and the consequent negative effects on her health.113 111 112

113

Pasternak 1985: 309–34. Dyson and Moore 1983: 35–60; Barakat 1993; Bittles 1994; Murthi et al. 1995: 745–82; Das Gupta 1997: 36–52; Hussain and Bittles 1999; Yount 2005: 410. Contemporary rural northern India has similar household patters as Egypt in the Hellenistic and Roman period, characterized by multiple family families, virilocal marriages, and partible inheritance. Das Gupta 1995: 481–91.

chapter 7

Childless old age The worst of all fates?

I raised you, children, but it was no use; No use, the way I toiled, how much it hurt, The pain of childbirth, piercing like a thorn. And I had so much hope when you were born: You’d tend to my old age, and when I died, You’d wrap me in my shroud with your own hands: Euripides, Medea 1029 361 An admirable fate for anyone.

7.1 Growing old without children In our analysis of support, transfers, harmony, and conflict between parents and children, we have so far taken the presence of the latter for granted. However, despite almost universal marriage and much higher birth rates in antiquity than today, a considerable number of individuals found themselves without children or with only daughters (which was almost equally unfortunate) when they were approaching old age. The likelihood of being childless was higher for elderly women than for elderly men because men remarried more often and fathered children also at an advanced age.2 According to model life tables, in antiquity on average about 25 percent of the elderly above the age of 60 did not have a living child. These numbers are more or less equivalent to today’s proportion of people in the modern Western world who find themselves without offspring when they reach the age of 60. However, these contemporary couples or individuals are childless for different reasons than in antiquity. Today, non- or later marriage for women, the desire to maintain a certain life standard, and participation of women in the upper-level workforce are responsible for voluntary 1 2

Trans. D. Arnson Svarlien, Indianapolis, 2008. Cf. the model life tables in Saller 1994: 49, 52. About 22 percent of all “ordinary” women at the age of 60 were childless; 44 percent did not have a son. About 16 percent of all men reaching 60 did not have a child, and 32 percent did not have a son.

162

Reasons for childlessness

163

childlessness and ever-decreasing birth rates in modern societies. In antiquity, sterility and high infant and child mortality left many people without children when they were approaching old age. While in the modern Western world parents rarely expect their children to support them financially, in Roman Egypt, in the absence of communitybased pensions and health care, children were considered crucial for ensuring their elderly parents’ economic welfare, as we saw in Chapters 3 and 5. Having no one to rely on for material security and emotional comfort in old age and, equally dreadful, no one to provide a proper burial and perform the customary funerary rites seemed like the worst of all fates to many.3 In this chapter I will have a closer look at the reasons for childlessness among the elderly and will analyze in detail its implications for their economic, emotional, and social status. Furthermore, we will also examine the availability and reliability of alternatives to the support that children provided. Closely related to the aspect of lacking children as caregivers is the question of the living situation of the childless elderly.

7.2 Reasons for childlessness Elderly men and women in Roman Egypt were childless for various reasons. Infertility is one of the first that comes to mind. Levels of sterility in societies before the advent of effective contraceptives were, however, in general rather low. According to demographers, only about 3 percent of all couples in a natural fertility regime were affected by sterility, defined as the “physiological incapacity to produce a live birth,” when the wife was 20 years old.4 With increasing age, sterility levels rose slowly over a woman’s thirties, and rapidly thereafter. Late marriage is thus often regarded as a more consequential check on fertility than actual sterility. Since human biology favors the fecundity of women between their late teens and early thirties, women who married later in life were more likely to remain childless. Men were less affected by this risk of sterility since they had, thanks to their lifelong fertility, higher chances to have children in a second and third marriage.5 However, since in Roman Egypt virtually all women were married when they reached their mid-twenties, late 3

4 5

Hom. Il. 5.152–8; Hes. Th. 602–7; Eur. Ion 618–20; Eur. Alc. 735; Eur. Tro. 1185; Eur. Andr. 1081; Eur. Med. 1030; 1394; Isaeus 4.30. Even in modern Egypt the primary reason for Egyptians wanting children is old-age security (Ammar 1966; Gadalla 1978; Morsy 1978: 599–616; Rugh 1984). Lockwood 1998: 51. For instance, the otherwise very happy marriage of Turia and her husband Quintus Lucretius in the second century bce remained childless because Lucretius spent many of the early years of their marriage away in exile. When, after many years of trying, Turia in middle age realized that she and her husband would never have children of their own, she was very depressed about her infertility and

164

Childless old age: the worst of all fates?

marriage cannot have been a major contributing factor for childlessness.6 Infertility due to malnutrition and heavy physical work (also seen in skeletal evidence) might have been more common.7 Malnutrition makes conception difficult or impossible because it leads to severe underweight and a low percentage of body fat, which can disrupt hormones and prevent ovulation. Moreover, malnourished women who are underweight who do get pregnant have a higher change of miscarriage or delivering a premature or low-weight baby that is less likely to survive the first years. Biological childlessness thus might not have been a quantitatively negligible factor. Childlessness due to lifelong celibacy, however, was not very common in antiquity. Almost all men and women married in pre-Christian times. Continuous celibacy was disapproved of among pagans and Jews alike. In later times, with the spread of Christianity and the ideal of monasticism, the numbers of those who never married surely rose, yet the percentage of the population who followed this ideal was probably still negligible.8 Reasons for non-marriage in Roman Egypt could include severe health problems, such as a mental handicap. Other reasons might have been extreme poverty. If a family belonged to the poorest segment of the population, such as landless laborers, it might not have been able to raise a dowry for a daughter or afford to take in and maintain a bride for a son. Marriage for the men and women recorded in the Roman census returns, however, seems to have been universal.9 Deliberate childlessness after marriage is unlikely to have been common, either. Raising children was the ultimate goal of marriage. According to model tables, the major contributing factors to a high percentage of childless elderly were very high infant mortality and a high risk of mortality over the entire life course.10 About 30 percent of all infants died in their first year, and just 50 percent survived up to the age of 5.

6 8

9 10

grieved for her husband because he was without an heir. She consequently offered her husband a divorce and even offered to arrange a new wife for him. Proving his love for her, Lucretius refused to let her go: “How could you even conceive in your mind of any reason why you should, while still alive, cease to be my wife . . . How could the desire or need of having children be so great that I would break faith with you?” (CIL 6, 1527 ILS 8393). 7 Bagnall and Frier 2006: 113. Cf. section 1.4. From the earliest period of Christianity we come across those who “renounced marriage for the sake of the kingdom” (Matt. 19:12), who “left everything” to follow Christ (cf. Matt. 19:27), men and women who lived a life of renunciation and celibacy, denying themselves marriage and a family. Nuns and monks were, however, no invention of Christianity. Almost every religion has its holy men and women who live a life of celibacy. “Nuns,” holy celibate women of the sun god Samas, lived in the cloisters in Sippar and Nippur, the religious centers of Mesopotamia, near Baghdad in the desert, under Hammurabi and his son Samsuiluna (1792–1712 bce). They did not marry and had no children. Another example is the vestals in Rome, who had to live in celibacy for thirty years (Wildfang 2006). Bagnall and Frier 2006: 113, 116. Bagnall and Frier 2006: 75–110; Scheidel, 2012a. Cf. I.Métr. 83; Anth. Pal. 7.261. See also Chapter 3.

De facto childlessness

165

Undernourishment and infectious diseases were the most common causes of high infant and child mortality.

7.3 De facto childlessness Sometimes, even if there were adult children, the construction of familial support networks nonetheless proved precarious. Some people that had children were de facto childless since they could not benefit from their assistance. Out-migration of children was probably one of the main reasons why parents were left behind without close-kin support. Sons found employment in the metropolis, or signed up for the army.11 This was especially the case among poorer parents who had not sufficient land to provide their sons with a working living and were thus more likely to be left behind.12 Other children were not in the position to provide economic or practical support because they were physically or mentally handicapped. Daughters regularly moved out upon marriage, and not all married in the same village. If the younger generation moved to the metropolis, old parents seemed to have stayed behind in their villages because housing conditions in urban areas were not suitable for an extended family.13 Even a distance of a few hundred miles must have been insurmountable for many from the lower social classes who did not have the means for traveling. Support from children living far away thus can only have been sporadic, inadequate, and unreliable.14 The elderly widow Aurelia Gl[?] from early fourth-century Oxyrhynchus complained to the prefect that she had been robbed by her estate manager while her sons were away as soldiers on foreign service, and thus could not protect her from this fraud.15 Normally, if a couple had several sons, all of them cared equally for their parents.16 We do not know how the system worked if only one or two sons stayed at home and the others lived somewhere else. We have read above some letters written by sons to their mothers at home, which often included vigorous admonitions to siblings at home to take good care of the parents.17 Comparative studies from rural India and Bangladesh show that in those cases sons who did not 11

12 14 16

17

See, e.g., P.Berenike 2.129 from the first century ce; P.Mich. 3.203 ( SB 4.7356) from 114–16 ce; P. Mich. 8.490 from the second century ce; SB 3.6263 from the second half of the second century ce. Cf. Chapters 3 and 5. Cf. Bagnall and Frier 2006: 160–9. 13 Bagnall and Frier 2006: 68–70. Cf. for instance BGU 3.948 discussed in Chapter 5. 15 P.Oxy. 1.71 from 303 ce. Cf., e.g., P.Oxy. 9.1210 from the first century bce to the first century ce; P.Lond. 3.932 from 211 ce. Cf. Chapter 3. P.Oxy. 2.294 from 22 ce; SB 3.6263, ll. 18–31, from the second half of the second century ce; BGU 2.380 from the third century ce.

166

Childless old age: the worst of all fates?

actually live with the elderly parents contributed financially to their support, and a larger piece of the inheritance was allocated to the son who stayed at home and took on the greater part of the care for the elderly.18 We might have a parallel for Roman Egypt from early third-century Hermopolis, where one son relinquished all rights to his inheritance while his two brothers agreed to take over the debts of their late father and provide their mother with maintenance and care in her old age.19 Estrangement might have been another reason why some parents found themselves without any support from their offspring late in life.20 Divorce surely alienated some individuals from their children. As opposed to modern Western societies where custody of children in the case of divorce is regularly assigned to the mother, children in Roman Egypt usually remained living with their fathers; only very small children might have stayed with their mothers while their fathers paid maintenance support.21 Some of them might even have lost contact with their absent parent, especially if he or she married again and moved away. We do not know if the divorced parent who had left was still entitled to support in old age by their children if they had left many years earlier. A husband at least was no longer responsible for maintaining his ex-wife; the return of the dowry usually constituted the security for a woman’s support in old-age supplemented with whatever she inherited from her parents.22 The question was just whether its amount was always sufficient to support the divorcée over the next decades.23 Moreover, a woman’s ex-husband had the right to retain parts of his ex-wife’s dowry for their children’s maintenance.24 We have an interesting case that shows how seriously a divorce could affect a woman’s old-age security. A certain Aurelia Senmoros from Panopolis had married a certain Paniskos around 317 ce.25 They had three children, of whom only one son survived. After eleven years of marriage her husband forced Aurelia Senmoros out of the household and she and her son took refuge with her 18 20

21 22

23

24

25

Cain 1981b: 377. 19 P.Lond. 3.932 from 211 ce. Cf. Kreller 1919: 134, 365. Cf. for instance the case of the old veteran from the Arsinoite nome discussed in Chapter 3: BGU 7.1578. For a parallel in nineteenth-century Egypt, see Tucker 1985: 57–60. For the dowry in the Ptolemaic and Roman period in Egypt, see Yiftach-Firanko 2003: 105–84. In the Roman period we have three different categories of dowries: the pherne, consisting of the wife’s personal trousseau (clothes, jewelry); the parapherne, comprising jewelry, clothing, and chattels; and finally the prosphora, consisting of land and slaves (Yiftach-Firanko 2003: 129–30). For the size of the dowry, which varied greatly in Roman Egypt, see Hopkins 1980 and YiftachFiranko 2003. P.Oxy. 3.497 from the early second century ce; P.Oxy. 3.603 from 169–76 ce; P.Oxy. 49.3491 fr. 2, fr. 4, from 157/8 ce; P.Oxy. 6.905 from 170 ce. See Yiftach-Firanko 2003: 211. P.Panop. 28.

The living situation of the childless elderly

167

father, who assumed the responsibility to support his daughter financially because she was destitute. Her ex-husband had squandered her dowry as security for a loan from a moneylender. Shortly afterward, her ex-husband also succeeded in kidnapping their little boy. At this point Aurelia Senmoros’ father, Aurelius Sansnos, decided to take legal action against his son-in-law and petitioned that the dowry was to be given back to his daughter. He did not ask for his grandson, most probably because it was the usual practice that children stayed with their father in case of a divorce. We do not know the outcome, but this case illustrates that without her father’s support this woman would have been in a desperate situation. In sum, having to cope without children in old age was thus more often an outcome of many events and processes over a person’s life course. Even if children survived to adulthood, they were not always a guarantee for support in old age.

7.4 The living situation of the childless elderly The situation reflected in the Roman Egyptian census returns shows, as we would expect, an immense heterogeneity and changeability of living forms for individuals over the age of 60 who were living without children, with no clear preference for any specific living constellation. Numbers of elderly people living without children are in general small since few individuals lived to the age of 60. Since their number in the census returns is so small (n=12), the distribution of residence patterns for the elderly without children is likely not statistically significant. We give them nonetheless. About 33 percent (n=4) of the elderly were living alone without any other close relatives.26 It could have been, of course, that their children or other relatives lived nearby and cared for them on a regular basis; we just cannot tell from the census returns that were drawn up by household. Another 8 percent (n=1) of the elderly were living with a spouse,27 and another 8 percent (n=1) with a spouse and nephews and the latter’s children;28 17 percent (n=2), two single elderly women, lived in an extended family household, with married siblings and the latter’s children.29 It would be interesting to know what it meant for the elderly’s status and well-being if they were dependent on their wider family network. Did unmarried men 26

27 29

173-Pr-8 (male, 70 years); 187-Ar-30 (female, 60 years, with six slaves); 215-He-1 (female, 80 years, lodging with a family and her own freedwoman); 215-Hm-3 (female, 72 years, with her two slaves). 131-Ar-8 (male, 66 years). 28 215-Ar-4 (male, 66 years). 117-Ar-1 (female, 70 years); 201-Ar-9 (female, 59 years).

168

Childless old age: the worst of all fates?

and women who resided with their siblings and the latter’s families accept a secondary position in the household? These two elderly women’s low rank in the hierarchy of the respective census returns implies this was most likely the case.30 We have another fragmentarily preserved household from 131 ce in which several childless elderly individuals (33 percent; n=4) lived together under one roof. They were all over the age of 60 and might have been unmarried or no longer married siblings who shared one household.31

7.5 Emotional consequences of childlessness Even if material circumstances permitted certain individuals to live their later lives without assistance by their offspring, what did it mean socially and emotionally for them to grow old without children? In general, we can assume that childlessness did not affect a person’s social and emotional well-being abruptly when he or she reached old age. Exclusion and isolation started to be felt much earlier in life. Children were important for establishing one’s social identity because they integrated their parents into the wider community. Children offered them the possibility to take part in networks and social events, such as rites of passage, marriages, or education. The childless have been looked down upon in all cultures over time. Parents have always played the central role in society, and childless individuals regularly occupied a subordinate role with lower social reputation and a less extended social network, a void that was felt especially later in life. Especially for elderly widows in traditional societies, the value of sons cannot be measured in purely economic terms alone. Their presence raised a widow’s social status in the community and the family as well as her emotional well-being.32 Being childless was, like being widowed, a deplorable situation that evoked pity throughout antiquity. In a court speech in fourth-century bce Athens, Isaeus called the possibility of adoption for the childless “the only refuge in their loneliness and the only comfort in life.”33 Eight hundred years later in the Arsinoite nome, a cruel fate had bereft the old Arion from early fourth-century Theadelphia of his wife and his sons. He tried to evoke the pity of the praeses Sabinianus and asked him to reduce his tax liability. Of his enemies he said: “they despise me because of my modest means, my 30 31

32 33

Cf. Chapter 5. 131-Ox-6. We have another household in which three siblings of advanced age were living together: a brother, 56 years old, and his two sisters, 48 and 44 years of age (173-Pr-3). Cf. P.Sakaon 36 from around 280 ce. Cf. Bagnall 1993: 206. Isaeus 2.13 (trans. Edwards 2007: 34).

Alternatives to care by children: the spouse

169

inexperience, my old age and my childlessness.”34 Likewise, in a papyrus from fifth-century ce Oxyrhynchus, an old man implored the praeses to feel sorry for him because of the injustice that had caused him to have “no brother, no relative, and no son to sympathize with him.”35 The old man had agreed to act as a substitute for someone else as a police official but had not been supplied with the necessary staff of helpers that had been promised to him. He was now subject to violent behavior and humiliation in the performance of his office. Losing one’s children to death or remaining childless was seen as a heavy stroke of fate. Men and women alike were anxious, and asked oracles whether or not it was granted to them to beget children.36

7.6 Alternatives to care by children: the spouse as principal source of assistance In cases of childlessness or if the children were unable or unwilling to meet their filial responsibilities, the spouse could be the principal source of assistance. Widowed or divorced men who did not have children from a previous marriage seem routinely to have remarried.37 Even if already of advanced age, if fate bereaved him of his wife and children, a man apparently married again and fathered children who would care for his later years. The 70-year-old Harpokration from the village Tanyaithis in the Apollonopolite nome had married his wife, who was younger by 31 years, when he was already over the age of 50.38 They had two children, a 17-year-old son and a 15-year-old daughter. It is clear that this was at least his second marriage because he already had several adult daughters who owned the house in which he and his family lived. The house had probably belonged to Harpokration’s first wife, who had bequeathed it upon her death to her daughters. The option of being cared for by a spouse in old age was more relevant for men than for women, since wives were generally younger than their husbands. Moreover, since women regularly did not inherit from their husbands but instead just received their dowries back, a childless marriage had especially adverse consequences for them. An example is the case of Lucretia Octavia, a young woman with Roman citizenship living in 34

35 36

37

P.Sakaon 41 from 322–4 ce: καταφρονοῦντες τῆς μετριότητός μου καὶ ἀπραγμοσύνης ὁμοῦ τε καὶ τοῦ γήρους καὶ τῆς ἀτεκνίας μου. P.Oxy. 6.904. P.Oxy. 12.1477; P.Oxy. 38.2832–3; P.Oxy. 47.3330; P.Leid.Inst. 8. Cf. Browne 1970: 95–100; 1974; 1983; Rowlandson 1998: 282–4. Bagnall and Frier 2006: 126–7. 38 117-Ap-7 from 211 ce.

170

Childless old age: the worst of all fates?

Philadelphia in the early third-century ce Arsinoite nome. She was married to Lucretius Ignatius Rufus, who was several years older. During the course of their marriage she had been busy caring for her ailing husband, as his will states: “Lucretia Octauia [c]oniugae meae, qui multum laborauerit in infirmitatem meam.”39 When her husband eventually died, however, he left his wife only a small plot of land and a half-share of a house with its furnishings, while his brother was designated as his sole heir. If Lucretia and her husband had children, they would have inherited his property, and Lucretia would probably have been made the trustee of her late husband’s property, as long as their children were still minors. However, since the marriage had remained childless, Lucretius, even though he explicitly acknowledged in warm words his young wife’s efforts and sacrifices during his lingering illness, made sure that his property fell back to the agnatic line of his family.40 We have another example from early second-century Oxyrhynchus of a widow who did not inherit: A certain Didymus confirmed to Taysorapis, his stepmother and the wife of his recently deceased father, that he had received all of his father’s worldly goods: “I acknowledge that I have received the goods and utensils and household furniture and all of the other things that were left by the above-mentioned and deceased Didymus, my father and your husband.”41 The continued use of these goods was only bequeathed to a wife if the couple had lived in an independent household and if there were joint children for whom she would administer it until they came of age.42 In later times Tisoïa, the wife of Dioskoros, from mid-fifth-century Antinoopolis, who had cared for her husband in his old age, fared much better.43 Her husband made her full heir of his property – even though the marriage had been childless – and acknowledges specifically her many acts of kindness and her care for him. This arrangement might only have been possible because there were no close relatives on the husband’s side who could have contested his will.

39 40

41

P.Diog. 10. Cf. Rowlandson 1998: 343–4, no. 270; Schubert 2000: 70–2, no. 8. Widows in Ptolemaic times seem to have been slightly better off than those in Roman times: when the husband of the childless Axiothea from third-century bce Boubastos in the Fayum died, not only did she receive her dowry back, but Peisias, her late husband, even though he had a son from a previous marriage, left her a house and a slave-girl, which built a solid basis for her old age (P.Petr.2 1.13 from 238/7 bce; cf. Bagnall and Derow 2004: 241–2, no. 148). According to his late father’s will, Peisias’ son received another house in Alexandria and four slaves and had to share the furniture in the house in Boubastos with his stepmother. Even better cared for was the widow of the Egyptian Pachnoubis in the late second century bce (SB 18.13168 from 123 bce). Tathotes’ husband left her all of his property, consisting of a house, land, cattle, and sheep, even though he had two sons from a previous marriage, who were to inherit only one coverlet and a bed from their late father. P.Oxy. 64.4433 from 130(?) ce. 42 Cf. Chapter 4. 43 SPP 1, pp. 6–7, no. 1.

Support by other kin in the household

171

7.7 Support by other kin in the household Children were morally and socially expected to provide for their aged parents’ needs. However, were siblings, cousins, nephews, and nieces also required to step in when their old childless sister, uncle, or aunt was in need of financial support or practical care? Could they be held liable to maintain their relatives? While in classical Athens there was a legal obligation to care for one’s grandparents and orphaned minor relatives,44 siblings, cousins, nephews, and nieces were apparently not obliged to assume this responsibility.45 Since the neglect of a nephew’s obligation is never mentioned in our petitions from Hellenistic and Roman Egypt and since it is not mentioned in Athenian forensic speeches or Roman law, we might assume that there was no enforceable right to support and maintenance from one’s more distant kin. Comparative studies on other traditional societies show, however, that the childless and unmarried often developed much closer relationships with parents, siblings, or other relatives than did their married counterparts.46 Even if there were no legal bonds, it is well documented that by transfers of wealth, actual or de facto childless elderly bought themselves support in old age from their wider kin group. Since the recipients were not morally or socially obliged to provide old-age care, it was more of an economical calculation. In order to entice a nephew or niece to take over the task of caring for an elderly relative, this chore needed to be met by an attractive exchange in the form of property. The expected gain had to be considered sufficient to outbalance the costs and labor of care and maintenance. A childless uncle or aunt of some means surely led nephews and nieces to cherish hopes of an impending inheritance. From several documents from the Theognostos archive from the third century we learn that a man and his three sons inherited from a childless brother.47 The latter left a house and a courtyard to his brother, who later handed them down to his three sons,48 and left another house and courtyard directly to his three nephews.49 Similar constellations are recorded in the census returns: Isarous, a 70-year-old elderly woman from early second-century Arsinoe, lived with her sister-in-law, Zois, 53 years of age, and the latter’s three adult children (a 32-year-old son and a 28-year-old daughter who were married to each other, and a 33-year-old daughter who had either not yet married or

44 47 49

Isaeus 1.39–40; cf. Rubinstein 1993: 64–5. 45 Rubinstein 1993: 64–5. 46 Foner 1984: 33. P.Lond. 3.935 from 217 ce; cf. Sijpesteijn 1989: 214–16. 48 P.Lond. 3.932 from 211 ce. P.Lond. 3.940; cf. Sijpesteijn 1989: 215–16.

172

Childless old age: the worst of all fates?

was already divorced).50 The head of the family, the brother of Isarous and husband of Zois, was already dead at this point. Childless women in Roman Egypt who were divorced or widowed routinely seem to have returned to their natal home and spent the rest of their lives there if they did not remarry. We have a case where a brother provided for his childless sister in his will. He wished that his sister, Theognoste, would stay with his wife and his three minor sons, assigning to her a yearly allowance of corn and wine and a lifelong right of abode in his house.51 Co-residence of more distant kin, as we saw in Chapter 2, seems to have been, however, less stable. Since an elderly uncle or aunt living in the household of their nephews wielded less authority than a father would have done, tensions were bound to occur. In second-century Oxyrhynchus the old Theomnestos petitioned against his nephew and niece about the ownership of a house of which he owned the larger part. In an effort to evoke the epistrategos’ pity, Theomnestos stressed that he was old and childless since his sons had died.52 Even though cross-cultural studies suggest that relationships with the wider family tended to be characterized by greater closeness and proximity over the life course in traditional societies than in the modern West, where retirement pensions allow ties with the wider family network to be rather loose and sporadic, the wider kin group was less dependable as a safety net in crisis situations.53

7.8 Care by slaves and freedmen Seeking support among one’s household attendants was an option only for the wealthier strata of society. Of the 233 households recorded in the Roman census returns Bagnall and Frier analyzed in the first edition of their Demography of Roman Egypt, only 36, or less than one-sixth, contained slaves. Nearly 60 percent of the households with slaves owned only one or two slaves. About 30 percent had three to six slaves, and only about 10 percent of these households had seven or more slaves. An example is the elderly Thermoution, who took a little slave-girl, born in her own household, into her care and loved and raised her like her own daughter, in the

50 52

53

117-Ar-1. 51 P.Oxy. 27.2474 from the second half of the third century ce. P.Oxy. 34.2708 from 169 ce or 201 ce: γέροντα ἄνθρωπον καὶ ἄ[τ]εκνον διὰ τὸ τοὺς υἱούς μου τετελευτηκέναι, ἀξι[ῶ] σε τὸν κύριον κριτὴν. Cicirelli 1994: 7–20; Wenger 1997: 1–21.

Care by slaves and freedmen

173

hope of having in her a support in old age.54 Two elderly, apparently childless widows in the census returns were living on their own, but owned slaves who were available to care for them. The 60-year-old Isidora from second-century Tebtunis lived with her two slave women, 45 and 20 years old, and the women’s four daughters, who were between the age of 1 and 8 years. The women and little girls probably served as her surrogate family.55 The 72-year-old Aurelia Demetria was widowed and childless as well.56 But she too was wealthy enough that she did not have to worry about adequate support and care in her old age. She owned the house in which she lived and, in addition, had two slaves, a 34-year-old (sex unknown) and a 28year-old male slave, who took care of her. Another option to organize support in one’s household, business, or farm was the so-called paramonē arrangement which is already testified for the ancient Near East. Paramonē comes from the verb παραμένειν (‘to stay with someone’) and denoted a legal obligation to remain in someone’s service for a designated period of time. For several places in the ancient Near Eastern and Hellenistic world we have evidence of this custom.57 In this special form of manumission contracts the owner stipulated that his former slave should stay with him and serve him until his death (and sometimes even until his wife had died too) before he became entirely free.58 Some of these obligations encompassed also the observance of a proper burial. Paramonē contracts are also well documented for Roman Egypt. Here it was usually not a freed slave but a free workman who contracted his services for a certain period of time, in return for maintenance and sometimes also accommodation and a cash wage. Under paramonē he did not have freedom of movement and had to remain available (παραμένειν) to work for the master.59 These contractual labor agreements did not last, however, until the employer’s death and did not explicitly mention old-age care. Therefore, the employment of a paramonē contract as an old-age strategy

54

55

56 57

58 59

P.Oxy. 50.3555 from the first or second century: θεραπαινίδιόν μου οἰκογενέ[ς], οὗ ἔστιν ὄνομα Πεῖνα, ἠγάπησα καὶ ἐτημέλησα ὡς θυγάτριο[.] ἐπ᾽ ἐλπίδι τοῦ ἡλικίας γενόμενον ἔχειν με γηροβοσκόν, γυναῖκα ἀβοήθητον οὖσαν καὶ μόνην. 187-Ar-30. Three further slave women, 38, 42, and 68 years old, most probably a mother and her two daughters, were classed as runaways in this declaration. 215-Hm-3. Cf. Parker 2005: 77–8; Hopkins and Roscoe 1978 for Delphi; Ricl 1995 for Phrygia; Darmezin 1999 for central Greece; Youni 2000: 54–120 for Macedonia; Petsas et al. 2000 for Macedonia; Veenhof 1998: 134 for the ancient Near East. Cf. Gnilka 2001: 272–3; Zelnick-Abramovitz 2005. Westermann 1948; Taubenschlag 1955; Adams 1964; Samuel 1965; Hengstl 1972.

174

Childless old age: the worst of all fates?

does not seem to have played a significant role in provision for old age in Roman Egypt.

7.9 Support networks beyond the household walls Informal patterns of kin support and kinship systems extending beyond the confines of households are difficult to detect in our sources.60 Since census returns are based on the household as registration unit, our ability to draw any conclusions about the social role of kin and affines who were not living in the same household, and about kinds and degree of mutual support, alliances, and conflicts, is limited. Neighborly support, for instance, is barely possible to trace.61 We know from studies on societies for which we have better evidence than for antiquity that networks of support often penetrated household boundaries. Besides the exchange between parents and children, wider family and social networks were of great importance in situations of crisis, such as the birth of a child or the death of a family member, e.g., in looking after small children or preparing meals together, lending small sums of money or goods, contributing to funeral expenses, or helping with property building or farming.62 Moreover, relatives often undertook joint economic activities together, even if they did not live in the same household but only in adjacent buildings, which in the census returns were evidently recorded as separate households.63 The same might be true for solitary households, especially of elderly persons. We just do not know if these seemingly isolated individuals had relatives living nearby who visited every day and lent help and support in various ways, as if they were living in the same household. The census returns are silent about these matters, and in addition present an image of clear-cut household boundaries that surely did not reflect reality. We must look at other kinds of sources to form a more complete picture of family and household life. Occasionally private letters allow us a glimpse into informal networks of support between households. When Aurelius Stephanus was away in the city of Alexandria, he wrote a letter to his brother mentioning the purchase of yokes and manure, and other domestic matters, and asked him to go over to his farm, look after his wife, and help her until he returned. “Do not fail to go there, both to help my wife until I return, and for the sake of the 60

61 63

For this more recent emphasis on support networks that reach beyond the household walls, see Hammel 1984; Kertzer 1984; Loefgren 1984; Netting 1984; Plakans, 1984. Levi 1990: 568. 62 e.g., Agree et al. 1999. For present-day Istanbul, see Duben 1982. For instance, for fifteenth-century Tuscany, see Klapisch-Zuber and Demonet 1972: 88.

Adoption

175

irrigation.”64 Letters sent between households document that news, greetings, and small goods were regularly exchanged. Thermouthas from the early second century ce wrote home to her mother thanking her for the cakes her mother had dispatched to her. She asked her mother to send her blankets and fine quality wool, remarked in passing that she was seven months pregnant, and conveyed greetings to all members of the family.65 Another woman from the fourth-century ce Arsinoite nome wrote to her son who was living elsewhere: “I have sent you a basket of parsley roots, a basket of shoots and a basket of some small raisins. Wash them and put them outside in the sun, wherever possible.”66 In another letter written on an ostracon in the second century ce, a certain Maximo(u?)s advised her (?) sister about the upcoming birth of her child: If you are coming to your day of giving birth, write to me so that I may come and perform your delivery, since I do not know your month. I wrote to you in advance for this reason, so that you might also act in advance and write to me so that I would come in the provision boat, so that I too may remain with you and perform your delivery. [. . .] If you don’t send word to me, you do me no favor. I was going to send you jars for your delivery. I didn’t send them for this reason, so that I might bring them when I come along with the two matia of lupines. [. . .] Send me leaves as for a small basket and I will make it for you . . . I pray for your health.67

Reciprocal exchange of goods and services created a local support system between families and provided relatives, friends, and neighbors with informal support in daily life. Living in nuclear families therefore did not necessarily mean “nuclear hardship.”68 However, these neighborly gifts were presumably too sporadic and random to bank on them in times of need or as an adequate and reliable support in old age.

7.10 Adoption The centrality of children in old-age support, finally, leads to the notion that the adoption of children was the best alternative to biological children. Adoption is a system of fictional kinship that occurred almost everywhere in Eurasian societies across time, cultures, and regions.69 In traditional 64 65 66

67 68

P.Oxy. 6.934 from the third century ce. SB 5.7572 from the early second century ce; trans. Bagnall and Cribiore 2006: 283. P.Benaki 4 from the fourth century ce; trans. Bagnall and Cribiore 2006: 326. The editors of the papyrus suggest that the mother–son relationship was rather spiritual than biological. O.Florida 14 from the second-century Eastern Desert; trans. Bagnall and Cribiore 2006: 167. Cf. Laslett 1988: 153–75. 69 Goody 1969: 66.

176

Childless old age: the worst of all fates?

societies it was a way of redistributing sons between households with too many sons and those without any sons at all. Since high mortality disrupted many families, adoption was an adequate means to make up for this loss. I have discussed adoption practice in the ancient Eastern Mediterranean with special regard to Graeco-Roman Egypt in detail in two previous studies and will refer to those in the following.70 The earliest adoption contracts from Roman Egypt date to the fourth century ce. We find, however, various earlier references to adoptive status in the papyrological evidence from the Roman imperial period.71 I have argued in a related study that the institution of adoption – despite its scarcity in our sources72 – must have been much more common than the census returns imply. Among the roughly 1,450 individuals recorded in the Roman census lists,73 not one declared that he had adopted a child or that he had himself been adopted. Nevertheless, almost every man above the age of 50 had a living male heir, a proportion that is virtually impossible to achieve under the ancient mortality regime. According to Saller’s model life tables we should expect a rate of around 40 percent of men who did not have a living son, and when occupational migration is taken into account, an even higher proportion.74 That means some elderly people appear to have biological sons, even though these children probably were not natural, but adopted sons. Furthermore, some age differences between parents and their children are suspicious. There is, for instance, the woman who is recorded as giving birth at the age of 9 (thus marriage at age 8 is implied), before the likely onset of puberty.75 We further have two mothers who apparently gave birth at the age of 4976 and another who was 51 at the birth of her daughter.77 While reported ages and attested ages at childbirth appear in general to be credible, Bagnall and Frier dismiss these ages as implausible. Moreover, we have a man who had a son 13 years his junior, which implies marriage at age 12.78 We have another two men whose sons were 17 years79 and 18 years80 their junior. Bagnall and Frier did not consider the option that these children were adopted. Adoption would, however, be the most obvious explanation. 70 71 72 73

74

75 80

Huebner 2007, in press a. Cf. Kurylowicz 1983: 61; Rupprecht 1998: 232. See below for further references. Huebner 2007, in press a. Some 1,108 individuals in Bagnall and Frier 1994 (p. xv) plus around 100 individuals in the supplement to Bagnall and Frier 2006, plus another 256 individuals in P.Oxy.Census 1997. The fact that these adoptions were not indicated in the returns is not particularly surprising when we recall that, unlike Roman adoptive nomenclature, Greek onomastic conventions did not generally record adoptive status (Modrzejewski 1955–6: 349; Salomies 1992). 131-Pr-1. 76 159-Hm-3; 173-Pr-5. 77 187-Hm-1. 78 187-Ar-4. 79 131-He-4. 117-Ar-2; 173-Pr-15.

Adoption

177

In addition, since my 2007 article, I have discovered that the census returns record far too few orphans.81 While according to model life tables there was a one-in-three likelihood that a child would lose his or her father before the age of 16, only around 16 percent of all freeborn children aged 15 or below recorded in the census returns were living without a father. This number is even lower if we look at just the boys: only 14 percent were not living with their father. Furthermore, even if children could enjoy their mothers for a few years longer because the age difference between them and their mothers was around 5–10 years lower than that between children and their fathers, at least one in four would have lost their mother by the age of 15. In the census returns, however, under 13 percent of all children were living without their mother. Finally, according to model life tables, around 11 percent of all children would be complete orphans on reaching the age of 15. In the census returns, however, under 7 percent of recorded children aged 15 and younger were without both parents. Again, just looking at the boys, this number was even lower: just 4 percent of all recorded boys were apparently complete orphans. We have to remember here that the census returns only record co-residence. Divorce or migration could also have been the reason why a child did not live with his or her parents in the same household. Fathers were more likely to be absent for work reasons, while mothers often left their children behind in their paternal home in case of divorce or the death of the husband. Numbers for apparent half and complete orphans should thus be even considerably higher than in the model life tables. The explanation for these discrepancies between model life tables and the census returns lies most probably again in the fact that adoptive status was not recorded in the census. It seems likely that Roman Egyptian children who lost their parents to death found foster or adoptive parents (boys seem to have been luckier than girls) and were registered in the census as their natural children. They were adopted by an uncle who might have been already living in the same household in any case, or by the new spouse of the surviving parent. Adoption by a step-parent is a common feature in many ancient societies: we hear of several cases from the ancient Near East where stepfathers, but also stepmothers, are recorded as adopting the children of their new spouse upon marriage, in order to assimilate marital ties to natural family ones and to secure support in old age in the absence of their own children.82 In classical Athens the orator Isocrates married a widowed mother of three sons, and adopted the youngest of his stepsons, Aphareus, because he had no 81

Cf. Huebner in press a.

82

Cf. Huebner in press a.

178

Childless old age: the worst of all fates?

child of his own.83 We also have several cases from imperial Rome. Tiberius was adopted in 4 ce by his stepfather, Augustus. Claudius adopted his stepson Nero and, in order to strengthen the family ties, married him to his daughter Claudia Octavia – all this even though he already had a natural son, Britannicus. Adoption of stepchildren is also indirectly recorded in the census returns from Roman Egypt: Chentmouphis, whose record we have in two consecutive census returns from 131 ce and 145 ce, registered in both returns his wife’s daughter from a previous union as his daughter and as the full sister of the son he and his wife later had together.84 We only know by chance that she was not his biological daughter: From a cover letter dated to 161 ce, which accompanied these two copies of the census returns, we learn that the girl was only the half-sister of her brother on the mother’s side and her father was unknown.85 Women also adopted the children of their husbands. In secondcentury ce Oxyrhynchus a childless woman adopted her husband’s two sons from a previous marriage and made them heirs to her property.86 In our adoption contract from late fourth-century Hermopolis it is the paternal uncle who adopts the orphan boy.87 While modern adoption practices center upon the welfare of the orphaned or abandoned child, it is the adopter in the Graeco-Roman world who benefitted at least as much as the adoptee.88 As Goody has remarked for East Asian societies, so also in the ancient Greek and Roman world, adoption was seen as an “heirship strategy” under demographic constraints.89 Adoption was a means for coping with the discrepancies that arose between ideals for succession and inheritance and demographic realities. In some preindustrial patriarchal societies adopted sons constituted up to one-fourth of all male heirs.90 For as Goody remarks, “there is a knifeedge between having too few heirs for comfort and producing too many claimants for limited resources.”91 83

84 86

87 88

89 90

91

Ps.-Plut. Mor. 838a, 839b; cf. Isaeus 8.40–2. See Corbier 1991b: 72–3; Rubinstein 1993: 87, 96, 101; Watson 1995: 39–42; Cox 1998: 90; Patterson 1998: 199. 131-Pr-1; 145-Pr-1. 85 Cf. Huebner 2007: 38; Rowlandson and Takahashi 2009: 104. P.Oxy. 3.583 descr. from 119/20 ce P.Sijpesteijn 43 (2007). See Kreller 1919: 176, 247, 266; Taubenschlag 1959: 264. P.Lips. 1.28 from 381 ce. “Adoption allowed the childless man, or the man without male heirs, to choose an heir for the sake of continuing his, the adopter’s, line” (Cox 1998: 148). Goody 1969: 68; cf. Befu 1971; Goody 1973; Wrigley 1978: 138; Wolf and Huang 1980. Among the samurai of seventeenth- to nineteenth-century Japan, one-fourth to one-third of all the successions were adoptions because a natural male heir was lacking (Moore 1970: 619–20). Another study focusing on nineteenth-century Japan in general shows that between 11 and 24 percent of the male population were adopted (Kurosu and Ochiai 1995: 261–88). Goody 1973: 7.

Adoption

179

Since in these societies usually all the adoptions took place to ensure succession in the family line, those who were orphans or who had no chance of succeeding in their own households of origin regularly became successors via adoption. The adopted sons were usually younger sons from families with several male offspring, so-called “surplus sons,” to whom the adoption provided a chance for advancement in an heirless household.92 Adoption thus functioned as a redistribution of sons between households to avoid the extinction of households and the defragmentation of others, and was mutually beneficial.93 Most of these sons were closely related to the adopter, such as a grandson, a younger brother, a nephew or a cousin.94 Not only did men without sons adopt. Some men also considered adoption when their own son was of weak health, physically or mentally disabled, or still too young when they themselves were approaching old age. The most important sources for adoption practices in the ancient Greek world are the law court speeches by Isaeus and Pseudo-Demosthenes from fourth-century Athens.95 As we would expect from the model drawn up above, adoption of a male heir in classical Athens usually took place when the adopter had abandoned hope that he would produce a natural son.96 The institution of adoption in classical Athens provided an individual or a couple with a son who could care for them in their old age, who continued his father’s business, who perpetuated the family cults, and who maintained the agnatic lineage, and so the continuation of the oikos. Also in classical Athens there was a strong preference for adopting an heir from one’s agnatic lineage.97 Families with more than one child would not have resisted when a (rich) heirless male relative requested one of their sons. Moreover, it was not considered dishonorable to give one’s child away in adoption, especially as both sides gained an advantage: the adopter received a legitimate child and heir, and the adopted son became the sole heir to his adoptive father, and 92 95

96

97

Moore 1970: 620. 93 Wrigley 1978: 147; Hanley 1985: 220. 94 Moore 1970: 622. Isaeus 1–7, 9–10; Ps.-Dem. Or. 41, 43, 44. The evidence found in the Athenian forensic corpus of the fourth century bce is supplemented by several of Menander’s comedies (Dyskolos, Samia, and Adelphoi; cf. Aspis, 258–69). See also Rubinstein and Bjertrup 1991: 141; Rubinstein 1993: 93. Isaeus records the old and childless Athenian Menekles, who had been married twice but both marriages had failed because of his sterility: “Menekles began to consider how to avoid being childless, and how to have someone who would look after him in his old age while he was alive, and when he was dead, would bury him and, in the future, perform the rites for him” (Isaeus 2.10; trans. Rubinstein 1993: 63). Cf., e.g., Isaeus 7.14. This is an apparently cross-cultural phenomenon, to which only modern Western society seems to be an exception (MacDowell 1978: 100; Lindsay 2001). There was no legal bar that hindered the adopter from adopting a complete stranger (Harrison 1968: 88; Lindsay 1998–9: 94), but adopting a foundling caused many problems, since Athenian citizens were allowed to adopt only another free citizen (Lindsay 1998–9: 92; cf. Gernet 1955: 129–31; Hunter 1993: 103–8, 117–19; Rubinstein 1993: 117–25; Cox 1998: 126).

180

Childless old age: the worst of all fates?

therefore no longer needed to share the estate of his biological father with his biological siblings.98 If this option, i.e., the adoption of a nephew, was unavailable, Athenians usually reached out to more distant relatives in order to perpetuate their lines. Sometimes it happened that the adopter was forced to look beyond blood relatives if close male agnates were lacking.99 Once a suitable candidate had been found, the adoptee was introduced to the family, phratry, and deme of his adoptive father and inscribed in the public register, in the same manner as a natural son. Everything was done in such a way as to secure for the adoptee a status that replicated that of a natural son.100 After an adoption inter vivos the adopted son was, in fact, on a par with a natural son.101 Along with the legal relationship came the expectation of “normal” filial and familial relations. According to the court speeches given by Isaeus and Pseudo-Demosthenes, it was expected that the adopter would treat the adopted son just as he would a biological son. In return, the adopted son was obliged to act like a real son, e.g., show filial love and affection, care for his adoptive father in his old age, provide him with a proper burial, and perform the accustomed rites at his grave.102 On the father’s death, a son adopted during his adoptive father’s lifetime had the same right as a natural son to come into his estate.103 For later periods of the Hellenistic age, adoption practice is mainly attested through epigraphic sources from Hellenistic and early Roman Greece104 and Asia Minor105 and through adoption contracts on papyrus from later Roman 98

99

100

101 102

103

104

In another of Isaeus’ speeches, an Athenian mother persuaded her husband, with whom she had several sons, to have one of their sons adopted into the estate of her childless brother for this very reason (Isaeus 11.49). Isaeus mentions a further rationale for adoption: “People get their children adopted into other oikoi, when they are unlucky in financial matters, so that the children will not have to share in their father’s atimia” (Isaeus 10.17; trans. E. S. Forster, Cambridge, MA, 1927). For instance, the old and childless Menekles from the first considered adopting a relative (Isaeus 2.10). But his only brother had only one son, and Menekles did not want to ask for him and thus deprive his brother of his only heir (Isaeus 2.10 and 21). Menekles therefore decided to ask the two brothers of his ex-wife, with whom he had parted on friendly terms, if one of them would consent to his being adopted. One of these brothers agreed and Menekles adopted him and introduced him to his phratry and deme (Isaeus 2.12; cf. Rubinstein 1993: 33; Lindsay 1998–9: 95). This father–son relationship lasted happily for 23 years after the adoption until Menekles’ death (Isaeus 2.15, 2.45). Only if he had produced a son with his stepsister in his adoptive oikos was he permitted to return to his own family (Isaeus 6.44, 9.33, 10.11; Ps.-Dem. 44, 64, 68; Rubinstein 1993: 57–8). Also, in modern India the adopted child is excluded from the succession to his biological father’s name and inheritance (Goody and Tambiah 1974: 81; Lindsay 2001: 194). Ps.-Dem. 44, 29; 44, 42–3; 53–5; Isaeus 2.2, 2.17, 5.16, 7.17; cf. Rubinstein 1993: 40–1, 45. Rubinstein 1993: 64–76. If an adopted child in ancient Mesopotamia said to a chamberlain or epicene who was his adoptive father and had brought him up “You are not my father,” his tongue was cut out (David 1927; Driver and Miles 1955: 75–7; Lindsay 2001: 192). An heir, designated by will, could not directly come into an estate but first had to state his claim before the archon (Harrison 1968: 95). Rubinstein and Bjertrup 1991: 139–51. 105 Poma 1972; Stavrianopoulou 1993.

Adoption

181

Egypt.106 For Asia Minor most of our evidence comes from the island of Rhodes and the Roman province of Caria. For Rhodes alone we have around 550 instances of adoption recorded in inscriptions dating from between the third century bce and the first century ce,107 but we should wonder whether we are dealing here with a local epigraphic habit with regard to recording adoption,108 which would mean that adoptions were probably just as common in other regions.109 An extract from the registry law of succession from Syrian Dura Europos from the second quarter of the second century ce states that adopted children were as entitled to inherit as natural children, and ahead of parents, siblings, and grandparents of the deceased.110 In Egypt the institution of adoption was already well developed in the New Kingdom period.111 Our earliest testimony in Greek for adoption in Egypt dates to the early Ptolemaic period (248 bce) and speaks of the register of adoption contracts.112 As mentioned above, for later times we have occasional references to adoptive status in the Roman Egyptian papyri. However, for the most part these documents contain little information.113 The term for adoption was θέσις. The usual adoptive formula in the papyri is θέσει followed by the adoptive father’s name in the genitive, which designates the adoptive relationship.114 A few times we have (κατὰ δὲ) υἱοθεσία(ν).115 Once we find κατὰ δὲ τεκνοθεσίαν.116 106 108 110

111 112

113

114

115

116

Kurylowicz 1983. 107 Poma 1972; Stavrianopoulou 1993. MacMullen 1982: 233–46; Meyer 1990: 74–96. 109 Cf., however, Poma 1972: 185–91. P.Dura 12. The adopted son in P.Lips. 1.10, however, inherited 1⁄12 of his adoptive father’s inheritance, while the adopter’s natural daughter inherited 11⁄12. Cf. Tacoma 2006: 201; Rowlandson and Takahashi 2009: 124. Gardiner 1941. P.Col.Zen. 1.58. For adoption in Pharaonic times, see Seidl 1968: 54, 80; Allam 1974: 277–95. For adoption in the Graeco-Roman period of Egypt, see Mitteis 1906: 179; Taubenschlag 1955: 261–321; Kurylowicz 1983: 61–75; Beaucamp 1990–2: 48–52; Krause 1994–5: iii.80–1. Cf. those epikriseis declarations that stress that the son was natural, not adopted: P.Oxy. 2.257 from 94 ce; P.Oxy. 10.1266 from 98 ce; SB 14.11271 from 117 ce; P.Oxy. 7.732 from 153 ce; P.Oxy. 18.2186 from 260 ce; PSI 5.457 from 268 ce; P.Mich. 14.676 from 272 ce. P.Oxy. 10.1266 from 98 ce (male); P.Oslo 3.114 from the first to early second century ce (male); P.Oxy. 1.46 from 100 ce (male); P.Heid. 4.329 from 105/6 ce (male); P.Oxy. 3.492 from 130 ce (male); PSI 7.732 from 153/4 ce (male); P.Oxy. 8.1123 from 159/9 ce (male); P.Mert. 1.18 from 161 ce (male); P.Oxy. 3.502 from 164 ce (male); P.Oxy.Hels. 36 from 167 ce (male); SB 18.13176 from 168 ce (male); SB 20.14395 from 181 ce (male); P.Oxy. 14.1721 from 187 ce (male); P.Oxy. 74.4990 from 188/9 ce (male); P.Oxy. 66.4531 from 196 ce (male); P.Fuad. 1 Univ. App. ii 321 from the first/second century ce (male); P.Oxy. 3.504 from the second century ce (female); P.Wash.Univ. 1.2 from the second century ce (male); P.Oxy. 31.2583 from the second century ce (male); SB 14.11337 from the second half of the second century ce (male); P.Oxy. 14.1719 from 204 ce (male); PSI 5.457 from 269 ce (male); P.Strasb. 1.4 from 550 ce (two sisters adopted by the same man). Cf. Kurylowicz 1983: 61. P.Oslo 3.114 from the first to early second century ce (male); P.Erl. 28 from the first half of the second century ce (male); P.Oxy. 50.3593 from 238 ce (male); P.Oxy. 9.1206 from 335 ce (male); P.Lips. 1.28 from 381 ce (female); P.Köln 7.321 from the seventh or eighth century ce (male). P.Oxy. 46.3271 from 47–54 ce (male).

182

Childless old age: the worst of all fates?

Θέτος is used in a papyrus from the third century ce for the adoptive son.117 In an early eighth-century papyrus from Aphrodites Kome the adoptive son is called υἱοποιήτος.118 Most of the documents that contain an adoptive formula date to the second century ce. This should not surprise us as it coincides with the general chronological distribution of the papyrological evidence. Both males and females were adopted in Roman Egypt, though there seems to have been a strong preference for males.119 Our earliest surviving contract of adoption comes from the second quarter of the fourth century ce. It is only from this and later evidence (two contracts from the fourth century, one from the sixth, and one from the seventh/eighth century) that we are able to gain more information on the procedure of adoption in Roman Egypt.120 Yet, this does not mean that adoptions necessarily occurred less frequently before this point, just that from the fourth century onward a written contract seems to have become more important.121 The adoption contracts provide us with detailed information on the procedure of and legal background to adoption. From these documents we learn that adoption was considered binding and permanent. The biological parents lost all rights regarding their child, who thus became the legal child of the adopting parents, who were responsible from then on for supplying all the child’s needs. The adoptee forewent all kinship ties with his or her biological parents. An adoption contract from the first half of the fourth century states: “We, Heracles and his wife Isarion, on the one part agree that we have given away to you, Horion, for adoption our son Patermouthis, aged about 2 years, and I, Horion, on the other part, that I have him as my genuine son, so that the rights proceeding from succession to my inheritance shall be maintained.”122 Further down in this document, the adopting father confirmed that he had registered the adoptee in the official records as his lawful, legitimate son. The adoptee even assumed the 117 119

120

121

122

PSI 4.294 from the third century ce. 118 P.Lond. 4.1422, l. 74. Adoption of females: P.Oxy. 1.46 from 110 ce; P.Oxy. 3.492 from 130 ce; P.Oxy. 3.502 from 164 ce: P. Oxy. 3.504 from the second century ce; P.Strasb. 1.4 from 550 ce. Cf. Taubenschlag 1955: 100 n. 15, who considers P.Oxy. 8.1123 from 58/9 ce as another instance of female adoption. The adoption formula θέσει refers, however, in my opinion back to Apion, not to his mother, Ammonous (Ἀπ̣̣ ίω̣ν̣ Σαραπίωνος τοῦ Ἰσχυρίωνος μη[τρὸ]ς Ἀμμωνοῦτος θέσει Βαλλάρου Πε[σουχίω]ν̣ ος ἀπ’ Ὀξυρύγχων πόλεως). P.Oxy. 9.1206 from 335 ce; P.Lips. 1.28 from 381 ce; P.Oxy. 16.1895 from 554 ce; P. Köln 7.321 from the seventh or eighth century ce. For the adoption contracts see also Taubenschlag 1955: 327; Kurylowicz 1983: 61–75; Beaucamp 1990–2: 48–52; Krause 1994–5: iii.80–1; Rupprecht 1998: 232. A written contract was merely a document of proof without any constitutive or dispositive power: “rather, a purely oral conclusion of a contract was also sufficient” (Rupprecht 2005: 331, 335–6). P.Oxy. 9.1206 from 335 ce; trans. Winter 1933: 58.

Adoption

183

right of the firstborn son. In another adoption contract from Heracleopolis, an adoptive father also declared that his adoptive son would take the status of a biological and firstborn son.123 Thus, a boy or a young man as adoptee became the legitimate son and heir of the adopter “as if he were engendered by you from your own blood,” as another adoption contract from fourthcentury Hermopolis affirms. The adopting father acknowledged his responsibility for “feeding and clothing the child in decent and appropriate fashion as my own legitimate and physical son, as though he were born from me.”124 A girl adopted in Oxyrhynchus in 554 ce likewise became the legitimate daughter of her adoptive parents. The biological mother, who gave her up, assured the adopting couple that she handed her over from then on until forever as the couple’s legal daughter and that she had no right to take her away again.125 The adopting couple took over the role and duties of genuine parents and agreed to fulfill all the girl’s needs. All our adoption contracts stress the care for the child. The adopting parents promised to provide the child with all the necessities of life and to act as their “true” parents; that is, also to name them as heirs to their property. In return they overcame their biological childlessness and won support for their old age. Horion from fourth-century Oxyrhynchus, already mentioned above, who adopted a 2-year-old boy called Patermouthis from a couple from the same city, pledged to hold this child as if he were his own, register him officially as his son, and make him his heir. He promised not to send him away or to reduce him to slavery since the child and his parents were of free status.126 Patermouthis’ parents and his adoptive father apparently were not related to each other. In another adoption contract from sixth-century Oxyrhynchus, the adoptive parents promised to “supply her needs and fulfill the position of parents to daughter.”127 In another case from early second-century Oxyrhynchus a couple, a certain Epikrates and his wife Thaisous, had adopted their slave-girl, Aphroditous, having freed her before. Here Thaisous is explicitly styled as the adoptive mother of Aphroditous, which provides us with evidence that in Roman Egypt women were seen as able to adopt.128 The reason for adopting their slave-girl seems to have been mutual affection. When the document that 123 124

125 127

P.Köln 7.321 from the seventh or eighth century. P.Lips. 1.28 from 381 ce; for a translation see Rowlandson 1998: 297–8. A Syrian papyrus from Dura Europos from the second quarter of the second century ce states that adopted children were as entitled to inherit as natural children, and before parents, siblings, and grandparents of the deceased (P.Dura 12). P.Oxy. 16.1895. 126 P.Oxy. 9.1206 from 335 ce. P.Oxy. 16.1895 from 554 ce; trans. Rowlandson 1998: 298–9, no. 234. 128 P.Oxy. 3.504.

184

Childless old age: the worst of all fates?

preserves this information was drawn up, Epikrates, the adoptive father, was already dead and his adoptive daughter grown up and married. She was selling a piece of land and her adoptive mother gave her consent to the transaction. Epikrates most likely had made Aphroditous heir of his estates and had assigned his wife a lifelong usufruct and right of abode. Taubenschlag and Kurylowicz believed that the practice of adoption in Roman Egypt was firmly based on Hellenistic law,129 and I followed their assumption in my 2007 article. However, having done further cross-cultural studies on adoption and made a closer reading of the sources I no longer believe that this assumption is supported by the evidence. In addition, details given in the adoption contracts from later Roman Egypt ostentatiously contradict classical Greek laws on adoption. Adoptions in Egypt did not follow classical Athenian adoption practice, where we hear only of the adoption of young men. In the contracts from Roman Egypt all adoptive children were minors: we have a 2-year-old boy, a 10-year-old boy, and a 9-year-old girl. Furthermore, even though all surviving adoption contracts date to post-Antonine times (after 212 ce), they conform little with Roman law concerning adoption. Adoption in Roman Egypt did not confer patria potestas on the adopter. Adoption contracts from Roman Egypt stress the child’s welfare and inheritance rights as their central focus, not the benefits (such as care in old age and social progeny) the adopter would gain from adopting. The parenting aspect of adoption and the assimilation of natural family ties seem to have been a dominant feature. The parents who gave away their child were most often motivated by extreme poverty and the consequent inability to raise their child.130 As far as the adoption of minors goes, adoptions in Roman Egypt resemble the practice in ancient Mesopotamia rather than in classical Greece or Rome. In ancient Mesopotamia children were usually adopted right after birth or at least as infants.131 Reasons to give up one’s own children for adoption were usually the inability to bring them up as a result of poverty. The law codes of Hammurabi reveal that candidates for adoption were not necessarily orphans. For, if the adoptee injured his adoptive parents or if the adoptive parents did not take care of the adoptee, the code stipulates that the adoptee should return to his natural father.132 Issues that 129

130

131

For the terminology used, see Taubenschlag 1955: 263; Kurylowicz 1983: 61. For the unity of Greek law, see Todd and Millett 1990: 8–10; Modrzejewski 2005: 344, 354; Rupprecht 2005: 329. P.Oxy. 16.1895. We might suspect, however, that in those cases where the adoption took place under different circumstances, e.g., among family members in order to provide a brother with a son, a contract did not always seem necessary. Code of Hammurabi §185. 132 Code of Hammurabi §§185–6, 190.

Adoption

185

are addressed in these Mesopotamian contracts between the adopter and the natural parents are the name the adoptive child was going to take, the status and inheritance rights of the adoptive child, the rights and obligations the adoptive child had toward his adoptive parents such as care and support in old age, and the impossibility of the child being reclaimed by its natural parents at a later point.133 If in ancient Mesopotamia the family property was based not on land, but rather on a workshop, men without sons would be looking not only for an heir to their property but rather a successor in their workshop. Thus a craftsman who did not have a son considered adopting a young boy to raise him in his household with the main purpose of teaching him his craft, so that the latter could support his adoptive father when he became too old and frail to continue working, carry on his trade, and succeed him in his business. If the adoptive father successfully taught the craft as he had promised in the contract, natural parents could not reclaim the child; nor was the child allowed to leave his adoptive father for his natural family. However, if the craftsman failed to teach his craft or the adoptive child failed to learn it, then the child had to return to his family of origin.134 Not only childless people resorted to adoption in the ancient Near East: it was not uncommon for the adopter to have his own sons and daughters and nonetheless adopt a son or daughter who inherited together with his biological children. In an adoption contract from Old Babylonian Nippur a man who already had a son by his wife adopted another son, probably because he was in financial need and the adoptive son promised to clear his adoptive father’s debt.135 The adoptive father took this as the occasion to revise his testament and make a new division of inheritance shares. The adoptee and the natural son inherited equally and obliged themselves to provide their father with maintenance in old age. Should one of them fail to do so, he would lose his heirship. As well as just the one son being adopted, we even have cases where a man adopted three brothers to whom he promised to leave his property in return for maintenance in old age. We also find that adult males were adopted under the condition that they married a natural daughter of the adoptee, i.e., an adoption as a son-in-law. Females were also adopted. Fathers, and if they were no longer alive, their widows or the girl’s siblings, gave them up 133

134 135

Driver and Miles 1952: 388; Stone and Owen 1991: 42; Greengus 1995: 469, 479; Stol 1995: 485, 491–2; Stol and Wiggermann 2000: 125–8. Code of Hammurabi §§188–9; cf. Driver and Miles 1952: 390–4; Paulissian 1999: 13–14. For references to this and similar cases where the adoptee cleared his adoptive father’s debt, see Stone 1991: 10.

186

Childless old age: the worst of all fates?

for adoption. Girls were adopted as daughters, for support in old age, as a means of alliance with another family, or as future daughters-in-law.136 Women could apparently adopt in conjunction with their husbands.137 In classical Greece, however, women could not adopt,138 and Roman law also forbade it.139 But in Hellenistic and Roman Egypt the status of women was significantly higher. A woman was allowed to give her daughter in marriage jointly with her husband or was free even to do it herself if she was widowed or divorced. Contrary to Roman law, females in Roman Egypt seem to have adopted children (alone and jointly with their husbands) regardless of whether this was legally recognized. In one case an adopted woman from second-century ce Oxyrhynchus calls the wife of her adoptive father θέσει μήτηρ.140 Another woman adopted the two sons of her husband and made them heirs of her property.141 Moreover, a woman was also in the position to give a child for adoption, which is unheard of in classical Greek or Roman law. In all three surviving adoption contracts it is a woman who is giving a child for adoption – in two cases alone,142 in one case jointly with her husband.143 Heracles and his wife Isarion from early fourth-century Oxyrhynchus jointly declared that they gave their son up for adoption,144 and in another case from early sixth-century Oxyrhynchus the widowed mother Aurelia Herais surrendered her 9-year-old daughter, who was then jointly adopted by a married couple.145 The normalization of adopted children regarding nomenclature within the family extended not only to father and mother but adoptive children even referred to the natural children of their new parents as “brother” and “sister.”146 Adoption in Roman Egypt was clearly based on Volksrecht and indigenous custom rather than on Roman law, and deviated in many respects from classical Athenian adoption practice. Rather we find many similarities with adoptions known from ancient Mesopotamia and pharaonic times of Egypt when minors were adopted and women were involved in adopting a child or ceding it to the adopting party. For later periods of Egypt under Arab rule we do not find any other further examples, probably because adoption of children is prohibited in Islamic law. There were surely

136 138 139 140 141 142 144

Paulissian 1999: 15–21. 137 Stone and Owen 1991: 41–8; Paulissian 1999: 11, 26, 33–4. Isaeus 7; cf. Harrison 1968: 94. Gaius, Inst. 1, 97, 104; CIust 7.33.8 from 294 ce. Cf. Krause 1994–5: iii.83. P.Oxy. 3.504 from the early second century. P.Oxy. 3.583 descr. from 119/20 ce ( P. Sijpesteijn 43). P.Lips. 1.28 from 381 ce; P.Oxy. 16.1895 from 554 ce. 143 P.Oxy. 9.1206 from 335 ce. P.Oxy. 9.1206. 145 P.Oxy. 16.1895. 146 SB 7871: ἀδελφὸν οῦ τῆ φύσει. Cf. Kurylowicz 1983: 63.

Uxorilocal marriages and adoption of son-in-law

187

various substitutes for formal adoption, but unfortunately the topic has not been the subject of any scholarly research so far.147 An adopted child provided a childless couple not only with an heir but also with social progeny and a relationship of reciprocity that could be useful at a time of crisis. In exchange for inheritance of property, care in old age was certainly expected by the adopting party,148 even if it is never mentioned explicitly in the contracts. However, this strategy of rewarding care in old age with succession to one’s estate could be used only if the adopter owned something to invest, land or a house. Those childless elderly with few material possessions were definitely disadvantaged because they had not much to offer. Furthermore, one of the major sources of conflicts could arise from the challenge of loyalty between adoptive and biological parents when they were still alive and lived in the same community. The adopted child might have felt the obligation to look after both his biological and adoptive parents, and dual allegiances might have existed that could result in conflicting interests. It is thus no surprise that the adoption contracts from Roman Egypt stress that the biological parents who had given away their child for adoption could not reclaim him or her later in life but had once and for all given it over to the adopting party.149

7.11 Uxorilocal marriages and the adoption of the son-in-law Parents without sons but who had daughters were in a situation in their old age almost as precarious as those without any children at all. We find examples in many societies worldwide of an elderly couple who had only daughters, or whose sons were sick or still too young to take over the headship, trying to find a husband for their daughter who would move into their home and take over the role as a son and son-in-law. Owing to the severe mortality regime up to 30 percent of fathers turned to negotiating an uxorilocal marriage for their daughters because they did not have a son.150 Instead of watching the daughter move out upon marriage, the parents won a young man for the household and could also keep their daughter at home. 147

148

149 150

One obvious way this was done was in the incorporation of mamluk slaves into a household, through marriage (Kenneth Cuno, by email correspondence). Cf. also Sonbol 1995 on adoption in Islamic society; Bargach 2002 on secret adoption in Morocco. Adoption contracts from the ancient Near East included clauses stipulating that the adoptee was obliged to care for the adopting couple or individual (Stol 1998: 61–4; Veenhof 1998: 123–7). P.Oxy. 9.1206; P.Oxy. 16.1895. The in-marrying son-in-law is a “recurrent feature of the major agricultural societies of Europe and Asia where it is usually associated with a daughter acting as heiress in the absence of sons” (Goody 1990: 4).

188

Childless old age: the worst of all fates?

Sometimes parents even formally adopted their daughter’s husband as son and heir. The couple’s children continued then the agnatic lineage of the household. We find the practice of adoption of the son-in-law in the absence of a biological male heir in the ancient Near East,151 classical Greece,152 Roman Asia Minor,153 Roman Syria,154 Byzantium,155 the medieval West,156 and, for the modern period up to the early twentieth century, France,157 England,158 Russia,159 China,160 and Japan,161 among others. In many other societies the son-in-law took over the role as only son without going through the process of formal adoption. If, for instance, a man in ancient Nippur had no sons but just daughters, he adopted an adult male and married him to one of his daughters. Through the adoption of a son-in-law he won a man for the household and by not only adopting an heir but also marrying him within the family he made sure that his property would pass to his own daughter’s children, and not to the children of his adoptive son by another woman. If the adoptive son later decided to leave his wife and marry another one, he lost all inheritance rights and had to leave his adoptive father’s household empty-handed. Surprisingly, we even have several instances recorded in the sources in which a man adopted a son (who was supposed to marry a daughter) even though he had sons of his own.162 Such adoptions circled around an economic purpose: the adopted sons-in-law usually came from propertied families and had already inherited from their own father. Their adoption brought their property into their wives’ families while they inherited more upon their adoptive fathers’ death, thus concentrating the property in one family. In classical Athens the adoption of the son-in-law was a strategy that ensured that the adoptee became adoptive son and son-in-law in the same 151

The so-called errebu marriage, the adoption of a son-in-law by a man who did not have a son, in Gen. 31; cf. Gadd 1926, no. 51. Cf. Burrows 1940: 3; Gordon 1940: 5–6. Isaeus 3.50, 3.68, 10.13; Dem. 41. Cf. MacDowell 1978: 100; Lindsay 1998–9: 91–2. 153 SEG 6.137, 6.139, 30.596; MAMA 1.232; TAM 2.1–3, 148; RECAM 2.303. Cf. Huebner 2007. 154 Apost. Const. 4.1 ( Syr. Didasc. 17). 155 Patlagean 1996: 476–80. 156 Cohn 1908: 102; Vismara 1958: 582–3; Macrides 1990: 109–18; 1999: 309. 157 Fauve-Chamoux 1988. 158 Wachter and Laslett noted the same custom, termed by them “patriline repair,” among English elites in the early modern period (1978: 116). 159 Czap 1982: 24: “An individual, kin or otherwise, was adopted by the household – a process described in the listings as priniatii vo dvor – ‘taken into the household’. Usually male, he could already be a married man with a family, or a youth destined to marry a daughter of the head, stay on as a ‘son-inlaw’, and eventually succeed to headship.” Cf. Kertzer and Barbagli 2001: i.52. 160 Goody 1976: 76; Wolf and Huang 1980; Pasternak 1983. For Sung China, see Ebrey 1991: 106. A nice case cited in Ebrey 1993: 237 is that of a Chinese man with a 3-year-old son who left 30 percent of the estate to his son and 70 percent to his uxorilocal son-in-law who managed and worked on the estate. 161 Dore 1999: 147; Dunn 1969: 71. 162 Stone 1991; Veenhof 1998; Paulissian 1999. 152

Uxorilocal marriages and adoption of son-in-law

189

place, keeping him physically in the home, and securing the continuation of the family name and lineage.163 In early modern Japan the adoption of the son-in-law (muko-yoshi) was the major form of adoption practiced.164 In such cases the groom received a dowry from his own family and became the heir of the bride’s household, where he moved after marriage. In sum, in the absence of a natural male heir adoption cum marriage served as a strategy to ensure labor, continuity, and succession. As we saw in previous chapters, in Roman Egypt the multiple family system was ideally virilocal: young brides went to live with their husbands’ families.165 An example for this pattern is the household of 75-year-old widower Peteamounis, an undertaker, from the village Moithymis in the Memphite nome, and his three sons, who also worked in this business.166 Each of the three sons was married and had brought his wife into his father’s home. The two eldest sons already had children. However, even in societies in which virilocality was the ideal, uxorilocal marriages took place quite frequently, since many families did not have a son who survived to adulthood. The proportion of uxorilocal marriages in these societies that are otherwise dominated by virilocal marriage patterns could reach in some regions and times nearly 30 percent.167 Against all expectations, uxorilocal marriages are, however, virtually absent in the Roman Egyptian census returns. However, I have argued at length elsewhere that we should read our evidence for brother–sister marriages in the Roman Egyptian census returns as the arrangement of an in-marrying son-in-law who was adopted by his father-in-law, especially in view of the fact that this custom of adoption and subsequent marriage of the adoptive child to a natural one is well documented for many historic societies around the Mediterranean and elsewhere, but is surprisingly “absent” from our Egyptian evidence.168 About 25 percent of all documented marriages in the Arsinoite nome were 163

164 165 167

168

An illustration of the custom outlined above is found in the personal history of the Athenian Polyeuctos, mentioned in Demosthenes’ speech Against Spoudias. Polyeuctos had only two daughters with his wife. So, he adopted his wife’s brother, Leocrates, and married him to one of his daughters. Leocrates thus became the adoptive son and son-in-law of his brother-in-law Polyeuctos, and the husband of his niece. After adoption and marriage, Leocrates moved into the household of his wife and her parents (Dem. 41; see Cox 1998: 35). An example from Homeric Greece for a husband who moved in with his bride’s parents is Bellerophon, son of Glaucus, from Epirus, who married one of the daughters of Iobate, king of Lycia. The groom who moved into his in-laws’ household did not receive the usual dowry, but was bequeathed half of the king’s landed property (Leduc 1994: 262). For the household registers of late nineteenth-century South-Tama, see Kurosu 1997. Bagnall and Frier 2006: 122. 166 173-Me-3. Cf. Goody 1990: 46, 106–7; Shimizu 1991; Bowen 1996; Dore 1999: 147; Hammel and Gullickson 2004: 151. Huebner 2007.

190

Childless old age: the worst of all fates?

brother–sister marriages,169 but we lack evidence for uxorilocal marriages that – judging from the comparative evidence – should have accounted for around 30 percent of all marriages. Thus, instead of reading these marriages as full-sibling marriages and assuming two peculiarities of Roman Egyptian society at the same time (the practice of brother–sister marriage and absence of uxorilocal marriage), we would have an uxorilocal marriage where the sonin-law moved in with his wife and her parents and was adopted by his fatherin-law to bind him closer to the family, and secure his inheritance rights to the household in which he invested his work and raised his children. This seems to me the more obvious explanation than declaring the society of Roman Egypt the only case in human history where full-sibling marriages were celebrated among the common people at large and on a regular basis. Apart from the lack of any record of adoptive status in the census documents, even though adoptions were undoubtedly practiced, the virtual absence of any uxorilocal marriage from these returns is another argument that speaks in favor of my interpretation. The presumable reason why the census placed no emphasis on recording adoptive status was that it had no relevance to any purpose for which the census existed. Adoptive children enjoyed the same rights and status as natural children; they were legally equal to natural children. As long as filiation by adoption had no legal or social status different from filiation by birth, there was no need to mention it in these official documents. However, my suggestion that apparent brother–sister marriages were in fact marriages between a natural and an adoptive child has met with vehement criticism.170 Critics concede that adoption cum marriage would be the most economical explanation for the peculiar Egyptian custom of brother– sister marriage as it would fall into the same pattern widely attested in the ancient Near East, pharaonic Egypt, classical Greece, and the Hellenistic world, whereby families aimed at binding an adoptive child closer to the family by marrying him to a natural one. They nonetheless adduce reasons why Roman Egypt should be considered the only exception, that is, the only society in world history where the incest taboo should not have existed and brother–sister marriages should have been practiced on a large scale among the common population. Arguments brought forward against my thesis, however, have been based on wrong assumptions; moreover, my critics have picked out just single aspects, but have not made the effort to 169

170

The exact numbers are 23.5% for the Arsinoite nome (n 102): 37% in the city of Arsinoe and 18.9% in the surrounding villages (Scheidel 1997). Remijsen and Clarysse 2008; Rowlandson and Takahashi 2009. Cf. now also Huebner in press a.

Uxorilocal marriages and adoption of son-in-law

191

deal with the central thesis. Remijsen and Clarysse (2008) argue that if those brothers who married their sisters were really adoptive and not natural brothers, these men should have been the only sons of their adoptive parents. And they point out that the census returns conflict with this hypothesis since many men who were married to their sisters had one or even more brothers living in the same household. However, as we have seen above, the presence of natural sons and even several of them was no bar to adoption in ancient Near Eastern societies or Hellenistic states. We have indeed evidence that individuals in ancient Mesopotamia and pharaonic Egypt even adopted two or more children simultaneously.171 Also the evidence for adoption of women from Hellenistic and Roman Rhodes shows us that the goal of adoption in the Hellenistic world was much more diverse than simply the continuation of an oikos.172 Furthermore, the adoption contracts that survive for Roman Egypt show that the welfare of the child stood at the forefront. Orphaned destitute children were given a new home and new parents who had the means to take care of them.173 The obligations of the adoptees such as care and maintenance in the adopter’s old age and the performance of funerary rites were not even discussed in the contracts. Adoptees were children and, by contrast to adoption practice in classical Athens or Rome, the parenting aspect seems to have been one of its main features. For this reason I see no compelling evidence why the possibility of adoption of several sons or the adoption of a son in the presence of natural sons should be entirely ruled out. If children were adopted under these circumstances, it nonetheless made sense to bind them later by marriage to the family, avoiding payment of a dowry and achieving consolidation of property. Remijsen and Clarysse argue that too many of the alleged adoptees in the census returns bore the same name as their adoptive paternal grandfather, an argument that in their opinion speaks against an adoptive relationship but for a natural one.174 However, considering the fact that a child was most likely adopted by a close relative, often by a paternal or maternal uncle (whether a childless uncle needed a son or a child had lost his parents), a preference widespread all over the ancient Mediterranean and basically in every society apart from the contemporary West.175 the same name of grandchild and grandfather is not very surprising if we assume that a child had been adopted by his late father’s brother. We have an example in the adoption contract from late fourth-century Hermopolis where the paternal 171 174

Huebner in press a. 172 Stavrianopoulou 1993. 173 Cf. section 3.4. Remijsen and Clarysse 2008, 59. 175 Cf. Huebner 2007.

192

Childless old age: the worst of all fates?

uncle adopts his late brother’s orphan boy.176 Thus the adoptee’s adoptive grandfather was simultaneously his natural grandfather. Likewise, the same name of Kronion and his firstborn son Kronion who married his “sister,” known through the extensive archive of this family from the village of Tebtunis, does not prove at all that they were natural father and sons.177 Firstborn sons were named after paternal grandfathers, while second sons were named after maternal grandfathers or fathers. Kronion the younger could very well have been Kronion’s adopted nephew, the son of a brother. Finally, Remijsen and Clarysse review once more references to brother– sister marriage in Egypt in classical Greek and Latin literature. However, in six references that they found I simply see the same stereotypes being repeated over and over again: in Athens it was permissible to marry one’s half-sister (no one has ever asked whether the authors are referring to contemporary Athens, or rather to classical Athens in the times of Solon and Pericles 600 years before their time – Philo of Alexandria is explicitly referring to Solon, the lawgiver of Athens),178 Persian magi married their mothers, and Egyptians married their sisters. Several of the classical authors who mention Egyptian brother–sister marriage allude to an indigenous Egyptian practice that existed before the arrival of the Ptolemies, who then followed their Egyptian subjects.179 On the other hand, however, all scholars engaged with the problem of apparent widespread brother–sister marriage in Roman Egyptian society agree that marriage between biological full siblings was not practiced in Greek, Roman, Macedonian, or indigenous Egyptian society on a regular and common basis, and that the practice only started sometime after the advent of Graeco-Roman rule,180 contradicting therefore the claim of the ancient authors of an indigenous Egyptian practice. Rowlandson and Takahashi (2009) agree with Remijsen and Clarysse (2008) that the papyrological evidence conflicts with my interpretation “because neither the family structures (notably the number of sons) nor the patterns of nomenclature in families where sibling marriage is recorded meet the expectations demanded by her hypothesis.”181 They seem to miss, however, the central point of my thesis when asserting that “only sibling marriage enabled a married daughter to remain in the parental home along

176 178 180

181

P.Lips. 1.28 from 381 ce. 177 Cf. Rowlandson and Takahashi 2009: 124–5. Philo, De spec. leg. 3.23. 179 Diod. 1.27; Philo, De spec. leg. 3.23; Pausanias 1.7.1. Černý 1954: 29; Thierfelder 1960: 7–9; Pestman 1961: 2–5; Hopkins 1980: 350–2; Shaw 1992: 274–5; Scheidel 1997a: 369. Rowlandson and Takahashi 2009: 105.

Uxorilocal marriages and adoption of son-in-law

193

with her brothers.” Quite the contrary, the in-marrying son-in-law who joined his bride in her parents’ home led to the same constellation and is by contrast to full brother–sister marriage a very common phenomenon in all patriarchal virilocal societies around the world. However, instead of just refuting my hypothesis, Rowlandson and Takahashi go one step further than Remijsen and Clarysse by trying to shed light on the reason behind the singular phenomenon of full brother–sister marriage in Roman Egypt. They argue that the strong legal and social position of women and their capacity to inherit and own property in their own right made the adoption of a son unnecessary in the presence of a daughter. Moreover, marriage of brother and sister would have thus avoided excess fragmentation of property. However, their explanation falls short. We know that in many societies and historical periods women could own property and were entitled to inherit which also led to fragmentation of property, just as in Roman Egypt. In virtually all patriarchal societies, a daughter required a dowry upon her marriage and then left her parents’ home, while a son stayed, brought a daughter-in-law in, and took care of his parents in old age. Nonetheless, we find no evidence that in any of these societies brother–sister marriage was even considered as an option to circumvent further division of family goods. Moreover, in Roman Egypt, where women had a stronger standing than in classical Athens, sons were still preferred over daughters. There cannot be any doubt that the proportion of female newborns that were exposed right after birth was considerably higher than that of male newborns.182 For Rowlandson and Takahashi, brother–sister marriage became “fashionable” in Roman times and was declared with “a positive exuberance and pride in their distinctively Egyptian heritage.”183 Unfortunately, however, we have no evidence supporting this assumption. They adduce the census returns as proof that the Egyptians flaunted “these relationships with an explicitness that goes beyond practical need. No anthropologist could ask for clearer responses from his informants.”184 However, easy as it is to adduce parallels from other societies for my explanation of these “brother–sister” marriages as the adoption of the son-in-law, it will be hard for them to find parallel cases where ordinary people made a conscious choice to follow a “fashion” when choosing a marriage partner for their sons and daughters “encouraged

182

183

Cf. Germain 1969: 177–97; 1975: 211–46; Harris 1980, 1994, 1999; Golden 1981; Pomeroy 1983: 207–22; 1986: 146–62; Garland 1985: 81; Patterson 1985; Boswell 1988; Motomura 1988; Huys 1989: 190–7; Brule 1992; Bagnall 1993: 202; 1997a; Scheidel 1997b, 2005, 2011: 287–310; Corbier 2001a: 52–73; Thompson 2002: 152–3; Bagnall and Frier 2006: 92–4, 151–3, 158–9; Evans Grubbs 2010; Harper 2011: ch. 10;. Rowlandson and Takahashi 2009: 129. 184 Rowlandson and Takahashi 2009: 129.

194

Childless old age: the worst of all fates?

by the historical connection of the city and its nome.”185 In Rowlandson and Takahashi’s opinion the Egyptian population “duly confirmed the expectations of their Roman rulers by registering their marriages to their full sisters.” In their marriage choice, so they believe, these peasants were even influenced and reinforced by the writings of Greek and Latin authors (starting with Herodotus, who was fascinated by Egyptian exotic cultural practice) as to what was natural for Egyptians and what constituted Egyptian identity.186 In my view, however, members of these social strata were concerned with holding their property together, strengthening their ties with kin near and far, retaining a sufficient labor pool for the farm, and securing care for their old age, but not with following a “fashion” set by some Ptolemaic rulers who lived several hundred years before their time, or Egyptian stereotyping of some foreign authors who had no relevance for their daily lives. Even now against the background of television and the Internet I wonder how much these small farmers in the Fayum know about what is written and thought about Egyptian society – all the more so in antiquity, when illiteracy was even more widespread than today, when one in every four Egyptians is illiterate.187 By contrast to antiquity, school enrolment in modern Egypt is free and mandatory; nonetheless parents in rural communities often remove their children from school to work at home or in the fields.188 I further doubt that their ancient counterparts cared to “fulfill their destiny as Egyptians,”189 especially as the evidence clearly shows that “brother–sister” marriages were practiced by Greek and Egyptian families alike.190 The Gnomon of the Idiologos, a handbook for the Roman chief financial administrator of Egypt dating to Augustan times but surviving in a copy from second-century Theadelphia in the Arsinoite nome, stressed that Roman citizens living in Egypt were not allowed to marry their sister or their aunts.191 What should we conclude from this passage that was simply reproducing Roman legislation on prohibited degrees of consanguinity and affinity?192 185 187 188 189 190

191 192

Rowlandson and Takahashi 2009: 116. 186 Rowlandson and Takahashi 2009: 129. Baines 1983; Harris 1989: 29, 276–80; Bagnall 1993: 240–51; Cribiore 1996, 2001. Oxford Business Group, The Report: Egypt 2011. London, Istanbul and Dubai, 2011: 183. Rowlandson and Takahashi 2009: 130. Bagnall and Frier 2006: 129. Note that Shaw (1992: 292) argues for exactly the opposite, that only the privileged ethnic-political class of Greeks in Egypt, overly concerned with blood ancestry to maintain their critical ethnic-political distinctions, practiced brother–sister marriage. BGU 5.1210, no. 23, after 149 ce. CIust 5.4.17: “Nemini liceat contrahere matrimonium cum filia nepte pronepte, itemque matre avia proavia et ex latere amita ac matertera, sorore sororis filia et ex ea nepte, praeterea fratris filia et ex ea nepte, itemque ex adfinibus privigna noverca nuru socru ceterisque, quae iure antiquo prohibentur: a quibus cunctos volumus abstinere.”

Uxorilocal marriages and adoption of son-in-law

195

That Egyptians were marrying their sisters and their aunts? No one has ever contended that marriage between nephew and aunt (or niece and uncle) was routinely practiced in the Roman Egyptian population and there is no reason to assume it was at all common. But the idea of widespread marriage between full brother and sister in Roman Egyptian society, which – as we have seen – in every respect resembles other traditional patriarchal societies most closely, still has not been laid to rest. As shown above, the status of adoptive children in the adoption contracts from fourth-century Egypt was assimilated as closely as possible to that of natural children; the formal adoption of a child meant full legal absorption into the new family. After registration the adoptee became the legal child of the adopter – he took over his patronymic and became his heir as “if he had been engendered by him by his own blood” – and exactly this legal status is what the census and epikriseis returns record. In addition, adoptive relationships were not only assimilated to natural ones between adoptee and adoptive father but also between him and his adoptive mother and adoptive siblings. We have adduced ample proof above to show that adoptions did occur in Roman Egypt but were not declared in the census returns.193 The census, however, was only interested in official legal relationships between family members, not in recording the origins of family composition. It was presumably the son-in-law, not necessarily the bride’s father, who had the greatest interest in formal adoption. Without a formal adoption, should his marriage end in divorce or should his wife die before him, he could be ejected without any property from his parents-in-law, even if he had contributed his labor to the household for many years.194 From firstcentury Oxyrhynchus we hear about a case where a man, Sarapion, had married into the household of his wife, Syra, without being adopted by her parents.195 Unfortunately, the marriage ended in court. Syra alleged that she had received Sarapion into her parents’ home since he lacked all means. During their marriage, however, Sarapion squandered her dowry, so Syra claims. In addition, she accused Sarapion of ill-treating, insulting, and beating her. In the end, according to Syra, he left her without any means, even though we know that she still owned at least the house they had been living in. Syra litigated against her ex-husband in order to make him pay back the dowry. In case she was successful, it would have been Sarapion, not Syra, who then left the marriage penniless, even though he had probably

193 195

Cf. also Huebner 2007, in press a. P.Oxy. 2.281 from 20–50 ce.

194

Cf. Farnsworth 1986: 57.

196

Childless old age: the worst of all fates?

invested several years of work in the household of Syra’s parents. We could learn a lot about this case, and about uxorilocal marriages in general, if we had also Sarapion’s side of the story. From the fourth-century Arsinoite nome comes the case of Aron and his son-in-law, Asep.196 Aron married his daughter Apia to Asep197 and accepted the couple into his household since (again the same rationale is given as in the case of Syra’s husband) his son-in-law was destitute.198 In a contract drawn up between him and his son-in-law, Aron laid down the terms and conditions of this union. Asep had to promise never to leave Aron’s daughter and, more interestingly, he also had to promise never to leave his father-inlaw!199 Should Asep transgress this agreement, he had to pay the substantial fine of 10 solidi. This arrangement reminds us of the paramonē contracts discussed above. Also here Asep contracted his labor for several years, or better, as long as his father-in-law lived, in exchange for a place to live and the prospect of becoming head of household in his own right upon his father-inlaw’s death. We have many parallels from other patriarchal societies that a position such as the in-marrying son-in-law, one that was associated with a certain stigma, was only attractive for a son from a poor family or for a younger son who had no prospects of becoming head in his own household.200 In China and India a man moving in with his in-laws was not in a very honorable position. He was called a “wild duck” (as opposed to a homegrown duck).201 In north India he was called a “house-bridegroom.” In Portugal a man who moved in with his parents-in-law was called a “woman” because he accepted a weak, subordinate position in a household that was not his own.202 Sarapion and Asep are similarly said to be destitute and without means, and seem to have held a rather weak standing in their wives’ homes. In sum, for elderly fathers without biological sons, adopting their daughter’s husband as an old-age strategy made sense and is, as I believe, amply, albeit indirectly, recorded in the Roman Egyptian census returns.

196 197

198 199

200

P.Ross.Georg. 3.28 from 343 ce or 358 ce. Cf. Montevecchi 1936: 18; Evans Grubbs 2002: 129–30. ll. 6–7: ἐπὶ τῷ [τ]ὸν Αὐρήλιον Ἀσ̣ὴ̣ π̣ συνέρχεσθαι τῇ τοῦ Ἄρωνος θυγατρὶ Ἀπίᾳ πρὸς γάμου κοινωνίαν. ll. 7–8: Ἄρων πεισμοσύνῃ ὑπεδέξατο τὸν Ἀσὴπ γυμνιεύοντα. ll. 13–19: οὐκ οὔσης δὲ ἐξουσίας [τῷ Ἀσὴ]π̣ καταλεῖψαι τὴν Σοφίαν μετὰ ταῦτα, ἐπιδώσει ὁ παραβὰς μετὰ ταῦτα τῷ ἐμμένοντι ὑπὲρ ἐ[πι]τίμου καὶ στροφῆς καὶ ἐπηρείας χρυσίου νομίσματα δέκα, οὐ μὴν οὐ καταλεῖψαι τὸν ταύτης πατέρα Ἄρωνα δ̣ιὰ τὸ οὕτως συντεθεῖσθαι κ̣ α̣̣ ὶ συμπεπεῖσθαι, καὶ ἐπερωτηθέντες ὡμολόγησαν. Cf. Huebner 2007: 47. 201 Xie 1997. 202 Poska 2000: 318.

Fostering

197

7.12 Fostering Fostering was a less formal way of substituting for a natural child than adoption, an option that was also open to women who could not legally adopt a child on their own according to Attic and Roman law.203 The decisive difference between fosterage and adoption was that fosterage gave the orphaned or abandoned child a new home but did not involve any change in filiation or status.204 We have evidence from all over the ancient Eastern Mediterranean that close kin were preferred candidates for adoption, often a son of a brother or sister. However, foundlings of unknown parentage were rather raised as foster children (Greek threptoi, Latin alumni) but not formally adopted. Interestingly, through raising a child and providing for him or her, the foster parents acquired a right over the child. The natural parents were able to regain their child only when they repaid the expenses the foster parents had incurred on the child over the years.205 That means that saving an exposed child from death did not automatically entail a right over the infant and his or her services in later life; however, a proper upbringing of the boy or girl did.206 From a will from the year 124 ce, we learn of a certain Tatastron, an old woman who owned considerable property but apparently did not have a husband or children.207 She decided to bequeath her entire estate to the beloved son of her freedman, Dionysios. The boy was still a minor when Tatastron drew up her will and was apparently not required to care for her (perhaps the boy’s parents were, but this is not explicitly stated in the will). His appointment by will was not made on any condition, and the only reason Tatastron gives for her decision to disinherit her more distant relatives and to make her freedman’s son her only heir is “his affection towards her.” Only if Dionysios himself should die childless one day the property of Tatastron was to go back to her own family line. Candidates for fostering were exposed infants that the fosterer picked up at a dunghill and or other much-frequented places, or the child of a slave or freedman of the household.208 An apparently single woman from Oxyrhynchus had taken in a slave-girl and raised her as her own daughter. She trained her as a musician “in the hope that when she came of age I would 203 204

205

206 207

Nielsen 1987: 141–88. For fosterage in the Greek East, see Cameron 1939; Corbett 1983; Bremmer 2000. For fosterage in the Roman West, see Rawson 1986; Nielsen 1987, 1999; Corbier 1999a. Seneca, Contr. 9.3; Quint. Inst. Orat. 7.1.14; Ps.-Quint. Declam. Min. 278; Dig. 45.1.107 (Iavolen); CIust 5.4.16; P.Oxy. 16.1895 from 554. For threptoi see the classic article by Cameron 1939. Cf. Boswell 1988: 63–5. P.Oxy. 3.490 from 124 ce. 208 Bieżuńska-Małowist 1971; Pomeroy 1986.

198

Childless old age: the worst of all fates?

have her to provide for me in my old age.”209 Non-married celibate women in later Christian times pursued similar strategies. Deaconesses and nuns took in orphans and raised them in place of biological children. By offering these orphans a home and education, they employed an alternative strategy to having their own children in order to be provided for in old age.210 On a late antique ostracon from the Theban area in southern Egypt we read that the female monk Maria had taken in a young boy whose mother had died.

7.13 Charity The emphasis the Christian Church placed on the care of the elderly as a distinctive group from the fourth century onward reveals the miserable situation of old people without sufficient family-based support. Beginning in the fourth century, the Church established almshouses for the aged, which provided accommodation, catering, and care for the childless single elderly who had no one else to take care of them.211 These institutions were financed by the town or local church. In comparison to adoption, reliance on charity was less proactive than defensive. Was charity seen as one of many strategies to find support in old age, or rather as an option of last resort? There can be little doubt that spending one’s old age within one’s own family, cared for by one’s own children and grandchildren, must have been a much more desirable prospect and more dignified option. The reliance on charity probably came with destitution, loss of social status, and social exclusion. It was not only a sign of poverty to be forced to rely on charity, but also a sign of a lack of social ties and networks. Charity certainly did not lift the recipient out of poverty and a socially disadvantaged position. Monasteries and nunneries also offered old people a refuge. The withdrawal to a monastery surely was seen as a more respectable option than continuing to live in society below the necessary income level, without children or other social ties. However, monasteries did not accept everyone. Over the course of the fourth and fifth centuries, retirement to a convent became more and more an option for the upper strata of society, because convents expected gifts and part of the inheritance. Withdrawal from the world in favor of a life of piety and prayer thus became reserved for wealthy widows and widowers.212 209 210 211

212

P.Oxy. 50.3555 from the first or second century ce. O.Brit Mus.Copt.Ad. 23; cf. Wilfong 2002: 107–8. e.g., SB 4845 from 323–642 ce: γηροκομεῖον. This instance cannot be seen, however, as evidence for the ecclesiastical institution. Cf. Taubenschlag 1956: 179; Rupprecht 1998: 225. See in general for ecclesiastical old age homes: Dagron 1974: 512; Finn 2006: 106. Constantelos 1968: 223–4; Lascaratos et al. 2004.

chapter 8

Conclusions

Modernization, urbanization, migration, the demographic transition, new lifestyle aspirations, and the spread of values which emphasized the individual rather than the collective familial interests have been accused of eroding traditional filial obligations. Further factors that have challenged traditional intergenerational support are changing definitions of family, increasing numbers of single-mother households, high divorce and remarriage rates, the establishment of blended families, and dramatic changes in the age structure and intergenerational composition of the family, which all affect solidarity and reciprocity on the household level. These factors test the ability of the family to continue to provide care for its members. This study pursued the question of intergenerational solidarity for antiquity, or to be more precise, for members of the lower and middle classes documented in our documentary sources from Roman Egypt, and revolved around all those factors that have been shown to have a direct or indirect effect on generational relations and support networks within the family and the wider kin group: the social, legal, and moral context governing the relations between the generations, motivations for having children and attitudes toward them, the demographic regime, household size and composition, marriage patterns, rates of remarriage in case of divorce or widowhood, residence patterns of the elderly, the availability and reliability of children for old-age support, and finally alternatives for care in old age in the absence of one’s own children. Regarding household size and composition, various ecological, socioeconomic, and demographic conditions and factors, such as urban or rural environment, agricultural forms, inheritance patterns, and mortality and fertility levels influenced the dominance of specific household patterns in different regions of Roman Egypt. The timing of certain family strategies, such as decisions about a suitable spouse, the place of residence of the young couple, the purchase of more land or the sale or lease of existing land, the exposure of a newborn child, or the adoption of another one was the 199

200

Conclusions

response to the opportunities or needs that arose from various household constellations. The stage of the household’s life cycle also affected the way an individual or family coped with situations of crisis, the death of a spouse, divorce, orphanhood, or old age. Household composition and family strategies were thus closely interdependent. On the other hand, certain cultural patterns of predominant living forms guided decision-making to achieve the household form that was considered the ideal for traditional reasons. While most of our evidence originates in Middle Egypt, regional differences might not make it representative for the whole of Roman Egypt. Differences in population density and the fertility of the land, accompanied by differences in agricultural patterns and urbanization, might have led to quite different household composition patterns between different regions of Egypt, such as the Delta, Middle and Upper Egypt, the Fayum, or the Dakleh Oasis, as the separate census register from one city (probably Lykopolis1) in Upper Egypt already suggests. Roman Egypt’s demographic regime, which was prevalent in many societies before the demographic transition, was characterized by high mortality, high fertility, and low population growth. People died mostly from epidemic infectious diseases, by contrast to modern patterns of death that result mainly from chronic degenerative diseases at the end of the natural lifespan. In particular infant mortality was high in Roman Egypt: almost half of all children did not live to celebrate their fifth birthday. Given the high mortality rates, the overlap between generations was small compared to modern societies. Most fathers did not live long enough to see their sons marry and have children, and two-fifths of all fathers did not have a son when they were approaching old age. In addition, many women found themselves widowed with children while still in their twenties or thirties. Likewise, it was not a rare event for a man to lose his wife in childbed or to an epidemic, and to marry again and start a new family with his second wife. Moreover, divorce was easy to obtain for either party and does not seem to have been uncommon, despite its adverse consequences for family ties, the fate of children, the economic situation of the family, and, especially for women, their social status and support in later years. In general, the result was a great variety of living constellations – many households headed by single mothers, multiple consecutive marriages, and patchwork families with children of several relationships – a diversity of family forms that surely had a destabilizing effect on a patriarchally organized society. 1

See Montevecchi 2000.

Conclusions

201

Household size in Roman Egypt was surprisingly small, according to the census returns, but in line with other societies where mortality and birth rates are high. Average household size was 5.3 members for metropolitan households, where family more often owned slaves, and 4.8 members for households in the countryside. Small household size is confirmed by archeological excavations in Roman Egypt. The modest houses were predominantly built of inexpensive mud brick. Wood was used for frames and doors. The most common type had two stories, access to a roof terrace, and a courtyard to the back as the center of activity where food preparation took place and domestic animals were kept. Prices for houses ranged considerably depending on their size, quality, and location. Houses in the metropoleis were considerably more expensive than in the villages. Few people lived as solitaries: these accounted for 16.2 percent of all households but only 6 percent of all recorded individuals. Most were already of advanced age, probably the surviving members of larger family complexes. The most common household type was the simple family household, which accounted for 43.1 percent of all households; 15 percent were extended households, usually a couple with their children and an elderly parent of the husband; and 21 percent of all households belonged to the multiple family type. However, as multiple family households regularly contained more members than simple ones, most individuals, 42.8 percent, lived in multiple family households. These different forms of family composition were not alternatives but each a stage in an ideal model of household progression. They all might have been experienced by a single family over the course of several decades. The fact that the multiple family household was so relatively common, even though demographic circumstances worked against achieving it, suggests that it constituted the desired form of residence, at least among the rural population. For considerable differences existed between urban and rural areas regarding predominant household forms. Living in an urban environment seems to have led to an increase in nuclear families, while in rural environments multiple family households were more common, a pattern that we find in many traditional patriarchal societies. Economic contingencies related to the profession of the head of household and higher costs of keeping up a large household in the cities apparently led to these differences. While families of artisans and shopkeepers were more likely to live in simple family households, the marital union building the basis of the family’s economic activity, peasant farmers lived in larger households in order to maintain a sufficient domestic workforce and to avoid excessive fragmentation of the paternal land.

202

Conclusions

Co-residence was restricted to patrilineal kinship and females marrying into the family. All the male members of the household were related to each other. In the Roman Egyptian family system, men had a lifelong membership in the family into which they were born, whereas females were only temporary residents of their natal family and joined their husbands’ families at marriage, when they then took up disadvantaged membership. As in such a system it did not require any savings on the side of the young couple or the transfer of the estate from father to son, marriage must have been possible at any point whether or not the groom was already financially independent. While in societies with simple family household patterns marriage is usually late, since the couple can establish an independent household and start having a family only after acquiring the necessary means to do so, in Roman Egypt, where the multiple family household was common, marriage could take place earlier. For this reason also more three-generational households were possible than in societies such as classical Athens, where sons had to wait until their fathers around their sixtieth birthday transferred the headship and patrimony to them. In general, the observed pattern of multiple family households being widespread in Roman Egypt goes together with the similarly noted early age at first marriage and nearly universal marriage for both sexes. Some men in Roman Egypt must have married into their brides’ families, i.e., made an uxorilocal marriage, but the way the census returns were drawn up apparently obscures this. The few recorded instances show that especially men from an impoverished family background considered this option. The succession and inheritance pattern prevalent in Roman Egypt followed indigenous Egyptian rather than Greek customs, and also the relatively high proportion of multiple family households in Roman Egypt resembles Egyptian rather than Greek household structures of earlier Ptolemaic times. Roman Egypt shows here an ideal of family formation that we also find in many early modern Mediterranean and Asian societies, for which the proportion of multiple family households has been reported at between 20 and 40 percent. Early modern societies of northwestern Europe, such as England, Germany, and northern France, on the other hand, show a level of multiple family households of between 0 and 4 percent, lending weight to the assumption that we have here over the longue durée a completely different system of family and household formation at play. Marriage had hardly any effect on the household as an economic entity in Roman Egypt, as the marriage of sons did not typically establish a new household but the new couple was rather integrated into a pre-existing

Conclusions

203

family structure and the ongoing domestic economy. It was thus not marriage that brought about the reorganization of the household in Roman Egypt, but rather the death of the patriarch or even later that of his wife. However, even if the multiple family household was considered the ideal, families did not stay together at all costs. Despite high mortality rates, a proportion of 21 percent of multiple family households ranks rather at the lower end of what would have been demographically possible if more distant relatives had been willing to stay together under one roof. From our evidence we can conclude that families in Roman Egypt preferred to discontinue keeping up a joint household at the latest when the multiple family household consisted only of the diagonal stage of married uncles and nephews. While the most common type of multiple family household was the vertical one (a couple with their married sons) and the second most common type the horizontal one (married brothers living together), the diagonal multiple family household consisting of cousins or uncles, or nephews, forming one household with their respective families was exceptional. As a rule, the oldest male of the household served as head. Occasionally, even if adult males were present, the oldest female held this position, usually the widow of the former head of household. Very rarely do we find the headship of the household transferred to the younger generation during the patriarch’s lifetime, a model quite common in classical Athens and early modern north-western Europe. Children shared in the estate of both parents, and in theory all sons and daughters received a portion. The eldest son often received a larger share and it was on him that the role as “father” devolved, a title that we find for the head of the family in private letters. Daughters received parts of their inheritance in the form of dowries. In the absence of a son, daughters took all the estate, and some women were substantial property owners and could dispose of it as they wished. Roman Egyptian inheritance patterns thus followed the indigenous Egyptian custom of partible inheritance recorded for pharaonic times whereby all children, sons as well as daughters, inherited equally from their parents, the eldest son assuming the role as head of the family and managing the landed property jointly for the household of brothers. Wives maintained control over the property inherited from parents even during their marriage. As we would expect in a patriarchal society, women often sought the assistance of a male relative in property transfers or legal petitions to enforce their claims, whether a brother, husband, son, or son-in-law, without his necessarily having the role of official guardian, but nevertheless acted autonomously.

204

Conclusions

In contemporary Western society parents are socially, morally, and legally expected to act as providers of economic support for their children. Parents raise their children, invest in their education, and transfer to them their property. Even in old age they not infrequently continue to contribute to their children’s welfare, and assist with childcare and monetary contributions. For Roman Egypt we also find an overwhelming consensus in the sources that parents were expected to support their young children. Fathers, in particular, were held responsible for materially maintaining them, for enabling them to learn a trade, for using their social influence to their advantage, for introducing them into society, and for supporting them in difficult situations. While we have quite specific instructions for adoptive parents or guardians about what constituted adequate support of a child, such as maintaining them with food and clothing according to the means available, and eventually transferring to them their property, more detailed regulations about what exactly natural parents were required to provide are lacking in our sources. We do not know what constituted a minimum subsistence level, up to what age a father was required to maintain his son, what was considered an adequate education, or what level of dowry a daughter could expect. While a wife secured adequate provision and other details of daily life in a marriage contract, the parents themselves had to decide what they wanted and were able to offer their children. Their support, however, was governed by social expectations. Classic demographic theories of fertility, such as the “intergenerational wealth flows” theory proposed by Caldwell, maintain that in preindustrial societies high fertility was economically advantageous for poor families, because raising children cost little while they in return could contribute to the family economy from an early age and thereby support their aging parents. According to this theory, net wealth flows were primarily upward from the younger to the older generations. However, our evidence from Roman Egypt gives strong indications that children were rather a net loss to their parents and, considering the entire life course, wealth flows were downward from parents to children. Even if sons in peasant families started to work full-time from the age of 14, it took them many years to repay their parents the costs of raising them. Moreover, infant mortality was very high, so that parents invested in vain in many children only to see them die before they reached adolescence. Sons who lived through adulthood thus had to make up for not only their sisters, who moved out early and in addition needed a dowry, but also their brothers who had died previously. Even the argument that sons provided labor to work the farm became obsolete: hiring wage laborers who earned little and could be employed when needed was

Conclusions

205

more economical and less risky. Having children was even less financially beneficial for older parents than for younger ones, as the older might not live long enough to reap the economic benefits. From a purely economic point of view, it was thus not advisable to have as many children as possible. However, children did seem to have added considerably to their parents’ perceived oldage security – and also to their social and emotional well-being. Moving on in the household life cycle, when children grew up and parents grew older, family expectations, religious norms, and community values also protected old parents from being abandoned, neglected, or mistreated by their children. A failure to reciprocate the love and care one had received in earlier years or even to abandon, neglect, or mistreat one’s elderly parents was met by strong social opprobrium. Proper filial behavior was understood as kind treatment and economic support of one’s elderly parents, father and mother, as long as they were alive, and provision of a proper burial and observation of the customary grave rites after their death. The public demonstration of such filial loyalty enhanced one’s social reputation, fostered one’s career, and secured one’s rightful inheritance. In practice, however, this norm was subject to negotiations which were shaped by several factors: the number of children, their availability, their economic capability, and the strength of emotional ties. We see children opting out of the inheritance in favor of siblings and thereby also relinquishing all obligations to provide old-age care. In other cases parents assigned a larger share to one of their children because he had shown himself especially kind and caring toward his elderly parents. Furthermore, failure from one side to support the other had the consequence that the other was no longer obliged to fulfill his role. If, for instance, parents did not provide their son with an education, the latter was no longer obliged to care for his parents in their old age – a norm documented at least for the early Greek period of Egypt, but not explicitly attested for later periods of Egypt. Parents were in any case not obliged to bequeath their property to their children when the latter failed to support them or otherwise severely mistreated them. Individuals were thus governed by varying motives for providing support. Affection, obligation, the feeling of reciprocity, the threat of social disapproval, the prospect of an inheritance, and, most often, a complex combination of all of these bound different generations of a household together and led them to provide support to its weaker members. Different types of intergenerational solidarity and support, such as structural-behavioral and cognitive-affective solidarity, as already set out in the introduction, can be identified for Roman Egyptian society. In practice, they are generally entwined and interdependent. In general,

206

Conclusions

cultural, social, and religious norms and values all over the ancient world expected mutual solidarity between generations, especially on the family level. Parents in Roman Egypt were expected to feed and clothe their children, watch over their moral development, celebrate their rites of passage, provide their sons with an education and their daughters with a dowry, and find suitable spouses for them. Co-residence in a household of three or more generations as the highest level of structural solidarity seems to have been considered the ideal by the population under study, and mutual respect, practical assistance, and support were expected. Associational solidarity, i.e., the frequency of contact between elderly parents and their children, was thus high, and children away from home sent letters and promised to visit soon. From their letters we get a glimpse of affective solidarity such as emotional assistance and the degree and reciprocity of positive sentiments among family members. Children ideally brought their old parents comfort and showed affection and deference. If they did not live in the same household, children supported their aged parents with allowances in cash and in kind. Having and raising children seems to have served as a lifelong strategy aimed at securing late-life emotional, physical, and material security. The potential for providing old-age security seems therefore to be a more likely hypothesis to explain high fertility rates than an expected economic net gain from the children’s early labor contribution to the household. There is no doubt that parents in Roman Egypt sometimes exposed, killed, sold into slavery, or prostituted their children and routinely expected them to contribute to the household income from a very early age, and that also children sometimes were negligent and did not pay back their frail elderly parents the support and care they had received during their youth. All in all, however, relationships between generations in Roman Egypt seem to have been dominated less by a patriarchal authority and subordination than by genuine affection and mutual devotion between elderly parents and adult children and young parents and their small children. As mortality remained high over the life course and women regularly married men older than them, many found themselves widowed after a few years of marriage. Also divorce seems not to have been unusual, with incompatibility of spouses rather than infertility apparently being the most common reason for it. This led to the fact that many women who were still in their childbearing years returned to the marriage market. However, these widowed or divorced women are not identifiable in the census returns as having been previously married. They appear as either unmarried in their fathers’ households, or as if in their first marriages with their new husbands. Remarriage of a

Conclusions

207

widow or divorcée is only recognizable in the census lists when children from a previous union lived with her, a constellation which was uncommon under the patrilocal household system prevalent in Roman Egypt. Children usually stayed with the father or the father’s family. Co-residence of a stepfather with children from his wife’s first union was the exception. In this study I have argued that a widowed woman’s choice whether or not to remarry was closely linked to the household’s composition at the time of her husband’s death. When a widow and her children were provided for by the deceased husband’s estate, or her own property, and this estate allowed her to live independently, the widowed mother usually refrained from remarrying. Women in Roman Egypt regularly did not inherit from their husbands, but merely received a usufruct and held in trust the patrimony until their children came of age. Remarriage forfeited these privileges. The papyri give vivid proof of how widows astutely managed their households and raised their children on their own, receiving revenues from the late husband’s patrimony managed by a guardian or the mother herself until the children reached maturity and could take over their paternal property. By contrast, if a woman had been living with her husband in his paternal multiple family household, upon his death her position among her husband’s agnates indubitably became very weak. Occupying one of the lowest positions in the household’s hierarchy without any of her own kinsmen, she was an easy victim for disrespect, mistreatment, and possibly even sexual abuse. In these household constellations, more common in the early stages of marriage, the young widow simply was given her dowry back and left. An analysis of household forms in the Roman Egyptian census returns suggests that, in fact, many more very young rather than older couples in rural Roman Egypt lived in such a multiple family household. A young woman was very unlikely to be assigned the management of the family estate should her husband die. This meant, however, that she also had to leave her children behind and return to her natal home. Many relatively young widows probably used the occasion of remarriage to regain economic and social security, if they did not want to or could not rely permanently on their natal family. Cross-cultural studies bring analogies from other patriarchal societies with similar household patterns: A widow who had been part of a multiple family household could not be assigned the management of her late husband’s estate because it was part of a multiple family household managed by the husband’s father, uncles, or brothers. This scenario must have been more common if a husband died in the early years of marriage, since at this stage it was less likely that the couple had already set up their own household.

208

Conclusions

All family systems traditionally face the problems of generational succession, economic support of the non-active elderly, and physical care of the frail elderly, but the structures that provided solutions to these problems were by no means uniform. We encounter many different living circumstances for the elderly in Roman Egypt. We see an elderly couple living in a multiple family household surrounded by their children and grandchildren, or an old man cared for by his young wife and minor children, or a widow or widower who lived with one of their married children, or even the solitary elderly who apparently did not have children or whose children had moved away. However, even if the household seems to have been the fundamental unit for organizing intergenerational support, co-residence was not necessarily an uncomplicated indicator or proxy for family support. Even if the census returns document an elderly individual living on his or her own, we do not know if children had set up independent households within the same compound so that they still had the ability to provide support, and interaction could remain constant and regular. Nevertheless, antiquity did not know the systematic separation of the elderly from adult society based exclusively on chronological age. This development only started with the enactment of mandatory retirement which reached its climax in the mid-twentieth century. Old men and women in Roman Egypt did not experience a sudden cutoff from manual labor on reaching a certain age; they continued to be active until prevented by physical incapability. This had the advantage that there was no sudden change in their daily activity at a defined point in life, regardless of their actual capability. Elderly individuals thus did not feel useless but were still able to contribute to the household’s income and felt socially involved. A son helped his old father working the land or practicing the business, and a daughter-in-law probably took over step-by-step the more strenuous of the domestic tasks from her mother-in-law who was growing old, but the old couple remained the heads of the family. While property transfer during the father’s lifetime to the younger generation was the rule in classical Athens, it was not common in Roman Egypt, a pattern that resembles that already recorded for earlier centuries of pharaonic Egypt. In Roman Egypt the old father usually controlled all the family property until his death and probably relinquished managerial authority only gradually to his sons. Cross-cultural studies have shown that in these cases relations between the generations in a household are more favorable and the elderly enjoy a better position than in a family system of pre-mortem property transfer, where the elderly couple, once deprived of their power, are mainly seen as a burden.

Conclusions

209

It was certainly easier for a father than for a mother to demand the fulfillment of children’s obligations because he was the owner of the family property as long as he lived. The patriarch’s wife enjoyed a comparably powerful position as long as her husband was alive; his death, however, often went hand in hand with an immediate loss of authority within the household, since ownership regularly did not go to the widow but to the couple’s children. Just in a few cases we see an elderly widow acting as head of household, even if she had adult sons. Women had full rights to hold property in Roman Egypt, and in theory a woman was protected against divorce or widowhood by her dowry, which was returned to her, but it was hardly sufficient to maintain her for the rest of her life if the woman did not inherit further property from her parents or was not assigned a usufruct of her late husband’s property. Thus, as we can see from private letters, many widows seem to have strived to develop close emotional ties with their sons to ensure support, proper treatment, and provision for old age. If the widow had not cared to establish a loving relationship of filial affection with her children earlier in life, it must have been very difficult for her to obtain the necessary assistance. Her own property could of course serve as a leverage to maintain some authority and to secure adequate support. Private letters complement here the information we gain from the census records. As with the census returns, we find the same two scenarios, the elderly widowed mother honored by her children as the core of the family, or neglected and mistreated with little support and low standing among the other members of the household. Very few actual contracts between parents and children that formally regulated old-age support have survived for Roman Egypt. These extant intergenerational contracts imply that relationships between parents and children were not always exclusively defined by filial love and selfless devotion. Sometimes it seems to have been necessary to lay down contractually the responsibilities, commitments, and liabilities of both parties, even if these contractual arrangements seem to have been rather the exception than the rule. They offered, however, the advantage of security for the elderly and justice for the child who acted as caregiver with respect to his siblings. These contracts that should ensure provision for old age were usually interwoven with property transmissions, whether it was before or after the parent’s death, partly or in total, in return for lifelong payments in cash or kind or both. Curiously, we do not find a single case in the papyri where children requested that the management of the property should go over to them because their elderly mother or father, owing to senility or insanity, had become incapable of supervising their own affairs.

210

Conclusions

While in modern nuclear families parents as they age live in an evershrinking household as their children grow up and move out, in the multiple family household of Roman Egypt all the couple’s sons usually stayed in the parental household even after marriage, repopulating the household with their spouses and children. Growing old thus did not entail isolation, but quite the contrary. The number of family members in the household ideally increased as the parents entered old age. As we have seen, both daughters and sons were involved in the support of their old parents, even if to different degrees and in different ways. Forms of support varied between financial, practical, and emotional support and depended on the respective living situation and the child’s economic and physical capability. Special attention has been paid here to the role of the daughter-in-law, who in patriarchal societies with virilocal marriage patterns often functioned as the primary caregiver for elderly household members. We have seen that in the Roman Egyptian patriarchal household pattern a young daughter-in-law had hardly any influence in her husband’s home, and ranked in one of the lowest positions in the household hierarchy. She was under the supervision of her mother-in-law and it was the latter’s task to familiarize the younger woman with her new duties and responsibilities. The young woman was expected to act demure and shy and to endure the hardships and constant criticism to work harder without complaint. Allusions to these issues are scattered in our evidence; just how common these strains between the generations of women in the households were is difficult to ascertain. Greek and Roman literary evidence confirms that the relationship between a mother- and daughter-in-law was not an easy one, and struggles between mothers- and daughters-in-law were common topoi in myth and comedy. A daughter-in-law could, however, work her way up to higher status and authority by bearing children to the household. The circle closed itself when her sons were old enough to get married and bring in young brides. From this perspective it becomes even more evident why it seemed vital for a woman to have sons, the more the better, since they provided her in old age with emotional and economic support, and they brought young women to the household who took over a considerable share of the domestic tasks and increased her standing, and paid her respect and deference. Indubitably, consanguineous unions, the marriage to a cousin or uncle, softened tensions that could arise between mothers- and daughters-in-law. And in fact consanguineous marriages seem to have been very common in Roman Egypt. As cousin marriages are not as easy to identify as “brother– sister” marriages in the census returns, since we would need the grandparents’ names for both spouses, we should expect an even higher

Conclusions

211

proportion of close-kin unions than previously assumed. In consanguineous marriages, close kinship ties accorded the daughter-in-law as a close relative a higher value within her husband’s family and a greater autonomy in household decision-making. The mother-in-law was instead of a stranger an aunt or otherwise closely related woman whom the young bride had known since childhood. The retention of family wealth and the dowry within the family and the strengthening of kinship solidarity were further motives that led parents to look for a groom for their daughters within their wider family circle. Having children was not always a guarantee for having access to them or their support in later life. Although co-residing sons and their families were probably the most important contributors to old-age care, this family model constituted only the ideal that often could not be achieved because of high infant mortality rates. Moreover, sons migrating away from home for work and daughters marrying and leaving the household had probably a deep impact on resources available for old-age support. A considerable proportion of parents therefore found themselves childless or de facto childless in old age. Not only did childlessness affect an individual’s financial security in old age; in traditional societies not having children went along with social isolation and stigma, which were already felt much earlier in life. Because women were more likely to be widowed in old age, childlessness and its consequences had an even deeper effect on them. We find several strategies employed by elderly individuals in the sources from Roman Egypt. Elderly couples or individuals without children lived with and were taken care of by more distant kin, such as nephews, probably in return for a proportion of the inheritance. Some others, especially solitary elderly women, owned slaves who were available to care for their owners and keep them company, a strategy that required, however, a certain degree of economic prosperity. Beyond the household walls, it is hard to estimate how much care was available from relatives or neighbors. In the pre-Christian period those who did not belong to a specific household and were lacking close relatives and economic means had severe problems finding care and support if they needed it. Communal charity was only marginally developed. The best alternative to having one’s own children seems to have been adoption. Adoption could serve as a means to cope with the discrepancies that arose between ideals for succession and inheritance and demographic realities. The most common motivations for adoption seem to have been (1) to provide a destitute or orphaned child with a home and family, (2) to provide a childless couple with an heir, and (3) to secure support in old age and provision of a proper burial. By contrast to the Roman West and

212

Conclusions

classical Athens, where the adoptee was usually a young man, in Roman Egypt, as in other ancient Near Eastern societies, it was predominantly infants and young children who were adopted. And not all adoptees were males; girls were also candidates for adoption, even if the adoption of males seems to have been far more common. Reasons to adopt a girl seem to have been to provide an orphaned or otherwise destitute girl with a home; to foster alliances with another family; or to assimilate marital and parental ties by adopting a spouse’s offspring from a previous union. Adoptive children were given a status that resembled as closely as possible that of a natural child. While women had been able to adopt according to indigenous Egyptian and Near Eastern tradition, they could not do so according to classical Greek and Roman law. In Roman law, women were not able to adopt simply because adoption would confer potestas over the adoptee and women could not hold potestas over anyone. Women could only foster a child but not equal natural mother–child ties by a legal procedure. Women in Roman Egypt, however, did adopt, mainly it seems to make up for the lack of their own children and to establish rights of inheritance of stepchildren. Only in the later Roman period did adoption by women became recognized by Roman jurists, apparently as a concession to local practices in the Eastern provinces. The rare evidence for adoption in Graeco-Roman Egypt is no proof that adoptions took place only rarely. On the contrary, more indirect evidence points to the fact that adoptions were at least as common in Graeco-Roman Egypt as elsewhere in the ancient Mediterranean. If the couple had no sons but only daughters, such a situation could lead an elderly couple to consider an uxorilocal marriage for one or several of their daughters, that is, the introduction of the son-in-law into their household. Even if evidence for Roman Egypt is only indirect, adopting a son and marrying him to a daughter was a strategy to guarantee the continuity of the family and to ensure support for old age employed in many other regions of the ancient world and beyond. Thus, instead of watching their daughter leave their house on her wedding day, the parents gained a male heir who could continue his adoptive father’s business, take over the role of son to the household, provide his adoptive parents with a proper burial, and perform the customary rites at their grave. Moreover, marrying the daughter inside the family meant the dowry also remained in the family. While according to Roman tradition intermarriage between adopted and natural children was forbidden, it was a desired outcome of adoption in societies all over the Eastern Mediterranean.

Conclusions

213

In sum, a flexible set of family strategies was at play in Roman Egypt which was adapted to the individual family and household situations. Cross-cultural comparative studies help us see recurrent patterns in historic household formations, and disperse the alleged uniqueness of the classical world. In this study I have described and analyzed prevalent family strategies and patterns of household composition in Roman Egypt and pointed out similarities with other traditional patriarchal societies in the Mediterranean and in Asia before the twentieth century. However, I have also stressed the considerable differences from early modern societies in north-western Europe. Interestingly, we can observe essential similarities in the structure of households, succession and inheritance patterns, and the employment of family strategies between much earlier and much later periods of Egypt. On the local level the continuity of social custom is striking.

Bibliography

Aboderin, I. 2004 “Modernisation and ageing theory revisited: current explana tions of recent developing world and historical western shifts in material family support for older people,” Ageing & Society 24: 29 50. Achenbaum, W. A. 1978 Old Age in the New Land. Baltimore. Adams, B. 1964 Paramone und verwandte Texte. Berlin. Agigian, A. 2000 “Infertility,” in: C. Kramarae and D. Spender eds. Routledge International Encyclopedia of Women: Women’s Global Issues and Knowledge. 4 vols. New York: iii.1138 9. Agmon, I. 2006 Family and Court: Legal Culture and Modernity in Late Ottoman Palestine. Syracuse, NY. Agree, E. M., A. E. Biddlecom and T. W. Valente 1999 “Multi generational exchanges in Taiwan and the Philippines: a social network approach,” paper presented at the 1999 annual meeting of the Population Association of America. Ahlawat, K. S. and A. S. Zaghal 1989 “Nuclear and extended family attitudes of Jordanian Arabs,” in: K. Boh, G. Sgritta, and M. B. Sussman eds. CrossCultural Perspectives on Families, Work, and Change. New York: 251 73. Al Haj, M. 1989 “Social research on family lifestyles among Arabs in Israel,” Journal of Comparative Family Studies 20: 175 95. Al Rashoud, R. and S. Farid 1991 Kuwait Child Health Survey. Ministry of Health. Kuwait. Alexander, R. T. 1999 “Households and communities in Yaxcaba, Yucatan, Mexico, 1750 1847,” in: D. B. Small and N. Tannenbaum eds. At the Interface of Households and Beyond. Lanham: 175 99. Allam, S. 1974 “De l’adoption en Egypte pharaonique,” Oriens Antiquus 11: 277 95. Allison, P. M. 1999 “Introduction,” in: P. M. Allison ed. The Archaeology of Household Activities. London: 1 18. Alston, R. 1997 “Houses and households in Roman Egypt,” in: A. Wallace Hadrill and R. Laurence eds. Domestic Space in the Roman World: Pompeii and Beyond. Portsmouth, RI: 25 39. 2002 The City in Roman and Byzantine Egypt. London. 2005 “Searching for the Romano Egyptian family,” in: M. George ed. The Roman Family in the Roman Empire: Rome, Italy and Beyond. Oxford: 129 57. 2006 “Settlement dynamics in third and fourth century Roman Egypt,” in: A. S. Lewin and P. Pellegrini eds. Settlements and Demography in the Near East 214

Bibliography

215

in Late Antiquity: Proceedings of the Colloquium, Matera, 27 29 October 2005. Pisa: 23 34. Altekar, A. S. 1959 The Position of Women in Hindu Civilization: Prehistoric Times to the Present Day. 2nd edn. Delhi. Amerise, M. 2008 Girolamo e la Senectus: età della vita e morte nell’epistolario. Studia ephemeridis Augustinianum 109. Rome. Ammar, H. 1966 Growing Up in an Egyptian Village: Silwa, Province of Aswan. New York. Anderson, M. 1988 The American Census: A Social History. New Haven, CT. Andreau, J. and H. Bruhns ed. 1990 Parenté et stratégies familiales dans l’Antiquité romaine. Rome. Angel, R. and M. Tienda 1982 “Determinants of extended household structure: cultural pattern or economic need?” American Journal of Sociology 87: 1360 83. APIS = Advanced Papyrological Information System, www.columbia.edu/cu/ lweb/projects/digital/apis/. Arjava, A. 1996 Women and Law in Late Antiquity. Oxford. 1997 “The guardianship of women in Roman Egypt,” in: B. Kramer, W. Luppe, H. Mähler and G. Pöthke eds. Akten des 21. Internationalen Papyrologenkongresses, Berlin 13. 19.8.1995. Archiv für Papyrusforschung Beiheft 3. Stuttgart and Leipzig: 25 30. 1998 “Paternal Power in Late Antiquity,” Journal of Roman Studies 88: 147 65. Ault, B. A. and L. C. Nevett 1999 “Digging houses: archaeologies of classical and Hellenistic Greek dwelling in the past. Domestic assemblages,” in: P. M. Allison ed. The Archaeology of Household Activities. London and New York: 43 56. eds. 2005 Ancient Greek Houses and Households: Chronological, Regional, and Social Diversity. Philadelphia. Backe Dahmen, A. 2006 Innocentissima aetas: Römische Kindheit im Spiegel literarischer, rechtlicher und archäologischer Quellen des 1. bis 4. Jahrhunderts n. Chr. [with English summary]. Mainz am Rhein. Bagnall, R. S. 1991 “The beginnings of the Roman census in Egypt,” Greek, Roman and Byzantine Studies 32: 255 65. Reprinted in: R. S. Bagnall 2006 Hellenistic and Roman Egypt. Variorum Collected Studies. Aldershot: ch. 13. 1992 “Landholding in late Roman Egypt: the distribution of wealth,” Journal of Roman Studies 82: 128 49. 1993 Egypt in Late Antiquity. Princeton. 1995 Reading Papyri, Writing Ancient History. London. 1997a “Missing females in Roman Egypt,” Scripta Classica Israelica 16: 121 38. 1997b “The people of the Roman Fayum,” in: M. L. Bierbrier ed. Portraits and Masks: Burial Customs in Roman Egypt. London: 7 15. Reprinted in: R. S. Bagnall 2006 Hellenistic and Roman Egypt. Variorum Collected Studies. Aldershot: ch. 14. 2007 “S. J. Matthews, The Journey of Theophanes: Travel, Business and Daily Life in the Roman East,” review, Journal of Roman Archaeology 20: 591. 2009 “Response to Elio Lo Cascio,” in: A. Bowman and A. Wilson eds. Quantifying the Roman Economy: Methods and Problems. Oxford: 107 14.

216

Bibliography

Bagnall, R. S. and R. Cribiore 2006 Women’s Letters from Ancient Egypt, 300 BC–AD 800. Ann Arbor, MI. Bagnall, R. S. and P. Derow 2004 The Hellenistic Period: Historical Sources in Translation. 2nd edn. Oxford. Bagnall, R. S. and B. W. Frier 1994 The Demography of Roman Egypt. Cambridge. 2006 The Demography of Roman Egypt. 2nd edn. Cambridge. Bagnall, R. S. and D. W. Rathbone 2004 Egypt from Alexander to the Copts: An Archaeological and Historical Guide. London. Bagnall, R. S., B. W. Frier and I. C. Rutherford 1997 The Census Register P.Oxy. 984. Brussels. Baines, J. 1983 “Literacy and ancient Egyptian society,” Man 18: 572 99. Baltrusch, E. 2003 “An den Rand gedrängt: Altersbilder im Klassischen Athen,” in: A. Gutsfeld and W. Schmitz eds. Am schlimmen Rand des Lebens? Altersbilder in der Antike. Cologne, Weimar, and Vienna: 57 86. Bang, P. F. 2009 The Roman Bazaar: A Comparative Study of Trade and Markets in a Tributary Empire. Cambridge and New York. Barakat, H. 1993 The Arab World: Society, Culture, and State. Berkeley. Barbagli, M. 1984 Sotto lo stesso tetto: mutamenti della famiglia in Italia dal XV al XX secolo. Bologna. 1991 “Three household formation systems in eighteenth and nineteenth century Italy,” in: D. I. Kertzer and R. P. Saller eds. 1991 The Family in Italy from Antiquity to the Present. New Haven: 250 70. Barbagli, M. and D. Kertzer 1990 “An introduction to the history of Italian family life,” Journal of Family History 15: 369 83. Barclay, G. W., A. J. Coale, M. A. Stoto and J. Trussel 1976 “A reassessment of the demography of traditional rural China,” Population Index 42: 606 35. Bargach, J. 2002 Orphans of Islam: Family, Abandonment, and Secret Adoption in Morocco. New York. Barker, D. C. 1997 “The place of residence of the divorced wife in Roman Egypt,” in: B. Kramer, W. Luppe, H. Mähler and G. Pöthke eds. Akten des 21. Internationalen Papyrologenkongresses Berlin, 13. 19. 8. 1995. Archiv für Papyrusforschung Beiheft 3. Stuttgart and Leipzig: 59 66. Barnow, F. 2001 City of the Divine King: The Urban Systems and Urban Architecture in Egypt, Mesopotamia, Indus, India, Nepal and China. An Outline of the History of Urban Development. Vol. i. Copenhagen. Bates, D. and A. Rassam 1983 Peoples and Cultures of the Middle East. Englewood Cliffs, NJ. Bavel, J. Van 2004 “Deliberate birth spacing before the fertility transition in Europe: evidence from nineteenth century Belgium,” Population Studies 58: 95 107. Beall, C. and M. Goldstein 1982 “Work, aging, and dependency in a Sherpa population in Nepal,” Biocultural Studies of Aging: Special Issue, Social Science and Medicine 16: 141 8. Beaucamp, J. 1990 2 Le statut de la femme à Byzance (4e 7e siècle). I . Le droit impérial; II. Les pratiques sociales. Paris.

Bibliography

217

2007 “Byzantine Egypt and imperial law,” in: R. Bagnall ed. Egypt in the Byzantine World, 300 700. Cambridge: 271 87. Befu, H. 1971 Japan: An Anthropological Introduction. San Francisco. Bellah, M. S. 1980 “Agricultural and carpentry implements from Karanis,” unpub lished paper, Kelsey Museum of Archaeology: 11 13. Bengtson, V. L. 2001 “Beyond the nuclear family: the increasing importance of multigenerational bonds,” Journal of Marriage and Family 63: 1 17. Bengtson, V. L. and R. E. L. Roberts 1991 “Intergenerational solidarity in aging families: an example of formal theory construction,” Journal of Marriage and Family 53: 856 70. Bengtson, V. L., R. Giarrusso, J. B. Mabry, and M. Silverstein 2002 “Solidarity, conflict, and ambivalence: complementary or competing perspectives on intergenerational relationships?” Journal of Marriage and Family 64: 568 76. Bengtson, V. L, E. B. Olander and A. A. Haddad 1976 “The ‘generation gap’ and aging family members: toward a conceptual model,” in: J. F. Gubrium ed. Time, Roles and Self in Old Age. New York. Benigno, F. 1989 “The southern Italian family in the early modern period: a discussion of co residential patterns,” Continuity and Change 4: 165 94. Bergamasco, M. 1995 “Le didaskalika nella ricerca attuale,” Aegyptus 75: 95 167. 1996 “Una nota su P. Heid. iv 327,” Zeitschrift fu¨r Papyrologie und Epigraphik 110: 174. 1997 “Una petizione per violazione di un contratto di tirocinio: PKell G 19.a,” Aegyptus 77: 7 26. 1998 “P. Kell. G. 19.a, Appendix. 197,” Zeitschrift fu¨r Papyrologie und Epigraphik 121: 193 6. Berkner, L. K. 1972 “The stem family and the development cycle of the peasant household: an eighteenth century Austrian example,” American Historical Review 77: 398 418. Bertier, J. 1996 “La médecine des enfants à l’époque impériale”, in: Aufstieg und Niedergang der Antiken Welt. II. Principat 37.3. Berlin: 2147 227. Bertman, S. ed. 1976 The Conflict of Generations in Ancient Greece and Rome. Amsterdam. Bhat, P. N. M. 1987 “Mortality in India: levels, trends and patterns,” PhD dis sertation, University of Pennsylvania, Philadelphia. Bieżuńska Małowist, I. 1971 “Die Expositio von Kindern als Quelle der Sklavenbeschaffung im griechisch römischen Ägypten,” Jahrbuch fu¨r Wirtschaftsgeschichte 2: 129 33. 1977 L’esclavage dans l’Égypte gréco-romaine. Seconde partie: période romaine. Archiwum Filologiczne 35. Wrocław, Warsaw, Krakow and Gdansk. Biraben, J. N. 1972 “A southern village: the inhabitants of Montplaisant in 1644,” in: P. Laslett and R. Wall eds. Household and Family in Past Time. Cambridge: 237 54. Bittles, A. H. 1994 “The role and significance of consanguinity as a demographic variable,” Population and Development Review 20: 561 84.

218

Bibliography

Bittles, A. H., H. S. Savithri, H. S. Venkatesha Murthy, G. Baskaran, W. Wang, J. Cahill, et al. 2001 “Human inbreeding: a familiar story full of surprises,” in: H. Macbeth and P. Shetty eds. Ethnicity and Health. London: 68 78. Bloch, M. 1998 How We Think They Think: Anthropological Approaches to Cognition, Memory, and Literacy. Boulder, CO. Boak, A. E. R. 1933 Papyri from Tebtunis. Ann Arbor. 1935 Soknopaiou Nesos: The University of Michigan Excavations at Dime in 1931 32. Ann Arbor. Bodel, J. ed. 2001 Epigraphic Evidence: Ancient History from Inscriptions. London. Bodel, J. and S. M. Olyan eds. 2008 Household and Family Religion in Antiquity. Oxford. Bonte, P. ed. 1994 Epouser au plus proche: inceste, prohibitions et stratégies matrimoniales autour de la Méditerranée. Paris. Boozer, A. 2007 “Housing empire: the archaeology of daily life in Roman Amheida, Egypt”, PhD dissertation, Columbia University. 2010 “Memory and microhistory of an empire: domestic contexts in Roman Amheida, Egypt,” in: D. Borić ed. Archaeology and Memory. Oxford: 138 57. Borgeaud, P. 2004 Aux origines de l’histoire des religions. Paris. Boswell, J. E. 1984 “Expositio and oblatio: the abandonment of children and the ancient and medieval family,” American Historical Review 89: 10 33. Republished in: C. Neel ed. 2004 Medieval Families: Perspectives on Marriage, Household, and Children. Toronto: 234 72. 1988 The Kindness of Strangers: The Abandonment of Children in Western Europe from Late Antiquity to the Renaissance. New York. Bourdelais, P. 1985 “Vieillir en famille dans la France des ménages complexes (l’exemple de Prayssas, 1836 1911),” Annales de Démographie Historique 24: 21 38. Bowen, J. R. 1996 “Equality, difference, and law in Indonesian inheritance practi ces: a Sumatran case study,” Political and Legal Anthropology Review 19: 83 90. Bowman, A. K. 1986 Egypt after the Pharaohs, 332 BC AD 642. London. 2005 “Egypt from Septimius Severus to the death of Constantine,” in: A. K. Bowman, P. Garnsey, and A. Cameron eds. The Cambridge Ancient History. XII. The Crisis of Empire, A.D . 193 337. Cambridge: 313 26. Bradley, K. R. 1984 Slaves and Masters in the Roman Empire: A Study in Social Control. Oxford. 1991a Discovering the Roman Family. Oxford. 1991b “Remarriage and the structure of the upper class Roman family,” in: B. Rawson ed. Marriage, Divorce, and Children in Ancient Rome. Oxford: 79 98. Brandes, S. 1996 “Kinship and care for the aged in traditional rural Iberia,” in: T. Hareven ed. Aging and Generational Relations: Life-Course and Cross-Cultural Perspectives. New York: 13 29. Brandt, H. 2002 Wird auch silbern mein Haar: Eine Geschichte des Alters in der Antike. Munich. Bremmer, J. N. 1976 “Avunculate and fosterage,” Journal of Indo-European Studies 4: 65 78. 1983 “The importance of the maternal uncle and grandfather,” Zeitschrift fu¨r Papyrologie und Epigraphik 50: 173 86.

Bibliography

219

1987 “The old women of ancient Greece,” in: J. Blok and P. Mason eds. Sexual Asymmetry. Amsterdam: 191 215. 2000 “Fosterage, kinship and the circulation of children in ancient Greece,” Dialogos: Hellenic Studies Review 6: 1 20. Breschi, M., M. Manfredini and A. Fornasin 2007 “Remarriage in a pre transitional Italian community,” Continuity and Change 22: 1 22. Brettell, C. B. 1991 “Kinship and contract: property transmission and family relations in northwestern Portugal,” Comparative Studies in Society and History 33: 443 65. Broude, G. J. 1995 Growing Up: A Cross-Cultural Encyclopedia. Ann Arbor. Brown, J. K. 1997 “Agitators and peace makers: cross cultural perspectives on older women and the abuse of young wives,” in: A. Sev’er ed. A Cross-Cultural Exploration of Wife Abuse. Lewiston, ME: 79 99. 2004 “Transitions in the life course of women,” in: C. R. Ember and M. Ember eds. Encyclopedia of Sex and Gender. New York: 163 74. Brown, J. K., P. Subbaiah and T. Sarah 1994 “Being in charge: older women and their younger female kin,” in: J. K. Brown and J. Dickerson Putman eds. Women’s Age Hierarchies: Special Issue, Journal of Cross-Cultural Gerontology 9: 231 54. Browne, G. M. 1970 “The composition of the Sortes Astrampsychi,” Bulletin of the Institute of Classical Studies 17: 95 100. 1974 The Papyri of the Sortes Astrampsychi. Beiträge zur klassischen Philologie 58. Leipzig. ed. 1983 Sortes Astrampsychi. Vol. i. Leipzig. Brule, P. 1992 “Infanticide et abandons d’enfants: pratiques grecques et comparai sons anthropologiques,” Dialogues d’Histoire Ancienne 18: 53 90. Brunt, P. A. 1987 Italian Manpower. 2nd edn. Oxford. Burgess, E. W. 1916 The Function of Socialization in Social Evolution. Chicago. Burnet, R. 2003 L’Égypte ancienne à travers les papyrus. Paris. Burrows, M. 1940 “The ancient oriental background of Hebrew levirate marriage,” Bulletin of the American School of Oriental Research 77: 2 15. Burton, A. 1989 “Looking forward from Ariès? Pictorial and material evidence for the history of childhood and family life,” History and Theory 19: 132 53. Burton, M. L., K. L. Nero and J. Hess 2002 “Who can belong to a Micronesian household: representations of household composition across social contexts,” Field Methods 14: 65 87. Bussi, S. 2002 “Mariages endogames en Égypte hellénistique et romaine,” Revue Historique de Droit Français et Etranger 80: 1 22. Caftanzoglou, R. 1994 “The household formation pattern of a Vlacj mountain community of Greece: Syrrako, 1898 1929,” Journal of Family History 19: 79 98. 1997 “Shepherds, innkeepers and census takers: the 1905 census in two villages in Epirus,” Continuity and Change 12: 403 24. Cain, M. 1977 “The economic activities of children in a village in Bangladesh,” Population and Development Review 3: 201 27.

220

Bibliography

1981a “Risk and insurance: perspectives on fertility and agrarian change in India and Bangladesh,” Population and Development Review 7: 435 74. 1981b “The consequences of reproductive failure: dependence, mobility, and mortality among the elderly of rural South Asia,” Population Studies 40: 375 88. Calderini, A. 1924 La composizione della famiglia secondo le schede di censimento dell’Egitto romano. Milan. Caldwell, J. C. 1976a “Fertility and the household economy in Nigeria,” Journal of Comparative Family Studies 7: 193 253. 1976b “Toward a restatement of demographic transition theory,” Population and Development Review 2: 321 66. 1982 Theory of Fertility Decline. London. 2005 “On net intergenerational wealth flows: an update,” Population and Development Review 31: 721 40. Cameron, A. 1939 “ΘΡΕΠΤΟΙ and related terms in the inscriptions of Asia Minor,” in: W. M. Calder and J. Keil eds. Anatolian Studies Presented to W. H. Buckler. Manchester: 27 62. Caplan, L. 1970 Land and Social Change in East Nepal: A Study of Hindu-Tribal Relations. London. Černý, J. 1954 “Consanguineous marriages in Pharaonic Egypt,” Journal of Egyptian Archaeology 40: 23 9. Champlin, E. 1986 “Miscellanea testamentaria,” Zeitschrift fu¨r Papyrologie und Epigraphik 62: 247 55. 1991 Final Judgements: Duty and Emotion in Roman Wills, 200 BC AD 250. Berkeley. Chang, K. S. 1990 “Socialist institutions and family wealth flows reversal: an assessment of post revolutionary Chinese rural fertility,” Journal of Family History 15: 179 200. Cheal, D. 1989 “Strategies of resource management in household economies: moral economy or political economy?” in: R. R. Wilk ed. The Household Economy: Reconsidering the Domestic Mode of Production. Boulder, CO: 11 22. Cicirelli, V. G. 1994 “Sibling relationships in cross cultural perspective,” Journal of Marriage and Family 56: 7 20. Clark, G. 1994 Women in Late Antiquity: Pagan and Christian Lifestyles. Oxford. Clarysse, W. 1995 “Ptolemaic wills,” in: M. Geller, H. Mähler and A. Lewis eds. Legal Documents of the Hellenistic World: Papers from a Seminar arranged by the Institute of Classical Studies, the Institute of Jewish Studies and the Warburg Institute, University of London, February to May 1986. London: 88 105. Clarysse, W. and D. J. Thompson 2006 Counting the People in Hellenistic Egypt. I . Population Registers (P. Count); II. Historical Studies. Cambridge. Coale, A. J. and P. Demeny 1983 Regional Model Life Tables and Stable Populations. 2nd edn. New York and London. Coale, A. and R. Treadway 1979 “A summary of changing fertility in the provinces of Europe,” paper presented to the Summary Conference on European Fertility, Princeton.

Bibliography

221

Cohen, M. R. 2005 Poverty and Charity in the Jewish Community of Medieval Egypt. Princeton, NJ. Cohen, S. J. D. ed. 1993 The Jewish Family in Antiquity. Atlanta, GA. Cohn, M. C. 1908 Breviarium Alaricianum: Römisches Recht im fränkischen Reich in systematischer Darstellung. Leipzig. Cokayne, K. 2003 Experiencing Old Age in Ancient Rome. London. Constantelos, D. J. 1968 Byzantine Philanthropy and Social Welfare. New Brunswick, NJ. Corbett, J. H. 1983 “The foster child: a neglected theme in early Christian life and thought,” in: P. Slater and D. Wiebe eds. Traditions in Contact and Change: Selected Proceedings of the XIV th Congress of the International Association for the History of Religions. Waterloo, ON: 307 21. Corbier, M. 1991a “Constructing kinship in Rome: marriage and divorce, filiation and adoption,” in: D. I. Kertzer and R. P. Saller eds. The Family in Italy from Antiquity to the Present. New Haven: 127 44. 1991b “Divorce and adoption as Roman familial strategies,” in: B. Rawson ed. Marriage, Divorce and Children in Ancient Rome. Canberra and Oxford: 47 78. 1991c “Family behavior of the Roman aristocracy,” in: S. Pomeroy ed. Women’s History and Ancient History. London: 173 96. 1999a “Adoptés et nourris,” in: M. Corbier ed. Adoption et fosterage. Paris: 5 41. ed. 1999b Adoption et fosterage. Paris. 2001a “Child exposure and abandonment,” in: S. Dixon ed. Childhood, Class and Kin in the Roman World. London and New York: 52 73. 2001b “Solidarité entre les générations et sécurité du cycle de vie: la société romain (iie siècle av. J. C. iiie siècle apr. J. C.),” in: P. Guillaume ed. Les solidarités: le lien social dans tous ses états (colloque de Bordeaux, 16 17 juin 2000). Bordeaux: 102 23. 2005 “Family and kinship in Roman Africa,” in: S. Dixon, ed. The Roman Family in the Empire: Rome, Italy and Beyond. Oxford: 255 85. 2006 “Parenté et alliance dans le monde romain (iiie siècle avant J. C. ive siècle après J. C.),” in: A. Bresson, M. P. Masson, S. Perentidis and J. Wilgaux eds. Parenté et société dans le monde grec, de l’Antiquité a l’âge moderne: colloque international, Volos, 19 21 juin 2003. Paris: 199 208. Corvisier, J. N. 1985 “La vieillesse en Grèce ancienne d’Homère à l’époque hellénistique,” Annales de Démographie Historique: 53 70. Corvisier, J. N. and W. Suder 2000 La population de l’Antiquité classique. Paris. Cowgill, D. O. 1974 “Aging and modernization: a revision of the theory,” in: J. F. Gubrium ed. Late Life: Communities and Environmental Policy. Springfield, IL: 123 46. 1986 Aging around the World. Belmont, CA. Cowgill, D. O. and L. D. Holmes eds. 1972 Aging and Modernization. New York. Cox, C. A. 1998 Household Interests: Property, Marriage Strategies, and Family Dynamics in Ancient Athens. Princeton. Crayen, D. and J. Baten 2010 “Global trends in numeracy, 1820 1949, and its implications for long term growth,” Explorations in Economic History 47: 82 99.

222

Bibliography

Cribiore, R. 1996 Writing, Teachers and Students in Graeco-Roman Egypt. Atlanta. 2001 Gymnastics of the Mind: Greek Education in Hellenistic and Roman Egypt. Princeton. Crook, J. 1967 “Patria potestas,” Classical Quarterly 17: 113 22. Cuno, K. M. 1995 “Joint family households and rural notables in 19th century Egypt,” International Journal of Middle East Studies 27: 485 502. 2005 “Demography, household formation, and marriage in three Egyptian villages during the mid nineteenth century,” in: M. Afifi, R. Chih, B. Marino, N. Michel and I. Tamdogan eds. 2005 Sociétés rurales ottomanes/ Ottoman Rural Societies. Cairo: 105 17. Czap, P. 1978 “Marriage and the peasant joint family in the era of serfdom,” in: D. Ransel ed. The Family in Imperial Russia: New Lines of Historical Research. Urbana: 103 23. 1982 “The perennial multiple family household, Mishino, Russia 1782 1858,” Journal of Family History 7: 5 26. 1983 “‘A large family: the peasants’ greatest wealth’: serf households in Mishino, Russia, 1815 1858,” in: R. Wall, P. Laslett and J. Robin eds. Family Forms in Historic Europe. Cambridge: 105 51. Dagron, G. 1974 Naissance d’une capitale. Paris. Daniel, R. W. 2010 Architectural Orientation in the Papyri. Paderborn, Munich, and Vienna. Darmezin, L. 1999 Les affranchissements par consécration en Béotie et dans le monde grec hellénistique. Nancy and Paris. Das Gupta, M. 1995 “Fertility decline in Punjab, India: parallels with historical Europe,” Population Studies 49: 481 91. 1997 “Kinship systems and demographic regimes,” in: D. I. Kertzer and T. Fricke eds. Anthropological Demography: Towards a New Synthesis. Chicago and London: 36 52. David, E. 1991 Old Age in Sparta. Amsterdam. David, M. 1927 Die Adoption im altbabylonischen Recht. Leipzig. Davies, J. K. 2007 “Linear and nonlinear flow models for ancient economies,” in: J. G. Manning and I. Morris eds. The Ancient Economy. Stanford: 127 56. Davis, K. 1941 “Family structure and functions,” American Sociological Review 8: 311 20. Davoli, P. 1998 L’archeologia urbana nel Fayyum di età ellenistica e romana. Naples. de Beauvoir, S. 1972 Old Age. London. de Luce, J. 1994 “Ancient images of aging: did ageism exist in Greco Roman antiquity?” in: D. Shenk and W. A. Achenbaum eds. Changing Perceptions of Aging and the Aged. New York: 65 74. DeLancey, V. 1990 “Socioeconomic consequences of high fertility for the family,” in: G. T. F. Acsadi, G. J. Acsadi, and R. A. Bulatao eds. Population Growth and Reproduction in Sub-Saharan Africa: Technical Analyses of Fertility and Its Consequences. World Bank, Washington: 205 34. Delaney, C. 1991 The Seed and the Soil: Gender and Cosmology in Turkish Village Society. Berkeley and Los Angeles.

Bibliography

223

Demand, N. 1994 Birth, Death, and Motherhood in Classical Greece. Baltimore. Dickerson Putman, J. and J. K. Brown eds. 1998 Women among Women: Anthropological Perspectives on Female Age Hierarchies. Urbana and Chicago. Dickey, E. 2004 “Literal and extended use of kinship terms in documentary papyri,” Mnemosyne 57: 131 76. Dixon, S. 1988 The Roman Mother. London. 1992 The Roman Family. Baltimore. Donker van Heel, K. 1990 The Legal Manual of Hermopolis: Text and Translation. Leiden. Dore, R. P. 1999 City Life in Japan: A Study of a Tokyo Ward. Berkeley and Los Angeles. (1st edn. 1958.) Douglass, W. A. 1991 “The joint family household in eighteenth century south Italian society,” in: D. I. Kertzer and R. P. Saller eds. The Family in Italy from Antiquity to the Present. New Haven: 286 304. Draper, P. and J. Keith 1992 “Cultural contexts of care: family caregiving for elderly,” Journal of Aging Studies 6: 113 34. Drexhage, H J. 1991 Preise, Mieten/Pachten, Kosten und Löhne im römischen Ägypten bis zum Regierungsantritt Diokletians. St. Katharinen. Drèze, J. 1996 India: Development and Participation. Oxford. Driver, G. R. and J. C. Miles 1952 The Babylonian Laws. I . Legal Commentary. Oxford. 1955 The Babylonian Laws. II. Laws of Hammu-Rabi. Oxford. Duben, A. 1982 “The significance of family and kinship in urban Turkey,” in: C. Kagitcibasi ed. Sex Roles, Family and Community in Turkey. Bloomington: 73 99. 1985 “Turkish families and households in historical perspective,” Journal of Family History 10: 75 97. 1990 “Understanding Muslim households and families in late Ottoman Istanbul,” Journal of Family History 15: 71 86. Duben, A. and C. Behar 1991 Istanbul Households: Marriage, Family and Fertility, 1880 1940. Cambridge. Duncan Jones, R. 1990 Structure and Scale in the Roman Economy. Cambridge. Dunn, C. J. 1969 Everyday Life in Traditional Japan. London. Dupâquier, J. and L. Jadin 1972 “Structure of household and family in Corsica, 1769 71,” in: P. Laslett and R. Wall eds. Household and Family in Past Time. Cambridge: 283 97. Durkheim, E. 1933. The Division of Labor in Society, trans. G. Simpson. New York. Dyson, T. and M. Moore 1983 “Kinship structure, female autonomy, and demo graphic behavior in India,” Population and Development Review 9: 35 60. Ebrey, P. B. 1991 Confucianism and Family Rituals in Imperial China: A Social History of Writing about Rites. Princeton. 1993 The Inner Quarters: Marriage and the Lives of Chinese Women in the Sung Period. Berkeley and Los Angeles. Eck, W. 1978 “Zum neuen Fragment des sogenannten testamentum Dasumii,” Zeitschrift fu¨r Papyrologie und Epigraphik 30: 277 95.

224

Bibliography

Edlund, L. and A. Rahman 2005 “Household structure and child outcomes: nuclear vs. extended families evidence from Bangladesh,” mimeo., Columbia University. Edmondson, J. 2005 “Family relations in Roman Lusitania: social change in a Roman province?” in: M. George ed. The Roman Family in the Empire: Rome, Italy, and Beyond. Oxford: 183 229. Edwards, M. 2007 Isaeus. Austin, TX. Ehmer, J., T. K. Hareven and R. Wall eds. 2001 Family History Revisited: Comparative Perspectives. Newark, NJ. El Mosallamy, A. H. 1994 “Upon veterans’ exemption from epikephalia,” in: A. Bülow Jacobsen ed. Proceedings of the 20th International Congress of Papyrologists. Copenhagen: 456 62. Eldem, S. H. 1968 Turk Evi Plan Tipleri (Turkish House Plan Types). Istanbul. 1984 Turkish House: Ottoman Period. 2 vols. Istanbul. Engelen, T. and A. P. Wolf eds. 2005 Marriage and the Family in Eurasia: Perspectives on the Hajnal Hypotheses. Life at the Extremes 1. Amsterdam. Engels, D. 1980 “The problem of female infanticide in the Graeco Roman world,” Classical Philology 75: 112 20. Erdkamp, P. 1999 “Agriculture, underemployment, and the cost of rural labour in the Roman world,” Classical Quarterly 49: 556 72. 2005 The Grain Market in the Roman Empire: A Social, Political and Economic Study. Cambridge. Evans Grubbs, J. 2002 Women and the Law in the Roman Empire: A Sourcebook on Marriage, Divorce, and Widowhood. London and New York. 2010 “Hidden in plain sight: expositi in the community,” in: V. Dasen and T. Späth eds. Children, Memory and Family Identity in Roman Culture. Oxford: 293 310. Eyben, E. 1977 De jonge Romein. Brussels. 1993 Restless Youth in Ancient Rome. London and New York. Eyre, C. J. 1984 “Crime and adultery in ancient Egypt,” Journal of Egyptian Archaeology 70: 92 105. 1992 “The adoption papyrus in social context,” Journal of Egyptian Archaeology 78: 207 21. 1999 “The village economy in ancient Egypt,” in: A. K. Bowman and E. Rogan eds. Agriculture in Egypt: From Pharaonic to Modern Times. Oxford: 33 60. 2004 “How relevant was personal status to the functioning of the rural economy in Pharaonic Egypt?” in: B. Menu ed. La dépendance rurale dans l’Antiquité égyptienne et proche-orientale. Cairo: 157 86. 2007 “The evil stepmother and the rights of a second wife,” Journal of Egyptian Archaeology 93: 223 43. Falkner, T. M. 1995 The Poetics of Old Age in Greek Epic, Lyric, and Tragedy. Oklahoma. Falkner, T. M. and J. de Luce eds. 1989 Old Age in Greek and Roman Literature. Albany, NY.

Bibliography

225

Farber, J. J. 1996 “Byzantine texts,” in: B. Porten et al. eds. The Elephantine Papyri in English: Three Millennia of Cross-Cultural Continuity and Change. Leiden: 438 549. Farnsworth, B. 1986 “The litigious daughter in law: family relations in rural Russia in the second half of the nineteenth century,” Slavic Review 45: 49 64. Fauve Chamoux, A. 1988 “Les structures familiales en France aux xviie et xviiie siècles,” in : J. Dupâquier ed. Histoire de la population française. Vol. ii. Paris: 317 47. 1996 “Aging in a never empty nest: the elasticity of the stem family,” in: T. K. Hareven ed. Aging and Generational Relations over the Life Course. Berlin and New York: 75 99. Finley, M. I. 1973 The Ancient Economy. Berkeley and Los Angeles. 1981 “The elderly in classical antiquity,” Greece & Rome 28: 156 71. Finn, R. D. 2006 Almsgiving in the Later Roman Empire: Christian Promotion and Practice, 313 450. Oxford and New York. Fischer, D. H. 1977 Growing Old in America. Oxford and New York. Flandrin, J. L. 1979 Families in Former Times. London. Foner, N. 1984 Ages in Conflict: A Cross-Cultural Perspective on Inequality between Young and Old. New York. 1989 “Older women in non industrial cultures: consequences of power and privilege,” in: L. Grau and I. Susser eds. Women in the Later Years: Health, Society, and Cultural Perspectives. London: 227 38. Fraad, H., S. Resnick and R. Wolff 1994 Bringing It All Back Home. London. Frederiksen, M. 1975 “Evidence and models,” review of Finley 1973, Journal of Roman Studies: 164 75. Freedman, M. 1979 “The Chinese domestic family: models,” in: G. W. Skinner ed. The Study of Chinese Society: Essays by Maurice Freedman. Stanford: 235 9. Fridell, E. 2003 “Tribal enterprises and marriage issues in twentieth century Iran,” in: B. Doumani ed. Family History in the Middle East: Household, Property, and Gender. Albany, NY: 151 71. Frier, B. W. 1994 “Natural fertility and family limitation in Roman marriage,” Classical Philology 89: 318 33. 2000 “Demography,” in: A. K. Bowman et al. The Cambridge Ancient History. Vol. xi. 2nd edn. Cambridge: 787 816. Frier, B. W. and T. A. J. McGinn 2003 A Casebook on Roman Family Law. Oxford. Gadalla, S. 1978 Is There Hope? Fertility and Family Planning in a Rural Egyptian Community. Cairo. Gadamer, H. G. 1976 “Das Vaterbild im griechischen Denken,” in: H. Tellenbach ed. Das Vaterbild in Mythos und Geschichte: Ägypten, Griechenland, Altes Testament, Neues Testament. Stuttgart: 102 15. Gadd, C. J. 1926 “Tablets from Kirkurk,” Revue d’Assyriologie et d’Archéologie Orientale 23: 49 161. Gallant, T. W. 1991 Risk and Survival in Ancient Greece: Reconstructing the Rural Domestic Economy. Stanford. Gallazzi, C. 1994 “Tebtunis: piecing together 3000 years of history,” Egyptian Archaeology 5: 27 9.

226

Bibliography

Gallazzi, C. and G. Hadju Minaglou 1989 “Tebtynis I: la reprise des fouilles et le quartier de la chapelle d’Isis thermouthis,” Le Bulletin de l’Institut Français d’Archéologie Orientale 89: 180 202. Gardiner, A. 1941 “Adoption extraordinary,” Journal of Egyptian Archaeology 26: 23 9. 1953 “The harem at Miwer,” Journal of Near Eastern Studies 12: 145 9. Gardner, J. F. 1986 Women in Roman Law and Society. London. 1998 Family and Familia in Roman Law and Life. Oxford. Gardner, J. F. and T. E. J. Wiedeman 1991 The Roman Household: A Sourcebook. London. Garland, R. 1985 The Greek Way of Death. Ithaca. 1990 The Greek Way of Life: From Conception to Old Age. London. 1995 The Eye of the Beholder: Deformity and Disability in the Graeco-Roman World. London. Garnsey, P. 1991 “Child rearing in ancient Italy,” in: D. I. Kertzer and R. P. Saller, eds. The Family in Italy from Antiquity to the Present. New Haven and London: 48 65. 1998 Cities, Peasants and Food in Classical Antiquity. Cambridge. Garnsey, P. and R. P. Saller 1987 The Roman Empire: Economy, Society and Culture. Berkeley. Gaunt, D. 1983 “The property and kin relationships of retired farmers in northern and central Europe,” in: R. Wall, J. Robin and P. Laslett eds. Family Forms in Historic Europe. Cambridge: 249 79. 1987 “Rural household organization and inheritance in northern Europe,” in: T. Hareven and A. Plakans eds. Family History at the Crossroads. Princeton: 121 41. Gazda, E. K. ed. 1983 Karanis: An Egyptian Town in Roman Times. Ann Arbor. George, M. 2005 The Roman Family in the Empire: Rome, Italy, and Beyond. Oxford. Gerber, H. 1980 “Social and economic position of women in an Ottoman city, Bursa, 1600 1700,” International Journal of Middle East Studies 12: 231 44. 1988 Economy and Society in an Ottoman City: Bursa, 1600 1700. Jerusalem. 1989 “Anthropology and family history: the Ottoman and Turkish families,” Journal of Family History 14: 409 21. 1995 State, Society and Law in Islam: Ottoman Law in Comparative Perspective. Syracuse. Germain, L. R. F. 1969 “Aspects du droit d’exposition en Grèce.” Revue Historique de Droit Français et Etranger 47: 177 97. 1975 “L’exposition des enfants nouveau nés dans la Grèce ancienne: aspects sociologiques,” in: L’enfant. Recueils de la Société Jean Bodi pour l’histoire comparative des institutions 35/1. Brussels: 211 46. Gernet, L. 1955 Droit et société dans la Grèce ancienne. Paris. Gilleard, C. 2007 “Old age in Byzantine society,” Ageing & Society 27: 623 42. Gilliam, J. F. 1952 “The minimum subject to the vicesima hereditatum,” American Journal of Philology 73: 397 405. Glanville, S. R. K 1950 “Notes on a Demotic papyrus from Thebes (B.M. 10.026),” in: D. Winton Thomas ed. Essays and Studies Presented to St. A. Cook. London: 59 65.

Bibliography

227

1955 The Instructions of Onchsheshonqy: British Museum Papyrus 10508. Catalogue of Demotic Papyri in the British Museum 2. London. Gnilka, C. 1983 “Greisenalter,” in: Reallexikon fu¨r Antike und Christentum 12: 995 1094. 2001 “Altersversorgung,” in: Reallexikon fu¨r Antike und Christentum: Supplementband I : 266 89. Golden, M. 1981 “Demography and the exposure of girls at Athens,” Phoenix 35: 316 31. 1990 Children and Childhood in Classical Athens. Baltimore and London. Goldsmith, R. W. 1984 “An estimate of the size and structure of the national product of the early Roman Empire,” Review of Income and Wealth 30: 263 88. Goode, W. J. 1963 World Revolution and Family Patterns. New York. Goody, J. 1969 Comparative Studies in Kinship. Stanford. ed. 1971 The Developmental Cycle in Domestic Groups. Cambridge. 1972 “The evolution of the family,” in: P. Laslett and R. Wall eds. Household and Family in Past Time. Cambridge: 103 24. 1973 “Strategies of heirship,” Comparative Studies in Society and History 16: 3 20. 1976 Production and Reproduction: A Comparative Study of the Domestic Domain. Cambridge. 1982 “Erbschaft, Eigentum und Frauen: Einige vergleichende Betrachtungen,” in: M. Mitterauer and R. Sieder eds. Historische Familienforschung. Frankfurt am Main: 88 122. 1983 The Development of the Family and Marriage in Europe. Cambridge. 1990 The Oriental, the Ancient and the Primitive: Systems of Marriage and the Family in the Pre-industrial Societies of Eurasia. Cambridge. 2000 The European Family: An Historico-Anthropological Essay. Oxford. Goody, J. and J. A. Tambiah 1974 Bridewealth and Dowry. Cambridge. Gordon, C. H. 1940 “Biblical customs and the Nuzu tablets,” Biblical Archaeologist 3: 1 12. Grauer, A. L. ed. 1995 Bodies of Evidence: Reconstructing History through Skeletal Analysis. New York. Greenfield, J. 1982 “Adi baltu: care for the elderly and its rewards,” Archiv fu¨r Orientforschung 19: 309 16. Greengus, S. 1995 “Legal and social institutions of ancient Mesopotamia,” in: J. Sasson ed. Civilizations of the Ancient Near East. Vol. i, part 4. New York: 469 84. Grenfell, B. P., A. S. Hunt, and D. G. Hogarth eds. 1990 Fayum Towns and Their Papyri. London. Guidicini, P. and C. Alvisi 1994 L’arzdaura: donne e gestione familiare nella realtà contadina. Milan. Guilmoto, C. and S. I. Rajan 2005 Fertility Transition in South India. New Delhi. Günther, L. M. 1993 “Witwen in der griechischen Antike: zwischen Oikos und Polis,” Historia 42: 308 25. Gutsfeld, A. and W. Schmitz eds. 2003 Am schlimmen Rand des Lebens? Altersbilder in der Antike. Cologne.

228

Bibliography

Gwynn, D. M., S. Bangert and L. Lavan eds. 2010 Religious Diversity in Late Antiquity. Leiden and Boston. Haas, L. 1998 The Renaissance Man and His Children: Childbirth and Early Childhood in Florence, 1300 1600. New York. Häge, G. 1968 Ehegu¨terrechtliche Verhältnisse in den griechischen Papyri Ägyptens bis Diokletian. Cologne and Graz. Hajnal, J. 1965 “European marriage patterns in perspective,” in: D. V. Glass and D. E. C. Eversley eds. Population in History: Essays in Historical Demography. London: 101 43. 1982 “Two kinds of preindustrial household formation system,” Population and Development Review 8: 449 94. 1983 “Two kinds of pre industrial household formation systems,” in: R. Wall, J. Robin and P. Laslett eds. Family Forms in Historic Europe. Cambridge: 65 104. Hammel, E. A. 1972 “The zadruga as process,” in: P. Laslett and R. Wall eds. Household and Family in Past Time. Cambridge: 335 73. 1984 “On the *** of studying household form and function,” in: R. M. Netting, R. R. Wilk and E. J. Arnold eds. Households: Comparative and Historical Studies of the Domestic Group. Berkeley, Los Angeles and London: 29 43. Hammel, E. A. and A. Gullickson 2004 “Kinship structures and survival: maternal mortality on the Croatian Bosnian border, 1750 1898,” Population Studies 58: 145 59. Hammel, E. A. and P. Laslett 1974 “Comparing household structure over time and between cultures,” Comparative Studies in Society and History 15: 75 9. Hammer, J. 1986 “Children and savings in less developed countries,” Journal of Development Economics 23: 107 18. Han, H. 2004 “Kinship, gender and mode of production in post Mao China: variations in two northern villages,” in: R. Parkin and L. Stone eds. Kinship and Family: An Anthropological Reader. Oxford: 408 23. Hanley, S. B. 1985 “Family and fertility in four Tokugawa villages,” in: S. B. Hanley and A. P. Wolf eds. Family and Population in East Asian History. Stanford: 196 228. Hansen, M. H. 1985 Demography and Democracy: The Number of Athenian Citizens in the Fourth Century B.C . Herning. 1988 Three Studies in Athenian Demography. Copenhagen. Hanson, A. E. 2000 “Widows too young in their widowhood,” in: D. E. E. Kleiner and S. B. Matheson eds. I, Claudia II: Women in Roman Art and Society. Austin: 149 65. 2005 “The widow Babatha and the poor orphan boy,” in: R. Katzoff and D. Schaps eds. Law in the Documents of the Judaean Desert. Leiden and Boston: 85 104. Hareven, T. K. 1991 “The history of the family and the complexity of social change,” American Historical Review 96: 95 124. Harlow, M. and R. Laurence 2002 Growing Up and Growing Old in Ancient Rome: A Life Course Approach. London.

Bibliography

229

Harper, K. 2011 Slavery in the Late Roman Mediterranean, AD 275 425: An Economic, Social and Institutional Study. Cambridge. Harris, W. V. 1980 “Towards a study of the Roman slave trade,” in: J. H. D’Arms and E. C. Kopff eds. The Seaborne Commerce of Ancient Rome: Studies in Archaeology and History. Rome: 117 40. 1982 “The theoretical possibility of extensive infanticide in the Graeco Roman World,” Classical Quarterly 32: 114 16. 1989 Ancient Literacy. Cambridge, MA, and London. 1994 “Child exposure in the Roman Empire,” Journal of Roman Studies 84: 1 22. 1999 “Demography, geography and the sources of Roman slaves,” Journal of Roman Studies 89: 62 75. Harrison, A. R. W. 1968 The Law of Athens. I . The Family and Property. Oxford. Hasegawa, K. 2005 The Familia Urbana during the Early Empire: A Study of Columbaria Inscriptions. Oxford. Haupt, H. G. 2001 “Comparative history,” in: International Encyclopedia of the Social and Behavioral Sciences. Vol. iv. Oxford: 2397 403. Haupt, H. G. and J. Kocka eds. 1996 Geschichte und Vergleich: Ansätze und Ergebnisse international vergleichender Geschichtsschreibung. Frankfurt am Main and New York. Henderson, J. 1987 “Older women in Attic comedy,” Transactions of the American Philological Association 117: 105 29. Hengstl, J. 1972 Private Arbeitsverhältnisse freier Personen in den hellenistischen Papyri bis Diokletian. Bonn. Herlihy, D. and C. Klapisch Zuber 1978 Les Toscans et leurs familles: une étude du “catasto” florentin de 1427. Paris. 1985 Tuscans and Their Families: A Study of the Florentine Catasto of 1427. New Haven and London. English abridged trans. of Herlihy and Klapisch Zuber 1978. Hermalin, A. 2003 Family Planning Impact Evaluation: The Evolution of Techniques. PSC Research Report 03 534, April. Herzig, H. E. 1994 “Der alte Mensch in der griechisch römischen Antike,” in: K. Burasélis ed. Unity and Units of Antiquity. Athens: 169 79. Hess, B. B. and E. Warren Markson eds. 1991 Growing Old in America. New Brunswick, NJ. Hezser, C. 2003 “The exposure and sale of infants in rabbinic and Roman law,” in: M. Schlüter et al. eds. Zwischen den Welten: Judaistik im Aufbruch und Wandel. Leiden: 1 26. Hin, S. C. in press The Demography of Roman Italy: Population Dynamics in an Ancient Conquest Society, 201 BCE 14 CE). Cambridge. Hionidou, V. 1995 “Nuptiality patterns and household structure on the Greek island of Mykonos, 1849 1959,” Journal of Family History 20: 67 102. Hobson, D. W. 1985 “House and household in Roman Egypt,” Yale Classical Studies 28: 211 29. 1989 “Naming Practices in Roman Egypt,” Bulletin of the American Society of Papyrology 26: 157 74.

230

Bibliography

Hoch, S. L. 1982 “Serfs in Imperial Russia: demographic insights,” Journal of Interdisciplinary History 13: 221 46. Hodkinson, S. 1983 “Social order and the conflict of values in classical Sparta,” Chiron 13: 239 81. Holmgren, J. 1985 “The economic foundations of virtue: widow remarriage in early and modern China,” Australian Journal of Chinese Affairs 13: 1 27. Holy, L. 1989 Kinship, Honour and Solidarity: Cousin Marriage in the Middle East. Manchester. Hombert, M. and C. Préaux 1952 Recherches sur le recensement dans l’Egypte romaine (Pap.Lugd.Bat. 5). Leiden. Hope, C. and H. Whitehouse 2006 “A painted residence at Ismant el Kharab (Kellis) in the Dakleh Oasis,” Journal of Roman Archaeology 19: 312 28. Hope, V. 1997 “Constructing Roman identity: funerary monuments and social structure in the Roman world,” Mortality 2: 103 21. Hopkins, K. 1965a “Contraception in the Roman Empire,” Comparative Studies in Society and History 8: 124 51. 1965b “On the probable age structure of the Roman population,” Population Studies 20: 245 64. 1980 “Brother sister marriage in Roman Egypt,” Comparative Studies in Society and History 22: 303 54. Hopkins, K. and P. J. Roscoe 1978 “Between slavery and freedom: on freeing slaves at Delphi,” in: K. Hopkins ed. Conquerors and Slaves. Cambridge: 133 71. Hubbard, T. K. 1989 “Old Men in the Youthful Plays of Aristophanes,” in: T. M. Falkner ed. Old Age in Greek and Latin Literature. Albany, NY: 90 113. Huebner, S. R. 2005 “Alte Männer im klassischen Athen (5 Jh.): An den Rand gedrängt?” Göttinger Forum der Altertumswissenschaften 8: 31 57. 2007 “Brother sister marriages in Roman Egypt: a curiosity of humankind or a widespread family strategy?” Journal of Roman Studies 97: 21 49. 2009a “Callirhoe’s dilemma: remarriage and stepfathers in the Graeco Roman East,” in: S. R. Huebner and D. M. Ratzan eds. Growing Up Fatherless in Antiquity. Cambridge: 61 82. 2009b “Therapeuteria reconsidered: female circumcision in Egypt from ancient times until today?” Journal of Egyptian History 2: 149 71. 2010 “Household and family in the Roman East and West: what do we know?” in: B. Rawson ed. Companion to Families in the Greek and Roman Worlds. Oxford: 73 91. 2012a “Brother sister marriage,” in: R. S. Bagnall, K. Brodersen, C. Champlin, A. Erskine and S. R. Huebner eds. Blackwell’s Encyclopedia of Ancient History. Oxford: 1194 5. 2012b “Census,” in: R. S. Bagnall, K. Brodersen, C. Champlin, A. Erskine and S. R. Huebner eds. Blackwell’s Encyclopedia of Ancient History. Oxford: 1410 12. 2012c “Pères et fils dans l’Antiquité tardive: l’expérience de Basile de Césarée,” in: B. Caseau ed. Les réseaux familiaux à la fin de l’Antiquité et au Moyen Âge: colloque en l’honneur d’A. Laiou et d’E. Patlagean. Paris: 45 68.

Bibliography

231

in press a “Adoption and fosterage in the ancient Eastern Mediterranean,” in: J. Evans Grubbs and T. Parkin eds. Oxford Handbook of Childhood and Education in the Classical World. Oxford. in press b “The ‘Mediterranean family’: a useful concept for analyzing family structures in the ancient Mediterranean?,” in: S. R. Huebner and G. Nathan eds. Household and Family in the Mediterranean from Antiquity to the 19th Century. Cambridge. in press c “‘It is a difficult matter to be wronged by strangers, but to be wronged by kin is worst of all’: inheritance and conflict in Graeco Roman Egypt,” in: B. Casean and S. R. Huebner eds. Family, Religion and Inheritance in the Ancient and Medieval Worlds. Paris. Huebner, S. R. and G. Nathan eds. in press Household and Family in the Mediterranean from Antiquity to the 19th Century. Huebner, S. R. and D. M. Ratzan 2009a “Fatherless antiquity? Perspectives on fatherlessness in the ancient Mediterranean,” in: S. R. Hübner and D. M. Ratzan eds. Growing Up Fatherless in Antiquity. Cambridge: 3 28. eds. 2009b Growing Up Fatherless in Antiquity. Cambridge. Hughes, D. O. 1977 “Kinsmen and neighbors in medieval Genoa,” in: H. A. Miskimin, D. Herlihy and A. L. Udovitc eds. The Medieval City. New Haven: 3 28. Hugo, G. 1997 “Intergenerational wealth flows and the elderly in Indonesia,” in: G. W. Jones, R. M. Douglas, J. C. Caldwell and R. M. D’Souza eds. The Continuing Demographic Transition. Oxford and New York: 111 33. Humbert, M. 1972 Le remariage à Rome: étude d’histoire juridique et sociale. Università di Roma. Pubblicazioni dell’Istituto di diritto romano e dei diritti dell’Oriente mediterraneo 44. Milan. Hunter, V. 1989 “The Athenian widow and her kin,” Journal of Family History 14: 291 311. 1993 “Agnatic kinship in Athenian law and Athenian family practice: its impli cations for women,” in: B. Halpern and D. Hobson eds. Law, Politics, and Society in the Ancient Mediterranean World. Sheffield: 100 21. Hussain, R. and R. J. Bittles 1999 “Consanguineous marriage and differentials in age at marriage, contraceptive use and fertility in Pakistan,” Journal of Biosocial Science 31: 121 38. Husselman, E. M. 1979 Karanis Excavations of the University of Michigan in Egypt, 1928 1935: Topography and Architecture. Ann Arbor. Huys, M. 1989 “Ekthesis and apothesis: the terminology of infant exposure in Greek antiquity,” L’Antiquité Classique 58: 190 7. Ikels, C. 1983 “The process of caretaker selection,” Research on Aging 5: 491 509. Inhorn, M. C. 1996 Infertility and Patriarchy: The Cultural Politics of Gender and Family Life in Egypt. Philadelphia. Isager, S. 1981 2 “The marriage pattern in classical Athens: men and women in Isaios,” Classica et Mediaevalia 33: 81 96. Janelli, R. L. 1982 Ancestor Worship and Korean Society. Stanford.

232

Bibliography

Janssen, R. M. and J. J. Janssen 2007 Growing Up and Getting Old in Ancient Egypt. London. Johnson, K. A. 1983 Women, the Family, and Peasant Revolution in China. Chicago. Jouanna, J. 2007 Sophocle. Paris. Jouguet, P. 1911 La vie municipal dans l’Egypte romaine. Paris. Just, R. 1989 Women in Athenian Law and Life. London. Kabeer, N. 1991 “Gender dimensions of rural poverty: analysis from Bangladesh,” Journal of Peasant Studies 18: 241 62. Kagitcibasi, C. 1982 “Introduction,” in: C. Kagitcibasi ed. Sex Roles, Family and Community in Turkey. Bloomington: 1 32. Kandiyoti, D. 1988 “Bargaining with patriarchy,” Gender and Society 2: 274 90. Kaplan, H. 1994 “Evolutionary and wealth flows theories of fertility: empirical tests and new models,” Population and Development Review 20: 753 91. Kaplan, H. S. and J. B. Lancaster 2003 “An evolutionary and ecological analysis of human fertility, mating patterns and parental investment,” in: K. W. Wachter and R. A. Bulatao eds. Offspring: Human Fertility Behavior in Biodemographic Perspective: Proceedings of a Workshop. Washington, DC: 170 223. Karabélias, E. 2002 L’épiclérat attique. Athens. Kaser, K. 1994a “The Balkan joint family household: seeking its origins,” Continuity and Change 9: 45 68. 1994b “The Balkan joint family: redefining a problem,” Social Science History Journal 18: 243 69. 1996 “Introduction: household and family contexts in the Balkans,” History of the Family 1: 375 86. Kaser, M. 1975 Das römische Privatrecht. ii. Die nachklassischen Entwicklungen. Mit Nachträgen zum ersten Abschnitt (Handbuch der Altertumswissenschaft. Rechtsgeschichte des Altertums. Band X,3.3.2). 2nd edn. Munich. (1st edn. Munich, 1959.) Keenan, J. G. 1989 “Roman criminal law in a Berlin papyrus codex (BGU iv 1024 1027),” Archiv fu¨r Papyrusforschung 35: 15 23. Kertzer, D. I. 1983 “Generation as a sociological problem,” Annual Review of Sociology 9: 125 49. 1984 Family Life in Central Italy, 1880 1910: Sharecropping, Wage Labor, and Coresidence. New Brunswick, N.J. 1989 “The joint family household revisited: demographic constraints and house hold complexity in the European past,” Journal of Family History 14: 1 15. 1991 “Reflections on the European marriage pattern: sharecropping and pro letarization in Casalecchio, Italy, 1861 1921,” Journal of Family History 16: 31 45. 1995 “Toward a historical demography of aging,” in: D. I. Kertzer and P. Laslett eds. Aging in the Past: Demography, Society, and Old Age. Berkeley, Los Angeles and Oxford: 363 83. 2001 “Living with kin,” in: D. I. Kertzer and M. Barbagli eds. The History of the European Family. II. Family Life in the Long Nineteenth Century, 1789 1913. New Haven and London: 40 72.

Bibliography

233

Kertzer, D. I. and M. Barbagli 2001 The History of the European Family. I . Family Life in Early Modern Times, 1500 1789; II. Family Life in the Long Nineteenth Century, 1789 1913. New Haven and London. Kertzer, D. I. and D. P. Hogan 1989 Family, Political Economy, and Demographic Change: The Transformation of Life in Casalecchio, Italy, 1861 1921. Madison. 1995 “Relations between older adults and their adult children in a nineteenth century Italian town,” in: T. K. Hareven ed. Aging and Generational Relations over the Life Course: A Historical and Cross-Cultural Perspective. Newark, DE: 120 39. Kertzer, D. I. and P. Laslett eds. 1995 Aging in the Past: Demography, Society, and Old Age. Berkeley, Los Angeles and Oxford. Kertzer, D. I. and R. P. Saller eds. 1991 The Family in Italy from Antiquity to the Present. New Haven. Klapisch, C. 1972 “Household and family in Tuscany, 1427,” in: P. Laslett and R. Wall eds. 1972 Household and Family in Past Time. Cambridge: 267 81. Klapisch Zuber, C. 1985 Women, Family and Ritual in Renaissance Italy. Chicago. 1998 “The ‘cruel mother’: maternity, widowhood and dowry in Florence in the fourteenth and fifteenth centuries,” in: L. K. Little and B. H. Rosenwein eds. Debating the Middle Ages: Issues and Readings. Princeton: 264 76. Klapisch Zuber, C. and M. Demonet 1972 “‘A uno pane e uno vino’: la famille rurale toscane au début de xve siècle,” Annales E.S.C. 27: 873 902. Kleijwegt, M. 1991 Ancient Youth: The Ambiguity of Youth and the Absence of Adolescence in Greco-Roman Society. Amsterdam. Knodel, J. E. and N. Chayovan 1997 “Family support and living arrangements of Thai elderly,” Asia-Pacific Population Journal 12: 51 68. 2008 Population Ageing and the Well-Being of Older Persons in Thailand. PSC Research Report 08 659. Kocka, J. 1999 “Asymmetrical historical comparison: the case of the German Sonderweg,” History and Theory 38: 40 50. 2009 “Comparative history: methodology and ethos,” East Central Europe 36: 12 19. Komter, A. E. 2005 Social Solidarity and the Gift. Cambridge. Krause, J. U. 1994 5 Witwen und Waisen im Römischen Reich. I . Verwitwung und Wiederverheiratung; II. Wirtschaftliche und gesellschaftliche Stellung von Witwen im fru¨hen Christentum; III. Rechtliche und soziale Stellung von Waisen; IV . Witwen und Waisen im fru¨hen Christentum. Stuttgart. 2003 “Antike,” in: A. Gestrich, J. U. Krause and M. Mittermauer eds. Geschichte der Familie. Stuttgart: 21 159. Kreller, H. 1919 Erbrechtliche Untersuchungen auf Grund der gräco-ägyptischen Papyrusurkunden. Leipzig and Berlin. Kurosu, S. 1997 “Adoption as an heirship strategy? A case from a north eastern village in pre industrial Japan,” Japan Review: 171 89. Kurosu, S. and E. Ochiai 1995 “Adoption as an heirship strategy under demo graphic constraints: a case from nineteenth century Japan,” Journal of Family History 20: 261 88.

234

Bibliography

Kurylowicz, M. 1983 “Adoption on the evidence of the papyri,” Journal of Juristic Papyrology 19: 61 75. Lacey, W. K. 1968 The Family in Classical Greece. London. Lang, O. 1946 Chinese Family and Society. New Haven. Lascaratos, J., G. Kalantzis and E. Poulakou Rebelakou 2004 “Nursing homes for the old (‘Gerocomeia’) in Byzantium (324 1453 ad),” Gerontology 50: 113 17. Laslett, P. 1965 The World We Have Lost. New York. 1972 “Introduction: the history of the family,” in: P. Laslett and R. Wall eds. Household and Family in Past Time. Cambridge: 23 89. 1977 “Characteristics of the Western family considered over time,” in: P. Laslett ed. Family Life and Illicit Love in Earlier Generations. Cambridge: 12 49. 1983 “Family and household as work group and kin group: areas of traditional Europe compared,” in: R. Wall, J. Robin and P. Laslett eds. Family Forms in Historic Europe. Cambridge: 513 63. 1988 “Family, kinship and collectivity as systems of support in pre industrial Europe: a consideration of the ‘nuclear hardship’ hypothesis,” Continuity and Change 3: 153 75. Laslett, P. and R. Wall eds. 1972 Household and Family in Past Time. Cambridge. Laungani, P. D. 2007 Understanding Cross-Cultural Psychology: Eastern and Western Perspectives. Thousand Oaks, CA. Le Bras, H. 2008 The Nature of Demography. Princeton. Le Play, F. 1871 L’organisation de la famille selon le vrai modèle signalé par l’histoire. Paris. Le Roy Ladurie, E. 1966 Les paysans de Languedoc. Paris. 1975 Montaillou, village occitan de 1294 à 1324. Paris. Leduc, C. 1994 “Come darla in matrimonio? La sposa nel mondo greco, secoli IX IV bc,” in: P. Schmitt Pantel ed. Storia delle donne: l’antichità. Bari: 246 314. Lee, J. Z. and C. D. Campbell 1997 Fate and Fortune in Rural China: Social Organization and Population Behavior in Liaoning, 1174 1873. Cambridge. 1998 “Headship succession and household division in three Chinese banner serf populations, 1789 1909,” Continuity and Change 13: 117 41. Lee, R. D. 2000 “A cross cultural perspective on intergenerational transfers and the economic life cycle,” in: A. Mason and G. Tapinos eds. Sharing the Wealth: Demographic Change and Economic Transfers between Generations. Oxford: 17 56. Lee, R. D. and K. L. Kramer 2002 “Children’s economic roles in the Maya family life cycle: Cain, Caldwell and Chayanov revisited,” Population and Development Review 28: 475 99. Leibenstein, H. 1975 “The economic theory of fertility decline,” Quarterly Journal of Economics 89: 1 31. Lemaître, N. 1976 “Familles complexes en Bas Limousin: Ussel au début du xixe siècle,” Annales du Midi 127: 219 24. Levi, G. 1973 “Famiglie contadine nella Liguria del settecento,” Miscellanea Storica Ligure 5: 207 90. 1990 “Family and kin: a few thoughts,” Journal of Family History 15: 567 78.

Bibliography

235

Levy, M. J. 1965 “Aspects of the analysis of family structure,” in: A. Coale, L. A. Fallers, M. J. Levy, Jr., D. M. Schneider and S. S. Tomkins eds. Aspects of the Analysis of Family Structure. Princeton: 1 63. Lewis, N. 1970 “On paternal authority in Roman Egypt,” Revue International des Droits de l’Antiquité 17: 251 8. 1984 “The Romanity of Roman Egypt: a growing consensus,” in: Atti del XVII Congresso internazionale di papirologia. Naples: 1077 84. 1986 Greeks in Ptolemaic Egypt: Case Studies in the Social History of the Hellenistic World. Oxford. Liao, T. 2001 “Were past Chinese families complex? Household structures during the Tang Dynasty, 618 907 ad,” Continuity and Change 16: 331 55. Lindsay, H. 1998 9 “Adoption in Greek law: some comparisons with the Roman world,” Newcastle Law Review 91: 91 110. 2001 “Adoption and its function in cross cultural contexts,” in: S. Dixon ed. Childhood, Class and Kin in the Roman World. London and New York: 190 204. Lions, P. 1967 “Un dénombrement de la population de Brueil en Vexin en 1625,” Annales de Démographie Historique: 521 31. Lippert, S. 2008 Einfu¨hrung in die altägyptische Rechtsgeschichte. Einführungen und Quellentexte zur Ägyptologie 5. Berlin. Liu, W. T. and H. Kendig 2000 eds. Who Should Care for the Elderly? An East West Value Divide. Singapore. Livi Bacci, M. 1992 A Concise History of World Population. Oxford. Lo Cascio, E. 2006 “Did the population of imperial Rome reproduce itself?” in: G. R. Storey ed. Urbanism in the Preindustrial World: Cross-Cultural Approaches. Tuscaloosa: 52 68. 2009 “Urbanization as a proxy of demographic and economic growth,” in: A. Bowman and A. Wilson eds. Quantifying the Roman Economy: Methods and Problems. Oxford: 87 106. Lockwood, M. 1998 Fertility and Household Labour in Tanzania: Demography, Economy, and Society in Rufiji District, c.1870 1986. Oxford. Loefgren, O. 1984 “Family and household: images and realities: cultural change in Swedish society,” in: R. M. Netting, R. R. Wilk and E. J. Arnould eds. Households: Comparative and Historical Studies of the Domestic Group. Berkeley: 446 69. Loewenstein, K. 1973 The Governance of Rome. The Hague. Lorenz, C. 1999 “Comparative historiography: problems and perspectives,” History and Theory 38: 25 39. Luckhard, F. 1914 Das Privathaus im ptolemäischen und römischen Ägypten. Giessen. Lumpe, A. 1959 “Eltern,” in: Reallexikon fu¨r Antike und Christentum 2: 1192 4. Lüscher, K. 2002 “Intergenerational ambivalence: further steps in theory and research,” Journal of Marriage and Family 64: 585 93. Lüscher, K. and K. Pillemer 1998 “Intergenerational ambivalence: a new approach to the study of parent child relations in later life,” Journal of Marriage and Family 60: 413 25.

236

Bibliography

Lymer, A. and J. Hasseldine eds. 2002 The International Taxation System. Norwell, MA. Ma, X. 1988 “Changes in Chinese household size and structure,” in: X. Dixin ed. Population in Modern China. Beijing: 343 74. MacDowell, D. M. 1971 Aristophanes: The Wasps. Oxford. 1978 The Law in Classical Athens. Ithaca, NY. MacMullen, R. 1982 “The epigraphic habit in the Roman Empire,” American Journal of Philology 103: 233 46. Macrides, R. 1990 “Kinship by arrangement: the case of adoption,” Dumbarton Oaks Papers 44: 109 18. 1999 “Substitute parents and their children in Byzantium,” in: M. Corbier ed. Adoption et fosterage. Paris: 307 19. Maddison, A. 2001 The World Economy: A Millennial Perspective. Paris. Maehler, H. 1983 “Häuser und ihre Bewohner im Fayum in der Kaiserzeit,” Aegyptiaca Treverensia 2: 119 37. 2005 “Greek, Egyptian and Roman law,” Journal of Juristic Papyrology 35: 121 40. Makinwa Adebusoye, P. 1994 “Changes in the costs and benefits of children to their parents,” in: T. Locoh and V. Hertrich eds. The Onset of Fertility Transition in Sub-Saharan Africa. Liège. Mandelbaum, D. G. 1970 Society in India: Changes and Continuity. 2 vols. Berkeley. Mangen, D. J., V. L. Bengtson and P. H. Landry 1988 Measurement of Intergenerational Relations. London. Mann, S. 1987 “Widows in the kinship class and community structures of Qing dynasty China,” Journal of Asian Studies 46: 37 56. Martin, D. B. 1996 “The construction of the ancient family: methodological considerations,” Journal of Roman Studies 86: 41 60. Martin, J. 1984 “Zur Stellung des Vaters in antiken Gesellschaften,” in: H. Süssmuth ed. Historische Anthropologie: Der Mensch in der Geschichte. Göttingen: 84 109. Masciadri, M. M. and O. Montevecchi 1984 I contratti di baliatico. Milan. Mattioli, U. ed. 1995 Senectus: la vecchiaia nel mondo classico. Bologna. Mattson, I. 2005 “Adoption and fostering: overview,” in: S. Joseph and A. Najmabadi eds. Encyclopedia of Women and Islamic Cultures. Leiden: 1 3. Mayer, K. U. and S. Hillmert 2003 “Neue Flexibilitäten oder blockierte Gesellschaft? Sozialstruktur und Lebensverläufe in Deutschland 1960 2000,” in: Jahrbuch/Max-Planck-Gesellschaft zur Förderung der Wissenschaften. Munich: 74 100. Mayer, K. U. and W. Müller 1986 “The state and the structure of the life course,” in: A. B. Sorensen, F. E. Weinert and L. R. Sherrod eds. Human Development and the Life Course. Hillsdale, NJ: 217 45. McArdle, F. 1978 Altopascio: A Study in Tuscan Rural Society, 1587 1784. New York. McCarthy, J. 1979 “Age, family, and migration in nineteenth century Black Sea provinces of the Ottoman Empire,” International Journal of Middle East Studies 10: 309 23. McEvedy, C. and R. Jones 1978 Atlas of World Population History. London.

Bibliography

237

McKie, L., S. Bowlby and S. Gregory eds. 1999 Gender, Power and the Household. London. Meek, T. J. 1969 “The Code of Hammurabi,” in: J. Pritchard ed. Ancient Near Eastern Texts Relating to the Old Testament. Princeton: 163 80. Melaerts, H. and L. Mooren, eds. 2002 Le rôle et le statut de la femme en Égypte hellénistique, romaine et byzantine. Studia Hellenistica 17. Paris, Leuven and Sterling, VA. Meriwether, M. L. 1999 The Kin Who Count: Family and Society in Ottoman Aleppo, 1770 1840. Austin, TX. Metzger, H. 1974 Nachrichten aus dem Wu¨stensand: Eine Sammlung von Papyruszeugnissen. Zurich and Munich. Meyer, E. 1924 “Die Wirtschaftliche Entwicklung des Altertums,” in: Kleine Schriften. Vol. i. 2nd edn. Halle: 79 168. Meyer, E. A. 1990 “Explaining the epigraphic habit in the Roman Empire: the evidence of epitaphs,” Journal of Roman Studies 80: 74 96. Minois, G. 1989 History of Old Age: From Antiquity to the Renaissance. Cambridge and Oxford. Mirkovic, M. 2005 “Child labour and taxes in the agriculture of Roman Egypt: ΠΑΙΣ and ΑΦΗΛΙΞ,” Scripta Classica Israelica 24: 139 49. Mitteis, L. 1906 “Adoptionsurkunde vom Jahre 381 n.Chr.,” Archiv fu¨r Papyrusforschung und verwandte Gebiete 3: 173 84. Mitterauer, M. 1996 “Family contexts: the Balkans in European comparison,” History of the Family 1: 387 406. Mitterauer, M. and R. Sieder eds. 1982 Historische Familienforschung. Frankfurt am Main. Modrzejewski, J., 1955 6 “Droit de famille dans les lettres privées grecques d’Égypte,” Journal of Juristic Papyrology 9 10: 339 63. 2005 “Greek law in the Hellenistic period: family and marriage,” in: M. Gagarin and D. Cohen eds. The Cambridge Companion to Ancient Greek Law. New York: 343 56. Montevecchi, O. 1935 “Richerche di sociologia nei documenti dell’Egitto greco romano, i: i testamenti,” Aegyptus 15: 67 121. 1936 “Ricerche di sociologia nei documenti dell’Egitto greco romano, ii: i contratti di matrimonio e gli atti di divorzio,” Aegyptus 16: 3 83. 1941 “Ricerche di sociologia nei documenti dell’Egitto greco romano, iii: i contratti di compravendita,” Aegyptus 21: 91 151. 2000 “Gli apo gymnasiou de Lykon polis,” in: S. Russo ed. Atti del V Convegno nazionale di egittologia e papirologia, Firenze, 10 12 dicembre 1999. Florence: 175 84. Montserrat, D. 1996 Sex and Society in Graeco-Roman Egypt. London and New York. Moore, R. A. 1970 “Adoption and samurai mobility in Tokugawa Japan,” Journal of Asian Studies 29: 617 32. Moreau, P. 1978 “Plutarque, Augustin, Lévi Strauss: prohibition de l’inceste et mariage préférentiel dans la Rome primitive,” Revue Belge de Philologie et d’Histoire 56: 41 54. 2002 Incestus et prohibitae nuptiae: l’inceste à Rome. Paris.

238

Bibliography

Moring, B. 1998 “Family strategies, inheritance systems and the care of the elderly in historical perspective: eastern and western Finland,” Historical Social Research 23: 67 82. Morsy, S. A. 1978 “Sex differences and folk illness in an Egyptian village,” in: L. Beck and N. Keddie eds. Women in the Muslim World. Cambridge, MA: 599 616. Motomura, R. 1988 “The practice of exposing infants and its effects on the develop ment of slavery in the ancient world,” in: T. Yuge and M. Dot eds. Forms of Control and Subordination in Antiquity. Leiden and New York: 410 15. Muhs, B. 2008 “Fractions of houses in Ptolemaic Hawara,” in: S. Lippert and M. Schentuleit eds. Graeco-Roman Fayum: Texts and Archaeology. Wiesbaden: 187 97. Müller, E. 1976 “The economic value of children in peasant agriculture,” in: R. Ridker ed. Population and Development: The Search for Selective Interventions. Baltimore: 98 153. Murthi, M., A. C. Guio and J. Drèze 1995 “Mortality, fertility and gender bias in India: a district level analysis,” Population and Development Review 21: 745 82. Nair, J. 1996 Women and Law in Colonial India: A Social History. Bangalore. Nathan, G. S. 2000 The Family in Late Antiquity: The Rise of Christianity and the Endurance of Tradition. London and New York. Netting, R. M. 1984 “Introduction,” in: R. M. Netting, R. R. Wilk and E. J. Arnould eds. Households: Comparative and Historical Studies of the Domestic Group. Berkeley. Nevett, L. C. 1994 “Separation or seclusion? Towards an archaeological approach to investigating women in the Greek household in the fifth to third centuries bc,” in: M. Parker Pearson and C. Richards eds. Architecture and Order: Approaches to Social Space. London and New York: 98 112. 1999 House and Society in the Ancient Greek World. Cambridge and New York. 2010 Domestic Space in Classical Antiquity. Cambridge. Nielsen, H. S. 1987 “Alumnus: a term of relation denoting quasi adoption,” Classica et Mediaevalia 38: 141 88. 1999 “Quasi kin, quasi adoption and the Roman family,” in: M. Corbier ed. Adoption et fosterage. Paris: 249 62. Ogburn, W. E. 1933 “The family and its functions,” in: W. C. Mitchell ed. Recent Social Trends in the United States: Report of the President’s Research Committee on Social Trends. New York: 661 708. Okawara, T. 2003 “Size and structure of Damascus households in the late Ottoman period as compared with Istanbul households,” in: B. Doumani ed. Family History in the Middle East: Household, Property and Gender. Albany, NY: 51 75. Oppenheim Mason, K. 1992 “Family change and support of the elderly in Asia: what do we know?” Asia-Pacific Population Journal 7: 13 32. O’Roark, D. 1996 “Close kin marriages in late antiquity: the evidence of Chrysostom,” Greek, Roman and Byzantine Studies 37: 399 411. Osborne, R. 1991 “Pride and prejudice, sense and subsistence: exchange and society in the Greek city,” in: A. Wallace Hadrill and J. Love eds. City and Country in the Ancient World. London: 119 45.

Bibliography

239

Palmore, E. 1987 “Cross cultural perspectives on widowhood,” Journal of CrossCultural Gerontology 2: 93 105. Panter Brick, C. 1992 “Coping with an affected birth: genetic counseling in Saudi Arabia,” Journal of Child Neurology 7: 69 72. Papanek, H. 1990 “To each less than she needs, from each more than she can do: allocations, entitlements and value,” in: I. Tinkor ed. Persistent Inequalities: Women and World Development. New York and Oxford. Parker, R. 2005 “Law and religion,” in: M. Gagarin and D. Cohen eds. The Cambridge Companion to Ancient Greek Law. Cambridge: 61 81. Parker, S. 1996 “Full brother sister marriage in Roman Egypt: another look,” Cultural Anthropology 11: 362 76. Parkin, T. G. 1992 Demography and Roman Society. Baltimore and London. 1997 “Out of sight, out of mind: elderly members of the Roman family,” in: B. Rawson and P. R. C Weaver eds. The Roman Family in Italy: Status, Sentiment, Space. Oxford: 123 48. 2003 Old Age in the Roman World: A Cultural and Social History. Baltimore. Parkin, T. G. and A. J. Pomeroy eds. 2007 Roman Social History: A Sourcebook. London. Parkins, H. ed. 1997 Roman Urbanism: Beyond the Consumer City. London and New York. Parsons, T. 1944 “The social structure of the family,” in: R. N. Anshen ed. The Family: Its Function and Destiny. New York: 173 201. Pasternak, B. 1983 Guests in the Dragon: Social Demography of a Chinese District, 1895 1946. New York. 1985 “On the causes and demographic consequences of uxorilocal marriage in China,” in: S. B. Hanley and A. P. Wolf eds. Family and Population in East Asian History. Stanford, CA: 309 34. Patlagean, E. 1996 “Families and kinship in Byzantium,” in: A. Burguière, C. Klapisch Zuber, M. Segalen and F. Zonabend eds. A History of the Family. I . Distant Worlds, Ancient Worlds. Cambridge, MA: 467 88. Patterson, C. 1985 “‘Not worth the rearing’: the causes of infant exposure in ancient Greece,” Transactions of the American Philological Association 115: 103 23. 1998 The Family in Greek History. Cambridge. Paulissian, R. 1999 “Adoption in Ancient Assyria and Babylonia,” Journal of Assyrian Academic Studies 13: 5 34. Pelting, M. and R. Smith eds. 1991 Life, Death and the Elderly: Historical Perspectives. London. Perdue, L. G., J. Blenkinsopp, J. J. Collins and C. Meyers eds. 1997 Families in Ancient Israel: The Family, Religion, and Culture. Louisville, KY. Perez Fuentes Hernandez, P. and A. Pareja Alonso 1997 “Aging alone or in a family: the case of Bilbao, 1825 1935,” Continuity and Change 12: 79 102. Pestman, P. W. 1961 Marriage and Matrimonial Property in Ancient Egypt: A Contribution to Establishing the Legal Position of Women. Leiden. 1969 “The law of succession in ancient Egypt,” in: J. Brugman, M. David, F. R. Kraus, P. W. Pestman and M. H. van der Valk eds. Essays on Oriental Laws of Succession. Leiden: 58 77.

240

Bibliography

1987 “Inheriting in the archive of the Theban Choachytes (2nd cent. b.c.),” in: S. Vleeming ed. Aspects of Demotic Lexicography: Acts of the Second International Conference for Demotic Studies, Leiden, 19 21 September 1984. Studia Demotica 1. Leuven: 57 73. 1995 “Appearance and reality in written contracts: evidence from bilingual family archives,” in: M. Geller, H. Mähler and A. Lewis eds. Legal Documents of the Hellenistic World: Papers from a Seminar Arranged by the Institute of Classical Studies, the Institute of Jewish Studies and the Warburg Institute, University of London, February to May 1986. London: 79 87. Petersen, H. 1964 “The earliest Christian inscriptions of Egypt,” Classical Philology 59: 154 74. Petersen, K. K. 1968 “Demographic conditions and extended family households: Egyptian data,” Social Forces 46: 531 7. Petsas, P., M. Hatzopoulos, L. Gounaropoulou and P. Paschidis 2000 Inscriptions du sanctuaire de la Mère des Dieux autochtone de Leukopétra (Macédoine). Athens. Petzl, G. 1994 Die Beichtinschriften Westkleinasiens, Epigraphica Anatolica 22. Bonn. Pfisterer Haas, S. 1989 Darstellungen alter Frauen in der griechischen Kunst. Bern. 1990 “Ältere Frauen auf attischen Grabdenkmälern,” Mitteilungen des Deutschen Archäologischen Instituts 105: 179 96. Phang, S. E. 2001 The Marriage of Roman Soldiers, 13 BC AD 235: Law and Family in the Imperial Army. Leiden and Boston. Pillemer, K. and K. Lüscher eds. 2004 Intergenerational Ambivalences: New Perspectives on Parent Child Relations in Later Life. Amsterdam. Pillemer, K. and J. J. Suitor 2002 “Explaining mothers’ ambivalence toward their adult children,” Journal of Marriage and Family 64: 602 13. Plakans, A. 1984 Kinship in the Past: An Anthropology of European Family Life, 1500 1900. Oxford. 1989 “Stepping down in former times: a comparative assessment of retirement in traditional Europe,” in: D. I. Kertzer and K. W. Schaie eds. Age Structuring in Comparative Perspective. Hillsdale, NJ: 175 96. Poethke, G. 1981 “Der Papyrus Kodex BGU 1024 1027 aus Hermupolis Magna,” in: R. S. Bagnall, G. M. Browne and A. E. Hanson eds. Proceedings of the Sixteenth International Congress of Papyrology: New York, 24 31 July 1980. American Studies in Papyrology 23. Chico, CA: 457 62. Poma, G. 1972 “L’adozione a Rodi,” Epigraphica 34: 169 305. Pomeroy, S. B. 1976 “The relationship of the married woman to her blood relatives in Rome,” Ancient Society 7: 215 27. 1983 “Infanticide in Hellenistic Greece,” in: A. Cameron and A. Kuhrt eds. Images of Women in Antiquity. London: 207 22. 1984 Women in Hellenistic Egypt from Alexander to Cleopatra. New York. 1986 “Copronyms and the exposure of infants in Egypt,” in: R. S. Bagnall and W. V. Harris eds. Studies in Roman Law in Memory of A. A. Schiller. Leiden: 147 62. 1997 Families in Classical and Hellenistic Greece: Representations and Realities. Oxford and New York.

Bibliography

241

Poska, A. M. 2000 “Gender, property, and retirement strategies in early modern northwestern Spain,” Journal of Family History 25: 313 25. Potter, D. S. 2006 A Companion to the Roman Empire. Malden. Puliatti, S. 2001 Incesti crimina: regime guiridico da Augusto a Giustiniano. Milan. Quadagno, J. S. 1982 Aging in Early Industrial Society. New York. Rathbone, D. 1990 “Villages, land and population in Graeco Roman Egypt,” Proceedings of the Cambridge Philological Society NS 36: 103 42. 1991 Economic Rationalism and Rural Society in Third-Century A.D . Egypt. Cambridge. 1997 “Prices and price formation in Roman Egypt,” in: J. Andreau, P. Briant and R. Descat eds. Économie antique: prix et formation des prix dans les économies antiques. Saint Bertrand de Comminges: 183 244. Rawson, B. ed. 1986 The Family in Ancient Rome: New Perspectives. Ithaca, NY. ed. 1991 Marriage, Divorce, and Children in Ancient Rome. Oxford. 2003a Children and Childhood in Roman Italy. Oxford and New York. 2003b “The Roman family in recent research: state of the question,” Biblical Interpretation: A Journal of Contemporary Approaches 11: 119 38. 2005 “The future of childhood studies in classics and ancient history,” in: K. Mustakallio, J. Hanska, H. L. Sainio and V. Vuolanto eds. Hoping for Continuity: Childhood, Education and Death in Antiquity and the Middle Ages. Rome: 1 11. Rawson, B. and P. R. C. Weaver eds. 1997 The Roman Family in Italy: Status, Sentiment, Space. Canberra, New York and Oxford. Reher, D. S. 1998 “Family ties in Western Europe: persistent contrasts,” Population and Development Review 24: 203 34. Relihan, J. C. 2007 The Golden Ass, or, A Book of Changes. Indianapolis. Remijsen, S. and W. Clarysse 2008 “Incest or adoption? Brother sister marriage in Roman Egypt revisited,” Journal of Roman Studies 98: 53 61. Revillout, E. 1912 Les origines égyptiennes du droit civil romain. Paris. Rheubottom, D. 2000 Age, Marriage, and Politics in Fifteenth-Century Ragusa. New York. Richardson, B. E. 1933 Old Age among the Ancient Greeks. Baltimore. Ricl, M. 1995 “The appeal to divine justice in the Lydian confession inscriptions,” in: E. Schwertheim ed. Forschungen in Lydien. Asia Minor Studien 17. Bonn: 67 76. Ring, R. R. 1979 “Early medieval peasant households in central Italy,” Journal of Family History 4: 2 25. Roberts, R. E. L., L. N. Richards and V. L. Bengtson 1991 “Intergenerational solidarity in families: untangling the ties that bind,” Marriage & Family Review 16: 11 46. Robin, J. 1984 “Family care of the elderly in a nineteenth century Devonshire parish,” Ageing & Society 4: 505 16. Roda, S. 1979 “Il matrimonio fra cugini germani nella legislazione tardo imperiale,” Studia et Documenta Historiae et Iuris 45: 289 309. Rostad, A. 2002 “Confession or reconciliation? The narrative structure of the Lydian and Phrygian ‘Confession Inscriptions,’” Symbolae Osloenses 77: 145 64. Rostovtzeff, M. 1926 The Social and Economic History of the Roman Empire. Oxford.

242

Bibliography

Rowlandson, J. 1996 Landowners and Tenants in Roman Egypt: The Social Relations of Agriculture in the Oxyrhynchite Nome. Oxford. 1998 Women and Society in Greek and Roman Egypt: A Sourcebook. Cambridge. Rowlandson, J. and R. Takahashi 2009 “Brother sister marriage and inheritance strategies in Greco Roman Egypt,” Journal of Roman Studies 99: 104 39. Rubinstein, L. 1993 Adoption in IV . Century Athens. Copenhagen. Rubinstein, L. and L. Bjertrup 1991 “Adoption in Hellenistic and Roman Athens,” Classica et Mediaevalia 42: 139 51. Rugh, A. B. 1984 Family in Contemporary Egypt. Syracuse, NY. Rupprecht, H. A. 1998 “Die Sorge für die Älteren nach den Papyri,” in: M. Stol and S. P. Vleeming eds. The Care of the Elderly in the Ancient Near East. Leiden: 223 39. 2005 “Greek law in foreign surroundings: continuity and development,” in: M. Gagarin and D. Cohen eds. The Cambridge Companion to Ancient Greek Law. New York: 328 42. Ruschenbusch, E. 1990 “Bemerkungen zum Erbtochterrecht in den solonischen Gesetzen,” in: G. Neci and G. Thür eds. Symposium 1988: Vorträge zur griechischen und hellenistischen Rechtsgeschichte. Cologne and Vienna: 15 20. Sallares, R. 1991 The Ecology of the Ancient Greek World. London. 2002 Malaria and Rome: A History of Malaria in Ancient Italy. Oxford. Saller, R. P. 1987 “Men’s age at marriage and its consequences in the Roman family,” Classical Philology 82: 21 34. 1991 “Roman heirship strategies in principle and in practice,” in: D. J. Kertzer, R. P. Saller, eds. The Family in Italy from Antiquity to the Present. New Haven and London: 26 47. 1994 Patriarchy, Property and Death in the Roman Family. Cambridge. 2007 “Framing the debate over growth in the ancient economy,” in: J. G. Manning and I. Morris eds. The Ancient Economy. Stanford: 223 38. Saller, R. P. and B. D. Shaw 1984 “Tombstones and Roman family relations in the principate: soldiers, civilians and slaves,” Journal of Roman Studies 74: 124 56. Salomies, O. 1992 Adoptive and Polyonymous Nomenclature in the Roman Empire. Helsinki. Samuel, A. E. 1965 “The role of paramone clauses in ancient documents,” Journal of Juristic Papyrology 15: 221 311. Sangari, K. and S. Vaid 1996 “Institutions, beliefs and ideologies: widow immolation in contemporary Rajasthan,” in: K. Jayawardena and M. de Alwis eds. Embodied Violence: Communalising Women’s Sexuality in South Asia. Delhi: 240 96. Sant Cassia, P. and C. Bada 1992 The Making of the Modern Greek Family: Marriage and Exchange in Nineteenth-century Athens. Cambridge. Sartre, M. 2005 The Middle East under Rome. Cambridge, MA. Scheidel, W. 1994 “Libitina’s bitter gains: seasonal mortality and endemic disease in the ancient city of Rome,” Ancient Society 25: 151 75. 1996a “Brother sister and parent child marriage outside royal families in ancient Egypt and Iran: a challenge to the sociobiological view of incest avoidance,” Ethnology and Sociobiology 17: 319 40.

Bibliography

243

1996b Measuring Sex, Age and Death in the Roman Empire: Explorations in Ancient Demography. Ann Arbor, MI. 1996c “The most silent women of Greece and Rome: rural labor and women’s life in the ancient world (ii),” Greece & Rome 43: 1 10. 1997a “Brother sister marriage in Roman Egypt,” Journal of Biosocial Science 29: 361 71. 1997b “Quantifying the sources of slaves in the early Roman Empire,” Journal of Roman Studies 87: 156 69. 2001a Death on the Nile: Disease and the Demography of Roman Egypt. Leiden. 2001b “Roman age structure: evidence and models,” Journal of Roman Studies 91: 1 26. 2002 “A model of demographic and economic change in Roman Egypt after the Antonine plague,” Journal of Roman Archaeology 15: 91 114. 2005 “Real slave prices and the relative cost of slave labor in the Greco Roman world,” Ancient Society 35: 1 17. 2009 “The demographic background,” in: S. R. Huebner and D. M. Ratzan eds. Growing Up Fatherless in Antiquity. Cambridge: 31 40. 2011 “The Roman slave supply,” in: K. Bradley and P. Cartledge eds. The Cambridge World History of Slavery. I . The Ancient Mediterranean World. Cambridge: 287 310. 2012a “Age and health,” in: C. Riggs ed. Oxford Handbook of Roman Egypt. Oxford: 305 16. 2012b “Epigraphy and demography: birth, marriage, family, and death,” in: J. Davies and J. Wilkes eds. Epigraphy and the Historical Sciences. Proceedings of the British Academy 177. Oxford: 101 29. in press a “Disease and death,” in: P. Erdkamp ed. The Companion to Ancient Rome. Cambridge. in press b “Real wages in early economies: evidence for living standards from 1800 bce to 1300 ce.” Schellekens, J. 1993 “Wages, secondary workers, and fertility: a working class perspective of the fertility transition in England and Wales,” Journal of Family History 18: 1 17. Schmitz, W. 2004 Nachbarschaft und Dorfgemeinschaft im archaischen und klassischen Griechenland. Berlin. 2005 “Altersklassen in Sparta?” in: T. Schmitt, W. Schmitz and A. Winterling eds. Gegenwärtige Antike antike Gegenwarten: Kolloquium zum 60. Geburtstag von Rolf Rilinger. Munich: 105 26. 2007 Haus und Familie im antiken Griechenland. Munich. Schmitz, W. and H. H. von Prittwitz 2009 “Einführung und Demographie,” in: Alter in der Antike: die Blu¨te des Alters aber ist die Weisheit. Ausstellungskatalog Bonn. Mainz: 19 22. Schott, J. and A. Henley 1996 Culture, Religion and Childbearing in a Multiracial Society: A Handbook for Health Professionals. Oxford.

244

Bibliography

Schubart, W. 1912 Ein Jahrtausend am Nil: Briefe aus dem Altertum verdeutscht u. erklärt. Berlin. 1923 Ein Jahrtausend am Nil: Briefe aus dem Altertum verdeutscht u. erklärt. 2nd edn. Berlin. Schubert, P. ed. 2000 Vivre en Égypte gréco-romaine: une sélection de papyrus. Vevey. Segalen, M. 1986 Historical Anthropology of the Family. Cambridge. Seidl, E. 1962 Ptolemäische Rechtsgeschichte. Glückstadt and New York. 1965 “La preminente posizione successoria del figlio maggiore nel diritto dei papiri,” Istituto Lombardo (Rend. Lett.) 99: 185 92. 1966 “Die Unterhaltspflicht der Töchter und die Kaufehe in den Papyrusurkunden,” Atti dell’XI Congresso Internazionale di Papirologia, Milano, 2 8 settembre 1965. Milan: 149 55. 1968 Ägyptische Rechtsgeschichte der Saiten- und Perserzeit. Glückstadt. Severy, B. 2003 Augustus and the Family at the Birth of the Roman Empire. New York. Shaffer, J. W. 1982 Family and Farm: Agrarian Change and Household Organization in the Loire Valley, 1500 1900. Albany, NY. Shahd, L. S. 2003 An Investigation of the Phenomenon of Polygyny in Rural Egypt. Cairo. Shaw, B. D. 1984 “Latin funerary epigraphy and family life in the later Roman Empire,” Historia 33: 457 97. 1987 “The age of Roman girls at marriage: some reconsiderations,” Journal of Roman Studies 77: 30 46. 1992 “Explaining incest: brother sister marriage in Graeco Roman Egypt,” Man 27: 267 99. 1996 “Seasons of death: aspects of mortality in imperial Rome,” Journal of Roman Studies 86: 100 38. 2006 “Seasonal mortality in imperial Rome and the Mediterranean: three problem cases,” in: G. R. Storey ed. Urbanism in the Preindustrial World: Cross-Cultural Approaches. Tuscaloosa: 86 109. Shimizu, A. 1991 “On the notion of kinship,” Man 26: 377 403. Shultz, W. J. 1926 The Taxation of Inheritance. Boston. Sieder, R. 1987 Sozialgeschichte der Familie. Frankfurt am Main. Sieder, R. and M. Mitterauer 1983 “The reconstruction of the family life course: theoretical problems and empirical results,” in: R. Wall, J. Robin and P. Laslett eds. Family Forms in Historic Europe. Cambridge: 309 46. Siegel, J. S. and S. L. Hoover 1984 International Trends and Perspective: Aging. International Research Document 12. Washington, DC. Sijpesteijn, P. J. 1989 “Theognostos alias Moros and his family,” Zeitschrift fu¨r Papyrologie und Epigraphik 76: 213. Silverstein, M. and V. L. Bengtson 1997 “Intergenerational solidarity and the structure of adult child parent relationships in American families,” American Journal of Sociology 103: 429 60. Simmons, L. W. 1945 The Role of the Aged in Primitive Society. Hamden, CT. Skinner, G. W. 1997 “Family systems and demographic process,” in: D. I. Kertzer and T. Fricke eds. Anthropological Demography: Toward a New Synthesis. Chicago: 53 95.

Bibliography

245

Sommer, M. H. 2000 Sex, Law, and Society in Late Imperial China. Stanford. Sonbol, A. A. 1995 “Adoption in Islamic society: a historical survey,” in: E. W. Fernea ed. Children in the Muslim Middle East. Austin, TX: 45 67. Sorensen, A. B. 1989 “Old age, retirement and inheritance,” in: D. I. Kertzer and K. W. Schaie eds. Age Structuring in Comparative Perspective. Hillsdale, NJ: 197 214. Souvatzi, S. 2007 “The identification of Neolithic households: unfeasible or just disregarded?” in: R. C. Westgate, J. Whitley and N. R. E. Fisher eds. Building Communities: House, Settlement and Society in the Aegean and Beyond. London: 19 28. 2008 A Social Archaeology of Households in Neolithic Greece: An Anthropological Approach. Cambridge and New York. Stavrianopoulou, E. 1993 “Die Frauenadoption auf Rhodos,” Tyche 8: 177 88. Stearns, P. N. 1982 Old Age in Preindustrial Society. New York. Stecklov, G. 1999 “Evaluating the economic returns of childbearing in Côte d’Ivoire,” Population Studies 53: 15 32. Stol, M. 1995 “Private life in ancient Mesopotamia,” in: J. M. Sasson ed. Civilizations of the Ancient Near East. Vol. i. New York: 485 502. 1998 “The care of the elderly in Mesopotamia in the Old Babylonian period,” in: M. Stol and S. P. Vleeming eds. The Care of the Elderly in the Ancient Near East. Leiden: 59 117. Stol, M. and S. P. Vleeming eds. 1998 The Care of the Elderly in the Ancient Near East. Leiden. Stol, M. and F. A. M. Wiggermann 2000 Birth in Babylonia and the Bible: Its Mediterranean Setting. Groningen. Stone, E. C. 1991 “Adoption in Old Babylonian Nippur,” in: E. C. Stone and D. I. Owen, Adoption in Old Babylonian Nippur and the Archive of MannumMesu-Lissur. Mesopotamian Civilizations 3. Winona Lake, IN: 1 33. Stone, E. C. and D. I. Owen 1991 Adoption in Old Babylonian Nippur and the Archive of Mannum-Mesu-Lissur. Mesopotamian Civilizations 3. Winona Lake, IN. Stone, L. 2005 Kinship and Gender: An Introduction. 3rd edn. Boulder, CO. Straus, J. A. 1988 “L’esclavage dans l’Égypte romaine,” in: W. Haase and H. Temporini eds. Aufstieg und Niedergang der römischen Welt 2.10.1. Berlin: 841 911. Strauss, B. 1993 Fathers and Sons in Athens. Princeton. Tacoma, L. E. 2006 Fragile Hierarchies: The Urban Elites of Third-Century Roman Egypt. Leiden. Taubenschlag, R. 1916 “Die Patria potestas im Recht der Papyri,” Zeitschrift fu¨r Rechtsgeschichte 37: 177 230. 1936 “Die Alimentationspflicht im Rechte der Papyri,” in: Studi in onore di Salvatore Riccobono nel XL anno del suo insegnamento. Vol. i. Palermo: 505 18. 1955 The Law of Graeco-Roman Egypt in the Light of the Papyri, 332 BC 640 AD. 2nd edn. Warsaw.

246

Bibliography

1956 “La γηροκομία dans le droit des papyrus,” Revue Internationale des Droits de l’Antiquité 3: 173 9. 1959 Opera Minora. Vol. ii. Warsaw. Tenney, J. S. 2011 Life at the Bottom of Babylonian Society: Servile Laborers at Nippur in the 14th and 13th centuries B.C . Leiden and Boston. Therborn, G. 2004 Between Sex and Power: Family in the World, 1900 2000. London. Thierfelder, H. 1960 Die Geschwisterehe im hellenistisch-römischen Ägypten. Münster. Thomas, Y. P. 1980 “Mariages endogamiques à Rome: patrimoine, pouvoir et parenté depuis l’époque archaïque,” Revue Historique de Droit Français et Étranger 58: 345 52. Thompson, D. 2002 “Families in early Ptolemaic Egypt,” in: D. Ogden ed. The Hellenistic World: New Perspectives. London: 137 56. 2006 “The Hellenistic family,” in: G. R. Bugh ed. The Cambridge Companion to the Hellenistic World. Cambridge: 93 112. Thompson, W. 1967 “The marriage of first cousins in Athenian society,” Phoenix 21: 273 82. 1972 “Athenian marriage patterns: ‘remarriage,’” Californian Studies in Classical Antiquities 5: 211 24. Timmer, J. 2008 Altersgrenzen politischer Partizipation in antiken Gesellschaften. Berlin. Todd, S. and P. Millett, 1990 “Law, society and Athens,” in: P. Cartledge, P. Millett and S. Todd eds. Nomos: Essays in Athenian Law, Politics and Society. Cambridge: 1 18. Tönnies, F. 1957 Community and Society. East Lansing. First published in 1887 as Gemeinschaft und Gesellschaft. Leipzig. Torti, C. 1981 “Struttura e caratteri della famiglia contadina: Cascina 1841,” in: G. Cherubini et al. eds. Contadini e proprietari nella Toscana moderna: atti del Convegno di studi in onore di Giorgio Giorgetti. Florence: 173 201. Treas, J. and B. Logue 1986 “Economic development and the older population,” Population and Development Review 12: 645 73. Treggiari, S. 1991a “Divorce Roman style: how easy and how frequent was it?” in: B. Rawson ed. Marriage, Divorce and Children in Ancient Rome. Oxford: 31 46. 1991b Roman Marriage: Iusti Coniuges from the Time of Cicero to the Time of Ulpian. Oxford. Troyansky, D. 1996 “Progress report: the history of old age in the Western world,” Ageing & Society 16: 233 43. Tucker, J. E. 1985 Women in Nineteenth-Century Egypt. Cambridge. 1988 “Marriage and the family in Nablus, 1720 1856: toward a history of Arab marriage,” Journal of Family History 13: 165 79. 1993 “The Arab family in history: ‘otherness’ and the study of the family,” in: J. Tucker ed. Arab Women: Old Boundaries, New Frontiers. Bloomington: 195 207. United Nations 2002 Madrid International Plan of Action on Ageing, 2002, Adopted at the 10th Plenary Meeting of the Second World Assembly on Ageing, Madrid, 12 April 2002 [part 3 of 5 of the Report of the Second World Assembly on Ageing,

Bibliography

247

Madrid, 8 12 April 2002]. United Nations publication, Sales No. E.02.IV.4 [A/CONF.197/9]. New York. Van Beek, B. 2006 “The archive of the engineers Kleon and Theodoros”, unpub lished PhD, KU Leuven. van Bremen, R. 1996 The Limits of Participation: Women and Civic Life in the Greek East in the Hellenistic and Roman Periods. Amsterdam. Van Dam, R. 2003 Families and Friends in Late Roman Cappadocia. Philadelphia. van den Braembussche, A. A. 1989 “Historical explanation and comparative method: toward a theory of the history of society,” History and Theory 28: 1 24. van Driel, G. 1998 “The care of the elderly: the Neo Babylonian period,” in: M. Stol and S. P. Vleeming eds. The Care of the Elderly in the Ancient Near East. Leiden: 161 97. Vandorpe, K. 2002 The Bilingual Family Archive of Dryton, His Wife Apollonia, and Their Daughter Senmouthis. Collectanea Hellenistica 4. Brussels. Vatin, C. 1970 Recherches sur le mariage et la condition de la femme mariée à l’époque hellénistique. Paris. Veenhof, K. R. 1998 “Old Assyrian and Anatolian evidence for the care of the elderly,” in: M. Stol and S. P. Vleeming eds. The Care of the Elderly in the Ancient Near East. Leiden: 119 60. Vervier, B. 1984 “Putting kin and kinship to good use: the circulation of goods, labour and names on Karpathos (Greece),” in: H. Medick and D. W. Sabean eds. Interest and Emotion: Essays on the Study of Family and Kinship. Cambridge: 28 76. Viazzo, P. P. 1989 Upland Communities: Environment, Population and Social Structure in the Alps since the Sixteenth Century. Cambridge. 2003 “What’s so special about the Mediterranean? Thirty years of research on household and family in Italy,” Continuity and Change 18: 111 37. Viazzo, P. P. and D. Albera 1990 “The peasant family in northern Italy, 1750 1930: a reassessment,” Journal of Family History 15: 461 82. Vismara, G. 1958 “Adozione (diritto intermedio),” in Enciclopedia del Diritto. Vol. i. Milano. Vuolanto, V. 2002 “Women and the property of fatherless children in the Roman Empire,” in: R. Beig, R. Hälikkä, P. Raitis, and V. Vuolanto eds. Women, Power and Property in the Roman Empire. Rome: 203 43. 2003 “Selling a freeborn child: rhetoric and social realities in the Late Roman World,” Ancient Society 33: 169 207. Wachter, K. W. and P. Laslett 1978 “Measuring patriline extinction for modeling social mobility in the past,” in: K. W. Wachter with E. A. Hammel and P. Laslett eds. Statistical Studies of Historical Social Structure. New York: 113 36. Wahl, H. M. 1999 “Ester, das adoptierte Waisenkind: zur Adoption im Alten Testament,” Biblica 80: 78 99. Wall, R. 1995 “Elderly persons and members of their household in England and Wales from preindustrial times to the present,” in: D. I. Kertzer and P. Laslett eds. Aging in the Past: Demography, Society and Old Age. Berkeley: 81 106. Wall, R., J. Robin and P. Laslett eds. 1983 Family Forms in Historic Europe. Cambridge.

248

Bibliography

Wallace, S. L. R. 1938 Taxation in Egypt from Augustus to Diocletian. Princeton. Waltner, A. 1981 “Widows and remarriage in Ming and early Qing China,” in: R. W. Guisso and S. Johannesen eds. Women in China: Current Directions in Historical Scholarship. Youngstown, NY: 129 46. Wang, D. 2004 “Ritualistic coresidence and the weakening of filial practice in rural China,” in: C. Ikels ed. Filial Piety: Practice and Discourse in Contemporary East Asia. Stanford: 16 33. Waring, M. 2004 Counting for Nothing: What Men Value and What Women are Worth. Toronto. Waterson, R. 1995 “Houses and hierarchies in island Southeast Asia,” in: J. Carsten and S. Hugh Jones eds. About the House: Lévi-Strauss and Beyond. Cambridge: 47 68. Watson, P. 1995 Ancient Stepmothers: Myth, Misogyny and Reality. Leiden. Weinberger Thomas, C. 2000 Ashes of Immortality: Widow-Burning in India. Chicago. Wenger, G. C. 1997 “Review of findings on support networks of older Europeans,” Journal of Cross-Cultural Gerontology 12: 1 21. Wenger, L. 1904 “Strafprozesse vor dem römischen Statthalter in Ägypten,” in: Archiv fu¨r Kriminal-Anthropologie und Kriminalistik 16: 304 23. Wente, E. 1990 Letters from Ancient Egypt. Atlanta. West, M. 1893 The Inheritance Tax. New York. Westbrook, R. 1998 “Legal aspects of care of the elderly in the ancient Near East: introduction,” in: M. Stol and S. P. Vleeming eds. The Care of the Elderly in the Ancient Near East. Leiden and Boston: 1 22. Westermann, W. L. 1948 “The paramone as a general service contract,” Journal of Juristic Papyrology 2: 9 44. Wheaton, R. 1975 “Family and kinship in Western Europe: the problem of the joint family household,” Journal of Interdisciplinary History 4: 601 28. Whitehorn, J. E. G. 1979 “Sex and society in Greco Roman Egypt,” in: J. Bingen and G. Nachtergael eds. Actes du XV e Congrès International de Papyrologie. Vol. iv. Brussels Louvain: 240 6. Wiedemann, T. 1989 Adults and Children in the Roman Empire. New Haven and London. Wilcken, U. 1899 Griechische Ostraka aus Ägypten und Nubien: ein Beitrag zur antiken Wirtschaftsgeschichte. Vol. i. Leipzig. Wildfang, R. L. 2006 Rome’s Vestal Virgins. Oxford. Wilfong, T. G. 2002 Women of Jeme: Lives in a Coptic Town in Late Antique Egypt. Ann Arbor. Wilk, R. 1991 Household Ecology: Economic Change and Domestic Life among the Kekchi Maya of Belize. Tucson. Wilk, R. and R. Netting 1984 “Households: changing forms and functions,” in: R. Netting, R. Wilk and E. Arnould eds. Households: Comparative and Historical Studies of the Domestic Group. Berkeley: 1 28. Willis, R. J. 1982 “The direction of intergenerational transfers and demographic transition: the Caldwell hypothesis reexamined,” Population and Development Review 8: 207 34.

Bibliography

249

Wilson, A. 1980 “The infancy of the history of childhood: an appraisal of Philippe Ariès,” History and Theory 19: 132 53. Wilson, C. and P. Airey 1999 “How can a homeostatic perspective enhance demographic transition theory?” Population Studies 53: 117 28. Winter, J. G. 1933 Life and Letters in the Papyri. Ann Arbor. Wolf, A. P. 1968 “Adopt a daughter in law, marry a sister: a Chinese solution to the problem of the incest taboo,” American Anthropologist, ns 70: 864 74. 1981 “Women, widowhood, and fertility in pre modern China,” in: J. Dupâquier, E. Hélin, P. Laslett, M. Livi Bacci and S. Sogner eds. Marriage and Remarriage in Populations of the Past. London, New York, Toronto, Sydney and San Francisco: 139 47. 1995 Sexual Attraction and Childhood Association: A Chinese Brief for Edward Westermarck. Stanford. Wolf, A. P. and C. Huang 1980 Marriage and Adoption in China, 1845 1945. Stanford. Wolf, M. 1972 Women and the Family in Rural Taiwan. Stanford. Wong, O. M. H. 2000 “Children and children in law as primary caregivers,” in: W. T. Liu and H. Kendig eds. Who Should Care for the Elderly? An East West Value Divide. Singapore: 297 321. Woods, R. 2007 “Ancient and early modern mortality: experience and under standing,” Economic History Review 60: 373 99. Wrigley, E. A. 1978 “Fertility strategy for the individual and the group,” in: C. Tilly ed. Historical Studies of Changing Fertility. Princeton: 135 54. 1982 “The prospects for population history,” in: T. K. Rabb and R. I. Rotberg eds. The New History: The 1980s and Beyond. Princeton: 207 26. Xiaonan, D. 1999 “Women in Turfan during the sixth to eighth centuries: a look at their activities outside the home,” The Journal of Asian Studies 58: 87. Xie, W. 1997 “Family changes in rural and urban China, 1950s to 1980s: a multilevel model analysis,” unpublished PhD dissertation, Catholic University of America, Washington, DC. Yanagisako, S. J. and J. F. Collier 1987 “Toward a unified analysis of gender and kinship,” in: J. F. Collier and S. J. Yanagisako eds. Gender and Kinship: Essays toward a Unified Analysis. Stanford: 14 50. Yang, C. K. 1959 A Chinese Village in Early Communist Transition. Cambridge. Yiftach Firanko, U. 2003 Marriage and Marital Arrangements: A History of the Greek Marriage Document in Egypt, 4th Century BCE 4th Century CE. Münchener Beiträge zur Papyrusforschung und antiken Rechtsgeschichte 93. Munich. 2006 “Spouses in wills: a diachronic survey (iii bc iv ad),” Journal of Juristic Papyrology 36: 153 66. 2009 “Law in Greco Roman Egypt: Hellenization, fusion, Romanization,” in: R. Bagnall ed. The Oxford Handbook of Papyrology. Oxford: 541 60. Youni, M. 2000 Provincia Macedonia. Athens and Komotini. Yount, K. M. 2005 “Women’s family power and gender preference in Minya, Egypt,” Journal of Marriage and Family 67: 410 28. Zelnick Abramovitz, R. 2005 Not Wholly Free: The Concept of Manumission and the Status of Manumitted Slaves in the Ancient Greek World. Leiden and New York.

Index of sources

Old Testament Gen. 29 157 Gen. 31 182, 188 Num. 1:1–4 22 Ruth 1:20–21 155 2 Sam. 24:1–9 22 1 Chr. 21:1–6 22 1 Chr. 23.3 22 1 Chr. 27:24 22 2 Chr. 2:17 22 Esther 2:7 104 New Testament Matt. 19 164 Luke 2:1–3 22 Literary Sources Aeschin. 1.28 82 Aeschin. 3.251 133 Ambros.Vid. 6 1, 155 Anth. Pal. 7.261 84, 164 Apost. Const. 4.1 188 Apul. Met. 4 157 Aristoph. Cl. 844–5 133 Aristoph.Wasps 133 Aristot. Ath. Pol. 55 81, 133 Athen. 6.103 272C 22 Aug. Serm. 14.1 95 Cic. Caec. 34 23 Cic. Off. 1.53 16 Cic. Senec. 7.22 3, 133 Cic. Ver. 2.1.153 95 Dem. 27 93, 94 Dem. 28 94 Dem. 29 93, 94 Dem. 30 94 Dem. 36 94, 100 Dem. 41 88, 188 Dem. 47 132 Dem. 57 94

Ps.-Dem. Or. 41 179 Ps.-Dem. Or. 43 100, 179 Ps.-Dem. Or. 44 179, 180 Ps.-Dem. Or. 53 180 Ps.-Dem. Or. 54 180 Ps.-Dem. Or. 55 180 Dio Chrys. Or. 7.10 51 Diod. 1.27 192 Diod. 1.31 21 Diog.Laert. 1.55 82 Epistle of Aristeas 248 75 Eur. Alc. 735 163 Eur. Andr. 1081 163 Eur. Ion 618–20 163 Eur. Med. 1030 163 Eur. Med. 1394 163 Eur. Tro. 1185 163 Gr. Nys. Vit. Macr. 6.11 93 Hdt. 2.35 83 Hes.Th. 602–7 163 Hom. Il. 5 163 Hom.Od. 19 150 Hor. S. 1.6 76 Isaeus 1 100, 171, 179 Isaeus 2 87, 94, 168, 179, 180 Isaeus 3 179, 188 Isaeus 4 163, 179 Isaeus 5 179, 180 Isaeus 6 133, 179, 180 Isaeus 7 157, 179, 180, 186 Isaeus 8 82, 100, 157 Isaeus 9 100, 179, 180 Isaeus 10 178, 179, 180, 188 Isaeus 11 179, 180 Jo. Chrys. Hom. Ep. I Corinth. 34 158 Jo. Chrys. Sacerd. 1.2 93 Joseph. Bell. Jud. 2 21 Joseph. Bell. Jud. 3 21 Lact. De mort. pers. 23.1–4 24 Lib. Ep. 25 105

250

Index of sources Lib. Or. 1.26 93 Liv. 3.3 23 Liv. 4.8 23 Liv. 4.22 23 Liv. 4.42 23 Lucian, Som. 1–4 75 Menand. Asp. 258–69, 179 Paus. 1.7 192 Philo, De spec. leg. 2.233 75 Philo, De spec. leg. 3.23 192 Plato, Laws 776a–b 143 Plut. Aem. P. 5 20, 34, 35 Plut. Crass.1 20, 34 Plut. Mor. 138a–146a 142 Plut. Mor. 143a–c 143 Plut. Mor. 289 d–e 158 Plut. Mor. 479f 82 Plut. Mor. 492a–b (De fraterno amore) 34, 35 Plut. Sol. 22 75, 89 Ps.-Plut. Mor. 838a 178 Ps.-Plut. Mor. 839b 178 Ps.-Plut. Vit. dec. Orat. 847c 157 Quint. Inst. Orat. 6 82 Quint. Inst. Orat. 7 82, 197 Ps.-Quint. Declam. Min. 278 197 Sen. Ben. 3.38.2 82 Sen. Ben. 5.5.2 95 Sen. De ira 2.21.2–6 82 Sen. Helv. Matr. de consol. 2.4 82 Seneca, Contr. 9.3 197 Thdt. Ep. 8 158 Val. Max. 4.4 34 Val. Max. 7.8 90 Varro, De ling. Lat. 6.86–7 23 Varro, Rust. 1.17 62 Varro, Rust. 2.10 62 Varro, Rust. 3.2 23 Xen. Equest. 2.2 77 Papyri Bagnall and Frier 2006 11-Ar-1 110, 111, 115, 127 33-Ar-1 128 33-Ar-2 95, 97, 125 33-Ox-1 125 33-Ox-2 125 59-Ar-26 128 61-Ar-1 126 75-Ar-2 125 75-Ox-1 110, 111, 132 89-Ar-1 104, 128 89-Hm-1 117, 125, 126 103-Ar-1 97, 125, 126 103-Ar-3 101, 102, 125, 126, 160

251

103-Ar-5 125 103-Ar-7 128 103-Ar-9 95, 97, 125 103-Ar-11 125 103-Ar-12 104 117-Ap-5 101, 102, 110, 111, 113, 125, 126, 128, 132, 145 117-Ap-6 43, 101, 102, 125, 126, 128, 145 117-Ap-7 94, 125, 126, 128, 169 117-Ap-8 126 117-Ap-10 125 117-Ar-1 47, 101, 102, 110, 115, 121, 127, 145, 160, 167, 172 117-Ar-2 125, 126, 176 117-Ar-4 125, 128 117-Ar-5 95, 97, 111, 125 117-Ar-6 25, 97, 110, 111, 115, 125, 127, 145, 146 117-Ar-7 95, 97, 115, 125, 127, 128, 145 117-Ar-11 101, 102, 110, 111, 125, 127, 128, 145 117-Ar-12 97, 98, 101, 102, 110, 125, 128, 131, 145, 146 117-Ar-13 101, 102, 125 117-Ox-1 128 117-Ox-2 110, 125, 132 131-Ar-3 47, 125, 126 131-Ar-8 110, 132, 167 131-Ar-11 104, 115 131-Ar-12 97, 98, 101, 102, 110, 115, 125, 128, 131, 145, 146 131-Ar-13 97, 115, 125, 131 131-Ar-14 110, 132 131-Ar-18 101, 102, 110, 111, 125 131-Ar-20 125 131-Be-1 110, 125, 132 131-He-2 97, 110, 112, 115, 117, 125, 126, 127, 145 131-He-3 110, 112, 117, 118, 127, 132 131-He-4 110, 117, 126, 132, 176 131-Me-1 95, 125, 126, 130 131-Oa-1 110, 111, 126 131-Ox-1 101, 110, 125, 126, 128, 145 131-Ox-6 126, 168 131-Ox-16 125, 128 131-Pr-1 43, 117, 126, 128, 176, 178 145-Ar-1 78, 97 145-Ar-2 78, 95, 97, 110, 125 145-Ar-3 97, 101, 102, 110, 115, 125, 126, 127, 145, 146 145-Ar-4 78 145-Ar-9 110, 111, 125, 126, 145, 160 145-Ar-12 78, 96, 101, 102, 110, 126, 132 145-Ar-17 125 145-Ar-18 128 145-Ar-19 128 145-Ar-20 97, 104, 127, 145, 160

252

Index of sources

Bagnall and Frier 2006 (cont.) 145-Ar-22 97, 98, 115, 125 145-Ar-23 125, 126 145-He-1 125 145-He-2 95, 125, 126, 127, 145 145-Ly-1 97, 101, 102, 104, 115, 126 145-Oa-1 110, 117, 126 145-Oa-2 126 145-Ox-1 126 145-Ox-2 110, 117, 132 145-Pr-1 43, 101, 102, 117, 178 159-Ar-1 110, 117, 125, 132, 145 159-Ar-4 101, 102, 128, 130 159-Ar-5 101, 126, 128 159-Ar-7 126 159-Ar-10 97, 102, 104, 115, 126, 127, 128, 145 159-Ar-11 101, 111, 125, 127, 145, 160 159-Ar-13 104 159-Ar-20 126 159-Ar-21 126, 128 159-Ar-26 125 159-Hm-3 78, 101, 102, 110, 125, 132, 176 173-Ar-3 125 173-Ar-7 117 173-Ar-9 97, 110, 115, 125, 127, 132, 145, 160 173-Ar-11 95, 97, 101, 104, 126, 127 173-Ar-15 125 173-Me-1 117 173-Me-3 110, 111, 113, 117, 127, 132, 145, 189 173-Pr-3 101, 126, 128, 168 173-Pr-4 95, 97 173-Pr-5 101, 102, 110, 111, 113, 125, 126, 127, 128, 132, 145, 176 173-Pr-7 125, 128 173-Pr-8 110, 114, 167 173-Pr-10 126, 128 173-Pr-11 97, 115, 126 173-Pr-13 104, 117, 126 173-Pr-14 97, 98, 102, 110, 112, 115 173-Pr-15 97, 98, 115, 126, 128, 145, 146, 176 173-Pr-16 125 173-Pr-17 97, 110, 115, 127 187-An-2 159 187-Ar-2 126 187-Ar-4 94, 97, 101, 102, 126, 127, 128, 145, 160, 176 187-Ar-5 110, 114 187-Ar-7 104 187-Ar-8 101, 102, 104, 110, 111, 113, 126, 127, 132, 145, 160 187-Ar-9 115, 126, 132 187-Ar-10 102 187-Ar-11 126

187-Ar-12 126 187-Ar-18 110, 117 187-Ar-21 102 187-Ar-22 95, 101, 110, 111, 113, 126, 127, 132, 145, 160 187-Ar-23 126 187-Ar-26 104, 127 187-Ar-27 41 187-Ar-28 117 187-Ar-29 78, 95, 97 187-Ar-30 110, 114, 167, 173 187-Ar-32 95 187-Ar-34 97 187-Ar-39 97, 101, 102 187-Hm-1 25, 125, 176 187-Me-1 97, 102, 115 187-Ox-4 95, 97 201-Ar-1 97, 98, 102, 115, 126, 128, 131, 145, 146 201-Ar-5 126 201-Ar-6 102 201-Ar-8 97, 101, 115, 128 201-Ar-9 97, 98, 101, 102, 110, 111, 112, 115, 126, 128, 131, 168 201-Ar-10 110, 112, 117, 118, 126, 128, 159, 160 201-Ar-14 102 201-Ox-1 101 215-Ar-1 97, 110, 115, 125 215-Ar-2 101, 102 215-Ar-3 101, 102, 104, 126, 128 215-Ar-4 110, 113, 127, 167 215-Ar-5 97, 115, 125, 127, 128 215-Ar-6 128 215-Ar-9 104 215-He-1 76, 126, 167 215-He-2 101, 126, 128 215-He-3 110, 111, 113, 128, 132, 145 215-Hm-3 110, 114, 117, 167, 173 229-Hm-2 104 243-Ar-1 95, 97, 126 243-Ar-3 95, 97, 126 243-Ar-4 97, 126 257-Ar-1 95, 97, 126 ???-Ar-2 97 ???-Ar-3 25, 101, 102, 110, 111, 126, 128, 148 ???-Ar-5 126, 128 ???-Me-1 126 P.Oxy.Census 89-Pt-1 125 89-Pt-5 95 89-Pt-7 125 89-Pt-8 125 89-Pt-10 125

Index of sources 89-Pt-13 125 89-Pt-14 125 89-Pt-16 125 89-Pt-19 125 89-Pt-21 126 89-Pt-24 126 89-Pt-27 97, 128 89-Pt-28 97 89-Pt-33 125, 126 89-Pt-35 125 89-Pt-36 97, 125, 127 89-Pt-37 97, 125 89-Pt-42 97, 125, 126 89-Pt-44 127 89-Pt-47 125, 128 89-Pt-51 95, 97, 125 89-Pt-56 125 89-Pt-60 125 89-Pt-63 126 BGU 1.76 133 BGU 1.86 136, 137 BGU 1.319 106 BGU 2.380 69, 165 BGU 2.388 104, 105 BGU 2.615 121 BGU 3.801 89 BGU 3.948 121, 165 BGU 3.970 104, 105 BGU 3.998 133 BGU 4.1013 136, 137 BGU 4.1024 116 BGU 4.1124 77 BGU 5.1210 194 BGU 7.1578 83, 84, 166 BM 10593 151 Chr.Mitt. 306 98 Codex Hermupolis 8.30–9.4 99 Codex Hermupolis 8.30–1 50 Codex Hermupolis 9.9–11 51 CPR 1.4 133 CPR 1.188 133 CPR 4.23 76 CPR 6.1 98 M.Chr. 88 104, 105 O.Brit Mus.Copt.Ad. 23 198 O.Florida 14 148, 175 O.Max.inv. 267 148 P.Alex.553 22 P.Bad. 4.48 151 P.Bas. 7 133 P.Benaki 4 175 P.Berenike 2.129 83, 164 P.Berl.Zill. 4 106 P.Brem. 63 149

P.Cair.Goodsp. 6 135 P.Cair.Isid. 62 104, 105 P.Cair.Isid. 113 106 P.Cair.Masp. 1.67005 106 P.Cair.Masp. 1.67028 106 P.Cair.Masp. 1.67121 106 P.Cair.Masp. 2.67153 106 P.Cair.Masp. 2.67154 R 106 P.Cair.Masp. 2.67154 V 138 P.Cair.Masp. 2.67156 98 P.Cair.Masp. 3.67305 106 P.Cair.Masp. 3.67311 106 P.Cair.Masp. 3.67314 83 P.Cair.Masp. 3.67340 106 P.Cair.Masp. 3.67353 V 90 P.Cair.Masp. 3.7305 79 P.Cair.Zen. 1.59028 116 P.Col. 7.181 106 P.Col.Zen. 1.58 181 P.Coll.Youtie 2.67 104 P.Coll.Youtie 2.83 85 P.Diog. 9 98 P.Diog. 10 170 P.Dura 12 181, 183 P.Eleph. 2 135, 136 P.Enteux.25 13, 86 P.Enteux. 26 13, 86, 117 P.Erl. 28 182 P.Fam.Tebt. 33 80 P.Fay. 124 136, 137 P.Fay. 127 121 P.Flor. 1.93 106 P.Flor. 3.312 13 P.Flor. 3.383 83 P.Frankf. 5 22 P.Fuad. 1 Univ. App. II 321 181 P.Giss. 1.20 149 P.Giss. 1.21 149 P.Giss. 1.24 149 P.Giss. 1.30 133 P.Giss. 1.78 149 P.Grenf. 136, 137 P.Heid. 4.329 181 P.Herm. 31 106 P.Köln 7.321 182, 183 P.KRU 116 138 P.Leid.Inst. 8 169 P.Lille 1.27 22 P.Lips. 1.10 181 P.Lips. 1.28 74, 105, 160, 178, 182, 183, 186, 192 P.Lips. 1.29 147 P.Lond. 3.904 24, 151 P.Lond. 3.932 88, 136, 137, 165, 166, 171

253

254 P.Lond. 3.935 171 P.Lond. 3.940 171 P.Lond. 3.1044 136 P.Lond. 4.1422 182 P.Lond. 5.1709 106 P.Lond. 5.1712 106 P.Lond. 5.1731 79 P.Lond. 5.1857 88 P.Lond. 6.1717 106 P.Lond. 7.1976 116 P.Mert. 1.18 181 P.Mich. 3.170 76 P.Mich. 3.171 76 P.Mich. 3.172 76 P.Mich. 3.203 118, 165 P.Mich. 5.321 136 P.Mich. 5.346a 77 P.Mich. 8.468 118 P.Mich. 8.490 119, 165 P.Mich. 14.676 181 P.Mich.inv. 121 122 P.Mich.inv. 5191a 77 P.Mu¨nch. 1.6 85 P.Mu¨nch. 1.8 88 P.Mu¨nch. 3.101 62 P.Oslo 2.32 111 P.Oslo 3.114 181, 182 P.Oslo 3.124 13 P.Oxy. 1.358 153 P.Oxy. 1.46 181, 182 P.Oxy. 1.71 165 P.Oxy. 1.119 67 P.Oxy. 2.237 133, 153 P.Oxy. 2.257 181 P.Oxy. 2.265 79 P.Oxy. 2.275 77 P.Oxy. 2.281 153, 195 P.Oxy. 2.294 84, 165 P.Oxy. 2.295 122 P.Oxy. 3.472 136, 137 P.Oxy. 3.490 198 P.Oxy. 3.492 181, 182 P.Oxy. 3.497 166 P.Oxy. 3.502 181, 182, 186 P.Oxy. 3.504 181, 182, 183 P.Oxy. 3.531 68 P.Oxy. 3.583 178, 186 P.Oxy. 3.603 166 P.Oxy. 4.713 133 P.Oxy. 4.716 57 P.Oxy. 4.744 152 P.Oxy. 5.889 13 P.Oxy. 6.903 153 P.Oxy. 6.904 169

Index of sources P.Oxy. 6.905 166 P.Oxy. 6.906 79 P.Oxy. 6.907 123 P.Oxy. 6.934 175 P.Oxy. 6.938 68 P.Oxy. 7.732 181 P.Oxy. 8.1121 84, 88 P.Oxy. 8.1123 181, 182 P.Oxy. 8.1148 67 P.Oxy. 8.1160 68 P.Oxy. 9.1206 74, 182, 183, 186, 187 P.Oxy. 9.1208 57 P.Oxy. 9.1210 83, 85, 165 P.Oxy. 10.1266 181 P.Oxy. 10.1272 41 P.Oxy. 10.1299 68 P.Oxy. 11.1381 69 P.Oxy. 12.1415 124 P.Oxy. 12.1477 169 P.Oxy. 12.1481 69 P.Oxy. 12.1484 129 P.Oxy. 12.1487 129 P.Oxy. 12.1493 69 P.Oxy. 14.1638 56 P.Oxy. 14.1647 77 P.Oxy. 14.1665 67 P.Oxy. 14.1679 121 P.Oxy. 14.1719 181 P.Oxy.14.1721 181 P.Oxy. 16.1895 182, 183, 184, 186, 187, 197 P.Oxy. 18.2186 181 P.Oxy. 27.2474 171 P.Oxy. 31.2583 181 P.Oxy. 31.2586 77 P.Oxy. 34.2708 172 P.Oxy. 34.2711 104 P.Oxy. 34.2713 51, 55, 99, 102, 104 P.Oxy. 34.2731 148 P.Oxy. 38.2832 169 P.Oxy. 38.2833 169 P.Oxy. 42.3015 133 P.Oxy. 43.3139 106 P.Oxy. 46 3271 182 P.Oxy. 47.3330 169 P.Oxy. 48.3403 153 P.Oxy. 49.3491 166 P.Oxy. 50.3555 173 P.Oxy. 50.3593 182, 198 P.Oxy. 64.4433 170 P.Oxy. 66.4531 181 P.Oxy. 67.4596 77, 78 P.Oxy. 74.4990 181 P.Oxy.Hels. 36 181 P.Panop. 28 166

Index of sources P.Par. 21ter 76 P.Petaus 29 147 P.Petr. 1.13 170 P.Petr. 2.11 70 P.Petr. 2.13(19) 71 P.Petr. 2.4(7) 71 P.Petr. 2.42(a) 72 P.Petr.2 1.3 134 P.Pisentius 26 159 P.Prag 1.42 106 P.Ross.Georg. 3.28 196 P.Sakaon 36 168 P.Sakaon 38 106 P.Sakaon 41 169 P.Strasb. 1.4 181, 182 P.Strasb. 7.666 79 P.Tebt. 1.104 151, 153 P.Tebt. 2.385 62 P.Tebt. 2.480 159 P.Tebt. 3.814 22 P.Wash.Univ. 1.2 181 P.Wisc. 1.4 76 P.Wisc. 1.5 77 P.Wisc.inv. 12 122 PSI 3.177 69 PSI 4.294 182 PSI 5.457 181 PSI 5.484 13 PSI 7.732 181 SB 1.4503 76 SB 1.4504 76 SB 1.4505 76 SB 1.4845 83 SB 1.5285 76 SB 1.5286 76 SB 3.6262 68 SB 3.6263 84, 120, 129, 165 SB 3.6264 146 SB 5.7572 122, 175 SB 5.7871 186 SB 6.9065 104, 105 SB 6.9374 62 SB 8.9642 98 SB 10.10239 153 SB 14.11271 181 SB 14.11337 181 SB 14.11646 89 SB 14.11881 104 SB 16.12226 106 SB 16.12265 106 SB 16.12742 80 SB 18.13168 170 SB 18.13176 181 SB 20.14395 181

SPP 1, pp. 6–7 83, 170 SPP 12.40 77 UPZ 1.1 88 UPZ 1.59 152 UPZ 1.60 153 UPZ 1.118 133 UPZ 1.148 119 W.Chr. 100 146 W.Chrest. 122 67 W.Chrest. 198 22 Inscriptions CIL 6.1527 164 CIL 6.10230 90 CIL 6.10299 90 CIL 8.1641 80 CIL 9.2845–6 158 CIL 14.2298 79 I.Métr. 83 84, 86, 164 MAMA 1.232 188 RECAM 2.303 188 SEG 6.137 188 SEG 6.139 188 SEG 30.596 188 TAM 2.1–3 188 Legal Sources CIust 2.18 89 CIust 3.28 90 CIust 5.4 158, 195, 197 CIust 5.24 78 CIust 5.25 66, 82 CIust 6.25 158 CIust 6.61 133 CIust 7.33 186 CIust 7.47 74 CIust 8.46 82 Code of Hammurabi §185 185 Code of Hammurabi §186 185 Code of Hammurabi §188 185 Code of Hammurabi §189 76, 185 Code of Hammurabi §190 185 Code of Hammurabi §200B 76 CTh 3.10 158 CTh 3.12 158 CTh 11.27 80 Dig. 1.18.4 (Ulpian) 82 Dig. 5.2.2–4 (Marcian) 90 Dig. 7.2.8 (Ulpian) 98 Dig. 23.2.3 (Paulus) 158 Dig. 25.3.5.10 (Ulpian) 82 Dig. 25.3.5.14 (Ulpian) 78 Dig. 25.4.1.1 (Ulpian) 78

255

256 Dig. 27.10.4 (Ulpian) 82 Dig. 28.7.23 (Marcellus) 158 Dig. 28.7.24 (Papinian) 158 Dig. 33.2.37 (Scaevola) 98 Dig. 35.1.62.2 (Clementius) 98 Dig. 35.2.1.pr. (Paulus) 90 Dig. 37.15.1 (Ulpian) 82 Dig. 37.15.2 (Ulpian) 82 Dig. 45.1.107 (Iavolen) 197

Index of sources Dig. 48.2.11 (Macer) 82 Dig. 48.9.1 (Marcian) 82 Dig. 48.9.9 (Modestin) 82 Gaius, Inst. 1.104 186 Gaius, Inst. 1.97 186 Inst. 1.10.4 158 NovIust 117 133 NovIust 155 82 Rules of Ulpian 14 94

Index

abuse, 66, 103, 144, 153, 207, 219 adoption, 15, 19, 34, 48, 59, 73, 105, 168, 175–96, 199, 211 adoptive parents, 74 contracts, 159, 176, 178, 181, 184, 191, 195 of a daughter-in-law, 155, 186 of a son-in-law, 185, 187–96 of females, 186, 212 affection, 4, 9, 19, 65, 66, 69, 70, 114, 119, 120, 122, 142, 154, 155, 180, 187, 197 Africa, 3, 25, 59, 60, 108 aging, history of, 10 agriculture, 5 Alexandria, 21, 25, 39, 67, 68, 70, 71, 84, 116, 118, 119, 146, 152, 170, 174 alimony, 62 Anatolia, ancient, 44, 54, 134, 135 anthropology, 5, 8, 10, 14, 50, 60, 132, 142 apatores, 97 Apollonopolite nome, 43 Tanyaithis, 43, 113, 169 apprenticeship, 31, 62, 69, 75, 76, 77, 78 Arab world, 6, 8, 49, 156, 187 archeology, 14, 20, 201 army, 23, 24, 67, 79, 118, 130, 147, 151, 165 Arsinoe, 22, 25, 70, 71, 130, 171, 190 Arsinoite nome, 26, 71, 86, 111, 113, 114, 121, 137, 146, 147, 148, 159, 168, 175, 190, 196 Bakchias, 121 Boubastos, 170 Euhemeria, 137 Karanis, 39, 40, 41, 83, 112, 118, 119, 131 Kerkeosiris, 135 Kerkesoucha Orous, 135 Magdola, 86, 117 Philadelphia, 26, 84, 111, 116, 122, 170 Ptolemais, 24, 42, 71, 114 Soknopaiou Nesos, 39, 113, 137, 148 Tebtunis, 39, 114, 135, 136, 138, 159, 173, 192 Theadelphia, 168, 194 artisans, 26, 76, 201

Asia, 6, 8, 49, 53, 85, 89, 103, 144, 147, 154, 178, 180, 202, 213 Asia Minor, 16, 188 Caria, 181 Lycia, 189 Assyria, ancient, 22 Athens, classical, 15, 37, 66, 74, 75, 76, 78, 82, 93, 99, 103, 123, 132, 134, 140, 157, 168, 171, 177, 179, 191, 192, 193, 202, 203, 208, 212 Augustus, 24, 94, 178 aunts, 30, 47, 85, 90, 97, 113, 117, 121, 127, 144, 156, 159, 160, 171, 172, 195, 211 Babylonia, ancient, 22, 76, 107, 108, 125, 185 Balkans, 32, 232 Bangladesh, 165 Bible, 23, 245 New Testament, 23, 25 birth order, 128 birth rates, 1, 39, 107, 162 breastfeeding, 58, 62, 79, 83, 84 brothers-in-law, 145, 150 burial, 74, 84, 85, 86, 87, 88, 89, 93, 103, 135, 136, 137, 163, 173, 174, 180, 191, 205, 212 funeral expenses, 88 Byzantine world, 15, 138, 188 Cambridge Group, 13, 31, 37 care intergenerational, 3, 4, 19, 29, 30, 66, 69, 71, 72, 74, 81, 82, 83, 84, 85, 86, 87, 108, 114, 134, 138, 154, 171, 185, 199, 205 practical, 9, 206 caregivers, 108, 109, 114, 154, 210 celibacy, 8, 49, 164 census declarant of a household, 23, 97, 104, 124, 126, 127, 128, 131 order of listed family members, 124–30

257

258

Index

census (cont.) returns, Roman-Egyptian, i, 12, 15, 17, 22, 24, 25, 26, 27, 28, 29, 36, 38, 39, 41, 42, 43, 47, 48, 55, 56, 59, 73, 77, 78, 94, 95, 96, 97, 100, 101, 102, 105, 107, 109, 110, 114, 115, 117, 122, 124, 127, 128, 130, 132, 141, 145, 150, 151, 158, 159, 164, 167, 171, 172, 174, 176, 189, 195, 196, 201, 206 charity, 198 childbirth, 1, 2, 12, 17, 39, 58, 79, 80, 100, 107, 118, 121, 147, 148, 149, 163, 174, 175, 193 age at, 58, 176 childlessness, 3, 4, 31, 33, 53, 59, 114, 127, 162–87, 211 children, 13 care of, 14, 174 childhood, 17, 30, 65–86 eldest son, 13, 50, 51, 85, 108, 111, 112, 113, 123, 129, 132, 136, 150 living situation in case of divorce, 95, 103 sale of, 66 sick, 69, 121 China, 6, 8, 12, 44, 46, 52, 56, 59, 60, 92, 93, 98, 103, 107, 125, 144, 154, 155, 188, 189, 196 Christianity, 14, 94, 106, 158, 164, 198, 211 climate, 5 community vs individual, 4 comparative approach, 1, 4, 5, 6, 7, 8, 12, 14, 16, 37, 39, 40, 53, 75, 141, 147, 150, 165, 171, 190, 213 conflict, intergenerational, 9, 10, 18, 30, 34, 57, 67, 113, 121, 132, 141, 143, 149, 150, 162, 172, 187 Constitutio Antoniniana, 21 cooking, 13, 41, 114, 115, 118, 143, 174 Coptos, 67, 159 co-residence, 3, 9, 32, 34, 40, 45, 51, 54, 83, 113, 114, 117, 127, 128, 208 couriers, 118, 121 cousins, 30, 43, 45, 53, 55, 56, 72, 73, 97, 99, 102, 104, 111, 126, 127, 171, 179, 203, 210 craftsmen, 5, 20, 94 Crocodilopolis (later Arsinoe), 135 Cyrene, 119 daily life, 5, 7, 17, 35, 129, 175, 194, 204 Dakleh Oasis, 39 daughters-in-law, 18, 30, 49, 64, 115, 122, 124, 126, 127, 141–61, 193, 208, 210 debts, 56, 63, 64, 77, 88, 136, 137, 166 Deir el-Medina, 22 demographic conditions, 2, 4, 5, 10, 11, 12, 23, 25, 28, 30, 32, 43, 44, 48, 53, 59, 130, 144, 145, 178, 199, 201, 204, 211 demographic transition, 2, 5, 44, 59, 92, 199 demography, 5, 11, 14, 18, 26, 60, 216

Demotic, 38, 119, 133 developing countries, 1, 3, 7, 17, 59, 60, 73 Diocletian, 25 disabled, 3, 179 disease environment, 2 diseases, 12, 165, 200 see also malaria divorce, 2, 28, 31, 43, 49, 78, 79, 92, 93, 95, 96, 101, 102, 105, 106, 109, 110, 111, 121, 123, 126, 130, 145, 166, 177, 195, 199, 200, 206 doctors, 5, 24 donatio mortis causa, 85, 138 donkey-drivers, 24 dowry, 29, 30, 49, 64, 66, 79, 84, 87, 88, 99, 102, 103, 104, 116, 123, 124, 153, 156, 164, 166, 169, 189, 191, 193, 195, 203, 206, 211, 212 Dura Europos, 181, 183 economic considerations, 3, 4, 7, 8, 10, 11, 13, 17, 27, 28, 32, 34, 48, 49, 54, 56, 57, 59, 60, 64, 65, 81, 86, 92, 105, 144, 168, 188, 199, 200 education, 3, 29, 59, 66, 70, 72, 74, 75, 76, 79, 86, 89, 168, 198, 204, 205, 206 Egypt Greeks in, 38, 45, 130, 135 medieval, 138, 213 modern, 163, 186 nineteenth century, 49, 52, 53, 54, 98, 100, 103, 166, 213 pharaonic, 8, 22, 105, 123, 138, 181, 186, 190, 191, 203, 208 Ptolemaic, 8, 15, 23, 27, 38, 39, 45, 46, 70, 86, 88, 115, 116, 124, 130, 131, 133, 135, 138, 166, 170, 181, 194, 202 Roman indigenous Egyptian population of, 21, 51, 74, 124, 133, 134, 139, 186, 192, 202, 203, 212 twentieth century, 64 elderly people, 1, 2, 3, 4, 9, 10, 11, 13, 18, 19, 28, 29, 30, 33, 35, 42, 43, 48, 54, 81, 83, 84, 85, 87, 90, 92, 108, 109, 110, 111, 112, 113, 117, 121, 127, 129, 139, 140, 162–98 living situation, 29, 108–18, 132, 167–8, 199 see also old age Elephantine, 135 elites, 5, 15, 16, 20, 29, 34, 35, 52, 82, 92, 94, 95, 157, 172 England, 10, 32, 46, 156, 188, 202 epigraphic sources, 11, 14, 15, 16, 17, 35, 36, 37, 89, 180 epigraphy, 35, 36 equity, intergenerational, 2, 3, 8, 81, 123

Index Ethiopia, 115 ethnicity, 38, 194 Europe, 1, 6, 13, 32, 37, 92 Eastern, 32 north-western, 1, 2, 8, 13, 17, 29, 32, 36, 45, 48, 49, 50, 107, 108, 109, 162, 163, 168, 172, 202, 203, 213 southern, 8, 32, 112 exposure of children, 58, 59, 66, 197, 199 eyesight, 13, 86 family, authority, 3, 18, 19, 29, 48, 55, 66, 109, 124, 125, 127, 129, 133, 139, 140, 142, 144, 148, 154, 155, 156, 172, 206, 208, 210 composition, 4, 38, 47, 195, 201 as nucleus of society, 3, 4 size, 3, 38 strategies, i, 4, 14, 29, 30, 31, 37, 54, 108, 109, 198, 199, 211, 213 woman’s natal, 49, 78, 98, 99, 103, 112, 121, 145, 147, 202, 207 see also household farming, 111, 115, 159 fathers-in-law, 111, 130, 145, 189, 196 Fayum, 21, 26, 39, 40, 41, 118, 135, 194, 215, 236, 238 fertility, 2, 5, 7, 12, 28, 44, 50, 56, 57, 58, 59, 60, 65, 108, 161, 163, 199, 204 fosterage, 15, 59, 105, 197–8 France, 14 Gaul, 14 northern, 32, 46, 202 southern, 32, 49, 112 freedmen, 26, 79, 197 frérèches, 34, 37, 43, 44, 47, 55, 96, 101, 113, 128, 149, 151 friends, 66, 70, 75, 89, 119, 137, 175 gardeners, 23, 24 gender differences, 11, 14, 17, 23, 29, 108, 109, 110, 113, 115, 118, 123, 124, 125, 142, 143, 150, 162, 168 gender studies, 14 generations, 1, 2, 3, 4, 5, 9, 17, 18, 28, 31, 32, 33, 43, 44, 47, 48, 50, 53, 54, 55, 81, 107, 108, 111, 113, 122, 127, 139, 199, 200 genital mutilation, female, 59 Germany, 2, 32, 202 Gnomon of the Idiologos, 194 goldsmiths, 24 grandchildren, 3, 13, 108, 111, 112, 113, 124, 127, 131, 135, 136, 149, 159, 167, 179, 198, 208 grandparents, 73, 97, 105, 108, 158, 171, 181, 210

259

grandfathers, 72, 73, 75, 191, 192 grandmothers, 122, 149, 159 granduncles, 104 graying of the Western world, 2 Greece, 14 classical, 2, 14, 23, 37, 83, 103, 123, 132, 173, 184, 186, 188, 190 Hellenistic, 180 nineteenth century, 37 Roman, 180 guardians, 23, 49, 66, 73, 77, 98, 105, 112, 130, 137, 203, 207 health, 2, 3, 11, 12, 59, 69, 70, 83, 84, 119, 120, 164 hearing, 13 Hellenistic world, 5, 14, 15, 30, 74, 123, 134, 157, 161, 171, 173, 180, 186, 190, 191 Heracleopolis, 76, 183 Heracleopolite nome, 112, 117, 121 Hermopolis, 26, 88, 116, 136, 137, 146, 166, 178, 183, 191 Hermopolite nome, 149, 159 houseful, 19 household composition, 7, 12, 13, 16, 29, 31, 35, 36, 38, 47, 57, 141, 199, 213 definition of, 26–7, 33–4 domestic economy, 37, 48, 203 domestic space, 21, 41 dynamics, 21, 37, 141 extended family, 2, 16, 23, 33, 35, 37, 38, 39, 42, 43, 45, 46, 47, 53, 72, 85, 96, 99, 105, 110, 111, 112, 122, 129, 131, 145, 151, 157, 165, 167, 201 fiscal domicile, 25 as a fluid concept, 19, 28, 47 head of, 13, 18, 23, 27, 50, 51, 52, 55, 72, 97, 98, 99, 100, 111, 112, 113, 117, 124, 125, 127, 128, 129, 130, 132, 139, 143, 150, 187, 196, 201, 209 female head of, 18, 115, 125, 127–8, 130–1, 146 joint family, 14, 141, 232, 238, 248 multiple family, 14, 30, 32, 33, 35, 36, 37, 39, 41, 43, 44, 45, 50, 51, 52, 53, 55, 56, 72, 73, 96, 99, 101, 103, 110, 112, 124, 141, 143, 150, 154, 161, 189, 201, 207 nuclear family, 16, 31, 32, 33, 34, 35, 36, 37, 38, 39, 42, 43, 45, 47, 50, 51, 52, 53, 72, 73, 95, 97, 100, 110, 125, 126, 131, 144, 175, 201 registration unit, 26 size, 15, 23, 38, 39, 199, 201 solitary household/single household, 33, 38, 42, 45, 46, 47, 84, 110, 114, 115, 167, 174, 201, 208, 211

260

Index

household (cont.) structures, 4, 7, 15, 16, 17, 20, 28, 29, 35, 36, 42, 161, 199, 202 studies, 13, 17, 21, 26, 27, 28, 32 walls, 19, 211 housework, 3, 19, 62, 117, 143, 147, 154 housing, 13, 21, 34, 39, 40, 41, 47, 48, 53, 55, 77, 99, 103, 105, 111, 112, 115, 126, 130, 138, 169, 170, 171, 172, 173, 195 oikos, 15, 100, 179, 180, 191 physical structure, 28 identity, 11, 168, 194 illness, 2, 3, 12, 19, 54, 69, 83, 84, 86, 91, 122, 138, 147, 152, 170, 179, 187 incest. See marriage, brother–sister India, 6, 8, 12, 72, 93, 98, 100, 103, 148, 156, 161, 165, 180, 196 Industrial Revolution, 2, 10, 31, 37 infanticide, 58, 59, 66 infections, 12 infertility, 59, 92, 163 inheritance, 4, 8, 19, 28, 29, 31, 32, 37, 49, 50, 51, 52, 53, 54, 55, 57, 64, 87, 88, 98, 99, 108, 111, 112, 114, 116, 123, 129, 131–4, 137, 147, 160, 171, 184, 198, 211 daughter as heir, 123 disinheritance, 89, 90, 197 inheritance patterns, 28, 29, 32, 49, 52, 199, 203, 213 lawful heir, 50, 51, 53, 74, 87, 88, 89, 98, 99, 101, 105, 123, 127, 135, 136, 137, 149, 160, 170, 176, 178, 183, 185, 186, 188, 195, 197, 211 partible inheritance, 49, 50, 51, 53, 57, 123, 161, 203 primogeniture, 132 usufruct, 98, 99, 111, 133, 135, 140, 184, 207, 209 isolation, 8 of daughter-in-law, 144, 156 of the elderly, 8, 13, 114, 168, 174, 210, 211 Israel, 15, 23, 103, 214, 239 Italy, 14, 16, 24, 29, 32, 35, 44, 46, 50, 52, 80, 100, 112, 127, 145 northern and central, 32, 37, 112 southern Italy, 32 Japan, 42, 103, 114, 154, 178, 188, 189 Jeme, 138 kinship, 3, 14, 19, 26, 30, 53, 87, 122 Korea, 154 labor demands, 13, 17, 19, 22, 26, 49, 51, 52, 54, 57, 60, 62, 63, 64, 75, 76, 83, 194, 196, 204 laborers, 24, 62, 63, 64, 100, 115, 164, 204

landholding patterns, 33, 49 landowners, 5, 26 late antiquity, 11, 20, 94, 106, 108 laundry, 114, 118 laws, 4, 9, 11, 15, 19, 23, 24, 35, 66, 77, 84, 87, 90, 98, 167, 199 ancient Near Eastern, 184 Egyptian, 133 Greek, 74, 82, 87, 99, 133, 135, 158, 171, 184, 186, 197, 212 Roman, 16, 21, 35, 73, 78, 82, 87, 94, 133, 134, 153, 158, 171, 184, 186, 197, 212 legal sources, 20 Leptis, 142 letters, private, 15, 22, 27, 28, 65, 109, 118, 119, 121, 141, 146, 174, 203, 209 Libya, 119 life cycle, 3, 17, 28, 47, 53, 66, 105, 122, 205, 234 life expectancy, 1, 2, 7, 12, 44, 85, 107 literary sources, 5, 11, 13, 14, 15, 16, 20, 30, 35, 66, 93, 95, 141, 192, 210 livestock, 22, 33, 38 lodgers, 19, 26, 34, 39, 97, 114, 125 longue durée, 106 love, 66, 69, 71, 82, 122, 143, 148, 152, 155, 180, 205 lower social strata, 7, 16, 20, 21, 26, 35, 65, 75, 77, 78, 92, 93, 94, 96, 165 Lykopolis, 24, 26 malaria, 12, 69 malnutrition, 12, 59, 72, 164, 165 marriage, 3, 8, 13, 14, 15, 17, 19, 24, 28, 29, 31, 34, 37, 44, 45, 48, 49, 50, 51, 52, 56, 57, 58–9, 78, 83, 106, 108, 109, 115, 121, 123, 126, 138, 199, 206 adoption cum marriage, 155, 159 age at first, 8, 48–50, 58, 94, 114, 123, 163 age gap between spouses, 48, 92, 94–5, 102, 113, 127 brother–sister, 47, 126, 127, 159–61, 187–96 contracts, 14, 15, 134, 151, 153, 204 cousin, 156, 157, 158, 159 endogamous, 30, 155–61 exogamous, 161 half-sibling, 158, 160, 192 law, Augustan, 94 marital status, 96 patterns, 14, 15, 19, 48, 49, 210 remarriage, 15, 18, 28, 29, 56, 58, 78, 79, 92–106, 109, 110, 113, 126, 127, 130, 162, 169, 172, 199 uncle–niece marriage, 158 medical care, 2, 11, 14, 19 Mediterranean, 5, 7, 11, 14, 16, 28, 37, 40, 49, 76, 95, 144, 156, 189, 191, 202, 212

Index Eastern, 8, 15, 28, 32, 48, 59, 133, 158, 176, 197, 212 Western, 35, 36, 197, 211 Memphis, 88, 130, 152 Memphite nome, 113 Moithymis, 113, 189 merchants, 5 metropolis, 22, 25, 26, 27, 39, 40, 45, 52, 58, 94, 121, 131, 165, 201 Middle Ages, 32, 188 Middle East, 8, 25, 40, 49, 52, 103 migration, 2, 30, 42, 44, 59, 83, 118, 130, 147, 151, 165, 176, 177, 199 minors, 23, 38, 45, 49, 56, 73, 77, 95, 96, 98, 105, 125, 129, 130, 170, 184, 186, 197, 200 mortality, 4, 5, 12, 28, 30, 39, 44, 53, 56, 58, 69, 102, 151, 156, 176, 187, 199, 206, 211 infant, 7, 12, 44, 64, 161, 163, 164, 200, 204 mortgaging, 135, 136 mothers-in-law, 30, 85, 117, 122, 127, 128, 141–61, 208, 210, 211 Near East, ancient, 11, 108, 164, 173, 177, 184, 186, 188, 190, 191, 212 neighbors, 41, 66, 174, 175, 211 Nepal, 108, 216 nephews, 33, 45, 55, 56, 76, 99, 104, 110, 113, 126, 127, 160, 167, 171, 172, 179, 180, 192, 195, 203, 211 Netherlands, 32 nieces, 55, 85, 90, 94, 117, 127, 128, 159, 160, 171, 172, 189, 195 Nile Delta, 25 North Africa, 14, 79, 142, 144 Nubia, 118, 119 nutrition, 60–2, 72 obligations, intergenerational, 2, 4, 8, 29, 36, 54, 66, 73, 78, 80, 82, 84, 86, 87, 89, 123, 132, 135, 137, 138, 140, 191, 199 occupation, 13, 26, 40 old age, 3, 4, 10, 11, 13, 14, 17, 18, 19, 28, 29, 30, 31, 37, 65, 74, 80, 81, 83, 84, 87, 90, 92, 107–40, 162–98, 200, 211 care in, 138 frailty, 4, 11, 13, 19, 154, 155, 185, 206 support in, 65, 74, 88, 109, 113, 114, 116, 124, 132, 134, 136, 137, 138, 140, 146, 154, 155, 159, 162–98, 199, 205 orphans, 2, 3, 4, 23, 31, 73, 97, 99, 105, 159, 171, 177, 179, 184, 197, 198, 200, 211 Ostia, 119 Ottoman Empire, 6, 44, 54, 55, 93, 98 Oxyrhynchite nome, 125 Oxyrhynchus, 18, 22, 25, 41, 55, 56, 67, 68, 69, 78, 84, 85, 88, 104, 124, 129, 137, 148, 152, 153, 165, 169, 170, 172, 178, 183, 186, 195, 197

261

Pakistan, 156 papyrological sources, 5, 11, 15, 21, 24, 25, 150, 176, 182, 192 paramonē, 174 parents-in-law, 18, 42, 47, 48, 102, 143, 144, 145, 146, 147, 150, 154, 155, 160 see also fathers-in-law; mothers-in-law patchwork families, 126, 199, 200 patria potestas, 35, 51, 66, 74, 124, 133, 153, 184, 212 patriarch, 35, 48, 51, 54, 66, 126, 127, 128, 129, 131, 132, 134, 139, 153, 154, 203, 209 patriarchal family systems, 4, 5, 29, 32, 49, 51, 52, 53, 98, 99, 103, 141, 147, 150, 156, 178, 193, 195, 196, 201, 206, 210 peasant population, 15, 19, 20, 21, 22, 24, 26, 32, 33, 35, 37, 51, 52, 60, 62, 63, 64, 79, 83, 94, 114, 141, 194, 201 Persia, ancient, 192 pietas, 66, 82 population growth, 12, 44, 200 postmarital residence, 29, 32, 48–50 neolocality, 49, 123, 144 patri-virilocal, 32, 48, 49, 156, 207 uxorilocal, 48, 115, 130, 161, 187–96, 202, 212 virilocal, 30, 48, 96, 123, 143, 144, 147, 161, 189, 193, 210 poverty, 3, 34, 35, 80, 81, 116, 155, 164, 184, 196, 198 pregnancy, 50, 99, 118, 122, 142, 147, 149, 152, 164, 175 prices, 60–5 property, 41, 43, 51, 98 division of, 56 transfers of, 4, 15, 122, 123, 131–9, 171 Prosopite nome, 112, 113, 114, 115 Thelbonthon Siphtha, 113, 114 prostitution, 66, 89, 116, 206 Pselkis, 118 public welfare, 2, 5, 163, 172 quality of life, 13 reciprocity, intergenerational, 3, 4, 5, 9, 28, 66, 78, 81, 82, 86, 137, 154, 155, 187, 199, 205 regional environment, 5, 6, 11, 14, 32, 37, 49, 200 relatives, 1, 4, 9, 27, 30, 33, 38, 42, 49, 56, 69, 73, 76, 77, 78, 81, 82, 89, 97, 103, 105, 114, 117, 122, 123, 124, 127, 133, 137, 143, 155, 170, 171, 174, 175, 179, 180, 191, 197, 211 religion, 7, 9, 105, 111, 152, 159, 164, 205 respect, 3, 6, 68, 69, 70, 71, 81, 84, 90, 120, 122, 129, 206, 209 retirement, 2, 5, 85, 198, 208 contracts, 29, 123, 132, 134–9, 209 Rhodes, 181, 191

262

Index

Roman citizenship, 21, 36, 139, 153, 169, 194 Roman Empire, 16, 21, 24, 25, 36, 37, 103, 191 Roman family, 14, 16, 36 Roman Republic, 20, 23, 35, 103, 184 Rome, city of, 14, 20, 21, 23, 35, 66, 76, 79, 82, 93, 94, 103, 119, 139, 164, 178 Russia, 32, 44, 46, 49, 55, 188 Scandinavia, 32 scribes, 5, 77, 124, 135 servants, 19, 144 sex, 15, 125, 126, 127 shepherds, 23 siblings, 5, 33, 38, 50, 51, 53, 55, 72, 73, 90, 104, 111, 123, 127, 129, 168, 171 eldest brother, special position, 85, 113, 129, 203 half-siblings, 56, 57, 90, 126, 128, 158 single parenthood, 3, 73 single fathers, 73 single mothers, 73, 97, 130, 199, 200 sisters-in-law, 32, 34, 44, 47, 48, 59, 94, 101, 102, 122, 128, 143, 145, 150–1, 154 skeletal evidence, 12, 164 slaves, 23, 24, 26, 27, 28, 34, 37, 39, 51, 55, 59, 74, 76, 79, 87, 89, 114, 117, 120, 125, 138, 149, 172–3, 183, 197, 201, 206, 211 social class, 11, 13 social exchange theory, 9 social expectations, 4, 204 social security, 2, 81, 220 sociology, 5, 10, 15, 18, 29, 232 solidarity, intergenerational, 2, 3, 5, 9, 10, 17, 18, 30, 57, 65, 81, 199, 206, 211, 217, 241 sons-in-law, 110, 112, 115, 117, 147, 153, 188 Spain, 14, 16, 112, 224, 241 state, 2, 3, 4, 14, 24, 27, 40, 81, 138 stem family, 14, 37, 50, 132 stepchildren, 177 stepsons, 126, 178 stepfathers, 15, 29, 103, 105, 126, 177, 207, 230 stepmothers, 160, 177 succession. See inheritance support financial, 2, 3, 18, 78, 83, 114, 116, 117, 144, 165–7, 171, 185 intergenerational, 1, 2, 3, 4, 5, 7, 9, 13, 18, 19, 20, 28, 29, 30, 56, 65, 74, 76, 79, 80, 81, 82, 83, 85, 86, 87, 88, 89, 107, 109, 114, 117, 134, 154, 155, 162, 167, 183, 198, 199, 205, 211 Syene, 79, 88 Syria, 21, 32, 44, 54, 55, 93, 158, 188

Taiwan, 103, 155, 214, 222, 249 taxation, 2, 18, 22, 23, 24, 25, 26, 28, 38, 61, 63, 75, 81, 85, 131, 136, 138, 139, 140, 168 inheritance tax, 138, 139 poll tax, 18, 24, 80, 85 teachers, 68, 75, 76 Thailand, 114 Thebes, 198 town councilor, 27, 116, 124 traders, 5, 21, 26, 79, 83, 94 traditions, 4, 108 transfers, intergenerational, 28 Turkey, 32, 103, 223, 229, 232 Tuscany, late medieval, 37, 39, 46, 49, 50, 51, 52, 103, 125, 127, 145, 174 uncles, 30, 45, 55, 56, 69, 72, 75, 90, 97, 99, 101, 104, 105, 111, 113, 127, 128, 144, 159, 171, 172, 177, 191, 195, 203, 207, 210 unemployed, 2 urban vs rural, 5, 12, 13, 15, 28, 29, 52, 131, 165, 201 urbanization, 2, 22, 33, 199, 200 usufruct. See under inheritance veterans, 5, 224 villages, 21, 25, 26, 27, 39, 40, 41, 52, 94, 96, 109, 131, 135, 201 visits between family members, 68, 88, 118, 121, 122, 143, 206 wealth, 3, 30, 34, 49, 51, 53, 60, 156, 211 wealth flows theory, 59–65, 204 weavers, 24, 76 weddings, 8, 49, 83, 129, 142, 144 well-being, 3, 9, 19, 65, 69, 70, 119, 120, 122, 148, 151 emotional well-being, 168 widowhood, 4, 29, 33, 38, 42, 47, 49, 55, 84, 92–106, 110, 111, 114, 115, 121, 123, 126, 129, 130, 145, 199, 200, 206 poverty in, 92 widow administrating the patrimony, 98, 130–1, 207 widowers, 45, 94, 112, 113, 117, 127, 189, 198, 236 widows, 2, 18, 29, 34, 45, 48, 85, 88, 94, 100, 109, 112, 115, 117, 118, 119, 121, 127, 128, 130, 136, 137, 140, 146, 148, 165, 170, 173, 198, 203 wills, 1, 2, 3, 4, 5, 7, 10, 14, 16, 22, 27, 28, 30, 39, 44, 47, 48, 56, 57, 58, 65, 66, 67, 69, 71, 74, 78, 81, 82, 89, 90, 92, 94, 98, 106, 109, 118, 119, 123, 125, 134, 136, 137, 140, 141, 146, 170, 172, 197, 235 wisdom of the elderly, 10, 11