A field of one's own: gender and land rights in South Asia 9780521418683, 9780511995095, 9780521429269

Table of contents :
Frontmatter
List of illustrations (page xii)
List of tables (page xiii)
Preface (page xv)
1 Land rights for women: making the case (page 1)
2 Conceptualizing gender relations (page 51)
3 Customary rights and associated practices (page 82)
4 Erosion and disinheritance: traditionally matrilineal and bilateral communities (page 153)
5 Contemporary laws: contestation and content (page 198)
6 Whose share? Who claims? The gap between law and practice (page 249)
7 Whose land? Who commands? The gap between ownership and control (page 292)
8 Tracing cross-regional diversities (page 316)
9 Struggles over resources, struggles over meanings (page 421)
10 The long march ahead (page 467)
Definitions (page 505)
Glossary (page 507)
References (page 510)
Index (page 553)

Citation preview

Economic analysis and policies concerning women have long been preoccupied with employment. In a radical shift of focus, Professor Agarwal argues that the single most important economic factor affecting women’s situation is the gender gap in command over property. In rural South Asia, the most significant form of property is arable land, a critical determinant of economic well-being, social status, and empower-

ment. However, few South Asian women own land, and even fewer control it. In a comprehensive and rigorous analysis that draws upon a wide range of historical, economic, legal, and ethnographic sources and her own field research, the author investigates the complex reasons for this gender gap, and examines how existing barriers to women’s land ownership and control might be overcome. Regional variations on these counts

across India, Pakistan, Bangladesh, Nepal, and Sri Lanka are also identified. The study extends the boundaries of economic analysis to explore the interface of economics, culture, and gender politics through an interdisciplinary approach. It examines women’s covert and overt resistance to gender inequality, especially in the context of land struggles. And it offers new theoretical insights by extending the ‘bargaining approach’ to illuminate how gender relations get constituted and contested, both within

and outside the household. |

A field of one’s own is the first major study on gender and property in South Asia. It makes significant contributions to current debates on land reform, women’s status, and the nature of resistance. Its compelling and original argument will interest scholars, students, policy makers, and activists. BINA AGARWAL IS Professor of Economics at the Institute of Economic Growth, University of Delhi. Educated at the Universities of Cambridge and Delhi, she has taught at Harvard University as visiting professor, and been a fellow of the Bunting Institute (Radcliffe College) and the Institute of Development Studies (University of Sussex). She has published extensively on poverty and inequality, rural development, environmental issues, and technological change, from a political economy and gender perspective. Her books include: Cold Hearths and Barren Slopes: The Woodfuel

Crisis in the Third World (1986), Mechanization in Indian Agriculture (1983), and Structures of Patriarchy: State, Community and Household in Modernising Asia, ed (1988).

BLANK PAGE

Cambridge South Asian Studies

A field of one’s own

Cambridge South Asian Studies

Editorial Board

C. A. Bayly, G. P. Hawthorr., Gordon Johnson, S. J. Tambiah

, A list of the books in the series will be found at the end of the volume

A field of one’s own Gender and land rights in South Asia Bina Agarwal

CAMBRIDGE

9 UNIVERSITY PRESS

Published by the Press Syndicate of the University of Cambridge The Pitt Building, Trumpington Street, Cambridge CB2 IRP 40 West 20th Street, New York, NY 10011-4211, USA 10 Stamford Road, Oakleigh, Melbourne 3166, Australia

© Bina Agarwal First published 1994 A catalogue record for this book is available from the British Library Library of Congress cataloguing in publication data

Agarwal, Bina A field of one’s own: gender and land rights in South Asia/Bina Agarwal

p. cm. — (Cambridge South Asian studies) Includes bibliographical references. ISBN 0-521-41868-2 — ISBN 0-521-42926-9 (pbk.) 1. Land tenure-South Asia. 2. Land reform—South Asia. 3. Right to property—South Asia. 4. Women’s rights—South Asia. 5. Rural

women-South Asia. 6. Rural women-South Asia. 7. South Asia— Social policy. I. Title. II. Series. HD860.3.Z63A35 1994 323.4'6'0959-dc20 93-37621 CIP ISBN 0 521 41868 2 hardback ISBN 0 521 42926 9 paperback

Transferred to digital printing 2003

SE

To my father for his wisdom and optimism To my mother for her generosity and vigour

BLANK PAGE

Contents

Preface XV I The backdrop 2

List of illustrations page xii List of tables Xill 1 Land rights for women: making the case l Il Gender, property, and land: some conceptual links Il

(1) Household property and women’s property 12

(2) The significance of land as property 17 (3) What do we mean by rights in land? 19

(4) Prospects for non-land-based livelihoods 24

II] Why do women need independent rights in land? 27

(1) The Theefficiency welfare argument 27 (2) argument 33 (3) The equality and empowerment arguments 38

IV Questions addressed, information base, and the book’s structure 45

2 Conceptualizing gender relations 51 I Gender relations within the household/family 53

(1) The bargaining approach 54 community, and the State . 71 III Interactions: the household/family, the community, and the State 80 (2) What determines intra-family bargaining power? 60

II Gender relations outside the household/family: the market, the

3 Customary rights and associated practices 82 I Which communities customarily recognized women’s rights in land? 83 II Women’s land rights in traditionally matrilineal and bilateral communities _ 100

(1) Northeast India: The Garos, Khasis, and Lalungs 101

Vellalars, and others 109

(2) South India: The Nayars, Tiyyars, Bants, Mappilas, Nangudi

(3) Sri Lanka: The Sinhalese, Hindu Tamils, and matrilineal Muslims 120

(4) Some cross-regional comparisons 132 III Women’s land rights, structural conditionalities, and gender relations 133

(1) Women’s land rights and associated practices 133

(2) Land rights and gender relations 146

x Contents

I(1)India 154 The Garos 154 II Sri Lanka 180 The Sinhalese III In conclusion 180 192 I India 199 bilateral communities 153

4 Erosion and disinheritance: traditionally matrilineal and

(2) The Nayars of central Kerala 168 (3) Matriliny and development 179

Appendix 4.1: A marriage proposal among the Christian Garos 194

5 Contemporary laws: contestation and content 198 (1) The formulation of contemporary Hindu law 199

(2) Anomalies resulting from existing land legislation 215

(3) Laws governing Christians and Parsis in India 223

II Pakistan, Bangladesh, and Muslims in India 227

(1) From custom to the Shariat 227

(2) Devolution under Islamic237 law 233 Ii] Sri Lanka IV Nepal 242

in South Asia 246

V Summary comments on women’s legal rights in landed property

6 Whose share? Who claims? The gap between law and practice 249 I The gap between law and practice in traditionally patrilineal communities 249 II Barriers to women inheriting land in traditionally patrilineal communities 260

(1) ‘Voluntary’ giving up of claims 260 (2) The necessity of male mediation 268

(3) Hostility from male kin: pre-emptive steps to direct violence 271

(4) Responses of village bodies and government officials 276

patrilineal communities 282

III Glimmer of change: women claim inheritance shares in some traditionally

V_ Some hypotheses 291 IV A look at traditionally matrilineal and bilateral communities 285

control 292

7 Whose land? Who commands? The gap between ownership and

I Women’s ability to retain their land 292

II Control over the transfer and use of land 294 III Barriers to women self-managing land 298 (1) The physical and social confinement of women 298

(2) Post-marital residence: village exogamy and patrilocality 311

(3) Male control over labour and technology 311

8 Tracing cross-regional diversities 316

I Some hypotheses II Information sources317 321

Contents xi

II](1)The cross-regional patterns 325 Marriage location and post-marital residence 325

(3) Purdah practices 344 (4) Sexual control over women 345 (2) Close-kin marriages, especially between cross-cousins 336

(5) Rural female labour force participation rates 355

(6) Rural female literacy rates359 358 (7) Total fertility rates (8) Land scarcity 361

IV An overview of regional patterns 368 9 Struggles over resources, struggles over meanings 421

I On women’s consciousness and individual resistance 422

II Group resistance: struggles over privatized land 438

(1) The The Telangana Tebhaga struggle 438 (2) struggle 44] (3) The Bodhgaya struggle 444 III Group resistance: claiming rights in public land 454 IV Further observations on gender construction and group contestation 458

10I The long march ahead468 467 Recapitulation II Some suggestions, some dilemmas 478 (1) Dowry Reforming the laws 478 (2) v. inheritance 480 (3) Establishing de facto inheritance rights in land 483

(4) Strengthening land claims through channels other than inheritance 486 (5) Exploring joint management and promoting infrastructural support 488

(6) Building group support among and for women 490 III The macro-scenario 493 (1) Interlinking diverse concerns 493

(2) Bargaining with the State 496 (3) Increasing women’s presence in public decision-making forums 499

Definitions 505 Glossary 507 References 510 Index 553 (4) Some recent developments and the road ahead 502

Illustrations

Diagram 4.1 Garos in transition: causes and effects page 166 Maps

1.1 South Asia: provincial/state divisions XXil 3.1. Traditional forms of land inheritance 100

3.2 Sri Lanka: provincial divisions 121

8.1 Village endogamy norms 329 8.2 Close-kin marriage norms 342

8.3. Purdah practices 348 8.4 Rural female labour force participation rates (1981) 357

8.5 Rural female literacy rates (1981) 359

8.6 Total fertility rates 361 8.7 Population density (1981) (persons per sq km) 365

8.8 Percent landless to total rural households 366

8.9 Percent geographic area under forest 367

8.10 A comparative perspective 372

Xil

Tables

Al.1 Village common lands in India by State: 1987-88 page 49

Al.2 Selected development indicators for South Asia 50 3.1 Some characteristic features of matrilineal and bilateral

communities in South Asia 141

4.1 Changes among the Garos of northeast India: a summary 167

4.2 Women in selected occupations: Sri Lanka, 1901-21 183 4.3 Ownership of paddy land and highland by gender among the ]

Kandyan Sinhalese 189

4.4 Post-marital residence among the Kandyan Sinhalese 190 5.1 Summary of oral and written opinions on the Draft Hindu

Code received by the Second Rau Committee, 1945 209

sition 287

6.1 Widows in India who inherited land as daughters 253

6.2 Dowry land by type among the Jaffna Tamils 287 6.3 Landownership in three Jaffna villages by source of acqui-

8.1 Village endogamy norms in South Asia 326 8.2 Marriage distance from a woman’s natal home in South Asia 332 8.3 Close-kin marriage norms in South Asia 338

8.4 Purdah practices in South Asia 346

8.5 Rural female labour force participation rates and rural

female literacy rates in South Asia (1981) 356

8.6 Total fertility rates in South Asia (1988) 360 8.7 Population density, landlessness, land concentration, and

forest area in South Asia 362

8.8 Distribution of ownership holdings in South Asia among

rural landowning households 364

8.9 Summary for cross-regional comparisons 370 home: detailed evidence from South Asia 379

A8.1 Women’s post-marital location and distance from the natal

from South Asia 390

A8.2 Close-kin marriage norms and practice: detailed evidence

Xi

XIV Tables A8.3a Sexual control over women in South Asia: norms of pre-

marital sex and adultery 403 A8.3b Sexual control over women in South Asia: norms and practice

of divorce and divorcee remarriage 407

A8.3c Sexual control over women in South Asia: norms and practice

of widow remarriage 414

Preface

This book has grown out of two long-standing involvements. One is my decade and a half of research on rural poverty, agrarian change, and the political economy of gender. The other is my association with the women’s movement in South Asia, and my interaction over the years with peasant women from across the region, and with the few grassroots activists who were beginning to raise the issue of women’s independent land rights within mass-based peasant movements in the late 1970s and early 1980s. Both my academic analysis and these interactions led me to investigate, and in that process recognize the central importance of women’s lack of effective rights

in property, especially land, in explaining their economic, social, and political subordination in South Asia. It became increasingly clear that it was critical for women to win those rights for establishing more equal gender relations, both within and outside the household. Indeed while the

link between property and class relations has been well established in political economy, the link between property and gender relations has remained largely unexamined. Land has been and continues to be the most significant form of property

in rural South Asia. It is a critical determinant of economic well-being, social status, and political power. However, there is substantial evidence that economic resources in the hands of male household members often do

not benefit female members in equal degree. Independent ownership of such resources, especially land, can thus be of crucial importance in promoting the well-being and empowerment of women. But as the present analysis shows, the issue is not just one of property ownership; it 1s also that

of property control. Historically, even in matrilineal communities where formal ownership of property (including land) was vested in women, its effective control was often vested in men, as was jural authority. And this scenario of women’s virtual exclusion, in most regions, from control over property and from most public decision-making bodies, continues to be a familiar one today. Although economic surveys typically do not give a gender breakdown of land distribution, the ethnographic evidence examined here indicates that XV

XVi Preface despite gender-progressive legislation, in practice few South Asian women inherit landed property, and even fewer control it. This book probes what underlies the vast gap between law and its implementation, and identifies a

number of factors constraining women in exercising their legal claims, including patrilocal post-marital residence and village exogamy, strong Opposition from male kin, the social construction of gender needs and roles, low levels of female education, and male bias and dominance in administra-

tive, judicial, and other public decision-making bodies at all levels. The analysis here points to the interactive effects of economic factors, cultural

norms, and gender ideologies and politics, in determining women’s property position, an interaction that has received little attention from most economists or other social scientists, each typically operating in a separate disciplinary domain. The constraints identified indicate that rural women’s struggle for effective land rights will not be an easy one. At the same time, the considerable regional differences within and between South Asian countries in the nature and degree of these constraints, suggests that they are subject to contestation and change. More generally, the framework of contestation and bargaining between actors with differential access to economic and political power illuminates the process by which the hierarchical character of gender relations is maintained and changed. In this context, the process of acquiring land rights is likely to be as important in

| empowering women, as the end result. And it is precisely the formidable nature of the obstacles to be overcome that gives the struggle for land rights

a strategic importance and transformative potential which perhaps no other gender-related issue singly possesses. As peasant women in Bodhgaya (Bihar, east India) said on first receiving land in their own names in 1982: We had tongues but could not speak. We had feet but could not walk. Now that we have the land we have the strength to speak and walk!

I first began exploring the issue of gender and land rights in 1985, as part

of a larger project on this subject (covering several parts of the Third

World) launched by the International Labour Organisation (ILO), Geneva. This issue had also emerged as significant in my own research on the gender dimensions of agrarian change, and I welcomed the chance of focusing on it in depth when Zubeida Ahmed invited me to participate in

the project. In that year I visited Pakistan, Nepal, Bangladesh, and Sri Lanka; talked with villagers, grassroots activists, lawyers, government officials and academics; and attempted to locate written material on the subject. There was hardly any material to be found. Indeed my explorations

revealed how little attention had been paid to this subject in South Asia,

Preface XVII either in research or in policy; researchers, policy makers, and most non-

governmental (including women’s) groups seemed preoccupied with employment as the indicator of women’s economic status, to the neglect of property rights. In 1986-87, for a separate project, I spent several weeks doing fieldwork on the survival strategies of poor, low caste women in a village in Rajasthan (northwest India). During the course of that work, the importance for rural women of having even a small field of their own as a security against poverty emerged clearly. In addition, I spent a few weeks visiting Khasi and Garo villages in Meghalaya, drawn by the need to geta first-hand feel of how women’s situations and self-perceptions in matrili-

neal communities differed from those in the strongly patrilineal northwestern states with which I had greater familiarity. A number of research papers followed (Agarwal 1988, 1989, 1990b, 1990c), but I began work on the book itself only in the winter of 1989 at Harvard University. My travels, both within India and to other South Asian countries, were financed by the ILO, and I am very grateful to them for this support. The

success of my field visits depended greatly on the generous help and hospitality of many friends and colleagues. In particular, for my trip to Sn Lanka I thank Kumari Jayawardena, Charles and Sunila Obeyesekera,

Newton Gunasinghe (a dear friend who is no more), and Radhika Coomaraswamy. Newton accompanied me to Kandy, and with his help I

was able to talk to a number of villagers and tea plantation workers in the , stunningly beautiful Kandyan Highlands. It was already a period of severe ethnic strife in Sri Lanka, which has escalated tragically since. What I have to say in the book about Sri Lanka’s social and economic relations must be read against this backdrop. There is no easy way of predicting the long-term effects of these conflicts on the norms that govern rural life there. I therefore make no attempt to speculate about future trends, and hope that much of

what I have said continues to be valid. In Pakistan, I especially thank Farida Shaheed, Khawar Mumtaz, Nighat Khan, Akmal Hussain, Nigar Ahmed, Samina and Anjum Altaf, and Akbar Zaidi for their hospitality, for useful discussions on women’s position in the country, and for helping

me make contact with others (lawyers, academics, activists) and locate research material. During that trip, meeting with members of the Women’s Action Forum in Lahore, Karachi and Islamabad was particularly illuminating. In Bangladesh, I owe special thanks to Khushi Kabir, Mahmuda Islam, Jahan Ara Huq, B. K. Jehangir, Shireen Hug, Muhammad Yunus

(founder of the Grameen Bank), F. H. Abed (founder) and the staff members of the Bangladesh Rural Advancement Committee (BRAC), Zafarullah Choudhury (founder) and staff members of Gonoshastheya Kendra, and the faculty of the Bangladesh Institute of Development Studies, for useful discussions, sharing with me their field experiences, and

XVIl1 Preface facilitating my travel to rural areas near Dhaka. For my Nepal trip, I am especially grateful to Bina Pradhan, Shilu Singh, Chandini Joshi, and Deepak Bajracharya for helping me to locate research material and organize my visits to villages on the outskirts of Kathmandu. I also thank Mr and Mrs Narendra Agarwal for their warm hospitality during my stay in Nepal. My exposure to village life in Rajasthan extends over many years and

numerous visits: initially during childhood and adolescence to see my maternal grandparents in Jhunjhunu district; subsequently in the mid1980s to meet with activists of the Social Work and Research Center (Tilonia, Madanganj district) and to participate in a large festive gathering (mela) of village women that they had organized; and later still in 1986-87 to undertake a spell of systematic fieldwork in a village (in Alwar district)

which the anthropologist Miriam Sharma and her research assistant Urmila Vijnani were also researching, and whose help in facilitating my entry into the lives of the villagers I gratefully acknowledge. Finally, I am extremely grateful to Yogeshwar Kumar, without whose help my 1989 field

visit to the Khasi and Garo Hills in Meghalaya would not have been possible. On that trip I benefited greatly from discussions with D. N. Majumdar, Gilbert Shullai, and R. T. Rymbai; and my visits to the Khasi villages with Helen Giri and to the Garo villages with Debila Marak were most illuminating. Although most of the village visits in various parts of

| South Asia were not long enough to permit a systematic collection of information, they were invaluable in giving me a broad sense of the perceptions of rural women and men, village leaders, and local bureaucrats on the question of women and land. And I have drawn upon these field trips in various parts of the book. My two and a half years at Harvard University (September 1989—March

1992) were critical in the shaping and writing of this book. Harvard’s superb library facilities and its inter-library loan services gave me access to a

vast body of historical and contemporary material including many unpublished social science doctoral dissertations on South Asia, submitted to American universities. Also invaluable were several rounds of stimulating discussions on parts of my work with faculty members at Harvard and other US universities, many of whom had done detailed fieldwork in South Asia. My stay at Harvard was made possible through the support of several

institutions: the Mary Ingraham Bunting Institute (Radcliffe College), which awarded me a Bunting Fellowship for 1989-90 and affiliation as a Fellow for 1990-91; the Harvard Center for Population and Development Studies, of which Lincoln Chen kindly invited me to be a member during 1990-91; the Faculty of Arts and Sciences and the Committee of Women’s Studies, Harvard University, where I taught as visiting professor in 1991-—

Preface XixX 92; the John D. and Catherine T. MacArthur Foundation, which awarded me a Research and Writing Grant for Individuals for 1990-91; and the

Norwegian Agency for Development Co-operation (NORAD, Delhi), which provided supplementary financial assistance and a travel grant to the

USA for my first year at Harvard. I am most grateful to all these institutions. In particular the Bunting Institute — its fellows, staff and Director, Florence Ladd — provided an atmosphere which was wonderfully conducive to discussion, reflection, and writing. I thank the administrative staff of the Harvard Committee on Women’s Studies for their warmth and helpfulness, Barbara Johnson for generously lending me use of her office in Pusey library during my teaching term, and my students for their lively and challenging discussions on parts of my work. I have presented some aspects of my research at several invited lectures

and seminars both in India and abroad, including the Ninth European Conference on Modern South Asian Studies (Heidelberg) in 1986; the Center for Development Studies (Kerala) in 1988; the National Seminar on

Women and Access to Land and Productive Resources (University of Delhi) in 1988; the Planning Commission of India (Delhi) in 1989; the Association for Women in Development International Conference (Washington) in 1989; the International Food Policy Research Institute (Washington) in 1989; the Bunting Institute Public Colloquium Series in 1990; and the Distinguished Lecture Series of the Center for Advanced Study of

‘International Development, Michigan State University, in 1990. The responses of the audiences following these presentations, and Amartya Sen’s comments as a discussant on my Bunting Colloquium presentation at Harvard, were encouraging, stimulating, and helpful. Comments from and discussions with many other friends, colleagues, and associates provided essential feedback as the book progressed. It is not possible to name them all, but a few need particular mention. One person whose meticulous comments on several drafts of the book were invaluable is Janet Seiz — friend and fellow-economist at the Bunting Institute. I am deeply indebted to her for so generously sparing time to read the manuscript and brainstorm over parts of it, and for her unflagging interest and belief in the value of this research. I also greatly value Geoffrey Hawthorn’s keen support for this study since its inception; and owe a very special thanks to him, Nancy Folbre, Gail Hershatter, Primila Lewis, and the reviewer of

Cambridge University Press, all of whom read the manuscript in its entirety, in one or other of its incarnations, and offered most useful, detailed and encouraging comments. Gail found time to do this while settling into a new teaching job and parenting two young children. To Amartya Sen and S. J. Tambiah I am most grateful for sparing time from their extremely busy schedules during the teaching terms at Harvard to

XX Preface comment on and discuss some of my chapters over several tasty lunches at the Harvard Faculty Club. Discussions with Nur Yalman on some aspects of his earlier research on Sri Lanka were also most helpful. Joan Mencher

offered a number of valuable suggestions and insights on the Nayars of Kerala along with a sumptuous south Indian dinner in New York; and with

my friend and colleague, Gillian Hart, I shared many an evening of challenging discussions on my work and hers. I also greatly appreciate the

comments on selected chapters by Michael Lipton, Lourdes Beneria, Pauline Peters, Gunanath Obeyesekere, John Mansfield, Paul Seabright,

Uday Mehta, Christopher Fuller, Raghav Gaiha, Kate Gilbert, and Patricia Uberoi. I thank Alice Thorner, Jean Dreze, Terry Byres, and Veena Das for their responses to an earlier paper on the theme of the book. And I

am grateful to Marty Chen for sharing with me some results from her ongoing study on widows in India; to Hilary Standing for digging out some of her unpublished fieldwork findings on Bihar; and to Victor de Munck, Dennis McGilvray, Hamza Alavi, Hanna Papanek, Savitri Goonesekere, and Jack Goody for discussions which helped clarify specific points relating to their own research findings. I would also like to thank Nick Stern at STICERD (London School of Economics) for offering me STICERD’s hospitality for discussions and library work at the LSE over two summers. In Delhi, I am immensely grateful to B. Sivaramayya for the care with which he went over the legal sections of my manuscript, helping to correct several subtle shifts in meaning resulting from my attempts as a social scientist to translate legalese into simple English, and for drawing upon his

prodigious knowledge of inheritance laws to point out a number of important cases and recent changes in legislation. Iam also very thankful to Lotika Sarkar for discussions on a number of legal aspects and for directing

my attention to some landmark judgements. Vina Mazumdar’s descriptions of her interactions with various government departments over issues of gender, and of her fieldwork experience in West Bengal, added in important ways to my understanding of government policy responses to the gender question in the late 1970s and early 1980s. With Manimala I had several illuminating discussions on developments in the Bodhgaya move-

ment over the last decade, and I am most grateful to her for sparing the

time. I also thank the staff of the Institute of Economic Growth for facilitating my research on this book in various ways. To my parents I owe very special thanks for their unending generosity, patience, and interest in my work. I am especially grateful to my father, S. M. Agarwal, for reading through a draft of the manuscript on trains and planes and late into the night, and offering many valuable suggestions for clarifying my arguments for the non-specialist reader. Finally, to my friends

in Cambridge (Massachusetts) and Delhi goes my gratitude for their

Preface XXl forbearance while I closeted myself over long stretches of time to reflect on and write the book. It is my hope not only that this book will provoke serious academic and policy debate, but that the issue of women’s land rights will be given the

centrality it justly deserves by government policy makers, by political parties and, most of all, by gender-progressive grassroots groups. For it is on collective action by the women peasants of South Asia that change is _ ultimately likely to depend.

North West | Jammu &

Frontier Province Kashmir

P. ‘7 Himachal Pradesh Arunachal

Baluchistan irr Sikkim Uttar Nepali if

Rajasthan J Nagaland rf fradesh ; dy frTs Meghalaya

a Banglaie Manipur , a then Tripura Madhya Pradesh rs A\ desh (7 Mizoram

est

Maharashtra | Bengal Andhra

Pradesh

e ro arnataka

Lakshadweep < 4° : islands ——— rg

Kerala : 0 500 kms.

Sri Lanka .

Map 1.1 South Asia: provincial/state divisions

] Land rights for women: making the case

To my brother belong your green fields O father, while I am banished afar. Always you said Your brother and you are the same

O Father. But today you betray me... My doli leaves your house, O father My doli leaves your house. These dowry jewels are not jewels but wounds round my neck, O father. My doli leaves ...'

Rural women in northwest India, married among strangers miles away from their natal villages, use folksongs to decry their estrangement from the green pastures of their childhood homes — homes to which their brothers,

who customarily inherit the ancestral land, have automatic access. In Maharashtra (west India), women divorced or deserted by their husbands

can be found working as agricultural labourers on the farms of their brothers who are substantial landowners (Omvedt 1981). Elsewhere in India and in Bangladesh there are similar cases of widows who, deprived of their rightful shares by prosperous brothers or brothers-in-law, have been left destitute and forced to seek wage work or even beg for survival.2 Many poor rural women from Rajasthan and Bihar told me: we must get some land to take care of our children ... even a little land. In Bodhgaya (Bihar), in 1979, landless labourer women, agitating alongside their husbands for

ownership rights to the land they had sown for years, protested the 1 These verses are taken from folksongs sung by Hindu women in northwest India when the bride departs from her parents’ home. The first was translated by me. The second was given to me by Veena Das (Department of Sociology, Delhi University). Doli means palanquin.

2 For India: personal observation in Rajasthan (northwest India) and Bihar (east India) during 1986-88; and personal communication on West Bengal (east India) from Vina Mazumdar, Center for Women’s Development Studies (CWDS), Delhi. For Bangladesh see Abdullah and Zeidenstein (1982), Cain et al. (1979), and Schendel (1981). In one case described by Cain et al. (1979: 5-6) a widow was reduced to beggary after her husband’s death, although her brother-in-law was the richest man in the village. l

2 A field of one’s own distribution of titles only to men, noting: ‘If these men who are today landless beat up their wives so badly, merely using the power derived from being men, then tomorrow when they get the land will they not become relatively even more powerful? We are part of the struggle so we should also get land’ (Manimala 1983: 8). And in the hills of Uttar Pradesh (northwest India), women in the Chipko movement have been working along with the men of their community to protect and restore the forests on which their livelihoods depend. At times they have even gone against the wishes of the village men (including their husbands), to resist income-generating schemes that would have destroyed a local forest. ‘Planning without fodder, fuel and water’, they assert, ‘is one-eyed planning’ (Bahuguna 1991: 152). These images, and these voices of lament, protest and assertion that are beginning to resonate across South Asia today, highlight the multiple facets of rural women’s relationship with land, and the importance many attach to having a field of their own. Fora significant majority of rural households, arable land (an increasingly scarce resource) is likely to remain for a long time yet, the single most important source of security against poverty in rural South Asia, even if it ceases to be the sole source of livelihood for many. Land defines social status and political power in the village, and it

structures relationships both within and outside the household. Yet for most women, effective rights in land remain elusive, even as their marital and kin support erodes and female-headed households multiply. In legal terms, women have struggled for and won fairly extensive rights to inherit

and control land in much of South Asia; but in practice most stand disinherited. Few own land; even fewer can exercise effective control over it.

Yet the voice of the disinherited female peasant has until recently gone largely unheard, not only by policy makers but also by grassroots groups and academics. Instead, employment is taken as the principal measure of women’s economic status, obscuring what has been commonplace in measuring the economic status of men or of households: property ownership and control. This book argues that women’s struggle for their legitimate share in landed property can prove to be the single most critical entry point for women’s empowerment in South Asia; and it seeks to bring this issue from out of the wings onto centre stage.

I. The backdrop ! Two decades ago, the question: ‘Do women need independent rights in land?’ was not even admitted in public policy discourse in most parts of South Asia. Today, the question is admissible, but the discussion on it is limited and the answers to it disputed. Indeed gaining acceptance for the

Land rights for women 3 idea that women need independent rights in land is itself an arena of struggle, an essential first step in the struggle to translate that need into effective rights in practice. To begin with, to argue that women’s economic needs require a specific focus, distinct from those of men, is to challenge a long-standing assumption in economic theory and development policy, namely, that the household is a unit of congruent interests, among whose members the benefits of

available resources are shared equitably, irrespective of gender. This assumption has (until recently) been shared widely by governmental and non-governmental groups, institutions, and individuals. To go further and argue that women need independent rights in /and — the most critical form of property in agrarian economies — 1s to challenge the assumption that

women’s economic needs can be accommodated adequately merely through the employment and other income-generating schemes that typify development planning. It means admitting new contenders for a share in a scarce and highly valuable resource which determines economic well-being

and shapes power relations especially in the countryside; and it means extending the conflict over land that has existed largely between men, to men and women, thus bringing it into the family’s innermost courtyard. The process by which the assumption of a unitary household, and more generally of the gender-neutrality of development, has come to be challenged over the past twenty years is a complex one, which will not be detailed here. What is notable is that it has been a process of negotiation and struggle involving multiple actors — academics and researchers, women’s activist groups, government policy makers and bureaucrats, and international agencies. It was set in motion by at least three interrelated factors: the building up of gender-specific empirical evidence and analysis, especially since the mid-1970s, which exposed a systematic gender gap in how the benefits and burdens of development were being distributed; the mushrooming of women’s organizations loosely constituting a women’s movement, since the late 1970s; and changes in the international context. This last included, in particular, the declaration of 1975-85 as the United

Nations (UN) Decade for Women, with associated fall-outs in terms of research funding and dissemination, media coverage, and pressure on countries to generate gender-specific data and status of women reports.* 3 ‘Independent’ land rights are defined here as rights that are formally untied to male ownership or control, in other words, excluding joint titles with men. By effective rights in

land I mean not just rights in law but also their effective realization in practice, as elaborated later in this chapter. + Documents (Reports, Action Plans, etc.) from various international and national Conferences, Symposia and Working Groups, that met during 1975-85 to focus on rural women, provide interesting insights into the changing nature of concerns over this period. For a selected compilation of such documents (international, and those relating to India), see

4 A field of one’s own . Indirectly, feminist scholarship and activism in the West were also facilitating factors in promoting the issue internationally.

Today, as a result, the idea that development is not gender-neutral has gained fairly wide acceptance in development enquiry and policy, even though there is no consensus on the causes of the gender gap or on how it

could be bridged. At the level of policy, this recognition of gender disadvantage has been reflected particularly in three types of developments: —the establishment of separate cells, departments or ministries in government bureaucracies to monitor and coordinate women’s concerns in the development process;

—the incorporation of policy directives on women and development in the planning process, as in the Indian Sixth Five Year Plan, 1980-85 (for the first time in the history of planning in India), with subsequent plans following suit; and

—the initiation of special programmes targeted at women, especially income-generating and literacy schemes.

However, the approach underlying these directives and programmes treats gender as an additive category, to be added onto existing ones, with women asa special focus or target group, rather than seeing gender as a lens through which the approach to development should itself be re-examined.

The programmes are essentially couched in welfare terms, under the umbrella of the ‘basic needs’ approach that gained currency in development

thinking in the mid-1970s. This approach emphasizes the provision of ‘basic’ goods and services (such as food, health care, education) to the economically disadvantaged, but usually without seriously questioning the

existing distribution of productive resources and political power, or the social (gender/class/caste) division of labour. Most governments typically deliver such programmes in a top-down manner, involving little dialogue with the people (especially women) themselves on the definition of their needs or the best means of meeting those needs.

In this scenario, the issue of women’s land rights has, until recently, received little attention in policy formulation. In India, the numerous CWDS (1985). The Report, Towards Equality, on the status of women in India, was also a significant landmark (Government of India (GOI) 1974). Brought out by a Committee set up by the Indian Ministry of Education and Social Welfare, the Report compiled evidence of gender gaps in virtually every sector and made recommendations on how to bridge them.

The issue of women’s land rights, however, was not raised in the Report, although it included a discussion on gender inequalities in inheritance laws. On the role of international aid agencies in pushing the gender question, see especially White (1992) for Bangladesh. In India, I understand, international organizations such as the UN Food and Agricultural Organization (FAO) played an important role in pushing the government to set up review committees on rural women, such as the 1979 National Committee to Review and Analyse Participation of Women in Agriculture and Rural Development, set up by the Ministry of Agriculture (personal communication, Vina Mazumdar, 1992).

Land rights for women 5 committees and working groups on the status of women that met between 1975 and 1979, focused almost exclusively on three elements: employment, education and health.? It is only in the Sixth Five Year Plan (1980-85) that we see the first limited recognition of women’s need for land (and then only in the context of poverty). Several factors appear to have contributed to this

recognition. In 1979, at a women’s conference in Calcutta, a group of elected women gram panchayat (village council) representatives from West Bengal put forward a demand for joint titles (with their husbands) on behalf of destitute Muslim women in their constituencies. They argued that many

Muslim women had been evicted by their husbands; women therefore needed the economic security that land provides. This is said to be among the earliest such public grassroots demands. A similar plea was made by landless women in 1980 to a sympathetic Land Reform Commissioner at a camp in West Bengal’s Bankura district.© Such demands were subsequently included in the recommendations (placed before the Planning Commission) of a pre-Plan symposium organized by eight women’s groups in Delhi in 1980.7 Additional pressure came from the 1979 FAO Report of the World Conference on Agrarian Reform and Rural Development (WCARRD) held in Rome, which recommended that gender discriminatory laws in respect to ‘rights in inheritance, ownership and control of property’ be

repealed and measures be adopted to ensure that women get equitable access to land and other productive resources (FAO 1979). These recommendations were incorporated (albeit in very diluted form) in the country

review follow-up to WCARRD undertaken by the Indian Ministry of Agriculture and Rural Development (CWDS 1985: 89-94). The result of all this was a policy statement which, as finally incorporated in the Sixth Plan (in a separate chapter on women and development), said that the govern-

ment would ‘endeavour’ to give joint titles to spouses in programmes involving the distribution of land and home sites. However, even this limited formulation, which stops short of granting women independent titles, remained only a promise on paper. In practice, government land-redistribution programmes continued to reflect the old assumption of a unitary male-headed household, and titles were granted principally to men. In India’s Seventh Plan (1985-90), although a separate chapter on women and development was retained, the directive on joint

titles was not restated, despite strong recommendations for entitling women by a governmental working group on women and development, 5 See various Indian documents compiled in CWDS (1985). ® Both incidents were related to me in 1992 by Vina Mazumdar.

7 The group brought out a memorandum entitled: ‘Indian Women in the Eighties: Development Imperatives’ (CWDS 1985: 95-8). The gender-sensitive response of some State planners in authoritative positions within the Planning Commission was of critical importance in ensuring that such recommendations were taken seriously.

6 A field of one’s own during the plan-formulation stage.? Meanwhile, the National Perspective Plan For Women: 1988-2000 A.D., drawn up at the initiative of the Indian Ministry of Human Resource Development, made a number of substantive recommendations for closing the gender gap in access to land, amongst other gender issues needing attention (GOI 1988a). And the report of a National Seminar on Land Reform called by the Planning Commission in 1989, in which I had presented the case for women’s land rights, incorporated most of my recommendations on this count (GOI 1989a).?

Reports, however, have a tendency to gather dust, their contents forgotten. It is in this context that the passing of the National Commission for Women Act, 1990, is an important step forward. The result of many years of sustained efforts by women’s organizations and gender-progressive

individuals, this Act has created a Commission with a wide mandate to investigate and monitor ‘all matters relating to the safeguards provided for women under the constitution and other laws’ (GOI 1990a: 4). In particular, itis mandatory on the government to place any recommendations made by the Commission before both houses of Parliament (or, where relevant, before the state legislatures), along with a memorandum of actions taken or proposed to be taken by the government, and to give reasons in cases of non-acceptance of such recommendations.!° Of course, it remains to be seen what issues the Commission will focus on, and how much weight will be given by the government to its recommendations. Certainly, the recently formulated Eighth Five Year Plan (1992-97) for

India has left much of the responsibility for monitoring gender-related issues (including keeping tabs on the enforcement of social legislation), on

the National Commission for Women, and on women’s groups; the appointment of a National Commissioner of Women’s Rights is also proposed (GOI 1992a—b). This Plan document (unlike the Sixth and Seventh Plans) does not have a separate chapter on women and development, but subsumes women’s concerns largely under the chapter on social welfare (which also deals with children, the disabled, the elderly, and the destitute), and these concerns are couched essentially in the language of § This group, set up by the Department of Social Welfare, Government of India, made four recommendations concerning women and land: that land and other property be registered

in revenue records in the joint names of both spouses; that single women be given preference in land distribution by the government; that all property acquired after marriage be in the names of both spouses; and that loopholes in the Hindu Succession Act of 1956 be plugged (GOI 1983a). ° These recommendations were based on some initial research I had done on the subject in 1985-86, and published in 1988 in a widely circulated paper (Agarwal 1988) which had previously been presented in several academic and other forums, within and outside India. 10 For further details on the Act and on how the National Commission for Women came to be set up, see Women’s Equality (1992).

Land rights for women 7 women as victims, rather than women also as agents of change and contributors to development.!! The Plan makes two specific points in relation to women and agricultural land: one, it recognizes that ‘one of the

basic requirements for improving the status of women’ is to change inheritance laws so that women get an equal share in parental property, inherited or self-acquired (GOI 1992b: 392). However, it does not lay down any specific directives to ensure that this is followed through.'!? Two, and this is the only concrete policy directive, state governments have been asked to allot 40 per cent of surplus land (i.e. land acquired by the government from households owning land more than the specified ceilings) to women alone, and to allot the rest jointly in the names of the husband and wife (GOI 1992b: 34). This sounds good until one recognizes how little land 1s involved: only 1.04 million hectares (mha) remain to be distributed (GOI 1992b: 34). This constitutes just 0.56 per cent of the country’s arable land.!% In other words, the process of incorporating the issue of women and land into public policy in India has been extremely slow, involving negotiations

between the government, women’s groups, individual women academics, and international agencies, as well as between different elements within the government. And today, despite the noted progress, it remains an issue of marginal not central concern. The situation in other South Asian countries is even more discouraging. Nepal’s Eighth Five Year Plan (1992-97) Summary highlights women’s

employment and the need to encourage their participation in various activities, but contains no reference to women’s need for land.'* In Bangladesh, the latest Fourth Five Year Plan (1990-95) contains two special chapters on women and development, and some others incorporate

women’s concerns (Government of Bangladesh (GOB) 1990). But the emphasis throughout is on issues such as female employment, literacy, health, nutrition and credit; there is no mention of land for women, not ‘1 In contrast, although the Sixth and Seventh Plans also mentioned women’s concerns in their chapters on social welfare, it was their separate chapters dealing with women’s programmes which outlined the primary thrust of policy in this regard; and these were framed much more in the language of equality and rights, and recognized women’s productive contribution to the economy. Of course, in these documents, as noted, the issue of women’s land rights received marginal (Sixth Plan) or no (Seventh Plan) attention. 12 Indeed, as will be seen in chapter S, Indian women of most communities already have considerable /ega/ rights of inheritance (although gender gaps remain on several counts). It is in the implementation of laws that action 1s especially necessary. ‘3 Taking the aggregate of net sown area, fallow land (current and other fallows), cultivable wasteland, and land under miscellaneous tree crops and groves, the country’s arable land in 1987-88 was 184.73 million hectares (GOI 1992c). This tallies with the Ministry of Agriculture’s method of estimating arable land. ‘4 The full Plan document has yet to be released.

8 A field of one’s own even in terms of government allocations for poor women.!> Similarly, although Pakistan’s Report of the Working Group on Women’s Development

Program for the Sixth Plan (1983-88) recommended that all land distributed under the land reform programme should be registered jointly in the names of both spouses, this recommendation was not incorporated into the formal plan document. And Pakistan’s Eighth Five Year Plan (1993-98)

Approach Paper, in its chapter on ‘Affirmative Action for Women and other Disadvantaged Groups’ promises women preferential treatment in education and employment, but does not mention implementing their property rights. It also casts gender relations in traditional terms, with the State explicitly undertaking to ‘protect the marriage, the family, the mother and the child’ and to forego any approaches ‘which [could] antagonise male members of the community ...’ (Government of Pakistan 1991a: 22, 24).

What is especially striking is the disjunction between public policy formulation and the rights encased in personal law. The idea of women having independent property rights (including rights in land) was accepted

by most South Asian countries in laws governing the inheritance of personal property in the 1950s (and even earlier in traditionally bilateral and matrilineal communities).'© But such acceptance remained confined to inheritance laws that affect private land; in development policy governing the distribution of public land, the issue of women’s land rights was not discussed (as we’ve noted) till the 1980s. Hence the redistributive land reform programmes of the 1950s and 1960s in India, Pakistan, and Sri Lanka, and of the 1970s in Bangladesh, continued to be modelled on the notion of a unitary male-headed household, with titles being granted only to men, except in households without adult men where women (typically 15 In 1991, however, a Task Force set up by the Bangladesh Ministry of Planning to review the

country’s development strategies made a modest recommendation that female heads of households, with or without adult sons, and women in households with incapacitated male heads, be given priority in the distribution of government land (see Report of the Task Force on Bangladesh Development Strategies for the 1990s (1991)). At present, the Report notes,

under the conditions laid down by the Bangladesh Land Ministry in 1987 for the distribution of government land to the landless, women can be given priority only if they are widowed or abandoned and have an adult son who is able to work. It remains to be seen whether the Task Force recommendations will be acted on. 16 Bilateral inheritance: ancestral property passes to and through both sons and daughters;

matrilineal inheritance: ancestral property passes through the female line; patrilineal inheritance: ancestral property passes through the male line. The specific, complex workings of such inheritance systems in South Asia will be discussed in later chapters. The

terms ‘matrilineal’, ‘bilateral’, and ‘patrilineal’ will be used throughout the book (unless otherwise specified) to relate to inheritance practices, and not to those of descent. In any case, in all the communities referred to in the book, those following any one of these inheritance systems also practised the same type of descent system, with the exception of the Nangudi Vellalars who practised matrilineal descent and bilateral inheritance.

Land rights for women 9 widows) were clearly the heads. This bias was replicated again in resettlement schemes, even in Sri Lanka where customary inheritance systems have been bilateral or matrilineal.

Underlying this disjunction between government policy in relation to public land distribution and the rights in private land granted to women under inheritance laws are likely to be a complex set of factors. These would

include the continued assumption in most public policies of gendercongruence in interests within the family; the dominant view that men are the breadwinners and women the dependents; strong male vested interests in all land, including public land; gaps between the central government’s policy directives and the shape these are given at the state/province level;'” and the belief that land distribution to women will further decrease farm size and fragment cultivated holdings, in turn reducing agricultural productivity. The farm size and fragmentation arguments have also been used in many Indian states to undercut post-independence, gender-progressive personal laws,!® by retaining age-old customary laws that disadvantage women in relation to agricultural land. The weaknesses in these arguments will be discussed later in this chapter. Here it suffices to reiterate the limited progress made in public policy towards entitling women with land and the ambiguities that continue to surround even the idea of doing so. A similar ambiguity toward this issue is found among groups which have otherwise been strong advocates of redistributive land reform, namely Marxist political parties and left-wing non-party organizations, most of whom still see class issues as primary and gender concerns as divisive and distracting.!° At the same time, most women’s organizations (whatever their political persuasion), with some recent exceptions, have been preoccupied with employment and non-land-related income-generating schemes as the means of improving women’s economic status and welfare, paying little '7 In India, the term ‘state’ relates to administrative divisions within the country and is not to be confused with ‘State’, used throughout the book in the political economy sense of the

| word. In Pakistan and Sri Lanka these administrative divisions are termed provinces. '8 The term ‘gender-progressive’, as used here and subsequently, relates to those laws, practices, policies, etc., which reduce or eliminate the inequities (economic, social, political)

that women face in relation to men. Individuals and organizations that work toward this end are also so described. ‘“Gender-retrogressive’ has the opposite meaning.

'9 It is noteworthy that in West Bengal when the CPI(M) (Communist Party of India (Marxist)) government carried out ‘Operation Barga’ (launched in 1978), a major land reform initiative which sought to provide tenants with security of tenure by systematically registering them, primarily men were registered. A similar male bias has characterized the programmes of most left-wing non-party groups, among the notable exceptions being the Bodhgaya (Bihar) peasant movement initiated in 1978 by the Chatra Yuva Sangharsh Vahini, a Gandhian-socialist Youth organization which also took up the issue of women’s land rights (see chapter 9 for details).

10 A field of one’s own attention to the issue of property rights.7° Several years ago, when I began research on this subject and raised the question of women’s land rights with a number of left-wing women’s groups across South Asia, the responses of

most were either that ‘we haven’t really thought about it’, or that advocating individual property rights went against their vision of a socialist society. Yet, to my knowledge, this latter argument has not been used in South Asia against redistributive land reform or peasant struggles through which (typically male) heads of landless households gain rights in land.?? This neglect of women’s land-related concerns by both governmental and non-governmental institutions mirrors a parallel gap within academic scholarship, where the relationship between women and property has remained virtually unattended and little theorized. For instance, the social science literature on rural South Asia of relevance to this discussion falls broadly into three categories. First, a vast body of economic development and political science studies document a strong interdependence between

the rural household’s possession of agricultural land and its relative economic, political and social position. Characteristically, these studies

focus on the household as the unit of analysis, neglecting the intrahousehold gender dimension. Second, there is a substantial body of (primarily descriptive) sociological and anthropological literature on South Asia, especially that relating to

kinship and marriage. From this, a picture can be constructed of some aspects of women’s position in different communities, socio-economic strata and parts of the subcontinent. But even in the best of ethnographies

up to the 1970s, the analysis is typically ungendered. Women appear 20 Among the exceptions is the Shetkari Sanghatana’s Mahila Aghadi, the women’s front of the Shetkari Sanghatana — a farmers’ organization founded in Maharashtra (west India) in 1980 (see chapter 9 for details). Also noteworthy is the role played by Manushi (a women’s journal from India) in reporting such initiatives, and by one of the journal’s founders, Madhu Kishwar, who in 1982 filed a petition in the Supreme Court of India challenging the denial of land rights to Ho tribal women in Bihar (see Kishwar 1982, 1987). 21 Joshi (1974) who explains the background to the formulation of land reform programmes

in post-independence India and Pakistan, makes no mention of any resistance to redistributive reform on these grounds. Rather he notes (1974: 167): ‘The fundamental question of land policy was the question of removing [the] discrepancy between ownership of land and its actual cultivation’; ownership being largely concentrated in the hands of a

minority of landlords and cultivation being done by peasants with usually limited or no proprietary rights. However, in a personal communication to me in 1992, Joshi added that a minute section of the left did express unease about measures that could strengthen individualistic tendencies among the peasantry, but this was not a widely shared concern: the preoccupation of most was with the need to break the stranglehold of ‘feudal’ elements. What was discussed widely, though, both by the Planning Commission and various political parties, was the need to encourage (largely voluntary) cooperation among the peasantry in various forms, including joint cultivation, the joint ownership of non-land assets, cooperative marketing and distribution, etc. (On this debate and the limited success of efforts in this direction, also see Frankel 1978.)

Land rights for women I] principally as objects of study and exchange, not as subjects; they are occasionally seen but rarely heard; their presence is registered but seldom their perspective; and gender relations are depicted as essentially unproble-

matic. Usually (if not universally) implicit in these descriptions is the assumption that the underlying basis of women’s social subordination (typically defined in terms of women’s roles) is the cultural values of the community to which they belong. In this emphasis on the ideological, the possible material basis of this subordination, or the dialectical link between the material context and gender ideology, is seldom recognized.2? And,

culture is often characterized as ‘given’ rather than in the process of constant reformulation, or as an arena of contestation. Third, over the last decade and a half a body of work has emerged which incorporates gender analysis in diverse ways. This includes some gendersensitive ethnographies which bridge important gaps (mainly on women’s work and roles) and a spectrum of other studies which can loosely be

characterized as ‘women and development’ literature. This literature examines gender biases in economic development, often giving primacy to women’s economic position as a significant indicator of gender inequality, and sometimes also as a causal factor underlying non-economic dimensions

of that inequality. But the measure of women’s economic status is still typically employment and labour force participation, not property rights. In giving centrality to the issue of gender and land distribution, the question which gets spotlighted before all others is the one we began with: why do women in South Asia need independent rights in land? An answer to this question is attempted in section III. Before ‘making the case’ however, it is useful to consider some of the wider conceptual links between gender and property, and to explain why a focus on /anded property is important, and what I mean by ‘rights’ 1n land.

II. Gender, property, and land: some conceptual links The relationship between gender, property, and land can be explored from

several angles.*° In the present discussion, six interrelated issues need particular focus: gender relations and a household’s property status; gender relations and women’s property status; the distinction between ownership

and control of property; the distinctiveness of land as property; what is meant by rights in land; and prospects of non-land-based livelihoods. The first three issues are discussed in interrelation in the subsection below, and the last three in separate subsections. 22 Two notable exceptions are U. Sharma (1980) and Kishwar (1987). 23 Also see ‘introduction’ in Hirschon (1984).

12 A field of one’s own (1) Household property and women’s property The links between gender subordination and property need to be sought in

not only the distribution of property between households but also in its distribution between men and women, in not only who owns the property but also who controls it, and in relation not only to private property but also to communal property. Further, gender equality in legal rights to own property does not guarantee gender equality in actual ownership, nor does ownership guarantee control. The distinctions between law and practice and between ownership and control are especially critical in the context of gender: for most South Asian women there are significant barriers to realizing their legal claims in landed property, as well as to exercising control over any land they do get (as will be elaborated in chapters 6 and 7).

This formulation departs significantly from standard Marxist analysis, particularly from Engels’ still-influential, though much-criticized, The Origin of the Family, Private Property and the State, where intra-family

gender relations are seen as structured primarily by two overlapping economic factors: the property status of the households to which the women belong, and women’s participation in wage labour. Engels argued that in capitalist societies, gender relations would be hierarchical among the property-owning families of the bourgeoisie where women did not go out to work and were economically dependent on men, and egalitarian in propertyless proletarian families where women were in the labour force.

The ultimate restoration of women to their rightful status, in his view, required the total abolition of private property (i.e. a move to socialism), the socialization of housework and childcare, and the full participation of women in the labour force. In the context of industrializing Europe, Engels (1972: 137-8) argued: ‘the first premise for the emancipation of women is the re-introduction of the entire female sex into public industry’.*+ In his analysis, therefore, the presumed equality of gender relations in a working class family rested on both husband and wife being propertyless and in the labour force, and the inequality in the bourgeois family rested on men being propertied and women being both propertyless and outside the labour force. This underlying emphasis on the relational aspect of gender is clearly important. So is the emphasis on women’s economic dependency as 24 This is not meant as a summary of Engels’ complex thesis, but merely of one part of his argument. Critiques of different aspects of Engels’ analysis abound: see especially, Sacks (1975), Reiter (1977), Aaby (1977), Barrett (1980, 1985), Coward (1983), MacKinnon (1989), Delmar (1976), Molyneux (1981), and various articles in Sayers et al., eds. (1987). In particular, Engels’ assumption that gender relations within propertyless groups such as the industrial proletariat or under socialism would necessarily be egalitarian has been widely

98S) in the literature: see especially, Delmar (1976), Molyneux (1981), and Barrett

Land rights for women 13 a critical constituent of the material bases of gender oppression. However, by advocating the abolition of all private property as the solution, Engels by-passed the issue of women’s property rights altogether, and left open the

question: what would be the impact on gender relations in propertied households if the women too were propertied as individuals? Entry into the labour force is not the only way to reduce economic dependency; independent rights in property would be another, and possibly the more effective way.

Engels’ emphasis on women’s entry into the labour force as a necessary

condition for their emancipation has been enormously influential in shaping the thinking of left-wing political parties and non-party groups, including left-wing women’s groups in South Asia.?°> As noted, they too give centrality to women’s employment, but the necessary accompaniments

emphasized by Engels, namely the abolition of private property in male hands and the socialization of housework and childcare, have largely been neglected, as has the question of women’s property rights. In my argument that independent property rights can play a pivotal role in women’s struggle for equality in gender relations, a critical additional point (missed out in Engels’ analysis and associated discussions) that needs

emphasis is that of property control. Property advantage stems not only from ownership, but also from effective control over it. In societies which underwent socialist revolutions, while private property ownership was legally abolished, control over wealth-generating property remained predominantly with men; any positive effects on gender relations that could have stemmed from the change in ownership if accompanied by genderegalitarian mechanisms of control, thus went unrealized.?° Indeed in most societies today it is men as a gender (even if not all men as individuals) who

largely control wealth-generating property, whether or not it is privately owned, including as managers in large corporations. Even property that is under State, community, or clan ownership remains effectively under the

managerial control of selected men through their dominance in both traditional and modern institutions: caste or clan councils, village elected 25 Also see Molyneux (1981) who describes how in socialist countries (including whose which were socialist until recently), the influence of Engels’ analysis led to a similar preoccupation with women’s entry into employment as the major means of eliminating gender oppression.

At the time of Molyneux’s writing, although such entry had taken place in significant degree in most of these countries, the types of jobs women held were largely at the lower end of the job hierarchy; progress toward the socialization of housework was extremely limited;

and the ideological basis of gender oppression (neglected by Engels) had persisted, in greater or lesser degree. 26 Women’s representation in the top political and economic decision-making bodies in such countries remained minimal. For instance, in the late 1970s, in the USSR, Czechoslovakia, Poland, and Yugoslavia, of some 557 top government posts only 27 (that 1s under 5 per cent) were filled by women (Molyneux 1981).

14 A field of one’s own bodies, State bureaucracies at all levels,2’ and so on. Also in most countries,

men as a gender exercise dominance over the instruments through which their existing advantages of property ownership and control get perpe-

tuated, such as the institutions that enact and implement laws,?* the mechanisms of recruitment into bodies which exercise control over property (private or public), the institutions which play an important role in shaping gender ideology, and so on. A second issue which arises in exploring the relationship between gender and property is: how do we define a woman’s class? Marxist analysis, for

instance, implicitly assumes that women belong to the class of their husbands or fathers. Hence women of propertied ‘bourgeois’ households

are part of the bourgeoisie and women of proletarian households are counted as proletarian, although they may also have a proletarian status by virtue of being workers themselves. However, as is now well-recognized in

feminist literature, this characterization 1s problematic in at least two respects: (a) A woman’s class position defined through that of a man — father, husband, etc. — is more open to change than that of a man: a wellplaced marriage can raise it, divorce or widowhood can lower it. As Millett (1970: 38) notes: “Economic dependency renders [women’s] affiliations with

any class a tangential, vicarious, and temporary matter’. (b) To the extent that women, even of propertied households, do not own property themselves, it is difficult to characterize their class position;?? some have even argued that women constitute a class in themselves.*° In fact, neither deriving women’s class from the property status of men

nor deriving it from their own propertyless status appears adequate, 27 See chapter 10 for figures and a further discussion on this. *8 Scandinavian countries have a better record than most others on this count: in Norway and Finland, for instance, women constituted 34 and 32 per cent of all elected and appointed

members of national legislative bodies in 1985-87. This contrasts sharply with the analogous figures for India, Bangladesh, and Pakistan which ranged between 8 and 10, as well as with those for the USA and UK which were 5.3 and 6.3 respectively (United Nations 1990).

29 Also, property differences alone do not distinguish classes. Education, life-styles, and so on, count as well (see especially, Bourdieu 1984). Similarly a group of persons wielding power or authority, which may or may not stem from property ownership, may be seen to constitute a class. On the concept of ‘class’ within Marxist and non-Marxist literature, also see Wolff and Resnick (1989). 39 While several feminist authors have denied the significance of class divisions between women, they have done so from different standpoints. Millett (1970) did so because she saw women’s class affiliations as basically impermanent. Firestone (1970) saw women as united by their biology; she substituted the term ‘sex’ for ‘class’ and categorized women as a ‘sex class’ as opposed to an economic class. Delphy (1977), in a more sophisticated analysis, rooted the problematics of defining women by class in their not owning the means of production and their economic vulnerabilities with marital break-up, even when married to men from the capitalist class. She located the material basis of women’s oppression in patriarchal exploitation which, she argued, cuts across classes. Also see the discussions in Barrett (1980) and MacKinnon (1989).

Land rights for women 15 although both positions reflect a dimension of reality. Women of large landed households in South Asia do gain from their husbands’ class positions in terms of their overall living standards, their typically lower work burdens, the social status and influence they can command in relation to other village women, and so on. Hence property mediates relationships not only between men and women but also between women. At the same time, there are significant commonalities between women which cut across derived class privilege (or deprivation), such as vulnerability to domestic violence; all women’s responsibility for housework and childcare (even if not all women are obliged to perform such labour themselves — the more affluent ones can hire helpers); gender inequalities in legal rights; and the

risks of marital breakdown due to which even women of rich peasant households can be left destitute and forced to seek wage work, reflecting their propertyless state and economic vulnerability as women. In other words there is an ambiguous character to women’s class position.

This complexity impinges with critical force on the possibilities of collective action among women, again in a double-edged way. Class differences among women, derived through men, can be and often are divisive in terms of relative economic privilege or deprivation, the associated ability (or lack of ability) to dominate women’s groups,*! perceptions about which aspects of gender relations need challenging, willingness to engage in collective struggle, and so on. At the same time, the noted commonalities between women’s situations and the relatively vicarious character of their class privilege make class distinctions between them less sharp and divisive than those among men, and could provide the basis for collective action on several counts (as will be elaborated in chapters 9 and 10).32

A third significant aspect of the relationship between gender and property concerns the links between gender ideology and property. Several types of interconnections impinge on our discussion, such as those outlined below: (a) Gender ideologies can obstruct women from getting property rights.

For instance, ideological assumptions about women’s needs, work roles, capabilities, and so on, impinge on the framing and implementation of public policies and laws relating to property (as noted earlier). Again, ideas about gender underlie practices such as female seclusion,

control of women’s mobility and sexual freedom, and so on. These ideologies and associated practices restrict women’s ability both to exercise their existing property claims and to successfully challenge 31 Qn this, see especially Dixon (1978: chapter 6) and Caplan (1985). 32 There are of course aspects of a person’s identity other than class which also can be divisive or adhesive, such as caste, ethnicity, and religion.

16 A field of one’s own persisting gender-inequalities in law, policy, and practice in relation to such claims (as will be detailed in chapters 6 to 8). Hence ideological

struggles are integrally linked to women’s struggles over property rights.

(b) How property ownership and/or control is socially distributed can significantly affect ideological constructions, including those of gender.

Those who own and/or control wealth-generating property can exercise considerable direct or indirect control over the principal institutions that shape ideology, such as educational and religious establishments and the media (defined broadly to include newspapers, TV, radio, film, theatre, as well as literature and the arts). These can be instrumental in shaping views in either gender-progressive or genderretrogressive directions. But to the extent that such institutions represent a plurality of views there are possibilities of contestation over ideological constructions through them. (c) The impact of gender ideologies on women can vary according to their households’ property status (other sources of ideological variations such as religion, caste, etc., being held constant). This variation could result from two points of difference between propertied and propertyless households: in what ideas about gender are dominant, and in how these ideas are put into practice. For instance, in both propertied and propertyless households the ideology of female seclusion may be espoused, but the former group would be in a better economic position

to enforce its practice, and in so doing reinforce its emulation by unpropertied households as a mark of social status. But would there also be differences in the ideologies espoused within propertied and propertyless households? Over the years there has been considerable debate on the degree to which ideological constructions of gender are autonomous of economic circumstances (excellently summarized in Barrett 1980). The view I take in this book is that gender ideologies and

associated practices are culturally specific, historically variable, and dialectically linked to property ownership and control. The form and practice of gender ideologies can differ among propertied and unpropertied households; at the same time, gender ideologies and associated practices are not derived from property differences alone, nor can they be seen in purely economic-functional terms. Rather they would tend to shift and change in interaction with economic shifts.?3

A fourth issue that arises in relation to women and property is the possible links of women’s property rights with control over women’s 33 For illustration, see the discussion on female seclusion in chapter 9.

Land rights for women 17 sexuality, marriage practices, and kinship structures. Engels argued, for instance, that in propertied households the need to ensure the legitimacy of heirs would necessitate strict control over women’s sexuality within marriage and provide the logic for monogamy, while such control would be unnecessary in propertyless families. That is, he saw the exercise of control over women’s sexuality in essentially economic-functional terms. The observed emphasis on monogamy and male supremacy over women even in European working class families and under socialism, that is even when the presumed material necessity for that control was absent, has been widely

used to criticize Engels’ formulation.3+ But his argument raises another question: would women with independent rights in property (a category of person which, as noted earlier, he did not consider) be subject to greater or lesser familial control than those without them? Goody (1976) argues the former; he proposes that in societies which recognize women’s inheritance rights in parental property, in order to keep the property intact and within their purview, families or kin-networks would tend to emphasize and ensure women’s pre-marital virginity, and control women’s choice of marriage partners and post-marital residence. Some of the problems with Goody’s formulation are elaborated in chapter 3. The relationships themselves are, however, significant ones to explore and will be examined in the book in some detail for South Asian communities.

(2) The significance of land as property We want [arable] land, all the rest is humbug. (A landless woman in Andhra Pradesh)?5

Thus far our discussion has revolved around property in general, but not all forms of property are equally significant in all contexts, nor are they equally coveted. In the agrarian economies of South Asia, for instance, arable land is the most valued form of property, for its economic as well as its political and symbolic importance. It is a productive, wealth-creating,

and livelihood-sustaining asset. Traditionally it has been the basis of political power and social status. For many, it provides a sense of identity and rootedness within the village; and often in people’s minds land has a durability and permanence which no other asset possesses.3° Although 34 Moreover, notions about ‘legitimate heirs’ vary widely across cultures and are not linked everywhere to monogamy or wedlock, even among propertied households (as will be discussed in chapter 3). 35. This was in answer to a query by Mies et al. (1986: 134), to a group of landless women in Andhra Pradesh, about whether they wanted better houses. 36 See e.g. Selvadurai’s (1976) observations on a Sinhalese village in Sri Lanka.

18 A field of one’s own other forms of property such as cash, jewellery, cattle, and even domestic goods (the usual content of, say, dowry in rural India and Nepal) could in principle be converted into land, in practice rural land markets are often constrained, and land 1s not always readily available for sale.>’ In any case, ancestral land usually has a symbolic meaning which purchased land does not: within some village communities, continuity of ancestral land also stands for continuity of kinship ties and citizenship (Selvaduri 1976); in some others it has ritual importance;>* and so on. Hence in land disputes people are often willing to spend more to retain a disputed ancestral plot than its market value would justify.°° Also inheritance systems often have different rules for the devolution of ancestral and self-acquired land (see chapter 3 for details).

In other words, both the form that property takes and its origin are important in defining its significance and the associated possibility of conflict over it. Hence, for instance (as will be elaborated in chapter 3), in terms of the possible relationship between a woman’s property rights and the contro] that her kin might seek to exercise over her sexual and marital

choices, we would expect families to be much more concerned about keeping /anded property intact and under their control, than about keeping

control over movables given say in dowry,*° even if such control over movables were a practical possibility. 37 Rosenzweig and Wolpin (1985: 978), in an all-India sample of 1,523 landowning farm households in 1970-71, found that less than 1.75 per cent had sold land during the survey year. Again, Shankar (1990: 27) found that in Uttar Pradesh (northwest India), over a period of thirty years (1952-53 to 1982-83), only 4.1 per cent of owned agricultural land was sold, viz. an annual average of 0.14 per cent. The sellers typically owned less than 2.5 acres. In Bangladesh, similarly, Wallace et a/. (1988: 112), observed that between 1983-86, | in two villages near Dhaka, only 9 and 12 per cent respectively of the land acquired was

through purchase; 83 and 86 per cent was through inheritance and the rest via gifts (including dowry). Rural land sales on any significant scale in South Asia have usually tended to be under distress circumstances. Historically custom restricted the sale of ancestral land (see e.g. Rattigan 1953, on the Punjab). From the mid nineteenth century, large-scale land transfers were largely in the nature of distress sales, say by severely indebted peasants to moneylenders in the Punjab and the Bombay Presidency (Barrier 1966, Charleston 1985); or by Starving peasants during the 1943 Bengal famine (Mahalanobis er al. 1946). Also see Binswanger and Rosenzweig (1986a) on the theoretical reasons why in areas with poorly

developed capital markets land is sold mainly for distress reasons, and why in such circumstances it accumulates with persons with already large holdings.

38 For instance, in northwest Nepal, Krause (1982) found that land inherited from an ancestor was ritually associated by some communities with the kuldevta (ancestral God). 39 Selvaduri (1976) in his study of a Sinhalese village, describes several bitter disputes around

ancestral land, including one where a woman fought her brothers for twelve years for a small plot of land, in the process spending more than its market value. 4° For a definition of dowry see the appendix on ‘definitions’ (p. 505).

Land rights for women 19 (3) What do we mean by rights in land? Rights are defined here as claims that are legally and socially recognized and enforceable by an external legitimized authority, be it a village-level institution or some higher-level judicial or executive body of the State.*! ' Rights in land can be in the form of ownership or of usufruct (that is rights of use), associated with differing degrees of freedom to lease out, mortgage, bequeath, or sell. Land rights can stem from inheritance on an individual or

joint family basis, from community membership (e.g. where a clan or village community owns or controls land and members have use rights to it), from transfers by the State, or from tenancy arrangements, purchase, and so on. Rights in land also have a temporal and sometimes locational dimension: they may be hereditary, or accrue only for a person’s lifetime, or for a lesser period; and they may be conditional on the person residing where the land is located, say, in the village. As distinct from rights in land, we could also speak of ‘access’ to land, a

term very loosely used in the development literature that needs some clarification. Access can be through rights of ownership and use, but it can

also be through informal concessions granted by individuals to kin or friends. For instance, a man may allow his sister to use a plot of his owned

land out of goodwill, but she cannot claim it as a right and call for its enforcement. There are thus several possible ways by which a woman may,

in theory, have access to land, but of these, having ‘rights’ provides a measure of security that the others typically do not. In relation to land rights, four additional distinctions are relevant. First we need to distinguish between the /ega/ recognition of a claim and its social recognition, and between recognition and enforcement. For instance, a woman may have the legal right to inherit property, but this may remain merely a right on paper if the law is not enforced, or if the claim is not socially recognized as legitimate and family members exert pressure on the woman to forfeit her share in favour, say, of her brothers. Second, as noted earlier, is the distinction between the ownership of land and its effective control. (Control itself can have multiple meanings, such as the ability to decide how the land is used, how its produce is disposed of, whether it can be

leased out, mortgaged, bequeathed, sold, and so on.) It 1s sometimes assumed incorrectly that legal ownership carries with it the nght of control in all these senses. In fact legal ownership may be accompanied by legal restrictions on disposal: for instance, among the Jaffna Tamils in Sri Lanka, a married woman needs her husband’s consent to alienate land which she 41 Also see Bromley (1991), Feder and Feeny (1991), and Friedman (1977) for some useful discussions on property rights, as well as on rights in land.

20 A field of one’s own legally owns. Or there may be no legal restriction on disposal but social constraints on doing so: for instance, the sale of ancestral land to strangers is often socially disapproved by kin and the village community. Third, it is

important to distinguish between ownership and use rights vested in individuals and those vested in a group; and fourth one might distinguish ~

between rights conferred via inheritance and those conferred by State transfers of land (as will be elaborated further on). Although the different possible forms (ownership or usufruct, as vested in individuals or in groups, and so on) that rights in land can take are not equivalent, it is not always obvious a priori which may be superior: much depends on the accompanying organization of production and distribution. Hence, for instance, in terms of fulfilling an individual’s consumption needs, communal ownership by a clan, where all members (irrespective of rank or sex) have equal use rights, cannot a priori be designated as inferior to a system where each member privately owns a plot. Similarly, legal ownership of land without effective control over its use and the disposal of

its produce need not leave a person better off than if s/he had control without legal ownership. Given the above complexities, it is not possible to specify with precision for all contexts what may be the most desirable form for women’s rights in

land to take, but a specification in broad terms is attempted here. When _ speaking of the importance of women having ‘independent rights in land’ I mean effective rights, that is rights not just in law but in practice. This needs emphasis, given the popular tendency to equate property rights only with legal rights. When referring to legal rights alone I will say so explicitly. By ‘independent rights’, as explained earlier, I mean rights independent of

male ownership or control (that is excluding joint titles with men). Independent rights would be preferable to joint titles with husbands for several reasons: one, with joint titles it could prove difficult for women to gain control over their share in case of marital breakup. Two, women would

also be less in a position to escape from a situation of marital conflict or violence; as some Bihari village women said to me: ‘For retaining the land we would be tied to the man, even if he beat us.’ Three, wives may have different land-use priorities from husbands which they would be in a better

position to act upon with independent land rights. Four, women with independent rights would be better placed to control the produce. Five,

with joint titles the question of how the land would subsequently be inherited could prove a contentious one. This is not to deny that having joint titles with husbands would be better for women than having no land rights at all; but many of the advantages of having land would not accrue to

women by joint titles alone. .

Here the distinctions mentioned earlier between rights vested in

Land rights for women 21 individuals and those vested in groups, and between privatized land transfers via inheritance and land transfers by the State, need elaboration. In relation to privatized inheritable landed property, by effective land rights for women today I mean inheritance as individuals linked with full rights of

control over land use (viz. sale, bequest, etc.) and the disposal of its produce. Where land transfers by the State to women are involved, the issue is more complex. Here effective land rights for women could either mean

individual titles conferring ownership and control rights exactly as with private land; or they could take the form of land transfers to groups of women (such as to those belonging to landless households) who could hold it in joint ownership or long lease, having full control over land use and disposal of its produce, but excluding the right to sell or bequeath the land. Although many of the potential advantages for women in having rights in land would accrue whichever of the above forms those rights take, some advantages are specific to the form: for instance, individually owned land

can be sold or mortgaged, which could be advantageous in distress circumstances. But group rights over land which women could use for joint or individual projects could, in certain contexts, strengthen women’s ability to retain the land and to use 1t more productively (as will be elaborated in chapter 10). When specific issues such as these are involved, the discussion will seek to make clear what form of rights I mean, even though it is not possible to so nuance every step of the argument. Finally, some further elaboration is necessary on a specific form of group

rights in land that has both historical and contemporary importance in South Asia, namely, usufruct rights in village common lands and in forests. In most South Asian village communities in the pre-British period and even

up to the mid nineteenth century, a significant percentage of land was available for communal use. This included forests, woodlots, pastures, and

other multiple-use land. Communal access to these lands took complex forms, but could be seen as linked in broad terms to the prevailing (also varied) systems of village land tenure. For instance, at the time of British accession to power, in villages of India where individual peasant proprietorship prevailed, cultivated land was separately held by each peasant household. But all uncultivated land of the village, often consisting of reportedly vast tracts extending beyond the fields, could be used by the villagers for grazing, for gathering multiple forest produce, as well as, in consultation with the village headman, to extend cultivation. This system (or variations thereof) was reportedly common in western and southern India.** In contrast were villages where group proprietorship prevailed +2 See e.g. Baden-Powell (1892, 1896). Whether one can see this form of tenure as a characteristic feature of these regions is more contentious, given the variety of land systems existing therein (see e.g. Fukazawa 1984, D. Kumar 1984, and Stokes 1978).

22 A field of one’s own (also termed as the ‘joint-village’ system):+3 here a founding family or clan

claimed proprietorship of the village’s cultivated and uncultivated land. Male members of the family/clan held ancestral shares in the uncultivated tracts which could technically be partitioned but, in practice, were left undivided for use by the family/clan members, as well as for the use of other village residents serving as tenants to or servicing the founding family(ies).

This system was to be found especially in northwest India.** Even here, however, there were additional common lands outside the village — scrub forests, hill grazing runs, riverain grazing tracts, open ‘primeval’ wastelands, etc. — to which the villagers had easy access (Kaul 1990). Hence, unlike individually held land, communal lands within and outside the village bounds were accessible, in one way or another (restrictions being

greater under the group-proprietorship system), to all members of the village, irrespective of sex, age, caste, and ethnicity. Where individual peasant proprietorship prevailed, women had grazing and gathering rights on these lands by virtue of their membership in the village community, rights typically neither mediated by dependency relationships on men, nor dependent on inheritance, purchase, or tenancy which governed access to private land. Where group proprietorship prevailed, although access to village common land was more restricted, women resident in the village still had significant rights of grazing and gathering.*5 Such communal land has always contributed in important ways to the livelihoods of rural households, especially of poor households, serving as pastures and as sources of a wide range of subsistence items: fuel, fodder, fibre, food products, medicinal herbs, materials for house construction, and so on. Even today, village commons (VCs) provide a significant percentage

of the needs of the landless and landpoor.*® Much of the gathering of 43 See e.g. Baden-Powell (1896), and Mukerjee and Frykenberg (1969). +4 For a detailed description of this system in undivided Punjab and the Delhi region, see Rattigan (1953) and Kaul (1990); also see Baden-Powell (1892, 1896). +5 For instance, male members of the village proprietary body (the maliken-deh) in Punjab had inheritance rights in the cultivated land, as well as ancestral shares in the unpartitioned

communal land. Widows and daughters of the proprietary families had no inheritance rights in this communal land, but did have use rights in it as long as they resided in the village (Kaul 1990). The use rights of women belonging to the families of tenants or service castes would have been circumscribed by their own and their households’ relationship with the proprietor families.

4° Jodha’s (1986) study of semi-arid regions in seven Indian states during 1982-85 showed that among the rural landless and landpoor households, VCs account for 9-26 per cent of total household income, 91-100 per cent of the firewood used, and 69-89 per cent of the grazing needs. Gathered food items are particularly important during extreme food crises. Forests serve a similar function. An estimated 30 m people in India depend on non-timber forest produce for a livelihood (Kulkarni 1983). The dependence on communal sources of firewood is especially critical, since this is the single most important source of domestic fuel in rural South Asia, and in the villages is still largely gathered rather than purchased (Agarwal 1986a, 1987).

Land rights for women 23 needed items is done by women and children. Where women’s access to the cash economy is limited, access to communal land can serve as a means of some independent income and economic support. However, the availability of communal land has been declining rapidly, both in quantity and quality, especially since the mid nineteenth century, and noticeably so over the last three decades: between 1950-84, the decline

in area under communal access (now largely in the form of village commons) ranged between 26 and 63 percentage points across seven Indian states, and the productivity of such land also fell (Jodha 1986). Much of the area decline was due to the dual processes of Statization (appropriation by

the State) and privatization (appropriation by individuals). In India, for instance, both under colonial rule and continuing in the post-colonial period, State control over forests and village commons has expanded, with selective access being granted to a favoured few.*’ Especially in the late nineteenth century, the British established State monopoly over forests, and cut large tracts for timber extraction and for the setting up of railways and shipbuilding. They also severely curtailed the customary rights of local populations over these resources,*® granted private contractors the right to clear forest land for establishing tea and coffee plantations, and encouraged

farmers to expand crop area to increase the government’s land revenue base. Post-colonial policies have continued this process, with forest land also being increasingly absorbed for commercial exploitation, urban/ industrial expansion, and large-scale irrigation and other development schemes.*? The nationalization of non-timber forest produce has further reduced the availability of such products to villagers. The privatization of community resources in individual hands has paralleled the process of Statization. State policy has typically acted to benefit selected groups over others: there has been legalization of encroachments by farmers; auctioning of parts of the commons to private contrac-

tors for commercial exploitation; and the distribution of VC land to individuals under various land reform and anti-poverty schemes ostensibly designed to benefit the landless but effectively endowing the already landed. 47 For elaboration see Guha (1983, 1985) and Arnold and Stewart (1989). In Nepal a similar process of Statization took place under the local monarchs, especially in. the late nineteenth and early twentieth centuries (Bajracharya 1983).

+8 Also see Thompson (1976) for interesting parallels in terms of state interventions in eighteenth-century England which curtailed the peasants’ customary rights to the commons. 49 In India, by the 1985-87 satellite survey, only 64.0 mha or 19.5 per cent of the geo-area is today under forests, and this is estimated to be declining at the rate of 1.3 mha a year. In much of Pakistan and Bangladesh, less than 10 per cent of the geo-area is under forest. The gender implications of this loss are particularly adverse, given the noted dependency of rural women on these resources, and their domestic responsibility for fetching fuel, fodder, and water (for elaboration, see Agarwal 1991).

24 A field of one’s own In some regions of India the poor received less than 20 per cent of the land so distributed.°° A similar process of declining village commons due to

illegal encroachments and government redistribution programmes is observed in Pakistan (Cernea 1981). Today in many parts of South Asia there are virtually no VCs left (see table A1.1 for India), and population growth has compounded the pressure

on what remain, speeding their degradation.>! This has both class and gender implications. The process of privatization has been accompanied by an increasing concentration of previously communal land in the hands of the male members of relatively few households. Poor households have thus lost out collectively while gaining little individually. Within poor households, women’s loss has exceeded men’s due to women’s greater depen-

dence on communal resources, given their little control over private resources (Agarwal 1991). In effect this has deepened both class and gender differentiation among the peasantry. With the decline in communal land, access to privatized land acquires a critical importance today which it did not have even a century ago. In India, for instance, by a rough estimate about 85.6 per cent of arable land is likely to be in private hands.°? Hence the case for women’s land rights spelt out in the next section, while couched in general terms, is especially focused on

rights in privatized land, with two caveats: one, given the importance of communal land to the rural poor, and especially to poor women, there is a strong case for protecting the communal nature of any land which still exists in that form. Two, it is necessary to explore the possibilities of new institutional arrangements for jointly owned/controlled land holdings by groups of women, rather than by groups of households (as is the usual focus). Joint ownership need not, however, imply joint cultivation.

(4) Prospects for non-land-based livelihoods It is commonly argued that as a country’s manufacturing and service sectors expand, the dependence on land as a source of livelihood is likely to decline. This may certainly be expected, and indeed is desirable for relieving

the population pressure on land. But at present agriculture is still either a °° For elaboration, see Jodha (1986). Also see Brara (1987) for Rajasthan state. ‘! There has also been significant degradation of both public and private land as a result of factors such as: (a) waterlogging and soil salinity associated with large irrigation works: in some Indian projects, half the irrigable potential created has been lost in this way (Joshi and Agnihotri 1984); (b) soil and wind erosion: an estimated 56 per cent of cultivable area in India is so affected; and (c) large-scale deforestation (for details, see Agarwal 1991). ‘2 This was calculated from India’s land-use statistics for 1987-88 (GOI 1992c) as follows: total arable area, as noted in footnote 13, comes to 184.73 mha. Of this, 158.09 mha, which is the aggregate of net sown area, land under current fallows, and land under miscellaneous tree crops and groves, could broadly be assumed to be in private hands.

Land rights for women 25 primary or an important supplementary source of income for the bulk of the rural population in South Asia. Only a small proportion of the total labour force, and even less of the female labour force, is employed in the manufacturing sector; and sectoral labour force projections, where available, do not predict any dramatic increase in labour absorption into formal industry in the near future. Also (as noted), whether or not engaged in agriculture, a significant proportion of the rural population (which in 1990 constituted 74 per cent of South Asia’s population)>3 is dependent on VCs and forests for its supplementary survival needs. Hence the importance of land-dependent livelihoods is unlikely to decline quickly. Consider, more specifically, the Indian situation; in other South Asian

countries the scenario would differ in degree but not dramatically in substance. In India, in 1987-88, 85 per cent of rural female workers and 75

per cent of rural male workers were employed in agriculture and allied activities such as forestry and fishing (GOI 1990b: 99). Only 11.1 per cent of the total (rural plus urban) labour force was employed in the manufacturing sector. Moreover, the growth rates of employment in the manufacturing and service sectors have fallen sharply over the past decade (GOI 1992b:

129); and employment in organized sector manufacturing was stagnant between 1983 and 1988 (GOI 1992b: 130).5* According to the Eighth Five Year Plan projections, some 94 million persons (new entrants plus unem-

ployed backlog) who will seek employment between 1997-2002 (GOI 1992b: 120) will have to find work largely as self-employed or casual

workers.°5 The bulk of these will still be located in the rural areas. Wasteland development through reforestation schemes etc., which has been emphasized in the Eighth Plan, is directly land dependent. Rural nonfarm employment is another outlet, but at present, although 25 per cent of male rural workers are engaged in non-agricultural activity as their primary or subsidiary income source, only 15 per cent of female rural workers are so engaged (GOI 1990b: 99). Moreover the non-farm sector is very heterogeneous, containing both high return/high wage activities and low return/ 53 This is taking an average for five countries: India, Pakistan, Nepal, Bangladesh, and Sri Lanka (for individual country figures see table A1.2). 54 Manufacturing enterprises, as defined by the Annual Survey of Industries, are enterprises employing ten workers or more if using power and twenty workers or more if not using power. Assessments of employment shifts in the ‘organized’ part of the manufacturing sector (and for the organized sector in general) are likely to be rather rough: all organized

sector employment data are collected by the Ministry of Labour, and the return of information by certain categories of private (non-agricultural) sector enterprises is on a voluntary basis and therefore tends to be incomplete (see GOI 199 1a: 1). 58 The National Sample Survey Organization (NSSO) gives a breakdown of total employment into regular employment, casual employment and self-employment. In 1987-88 only 18 per cent of the male and 8.3 per cent of the female labour force had regular and salaried employment, about half the workers of both sexes were self-employed, and the rest were in casual employment (GOI 1992a: 133).

26 A field of one’s own low wage ones. Existing evidence suggests that although in some highly agriculturally prosperous regions, such as Punjab and Haryana, there has been a marked growth in non-farm activities which provide high returns for the self-employed as well as high wages for the wage earners, in agricultur-

ally less-developed regions, where the bulk of the poor are located, low return, low wage activities (which people undertake for distress reasons) are

more typical.*° This appears to be the case in even greater measure for women than men: both the country-wide survey undertaken in 1987 by the

National Commission on Self-Employed Women and Women in the Informal Sector (see Shramshakti 1988), and micro-studies of women workers in individual occupations,*’ suggest that women are largely concentrated in the low-and-insecure-earnings end of the non-farm occupational spectrum. The picture for most other parts of South Asia appears to be similar.*® Moreover, how well someone does in the rural non-farm sector through self-employment is itself often significantly related to land access.°° Even a small plot can considerably expand the range of non-farm opportunities for °© See e.g. Islam (1986), Papola (1987), Hazell and Haggblade (1990), Saith (1991), and Basant and Kumar (1989). Hazell and Haggblade, in a cross-state analysis for India for 1981, found that it was the agriculturally most prosperous areas of Punjab and Haryana which had the best-developed service, commerce and factory manufacturing activities in the rural areas, while household manufacturing (which brings relatively low returns) was less important. In states with low agricultural development, household manufacturing was the main form of rural manufacturing activity. Papola’s (1987) cross-state analysis for rural industries also suggests this difference. And in a cluster of West Bengal villages, Islam (1986) similarly found that non-farm activities typically provided very low earnings (also see footnote 59). 57 See e.g. various case studies in Singh and Kelles-Vitanen, eds. (1987), and Mies’ (1982) study on the lace makers of Narsapur (Andhra Pradesh). 58 See e.g. Islam (1986, 1987). Taking a sample of villages in India (in West Bengal), Bangladesh, Pakistan, and Sri Lanka, Islam (1986) found that non-farm activities usually took the form either of very low wage employment, or of small enterprises doing petty

trading, servicing, and manufacturing with very low productivities and/or returns. Typically non-farm wages or earnings were lower than farm wages or earnings. Non-farm wage employment was also negatively related to farm size. Similarly, Hazell and Haggblade (1990) found a negative relationship between farm size and dependence on non-farm earnings (especially wage earnings), in an all-India analysis for 1970-71 and 1980~81. In addition, in relation to women, Islam (1987: 4) notes that cottage industries, which are the

main source of women’s employment in rural manufacturing in Bangladesh, Nepal, Pakistan, and India, have been doing poorly in the former three countries (in terms of both

employment and contribution to Gross Domestic Product), although India presents a mixed regional picture. 59 See e.g. Islam (1986), Papola (1987), Saith (1991), and Chadha (1992). Chadha, in a recent

survey of eighteen villages in three states (Bihar, Andhra Pradesh, and Uttar Pradesh), found that access to land made a significant difference to daily household earnings from self-employment in the rural off-farm sector: in all three states, small farmer households earned substantially more from this source than landless labour households; in Andhra Pradesh the former’s earnings from rural off-farm self-employment were eight times those of the latter.

Land rights for women 27 poor rural households. And for many rural non-farm wage workers, access to land as a supplementary source of income would be critical in reducing the risk of poverty. In addition, the growth and labour-absorbing capacity of the rural non-farm sector depend in considerable degree on how much local demand for labour-intensive products is generated; and South Asian

experience suggests that more egalitarian land (and farm income) distribu- | tion is likely to generate greater non-farm employment, especially through consumption linkages (Islam 1986: 172; Harriss 1991: 454—S). All said, therefore, it would be realistic to expect that for a significant majority of persons in rural South Asia, for a considerable time to come, viable livelihood systems are likely to need access to at least some land. For rural women this appears even more necessary. As noted, rural non-farm

activity does not, in itself, tend to be especially high paying for most women. Also there is the prevailing tendency of the non-farm sector to absorb more men than women; if this gender gap continues as this sector expands, and it is associated too with male outmigration, the number of de facto female-headed households in agriculture will tend to increase. At the same time, as argued below, an increase in male-earned income, whether from farm or non-farm work, does not automatically translate into equal benefits for women and children. Moreover (also as argued below), land rights have an importance for rural households in general and for women in particular not only for reasons of livelihood, but on many other counts as well.

Il. Why do women need independent rights in land? The case for women having independent rights in arable land rests on several interconnected arguments which can be grouped into four broad categories: welfare, efficiency, equality, and empowerment.°°

(1) The welfare argument To begin with, especially among poor households, rights in land could reduce women’s own and, more generally, the household’s risk of poverty and destitution. The reasons for this stem partly from the general positive effect of giving women access to economic resources independently of men;

resources. |

and partly from the specific advantages associated with rights in land

Consider first the general case. There is considerable evidence of

60 The discussion here will concern land linked in one way or another to rural livelihoods, especially arable land, but will exclude homesites, even though the available data on land ownership do not always separate land under homesites from the rest.

28 A field of one’s own intra-household gender inequalities in the sharing of benefits from the

household’s resources. For instance, in large parts of South Asia a systematic bias against women and female children is found in intrahousehold access to basic necessities such as health care, and in some degree

also food.®! This is revealed in gender differences in one or more of the

following indicators: malnourishment, morbidity, mortality, hospital admissions, health expenditures, and (especially) in female-adverse sex ratios (females per 100 males), although the evidence on food allocation per

se is less conclusive.°? The extent of this anti-female bias varies crossregionally,°> but it exists in some degree almost everywhere, particularly as revealed by the sex ratios which are female-adverse across all of South Asia, except Kerala in southwest India. The bias ts strongest in northwest India,

Pakistan, and Bangladesh, where sex ratios are particularly low,°* and much less in south India and Sri Lanka, where the sex ratios, although still female-adverse, are closer to parity (see table A1.2). Further, in many states of India notable differences have been found in how men and women of poor rural households spend the incomes under their control: women of poor households typically spend almost all their incomes to purchase goods for the family’s general consumption and for the

children; men usually spend a significant part on their personal needs (tobacco, liquor, etc.).°> Mencher’s (1988) findings are especially note°! Fora discussion on this bias and its possible causes in India, see Agarwal (1986b), Harriss (1990), Kynch and Sen (1983), Sen and Sengupta (1983), Behrman (1986), Bardhan (1974), Dreze and Sen (1989), Ghosh (1985), Miller (1981), Dandekar (1975), Gordon et al. (1965) and Dasgupta (1987a). For Bangladesh, see Chen ef al. (1981). For Pakistan, see Miller (1983); and for Sri Lanka, see Schrijvers (1988). Harriss (1990) covers several South Asian

countries. _

°2 Harriss’ (1990) extensive review of literature and detailed data analysis on intra-household food allocation in South Asia illustrates the difficulties of arriving at firm conclusions on

this count. Nevertheless, her tentative conclusions include the following two: one, ‘discrimination in energy and protein intakes through the allocation of food within the household seems to be greater in the north [of the subcontinent] than in the south’; two, ‘in

the north it is least “fair” for very young and very old females, and probably for adult women with special needs associated with pregnancy and lactation’ (p. 405). °3 Some studies also find a cross-class variation, the bias (measured in anthropometric indices) being higher among landless households relative to the landed (see e.g. Sen and Sengupta 1983, for West Bengal; and Levinson 1974, for the Indian Punjab). But contrary findings for some other regions make generalizations problematic (on this see Harriss 1990).

64 According to Dreze and Sen’s (1989: 52) estimate, India would have 36.9 million additional

| women today, and Pakistan and Bangladesh an additional 5.2 and 3.7 million respectively, if these countries had the same sex ratios as sub-Saharan Africa, namely 102. As they note, for most developed countries in Europe and North America the sex ratio averages about 105, essentially indicative of women’s survival advantages over men in the absence of serious anti-female bias in the distribution of food and health care. ©’ There is considerable evidence on this from different parts of India: see Gulati (1978) for Kerala; Mencher and Saradamoni (1982) and Mencher (1988) for Kerala, Tamil Nadu and West Bengal; Dasgupta and Maiti (1986) for Himachal Pradesh, Madhya Pradesh, Uttar

Land rights for women 29 worthy in this regard: based on a sample of landless and near landless agricultural labour households in ten villages each of Tamil Nadu and Kerala states in India, and taking a weighted average per village, she found that with both spouses earning the wife usually contributed 90—100 per cent

of her earnings to general household expenses, while the husband rarely gave over 75 per cent, spending the rest on himself.°° Where the wife

contributed somewhat less than 90 per cent of her earnings, it was accounted for, in most cases, by work-related transport and lunch expenses. The portions that men kept back were spent on food and drinks with friends and their own clothes. The women’s contributions in absolute terms were also substantial in all cases.°’ A corollary to the noted gender differentials in spending patterns are research findings which suggest that children’s nutritional status tend to be much more positively linked to the mother’s earnings than the father’s

(Kumar 1978; Gulati 1978). Kumar, for instance, found that among landless and marginal farmer households in Kerala, where the mother was in the wage labour force it was her own wages that primarily accounted for the positive household wage income effect on child nutrition; where the mother was not in the wage labour force an increase in household wage income showed no incremental effect on child nutrition.°® Of course the Pradesh, Maharashtra, and Assam; and Per-Lee (1981) for Rajasthan. (For evidence from outside South Asia, see especially the literature surveys in Bruce 1989; Blumberg 1991; and

Hoddinott 1991.) Moreover, male expenditure on liquor can have a high cost in poor households in terms of basic necessities foregone. In one survey in the Indian Punjab an average household with a heavy drinker was found to spend 40 per cent less on food per capita and less also on clothing, education, and medicine, than a household without this disability (cited in Harriss 1990). 66 The median contributions (calculated from Mencher by Blumberg 1991: 102) were 90 per cent of women’s earnings in Kerala and 98 per cent in Tamil Nadu, relative to male median contributions of 70 per cent and 74 per cent respectively. ©? Some other aspects of Mencher’s (1988) findings are also important. She found that: (a) a woman’s absolute contribution to household maintenance from earned income exceeded her husband’s in six of the twenty villages, was quite close to equal in five others, and substantial in the rest; and (b) the minimum contributed by all household males was less than by all females in thirteen of the twenty villages. This is despite women’s lower absolute earnings due to lesser access to employment and lower wages relative to men. It is also noteworthy that these contributions do not include the values of items such as fuel, fodder, food, etc. that are gathered by women. Moreover, in a subsample of six villages, three each from the three states, Mencher found that even in the month when the wives earned least, husbands kept back some earnings for personal expenses: in fact they kept back a larger percentage than in a month when the wife earned the most. 68 There is also considerable evidence from other Third World countries that resources controlled by mothers have a greater impact on children’s health than those controlled by fathers. For instance, in a sample for urban Brazil, Thomas (1990) found that the effect on child survival probabilities was almost twenty times greater when unearned income (from rent, physical and financial assets, gifts, etc.) accrued to the mother than the father. Also see Hoddinott (1991) and Blumberg (1991) for evidence on this relationship from several other countries.

30 A field of one’s own relationship between the mother’s wage earnings and the child’s health status is complex: for instance, the positive consumption effect of maternal earnings could be offset by reduced maternal time in childcare. However, Kumar’s results held, even when the length of the mother’s employment was taken into account.®? It is notable too that among the marginal farmer households, the mother’s cultivation of a home garden (the output of which she controlled) had a consistently high positive effect on child nutrition. (In any case, for certain types of land-related incomes (e.g. those from rent) there need be no conflict between maternal income and time devoted to

childcare; and where such conflict arises there is a case for providing substitute childcare support.)

In other words, the risk of poverty and the physical well-being of a woman and her children could depend significantly on whether or not she has direct access to income and productive assets such as land, and not just access mediated through her husband or other male family members. For female-headed households with no adult male support, the link between

direct access to economic resources and physical well-being needs no emphasis. Such households constitute an estimated (and by no means negligible) 19-20 per cent of all households in India and Bangladesh.7° Moreover, as noted earlier, a woman’s economic status cannot be judged

adequately by the economic status of her family. Even women whose parental or marital households are classified as rich peasant can be rendered economically vulnerable in the absence of independent economic resources in case of divorce, desertion, separation, or widowhood. At the beginning

of this chapter we noted how in the Indian states of Maharashtra and Rajasthan, not uncommonly, rural women — divorced, deserted or widowed — have been found working as agricultural labourers on the farms of their well-off brothers or brothers-in-law. We also noted how elsewhere,

as in West Bengal (east India) and Bangladesh, there are many cases of women, married into prosperous households, being left destitute and forced to seek wage work or even to beg after widowhood. ‘This fact’, as Omvedt (1981: 21) observes, ‘perhaps ... more than any other, shows the essential propertylessness of women as women.’ 69 Also see Leslie’s (1989) fairly comprehensive review of studies on Third World countries on

the effect of maternal employment on infant feeding practices and child nutritional status. She concluded that the research did not suggest that maternal employment would have a negative effect either on breast feeding or child nutritional status: several studies found better nutrient intake among children whose mothers worked, especially where the mothers were high income earners; and studies which found a negative association were open to more than one interpretation. The quality of substitute childcare support was also a significant factor in determining the effects. 70 See Buvinic and Youseff (1978) for India and Safilios-Rothschild and Mahmud (1989) for Bangladesh. According to the Indian census some [0 per cent of households are headed by women, but this is a significant underestimate (see Agarwal 1985b, on reasons for the undercounting).

Land rights for women 31 Within this general argument in favour of women’s independent access to economic resources, the case for rights in /and is especially strong. In a context of limited non-farm opportunities, land serves as a security against poverty — a means to meet basic needs. Consider, fora start, the relationship between a household’s access to land and poverty. In India, in 1982 an estimated 89 per cent of rural households owned some land (GOI 1987b: 9),

and an estimated 74 per cent operated some (GOI 1986a: 12).7! In Bangladesh, in 1978, the percentage of rural households owning some land

(arable or homestead) was 89, and those owning arable land was 67 (Jannuzi and Peach 1980: 101). In Sri Lanka, in 1982, 89 per cent of agricultural operators owned some land (including home gardens).’ Although, given the high degree of land concentration, the majority of these

households only have marginal plots,’> they are likely to have a significantly lower risk of absolute poverty than landless households. A negative relationship between the incidence of absolute poverty and land access (owned or operated) has been noted in several studies;’* and landless labourers are found to be worse off than the near-landless during famines 71 The estimates are based on the 37th round of the National Sample Survey (NSS) carried out in 1981-82. The figure for land ownership covers all land owned by the household, whether or not cultivated, including that used for non-agricultural uses. 72 In the Sri Lankan Agricultural Census, agricultural operators include land cultivators as well as purely livestock and poultry operators (Government of Sri Lanka 1984b:17). 73 In rural India, according to the NSS, 41.6 per cent of landowning households owned one acre or less and accounted for only 3.6 per cent of all land owned by rural households (GOI 1987b: S-18). The distribution of operational holdings was somewhat less skewed (GOI 1986a). High inter-household inequalities in land ownership and control also characterize other South Asian countries. In Bangladesh, in 1978, 43.3 per cent of the households owning some arable land owned I acre or less and accounted for only 9.6 per cent of total owned area (Jannuzi and Peach 1980: 100). In Sri Lanka, 42.8 per cent of landowning agricultural operators owned under one acre and accounted for 9.4 per cent of land owned (Government of Sri Lanka 1983: 17). Also see chapter 8 (table 8.7) for estimates of Gini coefficients measuring land concentration. 74 See, Ali et al. (1981), Sundaram and Tendulkar (1983), Gaiha and Kazmi (1981), and Lipton’s (1985) survey of issues and literature. Estimates for 1975 by Ali e¢ al. (1981) quoted in Sundaram (1987: 179) indicate a consistent decline in the percentage of rural population below the poverty line as the operational holding size increases. For instance, among those operating no land the percentage below the poverty line was 81.7; among those operating 0.0-0.5 ha the percentage was 75.4; for the 2-4 ha category of farms it was 45.3; and among those operating over 8 ha it was 4.5. Sundaram and Tendulkar (1983), on the basis of NSS data for 1977-78, found that the incidence of poverty among households dependent mainly on agricultural wages for a livelihood was almost twice that among cultivating households (58.8 per cent relative to 30.1 per cent). Gaiha and Kazmi (1981) again found the highest risk of poverty among agricultural wage labour households, on the basis of all-India data for 1970-71 from the National Council of Applied Economic Research. Lipton’s (1985) review of literature shows that the relationship between land and poverty varies regionally and by land quality, the negative relationship between land (owned and operated) and risk of poverty is strongest for good quality, irrigated plots (in which case even an acre can significantly reduce the risk of poverty), and weakest for very poor quality, rainfed land (in which case much larger plots are needed to make a noteworthy difference).

32 A field of one’s own (Sen 1981). Land access helps in both direct and indirect ways. The direct advantages stem from production possibilities, such as of growing crops, fodder, trees, or a vegetable garden (unless of course the land is of very poor quality), or keeping livestock, practising sericulture, and so on. In addition, land provides indirect advantages, such as facilitating access to credit from institutional and private sources, helping agricultural labour maintain its reserve price and even push up the aggregate real wage rate,’°> and, where the land is owned, serving as a mortgageable or saleable asset during a crisis. Moreover, for vulnerable groups such as widows and the elderly, ownership of land and other wealth strengthens the support they receive

from their relatives, by increasing their bargaining power within the household.’7© As an old man (cited in Caldwell et a/. 1988: 191) put it: ‘Without property, children do not look after their parents well.’

However, given the noted biases in the intra-family distribution of benefits from household resources, exclusively male rights in land, which would render the household less susceptible to poverty by some average measure, will not automatically provide this protection to all its members, and especially not to its female members. Thus on grounds of both women’s

and children’s welfare, there is a strong case for supporting women’s effective rights in private or public land, independently of men. Equally, there is a case for protecting and ensuring women’s interests in land which is

as yet non-privatized, possibly by granting groups of women usufructuary rights in public land (as will be elaborated in chapter 10). Although rights in private or public land are especially important as a poverty-alleviation measure for women in poor households, they are also relevant for women of better-off households, given the risk of poverty following marital breakdown faced by all rural women. It needs emphasis here that the welfare case for women’s land rights stands even if the plot is too small to be economically viable on its own. Indeed those opposing female inheritance in land often emphasize that women might end up inheriting economically non-viable holdings. In my view, this could be a problem where cultivation is seen as the sole basis of subsistence, but not where land-based production is only one element in a diversified livelihood system. For instance, a plot of land which does not produce enough grain to economically sustain a person or family could still support trees or provide grass for cattle. Many landless widows I spoke to in

the Indian state of Rajasthan said they could not take advantage of the government’s anti-poverty loans for cattle purchase because they had 75 See e.g. Raj and Tharakan (1983) and Bardhan (1984). For details of their findings see 76 Soe eg White (1992) for Bangladesh; and Caldwell et a/. (1988), Dreze (1990: 68-70), Sharma and Dak (1987), and Raj and Prasad (1971) for India.

Land rights for women 33 nowhere to graze the animals. With even small plots of land for growing grass, animal upkeep would have been more viable. Moreover, although forced collective farming is likely to be inefficient, cases of people voluntar-

ily cooperating to undertake land-based joint productive activities also exist: there are several successful instances of small women’s groups doing so in India and Bangladesh (as will be described in chapter 9). Of course the emphasis here on land for women is not to deny the need to expand women’s employment and earning opportunities in non-land, or non-farm activities. Indeed, thinking of land as one income source among others points toward the importance of planning along these lines as well.

But, as my earlier discussion suggests, this necessitates promoting a productive non-farm sector and ensuring women’s entry into the higherearning segments of that sector.’’

(2) The efficiency argument To trace the likely efficiency effects of women having land rights is a much

more complex undertaking than tracing the potential welfare effects. 7° Consider the issue situationally. In a variety of contexts women are today operating as household heads with the primary and sometimes sole responsibility for organizing cultivation and ensuring family subsistence, but without titles to the land they are

‘cultivating. For instance, due to long-term male outmigration many women are serving as de facto household heads, especially but not only in the hill regions of the subcontinent.’? Or widows are cultivating plots given to them out of joint family estates (as part of their inheritance claims to their deceased husbands’ lands) but the plots are still in their in-laws’ names.®° Again, tribal women cultivating communal land rarely hold titles to their

fields, which are typically given out by the State only to male farmers. Titling women in these circumstances and providing them infrastructural support could help increase output, by increasing their access to credit,®! 17 Also see the recommendations in Shramshakti (1988). 78 This includes both productive efficiency and efficiency in the more general sense of meeting

the target goals of a programme such as a poverty-alleviation programme. 79 Jain (1984: 1789) found that 20 per cent of households in Chamoli district (Uttar Pradesh hills, northwest India), were de facto female-headed due to male outmigration. For India, also see Jetley (1987) and for Pakistan, see Shaheed (1981). 80 ]T found several such cases in Rajasthan in 1987. 8! There is considerable evidence from Asia that titling can significantly enhance farmers’ access to credit (in terms of sources, amounts, and terms) by enabling them to use land as collateral (see e.g. Binswanger and Rosenzweig 1986b; Binswanger 1986; Feder 1989; Feder et al. 1986; and Feder and Noronha 1987). Also see Staudt (1975/6), and Saito and Weidenmann’s (1990) discussion (mainly in the African context) on the problems women farmers face in obtaining credit 1n the absence of titles.

34 A field of one’s own and to technology and information on productivity-increasing agricultural practices and inputs (in the dissemination of which both a class and gender bias prevails).°? Land titles could make it easier for women to adopt improved agricultural technology and practices as well as enhance their

motivation to do so, and so increase overall production. This is not dissimilar to the argument made in land reform discourse favouring security of tenure for tenants to encourage technical investments in land by increasing the tenants’ capacity and incentive to invest. A more general issue, however, is the likely efficiency effect of women inheriting land. Female inheritance is often opposed in South Asia on the grounds that it will further reduce farm size, increase land fragmentation, and thus reduce output. Is this fear valid? The efficiency implications of female inheritance could be separated analytically into three parts: (a) a farm-size effect: the average size of ownership holdings would be less than if only men inherit; (b) a land-fragmentation effect:3 fragmentation could increase in so far as the land is parcelled out to heirs say according to land quality; and (c) a gender-transfer effect: some of the land which would have gone only to men would now go to women. The concerns surrounding the farm-size effect are similar to those that

arise in relation to redistributive land reform, namely the effects of redistributing land from big to small farmers on farm output, on the adoption of new technology, and on marketed surplus. Those who oppose

: redistribution argue that it will have a negative effect on all three counts. However, existing evidence does not support this view. Studies relating to the 1950s and 1960s, that is the pre-‘green revolution’ period,®* clearly show that small-sized farms had a higher value of annual output per unit of cultivated area than large-sized ones, typically because small farms tended to have higher cropping intensities and a more labour-intensive and highervalue crop-mix.®> It is significant that this inverse size-productivity relationship, which some had predicted would disappear with the new agricul82 For a review of South Asian material on the class bias in agricultural extension, see Dasgupta (1977) and Byres (1972), and on gender bias see Agarwal (1985a), Goetz (1990), and Kilkelly (1986).

83 The term ‘fragmentation’ as used here relates to the division of a farm into several noncontiguous parcels of land, and farm size relates to the aggregate area of such parcels held by the cultivator. The analytical distinction between the farm size effect and the fragmentation effect is important, as will be seen from the discussion which follows. In popular parlance the term ‘fragmentation’ has come to be used rather loosely (and incorrectly) to refer also to the process of declining farm size. 84 The term ‘green revolution’ is popularly used to describe the dramatic increases in agricultural output in Asia during the late 1960s and early 1970s, following the adoption of a package of agricultural inputs and practices, principally high-yielding cereal varieties with chemical fertilizers and an assured water supply. 85 For a review of studies see especially Agarwal (1983: chapter 7), Boyce (1987: chapter 2), and Berry and Cline (1979: chapter 4). Also see Bharadwaj (1974).

Land rights for women 35 tural technology, has persisted, as evidence from India, Bangladesh, and Pakistan bears out.8° Small farmers have adopted the new technology in most areas where large farmers have done so, although after a time lag,®’

and the inverse size-productivity relationship remains strong, even if somewhat weakened in some areas in comparison with the pre-green revolution period.®® Again, the evidence on marketed surplus does not bear up to the sceptics’ claim that this will decline because small farmers will tend

to retain a larger percentage for self-consumption. For non-foodcrops the

marketed surplus is found to be very high on farms of all size groups (Lipton 1992), and on foodcrops the higher productivity effect of small farms may well outweigh their higher propensity-to-consume effect.®? In any case, an improvement in the consumption standards of the poor in the farm sector cannot in itself be seen as an inefficient outcome. In fact a diet

improvement among the very poor may add to labour productivity (as elaborated later). In other words, the existing evidence suggests that land redistribution from big to small farmers would probably increase agricultural output; certainly there is no reason to expect an output decline. Although these studies do not differentiate between male and female farmers, the arguments presented above can also justifiably be extended in response to the farm-size concerns raised by opponents of female inheritance. The same is true for the land fragmentation argument: it would be of relevance for both ‘male and female heirs, and in both instances there would be a case for land consolidation. 86 See Berry and Cline (1979) for India and Pakistan; Agarwal (1983) for India (Punjab); and Boyce (1987) for India (West Bengal) and Bangladesh. It is noteworthy that farms at the

lower end of the size spectrum in Bangladesh are on average much smaller than farms elsewhere in the subcontinent. The persistence of the inverse relationship in that country therefore adds further strength to the argument for redistribution. 87 See the considerable evidence for India, Pakistan, Bangladesh, and several other countries presented in Lipton and Longhurst (1989: 115-16). Large farmers usually adopt first, given their better economic and institutional access to inputs. But smaller farmers typically catch up, especially if State support is forthcoming. 88 See e.g. Boyce (1987: 213) who notes: ‘the agricultural census data reveal little tendency for the inverse relation to disappear with the advent of the new seed-fertilizer technology. As of 1976/7... the proportion of acreage devoted to HYVs was inversely related to farm size’. I

found the same for data relating to the Indian Punjab for 1971-72 (Agarwal 1983). Similarly Berry and Cline (1979: 111) conclude from Indian data relating to several states that ‘the inverse relationship persists in [1970-71] .. ., although its steepness is moderately reduced’.

It is argued by some (e.g. Bhalla and Roy 1988) that the inverse relationship is significantly weakened (although not eliminated) if land quality (treated as an exogenous variable) is controlled for (small farms having better quality land than large ones); but as

Lipton (1992) points out, land quality is not an entirely exogenous variable but itself depends in fair degree on the quality-improving labour of small farm families, in activities such as bunding, levelling, maintenance of irrigation channels, etc. 89 As found, for instance, in Kenya (Lipton 1992).

36 A field of one’s own However, female inheritance, as noted, raises an issue in addition to those of farm size and fragmentation, namely the possible output implications of land transfer between the genders, that is from potential male heirs to female heirs. (This issue also arises in relation to the transfer of public land to women.) In this context, it needs emphasis that ownership rights can co-exist with a variety of cultivation arrangements. In so far as women owners do not cultivate the land themselves, but lease it out to male relatives or to other male tenants, there need be no gender-transfer effect on

output. But women who seek to manage the land themselves tend to encounter several obstacles that a male farmer typically does not: gender biases in access to technical information and inputs, difficulties in commanding family labour, social constraints to operating freely in factor and product markets, and so on (for elaboration, see chapter 7). These could have a negative effect on output. The solution, however, lies in providing women farmers with the necessary technical and institutional support to reduce the constraints they face, and not in disinheriting them.°° After all, the ability of male farmers to adopt modern technologies and increase output is also dependent on access to support services.?! Indeed, the experience of several non-governmental credit institutions, such as the Grameen Bank in Bangladesh, which cater to the very poor, indicates that women usually have better repayment rates and are therefore often better credit risks than men (Hossain 1987).9? Here the perceptions of

poor peasant women in Bihar are also revealing. During a discussion on 99 Some African case studies are illustrative. For instance, in west Kenya, maize yields were almost 7 per cent more on female-managed farms relative to male-managed ones when they had the same access to extension services (cited in Dey 1992: 14). °' An additional question arises here, which impinges on the issue of marketed surplus raised earlier, namely: would poor peasant women’s immediate concern with ensuring family subsistence have implications for cropping patterns (with a possible bias towards subsistence crops and lower marketed surplus), if they had direct control over farm production?

This needs further probing. My own observation in Rajasthan and the hills of Uttar Pradesh, of widow-managed or de facto female-managed fields, does not suggest this need follow: in both places, cultivation for sale (onions in Rajasthan, potatoes in the UP hills) was popular with many women. A Sri Lankan village study too, found no preponderance of men growing crops for sale and women growing only subsistence crops (Gunawardena 1989). A more systematic examination of this issue is, however, possible using case studies for sub-Saharan Africa where men and women typically cultivate separate plots. In Dey’s (1992) literature review, for instance, cases of women not adopting higher value crops or improved varieties, could be traced either to specific constraints, such as inadequate labour at their command, or to their having greater control over the incomes from traditional crops. Again, Whitehead (1990: 459), from her extensive survey of the African material, concludes: ‘The historical record shows that rural wives have... changed their labour time use quite considerably in response to the development of a market for rural products ... they have adopted new crops, new crop mixes, and new techniques... they have marketed crops ... [while] seeking to maintain within their farming a viable food-producing element for self-consumption..’

°2 Since 1983, the Grameen Bank has been giving special preference to female household members in providing credit: only one member per household, from households owning less than 0.5 acres of cultivable land, is now eligible for a bank loan. In June 1988 the

Land rights for women 37 land access and government credit, some insisted: ‘If the land is in women’s names, the loan money cannot be spent on drink or frittered away’ (Alaka and Chetna 1987: 26). We might thus find greater efficiency in the use of

investment resources on farms managed by women. Also, supporting women as farm managers could make for a more talented and better informed pool, than one consisting solely of men.?3 An efficiency corollary stems from the welfare argument as well. In so far as the allocation of economic resources such as land to women within poor households improves their own and their children’s nutrition and health, it

could increase labour productivity as well, both immediately and in the future (through the children).°* Moreover, it is possible that land in women’s hands could lead to a different, more environmentally sound use of the resource. Where men’s and women’s land-use priorities fail to correspond, whose priorities prevail

can significantly affect the use patterns of both public and private land. Examples from the Chipko movement for forest protection and regeneration, which began in 1973 in the hills of Uttar Pradesh, reveal noteworthy gender divergence on this count: for instance, in tree-planting schemes, men have usually opted for fruit trees for cash and women for fuel and fodder trees for subsistence.?> Again, women successfully resisted the axing of the Dongri-Paintoli oak forest for setting up a government potato-seed farm,

In opposition to the village men who favoured the new scheme for its potential cash benefits. These women’s direct concern with the protection and regeneration of the forest as a source of ‘fuel, fodder, food, fibre, and fertilizer’ and of ‘soil, water, and pure air’, has had significant positive implications for ecological preservation in the region. However, as I have argued in some detail elsewhere,?® the basis of poor peasant and tribal scheme covered 9,000 villages, providing credit to 413,000 members, of which 83.8 per cent were women (Hossain 1988).

°3 Women in many rural communities, for instance, are often better informed about crop varieties than men. Burling (1963) found that among the Garos of Meghalaya (northeast India) the women knew of some 300 indigenous rice varieties and the men always deferred . to the women on this count. In Nepal, it is women who do much of the seed selection work

among almost all agricultural communities (Acharya and Bennett 1981). In Pakistan Punjab again, women play an important role in the selection of crops and seed variety (D. Merry 1983). 94 There is some empirical evidence to support this view: see, for instance, Strauss (1986) and

Deolalikar (1988) on the positive association between nutritional intake and labour productivity, although admittedly the interaction between nutritional intake and human functioning could be subject to interpersonal and intrapersonal variation (see discussion in Dreze and Sen 1989). °> Brara (1987) found a similar gender divergence in tree choices in Rajasthan (northwest India). °6 See Agarwal (1991, 1992). These papers also criticize the tendencies within ‘ecofeminist’ discourse to trace the connection between women and the environment either to female biology or to ideology (e.g. the symbolic identification of women with nature), while neglecting the material basis of the connection in the gender divisions of labour, property, and power.

38 A field of one’s own women’s concern with environmental protection in movements such as Chipko needs to be sought not in some biological connection between women and nature, but in the prevailing gender division of labour, which makes hill women primarily responsible for fetching fuelwood, fodder, and water. In the Dongri-Paintoli incident, for instance, the loss of the forest would have added several miles to women’s fuelwood journeys, while they feared that the scheme’s cash benefits would flow to men alone.

Finally, the provision of land to women could have other indirect benefits, such as reducing outmigration to the cities, both of the women themselves and of the family members dependent on them. Given the

already high pressure on urban facilities and the inability of cities to continue absorbing new immigrants in economically viable ways, a reduction in outflow of labour from the countryside would be a decided benefit. Also, improved farm incomes in women’s hands could generate a higher demand for non-farm goods that are produced locally and labour-intensively, in turn creating more rural jobs.?’

(3) The equality and empowerment arguments While the welfare and efficiency arguments are concerned with women having some land in absolute terms, especially in a situation of poverty when they have little or none, the empowerment and equality arguments are concerned also with women’s position re/ative to men, and particularly with women’s ability to challenge male oppression within the home and in the wider society.

The equality argument for land rights can be approached in several different ways, but two aspects are especially important here. One is the larger issue of gender equality as a measure of a just society, in which equality of rights over productive resources would be an important part. In

India, for instance, the Constitution of 1950 promises equality as a fundamental right. This includes equality before the law, and prohibition of discrimination on grounds of religion, race, caste, sex, and place of birth.

These provisions are commendable, as Beteille (1987) points out, for a society which has traditionally been hierarchical to an unusual degree, especially in terms of caste and gender. However a vast gap remains between stated principles and State practice, as revealed, for instance, in the gender biases in existing laws concerning landed property, the functioning of legal institutions, public policies and their implementation, and so on. °? This is partly because women’s lesser mobility tends to confine them more than men to local

markets, and partly derivative of the more general argument, made earlier, that villages with greater equality in land distribution in South Asia are likely to generate more demand for local non-farm products.

Land rights for women 39 Also obstructing women’s ability to achieve equality of land rights are gender disadvantages (some already noted, others discussed later in the book) in the social, ideological, and political spheres. These gender gaps need closing if indeed we are to move toward a more just society.

Two, there is the specific aspect of equality in land rights both as an indicator of women’s economic empowerment and as a facilitator in challenging gender inequities in other (e.g. social and political) areas. In the

present discussion, the links between gender equality in land rights and women’s empowerment are of special interest and importance. But first, what do we mean by empowerment? The term has been used variously (and often loosely) in academic writing and by social action groups across the world, including South Asia. In the present context, it could be defined as a - process that enhances the ability of disadvantaged (‘powerless’) individuals or groups to challenge and change (in their favour) existing power relation-

ships that place them in subordinate economic, social, and political positions.°° Empowerment can manifest itself in acts of individual resistance as well as in group mobilization. Entitling women with land would, on the one hand, empower them economically, and, on the other hand, strengthen their ability to challenge social and political gender inequalities. That is, land rights would enhance women’s ‘freedom to achieve’ (or ‘capability to function’) in non-economic spheres as well.?? A telling illustration is provided by the Bodhgaya movement (see chapter ‘) for details) that emerged in the state of Bihar in eastern India in the late 1970s, in which women and men of landless households jointly participated in an extended struggle for ownership rights in the land they cultivated, which was under the illegal possession of a local Math (a temple-monastery complex). During the struggle, women raised a demand for independent land rights, not only for reasons of economic security but also because this impinged on marital relations. As the quotation towards the beginning of

this chapter indicates, they feared that if land titles went only to their husbands, they would be rendered relatively even more powerless, and vulnerable to domestic violence. Their fears proved correct: where only men got titles there was an increase in drunkenness, wife-beating and threats: ‘Get out of the house, the land is mine now’ (Manimala 1983: 15); while where women got the titles they could now assert: ‘We had tongues but could not speak, we had feet but could not walk. Now that we have the °8 Bookman and Morgen (1988: 4) arrive at a very similar definition in the context of working-class women’s resistance and political activity in the United States.

°° See Sen (1992) for an exposition of the notion of ‘freedom to achieve’ which he distinguishes from actual achievement. He notes (1992: 31)‘ Achievement is concerned with

what we manage to accomplish and freedom with the real opportunity that we have to accompish what we value’. Freedom, in his schema, is reflected in a person’s ‘capability to

40 A field of one’s own land, we have the strength to speak and walk’ (Alaka and Chetna 1987: 26). Similar responses were noted in China, when the Chinese Communist Party

promulgated the Agrarian Reform law in 1947, which entitled women to hold separate land deeds for the first time. The women were quick to realize the law’s potential for freeing them from oppressive marital lives: ‘When I get my share I'll separate from my husband. Then he won’t oppress me any more’; and ‘If he divorces me, never mind. I'll get my share, and the children will get theirs. We can live a good life without him’ (Hinton 1972: 470). In Chaojia village, Croll (1978) notes, every poor peasant, man or woman, was allotted a piece of land. Where earlier women had been referred to as ‘soand-so’s wife’ or ‘so-and-so’s mother’, now their own names were written in

the land certificates: ‘They had acquired a name alongside a share of the land’ (Croll 1978: 215).

Land rights can also make a difference to women’s relationships (as indeed to men’s) with other family members. For instance, during fieldwork in Rajasthan (northwest India) in 1987, I found that landowning widows living with their adult sons were treated with much greater respect and consideration than those who were landless and economically depen-

dent. Dreze (1990) makes similar observations for some other parts of India. Although better treatment from family members can also result from an improvement in the women’s economic position through employment or other means,'°° in the rural context land offers a specific form of security which other sources of income do not — at the very least, a space of one’s own. In the Bodhgaya case, for instance, the women were already wage

labourers and were therefore not economically dependent; but their husbands were still able to threaten them with eviction. It is also notable that the Bodhgaya women saw intra-household gender relations being affected not just by their own propertyless state, but by their remaining propertyless while their husbands became propertied. In other words, land titles were important to women not only for improving their economic well-being in absolute terms (the welfare argument), but also for improving their relative position vis-d-vis their husbands: their sense of empowerment within the home was linked to economic equality.'°! Outside the household as well, having land rights can empower women in various ways. First, it can impinge on the social treatment received from other villagers: low caste women members of a rural voluntary organiza-

tion in Andhra Pradesh (south India), attributed their powerlessness mainly to their lack of land: ‘If we had a little land they [the upper-caste 100 For elaboration and examples see chapter 2.

101 The importance of women’s rights in land in strengthening their overall bargaining position within the household will be discussed in detail in chapter 2, which also seeks to provide a conceptual framework for characterizing gender relations.

Land rights for women 41 landlords] would not talk like this. But now they keep us at a distance because they say we are untouchables and shameless’ (Mies et al. 1986: 115).

Second, it can enable poor rural women to bargain with employers for higher wages from a stronger fall-back position. Third, land ownership is widely linked to rural political power. Traditionally, influential villagers usually came from landed backgrounds, and even today this is true, more often than not, of those elected to local village councils.!°? Of course land ownership by individual women, in itself, is not a sufficient condition for

their greater participation or influence in social and political institutions.!°> The ideological barriers to women’s participation in public decision-making bodies, even women endowed with land, could be considerable (see chapter 7 for details). In addition women’s participation in decision-making within local social and political institutions is likely to be linked to the class and caste (or ethnic) structure of the village.!°* The ability of low caste women to have a say in the decision-making processes of village councils, for instance, will not necessarily improve solely by their having small plots of land in a village dominated by upper-caste, rich, male landowners. Nevertheless, land rights

could play a significant role in facilitating such participation. And the barriers of individual isolation could be overcome by group solidarity among the women: for instance, an individual woman or a few women with landed property may yet find it difficult to assert themselves politically or

‘socially in the village, especially where social norms dictate seclusion. However, a large group of women acting in unity could be in a position to do so.'°> (Here there could be some congruence of interests even between women of diverse class and caste backgrounds.) Indeed the term ‘empowerment’ has been equated by some with collective action. However, as a definition of empowerment this would appear 102 For Bangladesh, see Alamgir (1981), Chowdhury (1978), Chowdhury (1987), Rahman (1979), and Solaiman and Alam (1977). For Pakistan, see Ahmad (1968), Alavi (1982), D. Merry (1983) and Raza (1969). For India, see Singh (1961), and R. Singh (1988). 103 It may not be a sufficient condition even for men. For instance, in Pakistan Punjab, Alavi (1982) notes that landowners typically build their political power by organizing factions of supporters consisting of their sharecroppers and other economic dependents, and forging alliances with fellow landlords who too are faction leaders. Here economic power becomes ‘the aggregated power of members of a landowning class who are in close alliance with each other, rather than the power of a single individual or a family’ (p.50). 104 This issue needs more probing. But it is notable that those environmental movements in India in which entire communities or villages have participated, and in which women’s participation in particular has been high, have emerged primarily in hill and tribal areas that are marked by relatively low levels of class, caste, and gender differentiation (also see Agarwal 1991). Further, the experience of women’s cooperatives suggests that class and caste homogeneous institutions are much more conducive to the participation of poor,

low caste women than are heterogeneous ones: mixed class/caste cooperatives, for instance, tend to be dominated by upper-caste, rich peasant women (see e.g. Dixon 1978). 105 For elaboration and illustrative examples, see chapter 9.

42 A field of one’s own too narrow. In a limited sense, collective action may itself empower women

by enhancing their self-confidence and their ability and willingness to challenge oppression, but in a larger sense it is a means to empowerment, wherein empowerment lies not only in the process of challenging gender inequality but in eliminating it. In this sense, collective action is one channel for change. What is important, nevertheless, is that it is often the most effective one. And it is likely to be critical in the struggle for land rights (as will be elaborated in chapter 9). While each of these arguments for women’s independent rights in land is of importance, are they of comparable weight? Do some merely serve what have been described as practical gender needs, while others serve strategic gender needs? This distinction between practical and strategic needs, first

made by Molyneux (1985) and elaborated by Moser (1989), is worth exploring since it also appears to define where, in public policy itself, a line is

drawn on questions of gender. Practical gender needs, as defined by these two scholars, are the needs of basic subsistence (such as food, health care, water supply, etc.); to satisfy them does not challenge women’s position within the gender division of labour, or a given distribution of property or political power. By contrast, strategic gender needs, they argue, are those needs that would help overcome women’s subordination, including transforming the gender division of labour, removing institutionalized forms of

discrimination, such as in rights to own and control property, and establishing political equality. In these terms, women’s land rights would fall under strategic gender needs.

However, the apparent analytical neatness of this distinction is confounded when we examine it from the perspective of practice, on several counts: First, certain strategic gender needs, such as for rights in landed property, are also, in specific contexts, necessary for fulfilling practical gender needs, as evidenced from the welfare and efficiency arguments spelt out earlier: for instance, entitling poor rural women with land may be a necessary component for improving female nutrition and health status. At the same time, we also noted the significance of land in empowering women to challenge unequal gender relations within and outside the home. In other

words, the case for women’s land rights has both a welfare-efficiency (‘practical’) component and an empowerment (‘strategic’) component.

Second, even meeting subsistence needs often requires challenging existing political-economic structures. For instance, a demand for wage increases by poor women workers is a practical need in that it would improve their living standards, but it is strategic in that it challenges existing production relations and requires confronting the (often violent) resistance of employers. Third, and relatedly, the same process, viz. group organization, is often

Land rights for women 43 necessary for fulfilling both practical and strategic gender needs: for instance, organizing into groups would often be just as necessary for poor women to effectively fight for wage increases, as it would be for them to fight for land rights or to confront gender violence. Fourth, action in pursuit of ‘practical’ needs may easily turn into action to meet ‘strategic’ needs. Group organization around economic issues often opens the door for women to raise questions about other aspects of their

lives, such as gender violence within and outside the home, their low representation in local decision-making institutions, and so on. In several

instances in India over the past decade, rural camps held to discuss problems of low wages and employment among village women workers have ended with the women marching to the concerned government office to report and protest cases of rape in their village.!°® Similarly women organized into groups for the better delivery of credit or other economic programmes by the Grameen Bank in Bangladesh, or the Bangladesh Rural Advancement Committee, or the Self-Employed Women’s Association in north India, have in many cases also been able to challenge gender violence or restrictive social practices such as female seclusion (see chapters 2 and 9 for details). Indeed for women to even participate in group meetings often requires them to challenge and overcome the constraints of social norms, to

face the disapproval and even wrath of their husbands, to negotiate the domestic division of labour and childcare with their husbands and other family members, and so on. In other words, the process of fulfilling ‘practical’ gender needs cannot always be delinked from that of fulfilling ‘strategic’ gender needs. To some

extent even elements in the State apparatus are beginning to realize this: in India a recent government attempt to promote adult female education (the Mahila Samakhya (Education for Women’s Equality) Programme launched in 1989) is not only couched in terms of female ‘empowerment’ but recognizes that organizing rural women into groups to discuss gender relations can be a necessary first step toward that end (GOI 1991c). The

eventual fall-outs of this process may indeed be more than the State bargained for. That it may often be more ‘politic’ to couch gender concerns in terms of practical rather than strategic gender needs, in that welfare and

efficiency arguments for focusing on these concerns appear to resonate more easily with State planners, should not detract from this linkage. We might of course ask why welfare and efficiency arguments resonate more with State planners. Part of the answer certainly lies in the fact that these arguments (especially those concerning welfare) focus especially on poor women, and can be subsumed within the poverty-alleviation compo10° Personal observation in Rajasthan and personal communication from several grassroots activists in India.

44 A field of one’s own nent of planning, with special targeting towards ‘the most vulnerable’ groups, identified as women and female children. But part of the answer must also lie in deep-rooted notions of appropriate gender relations shared by many men who make and implement policy, for whom empowering women to transform those relations would appear inappropriate and even threatening to existing family and kinship structures.'°’ Hence it would be easier to push for changes where the goal appears to be to give poor women a slightly better deal, than where the goal is to challenge basic inequities in

gender relations across classes. It 1s also the case that programmes for health and nutrition are more readily perceived in welfare terms than programmes which call for gender-redistributive land reform. It is not a coincidence that /and rights have yet to become a necessary component even of women-directed poverty-alleviation programmes.

Against this background, the question (which I am sometimes asked) — how will providing women ownership rights in land change gender relations, given the enormity of gender inequities in the economic, social and political power structures? — is misleading. It is not just an increase in women’s command over economic resources, but also the process by which

that increase occurs that has a critical bearing on gender relations. Land rights are not a ‘given’ and will not be ‘provided’ to most South Asian women without contestation. Acquiring those rights (as will be made clear from the body of this book) will require simultaneous struggles against many different facets of gender inequities embedded in social norms and practices, access to public decision-making bodies at every level, gendered ideas and representations, and so on. It will require shifts in power balances in women’s favour in several different arenas: within the household, in the community and market, and at different tiers of the State apparatus. Even to organize collectively often requires challenging existing norms, such as breaking the traditional bounds of female seclusion in some communities for attending public meetings. Land thus has a strategic importance that other gender concerns such as

employment and education appear not to share in equal measure. These latter concerns are already a part of the political agenda in most South

Asian countries, there being a broad consensus in public policy that women’s education and (in lesser degree) employment are important for the national good (even if these may still be resisted by many households and communities). But redistributive land reform, even in favour of men, does

not appear to be seriously on the political agenda at the moment, and the distribution of land in favour of women on any significant scale not at 107 See the many quotations in chapters 2, 5, 6, and 9.

Land rights for women 45 all.'°® To have the issue of land rights for women become one of the objectives which political parties are actively pursuing (as opposed to merely including in their rhetoric), to have the issue be given centrality in public policy, will itself need a major struggle against multiple obstacles; and to make those rights a reality, an even greater one. But it is precisely the complex and wide-ranging nature of these obstacles that gives the struggle

to overcome them a transformative potential; and this is also why a successful struggle by women for land is likely to have more far-reaching implications for gender relations in South Asia than possibly any other single factor.

IV Questions addressed, information base, and the book’s structure In further pursuing the issue of women’s independent rights in land, a number of related questions arise, such as: how should we characterize gender relations within the household, the community, the market, and the State, and what role might land rights play in constituting and changing those relations? What rights in land have women historically enjoyed, and

how and why did these change over time? Were gender relations more egalitarian within traditionally matrilineal and bilateral communities? Were there structural links between women’s land rights in these communities and other aspects of women’s lives, including post-marital residence, choice of spouse, and so on? What inheritance rights in landed property are

granted to women under contemporary law, and to what extent are they 108 In their 1991 election manifestos, all the major political parties in India, when spelling out their policy proposals concerning women, spoke of women’s employment, and most also spoke of female education, but very few mentioned the issue of land. And those that did appear only to be paying lip service, since they did little to promote women’s access to land in the states where they were (or continue to be) in power. The Communist Party of India’s manifesto did not mention the issue at all. The Congress (I) manifesto did so only in terms

of distributing titles to public land jointly or individually in the names of women, and although some steps have been taken in this direction under the Congress government, the

amount of land involved (as noted earlier), is minimal. The Bharatiya Janata Party manifesto promised to make women equal shareholders in the husband’s wealth, but ignored women’s rights in the patrimony, and in practice the party appears to have done little to promote even the rights of widows in the states in which it was elected in 1991. The manifesto of the National Front likewise held that women would be assured equal rights in family property, and that gender-discriminatory laws, including inheritance laws, would be reviewed, but the party’s policies during its tenure in the central government (December 1989—June 1991, including the Janata Dal period) give little reason to expect that this was more than rhetoric. Similarly the CPI (M), which alone explicitly promised to take measures to ensure women’s equal rights in /anded property, gives little hope that this

promise is meant to be taken seriously, since when it had had the opportunity of implementing such measures in its Operation Barga programme, say by registering both spouses, it paid little attention to women’s interests. It remains to be seen how these election promises will be dealt with by these parties in the future.

46 A field of one’s own able to exercise these in practice? What factors obstruct women from claiming or being able to exercise effective control over their shares, and what kinds of measures would strengthen their ability to overcome these obstacles? To what extent have women’s claims received recognition in organized land struggles? What are the likely future scenarios? These questions are addressed in this book in the context of five South Asian countries: India, Pakistan, Bangladesh, Nepal, and Sri Lanka. These countries constitute a region, given the many common elements in their histories, languages, and cultures that cut across national boundaries; but today they are also distinct entities, with different State structures, politics,

economic programmes, and welfare outcomes. This dual character of South Asia is reflected throughout the book, as the discussion shifts between a regional focus and a country-specific focus. (Table Al.2 gives a comparative picture of selected development indicators by country.) Most of the questions mentioned above have never been raised, let alone

adequately addressed, in the South Asian context. No country in South Asia, with the exception of Sri Lanka, even collects gender-disaggregated land ownership and land-use data in its agricultural or centennial censuses or its large-scale rural surveys. The available data are usually aggregated by household, so it is not possible to say how much land 1s owned by women relative to men. In Sri Lanka, although gender-disaggregated land ownership and use data were collected in the 1982 agricultural census, they were limited to agricultural operators and did not cover all rural households; and the published data I have seen do not give a gender-wise breakdown of land ownership even among agricultural operators. Similarly, although a large number of village studies have focused on questions of agrarian structure, land use, and land ownership, virtually none has gathered gender-disaggregated information on these aspects. Hence to gain an idea of where women have been given or have claimed their shares in landed property and under what circumstances, we have to probe historical, anthropological, and legal sources. To answer the larger body of questions raised above is equally difficult and again requires spreading the analytical and disciplinary net wide. The

historical story presented here of women’s customary land rights and changes therein has been woven from colonial documents, accounts by medieval travellers, historical ethnographies, and legal sources. The theoretical framework for understanding gender relations builds on and extends aspects of recent work by a few economists (including my own work), using the bargaining approach to conceptualize social relations. For delineating contemporary law and its formulation I have drawn on historical and legal

documents and commentaries, parliamentary debates, and accounts by women leaders of the struggles for property rights in the early twentieth

Land rights for women 47 century. The obstacles women face in realizing their rights and the regional variations in these have been conceptually inferred and empirically documented by sifting through a vast body of ethnographies and village studies, mostly available in the form of unpublished dissertations. And amidst all these texts, I have searched, not always with success, for rural women’s own VOICES.

Most ethnographies and village accounts on South Asia are studies of single villages and/or communities, and the majority date from the postindependence periods, especially the 1960s and 1970s. With occasional exceptions, their concern is not specifically with gender issues, and virtually none focuses on women’s rights in land. They are preoccupied variously with kinship and marriage practices, land tenure, village politics, and so on. A large number consist of unpublished doctoral dissertations in anthropology, sociology and, occasionally, economics. The questions I am seeking to answer using some of the material they provide are not the central questions (and at times not even the tangential questions) these studies ask. What this

has necessitated, therefore, is a gleaning (something village women are

adept at) for information, amidst a harvest from a different crop of questions. Out of several hundred doctoral dissertations and other village ~ accounts which I examined, only a small number provided much useful information. And even these few mostly focus upon norms rather than actual practice. For instance, the accounts are more likely to mention that ‘women are allowed to inherit in the absence of sons, than to say what proportion, if any, actually do so in the village under study. Also there is a noticeable regional and temporal clustering of ethnogra-

phies (as will be apparent from the tables in chapter 8). Regionally, Pakistan Punjab, the middle-hills of Nepal, Bangladesh, India’s northwestern states and the state of Karnataka, and the Kandyan Highlands in Sri Lanka, have been particularly favoured. But I found no ethnographies on Sind in eastern Pakistan, and few on central and eastern India, northern Sri Lanka, and the terai plains in Nepal. Time-wise, since the 1970s, there appears to have been a marked decline in ethnographies on India and Sri Lanka and a parallel rise of such research on Bangladesh and Nepal.!°? Despite these limitations and biases, however, it is possible to paint a broad picture and point out emerging trends. These materials have been supplemented by my fieldwork in northwest and northeast India, especially Rajasthan (in 1986-87) and Meghalaya (in 1989); discussions in each of the five South Asian countries with social 109 Among factors which probably contributed to this shift would be the cooling of Indo-US

relations after the 1971 Bangladesh war, the increased availability of international funding for research on Bangladesh, and the greater ease with which foreign scholars got governmental permission for research in parts of South Asia other than India.

48 A field of one’s own activists, village women, lawyers, government functionaries, and academics; and village visits in Sri Lanka, Bangladesh, and Nepal. In the chapters which follow, chapter 2 is theoretical and suggests how a bargaining approach may be used to analyse the process by which gender relations are constituted and how they are maintained and change over time, both within and outside the household. Chapters 3 and 4 are primarily historical. The former details what rights in landed property were customarily enjoyed by women under patrilineal, matrilineal, and bilateral systems of inheritance. It also traces the structural links between women’s land rights in matrilineal and bilateral communities and social practices governing post-marital residence, choice of spouse, and sexual restrictions on women. Chapter 4 analyses how and why these rights were eroded within these communities. The historical process of contestation by which contemporary property laws in relation to women came to be formulated, the contents of these laws today, and the gender inequalities still inherent in

them, are elaborated in chapter 5. Chapters 6 to 8 then identify, respectively, the obstacles women face in claiming their legal shares in landed property; what constrains them from exercising control over the land they do possess; and the cross-regional variations in the obstacles and constraining factors. These chapters also demonstrate how women’s struggle for effective land rights needs simultaneously to be a struggle to change existing ideological constructions of gender. Chapter 9 examines how women have

sought to resist gender inequities through both covert individual action, and overt collective action, the latter especially in the context of grassroots land-related movements. Finally, the issues which will need particular focus in the long and difficult struggle for realizing women’s land rights, and the macro-context which will impinge on such a struggle, are discussed in the concluding chapter 10.

Land rights for women 49 Table Al.1: Village common lands in India by State: 1987-88

VC area as a VC area? Total VC area! percentage of per rural

State (in °000 ha) geographic area person (ha)

Andhra Pradesh 5465 19.9 0.12 Arunachal Pradesh Til 1.3 0.16 Assam? 1913 24.4 0.09

Bihar 2623 15.1 0.04 Gujarat* 5515 28.1 0.21 Haryana 229 5.2 0.02 Himachal Pradesh 1549 27.8 0.35

Jammu and Kashmir 551 2.5 0.10 Karnataka 2811 14.7 0.10 Madhya Pradesh 759017.5 17.10.12 0.16 Maharashtra 5376

Kerala 219 5.6 0.01

Meghalaya 817 25.5 36.4 1.21 0.65 Mizoram® 538 Nagaland 201 12.1 0.25

Orissa 0.08 Punjab1945 114 12.5 2.3 0.01

Rajasthan 13794 40.3 0.79 0.44 Sikkim® 252 35.5 ‘Tamil Nadu 1830 14.1 0.05 Uttar Pradesh 3371 11.1 0.03 West Bengal 377 4.2 0.01

Other areas 1578 33.2 0.30 All India 58769 17.9 0.10

Notes: ' This includes the following categories of land: fallow land, other than under current fallow; cultivable wasteland; permanent pastures and other grazing lands; and barren and uncultivable wasteland. It excludes protected and unclassed state forests and private land to which people may have use access by local custom. The estimates for village commons given here are very rough and likely to be inflated since: (a) some part of the cultivable wasteland is likely to have been encroached upon and de facto privatized; and (b) in some states the barren and uncultivable wasteland would include rocky land, marshes, sand dunes, mountainous slopes, etc. which would effectively be inaccessible or likely to provide little of use to the village communities. In Rajasthan, for instance, the high figures for village commons are primarily accounted for by barren land, a significant part of which would consist of desert dunes. 2 Per capita estimates are based on the Rural Population Projections, as on | January 1988, made by the Expert Committee on Population Projections and reproduced in GOI (1990b: 16).

3 Area figures relate to the year 1981-82. 4 Area figures relate to the year 1985-86. > Area figures relate to the year 1974-75. © Area figures relate to the year 1984-85. Source: Computed from the Government of India’s ‘Land Use Statistics’ which classify

geographic area into various categories (see GOI 1992c). |

=~ ~ ~ > + S ON xs ”to io)+a= , one ON 22x ~ AN +- OA + a a

Ss m™ AQ 00 ™ Mm 3)

yet] Nm = A OO . $ “ m0 dASb ~ 2 2. ASRS

Fos S aa) Ss oR 8 oy a_

= very wn m og O fl, WMAMNM &a wee”

(ue

Q

ote ee DD AMA ZSSoOSsSrl agagags9)3 a

3 3 wn G x S

o «c 22 8 g “EE = Z £ 22a 2 SO x I2V2sree = 8352 % ime= | oe) =e ©= as Foo}

wy

om

om ON

oom MOON @) vom

3voy—m2 =GarARNe & => enese Sse& ee

y o . oO wt HAO ao) Sn &

in ©

= ZEaNFloececooll om — nn=-as+ WW ™SFveEs OOn ue WY GOso) STTili 30

ai ZAas = 3 ° om te DO te OO < 2 i -S||/=e2t 2 © Ss 2

ry = 3 86 & Ce ra) = wma S25 go" os C 2O |aEZaAnIlezauctH SoS S35 5" 8YB & am

2 Conceptualizing gender relations

Please go and ask the sarkar [government] why when it distributes land we don’t get a title? Are we not peasants? If my husband throws me out, what is my security?!

Economists of both persuasions [neoclassical and Marxist] tend to treat the household as though it were an almost wholly cooperative, altruistic

unit. Today, however, they are confronted by certain ‘anomalies’ —

hold. (Folbre 1986: 245)

empirical evidence of economic conflict and inequality within the house-

Gender relations are neither uniform across societies nor historically static. A vast array of studies of different cultures, regions, and communities bears this out, as does the material on South Asia in this book. However, these

call for a conceptual framework that would help us characterize gender relations, explain how they are maintained, and identify the processes by which they might change over time. This chapter is an attempt in that direction. The term gender relations as used here refers to the relations of power between women and men which are revealed in a range of practices, ideas, and representations, including the division of labour, roles, and resources between women and men, and the ascribing to them of different abilities,

attitudes, desires, personality traits, behavioural patterns, and so on. Gender relations are both constituted by and help constitute these practices and ideologies in interaction with other structures of social hierarchy such as Class, caste, and race. They may be seen as largely socially constructed (rather than biologically determined),? and as variable over time and place. ' A message conveyed by poor peasant women to the West Bengal Government at its seminar on Women and Development in January 1979, through the women they had elected to the village panchayat (personal communication, Vina Mazumdar, 1992).

2 This is not to deny the possible role of biology (pregnancy, childbearing, etc.) in the historical construction of some aspects of gender relations, such as the division of labour. But biology cannot explain the entire gamut of gender inequalities we observe today, nor even the perpetuation of an observed gender division of labour (e.g. technical developments have minimized the importance of muscular strength; contraceptive technology

reduces the disability of frequent pregnancies; and a variety of possible childcare arrangements make childcare a less binding constraint). In any case, the considerable variation of gender relations across cultures indicates the enormous importance of nonbiological factors. 51

52 A field of one’s own Further, although gender relations are defined here as relations between women and men, gender hierarchies also influence and structure relations between individuals of the same sex — e.g. how two women of the same household relate to one another is affected by the gendered character of their relations with the household men.? The relationship between a woman and her daughter-in-law is one example.

It is suggested here that gender relations are characterized by both cooperation and conflict, and that their hierarchical character in any given context is maintained or changed through a process of (implicit or explicit)

contestation or bargaining between actors with differential access to economic, political, and social power. This contestation can vary: —in form, ranging from women’s covert individual acts of resistance to overt group mobilization, with varying degrees of overt individual action and covert group resistance in-between;* —in content, relating to a spectrum of economic, social or political rules, practices, and institutions. For instance, contestation can occur over how women are perceived and ideologically constructed, or over what economic returns their work commands, or the interlinks between the two; and

—in the arenas within which it takes place: for instance, the household/family, the community, the market, and the State. These arenas are interactive rather than mutually independent and can reinforce or weaken each other’s impact. In suggesting this framework I do not seek to delve into the contentious question of how women’s subordination originated historically.* Rather my concern is to explore how gender inequities are currently structured and perpetuated, and how they could be subject to change. In particular, what

role might women’s rights in landed property (or lack thereof) play in structuring, maintaining, and changing these relations? The discussion below explores this issue both within the household/family and outside it — the latter especially in relation to the community and the State. The possible interactive effects of these three arenas are also discussed. Although a distinction is sometimes made between the terms ‘household’ and ‘family’,® for my purpose here I have used the terms in conjunction or 3 See Moore (1991) for a useful discussion on this point. + For a detailed discussion on covert and overt resistance, see chapter 9. > At the same time, the notion that gender relations were historically constructed by a process of contestation could be a useful one to explore, although existing evidence appears inadequate to trace the process concretely. © For instance see Shah (1973), who differentiates household from family, defining the household as a commensal and residential unit composed of members sharing one hearth and roof, and the family as a wider kinship group whose members may be living in more than one household.

Conceptualizing gender relations 53 interchangeably, given the considerable empirical variability of these units across regions, and their definitional variability across the literature. For instance, households can be commensal and residential units, and/or units

of joint property ownership, production, consumption or investment, or they can constitute some intersection of these dimensions. They also vary in membership composition from units of single persons, to those of parents and children, and those with additional relatives: siblings, grandparents, and so on.’ Consider then the issue of gender relations in the household/family.

I. Gender relations within the household/family Are you suggesting that women be given rights in land? What do women want? To break up the family? (Minister of Agriculture to the author at an Indian Planning Commission seminar on Land Reform, June 1989)

Whether or not so intended, the Minister’s reaction implies at least two assumptions about the family: —that the stability of the family as an institution is linked to the maintenance of unequal resource positions between women and men; and —that economic self-interest plays a significant role in intra-family

gender relations, which would be revealed with particular starkness in gender conflict over a critical form of property such as land. Such a picture of the family is a far cry from the assumption implicit in much of standard economic theory that the family is an undifferentiated

unit, or that it is governed primarily or solely by altruism.’ Indeed, as pointed out in some recent feminist critiques, there is a noteworthy similarity between neoclassical and Marxist economic theories on this count (Folbre 1986, Hart 1990). Yet the growing evidence (noted in chapter ’ For a useful review of regional variations in household composition across India, see Kolenda (1987). As Kolenda’s study shows, Indian families operate variously as joint consumption and/or production units; and joint consumption units themselves vary enormously in composition. For instance, in patrilineal contexts, there are cases of parents and one or more married sons maintaining a common hearth, dwelling, and landholding. There are others of sons establishing separate hearths and dwellings but cultivating jointly with the father. And there are yet others where each son has a separate hearth, dwelling, and landholding. In Rajasthan I have also seen families which live under one roof but subunits of which maintain separate hearths. In the African context, spouses often maintain separate farm enterprises and budgets and may even live in separate residential units (see various essays in Guyer and Peters, eds. 1987). 8 As Folbre (1986: 247) notes: ‘There is something paradoxical about the juxtaposition of naked self-interest that presumably motivates efficient allocation of market resources and perfect altruism that presumably motivates equitable allocation of family resources.’

54 A field of one’s own 1) of persistent intra-family inequalities in the distribution of resources and tasks, and of gender differences in expenditure patterns, as well as cross-

cultural anthropological descriptions of intra-family interactions and decision-making, indicate the need for a very different conceptualization of

the household: one that takes account of multiple actors, with varying (often conflicting) preferences and interests, and differential abilities to pursue and realize those interests.°

(1) The bargaining approach In the context of the present discussion, it would be useful to conceptualize the household/family as a complex matrix of relationships in which there is

ongoing (often implicit) negotiation, subject to the constraints set by gender, age, type of relationship (kinship association), and what could be termed ‘undisputed tradition’ (elaborated later). The nature of this intra-household interaction could usefully be des-

cribed as simultaneously containing elements of both cooperation and conflict.!° The members of a household cooperate in so far as cooperative arrangements make each of them better-off than non-cooperation. However, many different cooperative outcomes are possible in relation to who does what, who gets what goods and services, and how each household member is treated. These outcomes are beneficial to the negotiating parties relative to non-cooperation. But among the set of efficient cooperative outcomes,!! some are more favourable to each party than others — that is, One person’s gain is another person’s loss — hence the underlying conflict between those cooperating. Which outcome will emerge depends on the relative bargaining power of the household members. A member’s bargaining power would be defined by a range of factors, in particular the strength of the person’s fall-back position (the outside options which determine how well off he or she would be if cooperation ceased), and the degree to which his/her claim is seen as socially and legally legitimate. 12 The person who has

a stronger fall-back position (better outside options), and/or whose claim ® For interesting discussions on some of the problems associated with a unitary conceptualization of the household, also see the writings of economists Sen (1983, 1990a), Hart (1990), Kabeer (1991a), Folbre (1986, 1988), and Agarwal (1990a); anthropologists Guyer (1981),

Moore (1991), Harris (1981), and several others in Guyer and Peters, eds. (1987); and various articles in DS Bulletin (1991) on ‘Researching the Household: Methodological and Empirical Issues’. '0 The term ‘cooperative conflict’ to describe these intra-household interactions has been popularized by the writings of Amartya Sen (1983, 1985, 1990a), which will be discussed in more detail below. 1! An efficient outcome, in economic parlance, is one in which it is not possible to make any party better off without making some other party worse off. 12 By social legitimacy I mean that which is accepted and enforced as legitimate by the community (identified by kinship, caste, religion, or location) of which the household is a part. By legal legitimacy I mean that which is established in law. The two need not coincide.

Conceptualizing gender relations 55 enjoys greater legitimacy, would emerge with a more favourable outcome, although both parties would be better-off than if they did not cooperate. !% The application of a bargaining perspective involving cooperation and

conflict and the notion of a fall-back position to characterize intrahousehold dynamics is relatively new, but growing. Although he did not focus on the household, these concepts were at the core of Nash’s (1950, 1953) formulation of ‘bargaining problems’ within game theory, and have

recently been applied in various ways (within and outside the gametheoretic format) to the household by a number of economists, including Amartya Sen (1983, 1985, 1990a).!+ These efforts are in contrast to the

simplistic model of a unitary household that has typified economic

analysis. How we characterize the household — as an arena of bargaining and contestation, or as a place where all decisions are made by consensus or by an altruistic household head (as typically assumed under the unitary model) — 1s not merely of academic interest. It can impinge critically on policy decisions regarding whom resources and programmes get directed toward. As we had noted in chapter 1, policy makers in South Asia have (implicitly or explicitly) assumed a unitary household model and have tended to direct resources principally at male household heads, trusting that the resources will be shared equitably within the household; but empirical evidence shows considerable intra-household inequities, and not only in resource sharing. The bargaining approach, suggested here as an alternative conceptualiza-

tion, implies that government policies and resources would need to be directed differently. At the same time, as Sen has effectively pointed out, it is critical to think beyond the restrictions imposed by fully specified quantitative bargaining models that have characterized most economic work on this issue, and to move towards a less restrictive formulation which incorporates qualitative

aspects.'° In other words, the bargaining perspective or approach has 13 My emphasis on these factors is not to deny that feelings of love and concern are also important in shaping family relations and economic outcomes. '4 Among others, see especially Manser and Brown (1980), Jones (1983, 1986), McElroy and Horney (1981), and Seiz’s (1991) review piece. '5 Itis argued by some that since available data don’t usually permit us to test hypotheses that

would help choose between the bargaining and neoclassical household models, the analytically simpler neoclassical model is preferable. This, however, is not an adequate reason for dismissing the bargaining approach, but merely strengthens the case for more adequate data gathering. Also see Hoddinott (1991) for a useful summary discussion comparing the ‘altruistic’ and ‘bargaining’ models of the household. 16 The social legitimacy of claims, mentioned above, is one such qualitative aspect. Formal

bargaining models generally ignore these and focus only on fall-back positions as determining bargaining power. For some extensions and applications of such qualitative dimensions by economists in relation to South Asia, see especially, Sen (1990a), Agarwal (1990a), and Kabeer (1991a). Some recent anthropological writings also seek to elaborate on the concept of intra-household bargaining (see especially, Moore 1991).

56 A field of one’s own particular usefulness in examining gender relations, in the application

of which we need not restrict ourselves to formal game-theoretic formulations.!7 Sen’s point of departure from formal bargaining models stems from his

recognition that the outcome of bargaining will depend not only on a person’s fall-back position should cooperation cease, but also on what he terms perceived interest response and perceived contribution response. The outcome would be less favourable to a person (a) the less value s/he attaches to her/his own well-being relative to the well-being of others (perceived interest response), and (b) the smaller her/his contribution to the household economy Is perceived to be (perceived contribution response). Sen argues that both types of perceptions are usually biased in women’s disfavour, although these biases operate in varying degrees in different societies: they tend to be especially adverse in ‘traditional societies’ such as India, where women may tend not to think in terms of self-interest or their individual well-being, and where women’s economic contributions to the household are seriously undervalued. Since notions about legitimate shares in the family’s resources are linked to such perceptions, women will get less because they are seen (by both themselves and men) as ‘naturally’ deserving less. Sen places considerable importance on a woman’s earnings outside the home for giving her a stronger fall-back position, a clearer perception of her own well-being, and a higher valuation of her contribution. (Although he appears to emphasize only other people’s valuation of women’s contributions, clearly a woman’s own valuation of her contribution would also be relevant here.) Sen’s emphasis on notions of legitimacy as determinants of what a person

gets, in addition to the person’s fall-back position, is important, and I concur with it. But I extend his analysis and, in some respects, depart from

it, in five ways: (1) I extend his specification of the factors that might determine a person’s fall-back position in intra-household bargaining over consumption, by identifying additional factors. What factors might affect a person’s bargaining power over a share in family land are also specified here. I argue that there are several levels of bargaining, and some variables such as land rights are determinants of bargaining at one level but outcomes

at another. (2) I debate his conceptualization of ‘perceived interest response’ and its importance in determining the outcomes of bargaining within the family in terms of women’s welfare. (3) I argue that the notion of

legitimacy needs to be broader than that captured by Sen’s ‘perceived contribution response’. (4) I explore the effect of gender differentials in '7 Also see Seiz’s (1991) succinct exposition of this distinction between bargaining models and

the bargaining approach.

Conceptualizing gender relations 57 bargaining power not only on outcomes in relation to specified items/issues,

but also on what is bargained about. (5) I apply the bargaining approach also to gender relations in arenas outside the household. The first point, on the factors determining women’s fall-back position and bargaining power, is discussed in subsection (2) below; and the last

point on extra-household bargaining is elaborated in Section II of this chapter. On the second point, regarding women’s perceptions of their selfinterest, I see the problems as three-fold (further elaborated in chapter 9):

(a) That women tend to have a less sharp perception of their individual interests in societies such as India, that is that they may suffer from a form

of ‘false consciousness’ about their own well-being, is debatable. The empirical evidence on this is limited, and that which exists (described in chapter 9) points more to the contrary: it suggests that women’s overt compliance with practices which disadvantage them does not necessarily mean that they accept those practices as legitimate; their perceptions are better revealed in their many covert forms of resistance to gender inequities.

(b) To the extent that women do seek to maximize ‘family’ welfare, this could still be consistent with their long-term self-interest (even if it is at the cost of their immediate well-being), in so far as women are more dependent on the family for their survival than are men. This dependence can be both economic and social (the last especially in contexts where female seclusion is

important and women need male mediation to deal with outside-family institutions).'® Also women’s dependence on the family can extend over a long period in so far as they often tend to survive longer into old age than men, and widowhood carries with it social disabilities that widowerhood does not. In the circumstances, women may well come to believe that no

other option than compliance is possible. The appropriate conclusion would then be not so much that women need to realize they deserve better, but that they need to believe they can get a better deal, and to know how that would be possible. In other words, in explaining intra-household gender inequalities and gender gaps in measures of well-being, I would place much less emphasis than Sen does on women’s perceptions of their self-interest and much more on the external constraints to their acting on those interests.

Or, to put it another way, what may be needed is less a sharpening of women’s sense of self-interest than an improvement in their ability to pursue that interest, including by strengthening their bargaining power. A

part of this strengthening would come from improving their fall-back position, and a part from strengthening the legitimacy of their claims as perceived by others. (c) Women may of course sacrifice their individual interests for the interests of family members out of conscious choice, but '8 For elaboration see chapter 7.

58 A field of one’s own here the notion of ‘family’ also needs probing. The ‘family’ for whose well-

being women may be willing to make sacrifices may include only their children, on whose behalf women may still seek to strike a hard bargain with their husbands or kin. Indeed they may do so more overtly than if they were acting only on their own behalf. Again there is some evidence on this from South Asia, outlined in chapter 9, which suggests that women usually

see their interests as congruent to those of their dependent children, but often potentially antagonistic to those of their husbands. Also, sacrificing for one’s children may well reduce a woman’s material well-being, but this need not be a case of false perception on her part. Both altruism and the pursuit of self-interest can be self-aware actions. On the third point, notions of legitimate shares may stem from a variety of ethical principles, of which a person’s contribution is only one. Among others could be needs. Reducing the gap between women’s actual economic contributions and social perceptions about their contributions may not, on

its own, strengthen the legitimacy of their claims and improve their consumption shares (as Sen suggests) if, for instance, the criterion for justifying particular shares is needs rather than contributions. Indeed, contributions and needs are often posed as competing principles in justifying the shares of individuals: ‘to each according to her work’ vs. ‘to

each according to her needs’. To change perceptions about economic contributions may in fact be easier than to change perceptions about needs,

in so far as the former may be easier to quantify than the latter. But whichever criterion (contribution or need) one uses, the recognized social and legal legitimacy of a claim could be both a determinant of bargaining

power and an outcome of bargaining power. And, to establish the legitimacy of a claim would typically require not just a change in ideas, but their enforcement through some system of social and/or State sanctions. On the fourth point, gender differentials in bargaining power are likely to affect not only outcomes in relation to specified items/issues but also what is

bargained about. Not all issues may be accepted as legitimate ones for contestation. At any given time, for a given society, some decisions would fall in the realm of what the French sociologist Bourdieu (1977: 167-70) terms ‘doxa’— that which is accepted as a natural and self-evident part of the social order, which goes without saying and is not open to questioning or contestation — the ‘undiscussed, unnamed, admitted without argument or scrutiny’. A good deal of what is justified in the name of ‘tradition’ would

fall in this category: ‘the tradition is silent, not least about itself as tradition’. In contrast to doxa is the ‘field of opinion, of that which is explicitly questioned’, ‘the locus of the confrontation of competing discourses.’!9 19 Within ‘the field of opinion’ Bourdieu further distinguishes between orthodoxy and heterodoxy. He does not fully spell out this distinction, but implies that orthodoxy would

Conceptualizing gender relations 59 In the present context, doxa could include widely accepted practices that

favour group over individual interests (such as strategic marriage alliances), or that favour some groups over others (such as a given gender division of labour, or women eating last and the least nutritious foods in many regions of South Asia), or that favour some individuals over others (such as an older daughter-in-law who has sons being allowed to participate in household decision-making where a new bride may not), and so on. The acceptance of such practices also reflects the dominant perceptions of the needs and rights of people (say of women in relation to men, or of younger people in relation to older) prevailing in a community,?° and contestation may be necessary to establish the legitimacy both of alternative notions of needs and rights, and of specific persons as contestants. The interest of the dominant groups would be to maintain the space of doxa, while that of the dominated would be to reduce it by exposing the arbitrariness of the taken-for-granted. In other words, what constitutes doxa may itself be subject to challenge and change. In Bourdieu’s (1977: 169) schema, such change would come about ‘when the dominated have the material and symbolic means of rejecting the definition of the real that is imposed on them’; or, putting it in terms of our bargaining framework,

when the dominated have a stronger bargaining position. For our purposes, what is noteworthy here is that the strength of a person’s bargaining power: (a) is defined not only by material circumstances but also by symbolic meanings; and (b) not only affects the outcomes of bargaining __ over those issues/items which are admitted into the realm of contestation but also what issues/items are so admitted. The notion of contestation implies that the disadvantaged perceive the conflict between their self-interest and the socially defined order. A critical question then is: to what extent do dominated groups in fact perceive such conflict? Do they comply with the ideologies and practices of the dominant groups because they have internalized them through a process of socialization, or because of a lack of choice, or some combination of both? These questions are also relevant for Sen’s discussion on women’s perception of self-interest. Neither Bourdieu’s notion of doxa, nor Gramsci’s characteri-

zation of hegemony which closely parallels it, explicitly addresses or resolves these questions, although it can be inferred from Gramsci’s writings that his emphasis was on consent via internalization.2! Observa-

tionally this issue is not easy to settle, since it is difficult to infer from be at one end of the spectrum and heterodoxy at the other, the former representing one dominant system of beliefs and the latter representing several alternative systems of beliefs.

20 On how discourses about women’s needs tend to be structured by the power relations between women and men, see Fraser (1989). 21 Gramsci’s (1971) discussion on hegemony, scattered through his Prison Notebooks, has been subject to numerous commentaries. On this specific point, Femia’s (1987) lucid discussion is particularly useful. Gramsci distinguishes between ‘hegemony’ and ‘domina-

60 A field of one’s own people’s overt behaviour whether they are conforming because they fully accept the legitimacy of an unequal order, or partially accept it, or out of fear, or because they have (or believe they have) no other options. More clues are provided by examining the covert ways in which disadvantaged groups resist authority and power. As noted, observational evidence for South Asian women (that will be discussed in chapter 9) appears to conform more with women’s lack of (or belief that they lack) options than with their unquestioning acceptance of male dominance.” In any case, what is important here is that the outcomes of bargaining over work, economic resources, and so on, are critically and dialectically linked with the meanings attached to them, and with the accepted legiti-

macy (or otherwise) of certain claims, needs, rights, etc. Ideological contestations are therefore an integral part of contestations over material resources.

(2) What determines intra-family bargaining power? To identify the factors which are important in determining the outcomes of bargaining is complex for several reasons: First of all, as noted, a wide range of (not always quantifiable) factors could define a person’s bargaining power, such as individual economic assets and communal/external economic support systems which determine the person’s fall-back position, as well as social and legal recognition of the legitimacy of the claim. The complexity of bargaining would be considerably less if the social legitimacy of a person’s claim to some share of the contested item is recognized and only the size of that share 1s in dispute, than if the legitimacy of claiming any share at all in the item is unaccepted. Cases in point are the intra-household gender division of subsistence resources and of ancestral land. The rights of

female family members to basic subsistence are usually not in dispute, although the proportionate shares may be, while any share for them in arable ancestral land might be disputed in many communities. tion’, the latter constituting power exercised (by the dominant class or social group) through direct coercion, hegemony representing a more subtle and insidious exercise of power by obtaining the ‘consent’ of the dominated. However, as Femia elucidates, what Gramsci meant by ‘consent’ and what sort of conforming behaviour he had in mind, can be _ Subject to varied interpretations. Also see Williams’ (1977) discussion on hegemony. 22 Sen too recognizes that deprived groups may comply for many different reasons — habit, hopelessness, resignation, etc. ~ but appears to see this as resulting in their willingness to accept the legitimacy of the established order rather than in their covertly resisting that order. He writes (1990a: 127): Deprived groups may be habituated to inequality, may be unaware of possibilities of social change, may be hopeless about upliftment of objective circumstances of misery, may be resigned to fate, and may well be willing to accept the legitimacy of the established order.

Conceptualizing gender relations 61 Second, in bargaining for something like a share in arable land, in so far as the social or legal legitimacy of any share at all for women may first need

to be established, the outcomes of intra-household bargaining would be pre-conditioned by the outcomes of extra-household bargaining with the community and the State (as elaborated later in this chapter). Third, some resources would both be determinants of a person’s bargaining power vis-a-vis other resources, and themselves need to be bargained for. Again arable land is a good example. We could argue that women with independent ownership in arable land (which could be used productively

for subsistence) would have a stronger fall-back position and greater economic bargaining power than landless women, vis-a-vis the allocation of household subsistence. At the same time, to gain a share in arable land may

itself require bargaining, and another set of factors would determine women’s bargaining power in relation to land. This could also be seen as two levels of interlinked bargaining.

Fourth, the outcomes of bargaining at a given point in time, by strengthening or weakening a person’s fall-back position, could affect the outcomes of bargaining at a later point in time.?? Fifth, the outcomes of contestation or bargaining may or may not be the

result of an explicit observable process of discussion and negotiation between two or more parties. Nevertheless they could be seen to result implicitly from relative bargaining power. For instance, a man does not necessarily have to tell his sister in a north Indian village that he will break off all contact with her if she demands her share of ancestral land. That he can do so at low economic and social cost to himself, but at high potential cost to her, may be enough for her to forego her claim. Indeed the fact that one party can get a favourable outcome even without open contestation, suggests a considerable bargaining power. To help concretize the discussion, consider the issue of land rights as a factor in bargaining at the two levels mentioned: for household subsistence and for land itself.

Bargaining for subsistence within the family. What determines a woman’s bargaining power within the family in relation to subsistence needs such as food and health care? Sen (1981), in his entitlement approach to famine, highlights two factors as significant in determining a person’s (or

a family’s) ability to command goods (including food) and services: endowments (ownership of productive resources) and the exchange entitle-

ment mapping (that is, the exchange possibilities that exist through production and trade, which determine the consumption set available to a 73 Also see Sen (1990a) on this point.

62 A field of one’s own person with a given endowment).2+ For our purposes here (that is in the context of rural families), the most important determinants of exchange entitlement would be ownership of land and access to employment and other income-earning means (given the production and trade possibilities and the structure of factor and commodity prices). In addition, Sen’s approach can usefully be extended to cover three other types of entitlements which do not derive from private ownership nor usually from market exchange, namely those stemming from traditional rights in communal resources, from traditional external social support systems, and from the State and non-governmental organizations (NGQs). Social support systems are constituted of relationships between persons or social groups in which considerations other than the solely economic take precedence, falling under the rubric of what has been termed by some as ‘the

moral economy’ (see e.g. Scott 1976, Greenough 1982). These typically relate to non-market transactions, such as relatives or friends providing

informal credit without interest during an economic crisis, or intergenerational transfers between parents and children during times of need, and so on.

In other words, it could be suggested that a rural person’s fall-back position (and associated bargaining strength within the family vis-d-vis subsistence needs) would depend especially on five (although not always unrelated) factors: ——private ownership and control over assets, especially arable land; —access to employment and other income-earning means; —access to communal resources such as village commons and forests; —access to traditional external social support systems; and —access to support from the State or from NGOs.?° These five factors impinge directly on a person’s ability to fulfil subsistence needs outside the family. The premise here 1s that the greater a person’s 24 Sen, in this context, does not address the question: how do families, or individuals within them, arrive at a certain endowment position, a question critical to the present analysis.

28 The category, NGOs, used here very broadly, is vast and varied, containing within it organizations that differ in the size and social backgrounds of their membership, their objectives, ideological positions, the issues they take up and their approaches to those issues, their forms of operation, sources of funding, physical locations, and so on. For many NGOs, gender is not a concern at all; for others it is a tangential concern, and for yet others a central one. I will use the term ‘gender-progressive NGOs’ to include all those whose activities are centrally or partially aimed at reducing gender inequities. This could include organizations with mixed (male and female) membership but with a specific gender focus in their activities, as well as women’s groups, such as women’s cells or wings of political parties, women’s groups autonomous of political parties having a mass base, or those with small memberships promoting gender-specific programmes, gender-focused research institutions, and so on.

Conceptualizing gender relations 63 ability to physically survive outside the family, the greater would be her/his bargaining power (at least in relation to resource sharing for subsistence) within the family. Inequalities among family members in respect of these factors would place some members in a weaker bargaining position relative to others. Gender is one such basis of inequality, age another.?°

While my focus here will be on the determinants of well-being if cooperation should break down, it is important to note that many of the factors which determine a person’s fall-back position also influence her/his

ability to make contributions within the relationship. If a woman loses access to employment, for example, it both worsens her fall-back position and diminishes the income she can bring into the family. This dual effect may have devastating consequences for her in periods of severe crisis (such as a famine), as we shall see below.

These factors can be complementary to, as well as substitutes for, one another. The significance of the first two, namely ownership of privatized land and access to employment, in strengthening a person’s survival ability outside the family, is self-evident; that of access to communal resources and social support systems needs elaboration. In South Asia, village commons (VCs) and State forests acquire importance especially for two reasons: (a) the high overall dependence on them of rural households, and especially of the poor and tribal populations, for a wide variety of items essential for

daily use, a dependence which becomes critical during drought and famine;?’ (b) the fact that products from VCs and forests, which are primarily gathered by women and children,?® provide women with an independent source of subsistence unmediated by dependency relationships on men. As noted in chapter 1, women usually have rights to use the VCs by

virtue of their membership (through birth or marriage) in the village community, whereas their access to the cash economy and (in areas of 26 It is suggested by some that women’s reproductive responsibilities, such as their childbear-

ing role, could also weaken their bargaining position in the household (e.g. Anne Marie Goetz, personal communication, Institute of Development Studies, Sussex, 1992). In my view the likely effect of this factor is difficult to judge a priori in South Asia. On the one hand, frequent pregnancies and looking after a large number of young children could affect women’s bargaining power adversely, especially by limiting their employment options; on the other hand childlessness may be equated with ‘barrenness’ which carries a social stigma, and could even be a cause for divorce (as Cain 1988, found in Bangladesh). Again if a woman has a son (or sons) this would increase her bargaining power (e.g. by improving the social legitimacy of her claims), but if she has only daughters this may weaken it. Also

the extent of women’s reproductive responsibilities varies across cultures and much depends on the autonomy a woman can exercise in limiting the number of children she has and her access to childcare support (through relatives or the State). This last aspect could be

subsumed under ‘external support systems’ listed above. Moreover, the extent of a

husband's contribuiion to childcare is itself an item of contestation. 27 See Agarwal (1990a), Pingle (1975), and Banerjee (1988). 28 See Dasgupta (1987b), Brara (1987), and Agarwal (1991).

64 A field of one’s own strong female seclusion) to the market place itself, is constrained and often dependent on the mediation of males.?? Similarly, social support systems of patronage, kinship, caste groupings, and even friendships built up in different ways by male and female family

members can be extremely important means of economic support, especially in tiding individuals and families over social and economic crises.

However, not all the factors noted would carry equal weight in all contexts. In South Asia today, it can be argued that rights in private land hold a privileged position for several reasons. First, the rapid decline in

forests and VCs, especially in semi-arid areas, noted in chapter 1, is effectively eroding this source of supplementary economic support for the

poor in general and for women in particular. Second, erosion is also occurring in social support systems of patronage, kinship, friendship, and caste groupings.>° The decline in kin support is especially apparent among communities and groups which have become poorer over time, making it

increasingly difficult for families to offer such support. The effects are particularly dramatic in tribal communities traditionally characterized by a high degree of communal and intra-gender cooperation in work and social

life.2! The worst affected are women, especially the widowed and the aged.3 Third, access to wage employment and to other income-earning opportunities (such as in non-farm activities) may themselves be linked with access to land. We noted in chapter 1, for instance, that the opportunities for and earnings from non-farm activities tended to be substantially greater among households with some land, relative to the totally landless. Also there is

some evidence to suggest that women from landed households have a greater probability of finding wage employment than women from landless 29 Also see the discussions in chapters 6 and 7. 30 For discussions of the erosion of patron—client relationships, see Breman (1985), Dasgupta

(1987b), and Commander (1983). On declining support from kin, see Fernandes and Menon (1987) and Dreze (1990) for India; and Cain et al. (1979) and Jansen (1983) for Bangladesh. 31 This decline is linked partly to the shift from communal, swidden cultivation to settled individual farming (a shift to which State policies towards land use and forests have contributed in crucial ways), and partly to the growing impoverishment of these tribal populations. (For a case study of the effects and causes of the shift from swidden to settled farming among the Garo tribe in northeast India, see Agarwal 1990b, and chapter 4; and for a graphic and poignant description of the adverse effects, especially on women, of growing impoverishment and associated erosion of intra-community support among the tribal groups in Orissa, see Fernandes and Menon 1987.) The decline in customary systems of patronage, however, is associated particularly with the growth of capitalist farming (Breman 1985, Bhalla 1976, Banerjee 1988). 32 See Dreze (1990) and Fernandes and Menon (1987) for India; and Jansen (1983) and White (1992) for Bangladesh.

Conceptualizing gender relations 65 households.33 Families with some land, relative to the totally landless, would also tend to have a higher reserve price for their labour, which 1s likely to push up aggregate wage rates, as was found to have happened after

redistributive land reform in Kerala in the 1960s and 1970s (Raj and Tharakan 1983); and as is also indicated by Bardhan’s (1984: 54) finding in

his. district-level analysis for West Bengal for 1970-71, of a negative association between the percentage of asset-poor rural households in a district and the average daily wage rate of farm labourers in the district. Effective rights in land thus have the potential for strengthening women’s fall-back position not only directly but also indirectly by improving returns from other income sources. In terms of relative intra-family bargaining power, therefore, effective independent rights in private land could strengthen rural women’s fall-back position in ways that employment alone may not. Recognizing this does not

in any way weaken the case for also enhancing women’s employment opportunities or stemming the erosion of common property resources. Indeed, given the complex hurdles in the path of making women’s land rights a reality (as will be discussed in later chapters), there is a strong case for taking steps in all these directions. At the same time, as noted in chapter

1, land ownership provides more than employment can, including a stronger base for social and political participation, and so for challenging gender inequality on several other fronts. Alongside the factors mentioned above are external systems of potential

support in the form of the State and non-governmental (including women’s) organizations. These can add to a person’s intra-household bargaining power both by direct provision of subsistence, and indirectly by increasing access to some of the other mentioned factors such as employment, assets, etc. For instance, in recent years many NGOs in South Asia

have come to play a noteworthy, even if localized, role in enhancing subsistence possibilities by providing income-earning opportunities, credit,

and infrastructural support, including support for pressing legal claims; influencing the perceived legitimacy of claims (such as via the media); pushing for legislative change and the implementation of laws such as 33 See e.g. Ryan and Ghodake (1980), who in a study of six semi-arid villages in Andhra Pradesh, Karnataka, and Tamil Nadu found that in three of the villages the probability of employment was significantly greater for women from small and medium farm households, relative to landless households; in the other three villages the findings were not statistically significant. The authors do not give reasons for their observation, but do note that women from different socio-economic groups tend to compete for the same jobs. Presumably then,

landholding households have better connections with the (also landed) employers than agricultural labour households. Caste connections may also be a factor. See Lipton (1983) for a review of additional evidence.

66 A field of one’s own | minimum wage laws; supporting (in some cases) the landless or near landless in their struggles for land and higher wages; confronting and contesting biases in State policies and practice; and so on. However, whether these interventions increase or decrease women’s bargaining power

will depend on the gender bias inherent in the programmes and gender differences in the benefits which accrue from them. For instance, organizations which enhance credit and income-earning opportunities for women relative to men, as have the Grameen Bank in Bangladesh and the Self-

Employment Women’s Association and Working Women’s Forum in India (to name but a few), would tend to strengthen women’s bargaining power. But NGO interventions which enhance only men’s access to land,

credit, etc. would tend to reduce women’s intra-household bargaining power. The same would be true of State interventions (in the form of development programmes, policy directives, laws and their implementation, and so on): these can increase or decrease women’s intra-household bargaining power, depending on whether they are gender-progressive or gender-retrogressive. In other words, the mentioned five factors, by impinging on women’s and men’s subsistence opportunities and access to resources from outside the family, would affect their bargaining power and so their access to subsistence within the family as well.

Bargaining for family land. Contestation within the household over subsistence resources, given gender differences in endowments etc., is

only one level of bargaining. The second, more basic, involves women bargaining over the endowments themselves, such as for a greater share in the family land, or the freedom to seek wage work outside the home, and so

on. At this level, the factors determining women’s bargaining power become even more complex. For instance, in South Asia, a daughter’s ability to successfully claim a share in parental property (assuming she 1s not voluntarily given it) is likely to depend especially on the following factors:>4

—the existing inheritance laws (legal legitimacy);

—the woman’s literacy, including legal literacy (that is, her knowledge of her legal rights); —the social legitimacy of her claim, that is, whether the claim is considered a valid one in the community of which the household is a part; —her access to government officials who administer land-related matters; 34 For an empirical elaboration of the relevance of these factors see chapter 6.

Conceptualizing gender relations 67 —her economic and physical access to legal machinery (lawyers, law courts); and —her access to economic and social resources for survival outside the support systems provided by contending claimants such as brothers or kin. In other words, individual women’s struggles to acquire a share in family land would require interlinked struggles outside the household arena as well, such as struggles to legitimize women’s need for independent rights in land and to mobilize economic, social, and political support for the cause. A change in the law to make it more gender-equitable, for instance, would require contestation with the State; establishing social legitimacy for the claim would require contestation with the community; and so on. Gender differences in intra-household bargaining power are thus linked with the person’s extra-household bargaining power with the community

and the State. This would be especially so in contestation over landed property, since control over arable land helps define (and is also defined by) wider access to economic, social, and political power.

Revealing intra-family relative bargaining power: empirical evidence. In a limited sense, relative bargaining power within the family/ household could be revealed in who participates in decision-making and about what. Hence, women who participate in decision-making concerning, say, agricultural production or cash expenditure in the home, may be said to have greater bargaining strength than those who are excluded from such decision-making altogether. But more fundamentally, relative bargaining power is revealed in whose interests prevail in the decisions made, namely in final outcomes: in the intra-family distribution of resources, goods, services, and tasks, the treatment meted out by family members, the control exercised over resources, and so on. As noted, these outcomes cannot always be traced to an observed or observable process of discussion and negotiation, but nonetheless reflect implicit bargaining strengths. Since this approach to the issue of intra-household gender relations is relatively new, empirical evidence is rather limited. What does exist is indicative. First, several recent studies show a positive relationship between

women’s access to independent earnings from employment (or other income-generating activities) and (a) their participation in household decision-making processes;3> and (b) their self-esteem and the treatment they receive from husbands and other family members.*° For instance, 35 See e.g. Acharya and Bennett (1982) for Nepal; Bhatty (1980) for India; and Rahman (1986) for Bangladesh.

36 See Bhatty (1980) for India; Rahman (1986), M. Ahmed (1985), and Chen (1983) for Bangladesh; Shaheed (1988) for Pakistan; and Roldan (1988) for Mexico.

68 A field of one’s own Acharya and Bennett (1982), drawing on detailed data from seven village studies in Nepal, found that women who were largely confined to domestic and subsistence production played a much less significant role in major household economic decisions, than did women who participated actively in the market economy. They note (1982: ix): Women’s involvement in market activities gives them much greater power within the household in terms of their input in all aspects of household decision-making. At the same time, confining women’s work to the domestic and subsistence sectors reduces their power vis-a-vis men in the household. Two explanations are offered for this phenomenon. First, women who participate in the market activities make a measurable contribution to the household income and second, they are more likely

to control their own production assets, while women working in subsistence agriculture are generally laboring on land controlled by the male household head.

In her study of beedi workers in Uttar Pradesh, Bhatty (1980: 41) similarly observes: A greater economic role for women definitely improves their status within the family. A majority of them have more money to spend, and even more importantly, have a greater say in the decisions to spend money. Most women claim to be better

treated as a result of their contribution to household income ... A substantial proportion of women feel that they should have a recognized economic role and an independent source of income .. . Their attitudes evidence a clear perception of the significance of their work to family welfare and their own status within the family.

Also studies from outside South Asia indicate that access to independent

income tends to increase women’s leverage, especially over fertility decisions.?’ That women’s self-esteem and how they are treated by family members

likewise tend to improve, is illustrated by Rahman’s (1986) and M. Ahmed’s (1985) studies of the Grameen Bank in Bangladesh and Chen’s (1983) of the Bangladesh Rural Advancement Committee (BRAC). Both

organizations draw members from very poor households and provide credit to small gender-homogeneous groups. The women loanees in Rahman’s study reported that access to fixed assets through the loans gave them

a special status in the family, and other family members, including husbands, became more conscious about their comfort and well-being. Data collected on intra-household decision-making also showed a notably greater participation of women loanees relative to the women who were not

loanees themselves but were wives of male loanees. Ahmed found a significant reduction in verbal and physical abuse and threats of divorce by husbands after their wives joined the Grameen Bank. Similarly the women

members of BRAC told Chen that they were better respected by their 37 See e.g. Roldan (1988) for Mexico and Blumberg (1991) for Guatemala.

Conceptualizing gender relations 69 husbands and faced less violence in the home after joining the organization.38 Although the sources of women’s income in the studies cited above are mainly employment and non-land-related activities, we would expect the observed correlations to also hold for land-related income.*°

Second, where women have traditionally had rights in land, or have acquired land more recently, this is seen to impinge positively on several aspects of their position. For instance, in South Asian communities which

customarily practised matrilineal or bilateral inheritance, and among which women either continued to reside in their parental homes after marriage, or could return to them in case of marital breakup, women enjoyed a great deal of economic and social security, freedom of movement

and interaction outside the home, and relative equality in marital relations.*° They also often controlled the household food stores; and, as observed among the matrilineal Garos of northeast India, could eat before their husbands did, if the latter were out late,+! in contrast to the pattern in, say, northwest India where women are typically expected to eat after their husbands, and where patrilineal inheritance and patrilocal residence sever-

ely circumscribe women’s autonomy in many ways. Also among the bilateral Sinhalese today, Gunawardena (1989) found that spouses typically make household decisions jointly, other than fertility decisions, which are made by women alone. Similarly, landless women who in recent years have acquired independent rights in plots of land through struggle, report that this has enhanced their sense of economic security and self confidence,

and improved the treatment they receive from their husbands and other family members.*? 38 The emphasis on group formation in these organizations also played an important complementary role in strengthening women’s bargaining position (as will be elaborated in chapter 9). 39 For instance, in the Mwea irrigation-resettlement scheme in Kenya, women’s participation in household decision-making in the scheme villages was distinctly less than in the offscheme ones, a factor clearly linked to the scheme women’s lesser access to land and cash. In the off-scheme villages, women’s own plots were sufficient for providing family subsistence and also a surplus for sale, the cash from which they controlled, while free firewood could be gathered from the local forests. In the scheme villages, by contrast, the plots allotted to women were too small (and of poor soil) to produce enough food for family subsistence, and firewood too had to be purchased, increasing women’s dependence on their husbands for cash generated from the latter’s more substantial irrigated rice fields (Hanger and Moris 1973).

40 For elaboration on these and other aspects of gender relations in matrilineal and bilateral societies, see chapter 3. There were of course differences in degree between these societies in

relation to the noted aspects. 41 Nakane (1967), and personal observation. 42 The Bodhgaya women cited in chapter | reported this graphically; as did the women of Vitner village (in Maharashtra, west India) when they received shares in their husbands’ lands in 1990, as part of a movement for empowering peasant women spearheaded by the Shetkari Sanghatana’s Mahila Aghadi described further in chapters 9 and 10; also see Gala (1990) and Omvedt (1990).

70 A field of one’s own Third, famine-related evidence for South Asia suggests that the oftenobserved abandonment of women and children in poor households during such a calamity can also be explained through the bargaining approach. The assets that are first sold during the crisis, for instance, are typically jewellery, household utensils, and small animals. These are usually the only assets women own, while land, which as a productive asset is retained to the last, is typically in men’s names.+3 Women’s employment opportunities, more limited than men’s even in normal times, also tend to collapse during famine situations, while men are often 1n a better position to migrate out for work. Asa result, women in poor households tend to be left on the one hand with a much more weakened fall-back position relative to men, and on the other hand with a diminished ability to contribute to family income. While a deterioration in the wife’s fall-back position would improve the husband’s

bargaining position within the household, this may provide him little realizable advantage given the simultaneous (and severe) decline in her ability to contribute to joint well-being. An eventual outcome may be that the man (whose outside options have not deteriorated as badly as those of his spouse), chooses to abandon his wife and children and strike out on his own. This is brought out with particular starkness in an analysis of the 1943 Bengal famine (see Agarwal 1990a), but a similar pattern of famine-related

asset disposal and family breakup in some other regions and periods suggests that this process is not atypical.*+ Fourth, women have been known to sometimes explicitly use property and wealth to bargain for a better deal in the family, especially where they are elderly. In a Punjab village (northwest India), Sharma and Dak (1987: 49) found that women ‘had to resort to somewhat coercive techniques to ensure their care and control which included keeping property under lock,

secret purchasing of gold and ornaments and promising favour with valuables to those members serving [them] best’. Control over land and property was similarly found to confer greater authority and respect to the elderly in a three-village study in Uttar Pradesh, which also cited cases of old women using property control to ensure care from relatives (Raj and Prasad 1971). 43 Although retaining land makes good economic sense since it is a productive asset and necessary for the household’s economic security, this does alter men’s and women’s relative economic position within the household.

44 On this point, Dasgupta (1993: 329) misinterprets my paper on drought and famine (Agarwal 1990a) as arguing that it is only the collapse of a woman’s fall-back position relative to her husband’s which leads to her being abandoned in a crisis. As I have stressed here, a woman’s fall-back position diminishes simultaneously with her potential contribution to family income, since factors such as her asset ownership, access to employment, and access to support from kin, affect both her fall-back position and her ability to contribute economically to the family’s well-being. Hence in an extreme crisis, while the sharp decline in the wife’s fall-back position may improve the husband’s bargaining situation, she may have so little to offer that it would still be in his economic interest to leave her.

Conceptualizing gender relations 71 Although most of the available evidence indicates a positive association

between women’s independent access to economic resources and their situation within the family, there 1s also some which is inconclusive and which suggests that when women begin to earn, this may not necessarily translate into more control over the earnings or greater participation in

household decision-making. For instance, Standing’s (1991) study of working women in Calcutta gives a mixed picture on these counts. However, in my view, such findings do not in themselves negate the significance of women’s independent earnings as a factor determining women’s bargaining strength. Rather, we need to examine not just the fact

of earning but also a number of related factors which are likely to be important, such as the period over which such earnings are sustained, the level of earnings,4+> community attitudes and norms about women’s needs and rights (i.e. the social legitimacy of women’s claims) and, most importantly, the process by which an improvement in women’s earnings has been achieved. Although women’s outside earnings could be one factor, among others, which could have an impact on family and community attitudes, a few women acting individually would tend not to have the same potential for challenging gender-conservative biases as would many women from the same community doing so concurrently. And the latter action would not have the same potential for change as would the process of women acting in groups.*® Also, since norms and attitudes do not change overnight, we are more likely to observe improvements in intra-household gender relations in

communities which have had marked increases in women’s economic contributions over an extended period and not just in the recent past. Of course, all this is only by way of pointers. Much more empirical work in this direction is clearly needed.

i. Gender relations outside the household/family: the market, the community, and the State

We have noted that women’s bargaining power within the home ts clearly linked to their situation outside it. Outside the household/family, gender

interactions take place in a variety of arenas, of which three could be especially important: the market, the community and the State. It is suggested here that the bargaining approach can usefully be extended to characterize gender interactions in these arenas as well. To reiterate, the basic idea behind the bargaining approach for charac45 For instance, in her study of women home-based garment makers in Pakistan, Shaheed (1988) found that the level of women’s earnings made a distinct difference to the extent to which their treatment within the home improved. Roldan (1988) in her Mexico case study also notes the importance of women’s earning levels in this regard. 46 For elaboration and examples see chapter 9.

72 A field of one’s own terizing gender interaction is that these interactions simultaneously contain elements of cooperation and conflict. Two parties (either or both of which may be groups) cooperate in so far as cooperation leaves them better-off than non-cooperation. But there is conflict between them with respect to

which among a set of cooperative arrangements 1s arrived at, each arrangement of the set being better for both than a non-cooperative one, but some arrangements in the set being better for each party than others. The outcome is determined by relative bargaining power, which depends partly on each party’s fall-back position and partly on the perceived (legal, social) legitimacy of each party’s claims. Consider now how this approach can be applied to gender relations within the market, the community, and

the State. |

The market. Unlike the theoretical ambiguity that has surrounded the characterization of gender relations within the household (e.g. are they altruistic or dominated by self-interest?), market relations are unambiguously depicted in economic analysis as guided by self-interest. It is in this

arena that bargaining takes its most explicit form and has been widely focused on in the literature, especially in the context of the labour market and trade unions. Given the attention that market-related bargaining has already received, I will not discuss it in detail here. What does need mention,

however, is that (a) women’s ability to bargain in the market, as in other arenas, is critically affected by gender ideology and practices; and (b) owning land would strengthen women’s bargaining power in the market arena as well. The latter point has been discussed earlier, but the former needs elaboration. Consider for instance the labour market. Bargaining may occur over wages, the duration and intensity of work, the working conditions, and so on. But women’s bargaining power in the workplace (in comparison with that of male workers) is likely to be constrained not only by gender gaps in skills and education but also by their domestic responsibi-

lities which reduce their job options; by employers’ assumptions (which may be quite erroneous ones)*’ regarding women’s abilities, work commitment, efficiency, and needs; by cultural specifications of appropriate female behaviour (e.g. norms regarding female seclusion, or the view that public

bargaining or haggling is unfeminine and improper); by barriers to women’s entry into trade unions and the male biases within trade unions;*® 47 The depiction of women as only ‘supplementary’ earners and men as the primary ‘breadwinners’ is often used to justify gender differentials in recruitment and wages (see Barrett’s (1980), survey of the arguments in the context of western Europe, and Kumar (1989) on the specific form that the family wage debate took in India in the early part of this

48 Sec e Hensman (1988) on the experience of women in Indian trade unions.

Conceptualizing gender relations 73 and so on. Many of these factors are also likely to adversely affect women’s functioning in markets for land and agricultural inputs (as will be elaborated in later chapters).

In other words, gender ideology, crystallized in social assumptions, norms and practices, affects bargaining not just within the home space but also the public space. At the same time, social norms themselves can be and

often are contested and bargained over. For rural women, the village community, which also often defines their workspace locationally and socially, assumes particular importance in the construction of and contestation over gender ideology, as discussed below.

The community. A community could be defined in terms of a shared identity based on location (e.g. a village) and/or social grouping (religious, ethnic, caste, clan, and so on). A person can be a member of several communities simultaneously, for instance, of a caste or religious grouping within a village (or spreading across several villages) as well as of

the larger village community containing several castes or religious groupings. Like gender relations within the household, those within a community

can also be characterized as relations of cooperative conflict within a bargaining framework of analysis, although with some important differences from the intra-household context, as discussed later. Consider first the general case of the individual within the community and then the issue of gender. It can be argued that an individual is likely to cooperate with the

community in so far as it brings her/him greater economic, social, or political gain than possible otherwise. Community membership can provide individuals with economic support (jobs, credit, other economic help in a crisis), social support (during marriages, illnesses, deaths, etc.), and political support (say in conflicts with other communities), which are denied to non-members. Hence each individual may be better off and better able to survive economically and socially as a part of the community than outside it. Further, community members can cooperate in specific contexts for mutual benefit, such as by jointly managing a communal resource. What would cooperation with an individual on the part of the community mean? It could be argued that the community would want to retain the

loyalty of its members who, in aggregate, constitute the human and material resources of the community and its political strength. It could therefore seek to hold its individual members by the promotion of support networks, the formulation and enforcement of consensual rules, and so on. At the same time, there can be at least three types of inherent conflict between an individual and the community: (a) over the sharing of economic resources held in common (such as common land); (b) over positions of

74 A field of one’s own political power and decision-making authority; and (c) over community norms which dictate social behaviour. Implicit or explicit bargaining can occur between an individual and the community over the rules governing economic resource use, political positions, and social behaviour, and over the enforcement of those rules. The cooperation of an individual with the community could imply her/his (a) following the established rules; or (b) ‘bargaining’ to change the rules by discussion, protest, etc.49 Non-cooperation would mean opting out of the community altogether.°*° A person could opt out of a local community in a variety of ways, for different reasons, with varying implications. For instance, s/he may physically relocate permanently (e.g. migrate) for economic reasons, because of an inability to negotiate a satisfactory deal over community resources; or s/he may relocate for gaining greater social freedom. Even more drastic would be opting out of a community not just locally but by changing one’s social identity — for instance, changing one’s religion or caste, the latter via a

process that Srinivas (1955-56, 1965) termed ‘sanskritization’. By this process a low-caste Hindu family changes its name, customs, and overall way of life to match those of upper-caste families. Of course a person who opts out of a local community could assimilate into another, but this is not

always easy. In practice, therefore, opting out would not be an option available to all, and for most it may be the option of last resort. In some ways the nature of the inherent cooperative conflicts between an individual and the community is not dissimilar to that within the household. There would, however, be at least two critical differences. One, since the size of a community is larger than that of a household, the costs to the community of an individual member not cooperating would be smaller, and may even be insignificant. Two, the community, unlike the household,

would not necessarily be a unit of joint consumption, production or 49 Non-compliance with community rules could be seen as a form of implicit bargaining. But sanctions for some forms of non-compliance, such as breaking rigid sexual taboos, could be severe, even involving ostracization, in effect exclusion from the community. °° There are some interesting parallels here with Hirschman’s (1970) formulation in his book Exit, Voice and Loyalty, where he argues that individuals have two options for expressing dissatisfaction with an organization (a firm, a political party, a community, etc.) — exit and voice. That is, the person can stop dealing with/opt out of the organization altogether, or give voice to his/her dissatisfaction by protesting to the authorities. However, Hirschman notes that organizations which have a high price associated with the exit option — loss of life-long association, defamation, deprivation of livelihood, and so on (as could happen in relation to a community) — could repress or delay the use of the voice option as well: ‘Obviously, if exit is followed by severe sanctions the very idea of exit is going to be repressed and the threat will not be uttered for fear that the sanction will apply to the threat

as well as to the act itself” (pp. 96-7). |

In terms of my formulation here, voicing dissatisfaction could be seen as a form of contestation or bargaining; and a person’s effectiveness in doing so or ability to pay the price of exit could both be seen to depend on her/his fall-back position (as elaborated later in the chapter).

Conceptualizing gender relations 75 investment, although some or all members may cooperate in specific contexts, including investing in and using for production a communal resource such as land or water.>!

How does gender impinge on this formulation? As an illustration consider a woman belonging to a village community which is more or less homogeneous in terms of caste. She could benefit from caste support in the

ways mentioned earlier, such as receiving loans or other economic and social help during crises, being able to enter into labour-sharing or labourexchange arrangements for domestic or agricultural tasks, receiving help in arranging the marriages of children, being allowed access to a well or a piece of common land possessed by the caste group, and so on. At the same time, there would be underlying conflict with community arrangements over the sharing of communal resources, or in her being subject to caste rules about whom she may marry or have sexual liaisons with, the degree of seclusion she is expected to maintain (where female seclusion is emphasized), and so on. The last would constrain her earning options and could be a particular source of conflict in poverty contexts.*?

However, the ability of an individual woman to ‘bargain’ with the community for a greater share in communal resources, or for a better deal in various other community arrangements, or for greater personal freedom, would be more limited than that of a man, for several reasons. One, women in South Asia tend to be excluded from most traditional public decisionmaking bodies, such as caste councils, which enforce existing rules govern-

ing the community and play a role in modifying those rules. Two, a woman’s typically weaker (than men’s) intra-household bargaining power

would also impinge upon and weaken her extra-household bargaining power, if her husband and marital family oppose her stand. Three, where

patrilocal, inter-village marriages with non-kin are the norm, married women would not have the kin support that men can get within the village. Four, and relatedly, the loss to the community if a woman decides not to cooperate would be small, except where she commands significant econ-

omic or political influence by virtue of her property status or political contacts within or outside the village.*?

In general, women’s ability to bargain for a better deal within the 51 There is a growing theoretical and empirical literature on the question of whether and under what circumstances individuals would tend to cooperate with one another as a group for economic gain: for some interesting South Asian applications see Wade (1988) and Seabright (1993). Wade also provides a useful review of some of the theoretical literature. 52 The nature of these constraints would vary by a woman’s community: upper-caste or nontribal women are typically subject to greater social constraints than lower-caste or tribal women, as are Muslim women relative to non-Muslims. These will be discussed further in chapters 7 and 8. 53 In special cases, of course, even individual resistance or violation of social norms can shake up the community: this appears to be especially so when caste rules governing sexual unions are violated. See chapter 9 for illustrative examples.

76 A field of one’s own community would be greater if they operate as a group than as individuals. For instance, a caste group can penalize an individual woman who breaks

seclusion norms in various ways, including casting aspersions on her character, or shunning her. Such sanctions are much more difficult to apply if a group of women decide to transgress the rules. Similarly, it would be much more difficult for individual women acting alone to gain control over

parts of common land, or to have a greater say in public bodies, or to protest sexual harassment or assault, than it would be for a women’s group or fora woman supported by such a group. In other words, in the context of gender relations within a socially homogeneous community, a woman’s bargaining power for getting a better deal from the community is likely to stem only partly from her individual economic and political position, and

more particularly from gender-specific group cohesion within the community.>**

The same would hold for women in a multi-caste village community, but

with one important difference. Here the bargaining power of a group of women could depend not only on the size of their group but also on their caste and class position within the village community. On certain counts, such as in the sharing of communal resources, low caste or poor peasant women’s ability to negotiate would tend to be weaker than that of high caste or rich peasant women whose caste or class as a whole commands greater power in the village. On other counts, such as negotiating greater social freedom, lower caste/class women may have an advantage.

Non-cooperation in the sense of opting out of the local community altogether may not be an option for many village women; it would depend

on the woman’s fall-back position defined by her ability to survive, economically and socially, outside the local community. Among factors which could impinge on this are the following:

—her personal property position: women with, say, landed property would be less dependent on the community for econ-

omic survival than those without; also personal property positions could be translated into political strengths outside the village community, although less directly so for women than for

, men, as discussed in chapter 1; —the economic and _ social support provided by her household/family; —her skills (including education) and associated economic opportunities independent of the community; and

—material and social support from outside the community and family, such as that from women’s groups, other NGOs, and the >4 For elaboration and illustrative examples see chapter 9.

Conceptualizing gender relations 77 State: this could include provision of earning opportunities, housing etc., and also (say, from women’s groups) emotional and social support. In other words, here a woman’s fall-back position could depend on her direct rights in property, her access to extra-community economic opportunities and social support, her intra-household bargaining strength, and the inter-household political dynamics in the village.

The State. The framework of cooperative conflict and contestation 1s also relevant in characterizing women’s relationship with the State (although again not in the same way as for intra-household relations). This is perhaps most apparent in the relationship of gender-progressive NGOs to the State. For instance, the demands of women’s organizations (and of many NGOs in general) are typically directed both at the State and against

it.55 The State has the power to enact laws and formulate policies and programmes in women’s favour; to allocate financial resources for reducing

gender bias in access to productive resources, employment, information, education, and health; to provide protection from gender violence in the family and community; to counteract the force of doxa by influencing discourse on gender relations in specific directions through the media and the educational institutions; and so on. All these are potential areas of cooperation between the State and women’s groups. However, the same State can also use its resources and coercive apparatus to reinforce existing gender-retrogressive biases within the family and community, constituting a situation of conflict.

It could of course be asked: what would be the State’s interest in responding sympathetically to gender-progressive demands, and why should the State cooperate with gender-progressive groups? Several interlinked factors impinge on this: one could be political pressure built up by such groups, perhaps with the support of oppositional political parties, and/or the media, with implications for voting patterns. Two, international public opinion could matter, as could implicit or explicit pressure from international aid agencies: for instance, White (1992) argues that international aid agencies played a significant role in pressuring the government of Bangladesh to formulate some gender-progressive policies and programmes during the International Decade for Women, despite the counter-

weight of Islamization which the State was also promoting. She notes (1992:13): ‘[A]id-inspired studies and projects aim to “bring women out”’, *5 Also see Daniels (1993) who makes a somewhat similar point when exploring women’s relationship with the State in the United States. However, what she sees as a ‘contradictory’

relationship could, in my view, be seen as consistent with the cooperative-conflict

framework.

78 A field of one’s own to redress their “invisibility”; while the purdah culture ... of Bangladesh holds seclusion to be the highest ideal for women. That Bangladesh accepts this intervention ts indicative of its client status ...’ A third factor could be the State’s recognition of the inefficacy of both market mechanisms and of its own machinery in implementing particular programmes which are seen as essential for development, such as programmes for literacy, health improvement, poverty alleviation, and so on. In India, the State’s attempts since the mid-1980s to enlist the support of NGOs for this purpose (and

especially the support of women’s NGOs for literacy and health programmes), are a reflection of this recognition. In other words, on several issues the interests of the State and of gender-progressive groups could coincide. Cooperation by individuals or groups with the State could take the form of supporting it politically (say, via votes), providing it ideological legitimacy in international and national forums, desisting from ‘disruptive’ activities such as demonstrations, pickets and strikes, and so on. At the same time, the State may only cooperate with NGOs over certain

types of programmes. Typically in non-socialist regimes these are programmes which are welfare-oriented and fit into a ‘basic needs’ approach to

development, such as programmes for the better delivery of health and educational services or for providing income-earning opportunities to poor men and women. The State is less likely to support programmes which call

for a shift in basic economic resources and relations, such as land redistribution. In other words, within the framework of cooperativeconflict there can be a deep divide in the issues over which the State may be willing to cooperate and those over which there would be explicit conflict. A demand for gender-equitable distribution of arable land would tend to fall in the latter category.

Further, the State itself can be seen as an arena of cooperation and conflict which take place in many forms and on multiple levels. For instance, if we distinguish its role in enacting laws and formulating policies from that of implementation, the State may cooperate in passing gender-

progressive laws and policies, but the resistance offered by the local bureaucracy, judiciary, police, or other arms of the State apparatus in the implementation of these measures, or in the allocation of funds, would constitute aspects of conflict. There may also be departments or ministries within the State apparatus which seek to pursue gender-progressive policies and programmes within an overall gender-retrogressive State structure and development framework. Women’s Cells, Departments, or even Ministries, set up in different South Asian countries after 1975 (which marked the beginning of the United Nations Decade for Women), are cases in point. Likewise, there may be gender-progressive individuals within particular departments of the State apparatus — in every South Asian country, it is

Conceptualizing gender relations 79 possible to name individual bureaucrats (male and female) who have played a critical positive role in this respect, typically, but not only, in response to demands by women’s groups.°*°®

On the one hand, therefore, there would be gender-related contestation between elements of the State and non-State organizations, institutions, or individuals; on the other hand the State itself can be seen as an arena of gender contestation between parties with varying understandings of and commitment to reducing (or reinforcing) gender hierarchies. These contes-

tations can be between State officials within a department, between different tiers of the State apparatus (such as policy making and policy implementation bodies), and/or between different regional elements of the State structure (e.g. the bureaucracy in the Indian state of Bihar operates differently in this regard from that of Kerala), and so on. Such a conceptualization implies that the State is not being seen here as a monolithic structure which is inherently, uniformly or trans-historically

‘patriarchal’.*’ Rather it is a differentiated structure through which and within which gender relations get constituted through a process of contestation.5®8 Such a conceptualization does not deny the empirical realities of State-functioning in South Asia as having been, in greater or lesser degree, more gender-retrogressive than gender-progressive. But it does mean that 56 Sanyal (forthcoming) makes a similar point in his work on the relationship between the government and NGOs in South Asia. In his meetings with a number of bureaucrats and state planners, he found that the barrage of criticism against them as people interested only in furthering their own interests was not entirely justified; that a number of ‘these planners also cared about their countries’ well-being, were intensely critical of inefficiencies within the government, and were often very appreciative of PVOs [private/traditional voluntary organizations] and NGOs who had organized the poor, made demands on the government on their behalf, and thus, had facilitated social reform’ (p.23). NGOs, likewise, while complaining about many obstructive social officials also occasionally mentioned ‘good bureaucrats’ who helped NGOs in many ways, even against the recommendations of fellow bureaucrats. Goetz’s (1990) discussion on the functioning of field-level bureaucrats in Bangladesh 1s also revealing. Among other things, she finds interesting differences in the approaches and attitudes of male and female bureaucrats. In village-level credit programmes, for instance, women bureaucrats were much more sympathetic to the constraints faced by village women and (because they were women and therefore excluded from most male networks) were less susceptible than their male colleagues to being coopted by the local male elite.

57 Among those who see the State as an inherently male institution, embodying and protecting male interests, see especially MacKinnon (1989), who argues: “The state is male in the feminist sense: the law sees and treats women the way men see and treat women’ (pp. 161-2). Again: ‘Over and over again, the state protects male power through embodying and ensuring existing male control over women at every level — cushioning, qualifying or de jure appearing to prohibit its excesses when necessary to its normalization. De jure relations stabilize de facto relations’ (p. 167). 58 Here I come close to Connell’s (1987: 130) conceptualization of the State in the context of Western democracies: ‘The patriarchal state can be seen, then, not as the manifestation of a patriarchal essence, but as the center of a reverberating set of power relations and political processes in which patriarchy ts both constructed and contested’.

80 A field of one’s own the State could be and has been in some degree subject to challenge and change in this respect. In this process of contestation, women’s bargaining strength with the State could depend ona complex set of factors. For instance, women would have much greater strength if they were functioning as a group, say as a women’s organization, than as individuals (as also noted in the context of community-level bargaining). Their bargaining power would be greater, the larger the group and the greater its ability to muster support from the media, oppositional political parties, and gender-progressive individuals and groups in the State apparatus. The degree to which judicial institutions within the country can act autonomously of the ruling political party would also impinge on the outcomes of women’s interactions with the State, as would the extent of sensitivity to gender-related concerns prevailing within the country and internationally.

Hl. Interactions: the household/family, the community, and the State So far we have noted how the household/family, the community and the State can be characterized as three principal arenas of contestation.°? Gender relations get constituted within and by each of these. At the same time, the State, the community, and the family are also interacting arenas, embodying pulls and pressures which may, at specific junctures and in different country contexts, either converge (reinforcing each other) or move

in contradictory directions (providing spaces for the building of countervailing resistances). For instance, a State may take a progressive prowomen stance, passing laws, defining policies, and promoting programmes that favour women’s interests, while communities within the country may resist the implementation of these measures: the situation in post-Independence India and some other parts of South Asia, at several points in time, could be so characterized. Or the State, the community, and the family may reinforce each other in strengthening, say, the strictures on women’s social and sexual conduct and dress, as happened especially under Zia-ul-Hugq in Pakistan. Or State policies may be congruent with the dominant interests of

the community but individual families may find that their economic interests are in conflict with the norms set by local communities. Many poor rural households in Bangladesh today are cases in point: here Islamization

drives launched by the State and supported by local communities have

dictated greater female seclusion, but such strictures are now being contested by many poor women (often with the tacit support of their husbands) who find that these norms seriously constrain their attempts at economic survival (see chapter 9 for details). 59 The market is a fourth, which would also interact with these three institutional arenas in various ways, but, as noted earlier, is not our principal focus here.

Conceptualizing gender relations 81 Essentially, the local communities can be seen as playing an intermediate

role between the State and the individual or household, in defining and enforcing people’s obligations and rights in different areas, including appropriate forms of social behaviour, economic activity, and sometimes even dress. At the same time, not all members of a given community need

conform to what is specified by the community’s economically and politically influential members. To the extent that the State as a whole (or significant elements within it) maintains a relatively gender-progressive position in policies, legislation and implementation, it provides space to

individual women or individual households to escape from or openly contest a community’s gender-retrogressive stranglehold. It also provides scope to women for building organized resistance against specific genderretrogressive practices prevailing in the community and/or household. This is especially critical in women’s struggle for a share in arable land, since many men both in their local communities and in their households are likely to have a considerable stake in preserving the status quo. Among these contradictory pulls and pressures (as also emphasized in chapter 1) for women to acquire and exercise control over land will require simultaneous struggles in all the different arenas of household, community,

and State. It will involve contesting the inequities inherent in existing distributions of material resources as well as in gendered ideologies and social practices. The very scale of the struggle that will be necessary constitutes a formidable barrier, but it is precisely this which also gives this struggle its unique potential to transform women’s lives.

In the chapters which follow, the framework of bargaining or contestation

spelt out here will constitute a running thread. For instance, chapters 3 and 4 will help illuminate how despite the considerable bargaining power that women of matrilineal and bilateral communities historically enjoyed in

marital relationships, their limited access to and bargaining power in community and state decision-making bodies left them vulnerable to the erosion of their property rights and sexual freedoms. Again the process of reform in women’s legal rights in property in the twentieth century may be seen as one of contestation between individuals and groups with differing ideologies and interests within and outside the State arena (chapter 5). The

framework of bargaining is similarly useful in understanding (a) the difficulties women face in claiming, retaining, and controlling their legal shares in land, in so far as these are traceable to the weakness of women’s fall-back positions and overall bargaining power (chapters 6 and 7); (b) the regional variations in these difficulties (chapter 8); and (c) the ways in which women resist and contest prevailing gender inequalities in practices affect-

ing their rights in arable land and their social freedoms (chapter 9).

3 Customary rights and associated practices

The natives of Ceylon are more continent with respect to women, than the

other Asiatic nations; and their women are treated with much more attention. A Ceylonese woman almost never experiences the treatment of

a slave, but is looked upon by her husband, more after the European

manner, as a wife and a companion. (Percival 1803: 176) {I]f she is weary of a man, she tells him to go, and he does so, or makes terms with her. Any children they may have stay with the mother who has to bring them up, for they hold them not to be children of any man, even if they bear his likeness, and they do not consider them their children, nor are they heirs to their estates... (Barbosa c.1518, on the Nayars, translated from the Portuguese by Dames 1921: 42)

Prior to colonial rule, the inheritance of property, including land, was governed by local customs in South Asia. These customs varied by region, religion, caste, and sometimes even family, forming a complex mosaic.! But to what extent did they give women inheritance rights in land? Did such rights, where they existed, make for greater equality in gender relations, as suggested by the above quotations on the bilateral Sinhalese and matrilineal Nayars? And were these rights structurally linked to (or conditional upon) certain social practices? For instance, did those communities which recognized women’s rights in land have particular preferences about whom women should marry, where they should live after marriage, whether and whom they could remarry on widowhood, the degree to which they could take decisions about the use and disposal of the land, and so on? It could be

hypothesized that given the critical importance of landed property in agrarian economies, families or communities which recognized women’s land rights would have had marked preferences on these counts, to ensure that the land so inherited would remain within their overall purview. Closekin marriages, for instance, would be conducive to ensuring this, as would a woman’s post-marital residence in her natal home or village. These questions and hypotheses will be addressed in this chapter through ' See Cohn (1965), Mayne (1900, 1953), Derrett (1968), and Lingat (1973). 82

Customary rights and associated practices 83 a historical analysis of communities which indisputably gave women customary rights in land. In particular, an attempt will be made to probe whether there was any systematic association of women’s land rights with

specific marriage practices and the sexual freedom granted to women. Through this analysis I also hope to illuminate why communities and regions in which today there is a significant disjunction between contemporary laws governing women’s inheritance and customary rules of marriage, are likely to be particularly resistant to the de facto realization of women’s legal rights. Such a disjunction could occur, for instance, in cases where current laws framed by the modern State extend inheritance rights in land to women who did not have those rights before, but customary rules of marriage (such as prohibition of village endogamy and close-kin marriage) do not allow the kin group to exercise control over the land so inherited. The three sections below will address these issues.as follows. Section I will

examine the evidence on gender and property (including land) rights in patrilineal communities in the pre-colonial period, as well as identify the regions and communities where matrilineal and bilateral inheritance systems prevailed which customarily recognized women’s rights in land. Section IT will focus in detail on the traditionally matrilineal and bilateral communities, describing the form that women’s land rights historically took in each case and the conditions under which these were granted. And

in section III I will draw upon these case studies to: (1) identify the structural links between women’s land rights and other social (especially marriage) practices; and (2) comment on the nature of gender relations within these communities. In identifying the structural links I will also critically examine and spell out the ways in which I differ from Goody’s

(1973, 1976) propositions on the nature of these links, including his proposition that dowry, by definition, is a form of pre-mortem inheritance. (The appendix on ‘definitions’ defines the different types of post-marital residence that will be mentioned in the discussion below.)

I. Which communities customarily recognized women’s rights in land?

There is undisputed evidence that in at least three regions of South Asia there were significant (although localized) pockets of communities customarily practising matrilineal and/or bilateral inheritance. These are: —Northeast India, the home of three matrilineal tribal communities: the Garos, Khasis, and Lalungs; —South India: here the Nangudi Vellalars of Tamil Nadu state practised bilateral inheritance, and several other groups practised matrilineal inheritance, including the Nayars and Tiyyars

84 A field of one’s own of Kerala, the Mappilas of north Kerala and the Lakshadweep (earlier called Laccadive) Islands, the Bants of south Canara

(now in Karnataka state), and the Phadiyas and Chettis of Wynad district (bordering northern Kerala and Tamil Nadu); and —Sri Lanka: here all the major communities practised bilateral or matrilineal inheritance — the Sinhalese and the Jaffna Tamils were bilateral, and the Muslim ‘Moors’ were matrilineal. Information on the customary practices prevailing among these communities, while not extensive (the best documented are the Garos, Nayars, and Kandyan Sinhalese), is adequate for tracing their inheritance and marriage

customs historically. However, for communities other than these, such information is relatively scarce, especially but not only for the period prior to the 1950s, when a range of ethnographic evidence on local norms and practices began to emerge through village studies. These ethnographies (to be discussed in chapter 6) suggest that in South Asian communities, other than those mentioned above, inheritance practices were essentially patrilineal, and women had few and highly restricted rights in land. But might women’s rights in these patrilineal communities have been greater at some earlier points in time, say before the advent of the British? It is not possible

to probe this question here in depth or detail, but some pointers are attempted below.

For tracing Hindu inheritance and marriage practices in ‘ancient times, reference is often made to the detailed textual information provided by the ancient legal treatises — Dharmashastras — and the many commentaries on

them. Although it is now widely recognized that this classical shastric literature provided prescriptions about appropriate practice rather than descriptions of actual practice, the shastras are a useful reference point since

they did draw upon custom in some degree and in turn shaped custom (Lingat 1973, Derrett 1968). And especially as formalized under the Mitakshara and Dayabhaga legal doctrines (described below) they significantly influenced legal practice in the British period and the formulation of contemporary Hindu law. Dated sometime between 200 BC and AD 300 (Kane 1930), the shastras are said to have been constituted from the smritis, that which the renowned sages (especially Manu, Narada, and Yajnaval* Although the nomenclature ‘Moor’ (given to the Sri Lankan Muslims under Portuguese rule) is no longer commonly used in Sri Lanka and has been subsumed under the general category ‘Muslim’, I have retained the term in my narrative to distinguish the group both from other Muslims in Sri Lanka today who did not traditionally practise matriliny, and from the matrilineal Muslims (such as the Mappilas) of southwest India whose inheritance practices, as will be noted, were different.

Customary rights and associated practices 85 kya)? ‘remembered’ from what was explained to them by the ‘Self-existent’ one. The Manu smriti was the most orthodox, and the Yajnavalkya smriti among the more liberal in granting women rights. Commentaries on each of the smritis followed, some taking the form of digests and legal treatises that gained influence in different regions of the country. The most important of

these were the Mitakshara and Dayabhaga legal doctrines, dated around the twelfth century AD (and referred to as ‘schools’ of law by the British). The Dayabhaga system, based on a digest composed by Jimutavahana (a

Bengali Brahmin), held sway mainly in Bengal and Assam, and the Mitakshara system, based on a commentary on the Yajnavalkya smriti by Vijnaneshvara (from south India), held sway in the rest of the country.

Between the thirteenth and sixteenth centuries, the Mitakshara school branched into four sub-schools which came to be known as the Mithila, Bombay, Madras, and Benaras schools. These sub-schools differed in particulars while being bounded by the general principles of the parent doctrines, which form the basis of contemporary Hindu inheritance laws (although now much modified). What rights were granted to women under the Mitakshara and Dayabhaga systems?* Without going into all the complex details of the two systems, their main features that are of relevance here were as follows.5 The Mitakshara system distinguished between two types of property: joint family property® and separate property. The former consisted principally of ancestral property

(that is, property inherited from the father, paternal grandfather or paternal great-grandfather), plus any property that was jointly acquired or was acquired separately but merged into the joint property.’ A community

of interests and rights was recognized in the joint family property, held + Others often cited are Brihaspati and Katayana. There is a considerable literature on the origins and contents of the Dharmashastras, but see especially Kane (1930). By Kane’s assessment, the Manusmriti was composed between 200 BC and AD 200, and the smritis by Yajnavalkya and Narada between 100 BC and AD 300. + For discussions on women’s inheritance rights in property as enunciated in shastric texts composed prior to the twelfth-century Mitakshara and Dayabhaga texts, see especially Kane (1946), Vishnoi (1987), and Altekar (1956). As the discussions by these authors indicate, although there were some notable differences in the approaches of different smriti writers and subsequent commentators, the inheritance rights granted to women in the earlier texts were not more than (and typically were less than) those recognized under

Mitakshara and Dayabhaga. As noted earlier, Mitakshara is a commentary on the Yagnavalkya smriti which was among the more liberal of the smritis. ° For these features I draw primarily on Kane (1946), supplemented by Mayne (1900, 1953), Altekar (1956), Carroll (1991), and Mulla (1982). © The joint family here is a legal concept and need not coincide with joint residence or any other aspect of a common household economy that may be implied in a sociological use of the term. ” On how the share obtained on partition of joint family property was treated, see discussion further on.

86 A field of one’s own jointly by (a maximum depth of) four generations of male members — a man, his sons, sons’ sons, and sons’ sons’ sons — who became coparceners

on birth. Women could not be coparceners. Devolution was by survivorship: the living coparceners had an interest in the property of deceased ones,

and the individual shares could be determined only at partition; these shares decreased in case of births and increased in case of deaths among the members of the coparcenary. There were severe restrictions on the aliena-

tion of coparcenary property. Even the rights of the Karta (the father or other senior coparcener who served as manager of the property) to gift, sell,

or mortgage any part of the undivided ancestral property were strictly limited to contingent circumstances, the restrictions being especially severe for immovable property; while other coparceners could not alienate any

part of their undivided interests except with the permission of all the coparceners. However, each coparcener had the right to demand partition

unilaterally at any time, while the remaining coparceners could stay undivided.

Over his separate property, by contrast, a man had absolute rights of ownership and disposal. This property included that which was selfacquired (if acquired without detriment to the ancestral estate) and any property inherited from persons other than his father, paternal grandfather, or paternal great-grandfather. In addition, if he had no sons, sons’ sons, or sons’ sons’ sons, his share of the ancestral property obtained on partition was also counted as his separate property. In the presence of these male lineal descendants, however, the partitioned share was still ancestral property, as far as he and any of these descendants were concerned, and his rights of disposal over it stood curtailed (Kane 1946: 576, 639-40). Under Mitakshara, from the joint family property women were only entitled to maintenance as incoming wives (including as widows) and as

unmarried daughters; upon marriage a daughter was also entitled to marriage expenses and associated gifts. In a man’s separate property, however, his widow could inherit a limited estate, but only in the absence of

sons, agnatic grandsons, and agnatic great-grandsons,® and only if she remained chaste. A limited estate (also termed a limited interest) meant that

the woman could enjoy the property for her lifetime, but she could not normally alienate it (such as by gift, sale, or mortgage), except in a period of

severe necessity (later termed ‘legal necessity’), and (within reasonable limits) for performing pious and religious acts, especially those seen as 8 Agnates are individuals of either sex who have descended through the male line and trace descent from a common male ancestor. In other words, they are so related to the deceased

that there is no intervening female link. For instance, a daughter is an agnate but a daughter's child 1s not, while a son’s son and son's daughter are both agnates.

Customary rights and associated practices 87 conferring spiritual benefit on the deceased.’ There appears to be no clear agreement among scholars on the extent to which women’s alienation of property for these purposes needed the permission of the reversioners,!° but there does appear to be consensus that greater freedom was granted to women in relation to expenditures for religious and spiritual purposes than

for worldly purposes.!! A daughter (an unmarried one got preference) came even after the widow, and a daughter’s son after the daughter.!* That is, for a daughter to inherit her father’s estate required the absence both of the noted male heirs and of the widowed mother. And when she did inherit, a daughter, like the widow, could receive only a limited estate, except under

the Bombay (Mayukha) sub-school of Mitakshara which allowed the daughter an absolute estate.'* Kane (1946) and Altekar (1956) note that for several centuries prior to this even these limited rights of the widow and daughter were disputed by orthodox jurists. In the early shastric texts, however, for sonless families the practice of putrikaputra or ‘appointed daughter’ was recognized, interpreted either as

‘the daughter appointed as a son’ or ‘the son of an appointed daughter’ (Kane 1946: 647, 657-9). In either case, the idea (as several commentators argue) was that the daughter would raise a son for her sonless father. She and her husband were expected to reside with her parental family, her son

inheriting her father’s estate and taking his name and so continuing the latter’s line.!* But this system became obsolete over time, as (according to ° The term ‘limited estate’ is legally used only in relation to Hindu women. An analogous term, derived from English law, is ‘life estate’ (or ‘life interest’). Both terms (viz., limited estate and life estate) mean that the woman can enjoy the property only for life, but there are differences in the restrictions on transfers. A Hindu woman’s limited estate did not allow her to alienate the property, except in the specific restricted contexts mentioned above. A life estate meant that the woman could freely transfer the property, viz. lease out, mortgage or sell it, although such transactions were only valid for the duration of her life. (Personal communication, B. Sivaramayya, Professor of Law, Delhi University, 1992.)

'0 Those to whom the property ultimately reverted if the present heir died or relinquished her/ : his rights to it. '! See discussions in Kane (1946: 710), Mayne (1953: 767-70), and Altekar (1956: 264—5). Mayne also mentions that alienation for the benefit of the estate was permitted under Mitakshara, but there are conflicting views on how this was to be interpreted. '2 The daughter’s son, unlike the daughter, received the property as an absolute estate. In the absence of the daughter’s son, the property went to the deceased man’s parents and then to his brothers and their sons. 13 See e.g. Banerjee (1984: 355), first published in the 1890s. Also see Mayne (1953: 647-8), Kane (1946), and Roy (1911: 264, 266). In cases dealt with by the Bombay High Court in the nineteenth century, an attempt was made to ascertain Mayukha law, and it was noted that under that sub-school daughters were not excluded from the patrimony, and sisters rather than paternal relatives were treated as heirs to the brothers, on the understanding that this had long been the general rule in Bombay Presidency. Daughters and sisters inheriting in this way were also said to get an absolute estate. '* Hence the ‘daughter appointed as a son’ appears not to have meant that she inherited the estate like a son, but that her son was equivalent to what a son’s son would have been.

88 A field of one’s own Kane 1946: 714) the ordinary daughter came to be recognized by analogy (after the widow) as the heir of a sonless man. A man could also adopt a son in the absence of one, there being a strong preference for the brother’s son. The Dayabhaga system was different from Mitakshara in some important

respects.!> A man was deemed absolute owner of all his property (no distinction being made between ancestral and self-acquired property) and could dispose of it (that is sell, mortgage, or gift it)'!© as he wished. The son

did not acquire an automatic interest by birth in the father’s ancestral property, nor was there any rule of survivorship: each heir took a definite and nonfluctuating share. Division of property among heirs could take place only at the man’s death, and the property went in the first instance equally to his sons. The share of a predeceased son would devolve on the son’s sons, or failing this on the son’s sons’ sons. A chaste widow could inherit in the absence of these male heirs, but only as a limited interest, with the right to manage but not alienate the property. Daughters came only after the widow, again unmarried daughters getting first preference and inheriting only a limited interest. However, in contrast to Mitakshara law, Dayabhaga recognized the widow and (after her) daughters as heirs even when the man’s ancestral estate had not been separated before his death.

Hence, unlike under Mitakshara, women inherited an interest in all property, irrespective of whether it was ancestral or separate.!’ This also meant that the probability of a widow or daughter inheriting some property was somewhat greater under Dayabhaga than Mitakshara.'® 'S See especially Kane (1946) and Mayne (1953). '6 Later, under the British, testamentary disposition was also expressly recognized. 17 The undivided ancestral estate being referred to here is an estate that may be held jointly, say by a man and his brothers. Under Dayabhaga, the father and sons were not coparceners in a joint estate (as they were under Mitakshara), but sons who inherited on the father’s death could hold their inherited property jointly as coparceners, each holding a clearly defined share. Each such coparcener had full rights of disposal over his share, and his interest in which, while still undivided, could pass on his death to his own heirs, male or female.

'8 It needs mention here that succession under the Mitakshara and Dayabhaga systems was based on two different principles. Under Mitakshara it was propinquity (that is, nearness of blood relationship to the deceased) and under Dayabhaga it was religious efficacy or spiritual benefit conferred on the deceased. The principle of propinquity meant that among the admitted heirs, those nearer in terms of the blood relationship would inherit before the more distant ones. The principle of religious efficacy meant that a person conferring more spiritual benefit on the deceased was preferred as an heir over one conferring less spiritual benefit. Under both systems sons were deemed to occupy an overarchingly superior position over all other heirs. The notion of sapinda underlay both principles but was differently defined. Under Mitakshara, sapinda was defined as a relationship established by ‘shared body particles’. Under Dayabhaga, the sapinda relationship was linked to the obligation to offer a ball of cooked rice (pinda) to the deceased, and to the deceased’s father and paternal grandfather during the ritual feasts of the dead called sraddha.

Sapinda also governed permissible marriage partners and those subject to death

Customary rights and associated practices 89 Under both systems, there was also some recognition of female property rights in the concept of stridhan (literally meaning a woman’s property), although there were varied and changing interpretations of what stridhan could include, how much control a woman could be allowed over it, and how it would devolve on the woman’s death. The discussions around this are too complex and contentious to consider here in detail, and I will seek only to outline some of the overall conclusions arrived at by a number of scholars.'9 Broadly, it appears that in the very early shastric texts, stridhan could consist only of movables (such as ornaments, clothes, and household utensils) given to a woman by her parents, brothers, or relatives before or at

the time of her marriage and by her husband after marriage. Over this property she was allowed absolute control, and it devolved on her female heirs in the first instance. In some later texts, especially from the seventh century AD onwards, there was a tendency to enlarge the scope of stridhan in terms of the content and source of the gifts. This also led to considerable controversy among the commentators on whether landed property should be included in stridhan and what control women should be allowed over it. In particular, allowing women full control over any land received from the husband, for instance, would have meant that his lineage risked losing the land if the wife sold or mortgaged it or if it was inherited by her heirs. Similar considerations arose in the case of land inherited from a father. Vijnaneshvara, the proponent of Mitakshara law, proposed the most extensive additions to stridhan, including in it any property (movables and immovables) that the woman received, whether by inheritance, purchase, partition, or chance. He was, however, silent on the question of a woman’s power of disposal over any property included in her stridhan which was

acquired by inheritance or partition; and there was a good deal of speculation as to what he may have intended. Altekar (1956: 226—7) persuasively argues that Vijnaneshvara could not really have intended to pollution. Sapinda-exogamy prohibited marriage with all near relatives by birth, which under Mitukshara meant that the prospective bride and groom could have no common ancestors within seven generations when tracing the relationship through their fathers and

within five generations when tracing the relationship through their mothers. Death impurity was seen as occurring within the prescribed degrees and marriage was permitted outside them. In addition, marriage was subject to the gotrarules of exogamy. Two persons of the same gotra were prohibited from marrying. Under the Davabhaga school, sapindaexogamy was less restrictive than under Mitakshara: two persons could marry even within the prohibited degrees provided three gotras intervened between the bride and the common

ancestor. Gotras are exogamous patrilineal clans whose members are thought to share patrilineal descent from an eponymous ancestor who is deemed to be a primeval seer. (For further discussion on the above issues see especially Trautmann 1981; Kane 1946; and Mayne 1953.) '9 For details see especially Kane (1946), Banerjee (1896, reprinted 1984), Mayne (1953), Altekar (1956) and Tambiah (1973), all of whom examine the diversity of views on these aspects and the shifts in these views over time.

90 A field of one’s own invest the widow with the absolute right to dispose of landed property included in her stridhan but acquired by inheritance or partition, since he was not prepared to concede full powers of alienating such property even to the male manager of the joint family property: Could [Vijnaneshvara] have ever dreamt of investing women with a right, which he was not prepared to grant even to the male manager? His silence on the point may be simply due to the fact that he tacitly accepted the general principle that women are limited heirs, a principle which was approved even by Brihaspati, the most well known advocate of women’s rights. (Altekar 1956: 226-7)

Mayne (1953: 728-9) arrives at a similar conclusion, but on the basis of a different argument: [Vijnaneshvara] did not intend... to include in stridhana the property inherited by a woman as heir to her husband or to her son. The very rules of stridhana succession which he lays down postulate as a condition the legal possibility of the acquirer’s

male issue or her husband succeeding to her property on her death. For, in the absence of the daughter and the daughter’s children, her son and son’s son are to take it and in their default, her husband. But obviously there can be no conceivable possibility of her male issue or her husband taking on her death the property which a woman inherits on her husband’s death only in default of male issue.

It is difficult to say how the matter was resolved in practice in the medieval period. Later, under the British, the Privy Council rulings took

the view that a woman could hold only a limited interest in property inherited from a male, and after her death 1t would pass not to her stridhan heirs but to the heirs of the male from whom she inherited it (see Mayne

1953: 728-30). The Bombay sub-school again appears to have been an exception: it held that property which a woman inherited from a male of the family in which she was born (e.g. as a daughter succeeding her father), or inherited from a female, became her stridhan and her absolute estate (Kane 1946: 783).

Under Dayabhaga the issue was resolved differently. In this system, by definition stridhan was that over which women had full rights of disposal. It included gifts received from her parents, relations and even non-relatives before the nuptial fire or on the bridal procession, as well as movable gifts received from her husband after marriage. All these she could gift, sell, or

use independently of her husband. By this definition, stridhan did not include any property that a woman had inherited (whether from males or females) or obtained by partition of her deceased husband’s joint estate, or gifts made by non-relatives after her marriage, or the earnings from her own labour.?° Effectively (if not explicitly), therefore, stridhan property allowed under Dayabhaga was limited essentially to movables. 20 See Banerjee (1984: 307, 314), and Mayne (1953: 729-30).

Customary rights and associated practices 9] The issue of how stridhan was to devolve is again highly complex and will not be detailed here: it depended especially on who had given the gifts and

on what occasion, and differed between Dayabhaga and the different Mitakshara sub-schools. But, in general, gifts given by a woman’s parents and relations during the marriage festivities went firstly to the female line, namely to her daughters (unmarried daughters getting preference). Other

categories of stridhan (depending on their source and context) passed variously to her parents, brothers, and children of both sexes.?! According to both Mitakshara and Dayabhaga, therefore, Hindu women could inherit immovable property such as land only under highly restrictive circumstances and (barring a few exceptions linked to the Bombay subschool) at best could enjoy a limited interest in it. In contrast, men enjoyed a primary right to inherit and control immovable property; and although

under Mitakshara they too faced certain restrictions in their power of disposal over joint family property, these restrictions related to their rights

as individuals (as vs. group rights), but not to their rights as a gender. Women were restricted by virtue of their gender. To what extent did actual practice conform with the above-noted prescrip-

tions of the shastras in pre-colonial India? An in-depth probing of this question, including tracing changes over time, would constitute a vast historical project which cannot be undertaken here. But some broad surmises can be attempted from the more readily available (albeit fragmentary and diverse) material, such as studies of inscriptions from the medieval

period, historical ethnographies, the inferences of legal scholars, and compilations regarding customary practices by the British. Such compilations are also few. Although over time the colonial State recognized the importance of local custom in its framing and implementation of inheritance and marriage laws, little was done in most regions to systematically collect detailed information on customary practice in its many variations. 21 For details on the devolution of different categories of stridhan see especially Tambiah (1973), Kane (1946), and Banerjee (1984). Kane (1946) also notes that the inclusion of sons as heirs along with daughters (or even in preference to daughters) in certain categories of stridhan represented a shift from the primacy given to daughters earlier, and probably related to the growing value of stridhan property which women alone could no longer be allowed to inherit. 22 There were a few regional exceptions. For instance, in Punjab, an elaborate questionnaire was developed for systematically collecting information on local customs (riwaj-i-am) relating to a variety of practices: inheritance and adoption, the use of common land, the rights of proprietors, etc. The information was gathered by British revenue officers engaged in settlement operations in various districts; and based on this, C.L. Tupper (a prominent British administrator) and W.H. Rattigan (a distinguished member of the Punjab bar) prepared compilations of customary practices in different parts of Punjab. Rattigan’s Digest of Customary Law in the Punjab, first published in 1880, also incorporated judicial rulings by the British courts, updated over the years in various editions of the Digest.

92 A field of one’s own As will be elaborated in chapter 5, as and when deemed necessary, the British drew upon the local (upper caste) elite and village elders to serve as

informants and interpreters regarding local custom, thus introducing an elite and upper caste (often Brahminical) bias in recording and interpretation and a tendency to homogenize existing diversity. Even in the rare cases when a systematic recording of customs was attempted, as in the Punjab, the recordings were still based on information provided by village leaders and were not entirely free from such biases.?3 Additionally, there

appears to be a male respondent bias: for instance, in Punjab, when enquiring about the widow’s right in some contexts to seek partition of her husband’s share 1n his joint family estate, the British found that in some districts informants denied the existence of this custom even when such partitions commonly took place in practice (Rattigan 1953: 316).

Despite their limitations, however, the various sources mentioned do provide important pointers regarding women’s property rights in the precolonial period. We know, for instance, that in several parts of India, and among a number of communities, local customs relating to marriage and property devolution, at variance with the shastras, prevailed in practice. For a start, some communities fell outside the purview of the shastras altogether: the texts did not address themselves to non-Hindus, and turned

a blind eye to the customs of the hill tribes and of the matrilineal communities of southwest India. But even for patrilineal Hindus, as Lingat (1973) argues, the law promulgated by the shastras differed from custom both in object and by origin. The intention of the shastras was to prescribe

duties and obligations of a religious character for acquiring spiritual benefits. Customs could either be in accord with the precepts of the shastras

or in conflict with them: ‘Custom is a social phenomenon, while [the shastric] law has a transcendent character’ (Lingat 1973: 177). At the same

time, the shastras were themselves influenced by custom, although the borrowings from custom were selective and in keeping with the smriti writers’ notions of morality.2+ The interpreters, Lingat argues, were similarly influenced by custom, choosing some and rejecting or amending

others: ‘We are confronted by works which doubtless owe much to 73 Originally, in the Punjab, records of customs are said to have been collected by the settlement officers from each village within a district, but after 1873 the officers called together local notables from various parts of a district to report on their customs (Gilmartin 1981). This would also have had some homogenizing effect in the recording of

24 Lingat (1973) gives several illustrative examples. For instance, he notes that Manu was hostile to certain types of marriages, such as widow remarriage or marriage between uppercaste men and Sudra women (Sudra being the lowest category of the four main categories (varnas) into which Hindu society is traditionally divided — viz. Brahmins, Kshatriyas, Vaishyas, and Sudras). However, confronted with deeply rooted customs that would have made prohibition ineffective, he limited their acceptability to specific contexts, emphasizing their overall undesirability.

Customary rights and associated practices 93 customs, but of which the least one can say is that they are compositions’ (Lingat 1973: 183). As noted, different ‘schools’ of jurisprudence thus emerged, although each was based on and drew its authority from the same totality of texts. Also, it appears that the king, as the traditional custodian of law, could censor custom and decide whether and to what degree his kingdom was to be governed by orthodox Hindu principles (Derrett 1968). In practice, all this meant that customary practices and shastric prescriptions would have converged in some regions and communities and diverged in others. But even where they diverged, shastric influence was not absent:

‘We must take account of the attraction which the dharmic rule has exercised upon custom by virtue of its religious significance and the veneration always attaching to the sacred books which supply it’ (Lingat 1973: 203). The prescriptions of the shastras would have been obeyed most closely by the higher castes, less by the lower castes, and not at all by the

tribal communities. Regionally, tension between the shastras and local custom on marriage, divorce, and inheritance practices is argued to have been most severe in the Deccan and the south and least 1n the north and east (Derrett 1968). Yet even in the latter regions, Derrett (1968: 221) argues,

deviations from shastric rules were tolerated: ‘[T]he hard core of convenience stood out against theory, and to this day some ancient customary elements have succeeded in defying shastric pronouncements — even those which were never compromised by dilution and customary material.’

In marriage practices the regional and community-wise divergences between custom and shastric prescriptions were clearly considerable: for instance, the prevalence of cross-cousin marriage?> throughout south India and parts of central India was seen by most commentators as a divergence,

although some south Indian commentators attempted to show that this practice did not contradict the shastras.?° Similarly, divorce, and divorcee and widow remarriage, were widely practised among the lower caste and

tribal communities in India, although the shastras typically prohibited them.?7 In terms of women’s customary rights 1n land, the divergence appears to

have been less sharp, although still apparent. For instance, there is some evidence that women’s rights were greater than we might expect from the 25 For a definition of cross-cousins, classificatory cross-cousins, and parallel cousins see the appendix on ‘definitions’. 76 See discussions in Trautmann (1981) and Lingat (1973). >” Under exceptional conditions, such as non-consummation of the marriage, some early shastric texts allowed widow remarriage (Kane 1941: 611). There was also an ancient practice — niyoga — whereby a widow could bear a male heir through an appointed male (usually the brother-in-law) if her husband had died sonless. But this was not recognized as remarriage in that no further cohabitation with the man was allowed (Kane 1941: 600). According to Altekar (1956: 146-8) niyoga was fairly common up to c.300 BC, but in later periods it came to be opposed increasingly by the orthodoxy, and went into disrepute over time for various reasons, including a growing emphasis on the widow leading an ascetic life.

94 A field of one’s own shastric prescriptions in parts of India, notably the south and west. In south

India, inscriptions (temple inscriptions in particular) found at several | different locations, dating especially from the tenth to the seventeenth century, record donations of land and other wealth by women, mainly but not only as widows.?° In Andhra Pradesh, Talbot (1991: 321-3) found that 14.8 per cent of the 391 individual donations to major and minor temples between AD 1175—1435 were by women, and 23 per cent of these gifts from women were of land. In Tamil Nadu as well, women were important donors to temples in the early medieval period (Mukund 1992). Although, as may be expected, a large percentage of women donors came from royal families,

a significant percentage did not: many (50 per cent of the donors in the Andhra Pradesh temple inscriptions studied by Talbot) belonged to the households of peasant leaders, warrior chiefs, herders, and merchants. Temple dancers (devadasis) also made important contributions of land and wealth (Mukund 1992). Indeed, devadasis in parts of south India were often endowed with considerable landed property: although much of this consisted of temple land over which the dancers only had hereditary use rights by virtue of their services to the temple, some land also appears to have been owned by them outright, possibly purchased or obtained as a gift from a rich patron.?? In the inscriptions cited by Altekar (1956) which record widows gifting

land to temples in south India, some clearly imply or state that the permission of the reversioners had to be obtained for making the gift, suggesting that these widows only enjoyed a limited interest in the property; but other inscriptions give no indication that such permission was needed.

For instance, a twelfth-century inscription in Mysore (now Karnataka) records a donation to a temple by a widow along with her brother-in-law (the next reversioner) and her caste group, suggesting that not only the reversioner but also the caste group had to approve the transaction. Againa thirteenth-century inscription in Madura district describes how the gift ofa garden to a temple by two childless widows was made possible only when some reversioners gave permission, even while others objected. In contrast

are several other inscriptions from south India, dating to the twelfth, 28 See Altekar (1956) for south India as a whole; Talbot (1991) for Andhra Pradesh; Mukund (1992) for Tamil Nadu; and Prasad (1988) for Karnataka. Temple inscriptions record endowments made to temples, and are typically found on the stone walls of the main shrine or on stone slabs and pillars on the temple grounds (Talbot 1991). Inscriptions are also found on copper plates, especially those recording land grants to individuals (such as to Brahmins), and sometimes to temples or other institutions (R.S. Sharma 1980). 29 Also see Srinivasan (1988), who describes how in Tamil Nadu, a campaign calling for a ban of the profession (which was labelled as ‘prostitution’), spearheaded under the social reform movement by the educated Indian elite from the late nineteenth century onwards, led to a decline both in the wealth and prestige of the devadasis in that state; and in 1947 the profession was legally banned there.

Customary rights and associated practices 95 fifteenth, and seventeenth centuries, which record rich Brahmin widows gifting land, and even an entire village, to the temple: these inscriptions give no indication that the consent of the reversioners was necessary for making the gift. Altekar suggests that these variations reflect differences in prevail-

ing customs among different castes and localities of south India. Also according to a twelfth-century inscription in Tanjore district (Tamil Nadu), the Chola king Rajadhiraja II decreed that a lawful wife should inherit her husband’s property including land, cattle, slaves, jewels, and other valuables (Kumar 1985, and Altekar 1956). In other words, a number of south Indian women of wealthy families in the medieval period were endowed with landed property, whether obtained through inheritance or as gifts (and in most cases we don’t know which), and many were able to alienate it for religious purposes. In west India, again, there are indications of a more liberal approach to women’s property rights. For instance, in Bombay Presidency, as noted earlier, even in formal law (under the Bombay sub-school of Mitakshara) the rights of daughters in sonless families were not restricted to a limited estate. And inscriptional evidence suggests this is also likely to have been the practice: for instance, a thirteenth-century inscription refers toa woman

selling some land she had so inherited from her father (Altekar 1956: 238-9). Of course, from the inscriptions it is not possible to say how widespread these practices were, that is, what proportion of women possessed landed property, nor to what extent women could use the property as they wanted. Donations to temples constitute a special category of wealth alienation and tell us little about women’s freedom to use that wealth in other ways. The ‘merit’ of donations for religious purposes is usually seen to extend to other family members — indeed the noted donations were often made for the spiritual benefit of specific relatives (usually husbands, sons, or brothers

(Mukund 1992)). Temple donations were also not out-of-keeping with shastric prescriptions. As noted, Mitakshara law itself allowed women to restrictedly alienate some part of the deceased’s estate for pious and religious acts, especially where it was seen as benefiting the deceased in spiritual terms. In other words, temple donations by women had a social (and legal) acceptance that would not necessarily have extended to the use

of that wealth for solely personal gain.*° That propertied women, more frequently than men of wealthy families, donated their wealth to temples (as

found by Talbot 1991), could again mean that women were typically not °° Even in the Punjab, under customary laws, a widow who otherwise held only a limited interest in the husband’s estate had the power to gift a small part of the estate for pious or religious purposes that would bring spiritual benefit to her dead husband. Such gifts were binding on the reversioners (Rattigan 1953: 790; Rustomji 1942: 254).

96 A field of one’s own free to use that wealth in other ways. At the same time, the facts that women

donors in the south although not numerous were by no means a rarity or confined to royalty, and that in many cases the permission of reversioners does not appear to have been needed for making the donations, do suggest that women in south India customarily had greater rights of alienation (even if for specific purposes) in the estates they inherited or were gifted, than the shastras were deemed to allow. That women’s property rights customarily exceeded shastric prescriptions in southern and western India is also argued by Mayne (1900) and Derrett (1968), based on the situation observed in the nineteenth century, and is indicated by the evidence presented by Roy (1911). Mayne (1900: 41)

notes, for instance, that a sister who was excluded in the Benaras and Bengal sub-schools of Mitakshara ranked high in the order of succession in the Bombay Presidency and comments: [I]t seems probable that the doctrine, which prevails in other districts, that women are incapable of inheriting, without a special text, has never been received at all in Western India. Women inherit there, not by reason, but in defiance, of the rules

which regulate their admission elsewhere. In their case, written law has never superseded immemorial custom.

Mayne insists that such variations cannot be seen as the operation of different schools of law, since the basic principles of the Mitakshara subschools were the same; rather they reflect variations of local customs.

Evidence from studies which touch on other regions is extremely fragmentary. Prasad (1988), for instance, describes some medieval inscrip-

tions (from the twelfth and thirteenth centuries) which record queens | granting entire villages to temples in parts of eastern India (Orissa) and central India (Madhya Pradesh), but this was usually with the king’s consent, and grants given by royalty tell us little about the land rights of women in the general population.*! Again (in so far as this might reflect an older, accepted tradition) some inferences can be drawn from late nine-

teenth-century cases of Hindu widows from wealthy backgrounds in Bengal inheriting zamindari estates? from their husbands, as a limited interest in the absence of sons, and managing these estates through male agents while themselves remaining in purdah (Borthwick 1984). This practice would be in keeping with the noted better recognition accorded toa widow’s claims under Dayabhaga law (prevalent in Bengal) compared to Mitakshara \aw, and was therefore probably not just a late nineteenthcentury phenomenon but stemmed from an earlier period. All said, though, 31 Also see Sircar’s (1983) compilation of select inscriptions for additional examples of land grants by queens from Himachal Pradesh and Uttar Pradesh. 32 Estates belonging to zamindars or landholders, who collected revenue on behalf of the government.

Customary rights and associated practices 97 these examples do not imply that the average woman’s rights were substantial, although the flexibility of custom allowed communities to take account of situational factors under which women might be given certain rights denied to them asa rule. For instance, in the Punjab (again in so far as customs recorded in the second half of the nineteenth century are indicative

of earlier practices) a Hindu widow who could prove that she was not getting adequate maintenance from her husband’s kin could seek partition of the husband’s share in his joint family estate. Normally she could.enjoy only a limited interest in this property, but in certain restricted circum-

stances she could alienate it.>> The custom of a brotherless woman inheriting her father’s estate, to be then passed on to her son, with her husband residing uxorilocally (that is with her and her parents), also appears to have been practised among some communities in India,** a practice similar to the ‘appointed daughter’ arrangement which, as noted,

was recognized in the early shastric texts but later became obsolete. (However, as will be argued further in chapter 6, this was essentially an

arrangement for sonless fathers to obtain a male heir (the grandson) through the daughter and not a recognition of the daughter’s independent right to the patrimony.) Among a number of patrilineal tribal communities in eastern and northeastern India, again, women as daughters, wives, and widows enjoyed only usufruct rights in land; where this land was inherited it was usually only in terms of a limited interest.?> When we weigh this (rather sparse) evidence as a whole, it suggests that among patrilineal Hindu communities in pre-colonial India, especially in the south and west, a number of women, of elite backgrounds in particular, did possess landed property. But we must be cautious against reading too much into this. Even in the relatively more favourable contexts of south and west India, there is little to suggest that the average Hindu woman was commonly endowed with immovable property; while the position of elite women in this regard appears to have been far from one of equality with men, either in their inheritance rights or in their freedom to use the property as they wished. As noted, temple donations, one of our strongest pieces of

evidence that women from non-royal backgrounds could possess and alienate land, fall into a special category of religious and pious acts which

enjoyed social and legal sanction; from this we cannot infer with any confidence that the women who made these donations were as a rule equally

free to dispose of their wealth in other ways, even if an occasional very wealthy woman was able to do so. As regards matrilineal or bilateral inheritance practices, to my knowl33 Rattigan (1953); also see the discussion in chapter 5. 34 See Altekar (1956) and Rattigan (1953). 35 See e.g. Archer (1984) on the customary practices of the Santal tribe in eastern India.

98 A field of one’s own edge there is no solid evidence to suggest that these ever existed in north India or in any regions far from where they were recorded in the British

period, although in the proximity of some of these regions (such as in southwest India) such practices were probably more widespread than these accounts indicate. For instance, on the basis of documents describing the

customs of Kerala, Gough (1973) suggests that matriliny probably embraced much larger sections of the population in that region sometime prior to and up to the early sixteenth century, than noted in the subsequent periods.

Muslim women’s inheritance rights in South Asia also showed a marked divergence historically between scriptural texts and custom. But unlike the Hindu texts the Koran gave women significant inheritance rights, including in land, although these rights were still unequal to men’s. Daughters and widows, for instance, could inherit in the presence of sons and were entitled to an absolute estate, although the daughter’s share was half that of a son,

and a widow could only receive one-fourth of her husband’s estate if childless or one-eighth if there was a child or a son’s descendants. A widower in the same circumstances was entitled to half and one-fourth of his wife’s estate.*° Customs deviated from these textual rules, among the Muslims, in two quite different ways. In parts of southwest India and in Sri Lanka’s Eastern province, a few Muslim communities practised matrilineal inheritance (as detailed in section II below), under which as a rule it was women who inherited landed property to which men had use rights. In the rest of the subcontinent, many Muslims appear to have followed customary

practices of inheritance similar to those of the local patrilineal Hindu communities among whom women’s rights were highly restricted. For instance, the evidence compiled by Tupper (1881) and Rattigan (1953) on customary practices for a number of districts and communities in the Punjab, indicates that among the Muslims (like the Hindus) in that region, widows could only inherit the husband’s ancestral landed property in the absence of sons, and only as a life interest. Daughters typically came after the widow, and sometimes even after the father’s male collaterals, although among a number of communities daughters received preference over the latter; and there is a rare mention of an unmarried daughter in some regions

being allowed a usufruct share (to be managed by her brothers) till she married. Similarly, the Khojas of northwest India, the Cutchi Memons, the Sunni Bohras of Gujarat, the Moleslam Girasias of Broach, and the Halai

Memons of Porbander (Gujarat) are all said to have followed local customs, like those followed by the Hindus, which were at considerable °° Fora fuller description of devolution rules under Muslim law, see chapter 5.

Customary rights and associated practices 99 variance with the Shariat.37 It is only occasionally that affluent Muslims appear to have endowed daughters (typically in sonless households) with landed property, either by recognizing their rights in the patrimony under the Shariat,** or via gifts, or by creating a wagf (endowment) (Kozlowski 1989).3° There were also cases of talented Muslim courtesans, dancing girls,

and singers in north India being given substantial gifts by male patrons in the form of land and houses; or of movable wealth which some invested in this way (Kozlowski 1989; Oldenberg 1991): this wealth was often inherited by daughters and granddaughters (Oldenberg 1991). And some Mughal rulers gave land grants to widows or to very young girls who had no other means of livelihood to protect them from destitution; in some cases these grants were inheritable (Bilgrami 1988). On the whole, therefore, it does appear accurate to infer that customarily in much of South Asia, inheritance and control rights in land rested largely (but not entirely) with men: women of wealthy patrilineal households in

some regions sometimes possessed and transacted in landed property, although in most such cases the extent of this property and its sources are unclear, and women’s degree of control over it typically appears to have

been limited. There were, however, some significant although locally confined pockets of matrilineal and bilateral inheritance in southwest and northeast India and Sri Lanka, where women’s property rights were not the exception but the rule (see map 3.1). It is to these communities that we now turn.

37 See e.g. Patel (1979), Sivaramayya (1973), Kaul (1990), Tyabji (1968), Derrett (1968), Roy (1911), and Hashmi (n.d.). 38 The Shariat or Muslim religious law is derived from the Koran, the sunna (practice of the

Prophet), ijjma (consensus of opinion among the learned of the community) and givas (analogical deduction) (see Tyabji 1968: 1). There are two principal schools of Islamic law prevalent in South Asia: the Hanafi school governing Sunni Muslims and the Ithna Ashari Shiite School governing Shia Muslims. For a further discussion on these schools of law see chapter S.

39 A wagf often provided the economic base for mosques, schools, and various Muslim religious institutions. It also gave individuals more choice than allowed by the rules of the Shariat in selecting heirs for their property. For instance, if a man had only daughters, by Koranic dictates (noted earlier) kin other than daughters took what would have been the sons’ portions. The Shariat also limited testamentary powers. Turning one’s property into a wagf and directing that the endowment’s income be given only to daughters precluded the possibility of having a part of the wealth divided up among distant kin. Also the founder of the endowment could name a daughter as custodian, leaving her in effective control of an undivided estate. Of course where a man had sons a wagf could also be used to completely cut off daughters, and families sometimes did do so. Nevertheless evidence relating to the early nineteenth century indicates that in a significant number of cases these endowments benefited daughters, occasionally to the exclusion of sons who were wastrels or incapable of managing property (Kozlowski 1989: 128-9).

r rs

: é|; 100 A field of one’s own

|

i|| wim ): )%asta nee )oP op ‘& ; tt Eeata‘|i Zw A) ! ; wd; i,4g3j ! oy [~~] Patrilineal :

:,§ +

:

“Sp Matrilineal ee ey }

|

Map 3.1 Traditional forms of land inheritance

II. Women’s land rights in traditionally matrilineal and bilateral communities

What was the nature of rights in land enjoyed by women in the customarily matrilineal and bilateral communities of South Asia? Were there certain social practices which defined the conditions under which these rights might prevail? What form did gender relations take within these groups?*° The case studies below outline the traditional inheritance, marriage and

livelihood patterns in the main matrilineal and bilateral communities of South Asia. Although these communities were limited in number and geographic spread, an examination of their customary practices is important for at least three reasons. First, they can help reveal the structural 40 My interest here is essentially to examine the gender implications of these inheritance systems. I therefore do not plan to delve into the contentious and speculative question of their origins.

Customary rights and associated practices 101 conditions associated with women’s traditional land rights, and so illumi-

nate why, in the absence of these conditions, women are likely to face particularly strong resistance in gaining shares in landed property, despite contemporary legal changes in their favour. As will be argued in chapter 6,

the divergence in many regions of South Asia between contemporary inheritance laws and local practices governing marriage and residence makes for a considerable tension, leading either to women forfeiting their claims or to social and legal conflict. Second, these case studies provide the basis for subsequently tracing (in chapter 4) the historical process by which many of women’s customary rights have been eroded, especially the role played by the colonial and post-colonial State in this regard. And third,

these case studies can throw some light on the possible links between women’s rights in land and other important aspects of gender relations. In other words, the descriptions of these communities (which may at times seem overly detailed) provide the essential empirical base for addressing the analytical questions posed above, including examining the validity of a set of propositions to be presented in section III of this chapter. A general comment about the information sources used here to recon-

struct ‘traditional’ practices is relevant, although additional details are given as each community 1s discussed. For most communities the earliest systematic accounts of inheritance, marriage and livelihood systems date around the nineteenth and early twentieth centuries. Even for groups such as the Sinhalese and Nayars on whom there are some descriptions that date several centuries prior to this, most of the detailed accounts stem from the

nineteenth century onwards. And for a few groups the only available sources are recent ethnographies undertaken since the 1950s. However, these accounts suffice for our present purpose, although a scholar particularly interested in the history of one or more of these communities would no doubt wish to probe further.

(1) Northeast India: The Garos, Khasis, and Lalungs

The Garos. The Garos, said to be of Tibeto-Burman origin, inhabit the Garo Hills in what are today the state of Meghalaya in India and the Mymemsingh district of Bangladesh.*+! In 1901 they numbered 0.16 41 This section draws especially from Playfair (1909), who provides the earliest systematic

account of the tribe, supplemented by Hunter (1879) and Allen ef al. (1906). For reconstructing some aspects of the traditional situation which these authors do not touch upon at all or deal with only in passing, I have drawn on the descriptions of that earlier period in the studies by Majumdar (1978) and Kar (1982). These authors, through their long association with the Garos, know the people and language intimately, and have had the advantage of talking to Garos who lived in that period. For a more detailed discussion on traditional Garo practices, see Agarwal (1990b).

102 A field of one’s own million, two-thirds of whom lived in the Garo Hills at altitudes ranging from 1,000 to 3,000 feet in the midst of thick semi-tropical forests. The population consisted of a number of exogamous matrilineal descent groups practising matrilineal inheritance and matrilocal residence.

Subsisting entirely on shifting cultivation (jhum) and gathered forest produce, the Garos at the turn of the century lived in widely dispersed villages, each with well-defined boundaries, located in valleys or hillside depressions close to running water. The settlements (a’kings) consisted of jhum and homestead land, owned and controlled communally by one dominant clan composed of the descendants of a common ancestress. All clan members (including the village headman, the nokma) had equal usufructuary rights to the jhum land but none had individual proprietary rights. The area to be cleared for jhum was identified and plots apportioned to each household in accordance with the number of its members. However land was plentiful and a family was free to open up more within the a’king

boundaries. Homestead plots reverted to the village community if the household shifted residence within the village or moved to another village. In other words, use rights in land were directly tied to residence, and land was not private property or individually transferable. However, a system of ‘preferential use privilege’ enabled households by custom, although not by right, to usually return to the same plot every time

an earlier-jhummed area was taken up for cultivation. Similarly trees planted by individuals belonged to them. Land improvements undertaken by a household thus created a ‘realm of preferential or mutually exclusive privilege of using a particular plot, or a plant, or a tree on the land’ (Kar 1982: 236).

Although land was not individually owned, the title of the a’king was vested in one household (usually the oldest a’king-founding household in the settlement), and was held in custody by the a’king-nokma — the husband of the inheritress of the founding household — on behalf of his wife and her

clan. He was the recognized custodian of the land but had no authority to gift or otherwise transfer any part of the a’king on his own without the consent of his wife and her matrilineal kin group (the mahari). All decisions regarding any land transfer or the granting of use rights in a ‘king land were made communally by the members of the founding lineage. The mahari, and especially its elderly male members, enjoyed considerable authority in this regard. Lineages other than the founding lineage had use rights, but not possessory rights to the a’king land. All property (movables and ancestral house) other than land passed from

mother to a chosen daughter, usually the youngest. Parents without daughters could adopt one from the wife’s nearest female relative, such as her sister. Marriages within the mother’s clan were forbidden. The heiress’s

Customary rights and associated practices 103 husband (the nokrom) was chosen from her father’s clan, and was ideally the father’s sister’s son, who moved in to live with his wife in her parental

house. Non-inheriting daughters too continued to live in the mother’s village after marriage, but in separate houses constructed with their family’s help. They were also allotted some jhum land for their use. Sons, likewise, left their parents’ house after marriage. For most men this meant moving to a different village (since most villages contained women of mainly one clan) and giving up the use rights to jhum land in their natal village which they had enjoyed prior to marriage. Only the trees they had planted continued to be theirs after marriage.

No man could inherit property under any circumstance. The selfacquired property of an unmarried man belonged to his mother and her female descendants, and that of a married man to his wife and her female descendants. In case of divorce, the husband could depart with his clothing but little else. However, a husband had rights to use and manage his wife’s land. Hence, although none individually owned land, men (as husbands and matrilineal kinsmen) enjoyed considerable rights in land management and control, and jural authority was vested in men alone.

Marriages were governed strictly by the principle of clan exogamy. Proposals always originated from the woman and her household. It ts of note that for the heiress daughter, the father took direct responsibility for finding an appropriate groom, preferably her father’s sister’s son. If there was more than one suitable suitor, the girl could indicate her preference but could not decline all suggestions without jeopardizing her heirship. In the absence of an actual matrilateral cross-cousin, a classificatory one was selected from the father’s clan, and if possible from the father’s immediate lineage and village. Non-heiress daughters, however, could choose anyone outside their mother’s clan. The bride’s and groom’s families were equal in status and no dowry or brideprice was exchanged. A particularly popular method of proposing to the groom was bridegroom capture: the girl identified her choice to the boys of her clan, who forcibly brought the prospective groom to her house at night and kept a close watch to prevent his escape. It was customary for the boy to express

reluctance even if the proposal was to his liking, and to escape and be captured thrice before accepting. If he rejected the proposal altogether, the girl had to choose another and the process was repeated. Burling (1963: 834), who witnessed this in the 1950s, describes it graphically: I was sitting in Rengsanggri one afternoon when three shy-looking youths from another village wandered in and inquired where they might find Unon. Everybody

chuckled, and somebody replied that he might be out in the fields ... The boys walked out in the direction of the fields ... Here the boys split up, so as to close in on him from all sides. Unon did not realize his peril until one boy was almost next to

104 A field of one’s own him. He started to flee but was caught, and after a brief struggle he recognized the uneven odds, surrendered, and let himself be led calmly to Waramgri, where a gir! was waiting, hoping to become his bride ... I had just witnessed one of the most exciting events in the life of every Garo man, the bridegroom capture, which 1s considered the only decent way to invite a man to become a husband.

Occasional exceptions to this general pattern of thrice-proposal twicerefusal did occur, of course, sometimes with tragi-comic results.4 At the turn of the century, jhum was virtually the only form of cultivation

practised by the majority of Garos, with a ten- to thirty-year rotation. A two-year agricultural cycle was maintained: a mixture of crops was grown in the first year and rice in the second, after which a new area was cleared. Cultivation was by simple implements such as the hoe. A wide range of jungle products supplemented the crops grown. Family labour was supplemented by a variety of reciprocal labour exchange arrangements between households. In general the Garos were self-sufficient in food and also able to produce a surplus for sale. Trading was done at periodic markets held in the foothills. Women played a major role in crop production and the gathering of forest produce. Their labour input in jhum was greater than that of the men;

and their knowledge of indigenous crop varieties was extensive.*3 They controlled the household’s food surplus, and any cash they earned from the

sale of small items in the village or the weekly market was considered exclusively theirs. They were subject to the same rules of sexual behaviour as the men: pre-marital sex was tolerated but adultery was punishable for both sexes.*+

The Garo community at the turn of the century thus had considerable equality in class and gender terms, although there were certain spheres of authority (e.g. jural) which men alone enjoyed. 42 Playfair (1909) describes one such case of a boy who petitioned the District Court in Tura town claiming compensation from the father of a girl for failing to give him his daughter in marriage. The complaint was that he had run away on first capture as was the custom. But nobody came to seek him again, and the girl married another man who was less firm on Garo etiquette! 43 As noted in chapter 1, some women knew of over 300 indigenous varieties of rice (Burling 1963).

44 An especially striking feature of traditional Garo society was the finely drawn code of conduct towards women, with a detailed listing of punishments for transgressions. For instance, a man assaulting a woman while she was lying down or forcing her to lie down, grasping her hand tightly or touching her breast while she was lying down and unable to defend herself, or entering her house uninvited at night, were all considered acts of violence

for which she could demand compensation. Unwanted sexual attention in the form of words, winking, whistling, hissing, caressing, suggestive letters, songs or poems, or any other acts a woman found objectionable, were also punishable by fines for the shame or embarrassment they had caused her, if she appealed to the clan elders (Playfair 1909). Hence what has taken the women’s movement in westernized urban cultures so much time to name as ‘sexual harassment’ was already recognized for what it was by the tribal Garos!

Customary rights and associated practices 105 The Khasis. Like the Garos, the Khasis are a matrilineal tribe dwelling in the hills of northeast India, occupying the Khasi-Jaintia Hills of

present-day Meghalaya.*> Unlike the Garos, in the early part of this century each major Khasi clan, composed of the descendants of a common ancestress, was usually scattered across several villages. Each village thus had segments of several clans.*° Also unlike the Garos, the Khasis had a complex system of local administration that extended beyond individual clans. Settlements were formed into Khasi ‘states’ through the voluntary association of groups of villages headed by hereditary or elected (usually male) chiefs, who exercised judicial and executive authority within multitiered councils.*’ The lowest order of clan segmentation was the iing, consisting of ‘a set of

strictly matrilineal descendants, who have a common right over the ancestral property, are subject to a common authority and practise joint ritual’ (Nakane 1967: 120). The central core of the iing consisted of two persons — the mother in whom the ownership of ancestral property vested, and her brother (usually the eldest) who managed the property for her and represented jing authority. A husband was excluded from the wife’s jing

although he was part of the common household. Sons (whether or not married) continued to be members of their mother’s ting, as did the heiress daughter. Non-heiress daughters branched off after marriage and initiated new iings, while also continuing as members of their mother’s iing. The children of such daughters thus belonged to the new iings of their mothers. Customarily, there were two main classes of land in the Khasi-Jaintia

Hills, Ri Kynti and Ri Raid: the former was that over which the clan members, either jointly or separately, held absolute proprietary rights 45 This section draws especially from Gurdon (1907) and Cantlie (1934), supplemented by Hunter (1879), Allen e¢ al. (1906), Dalton (1872), Nakane (1967), and personal conversations with Khasi academics and senior Khasi government officials on my visit to the Khasi Hills in 1989, Cantlie is a valuable source of information on traditional inheritance laws. For a more detailed account of the community than presented here see Agarwal (1990b).

*° In broad terms the Khasis are comprised of four main tribal groups — the Khynriams, Pnars, Bhois, and Wars, with the Lynngams (said to be part Khasi and part Garo) on the periphery. These groups are geographically concentrated in different parts of the Khasi and Jaintia Hills, with differing ecological conditions and cultivation practices. They have also been subject to varying forms and degrees of outside influences. In terms of inheritance and marriage practices the customs described here are those that were being followed by the Khynriams and Pnars at the turn of the century and which are said to typify traditional Khasi practices. The other groups appear to have assimilated non-matrilineal customs: among the Wars, for instance, Gurdon (1907) found that both sons and daughters were inheriting from the mother, although the youngest daughter was still the principal heir. +7 In Khasi ‘states’ with hereditary chiefs, there was provision in theory for a woman to succeed to office in the absence of specified categories of men. In practice, the chance of this happening was small, given the long list of men who preceded her. Only in one ‘state’ was the spiritual head a woman — a High Priestess (Gurdon 1907).

106 A field of one’s own which were heritable and transferable. Ri Raidland belonged to all the clans

comprising a village community. A clan’s landed property could be apportioned by it to its individual branches and by each branch to its constituent families. Division into separate portions held by families was more common than joint possession by the clan. The family’s ancestral property — land, the ancestral house, and movables — passed from the

mother to the youngest daughter (the heiress), who was essentially its custodian. Non-heiress daughters had usufructuary rights in ancestral land, a part of the produce of which was earmarked for the maintenance of the ancestral home. Typically, as with the Garos, a new house was built for each nonheiress daughter when she married, and a piece of family land was given to

her on usufruct, while the heiress and unmarried siblings continued to reside in the ancestral home. A non-heiress daughter could, if she so wished, lease out her share of the land with the consent of the heiress and leading male family members. Family members had the right of first refusal to such leases.

However, the mother could also, in consultation with her brothers and maternal uncles, divide ancestral property among all the daughters, usually when they married, the largest share going to the heiress who took it as a custodian. Sometimes, in addition to this, a separate share was given to the heiress for discharging her special duties. The divided shares of non-heiress

daughters were entirely separate and outside the control of the heiress. Where the family had insufficient land, only the youngest inherited. In the absence of daughters, a girl could be adopted from the nearest maternal descent group. Alternately, a woman’s separated share of ancestral property, if she had no daughters and female descendants, could be held by her son, but only as a life interest, after which it would pass back to his mother’s clan. Women’s self-acquired property, however, in the absence of

daughters, could be inherited by a son, eventually going to his wife and children. An heiress’s brother, married or otherwise, had maintenance rights in his

parental home. His earnings before marriage belonged to his parental family. After marriage, he could, with the clan’s permission, enjoy usufruc-

tuary rights to the clan land on payment of a rent. His self-acquired property, acquired while living with his wife and children, belonged to them

and not to his maternal clan; nor was it his to dispose of if he left his children. This property, along with his wife’s inherited and self-acquired

property, personal possessions, and gifts, formed the nucleus of the ancestral property of that branch of the clan. However, Cantlie (1934) argues, ifa man acquired property through a profession or trade outside the place where his wife or parents lived, he could gift it to anyone he wished in

Customary rights and associated practices 107 his lifetime. In any case, men had considerable control over the management of property as brothers and maternal uncles. The formal managerial authority over a woman’s ancestral property lay with the eldest brother, but since women played significant roles in the household’s economy and in property division, it is not unlikely that they also had a say in the actual

management of the estate. :

Strict clan exogamy and preferred village endogamy governed marital choices. Marriage within the village was possible because clusters of several clan groups resided within a single village. In contrast to the Garos, there was an aversion to cross-cousin marriage. Most Khasis practised matrilocal post-marital residence. The husband of the heiress moved in with her in her ancestral home, while the heiress’ sisters set up separate residences with

their husbands in the vicinity. But, some groups (such as the Pnars) practised duolocality — the husband maintaining a visiting relationship and working in the wife’s fields while continuing to live in and eat at his mother’s

house. This was clearly more feasible with in-village marriages (the common pattern) than with inter-village marriages.*® Women and men enjoyed considerable sexual freedom before marriage, but post-marital sexual norms were stricter for women than men. Adultery by a woman was subject to a heavy fine, could lead to divorce, and could

deprive an heiress of her rights to ancestral property. Men’s adulterous affairs were tolerated; at worst they could lead to divorce (Nakane 1967). Although divorce was common among the Khasis ‘for any sufficient cause and often without any assignable reason except mutual dislike or want of issue’ (Hunter 1879: 217), it was more frequent among heiresses who also usually initiated it. Non-heiress daughters were usually more dependent on their husbands since they could expect only limited help from

their maternal jing, and their marriages were more stable. Their stronger fall-back position clearly made heiresses less tolerant of the husbands’ authority or shortcomings. The Khasis traditionally practised both settled (including irrigated) and shifting agriculture, but they were not self-sufficient in foodgrains and imported rice from outside, even at the turn of the century. However, their jhum cultivation was on more or less privatized land, unlike the communal character of land held and cultivated under jhum by the Garos. Unfortunately there is little information on the exact gender division of labour in settled agriculture among the Khasis, but women’s role in jhum cultivation is likely to have been important as was the case under jhum elsewhere. Jungle products (gathered by women) and fishing supplemented cultivation. Trading was common and women were the primary traders. 48 Nakane (1967: 123) suggests that duolocal residence may well have been the pattern at some point in time, even among groups no longer practising it.

108 A field of one’s own The Lalungs. On this third matrilineal group, the Lalungs, there is

little historical information. An understanding of their customary practices, therefore, has to be gleaned from a 1970s ethnography by Syamchaudhuri and Das (1973). This means of course that departures from older traditions cannot be identified precisely. Located in the hills and plains of Assam state, traditionally the Lalungs appear to have been primarily jhum cultivators 1n the hills, although today they mainly practise settled agriculture. Women played a significant role in agricultural production. Inheri-

tance was matrilineal and property passed from mother to whichever daughter the mother chose to reside with. This was usually but not always the youngest. The heiress had no right to alienate ancestral land without the approval of her mother and the borjela — the eldest male in the family,

usually the mother’s eldest brother. Cultivation was managed by the woman’s husband, who resided matrilocally. However, he had no right of disposal over the land or its produce, nor any claim to his wife’s property. Also, it was the woman and not her husband who was the recognized head of the household. Non-heiress daughters had usufructuary shares in the land, as did sons as long as they resided in the mother’s home. Traditionally cross-cousin marriages are said to have been accepted, although today they are prohibited (attributable to the influence of local Hindu communities among whom such marriages are forbidden).*? Village endogamy was preferred. Sexual freedom before marriage was tolerated if the liaison was conducted discreetly, but it was not socially approved. Marriage was by

choice between the individuals concerned, although the initiative was usually taken by the girl.

In comparing these three tribal matrilineal communities, there are some noteworthy differences as well as similarities. The Garos were the most egalitarian in terms of land access; all clan members (irrespective of rank or

sex) had equal use rights to land but land was not individually heritable. Among the Khasis, land access was much more privatized and unequal —

families held heritable and transferable rights to the land under their occupancy and use. This also meant that when the family had insufficient land, non-heiress daughters and sons did not get any usufructuary shares.

Hence unlike the Garos, the Khasis had in-built tendencies in their inheritance structure towards intra-family and inter-family economic inequality. Although the Khasis too had certain categories of land, such as

undivided clan land and village community land, on which rights of possession were linked to use and were not heritable or transferable, these lands constituted a part and not the whole of the community’s land. Among the Lalungs, again, land access appears to have been much more privatized 49 Personal communication, D.N. Majumdar, Professor of Anthropology, Guwahati, 1989.

Customary rights and associated practices 109 than among the Garos, and more so than is usually characteristic of jhum cultivation, although in the absence of historical data it is difficult to assess how widespread privatization had been traditionally. What was common to all three communities was that women’s inheritance of land did not give them rights of free disposal or sale.

Four additional features of these groups merit emphasis at this point. One, there was a clear link between land rights and post-marital residence

for both women and men. Two, especially among the Garos, group ownership of land was linked with individual use rights based on residence. This could be seen as an interesting alternative institutional arrangement to individual ownership and use, and one which needs more consideration in discussions on land reform today (see chapter 10 for elaboration). Three,

customary rules vested women with significant rights in land but formal managerial control over the land was vested in men, although in practice women in these tribal matrilineal communities appear to have played a larger role in land management than they seem to have done in some of the matrilineal communities of southwest India (discussed below). Four, jural authority — the power to make rules and enforce them — was typically vested

solely in men, a feature we shall find repeated in the other case studies of

matrilineal communities in South Asia. These last two features have particular implications for our assessment of gender relations under matriliny (as discussed in section III of this chapter). Let us now consider matriliny and bilaterality in south India. Unlike the relatively isolated northeastern communities described above, many of the matrilineal communities in the south (mostly concentrated in and around Kerala) have been widely written about. The Nayars, in particular, have fascinated travellers and scholars for several centuries. Also unlike the northeastern tribes, most of the matrilineal communities in south India were an integral part of the Hindu caste system, and operated within an entrenched class hierarchy, as discussed below.

(2) South India: The Nayars, Tiyyars, Bants, Mappilas, Nangudi Vellalars, and others

The Nayars of central Kerala. In the pre-colonial period, the Nayars of central Kerala were a matrilineal community living alongside a number of patrilineal groups.5° They were divided into several subcastes 5° As outlined in Gough (1961 a: 305), central Kerala includes what in British times was south Malabar district and Cochin state, north Kerala includes the northern part of Malabar and

the southern part of south Canara district. and south Kerala covers Travancore. There is considerable ethnographic documentation on the traditional customs and practices of the Nayars of south Malabar and Cochin, but relatively little on the Nayars of north Kerala

110 A field of one’s own

according to the specialized functions they performed, the important subcastes being those of district chiefs, village heads, retainers to the royal

lineage or to the Nambudiri Brahmins, and the palanquin bearers to the royal lineage. All these subcastes, according to Gough (1952), provided soldiers to the Raja’s army in war, and the account here relates largely to their inheritance and marriage practices.*!

The system of landholding in the region was complex, involving a hierarchical and varied categorization of land rights. Landlord families (calling themselves janmis) held almost all the land in the village, some of which they retained to be cultivated by serfs, while the rest was given out to

high caste tenants on hereditary tenure, who in turn sublet to (typically) lower caste tenants, often leading to several layers of subletting.** Who the landlords were, differed between villages: in some villages the land was held

by a Nambudiri patrilineal joint family, in others by a branch of a royal lineage, in yet others by a large temple or by the lineage of a feudal chief of royal rank or by the matrilineage of a hereditary Nayar village headman originally appointed by the king, and so on. Some Nayars were thus village and even less on those of south Kerala. North Kerala Nayars will be discussed later in the chapter. On south Kerala, existing information suggests that the inheritance and marriage patterns in central and south Travancore were different from those of north Travancore. The practices in north Travancore were quite similar to those of central Kerala and will not be discussed separately. Also, the paucity of information on traditional practices in most other parts of Travancore does not permit an adequate reconstruction of such practices in that region. (Fuller's (1976) reconstruction of traditional practices is also based largely on material relating to central Kerala, although his fieldwork, undertaken in the early 1970s, was in central Travancore.) For my narrative on the central Kerala Nayars, I will draw especially on Gough (196la), Mencher (1965, and personal conversations), Buchanan (1807), lyer (1912), Thurston and Rangachari (1909) and Logan (1887, reprinted 1951). Some of the discussion in Fuller (1976) is also useful. The picture presented here relates to the dominant pattern, some deviations from which no doubt existed in practice. >! In addition there were a number of lower subcastes of washermen, potmakers, funeral chiefs, etc. about whom not much appears to have been written. $2 Basically, the landlord family enjoyed significant rights and privileges in the land, including the right to a fixed share of the produce, but not including the right to sell the land itself, except with the consent of the king or of a chief, and then only to selected caste groups: nor could tenants, village servants, or serfs be evicted without such consent. In this sense the rights held were not those of ‘ownership’ as understood in the modern western usage of the term (Gough 196la: 314; Logan 1951: 602-8). Hence, as Logan (1951: 602-3) argues, although land transfers of hereditary property appear to have taken place in the pre-British period, what was transferred was not the land itself but certain rights and privileges, and certain types of authority over the people located on the land. The British mistakenly assumed that the landlords held full property rights in the land itself, and recognized them as the sole owners (also see, Baden-Powell 1892: 166, on this). It needs mention here that the terms ‘landlord’ and ‘tenant’ are functional translations which do not adequately convey the meanings of the original terms, over which there is considerable debate; on this, and for additional details about the land tenure system in Malabar, see e.g., Baden-Powell (1892), Logan (1951), and Varghese (1970).

Customary rights and associated practices 111 landlords but most were non-cultivating tenants holding long-term hereditary rights in land. Although there was variation in pattern across central Kerala, typically the Nayar estates (consisting of gardens and paddy lands) were sublet to Tiyyars or cultivated by Cheruman or Pulaya serfs. More prosperous Nayar lineages also sometimes leased out land to less prosperous Nayars, who in turn further leased it out. The Nayar matrilineal joint families, termed taravads, were comprised of the matrilineal descendants of a common ancestress, and usually contained a set of sisters and brothers, their mother (if alive) and the sisters’ children and sisters’ daughters’ children. They normally shared a common residence (the taravad house) and enjoyed property collectively.>3 Some taravads were constituted of more than one homestead. A pre-puberty rite — the za/i-tying ceremony — ritually married the Nayar

girl to a suitable man of the same or higher caste, the man having no necessary further role in her life.>+ Among the royalty or other aristocratic

lineages and district chiefs, the tali was usually tied by a Nambudiri Brahmin. Subsequently, a woman could enter into sexual unions or sambandhams with one or more men belonging to her own or higher subcaste or caste. Among Nayars who were military retainers, the majority of such unions were between persons of the same caste. Bilateral cross-

cousin liaisons were freely permitted and those with matrilateral crosscousins preferred. These unions did not imply co-residence — the man continued to live in his natal home, visiting his wife at night and leaving at dawn. The children belonged to their mother’s taravad. A woman could have more than one liaison ongoing, and the man she was with for the night would leave his arms outside the door to indicate his presence to the others. Women were usually free to reject particular men, and the relationships could be informally terminated by either party: ‘.. . if she is weary of a man, she tells him to go, and he does so, or makes terms

with her’ (Barbosa 1921: 42). Indeed the Nayar marriage system represented a degree of sexual freedom rarely granted to non-tribal Hindu women in India. It was this feature that most fascinated observers, some of whom expressed shock, others approval (including the sixteenth-century Portuguese traveller, Duarte Barbosa): “The Nayre women of good birth

are very independent and dispose of themselves as they please with °3 The word taravad is also used to mean a matrilineal clan, consisting of all those who trace matrilineal descent from a common ancestress. More commonly, and as used here, it refers to the matrilineal joint family whose members held property incommon and usually shared a common residence; the house they lived in was termed a taravad as well (Gough 196la). ** Barbosa (1921: 41) writing in 1518 describes a ‘ali’ as ‘a small jewel, which would contain a half ducat of gold, long like a ribbon with a hole through the middle which comes out on the other side, strung on a thread of white silk’.

112 A field of one’s own Bramenes, and Nayres ...’ (Barbosa 1921: 40). Both female and male education were emphasized, and some Nayar women became famous as poets in the eighteenth century (Iyer 1938). Ancestral property was inherited through the female line, in accordance with what was termed the Marumakkatayam system of matrilineal inheritance. All taravad members had equal claims to maintenance from the joint

property, but no individual member could ask for a share by partition, which could only be brought about with the consent of all adult members. Management of taravad land was in the hands of the seniormost male, the karanavan, while many of the younger men were away serving in the armies of the kings or chieftans. The self-acquired property of the karanavan passed to the taravad on his

death, and that of the female members passed to their children and descendants, solely in the female line. Occasionally, a new taravad branch (tavazhi) was formed.** For instance, if the membership of a taravad grew

too large, a new branch was established as a separate property-owning group, through the partitioning of the ancestral property of the parent taravad, if all the adult members agreed. Or segments of the family, settled in outlying ¢aravad lands, might over time separate from the parent group and form an independent branch. Independent land gifts to Nayar women by sambandham partners could also constitute the property bases of such branches. *°

Typically Nayars did not themselves cultivate the land. And even in the

less-prosperous households women’s role in cultivation was minimal. Occasionally some may have assisted in supervisory capacities (Thurston

and Rangachari 1909). The karanavan usually allocated a share of the estate’s paddy crop and other consumption items to the seniormost woman of the taravad who kept the keys to the storehouses; where there was more than one homestead, separate allocations were made to each homestead. Gough (1961a) notes that among retainer Nayars, effective daily management of the taravad’s activities was divided between the karanavan and the oldest woman: ‘the former carried out all formal transactions outside the

group, controlled the estate in toto, and disciplined men and boys. The latter organized feminine tasks and held more informal authority over women and small children’ (Gough 1961a: 341-2). Where the karanavar 55 For discussions on the various circumstances under which this could happen, see especially Moore (1983) and Mencher (1965). 5© See e.g. Mencher (1965: 169) and Moore (1983). It is not apparent how common this might have been. Buchanan (1807: 411), for instance, observed when discussing south Malabar that the gifts bestowed on women by their samzbandham partners were ‘never of such value,

as to give room for supposing that the women bestow their favours from mercenary motives’.

Customary rights and associated practices 113 was the son or a younger brother of the taravad’s seniormost woman, she had some say in estate management as well (e.g. counselling the kKaranavan in his transactions with outsiders), but this was less likely where he was her older brother or uncle. There also appear to have been rare cases of women managing the taravad estate and even enjoying some degree of political power (see e.g. Moore 1983), but typically it is the karanavan who appears to have commanded paramount authority and held basic responsibility for the upkeep of the taravad and for its public dealings. He oversaw various

transactions with the Tiyyar sub-tenants and serfs, dealt with market functions, represented the taravad in the subcaste assembly and in interactions with higher authorities in the kingdom, managed the taravad’s ritual affairs, and was legally responsible for the junior members of the taravad.>’

Parallel to the Nayar system of matrilineal kinship was that of the patrilineal Nambudiri Brahmins who also lived in large joint families (illams) similar to the taravads but comprised of the patrilineal descendants of a common ancestor. They were typically landlords, and in the villages

where they were dominant their principal tenants were usually Nayar lineages holding hereditary tenurial rights. Among the Nambudiris, the eldest brother alone could marry and only within his caste; the younger ones established sambandham unions with

women of high-ranking matrilineal castes, especially the Nayars, thus limiting Nambudiri numbers and keeping the joint estates intact. This meant, however, that the majority of Nambudiri women remained unmarried and were kept under strict seclusion to prevent illegitimate unions. Some nevertheless dared to defy these strictures and seek out lovers (see chapter 9 for details). The Nayars of north Kerala. Nayars in north Kerala, as in central Kerala, followed matrilineal inheritance practices, but there were some important differences between the two regions in marriage and residence 57 In her doctoral dissertation, Arunima (1992) has argued that in the pre-colonial period the powers of the karanavan were much more limited and women enjoyed considerably more authority in ¢aravad affairs, than is usually recognized in the literature. She argues that the legal authority held by the Aaranavan over taravad members, as described by Gough

(196la) and others, was essentially vested in him by the British. Gough (196la: 340), however, has argued that ‘[t]he karanavan’s day-to-day authority over his juniors was conferred upon him by his feudal lord, backed by the judicial authority of the king’. It is also not clear whether the authority which Arunima argues women enjoyed in the precolonial period was authority over the domestic affairs of the household (which Gough (196la) and others also speak about), or authority also in estate management and public affairs. To establish that in the pre-colonial period Nayar women wielded authority in these public domains as a common occurrence, and not just as an exceptional case (as was noted above), much more supportive evidence (than Arunima provides) would be needed.

114 A field of one’s own customs.°® Although the chiefs and village headmen controlled considerable land in the north as well, many commoner property-owning Nayars of

north Kerala were independently small landlords in their own right. However, there was less subletting of land in this region, compared with central Kerala. Some families sublet a part of their land to Tiyyars, but many managed their own fields, employing Irava or other lower caste labourers. Marriages with matrilateral cross-cousins were preferred although those

with patrilateral cross-cousins were also allowed. The marriage of two sisters to two brothers, however, was forbidden. At one time pre-puberty tali-tying is believed to have been practised, and among the aristocratic families the ta/i-tier was an elderly Nambudiri Brahmin who had no rights of cohabitation. Plural marriages were forbidden for women. Marriages were expected to last (although divorce and remarriage were allowed) and the spouses had much greater rights and obligations toward one another

than among Nayars in central Kerala. Adultery by the woman was punishable and could lead to divorce, although discreet extra-marital relationships between male and female cross-cousins were sometimes tolerated. Polygamy was permitted for men and was noted to have been common among the elite. Residence was ideally avunculocal, with women joining their husbands in the latter’s matrilineal estates for the duration of the marriage. Cases of matrilocality were rare and usually limited to younger and poorer men. There were occasional cases of duolocal marriages where the spouses lived in close proximity in the same village. But in most cases the rugged, densely forested and inaccessible terrain, the associated geographic scattering of houses and villages, and the fact that men often cultivated their own estates, made a visiting relationship impractical. Sons normally moved back to

their own matrilineal homes at puberty, and daughters moved to their husbands’ taravad on marriage. In large estates, the karanavan allotted separate gardens and rice fields to

married men for usufruct in undivided taravad land, and men lived in separate houses with their wives and children on the men’s matrilineal estates. Sometimes a man could lease in land from his wife’s taravad with

the consent of his own and his wife’s karanavan. The karanavan had considerable freedom to build a house and develop new gardens on his wife’s taravad estate, which would go to his wife’s matrilineal descendants.

Occasionally men gifted landed property to their wives and the latter’s matrilineal descendants: chiefs and kings, in particular, were known to gift ‘8 The description below is based especially on Gough (1961a), Fuller (1976), Buchanan (1807), and personal conversations with Joan Mencher.

Customary rights and associated practices 115 large domains, although this was contrary to accepted practice. Such gifts led to the formation of branch property groups, tavazhis.

Even after marriage, women were often partly maintained by their maternal taravads, and received a share of the harvest as their right. Moreover, the ancestral jewellery which they took with them on marriage remained part of the joint property of their maternal taravads, and could be

recalled if needed. A married woman could visit her ancestral home for extended periods of several weeks at a time, while a divorced or widowed woman had a right of permanent residence there; indeed she had no option but to return there since on divorce or widowhood she ceased to have any

claims whatsoever in her husband’s taravad. However her claims in her maternal taravad (as noted) were strong; and older women, when they returned, could take over the management of household affairs. Widow remarriage was freely allowed, although it was more likely if the woman was young. In such cases the children from the first marriage remained 1n the mother’s taravad which maintained them.

In short, among Nayars of both central and north Kerala, land was traditionally held in joint family estates and was not individually inheritable. Inheritance was through the female line, but most women had little

control over the management of property, although they were equal participants with men in any major joint family decisions such as those involving the partitioning of any part of taravad land. Among the central Kerala Nayars, women continued to reside in their natal homes after marriage. In north Kerala, although women moved to their husbands’ homes on marriage, and occasionally husbands had use rights in the land of their wives’ taravads, such land remained within the overall control of her natal family, increasing in value through any improvements brought about by temporary usufruct arrangements. And it was in their maternal taravads that women had rights of inheritance and maintenance outside marriage.

The Tiyyars of north Kerala. The Tiyyars of north Kerala and parts of Travancore had matrilineal systems very similar to those of the north Kerala Nayars, although the Tiyyars in central Kerala had been largely patrilineal for some time.°? From early sixteenth-century accounts of local customs, Gough (1973) argues that sometime prior to and up to that period the Tiyyars of north as well as central Kerala must have been matrilineal, practising duolocal residence like the Nayars, whose serfs they were at that time. At some point in the sixteenth century, she argues, the

central Kerala Tiyyars adopted virilocal residence and tended towards 5° The description in this section is based largely on Gough (196la) and Thurston and Rangachari (1909).

116 A field of one’s own patrilineal inheritance patterns. However, in parts of Travancore and in north Kerala, matrilineal systems persisted, with landed property being held in joint family units. Among the north Kerala Tiyyars Gough suggests, on the basis of later writings, that over time the pattern of residence became

avunculocal (from duolocal), with wives joining their husbands in the latter’s matrilineal estates. Sons returned to their matrilineal kinsmen at puberty or marriage. Divorced or widowed women also returned to live in their matrilineal homes. Divorce could be initiated by either spouse and involved little formality. Traditionally the Tiyyars of north Kerala practised the pre-puberty talitying ceremony, but the tali was tied by a woman, usually in the role of a potential mother-in-law: for instance, she could be the girl’s father’s sister, or the mother of the man she was already engaged to. Marriages between matrilateral cross-cousins were preferred. Bridewealth was given by the groom’s family to the bride’s karanavan, while the bride’s party provided a dowry in jewellery and cooking vessels. Gough (196la) argues that the bridewealth was probably to compensate the bride’s family for the loss of her labour, since women did regular outdoor work: they grew vegetables, milked cows, and transplanted, weeded, and harvested rice. Among the

north Kerala Nayars, by contrast, neither bridewealth nor dowry was exchanged, the jewellery given to them on marriage (as noted) was considered part of their maternal taravad’s property, and women of landed households usually did little outdoor work. The Tiyyars ranked below the Nayars in caste and traditionally also in economic status, serving as sharecropping tenants on the gardens and rice

fields of the Nayars and Brahmins. In Gough’s (196la) judgement, customarily these sharecropping rights must have been hereditary. The Bants of south Canara. Unlike the Nayars, relatively little has been written on the matrilineal Bants of south Canara (now in Karnataka state).°° Thurston and Rangachari (1909: 149) describe them as a largely ‘independent and influential landed gentry’. They were the most significant agricultural community in south Canara. Ancestral property was inherited in the female line, in accordance with the Aliyasantana system of matrilineal inheritance. Their residence patterns were similar to those of the Nayars and Tiyyars of north Kerala. Women were expected to join their husbands in the latter’s matrilineal homes after marriage. In practice, however, in

households where younger men were away in military service, married women often remained in their natal homes. Agricultural cultivation, mainly rice, was critically dependent on female labour. Thurston and 60 The few existing accounts include Thurston and Rangachari (1909) and Claus (1975).

Customary rights and associated practices 117 Rangachari (1909: 150) note: ‘There is, in fact, not a single thing about agriculture which the South Canara man knows, and which the South Canara woman does not know’. Indeed in peasant households transplanting, weeding, and harvesting were done mainly by women, as was postharvest rice processing and care of cattle. Women of wealthy households sometimes managed the large estates ‘as efficiently as men’ (Thurston and Rangachari 1909: 150). Women resident avunculocally were still obliged to contribute agricultural labour to their natal homes when needed. Crosscousin marriages of either category were favoured. Widowed or divorced women returned to their maternal homes with their children and had no further rights to maintenance tn their marital homes, as we noted was also the case among the north Kerala Nayars and Tiyyars. Younger widows usually remarried, but divorcee remarriage was rare. Other non-Muslim matrilineal groups in the region: the Phadiyas and Chettis of Wynad district. The regions around northern Kerala and the southern tip of present-day Karnataka contained a number of matrilineal groups with customs similar to those of the dominant matrilineal groups in

their proximity. According to Gough (1973), the pattern of matrilineal inheritance and duolocal residence was probably widespread in this area several hundred years ago, especially before cash cropping and timber procurement from the forests enabled men to make money outside the joint éstates and to set up separate households. The Phadiyas and Chettis of Wynad district, located in northern Kerala and bordering Tamil Nadu, would be among such groups. From the brief

description provided by Furer-Haimendorf (1952), the Phadiyas were organized around matrilineal joint families termed tarwars, similar to the Nayar taravads. The oldest woman of the joint family headed the household and managed its domestic affairs; she was succeeded by the woman next to her in age, usually the oldest daughter. The karnar, or oldest man in the tarwar, managed the estate. Tarwar property usually consisted of paddy

and dry land, gardens, the ancestral house, cattle, ritual objects, and sometimes cash. As among the Nayars, no part of the tarwar estate could be

transferred without the unanimous consent of all adult tarwar members. However, a man could gift his self-acquired property either to his sons or to his own tarwar. A man received the brideprice for his daughters but that for his sons’ wives was paid by their maternal uncle. Post-marital residence was

avunculocal, but women returned with their children to their maternal homes on widowhood or divorce. A woman would thus grow up in her father’s tarwar, move to her husband’s farwar on marriage, and return to her own (that is her maternal) tarwar on widowhood or, sometimes, in old age. Women also enjoyed considerable pre-marital sexual freedom.

118 A field of one’s own The inheritance and residence customs of the Wynad Chettis were very

similar to those of the Phadiyas. Rice land and cattle formed the most important part of their tarwar property. A member of the tarwar who wanted to separate could be given a share of tarwar cattle and unhusked rice for his family’s maintenance till the next harvest but no share of the tarwar land. Cattle could also be sold by the karnar for tarwar expenses. Marriage with cross-cousins was allowed and post-marital residence was avunculo-

cal, with the woman joining her husband in his matrilineal estate; she returned to her own tarwar on widowhood or divorce.

Thus far we have focused on matrilineal inheritance among the nonMuslims of south India. However, located in this region are also pockets of matrilineal Muslims: the Mappilas of north Kerala and the Lakshadweep Islands.

The Mappilas of north Kerala and the Lakshadweep Islands. The Mappilas of north Kerala are Sunni Muslims, believed to have originated

partly from Arab traders who landed in Kerala sometime in the eighth century AD and partly from indigenous converts to Islam (mainly from among the Tiyyars). Gough (1973) notes that although the traders were present in central Kerala in the eighth century, it was probably not until the late thirteenth and early fourteenth centuries that large-scale conversions took place, and by the mid fourteenth century Mappilas were widespread in

south and north Malabar. She observes that sometime afterwards, those located in south Malabar must have become patrilineal (although she provides no specific reason for such a shift), while those who had earlier migrated to the north remained matrilineal. The economic status of the Mappilas of north Kerala ranged from large property-owning Muslim aristocrats to those who owned no land at all. Among the rich households, Tiyyar labourers attached by hereditary right to the taravad helped in cultivation. Some Mappilas were also prosperous traders. Property-owning Mappilas, whether of commoner or aristocratic descent, had lineage groupings similar to those of the Nayars. They were predominantly matrilocal: the husband moved to his wife’s ancestral home

and contributed to her and their children’s livelihood. Among poorer families even the karanavan resided matrilocally, but in rich households

men who succeeded to the headship of the taravad returned to their matrilineal ancestral homes to live in or near their kinswomen and to manage their estates. In choosing marriage partners, cross-cousins (especially matrilateral cross-cousins) were favoured. Women did not veil themselves before male

Customary rights and associated practices 119 paternal parallel cousins, but did do so in the presence of older male crosscousins except their husbands. Divorce was possible for both parties but rare in practice. The Mappilas of the Lakshadweep Islands appear to have had a dual origin. The earliest settlers are thought to have been Nayars, Tiyyars, and Mukkuvans from central Kerala, with matrilineal descent and duolocal residence. Subsequently, probably around the fourteenth century, according to Gough (1973), there was an influx of Mappilas from north Kerala. The customs of the Island Mappilas were close to those of north Malabar, except that the Islanders practised mainly duolocal residence with the husband maintaining a visiting relationship.°! This was probably facilitated by the greater physical proximity of households there than was found in north Kerala. Property (the most important form of which was coconut trees) was inherited through the female line, but use rights to it could be obtained by individual male or female members of the taravad; it reverted to

the taravad on their deaths. Today, as will be discussed in chapter 6, the growing influence of Islam has created a mixed form of property devolution: some is inherited in accordance with matrilineal rules and some in accordance with the Shariat, that governed by the former still being the dominant form of property held. All the communities described above practised both matrilineal descent

and matrilineal inheritance. However, there was at least one agrarian community of south India which practised matrilineal descent but bilateral inheritance (that is ancestral property passed to and through both sons and daughters): the Nangudi Vellalars of Tamil Nadu.

The Nangudi Vellalars of Tamil Nadu. This was a small agricultural community, localized in some of the villages of Tinnevelly district (Tamil Nadu), practising matrilineal descent and bilateral inheritance.®°? As described by Dumont (1957: 14-15), in the 1950s about half the property

(including land) was held by women and transferred from mother to daughter, and about half was held by men and passed from father to son. Women had exclusive inheritance rights to houses. The daughter received her inheritance at the time of marriage through a dowry, which was strictly the woman’s property. Marriage between patrilateral cross-cousins was preferred, and involved no exchange of ceremonial gifts. Residence was matrilocal, in a separate house provided by the wife’s parents to the newly married couple. In a few cases (among the chiefs), when their own daughter °! See Kutty (1972), and Dube and Kutty (1969). °? For details, see Dumont (1957), Tambiah (1973), and Thurston and Rangachari (1909).

120 A field of one’s own married and thereby inherited the mother’s property, the couple moved out

to live in the house of the man’s parents, whom he replaced in his natal home on their death. The above pattern of inheritance has a few features in common with the Thesawalami customs of Sri Lanka’s Jaffna Tamils (described below), and

lends some support to Mayne’s (1900) suggestion that in the past the Thesawalami may have been practised by some Tamil groups of south India, from whence the Jaffna Tamils are said to have originally migrated.

The Nangudi Vellalars were closer to the Sri Lankan than the south Indian matrilineal pattern in another respect: property was held and managed by nucleated households and not as joint estates, as was the case in

greater or lesser degree among the south Indian matrilineal communities discussed above. In communities holding joint family estates, as we have seen, the partition or alienation of land was circumscribed and subject to some form of family or communal consent. Also, as noted, there was a structural divergence between the transfer of property (which was through women) and its effective control (which was largely through men). Within a more nucleated setting, women were in a better position to exercise effective

control over their inherited shares. This was also the case among the bilateral and matrilineal communities of Sri Lanka, to which we now turn.

(3) Sri Lanka: The Sinhalese, Hindu Tamils, and matrilineal Muslims Agro-climatically, the island of Sri Lanka is divided into two distinct zones

~ the Wet and the Dry — which have different land use and settlement patterns (see map 3.2 for provincial divisions and the Dry/Wet Zone demarcations within Sri Lanka). About 70 per cent of the population is today concentrated in the Wet Zone, which accounts for some 30 per cent of the land area and has an average annual rainfall three times that of the Dry Zone. The economics of the Dry Zone settlements is crucially dependent on

water availability which has dictated their size, location, and cropping patterns. Yalman (1967) describes four types of agricultural settlements which characterized the Dry Zone: those found in the northern plains with

elaborate tank irrigation systems; those located in the mountainous districts of the Central Province using rain-fed streams for terraced paddy cultivation; those found in the more backward parts of the eastern districts practising rain-fed paddy cultivation; and those spread in the dry jungles where people subsisted entirely on rain-fed swidden cultivation (termed

chena) and hunting-gathering. Ethnic and religious clustering superimposed on ecological variations tended to produce some notable regional

patterns of inheritance, marriage, residence, land use, and the gender

|\

|f5i_3¢|f || |!; Net % | :'”}SS eo S\

Jatt \ i i!| VT PON ne i a) ‘ Zz REN ! Mond fey, . 4; $ tev Ey,

Customary rights and associated practices 121

e—e—e— OPT CAS OPT OPT CLT LES PTEPT Pe OPECPS ALP OP OAL OILS OIC ORT O BLO DS ORE LP EITL PD 2 HPO BPD OPT REALS LIS OPE BIO PEPE EGE REO PF ERP EOPE MF ERPERPLBS EDGE PFEDPE DI ERRERPE DIL RS A DPEDPE PERI E OPE RSE RPL OPE BPE RIE RP ERPENPE APE OPS DIG OPE O PE eieainaiaaiiaiaeiete:

4

Braddon sy Be, “—n,

5! onon %. d ey 3 f Northern oh By A Province H peer Na ( ae a) ‘ hn Central ( ! : Province !| bGy ms yar | a 1 Mie he,

%

Ta te

‘ ey : g ' 3 + ‘ :%,F

f

*

§

™

{\| North i —, | Ay Western ;P Province |. SK 4 North Centra x

z

PeOH LinnsFt 4% }, b.., pn otf Lyite, | Faster

\ CeLLY, lwdUCULT CE EOELLY EEE EEE i| VL), TT

eee

Ma LiMn 5th

PHT 2 nf. Ws es wy tes ay Lt Oe i) ie uf 4 Ms hes 7. Ms id : f

SL gay METH YY iy !

OLEATE YY LU EES Wy, iggy LYE Wie HE

|| nColombo VEE LD 8 YL7LEevince EE EL *

i$; ’LMM CEH, TELAT ESE EEE: YUE Yet 4i VE LE LY, Ly Yi yyy Li yeh Va } 1) t 7? LI YEE ih iy Wy ty ay GUM Lyi Ly ies ere ihsheYE, ip eeue py oy % H i; ELI TS ME LLY Hie Weep Hy Ge ty ee is q NL, CEE ARN I dae WEG 5 CLE eta Uns LEAs CARREY 3 Wi WI ip mine , ,

yy Ly Dry yy LMM a |i$j‘ QAnce : (“7] Zone 40 kms “VC tse RS OA ILS . ‘oi . RI nee“ah UE lLELLA Yi peat igen OM SABO A vey : yyy LY EL TL Pe

; philly

Searce of zonal

division: Yalroan (4987)

i ;

i BER PLEPLOPLERL OPE AIIR RE OPER PED OL PTO LE PLE OLR OPE OID OLS BPE REE DPE CPU OLE RIS Ot be DIET Die IO Ee OPE APL EOS DHE OPE RPE EL OLE OIE OPE ELE CES EPEROE IE HIE RPS OPE OPS EPS BILAPV OPE OLS EPS OBS OLS ODL EPL APD OPE OPE EPL ELS BLL OIE Bed BPE AIDED DPE ELL EPP CLD OILOPE DE PG DBS oP DOE secuneesetoveesunpusesssancuennessessonsenseerh

Map 3.2 Sri Lanka: provincial divisions

122 A field of one’s own division of agricultural labour. But despite these variations (which will be apparent in the descriptions below), one of the striking features about Sri Lanka 1s that women customarily had significant rights in landed property among the major communities of all regions and religious persuasions: the Buddhist Sinhalese, the Hindu Tamils, and the Muslim Moors. Consider

each in turn.°3 |

The Sinhalese.°* Under the traditional Sinhala laws of inheritance

(and also the contemporary Kandyan law), an important distinction attached to whether a woman married in binna or in diga. In a binna marriage, the husband came to live uxorilocally with his wife and in-laws and, jointly with his wife, managed the land she inherited. Their children

had rights and obligations in their mother’s household and took her ancestral name. Under diga, the woman moved out to live virilocally with her husband, and the children took their father’s name. A binna-married daughter had the same rights of inheritance in the ancestral estate (praveni)

of her father as her brothers and unmarried sisters. A diga-married daughter, however, had to forfeit her share in her father’s praveni, sometimes but not always receiving a dowry instead. Most writers note that

a dowry in movables, cash, and occasionally land was given among the wealthy, especially the nobility, but rarely among the middle and poorer peasants. Dowry was clearly not a daughter’s right in the way inheritance was, and could not be demanded if the family chose to give her none. Moreover, even a binna-married daughter had to forfeit her rights if her

marriage broke up and she subsequently married diga, although her children by her binna marriage retained their rights to their maternal grandfather’s estate. Binna marriage usually implied residence with the woman’s parents, but in some instances the couple could live separately on the woman’s land. It is noteworthy that the patrimonial inheritance of sons was not conditional on the type of marriage they contracted. For a woman, only if she was an only child could she inherit irrespective of the kind of marriage she contracted.

A diga-married daughter who subsequently returned to her parents’ home could re-establish her rights to the patrimony in the following circumstances (Hayley 1923: 389-90): (a) if she returned during her father’s

lifetime and was allowed to settle on his estate in binna, either with the husband with whom she had previously departed in diga, or with a new °3 For a detailed discussion on inheritance and marriage practices in Sri Lanka, see Agarwal

9,

04 The decctiption below is based especially on Knox (1681), Hayley (1923), Obeyesekere (1967), Yalman (1967), and Tambiah (1965, 1966).

Customary rights and associated practices 123 husband; or (b) if she returned after her father’s death and the other heirs expressly consented that her marriage (either to her former diga husband or to a new man) could now be considered a binna marriage.

Also in case of widowhood, divorce, or destitution, diga-married daughters who were in need could return to their parents’ house. At a very minimum this provided them rights to maintenance: ‘to have lodging or support and clothing from their parents’ estate...” (Sawers 1826, quoted in

Obeyesekere 1967: 42). There is some debate, however, whether such daughters had claims to parental property beyond maintenance if they chose not to remarry in binna. Hayley (1923) argues that they had none. But

other evidence suggests that women could re-establish their inheritance claims to the father’s estate by keeping in close touch with the parental home and rendering the parents assistance. Tambiah (1965) argues that the

distinction between binna and diga became critical, in terms of the daughter’s claim to her father’s praveni, mainly in cases where a diga marriage involved her moving out of the village. This was presumably because village exogamy perforce reduced her contact with her parental home. Obeyesekere (1967), for instance, emphasizes that the concept of rendering assistance to the father was an important determining factor in how these rules were situationally translated, as evidenced by the following statements: If the father left a son and a daughter, minors by one wife, and a son and a deega married daughter by another wife, if that deega married daughter came back and attended and assisted her father during his last illness, and if the father had therefore on his deathbed expressed his will, that his deega daughter should have a share of his lands, notwithstanding her being settled in deega, in that case the deega daughter will be entitled, by virtue of such nuncupative will, to participate equally with her uterine brother and their parental half-brother and half-sister in the father’s estate. (Armour 1860, quoted in Obeyesekere 1967: 50) Some of the chiefs are of the opinion that the daughter previously married in binna may preserve for herself and her children her own and their claim in her parents’ estate, by visiting him frequently and administering to his comfort, especially being present, nursing and rendering him assistance in his last illness. (Sawers 1826, quoted in Obeyesekere 1967: 49)

Essentially, as Obeyesekere (1967) notes, the fundamental rules of Sinhalese intestate inheritance embodied two basic principles: (a) equality of division among all children; and (b) the ultimate reversion of property to its source. He argues that the apparent contradiction in these two principles, the first emphasizing the bilateral character of Sinhalese inheritance and the second emphasizing its ‘unilinearity’, was resolved by a distinction between temporary and permanent rights to the praveni. Sons and binna-

124 A field of one’s own married daughters and their children who also carried the grandfather’s name were meant to enjoy a permanent right, and diga-married daughters whose children bore another family’s patrilineal name, a temporary right. Thus, Obeyesekere (1967: 42) notes: [While all members of the family [were] a man’s immediate heirs, the rights of some heirs [were] restricted in a manner that [would] ensure the continuity of the ancestral line, and the convergence of property among a body of agnatic kinsmen.

In practice, the rights enjoyed by a daughter and her children in her father’s estate were defined on the basis of multiple, interconnected criteria: her degree of contact with her parents and the assistance she gave to them, post-marital residence in terms of not only uxorilocality or virilocality but particularly village endogamy, and the taking of the maternal grandfather’s name by the children. Here post-marital residence with or near the parents was clearly a significant enabling factor in rendering assistance to them in their old age even though, in itself, as Obeyesekere emphasizes, it would not have been sufficient in the case of an uncaring child. In a deceased woman’s estate, if she was married in diga, all children from

such a marriage had equal rights no matter what their marital status (including diga-married daughters). But if the deceased woman was mar-

ried in binna on her father’s property, it appears that her diga-married daughters from such a marriage were excluded from a share in her ancestral estate (Hayley 1923: 467). During her lifetime a woman had full rights to

manage or alienate her separate property, whether inherited from her parents or received as dowry. There was no concept of a community of property after marriage (viz. property held in common by the spouses) in either diga or binna alliances.

Typically, a widow had no permanent rights to the deceased husband’s

ancestral estate, nor could she dispose of it to outsiders (except in Sabaragamuva Province, as discussed later). According to Obeyesekere (1967), historical sources disagree on whether, on a man’s death, his sons and binna-married daughters took over his praveni at once or only after the death or diga marriage of the widow. The likelihood was that these issues were judged according to their context by the villagers and the traditional courts. In a man’s acquired property both male and female children shared equally, subject to the widow’s life interest.°> A widow had a right to maintenance from her deceased husband’s ancestral estate if his acquired property was insufficient to provide adequate support, whether she was married to him in diga or binna. This right was not affected even if she had °° Widows and mothers could in some instances also inherit acquired property absolutely.

Customary rights and associated practices 125 means of her own. In contrast a binna widower could not inherit his wife’s

immovable property; and a diga widower had no interest in the wife’s

praveni, but did have a life interest in his wife’s acquired property (Goonesekere 1980). There was also a noteworthy local variation of the traditional Sinhalese

law. In Sabaragamuva Province, Obeyesekere (1967) notes that widows had much greater rights of inheritance than in other Kandyan areas. For instance, a woman married in diga succeeded to the whole of the husband’s estate if he died intestate and without issue, even if his brothers and other kin were alive. Her share fell to half if the man had children by a former marriage, since they inherited the other half. The same arrangement applied if she and the deceased had children. The widow had full alienation rights

over the portion allotted to her, but lost this right if she contracted a subsequent diga marriage. She could, however, dispose of her share if she wished, prior to such a marriage. Obeyesekere argues that these differences between Sabaragamuva and Kandyan law were not due to the influence of Roman-Dutch law on the former under colonial rule, but to the greater strength of marriage ties, the overall rarity of divorce and desertion, and the

strong ideology of wifely devotion, chastity, and fidelity prevailing in Sabaragamuva as well as in the Maritime Provinces.°° This was in contrast to the much looser marriage ties prevailing in the Kandyan areas. Cross-cousin marriages of both types were preferred among Sinhalese

everywhere, although their actual incidence varied across regions and classes. The practice was more rigidly adhered to among the aristocracy than among the commoners (Pieris 1956). In the Kandyan Highlands, except among the wealthy or aristocratic households, or where the families

of the bride and groom were relative strangers to each other, wedding ceremonies were usually dispensed with altogether (Yalman 1967). ‘In early times, the conducting of a daughter by a man of equal caste with the consent

of her relations constituted a marriage ...’°’ This also made it difficult to distinguish between ‘legitimate’ and ‘illegitimate’ children in the sense of those born within or out of wedlock. In fact all children were accepted as fully legitimate (with rights in the estates of both parents), except those born of specific types of unions. For instance, a child born of a union between a high caste woman and a low caste man, or of a union contracted contrary to

the wishes of parents, was considered semi-legitimate and had restricted °° The Maritime Provinces (also called the ‘Low Country’) were constituted of Ceylon’s Southern and Western Provinces that were occupied successively by the Portuguese, the Dutch, and the British from the early sixteenth century to the mid twentieth century. The Kandyan Provinces, which came to be known as the ‘Up Country’, consisted largely of

enn Central, North Central, Uva and Sabaragamuva provinces (see Obeysekere °7 The Kandyan Law Commission, appointed in 1927, quoted in Yalman (1967: 160).

126 A field of one’s own rights of inheritance; while the child of an incestuous union was considered illegitimate and had no rights of inheritance (Hayley 1923: 200-1). Extra-

marital relations were tolerated for both sexes, as long as they did not involve a high caste woman with a lowcaste man: ‘They do attempt to catch the culprits, but if the man is a friend or a good relation, nothing more will be said. Otherwise, he may escape with a beating. Of course, if low-caste

men were to be caught at such adventures they would be very severely handled’ (Yalman 1967: 188).

The Kandyans, men and women, commonly married and divorced several times during a lifetime: ‘For if they disagree and mislike one the other, they part without disgrace’ (Knox 1681: 93). Divorce involved no formalities: no payments to be returned, no authorities to be consulted. A couple who cannot get along may simply separate; their property is separate in any case, though property acquired during the marriage ought to be divided. Young children normally go with the mother; older children may live with either parent or other relations. (Yalman 1967: 187)

Although monogamy was the usual pattern, polyandrous unions were not uncommon among the Sinhalese, especially in the Dry Zone.°® The commonest form was fraternal polyandry (with one woman being married to two or more brothers), although there were also cases of such partnerships between step-brothers, cross-cousins, distantly related men and, in rare cases, even unrelated men. Polyandry usually did not involve simultaneous marriage to several men; more commonly the woman took a second husband after a year or two of monogamous marriage, either as a deliberate decision by the couple or to legitimize the wife’s extra-marital relationship (Tambiah 1966). Various economic explanations have been extended for such polyandrous unions: to enable land consolidation among brothers with very small parcels of land (Knox 1681, Tambiah 1966); to provide extra labour where fields were dispersed or where both shifting and settled cultivation was undertaken (Tambiah 1966); to limit the number of heirs and so prevent the subdivision of estates via inheritance (Pieris 1956), and so on. Be that as it may, what Is especially interesting is that the woman was

an active party in the decision leading to such a union, which was not necessarily the case among South Asian patrilineal groups practising fraternal polyandry.°? Women’s contribution to agricultural cultivation among the Sinhalese °8 See e.g. levers (1899), Knox (1681), Leach (1955), Pieris (1956) and Tambiah (1966). ©? Among such groups are the Paharis of Jaunsar Bawar in northwest India (Berreman 1962), and the Tibetan Chumiks (Schuler 1987) and Nyimba (Levine 1988) of Nepal.

Customary rights and associated practices 127 varied by the type of agriculture practised and the affluence of the household.7° In chena their labour was indispensable and widowed and divorced women sometimes cultivated their own chena independently, drawing upon male relatives for the initial clearing of land. In settled paddy cultivation women’s role was limited to certain tasks, especially weeding and harvesting. Given the importance of chena in the Dry Zone, women’s contributions to agriculture among the Sinhalese located there are likely to have been greater than among those settled in the Wet Zone.

The Hindu Tamils. As among the Sinhalese, Tamil women in Sri

Lanka, unlike most of their Tamil counterparts in India, had strong inheritance rights in land. However there were some notable differences among the Tamil groups settled in different parts of Sri Lanka, both in

terms of customary practices and the codified laws of marriage and inheritance. The discussion below will focus on two important Hindu Tamil

communities: the Jaffna Tamils, and those settled in the Eastern Province.7!

The traditional law of inheritance (and marriage) of the Jaffna Tamils — the Thesawalami — differed from the Indian shastric systems in significant ways (as will be apparent below). The Thesawalami was first compiled into a code of civil law under Dutch administration in 1706—07 by twelve leading Jaffna Tamil citizens chosen for their knowledge of the law. The Code came into full legal force after the British passed Ordinance No.5 of 1869. Prior to

Dutch codification, the Thesawalami is said to have undergone some modifications under the Portuguese. However, the text of the code provided

some information on inheritance patterns prior to Portuguese rule, in addition to describing Jaffna customs prevalent when the Code was first compiled (Banks 1957). Under the pre-Portuguese system three categories of property were recognized:’? chidenam, the hereditary property brought

as dowry by the wife; mudisam, the hereditary property brought by the husband; and thediathetam, or property acquired during the course of the marriage. This last included any income in cash or kind from the chidenam or mudisam. Hence grain produced from either chidenam or mudisam land was part of thediathetam; similarly income from the sale of animals (but not the animals themselves) fell into this category. Any capital investment on 70 See Brow (1978), Gunawardena (1989), Yalman (1967).

7! For an interesting account of the history of Dravidian settlements in Ceylon and the beginnings of the Kingdom of Jaffna, see Indrapala (1965). 72 The description below of the pre-Portuguese system draws especially on Banks (1957) and Tambiah (n.d.).

128 A field of one’s own land, however, became part of the category of property on which it was made, the chidenam or the mudisam.

For ancestral property the guiding principle was for inheritance to strictly follow gender lines. A man’s mudisam property was inherited on his death by his sons, and a woman’s chidenam property went to her daughters as dowry on marriage. Thus the man’s property always remained with the male heirs and the woman’s with the female heirs, but the thediathetam went to children of both sexes equally, after the death of both parents. Prior to its distribution, however, the size of the thediathetam was subject to deduction for any decrease in the chidenam and mudisam that had occurred during the lifetime of the parents. If this value was insufficient to make up the loss, the heirs of the chidenam or mudisam, as the case may be, had to bear the loss.

The remaining thediathetam was first divided into two halves, one half going to girls, the other to boys. Each half was then divided equally among the members of each sex. Acquired property could not be used as dowry

prior to the deaths of the parents. In the absence of children of one sex, those of the other inherited all three categories of property. The property of a childless woman passed after her death in the first instance to her sisters, and in their absence to their undowered female descendants (daughters and granddaughters). If the latter were already dowered, the heirs were the deceased woman’s brothers and their male descendants. Under these special circumstances, property could move from

One sex to another. In the absence of siblings, the property went to the woman’s parents, if alive; if the parents too were dead, it went to the parents’ siblings and their descendents. A widow with young children retained both her husband’s estate and her own, with which she dowered her daughters, the sons inheriting their dues

on her death. If she remarried, of the children from her first marriage, daughters received dowries from her chidenam, sons their natural father’s mudisam, and both sons and daughters shared the thediathetam. Of children from her second marriage, likewise, daughters were dowered from her chidenam, while the sons received their own father’s patrimony. A widower dowered his daughters from his wife’s chidenam, his sons received his mudisam, and all siblings shared the thediathetam. Daughters by different mothers but a common father derived the major part of their dowries from their own mothers, while sons by different mothers inherited their main portions from their common father. This provided for a scrupulous division of property along gender lines and also between half and full siblings. Typically a man could not inherit any part of his parents’ estate premortem (unless they were incapable of managing it themselves), and all that he earned prior to his marriage went into the common estate. What a woman received as dowry was written into a doty ola (dowry

Customary rights and associated practices [29 deed) in her name. As the Code cynically put it, ‘a well drawn up and executed doty ola’ was necessary since ‘it is by ... means [of the dowry] that

most of the girls obtain husbands, as it is not for the girls but for the property that most of the men marry’ (quoted in Tambiah n.d.: 172). The right to dowry property, once created, could not be lost, but the dowry claim could lapse and revert to the common estate if the woman did not take possession of it within a prescribed period. In case the parents prospered after the daughter’s marriage, she could ‘induce’ them to subsequently add to her dowry, although she could not demand this as a right. However, if the dowry land turned out to be a bad deed and was later lost to the daughter through a lawsuit, the parents or (after them) their sons had to make good the loss. Of course not all sections of the population could give a dowry. The

practice was virtually absent among the poor and untouchable castes. However among the Vellalas, the principal agricultural community of the Jaffna Tamils, dowry was customary and usually consisted of three parts: land, cash, and jewels. Land was mandatory, while cash and jewels were

optional. Dowry land was an assortment of garden land, paddy land, a home compound usually with a house, and rocky, uncultivated land. Garden land, in turn, could be of various types: gardens of palmyra palm, coconut palm, vegetables, and fruits. These were to some extent interconvertible, depending on soil type. Receiving garden land was important since paddy cultivation (partly rain-fed, partly irrigated) was characterized by low yields and Jaffna was a rice importing region at least after the mid eighteenth century (Banks 1957).

On marriage the husband acquired operating control over his wife’s dowry land, but could not alienate it without her consent. He could, however, dower his unmarried daughters with their mother’s chidenam if she died. The dowry property or any rent or profit from it was not liable for a husband’s debts. But a woman’s rights of disposal over her dowry and over her share of the thediathetam were restricted. While still living with her husband she could not lease, mortgage, sell, or donate her dowry property

or her part of the thediathetam without her husband’s consent. Only if maritally separated could she independently operate her property (dowry as well as her half of the acquired wealth) without her husband’s consent or control. And as a widow she could sell her property without her children’s consent. By contrast, the husband, even while living with his wife, could

lease out, mortgage, or sell without her consent not only his ancestral property and his share of the thediathetam, but even her share of the thediathetham. In the latter case, his wife could bring legal action for compensation only after a marital separation. The husband could also gift up to one-tenth of his ancestral property without the consent of his wife and

130 A field of one’s own

heirs.73 In terms of property control, therefore, although separated/ divorced or widowed women could manage their own estates, a married

woman’s legal position was constructed as subordinate to that of her husband. Some aspects of this legal disability continue even today (for details see chapter 5). Marriages among the Jaffna Tamils were usually within the village and between persons related to one another, with a strong preference for cross-

cousins (both patrilateral and matrilateral). Marriages were arranged, absolute pre-marital chastity was expected from girls and preferred for boys, and post-marital fidelity was obligatory for both partners. Although divorce was permitted, the chances of remarriage were small, especially for women. Similarly widow remarriage was permitted but was more likely if the woman was young and childless. Post-marital residence varied between uxorilocal and virilocal during the couple’s lifetime. The location was dictated by property ownership. Usually at the beginning of their married life the couple went to live in the house the wife had received as dowry and cultivated her land. Subsequently, when the husband inherited land from his father they shifted there, by which time

much of the wife’s property would have passed to her daughters as chidenam. During his lifetime, therefore, a man could expect to move residence thrice — once to his father’s mudisam estate when his sisters married, then to his wife’s chidenam estate when he himself married, and then again to his own mudisam when his daughters married and inherited the chidenam, while he inherited his patrimony. A woman only moved twice, once when she married and then when her daughters married. The Tamils of Batticaloa region in the Eastern Province, among whom the Mukkuvar community (land-owning and high-ranking in caste terms) was the most dominant, had strong similarities to as well as some differences from the Jaffna Tamils. According to McGilvray (1973, 1982), one of the few scholars who has studied the community, the Mukkuvars of Sri Lanka

probably originated in Kerala and in turn ‘encouraged, perhaps even enforced, the structural replication of their matrilineal clan system in all the 73 It is not clear from the accounts of either Tambiah (n.d) or Banks (1957) whether the husband's ability to lease, mortgage, or sell his ancestral or acquired property was also restricted to a maximum of one-tenth of it. Both authors only mention the one-tenth restriction in relation to donations. Banks does not discuss other forms of disposal, while Tambiah (n.d.: 121, 194) does so without mentioning any such restriction. In my view, there is no apparent reason why this restriction should have applied only to donations, since the

principle that the heirs should not be deprived of a share in the property would apply equally to other forms of alienation. But be that as it may, the man’s freedom of disposal was still greater that his wife’s. And after the Jaffna Matrimonial Rights and Inheritance Ordinance of 1911 (to be discussed in chapter 5) was passed by the British, the man’s rights of disposal extended explicitly to the whole of his property.

Customary rights and associated practices 131 lower castes’ (1982: 87).7* In some ways, however, their inheritance and marriage practices also resemble those of the Nangudi Vellalars, in that inheritance (although matrilineal) was via dowries, and post-marital residence was matrilocal. Unlike the Nangudi Vellalars, though, virtually all wealth was pledged and transferred to the daughters as dowries, the eldest one getting the largest share. Land and other immovables were deeded either solely to the daughters or jointly to the daughters and sons-in-law. In other words, among the Hindu Tamils in all regions of Sri Lanka, women’s inheritance claims received strong recognition, as did the claims of Muslim women, as discussed below.

The matrilineal Muslims. Like Kerala and the Lakshadweep Islands in India, Sri Lanka has a pocket of traditionally matrilineal Muslims. Most belong to the Sunni sect and are today governed by the Hanafi school of Islamic law (as will be described in chapter 5). The majority trace their descent to Arab traders who came to Sri Lanka around the eighth century AD, although it is believed that some Arabs resided in Sri

Lanka even in pre-Islamic times; and there has also been some Muslim immigration from the west coast of India (Mohan 1985: 5). Historically concentrated especially in the Eastern Province, and termed ‘Moors’ under Portuguese rule, they assimilated the customs of the local Tamil popula-

tion.’?> Many adopted Tamil as their language and began to practise matrilineal inheritance and succession. According to McGilvray (1973), the Muslim population in the region ‘is largely the product of marriages with

and conversions from the Hindu castes. They consequently share the kinship patterns, matrilineal clan organization, matrilocal marriage systems and many other customary practices of the Hindu Tamils ...’ The two communities maintain a strict residential separation today, but show considerable similarity in their social practices, including passing on landed property to daughters via dowry and tying the tali as a part of the marriage ceremony. 7°

Yalman (1967) described the Moors as a community subsisting on a mix of swidden and settled paddy cultivation, and practising matrilineal descent

and inheritance and matrilocal residence. Property was transferred to 74 In India, Mukkuvar villages stretch along the coast of Kerala from Malabar down to the tip of Kanyakumari district, now in Tamil Nadu state (Ram 1991). Their systems of inheritance, residence etc. would thus have been subject to quite differing influences over time in different parts. In Kanayakumari district, Mukkuvar fisherfolk, as observed today, are not matrilineal (Ram 1991). 7S The Malays in Sri Lanka, who also follow Islam, havea different history and are not termed Moors. 76 Indeed McGilvray (1973) provides a single description of the practices of marriage, inheritance, and residence for both the Tamils and the Moors in his area of fieldwork.

132 A field of one’s own daughters through dowries consisting of cash, land, cattle, etc. Typically the eldest received the largest dowry and her husband was expected to render the most help in finding suitable husbands for her sisters. Sons moved out to live in their wives’ natal homes on marriage, but retained usufructuary rights in their own natal homes to any undivided property

(including land) left over after their sisters had been dowered. (This sometimes included property that had been promised as dowry to a daughter but had not yet been divided and transferred to her on marriage.) The children belonged to their mother’s family. The girl’s family had no control over the cash she was given in dowry (the cash was not returned on divorce: Goonesekere 1980), but it could exercise control over the immovable property — land, houses, and trees. The resident son-in-law managed his wife’s fields, but needed her consent to sell any part of the land and, in the case of undivided property, also the consent of his inlaws. Women’s managerial control over their land was thus much more limited among the Moors than the Kandyans. Women’s role in cultivation was also more restricted, both in chena and in settled paddy cultivation. In overview then, traditionally among all major communities in Sri Lanka, irrespective of religious or ethnic differences, both sexes had inheritance rights in land, even though the forms of devolution differed. This strikingly highlights the strength of cultural commonalities over religious differences. For instance, in terms of their rights in landed property, Sri Lankan Hindu women of the Jaffna Tamil community had much more in common with

Buddhist Sinhalese women and with women of the matrilineal Moor community, than they did with Hindu women in northwest India. Also, in

contrast to the matrilineal communities of India, Sri Lankan women inherited land as individuals rather than as members of a joint family and, especially among the Sinhalese, could exercise a considerable degree of control over its management.

(4) Some cross-regional comparisons In very broad terms, the above descriptions suggest three types of matrilineal and bilateral communities embodying three different patterns of land rights customarily enjoyed by South Asian women. One category is of communities among which land was aclan’s communal property and could not be inherited either by individuals or by joint family units. Both women and unmarried men had usufruct rights in this clan land, as individuals or as members of nuclear families. Responsibility for managing the land vested with the husband, but a woman’s labour was critical and she could control the produce. The Garos are the only ones that fall in this category.

Customary rights and associated practices 133 A second category is of communities where land, although inherited in

the female line, was held as joint family property, and women had no individual rights of alienation. Responsibility for the overall management of land again rested principally with men, typically brothers or maternal uncles. However, in decisions relating to the partition or transfers of landed property, women’s concurrence was necessary. This category of communities would include the Khasis of northeast India, and the Nayars, Tiyyars, Mappilas, Phadiyas, and Chettis of southwest India. A third category is of communities where both women and men had

inheritance rights in land, which could be held individually. Women inherited either via dowries (as among the Nangudi Vellalars, Jaffna Tamils, and matrilineal Moors) or post-mortem (as among the Sinhalese). Women’s control over the inheritance varied from considerable among the Kandyan Sinhalese, to restricted among the matrilineal Moors.

Many of these traditional patterns have changed under colonial and post-colonial rule. In the next chapter, the nature of these changes and the factors underlying them will be examined for three of the communities described here — the Garos, Nayars, and Sinhalese — one from each of the categories identified above. But let us now consider, in the light of these case studies, two questions posed at the beginning of this chapter. One, what structural conditions were linked to women’s rights in land? Two, what can we deduce about gender relations and women’s ‘status’ in these societies?

Il. Women’s land rights, structural conditionalities, and gender relations

Let us first examine what structural conditions or norms were linked historically to women’s land rights, in terms of post-marital residence, choice of marital partner, and sexual freedom. Below I specify some hypotheses on the likely nature of these links, and use the case study material presented above to examine the validity of my hypotheses. My differences with Goody’s (1973, 1976) propositions on these links are also discussed in this connection.

(1) Women’s land rights and associated practices Since land is the most important form of property in agrarian economies, we would expect families and communities which recognize women’s inheritance rights in landed property to try and ensure that the land ts not lost to persons outside the kin group. The granting of such rights could thus be subject to conditions, or at least to marked preferences, regarding whom

134 A field of one’s own women marry and where they live after marriage. For instance, if daughters were to marry close relatives such as cross-cousins, rather than unrelated persons, the land would remain within the overall control of the extended kin group. Again if a married daughter were to reside within the village, and

especially in the natal home, the parental family would be in a better position to exercise control over any landed property she inherits from it, than if she went to live in a distant village. Women themselves would also be

in a better position to use and manage such land if they lived within the village.

The idea that there could be links between women’s property rights and

marriage practices is not new. Goody (1973, 1976)" presents several hypotheses in this regard. He proposes, for instance, that in societies practising ‘diverging devolution’ (that is where property is transmitted to

both daughters and sons), there would be a strong tendency to control women’s marriages. In particular, he hypothesizes that such societies would have a greater incidence of close-kin marriages and of uxorilocal, ambilocal, or neolocal post-marital residence, and a greater stress on pre-marital

virginity (to prevent early attachment to an unsuitable suitor). However, Goody’s formulation, while seemingly consistent with mine, is problematic on at least four counts.7® First, for reasons that are not apparent, Goody takes no explicit account of societies practising matrilineal inheritance. For a comprehensive analysis, the hypotheses need to be tested for all those societies (both matrilineal and bilateral) which traditionally recognized women’s rights in property, since in both sorts of societies, families would have an interest in preventing

the dispersion of property inherited by women outside the kin group.’ Second, Goody includes in his definition of property both movables and immovables, and aggregates societies which only give women dowries in movables with those which give them inheritance rights in land (either through dowry or post-mortem). The distinction is critical in the present discussion, since although we might expect societies which give women landed property to favour close-kin and in-village marriages in order to keep control over the land, it is not apparent why societies which only give 77 Some parts of Goody's (1990) discussion are also relevant here and will be referred to from

time to time. However, much of the discussion below is focused on his earlier writings which spell out in detail his propositions and analysis on these issues, and from which he does not radically depart in his recent work. 78 This is not meant to be a comprehensive critique of Goody’s formulation, which has been criticized also from other angles (see e.g. Whitehead 1977). But the points discussed below have a specific bearing on my own analysis. 79 In his (1976) three-fold classification of societies into those where diverging devolution is

present, absent, or where no individual property rights or rule of transmission exist, matrilineally inheriting groups presumably get subsumed under the second and third categories.

Customary rights and associated practices 135 women a dowry in movables would also need to favour such marriages. Cash, jewellery, clothes, utensils, and possibly small animals (the traditional constituents of movable dowries in South Asia) are not easy for relatives to control, nor do such items usually have the economic, political, and symbolic significance that ancestral land possesses that could motivate attempts at such control.®° (Of course, communities may still prefer invillage and close-kin marriages for reasons apart from property control, as will be elaborated later.) At a more general level, the issue of dowry vs. inheritance is important

and has also been the subject of recent debate among Indian women scholars (as will be discussed in chapter 10); hence the problems with Goody’s definition of dowry deserve some discussion here. Goody (1973: I,

1976: 6) sees dowry as ‘a type of pre-mortem inheritance’, ‘a process whereby parental property is distributed to a daughter at her marriage ... rather than at the holders’ death’, and as technically the woman’s property and in her control. He sees the difference between female inheritance and dowry as essentially ‘one of timing and flexibility’ (Goody 1973: 17), and (as

noted) does not distinguish between dowry in movables and dowry in immovables.®! The alternate view, presented by a number of authors on the

basis of field research among north Indian Hindu communities, is that dowry typically cannot be equated with inheritance in this way.® Neither view is accurate as a generalization for South Asia. Each reflects a regional perspective, Goody and Tambiah a south Indian/Sri Lankan one, the others a north Indian one. For certain communities in southern South Asia, dowry was customarily a form of pre-mortem inheritance, as among the Nangudi Vellalars of Tamil Nadu and the Jaffna Tamils and matrilineal Moors of Sri Lanka. These communities, as noted, used dowry to transmit to the daughter her share in the mother’s estate. The dowry was explicitly recognized as an inheritance share in ancestral property, and included both movables and land. However, for South Asian communities other than these, dowry was and is clearly not equivalent to inheritance for several reasons. 80 Yalman’s (1967: 291) comment from his ethnography on the matrilineal Muslims of Sri Lanka Is pertinent here: [They] are quite aware of the great distinction between the transfer of cash dowry to the son-in-law and the transfer of rights of management over landed property to the daughter. They realize full well that once cash has been given, little further control can be exercised over it. With immovable property, however — land and houses and trees — it is clear that the mother of the girl (the original holder) and the father of the girl (the original manager) give

up only a minimum of their rights when, in the marriage majlis (meeting) they agree to transfer these to the name of their daughter, and to allow the son-in-law to manage them. 81 Tambiah (1973) concurs with this view, although in a recent article he presents a modified and somewhat weaker version of it (see Tambiah 1989). 82 See especially U. Sharma (1980, 1984) and Madan (1975, 1989).

136 A field of one’s own To begin with, whether or not dowry is given, and how much is given, has always depended on the discretion of parents or brothers. Even customarily, daughters could not demand dowry as a right in the way that sons could demand their inheritance shares (and this would be even less possible today since the practice of dowry is illegal in most parts of South Asia).°3 As seen

from an examination of ethnographic evidence on dowry in India (especially Miller’s 1981 review), customarily dowry was not a universal

practice even among propertied groups, nor was its incidence uniform across the country. It was typical among the propertied in north India, but less common in the south, and virtually non-existent in the northeast where brideprice or no-payment was the norm. In the south there was also much greater equality in the marriage expenditures of the bride’s and groom’s families, as well as a greater incidence of marriages with no payments or token payments. In any case, dowry for women is linked to whether or not they marry, while inheritance for men is not. Further, except in the (matrilineal and bilateral) communities noted, the content of dowry has almost always been in the form of movables, while inheritance by men includes immovables, if the family has any.2+ Moreover

even movable dowry is usually not entirely a woman’s own or in her control, although here again there is regional variation. In south India, dowry has tended to be more commonly in the woman’s control,®> as it has

among some of the Tibeto-Burman communities in northern Nepal: the Kham Magars, for instance, draw up dowry deeds in the daughter’s name (Molnar 1981). However, in most parts of northern India a portion of the woman’s dowry is customarily taken away by the parents-in-law,®° who may then use it themselves or give it as dowry to their own daughters (the groom’s sisters). In fact, dowry in cash and goods may even be handed over by the bride’s father directly to the groom and his family, becoming a part of their joint estate.27 A woman’s jewellery more typically remains with her, but again (especially in north India) it is usually not hers to dispose of as she 83 Both the giving and taking of dowry have been illegal since 1961 in India and since 1980 in Bangladesh, although in practice the prohibition is widely disregarded in both countries.

84 Within patrilineal communities the exceptions to this (that is, cases of daughters getting immovables in dowry) are few and largely confined to south India (for examples see chapter 6).

85 But even here this is not guaranteed. Ram (1991: 196) in her study of the Mukkuvars in Kanyakumari district (Tamil Nadu) found that in 42 per cent of her sample households the bride’s in-laws had appropriated some part of the girl’s dowry (especially jewellery). In 16 per cent of the households it was used to finance marriages in the husband's family. 86 See e.g. Hooja (1969), Khare (1972), MacDorman (1987), Madan (1989), Miller (1981), Minturn and Hitchcock (1966), and Sharma (1984), for examples from India. Bennett (1979) notes that this is also observed among the Brahmins and Chetris of Nepal, although usually not among other Nepalese communities. 87 Vatuk (1975) found this among the Gaur Brahmins in Uttar Pradesh.

Customary rights and associated practices 137 wishes and it can be drawn upon by the family during times of economic need. What remains may be gifted by her to her daughter or daughter-inlaw. In other words, Goody’s assumption that dowry usually goes to the bride herself and serves to establish some form of conjugal fund which ‘ensures her support in widowhood ... and eventually goes to provide for her sons and daughters’ (Goody 1976: 6) is not a valid generalization, and Is contradicted especially by the north Indian evidence. In fact the part of dowry that passes to the girl’s in-laws cannot really be seen as a form of ‘devolution’ since it does not get transmitted vertically as inheritance to the couple’s children, but rather takes the form of a ‘circulating fund’ that moves from the bride’s family to the groom’s family, within the endogamous group.®®

In rare cases, both dowry and female inheritance in land have been practised by the same community and even family, as we noted was customary among most classes of Kandyan Sinhalese in Sri Lanka. Here daughters marrying diga could receive a dowry (sometimes even in immov-

ables) but forfeited their rights to inherit from the father’s estate, while those marrying binna could inherit patrimonial property, including land, on the same basis as their brothers and unmarried sisters. But a diga married daughter, by subsequently returning to her parental home and converting her marriage to binna, or contracting a subsequent binna marriage, or keeping in close touch with and rendering assistance to her . parents, could reclaim her rights to a share in ancestral property, indicating

that the dowry she had received was not regarded by the community as equivalent to inheritance. Nor was there any necessary equivalence between the value of dowry received by a diga married daughter and the inheritance shares of other children (Obeyesekere 1967, Yalman 1967, Tambiah 1973). Related to this last point is the absence of any clear rules concerning what share of the family wealth should be given as dowry, unlike the usually clear specification of rules governing inheritance in the family estate. There was a

suggestion in some early shastric literature that if the father died leaving behind unmarried daughters, brothers should set aside for each sister’s marriage an amount equal to one-fourth share, but the language was vague and open to varying interpretations. Altekar (1956) persuasively argues that the interpretation probably intended was that the marriage expenditure for each sister should be equal to one-fourth the share of a brother, but that this specification was meant to ensure an adequate provision for the sister’s marriage and was not to be taken literally. In any case, marriage

expenses cover many things, of which dowry could be but one. The empirical evidence on South Asia for recent decades suggests that the amount of dowry given to daughters among patrilineal groups has no 88 J am grateful to Patricia Uberoi for calling my attention to this last point.

138 A field of one’s own obvious relation to the shares of sons in ancestral property on partition, and depends on a range of factors such as the economic situation of the bride’s family, the affluence and qualifications of the bridegroom, the social

status of his family, the physical attractiveness of the bride, the marriage alliances contracted by the bride’s sisters (a high status marriage facilitated by a large dowry for the eldest daughter can help younger ones contract prestigious marriages for much smaller amounts), and so on.®? Unfortunately there is a dearth of studies which quantitatively compare the values of dowries with inheritance shares, but Schuler’s calculations for the Chumik in Nepal are revealing: she found (1987: 103) that for families with children of both sexes, the value of daughters’ dowries on average came to 10 per cent of the family’s assets, while the sons’ inheritance shares came to 90 per cent. Goody does not deal with the relative values of dowry and inheritance, or with the rules governing either, but clearly these issues have a bearing on

the nature of rights embodied in the two forms of transfers and on their ~ economic importance, and so impinge on the more general issue of gender equality that centrally concerns us here. All said, therefore, except among a few communities such as the Nangudi Vellalars, the matrilineal Moors and the Jaffna Tamils, inheritance and dowry cannot be equated in legal or economic terms; and movables (except

possibly some heirlooms) that are transferred through either practice cannot be equated in economic or symbolic terms with the transfer of landed property. Communities which give daughters dowries in movables but exclude them from inheritance therefore need to be distinguished from those which recognize the daughter’s inheritance rights, especially in land, a distinction which, as noted, Goody does not make in his 1973 and 1976 writings, and continues to downplay in his recent 1990 work.?°

Third, in Goody’s formulation the relationships between property transmission and marriage practices are set out as unidirectional, causal

ones. Diverging devolution, he argues, will tend to lead to close-kin marriages, endogamy (marriage within the same kindred, clan, community, or caste), monogamy, and the prohibition of pre-marital sex (Goody 1976: _ chapters 2-3). However, in his empirical work, Goody does not examine 89 See especially, Madan (1989), Pocock (1972), and U. Sharma (1980). 9° For instance, Goody (1990: 287-8) recognizes that there are regional variations across South Asia in women’s access to land, but still maintains that societies which give dowry and those which allow women to inherit land are part of the same overall pattern: {While throughout South Asia women are endowed with property, in Sri Lanka this claim extends to landed property, representing the fuller working out of the ‘bilateral’ division of wealth, that is, the diverging devolution which maintains the status of females as well as males in a hierarchical society. Also see chapter 6 on my differences with Goody on what can be inferred from the ancient practice of putrikaputra (‘appointed’ daughter) mentioned earlier in this chapter.

Customary rights and associated practices 139 other factors which could also influence his dependent variables. For instance, religious and cultural ideologies are known to play a significant

role in determining marriage partner preferences: most north Indian Hindus taboo close-kin (cross-cousin, uncle—niece, etc.) marriages, a rule

which is not broken among such communities even in choosing an uxorilocal son-in-law for brotherless daughters who may inherit the parental estate. Similarly many communities prefer and practise close-kin

marriages, even when they do not endow women with land or other property: as we will see in chapter 8, close-kin marriage preference is geographically much more widespread than is female inheritance in land.?! This is not to argue that we would expect no correlation between women’s

land rights and close-kin marriages. In fact (as I and some others have argued) such marriages would be an important way of keeping landed property within the kin group. Hence we would expect to find a positive correlation between the two variables. But this tells us little about the possible direction of the link. It may well be that close-kin marriage is the independent variable and the traditional recognition of female inheritance in land the dependent one, in that communities which (for one reason or another) practise close-kin marriages would be more open to endowing daughters with property than those which don’t. Fourth, I see no necessary reason to expect societies that transmit land to women to Insist on pre-marital virginity, as Goody suggests is likely to be the case. For instance, if a daughter’s rights in land were strictly conditional on her residence in the natal home, and she would lose her rights if she resided elsewhere, the family would not need to additionally restrict her sexually, merely to keep control over the land. I believe the reasons for particular sexual mores in South Asia need to be sought elsewhere, say in notions of caste purity, than in property concerns per se. My analysis departs from Goody’s formulation especially in four ways:

First, I focus specifically on societies which as a rule (and not just exceptionally) recognized women’s inheritance rights in land (whatever the transmission mechanism — dowry or post-mortem inheritance). Second, in this group of societies I include matrilineal ones, since they would be just as concerned as bilateral ones with preventing the loss of family land. The

Garos are included in this discussion even though arable land was not

customarily inheritable among them, because (a) other immovable property such as the ancestral house was inherited by Garo women; and (b) over time, land too became inheritable (as will be discussed in chapter 4). Third, I hypothesize that women’s rights in land are likely to be correlated °! In theory Goody (1976) recognizes that close-kin marriage could serve different functions

of which property considerations may be but one. However, this is not built into his statistical analysis.

140 A field of one’s own with marriages located near the natal home and with close kin, but this need not imply any simple causality. Fourth, I hypothesize that there need be no

necessary relationship between recognizing women’s rights in land and exercising control over women’s extra-marital relationships. Let us consider these hypotheses in the light of the case studies to see what they reveal about the relationship between female inheritance rights in land and the following three factors: post-marital residence, choice of marriage

partner, and sexual control over women. The major characteristics of the matrilineal and bilateral communities discussed above are summarized in table 3.1.

Post-marital residence. It is noteworthy that wherever women in South Asia have customarily had rights in landed property, it has been associated with their typically residing, and often having to reside, within the natal village and often in the natal home. Among the Garos, Khasis,

and Lalungs, the three main matrilineal tribes of northeast India, all daughters — the heiress who inherited ancestral property, as well as the others who had usufructuary land rights - were expected to reside near the maternal home after marriage, the heiress living in the ancestral home itself. And recent trends towards virilocality and neolocality among these com-

munities are associated with an erosion of women’s land rights. For instance, among the Garos and Khasis, as land scarcity has increased and non-heiress daughters are no longer allotted usufructuary plots, so also the rules of matrilocality are no longer binding, nor are they easily enforceable as the men seek to move out in search of other sources of livelihood. Again among those communities of southwest India which customarily practised matrilineal or bilateral inheritance, in most cases women continued to reside in their matrilineal joint family estate throughout their lives, including after marriage: the Nayars of central Kerala and the Mappilas of the Lakshadweep Islands practised duolocal residence, while the Mappilas of north Kerala and the Nangudi Vellalars of Tamil Nadu mainly practised matrilocality. Indeed, Tambiah (1973: 109) argues that among the Nangudi Vellalars it is matrilocal residence which ‘allows the concept of female property to include strong rights in land’. He further proposes (1973: 109): [T]he most general Indian pattern of dowry in movable wealth and of an intricate pattern of affinal prestations, etc. is associated with and occurs in conjunction with

the norm of patrilocality (or avunculocality). In contrast, the definition that a woman’s dowry right includes inheritance of land (and other patrimonial immovable wealth) which she shares equally with her brother can only occur in conjunction

with matrilocality or duolocality ... or ambilocality ... as the main form of residence.

Customary rights and associated practices 14] Table 3.1: Some characteristic features of matrilineal and bilateral communities in South Asia Pre-

Cross- marital

Post-marital cousin sex by Adultery Community Location: Inheritance residence marriage women by women India

Garos Meghalaya matrilineal matrilocal preferred accepted punished K hasis Meghalaya matrilineal matrilocal aversionto accepted punished and duolocal

Lalungs Assam matrilineal matrilocal aversion accepted n.i. today, tradi-

tionally . Nayars C.Kerala matrilineal duolocal * * * accepted

N. Kerala matrilineal avunculocal _ preferred n.i. punished and duolocal

Tiyyars N.Kerala matrilineal avunculocal preferred n.i. nl. and duolocal

Bants Karnataka = matrilineal avunculocal preferred n.1. n.i.

Mappilas N.Kerala matrilineal matrilocal preferred n.i. ni. Lakshadweep

Islands matrilineal duolocal preferred nt. n.1. Phadiyas N. Kerala matrilineal avunculocal accepted accepted _n.i.

Chettis N. Kerala matrilineal avunculocal accepted n.i. nl. Nangudi

Vellalars T.Nadu bilateral matrilocal —_ preferred nui. nt. Sri Lanka

Sinhalese Allregions __ bilateral uxorilocal __ preferred tolerated tolerated and virilocal

Tamils Jaffna bilateral uxorilocal preferred forbidden forbidden to ambilocal Eastern Province

(Mukkuvars) matrilineal miatrilocal n.i. ni. n.l. Moors Eastern

Province matrilineal matrilocal — preferred ni. ni.

142 A field of one’s own Table 3.1: (cont.) Divorced

Divorce women’s Widow

Community Location¢ by women remarriage remarriage India

Garos Meghalaya accepted accepted accepted Khasis Meghalaya common common common

Lalungs Assam n.1. nl. ni. Nayars C.Kerala * * *

N.Kerala accepted accepted accepted

Tiyyars N.Kerala accepted n.i. ni. Bants Karnataka rare rare accepted Mappilas N.Kerala rare n.i. accepted

Islands rare n.i. accepted Phadiyas N. Kerala rare ni. n.i. Lakshadweep

Chettis N. Kerala accepted n.i. n.l.

Vellalars T.Nadu ni. n.i. n.i. Nangudi

Sri Lanka

Sinhalese All regions common common common

Tamils Jaffna accepted rare rare (Mukkuvars) n.i ni. n.i. Eastern Province

Moors Eastern

Province accepted rare accepted

« State/provinces given here relate to present administrative units. * Not strictly applicable; n.i.= no information

These propositions appear valid, although the observation on avunculocality needs some qualification. For instance, among some of the groups we

have examined (such as the Bants of south Canara, and the Nayars, Tiyyars, Phadiyas, and Chettis of north Kerala), matrilineal inheritance in land was also found to coexist with avunculocal residence, that is, women

typically moved to their husband’s matrilineal estates after marriage. However, in these groups a woman’s rights in land took a specific form: neither she nor her children had inheritance rights in her husband’s matrilineal estate; her property rights (and those of her children) continued to be linked to her own matrilineal estate, managed and controlled by the

Customary rights and associated practices 143 seniormost male in her matrilineal household. She was away from that home for the duration of her marriage on the termination of which (due to widowhood or divorce) she would return to her maternal home, while the children could return even earlier. In other words, a woman did not lose her rights in her matrilineal estate by moving elsewhere on marriage, but the

land remained under her maternal family’s control during her period of absence. Occasional cases of matrilocality or duolocality were also found among these communities. The close correlation between female inheritance in land and norms of post-marital residence observed in India stands out even more strikingly in the Sri Lankan context. As noted, among the Sinhalese, a binna-married daughter, that is one continuing to reside in her natal home after marriage with her husband joining her there, had an equal right to her father’s estate as her brothers and unmarried sisters, but one residing virilocally (marrying diga) forfeited her inheritance. However, she could re-establish her claims if she subsequently remarried binna and assisted her father in his old age, just as a binna-married daughter could lose her claims by subsequently remarrying diga.

Among the Jaffna Tamils and the matrilineal Moors (as among the Nangudi Vellalar in India), daughters received land in dowry not as a privilege but as a right (as distinct from the occasional gifting of land to daughters in dowry by wealthy parents in south India). Among the Moors, ‘residence was matrilocal, while among the Jaffna Tamils it was customarily

uxorilocal to start with, and then ambilocal: where the plots of the two spouses were located in different villages, they would decide on residence location according to the ease of management of both properties. As noted, however, in general there was a very high incidence of marriages within the village.

All this broadly supports Tambiah’s (1973: 121-2) hypothesis (again with a qualifier on the issue of avunculocal residence) that: wherever there are strong patrilocal or avunculocal groupings (whether the descent system is patrilineal, or matrilineal, or bilineal), there women’s property, where it prevails, will be in movables only; where there are matrilocal-uxorilocal groupings, women will be invested with land, not necessarily to the exclusion of males (as among the Nangudi Vellalar). Where there are bilateral ambilocal local groupings and a free mixture of virilocal and uxorilocal residence, there both men and women

can plausibly be endowed with equal property rights in both movables and immovables.

It may be recalled that even in predominantly patrilocal groupings, if Hindu women inherited as residual heirs (that is in the absence of sons), uxorilocal residence was a pre-condition.

144 A field of one’s own Choice of marriage partner. Among most communities in India and Sri Lanka where a daughter’s land rights were customarily recognized, marriages with close kin, especially (one or both categories of) crosscousins, were preferred and promoted. The only exceptions to this among our case studies are the Khasis who strongly disapproved of such marriages, and the Lalungs who are averse to them today but are said to have practised cross-cousin marriages earlier. Of course the underlying basis of close-kin, including cross-cousin, marriage preference (or its absence) are multiple and complex. Although some of the considerable anthropological literature on the subject links a preference for close-kin marriage to the desire to keep property within the family, other studies give alternative or additional reasons, such as the ideology of descent prevailing in the community, the attempts by a social group to create and perpetuate certain types of solidarity ties and alliances, the desire to ensure purity of caste or status, or to ensure stable marriages between those with similar backgrounds, and so on.?* The south Indian Hindu pattern is often characterized as one where families seek to consolidate kinship ties by close-kin marriages, and the north Indian pattern as one

where families seek to extend their network of social ties over wider geographic spaces through marriage alliances with non-kin (for elaboration see chapter 8). Whatever the functional or other origins of close-kin marriage prefer-

ences, such marriages do have the effect of containing the dispersal of property outside the extended kin network: for instance, as Banks (1957: 260) notes, cross-cousin marriages, among other things, enable ‘the recombination of the properties of a man and his sister through their children’. We would thus expect communities which allow close-kin marriages to be less resistant to recognizing women’s rights in land than communities which don’t allow such marriages. Sexual control over women. The link between sexual control over women and their inheritance of landed property appears to be somewhat tenuous. On the one hand we have patrilineal Hindu communities in South

Asia which customarily recognized women’s inheritance rights only in highly restricted circumstances, but which placed a strong emphasis on pre-

marital virginity and post-marital chastity. And in the restricted contexts °2 Fora useful review of some of the theoretical debates on the subject see Fox (1967). Also see

Yalman (1962, 1967) on cross-cousin marriages in the context of south India and Sri Lanka, Tambiah’s (1965) critique of Yalman, Banks (1957) on the Jaffna Tamils, and Trautmann (1981). Tambiah (1965), Leach (1961), and Banks (1957), like Goody (1976), place emphasis on property considerations underlying marriage preferences. But unlike Goody, they confine themselves to communities which do not forbid cross-cousin marriages on other grounds.

Customary rights and associated practices 145 and forms in which they recognized women’s property rights (e.g. a limited estate for widows without sons), a variety of constraints were placed on the woman. For instance, widows were generally prohibited from remarrying, and 1f unchaste had to forfeit their rights in the property.

On the other hand, these and other sorts of limits on women’s sexual choices, which could be prescribed in order to keep property intact, were not a necessary feature of communities where women could customarily inherit land. In fact groups which recognized women’s land rights customarily (and not just in exceptional circumstances) exercised quite differing degrees of sexual control over women, as revealed in their varying emphasis on pre-marital virginity and chastity and tolerance of divorce and remarriage. Pre-marital sex was accepted among the matrilineal tribes of north-

east India (as it was among most tribal groups in India) and strongly disapproved of among the Jaffna Tamils (virginity being fully expected). Norms regarding extra-marital sexual relations likewise differed, disapproval being stronger among some groups than others. However, divorce and divorcee remarriage were relatively easy among most groups, although some such as the Jaffna Tamils and the matrilineal Moors allowed it in principle but rarely practised it. All the communities permitted widow remarriage. On the whole, therefore, the degree of emphasis on female chastity was highly variable across communities and appears to have had less to do with female inheritance practices than with notions of morality and/or purity of caste or race. The caste-purity issue is especially important in South Asia, where even communities which tolerated women’s extra-marital liaisons with same-caste or higher-caste men forbade them with lower-caste men. Here it appears pertinent to also consider Engels’ (1972) argument that ensuring the legitimacy of heirs would necessitate an emphasis on female

chastity and monogamy, although (as noted in chapter 1!) Engels was speaking of propertied households and not of propertied women. The case

of the Kandyan Sinhalese indicates that these links are not a given, but depend on how a society views the institution of marriage and what system of property devolution it follows. Among the Kandyan Sinhalese, whose marriage arrangements (as noted) were usually completely informal and the wedding ceremony was often dispensed with altogether, it was generally difficult to distinguish between wives and lovers and between ‘legitimate’ and ‘illegitimate’ children (Yalman 1967: 160). [legitimacy was related not

to children born out of wedlock, but to children born from unions considered inappropriate, such as those that violated caste or incest taboos.

Since male and female properties were separate, the children inherited separately from the mother and the father; and ‘whether the parents were ‘married’ or not [was] normally ... immaterial for inheritance claims’

146 A field of one’s own (Yalman 1967: 172). The formal association of legitimacy with wedlock was

a legal imposition on the community through British legislation in 1859, which made marriage registration compulsory and recognized only the children born from registered marriages as legitimate. In the early years

when few marriages were registered this led to the absurd situation commented on by the British Governor, ten years after the law was passed: It is probably within the mark to assume that 2/3 of the existing unions are illegal and that 4/5 of the rising generation born within the last 8 or 9 years are illegitimate. (Cited in Risseeuw 1988: 37)

In overview then, among the communities which customarily recognized women’s rights in landed property, families sought to keep the land within the purview of the extended kin either by strict rules against land alienation

by individuals, or, where such alienation was possible (as among the bilateral communities), by other means: these included post-marital residence in the village which often took the form of uxorilocality or matrilocality, and close-kin marriages, but did not usually include a rigid sexual control over women. In fact, proximity of the post-marital residence to the

natal home appears to have been virtually a necessary condition for recognizing a daughter’s share in landed property. Let us now turn to the second question posed earlier: what can we say from our case studies about the nature of gender relations among matrilineal and bilateral communities?

(2) Land rights and gender relations To assess gender relations in the communities described is not easy. Most

historical records and more recent accounts of matrilineal or bilateral communities (like accounts of patrilineal ones) give little space to the question of gender relations or to women’s own perceptions of their situations. Matriliny, in particular, has been described primarily in structural terms rather than in terms of people’s (especially women’s) everyday experience of it. Also, there is much more information on some groups,

such as the Nayars, Khasis, Garos, and Sinhalese, than on others. Nevertheless, some broad inferences can be drawn which, although not definitive, are certainly indicative.?? The picture we get is a mixed one. On the positive side, most historical accounts remark on the considerable independence of women in several of these communities, especially the central Kerala Nayars and Sinhalese (as °3 To supplement historical accounts I have also drawn on some recent ethnographies, where the descriptions do not appear to represent a recent change.

Customary rights and associated practices 147 indicated, for example, by the opening quotations in this chapter), and on the relative equality in marital relations among them. Descriptions of Garo and Khasi women also suggest this. Indeed, in all the communities studied, a daughter’s rights in land, and the fact that she either remained in her natal home after marriage or had inviolable rights to return to it if she so chose, provided her with a strong fall-back position within marriage. Women could choose their husbands (although heiresses faced some constraints) and initiate divorce. Where uxorilocality or matrilocality was the norm, as

it was in many of these groups, marital breakdown led to the husband departing, sometimes (as among the Garos) with only the clothes on his back. A binna-married Sinhalese woman wanting to terminate her marriage

merely had to place the husband’s personal belongings on the doorstep (Tilakaratne 1986). According to jocular folklore a binna husband was advised to always keep a walking stick, an umbrella, and an oil lamp handy in case he was evicted by his wife when he was unwell, in the rain, or in the middle of the night! (Yalman 1967). (In contrast, in patrilineal, patrilocal contexts, it was women (especially if they violated sexual norms) who faced

the very real risk of being evicted and being left with little means of support.) That the husband’s authority was likely to vary inversely to the material and other support a woman could get from her natal home is especially apparent in the Khasi case: here (as noted) marital instability was much higher among heiresses, who usually initiated the divorce, than among nonheiress daughters, who were more dependent on husbands for their material well-being.

Everyday marital relations also reflected the limited authority of a man married to a Khasi heiress. Nakane (1967: 125) found, for instance, that the last person to be introduced to visitors was usually the husband: When we visited the Khasi household of a youngest daughter [the heiress], if a man (obviously the husband) came first to greet us, he always said ‘please wait, my wife (or mother-in-law) is coming.’ And it was the wife who entertained us, offering tea or pan and discussing matters with us, while her husband remained silent in the corner of the room, or in the next room. If her uncle or brother was there, he would be the person who talked to us... In such a case, we were seldom introduced to the husband.?4

This is in marked contrast to the situation in upper-caste patrilineal rural households in northwest India where it 1s the woman who may be seen standing silently near the door in the adjacent room while the husband receives visitors. 94 My own observations were similar ina field visit to some Khasi villages in 1989: the women offered us pan and betelnut inside the hut while the husbands (where present) left the hut

and sat outside.

148 A field of one’s own Nakane also notes that a Garo woman rarely waited up for her husband if he was late for supper, but ate her meal and kept his share aside, once again in contrast to observed behaviour in the patrilineal northwest where the wife often eats last. To some extent, of course, the relative equality in marital relations that Nakane describes (and which I also observed) among the Khasis and Garos, would be a general feature of tribal communities, but

the contrast between the heiress and non-heiress daughters among the Khasis, suggests that the specific property rights enjoyed by the youngest daughter provided her additional bargaining power within marriage. Of course even where the husband did not play an authoritative role male

authority could prevail: for instance, among groups with joint family estates, the kinsman who managed the estate wielded considerable auth-

ority, as the karanavan did among the central Kerala Nayars and the wealthier north Kerala Mappilas. But among groups where women and men held property individually, as in Sri Lanka, there was more equal sharing of household authority and decision-making between the genders.

Gunawardena (1989), for instance, found in her fieldwork among the Kandyan Sinhalese in Kande village that neither dominance nor authority was vested solely with the husband. In fact, both spouses were reckoned as household heads. There was equal participation in decision-making on most matters, while fertility decisions were made solely by the women. In her interviews in which both husband and wife were present, she found that ‘rather than absolute agreement with the ideas expressed by their husbands, most women tended to challenge, interject, qualify, add to and contradict statements made by their husbands, and clarify an issue by saying: “wait, that’s not exactly the way it was. Now don’t mislead her (directed to the husband). Let me explain (to the anthropologist)’ (p. 114, insertions as in the original). Also, among many of the communities where women contributed to the household economy through agricultural work and trading (as among the

Sinhalese and the tribal matrilineal groups of northeast India), they typically controlled the crop surplus and their own cash earnings. In all the groups, women had considerable freedom of movement and of public interaction. On the Sinhalese in Kande village, Gunawardena (1989: 306) comments: ‘I have observed women who simply announce to their

husbands their plans for a journey be it far or near and simply leave (without waiting for assent, acknowledgement or permission). Men do likewise.’ Again, among the matrilineal (Muslim) Moors: ‘Women move

freely about the village without veils covering their faces. It is not uncommon to witness men and women disputing or conversing in public places ... Interaction between men and women is frequent and casual and

Customary rights and associated practices 149 often sexual comments are exchanged publicly’ (Munck 1985: 8, 108). This is in striking contrast to women’s situation among most patrilineal Muslims of the subcontinent. Similarly, Nayar women ‘could move freely about the village, [and] travel alone on pilgrimages ...’ (Gough 196la: 356). British

anthropologist, C.F. Fuller, found that Nayar village women were never embarrassed to answer his questions directly, or to take active part in his conversations with the family (personal communication, 1992). This freedom of movement and public interaction would also have made for greater awareness of public affairs among women. I found in my 1989 field visit (along with a Khasi woman sociologist) to some rural Khasi households near Shillong that many village women were well aware of the intricacies of local politics. On one occasion, forgetting our presence, two of the women we were visiting had a long and involved argument about the suitability or otherwise of particular candidates in the forthcoming local elections! This awareness probably took other forms in earlier times. Norms of sexual behaviour outside marriage, however, varied considerably, ranging from relatively gender egalitarian (as among the matrilineal

tribes of northeast India,°> and among the Sinhalese), to restricted for women (as among the Jaffna Tamils). But in comparison with the Hindu and Muslim women of patrilineal groups, especially those shackled by seclusion practices in northern South Asia, women among all matrilineal and bilateral groups enjoyed greater sexual freedom: social interaction between the sexes was relatively free, divorce and widow remarriage was permitted among all, and even among the stricter groups punishments for extra-marital affairs did not take the extreme and violent forms that they could take among some patrilineal groups, especially those of the north-

western frontier regions of the subcontinent (see chapters 6 and 7 for details).

Women’s status in these contexts also affected attitudes towards girl

children, who were specially desired among matrilineal groups. For instance, recently, when recording life histories of older Nayar women, anthropologist Joan Mencher found that many of them mentioned how in

the past (even forty to fifty years ago) Nayar mothers got angry with daughters who bore male children: “They wanted girl babies to increase the

number in the house’,?® in marked contrast to the strong preference for male children found in patrilineal communities across much of South Asia.

Relationships between women of the taravad also appear to have been supportive and close. Among Khasi households too, on my field visit I was °> The Garos (as noted) customarily had a well-defined code of behaviour towards women under which any form of sexual harassment of women was subject to punishment. °6 Personal communication, Joan Mencher, 1992.

150 A field of one’s own struck by the closeness and strong support networks that appeared to exist among married sisters. However, the features which favoured women in these communities were counterbalanced by other, less favourable, ones. One, property rights did not alter the overall gender division of labour: domestic work and childcare were still a woman’s responsibilities. Two, the range of sexual mores found among these communities indicates that rights in land did not guarantee

women the same sexual freedom as men. Three, formal managerial authority over land ina number of matrilineal communities lay mainly with men (as husbands, brothers, and maternal uncles). In practice, this would have worked in various ways, depending on the role women played in the household’s economy, the form (individual or joint) in which property was held, and the size of the estates involved. Where women’s role in production

and market activities was important (as among the tribal groups of northeast India), or where women held individual property rights (as among the bilateral groups in Sri Lanka), they also exercised a degree of control over the land and its management. Even among the Khasis who held property in joint family estates, although the brother or maternal uncle

was an important authority figure in estate management, it is mainly women (who were also involved in farming and trading) who appear to

have decided (albeit in close consultation with maternal uncles and brothers) how ancestral property would be divided among daughters, or what part of it could be leased out. The shallow generational depth of most Khasi households would also have facilitated this process. However, where women played little role in farm production, and property was held in large joint family estates collectively owned by several generations of a woman’s

matrilineal descendants, as among the Nayars of central Kerala and the wealthy Mappilas of north Kerala, men’s managerial control over property and their overall authority appears to have been especially strong. Here, as noted, it was the karanavan who managed the estate and was the representative authority in the taravad’s public dealings. Among the central Kerala

Nayars, for instance, although older women exercised control over domestic affairs, and wielded moral authority and influence over younger women, sons, and younger brothers, this was not the same as the overall authority and managerial control over property exercised by the karana-

van. And although some individual Nayar women may have played important managerial roles by virtue of circumstance, age or personality,

the system did not provide for this as a rule. (This also highlights an important difference between matrilineal and patrilineal inheritance systems: in the former there 1s often a gender divergence between property ownership and its control, while in the latter there is none: men (as a gender) both own and control the property.)

Customary rights and associated practices 151 Four, most importantly, customary institutions with jural power (such as the tribal and caste councils) were monopolized by men, not only in all the matrilineal communities, but also in the bilateral ones in Sri Lanka. In the

latter, although land rights were defined on an individual and not joint family basis, and women could exercise greater control over land manage-

ment, jural power and authority in public decision-making bodies still , rested with men as a gender (even if not with all men as individuals). As noted, women as a gender were typically excluded from such bodies. Among matrilineally inheriting communities, this meant that despite men’s restricted access to property ownership, their rights (as a gender) of control over that property on the one hand, and their access to public bodies on the

other (with links between the two domains), often enabled them to consolidate social prestige and political power. The Nayar karanavans of wealthy taravads and the Khasi chiefs commanded considerable local influence in ways that the women heiresses of these communities appear not to have done as arule. Also, among all groups, men’s control of the public decision-making domain gave them critical influence over the modification

of legal and social rules when external conditions began to change in significant ways, especially under British colonial rule.

In short, although their rights in landed property clearly conferred important benefits on women, their virtual exclusion from property management (1n some groups) and from jural and overall public authority (in all groups) circumscribed the power they could derive from those rights. This suggests that we cannot predict the full advantages of women having

land rights for gender relations merely by examining women’s position within customarily matrilineal or bilateral communities. For women to derive fuller benefits from land ownership they would need to enjoy managerial control over it as well. Moreover, women’s exercise of effective control over property is likely to be interlinked in a two-way process with

their entry into public decision-making bodies (as will be elaborated in chapter 10).

Our analysis of the detailed case studies of matrilineal and bilateral communities thus yields at least two notable insights. First, there was a significant link between women’s property rights in land and certain marriage customs which reduced the risk of the land passing to non-kin. And today, where there is an absence of congruence between these two factors, granting women possession of land is likely to be strongly resisted. Traditionally, localized communities played a significant role in the evolution and enforcement of both inheritance and marriage customs. Hence it would have been possible to ensure congruence between the two sets of practices. However, contemporary laws, as framed by the modern State,

152 A field of one’s own

recognize the inheritance rights of females as individuals among most communities, including traditionally patrilineal, patrilocal ones. The enforcement of these laws is also in the domain of the State. But marriage customs are still under the purview of the local community and, on the relevant counts, have remained largely unchanged. In other words, today the interface is between individual property rights granted under contemporary law by a State which ts relatively autonomous of localized kin networks, and the customs of marriage and residence which are still locally governed in considerable degree by those networks. Our observations would suggest that this contemporary recognition of female inheritance in land, where it was not recognized earlier, is likely to produce much greater conflict and opposition among communities which forbid both close-kin marriage and village endogamy, than among those

which allow both. As will be shown in chapter 8, most south Indian communities permit both practices, but most north Indian ones forbid marriages with close relatives and insist on village exogamy — conditions

under which any land inherited by a daughter could be lost to her natal family. This lays the ground for considerably greater opposition in north India than in south India to giving women the property rights promised to them by contemporary law. Second, given the noted limitations of being vested with ownership rights

in landed property while being excluded from managerial and jural authority, the arenas of contestation over land rights for women extend, as

noted in chapter 2, much beyond the courtyards of the household to encompass the complex institutions of community and State — the arenas where legal, social, and political rules are made and unmade. The next two chapters illustrate this further.

4 Erosion and disinheritance: traditionally matrilineal and bilateral communities

Apart from being a ‘more wanted fellow’, the [Nayar] husband now also feels that it is ‘in keeping with his dignity’ that he be the one to run things for his wife, and hischildren . .. Today the wife is expected to be faithful, to

look after her ‘husband’s needs’ and to listen to him, though she can oppose his decisions if she feels they are not in the best interest of her

tavari. (Mencher 1962: 241)

Matrilineal and bilateral systems of land inheritance advantaged women in many respects, especially in granting them economic and social security,

and considerable autonomy and equality in marital relations. These systems, however, did not remain fixed over time. Interventions by the colonial and post-colonial States, particularly in the legal and economic spheres, and the complex processes of social and cultural change which these set in motion, eroded customary practices. The large joint family estates came to be partitioned; formerly egalitarian tribal societies grew economically differentiated; there was an increasing penetration of market forces; there were notable shifts in the techniques of production, the social division of labour, and land relations; sexual mores altered; and patriarchal

ideologies spread in influence. Women, in particular, were profoundly affected by these changes. At the same time, their customary exclusion from

major authority in public bodies meant that they were unlikely to be the ones directing the change, and were seldom in a position to effectively protect their interests. This chapter will examine these gender effects by looking at the historical processes of change in three of the main matrilineal and bilateral communi-

ties in South Asia for which adequate information is also available: the Garos of northeast India, the Nayars of southwest India, and the Sinhalese of Sri Lanka. Although in terms of their populations the Garos and Nayars constitute but tiny dots in the largely patrilineal landscape of India, their

experiences, as well as those of the Sinhalese, are illuminating in many important respects. In particular they show that the ways in which the changes affected women were not always the unintended consequences of State policies, but that many of those policies embodied and promoted a view of gender relations that was deeply steeped in patriarchal mores. 153

154 A field of one’s own The Garo case study also demonstrates the limitations of Ester Boserup’s

(1970) influential argument that it is population growth and the resultant

land scarcity which primarily explains shifts from swidden to settled cultivation, the associated decline in women’s role in agriculture, and women’s loss of land rights in Asia and Africa. Especially in South Asia, the role of the State is seen to have been primary in determining the nature and direction of change.

An examination of the three communities mentioned also helps us to focus on the three regions in South Asia where matriliny or bilaterality has

traditionally prevailed. The experiences of change in matrilineal and bilateral communities other than these were not identical, but the direction of the shifts, and the types of factors that led to them, appear to have been similar.

I. India (1) The Garos The Garos in India today reside in what is now the state of Meghalaya, which has five administrative divisions (districts): East and West Garo

Hills, East and West Khasi Hills, and the Jaintia Hills.1 These are administered by a three-tiered system — the state government, District Councils (one for each of the five districts), and what remains of traditional

tribal institutions — each tier creating pulls and pressures in different directions. In 1981, the state’s population was 1,336,000, representing an almost four-fold increase since 1901. Thirty-eight per cent of it was located in the Garo Hills and 62 per cent in the Khasi-Jaintia Hills; 85.0 per cent was

rural, 80.5 per cent tribal, and 50 per cent Christian (GOI 1986b). At the turn of the century, the Garo Hills were covered with thick forests swarming with wildlife, and were said to have the richest soils in the region. Garo society, as described in the previous chapter, was characterized by economic self-sufficiency and considerable class and gender equality. Jhum cultivation yielded enough for subsistence. Land was communally held, but women inherited all other property. Overall they occupied a position of considerable economic and social strength: they had independent usufructuary rights to communal land, enjoyed the security of matrilocal residence,

played a primary role in agricultural production, controlled cash and agricultural surplus, and enjoyed a considerable degree of sexual freedom. Jural authority, however, was held by men. ' The discussion below is confined to the Garos in India. For information on changes in traditional practices among the Garos in Bangladesh, see Khaleque (1983, 1984). The directions of these changes were similar to those noted for the Indian Garos.

Erosion and disinheritance 155 The establishment of British control over the Garo Hills in the second half of the nineteenth century led to substantial changes in the economic and social fabric of the community, which were carried further by postIndependence policies. Many of these changes adversely affected the community in general and women in particular.

Shifts from jhum to wet rice cultivation. The most significant change, which in turn triggered others, was the shift away from almost exclusively jhum cultivation towards wet rice settled agriculture. The importance of fruit and nut orchards also grew. The adoption of wet rice farming was induced by a combination of factors which made it difficult to survive on jhum alone: (a) the shrinkage of area available for jhum asa result of State policies; (b) a growing population impinging on this shrinking land base; and (c) the direct encouragement of wet rice cultivation by the State. Although the Garos must have known of wet rice techniques since at least the eighteenth century (and possibly earlier) through their contact with the

plains people via markets situated in the foothills, virtually none were practising settled agriculture even by the early twentieth century (Playfair 1909).

However, beginning in the last two decades of the nineteenth century, the colonial State’s land control and land use policies began to make significant

inroads into jhum land. The British created large tracts of forest reserves from land belonging to different a’kings (settlements), and declared the remaining forest areas as unclassed State forests.* With the Garo Hills Regulation Act of 1882, they also acquired the power to regulate the use of forest resources by instituting a system of licences, fees, royalties, and so on. The land of some a‘kings was declared as khas (i.e. land not under private

ownership and subject to the government’s right of disposal), and that of 2 The changes have been traced primarily from village studies by Burling (1963) and Nakane (1967) in the early 1950s, Majumdar (1978) in the 1950s and mid-1960s, and Kar (1982) in the mid-1970s. This has been supplemented by less detailed village surveys undertaken by

the Anthropological Survey of India, the Agro-Economic Research Center (Jorhat) and

the Census of India Village Monograph Series 1961, as well as by my field trip to Meghalaya in 1989, during which | visited a number of Garo villages, spoke to local government officials, and had a useful discussion with D.N.Majumdar, who has known the Garos intimately since the 1950s: during 1953-62, when he did his initial survey in the area, Prof Majumdar (who is married to a Garo) was So well accepted by the community that his opinion was even sought in settling inheritance disputes. 3 Unclassed State forests are those for which specific rules and regulations regarding the use rights of people have not yet been specified, but over which the State retains the option of

defining such rights and of converting these forests into Reserved forests if deemed necessary. In Reserved forests the use rights of local dwellers are clearly defined and highly restricted. Through their takeover of the forests, the British deprived the local population of its control over this valuable resource, and also of most of its customary rights in forests,

and established their own claims to undertake unrestricted timber extraction.

156 A field of one’s own some others was forcibly divided and gifted to selected beneficiaries (Kar 1982: 238), although the colonizers refrained from settling non-tribals in the area, which protected the Garos from significant encroachments by outsiders. From the 1940s onwards the area under arecanut orchards also began to expand, further reducing that available for jhum (Kar 1982: 146). But in the early 1950s these were still tendencies. Burling (1963: 28) described the situation thus: ‘[L]and has been left fallow for ever shorter periods, and presumably it is becoming less capable of controlled cultivation. This may result in a crisis in the future but the crisis 1s not yet there.’ However, after 1947 with post-Independence policies the shrinkage of jhum land gathered

| momentum. A Garo Hills (Autonomous) District Council was set up in 1952, with powers to legislate on matters relating to land use and land revenue, inheritance, marriage and divorce, the management of unclassed State forests, the regulation of jhum, and village and town administration. Retaining the basic features of forest administration inherited from the British, the District Council reserved the right to give permits to selected

individuals for exploiting forest resources, and allowed Garo aking | members free access to forest produce only for domestic use. The Council itself undertook large-scale tree felling for revenue purposes. Majumdar (1978: 117) notes: ‘Practically anarchy prevailed in the forests managed by the District Council, resulting in indiscriminate felling of trees without any

serious effort at regeneration of forests’. This in turn contributed to environmental degradation. Although various parts of the forests were placed under plantation/regeneration/afforestation schemes, actual replanting was limited and mainly commercial species were grown to which a’king members had no access. In addition, the Jhum Regulation Act of 1954 (which is still in force) restricted shifting cultivation near water sources

and catchments, in areas identified as containing valuable timber, and in village or community forests. In 1963 the District Council’s land reforms branch also declared several categories of land as khas, including a good deal of land used for jhum. Added to this was the state government’s largescale land acquisition for setting up its administrative infrastructure. As a result of these policies, a growing population (increasing from 160,000 in 1901 to 307,000 in 1961) was forced to survive on shrinking tracts

of jhum land, leading to ever-shortening jhum cycles and declining land productivity, to which soil erosion and the drying up of many perennial streams also contributed. By the 1960s a crisis of subsistence was developing. A 1963-64 study in Banshidua village showed that the average annual per capita income of households depending solely on jhum was Rs. 168.2, which was below the officially defined all-India poverty line of Rs.180 at 1960-61 prices. Wet rice settled agriculture was clearly more rewarding: in

Erosion and disinheritance 157 some villages, it yielded four times more income per capita than jhum. In addition, it was being directly promoted by the administration. Under its jhum control scheme, the state government developed land for permanent cultivation, especially by employing local Garos to make terraces. Also the District Council passed laws providing that if local villagers did not take advantage of suitable land, others (even non-Garos) could convert it into paddy fields (Burling 1963: 305). The result was that substantial shifts to wet rice took place in the first decade and a half after Independence. By the mid-1960s many Garos had almost entirely switched to wet rice cultivation, practising jhum on the

margin (as in Wajadagiri village in Majumdar’s study); others were practising jhum along with wet rice cultivation (on flat lands or terraces) and orchards (GOI 1967a; Saha and Barkataky 1968). However, the shifts were constrained by shortages of flat lands and irrigation water, the drying up of perennial streams in some areas, and the need for additional labour to

construct and maintain hill terraces, which not all households could provide. In several areas, even the terraces created by the state government were abandoned after initial use (Saikia and Borah 1979). Hence not all

who wished to do so could shift to wet rice and orchards, and many continued practising only jhum. In 1981, an estimated 35.1 per cent of the total population (half of it female) in the Garo Hills was dependent on jhum as the only form of cultivation, which, as noted, was insufficient for survival and needed supplementing by income from other sources, such as trading in

forest products, working as agricultural wage labour, or seeking nonagricultural incomes.*

Increasing economic inequalities. Variations in the extent of shift

out of jhum into diversified production led to two forms of economic differentiation in the Garo Hills: inter-village, and inter-household within

villages. Inter-village variations stemmed essentially from differential availability of irrigable flat land for settled rice cultivation. In some villages, such as Wajadagir1, virtually all households were able to shift to wet rice

while practising some jhum alongside. Such villages remained relatively self-sufficient in food production.* Other villages, such as Machakolgiri, remained entirely dependent on jhum cultivation and became increasingly + In Agalgri (a settlement cluster of three villages), Saha’s (1976) study in the late 1960s found

that 40 per cent of the settlement income’came from non-agricultural sources, 57 per cent from jhum, and only 3 per cent from other sources, in contrast to Banshidua, where 44 per cent of village income came from settled farming and horticulture. * Rinsibara village studied by Nakane (1967) in the early 1950s, and Resubakrapara studied under the Census of India 1961 village survey project (GOI 1967a), were very similar to Wajadagir1 in this respect, in that in these too almost all families had adopted wet rice along with jhun.

158 A field of one’s own impoverished as a result, In both these categories of villages, intra-village economic differentiation between households remained low. A third category of villages was that where only some households, with State encoura-

gement and support, opened up wet fields or orchards, while others continued to depend on jhum alone. Here economic inequalities between households practising settled rice cultivation and those subsisting only on jhum were noticeable (Nakane 1967; Kar 1982: 197). Families that opened up wet rice plots beyond their capacity to cultivate with family labor alone, either sharecropped out some land or employed wage labour. Wage labour, which was just beginning to appear among the Garos in the 1950s (Burling 1963), was common by the mid-1960s and widespread by the 1970s (Saikia and Borah 1979, Lahiri 1979). In 1981, there were over 20,000 agricultural labourers in the east and west Garo Hills combined; of these, about 38 per cent were women dependent primarily on agricultural wages for a livelihood (GOI 1986b). For many others this was a

secondary source of income. Inter-village labour migration became common, reciprocal labour arrangements declined, and indebtedness to moneylenders increased (Kar 1982; GOI 1967a). Changes in the gender division of labour. Wet rice cultivation also led to significant changes in the gender division of labour. Where practised as in the plains, it required substantial changes in agricultural techniques

and practices, the timing of operations, and labour organization. Land preparation for wet rice needs more care than for jhum rice since ploughing and levelling are often necessary, requiring new skills and draft animals. Irrigation ditches and dikes have to be built around fields located near an

assured water supply (e.g. perennial streams); the crop usually has to be transplanted (and not merely sown with the dibble stick method as in jhum); its harvesting requires a sickle; and threshing has to be done with cattle. The

timings of planting and harvesting raise peak labour needs and require much more coordination among family labour, while reducing the need for labour cooperation between villagers (which is essential in jhum, especially for clearing forests). In themselves these changes need not reduce women’s

contribution to production, but they have tended to do so since the new skills are taught mainly to men. Women’s role, which was primary in jhum, has thus been made secondary

under permanent cultivation with the plough. In operations other than transplanting (which uses mainly female labour), women have been reduced to being ‘helpers’ to men. This includes harvesting, an operation in which, under jhum, men were women’s ‘helpers’. The decline in jhum and forests has also reduced women’s access to the cash they earned from selling produce gathered from the uncleared jungle

and jhum land. The proceeds from such small trading had belonged

Erosion and disinheritance 159 exclusively to the women. Now men control the crop and the cash generated from wet rice.

Privatization of communal land. The spread of wet rice cultivation

and orchards, coupled with direct measures by the District Council and state government, eroded the communal pattern of Garo land ownership and led to its increasing privatization in individual hands. The District Council’s power to regroup existing villages into larger ones for so-called stability, and for implementing development schemes, disrupted the communal basis of land ownership, and contributed to privatization. But more significantly, wet paddy settled agriculture involves cultivating the same field year after year. Customarily, as with the traditional recognition of individual rights over planted trees, the continuous use of a plot (along with

frequent cash and credit transactions relating to cash crops) bestowed a public sanction on the concerned household’s possession of the land. Similarly, over time, the establishment of orchards gave the household rights not only over the trees but effectively also over the land on which they grew. This could cover substantial tracts: for instance, in Darengri village, 40 per cent of a’king land was brought under private possession by eleven out of the fifty-three households through arecanut gardening (Kar 1982).

This customary privilege acquired legality when the British, and after Independence the state government, began to grant pattas (titles) to individuals, and so accelerated the shift to settled cultivation and privatiza-

tion. This policy continues today. Initially only temporary pattas are granted, under which land can be inherited in the usual way but not sold, nor can trees on it be cut; and such lands can be acquired by the government without compensation for road building. However, after some years a patta can be made permanent, giving the owner the right of free disposal. The pattas are typically granted in male names. In Wajadagiri, Majumdar (1978) found that out of twenty-three households, fourteen got pattas. Of these, eleven were granted to men, one to a widow, and two to unmarried

daughters. Although there are no published records of the aggregate number of pattas granted by the District Council or the gender of the grantees, my discussions with District Council officials in 1989 indicate that

a bias in favour of males is the common pattern. With increasing land scarcity, the development of a cash economy, and the growing tendency of

parents to gift land to sons, the absence of patta ownership in women’s names is likely to be a significant factor in the long-term erosion of women’s

land rights in the community.

Erosion of matrilineal inheritance practices. With privatization, land has become part of family property — indeed, over time, it has replaced

heirlooms as the most important form of property. In the 1950s, when wet

160 A field of one’s own

rice cultivation was still a recent innovation in the Garo Hills, Burling found that parents with a modest holding usually gave the bulk or even the whole of the land to the heiress-daughter to prevent land fragmentation. Those with extensive holdings sometimes gave small plots to non-heiress

daughters and even to sons, although for most part, this practice was restrained: ‘Garos from wet rice areas expressed to me the sentiment that it was much better to keep the land together because if it split up, everyone would become poor’ (Burling 1963: 308). Other children were advised to open up new land elsewhere or to work hard and purchase some. Majumdar’s observations in the 1960s and Kar’s in the 1970s indicate that till then the basic features of matrilineal descent and inheritance had not changed substantially and ‘lands held under annual patta in the name of the household head are for all practical purposes kept within the orbit of

{the] traditional descent pattern’ (Kar 1982: 249). However, there are clearly pulls in the opposite direction now, with parents increasingly favouring land transfers to sons. The extent to which this is occurring appears to be linked to the following factors: (a) the relative prosperity and overall economic self-sufficiency of the village; (b) the degree of economic differentiation between households within the villages; and (c) the extent of decrease in women’s field labour participation associated with the degree of shift away from jhum. The erosion of matrilineal inheritance and uxorilocal post-marital residence have been least in villages which are still relatively

self-sufficient, have low inter-household economic differentiation, and where the decline in women’s labour contribution has been limited. In Wajadagiri, for instance, village self-sufficiency continues to be high, economic differentiation between households is low, and most villagers practise both settled wet rice cultivation and jhum. Here the families have been able to maintain and indeed strengthen traditional ties of marriage, kinship, and inheritance. Men marrying heiresses have a particular stake in ensuring inheritance along traditional lines, and residents of Wajadagiri have no difficulty in finding sons-in-law from other villages. The village elders said to Majumdar (1978: 130-1) that if sons inherited, ‘the property would pass to another mahari. This is not desirable. If we give land to our sons, our daughters who are the actual owners of land will have to move

away.’ Again, in Resubakrapara (where too all households practised a combination of jhum and wet rice) women alone inherited land among the

landowning households (GOI 1967a). Nakane likewise noted that in Rinsibara village, despite the adoption of wet rice, the change in social organization was very limited, and attributed this to the low economic differentiation among its households. In my view, there are two additional factors of significance, shared by all three villages mentioned: their relative economic prosperity (which makes

Erosion and disinheritance 161 them attractive to in-marrying families) and the fact that jhum cultivation coexists with wet paddy, which means that the role of women in cultivation continues to be visibly significant, leaving less justification for by-passing women’s traditional rights. That low intra-village economic differentiation alone cannot sustain traditional forms of marriage and inheritance is also indicated by the case of Matchakolgiri, where the entire village economy is dependent only on jhum and has become increasingly impoverished. Such villages are unable to attract sons-in-law easily to maintain matrilocal/

uxorilocal marriage patterns, and both men and women have to look outside to non-agricultural sources of income. Villages such as Joshipara and Darengri provide a notable contrast to

Wajadagiri. In Joshipara, inter-household differentials are high. Some families remain dependent solely on jhum while others have shifted exclusively to settled rice. Among the latter, parents want to bequeath some private land to sons, since sons become familiar with wet rice cultivation techniques by working with their fathers in the fields, while the nokrom son-

in-law (heiress’ husband) may come from among the jhum cultivators of another village and be unacquainted with wet rice cultivation. Also in such households, women’s labour input has declined, both because they no longer practise jhum and because there is a greater possibility of getting hired labour to replace family labour among wealthy households in villages

with higher economic differentiation. Some fathers, therefore, say: “The daughters, being women, cannot do anything except rely on their husbands. To let them succeed to the land is a great risk’ (Nakane 1967: 89). They thus try and find ways of letting a son rather than the heiress-daughter succeed to

their private land. In Darengri village again, although the traditional system of female inheritance persists, some parents would prefer to give shares to sons as well, and some of the village women are apprehensive that one day the boys will demand an equal share of property (Saikia and Borah 1979). Bose (1980) likewise records several cases where mothers have gifted land to sons who settled in the mother’s village after marriage. The weakening of traditional Garo political institutions has accentuated

the tendency of households to by-pass traditional rules governing land access and use. The nokma and mahari, who formerly had virtually total control over village affairs, have seen a steady decline in their authority. Several factors have contributed to this erosion: the decline in the communal basis of land ownership, the limiting of the nokma’s role with the lapsing of many social and religious rituals due to conversions to Christianity, and

especially the installing by the British of a new hierarchy of officials responsible for settling local disputes and collecting house-tax from the nokmas on behalf of all a’king households. Post-Independence administrative changes have furthered this process.

162 A field of one’s own Emergence of intra-family inequalities. With the shrinkage of jhum land and expansion of wet rice cultivation, unmarried sons and daughters have less access to independent plots and their dependence on parents has

grown. Even in the 1950s, Burling found that while sons and daughters could still open up some dry land in addition to the family’s wet land, it was no longer possible for them to be virtually independent. Parents now do not

always take responsibility for settling their non-heiress daughters after marriage, and these daughters are often forced to seek other sources of livelihood. In Rinsibara village, out of twenty households, seven did not possess wet paddy fields, all seven being non-heiress households (Nakane 1967). This economic imbalance between daughters is another facet of growing economic inequalities in the villages.

Propertyless sons married to non-inheriting wives are somewhat less badly off than non-heiress daughters. First, the family’s demand on the labour of sons has grown. Second, unmarried sons, who traditionally forfeited their claims to their independent jhum plots after marriage, now try and gain access to some land by bringing it under wet rice or planting orchards and thus establishing customary possession over it, which they try

and retain after marriage. This is facilitated by the District Council’s granting of pattas to men and the increasing possibility of in-village marriages as the villages have become more multi-clan (as discussed below).

Changes in marriage patterns. The principles of clan exogamy are still strictly maintained in choosing marriage partners, but the pattern of post-marital residence has changed. In the late 1950s and even mid-1970s, it was still predominantly matrilocal, with most husbands and wives belong-

ing to different villages and most men moving to their wives’ villages (Nakane 1967; Burling 1963; Majumdar 1978). Although cases of in-village marriage were not unknown earlier, now the growing multi-clan nature of villages has increased the prospects of finding a spouse from another clan within the village itself.° The tendency to retain sons with the family has also increased. The desire to do so was expressed by several women I spoke to in the villages near Tura town in the Garo Hills. Bose (1980: 111) records cases of families who ‘are trying to imitate the custom of keeping sons in the family like the patrilineal

people in the plains. They arrange to marry their sons to girls who are willing to live in the house of the husband’s family.” Goswami and © Majumdar (1978) found that in the 1960s, 16 per cent of the male marriages in Wajadagiri

and 22 per cent in Matchakolgiri were within the village. Nakane (1967) found that in Rombagiri, apart from the oldest dominant clan, wives belonged to seven different clans and husbands originated from eight different clans. In Emangiri, for forty-one out of fiftyseven couples, both spouses were from within the village, in thirteen the husbands were from outside, and in three the wives were from outside.

Erosion and disinheritance 163 Majumdar (1965: 28) likewise quote fathers as lamenting: ‘My sons are bright students, but not so my daughters. My educated sons for whom I have suffered so much in defraying the expenses of their education will earn bread for others.’

Sons are also more reluctant to leave well-to-do households and selfsufficient villages (such as Wajadagiri) because of the paucity of jhum land elsewhere, and may seek to in-marry or bring their wives to the village. This

has other effects. Reduced marriage-migration of sons increases the possibility of their gaining possession of village land over time. Further, as Kar (1982: 250) notes, the settlement of sons’ wives and females of migrant households brings about ‘a vertical proliferation of owning and inheriting rights of those alien females. When land becomes scarce these alien females act as potential sources of descent lines in respect of land possession’. In Darengri village, in thirty-four out of fifty-three households, the principal females were from other maharis. All households, however, had alienated

a‘king land as gardens, without the prior permission of the nokma (Kar 1982: 250; see also Nakane 1967). There is still the customary preference for

the heiress daughter to marry her father’s sister’s son, but the practice is stronger in economically stable villages such as Wajadagiri (which are able to attract desired sons-in-law) than in poorer and less stable ones such as Matchakolgiri: 80 per cent of heiress marriages in Wajadagiri, but only 44 per cent in Matchakolgiri, were between actual cross-cousins (Majumdar 1978).

Other features of customary marriage practices are also changing. For instance, while the marriage proposal is still initiated by women, it is usually through an exchange of letters, especially among the Christian Garos. The practice of bridegroom capture-escape-recapture has been replaced by the girl proposing (often more than once) and the boy initially refusing before

accepting. Nakane provides a poignant sample of some typical letters in such an exchange, which are reproduced in appendix 4.1. Male marital violence was rare in the 1950s. Matrilocal marriages and the

presence of the wife’s matrikin in the villages were clearly a protection. Nakane’s (1967: 76) observations on some aspects of marital relationships in that period are revealing: If the husband is not at home when a meal is ready, she eats with her children without waiting for him, leaving his share wrapped in banana leaves. If the wife is engaged in feeding her baby when the husband asks her to bring some firewood to burn in the fireplace she simply says ‘I cannot, I am engaged’. Or, if the time is unreasonable, she may say ‘I am tired now, I am going to sleep’ ... So far as my observation goes, neither the authority of the husband nor the subordination of the wife are ever exercised to extremes.

Whether this observation would still hold today — almost three decades after Nakane’s visit — needs exploring.

164 A field of one’s own Conversion to Christianity and changing sexual mores. Theconver-

sion of large numbers of Garos to Christianity by Christian missions has clearly affected traditional practices and attitudes in many ways. In 1881, 7.6 per cent of Garos were Christian, in 1901, 30.8 per cent, and in 1971, over 50 per cent. The spread of Christianity, while associated with the advance of education for both sexes, had also led to a decline of customary communal and ritual practices, eroding the nokma’s ritual role and overall importance. Women’s sexual freedom has likewise declined. Attitudes to pre-marital sex among the non-Christian Garos were still liberal in the 1950s. Burling (1963: 73) notes that a typical response in Rengsanggri village was: ‘After all they are young, so what can we do?’ He found that ‘no one was greatly surprised when a girl became pregnant, even though no one really approved either’, and further that: Neither [premarital intercourse nor adultery] is fully condoned. The societal ideal of

desirable behaviour is certainly to confine sexual relations within the limits of marriage, and a girl who has been pregnant is markedly less desirable as a wife than one who has not. Virginity itself, however, is not prized, and most people have had some sexual experience before marriage. (Burling 1963: 74)

But conversions to Christianity, Christian influence on the unconverted, as well as exposure to the Hinduized culture of the plains are likely to have affected this. Again, while divorce continues to be common among those who have not converted, the Christian Garos are governed by the Indian Divorce Act of 1869, under which divorce involves court proceedings and is

therefore difficult to obtain except by the wealthy. Separations are thus common, but remarriage 1s difficult. Adultery, while tolerated among the non-Christian Garos, is not accepted by the Christians.

To sum up, over the past century, traditional Garo society has been subjected to a wide-ranging set of policies and influences, leading to substantial economic, social, political, and ideological shifts. Today it exhibits a considerable erosion of the communal basis of land holding, cultivation and control, and a slow but sure undermining of the basis of matrilineal inheritance and matrilocal residence. There have been (a) substantial shifts away from jhum to settled wet rice cultivation, usually also involving a shift from hoe to plough; (b) an overall decline in female labour input, associated with the decline in jhum and the advent of new paddy cultivation practices and techniques managed largely by men, who provide the greater part of the wet rice labour and control the crop and cash

so generated; (c) the privatization of land and, alongside it, increasing economic differentiation and landlessness, although in varying degree across villages; (d) the emergence of land as individually inheritable

Erosion and disinheritance 165 property; (e) the growing tendency for parents to retain sons after marriage to help with the new form of cultivation, and for sons to bring wives to live

with them (thus promoting virilocality); (f) an emergent tendency for parents to pass land to sons, especially via gifts; and (g) the enforcement of more restrictive sexual mores. These changes are presented in summary form in table 4.1 and in diagram 4.1. Similar changes have affected the Khasis and Lalungs, although there are

differences in the extent and causes of shifts in these communities (as detailed in Agarwal 1990b). In general, the histories of the northeastern matrilineal tribes strikingly illustrate how the shift from land as a communal resource to land under individual possession has been associated not only with the well-recognized process of class differentiation, but with the

equally critical (and little recognized) process of gender differentiation among the peasantry. Our discussion of the Garos demonstrates two especially important general points. First, there are close interlinks between ecology, economy, techno-

logy, the social, and the political. Economic changes and the erosion of institutions on one front can set off reactions on other fronts, including in social and family relations, leading to differential gender effects. State interventions in the name of development have seldom taken this into account. Second, not all the noted changes were inevitable. The role of the State in directing change, often listed as only one element among many, was critical, and the common emphasis on population growth as the primary factor inducing change appears to have been misplaced. For instance, in

1970 Ester Boserup advanced the thesis that population growth and associated land scarcity leads communities to shift from swidden-hoe cultivation to settled plough farming, and that this, in turn, reduces female

labour participation in agriculture or, as she put it, causes shifts from ‘female farming systems’ to ‘male farming systems’; it also causes women to lose their rights in land: Female farming systems seem most often to disappear when farming systems with ploughing of permanent fields are introduced in lieu of shifting agriculture. In a typical case, this change is the result of increasing population density which makes it impossible to continue with a system necessitating long fallow periods ... And the advent of the plough usually entails a radical shift in sex roles in agriculture ... (Boserup 1970: 32-3)

Further, a sort of de facto private property in land may emerge in regions where land is gradually becoming scarce, through population increase or the expansion of the cultivation of cash crops. This scarcity of land may result in the loss of women’s rights to land. (Boserup 1970: 58)

' 1t

1 'A! 17) !! t'{! CS!{1!{v: A4f!!rama ee te alienate Be Pome e rr n renter enna ween ann mn amon anne wwe nnmen at

4 t cx _— Z-{c 2 o=) 3 =z 3 D= cac@ = 6262 cs = @ < ca

Qi-yeS&PSec it SSCL Q. o- = Oo QO .- Pe8S Cc-wogl~*t= 5G 2 ui zome @®&£

:''!t {!!i1 !wea---4 -t34 ! ! ‘!{i1 ;1 1 ' ' ' 1 ’ ' ’ ‘ i ' { ! nls ee | | Ojo; UD -=t< Py ®

‘o)| j3328| c* se oO|982a| 2 © o.[=e] @ 8 [83 0] Qac¢|38 Oo 2 |22] 02202355 © ‘x ra) & Lu O ® 2 Qa838 Seuss ne oS @ & (5 j— Oe

A A A Wn Wee

' i 1 ‘ a aia: Clea ealenetalie |

Lu 2 ' ® ® 2 = ' ;|uw £ @ re] Of 5 9. Zo te] n Lu c ! OgiF jJhum: F)M (for food) hunt: M only, limited collect: W and C only orchard: M>F

Land rights all clan members equal Increasingly privatized, (including nokma class differentiated, household, and whether male biased men, women or older children)

Land management husband and wife wet rice: sons, husband jJhum: husband and wife

Control of produce wife mainly husband

Inheritance of usually youngest usually youngest property daughter: land not daughter, but tendency inheritable to gift land to sons

Post-marital matrilocal matrilocal, but residence tendency toward virilocality

Choice of marriage women’s initiative women’s initiative partner

Women's sexual considerable limited freedom

Jural authority male male

Religion traditional/animistic mainly Christianity

F = Females; M = Males; C = Children; W = Women

The Garo case shows that many of these shifts and the erosion of women’s rights in land were not inevitable. First, land scarcity was created

not just by population increases but by specific State policies of land appropriation and regulation. And the shift to settled plough agriculture was a result partly of land scarcity and partly of direct State pressure to ensure such shifts. Second, the reduction of women’s labour contribution to cultivation with the shift to wet rice cannot be disassociated from the male

168 A field of one’s own bias shown by State agencies in the transfer of skills and technology related

to settled farming. Third, this decline in women’s contribution to the household economy eroded the social legitimacy of women’s traditional claims to land. Male bias in the distribution of land titles by the District Council further dispossessed women. Fourth, it was the emergent inter-

village economic inequalities, rather than settled agriculture or land scarcity per se, which was significant in eroding the system of matrilocal residence. That this and other features of the traditional livelihood system were not antithetical to village prosperity and change in technology, is clear

from the several examples noted of villages that have, by virtue of fortuitous circumstances, achieved growth with equity. Altogether, if State interventions relating to technological transfer and land ownership and use had been premised on more gender- and class-egalitarian assumptions, a

different social and economic structure could have emerged. Heavily influenced by a development model in which only productivity increases have centrality and individual private property is seen as a necessary and positive aspect of change, State policies to date have been eroding precisely those communal and social institutions which they should have strengthened. For instance, while the limits of jhum would have been reached sooner or later with an increase in Garo population, the shift to settled agriculture could have been encouraged through a more egalitarian organi-

zational structure such as cooperatives (at least in ownership if not in production) in a context where land was not traditionally private property and reciprocal labour exchange was the norm. Women’s rights could also have been explicitly protected. Today it is difficult to say whether the better features of the traditional social and political institutions that remain can still be preserved. But the traditional organizational structure followed by the Garos of communal land ownership and individual or family-based farming, with a clear recognition of women’s rights in land, can still serve as a model of an alternative institutional arrangement for land ownership and use, when promoting women’s land rights today (as will be discussed in chapter 10). Let us now consider the Nayars.

(2) The Nayars of central Kerala Like the Garos, the Nayars of central Kerala today present a very different

picture from the traditional one described in chapter 3.’ The customary 7 In 1981 the population of Kerala was 25.4 million, of which 20 per cent are estimated to be

Nayars (assuming the percentage to be the same as reported in Government of Kerala 1969): exact figures are not available since data by caste were not comprehensively collected in the Indian censuses after 1931. There was selected census coverage by caste in 1941 and

none after that.

Erosion and disinheritance 169 systems of marriage, residence, land holding, and inheritance have undergone dramatic changes during and since the British period.® Today all the traditional large taravads have been partitioned (on an individual basis), both sons and daughters inherit property, the tali-rite has disappeared, the flexible sambandham union has given way to a more stable monogamous marriage, and visiting husbands are becoming a rarity. Pre-marital virgi-

nity and fidelity in marriage are emphasized, and divorce and widow remarriage are uncommon. Nambudiri men no longer take Nayar women as lovers on a visiting basis, but marry within their own caste. Post-marital residence too has tended towards neolocality. There have been changes in these directions in all parts of Kerala, although their extent and the speed with which they have occurred vary regionally. Underlying these changes are a complex set of interacting factors, the most important of which could broadly be divided into four categories: the

demobilization of the Nayar armies in the late eighteenth and early nineteenth centuries; changes in the local economy; land and marriage legislation passed by the British; and shifts in the ideological and social climate. These affected both marriage and inheritance practices. Although my focus here will be primarily on changes in these practices as

they existed in central Kerala, for comparative purposes I will also draw upon some of Fuller’s (1976) analysis for central Travancore and Jeffrey’s (1976) for Travancore as a whole. Changes in marriage practices. The British disbanded the Nayar armies after their annexation of Malabar in 1792, and those of Cochin and

Travancore were disbanded by the Raja soon after. Although scholars differ in the weight they attach to this factor in what was to follow, there is general agreement that it created, at the very least, ‘a crack in the fabric of the system’ which was to widen over time (Fuller 1976: 125). To begin with,

the return home of the younger Nayar men on a permanent basis led to a strengthening of sambandham bonds and a tendency towards monogamous, more permanent unions. The ceremonial and social importance of sambandham increased, while that of the tali-rite declined; the latter practice eventually disappeared around the turn of this century in central Travancore, and somewhat later in Cochin and south Malabar. Strong ideological and social pressures also induced changes in marriage practices.

British Victorian notions of morality and how a proper family should be constituted clearly influenced their growing condemnation of Nayar practices. But particularly, elite Nayar men, educated in the colleges of Madras ’ The story of these changes, as presented here, has been constructed by drawing particularly on Gough (1952, 1961a), Mencher (1962, 1965, and personal conversations), Fuller (1976), Jeffrey (1976, 1993), and Government of Madras (GOM 1891).

170 A field of one’s own and some in England, and steeped in a mixture of Tamil-Brahmanical and

western ideas, were embarrassed by the polyandrous unions of their mothers, and especially susceptible to the ‘barrage of European criticism of their sexual morality’ (Gough 1952: 83). The 1891 Report of the Malabar Marriage Commission (a Commission set up by the Madras government to examine the case for legislating a marriage law among matrilineal Hindus) was a good indicator of their views. The Commission severely indicted the traditional sambandham relationships as ‘based on a doctrine that there is

no merit in female virtue, and no sin in unchastity’, of which ‘the very defenders of the system are heartily ashamed’ (Government of Madras

(GOM) 1891: 36). It concluded that Marumakkatayam law did not ‘recognize the institution of marriage’ (GOM 1891: 26). Nambudiri-Nayar sambandhams were strongly condemned as exploitative of Nayar women and unfair to Nambudiri women: ‘An institution which by debauching the women of one class [the Nayars], condemns the women of another [the Nambudiris] to life-long and enforced celibacy, is not one which justice need hesitate to condemn’ (GOM 1891: 9). The matrilineal joint family system too came under heavy attack as ‘[making] home-life (in the best sense of the word) impossible’ (GOM 1891: 36). Prior to bringing out the Report, the Commission had solicited opinions from a wide range of men considered ‘fairly representative of the more

intelligent section of the community’ (GOM 1891: 35): officials, the educated classes, representatives of influential taravads, members of the Bar in the district, royalty, and important Nambudiris (GOM 1891: 1).

Opinions ranged from those which severely condemned the existing marriage practices to those which saw no need for change or opposed legalization on the ground that it would undermine the matrilineal system and ruin the taravads, as most men would give their earnings to their wives

and children. The Commission itself recognized that ‘the proposed legalization is not at present desired by the majority’, but took an elitist view that ‘the uninstructed majority will rapidly follow the lead of the enlightened classes’ (GOM 1891: 34). The views of those supporting legislation thus prevailed. Women’s opinions, however, were not even solicited. Although it was recognized that the views of ‘respectable Nayar ladies’ could provide insights, it was not considered ‘practicable to do anything in furtherance of this object’, on the grounds that Nayar women in

north Malabar were ‘conservative’ and lived in seclusion, while Nayar women in south Malabar, though ‘allowed greater freedom’, could not express themselves without the mediation of karanavans or husbands ° Fora detailed and interesting discussion on the divergence of views expressed, see Arunima (1992).

Erosion and disinheritance 171 (GOM 1891: 2).!° Hence a Report which formed the basis of a critical enactment was signed and sealed without having taken account of the experiences and understandings of those whom the proposed legislative change most centrally affected. Indeed the casualness with which the Commission dismissed the need to solicit women’s opinions, their exclusion from the list of those considered ‘representatives’ of the community, clearly indicate Nayar women’s lack of bargaining power in the public sphere.

In 1896 the Malabar Marriage Act was passed, followed by the first Travancore Nair Act of 1913, and the Cochin Nair Regulation 13 of 1920, all of which recognized the sambandham union as legal marriage.!! These Acts also gave a Nayar male the right to pass half his self-acquired property to his wife and children, if he died intestate.!? Earlier such property would have become part of the collective property of his matrilineal joint family estate. A number of other Acts followed, which increasingly expanded the rights of individuals not only over their acquired property but also to seek division of joint family property.!?

Changes in property status: taravad partitioning. From the nineteenth century onwards, taravad property came to be divided with increasing frequency (where earlier it was an occasional occurrence). Initially these partitions were between matrilineal branches — tavazhis (or tavaris) — with each such segment continuing to operate as a joint-property unit, but over

time the segments became ‘shallower and narrower’. From the 1930s onwards, the partitions were more on an individual basis. Joan Mencher (personal communication, 1992) comments: ‘In 1958 I travelled far and wide looking for traditional, large, unpartitioned taravads. The only ones I found were in the court. What could be found, though, were tavaris of much lesser depth.’

A number of factors led to a growing demand for taravad partitioning. Younger Nayar men released from army duties had ambitions of becoming karanavans at an early age, and this was only possible with partition. Also, from the late nineteenth century onwards, expanding income-earning '° Despite the absence of any attempt to solicit their views, some women did send in petitions: just prior to its dissolution the Commission received four, signed by 245 women supporting legislation and 387 women opposing it (GOM 1891: 2). '! The Malabar Marriage Act, however, did not apply to the Nayar- Nambudiri sambandham,

on the ground that the personal law of the Nambudiris made it impossible for them to contract valid marriages outside their caste. The Nayar-Nambudiri union was subse-

ee legal through the Madras Marumakkathayam Act of 1933 (see Puthenkalam '2 However, under the Malabar Marriage Act of 1896, this possibility was only available to those who voluntarily registered their marriages. In fact, between 1896 and 1903 only 83 marriages were registered (Fuller [976: 133-4). 'S For details of other Acts, see especially Jeffrey (1993: 43-4) and Puthenkalam (1977).

172 A field of one’s own

opportunities in several regions created new possibilities for property acquisition. Especially in central Kerala, a growing number of Nayar men found government employment or became professionals: lawyers, doctors, teachers, etc.!4 Many bought houses and land which they bequeathed to wives and children. The strengthening of sambandham unions also accentuated the tendency to bequeath property in this way. Earlier such property

had been at the owner’s disposal during his lifetime but had typically reverted to his taravad after his death. However, in the late nineteenth century such bequests still took place within an overall matrilineal structure, with the men founding tavazhis headed by their wives. The property became part of a new matrilineal segment rather than individually owned and disposable. Nevertheless, the bequeathing of his self-acquired property by a man to his wife and children became a source of growing contention with his matrilineal relatives.

In addition, karanavans were increasingly accused by junior taravad members of mismanaging or misappropriating the joint-family property by their ‘recklessness, extravagance, and neglect of duty’, and of favouring

their own tavazhis. Court cases against them multiplied (Puthenkalam 1977: 143). In Travancore between 1887-1906, junior members filed 4,365 court cases for the cancellation of property alienation made by karanavans,

142 for their removal from office, and 67 against their executing deeds attaching taravad property to pay off their personal debts. In the same period, petitions pending in Travancore courts included 1,161 for taravad partition, 1,367 against gifts made by karanavans to their children, and 871

asking for allotment of maintenance to junior members (Puthenkalam 1977: 148-9). There were similar disputes in Malabar; a number of the petitioners were women.!>

The types of pressures which led to the splitting of taravads varied regionally, affected especially by differences in agrarian structures and the social composition of the local populations. In the mid nineteenth century

the agrarian structure of Kerala showed considerable regional variation (Varghese 1970: 37-50). In Malabar virtually all land was owned by a relatively small number of very wealthy landlords, most of them Brahmins, Rajas, or high-ranking Nayars. In Travancore, by comparison, 80 per cent of cultivated land and all uncultivated land in the 1850s was held by the State. Cochin came in-between: here about 60 per cent of the cultivated land '+ In 1941, Nayar men constituted 11 per cent of the total male population in Cochin but represented 40 per cent of gazetted and 24 per cent of non-gazetted government officers, and 2! per cent of the professionals (Fuller 1976: 131). 'S See e.g. Arunima (1992: 123-4, 152): in one case cited by her, a woman who demanded her

share of the raravad property (while living away from the taravad with her sambandham partner) had to forfeit her share because the court accepted the karanavan’s argument that her relationship with her lover was one of ‘promiscuous intercourse’ (Arunima 1992: 152).

Erosion and disinheritance 173 was owned by landlords, and the rest, along with the uncultivated land, was

held by the State. Soon after their annexation of Malabar in 1792, the British recognized the landlords as the sole owners of landed properties in Malabar, leaving the tenants highly vulnerable to eviction.'!® By contrast,

in Travancore, the government conferred full ownership rights on all tenants in 1865 (Fuller 1976: 19-20). Both the Malabar and the Travancore

proclamations placed land at the disposal of the owner and allowed the

entry of land into the market; but, as Fuller persuasively argues, the implications of this were quite different in central Kerala compared with Travancore, due to the noted differences in their pre-existing agrarian structures, and the nature of economic and ideological shifts that took place in the nineteenth century. In Travancore the 1865 proclamation accentuated the tendency toward the break-up of taravads and the sale of land. The Nayar community was experiencing an economic decline, and this made partitioning attractive for many Nayars; the prevailing belief that it was their matrilineal joint family structure that accounted for their poor economic performance added weight to this.!7 On the demand side, there were ready buyers for any land sold, especially among the Syrian Christians

(Jeffrey 1976). As a result, a good deal of theoretically impartible and inalienable land was sold or mortgaged, particularly by the smaller and less wealthy families.!® These practices subsequently acquired legal legitimacy with the passing of laws permitting partition, especially the second Travancore Nayar Act of 1925 which allowed almost unrestricted partitioning. '© As noted in chapter 3, traditionally the landlords had enjoyed significant rights and privileges in the land they controlled, but were not full owners in the modern western sense of the word, and could not readily evict the tenants. Such eviction was made possible with the British recognition of the landlords as sole owners. A series of Mappila outbreaks during and after the 1830s are attributed to (among other factors) a marked rise in the eviction of Mappila tenants by Hindu landlords in the first half of the nineteenth century (Radhakrishnan 1989). '7 During the second half of the nineteenth century, there was a rapid development of cash

cropping and plantations, with Travancore’s exports doubling in value. The Syrian Christians, who were already in control of a large proportion of the trade in the region at that time, and had know-how, experience, and access to finance, gained significantly as a result. This prosperity stood in stark contrast to the stagnation of the rice economy on which the Nayars primarily depended. There was widespread Nayar propaganda attributing the Nayars’ economic failure to their matrilineal joint-family system and the Syrian success to their patrilineal nuclear-family system. While the functional connection between the Syrian family system and their economic success appears debatable, what is important, as Fuller (1976) points out, is that many Nayar leaders themselves believed this to be so. This provided additional fodder to the noted growing attacks on the Nayar joint-family system on other grounds. '8 Whatever may have been the nature and extent of land transfers in the pre-British period (some writers maintain that land was commonly sold: e.g. Arunima 1992, while others express doubts about this: e.g. Joan Mencher, personal communication, 1993), certainly during the British period various enactments and the prevailing economic climate led to large-scale land sales.

174 A field of one’s own Many Nayar leaders carried out an intensive campaign in the early part of the twentieth century in support of this legislation. Within five years of the 1925 Act, some 32,903 taravads are said to have been partitioned (Fuller 1976: 135).

Similar Acts were put into the statute books of Malabar and Cochin in 1933 and 1938 respectively; the former allowed legal partitioning of the taravad into matrilineal segments and the latter allowed partitioning on an individual basis. Here too there was a break-up of taravads, although not

on as extensive a scale as in Travancore. In all regions, the smaller, relatively poorer and lower-ranking taravads were the first to split on an individual basis soon after the 1930s Acts, while the very wealthy and prestigious ones remained unpartitioned till the 1950s, especially in Malabar. By 1960, however, all the large taravads in Malabar too either had been partitioned or there were cases for partition pending in court. Although the demand for legal reform which led to statutory enactments in the early twentieth century allowing taravad partitioning came from the Nayar community itself, it clearly had the endorsement of the British, who felt it was in keeping with a man’s ‘natural’ instincts and affections to leave his self-acquired property to his wife and children rather than have it revert to his natal taravad on his death. In addition, the enactments made a man

legally responsible for maintaining his wife and minor children. Hence although a woman could now own a house and land independently of her taravad and her husband, she was legally treated as a dependent. Under the traditional sambandham union(s) of the central Kerala Nayars, a woman

had enjoyed relative sexual and economic independence vis-a-vis her husband(s), which gave her a significant advantage in marital relations,

even though the karanavan’s authority had been dominant in estate management and tn public dealings. Also women were not in the same position as their husbands and brothers to take advantage of the professional employment opportunities and associated possibilities of accumulating self-acquired wealth that were then opening up; most Nayar women appear not to have received professional education,!° and social prejudice against women taking up public employment is also likely to have

been a deterrent. This gender disparity in access to earned wealth would have negatively affected women’s bargaining power within the home. Post-Independence changes. Post-Independence legislation relating to joint estates sustained the trend toward property partitioning. Land reform legislation had additional effects.2° The most widely debated was 19 See e.g. Arunima (1992: 125, 156).

20 For a detailed review of such legislation see Radhakrishnan (1989). For the gender implications of some pieces of legislation see Saradamoni (1983).

Erosion and disinheritance 175 the Kerala Agrarian Relations Bill of 1957 (eventually passed in somewhat

modified form as the Kerala Agrarian Relations Act (KARA), 1960) introduced by the Communist Party then in power in the state. This provided (among other things) for fixity of tenure and the purchase of ownership rights by the tenant who directly managed the land; it also set a ceiling of fifteen acres of double-cropped paddy land or its equivalent per

family of five members, plus an acre each for every additional member subject to an overall ceiling of twenty-five acres. The ceiling-surplus land was to be distributed among the landless and landpoor. The Bill (on which public opinion was elicited) was widely opposed by the large landowners, including the Nayars, whose interests as essentially non-cultivating landowners, with holdings well above the ceiling, were seriously threatened.?! The Bill also placed Nayar women at a particular disadvantage: while the men could have shifted to self-cultivation/management, Nayar women

(most of whom had never directly worked on or managed a farm) had limited options. A number of women heads of households submitted written opinions on the Bill, pointing out that their only source of income was from land leased out to tenants. But such concerns did not carry weight in the discussions on the Bill (Saradamoni 1983). KARA, 1960, underwent several amendments until the passing of the Kerala Land Reforms Act,

1963 (which is currently in force, although again after many amendments).2? The actual implementation of the laws also had a chequered history. But it is noteworthy that in three Palghat villages studied by Saradamoni (1983) in 1977-78, over a fourth of those who lost land following various land reform enactments were widows (p. 137). The way in which the ceiling laws were implemented had an additional adverse effect on women’s ownership of land. Surplus land above the ceiling was assessed on the combined property of husband and wife, and in many

instances it was the wife’s land that was forfeited. Petitions by women arguing that it is their husbands who should forfeit land above the ceiling where the latter also have non-land sources of income, came to no avail (Saradamoni 1983). In terms of inheritance, the Hindu Succession Act of 1956 made special

provisions for those governed by the Marumakkatayam laws. The Act sought to eliminate prevailing differences both in the devolution rules of customarily matrilineal Hindus across Kerala, and in different types of landed property — self-acquired, separated, and joint family property. For

purposes of devolution, a person’s share in joint family property (viz. taravad or tavazhi property) was deemed to be that which would have fallen 2! For a discussion on the opposition the Bill generated, see Radhakrishnan (1989). 22 For details on various amendments see, Radhakrishnan (1989) and Raj and Tharakan (1983).

176 A field of one’s own

to her/him if immediately before her/his death the property had been partitioned per capita among all the members holding an interest in the taravad or tavazhi. Hence under the Act, if a Nayar male died intestate, all his property (including that self-acquired, separated, and his share in joint property decided as above) would devolve very similarly to that of other Hindu males (as will be described in chapter 5). Where earlier statutes had favoured the lineal descendants of predeceased daughters and given few or no rights to descendants of predeceased sons, there was now a shift to a system which placed the children of predeceased sons and daughters on a par, and recognized some additional descendants of the sons alone. For a Nayar woman dying intestate, all her property in the first instance would devolve on her children, mother, and children of predeceased children (and

not, as in some earlier statutes, to children of predeceased daughters only).23 This enactment thus reduced the previously recognized shares of females in both parents’ property. It also conferred testamentary rights for

all property on both men and women. With the further passing of the Kerala Joint Hindu Family System (Abolition) Act of 1976, the ‘corporateness’ of families was no longer legally recognized: all surviving matrilineal joint family estates were deemed as partitioned on a per capita basis. Many families sought to preserve their joint property nature by setting up trusts, but the earlier legal validity of the matrilineal joint estate was gone (Jeffrey 1993: 46).

Where have all these changes left Nayar inheritance and marriage practices? Presently, as noted earlier, the Nayars constitute about 20 per cent of Kerala’s population. However, there is little ethnographic information on their situation after the early 1970s. Fuller’s ethnography relates to data collected from Ramankara village in central Travancore in 1971-72, while those of Gough, Mencher, and Unni relate to Cochin and south Malabar in central Kerala and go back to the 1950s and 1960s. Nevertheless, the directions of change noted by these studies are clearly indicative, even if they do not allow us to assess the extent of the change. The post-Independence inheritance practices described by Gough, Unni, and Fuller, for their respective study areas, differ from one another in some degree, but all three present significant bilateral features. In her Cochin village, Gough (1952) found that around 1950, ancestral land and large joint-family houses were still inherited matrilineally, but small houses, cash, and much of the garden land passed from the father to all children. In south Malabar, also in the 1950s, Unni (1956) noted that property which descended through the mother continued to be inherited matrilineally, 23 For details about other categories of heirs see chapter S.

Erosion and disinheritance 177 whereas any property given by the father was divided equally among his sons and daughters. And Fuller (1976) found that in central Travancore in the 1970s, ‘family land’ and ‘individual land’ devolved differently. The family land was that inherited from any matrilineal relative and was held by the woman on behalf of all matrilineal descendants as well as herself. This

could not be alienated without the consent of her adult matrilineal descendants, and it devolved equally among all descendants (usually daughters and daughters’ children). Individual land was any land not inherited from a matrilineal relative. As a rule, this was divided equally among all children, and was freely alienable. However if a woman passed on individual land, inherited say from her father, to her daughter, for the latter this became family land which could then only go to matrilineal relatives. Recent research in central Travancore by two doctoral students

indicates that although women’s property rights are still recognized, women usually inherit a house but are less likely to get agricultural land. Dowry too is becoming common among the Nayars, although unlike in north India the items given are typically registered in the woman’s name.?* In Nayar marriage practices also, regional differences, although still

noticeable on some counts, have been narrowing on others. In south Malabar, Mencher (1962: 241) found an increasing tendency toward crosscousin marriages (which could be seen as reunifying property divided by inheritance); and 34 per cent of marriages in the villages studied by Kala (1982: 311) in 1975-77, were between cross-cousins. But in Cochin, both

Gough (196la: 366) and Nakane (1962: 25) noted a decrease in such marriages; and in Ramankara village in central Travancore, Fuller (1976: 83) argues, the Nayars never had a cross-cousin marriage system. Also the visiting husband custom has been declining everywhere. It had survived toa limited degree in south Malabar into the 1950s (Unni 1956) and even up to the 1970s (Kala 1982). But in Ramankara, in 1971, Fuller observed no cases, possibly because even traditionally the practice was not common

there. For Kerala as a whole, only 16.5 per cent of the 745 families Puthenkalam (1977: 162) interviewed in 1961 had cases of visiting husbands, and these, he comments, were men ‘of the old generation. One can predict they will be the last of their class.’ Even where it had survived into the 1960s and after, the visiting husband system had undergone changes. The man was no longer just a visitor for the

night, but spent longer periods (sometimes several days) in his wife’s household, taking an interest in his children’s upbringing, and at times even

managing his wife’s estate (Unni 1956, Mencher 1962). The majority of 24+ Personal communication in 1992 by Caroline and Filippo Osella, doctoral students in anthropology at the London School of Economics. Their fieldwork in 1991 was on the Izhavas, but they had these observations to make about the Nayars.

178 A field of one’s own Nayars now marry only once, but divorce rates were still relatively high in the 1950s: in Unni’s (1958: 128) study, 20 per cent of the Nayar women over the age of forty-five had divorced their first husbands to marry other men. These were mostly women from wealthy families. There were also cases of

extra-marital affairs by women whose husbands were away in distant towns. The women suffered no social stigma as a result, as long as they did not violate subcaste rules in their choice of lovers.

Tendencies towards nuclear households and neolocal residence have grown, again in varying degree in different regions. In Fuller’s central Travancore village, 90 per cent of the households were nuclear or variants on the nuclear type. In Gough’s Cochin village, a majority were nuclear in 1964, but in south Malabar, Unni and Mencher found such households to be ina minority. The shift toward neolocality is likely to reduce the support

women can get from their matrilineal kin. The nuclearization of households, however, need not in itself have this effect. In the villages Mencher studied in 1959, many nuclear family units continued to live close to one another after a taravad split. Here, she notes (1962: 243): [T]he daughter may frequently be seen coming from or going to the mother’s house for advice, to exchange food cooked in one or the other of the houses, or to leave the baby girl to be cared for by the grandmother or the mother’s younger sisters. Indeed, the women and children constantly visit one another. One does not get a sense of great distance between the two households.

Mencher also observed that the majority of families in the village maintained close ties with their matrilineal kin. During recent visits to central Kerala, she found this still to be the case (personal communication, 1992). Thus the breakdown of the Nayar taravad into smaller family units has not meant a shift towards the patrilineal, virilocal pattern found in other parts of India. There is an increase in the authority men exercise in their wives’ homes, but women’s close contact with their matrikin clearly strengthens their fall-back position, which helps to limit the extent of male marital dominance. Of course, this could change as the observed tendency toward city-based male employment requires a growing number of women to move to more distant locations. The effects on women of the changes in inheritance practices are mixed. On the one hand, it can be argued that women have gained in that they now

have individual rights in landed property over which legally they have absolute control, while earlier control over management was essentially in

male hands through the karanavan. However, class differentiation and poverty among the Nayars have increased with the break-up of taravads, and many Nayar women whose economic needs would previously have been taken care of by the joint-family estate are today landless and poor.

Erosion and disinheritance 179 Yet women in Kerala by several indicators were still better-off in 1991 than

their counterparts (in class terms) elsewhere in the country: Kerala at present is India’s only state with a female-favourable sex ratio (104 females for every 100 males) and a female literacy rate of 86 per cent (for females

above seven years of age) compared with the all-India average of 39 per cent. Hence today: Rain seeps through many a faravad mansion, but the women in white, barefoot, walk literate on the metalled road.?°

(3) Matriliny and development There has been a popularly held view in the social sciences that economic

development and modernization would inevitably lead to a decline in matriliny.2° However, in our study of the Garos we saw that the erosion of matriliny was by no means inevitable: in particular the technological and population determinism characterizing Ester Boserup’s arguments did not

hold. A significant role was played in this regard by State intervention in terms of land acquisition, development policy, legislation, and the gender ideology underpinning these interventions. Interventions based on a different set of assumptions and on an agenda for strengthening the communal and matrilineal character of the community could have had very different results, even with modernization. Fuller makes a similar point in relation to Nayar matriliny, also arguing that the decline was not an inevitable result of modernization. In particular,

he takes issue with Gough’s (1961b) contention that the ‘root cause’ of modern change in matrilineal kinship systems, including in that of the Nayars, is incorporation of the societies into a unitary, capitalist market economy, the entry of land into the market being especially crucial. He notes that land had become freely marketable in Malabar soon after British annexation of Malabar in 1792, but this had led to no significant break-up of the taravads, while in Travancore the conferring of ownership rights on tenants in 1865 led to a splintering of joint estates on a massive scale. He argues (1976: 146), ‘if the entry of land into the market was determinant, why did taravads in Malabar persist as joint-property holders for so much longer than those in Travancore, when land entered the market in Malabar seventy years before it did in Travancore?’ In his view the critical factors were: on whom the land rights were conferred; the prevailing economic and

social conditions in Travancore which made Syrian Christians so much -> From the author's unpublished poem entitled ‘Kerala’. 76 See Gough’s (1961b) and Fuller’s (1976) discussion on this.

180 A field of one’s own more prosperous than the Nayars; and the view current among European scholars and accepted by the Nayars that both matriliny and joint families were obstacles to economic progress. He thus concludes (1976: 147): ‘no single determinant, such as increasingly complete incorporation into the capitalist economy, can be regarded as the cause of the process. On the one hand, unique factors like disbandment of the armies played a crucial role; on the other, economic development was not identical in all regions and, in any case, by itself it did not invariably bring about alterations in the family and kinship system.’ Unlike the situation described for the Garos, however, it is difficult to say

to what extent the overall decline in Nayar matriliny could have been

, stemmed if State interventions had taken other forms. But women’s traditional rights in land could certainly have been given better recognition in the State’s post-Independence land reform programmes, which showed the same insensitivity to women’s needs in Kerala, as elsewhere in the country.

II. Sri Lanka The Sinhalese

The Garo and Nayar case studies provided an idea of the historical processes of change among matrilineal communities in two quite different regions of India. In Sri Lanka, I will focus on the Sinhalese. The changes in marriage and inheritance practices here are less dramatic than those noted for the Garos and Nayars. Unlike those two communities, the Sinhalese are not a small pocket of matriliny amidst a larger culture of patriliny. Rather they constitute the dominant community in an area where the minorities

too are mostly bilateral or matrilineal. (In 1981 there were 11 million Sinhalese in Sri Lanka, constituting 74 per cent of the country’s population.) However, among them too, colonial and post-colonial interventions in marriage and inheritance laws played a significant role in the changes that occurred. Sri Lanka was brought under colonial rule much before India, and parts of it were ruled by a series of colonial powers. For instance, the Maritime Provinces along the coast were governed first by the Portuguese (15051658), then by the Dutch (1658-1796), and finally by the British (1796—

1948). Neither the Portuguese nor the Dutch, however, could break Sinhalese resistance in the kingdom of Kandy, which remained independent till its annexation by the British in 1815. The British alone gradually established political and administrative control over the whole island. The discussion here relates basically to the British period and after: it was then that legal and administrative changes were most systematically effected.

Erosion and disinheritance 18] Changes under the British. British interventions in relation to land

took a number of forms: direct appropriation of land by the State; encouraging appropriation by Europeans and the local elite for establishing plantations; changes in inheritance laws; and changes in laws governing marriage and divorce. As will be seen below, these interventions were linked by one central concern: the desire to establish control over land for the extraction of maximum economic benefit for the colonial powers (although in relation to marriage, British responses were also guided by their unease with the liberal sexual mores among the Sinhalese). From the 1840s onwards, through a series of Ordinances, the British established control over large tracts of land. In addition a good deal of the Crown land was sold to European capitalists for setting up coffee, tea, and

rubber plantations. As a result, villages in the Kandyan areas were gradually circumscribed by large tea estates, and those in many parts of the

Western, Southern and Sabaragamuva Provinces by rubber plantations (Obeyesekere 1967: 100). Plantations necessitated the clearing of forests and also reduced the availability of irrigation water and organic manure for the local population: all this adversely affected the local ecology and the stability and productivity of village farming (Sarkar and Tambiah 1957: x1).

In non-plantation areas, a Mudaliyar (a Sinhalese revenue-cum-police officer) was appointed with considerable powers to reallocate village land between Crown land and the private land of the villagers. A Grain Tax was instituted, and villagers who defaulted, or who in the officer’s judgement would be unable to pay, were forced to sell their land. Where unpartitioned land was held jointly and cultivated on a rotational basis (the tattumaru system),?7 if one partner defaulted the entire plot had to be forfeited. Also Ordinance No.10 of 1863 made it possible for shareholders to force the partition of an estate held in common, and to gain possession of the whole if other owners were unable to afford the cost of partition proceedings and/or if the portions each owned were too small for cultivation. Land so acquired could also be sold. Through the emerging land market, rich Ceylonese

landowners and European capitalists were able to acquire large tracts, increasing the tendency towards land inequalities (Risseeuw 1988; Obeyesekere 1967). Between 1878-88, in five provinces (Western, Central, Southern, Eastern, and Sabaragamuva), one in every twenty-six acres was sold in this way (Obeyesekere 1967: 122). With the increasing appropriation of land by the British colonial State, individual Europeans, and the new Ceylonese elite, and with the growing

partitioning and sale of land hitherto held jointly, land scarcity and 77 Tattumaru ‘is a system of rotation of plots in an undivided tract of paddy land so that access

to the whole stretch is made available to all shareholders during a fixed period of time’ (Obeyesekere 1967: 35).

182 A field of one’s own landlessness grew apace. In 1937, an analysis of seven districts showed that 45 per cent of the population was landless, and another 21 per cent had less than one acre per person (Sarkar, quoted in Risseeuw 1988: 79).2® British administrators, probably concerned less with the problem of landlessness

than with the ways in which growing land fragmentation could constrain the development of capitalist agriculture, suggested as one solution the introduction of primogeniture, which would effectively exclude younger sons and all daughters from rights in land. The scheme was eventually shelved because of enforcement difficulties and the recognition that exclud-

ing other children, especially younger sons, would be unacceptable in a system where all children traditionally had inheritance rights (Risseeuw 1988).

However, as land scarcity grew women increasingly lost land: between 1901 and 1921, the number of Sinhalese women paddy landowners fell to -

half, while the number of women wage earners and other labourers increased dramatically (see table 4.2). Land disputes grew alongside, including disputes over the division of family land. Many of the petitioners

were women. But their petitions seldom received the attention they deserved. For instance, in a 1910 diary notation, a British officer (quoted in Risseeuw 1988: 82—3) made the following observation on an undivided land case: One of them made me regret that one cannot enquire into every petition personally. In this particular case the petitioner appeared prima facie to have no case at all and I had already told her so. She petitioned again and I put it down for circuit. It was only after minute enquiries on the spot that I found that she was really losing her land in a most unjust and unfortunate manner.

Women also lost land as a result of changes in marriage and divorce laws:

these led to shifts in post-marital residence towards virilocality, which adversely affected women’s access to inherited land. As Risseeuw (1988) argues, British enactments regarding marriage were guided on the one hand by notions of morality and the assumed superiority of their own marriage

practices, and on the other hand by colonial material interests in the devolution of landed property via inheritance. Traditional Sinhalese customs of polyandry, polygamy, easy divorce, several marriages in a lifetime, and a liberal definition of legitimate heirs, conflicted with the British notion of marriage as a monogamous, lifelong union sanctioned by the Church and the State, with clear lines separating legitimate and illegitimate children. Hayley (1923: 174) comments on this as follows: 28 A 1955 survey of six Kandyan villages in Pata Dumbara by Sarkar and Tambiah (1957) revealed a similar story.

So

Erosion and disinheritance 183

Table 4.2: Women in selected occupations: Sri Lanka 1901-21

ee eeee———eeeeoe————————

1901 1911 1921 Community census census census Wage earners’

Lowland Sinhalese 171,358 174,355 280,199 Kandyan Sinhalese 99,189 65,808 181,507

ee Tamils? 199,881 38,745193,326 59,869 Indian Tamils — 176,728 Totals 490,814 478,055 745,925

Paddy landowners

Lowland Sinhalese 10,413 12,315 4,498 Kandyan Sinhalese 26,891 23,232 14,590

Tamils 1,600 2,824 Indian 7,910 Tamils — 4] 37

Total 47,553 39,106 23,349

Tea labourers

Lowland Sinhalese 5,146 5,203 5,964 Kandyan Sinhalese 4,286 6,317 7,261

Tamils 124,391 31] 144,440 3,549 Indian Tamils — 148,419

Totah 135,392 166,256 160,596

General labourers

Lowland Sinhalese 5,556 10,580 11,807 Kandyan Sinhalese 2,081 5,565 10,131

Tamils 5,984 — 3,432 10,619 Indian Tamils 3,124 2,209 Total 14,397 23,989 36,142

Notes: | The 1901 category includes all women wage earners, while the 1911 and 1921 categories include women wage earners in selected occupations. 2 Tamils in the 1901 census included members of the resident community as well as migrants; Indian Tamils in the later census reports are the immigrant South Indians. 3 Totals include European, Burgher, Malay, and Moor communities as well. Source: Grossholtz (1984: 116); based on data from the Census of Ceylon for the years cited.

[T]here was ... prior to recent legislation on the subject, a remarkable vagueness of ideas with regard to the inception, maintenance, and dissolution of matrimonial alliances.

Legislation was enacted to set this right. In 1859, polygamy and polyandry were made penal offences, and unregistered marriages were made punish-

able by a fine. Also divorce — unilateral or by mutual consent — was restricted, the only recognized grounds for it being adultery and desertion; and from being a private affair, it now needed a court decree.

184 A field of one’s own The failure to register a marriage or to obtain a divorce in court (as happened frequently when the law was first passed) created a category of illegitimate children who were not legally entitled to a share in the parents’ estate. In case of two marriages, one registered and the other not, heirs from the former could deny rights to those from the latter, as indeed was found to be happening in the 1960s (Obeyesekere 1967: 141). Also, given the customary ease of divorce, it had been simple for either partner to move out of a diga marriage to a subsequent binna one, or vice versa. The Ordinance took away this flexibility. This meant that a daughter first married in diga had considerably reduced chances of subsequently contracting a binna marriage which would enable her to re-establish a claim on the parental estate. In addition, the opening up of new avenues for nonagricultural incomes primarily for men further reduced the attractiveness of a binna marriage for them, in so far as access to the wife’s land was one of

the important inducements for less endowed men to contract such marriages.

Statistics from the early part of this century are noted to show a clear decline in the proportion of binna marriages, and from 1962 onwards the

distinction between binna and diga was no longer maintained when recording the figures (Risseeuw 1988). Polyandry also declined, although it

did not entirely vanish, and its ideological acceptability suffered: such unions came increasingly to be viewed as an embarrassment by younger, educated, and westernized males (Tambiah 1966), a reaction not dissimilar to that of many educated Nayars in Kerala during this period. Some of the changes in marriage laws were supported by the local male elite, who saw in it a chance to reduce their responsibility towards married daughters: That in numerous instances parents are reduced to poverty solely in consequence of

their married daughter, and her issue, being thrown upon them for support and maintenance, ... is aclear result of the existing law regulating the marriage contract

among them; and there is nothing whatever to prevent the husbands of married

daughters discarding their wives at any moment they choose and betaking themselves to other women in their stead. (Petition by the Chiefs of Kandy to the British in 1855, quoted in Risseeuw 1988: 33)

The decline in the incidence of binna marriages and the expressed reluctance to support diga-returned daughters, at least by sections of the elite, would also have had negative consequences for women’s land access. Moreover, some major legal changes specifically affected the Sinhalese in

the Maritime Provinces, who constituted a significant proportion of the Sinhalese population. Customarily, the Sinhalese, irrespective of their location on the island, were governed by a uniform set of laws. Colonial legal intervention, especially under the British, led to the emergence of a

Erosion and disinheritance 185 significant legal differentiation between the Sinhalese resident in the Kandyan Provinces and those resident elsewhere, notably those in the Maritime Provinces. The British established control over the Maritime Provinces in 1796 but

(as noted) only annexed the Kandyan Provinces in 1815. After their conquest of the Maritime Provinces, they proclaimed that the Sinhalese had

the right to keep their own laws and customs. However, they failed to ascertain the content of the Sinhala laws prevailing in the Maritime Provinces and wrongly supposed that the people there were governed by the

Roman-Dutch law since Dutch times. Although the Dutch did introduce Roman-Dutch law in Ceylon, Obeyesekere (1967) argues that it probably applied only to Dutch residents of Ceylon and to converted Sinhalese Protestants, while the majority of citizens were governed by their traditional laws. Between 1803 and 1833 the Roman-Dutch law was sporadi-

cally applied to the Sinhalese, but by the Charter of 1833 the British repealed their right to be governed by their own laws and fully imposed the Roman-Dutch law on the Sinhalese of the Maritime Provinces. By 1860 this had become the General Law of the country, applicable to the Maritime Provinces and to all inhabitants ‘except in those instances in which such an

inhabitant is by privilege under the sanction of another form of law’ (Thomson 1866, quoted in Obeyesekere 1967: 131). In effect, therefore, all Sinhalese, other than those identified as Kandyan (who were to be governed by traditional Sinhala law), now came under the purview of the RomanDutch law. ‘In the process’, Goonesekere (n.d.: 15) observes, ‘the Kandyan and the Low Country Sinhalese, a people sharing what may be described as

“ethnic homogeneity, a common historical tradition, and experience of successful political unity, though subsequently divided into separate political units’’, acquired different types of personal law.’ The British also brought the Sabaragamuva laws into line with laws in

other Kandyan areas. As a result, a diga widow’s right to the husband’s estate in Sabaragamuva was repealed, and the people there were subjected to the rules of Kandyan law under which the widow had neither temporary nor permanent rights in the praveni. The introduction of the Roman-Dutch law for the non-Kandyan Sinhalese made three critical changes in relation to their marriage and inheritance practices (Obeyesekere 1967). It established a community of property at

marriage, strictly bilateral rules of inheritance, and relatively inflexible divorce procedures. All property owned at the time of marriage or acquired during marriage by whatever means by either spouse became merged into a common pool as their ‘community of property’. To keep one’s pre-marital property separate, an ante-nuptial contract to this effect was needed (Goonesekere 1980).

186 A field of one’s own On the death of either spouse, the surviving one received half the property

and the children received the remaining half. In theory the concept of a community of property could be advantageous to women, in that men were

likely to acquire more property after marriage than their wives. But the disability lay in the husband’s absolute rights of disposal over this property, without the consent or approval of the wife and even in the face of her direct

opposition. The wife had no power to appear in court or enter into any contract without the husband’s consent, and was liable to his debts, while he was not liable to hers. She could dispose of her property without his consent only by a last will.

Under the strict bilateral rules of inheritance, men and women could , share equally in either parent’s estate, irrespective of the kind of marriage they contracted or where they resided. This initially benefited women who earlier had to forfeit their inheritance rights under diga marriage. But since this also meant that property could now be alienated even to those who were not recognized members of the village, in the long term this laid the

basis for overall family resistance to any land inheritance by women (Obeyesekere 1967: 35). In 1876, the British, through an ordinance, abolished the community of property. Later, under the Married Women’s Property Ordinance No. 18 of 1923, non-Kandyan Sinhalese women regained many of the rights they had lost under Roman-Dutch law. A married woman could now hold property

and enter into contracts in the same way as could an unmarried woman, bring or be brought to court action without the husband being cited as a plaintiff or defendant, and freely dispose of all her movable and immovable property held before marriage as well as that acquired during marriage. At

the same time, she had to maintain her husband if she had sufficient property and he did not. These improvements were enacted only after considerable debate. The debate preceding the 1923 Ordinance is especially interesting in what it reveals about prevailing attitudes toward women. Some of the objections to the Ordinance raised by male legislators in the Legislative Council included the following (quoted in Risseeuw 1988: 62): Our girls have not reached a stage of civilization where they can benefit by the extended rights which the Bill seeks to give them. [I]t will open the door to misconduct on the part of our wives. In every family there are occasions where there is unpleasantness between a husband and a wife, and at such times an unscrupulous neighbour or near relative might induce the wife to leave the protection of her husband, and all her property will be at the mercy of the person who takes her away ... [O]wing to the prohibition against alienation by a married woman, many a married woman has been saved.

Erosion and disinheritance 187 Some male legislatures, however, did support the Ordinance, arguing:2? 99% of the Sinhalese women are level-headed, and will always be guided by their husbands rather than be led astray by the first adventurer who comes along. [A]fter all, surely a woman can do what she likes with her property, what right has anybody to talk of sex disability in the 20th century?

It is notable, though, that even these men who spoke out in women’s favour made inferences on behalf of women, rather than soliciting women’s own views on the matter. Ultimately, several factors contributed to the passing of the 1876 and the

1923 Ordinances:*° the British administration’s unfamiliarity with and unsympathetic approach to the concept of a community of property (which

did not exist under English common law); the demands being raised by women in Britain at that time for greater equality in men’s and women’s rights and the passing of more progressive legislation there;?! and, in the case of the 1923 Ordinance, the support it received from fathers — after 1876

even the dowry given by a woman’s father had become the absolute property of the husband and a source of conflict between the father-in-law and the son-in-law. More generally, of course, legal changes instituted by the British need to be viewed in the light of their overall economic self interest, although the form this took was not static over time. Risseeuw (1987, 1988) usefully sums up the underlying thrust of British

intervention in the areas of land policy, inheritance, and marriage as follows: The overall land policy of the British entailed alienation of land ona large scale from its original (multiple) owners tn order to facilitate the economic exploitation of land

in the form of plantations ... This policy necessitated the simplification of the Sinhalese laws of inheritance. In the traditional system no secret will was recognized and it was customary for a parent to speak his/her will on the death bed, based toa large extent on the amount of care received from the children. This system obviously

kept matters pending over far too long a period when viewed in an economic perspective. The new economic system required unambiguous (individual) owners of land, capable of swift decisions concerning its exploitation. (1988: 29) When seen in the light of this land policy, it becomes clear that a form of enduring monogamy, preferably limiting economic power to one of the spouses and reducing

the responsibility networks from a large extended family to a smaller unit, is a necessary prerequisite for a successful transition to the private initiative and ownership envisaged by the British. (1987: 15) 29 The citations below are both taken from Risseeuw (1988: 64-5). 30 See discussion in Goonesekera (1980), Obeyesekere (1967), and Risseeuw (1988).

31 See e.g. Holcombe (1983: 148-234) on the debate leading to the Married Women’s Property Act of 1882 in England.

188 A field of one’s own In such a development ‘diffuse’ marriage patterns and forms of bilateral inheritance

form an obstacle, which was more successfully approached by modifying the position of the weakest link in the system. This was, as in the traditional system, the position of the women. (1988: 29-30)

Changes after Independence. What then is the situation among the Sinhalese today? Only a few ethnographies give an idea of actual inheritance by gender. They indicate that women do inherit land as daughters, sometimes even when married in diga, but (a) they usually inherit less land than their brothers; and (b) the chances of inheritance are greater for digamarried daughters if they remain in the natal village than if they leave it.

Interestingly, these broad conclusions hold not only for the Kandyan Sinhalese but also for the Low Country Sinhalese. Consider first the Kandyan Sinhalese. Two studies give information on actual inheritance by daughters among

the Kandyan Sinhalese: Sarkar and Tambiah’s (1957) and Tambiah’s (1965). Sarkar and Tambiah (1957) examined how paddy land and highland devolved between men and their first sisters in six Kandyan villages located in the Central Province and surveyed in 1955 (see table 4.3).

They found that while children of both sexes inherited land, the bias was distinctly in favour of sons. In landed families, 68 per cent of the male respondents and 56 per cent of their first sisters received some paddy land, and a much larger percentage of both men and women received some highland. Men received much more land than women on average, gender differentials being especially marked in paddy land (on average males received 0.375 acres and their first sisters 0.068 acres). Between pairs of brothers and sisters, the sisters received about one-third of the average area (of paddy as well as highland) given to brothers. Between themselves, all brothers received equal shares as did all sisters. Tambiah’s (1965) survey of Rambukkoluwa village (also in the Central Province) in the 1950s revealed that binna-married sisters were more likely to receive the same shares as their brothers, and diga-married sisters to inherit less. Of the forty married women in the village, twenty-two had been born in the village. Of these, eleven married binna, eleven diga.** All but two of these twenty-two who continued to reside in the village after marriage

inherited land. Of the eighteen who came from outside (that is who contracted inter-village diga marriages), thirteen had some property rights in their natal village. Tambiah’s evidence thus suggests that as long as a woman remained in her natal village after marriage, even if she did not reside in her parents’ home, she usually retained rights to land in practice. It 32 Diga marriage involves shifting to the husband’s home, but does not always involve leaving the village.

Erosion and disinheritance 189 Table 4.3: Ownership of paddy land and highland by gender among the Kandyan Sinhalese

Paddy land Highland

First First

Male sisters of Male sisters of Type of parental respondents respondents’ respondents respondents! household No. % No. “%o No. % No. % Landless households 94 — 86 — 42 — 46 — Landed households, but children yet

to receive land 26 32.1 23 44.2 39 29.3 30 32.6 Landed households, land received by

children 55 67.9 29 $5.8 94 70.7 62 67.4 (Average acres inherited) (0.375) (0.068) (0.721) (0.395) Pairs of sisters and brothers either or both of whom inherited.

Paddy Land Highland

No. of sibling pairs 49 87 Male ownership (average acres) 0.625 1.450

‘Female ownership (average acres) 0.221 0.562 Note: ' land inherited or received as dowry. Source: Sarkar and Tambiah (1957: 58-9).

is women contracting diga marriages outside the village who typically risked forfeiting their rights, and even then not always. Tambiah notes that of the women who inherited land, half received shares equal to their brothers and

half received lesser amounts. The latter were probably those contracting outside-village marriages.

Several additional studies, while giving no indication of actual land inheritance by gender, do give information on post-marital residence and the extent of in-village and outside-village marriages for Kandyan women, from which the likelihood of these women inheriting land can be inferred. Table 4.4 gives information on post-marital residence of married women relating to two Kandyan villages. In Yalman’s 1954-55 study of Terutenne village, half the women stayed in the village after marriage and half moved out, while in Rambukkoluwa village studied by Tambiah (1965), of all the women who married during 1890-1958, about 66 per cent remained in their

natal village and 34 per cent moved out. In contrast, of the men who

190 A field of one’s own Table 4.4: Post-marital residence among the Kandyan Sinhalese

Terutenne! Rambukkoluwa

(1954-55) (1890-1954) Post-marital residence women men women men

of married persons a from village No. % No. % No. % No. % Both spouses from village 15 36.6 15 67.3

Resident in village after 8652.4 93 92.1

outside village) 12 29.3 31 32.7 marriage (spouse from Left village after

marriage 78 8647.6 8 7.9 14 34.2 — —

Total 164 100.0 101 100.0 41 100.0 46 ~=100.0 Note: ' Based on study of eighty-nine full sibling groups at least one of whom is married. Sources: For Terutenne, Yalman (1967: 41-2); For Rambukkoluwa, Tambiah (1965: 145).

married, 92 per cent in Terutenne and 100 per cent in Rambukkoluwa stayed on in their natal village after marriage. In addition, Ryan (1953: 153)

found in his study of a Kandyan Highland village that only eleven out of seventy-four marriages (or 15 per cent) were binna marriages. (He does not indicate what percentage of the diga marriages were intra-village.) The higher incidence of women contracting diga marriages and leaving the village relative to men indicates that women’s chances of inheriting were and are considerably lower than for men, to the extent that residence rules still dictate inheritance patterns.

Among the Low Country Sinhalese, the marked preference for diga marriages historically produced a tendency towards predominantly male inheritance. Tracing inheritance patterns backward for several generations in Madagama, Obeyesekere (1967) found that up to the beginning of the twentieth century, women were almost entirely excluded from inheriting landed property, residence almost always being virilocal. From the first decade of the twentieth century, for the first time even diga-married women began to inherit intestate. He attributes this to the diffusion of new legal ideas and the acceptance over time by villagers of the Roman-Dutch law, which although it had been in force since 1833 had not immediately affected

traditional practice. Now, however, bilateral inheritance and intestate succession increasingly became the rule. As a result, at the time of his village study in 1961, Obeyesekere found

that men resident in the village were constantly fearful that their diga married sisters, on the instigation of their husbands, would sell the ancestral land to non-kin, especially after the deaths of their parents. This fear was

Erosion and disinheritance 19] not entirely misplaced. In several cases, sisters had in fact sold their shares to non-kin, against the wishes of their brothers who wanted to buy the land

but for less than that offered by other buyers. Out-resident males living virilocally in the rural areas or neolocally in cities also often induced their wives to sell their shares. This led to a growing conflict between women’s

brothers and their husbands (who could be cross-cousins), and in some cases even to the severance of familial ties between siblings (Obeyesekere 1967: 259-60).

Of course disputes could also occur in cases of uxorilocal marriage, not

dissimilarly to those described in the next two chapters in relation to patrilineal communities. Selvadurai (1976) quotes two such instances where the brother did not give the sister her right to the ancestral land or to its produce. In one case (also mentioned in chapter 1), after an extended

dispute lasting over twelve years, the woman received half her rightful share, but in the process she spent more than the market value of the land that she and her husband had fought for. In the second case, the father had left only a small part of the rightful share to the woman and she received but a portion of the bequest. Selvadurai observes that people fought over rights to ancestral land through court cases, violence, and even sorcery, because

such land was valuable not only economically but also symbolically: continuity of land ownership stood for continuity of kinship and these together defined citizenship rights in the village. Mayne (1977), who surveyed a village near Colombo in 1977, provides some idea of women’s actual land ownership in the more recent past among the Low Country Sinhalese. Here the women owned only 23 per cent of the village paddy land (as residents or absentee landlords), 14 per cent of the highland (again as in- or out-residents), and 14 per cent of the rubber land (only as residents). If land was available, even virilocally married daughters were given some at marriage or were promised it at the death of the parent. In other words, women did inherit land, but not on an equal basis with men. A study by Weerasinghe (1985), who surveyed female-headed households

(FHHs) in 1984-85 in two villages near Colombo, also bears this out. Weerasinghe’s survey included women who were widowed, deserted or single, or had migrant or incapacitated husbands, or were married polygamously. Of her total sample of 202 FHHs in the two villages together, 25 per cent had some land (highland and paddy land), but half had less than an acre each and many had rights in unpartitioned plots. Obeyesekere (1967) has argued that the relative incidence of binna and diga marriage in a region 1s linked to ecology and reflects the attempts by communities and families to match available land resources with available family labour. Hence families with a scarcity of land and an abundance of adult males tend to expel their men in binna marriages, and those with enough land buta scarcity of family men seek to attract sons-in-law in binna

192 A field of one’s own marriages for daughters, and to keep their male members via diga marriages for sons. However, implicit in and critical to Obeyesekere’s schema is the assumption that it is male labour alone which is of importance in agricul-

tural cultivation. In fact he goes so far as to say ‘if there is a “family” without steady male help to cultivate it, it may mean indigence or starvation, given the traditional economic importance of rice’ (Obeyesekere 1967: 60, emphasis mine). In other words, he ignores the importance of female labour in cultivation.

As a generalization for the Sinhalese peasantry as a whole, and as a historical explanation, this is problematic on at least two counts. First, traditionally, chena or swidden cultivation was of considerable importance in the Dry Zone and provided the staple food, while settled paddy (a more high-risk crop in the area) was being cultivated alongside it as a supplementary crop. In chena, female labour was critical: Yalman (1967: 107) notes that ‘women do assist in the harvest, and they are felt to be indispensable on chena’. In settled paddy cultivation, there was more dependence on male

labour, but here again women’s input in weeding, harvesting, and the processing and storing of paddy, has always been significant.+> Even in Obeyesekere’s study village, which is located in the Maritime Province of Mattara and where settled paddy cultivation dominates, female labour is likely to have been of some importance, as borne out by other studies of the Low Country.37* Obeyesekere himself does not touch on female labour

input at all. In any case, he presents his thesis as a general historical explanation for the relative incidence of binna and diga marriage, irrespec- , tive of the region and period. Recognizing the significance of women’s labour contribution, however, would weaken the logic of his schema. Second, his explanation does not take account of the traditional instability

of Sinhala marriages, and the shifts between binna and diga marriages during a person’s lifetime. It appears doubtful that people would have been able to keep to a neat economic-ecological rationale in all marriages. This is

not to argue that the prevailing land/person ratio has no bearing on marriage preferences — clearly one would expect people to prefer marrying into villages where cultivable land is more plentiful and into families with

more land, but it does not provide an adequate explanation.

HT. In conclusion One of the striking features of the three case studies presented above is the vulnerability of women’s customary rights in land, even in matrilineal and 33 See Schrijvers (1988), Fellenberg (1965), and Ryan er al. (1955). 34+ See especially Abeyewardena (1986), Weerasinghe (1985), Moore and Wickramasinghe (1980), and ESCAP (1983).

Erosion and disinheritance 193 bilateral communities, to exogenous forces over which women could exercise little direct control. These forces included, in particular, changes in laws and the growing scarcity of land as a direct or indirect result of State policies in both the colonial and the post-colonial periods. Women’s lack of jural authority in traditional public forums such as caste and clan councils was replicated in the modern State’s judicial and executive structures. Even

ideological discussions around issues such as the appropriateness of polyandrous unions among the Nayars and the Sinhalese were carried out mainly in elite male forums and women’s opinions were not solicited. This leaves open the question: how did women themselves feel about the shifts? Unfortunately, the ethnographies mostly do not include women’s voices, although the court petitions to protect their land shares filed by women in Sri Lanka at the turn of the century, and by women in Kerala in the colonial period as well as during the 1950s’ land reform programme, indicate their resistance to the erosion of their rights. The effectiveness of such resistance was nevertheless limited by the facts that (a) women acted as individuals and (b) they were appealing to a legal and administrative machinery which excluded their direct participation in the process of decision-making. A somewhat different picture is provided in chapter 5, where the exertion of political pressure by women in groups, as in the context of the Hindu Code Bill, was critical in pushing forward gender-progressive legal reform. It enabled women’s voices to be heard and their perspectives to be noted. At the same time, now as then, women’s limited participation in the institutions where laws and policies are formulated and implemented means that

they still largely remain takers and not makers of many decisions that deeply affect their lives.

Appendix 4.1 A marriage proposal among the Christian Garos* (A variation on the traditional custom)°°

I A girl sends a letter to the boy whom she likes, requesting marriage. The first letter usually accompanies one written by her father. Rombagiri. Ist Dec. 1955 Dear nephew, Repraksogiyil Sangma, I am writing a few lines to you, enclosing my daughter’s letter to you, with

an idea of knowing your intention in your reply. If you could marry my daughter, Ganjak, I would be very happy. I don’t mean to say that my daughter is a wonderful girl, but I hope that you would kindly help your maternal uncle and consider my request. For a man it is necessary to takea partner in this world and to have a happy married life. If you receive this letter, kindly reply to us, as I am anxious to hear from you. With best wishes, I remain, Yours sincerely, Your uncle, Gimbilpa Sangma

35 See Nakane (1967: 71-2). Nakane writes: ‘I tried to find whether any couple might have kept some original letters after their marriage, but I unfortunately failed to find any. But I happened to know a man from a neighbouring village of Rombagiri, whom the people regard as a good writer as well as a narrator of old Garo stories. He willingly agreed to produce for me a set of typical letters of the Garo courtship. These are the letters I present here. The Christians of Rombagiri highly appreciated them, saying they were exactly the same style as theirs, but rather more beautiful than the letters which they used to write. The English translation was done by one of the Garo intellectuals at Tura.’ 36 It may be recalled (from chapter 3) that traditionally a popular way for a girl to propose toa boy was to have him captured by other village boys; and the prospective groom had to express reluctance by escaping and being recaptured twice more before accepting the proposal. 194

Erosion and disinheritance 195 Rombagiri. Ist Dec. 1955

Dear friend, Repraksogiil Sangma, I am writing a few lines to you today, and I hope you will pay attention to my words. As my father has written to you, I request you to take me as your wife. I am ready to serve you. I don’t claim to be clever and beautiful, but one must have a partner in this world. God created man and woman to live together. So we too must have a partner to live in this world. I hope that you will choose me as your partner. Please let me know if you have any other plan in your future in your reply. With best compliments, I remain,

Yours sincerely, Your friend, Ganjak Marak Reply:

Chokagiri. 5th Dec. 1955.

Dear friend, Ganjak Marak, I have received your letter and I like to let you know that I have no idea of choosing a wife for myself yet. Therefore, I would like to advise you to try

someone else instead of me. I am intending to go for further study and afterwards I have to support my own parents. Kindly excuse my inability to take you as my wife. With best wishes, Sincerely yours, Repraksogiyil Sangma

I] Rombagiri. 10th Dec. 1955

Dear friend, Repraksogijil Sangma, I have received your letter, dated Sth, inst. Iam sad, you have not shown any intention to have me as your wife. I request you once more to consider

this matter. I understood that there is no other man better than you who could be my ‘husband, from my father’s words. You are so capable in managing affairs and working in the field. Iam a helpless girl, without your protection, itis very difficult to carry on my life. I shall obey all your orders. This is my duty. I have no property. I am poor. Although I am in such a position, if you consider taking me as your wife, I shall be very grateful. But

196 A field of one’s own it you do not consider this matter and do not take me, I shall be in great sorrow. I shall be nowhere. Expecting your kind reply, with best regards, Yours sincerely, Ganjak Reply:

Chokagiri. 15th Dec. 1955

Dear friend, Ganjak, I have received your letter, dated 10th, inst. I have read all what you have

written. But I am in difficulty to take you as my wife, because I have no property or education. Your family is big. I don’t know how to work. If you

have me, you will lose all your property. There will be many difficulties under such circumstances for both of us. Please don’t write any more, give up your writing to me and try to have another better boy. It will be useful for you. With best wishes, Yours sincerely, Repraksogiyil Sangma ITI

Rombagiri. 20th Dec. 1955

Dear friend, Repraksogijil Sangma,

I have received your letter and I feel so sad about it. What you have written shows your indifference to me. I have already chosen you as my partner. It is impossible to think of another man. Though we are poor, if you come to my house, you would feel very much at home with my father, your maternal uncle, and my mother and uncles and brothers will be very kind to you. I shall do my best to serve you. Please think over it once again!

In the night when stars are shining in the sky, sorrow fills my heart, thinking of you. I dream of happy days when you come to my home and we work in the jhum field together. But I know you have no mind to take me. Oh! when we will be united! Please do make a fresh decision to make me

your wife, rather than to continue to work for your parents. I have lots to tell you, but please let me stop here. Your lost friend, Ganjak

Erosion and disinheritance 197 Reply:

Chokagiri. 28th Dec. 1955

Dear friend, Ganjak, Just a few words to tell you that I was extremely glad to receive these three letters which you have written to me. I can now realise all the facts. So I have decided to marry you. I shall come to your house on 2nd January 1956. But

please let me know whether your uncles, brothers and elders are really agreeable to our marriage or not. I was very pleased with the way you wrote to me. Please let me know when you desire to have our marriage ceremony.

In the moonlit night I think of you very much. Please keep to your statements. I believe you are the only one with whom I can share my heart. I am closing here with best wishes and sweet expectation.

Yours lovingly, Repraksogiil Sangma

5 Contemporary laws: contestation and content

[T]housands of sensitive Hindu women ... for the first time in their lives

left the precious sanctuary of their sheltering homes [during India’s freedom struggle]. They came to the battle field and stood beside their brothers and faced jail and lathi charges and often enough, humiliation worse than death. If today ... [they] who fought for the independence of India are to be denied their just rights, then our hard-earned freedom is no more than a handful of dust. (Padmaja Naidu (Congress legislator), Parliamentary debates over the Hindu Code Bill in 1951)!

May God save us from ... having an army of unmarried women. (M.A. Ayyangar (Congress Legislator) predicting the result of daughters getting property during the Parliamentary debates on the Hindu Code Bill in 1951)

The formulation of contemporary inheritance laws on landed property has involved a complex process of interaction between the (colonial and postcolonial) State and different segments of the population, the interplay of varying ideologies and interests, and the conflicting pulls of scriptural rules and local custom. Around the early part of this century, these interactions increasingly took the form of explicit contestation, especially over women’s property rights, revealed most prominently (but not only) in the pre- and

post-Independence debates surrounding the Hindu Code Bill in India. Large numbers of Indian women participated in an organized campaign to

expand women’s rights, including property rights. Many Indian men supported the cause; the majority adamantly opposed it. In their responses, British functionaries of the colonial State, although primarily concerned with the regime’s economic and political interests, were also influenced by the notions of family, morality, and proper gender relations prevailing in England, with mixed effects on women’s position. Today, laws governing the inheritance of landed property form a collage across the legal map of South Asia. They vary by religion and region, both ' Parliamentary Debates, 20 September 1951 (see GOI 1951b: 2930). Lathi charge: policemen armed with wooden batons charge a crowd to break up a demonstration. 2 Parliamentary Debates, 7 February 1951 (see GOI 19Sla: 2530). 198

Contemporary laws 199 within countries and between countries across the subcontinent. In part these variations originate in the colonial State’s characterization of inheri-

tance and marriage laws as ‘personal laws’ applicable to members of particular communities; this characterization has endured in post-Indepen-

dence codifications that have attempted to accommodate differences in religions and local cultural traditions, while also seeking to establish a degree of uniformity. And partly (in India) the variations stem from different rules in relation to the inheritance of agricultural land, in various tenurial enactments passed by provincial (state) legislatures, which in 1935 were given legislative powers over such land. The discussion below will focus first on the process of contestation by which contemporary laws were formulated, and then outline the existing

laws and identify persisting gender inequalities therein, especially in relation to women’s rights in land. Inheritance laws in India (for nonMuslims), Sri Lanka, and Nepal will be considered separately by country. The laws governing Muslims in Pakistan, Bangladesh, and India will be discussed jointly, both because of their common historical evolution under colonialism, and because the majority of Muslims in the three countries are governed by the same school of law, the Hanafi school. At the same time,

significant legal differences between the three countries will also be highlighted.

I. India (1) The formulation of contemporary Hindu law Late eighteenth to early twentieth century. Before British rule, local customs, influenced in varying degree by the shastras and the ancient commentaries, formed the basis of Hindu law (as outlined in chapter 3). Shastric influence was strongest among the upper castes, especially the Brahmins, and appears to have been greater in the northern and eastern parts of India than elsewhere (Carroll 1989; Derrett 1968). Most disputes were settled locally by village or caste councils serving as local tribunals

(Galanter 1989). Under the British, the adaptability associated with customary systems declined. In 1765 the East India Company established control over eastern India (Bengal, Bihar, and Orissa); and soon after, in 1772, Governer-General Warren Hastings directed that the scriptural texts of the Hindus and Muslims would be the basis of legal governance: the Koran would apply to the Muslims and the shastras to the Hindus (Derrett 1968: 288-9). To implement this, for the Hindus the British drew partly on Brahmin pandits to help interpret and administer Hindu law, and partly on a diverse collection of juridical texts, including translations of the shastras

200 A field of one’s own and the ancient digests and commentaries, and a variety of new legal texts (some written by the pandits) forming what Derrett terms ‘modern shastric literature’.? Judicial decisions also set precedents which guided subsequent

judgements. By the mid nineteenth century, the texts had replaced the pandits as repositories of authoritative Hindu law (Carroll 1989). Over time, the British also came to recognize that traditionally there had

been no fixed Hindu law applicable uniformly to all Hindus; quite the contrary.* Customs, often at variance with the shastric rules, prevailed in

most parts. To take cognizance of this, regulations were passed giving precedence to custom over the shastras. But custom, being unwritten, was usually difficult to establish in court. To prevail over written law, custom

had to be ‘proved to be immemorial or ancient, uniform, invariable, continuous, certain, notorious, reasonable (or not unreasonable), peace-

able, obligatory and ... not ... immoral nor opposed to an express enactment ... or to public policy’ (Kane 1950: 44).> Little appears to have been done to collect authentic records of local customs on a systematic basis, although there were some detailed records available for selected parts of the country, especially the Punjab. In that region, for instance, the works of Rattigan (1953, first published in 1880) and Baden-Powell (1896) show that there were a variety of local customs in vogue which differed from village to village and caste to caste, ‘sometimes quite opposed to the later Hindu ideals’ (Baden-Powell 1896: 102). Typically, the source of British information on customs was the testimony of village elders or members of local elites who served as spokespersons and often also as interpretors of local custom for the colonizers, ‘but when these elders [were] once called upon to give their evidence, [the customs] necessarily [lost] their old position

... (Maine 1889: 72). The very process of collecting, recording, and juridically interpreting customs changed their character, fixing what had been flexible (Cohn 1965). It also privileged some customs over others. The patchwork information on customs, the strict proof required to establish customs, and their very diversity meant that in practice the judges were forced ‘more often than not, to fall back on the precepts of the smritis under the influence of their pandits, or rather to the interpretation of these precepts which had been established locally’ (Lingat 1973: 137). Also, as Carroll (1989) points out, the majority of judges in the court, both British and Indian, shared a Brahminical view of Indian society, derived by the 3 Derrett (1968) describes this process, detailing the ways in which the British served as patrons of the shastras. 4 Among the writers who had pointed this out most forcefully was Nelson (1877, 1886). > Carroll (1989) argues that while strong evidence was needed to prove custom which differed from orthodox Hindu law, such evidence was not insisted upon if the custom coincided with that law. This would have strengthened the shift of legal practice in the direction of orthodox Hinduism.

Contemporary laws 201 British from the legal texts they had studied and by the Indians from their own typically upper-caste origins. This emphasis on the shastric view on the one hand, and on selected incorporation of custom on the other, is argued to have privileged the shastric and elite interpretation of law, introducing

thereby a Brahmanical and elite bias in the settlement of legal disputes.° And oncea line of decisions had been established, the court was reluctant to depart from it. A body of case law or ‘judge law’ was thus constituted. From the nineteenth century onwards, a series of legislative acts were also passed.

Essentially what emerged was a cross-breed of classical Hindu law, customary law, and statutory law, broadly termed as Anglo-Hindu law. The law enforcement apparatus also underwent a major change with the

establishment of western institutions and procedures, including lower courts, District Courts, High Courts, and a system of appeals ultimately _ leading to the Privy Council in London. Initially, untrained Company officials served as judicial officers, but after 1790 English judges with legal training were brought in (Parashar 1992). Also initially the British held all

the important judicial positions, although over time Indian men too became significant practitioners in the system.

These new legal institutions eroded local ones operating on a caste, religious and regional basis. Moreover, as Cohn (1965: 109) notes: Every time new regulations, interpretations, or legislative enactments came into force, the structure of social relations for the bulk of the population was affected, usually in ways not anticipated by the lawmakers. Landed property became a commodity, new groups — urban-based landlords, bankers, merchants, money-

lenders — who previously had had minor roles in the rural society, came to prominence.

What were the effects of these changes on women’s property rights? To begin with, shifts away from custom in favour of the shastric versions of law, both prior to and during British rule, would have had some negative effect on women’s rights in landed property at two levels: (a) regionally, notably in the south and west, where custom favoured women in several respects, and (b) within some communities, principally the lower castes and tribes. For instance, under shastric law a Hindu widow had to forfeit any interest in her husband’s estate in case of unchastity, and widow remarriage was prohibited. But among the lower castes and tribes widow remarriage

was widely practised, and this did not always necessitate forfeiting the husband’s estate.’ The application of shastric rules would have led to the disinheritance of widows in such communities. Judicial decisions following © See Derrett (1968), Lingat (1973), Carroll (1989).

” See Carroll (1989) for some evidence, albeit limited, that among some lower castes, in a number of circumstances, custom did not entail forfeiture on remarriage.

202 A field of one’s own the Hindu Widow’s Marriage Act of 1856 graphically bring this out. Under this Act, Hindu widows were allowed to remarry, but they thereby lost their

limited interest in the husband’s estate. The Act was necessary to legally enable upper-caste Hindu women to remarry. However, Carroll’s (1989) examination of High Court cases shows that all High Courts, except the one

at Allahabad, interpreted the Act as applicable to all castes, and directed forfeiture on remarriage even for those lower castes for whom the Subordinate Judge (that is the judge in a lower court) had established that custom permitted the widow to remarry and to retain the deceased husband’s property after remarriage. Such women lost out in High Court judgements (except those of Allahabad).® It is not possible to say how widespread this effect was, since there is inadequate information on the extent of such customs or on how many such communities were involved in litigation. But, as Carroll (1989: 17) argues: [T]he result was undoubtedly the displacement of Customary Law as regards remarriage and the establishment in its stead of Brahmanical values which held widow marriage in disrepute and insisted on some penalty (in this case forfeiture of inheritance rights) being imposed for a breach of the preferred norm of the chaste, prayerful widow.

After 1880, by a Privy Council ruling, chastity ceased to be a condition for the widow to maintain her life interest in the husband’s estate. Ironically, this also meant that a remarried Hindu widow was now in a worse legal position than one ‘who lived in “notorious unchastity”’ but wisely did not permit her paramour to make her an honest (and property-less) woman’ (Carroll 1989: 26).

British endorsement of widow remarriage took a different turn in the Punjab, where they sought to protect local customs in order to stabilize their own political interests in the region.? Here leviratic unions were customary among landowning castes, most notably the Jats.!° Through such a union (locally called karewa), the husband’s kin retained control over the land in which the widow (in the absence of male lineal descendants) 8 The Allahabad High Court took the position that the Act did not apply to castes customarily allowing widow remarriage, and that forfeiture could not be enjoined in the absence of proof that such forfeiture was needed by custom. ° For fuller elaboration, see section II of the chapter. 10 This form of marriage was formalized by a simple ritual, such as the prospective husband placing glass bangles on the widow’s wrist before a public gathering, sometimes accompanied with the distribution of sweets. It was considered a valid but inferior form of marriage, and did not involve a religious ceremony, which Hindu women could not customarily undergo twice. Although typically karewa was with the husband's younger brother, remarriage with some other male relative of the deceased, such as a cousin, was not unknown. Today such unions continue to be practised in northwest India (see Dreze 1990, and chapter 8).

Contemporary laws 203 had a limited interest.1! Otherwise there was a clear risk that such land would be lost to the family. Customarily (as noted in chapter 3) the widow

could ask for the partition of her husband’s share from the joint family estate if she could prove that she was not getting adequate maintenance from her husband’s family.!2 Although she was meant to have only a limited interest in such an estate, the property could in fact be alienated by her (e.g. mortgaged or sold) if she could prove the ‘legal necessity’ of doing so. This meant proving that the income from the estate was insufficient for such purposes as her own maintenance, a daughter’s marriage, paying land revenue to the State, or repaying a ‘just’ debt (Rattigan 1953, Rustomyi 1942).}3

In the early twentieth century, a number of widows sought to use this proviso to seek partition and the freedom to alienate their land if necessary, but their appeals to the British administration were usually received with little sympathy. The Punjab Land Administration Manual of 1908 noted that to contain women’s fast-growing claims to partition, it was necessary to ensure ‘a firm anchoring of the widow in remarriage’. This could be ‘the only satisfactory arrangement against which she had no appeal’ (cited in Chowdhry 1989: 316). Chowdhry argues that in the interests of consolidating their political position in a region that was of economic and strategic interest to them, the British felt it necessary to protect the local peasantry’s hold over land and so minimize the danger of social disequilibrium; they feared that allowing widows to have their way would weaken this hold and antagonize the landowning groups. Petitions by widows against forced levirate also became common at the turn of the century, indicating their resistance to the custom. However, it was not easy for women to establish that such a union had not taken place, since even cohabitation with the brother-in-law was recognized as karewa, '! Among certain communities, such as the Sikh Jats, the widow did not have to forfeit her limited interest in her first husband’s estate if the Aarewa form of remarriage was to the husband’s brother, but such a forfeiture was necessary if she remarried some other relative of the deceased. In such communities, either way, the husband’s brothers kept control of the estate (Rattigan 1953: 480). '2 There were slight differences between different districts of the Punjab in the conditions under which they allowed partition by the widow. Most districts recognized the Hindu widow’s right to ask for partition if she was sonless but not if she had sons (Rattigan 1953: 311-17). Some groups denied allowing partition, but actually practised it. For instance, in Rawalpindi district, when Rattigan enquired about the prevalence of the custom, the answers received were at variance with actual practice. He noted: ‘the examples of such partitions are so numerous that there can be no doubt that the right to claim partition ts well established by custom’ (Rattigan 1953: 316). 13 A ‘just debt’ was defined as a debt that was not ‘immoral, illegal or opposed to public policy’ or ‘contracted as an act of reckless extravagance or of wanton waste or with the intention of destroying the interests of the reversioners’ (Rustomji 1942: 331).

204 A field of one’s own and it was only the widow’s word against that of the husband’s kin. The

district officers of the British administration were instructed to have ‘nothing to do’ with such petitions. Thus ‘the customary law of... . [Punjab],

backed by the full force of the colonial administrators, safeguarded the landed property from a woman’s possession’ (Chowdhry 1989: 319).!4 A related issue was the restriction of women’s inheritance rights to only a limited estate. Sarkar’s (1991) compilation of some of the cases dealt with

by the Privy Council and High Courts during the latter half of the nineteenth century gives us an idea of official leanings on this count, although the cases discussed are too few for generalization and relate largely to women of wealthy families. The courts’ decisions present a mixed

picture: the rights of daughters and widows appear to have been upheld only with respect to the deceased’s acquired property, not the joint estates;

and typically, in that period, the women were entitled to only a limited interest.!5 But in the early part of the twentieth century some rulings began to recognize widows and daughters as absolute owners of any immovable property that had been willed to them by husbands or fathers. At that time, some jurists had also begun to argue that restricting the widow’s estate toa

limited interest was not demanded by the shastras, but had only been emphasized by a few of the later commentaries which British judicial decisions had helped entrench (Sarkar 1991). A more detailed study of cases would be needed to establish whether the rulings varied regionally and at different levels of legal decision-making, including the District Courts. In the nineteenth century, a number of factors, but especially the social

reform movements led by several prominent Indians, and pressure for appropriate legal intervention, resulted in the British passing several pieces of legislation that impinged on women’s status, including the abolition of

sati in 1829, the legalization of widow remarriage in 1856, and the prohibition of infanticide in 1870.!° Criminal and commercial law too was codified by the British in the latter half of the nineteenth century. However, most aspects of laws governing inheritance and marriage among Hindus were left untouched. 14 The attitudes of some British male administrators towards women’s property rights in India were not independent of their disapproval towards British women’s steady acquisition of property in England in that period: ‘Now, whereas. a man has through the force of tradition and social custom, a tendency to spend his money for the benefit of the woman. the woman has no traditional tendency to spend her money for the benefit of the man. The

consequence Is that, in enjoying the benefits of little comforts and luxuries, woman in | England 1s steadily increasing her advantage over the man and the effect of this process on the relative male and female mortality can hardly be negligible’! (views of Col Forster. Director of Public Health, Punjab, quoted in Chowdhry 1989: 319). '5 The cases reviewed by Chaudhary (1961) relating to this period also indicate this. '© For a broad overview of the nineteenth-century social reform movements in India and also of changes in women’s rights in Sri Lanka during this period, see Jayawardena (1986).

Contemporary laws 205 The early part of the twentieth century saw both an increasing assertion by women (including peasant women) within patrilineal communities of

their limited legal rights in property, and growing efforts by women (especially the urban educated) to expand those rights, most notably through the codification of Hindu personal law. This demand was subjected to intense contestation, as outlined below.

Early to mid twentieth century. By the early twentieth century, several factors led to an intensification of pressure for changes In women’s legal status.!7 Among them was the founding of several women’s organizations, most notably the Women’s Indian Association (WIA) in 1917, the All India Women’s Conference (AIWC) in 1927,!8 and the National Council of

Women in India (NCWI) in 1925. Of these the AIWC was the most prominent during this period. By the mid-1930s, the ATWC and WIA jointly had over 10,000 members. These women’s organizations worked for social reform legislation, opened schools for girls, and demanded women’s suffrage. After a successful effort to get the Child Marriage Restraint Act of 1929 passed, they focused more directly on women’s rights to divorce and to inherit and control property.!? Forbes (1981: 71) notes: Throughout the 1930s the women’s organizations formed committees on legal status, undertook studies of the laws, talked with lawyers, published pamphlets on women’s position, and encouraged various pieces of legislation to enhance women’s status. At first these demands were presented as part of the organizations’ general '7 The story below, of the role played by women’s organizations and progressive male reformers in pushing forward legal reform for the benefit of women during this period, has been constructed from a range of sources: Constituent Assembly Debates (GOT 1949), Parliamentary Debates (GOI 195la-b), The Rau Committee Reports (GOI 1941, 1947), Basu and Ray (1990), Everett (1979), Forbes (1981) and Mies (1980), supplemented by accounts by some of the women who participated in the struggle for reform, including Chattopadhyay (1983). '8 British women, notably Margaret Cousins, Agatha Harrison and Eleanor Rathbone, played an important role in the establishment of WIA and AIWC, and continued to

support Indian women’s efforts at legal and social reform throughout this period (Ramusack 1981; Cousins 1947). 19 Apart from national women’s organizations, such as the AIWC and the NCWI, local ones also emerged during this period, often despite considerable male opposition. For instance, in the first two decades of the twentieth century, as demand for social reform grew in Kerala

(discussed in relation to the Nayars in chapter 4), some Nambudiri women formed organizations demanding the right of every woman within the community to marry, to inherit a share of the property and to education. These women were ridiculed by some of the conservative men of the community. One argued: [T]he development of women’s organizations and newspapers is a meaningless farce .. . the doctrine that ‘women must speak about their interests themselves’ will be shown up to be nonsense ...if women stake their claims separately from men and wish to achieve them by

contesting against men, then men will need to protect their interests from women by forming ‘men’s organizations’. (Arunima 1992: 295)

206 A field of one’s own efforts to uplift women, but by 1934, the AIWC passed a resolution demanding a . Hindu Code that would remove women’s disabilities in marriage and inheritance.

Also, among the Indian lawyers elected to the government’s Central Legislative Assembly after its establishment as a legislative body in 1935 was a group of liberals concerned with social and legal reform. These reformers sought to introduce a number of bills, including bills supporting Hindu women’s right to divorce and Hindu widows’ right to a share in their husbands’ property. These bills encountered strong opposition from the orthodox Indian members of the Assembly and were defeated a number of times. The liberals had to seek the support of the colonial government to bypass this opposition, and The Hindu Women’s Rights to Property Act of 1937 was a compromise.?° It gave the Hindu widow, who had previously been excluded from inheritance by the son, agnatic grandson and agnatic great-grandson of her husband, a right to intestate succession equal to a son’s share in separate property among those governed by Mitakshara, and in all property among those governed by Dayabhaga. It also gave her the

same interest as her deceased husband in the undivided Mitakshara coparcenary, with the same right to claim partition as a male coparcener, but she could hold this share only asa limited interest.2! Three categories of widows were recognized: the intestate’s widow, the widow of a predeceased son, and the widow of a predeceased son of a predeceased son. However, the widow’s share (as noted) was only a limited estate which she could enjoy

during her lifetime, after which it went to her deceased husband’s heirs; it was subject to forfeiture on remarriage; and it explicitly excluded agricultural land. (After the 1935 Government of India Act, agricultural land, as noted, came under the jurisdiction of the provincial legislatures.)?? Also daughters were excluded from the purview of the Act. ?° For details of the Act see Kane (1946). 71 As explained in chapter 3, under the traditional Mitakshara system ancestral property was held jointly by four generations of male lineal descendants in the male line of descent who became coparceners on birth. Any coparcener could unilaterally sue for partition. Women, however, could not be coparceners. Under the Dayabhaga system a man was absolute

owner of all his property (ancestral or self-acquired) with full rights of disposal, and division could take place only on his death. Under both systems a widow could not inherit any property in the presence of sons, sons’ sons, and sons’ sons’ sons. In their absence she got a limited estate in her husband’s separate property under Mitakshara and in all of her husband’s property (including his share in joint family property) under Dayabhaga. A daughter came after the widow under both systems and also got only a limited estate, again

only in the father’s separate property under Mitakshara and in all his property under Dayabhaga. 22 A few states subsequently (some prior to, others soon after Independence) passed laws extending the Hindu Women’s Rights to Property Act 1937 to include agricultural land. These included Bihar, Hyderabad, Orissa, the United Provinces, and Bombay. However, the efficacy of this extension was undercut by a clause that such legislation would not affect any rule of succession prescribed for tenants’ rights in agricultural land by any special law

then in force.

Contemporary laws 207 Somewhat earlier, the states of Mysore and Baroda passed their own legislation enhancing Hindu women’s inheritance rights, especially by giving them absolute rights to stridhan.*3 V.V. Joshi, a leading sanskritist and member of the Baroda Committee for Hindu law reform, also wrote an influential pamphlet arguing for comprehensive legislation on women’s property rights. It is noteworthy that these early reforms came from south and west India where women’s property rights (among patrilineal groups) had traditionally been stronger than elsewhere.

These developments formed the backdrop to the intense contestation which continued into the 1950s over the codification of Hindu law.24 In the earliest stages of this campaign, the noted women’s organizations focused mainly on child marriage and its negative effects on women’s physical wellbeing and education; they called for raising the age of consent.?°> Subsequently, the demands were broadened to include wider reform of marriage laws as well as improvements in women’s property rights, although there

was no demand yet for equal property rights. Till this point Hindu law reform was being sought on the grounds that it would enable women to make a larger contribution to society as well as relieve women’s sufferings —

consistent with what I termed the efficiency and welfare arguments in chapter 1. However, soon afterwards the calls for inheritance and marriage reforms were made explicitly on grounds of gender equality.7° As a first step towards codification the women called for the setting up of

a government commission that would examine women’s position in personal law and suggest measures to remove existing gender disabilities. Towards this end, they sought to mobilize public opinion by publishing articles in English language periodicals, meeting with politicians, attending

Legislative Assembly sessions when bills concerning women’s status in Hindu law were introduced, and presenting resolutions to government officials. Interestingly, in a set of resolutions presented in 1940 by the Women’s Subcommittee to the National Planning Committee (a group appointed by the Congress Party to think about directions for postIndependence economic and social development), one member, Kapila Khandwalla, wrote a dissenting note from a communist perspective, 23 See The Mysore Hindu Law Women’s Rights Act, 1933 (Mysore Act 10 of 1933). A similar Act was passed in Baroda state in 1937 (Everett 1979: 146). 24 Although in the 1940s the women’s organizations also called for a Uniform Civil Code for all religious communities, much of the discussion in fact focused on the Hindu Code. Fora detailed discussion on the issue of the Uniform Civil Code, see Parashar (1992).

25 An earlier effort to prevent the consummation of marriage before the age of twelve, through the Age of Consent Act of 1891, had had little impact. 26 In this shift, the noted legal reforms in Mysore and Baroda and V.V. Joshi’s 1933 pamphlet advocating gender-progressive reform, are believed to have been especially influential (see Basu and Ray 1990, and Everett 1979, who also give more details about the positions taken by women’s organizations and male lawyers).

208 A field of one’s own Opposing property law reform on the grounds that private property should be abolished altogether in India (Khandwalla 1947: 238-9). This type of argument, as noted in chapter 1, was echoed by some left-wing women’s groups in the 1980s. In 1941 the government set up the Rau (Hindu Law) Committee, a move that the women’s organizations supported even while they protested the absence of women on the Committee. The Committee in the first instance was appointed to suggest how the Hindu Women’s Rights to Property Act of 1937 should be amended especially in order to clarify the nature of rights conferred by it upon the widow and to remove any injustice done by it to the daughter. Noting the Act’s many technical defects and ambiguities which could lead to varying interpretations of women’s rights, the Committee felt that any attempt at piecemeal amendment would raise ‘all the controversies latent in the Act’ (GOT 1941: 23). Instead it strongly recommended that a

complete code of Hindu law be prepared, beginning with the law of inheritance and followed by the law of marriage and other aspects of Hindu law. The code as envisaged by the Committee would be one ‘which ...

recognize[d] that men and women are equal in status with appropriate obligations as well as rights’ (GOI 1941: 24).

The timing of the Committee’s appointment appears to have been unfortunate. Soon afterward the Congress Party intensified its civil disobe-

dience movement and boycotted the Legislatures, and large numbers of Congress members were jailed. To support a Committee appointed by the British government implied cooperating with the colonizers. This presented women with a difficult choice between a struggle for their gender-specific rights and the call of nationalism, and posed a special dilemma for women

who were members of both Congress and the AIWC. At the same time, AIWC members had learnt that not many among the nationalists were their allies when it came to codifying the Hindu law, since giving women legal rights in property and divorce posed a serious threat to male authority. Some women argued: ‘Today our men are clamouring for political rights at the hands of an alien government. Have they conceded [to] their wives, their

own sisters, their daughters, “flesh of their flesh, blood of their blood’, social equality and economic justice?’ (Forbes 1981: 74). Many women went on to support the Committee. In January 1944 the government reconstituted the Rau Committee, this

time for preparing a Hindu Code. AIWC carried out a countrywide campaign in favour of codification and submitted a draft memorandum to the Committee. In August of the same year, the Committee came out witha

Draft Code. Its main provisions were (GOI 1947): abolition of the Mitakshara right by birth and principle of survivorship; equal property shares for the sons and widow of the deceased, and half the sons’ shares for

Contemporary laws 209 Table 5.1: Summary of oral and written opinions on the Draft Hindu Code received by the Second Rau Committee, 1945

Draft Absolute Hindu estate

Code for widows Monogamy Divorce!

No. % No. % No. % No. %

Total Against For 22437 49 31 75 43 108 36 375 63 107 69 99 57 195 64 Women? For 32 71 10 59 21 68 Against 13 29 7 41 10 32

Men? For 192 35 39 28 54 38

Against 362 65 100 72 89 62

Notes: ' On this clause the data from most regions were not disaggregated by sex. 2 Includes both individual women and women’s organizations. 3 Includes both individual men and organizations other than women’s organizations. Source: Report of the Hindu Law Committee (see GOI 1947: compiled from pp. 82-181).

the daughters in all intestate inheritance; an absolute estate for the widow (as opposed to a limited interest); introduction of monogamy asa rule; and legalization of divorce under certain circumstances. Succession to agricultural land was, however, excluded from the scope of the Draft Code on the gtound that this issue fell within the purview of the provinces. There were ‘black flag’ demonstrations opposing the Code in five cities. Reactions from women were mixed: the AIWC supported the Draft Code, while advocating equal inheritance for sons and daughters. The NCWI and several other women’s groups, especially from Bombay and Calcutta, as well as many individual women (including a number of Advocates) also supported the Code; but women in orthodox associations such as the All India Hindu Women’s Conference opposed it. Among men, although some

supported the Code, the majority (including prominent lawyers and pandits) argued against it on grounds such as: abolishing the Mitakshara would adversely affect commercial enterprise; the divorce provisions would

undermine the family; women were incapable of managing property and were likely to be duped by male relatives if given an absolute estate; married daughters already received a share as dowry, and unmarried daughters only needed maintenance and provision for their marriage expenses; and so on. Only a small percentage of those whose views were recorded by the Second Rau Committee were women or women’s organizations, but the gender divergence in those views was marked (see table 5.1): 71 per cent of the women (or women’s organizations) and only 37 per cent of the men (or

organizations other than women’s organizations) supported the Bill.

210 A field of one’s own Support did not mean agreement with all the provisions in the Draft Code, and on specific clauses the gender gap was equally glaring. Despite the strong opposition the Rau Committee submitted a revised draft of the Hindu Code Bill (HCB) which was introduced in the Legislative Assembly in April 1947. India became independent four months later. In April 1948, a further revised Hindu Code Bill was introduced in the new Parliament and was again the subject of intense debate. While the ATIWC continued to campaign actively for the Bill, its constituency was largely confined to the literate, urban population, as it had been before Indepen-

dence. This reduced the effectiveness of its campaign and left the Bill vulnerable to the opposition labelling it as an elite demand. One Congress legislator from West Bengal, who was especially vociferous in his oppo-

sition, characterized those supporting the Bill as ‘a few ultra-modern persons who are vocal, but have no real support in the country’, and implied

that only women of ‘the lavender, lipstick and vanity bag variety’ were interested in the Bill.2” He argued: ‘[If the daughter inherits] ultimately the family will break up’ and queried: ‘Are you going to enact a code which will

facilitate the breaking up of our households?’ (GOI 1949: 1011). In September 1951, of the legislators who spoke on the Bill ten supported it and nineteen opposed it (Everett 1979: 172). Most top Congress leaders of

independent India were against the Bill, including the Home Minister, Vallabhbhai Patel, and India’s first President, Rajendra Prasad. Dr Prasad threatened to withhold his signature on the Bill — an action that could have resulted in a constitutional crisis.

In the face of such opposition, Prime Minister Jawaharlal Nehru, although committed to the Bill, shelved it in 1951. The Law Minister and framer of India’s constitution, Dr B.R. Ambedkar, resigned in protest. However, after 1951, riding on the strength of a Congress electoral victory,

Nehru was finally able to win passage for the important aspects of the Hindu Code Bill in four separate Acts.?® Of these, the Hindu Succession Act of 1956 forms the basis of Hindu succession laws today. Everett (1979: 166-7, 181) provides some interesting insights into the

contrasting images of the ideal Hindu woman that the supporters and opponents of the Bill appeared to hold: From the (1940s and 1950s] debates [on the Hindu Code Bill] two different images of ideal Hindu women emerged. The opponents’ image resembled the view of women presented in the Manusmnriti: she needed the protection of men during all the periods 27 See statements by Pandit Lakshmi Kanta Maitra in the Constituent Assembly of India (Legislative) Debates on the Hindu Code, | March 1949 (GOI 1949: 996-7). 28 These were The Hindu Marriage Act 1955; The Hindu Succession Act 1956; The Hindu

eee and Guardianship Act 1956: and The Hindu Adoptions and Maintenance Act

Contemporary laws 211 of her life (thus never capable of independently looking after property), and in this position of dependence she was worshipped as a goddess. The proponents’ image of the ideal Hindu woman was a competent, autonomous human being interacting with others on the basis of equal rights and individual freedom. This image stemmed from Western liberal thought, however imperfectly it had been achieved in practice in the West.

The HCB opponents believed that the interests of men and women were better served when women occupied a dependent position and men and women played different social roles. The HCB supporters believed that everyone’s interests were better served when men and women were independent and enjoyed equal rights ...

The HCB supporters operated within the equal rights perspective which had emerged as the dominant women’s movement ideology since the 1930s.

Women’s organizations also mobilized to win constitutional guarantees in the area of personal law. In 1945 while the Rau Committee’s draft Hindu Code was being discussed around the country, a new Constitution was being drafted by the Constituent Assembly. Hansa Mehta, in her Presiden-

tial address to the AIWC in December 1945, formulated a ‘Charter of Indian Women’s Rights’ advocating that equality between the sexes should be the basis of citizenship in India (Mehta 1981). This was again an area of intense contestation, since there were possible contradictions between the proposed constitutional clause promising freedom of religious practice and propagation (which could be read to include religiously sanctioned inegalitarian property and marriage laws), and the aim of social reform towards gender (and caste) equality. The matter was finally resolved by an explicit

government statement that freedom of religion did not preclude social reform.?° However, in the 1950s the struggle for gender equality in property laws was by no means over, and that for ensuring women’s de facto property rights had not even begun. Hindu personal law today. Today the property rights of Hindus are governed by the Hindu Succession Act of 1956 (applicable to all states other than Jammu and Kashmir?° and covering about 86 per cent of the Indian population).3! In the Act, ‘Hindus’ were defined as including Sikhs, Jains, and Buddhists. However, the Act has special provisions for Hindu matrili-

neal communities customarily governed by the Marumakkatayam and Aliyasantana systems, as well as for the Nambudiri Brahmins. (Tribal 29 The Indian constitution’s guarantee of no discrimination on the basis of sex has, in fact, been used to challenge continued gender-related inequalities in property laws: a case in point is that of Mary Roy, discussed later. 3° In this state, the Jammu and Kashmir Hindu Succession Act, 1956 (Act No. 38 of 1956) applies, which (with some modifications) contains most of the provisions of the Hindu Succession Act of 1956. 31 For details of the Act see especially Mulla (1982).

212 A field of one’s own communities of the northeastern States of Arunachal Pradesh, Manipur, Meghalaya, Mizoram, and Nagaland are not covered by the Act and they continue to be ruled by their local customs, which are still in large part uncodified.)** The Act sought to unify the Mitakshara and the Dayabhaga systems, and purported to lay down a law of succession whereby sons and daughters would enjoy equal inheritance rights, as would brothers and sisters. In fact, significant gender inequalities remain. Under the Act, in the case of a Hindu male dying intestate, all his separate

or self-acquired property, in the first instance, devolves equally upon his sons, daughters, widow, and mother. In addition (and simultaneously with the mentioned four categories of heirs), if there is a predeceased son, his children and widow get the share he would have received if alive; the children of a predeceased daughter get her share likewise; and the children and widow of a predeceased son of a predeceased son similarly inherit a share as representatives of the deceased in question. All these are the primary or Class I heirs under the Act.33 In the absence of Class I heirs, the property devolves on Class IT heirs and in their absence first on agnates and

then on cognates.3* For joint family property, if the deceased male was earlier governed by the Dayabhaga system the same rules of succession as relate to other types of property apply to this as well. However, for those previously governed by Mitakshara law, the concept of Mitakshara coparcenary property devolving by survivorship continues to be recognized, with some qualifications: in the case of an intestate male who has an interest in Mitakshara coparcenary at the time of his death and who leaves behind Class I female heirs, or male relatives specified in Class I as Claiming through Class I female heirs, his interest devolves not according to the Mitakshara principle of survivorship but according to the 1956 Act,

and his share in the joint property and hence the shares of his heirs are ascertained under the assumption of a ‘notional’ partition (that is, as if the 32, See GOI (1983b). The Planning Commission Working Group, set up to examine the legal systems of the northeast, recommended even in 1983 that on family, inheritance, and land laws ‘there should be as little interference ... as possible’ (GOI 1983b: 23). Although this is an advantage for women among the matrilineal communities of Meghalaya and Assam, among patrilineal tribes of the region the traditional laws tend to be gender-inegalitarian.

33 Class I heirs are those who have the first right to the property of the deceased. Other ‘classes’ of heirs follow. 34 An agnate (as noted in chapter 3) is a person related to a deceased through male links alone: for instance, a son’s son or son’s daughter are agnates. A cognate, in the legal literature, is defined as a person related to the deceased ‘through one or more female links’ (Fyzee 1974: 403, in the context of Islamic law) or ‘not wholly through males’ (The Hindu Succession

Act, Mulla 1982: 913). Hence a daughter’s son or daughter's daughter are cognates. In other words, in the legal literature, agnates and cognates are treated as mutually exclusive. In the anthropological literature, however, the category cognates includes agnates (see e.g. Fox 1967: 49; and Fortes 1969: 267). In the context of our discussion in this chapter the legal definition is applicable.

Contemporary laws 213 partition had taken place just prior to his death). If the deceased does not leave behind Class I female heirs or claimants through such female heirs, the devolution is according to the Mitakshara rules. Either way this does not affect the direct interest in the coparcenary held by male members by virtue of birth; it affects only the interest they may hold in the share of the deceased.

In the case of a Hindu woman dying intestate, if she has children or grandchildren from predeceased children, all her property in the first instance devolves equally upon her sons, daughters, children of prede-

ceased children, and husband. If she has no living children and no grandchildren from predeceased children, the property devolution differs according to the source of acquisition: that inherited from her parents goes to her father’s heirs; that inherited from her husband or father-in-law (as a widow of a predeceased son) goes to the husband’s heirs; and that acquired in ways other than these passes to her husband, and failing him to his heirs.

Under the Act, all female heirs have absolute ownership and full testamentary rights over all property, not just a limited interest in it. The Act also gives unrestricted testamentary rights to males in their separate or self-acquired property, as well as in their share of the joint family property; but in relation to ancestral agricultural land there are restrictions on the testamentary rights of males in some states, by virtue of custom, as in Punjab.?>

For Hindus customarily governed by the Marumakkatayam and Aliyasantana systems, under the special provisions of the Act the devolution of

property is specified as follows. If a man dies intestate, his property (including all separate or self-acquired property as well as any share he may hold in joint family property)?° devolves on the Class I heirs and Class IT heirs in the same way as for other Hindus. In the absence of heirs in these

two categories, the property goes to both agnates and cognates without distinction or preference between them. If a woman dies intestate, all her property (again including separate or self-acquired property, as well as any share in joint family property) devolves in the first instance equally on her

sons and daughters (including children of any predeceased sons or daughters) and mother. If the intestate leaves no children or children of

predeceased children, her mother takes all the property, except that 35 In Punjab, the customary law has been upheld, under which a male cannot will away his share of ancestral agricultural land (see ‘Kaur Singh Gajjan Singh v. Jaggar Singh Kehar Singh’, AIR (1961), Punjab 489). Also see the decision in ‘Joginder Singh v. Kehar Singh’ (AIR 1965, Punjab 407). 3¢ As noted in chapter 4, a person’s share in unpartitioned property was deemed to be that which that person would get if the property were partitioned just before his/her death on a per capita basis among all family members holding an interest in the joint property.

214 A field of one’s own inherited from her husband or father-in-law, which goes to her husband’s heirs. If there are no children or grandchildren but the intestate’s husband is alive, the property goes to the intestate’s father and husband. In the absence of children, grandchildren, and husband, all the intestate’s property, other

; than that inherited from her husband and father-in-law (which devolves on the husband’s heirs), will go first to her mother’s heirs, and in their absence to her father’s heirs and lastly to the heirs of her husband. Again both men and women have full testamentary powers over all their property.

Since the passing of the Act, some states have enacted legislation affecting joint family property. For instance, the Kerala Joint Hindu Family System (Abolition) Act of 1976 deemed all family members with an

interest in the Hindu undivided family estate as holding their shares separately as full owners from then onwards. This Act (as noted in chapter 4) struck a final blow to the remnants of matrilineal joint estates, but it also eliminated any advantages that sons enjoyed over daughters in joint family

property among patrilineal Hindus in Kerala. More recently, Andhra Pradesh in 1986 and Tamil Nadu in 1989 have amended the Hindu Succession Act to recognize unmarried daughters (that is, daughters still unmarried when the Act was passed) as coparceners by birth in their own right, giving them claims equal to those of sons in joint family property, including the right to a share by survivorship.’ However, in most states the 1956 Hindu Succession Act, as originally enacted, continues to be in force. And for customarily patrilineal Hindus the Act has reduced but not eliminated pre-existing gender inequalities. Several major sources of inequality persist: (1) Since the concept of the Mitakshara joint family succession continues to be recognized (except, as noted, in a few states such as Kerala, Andhra Pradesh, and Tamil Nadu), some of the basic gender inequalities inherent in relation to unpartitioned coparcenary property remain, such as those below: —Since only males can be coparceners in the joint family property, sons have an indefeasible right in such property,

but daughters don’t. In addition, sons have a right to succeed to the deceased father’s share of the coparcenary if

the father dies intestate. Daughters have only the latter right, that is the right to succeed to the father’s share of the coparcenary. —A coparcener can renounce his rights in the coparcenary property. In such cases his sons would continue to maintain their independent rights to the coparcenary, but daughters 37 For Andhra Pradesh, see Sivaramayya (1988); and for Tamil Nadu see The Hindu Succession (Tamil Nadu Amendment) Act 1989 (Act No. | of 1989).

Contemporary laws 215 and other Class I female heirs would lose the possibility of benefiting from such property. Similarly, after partition, the

father can make a gift of his share in the coparcenary property to his sons, thereby defeating the rights of female heirs. —A man can convert his separate or self-acquired property to

coparcenary property, in which case daughters, who would have enjoyed equal shares with sons in such separate and self-acquired property, lose out.

—Unlike sons, married daughters (even if facing marital harassment) have no residence rights in the ancestral home.

And while daughters who are unmarried, separated, divorced, deserted, or widowed do have residence rights, they cannot demand partition if the males do not choose to partition. (2) The children of a predeceased daughter of a predeceased daughter do not figure among the Class I heirs, while the children of a predeceased son of a predeceased son do.

(3) The right to will away property is not restricted: a man has full testamentary power over all his property, including his interest in the

coparcenary. In principle the provision is gender-neutral, but in practice it can be and often is used to disinherit females. (4) Although the Act covers owned agricultural land, certain other types of interests in agricultural land, such as those stemming from ‘tenancy rights’, are exempted. Section 4 (2) of the Act provides that: ... nothing contained in this Act shall be deemed to affect the provisions of any law for the time being in force providing for the prevention of fragmentation of

agricultural holdings or for the fixation of ceilings or for the devolution of tenancy rights in respect of such holdings.

The gender implications of this exemption are crucial to the present discussion, as elaborated below.

(2) Anomalies resulting from existing land legislation

Two factors in particular have led to a disjunction between women’s general legal rights in property and their rights in agricultural land. First, legislative powers are divided between the Union and Provincial (now

termed state) legislatures under the principle of Federalism; and on agricultural land-related enactments the state governments have had, and continue to have, considerable legislative powers: thus legislation affecting women’s rights in certain categories of agricultural land varies by state, reflecting regional differences in social attitudes and legal approaches.

216 A field of one’s own

Second, land reform policies have been based both on the principle of redistributive justice and on arguments regarding efficiency (land to the tiller, fixation of ceilings, prevention of fragmentation, etc.); but on neither count are gender inequalities taken into account.

The Government of India Act 1935, mentioned earlier, vested all legislative powers in relation to agricultural land exclusively in the provincial (state) legislatures. Thenceforth women’s inheritance rights were to be

determined by personal law on all matters of property other than those relating to agricultural land, their rights in which would depend on the land-related laws prevailing in the province in which the land was located. Hence, for instance, The Hindu Women’s Right to Property Act 1937 (as

noted) did not apply to agricultural land. And although a few provinces (mentioned earlier) subsequently extended the Act to cover agricultural land, the extensions left unaffected any rule of succession that related to a tenant’s rights in agricultural land by any special law then in force. Since Independence, state legislatures have continued to enjoy the power to enact land laws, but subject to some restrictions. Under the Constitution of India adopted in November 1949, if the state legislature wants to modify any laws which have been included in the ‘concurrent list’ of laws under the

Constitution, and which have already been enacted by Parliament, the

modifications need the assent of the President of India. The Hindu Succession Act (HSA) of 1956 is one such piece of legislation. Hence if states want to pass laws modifying the Hindu succession rules for owned agricultural land, this will need the President’s consent. However, state legislatures can continue to enact laws relating to tenancy rights, ceiling laws, etc. (which, as noted, were excluded from the purview of the HSA), without needing such assent. What this has meant 1s that women’s legal rights in agricultural land still show a vast disparity by region, especially in relation to (a) devolution rules for land deemed to be under ‘tenancy’; and (b) rules regarding the fixation of ceilings and the forfeiture of surplus land above the ceiling limit, as discussed below. Devolution of agricultural land under ‘tenancy’. The Hindu Succession Act of 1956 specifically exempted tenancy rights in agricultural land from the scope of the Act. Asa result, there is today a major disjunction in

several states between state land enactments affecting the devolution of certain categories of agricultural land and the personal laws affecting the devolution of all other property, on at least two counts. First, in a number of states the succession rules relating to land held under tenancy have a different order of devolution than the personal laws specify. For example, in the tenancy laws of Punjab, Haryana, Himachal Pradesh, Jammu and Kashmir, Delhi and Uttar Pradesh (all located in northwest India), the

Contemporary laws 217 rules of devolution specified show a strong preference for succession among agnates, with a priority in favour of agnatic males.*° In all these states the tenancy devolves in the first instance on male lineal descendants in the male

line of descent. In the first four states the widow comes next, and in her absence the widowed mother. Daughters and sisters are totally excluded as heirs. In Delhi and Uttar Pradesh, daughters are recognized but come very low in the order of heirs. Moreover, in all these states, a woman (in any capacity) gets only a limited estate in the land, after which it goes not to her heirs but to the heirs of the last male landowner. She also loses the land if she remarries or abandons it (that is, fails to cultivate it for a specified period, usually a year or two). States where the tenancy laws explicitly mention that the devolution of tenancy land will be according to personal law are very few: they include

Rajasthan and Madhya Pradesh, where the personal law applies for all communities, and parts of Andhra Pradesh, where for Hindus the Hindu

Succession Act applies.7° In practice, however, even in Rajasthan daughters have been recognized as heirs only in some judgements, while in others male heirs alone have received recognition.*° In addition, there are states which do not mention the order of devolution at all in their tenancy laws, such as Maharashtra and Karnataka.*! In this last category of states, we could presume that the personal laws automatically apply.

Second, although most states in their land reform laws dealing with owned land do not mention the order of devolution at all,+* some states 38 See e.g. The Punjab Tenancy Act 1887 (Act No. 16 of 1887) applicable also to Haryana in identical form and under the same title; The Himachal Pradesh Tenancy and Land Reform Act 1972 (Act No. 8 of 1974); The Jammu and Kashmir Tenancy Act 1980 (Act No. 2 of 1980); the Delhi Land Reforms Act of 1954 (Act No. 8 of 1954); and the Uttar Pradesh Zamindari Abolition and Land Reforms Act 1950 (U.P. Act No. | of 1951). 39 See The Rajasthan Tenancy Act 1955 (Act No. 3 of 1955), the Andhra Pradesh (Telangana Area) Tenancy and Agricultural Lands Act 1950 (Act No. 21 of 1950), and the Madhya Pradesh Land Revenue Code, 1959, as amended in 1961. Prior to the amendment, the Madhya Pradesh Land Revenue Code specified an order of devolution wherein the Class I

heirs (for both owned and tenancy land) were as follows: son; widow (or widower); predeceased son’s son and widow; son and widow of predeceased son's predeceased son; and widow of predeceased son’s predeceased son’s predeceased son. Since the 1961 amendment, however, devolution is according to personal law for the lands of both tenure holders and occupancy tenants. 40 See the cases cited in the commentary to Section 40 in the Rajasthan Tenancy Act 1955 (Act No. 3 of 1955).

41 See The Bombay Tenancy and Agricultural Lands (Vidarbha Region) Act, 1958 (Bombay | Act No. XCIX of 1958); and The Karnataka Land Reforms Act 1961 (Act No. 10 of 1962). 42 This includes: The Andhra Pradesh Land Reforms (Ceiling on Agricultural Holdings) Act 1972: The Bihar Land Reforms (Fixation of Ceiling Area and Acquisition of Surplus Land) Act 1961 (Bihar Act No. 12 of 1962); The Bihar Land Reforms Act 1950 (Bihar Act No. 30 of 1950); the Saurashtra Land Reforms Act 1951; The Punjab Land Reforms Act 1972 (Punjab Act No. 10 of 1973); The Karnataka Land Reforms Act 1961 (Act No. 10 of 1962); The Madhya Pradesh Ceiling on Agricultural Holdings Act 1960 (No. 20 of 1960); The

218 A field of one’s own define ‘tenants’ in very broad terms so that the category tends to include all interests arising out of agricultural land, as is the case in Uttar Pradesh and

Delhi. For instance, under the Uttar Pradesh Zamindari Abolition and Land Reforms Act 1950, the devolution rules mentioned above, which favour male lineal descendants, apply to various categories of cultivators: the bhumidhar, sirdar and asami.** Over time, according to the Statement of Objects and Reasons in the Act, ‘it is expected that the vast majority of cultivators will become bhumidhars’. Bhumidhars, however, are persons liable to pay land revenue to the government and not rent. Counting them as tenants and exempting them from the provisions of the Hindu Succession

Act means that most agricultural land in Uttar Pradesh, a state which contains one-sixth of the country’s population, is legally inheritable principally by males. In the Delhi Land Reforms Act, 1954 (Act No.8 of 1954), likewise, the specified order of devolution (which, as noted, favours male agnates) applies to both bhumidhars and asamis. Fixation of ceilings and assessment of surplus land. Gender inequa-

lities also arise from land reform enactments relating to the fixation of ceilings. (This part of the discussion is equally relevant for non-Hindu communities in India.) The enactments are characterized by certain general features as follows: A ceiling is fixed in relation toa family unit consisting of up to five members. Additional land is, however, allowed to families of over five members, subject to a specified maximum. In addition, in most states adult sons get special consideration (as elaborated shortly). Also, usually where the husband counts as a unit the wife cannot count as an independent Maharashtra Agricultural Lands (Ceiling on Holdings) Act 1961 (Act No. 27 of 1961); The

Orissa Land Reforms Act 1960 (Orissa Act No. 16 of 1960); The Tamil Nadu Land Reforms (Fixation of Ceiling on Land) Act 1961 (Act No. 58 of 1961); The Uttar Pradesh (Imposition of Ceiling on Land Holdings) Act 1960 (Uttar Pradesh Act No. | of 1961); The

West Bengal Land Reforms Act 1955 (West Bengal Act No. 10 of 1956). The Madhya Pradesh Land Revenue Code 1959, however, appears to be an exception: as noted, it explicitly mentions the order of devolution, which, after the 1961 amendment, is according to personal law for both owned land and land under tenancy. 43 Neither the Uttar Pradesh Zamindari Abolition and Land Reforms Act 1950 (termed below as UP (ZALR) Act) nor earlier acts such as the Uttar Pradesh Revenue Act 1901 and the Uttar Pradesh Tenancy Act 1939 give a clear definition of bhumidhar. What is indicated, however, is that the bhumidhar is a tenure holder or full proprietor with permanent heritable and (in most cases) transferable rights to the land, and whois liable to payment of land revenue to the government. This last point is clearly stated (see S.242 of the UP (ZALR) Act) and this is what is of most relevance to the present discussion. Sirdar is a landlord or under-proprietor or a permanent tenure holder who possesses sir and who has hereditable rights to the land. Sir is a name given to village lands cultivated by hereditary proprietors or village zamindars themselves. An asami is a tenant, and includes nonoccupancy tenants of land with no stable rights (UP (ZALR) Act). For further details and definitions see the original Acts; also see Srivastava (1976), Sivaramayya (1973), and GOI (1976).

Contemporary laws 219 unit even where she owns land in her own right. In these enactments, gender

inequalities and anomalies arise on at least three counts: one, in the definition of ‘family’; two, in the additional allotments for adult sons but not adult daughters; and three, in not allowing the wife to be counted as an independent unit where the husband is counted as one. The definition of ‘family’ varies widely across states. For instance, in Uttar Pradesh, Delhi, Punjab, Haryana, and Andhra Pradesh, the family is defined as the cultivator and his/her spouse, minor sons, and unmarried minor daughters.** In Bihar, Himachal Pradesh, and Madhya Pradesh it includes the cultivator, his/her spouse and minor children; in Tamil Nadu it includes the cultivator, his/her spouse, minor sons, unmarried daughters, and orphaned minor grandsons and unmarried granddaughters in the male

line. In Kerala it includes the cultivator, his/her spouse and unmarried minor children. Moreover, in virtually all the states, adult sons (as noted)

receive special consideration. For instance, in Uttar Pradesh, Delhi, Punjab, and Haryana the parental household can hold additional land on account of each adult son, subject to a specified maximum.*> In Bihar, Himachal Pradesh, Madhya Pradesh, Andhra Pradesh, and Tamil Nadu, each adult son counts as a separate unit and is entitled to hold a specified extent of land in his own right. Only in Kerala do both the unmarried adult

| son and unmarried adult daughter count as separate units. In these enactments, therefore, with the exception of Tamil Nadu and Kerala, unmarried adult daughters receive no recognition at all in the states mentioned: they do not count either as part of the family unit or as separate units; and in Uttar Pradesh, Delhi, Punjab, Haryana, and Andhra Pradesh,

married minor daughters also receive no recognition. Underlying the ceiling specifications is clearly the assumption that those who are recognized either as part of the family unit or separately (as with adult sons) will be maintained by the land allowed within the ceiling regulations. Under these enactments we thus have an extraordinary situation where most states do not give any consideration, when fixing ceilings, for the maintenance needs of unmarried adult daughters and married minor daughters, while giving consideration to all sons, whatever their age or marital status. And, as noted, in most states adult sons receive special recognition in that their parents, or they themselves, get additional land, while adult daughters receive no such recognition. Although in a state such as Uttar Pradesh, it is the tenure holder who is

allowed additional land on account of adult sons rather than the sons +4 Tam using the term ‘cultivator’ here in a general sense. The actual term used differs by state:

for instance, in Uttar Pradesh the term used is ‘tenure holder’. +> In Haryana, the allotment for the adult son is made to the parents if the son is living with the

parents, but he counts as a separate unit if living separately.

220 A field of one’s own themselves, given the land devolution rules prevailing there, any such land will ultimately pass to the sons. Even if we were to assume that married

daughters would be taken into account in their marital homes, the land ceiling rules along with the devolution rules in the state leave the unmarried

adult daughters out in the cold, as they do daughters whose marriages break down. The situation in Delhi, Punjab, and Haryana is analogous. In 1971, following the 1970 Conference of Chief Ministers on Land Reforms, a Central Land Reforms Committee was constituted under the chairmanship of the Union Agricultural Minister. It recommended, among

other things, that the definition of the family should be made uniform across the states and include a man and his wife and minor children, with additional land being allowed for extra members in excess of five, up to a

maximum of twice the ceiling limit for the family. A high-powered Committee set up to review these recommendations disagreed with them and suggested that the family should be defined to include a man and his wife and three children, including any major sons. However, the Chief Ministers’ Conference, held in July 1972, drawing its guidelines from the recommendations of both the above Committees, laid down (among other things) that a family should be defined as including a man, his wife, and

minor children, and additional land should be granted for members in excess of five, up to a maximum of twice the ceiling limit set for a family of five members. Further, each major son should be treated as a separate unit. In other words, the guidelines incorporated gender-inequitable rules;*° and

the situation, as it stands today, continues to be one where there is no uniformity across states on these counts and most states continue to have gender-discriminatory ceiling laws.*7 Over the years, some of these ceiling Acts have been challenged (unsuccessfully) in court, one of the grounds for challenging them being that they discriminated against women and were therefore unconstitutional.*® However, the First Amendment to the Constitution of India, enacted in 1951,

had introduced a device to protect the validity of all Land Reform legislation. Under Article 31b it provided that none of the Acts mentioned in the Ninth Schedule of the Constitution would be deemed to be void on the ground that they infringed on the fundamental rights granted by the Constitution of India.*? All the noted ceiling laws are included in the Ninth 46 For details of the recommendations made by the two Committees mentioned and the guidelines specified subsequently, see Government of India (1976). 47 It is noteworthy that Bangladesh too, in its ceiling laws for agricultural land, counts as a separate family an adult married son who had an independent means of livelihood prior to 20 February 1972 (Siddiqui 1981: 74). Adult daughters get no such consideration. 48 Article 14 of the Constitution of India promises equality before the law, and Article [5 prohibits discrimination on the basis of sex, etc. Both constitute part of fundamental rights (see GOI 1990c). 49 See GOI (1990c) for a list of the Acts included in the Ninth Schedule of the Constitution.

Contemporary laws 221 Schedule. This provision provides a basis for dismissing pleas challenging the ceiling laws on various grounds, including grounds of gender discrimi-

nation.°° The Andhra Pradesh Land Reforms (Ceiling on Agricultural Holdings) Act 1973, in fact, explicitly mentions in its Explanation to. Section 4 (3): ‘The constitutionality of discriminating against unmarried [major] daughters cannot be questioned as the Act is now included in the Ninth Schedule to the Constitution’. The judgement that was delivered when The Uttar Pradesh Zamindari Abolition and Land Reforms Act 1950 was challenged in the Supreme Court in 1980 is also noteworthy for its comments on the gender aspects of the Act.>! Among the grounds for the challenge were that it discriminated against major unmarried daughters by not providing extra land to their father as it did for adult sons; and that it discriminated against wives in the fixation of ceilings, by regarding the husband as the tenure holder even when the wife was the owner. Justice Krishna Iyer (although reputed to be an advocate of women’s rights) rejected the pleas in his judgement. He argued: ‘no submission to destroy this measure can be permitted using sex discrimination as a means to sabotage what is socially desirable’ (p.729).

While admitting that the advantage granted to major sons and not to daughters was sex-discriminatory, he nevertheless justified the rule on the ground that in effective terms the entire land goes to the father as the tenure

holder (and not to the son) ‘for feeding this extra mouth’ (p.729). Presumably adult daughters didn’t need to be fed! Moreover, the Act’s exclusion of women as tenure holders where their husbands are also tenure holders was explained away in the judgement as ‘a legislative device for simplifying procedural dealings’, and the judgement argued that: ‘When all 1s said and done, married women in our villages do need their husbands’ services and speak through them in public places...’ (SCC 1980: 730). Underlying these justifications was clearly a prioritizing of class interests at the cost of gender interests: ‘large land holders [cannot] be allowed to outwit socially imperative land distribution by putting female discrimination as a mask’ (SCC 1980: 730). Ironically, the land reform programme of the government was not successful even in significantly redistributing land between households. +? Another source of gender inequity has emerged in the assessment of ceiling surplus land. In most states, the holdings of both spouses (if the wife too has land in her name) are aggregated in assessing ‘family’ land, and °° For citations of some cases challenging the Ceiling Acts see: The Haryana Ceiling on Land

Holdings Act 1972; The Punjab Land Reforms Act 1972; The Andhra Pradesh Land Reforms (Ceiling on Agricultural Holdings) Act 1972; and GOI (1976).

o1 See (1980) Prasad Mishra v. the State of UP and Others’ in Supreme Court Cases (SCC) 52 The literature on this issue is vast, but see especially Frankel (1978), Joshi (1974), and Dasgupta (1977).

222 A field of one’s own there is considerable arbitrariness in deciding whose portion of this land will be declared surplus and forfeited. Typically the husband alone 1s consulted in this matter. This has been known to lead to the wife’s land being forfeited without her having a say in the matter, causing her to lose her primary or only source of economic independence. Given that so few women in any case have land in their own names, this practice compounds existing gender inequalities in land rights. In Kerala (as noted in chapter 4) when ceiling surplus land was being assessed, many women asked that they be allowed to retain their property where their husbands had independent

regular sources of income. But the demand went unheeded (Saradamoni 1983). Cases in which the wife has been able to establish her claim are rare.

For instance, in a 1986 court case in Orissa, the government Revenue Officer, in assessing ceiling surplus land, aggregated the land of both spouses as ‘family land’, including land separately registered in the wife’s name and inherited from her father. But he gave notice only to the husband as the ‘person interested’. The two men settled the matter between them, and the wife’s land was declared surplus. The wife appealed the order to the

High Court, asking that her separate land be excluded from the ceiling surplus, on the ground that since the land concerned was her separate property she was the ‘person interested’ to whom prior notice should have

been given. This, she argued, would have given her a chance to ask the Revenue Officer to let her retain her land and instead declare some part of her husband’s land as surplus. Her appeal was accepted by the Court under the constitutional principle of ‘natural justice’.>? Over the years there have also been explicit attempts by some states to amend the Hindu Succession Act of 1956 in such a way as to legally exclude daughters from inheriting agricultural land. For instance, in 1969, in a bill which came before the Punjab Legislative Assembly, it was argued that

daughters should not have a legal share in agricultural land (as allowed under the Hindu Succession Act) because it would cause fragmentation and because they got a dowry anyway. This was strongly opposed by women in

the state. Tara Ali Baig, then President of AIWC, pointed out that fragmentation took place even when sons inherited, and that dowry had been legally prohibited since 1961 (Mies 1980). The bill was not passed. Other (unsuccessful) attempts in this direction have been made subse-

quently. For instance, around 1979 the Haryana legislature sought to amend the Hindu Succession Act 1956 so as to deprive daughters of rights in agricultural property, but the President of India refused his assent to the proposed amendment.>** 53 See ‘Kunjalata Purohit v. Tahsildar, Sambalpur and Others’ (AIR 1986a), Orissa 115. 54 Personal communication from Prof Sivaramayya, 1992.

Contemporary laws 223 Moreover the special consideration given to major sons in fixing land

ceilings under the land reform enactments is being replicated in land resettlement schemes. For instance, in Gujarat state, farming families displaced by the Narmada Valley dam project have been promised (at the resettlement site) two hectares for each adult son, in addition to the two hectares for the family unit as a whole, but nothing for adult daughters. This is not only gender-discriminatory but also implicitly biased against families with no adult sons. As some of the displaced tribals at the dam site asked me: ‘What about those of us who have only adult daughters? Don’t we have to feed them?’ This project, the largest river valley project in South Asia, has become a focus of national and international controversy and agitation, especially around the issue of resettlement. The terms of resettle-

ment established here are likely to set a precedent for future projects involving displacement, making it even more imperative to correct the

gender bias in this scheme.55 |

In general, there is a strong case for re-examining the existing landrelated laws across the country to ensure gender equality in rights to this critical economic resource. For a start this would mean at least two types of changes: (a) eliminating the existing gender inequalities in the land ceiling

laws; and (b) bringing devolution rules relating to agricultural land (whether under tenancy or otherwise) in line with those governing other forms of property, that is in line with the personal laws. Although this would not take care of all the existing gender inequalities in personal laws, it

would at least reduce some of the anomalies which exist in relation to agricultural land.

(3) Laws governing Christians and Parsis in India The 1956 Hindu Succession Act covered Sikhs, Buddhists, and Jains under the definition of ‘Hindu’ but excluded Indian Christians, Parsis, Jews, and other minority communities. The laws pertaining to Christians and Parsis in India are briefly outlined below.

The Christians. The laws for Christians (who constituted 2.4 per cent of India’s population in 1981) vary according to domicile for all movable property and by location of property in the case of immovables. For instance, Christians from Goa are governed by the Portuguese Civil Code; those from Cochin and Travancore (Kerala) until recently by the Cochin Christian Succession Act 1921, and The Travancore Christian 55 The otherwise insightful ‘Morse Report’ (Morse and Berger 1992), in its review of the project, also failed to take this into account.

224 A field of one’s own Succession Act 1916, respectively; those in Punjab by the customary laws of Punjab (see Sivaramayya 1978); the Christian tribal populations of north-

east India also by their customary laws; and the rest by the Indian Succession Act (ISA) of 1925. Both the Cochin and the Travancore Acts contained significant gender inequalities. Under the Travancore Act, for example, a widow or mother inherits only a life interest in land and other immovable property, which too must be forfeited on remarriage; and a daughter gets ‘stridhan’ which is fixed at one-fourth the value of the son’s share or Rs. 5000/-, whichever is less. Under the ISA of 1925 (Section 33), however, if a man dies intestate leaving lineal descendants, his widow gets one-third of his estate, while sons and daughters get equal shares in the rest.

If there are no lineal descendants, but there are other kindred who are eligible to inherit, the widow gets half the estate; and in the absence of both lineal descendants and kindred, she gets the whole property.*° There are no restrictions on testation. In 1949 the former princely states of Travancore and Cochin merged to form what is termed as the Part-B state of Travancore-Cochin within the

Indian Union; and in 1951 certain Acts prevailing in the rest of India became applicable to Part-B states. For Travancore-Cochin this meant that the ISA Act should thenceforth have applied. However, a court judgement

in 1956 upheld the authority of the pre-existing Travancore-Cochin succession laws for Christians (viz. the Cochin Act 1921 and the Travancore Act 1916).57 It was only in 1983, more than two and half decades later, that the Travancore Act 1916 was challenged in the Supreme Court by Mary Roy, a Syrian Christian from Travancore and daughter of wealthy

parents, on the ground that it violated the Constitutional guarantee of equal rights for both sexes.>® As a result of this petition, in 1986 the ' Supreme Court (although it did not address the specific issue of gender inequality) held that after the inclusion of Travancore and Cochin in the Indian Union, the relevant law governing Christians in those regions was the ISA 1925. By this judgement, therefore, the Travancore Christian Succession Act of 1916 stood superseded by the ISA 1925, with retrospective effect from 1951.>°

One effect of this judgement has been that daughters and sons can now

share equally in their father’s property. The judgement was met with °° In 1926 Section 33—A was added to the Act giving a widow additional rights if the intestate

left no lineal descendants. However, this section does not apply to Indian Christians and several other categories of persons otherwise governed by the ISA. 57 See ‘Kurian Augusty v. Devassy Aley’ in AIR (1957), Travancore-Cochin 1; also see, Sivaramayya (1978).

°8 For details of this case see ‘Mrs Mary Roy v. the State of Kerala and others’ in AIR (1986b), SCC I0OL1.

59 Asa result of this judgement the Cochin Act of 1921 also now stands superseded by the ISA 1925.

Contemporary laws 225 immediate protest from the Kerala Christian community on a number of grounds, some long-familiar: for instance, they argued that it would cause

land fragmentation and would ‘open up a floodgate of litigation and destroy the traditional harmony and goodwill that exists in Christian families’.©° In addition, there was concern that the retrospective effect would invalidate bona fide land and money transactions, nullify the Land Ceiling Act under which land had been distributed to the landless, and

encourage thousands of Kerala nuns to claim their shares of ancestral property. Most of these fears appear to be unfounded; in any case, only intestate property will be affected. After Mary Roy a few other Syrian Christian women have filed cases.°! Meanwhile the protest against the Supreme Court ruling has been supported by the Synod of Christian Churches, which conducted a ‘pulpit campaign’ and arranged legal counsel to help draft wills to disinherit female heirs (Gandhi and Shah 1991: 247).°

The Parsis. The Parsis are today governed by the twice-amended provisions of the Indian Succession Act (ISA) of 1925. Before 1991 the ISA, as amended specifically for Parsis in 1939, was applicable, under which the property of male and female intestates devolved according to separate rules (Paruck 1977). In the deceased man’s property (other than agricultural land), the widow and each son got double the share of each daughter. If the man left behind parents in addition to a widow and children, his father teceived a share equal to half that of his son, and his mother half that of his daughter. In agricultural land, the devolution was in accordance with the Parsi rules of succession applicable prior to the 1939 enactment. Here the widow and daughter respectively got one-half and one-fourth of what the son got, and the parents of the deceased got no land. In a deceased woman’s

property, the husband, son, and daughter got equal shares. Hence daughters shared equally with sons in the mother’s property but unequally in the father’s property. There were no restrictions on testation by either the man or woman. With the passing of the Indian Succession (Amendment) Act 1991 (Act No. 51 of 1991) the difference in the succession of male and female property 6° P.J. Kurien, cited in Gandhi and Shah (1991: 247). 6! For instance, Aley Chacko, an eighty-four-year-old Syrian Christian woman was turned out of her ancestral home by her son, who seized the seventy cents of land she had been given as dowry by her parents. Like Mary Roy, Aley Chacko and her five daughters have petitioned the Supreme Court of India (Manushi 1987) 62 In some cases, wills have also been forged retrospectively. Arundhatti Roy, Mary Roy's daughter, recounted the following example to me. Aleykutty Thommen filed a case for her share in her family’s property after her brother refused to part with it. He forged a will which was rejected by the court. In 1991 she fenced off a part of her brother's rubber estate, built a shelter there, and claimed she was tapping the rubber, which (I was told) protects her from eviction under the occupancy laws of the state. She is now said to be making a significant profit from the land.

226 A field of one’s own among Parsis has been removed. Now devolution of intestate property of a Parsi male or female is as follows. If the intestate leaves behind a widow/ widower and children, these heirs get equal shares. Where there are no lineal

descendants and no widow/widower of lineal descendants, the widow/ widower of the deceased gets half the property. And if there are no lineal descendants, but there is a widow/widower of a lineal descendent, the latter gets one-third and the widow/widower of the deceased gets one-third, the remainder going to other relatives.

All Indians (except in the context mentioned below) can opt out of their personal succession laws if they have a civil marriage under the Special Marriage Act of 1954, or have been married under the Special Marriage Act of 1872, or if their marriage is registered under the Special Marriage Act of 1954, even if it was contracted in another way. For such persons, the Indian Succession Act of 1925 applies,°> under which (as noted) the widow of an intestate male with lineal descendants gets a fixed share of one-third in his

property and his children of both sexes inherit equally in the rest.°* The exception Is that after an enactment in 1976, two Hindus (or two Buddhists, Sikhs, or Jains) marrying under the Special Marriage Act of 1954 will be

governed by the Hindu Succession Act 1956, and not by the Indian Succession Act.°> In practice, very few persons opt for a civil marriage which would bring them under the purview of this Act (GOI 1974: 114). In overview, therefore, today non-Muslim women of most communities in

India (excluding some of the southern states) have fewer legal rights to property in general, and to agricultural land in particular, than do men. In this context, on the one hand there have been some significant cases, such as that of Mary Roy, which by drawing upon constitutional guarantees of sex equality to challenge persisting gender inequalities in inheritance laws have opened up yet another chapter in the contestation over women’s legal rights in land that began almost a century ago. This was important even though (as noted) the judgement on the case evaded the issue. On the other hand, the placing of land reform legislation beyond challenge on constitutional

grounds (through its inclusion in the Ninth Schedule) has taken away a major potential means of changing gender-unequal land-related laws still prevailing in many states. 63 A Hindu male marrying under the Special Marriage Act is automatically severed from the joint family estate. 64 As noted, Section 33—A of the Act, inserted in 1926, gives some additional advantages to widows, but it does not apply to several categories of persons, including Indian Christians as well as Hindus, Buddhists and Jains whose property would be governed by the ISA. 65 See Marriage Laws (Amendment) Act 1976 (Act No. 68 of 1976).

Contemporary laws 227 IT. Pakistan, Bangladesh, and Muslims in India Let us now consider the history of legal change that affected Muslims in Pakistan, Bangladesh, and India.°®

(1) From custom to the Shariat As among the Hindus, property inheritance among Muslims in undivided India (that is, including Pakistan and Bangladesh) showed a considerable gap between scriptural dictates and actual practice. These dictates gave Muslim women significant inheritance rights in landed property, although

unequal to men’s. But custom usually directed practice. Among many Muslim communities, which (as noted in chapter 3) followed customs similar to those prevalent among Hindus in their region of residence, this meant that (a) a daughter was either totally excluded from the inheritance of any landed property, or came very low in the order of heirs; a widow likewise came low in the order of heirs; and (b) both widow and daughter if they inherited took only a limited interest, rather than an absolute interest as would have been their due under Islamic law. Among the Mappilas of Kerala and the Lakshadweep Islands, however, customary practice meant

matrilineal inheritance. In between were Muslim communities among whom affluent families sometimes gave daughters a share of the patrimony under the Shariat, or occasionally (especially if sonless) created an endow-

ment (wagf) for them, both of which resulted in a number of Muslim women from such families controlling property, including land. The impact of British-Indian court rulings on this situation was mixed, partly because the British application of Shariat rules was not consistent

across regions and communities, and partly because (as noted above) custom itself disfavoured women in some regions and favoured them in others, so that a shift to the Shariat (and away from custom) had mixed regional implications. On the positive side, among patrilineal groups, in instances where the Islamic law was strictly applied it strengthened Muslim women’s access to property in two ways: (a) in recognizing their inheritance shares as dictated by the Shariat where custom disinheriting women would have ruled otherwise; and (b) in enforcing the woman’s right to mehr — the

sum of money or property normally promised to the bride by the bride66 T will not be discussing laws governing non-Muslim minorities in Bangladesh and Pakistan,

since they constitute very small sections of these countries’ populations. It.may be mentioned, however, that laws concerning such minorities have typically undergone little

reform since the colonial period, and include considerable gender inequalities. For instance, Hindu women in Bangladesh are governed by the traditional Dayabhaga rules of Hindu law under which widows and unmarried daughters still have only a limited interest in inherited property (Ahmed 1978).

228 A field of one’s own groom as part of the marriage contract, but not necessarily paid at the time of marriage.©’ Customarily, many Muslims treated mehr as a symbolic amount which they did not actually expect to have to pay in full; hence, especially among the wealthy, large sums were often pledged for prestige reasons (Ellickson 1972a; Kozlowski 1989). British court rulings, however, tended to support the claims of widows, granting them control over their husbands’ property until mehr was paid. Where large amounts had been pledged, this effectively gave such women lifetime control over much or all

of the deceased husband’s property. Once such rulings became known, many widows filed suits to claim their mehr. The decisions were frequently in their favour (Kozlowski 1989). On the negative side, however, was the overriding importance given by the British to local customs in their dealings with certain communities and regions which practised patrilineal inheritance. In the Punjab, for instance,

custom (rather than Islamic law) was made the ‘first rule of decision’ (Rattigan 1953: 42) in British court rulings in the late nineteenth century, and daughters were widely excluded from inheritance of land. As Gilmartin convincingly argues, court decisions relating to the Punjabis stemmed from two assumptions: (a) a theoretical one, enunciated by Tupper (1881) that the stability of ‘tribal’ Punjabi kinship rested on the exclusion of women from any share in their father’s land over which only male agnates were presumed to have claims (this came to be called the ‘agnatic theory’); and

(b) a political one, that it was critical to maintain the Punjabi kinship structure to ensure the stability of British rule in that region. The Punjab Alienation of Land Act of 1900 directly identified British political interests

with the protection of ‘tribal’ customs. Passed to stem the large-scale transfer of land from increasingly indebted peasants to moneylenders in the late nineteenth century, which, it was feared, would cause rural instability,

the Act identified some groups, such as the Jats, as ‘agricultural tribes’ whose landed interests had to be protected against others, especially moneylenders (Barrier 1966). The Act thus defined the dominant social categories with which the British had in large part identified their rule. Women’s interests were subordinate to these political considerations.°® The ‘agnatic’ theory expounded by Tupper was based on his compilation 67 The payment of mehr may be deferred in part or in full till the marriage dissolves or till the

husband's death, although in theory it is payable on demand at any time (unless a particular mode and timing of payment is specified in the marriage contract). Essentially it is meant to serve as an economic security to the wife in case of divorce or desertion by the husband, giving her something to fall back on. In practice not all marriage contracts specify an amount, or an economically adequate amount; nor do most women demand payment, or receive it if they do. (See discussions in Qureshi 1992, Patel 1979, and chapter 6.) 68 As noted earlier, similar political considerations probably underlay British support for forced leviratic unions for peasant (especially Jat) widows in the Punjab.

Contemporary laws 229 of extensive empirical information indicating the widespread disinheritance of daughters among Punjabi communities (Hindus and Muslims).°° But, as

Gilmartin (1981: 156) argues, ‘a description of popular practice’ was translated into ‘a normative formula supporting the ‘“‘tribal’’ system of kinship, underlying the British administration’; and this increasingly became the judicial presumption in court decisions, even in the absence of positive evidence of custom to this effect. In practice, not all daughters among Muslims in Punjab were customar-

ily disinherited, although it must be recognized that most undoubtedly were, since the conditions under which they could inherit were highly restricted, being typically confined to sonless families. For instance (as we

also noted in chapter 3), both Tupper’s (1881) and Rattigan’s (1953) compilations of customs indicate that among Muslim communities in the Punjab daughters did not inherit in the presence of sons, and, among most

groups, not even in the presence of a widow, but a number of groups (including some Muslim Jats) did give preference to daughters over male collaterals. Such inheritance was usually in the form of a life interest, and a daughter was more likely to receive it if married to a near collateral, since the property would then remain within the family.7° However, under the agnatic principle of succession formulated by the British, it was assumed that a daughter could never inherit; that she was excluded as a rule by the man’s male collaterals, and in their absence the estate would pass to the tribe or the village community (Rattigan 1953: 350; Gilmartin 1981: 155).

Although not all British officials or judges accepted the assumptions implicit in the ‘agantic theory’,”! most appear to have done so; as a result, ‘in relation to the rights of women ... [there] was the increasing tendency of the courts to seek uniform “tribal” standards of customary succession in their decisions’ (Gilmartin 1981: 157). °° As noted in chapter 3, in the nineteenth century, British settlement officers collected extensive information on customs in Punjab, which Tupper largely drew upon. 7° Gilmartin (1981: 164), drawing upon evidence of a high incidence of cousin marriages among Muslims in parts of Punjab (60 per cent in Attock district in 1921), suggests that

‘social foundations may already have existed in the Punjab for an interpretation of inheritance far different from the ‘‘agnatic theory” of customary law propounded in the late nineteenth century by the British courts’. In my view, this may be overstating the case: close-kin marriages can at best be seen as a facilitating factor in women’s inheritance rights,

in that communities/families practising kin marriages would be more open (or less resistant) to endowing daughters with property shares; but (as will be elaborated in chapter 8), close-kin marriages are also favoured by many communities in South Asia which are known not to endow women with shares in landed property. In the examples Rattigan (1953) cites, close-kin marriages did not enable daughters to inherit in the presence of sons, but only made such inheritance more likely in the absence of sons. 7) See e.g. Rattigan (1953: 350-1), for comments by some judges who disagreed with a strict presumption that daughters could not inherit in the presence of male collaterals. Also see Gilmartin (1981).

230 A field of one’s own

The British legal interpretation of Punjabi kinship and inheritance practices (an interpretation which also had its supporters among the local population) came to be challenged increasingly in the 1920s and 1930s by

Muslim reformers, who sought to establish the Shariat as the basis of Muslim personal law. Although political considerations clearly underlay such attempts at reform, in which women’s legal status was in large part only of symbolic importance, this had the potential for a positive fall-out for women, given the recognition accorded to female inheritance rights in the Shariat.’? There was a major move in this direction when a Muslim member of the Punjab legislature had a bill introduced in the Federal Legislative Assembly of India calling for the supersession of custom by the Shariat as the basis of personal law for all Muslims in undivided India (Parashar 1992).73 The bill, which was taken up for consideration in 1937, generated substantial controversy.’* In particular, there was strong opposition from the landowning classes of the Punjab on the ground that the bill would ruin agriculturists. However, unlike the Hindu Code Bill, whose clauses went directly against orthodox Hindu opinion, a bill to enforce the Shariat — the holy law of Islam — could not be opposed openly by the

orthodoxy. Indeed such a bill could be used as a symbolic means of politically affirming Muslim identity and solidarity. A number of scholars suggest that it was such considerations, rather than gender equality, which principally motivated the Muslim League under Mr M. A. Jinnah to get the Muslim Personal Law (Shariat) Application Act, 1937, passed.’75 Under the Act, prevailing customs or usage were abrogated in favour of Muslim personal law (the Shariat), but the Act explicitly excluded from its purview agricultural land, the devolution of which would continue to be governed by local customs. The Act was valid for all Muslims of undivided India, except those in Jammu and Kashmir. While ostensibly meant to extend women’s property rights, effectively the Act hardly served that purpose, since agricultural land, which constituted the bulk of property held by the Muslim community, was excluded. The Act therefore served the political purpose of affirming Muslim identity, without antagonizing the powerful Punjabi landlords.’° 72 Only a few Punjabi reformers were explicitly concerned with women’s rights: Gilmartin (1981: 168) mentions two by name, including a woman: Baji Rashida Latif. 73 Just prior to this, the North West Frontier Province (NWFP) had enacted the North West Frontier Province Muslim Personal Law (Shariat) Application Act, 1935, establishing the Shariat as the basis of inheritance rules (among other things) and superseding customs prevailing in the province. On this Act also see footnote 76. 74 For a flavour of the debate in the Assembly, see Jafar e¢ al., eds. (1977: chapter 6). 75 See e.g., Gilmartin (1981), Jalal (1991), and Lateef (1990). 76 As noted earlier, legislation relating to agricultural land had been made a provincial subject in 1935. In this context, it is interesting to note that the North West Frontier Province Muslim Personal Law (Shariat) Application Act of 1935 (which established that in this

Contemporary laws 231 The passage of the 1937 Shariat Act differed from that of the Hindu Code Bill (HCB) in at least two important respects: one, the shift here was toward

and not away from the scriptures, and therefore the Act encountered less widespread opposition than did the HCB. Two, although the reform was supported by urban-educated Muslim women and some Muslim women’s groups,’’ they do not appear to have played the kind of active mobilizing and lobbying role that was played by the AIWC and other predominantly Hindu women’s organizations for the codification of Hindu law.’* The situation changed somewhat after Pakistan was formed, with Pakistani women’s organizations playing a more active role in campaigning for women’s rights, including the reform of Muslim personal law (as discussed later in this section). Legal reform affecting Muslim women in the subcontinent took different regional directions after India’s Independence in 1947 and the country’s partition which led to the formation of Pakistan (constituted of a predominantly Muslim population located in two geographically separated parts, East and West Pakistan: East Pakistan later became Bangladesh). An early

step taken by the Pakistani State was the passing of the West Punjab Muslim Personal Law (Shariat) Application Act of 1948, which included agricultural land in its purview. For Muslim women in West Punjab, this removed one of the main gender disabilities embodied in the 1937 Shariat Act. (The basic gender inequalities embedded in Islamic inheritance laws (e.g. a daughter’s share is half that ofa son) could not, of course, be touched as long as the Shariat was the basis of law.) This Act (like the 1937 Act) encountered considerable resistance from the landlord-dominated West Punjab Provincial Assembly. Its delaying tactics caused several hundred province the Shariat would be followed in matters concerning inheritance, etc., two years prior to the 1937 Shariat Act), did include agricultural land within its purview (Mulla 1972: 14). There appears to have been little written about the imperatives behind the NWFP Act,

and whether its enactment involved any significant controversy. The issue of affirming Muslim identity, rather than any notable concern with women’s position, was probably the

main motive behind the Act, since in terms of women’s status the NWFP, by all ethnographic accounts, was and continues to be perhaps the most gender-inegalitarian region in the subcontinent. From these accounts (see chapters 6 and 7 for details), even in the late 1970s the province was characterized by severe purdah restrictions on women, extreme control over female sexuality with any transgressions leading to violence, and little adherence in practice to the law promising property rights to women. 77 For instance, the Anjuman-e-Khawatin-e-Islam (the All-India Muslim Ladies’ Conference) included women’s rights to property as one of its demands and its Punjab wing supported the Shariat Bill in 1937 (Jalal 1991; Minault 1981). The issue of women’s (including Muslim women’s) property rights had also come up indirectly in the 1920s during women’s struggle for enfranchisement. As both Hindu and Muslim women pointed out, linking the franchise

with property and income qualifications made for a significant gender gap in electoral representation, since few women owned property in their own names (Everett 1979). 78 The concerns of Muslim women’s groups during the early part of this century centred primarily on women’s education (Minault 1981; Mumtaz and Shaheed 1987; Lateef 1990).

232 A field of one’s own urban-educated women to demonstrate outside the assembly chambers. When the law was finally passed, loopholes remained, allowing families considerable personal discretion in its application (Jalal 1991). Two years later, in 1950, the province of Sind also brought agricultural land within the purview of the Shariat by amending the 1937 Shariat Act for that province. But it took another twelve years before the West Pakistan Muslim Personal Law (Shariat) Application Act of 1962 extended the Shariat as the basis of personal law to the whole of West Pakistan, except the ‘Tribal Areas’ in the NWFP.’® This Act abrogated custom as the basis of law and legally entitled Muslim women to inherit agricultural property as full owners (and not just as a life interest), as prescribed by the Shariat. Women leaders and women’s organizations, especially the All Pakistan Women’s Association, played a

significant role in pushing for these changes by agitating for reform in family law and against the gender inequalities of customary laws.®° These changes did not affect the Muslims of East Pakistan (Bangladesh since 1971), who continued to be governed by the 1937 Shariat Act. But

here women have not been legally disadvantaged by the absence of subsequent amendments, since even prior to this enactment customs at variance with Islamic law were not required to be enforced in (undivided) Bengal: here the Shariat therefore applied also to agricultural land.®! In India too, the 1937 Shariat Act has continued to be applicable even after Independence. Although this is not a disability for Muslim women in

some states, it is in others. For instance, in Tamil Nadu and Andhra Pradesh, the Act was amended in 1949 to include agricultural land; in 1963, Kerala did the same.®? In Assam, Maharashtra, and Gujarat, as in Bengal (noted above), there was no strong presumption in favour of custom even before the Act was passed; hence here the Shariat rules could be applied to agricultural land as well.§> But in several states of northwest India, such as

Uttar Pradesh, Punjab, Himachal Pradesh and Haryana, the Act has not 79 The NWFP of Pakistan is administratively divided into ‘Tribal Areas’ and ‘Settled Areas’ (see Ahmed 1980: 9-10). In the Tribal Areas: (a) no criminal or civil procedure codes of Pakistan apply: (b) they are not subject to taxes or rents of any kind; (c) they are federally administered but are attached to the Provincial Government of the NWFP for administrative purposes because of geographic and historical reasons; and (d) no political parties are allowed, and the voting right is reserved for selected groups. 80 See e.g., Patel (1979), Mumtaz and Shaheed (1987), and Rashid (1987). 8! In Bengal, Agra and Assam, the 1887 Bengal, Agra and Assam Civil Courts Act XII of 1887, along with the Bengal and Assam Laws Act, 1905, established that Islamic law would

prevail on all questions relating to succession, inheritance, marriage, and religious institutions, except in so far as such law had been altered or abolished through legislative

enactment (Mulla 1972: 9). Also see Tyabji (1968: 18) and Patel (1979: 5). Personal communications from several Bangladeshi legal experts on laws affecting women, including Salma Sobhan and Rabia Buyian, confirm this reading. 82 See The Muslim Personal Law (Shariat) Application (Madras Amendment) Act 1949, and the Muslim Personal Law (Shariat) Application (Kerala Amendment) Act 1963. 85 See Tyabji (1968: 18) for Assam and Bengal, and Mulla (1972: 11) for Maharashtra and Gyjarat (earlier in Bombay state).

Contemporary laws 233 been amended to include agricultural land, and customs prevailing prior to the Act (along with land reform laws) still govern the succession of such land, as they also do in Jammu and Kashmir (which was not covered by the Act). Under these customs male lineal descendants in the male line are typically the first order heirs in the inheritance of agricultural land. In these states, therefore, severe gender inequalities continue to characterize succession among Muslims in relation to the most important form of property in the countryside. While the primacy given to custom in northwest India has worked to the detriment of Muslim women there, so that a shift to the Shariat in relation to the succession of agricultural land would be a positive step, such a shift

has had the contrary effect for Indian Muslim women belonging to matrilineal communities, namely the Mappilas of north Kerala and the Lakshadweep Islands. The shift has been gradual. British enactments in the

early part of the twentieth century only brought some categories of Mappila property under Islamic law, leaving other categories of property untouched. For instance, in relation to undivided joint family estates, the Mappilas continued to be governed by the Marumakkatayam laws (similar

to those governing the Nayars and other matrilineal communities of southwest India), under which women had primary inheritance rights in ancestral property, including land. But for property that was individual in nature, they were subject to the Shariat. Subsequent enactments in the 1960s, however, extended the jurisdiction of the Shariat to all property held

by the Mappilas. The Mappilla Marumakkathayam (Amendment) Act, Kerala Act No. 32 of 1963, substituted the word ‘Muslim’ for Mappila (Derrett 1968: 528; also see 525-30). Under this Act, the share of any member of the Mappila taravad would devolve according to the Shariat rather than the matrilineal law. These amendments subordinated Mappila women’s rights, even in ancestral property, to those of men.

(2) Devolution under Islamic law Consider now the rules of succession under Islamic law (which were only touched upon in chapter 3). Today a vast majority of Muslims in India (where they constituted 11.4 per cent of the population in 1981) as well as those in Pakistan and Bangladesh belong to the Sunni sect, governed by the Hanafi School of Sunni law, while a small percentage are Shiites, governed by the Ithna Ashari School of Shia law. The inheritance rules under both systems are quite complex, and since they have been excellently spelt out in legal texts I will not detail them here.** But the broad aspects of the Hanafi 84 See, e.g. Fyzee (1974), Mulla (1972), and Tyabji (1968). For a clear exposition, accessible to a general reader, also see Carroll (1983, 1985).

234 A field of one’s own

school, some of its differences from Shia law, and the variation in its application across the subcontinent, are outlined below. Broadly, heirs are divided into three major categories: agnatic heirs who are almost all male, Koranic heirs who are mostly female, and ‘distant kindred’ who include all blood relations who are neither agnatic heirs nor Koranic heirs. The ‘distant kindred’ tend to be either women or connected

to the deceased through a female link: e.g. daughter’s children, son’s daughter’s children, daughters of male agnatic collaterals, children of female agnatic collaterals, paternal and maternal aunts and their children, maternal uncles and their children, and so on. These relatives are ‘distant’ not necessarily in terms of their blood relationship with the deceased, but in terms of the likelihood of their ever coming into a share of the inheritance. The three categories of heirs - agantic heirs, Koranic heirs, and distant kindred — together comprise the blood relations of the deceased and one relation by marriage, namely the husband or the wife. In terms of shares allotted, the implicit rule is : “Keep the bu/k of the property for the [male] Agnatic Heirs... the persons whose rights were always recognized by tribal [pre-Islamic] law, and respect the Koranic provisions by giving specific shares to the persons mentioned in the Koran’ (Fyzee 1974: 399), In specific terms, the shares of particular heirs under the Hanafi school are as follows: a daughter who is an only child receives a half share of the deceased parent’s estate as a Koranic portion and is excluded by no other heir. If there are two or more daughters and no sons, they jointly get a twothirds share which is divided equally among them. The presence of a son who 1s an agnatic heir, however, converts a daughter’s right from that ofa Koranic heir to an agnatic co-sharer, which means she gets half of what the son gets. Sons and daughters are excluded by no other heirs. Similarly a husband and wife, as Koranic heirs, are excluded by none: the husband receives a one-fourth share of his deceased wife’s property if there is a child or a son’s descendants, and a half share if there are no such heirs. A widow likewise receives either one-eighth or one-fourth of the husband’s estate, depending on whether or not there is a child or son’s descendants. If there is more than one widow, their collective share is one-eighth (or one-fourth), shared equally among them. Full sisters and consanguine sisters also share as Koranic heirs but can get excluded by male agnatic descendants and

ascendants, as can uterine sisters under specific circumstances.°> The mother gets a basic Koranic share of one-sixth, as does the father.

The Shia law of succession is noted to differ from the Sunni law, especially in the following respect:5® no relative of the deceased male is 85 Full sisters are daughters from the same set of parents. Consanguine relationships result when a man has children from two or more wives; uterine relationships result when a woman has children from two or more husbands (see Carroll 1983: 632). 86 See Tyabji (1968) and Carroll (1985).

Contemporary laws 235 excluded merely on grounds of his/her sex or because s/he is related to the

deceased through a female link. Cognates and agnates are placed on an equal footing. Hence males and females who are linked to the deceased in equal blood or degree inherit together, although female shares continue to be half those of males. For instance, if the deceased leaves a son’s son’s son

and a daughter’s daughter, under the Hanafi school the former as a male agnatic heir excludes the latter. Under the Shia system the daughter’s daughter has precedence, being deemed a higher ‘class’ of descendent (that

is, closer by blood to the deceased). Tyabji (1968: 897) clarifies the underlying differences between the two systems as follows: [T]he Hanafis take the Quranic alterations of the Pre-Islamic customs literally, and the Shiites take them as illustrations of underlying principles. The former let the substratum of the customary law stand unaltered except to the extent to which it is definitely altered by express provisions of the Quran. The Shiites take each instance mentioned in the Quran as speaking not only for itself but as indicating the possible principles.

The Shia system thus has more positive implications for women’s inheritance. In general, though, under all schools of Islamic law Muslim women have

inheritance rights in immovable property, although unequal to those of men. These rights also have some degree of protection from testation. Among the Hanafi Sunnis, for instance, an estate (in full or in part) cannot be willed to an heir without the consent of all the heirs, but a maximum ofa

third of the estate can be willed to a stranger without the consent of the heirs. Under Shia law, bequest to heirs or non-heirs of up to a third of the property is permitted without the consent of other heirs. There are, however, some differences between India, Pakistan, and Bangladesh in the application of Hanafi law, the main ones being the following (see especially, Carroll 1983). First, Indian Muslims, without renouncing Islam or converting to another religion, can opt out of the intestate succession rules mandatory under their personal law, by either marrying under or registering the marriage under the Special Marriage Act of 1954, in which case (as noted) the Indian Succession Act of 1925 applies

to the couple and to their children born thereafter. This option is not available to Pakistani and Bangladeshi Muslims. Second, under classical Hanafi law no non-Muslim could be an heir toa deceased Muslim. Ifa Muslim converted to another religion, s/he could not inherit from Muslim relatives. If a non-Muslim converted to Islam, his/her non-Muslim relatives who did not convert were denied a share in his/her estate. However, in India and Bangladesh, with the application of the Caste Disabilities Removal Act of 1850, an apostate from Islam, or a convert to Islam, retains the rights of succession s/he enjoyed under the law applicable

236 A field of one’s own to him/her prior to the apostasy or conversion. The Act, however, does not

protect the rights of such persons’ relatives (including immediate kin). Hence a convert from Islam retains her/his rights of succession to her/his Muslim relatives, but whether or not these relatives can succeed to the convert’s property (and if they do, to what share of it) would depend on the personal law applicable to her/him at the time of her/his death. Similarly, a convert to Islam retains her/his pre-existing rights of succession to his nonMuslim relatives; but the non-Muslim relatives of such a convert would not

have any rights of succession to the convert’s property, which would descend on her/his death according to Muslim law. In Pakistan, however, the 1850 statute relating to Caste Disabilities Removal was amended in 1963 so as to make it inapplicable to the property of a Muslim. Apostates from any other religion are not affected by a change of religion, but an apostate from Islam is disinherited in Pakistan.®’ Third, under traditional Sunni Law a predeceased son’s children are excluded by a surviving son. And a predeceased daughter’s children are excluded by any blood relative who is a male agnate or a Koranic heir. This is still applicable in India. In Pakistan and Bangladesh, however, with the passing of the Muslim Family Laws Ordinance of 1961, the children of a predeceased child (of either sex) are guaranteed the share of their grandfather’s estate that their parent would have received if alive at the time of the

grandfather’s death. This particularly helps a predeceased daughter’s children, who under traditional Sunni law seldom get a share in the maternal grandparents’ estate, while the children of a predeceased son are often heirs, since in the absence of a surviving son, the son’s son ts the highest male agnate, and 1n the absence of a surviving son or two (or more)

surviving daughters, the son’s daughter is a Koranic heir. The 1961 Ordinance allows a predeceased daughter’s children (in Pakistan and Bangladesh) to be admitted as heirs on the same terms as the predeceased son’s children; they are not disqualified merely because their link with the deceased is through a female. If the predeceased child leaves more than one child, those of the same sex share equally, and a female gets half the share of a male.

Fourth, in India, as noted earlier, both Sunni and Shia laws are inapplicable to the inheritance of certain categories of agricultural land in many of the states of northwest India, in which customary laws continue to prevail in this respect. Also the gender biases inherent in the ceiling laws of most states, which were discussed in detail in the subsection on Hindus, apply equally to Muslims in India. In Pakistan and Bangladesh, as noted 87 On the issue of inheritance rules governing apostates in India, Pakistan, and Bangladesh, see Carroll (1983).

Contemporary laws 237 earlier, agricultural land is today legally treated like any other property in matters of succession.

Let us now turn to Sri Lanka where, unlike in India, Pakistan, and Bangladesh, customary practices endowed women of all the major communities with significant rights in landed property. Here the shifts from custom to contemporary law were less dramatic and the process less contentious, since women’s inheritance rights in landed property were never at issue, although their extent, and the degree of control women could exercise over

the property, were. A significant part of this process, as it related to the Sinhalese, was described in chapter 4. Below I will focus on contemporary laws.

Il. Sri Lanka Today, the legal systems operating in Sri Lanka, like those in India, reflect cross-community differences.°® There are four separate systems in relation

to property and inheritance rights: (a) the Kandyan law, applicable as a personal law to the Kandyan Sinhalese (identified as the descendants of those domiciled in Kandy at the time the Kandyan Provinces were annexed

by the British in 1815); (b) the Thesawalamai or Tamil customary law, applicable to Tamils with a permanent home in the Jaffna province; (c) the Muslim law which, unlike the other customary systems, is a religious law and applicable to all adherents of Islam by birth or conversion, and (d) the General Law, which is an amalgam of Roman-Dutch and English law and applies to all those who do not fall within the purview of any of the above, including non-Kandyan Sinhalese, who constitute the bulk of the Sinhalese

population. :

The recorded versions of Sri Lanka’s legal systems are of relatively recent origin. The legal customs of the Jaffna Tamils were compiled under Dutch

colonial rule in the early eighteenth century, and those of the Kandyan Sinhalese by the British in the early nineteenth century. The absence in Sri Lanka of ancient legal treatises such as the Hindu shastras did mean that the mistake made by the British during the early part of their rule in India, of assuming the universal applicability of scriptural texts, was not repeated in Sri Lanka. Custom received a greater due, not just in stated intent but in fact. However, biases in the recording and application of custom surfaced here as well. Both the Dutch and the British sought to establish unambiguous and clear rules. The Dutch did so ‘as heirs to a Roman law tradition of clearly spelled out and codified legal rules’, and the British, despite their familiarity with an uncodified common law, did so ‘in an effort to ensure 88 For a discussion on these legal systems as they affect women, also see Goonesekere (1980).

238 A field of one’s own objectivity and impartiality in the administration of justice’ (Goonesekere n.d.: 3). As in India, this approach eroded the flexibility that local custom

allowed. Roman-Dutch law became the road to introduce this clarity wherever convenient. It influenced what was recorded and how ‘gaps’ in customary law were filled. Roman-Dutch law was used where ‘an institu-

tion [was] unknown to native custom, or custom was silent, or the customary rule was hostile to the prejudices of the rulers’ (Derrett 1968: 284).

The British inherited the Dutch compilations of Tamil and Muslim customary law and gave them official recognition through legislation. However, their attempts to compile Sinhalese customs were partial: those of the Kanydan Provinces were recorded but not those of the Maritime Provinces. Moreover, there was a gradual confining of recorded customary

laws to selected categories of persons: the Sinhalese in the Kandyan Provinces and the Tamils in Jaffna thus came to be governed differently from the Sinhalese and Tamils elsewhere. Islamic law was an exception in

that it ‘both consolidated and expanded its significance as a distinct personal law’ (Goonesekere n.d.: 9). These legal recordings impinged on diverse social and economic contexts and had varied effects on women’s

property rights among the Sinhalese, the Tamils, and the matrilineal Muslims, as discussed below.

The Sinhalese. In chapter 4 (and above) we noted how legal interventions under the British had led to a legal bifurcation of the Sinhalese community into the Kandyan Sinhalese and Sinhalese elsewhere. The Kandyans came to be governed by one set of laws (which drew a good

deal from custom) and the non-Kandyans by another set, viz. first by Roman-Dutch law and subsequently by the General Law. The historical shifts in these laws and the factors underlying them were detailed in the previous chapter. Here it suffices to reiterate that British interventions were guided by a variety of considerations which often could not be reconciled: their desire to give weight to local custom and at the same time to come up with an unambiguous set of rules, their own social history and prejudices

regarding appropriate marriage practices, and their need to establish control over land for economic gain. Today property devolution among the Kandyan Sinhalese continues to be in accordance with customary law (described earlier), which makes a distinction between the binna and diga forms of marriage, disadvantaging diga-married daughters in patrimonial inheritance. Non-Kandyan Sinhalese are governed by the General Law, under which women (irrespective of the form of marriage they contract) have the same rights as their brothers in their father’s estate. Also children of both sexes have equal inheritance

Contemporary laws 239 rights in the mother’s property. When either spouse dies intestate, the surviving spouse inherits a half share of the property. A widow may inherit the whole if the husband leaves no descendants, ascendents or collaterals capable of inheriting his property. She can also sell her deceased husband’s property (movable or immovable) to pay his debts. Moreover, a married woman (unlike among the Jaffna Tamils discussed below) has complete freedom to acquire, possess, and dispose of her separate property, including land and other immovables and any assets received in dowry (Goonesekere 1980).

The Jaffna Tamils. The original Thesawalami customs governing the Jaffna Tamils described in chapter 3 underwent a number of changes in the colonial and post-colonial periods. The changes were the result of a complex set of interactive factors: legislation, judicial decisions, replace-

ment of obsolete customs by new ones accepted in courts as law, the application of Roman-Dutch law to certain issues, and so on. Banks (1957) lists five main changes that resulted from colonial interventions (also see

Tambiah n.d.). First, sometime under Portuguese rule, sex-divided property distinctions were abolished and parents began to give dowry out of all three categories of property: chidenam (the wife’s ancestral inheritance received as dowry), mudisam (the husband’s ancestral inheritance), and thediathetam (property acquired by either spouse after marriage). The residual after the marriage of all daughters continued to pass, as before, to the sons on the death of both parents. The abolition of sex-divided property gave rise to the possibility of significant shifts in the proportion of property held as chidenam and mudisam. Earlier, by Banks’ assessment, landed property was more or less equally divided between chidenam and mudisam

lands; and, he argues, the fact that this change occurred without any apparent difficulty suggests that there was no corporate interest in sexdivided property among the Jaffna Tamils. Second, the British, by the Jaffna Matrimonial Rights and Inheritance Ordinance of 1911, introduced a new order of reversion in case of a person dying without issue. Now parents could succeed before siblings, grandparents before parents, and great-grandparents before grandparents. Further, brothers and sisters were to share equally when the shares were decided. Since sons inherited post-mortem, if all the daughters were already dowered the property of a deceased woman would now revert to her parents and then

go to her brothers (and not to her sisters’ descendants as earlier), thus enabling chidenam property to be diverted to the male line. Third, earlier, sets of half siblings had inherited per stirpes; hence each set of children, say by two marriages, got a half each of the property, which was

then divided equally within each set. Now each child could inherit per

240 A field of one’s own capita. Fourth, sons could now retain their pre-marriage earnings. Fifth, people could leave wills, whereas under the original code they did not have the right of testamentary disposal of property. These inheritance rules continue today, as does the basic principle that

each spouse’s ancestral property returns to its source. Neither spouse succeeds intestate to the other’s ancestral property. The widow holds a life

interest in the husband’s mudisam, inheritance rights vesting with the husband’s heirs, although she succeeds to half of the thediathetam property which has not been disposed of by a will or otherwise. However, a woman’s rights to exercise control over her property are only sightly less restricted than under the old Code. Under contemporary law, a

woman subject to Thesawalami law has complete power to deal with movable property (which belongs to her separate estate) without her husband’s consent, but the disposition of immovable property inter vivos, including land obtained in dowry, still requires his consent as long as she is

in the marital union. Also thediathetam property 1s now the separate property of the acquiring spouse; but here again the wife needs the husband’s consent to dispose of her portion of the immovables. It is only ifa

woman 1s maritally separated that she has full control over both her movable and immovable property. The man, however, even while in a marital union, has full powers of disposal over the whole of his ancestral and acquired property without his wife’s consent. This remains a major source of gender inequality in current laws.

The matrilineal (Muslim) Moors. Muslim Moors in the Eastern Province customarily followed a matrilineal system of inheritance, as described in chapter 3. Increasingly, though, they have been subject to

orthodox Islamic influence. The history of legislative change for the Muslims, as for the Sinhalese and Tamils, goes back to the early colonial period. The initial codification of laws governing the Muslims was undertaken by the Dutch who, under direction from Batavia, compiled a code framed by the ‘best informed and most learned of the Mohammaden priests who resided within the government’ (Mohan 1985: 54). This Mohammedan

code was circulated for acceptance among the headmen of all Muslim settlements and became the recognized law in Dutch courts of justice. According to Goonesekere (n.d. and 1980) the initial emphasis was on the

customary basis of Muslim law; it was thus possible for Muslims to conform to legal norms which were not always endorsed by Islam. However, dissatisfaction among the Muslim elite with the Mohammedan

Code led to legislative reform, especially from the second half of the nineteenth century, enshrining Islamic law as the primary source of law for the Muslims. Today it 1s the Muslim identity of the Moors rather than their cultural

Contemporary laws 241 identity which has legal primacy, and Islamic law applies uniformly to all Muslims in the country. This is the only system of personal law in Sri Lanka whose application is linked with adherence to a particular religion. Most Sri Lankan Muslims belong to the Sunni sect and are subject to the Hanafi school of intestate inheritance (discussed earlier). Under Islamic law, unlike other personal laws in Sri Lanka, women have an in-built disadvantage, in that their shares are always less than those of men related to the deceased in equal degree. However, Muslim women can dispose of their property as they wish, without seeking their husbands’ permission. Since the introduction of Islamic law, the customary rules of matrilineal inheritance are no longer legally applicable. Savitri Goonesekere, a legal advocate and professor of law in Colombo, in a personal communication to me in 1985, noted: Ever since Islamic law was introduced in regard to aspects like marriage and inheritance, our courts have been disinclined to recognize a concept of ‘Muslim customary law’. Consequently, the trend in our courts in recent years is to reject customary concepts that are in conflict with Islamic law. It seems unlikely that Sri Lankan courts will recognize matrilineal customs that seek to modify principles of Islamic law.

In their everyday practice, however, matrilineal customs continue to be followed by a number of Moorish communities (as will be described in chapter 6). In summary then, in Sri Lanka today, the General Law provides both sexes with equal rights in land and other property. However, the other personal laws embody inequalities. The Kandyan law, for instance, disadvantages the diga-married daughter, who forfeits her right to the father’s praveni (except under special circumstances when this right can be re-established). Since dowries are not mandatory, a diga-married daughter is not necessarily compensated for this loss. Diga marriage being the commonest form of marriage today, many married daughters among the Kandyan Sinhalese thus risk losing their shares. Among the Jaffna Tamils, under the Thesawa-

lamai, the biggest source of gender inequality lies in the jural control exercised by the husband over the wife’s immovable property. Under Islamic law, the inheritance shares themselves are gender-unequal. However, under all systems, women have a legal right to land, and bilaterality is

today the dominant underlying principle among all Sri Lankan communities.°° 89 The existence of several personal laws in Sri Lanka (as indeed elsewhere in the subcontinent) does, of course, create complications where the spouses are governed by different laws prior to marriage. Here, according to varying interpretations, either the husband’s

242 A field of one’s own We now come to Nepal, a part of the subcontinent that was never directly under colonial rule, although British rule in neighbouring India may have had an influence on the policies of the Nepalese monarchs. Hinduism was

promoted by different rulers and is today Nepal’s national religion. However, the population is ethnically heterogeneous and a significant belt of Tibeto-Burman communities, located especially in the hills, did not (and

still do not) follow the Hindu caste system. Customary practices of marriage and inheritance were therefore diverse, although, as described below, the local rulers, like the British in India, sought to bring some degree of uniformity in legal procedures in the nineteenth century.

IV. Nepal Nepal’s legal code (Maluki Ain) dates back to 1854. At that time, there were

several customary laws in existence, being followed by the many ethnic groups who were geographically scattered across the country and often quite isolated from one another. The Maluki Ain sought to establish a degree of uniformity in civil, criminal, and procedural law across the country, drawing heavily upon the shastras but with some modifications to take account of prevailing customary laws. As initially drawn up, the Code, described by Hoefer (1979), was structured around a hierarchical system of caste groupings. Every named caste and ethnic group, including groups of Tibeto-Burman origin, was assigned to one of five broad caste categories. At the same time, in the practice of civil law: various groups of the population were openly or tacitly granted a certain degree of

autonomy. Local traditions regulating marriage, inheritance, etc. ... were often tolerated as a kind of customary law, and jurisdiction was the concern of ad-hoc councils composed of village notables. (Hoefer 1979: 40)

Legal changes since 1951 have led to the formulation of a Code which does not recognize caste as the basis of legal difference. The individual has replaced caste in the new Maluki Ain which 1s applicable today. The Maluki Ain of 1854, as well as the version that exists today, were clearly superimposed on an array of local customs. This is indicated by the considerable variation in legal practice across ethnic groups revealed in recent ethnographies, even though, as Gilbert (1991) argues, what ethnographers observe today is itself a modification of earlier practices, which would have been affected over time by the Maluki Ain coming into force: personal law or the General Law of the country could apply to the wife. For instance, where

a man married to a non-Muslim under the General Law converts to Islam and takes a second wife, it is not clear whether, on his death, the first wife’s claim to his property will be decided under the General Law or under Islamic law (Goonesekere 1980: 18).

Contemporary laws 243 Studying intra-family dispute over property, persons and inheritance [among the Brahmin-Chetri immigrants and the Limbu in Eastern Nepal], I discovered that one of the side effects of the uniform legislation of the modern family code is that such disputes, patterns of property management, and gender relations within the family (which were previously ethnically differentiated) are beginning to conform to a generic ‘Nepali’ pattern. (Gilbert 1991: 2)

Be that as it may, the discussion here will focus on the central features of the Maluki Ain as it exists today in relation to women’s property rights.?° Broadly five categories of property are recognized — ancestral property, daijo, pewa, self-earned property, and juini—and different rules govern each category, as described below.

Ancestral property. Under the Maluki Ain, wives as well as unmarried daughters over the age of thirty-five have claims in a man’s joint family property. For instance, in a two-generational coparcenary (the usual depth of a Nepalese joint family estate in practice), a man, his wife, his sons, and his unmarried daughters over the age of thirty-five are all coparceners entitled to a single equal share on partition. (Prior to 1975, the unmarried

daughter’s share was half that of her brother’s.) If the property remains undivided for three generations, a man’s agnatic grandsons have a claim only in their father’s share of the coparcenary and not in the entire joint estate. Similarly the wife of a son in a three-generational coparcenary has an

interest (equal to that of her own sons and any unmarried daughters over thirty-five) only in her husband’s share of the joint family estate. Married

daughters have no claims in their father’s ancestral property. If the unmarried daughter over thirty-five subsequently marries, the unconsumed parts of the inheritance (after deducting her marriage expenses) revert to

the heirs of her father. A sonless man can adopt a son or institute his unmarried daughter as a dolaji and pass his ancestral property to her, rather than to his brothers. When she subsequently marries, her children and not her future husband are heirs to the property. Sons cannot force partition on the father unless he denies them mainten-

ance. An unmarried daughter likewise cannot force partition and is expected to live under the guardianship of her parents and her brothers until her death. A wife, however, can force a partition in her husband’s lifetime, in certain circumstances, such as if she is denied maintenance by her parents-in-law or husband, or is expelled from home or otherwise ill-

treated, or if her husband takes a second wife. But these provisions explicitly cover only those situations where the husband has partitioned °° This discussion draws especially on Bennett (1979) and Gilbert (1991). It includes modifications resulting from the enactment of the 1963 Nepalese National Code and the Sixth Amendment passed in 1975, the International Women’s Year.

244 A field of one’s own and is head of acoparcenary. Where the husband still holds property jointly with his father or brothers, it is unclear whether the wife can claim a share or is merely allowed maintenance. Moreover, a sonless widow whose husband has died before partition cannot claim a share and live separately from her husband’s family until she is thirty: till then she only gets maintenance. A divorced woman has rights neither in her father’s nor in her husband’s ancestral property. Certain specified causes of divorce, including mutual

consent, allow her to claim maintenance for five years or until she remarries. But a woman divorced for infidelity or elopement forfeits even this limited right to material support. A widow is similarly divested of her share if she remarries or indulges in any sexual activity. Only in case of rape is her property claim protected. In other words, a woman does not have the same rights in coparcenary property as a man. As a daughter she acquires 2 right only if she remains unmarried till the age of thirty-five, and loses it if she subsequently marries. On marriage a woman acquires a right in her husband’s coparcenary, but loses this right too on divorce, or as a widow if seen to be sexually immoral. A woman’s rights are thus fragile and primarily defined by her age, marital status, and sexual conduct. A man’s rights in the coparcenary property, however, remain unchanged in all circumstances, since he acquires such rights by virtue of birth.

In addition to women’s unequal claims in ancestral land, there are restrictions on the control they can exercise over what they do get. First, although a woman who is married, unmarried, or widowed and living separately can, without anyone’s consent, freely dispose of all-movable property that she has inherited, she cannot dispose of more than half the inherited immovable property. To dispose of any part of the remaining half, an unmarried woman needs the consent of her father, if he is alive, while a married woman living separately or a widow need the consent of their adult

sons. Second, as long as a woman is still a member of her husband’s undivided coparcenary, except when she is acting as head of the family, she cannot undertake financial transactions or enter into contracts on the basis of her eventual shares in her husband’s ancestral property since a creditor, in case of default, cannot make any claims on that property. She can enter into such contracts only when she actually succeeds to her share.

Daijo, pewa, and self-earned property. Daijo or dowry property for women consists of assets (both movable and immovable) gifted to a woman on the occasion of her marriage by members of her natal family and

by neighbours and friends. It usually consists of jewellery, clothing, and household items. Pewa (woman’s own property) can also include movables and immovables (and any increment thereof) gifted by the husband or his

Contemporary laws 245 coparceners (with the consent of all coparceners), as well as gifts from other friends and relatives on the husband’s side. However, both daijo and pewa need documentary evidence as to their origin, to protect such property from being included in the ancestral property of the husband’s lineage. A woman can use both daijo and pewa as she wants, including willing the property to anyone she chooses. If she leaves no will, however, the assets accrue first to the woman’s sons living jointly with her, if any; then to sons separated from

her, and failing these to her husband, unmarried daughters, married daughters, sons’ sons, daughters’ sons, or the nearest relative on the husband’s side.?! In other words, sons and husband inherit from a woman’s

property before daughters, and it is the heirs of her husband, rather than relatives from her natal lineage, who come into the order of succession. Over self-earned property®* similarly a woman has absolute rights of testation or disposal, again provided she can produce documentary proof of its origin and show that it did not belong to or derive from ancestral property. Otherwise the law courts assume that all her property, even that which is self-earned and in her own name or bank account, is a part of the coparcenary or family property rather than exclusively hers. Effectively, therefore, if a woman has no documentary evidence, her property can be deemed joint family property in which all the coparceners have equal claims, while she has rights only in the husband’s share of the joint estate.

Juini. This is a share in the ancestral property conferred on a person for her/his maintenance, in lieu of the ancestral share. In theory, it is identical in size to an official share of family property, but it can be greater or less than that share by 5S per cent without being legally challengeable. If it deviates from the official share by more than 5 per cent, the recipient or other claimants to the disputed amount can challenge it in court.°? It isa limited right which guarantees the person economic support but does not

allow her/him the freedom to dispose of the property, and is usually conferred on widows, old men, or elderly parents. In popular parlance, the term juini bhag is also used to mean the portion of coparcenary property a person takes to sustain herself or himself for the rest of her/his life. A father can, for instance, allot a juini share for himself when he divides his property among his children. The share subsequently passes to whichever relative (or spouse) has cared for him in old age. Today, a daughter too can inherit the °! The daughters were included as fourth and fifth in line only after an amendment in 1975 (Bennett 1979: 23). 92 This is property earned by an individual, before or after marriage, through the exercise of some skill, or as an heir of another person, or as the recipient of a legal gift or award (see Gilbert 1991: 13). °3 Personal communication, Kate Gilbert (faculty, Amherst College, USA) 1991.

246 A field of one’s own parents’ entire juini portion, and a sister can likewise inherit the brother’s entire juini portion. In sum, there are several basic inequalities faced by Nepali women under existing property laws. First, unless a woman has a large (documented)

daijo, pewa, or self-earned property, her access to land is essentially dependent on marriage and chastity. Men inherit landed property mainly as sons and their rights are not conditional (like women’s are) upon their age, marital status, or sexual conduct. Second, a woman’s right of disposal over what she inherits is restricted in ways that men’s 1s not. Third a woman

can only obtain credit or make a binding financial transaction if she has some exclusive property (again, such as documented daijo, pewa, or selfearned assets), or if she has succeeded to a share in her husband’s ancestral property through partition (in her husband’s lifetime), or on widowhood. In contrast, an indebted husband can draw upon his wife’s coparcenary share in the joint estate and thereby reduce its amount. Also a woman can lose even her daijo, pewa, and self-earned property unless she can prove its origin through documents. Fourth, divorce, legal separation, or, in the case of a widow, the failure to remain chaste, requires the woman to forfeit the share she gets from her husband’s estate (and any increment thereof) in favour of her husband’s nearest relative. Hence even though in theory a Nepalese woman can claim a share in her husband’s joint family estate, in practice this becomes a concrete reality only if she remains married to him.

V. Summary comments on women’s legal rights in landed property in South Asia

In overview, several features stand out when we examine women’s legal rights in landed property in the five countries under study. Traditionally, prior to the 1950s, women of most South Asian communities had few and restricted inheritance rights in landed property. Noteworthy exceptions were communities practising matrilineal (tribal and non-tribal) and bilateral inheritance in India and Sri Lanka. Today, legally in all five countries,

most women have significantly greater inheritance rights in landed property than they did before. However, gender inequalities in laws persist, with the exception of the General Law in Sri Lanka. These inequalities vary

across regions, being greatest under the Maluki Ain laws applicable in Nepal. In relation to succession the gender inequalities are of five kinds. First, among several religious groups women are entitled to smaller shares than men. For instance, under Islamic law in all of South Asia, and under the laws governing some categories of Christians in India, daughters inherit

Contemporary laws 247 only a portion of what the sons do. Also under Hindu law in India, the vestiges of the Mitakshara system give sons but not daughters rights in certain categories of property. Second, in some cases women are legally entitled only under restrictive conditions and can lose even those rights in

ways that men cannot: for instance, in Nepal women can inherit as daughters only if unmarried and over thirty-five, and have to forfeit their claims if they subsequently marry. Married Nepalese women lose their rights in their husband’s property on divorce or if they are unchaste. And in Sri Lanka, among the Kandyan Sinhalese, daughters lose their rights in the patrimony if they marry diga. Third, under several legal systems there are restrictions on women’s ability to dispose of what they might inherit: for instance, under the Maluki Ain women need the consent of men (fathers, brothers, or sons, as the case may be) to dispose of part of their inherited immovable property; and under the Thesawalami Jaffna Tamil women who are married need the husband’s consent for disposing of any part of their immovable property. Fourth, in India, tribal communities of the northeastern states continue to be governed by uncodified customary law under

which, among patrilineal tribes, women’s rights in land are severely circumscribed and typically limited to usufruct. Fifth, in India there are specific gender biases pertaining to the devolution of agricultural land: for example, in several states, especially of northwest India, rules of devolution specified in various land enactments (in the form of tenancy and other land ‘reform Acts) which give priority to male agnatic heirs, supersede the rules of devolution spelt out for Hindus in the Hindu Succession Act of 1956.

Similarly for Indian Muslims, again especially in several northwestern states, customary practices (and land reform laws) which are more genderunequal than Islamic law supersede the latter in relation to the inheritance of agricultural land. Apart from the inheritance rules, land reform enactments contain other serious gender inequalities. For instance, in laws pertaining to the fixation

of ceilings in India, many states allow additional land to a cultivator on account of adult sons, or adult sons are directly allowed to hold such additional land in their own right, but adult daughters are given no such consideration; and in several states, unmarried adult daughters do not figure either as part of the family unit, or as independent units. Also in the assessment (and confiscation) of ceiling surplus land, a woman does not count as an owner in her own right, and her land is clubbed with that of her husband, leaving her vulnerable to disproportionately losing her land. These ceiling laws affect all religious communities. Again it is the states of northwest India which are the worst (although not the only) offenders on this count.

To bring about gender equality in laws pertaining to property, and especially those pertaining to agricultural land, these inequalities and

248 A field of one’s own

anomalies will clearly need to be addressed. In India their persistence violates both the letter and principle of gender equality that is promised as a

fundamental right under the Constitution. In addition, certain aspects of the laws which are not technically gender-discriminatory but become so in practice, need amendment. For instance, the power of unrestricted testation under Hindu law in India could, in theory, affect sons and daughters equally, but given the social bias in favour of endowing males it works to women’s disadvantage. There is thus a case here for limiting testamentary power to protect female heirs. Doing so would not be without precedent: a number of legal systems in Europe, for instance, also restrain testamentary freedom (Sivaramayya 1973: 64), as does Islamic law. Again, there is no concept of community of property after marriage under Hindu law. Hence, on divorce, a woman gets no property benefit from any direct or indirect economic contribution she makes during her marriage toward increasing her husband’s wealth. There is a case here for establishing equal rights for both spouses in the ownership and control of property acquired by either spouse after marriage, and for an equal division on divorce. Removing the noted legal inequalities will involve a continuing process of contestation and struggle. But if women are to play a significant role as

law makers and not just as law takers, this struggle will need to be increasingly conducted not only from outside the State apparatus, but also from within it. Greater representation by gender-progressive women in the legislatures, the judiciary, and the many tiers of the legal machinery, would be a necessary part of that process, as would be the creation of a favourable ideological climate, the absence of which can lead to retrogressive shifts, as illustrated especially by women’s experience under regimes dominated by religious fundamentalist forces in the subcontinent. Moreover, legal restrictions are only one part of the story. The other part (told over the next three chapters) relates to the many formidable obstacles that constrain women from exercising even the rights they currently have.

From that perspective, although legal transformation is a critical step forward, it is only one of many steps that must be taken.

6 Whose share? Who claims? The gap between law and practice

To my brother belong the storied palaces; But, alas, for me, the foreign land... O my father.! If I take my inheritance, my brothers will forget they have a sister. If I give it to them, they will remember me and take care of me if I need them.

Women of most South Asian communities today have considerable legal rights to inherit landed property. But to what degree can they exercise their rights in practice? What factors constrain them from doing so fully? These questions are addressed here primarily in relation to traditionally patrilineal communities, with a brief contrasting look at traditionally matrilineal and bilateral ones. Although regional differences are highlighted at various points in the discussion below, a more systematic examination of the crossregional variations is deferred to chapter 8.

I. The gap between law and practice in traditionally patrilineal communities

Ethnographic information, although it is extremely fragmentary, consistently indicates that women in traditionally patrilineal communities of South Asia rarely realize the rights that contemporary laws have promised them. Custom still dominates practice. Hence the vast majority of women

do not inherit landed property as daughters, most don’t do so even as widows, and few women inherit in other capacities. To the extent women inherit, it is usually under very restricted conditions. As daughters, women’s claims appear to enjoy little social legitimacy, and the greatest likelihood of daughters inheriting is still in sonless families, ' From a popular Urdu song sung on the bride’s behalf by her young companions when a Muslim bride in northern India departs from her father’s house (Stuers 1968: 25; she cites a number of such songs sung to her by several different informants). ? A woman in Katni village, Bangladesh, to Hartmann and Boyce (1983: 92-3). 249

250 A field of one’s own usually involving uxorilocal post-marital residence.* Even then, the woman does not typically gain full ownership of the land, but serves as a trustee on

behalf of her son who ultimately inherits; occasionally her husband (the son-in-law) is designated heir.* However, endowing even brotherless daughters is explicitly forbidden by custom by some communities, such as the Gaddis of Himachal Pradesh (India), among whom women cannot hold any land, not even self-acquired (Newell 1962). Sonless Hindu couples may also adopt a male child (usually an agnate’s son) and designate him as heir, thus by-passing the daughter.> Cases of daughters inheriting land directly

and unconditionally from parents are therefore rare, especially in the northern parts of South Asia, and most relate to sonless families.° In Ramkheri village (Madhya Pradesh, central India), out of 146 persons with land registered in their names, 121 (83 per cent) were sons inheriting from fathers, eight were adopted male heirs, and only seventeen were women. Of

the women, only five were daughters (three inheriting from widowed fathers and two from widowed mothers), while nine were widows, and three were sisters inheriting from childless widower brothers (Mayer 1960). In Nepal, among the Tibetan Chumiks, cases of daughters inheriting land are almost entirely confined to sonless families, and in these usually the eldest daughter alone inherits the estate: Schuler (1987: 98) found that 11-12 per

cent of married daughters in her study had inherited estates from their fathers; all belonged to sonless families. Among them in 8-9 per cent of the cases the husband was living uxorilocally on the wife’s estate, and in the

remaining 3 per cent the woman had combined her inherited estate with 3 The studies listed below all mention one or a few cases of daughters inheriting in sonless families in their study area: For Bangladesh, see Ellickson (1972a, 1972b), Hoque (1987), Islam (1974), Jansen (1983), Nath (1984), and Schendel (1981). For India, see Ahuja (1966), Bailey (1957), Beck (1972), Bradford (1985), Cohn (1961), Fukutake er a/. (1964), Gupta (1974), Harper (1971), Hill (1982), Katiyar (1967), Kolenda (1983), Madan (1989), Mathur (1964), Mayer (1960), Nag (1960), Nicholas (1961), Orenstein (1965), Phylactou (1989), and U. Sharma (1980). For Nepal, see Fisher (1987), Hitchcock (1980), Krause (1982), Levine (1982), Macfarlane (1976), Molnar (1978), Ross (1981), and Schuler (1987). And for Pakistan, see Aschenbrenner (1967), Elgar (1960), Khan et al. (1984), and C. Pastner (1971). In Nepal the formal rules of inheritance under the Maluki Ain, as noted, do not allow married daughters to inherit, but custom has been known to deviate from those rules in rare cases, especially in sonless families. 4 For India, see Bradford (1985), Cantlie (1984), Gupta (1974), Harper (1971), Parry (1979), and Sengupta (1966). For Nepal, see Acharya (1981), Furer-Haimendorf (1956), Holmberg (1989), and Molnar (1981). And for Pakistan, see Aschenbrenner (1967). 5 For India, see Beck (1972), Bradford (1985), Gupta (1974), Hershman (1981), and Sharma (1973). © Studies mentioning such occasional cases in India are Bailey (1957), Fukutake e¢ al. (1964),

Mayer (1960) and Wolkowitz (1984). Interestingly, some of Wolkowitz’s informants told

her, during her research in Andhra Pradesh (south India) in 1975-76, that women of landholding households of the Kamma caste sometimes inherited land from their mothers, and were trained in agricultural management by their parents, an aspect of inheritance that is little documented (Wolkowitz 1984: 121).

Whose share? Who claims? 251 that of her husband. In Bangladesh, only four out of forty Hindu and Muslim women in Kumirpur village whom White (1992: 129) chose for detailed study had inherited land as daughters, and they had received less than their entitled shares. Nath (1984) mentions a somewhat larger number of cases in Natunpur village (Bangladesh), but again these constituted only a small percentage of those with legal claims, and many involved uxorilocal residence by the husband. In this context, it is important to examine Goody’s (1990: 287) interpretation of the fact that daughters among patrilineal Hindu families in India, ‘even if not normally allowed to inherit, may indirectly act as residual heirs when in the absence of brothers, they pass on the estate to their sons’. He argues: Great significance should be given to this institution of the ‘appointed daughter’ who acquires a marrying-in husband or an adopted son. Because once the claim of a brotherless daughter is admitted, and it is an extension from her claim on movables in the shape of the dowry, the door is opened for women to share in the family lands themselves even when they have male siblings.

This argument is problematic on at least two counts. First (as also noted in

chapter 3), the daughter is ‘appointed’ to receive the father’s property essentially to pass it on to her son (the grandson). The property she receives in that capacity thus cannot be seen as ‘her claim’ over the father’s property,

but is merely a means used by her natal family to keep landed property within the patrilineage. From the woman’s point of view, there 1s a critical difference between having a right to a share in family land which she can control and use as she pleases (including, if she wishes, passing it on to

daughters), and being used essentially as a vehicle for the transfer of property from one generation of men to another, namely from her father to

her son. Second, as noted in chapter 3, a dowry in movables is not equivalent to an inheritance share in land; and although a popular fiction is maintained that dowry constitutes a daughter’s inheritance share, in fact the difference between a dowry in movables and a share in landed property is implicitly well recognized by communities throughout South Asia. As elaborated later in this chapter and the next, there is widespread resistance to the idea of daughters getting any share in landed property in northern India,’ even in communities which may give daughters dowries or recognize the custom of ‘appointed daughters’, an empirical reality which does ’ The terms ‘northern India’ and ‘north India’ are used here and elsewhere very broadly to distinguish the northern parts of India from the southern peninsula (viz. ‘south India’) and the northeastern (mainly tribal) states. In much of the discussion in the book, however, finer geographic distinctions have been drawn between the northwestern, western, central, and eastern regions of ‘northern’ India. The states included in these four regions, as well as in the southern and northeastern ones, are as follows: northwestern includes Haryana,

252 A field of one’s own not suggest that this custom has opened any doors for daughters in general to shares in the family land. In families which do have sons, the chances of daughters inheriting land are usually very low, although within this general pattern there are regional and community-wise variations. For instance, there is some indication that among Muslim communities, with the operation of the Shariat, daughters with brothers now inherit occasionally.® From the few studies that discuss the conditions under which this takes place, it appears to be more likely where the daughter is married within the village or is married to a crosscousin:? both in-village and close-kin marriages keep the land within the overall control of the extended family. The amounts inherited, however, are usually less than the daughters’ legal shares. Also many more appear to be

formally recorded as landowners than acquire the land in practice. For instance, Aschenbrenner (1967: 61) found in Pakistan Punjab that although

many daughters were listed as landowners in the village land records, indications were ‘that they usually do not claim their rights and that trouble results when they do’. Similarly, in the Makran region of Baluchistan, C. Pastner (1971: 157) found that in 1961, out of 16,157 registered owners of agricultural land, 4,017 (that is 25 per cent) were women; but she also notes that the figures do not differentiate between daughters and widows. Hence the percentage of daughters among the landowners would be less than this.

An additional possibility is that in a number of these cases noted by Aschenbrenner and Pastner, families have registered land in women’s names to evade land ceiling laws, without involving real land transfers to the women. Be that as it may, all the above instances of daughters inheriting land add up to only a small proportion of those so eligible; most village studies (especially those relating to Hindu communities) mention only one or two cases or none at all. In Chen’s (1992) recent survey on widows in seven states Himachal Pradesh, Jammu and Kashmir, Punjab, Rajasthan, and Uttar Pradesh: western includes Gujarat and Maharashtra; central includes Madhya Pradesh; eastern includes Bihar, Orissa, and West Bengal; southern includes Andhra Pradesh, Karnataka, Kerala, and Tamil Nadu; and northeastern includes Arunachal Pradesh, Assam, Manipur, Meghalaya, Mizoram, Nagaland, Sikkim, and Tripura. 8 For Bangladesh, see Arens and Van Beurden (1977), Cain et al. (1979), Jahangir (1979), Nath (1984), Qadir (1981), Westergaard (1983), and White (1992). For India, see Roy (1984). For Pakistan, see Alavi (1972), Aschenbrenner (1967), Lindholm (1982), C. Pastner (1971), and Rouse (1988). ° The village headman in Aschenbrenner’s (1967) study for Pakistan Punjab emphasized that the daughters who inherit are those married within the village; and in the one case of such inheritance mentioned by Lindholm (1982) for the Swat region (which falls in the ‘Settled

Areas’ of the NWFP), the woman married her father’s brother's son. (It may be recalled from chapter 5 that the NWFP of Pakistan is administratively divided into two regions: ‘Tribal Areas’ and ‘Settled Areas’.)

Whose share? Who claims? 253 Table 6.1: Widows in India who inherited land as daughters

Total Those who sample inherited!

Region/State No. No. % Northern India 262 15 6

Bihar 71492I 32 Rajasthan Uttar Pradesh (hills) 50 l 2

West Bengal 92 1] 12

Southern India 283 40 14 Andhra Pradesh 79 12 15

KeralaNadu 104100 26 2252 Tamil

All regions 545 pp) 10

Note: ' Unfortunately, there is no information on how many of the widows’ natal households possessed land. The percentage inheriting would be somewhat higher if calculated only for widows who came from landed parental households, but they would still constitute only a small proportion of those with legal claims. Source: Chen (1992)

of India, in only 10 per cent of her sample of 545 Hindu households with widows had the widows inherited as daughters (table 6.1).!° In four states (Bihar, Rajasthan, Uttar Pradesh hills, and Tamil Nadu) the percentage was less than three, while in three states (West Bengal, Andhra Pradesh, and Kerala) it was greater. The north-south contrast was also marked: 6 per

cent of the women in the northern states, relative to 14 per cent in the southern ones, inherited. Kerala, as might be expected (given a significant presence of traditionally matrilineal groups), had the highest percentage inheriting. What is less expected is the noticeable (albeit not dramatic) deviation of West Bengal from the typical northern Indian pattern of very

few daughters inheriting. The possible reasons for this deviation are discussed further on. What is striking in overall terms, however, is that in all

the states, including Kerala, a significant majority of women, although legally eligible, do not inherit as daughters. '0 Information on what percentage of widows inherited from their deceased husband's estates is not yet available.

254 A field of one’s own Inheritance aside, in rich families fathers or brothers sometimes grant women usufruct rights to small plots,!! or gift them some land, especially but not necessarily as a part of dowry.!? In India, land gifts in dowry are rare and tend to be confined to the southern states.'? Among the Chumik of Nepal, on the other hand, they are not uncommon, but the proportion of family land given as dowry is neither large nor equivalent to the shares that sons inherit: in 46 per cent of the marriages examined by Schuler (1987: 102-3), daughters had received land as dowry, but (as noted in chapter 3) in

families with both sons and daughters the total value of the dowries, including land, on average came to only 10 per cent of the family’s assets, while the sons received the rest. In some tribal communities, unmarried daughters customarily have usufruct rights to land, as among the Ho and

Santal tribals of Bihar in eastern India. In fact, a number of Ho women today choose to remain unmarried for the sake of this access, although they lose these rights if made pregnant by a Ho man or if raped by or discovered

to have had a sexual liaison with a non-tribal (Kishwar 1987). Similarly Phylactou (1989) notes that one of the important reasons why many women in the Tibetan community he studied in Ladhakh remained unmarried and became Buddhist nuns, was because they were then entitled to an independent house and independent usufruct rights in fields belonging to their natal families. A few also cultivated land leased to them by the monastry. As widows, women’s claims enjoy somewhat greater social legitimacy than their claims as daughters. In India, for instance, the perception that a widow has a right to a share in the deceased husband’s land appears to be fairly widespread.'* In practice, however, the fragmentary available evidence suggests that many of those who are eligible to inherit do not, and those who do inherit do so mostly on severely restricted terms. In a rural Hindu household in India, for instance, the extent and nature of rights that a widow enjoys in her husband’s land are contingent in practice on a variety of factors, such as whether or not she remains single and chaste; whether she 1! For India, see Archer (1984), Furer-Haimendorf (1979), Haekel (1963), Kishwar (1987), and Standing (1987). For Nepal, see Molnar (1981), and Schuler (1987). And for Pakistan, see Lindholm (1982). 12 For Bangladesh, see Hartmann and Boyce (1983), and Wallace et a/. (1988). For India, see Epstein (1973), Furer-Haimendorf (1979), Gough (1981), Karve (1965), Lakshmanna (1973), Laxminarayana (1968), Madan (1975), Tambiah (1973), and Shobha Srinivasan (personal communication based on fieldwork in Tamil Nadu). For Nepal, see Andors (1976), Fricke et al. (1986), Hitchcock (1980), Rajaure (1981), and Schuler (1987). And for Pakistan, see D. Merry (1983). 13 The following studies which mention land gifts in dowry all relate to south India: Epstein (1973), Gough (1981), Lakshmanna (1973), and Shobha Srinivasan (personal communication). 14 Personal communication from Marty Chen (Harvard Institute for International Development, 1993) and from a former sarpanch (head of the village council) in Jhunjhunu district, Rajasthan, 1993.

Whose share? Who claims? 255 has sons and her sons (if any) are minors or adults; whether the deceased husband has partitioned from the joint family estate before his death; and so on. To begin with, as under traditional Hindu law, a widow usually loses her right if she remarries, is unchaste or leaves her husband’s village on his

death.'* If she has only daughters or is childless she often gets only maintenance,'° although there are studies relating to some communities of northwest and central India that mention Hindu widows with no children

or only female children inheriting a limited interest in the husband’s estate:'’ unfortunately, with the exception of Mayer (1960), discussed below, none indicates the actual number of such cases. A woman with minor sons is usually allowed use of the husband’s estate as a trustee on behalf of her sons till they grow to adulthood, after which she is expected to

live with one of them.!® On adulthood the sons are usually likely to partition the land during their mother’s lifetime; in such cases a part of the

land may well be marked out as hers, but it is generally expected to be

cultivated by the son she lives with, rather than designated for her independent use.!? In most cases women do not inherit the absolute estate they are entitled to under contemporary law. (The regional variations in this are discussed below.) A widow whose husband had not separated from the joint estate before death is likely to be given only use rights to a part of his share without her name being entered into the records: I came across more than one case

in Kithoor village (Rajasthan) of a widow with a minor son cultivating a small portion of her deceased husband’s share in the joint estate which was formally still in the name of the father-in-law.?° If the joint property is partitioned before the husband’s death, a widow with sons is more likely to be able to get a formal registration of her rights in her husband’s land, but usually this is done jointly with the sons: for instance, in one recent survey of two villages in Rajasthan (northwest India), sixteen out of the fifty-seven Hindu widows surveyed (that is, 28 per cent) had their names in the land

records in relation to their deceased husbands’ lands. Of these, one was sonless and the land was solely in her name; the remaining fifteen were registered jointly with their adult sons (Nandwana and Nandwana 1992). The same picture obtains from my survey in March 1993 of land records in "> See Harper (1971), Mayer (1960), Parry (1979), and Nandwana and Nandwana (1992).

1¢ This was the widespread perception among widows and other villagers I spoke to in Kithoor village (Rajasthan) in 1987. Also see Hershman (1981) for the Indian Punjab and Harper (1971) for Karnataka. 17 See Madan (1989), Mayer (1960), and Parry (1979). 18 See Gupta (1974), Harper (1971) and Hershman (1981). 19 Jean Dreze (London School of Economics), personal communication, 1992; and Chen and Dreze (1992).

20 In general, village land records are poorly maintained and seldom kept up-to-date (see Wadhwa 1989, and the discussion on this later in this chapter).

256 A field of one’s own three villages in Jnunjhunu district (Rajasthan): this showed that of the thirty-six women with land in their names, thirty-four were widows and two

were daughters. Of the widows, all the twenty-seven who had sons were registered along with their sons. I understand from a former sarpanch of this region that the registering of widows’ claims has only become common here over the past four or five years.

Having land in her name does not mean that the widow is allowed full control over it, to use, mortgage, sell, or will it as she wants. In the above surveys in Rajasthan, the popular perception among the villagers was that

this land was meant for the widow’s maintenance rather than for her independent control, use, or transfer. Nevertheless, formal registration is a step forward in establishing a woman’s legal claim. Within this overall quite restricted inheritance of land by Hindu widows in India, there are, however, some noteworthy regional variations within north India and between north and south India. For instance, Dreze (1990 and personal communication) found that in his 1988 survey of three villages in north India (one each in West Bengal, Gujarat, and Uttar Pradesh), in the West Bengal village, in five out of nineteen Hindu landed households with widows, all the land owned by the family was in the widow’s name;?! and in two others, the land was still in the deceased husband’s name but seen as belonging to the widow. These together constituted 37 per cent of

the cases. Five of these seven widows with land had adult sons. Dreze contrasts this with his Gujarat and Uttar Pradesh villages, where there were

virtually no cases of widows with adult sons who had land in their own names, although there were cases where a small share of the family land was allotted for a widowed mother and cultivated by whichever son looked after her. In terms of daughters’ rights also, we had noted from Chen’s (1992) data that West Bengal was somewhat more progressive than other states in northern India. If this is indeed a pattern, it warrants further investigation.

The reasons could lie in factors such as the following: the Dayabhaga inheritance system that historically prevailed in Bengal, which was somewhat more favourable toward widows and daughters than the Mitakshara system prevailing elsewhere in the North;? the social reform movements in

Bengal in the nineteenth century; and the less strict norms of female seclusion in Bengal (for instance, veiling is not customary among Hindus in 21 Unfortunately, in four of these cases there was no information on how the women got the land; the fifth one had inherited it from her husband. 22 As detailed in chapter 3, under Dayabhaga a sonless widow was entitled (as a limited interest) to her husband’s entire estate, while under Mitakshara she had a limited interest only in his separate property. We had also noted how in late nineteenth-century Bengal, some wealthy sonless widows were found to be managing zamindari estates in which they had inherited a limited interest.

Whose share? Who claims? 257 Bengal, as it isin Uttar Pradesh and Rajasthan) which would allow women greater freedom in affirming their claims and in managing land.??

The above patterns relating to north India, in turn, contrast in overall terms with south India, in the likelihood of widows inheriting. Preliminary results from Marty Chen’s recent survey on widows indicate that widows’ land claims are relatively better recognized in the south than in the north (personal communication, 1993).2* Also in chapter 3 we noted that even in the medieval period a number of sonless widows of wealthy families in south India had possession of their husbands’ estates and enjoyed some

degree of freedom in using their wealth for endowing temples. Again, Gough (1981: 208-9), on examining land records in Kumbapettai village of Tamil Nadu, found that in 1952, 19.4 per cent of all registered landowners were women (including both widows and non-widows). This represented a

notable increase over 1827, when there were no female registered landowners in the village; land shares were then generally held in joint estates

by male heirs, and women had rights only of maintenance. It also represented an improvement over 1897, when 12.2 per cent of registered landowners were women: by then the partitioning of extended joint family estates had become more common and land had begun to be registered in individual names. Also as land shares became individual property, men sometimes gave small plots to daughters as dowry.

Among many tribal groups in India and among Hindu and TibetoBurman communities in Nepal today, the pattern appears to be one of recognizing, at best, the widow’s right to a limited interest in her deceased

husband’s estate if she has no sons,?° and to allow the widow who has minor sons to hold the land on their behalf. Among the Kham Magar studied by Molnar (1978), a sonless widow gets half the husband’s estate as 23 Also see the discussion in Dreze (1990). Dreze gives more credit than appears warranted to communist rule in West Bengal in promoting women’s property rights. Certainly (as noted

earlier) the land reform programme (Operation Barga) launched by the West Bengal government in 1978 showed no significant deviation from the male-biased land distribution policies followed elsewhere in the country.

24 The percentage of widows in the rural female population is also greater in the southern states relative to the northern, especially the northwestern ones. Possible reasons for this, suggested by Dreze (1990: 28-33), include northern India’s lower female life expectancy rates, greater age differences between the spouses, and acceptance of levirate. Also, he hypothesizes, widows may have lower survival chances than other categories of women in northern India. *5 For Nepal, see Krause (1982), Macfarlane (1976), March (1979), Molnar (1978), and Rajaure (1981). There is also an occasional case of a Nepalese woman living separately from her husband but continuing to use part of the latter’s land which, in due course, would be inherited by her son (see e.g. Schroeder and Schroeder 1979). For tribal groups in India, see Furer-Haimendorf (1979), Haekel (1963), Kishwar (1987), and Sachchidananda (1978). In Furer-Haimendorf’s (1979) study of the Gonds of Andhra Pradesh, a sonless widow keeps the husband’s estate only if she marries a member of the husband’s kin, and forfeits it if she marries anyone else.

258 A field of one’s own a limited interest, but one with minor sons can hold the whole of it. Among

the Tharus, Sherpas, Tamangs, Thakuris, and Chetris of Nepal, the chances of a widow without sons getting any land at all are noted to be small.?° A sonless or childless Hindu widow (in both India and Nepal) sometimes

adopts a son, usually her husband’s brother’s child, and designates him as heir. Among the Kham Magar, Molnar (1981) found that six widows in her study village had done so. Gough (1981) notes several cases in her village study in Tamil Nadu.*? And in Ramkheri village (central India), of the nine widows who had inherited land, Mayer (1960: 244) found that three had adopted their husbands’ agnates as heirs, two others had young sons in whose names the property would be registered when they were old enough to work it, two had invited a daughter’s husband to live uxorilocally to work the land which the daughter and son-in-law would later inherit, while only two had no heirs living with them and farmed the land through tenants or labourers. In other words, land in the hands of widows usually tends to pass to male heirs in the next generation. Of course even a limited interest in land, although not equivalent to full property rights as granted under contemporary law, is of significance, since

it gives the woman a right to the produce from the land and can be an important source of economic security. It also restricts the access of her deceased husband’s relatives to that piece of property for the duration of the woman’s lifetime. Therefore, even women’s limited interest in land can be resisted strongly by their husbands’ kin, as will be discussed later in this chapter.

In Muslim communities, although most widows again inherit only as legal custodians on behalf of sons,?® there are also cases of widows with land in their own names.*° On the whole, though, the women who inherit would still be only a small percentage of those legally eligible. In Bangladesh, in a study of Shaheenpur village, only five out of nine widows were found to have received their shares (Ellickson 1972b); and in the villages of 26 See March (1979) on the Sherpas and Tamangs; Krause (1982) on the Thakuris and Chetris, and Rajaure (1981) on the Tharus. 27 She found that sometimes disputes arose between widows and their adopted sons over control of the property: in such cases the court usually decided in favour of the adopted son,

or an out-of-court settlement took place, with the land being divided between the contending parties or with the woman agreeing to pay income to the son.

28 For Bangladesh, see Aziz (1979), Hartmann and Boyce (1983), Jahan (1975), and Taniguchi (1987). For Pakistan, see Ahmed (1980). Ahmed notes that in the ‘Settled Areas’ of the NWFP (Pakistan), a widow can only keep her husband’s land and house if she has a son, and tn the name of the son. *9 For Bangladesh, see Abdullah and Zeidenstein (1982), Aziz (1979), and Ellickson (1972b). For Pakistan, see Akhtar and Arshad (1958), Aschenbrenner (1967), Asha (1971), D. Merry (1983), Qadir (1981), and Rouse (1988).

Whose share? Who claims? 259 Bogra and Pautakhali, 14 and 28 per cent respectively of the sampled widows and deserted women had some land, most owning under 0.5 acres (Qadir 1981). Of the twenty-five rural widows interviewed by Aziz (1979), only nine had sufficient agricultural land for subsistence, while the rest were reduced to beggary. However, there is some indication from recent research that where fathers or brothers do not voluntarily give them their shares, village women in Bangladesh are beginning to contest this, albeit in limited degree, as discussed later in this chapter. In capacities other than as daughters and widows, women (especially among Hindus) virtually never get land. It is very rare for a sister to inherit. Mayer’s (1960) ts one of the few ethnographies that mention such cases, in this instance of three sisters inheriting from brothers who were widowers and issueless. There are also likely to be cases in Maharashtra and Gujarat (formerly part of Bombay Presidency), where (as noted in chapter 3) the rights of sisters were traditionally recognized, even while they were denied elsewhere. The possibility of rural women acquiring land through means other than inheritance are usually small. In rare cases, women are observed to have received gifts of small plots from husbands, typically in the form ofa limited interest.°° Theoretically it is of course possible for women to buy land from

their own resources. But the barriers to this can be three-fold: limited economic resources in women’s hands, even in households where their husbands’ families or natal kin are well-off; difficulties in undertaking land transactions without male mediation; and (as noted in chapter 1) the dearth

of arable land for sale except in a generalized crisis context, as during a drought or famine. Hence cases of women purchasing land are few, and the plots purchased are usually very small.3! Against this backdrop, it is necessary to emphasize that the widespread

disinheritance of women as daughters is a critical gender disadvantage which cannot be made up even if women’s rights as widows are fully recognized. Widows constitute only a small percentage of the rural female

population in South Asia, and most belong to the upper age groups: in India, according to the 1981 census (see GOI 1987c: 47-8), widows constituted only 11.3 per cent of the rural female population over the age of ten, 53 per cent of whom were sixty years of age or older, and 76 per cent were over fifty.** Hence, once disinherited as daughters, most rural women 39 See e.g. Nag (1960), Plunkett (1973), and Sachchidananda (1968). 3! Hartmann and Boyce (1983), Jones (1977), and Nath (1984) mention such cases. 32 In Bangladesh, Nepal and Pakistan, likewise, by these countries’ 1981 censuses, widows constituted 12.2, 5.5, and 6.8 per cent respectively of the rural female populations over ten years of age (see, Government of Bangladesh 1984: 65, Government of Nepal 1987b: 39, and Government of Pakistan 1985b: 41).

260 A field of one’s own for the major part of their lives would have no land of their own, while males whose inheritance claims as sons are well recognized, would. This places women in a Significantly weaker bargaining position vis-d-vis men, both within and outside the family.

Also, as a result, women’s situation in case of divorce can be one of extreme vulnerability, as is apparent especially in Bangladesh today, where the rate of male-initiated divorce and desertion is high. The Bangladesh Fertility Survey of 1975 showed that one-fifth of all first marriages ended in

divorce or separation. And while most of the men remarried, a large percentage of the women did not (Miranda 1980: 83). The chances of women not remarrying are especially high if they have children. Several divorced and widowed women in Bangladesh told Arens and Van Beurden (1977: 57) that a new husband would refuse to take their children, since ‘the

children would only mean more mouths for him to feed’, and that they would never leave their children to remarry. At the same time divorced women, Hindu or Muslim, and not only in Bangladesh but in much of South Asia, have no effective means of making claims on the incomes of their children’s fathers for child support. Islam recognizes a woman’s right to ask for payment of the sum pledged as mehr if the husband divorces her, but in practice mehr is seldom demanded or paid.*? Also, there is no legal

recognition of a community of property, whereby property acquired by either spouse after marriage would be owned jointly by both. Hence a

divorced rural woman may easily be left with little or no economic support.3* In such a context inheritance of land would make a critical difference to her ability to sustain herself and her children. What then prevents women from claiming their shares, especially as daughters?

Il. Barriers to women inheriting land in traditionally patrilineal communities

(1) ‘Voluntary’ giving up of claims Typically women give up their claims in parental ancestral land in favour of their brothers, for a complex mix of reasons. Village women’s explanations 33. For Bangladesh, see Alamgir (1977), Kabeer (1985), and Lindenbaum (1981). For India, see Jeffery (1979). For Pakistan, see Ahmed (1986), Kurin (1981), and Lindholm (1982). Also see the general discussion in Qureshi (1992). Indeed it is on the assumption that mehr will not need to be paid that sometimes quite large sums are promised by the groom’s side among wealthy families (see e.g. Ellickson 1972a). A similar custom of /ub is found among the Baluch (C. Pastner 1971): this represents a promise of payment by the husband in the event of divorce, and often includes land, trees, water rights, etc.; effectively this is a part of the husband’s expected inheritance from his father, which may remain in the father’s name till the latter’s death. 34 The issue of a divorced Muslim woman’s right to maintenance became a source of major legal debate and agitation in India in 1985-86, in the context of the Shah Bano case. Fora

Whose share? Who claims? 261 for this have a common refrain across the subcontinent. In north India, an old woman from Punjab says: If a brother and sister are on good terms then ... she will tell her brother that she does not want her share of the inheritance. After all, if he eats, then she can eat (U. Sharma 1980: 57).

In Rajasthan I heard the following views commonly expressed: A sister gives up her claim in order to keep the passage to her natal village open.

If the sister claims the land, then she will have only the land. But if she maintains good relations with her brothers, she will have a constant flow of gifts each time she visits them.

If a woman stakes her claim, her brother’s wife will refuse to invite her home or speak to her.

Women in Katni village (Bangladesh) expressed very similar views to Hartmann and Boyce (1983), as indicated by the quotation at the beginning of this chapter. The crucial importance these women place on their relationships with brothers and on access to their natal homes can only be understood in the context of their overall life situations: early arranged marriages; patrilocal residence and village exogamy; economic, social, and physical vulnerability in case of marital discord, ill-treatment, marriage break-up or widowhood; and ritual connections and strong emotional ties with brothers. Access to the natal home can be a significant element in women’s economic security and fall-back position, and brothers are a critical link to the natal home even when the parents are alive, but especially after their deaths. These

factors vary in strength across communities and regions (as will be elaborated in chapter 8), and with them the criticality of the brother’s Support in a woman’s life, the situation described below being especially applicable to the northern parts of the subcontinent. Visits to the natal home often serve as women’s main or only respite from the drudgery of housework, childcare, and fieldwork, and represent periods of rest and personal freedom: {A]s each woman reaches her father’s house, she will uncover her face, shedding not

only the veil of purdah but also the burden of her daily worries ... A visit to her father’s house gives a woman her only chance of a holiday — there she can escape the

tensions of married life, the monotony of constant work and the demands of her children ... For a brief moment, she recovers her girlhood identity. (Hartmann and Boyce 1983: 92, on rural women in Bangladesh)

brief but useful overview of this case, and more generally of the problems of obtaining maintenance faced by divorced women, Hindu or Muslim, see Gandhi and Shah (1991: 237-42).

262 A field of one’s own These visits become particularly important where village exogamy, long distance marriages, and marriages with strangers are the norm, as they are

in much of northwest India. Women married into another village are treated with indulgence and as guests during their visits to their natal homes, and the behavioural strictures on daughters-in-law are relaxed for daughters.?> In villages in northwest India, it is generally easy to distinguish between daughters and daughters-in-law, since the former leave their faces uncovered on the streets and the latter usually cover their heads and faces

with the edge of the sari or dupatta.°° A father in Uttar Pradesh voices a commonly felt sentiment in the words: ‘O, she is with us for a while. Let her

play, sing and dance, enjoy life as she must, for she may have no rest or leisure in her husband’s place’ (Majumdar 1954: 87).37 This freedom takes a

most unusual form in the Jaunsar-Bawar region of Utter Pradesh: here women are free to take lovers during visits to their natal homes, but are expected to be strictly faithful in their marital villages (Majumdar 1955). Visits to the natal village usually provide women who have married into distant villages the only occasion when they can be with those with whom

they have the closest emotional ties. Particularly in the early years of marriage, a month-long stay every year is not uncommon.?? In some communities of both India and Pakistan, the birth of the first child also takes place in the woman’s parents’ home.*° Brothers play a significant role

in the maintenance of this link. Especially where long distances are involved, it is the brother who comes to escort a woman to and from her natal home.*° In many Hindu communities of northwest India, and among the Brahmins and Chetris in Nepal, parents and elder relatives are barred from visiting a woman in her marital home because accepting the hospitality of a son-in-law or of his family would go against the strict logic of kanyadan, ‘the gift of the virgin daughter’.*! Lewis (1958: 188-9) quotes Ibbetson’s 1883 report as follows: 35 For India, see Wadley (1976), Minturn and Hitchcock (1966), and Palriwala (1991); also personal observation in Rajasthan, Uttar Pradesh and Haryana. 36 This is graphically illustrated in an ethnographic film on a Haryana village, entitled “Dadi’s Family’, directed by Michael Camerini and Reena Gill (1982). Also see Wadley (1976). A dupatta is a long piece of cloth covering the bosom which can also be used as a veil. 37 Minturn and Hitchcock (1966: 29) similarly comment: ‘She is treated as a special guest with no household duties. She spends her time visiting friends, chatting, singing and playing games.’ Also see Ali (1982) for Pakistan. 38 For India, see Bradford (1985) and Wadley (1976).

39 For India, see Bradford (1985), Cantlie (1984), Harper (1971), Karve (1965), and Unninathan (1990); and for Pakistan, see Aschenbrenner (1967). 40 See Hershman (1981), Narain (1970), Unninathan (1990), and Wadley (1976). +! For India, see Hershman (1981) and Leaf (1974) for Punjab; Pocock (1972) for Gujarat;

Lewis (1958) for Delhi; Gupta (1974) for upper castes in Rajasthan; also personal observation in Rajasthan and Uttar Pradesh. For Nepal, see Bennett (1983). Trautmann (1981: 26) writes: ‘For the wife-givers to accept the smallest return would constitute taking visible “payment” for their daughter, destroying the invisible merit of the gift and making it no better than a commercial transaction.’

Whose share? Who claims? 263 The village into which his daughter is married is utterly tabooed for the father and her elder brothers and all near elder relations. They may not go to it, even drink water from a well in that village, for it is shameful to take anything from one’s daughter or her belongings.

Many Hindu families of northern India still adhere to these taboos strictly. Others avoid visits unless absolutely necessary and make a formal payment

for any items they use. Thus younger brothers may be a woman’s only regular visitors from her natal home. Women’s songs in northwest India poignantly capture the contradiction that underlies their relationships with their brothers: resentment at being disinherited by brothers on the one hand, and dependence on their support on the other: To my brother belong your green fields

O father, while I am banished afar... This year when the monsoon arrives, dear father, send my brother to fetch me home.

When my childhood companions return O my father . send me a message, O.*2

Even where village exogamy and long-distance marriages are not the rule, the brother occupies a place of fundamental importance. If the age difference is small, the relationship is likely to be one of companionship and

relative ease. Indeed it is usually the only male-female relationship in a woman’s life which can be so viewed (except possibly, but not necessarily,

that with her husband in later life). And after the parents’ deaths, a brother’s home becomes the woman’s natal home. He is held responsible

for arranging an unmarried sister’s wedding, and for maintaining the tradition of gift-giving to the married sister and her children when she visits or on ceremonial occasions.*3 In Bangladesh, in fact, the woman’s access to her brother’s home is seen as a right (naior), and several scholars argue that women exchange their inheritance in land for a continuation of this right:44 A sister, instead of claiming her inherited share from her father’s or mother’s property, enters into an informal agreement with her brother or brothers whereby

she gains the annual or semi-annual right to visit her family homestead. This agreement 1s not legally binding upon either party unless it is put in writing.

(Alamgir 1977: 15) :

42 A Hindi folksong sung by Hindu women in north India (my translation). A fragment of this

song was also reproduced at the start of chapter 1. 43 This is true in large parts of the subcontinent. For India, see especially Das (1976), Goody (1990), Karve (1965), and Wadley (1976). For some discussions in the context of Pakistan, see Elgar (1960) and C. Pastner (1971). 44 See Alamgir (1977), Ellickson (1972b), Jahan (1975), Sobhan (1978), and Tadahiko (1985).

264 A field of one’s own Deferring their claims, as White (1992: 131) notes, also allows women ‘to

keep some material stake in their natal family without souring relations there, and so mediate their dependence on their husbands’ families’. These

observations suggest the possibility of individual negotiation within custom, something not typically noted elsewhere in the subcontinent. Usually the fulfilment of customary obligations by the brother towards the

sister is part of a complex network of kinship and religious ties; the obligations may be subject to individual variation in the degree to which they are fulfilled, but they are typically not open to individual or explicit negotiation (of which more later). Even while the parents are alive, brothers are expected to play important

roles on various ritual occasions in women’s lives. Especially in the marriages of sisters’ daughters in north India, they bear responsibility fora special component of the gifts (the bhat) that the bride receives.** In south India, brothers may serve as authority figures for the sister’s children and they may even marry the sister’s daughter. (In the north, by contrast, most Hindu communities forbid uncle—niece marriages, and a brother’s relationship with the sister’s children is typically one of indulgence.) Ross’s (1961:

137) grading of the degree of emotional closeness in eleven types of relationships within Hindu joint families in Bangalore city (in Karnataka state), although based on an urban sample, is revealing: she found that the

mother-son and brother-sister relationships occupied the top two positions, and that between husband and wife came second to last.

Emotional and ritual ties apart, a brother is expected to provide economic and social support. Brothers (even younger ones), and natal kin in general, are seen as women’s potential protectors. In northern India and Nepal, this is ritualized in festivals such as raksha-bandhan (literally the tie

of protection) and symbolized by sisters tying a thread (rakhi) on the brother’s wrist. Thus Mayer’s (1960: 219) observation for central India would not be inaccurate for most communities in the subcontinent: A man’s tie with his sister is accounted very close. The two have grown up together, at an age when there is no distinction made between the sexes. And later, when the sister marries, the brother 1s seen as her main protector, for when her father has died to whom else can she turn if there is trouble in her conjugal household.

The parental home, and after the parents’ death the brother’s home, often offers the only possibility of temporary or longer-term support in case

of divorce, desertion, and even widowhood, especially but not only for a woman without adult sons. Her dependence on this support is directly 45 For India, see Miller and Archer (1985), Phylactou (1989), Reddy (1956), Vatuk (1975), and Wadley (1976).

Whose share? Who claims? 265 related to her economic and social vulnerability. Economically, limited access to personal property (especially in the form of productive assets), illiteracy, limited training in income-earning skills, restricted employment and other income-earning opportunities, and low wages for available work, can all constrain women’s access to earnings and potential for independent economic survival. This would be true, albeit in varying degree, for rural

women of most classes and communities in the subcontinent. Socially, women’s vulnerability is associated partly with the strength of the prevailing ideology of female seclusion (discussed further below), and partly with the extent to which a social stigma attaches to widowhood or divorce. Both

these factors vary in strength by community, region, and circumstance. Although a woman may be somewhat less vulnerable if she has adult sons, a brother’s home can be a social and physical refuge if she has none. Hence if women don’t have a natural brother they may induct one by tying a rakhi around a relative’s or family friend’s wrist, or in other ways. Elgar (1960: 188-9) recounts a story of two brotherless sisters, one married and the other widowed, who each inherited land from their father. The one still married, rather than allowing her husband to profit from her share, transferred it to

her widowed sister’s son, thereby placing on him ‘the responsibility of keeping up the tradition of the parental household’, in other words, of playing surrogate brother. Few women therefore wish to sour or break their relationships with their brothers; and Kabeer’s (1985: 88, 90) observation for Bangladesh has wider relevance, namely that many give up their claims for a promise of support

and protection in times of distress since: ‘[aJlthough such support 1s prescribed by religious and cultural norms it is more likely to be forthcoming if women renounce their rights to inheritance’. When Luschinsky (1963: 575) talked to women in a village in Uttar Pradesh (northwest India) soon after the 1956 Hindu Succession Act was passed, they said: Equal inheritance by brothers and sisters ... would have only one result. Brothers and sisters would quarrel. Brothers would want to obtain their sisters’ shares and the sisters might feel that they were not being adequately reimbursed. The close protective relationship of brother to sister would be in jeopardy.

Cultural constructions of gender, including the definition of how a ‘good’

sister would behave, also discourage women from claiming their rights. Women in Bangladesh told Westergaard (1983) that it would be ‘shameful’ to claim their shares. Albrecht (1974) found the same in Pakistan Punjab, as did Elgar (1960: 188): To ask their share of inheritance from their brothers would go entirely against the love which a woman traditionally cherishes for her brother and against the picture of a respected and much loved phuphi [father’s sister].

266 A field of one’s own

In an Indian Punjab village, a married sister who staked her claim and

threatened to pursue it in the courts had, in the community’s eyes, ‘performed so shameful an act that she might never dare show her face in her father’s village again’ (Hershman 1981: 74). Villagers in Kithoor village (Rajasthan) told me that even leasing or selling the land to a brother was not

desirable, since it would introduce an element of commerce into the relationship. What we therefore see in the sister—brother relationship is an idealized and complex construction of roles and expectations — ceremonially ritualized, culturally elaborated, economically necessitated, and ideologically reinforced.

How does the idealized image of this relationship match reality? In particular, to what extent do brothers meet women’s expectations of practical help in times of need? The evidence is mixed. Brothers generally appear to fulfil their ceremonial roles, such as giving marriage gifts at the weddings of sisters’ children. Stories about their reliability in an economic crisis are less consistent. Some indicate that brothers are often the only ones willing to help. Bailey (1957: 83) describes two cases from his Orissa village study. In one, involving a widow with two daughters and a small son, he found that: Atevery crisis in the life of that family, the widow’s father and then her brother have come forward to help. They financed the weddings of the two girls. When the boy grew up they helped to find him a wife — from their own village — and they helped build a new big house. The man is now grown with children of his own, but his mother’s people still come to help him and he is a frequent visitor at their house.

In the other case, when the house of a widow with small children needed roofing, her brother and his friend came from another village to assist her, while her husband’s agnatic kin provided no help whatsoever. Several other studies cite examples of brothers coming to the rescue of sisters, especially those widowed (Wadley 1976; Burkhart 1976). In south India, women’s natal families are often significant sources of support during periods of economic crisis such as a drought, when families might provide interest-free loans (Caldwell et al. 1988). The extent and nature of assistance offered by a brother may also depend on the distance of the sister’s conjugal home from her parental village and on how cordially he is received by her husband’s

family, both of which influence the frequency of contact between the siblings. Greater support is possible if the woman lives close by (Mayer

1960). , However, instances of neglect and duplicity by brothers also abound, as

recounted later in this chapter and in chapter 7. Moreover a sister whose marriage has dissolved, or who has been widowed, may be welcome in her

Whose share? Who claims? 267 brother’s home for a short while, but usually not for extended periods. Her presence is less problematic among communities where women contribute significantly to cultivation, as among the Tibeto-Burman communities of Nepal: Limbu, Gurung, and Tamang women spend long stretches of time in their natal homes, usually contributing labour in the parents’ fields, even after marriage.*° A woman is also less unwelcome among communities which commonly practise widow and divorcee remarriage and in circumstances where such remarriage is more likely (e.g. if the woman is young and childless).47 Among the Newars of Nepal, the easy remarriage of divorced and widowed women is one of the significant factors accounting for ‘the total and unquestioning support at all times and under all circumstances of the natal home, so that a woman can retreat there at anytime for as long as

she wishes without pressure being put on her to return to her husband’ (Pradhan 1981: 75). The same is true of the Tibeto-Burman communities of Nepal’s middle-hills.*®

But where status considerations and female seclusion practices preclude women’s active labour contribution, and/or where widowhood and divorce carry a strong social stigma (as among many upper-caste communities of northern India), the returning sister is viewed as an economic burden. Her presence is also a source of potential conflict with the sister-in-law over household management and decision-making. Hence, for instance, among the Brahmins and Chetris of Nepal: ‘while short visits ... are a cherished

delight ... prolonged or permanent stays can place severe strains on the filiafocal relationship’ (Bennett 1983: 245). Elgar (1960) observes the same for Muslims in Pakistan Punjab, as does Ali (1982) for the Burushos. Ali (1982: 252) says: ‘Although her brothers are legally and morally obliged to take her in should she be forced to leave her husband’s home, she is often made to feel unwelcome and pressured to leave as soon as possible.’ At the

same time, women there ‘simply lack the wherewithal and the personal autonomy with which to activate and reciprocate external support relationships in their own right’. Hence it is precisely in communities where women are economically and socially most vulnerable, and most in need of substantial natal support if

their marriages fail, that such support appears least likely to be readily forthcoming for any extended period. Such help may of course be provided grudgingly. In Dreze’s (1990: 116A) three study villages, 10.5 per cent of all widows (including several with adult sons) were living alone. In her ongoing

study on widows, mentioned earlier, Chen found that out of a 46 See Andors (1976), Jones (1977), and March (1987). *” On widow remarriage, see Dreze (1990), Dubey (1965), Harper (1971), and chapter 8. +8 See Holmberg (1989) on the Tamangs, Jones (1973) and Jones (1977) on the Limbus, Messerschmidt (1976) on the Gurungs, and Vinding (1979) on the Thakali.

268 A field of one’s own sample of 262 widows in four northern states of India, 40 per cent were living in households where their adult sons were household heads, 47 per cent in households which they themselves headed (most but not all of these women were sonless), and 13 per cent in households headed by ‘other’ relatives (presumably including brothers); very few thus used a brother’s home as refuge (see, Chen and Dreze 1992: WS-87). In Bangladesh, Begum and Greeley (1979) found that out of a sample of twenty-three women wage

labourers who were widowed, divorced, or separated and virtually assetless, six were living with brothers; in all six cases the brothers’ wives did not go out to earn a living. In other words, the sisters alone were compelled to

break the norms of seclusion and suffer the associated decline in social status. As Bangladeshi women said to Hartmann and Boyce (1983: 92): ‘A brother’s love is not as strong as a father’s.’ For the assurance of receiving it, women need to relinquish what is legally theirs, namely their inheritance. Weighted against these multiple considerations is the fact that a share in

the parents’ land is itself a crucial means of economic security and of significantly reducing the risk of destitution for women in poor households.

Also (as noted in chapter 2), widows with some land tend to be better looked after by sons or other relatives than those who have nothing to increase their bargaining power. To claim or not to claim must thus pose a painful dilemma.

(2) The necessity of male mediation Impinging on this dilemma are other practical considerations. Both economically and socially, rural women’s relationship with the world outside the family is typically mediated through male relatives: fathers, brothers, husbands, and male extended kin. This especially affects women’s

access to economic institutions (markets, banks, and so on), and judicial and administrative bodies. Such mediation can be necessitated by a variety of factors (the nature and strength of which vary regionally), but particularly by the physical and social restrictions on women’s mobility and

behaviour. In many South Asian communities, these restrictions are explicit and relate to the norms and ideology of purdah or female seclusion.

In many others, they are implicit and subtle, but nevertheless effectively restrict women. The complex manifestations and implications of these gender restrictions

will be discussed in detail in the next chapter, but a few points need mentioning here. To begin with, purdah (literally meaning curtain) as an ideology is manifest not just in the veiling of women, but also in the gender segregation of space and the gendered specification of behaviour. In fact,

Whose share? Who claims? 269 veiling is confined to a limited number of communities and regions — it is stronger among Muslims in northern South Asia, and among upper-caste Hindus in northwest India, than elsewhere. And even within these regions and communities, it varies in form (whether a woman wears a burga*? or merely veils her face), in extent (e.g. which men are to be avoided), and by caste, class, and age. But much more widespread than the practice of veiling is the gender division of space, especially outside the home. Certain parts of the village — those where men congregate, such as the market place — are

spaces which women are told to avoid or to use minimally. Even more pervasive and pernicious are the behavioural strictures imposed upon and

internalized by women from late childhood, which define whom they should speak to and in what manner, how they should dress, move, and so on. Although such gendering of space and behaviour is strongest and most apparent in communities which explicitly endorse purdah, in their more subtle manifestations they constitute an implicit code of expected female

behaviour in large parts of the subcontinent, even where purdah is not endorsed. Effectively, the ideology of purdah, and more generally the implicit gendering of space and behaviour, restrict women’s interaction with unrelated men and institutions, their physical and social mobility, their domain of activity and knowledge, their access to education, and so on. This has serious consequences for women’s ability to claim land and (as will be elaborated in the next chapter) to control it as well. First, it means that a brother’s potential support is important not just

economically but also socially. Brothers can be critically important as social mediators not only in the absence of adult sons and husbands, but also if women wish to undertake transactions independent of their marital homes. Second, these restrictions affect women’s access to information on the laws, a crucial prerequisite for exercising legal rights. Over half the women interviewed by Luschinsky (1963) in an Uttar Pradesh village in the early 1960s knew nothing about the Hindu Succession Act of 1956. In Panjgur (NWFP of Pakistan), many women to whom C. Pastner (1971) spoke disclaimed any knowledge of their legal property rights. Third, given that most officials in judicial and administrative institutions are men, the noted restrictions can make male mediation imperative in a land dispute. A widow’s dispute against her husband’s kin could need the mediation of her brother or father; a dispute against a brother could require the woman’s husband or an adult son in a mediatory role.°° A woman’s 49 An overgarment that covers all parts of the woman’s body and dress, other than her eyes and hands. Sometimes even her eyes are covered by a net-like veil. 50 Alamgir (1977: 17) notes in the Bangladesh context: ‘A male representative is almost an absolute necessity for any woman who wishes to seek a divorce or any widow who wishes to claim her share of her husband’s estate. Men often complete most of the formalities as well as help to pay or give a loan to pay for the legal expenses involved.’

270 A field of one’s own ability to draw upon male support to mediate in her dealings with judicial and administrative institutions and procedures would thus be a factor in her decision to claim land. For the same reason, the odds against her being able to keep control over the land, if she eventually gets any, are high, and this too would be a deterrent: even for self-managing land, women often need male support such as that of brothers, sons, or other family members. The need for male mediation in many spheres of women’s lives thus

circumscribes them in complex ways, reducing their ability to act as independent agents in relation to their legal rights. At the same time, this very dependency on male relations also leaves a woman especially vulnerable to being duped by them when they are the interested parties. Manipulating a woman’s statement or claiming that she has given up her rights is made especially easy where norms of female

seclusion are strong. As C. Pastner (1971: 162-3) notes, referring to Baluchistan (Pakistan): Since women are represented by their male kin in most dealings with the non-kin sector of society, particularly the governmental sector, women ordinarily do not have the opportunity to speak for themselves and make viable claims when they are aware of them.

Also daughters, if they are quite young at the time of their fathers’ deaths, would have to be informed of their rights by male kin. Stuers’ (1968: 49-50) comment is even more graphic: A pardanashin [woman under purdah], for whom all contact with the outside world had to be through a masculine intermediary, depended entirely on the integrity of the one who transacted business in her name. ‘The purdah’, wrote two jurists, judges by profession, ‘... exposes women to fraud, deceit, and undue influence... it makes

women incapable of transacting business.’ Even under the most favourable circumstances, when a Muslim woman was aware of her rights, how could she obtain the advice of men who were expert as well as honest when she could consult in

person only her nearest relatives? For handling her affairs, the custom was for the woman in parda to sign blank documents... which the agents... named by her then completed at their discretion. The pardanashin client often had confidence in the integrity of the agents solely on the recommendation of a relative. There were numerous cases in which agents and relatives conspired for their own personal interest and benefit by manipulating or forging these documents. Deliberately or

not, the pardanashin was never told that the Qur’anic law recognized her full authority over her personal possessions. Thus she was generally treated as a minor incapable of managing her own property.

In 1976 the Pakistan Women’s Rights Committee thus recommended that in order to protect female heirs, any statement where a woman relinquishes her share should be attested by a civil judge (Patel 1979).

Whose share? Who claims? 271 (3) Hostility from male kin: pre-emptive steps to direct violence From the very beginning, progressive legislation in the 1950s, which gave. South Asian women rights to inherit land, was not viewed with favour by men in patrilineal communities, either Hindu or Muslim. Several ethnographers undertaking village studies in the early 1960s, soon after the passing

of such laws, commented on this. In Pakistan Punjab, for instance, the initial response of land-owning villagers to the Shariat law was one of ‘uneasiness’, and ‘by new means they attempted to maintain the old equilibrium in which men were the holders of productive property and women shared in it through gifts which were their right ...’ (Elgar 1960: 189). Across India, traditionally patrilineal Hindu communities reacted similarly to gender-progressive legislation giving Hindu women rights of inheritance and divorce. The residents of Shivapur village in Mysore (south India) viewed such

laws as ‘a deliberate and sinister attempt to destroy the family and morality’, and concluded that: this equality must have the inevitable consequences of increasing divorce, desertion, adultery, destroying the love between husband and wife, depriving children of the certainty of a normal home life, and setting brother against brother, son against father, and man against man; that it [would] in a word, atomize society by gnawing at the foundations of the social bonds. (Ishwaran 1968: 183)

In Ramkheri village (central India), Mayer (1960: 242) observed: ‘the villagers assert that the present rules are new and view them with concern,

saying that the increased power of the daughter and widow mean more chance for the land to leave agnatic hands’. In a village in Himachal Pradesh (northwest India), the inheritance law ‘struck the valley as so unfair that they petitioned the government not to introduce the law, but without avail’ (Newell 1970: 51). Every single household surveyed in Jhatikra village near Delhi, after the 1956 Hindu Succession Act was passed, disapproved of its provisions allowing daughters to inherit the patrimony (Freed and Freed 1976: 197). In the Konku region of Tamil Nadu (south India), Beck (1972: 295) observed: ‘Everyone has heard rumours that both daughters and adopted children can now take their claims to court, although few have actually tried to. My informants still thought of these laws as just a new way “to get around” the old tradition.’ The Rajputs of Khalapur (in Uttar Pradesh, northwest India) considered even inheritance by a brotherless daughter with an uxorilocally resident husband, unacceptable: Regardless of its legality, such a situation is a very serious breach of village customary law, which has always held that no wife, daughter, or daughter’s

272 A field of one’s own husband could inherit land. This rule was a very important one and still is adhered to with deep emotion. It was a means of excluding persons who were not members of the village land-owning caste group, and in large part it accounts for the fact that this group has been able to maintain almost complete control of the Khalapur land through the vicissitudes of the past 400 years. (Minturn and Hitchcock 1966: 28)

The emotional strength of these responses suggests that male heirs were unlikely to relinquish their privileges easily. Not surprisingly, male relatives of potential female heirs began to take pre-emptive steps to circumvent the rights of women, especially of daughters and sisters. For instance, to ensure

that only sons would inherit, fathers began to leave wills explicitly disinheriting daughters,>! or to make a pre-mortem legal transfer of their land in favour of sons.** Elgar (1960: 187) describes a typical case where a father in Pakistan Punjab transferred his property to his sons in his lifetime: Though it was an expensive procedure, he preferred to undertake the expense rather than expose his sons to whatever inconveniences might be caused by his daughters or sons-in-law. He died in 1954, and, because two of his sons-in-law who were brothers had been troublesome all along and were not on good terms with their family-in-law, everybody now praised him for his wisdom and foresight [!].

Parents in Pakistan Punjab also became more inclined to give their daughter in marriage in exchange for a bride for their son from the same family, even though in social terms, such ‘exchange’ marriages were not highly regarded. The parents felt that in such cases neither daughter would claim her inheritance, especially where the two families were of equal economic status and the daughters’ shares would have been about equal (Elgar 1960: 188).

Over the years, sons (or other prospective heirs) have been known to forge wills not only after a father’s death (Parry 1979), but even during his lifetime: In Bhaimara village, Jansen (1983:69) heard ‘several stories of how thumbprints of the old father on a piece of paper were taken while he was asleep at night, as a signature of the will his sons had made, unknowingly, on his behalf’. The rare fathers who may seek to give land to daughters are likely to encounter the hostility of sons (personal observation, Rajasthan). Brothers sometimes appeal to revenue authorities who maintain the land registers, arguing that their sister is wealthy and does not need the land, or that she is an absentee landlord as she is living with her husband in another village (Mayer 1960). This last can become an important way of preventing women from claiming land where village exogamy is a rule and government

policy opposes absentee farming. Brothers have also been known to 51 See Parry (1979) for Madhya Pradesh (central India) and Elgar (1960) for Pakistan Punjab. 52 See Elgar (1960), C. Pastner (1971) and Patel (1979) for Pakistan, and Aziz (1979) for Bangladesh.

Whose share? Who claims? 273 persuade or coerce a sister into signing a deed gifting her share of the land to

them. Sometimes such subterfuges are later challenged in court (Alavi 1972). Male relatives have also been found to suppress information about the existence of a female heir, or (as noted earlier) to manipulate statements before the revenue authorities to make it appear that she has relinquished her right, or to compel the woman to make such a statement (Patel 1979). In some cases women claimed to have sold their shares to brothers or to their father’s male relatives for cash, but it is unclear whether this was indeed so or they were only saying so to avoid family conflict (C. Pastner 1971; Elgar 1960).

Natal kin are particularly hostile to the idea of daughters and sisters inheriting land, since the property can pass outside the patrilineal descent group. Strenuous efforts may thus be made by a woman’s brothers and her father’s collaterals to prevent this. A widow’s claims (as noted earlier) are often viewed with less antagonism than those of a daughter, since with a widow there is a greater chance of the land remaining with agnates: she can be persuaded to adopt the son of the deceased husband’s brother (whose potential hostility is thereby neutralized), made to forfeit the property if she remarries outside the family, or persuaded or forced into a leviratic union with the husband’s (usually younger)*? brother.*+ Levirate appears to be most easily accepted when the widow is young and childless or has only one child and the brother-in-law is unmarried, but cases of unwilling widows with several children being forced to cohabit with married brothers-in-law 53. Marrying the older brother is often forbidden. 54 In Pakistan, Lindholm (1982) notes that levirate is usually obligatory among the Swat Pukhtuns of the NWFP if the widow has no sons, while a widow with sons cannot remarry. Among the tribal Pukhtuns, Ahmed (1986: 295) finds that even a woman with a son may marry the husband’s brother if her son is ‘not old enough to carry a gun’. Among the Baluch, levirate was once mandatory, although there is greater flexibility today (C. Pastner 1971); while among the Kohistanis (non-Pakhtuns) in Swat, refusing a leviratic union and marrying someone else is classified as adultery, for which blood revenge is sanctioned by custom (Barth 1956). In Punjab province, however, although junior levirate is practised, it is neither mandatory nor common (see e.g. Elgar 1960, and Aschenbrenner 1967). Again, in India, the practice is found among many communities, especially in northern India: see table A8.3c in chapter 8. Some actively encourage it. Among the Garwalis levirate is the rule (Berreman 1962). It has also been a long-standing practice among the Punjabi Jats, as noted in chapter 5. A number of other communities customarily favour the practice as well: see Das (1979), Freed and Freed (1976), Haekal (1963), Hu (1957), MacDorman (1987), Mathur (1967), Murray (1984), Pettigrew (1975), and Sarkar (1988).

However, in south India, the practice is rare, and communities such as the Coorgs of Karnataka, who were noted by Srinivas (1965) to have a strong preference for leviratic unions, are exceptional. In fact, Karve (1965: 224) argues that levirate is normally taboo in

south India. Nepal is closer to the north Indian pattern and several communities here commonly practise levirate, although, unlike in parts of Pakistan and India, there appears to be no suggestion of women being forced into it. Leviratic unions are also found among Muslims in Bangladesh, although not commonly: of the fourteen cases noted by Karim (1988: 145) in two villages, twelve were among rich and middle farmers.

274 A field of one’s own

who then take over their land are not unknown.** A case in point is a Punjabi Jat widow I spoke to in Kithoor village (Rajasthan, northwest India) with five minor children (one son and four daughters) who inherited

3.2 acres from her husband, and who was strongly pressured by her husband’s younger brother (already married, but with no sons) to marry him. But when a daughter was born from this alliance, he abandoned her, enticed away her fourteen-year-old son (his nephew) who now lives with him, and through forgery got the widow’s land transferred to the boy’s name, thereby gaining effective control over it. He now gives her a part of the wheat grown on her land, but not any part of the crops grown for cash, leaving her to fend ineffectively for herself and her daughters. I found her in a state of extreme depression: she said it was only the thought of her minor daughters being left destitute that kept her from suicide. In communities where women have never before been given a share in

land, an attempt by any one member to do so voluntarily can meet with strong social disapproval. Ahmed (1980: 296-7) describes such an exceptional case, involving Pukhtuns from the non-tribal areas of the NWFP, which came for consideration to the Council of Pukhtun elders in 1977, that is, well after the passing of the NWFP Shariat Act of 1935 and the West Pakistan Shariat Act of 1962. In this instance, when A’s brother B died, A nominally gave plots of land to B’s widow and daughter (who was also his daughter-in-law), although de facto possession remained with A’s sons. On A’s death, a formal transfer of the women’s shares was strongly opposed by A’s relatives who stood to gain otherwise. The Council granted a share to B’s widow but not to his daughter. When the latter’s husband (A’s son) challenged this, the Council granted B’s daughter land as well, since she was

married to A’s son (her parallel cousin) and the land would therefore remain in the family. Nevertheless, most villagers felt that this decision had ‘stamped on riwaj [custom]’. Clearly there was a fear that this would set an undesirable precedent. Where pre-emptive tactics are not successful, and daughters and sisters also do not voluntarily give up their rights, male kin may try various forms of intimidation. A common tactic is to initiate expensive litigation which few women can financially afford (Kishwar 1987). Some women drop their claims as a result; others press on, with the risk of having to mortgage the land to pay the legal expenses, thus losing it altogether. Land disputes are found to be increasing in parts of the subcontinent, and they usually centre

around male attempts to prevent sisters or daughters from inheriting.°° 55 Even in the late nineteenth century, as noted in chapter 5, a number of communities in northwest India favoured levirate, especially in order to gain control over land which the widow inherited as a limited estate in the absence of sons. At that time too, many widows objected to the practice, and petitioned against it to the British, usually in vain. 56 See Mayer (1960) for Madhya Pradesh (India).

Whose share? Who claims? 275 Threats to kill those who still insist on exercising their rights are often made. Single women (unmarried or widowed) are particularly vulnerable to such

harassment. Direct violence is also being used increasingly to prevent women from filing claims or exercising their customary rights: beatings are

common and murder not unknown. Indeed in eastern and central India, murder, following accusations of witchcraft, is on the rise.>” The erosion of women’s customary rights and the increasing incidence of land-related ‘witch’ killings is particularly apparent among a number of

tribal communities in Bihar (eastern India), such as the Santal, Ho and Munda. According to the writings of W.G. Archer, a British officer who was an administrator in Bihar for sixteen years (1931-46), a Santal widow by custom had a right to maintenance from her deceased husband’s land. This had earlier been interpreted to mean that a widow with sons had a limited interest in the entire estate of the husband, which she supervised and managed ‘exactly as if she were their father’ if the sons were minors; and if they were adults she was still recognized as the head of the household. Also, if her major sons had separated from the joint estate, she kept her husband’s share as a limited interest. Even when she had only daughters, as long as she lived in the village and did not remarry, she inherited a limited interest in all of her deceased husband’s land. She could also adopt a son or bring in an uxorilocal son-in-law (Archer 1984). However, it appears that over time this right of maintenance began to be interpreted within the community asa night to only a plot of land sufficient to maintain the widow, and not to the

entire estate; and now, increasingly, it is being interpreted as a right to subsistence provided by the male heirs of the husband, rather than asa right to manage land through which the widow could directly maintain herself (Kelkar and Nathan 1991). Recent evidence provided by Chaudhuri (1987) and Kelkar and Nathan (1991) for the Santal, by Kishwar (1987) for the Ho, and by Standing (1987) for the Munda, suggests that the incidence of witch killing (whatever its traditional roots) has today become a means of preventing women in these communities from exercising their customary claims in land. For instance,

Chaudhuri’s compilation of police records relating to Malda district in West Bengal shows that over the three decades since 1950, forty-two out of the forty-six witch killings were of women. Of the twelve victims in 1982, ten were women. All the fifty-two victims rescued by the police since 1972 were women, most of them widowed and elderly, lacking ‘protection or coverage from powerful relations’ (Chaudhuri 1987: 160,156). The women accused

of being witches typically belong to the same tribe and often the same lineage as the accusers. Many of the accusers are close relatives who stand 57 See Kelkar and Nathan (1991), Minturn and Hitchcock (1966), Sachchidananda (1968), and Standing (1987).

276 A field of one’s own to gain materially from the women’s deaths.°® If the woman accused of

witchcraft is driven out of the village, she can usually find a job as a domestic servant in a nearby village, suggesting that the real reason for the accusation was not a fear of her evil influence, as popularly claimed, but an intent to deprive her of her usufructory rights in land (Kelkar and Nathan 1991).°°

Kishwar (1987) comes to the same conclusion in her study among the Ho tribals of Bihar. She describes a number of cases in some detail, including the following one: [A] man who worked as a veterinary doctor in a government hospital was accused of having murdered his two paternal aunts. The two old women were unmarried and lived together, working the land over which they had a usufructory right. He had accused them of being witches and of having caused the death of his wife. A child had seen [the doctor] murdering them with an axe but he was acquitted by the court.

He inherited the land which the two old women had been cultivating. (Kishwar 1987: 101)

The concerns here appear to be not merely economic but also ideological,

involving male fears of how gender relations might be altered if women have land.°° For instance, a Santal myth traces the origins of witch killing to a growing male concern in times long ago that women were no longer obeying them or recognizing their authority. The men approached Maran

Baru (the great spirit of the Santal tribe), who agreed to teach them witchcraft the following day to help subdue the women. But the women impersonated their husbands and learnt the craft instead. On learning of the deception, Maran Baru in recompense taught the men how to witchfind (Archer 1974: 292-3). Witchkilling thus became a means of controlling the mysterious powers of women and preserving male supremacy.

These are but the more extreme examples emanating from a fairly widespread climate of hostility and opposition to the idea of women inheriting land.

(4) Responses of village bodies and government officials

Apart from the attitudes of kin, a significant determinant of women’s ability to exercise their legal rights is the male bias in administrative and judicial bodies and processes. Traditionally, at the village level, both 58 Evidence from several other studies on witch-killings among the Santal, cited in Kelkar and Nathan (1991), supports this view as well. 59 Also see Sachchidananda’s (1968) observations on the practice of witch-hunting among the Oraon tribe of Bihar. He too notes that if the person identified as a witch is driven away, her land is confiscated by the village panchayat.

69 On this, also see Kishwar (1987), and Kelkar and Nathan (1991).

Whose share? Who claims? 277 legislative and judicial functions were served by local councils which took a variety of forms: caste panchayats (usually consisting of the prominent men

of the caste) in much of India and Nepal, tribal councils among tribal communities tn the subcontinent, village samaj (community) groupings and

the salish or village court in Bangladesh, and so on.°! Although these bodies differed somewhat in their membership composition and the level at which they operated (caste/tribe, single village, or multi-village), a common feature was their exclusion of women.°* Women had little say either in framing the rules made by these councils or in the process by which these rules were enforced. Nor did they have much control over the ideological underpinnings of such rules and their implementation. (As noted earlier, even in matrilineal communities jural authority rested with men.) Basically this exclusion meant that disputes which involved women were settled by

male authority and male-made rules. Only rarely, as among the Santal tribals or the Bhats (who were Muslim entertainers) of Uttar Pradesh, were women even allowed to attend tribal council meetings.°? Under colonial rule, with the setting up of British legislative and judicial

machinery, the role of traditional councils was eroded in some degree: people could now take recourse to higher level courts if their disputes were not satisfactorily resolved by local bodies such as the village courts. But it was mainly the rich and powerful who were in a position to go to the higher courts (which too were male dominated); for the majority, the traditional institutions still continued to be the instruments of justice. In the post-colonial period, there have been attempts to democratize the

system in the subcontinent.°* For instance, in India there have been specific | attempts to increase female representation in local level bodies. Here, in the

1950s, a Panchayati Raj structure was instituted, consisting of district, block and village level bodies for local self-governance (termed respectively, zilla parishad, panchayat samiti, and gram panchayat), with elected functionaries.°* These could be either men or women belonging to any caste or religious group. The idea that all castes and both sexes should have a voice in governance went against existing custom, and was a significant ®t See e.g. Jahangir (1979) for Bangladesh; and Minturn and Hitchcock (1966), and Cohn (1965) for India. °2 For Bangladesh, see Arens and Van Beurden (1977), and Hoque (1987). For India, see Bailey (1957), Mathur (1964), Newell (1962), Per-Lee (1981), and Luschinsky (1962). And for Pakistan, see Ahmed (1986). 63 See Archer (1984) and Sachchidananda (1968) on the Santal and Luschinsky (1962) on the 04 See McCarthy and Feldman (1987) for a useful discussion on Bangladesh’s attempts to democratize village bodies, and Frankel (1978) for India. °5 There are some exceptions, such as the northeastern states with primarily tribal populations where traditional councils, constituted by tribal leaders rather than elected individuals, typically continue to function (see GOI 1983b).

278 A field of one’s own

step forward. A few states also initiated nyaya panchayats (judicial councils), separating judicial powers from the executive powers vested in the gram panchayats. These were established to settle village disputes with jurisdiction over groups of villages, and their members were to be chosen by the gram panchayat either all through election or some through nomination and others through election. In some states, such as Karnataka, at least one

member of the nyaya panchayat has to be a woman.°® Moreover, the Seventy-third Amendment to the Constitution of India, which came into force in 1993, provides that one-third of the seats in Panchayati Raj institutions be reserved for women. In fact, for several years now, most states have reserved some seats for women, with places for women not filled by election to be filled by nomination. The potential advantages to women of these institutional changes are discussed in chapter 10. Much will depend

on whether the women who come to occupy these seats are genderprogressive and able to focus on women’s concerns. So far, however, in most parts of the country these bodies continue to be dominated by the economically and socially powerful men in the village, and elected women representatives, although increasing in number, are still the exception.®’

The ideology of female seclusion and more generally the cultural construction of gender roles also continue to restrict women’s attendance in panchayat meetings, although there is no jural bar to their attending. As an illustration, consider the following conversation between the anthropologist Erin Moore and a widow in the Alwar district of Rajasthan in 1988:6°

Q: Why not? | Q: Can you also, as a woman, call a panchayat?

A: No. ,

A: The men don’t call the women. Being an old woman, I can talk in the panchayat, but the women who are married can’t go to the panchayat. The young wives can’t go... Q: Why can’t the wives talk in the panchayat? A: The panch think it is bad... Q: How much land does your son have? A: Twenty bighas.°? 66 In practice, nyaya panchayats have not done well in most states, and some states have abolished them (Galanter and Baxi 1989). 67 Also see Arens and Van Beurden (1977), Gardner (1990), and Hoque (1987) on the continued domination of the samaj organization and salish bench by rich and powerful

village men in Bangladesh. 68 Personal communication from Erin Moore, Michigan (USA), 1989. 69 There is no uniform conversion rate of bighas into acres. Different regions of India and of the subcontinent typically use one of the following two rates: 1 bigha = 0.2 acres or 0.33 acres.

Whose share? Who claims? 279 Q: When your husband died, did half the land go into your name and half in your son’s name?

A: No, all in the son’s name... Q: Can’t the panchayat help you? A: There is no panch who can help me. Q: Have you gone and asked any of them?

A: We have said it many times, but no one helps us. They don’t say anything... Q: If your son gave you five bighas, then you could get the crop. A: They don’t give it, don’t give it. Gram panchayat rulings in northwest India have been observed to favour

the view that family property should be inherited by sons and not daughters,’° or that a woman must Stay in the village if she is to inherit her

husband’s land (Standing 1987).7! The views of the gram panchayat secretary of Kithoor village in Rajasthan, who was somewhat more progressive than most villagers I met there, are indicative: he said to me that he usually pressured daughters to sign away their shares in favour of their brothers, but sought to persuade widows to keep their shares. More generally, existing evidence from northern India suggests that the

patwari (village land records official), who maintains the land record register and plays a critical role in determining whose claim gets formally registered, usually tends to favour custom (which gives priority to male heirs) over existing law. The same 1s usually true of the tehsildar who oversees the work of several patwaris within the administrative unit of the

tehsil.’ Also revealing are government officials’ responses to recent attempts by

women’s groups to acquire land rights in parts of northern India. In the Bodhgaya peasant movement of landless labourers and sharecroppers 70 Personal communication by villagers during my fieldwork in Rajasthan in 1987. 71 Also see Furer-Haimendorf (1985), who notes that in tribal communities sometimes the village panchayat has no clear guidance on these matters in terms of customary law, since two generations ago land had been plentiful and was not considered to be a vital part of a man’s estate. 72 Record keeping on land rights is typically poor. Nandwana and Nandwana (1992), for instance, observed in two villages of Rajasthan that the patwari usually entered only the name of the eldest son, even if there was more than one successor; sometimes entries were

made 5-10 years after the death of the household head. Although in some parts of Rajasthan (as noted earlier), widows’ claims are now being better recorded, this is clearly not the case everywhere in Rajasthan, or elsewhere. See, e.g. Wadhwa (1989) for a general comment, based on a country-wide survey, on the dismal state of agricultural land records in India. He emphasizes the need for a systematic registration of titles to land. As he notes,

with the increasing pressure of population on land, the importance of formal titles will increase. He does not touch on gender, but clearly, in this context, ensuring that the land to which women are entitled gets registered in their names becomes especially important.

280 A field of one’s own (mentioned in chapter 1),’3 the women’s struggle was not only (along with the men of their community) against the religious institution (the Math) which held illegal possession of the land these households cultivated; it was also against the prejudices of men in their own community and in the local

government administration, when the women wanted land in their own names. Through an extended struggle, in which women played a crucial role, the peasant households were able to establish their rights over the land held by the Math. In some villages the women were also able to prevail upon the men of their community to allow the land so gained to be registered in the women’s names. But the district officer initially refused to do so on the

ground that titles could only be granted to men, since they were the household heads (Manimala 1983). In the end, many women did get land

titles, but only after considerable contestation. Again, when landless women in Udaipur district (Rajasthan) claimed a part of their village wastelands for growing herbs, fodder, etc., the bias of the local official was clear: ‘But we do not allot to women.’ When asked why not, he said with unbeatable logic: ‘Because we never have ... so that is why we won’t’ (Lal 1986).

Similarly illustrative is the case of a patwariin Pakistan Punjab who tried to allot a widow her inherited share in four separate parcels scattered across

the village. It was only when she paid a fee, and some of her husband’s associates threatened violence on her behalf, that the patwari agreed to allot her a consolidated plot; but even after a year of this agreement she had not

been given formal possession of the land (D. Merry 1983: 715). Other evidence from Pakistan (noted earlier in this chapter) suggests that in parts of the country the registration of a daughters’ claims, and more generally of women’s claims, is no longer uncommon, but the actual transfer of the land to women remains rare. Official actions both reflect and reinforce traditional attitudes. Prevailing biases affect the formulation and implementation of government policies, including land reform programmes. As noted in chapter 1, even in the late 1970s and early 1980s when the government of West Bengal (east India), in

an important land reform initiative (Operation Barga), undertook the registration of tenants, primarily men (rather than, say, both spouses) were registered.’* Ironically, in the process, widows who owned small amounts of land which they were sharecropping out to male tenants would have lost 73 See chapter 9 for a detailed account of the movement. 74 The very few women who were registered were typically widows in households without adult males, who had been able to continue leasing the land their deceased husbands had sharecropped (personal communication in 1993 from Nipen Bandyopadhyaya, who evaluated the programme in 1985).

Whose share? Who claims? 281 their land to the sharecroppers, a possibility which Dasgupta (1984: A-90), who played a significant role in the implementation of Operation Barga, saw as unimportant: ‘ [T]he number of such widows left alone without any adult male relatives looking after them cannot be very large.’ This view

unquestioningly endorses women’s dependency on male relatives, and assumes that widows without independent sources of income will be well-

treated by those relatives.’ Also illustrative of the general official attitude to women’s land rights is my own experience with the bureaucracy. An invited presentation by me on the question of gender and land rights at the Indian Planning Commission

in June 1989, to an almost all-male gathering of high-level government officials, elicited the following response (also cited in chapter 2) from the then Minister of Agriculture, who came from northwest India: ‘Are you suggesting that women should be given rights in land? What do women want? To break up the family?’ Indeed the issue of women’s rights in immovable property hits at the very fundamentals of class and gender relations in most Third World societies. Not surprisingly, the resistance to it is strong. These official attitudes also impinge on matters of dispute settlement,

including court judgements. One commentator argues that if a female litigant in a land dispute in Bangladesh is not closely identified with and supported by a man, she will probably lose, regardless of the merits of her case (Cain et al. 1979). Gender biases may be reinforced by class and caste biases. Poignantly illustrative of how the procedures of British courts in colonial India could prevent tribal women from obtaining justice is a case of two Santal women cited by Archer (1984). The women gave the following reason for failing to file appearance in a land inheritance suit in which they were both potential heirs: ‘We went to court. We had no money. We saw the other party with his Diku pleaders. We did not like it. We went back home’ (Archer 1984: 678). On this Archer comments (1984: 678): The two women saw stretching before them a hearing conducted by pleaders who knew neither their language nor their law, before a judge who did not know Santali in a court room, the very antithesis of a Santal village. They knew that D was wealthy and that if he failed at Dumka he could go to Patna. They themselves had never left the Santal Parganas. They saw demand after demand for lawyer’s fees. They saw no end to the hearing. They took one look. They did not like it. They went home ... No one, I think, will blame them but no one, I think, will call this justice. 78 Recent research in Bangladesh on mortality rates among widows living in different household arrangements in fact shows that those living as dependents of male relatives, other than adult sons, are at significantly greater health risk than widows who are heads of households (Rahman and Menken 1990), and who presumably have some independent sources of income.

282 A field of one’s own I understand from legal activists that the situation today would differ only in detail, not in substance.

The critical point here is that village women’s illiteracy and lack of education, the ideology of female seclusion, and the restrictions on women’s interaction with the extra-domestic sphere, necessitate male mediation in disputes and claims, especially but not only in cases that are not settled locally. Not only are the local village councils constituted largely _ of men, but so are the government administration and the judiciary. Male domination of the administrative and judicial bodies at every level, as well as of the social and other public networks of access to these bodies,’° and the complicated procedures and red tape involved in dealing with them, all work to women’s disadvantage, as does women’s relative lack of financial resources. As a Pakistani woman lawyer notes, ‘the lack of knowledge of the assets, the stamp duty, the cost and length of litigation and customary stigma, usually deters the sharer [who has been] denied her rights from going to court’ (Patel 1979: 139).

Il. Glimmer of change: women claim inheritance shares in some traditionally patrilineal communities

Despite the many obstacles to their claiming land, women in some regions

are beginning to do so in noticeable degree. Here the contrast between Pakistan and Bangladesh, both formally governed by Shariat laws, is interesting. In Bangladesh increasing numbers of women (especially married ones) are now reported to be claiming (or intending to claim) their shares; in Pakistan this still appears to be an uncommon occurrence.’’ In fact there is little to suggest in the available village studies that in Pakistan the situation has changed significantly from that described by Elgar (1960: 186-7) in her village study for Pakistan Punjab:78 In a few cases, daughters took advantage of the new laws and, although they had been married for years and had received their dowries ... they now wanted their share in land as well. A few cases were brought to court, and the brothers of these women tried to show that their sisters had in reality already received more than their share in their father’s property ... But such cases were few in number, and it was felt that they came up when the relationship between a brother and a sister was already

strained, so that the new law provided them with an opportunity for bringing trouble into the open. Otherwise such cases would not come to court and sisters would not claim their property. 76 In this context, female functionaries in such bodies could make a difference (as will be elaborated in chapter 10). 77 Studies which mention such cases are few: see D. Merry (1983) on Pakistan Punjab, and Ahmed (1980) on the ‘Settled Areas’ of the NWFP. 78 A systematic investigation of this aspect today would be useful.

Whose share? Who claims? 283 Bangladesh today provides a contrast. Aziz (1979) and Abdullah and Zeidenstein (1982) note that Bangladeshi widows with sons always file for their claims on behalf of their sons. In Bhaimara and neighbouring villages,

Jansen (1983) heard that women from poor families were now claiming their shares of the patrimony. Almost every woman Nath (1984: 229) interviewed was planning to claim her share in her father’s property after his death: ‘I encountered only one woman in Natunpur who said that she had no intentions of ever claiming her land entitlement.’ This woman’s natal home was 200 miles away, which would have made it difficult to make such a claim effective. All other Natunpur women made it clear that even if they allowed brothers to use their shares for a few years after the father’s death, they were determined to stake their claims sooner or later, the timing depending on the economic status of their husbands and on what phase of his domestic cycle the brother was in. Villagers similarly told Taniguchi

(1987: 30) that ‘formerly daughters used to give up their rights to their father’s assets in favour of their brothers, but nowadays 80% of them actually demand their shares’. Although such assessments are probably on the high side, the overall trend certainly appears to be towards a significant number of Bangladeshi village women seeking to exercise their rights. Ellickson (1972b), Hoque (1987), Jansen (1983), Nath (1984), Taniguchi (1987), and Zaman (1981) all mention one or more cases of daughters who had staked their claims in the

study villages. Indeed, given the small percentage of women who are voluntarily granted their shares as daughters or widows, and the noted field observations by a number of writers that a large proportion of widows and daughters in their study villages were planning to stake their claims, we have a situation of considerable potential conflict within families. Some writers on Bangladesh argue that the giving up of claims by women

is not recognized under Islamic law. Ellickson (1972b: 52) notes, for instance: ‘Though women usually give up their claims to a share of an inheritance in favour of their brothers, Islamic law did not recognize such a

renunciation. The woman’s rights to claim her share of the inheritance remained in abeyance.’ However, prospective male heirs do not always accept this view, and women’s attempts to later claim the ‘temporarily’ relinquished land can lead to bitter disputes. Hence the circumstances which underlie the noted increase in women filing claims are often grim. To begin with, women don’t always claim voluntarily. Sometimes they do so due to pressure from and even intimida-

tion by their husbands. Nath (1984: 227) notes: ‘Among small farmer households, husbands often pressurise their wives and sometimes torture them to get them to claim their parental heritage. Mostly in cases where the daughter’s husband is considerably poorer than her parental household she

284 A field of one’s own tries to exercise her legal rights to get a share ...’ In another Bangladeshi village, Kabeer (1988) found that: Some of the women interviewed had been driven by family poverty or by their husband’s threats to lay claim to their share of the patrimonial property. From the

information offered it appeared that husbands frequently used the threat of desertion to force wives to sell off their share of inherited land. Sometimes the husbands left anyway, having appropriated the proceeds of the sale. One abandoned woman had incurred her husband’s wrath for refusing to sell her share of land, at the same time as alienating her brothers by agreeing to sell trees from the same piece of land.

Even without the husbands’ instigation, given the widespread landlessness and poverty prevailing in Bangladesh today, poor rural women are faced with a difficult choice: on one side lies the economic security that a piece of land (even a small one) can provide; on the other side lie the risks of losing the social and economic security that the brother can offer in case of marital breakdown. Under extreme poverty the advantage of filing a claim

may outweigh that of not doing so on at least two counts: one, where the woman has an adult son, he could substitute to some degree for the loss of the brother’s support; and two, if the brother is also poor, he would be less willing or able to provide economic security. Essentially what we appear to be witnessing in Bangladesh today is a conflict over scarce resources where

women too are increasingly willing to stake a claim, either on their own account or under pressure from husbands and sons. This has probably been accentuated by the severe erosion of traditional kinship support systems. Stories of widowed mothers being abandoned by sons abound.’? As White (1992: 135) notes on the basis of her fieldwork: Family relationships alone are no longer enough: women are much less likely to be

abandoned by their children if they own some land from which to draw their subsistence. In circumstances of declining family solidarity, increasing numbers of women may be caught in the scissor action of norms which largely preclude material

independence through expectation of other (social) sources of support, and the actual failure of those forms of support to sustain them.

We might expect similar trends to be developing elsewhere in the subcontinent. Mayer (1960: 243), for instance, argues that in central India, women seldom claimed their shares when land was plentiful, ‘but now land is in short supply, and any daughter inheriting it will try to gain possession’.

This also provides a clue as to what may underlie the noted contrast between Bangladesh and Pakistan in terms of women staking their claims. I

suspect the answer lies in the much greater poverty and landlessness prevailing in Bangladesh today compared with most parts of Pakistan, especially the agriculturally prosperous Pakistan Punjab. 79 See e.g. Cain et al. (1979) and White (1992).

Whose share? Who claims? 285 It is worth noting, though, that most of the Bangladeshi women who (in

the studies cited earlier) had claimed or had indicated their intention of claiming their shares were married women or widows with sons: husbands

and sons can provide the mediation necessary for dealing with local or distant institutions and authorities. Also the religious legitimacy enjoyed by Shariat law probably helps women’s bid for land shares, and must to some

extent neutralize the ideology of female seclusion which too is given religious approval. In other words, both the material context of poverty and prevailing ideological conditions would be determining factors in whether or not women stake their claims and in the resistance they encounter. In Pakistan, women faced with religious and gender ideologies

similar to those in Bangladesh perhaps do not have the same material imperatives or pressures from kin for claiming their shares.

Consider now the situation in matrilineal and bilateral communities. In these communities, since women’s rights in landed property are recognized by custom, women don’t need to establish the social legitimacy of their claims, nor do they encounter the hostility toward their legal rights in land that women in the northern part of the subcontinent face. Women in these

groups are therefore more likely to inherit land in practice. Chapter 4 focused in detail on the current situation in three of these communities — the Nayars, Garos, and Sinhalese. But a further discussion (especially on some

of the other such groups) would provide an interesting contrast to the patrilineal context discussed above.

IV. A look at traditionally matrilineal and bilateral communities To gain an idea of the present situation in traditionally matrilineal and bilateral communities, we have to depend on ethnographies undertaken mostly in the 1960s and 1970s. Since these focused mainly on issues of kinship and household structures and neglected the intra-household gender distribution of property on which these structures directly impinged, the evidence on land ownership by women is fragmentary. What exists gives a mixed picture. Among the Nayars and Garos, daughters often inherit some landed property, but available evidence (cited in chapter 4) does not permit an assessment of the extent of the prevailing gender gap. In broad terms property division among the Nayars has moved toward bilaterality, with

both sons and daughters receiving some land: typically the father’s property devolves equally on both sexes, but there are regional differences in the extent to which sons share in property held by the mother. There are also regional differences in the degree to which tendencies toward nuclear households and neolocal residence are manifest. The Garos remain predo-

286 A field of one’s own minantly matrilineal, but Garo women are adversely affected by two types

of trends. One, there is a widespread decline in land available to the community, leading to inter-household and inter-village inequalities, and many women who earlier had use rights to communal land no longer do so. Two, there is a shift toward gifting land to sons and toward virilocal and neolocal post-marital residence patterns. Trends among the Khasis are in some ways similar to those found for the Garos, but as we noted in chapter 3, there were in-built tendencies in Khasi communities toward inter-household and intra-household inequalities, which over time have been accentuated in a number of ways. To begin with, some households have managed to bring under their possession tracts of Ri Raid land, that is public land belonging jointly to a village or to a group of villages. A typical way in which this is achieved is by planting trees on a tract

of public land and then claiming that the trees and the land belong to the planter. In addition, among families facing land shortage only the heiress daughter today inherits land: in a study of Mawnai village in the Jaintia hills, all the seventy-two persons interviewed indicated that this was the case (GOI 1967b). The remaining daughters, who earlier had received some part of the family land, thus get excluded. Nakane (1967: 105) found that the best houses in the Khasi villages she visited in the late 1950s were those of the heiress (usually youngest) daughters: ‘[They] are usually landowners, while their elder sisters have often to become day-labourers after their marriage, unless their husbands are capable men, or their fathers wealthy ... Also, although village endogamy and uxorilocality were still dominant in the 1950s, virilocality was a growing trend. For the Lalungs, since there is no direct evidence on land inheritance by gender, inferences about trends in inheritance patterns have to be drawn from changes in post-marital residence practices: these are found to be shifting toward virilocality. In one of their study villages, Syamchaudhuri and Das (1973) found that in forty-six out of a hundred first-generation marriages, the men had gone to live with their wives, while in the next generation in only seven out of fifty-eight marriages had the men done so.

When women continued to reside in their natal homes, non-heiress daughters as well as sons received a usufruct share in the land. But when women went to live with their husbands, they forfeited their claims. The shift toward virilocality is thus likely to mean a shift of land in favour of males.

The link between women’s post-marital residence and the likelihood of their inheriting land is again noted in the context of the Sinhalese in Sri Lanka. Here (as we observed in chapter 4) many women today do receive some part of the parental property, but not all who are legally eligible do so; and the shares of those that inherit are usually smaller than the shares of

Whose share? Who claims? 287 Table 6.2: Dowry land by type among the Jaffna Tamils

Type of land Paddy Garden Compound Palmyra Coconut Rocky No. of dowries with

given type of land 82 74 79 43 22 10

Total area (acres) 103.0 58.2 29.4 25.4 9.4 27.5 . Average land per dowry

in whole sample (acres)' 0.981 0.554 0.280 0.242 0.089 0.262 Average land per dowry, where dowries actually

contained land (acres) 1.256 0.786 0.372 0.591 0.427 2.75 Note: ' The sample size (that is, the total number of dowries) is 105. Source: Banks (1957: 189)

Table 6.3: Landownership in three Jaffna Villages by source of acquisition Source of acquisition

Dowry Inheritance Purchase Total

Sector No. % No. % No. % No. % Agriculture 6478 6738 2019 21 612312 96 100 100 Fishing 159 203 Artisan 20 44 15 33 10 22 45 100

All sectors 243 71 73 21 28 8 344 100 Note: Land listed under dowry is owned by women alone; that listed under inheritance and purchase is likely to be owned largely by men. Source: David (1980: 114)

their brothers. Women’s chances of inheriting are also weakened if they reside outside their natal village.

However, one Sri Lankan community where women were found to receive significant amounts of parental landed property is the Jaffna Tamils. Both Banks’ (1957) study in the 1950s and David’s (1980) in the 1960s establish this (see tables 6.2 and 6.3). Indeed 71 per cent of the land in

David’s three study villages was dowry land owned only by women. The

remaining 29 per cent was obtained by the owners via inheritance or purchase. Even if we assumed that this latter was exclusively male property, it would still constitute less than a third of all landed property in the village.

Traditional marriage preferences have also persisted among the Jaffna Tamils: in Chirripudi village studied by Banks (1957: 35), 69 per cent of the

288 A field of one’s own marriages were between relatives, and 35 per cent between cross-cousins (actual or classificatory); and David (1973a: 26) found that 90 per cent of the marriages in his study area were within the village and 65 per cent in the same ward.®° This would be conducive to women protecting their interests in any property disputes and exercising control over their land. It 1s difficult to say, however, in what ways the political disruptions of recent years have affected these patterns in Jaffna or in the Eastern Province of Sri Lanka where the Muslim Moors, discussed below, are concentrated. For the matrilineal Muslims of South Asia — the Moors of Sri Lanka, the Mappilas of north Kerala, and the Lakshadweep Islanders — the picture is mixed. Legally, these groups are governed by Islamic law. In practice,

Islamic inheritance laws have made only limited inroads in all three communities. Munck (1985), who did fieldwork among the matrilineal Moors in Kotabowa village (Uva district, Sri Lanka) in the early 1980s, describes a community still following customary practices: daughters inherited paddy lands and houses via dowries and mostly lived matrilocally. The dowry land was always written into a deed in the woman’s name; and where the family had insufficient land for all daughters, some daughters

were leased land on a sharecropping basis. In 79 per cent of the 119 marriages surveyed by Munck, daughters received paddy land in their dowries, and in 71 percent of the cases they also received cash. Thus the gap between contemporary law and customary practice in this community has

favoured women. Munck’s work suggests that no major erosion in women’s rights had occurred till the time of his study. McGilvray’s (1989)

fieldwork in 1969-71, 1975 and 1978 among the Muslims of Amparai district similarly showed that the women still inherited much of the landed

property through dowries. What we might expect, however, 1s that any disputes which come to court, as could happen with growing frequency as land scarcity increases, would be settled in accordance with Islamic law and favour male over female inheritance.

Information on the Mappilas of north Kerala is, unfortunately, very dated. In 1948 when Kathleen Gough did fieldwork in Kottayam village,

although 71 per cent of post-marital residences were matrilocal and duolocal, neolocality had also become significant. Ancestral property was being inherited according to the Mappilla Marumakkatayam Act, 1938

(Madras Act 17 of 1939), under which the shares of female and male members of the matrilineal joint family in the taravad were determined ona per capita basis. But a man’s self-acquired and separate property, if he died intestate, devolved according to Islamic rules (under the Mappilla Succes-

sion Act of 1918), whereby daughters received half the shares of sons. 80 A ward isa section of the village constituted ofa cluster of houses, gardens, and often also a temple.

Whose share? Who claims? 289 However, daughters were also receiving shares in men’s newly acquired lands in dowry, and among the aristocracy, houses and gardens were being gifted to wives near the women’s natal homes. In other words, Islamic law

had not entirely displaced custom, even for self-acquired land. Today, however, Islamic law applies to all Mappila property, although without upto-date ethnographic evidence it is difficult to say to what extent the law has altered practice. In the Lakshadweep Islands, again, custom still dictated practice in the early 1960s, when Kutty (1972) found a predominance of duolocal (77 per

cent) and uxorilocal (19 per cent) residence in Kalpeni (one of the nine Islands in Lakshadweep proper).®! Trees were still the main form of property. Although, like the Malabar Mappilas, the Islanders followed a dual system of inheritance, there was very little property which could be subjected to Islamic law: 80 per cent of the taravads in Kalpeni only possessed matrilineally inheritable trees, and another 17 per cent possessed both types of trees, viz. those matrilineally inheritable and those inheritable

according to the Shariat. Other aspects of Islamization were also very limited in the 1960s. As among the Moors, Lakshadweep women did not

veil themselves (although in Kerala Mappila women did so in limited degree), and divorce, initiated by both men and women, was fairly frequent. There are, however, recent winds of change: a small but growing number of Islanders, influenced by outside contacts and with the support of government officials (of largely patrilineal backgrounds), are suggesting that Islam and matriliny are incompatible and arguing fora stricter enforcement of the Shariat (Dube and Kutty 1969).

In short, while custom still prevails in significant degree to women’s advantage among the matrilineal Muslims of the subcontinent, the directions of change are adverse. So far, changes appear to be most apparent among the Mappilas of Kerala, who are also likely to be more susceptible to the influence of Muslims elsewhere in the country (Miller 1976). In general, the barriers to women inheriting land among the traditionally matrilineal and bilateral communities, compared with traditionally patrili-

neal ones, are much less formidable, but not absent. As noted, at the ideological level, women’s claims receive clear recognition among the former communities. The typically high levels of female literacy among these groups are also conducive to legal awareness and action. Compared with northern South Asia, the absence of female seclusion practices makes

the women less dependent on male relatives. And unlike the Hindu communities of northwest India, village endogamy is allowed and common. At the same time, women are still not on a par with men in their 81 Also see Dube and Kutty (1969).

290 A field of one’s own ability to exercise their claims. First, the growing tendency for women’s post-marital residence to be outside their natal villages, noted for several of these communities, is likely to reduce a woman’s chances of being given her share of the ancestral estate; and women living outside the village are also in

a less-strong position to protect their interests in inheritance disputes, especially where village-level institutions determine the decisions. Second, women of matrilineal and bilateral communities are not invulnerable to the male bias in government policy and in the bureaucracy which affects women in patrilineal communities. For instance, among the Garos, under the land privatization being encouraged by the State, the title deeds granted to individual households are typically in male names. In 1988 when I visited the Garo Hills and asked the male officials concerned why even ina matrilineal community they allotted the titles to men, they said: ‘Because women cannot come to our offices to fill out papers.’ Yet two streets away there were women traders to be seen everywhere! In Sri Lanka, similarly, the land allotted to Sinhalese couples in irrigation resettlement schemes, such as the Mahaweli scheme, is usually registered only in the names of the husbands, who are assumed to be the household

heads. Moreover, each household can nominate only one heir, who is almost invariably a son if the family has one. This undermines the bilateral rules of inheritance recognized by customary as well as contemporary law, whereby married Sinhalese women have independent rights to own and

control land. Under the Mahaweli scheme, if a woman divorces her husband she is deprived of any means of subsistence from the land, underlining her dependent and subordinate position. An anthropologist who studied one of the Mahaweli settlements found that ninety-six out of 112 (or 86 per cent) of the land allocations were made to men. Of the sixteen women who were granted land, only two (a widow and a separated woman) were living in the project area and managing their own farms. Typically ‘a woman only applied for land if there was a minimum chance for male members of her family obtaining a plot’ (Schriyvers 1988: 44-5). Given the trend toward village exogamy and the male bias in access to legal and administrative institutions, the noted disadvantages women face in protecting their interests are likely to be felt more and more acutely, as

land scarcity grows and the interests of all contenders (male or female) come increasingly into conflict.

All said, therefore, although in legal terms we see a movement toward bilateral forms of inheritance everywhere in the subcontinent, in practice there continue to be striking differences between traditionally patrilineal communities and traditionally matrilineal and bilateral ones, in the extent to which women actually realize their legal rights. This divergence between

Whose share? Who claims? 291 contemporary law and actual practice stems from the continued dominance of custom. A weakening of custom and a greater adherence to prevailing laws is likely to benefit women within patrilineal settings, but to have the

opposite effect for women in several of the traditionally matrilineal communities, especially the Muslim ones.

V. Some hypotheses On the basis of our discussion we can hypothesize that the likelihood of individual women exercising their inheritance claims to land would depend especially on the following factors: (a) the strength of purdah norms and practice; (b) post-marital residence and marriage distance; (c) the extent of male support; (d) women’s level of education; and (e) the extent of women’s economic vulnerability. The stronger are the practices of purdah and village exogamy and the greater the marriage distance, the higher are the chances of a woman giving up her claim. However, the greater the male support she can fall back on, the more likely is she to file a claim. For instance, if she has adult sons a woman will be in a stronger position to fight for her share both in her partrimony and in her husband’s land if widowed, than if she has no male-mediatory support. Economic vulnerability, however, could work in either direction: it could cause a woman to relinquish her claim because she does not have the financial means to exercise her rights, or it could induce her to stake a claim because the potential economic security of a piece of land could outweigh other considerations. Three of these factors (namely a, b, and d) show a systematic variation across the subcontinent and will be mapped in chapter 8, while chapters 9 and 10 will seek to provide pointers

on how the noted constraints on women could be reduced. Let us now move, however, to the difficulties women tend to face in controlling and self-managing the land they do come to possess.

7 Whose land? Who commands? The gap between ownership and control

When my husband died, my neighours wanted my land. They beat me on

any pretext, they tried to chase me out. Then they started to say I was having an affair with my brother-in-law, that was why I wouldn’t leave despite their harassment. My homestead is like a jungle. I cannot grow any crops on it because my neighbours let their goats graze on my land. When I go to the bazaar, they steal what few crops I have. (A Bangladeshi widow to Kabeer 1988:20)

The gap between women’s de jure and de facto ability to own land, discussed

in chapter 6, is only half the story. The other, equally significant, half concerns the gap between ownership and control. The issue of control has several dimensions, three principal ones (in the context of individual ownership) being the following: women’s ability to retain title to the land they inherit or otherwise acquire; their ability to take

decisions regarding the disposal of the land through sale, mortgage, bequest, or gift; and their ability to take decisions regarding the use of the

land, including leasing it out or self-managing it, and disposing of its produce. By self-management I mean directly cultivating the land with one’s own labour and/or cultivating it through hired labour under personal supervision. Each of these dimensions of control is important if women are

to benefit from their land, and their advantage is greatest if they have control in all three ways. None of these forms of control, however, is guaranteed to a woman by virtue of legal ownership alone. This chapter examines the odds against which women have to labour to exercise control over land in practice.

I. Women’s ability to retain their land Even after a woman has inherited land, her brothers, other relatives, and even neighbours may continue to seek ways of dispossessing her. This is especially noticeable in Bangladesh today where, as noted, conflict over small parcels of land is intensifying under conditions of extreme land

scarcity. Those interested in acquiring the land resort to all kinds of methods to achieve this end. Many Bangladeshi village studies provide 292

Whose land? Who commands? 293 vignettes of cases from which a larger picture can be constructed. To begin

with (as also noted in chapter 6), land inherited by a woman may not formally be registered in her name: there are examples of sons falsely registering in their own names land belonging to their widowed mothers (Zaman 1981). Where the woman’s share is registered, close kin may yet get the land records changed by bribing the concerned village officials (Hoque 1987). There are also cases of men coercing their sisters into selling the land to them at a low price, and sometimes defaulting on the promised payment.

One woman cheated of her inheritance, unable to bear being betrayed by

her own brother, committed suicide (Hartmann and Boyce 1983). In another instance, a widow with a minor daughter sold the land she had inherited from her husband, and went to live with a brother who promised

to help her buy another plot in his village. Instead, he kept the money (Abdullah and Zeidenstein 1982). A woman’s illiteracy and trust can both leave her vulnerable to fraud, sometimes through ingenious means. Arens and Van Beurden (1977) relate how a man, who had himself inherited nine acres, contrived to appropriate an additional three and a half acres that his sister had inherited in accordance with the Shariat after their father’s death. Inviting her to a film in a nearby town, he obtained her thumbprint on a piece of paper, saying that this was needed to get the cinema ticket. In fact the document was an agreement transferring the legal ownership of her land to him. Divorce threats by husbands, and even torture, if wives refuse to transfer their land to the man’s name, are not uncommon (Yunus 1984). A widow

who inherits is likely to be subjected to pressure from the husband’s relatives to give her share up; and where the relatives are powerful, she can

be dispossessed. In one such case, the woman is today working as a domestic servant (Islam 1985). In another, described by Bertocci (1972), a widow vigorously resisted all attempts to deprive her of her inheritance, earning the title of ‘pagali’ (madwoman) because she hung on tenaciously despite the litigation that ensued. Indeed, involving a woman in a court case is a common way of forcing her to mortgage or sell her share when legal expenses become unaffordable.! Pressure may also come from the woman’s natal family where its interests are involved. In 1985, Rowshan Qadir, a Bangladeshi sociologist, told me about a woman who had inherited a total of forty acres from both parents and was living abroad with her husband. She was being pressured by her mother’s relatives to divorce her husband and marry a relative from her natal village. They planned to kill her if she ever sought to directly control the land, which was being farmed by her maternal uncles.

Neighbours may also be part of the fray. Some have been known to 1 See Bertocci (1972) and Cain (1978).

294 A field of one’s own appropriate a woman’s land by bribing the land records officer;? others by

forcibly occupying her land; and still others by physical violence, as indicated by the story of the Bangladeshi widow quoted at the beginning of this chapter and by similar cases documented by several other writers.2 The frequency of these stories suggests that they are not just isolated instances, but part of an overall pattern. Within this general climate of hostility, one may expect that widows who have adult sons would be better able to retain control over their land, although, as noted, sometimes sons too defraud the mother.

Although all the cases described above relate to Bangladesh, I would hypothesize that elsewhere too similar gender conflicts over land are likely to be developing, as land becomes increasingly scarce. The land-associated resurgence of witch-killing in tribal Bihar and West Bengal, described in the

previous chapter, is a case in point. (There is a strong case here for the systematic collection of data on such conflicts in the subcontinent.) In addition to these pressures, conditions of poverty can make it difficult

for households to retain land. Although both sexes are affected by this, women (and especially women heads of households) are more likely to be forced to sell land in a crisis, given their greater economic vulnerability. Some micro-studies from Bangladesh bear this out. In land sales in Sherpur thana (Bogra district) during 1976-77, women constituted 7 per cent of the sellers and 2 per cent of the buyers, contributing 5.3 per cent of all land sold but only 0.7 per cent of all land bought. As a group, women thus lost out (Sultan 1982). Again in two villages of Comilla district, of the fourteen women who had inherited shares, only six still retained them at the time of Westergaard’s (1983) study in the late 1970s.* Similarly, 26 per cent of the 146 women who had inherited land from their fathers in Kabeer’s (1985: 88-90) study village had sold it; another 48 per cent had waived their claims in favour of brothers. Only a small percentage had retained their inheritance, the land being cultivated by their husbands or sharecropped out. Let us now turn to the other aspects of control.

IT. Control over the transfer and use of land What autonomy do women have in relation to the land over which they do retain ownership? This question becomes necessary because of potential differences in the rights that land ownership confers on women and men. For men, ownership tends to imply full control over the transfer and use of 2 See e.g. Hoque (1987): in this case the woman had inherited as an only child. But the land was owned jointly with another villager who had her name erased from the land records by bribing the concerned officer. The matter went to court but is yet to be decided. 3 See Arens and Van Beurden (1977), Hoque (1987), and Kabeer (1988). * On the greater likelihood of women than men losing land in the context of increasing landlessness in Bangladesh, also see Jansen (1983).

Whose land? Who commands? 295 land, if not as individuals (in case the property is joint) then certainly as a gender. But when women have ownership, control over the land may still vest in varying degree with men. The limitations on women’s control may be both legal and social. Legally, in some parts of South Asia, women are

not free to dispose of their landed property as they wish. In Nepal, for instance, a woman can dispose of only half the land she inherits. Disposal of the other half requires the permission of her father (if she is unmarried) and of her adult sons (if she is widowed or living separately from her husband).

Among the Jaffna Tamils in Sri Lanka, a married woman living with her husband needs the husband’s consent for disposing of any land she owns, whether inherited or self-acquired after marriage. Laws apart, male kin may also attempt to intimidate women against bequeathing their land as they wish. For instance, a sonless widow in Tamil Nadu was prevented from bequeathing to her daughter’s sons the fields she

had inherited from her husband, and instead was forced to adopt an agnate’s son as heir (Beck 1972). In another case, a widow’s attempt to bequeath her land to her sister’s sons led to murder in Pakistan Punjab (D. Merry 1983).

The right to alienate the land is an important aspect of control. The freedom to mortgage or sell the land can prove critical in an economic crisis, and is important even in non-crisis situations where land may be needed as collateral for a loan. This right also gives the owner extra leverage and a stronger fall-back position than lifetime usufruct rights over the land are likely to provide. For example, in old age women are more likely to receive

assistance and good treatment from their relatives if the latter are also potential heirs (White 1992; Dreze 1990).

The freedoms to decide how to manage and use the land and how to allocate the produce are also crucial ones, and are applicable to both owned

land and land over which individuals only have use rights. Women have typically been disadvantaged in these respects even in some of the customarily matrilineal communities, especially where large property holdings were involved. For instance in Kerala, among the Nayars and matrilineal Mappilas, although women were the legal owners, in that inheritance passed through the female line, men were the formally designated managers of the estates. Management could involve, among other things, exercising

control over the distribution of produce from the land: among Nayar households of central Kerala, for instance, the allocation of farm output is noted to have been strictly controlled by the karanavan,* who also enjoyed > Gough (196la: 337) notes that normally no crops were sold and no produce allocated without the karanavan’s consent. He decided on the quantities of unhusked rice, vegetables, etc. that were to be transferred from the granary to the storerooms within the house. He also controlled purchases from the town and twice a year distributed clothing to both male and female members.

296 A field of one’s own the advantages of public authority associated with property management. Again, among patrilineal groups governed by Mitakshara law, although individual men could not freely dispose of joint family property which they collectively owned, men as a gender still had overall managerial control over it. Today, even where women individually own land or have use rights over it, cases of self-management are rare, although they do exist.° Women are subject especially to two types of constraints if they seek to self-manage the land: direct ones in the form of pressure from relatives, and indirect ones defined by women’s social context (as discussed in the next section). Both types of constraints tend to discourage women from self-managing land and push them to either rent it out or let male relatives manage it. Typically, their husbands or adult sons manage their land.’ A widow without grown sons may get a son-in-law to settle uxorilocally, but usually she sharecrops out the land inherited from her husband to his relatives, while daughters who inherit usually sharecrop to a brother.® In theory, of course, leasing out land need not be a bad deal in itself, and men often do so as well. The lease money (in case of cash renting), or a share

of the harvest (in case of sharecropping), can improve women’s fall-back

position in the family both by giving her some independent means of survival (crop produce can be consumed directly or sold for cash), and by making her economic contribution to the family’s welfare more visible (and thereby reducing the ‘perceived contribution’ bias discussed in chapter 2). If a woman leases out her land to her brother, sometimes this too can work to her advantage by enabling her to accumulate savings in her natal home outside the control of her marital family (Kabeer 1988). In practice, however, these advantages from leasing out don’t always accrue to women. They are vulnerable to being defrauded even if they lease out to brothers and are often not in a position to ensure they will get the agreed-upon share of the harvest (as detailed below). In fact, the decision to lease out her land is not always a voluntary one for the woman. She 1s likely © Among studies which mention cases of women managing the cultivation of their own land are the following: for Bangladesh, see Abdullah and Zeidenstein (1982) and Nath (1984);

also personal communication in 1985 from Lily, a member of Nari Pokkho, a woman's group. For India, see Bailey (1957) and Murray (1984). And for Pakistan, see Asha (1971), Khan et al. (1984), and D. Merry (1983). 7 For Bangladesh, see Abdullah and Zeidenstein (1982), Cain (1978), Nath (1984), and Qadir (1981). For Pakistan, see Pastner (1978) and Rouse (1988). Also see footnote 3 in chapter 6 for cases where daughters in sonless families have inherited land; here the uxorilocal sonsin-law generally manage the land. 8 A number of studies mention cases of women renting out their land: for Bangladesh, see Begum and Greeley (1979), Gardner (1990), Hartmann and Boyce (1983), Jansen (1983), Kabeer (1985), Nath (1984), and personal communication from Lily, Nari Pokkho. For India, see Furer-Haimendorf (1985), Kessinger (1979), and Minturn and Hitchcock (1966). And for Pakistan, see Aschenbrenner (1967) and Young (1984).

Whose land? Who commands? 297 to face considerable pressure from brothers or the husband’s agnates (as the case may be) to lease the land to them,’ and a refusal to do so may even lead

to violence: Minturn and Hitchcock (1966: 28) report how in Khalapur

(northwest India), ‘a widow with an only daughter who insisted on managing her own estate and let it out on shares was severely beaten by her husband’s kinsmen’. In another case, in the Indian Punjab, a widow witha young son continued farming her husband’s land, resisting pressure from

her brothers-in-law to turn over its management to them and to accept room and board with them. As the conflict escalated, she retreated to her parents’ home and rented out the land, whereupon the enraged brothers-inlaw had her murdered and appropriated her land, depriving her young son of his share (Murray 1984: 359n). Leasing out land, whether to the woman’s natal kin or to her husband’s relatives, is likely to be on below-market terms.!° Where the lease is on a

sharecropping basis (as is the common pattern), this disadvantage is compounded by the limitations on a woman’s ability to ensure that she receives the harvest share agreed upon. First, if she inherits from her father but is married into another village, she will find it difficult to keep track of how good the harvest has been. Norms of female seclusion (discussed tn the

next section) also impinge on this. She will thus have to accept on trust whatever share she Is given. Second, if the sharecropper is a brother or other

relative, even if she suspects she is being cheated, it will be difficult to confront him, if she needs to maintain cordial relations with him. Instances of brothers cheating are not uncommon. Field workers on the Bangladesh Rural Advancement Committee (BRAC) project told me of several cases

where a brother had forced a married sister, living in another village, to | lease him her portion of their inherited parental land on a sharecropping | basis, and had then kept defaulting on her harvest share by pleading that the crop had failed or making some other excuse. In one instance, the woman was finally forced to sell the land to her brother at a very low price. In another case, described by Cain (1978), a Bangladeshi widow was pressured by her husband’s brothers to lease her land to them until her minor son was old enough to manage the land, on the promise that they would give her a part of the harvest. But after a while the payments stopped.

The agnates also failed to pay property tax on the land. The government therefore seized the land, but the agnates purchased it back at a nominal price by bribing the authorities, thus becoming its owners. Cases of women self-managing land appear more common in tribal ° For Bangladesh, see Qadir (1981). For India, see Mayer (1960), Minturn and Hitchcock (1966), and Standing (1987). And for Pakistan, see Aschenbrenner (1967). 10 See Qadir (1981). It would be useful to have systematic data on the extent to which the rental terms a woman gets deviate from prevailing market rates.

298 A field of one’s own communities, although today this is often under extremely hostile social conditions, as noted in the last chapter for the Ho, Santal, and Munda tribals. Female management is also common in the hill regions of India and

Nepal where, due to long-distance male outmigration, women are left to cultivate on their own (even while the legal titles are held by men) as de facto heads of households. However, a range of factors, as discussed below, can

restrict women’s ability to function effectively as independent farmers. These factors can also limit their ability to lease in land where they own little

or none. Although most of these obstacles would apply to women as a gender, their importance and implications vary by class and region.

II. Barriers to women self-managing land

(1) The physical and social confinement of women | Critical to women’s ability to self-manage land is the freedom they can exercise in their interaction with men, embodied in the cultural practices which define what sorts of interaction are permissible, with which men, in

what contexts, within which spaces, and using what modes of conduct. These cultural norms vary a great deal across the subcontinent, ranging from severe restrictions in Pakistan, northwest India, and Bangladesh, to much subtler ones in South India and Sri Lanka, with barely discernible restrictions among tribal communities in northeast India and the TibetoBurman communities of Nepal. The forms these restrictions can take and their implications for women’s autonomy vis-d-vis land use are described below. A mapping of their regional gradations is deferred to chapter 8. In broad terms, restrictions on male-female interactions fall into three interrelated categories: the veiling of women, the gender segregation of space, and the gendered specification of behaviour. Effectively, the first two

work towards the physical containment of women and the third toward their social containment. In some regions and communities, strictures relating to all three overlap and reinforce each other. In others, only the last may come into play. Indirectly, however, norms of ‘feminine’ behaviour can also work in subtle ways to physically restrict women’s movements even in societies (including western ones) which do not seek to gender spaces in explicit terms. Purdah ideology, or the ideology of female seclusion, 1s embodied in all

three categories of practices. At the same time, not all societies which gender spaces and behaviour can be labelled purdah societies, since not all forms of gendered behaviour are variations of ‘purdah’. The tendency, in some recent discussions on purdah in South Asia, to conflate all observed

Whose land? Who commands? 299 forms of gendering as facets of purdah,'! in my view obscures some notable differences between purdah practices and the more general social construc-

tion of gender that cuts across cultures. In terms of the implications for

women it is important to recognize both the commonalities and the differences in how communities and societies gender behaviour across South Asia and in other parts of the world. It can be suggested that a possible, admittedly crude, way of distinguishing purdah regions from others could be to see whether or not there are explicit strictures concerning the physical confinement of women.!? By this count, south and northeast

India, Sri Lanka, and Nepal would not be labelled purdah regions, although these societies do gender behaviour in a number of ways (as discussed later in this section). Common to all three sets of practices embodying purdah ideology is the principle of avoidance in interaction with men, although the specification of which men and 1n which social contexts varies across groups and communities. The rationalizations for this avoidance are cloaked in terms of izzat (family and personal honour), female chastity and modesty, the need to control female sexuality, and so on; and the precise nature of these practices varies by region, religion, caste, class, and circumstance.

Veiling, the most visible aspect of purdah and that which is most commonly associated with it in the popular imagination, is in fact not very widespread: it is not universal even among Muslims, and among Hindus is

restricted to northern India (being more common in the northwest than elsewhere). Its norms vary between Muslim and Hindu communities. Muslim women in patrilineal societies are expected, from soon after puberty, to veil before all men defined as outsiders (strangers, distant relatives)'? but usually not before near kin, close family friends, and certain

categories of high and low-ranking men (such as religious leaders and servants). In contrast, a Hindu woman is usually required to veil only from older male affines, although where all members of a woman’s marital village are notionally considered her in-laws, she is expected to veil before all older male members of that village.'+ Overall, the range of men before whom women are expected to veil themselves is narrower among Hindus than Muslims. '° 1! For instance, see Chakravarti (1986), Kabeer (1988), U. Sharma (1980), and Vatuk (1982). 12 Crude because what is actually observed is a gradation of purdah practices and not a clearcut dichotomy. 13 In practice who gets defined as ‘outsider’ varies a great deal: on this, see especially, Vatuk

14 Occarionally young Hindu brides veil or cover their heads before mothers-in-law or older female affines as a sign of respect (Luschinsky 1962). 15 For useful discussions of the differences and similarities in Hindu and Muslim veiling norms and practices, also see Mandelbaum (1988), Papanek (1982), and Vatuk (1982).

300 A field of one’s own In practice, the extent and form of veiling is highly variable among both

Muslims and Hindus. A graphic description of veiling among rural Muslims in the 1950s, in one of its more extreme forms, is provided by Barth

(1956: 46-7) for the NWFP in Pakistan: A woman, walking through the fields or on the paths in the company of her husband, will leave her husband’s side whenever a man appears, seek the shelter of a bush, and cover her head and face completely with her heavy black sheet, till the

stranger has disappeared. Similarly, groups of women working in the fields discontinue their work, and squat, totally covered by their sheets by the side of the terrace wall, when a man approaches.

Elsewhere in Pakistan, veiling has been used as a signifier of underlying hierarchical relationships: for instance, in Baluchistan in the late 1960s

lower-ranking Hitmatkar women traditionally did not veil before the upper-ranking Hakim, as a sign of respect towards the latter (C. Pastner 1971).

The extent of covering varies as well. For most women, it means covering the face with the end of the sari, shawl, or dupatta. The burqa or chador is worn only by Muslim women and then not by all, being more common among the better-off households and outside the home.'° Age permits some relaxation in these norms among both Hindus and Muslims.!’ Also the often emphasized differences in veiling practices between Hindus and Muslims obscure the many similarities which point to the subtle intermingling of religious and cultural prescriptions.!*® In fact, among the matrilineal Muslims, both in Sri Lanka and on India’s Lakshadweep Islands, women do not veil, and they only do so in limited degree (especially as a result of recent Islamic influences) in Kerala. Geographically, more pervasive than the practice of veiling is the related notion of ‘territorial’ purdah, or the gender segregation of space. Again this

can take various forms: within the house it can mean allocating the innermost parts of living spaces almost exclusively for women’s use and the

outer quarters almost entirely for male use. Or it can take the form of confining women to the family compound: in rural Bangladesh, this is the bari, typically consisting of an inner courtyard surrounded by a cluster of huts, with vegetation and sometimes also screens of woven rushes protecting it from outside view (Abdullah and Zeidenstein 1982). Most commonly, however, territorial purdah relates to spaces outside the home. Particularly in the villages of northwest India, Pakistan, and Bangladesh, certain spaces 16 For Pakistan, see Asha (1971), K. Merry (1983), Rouse (1988), Shaheed (1984), and Weeks

17 For Bangladesh, see Gardner (1990); and for India, see Minturn and Hitchcock (1966). 18 On this see especially Vatuk (1982).

Whose land? Who commands? 301 are defined as public, open to men but restricted for women. There is of course some variation in the spaces that are deemed public, but in general, places where men congregate (such as tea stalls, the panchayat house, and

the market place) are spaces which women must avoid, strictures being strongest in relation to the bazaar or market place. Minturn and Hitchcock (1966: 27) note for the Rajputs of Khalapur in northwest India: The life of a woman is surrounded by restrictions imposed by purdah. Women may visit neighbours, particularly if their houses connect with each other and they can go over the roofs unseen by men; but for visits to more distant neighbours they must wait for ceremonial occasions ...

Although a woman’s seniority, her age, whether she is a daughter or a daughter-in-law, and her class and caste all affect her freedom of movement, so that older women with grown-up sons, village daughters (among Hindus), and women of poor and low caste families are less restricted, even

for these women there are restrictions in relation to spaces of predomi-_ nantly male presence. Purdah restrictions are maintained in complex ways. To begin with, a woman’s character and chastity may be associated with compliance to purdah norms, so that women who observe the norms are assumed to be chaste and good and those who transgress them to be of questionable moral character. Girls are socialized into this way of thinking from an early age

and generally require no external policing. For instance, in the NWFP of Pakistan, where purdah is extremely strict, eight- and nine-year-old girls are completely separated from male society and must not show their faces to males other than near relatives (Barth 1956). In Pakistan Punjab, girls close to puberty are expected to cover their heads in front of male strangers and

to lower their eyes when speaking to men, and as they grow older are increasingly confined to the home (Kurin 1981). Modesty and submissive-

ness are often the two most emphasized characteristics of ideal female behaviour. These ascriptions, internalized by women over long years of socialization within the family, manifest themselves 1n docility of demeanor

and respectfulness toward male authority. The codes of modesty can include such explicit standards of feminine behaviour as: ‘Shyness’ of demeanor, avoidance of eye-contact with males, avoidance of loud speech and laughter (particularly in the presence or within earshot of males), and the

limitation of conversation with non-family males to necessary, work-connected topics. They include also such gestures as rising (or crouching on the floor) in the presence of male visitors or family members, turning one’s face aside and refraining from participation in male conversation when one is unavoidably present, and

covering the mouth with the hand or a corner of the sari or head-shawl when speaking. The particular form of these gestures varies from one part of the subcontinent to the other, but their function is similar. (Vatuk 1982: 70)

302 A field of one’s own A graphic illustration of how women internalize these values of avoidance is provided in a study of an Indian Punjab village, where American anthropologist Sandra Murray (1984: 269-70) reports the following conversation between herself and thirty-four-year-old Kiran Kaur (KK) as they were returning from the gurdwara (the Sikh place of worship):

KK (not wishing to be seen): Let’s take this lane. Murray: Why?

KK: Can’t you see that group of men outside the /ambardar’s (headman’s] house?

Murray: Yes, but what of it?

KK: No, I don’t want them to see me. I don’t want to give them anything to talk about. Murray: What could they talk about? KK: Oh they may say Sat Sri Akal [a greeting], to us, but when we are gone, they will say: ‘Where did she go? What did she have to go out for?’ And then they may tell someone else they saw us on the street.

Murray: Oh?

KK: Or they may ask me: ‘Where have you been? Why did you go there?’

Murray: Yes, but we have only been to the temple. KK: But I don’t want to talk to them. I don’t really know them, and if] have to talk to them then someone else may see us doing so, and then they will go around talking about it. The assumption that even innocent encounters will lead to gossip about her character — something to be avoided at all costs — can thus lead a woman to self-restrict her movements.

| The threat of gossip and of being labelled a woman of loose character, however, is only one means of controlling female behaviour. Among the Pukhtuns in the tribal belt of the NWFP of Pakistan, the restrictions of purdah are the severest in the subcontinent, and an actual or suspected transgression can lead to death. Lindholm (1982) and Ahmed (1980, 1986) describe several such cases. An illustrative one is the case of a woman whose

fiancé, on seeing her speak to a young male cousin of hers in the fields, complained to her family, whereupon her father and brother shot her dead.

In speaking to a potential sexual partner other than her fiancé, she was suspected of being unfaithful (Ahmed 1980: 207). In another instance a man shot his wife dead on the mere suspicion of her infidelity aroused by village

gossip (Lindholm 1982). In this community, Lindholm (1982: 220, 222)

notes, the notion of purdah goes far beyond anything enjoined in the

Whose land? Who commands? 303 Koran: ‘It is stretched to signify a prohibition on divorce, a taboo on female inheritance of land, and the complete dominance of husband over wife...’

To a greater degree than perhaps anywhere else in the subcontinent, among the Pukhtuns the honour of men is integrally tied to the chastity and seclusion of women as part of Pukhtunwali (the Pukhtun code of honour which precedes Islamic tenets). And ensuring that a woman observes these

brothers and father.!? |

prescribed norms is seen as the direct concern of men, particularly her

However, interaction between the norms of purdah and its actual

practice is complex. In ideological terms, purdah is associated with social status and deviations from it with the loss of status, albeit in varying degree.

But not all households can afford the strict confinement of women. Economic necessity constantly pushes and strains against the ideological wall of purdah, testing how far it will give with no loss or minimal loss of social status. Among those peasant households in Pakistan Punjab which cannot afford to substitute family female labour by hired labour, there is social acceptance for women working on the family fields. Rouse (1988) found that except among families which came very high in the religious and social hierarchy (the Syed and Miane), most women worked in the fields and did not observe strict veiling. Only sixteen out of 287 households in K. Merry’s (1983) study village adhered to the strict ideals of purdah. In fact, in Pakistan Punjab, the family fields are not considered public spaces for

‘tenant and subsistence farmer households, although they are out-ofbounds for large farmer and landlord households (Shaheed 1984). However, what this also makes clear is that the ideal still remains one of female seclusion. The families with the most social prestige in the village practise

strict purdah, be they the landlords of Shaheed’s study or the religious leaders of Rouse’s study. Hence if a household prospers economically, it seeks to emulate this ideal. In Naveed-I-Rahat’s (1979) study village in Pakistan Punjab, an increase in remittances due to male migration to the Middle East led women who had earlier worked in the family fields to retreat into stricter seclusion.

Other people’s fields, in any case, are considered public spaces in Pakistan, Bangladesh, and much of northern India, and it is only severe economic necessity which forces women among Muslims and upper-caste Hindus to do agricultural wage work. Indeed, in Bangladesh, all fields are considered public spaces, and traditionally the only major agricultural task 19 Qn the strictness of purdah practices among the Pukhtuns, also see Spain (1957) and Vreeland (1957). Some other communities too associate male honour with the seclusion of women, although among them this does not usually lead to bloodshed: see e.g. Jahangir (1979) and Abdullah and Zeidenstein (1982) for Bangladesh, and Hershman (1981) for the Indian Punjab.

304 A field of one’s own undertaken by women was post-harvest paddy processing that could be done in the family compound. Today as landless and near-landless women are being forced by economic necessity to seek wage work, as far as possible they work for better-off relatives within the latter’s bari. Such work can still be considered ‘inside’ work, and does not violate purdah norms in the way that working in the fields does (Westergaard 1983). Also the women travel out in the early mornings and return only when the dusk can render them as invisible as shadows (Abdullah and Zeidenstein 1982). Even so, such work is sought reluctantly and usually involves some sacrifice of social status. Off-bari work is generally disapproved: the majority of men interviewed by Westergaard (1983) in two villages of Comilla district in Bangladesh did not

approve of women seeking such employment. Hence the freedom of movement which low caste or poor women appear to enjoy needs to be weighed against this loss of social status which affects both them and their families. As a poor widow said to Hunt (1983: 27) in Bangladesh: I was married when I was 13. As long as my husband was alive, I never went out of the house. When my husband died, I still didn’t go out and I wore the burqa. I had some help from relatives. But as my children grew, they needed more to eat and what little | got was no longer enough. I then decided I had to go out and work to earn money. They said you will lose your self-respect, we will make an outcast of you. But I didn’t care. It was a matter of the stomach. I couldn’t worry about self-respect any more.

The decision was clearly a painful one. The strictures on women’s visibility, mobility and behaviour, whether internalized by women or imposed on them by threat of gossip, reprimand or violence, impinge directly on their autonomy and ability to claim and

control land. The degree of restriction and its adverse implications are greatest where purdah is most explicitly and strongly advocated. First, purdah affects women’s overall development, including their access to education. In strict purdah-practising communities, even households that can afford to forego the girl child’s labour withdraw her from school before puberty, or never send her at all. This, coupled with the generally low value placed on female education, makes for extremely low literacy rates in the northern part of the subcontinent, the 1981 rates being 6 and 15 per cent in Pakistan and Bangladesh respectively for rural females over fifteen. (By contrast, literacy rates are significantly higher in south India (especially Kerala) and in Sri Lanka. In Sri Lanka, 79 per cent of rural females over fifteen were literate in 1981.) Illiteracy compounds women’s disadvantage in gaining access to legal and other information, filing property claims, or dealing with public institutions. Also, as noted in chapter 6, it leaves them vulnerable to being cheated, such as by signing away their land rights on documents they cannot read.

Whose land? Who commands? 305 Second, purdah adversely affects women’s self-confidence in dealing with

the non-domestic sphere. A widow in northern India, suddenly forced to survive alone, put it graphically: If a woman wants to cultivate her piece of land or wants wages to be increased or to make an official enquiry, she has to talk to men which she has never done before. She cannot argue. She cannot bargain ... All these days you have never come out in the

open to talk to any men except your father, brother or husband, and that too occasionally, how can you suddenly lift your eyes and start discussing anything with a man? (Cited in Dreze 1990: 86)

Third, purdah restricts women’s participation in activities outside the home, including working in the fields, supervising cultivation, or interacting in the market place. To directly manage land by hiring labourers, or even to ensure that she is not cheated by a tenant, a woman would usually need to move about alone in the village or between villages if the land is located elsewhere.?° In purdah-practising communities, such as among the

Rajputs of Khalapur, ‘even if she is a fairly old woman, this always stimulates gossip, and her affinal relatives are annoyed because it damages

the family reputation’ (Minturn and Hitchcock 1966: 28). Bangladeshi village women who move out of the homestead into what is seen as ‘male space’ are considered to be both provocative and offensive (Abdullah and Zeidenstein 1982). And in a recent study in West Bengal, 80 per cent of the rural Muslim women interviewed said purdah was a major hindrance to their taking up employment (Jehangir 1991: 111).

Fourth, purdah affects a woman’s familiarity with the outside world, especially that beyond the village. Many women of the Hindu community in Sind (Pakistan) studied by Young (1984: 255) told her that ‘they did not know what the other side of the sand-dunes looked like’, and ‘few knew in what direction all but the nearest villages lay’. Most young married women

in the Jat-Sikh community of an Indian Punjab village similarly told Murray that ‘they [did] not know the lay of the village, because they [had]

seldom been in its lanes’. Some had only seen the location of visiting neighbours’ houses ‘by being shown these from their own rooftops’ (Murray 1984: 269). These may be relatively extreme examples, but they are not atypical in their reflection of the bounded nature of women’s existence

under seclusion. This is not to say that women do not move out of their homes, but that the paths they traverse are strictly defined and not open to exploration. They are especially likely to be unfamiliar with the village and its environs where the village is large and they have only entered it as brides. 20 Ofcourse a wealthy woman landowner could employ a manager to supervise her estate: for instance, in chapter 3 we noted cases of sonless Hindu widows in Bengal who had inherited limited interests in their husbands’ estates in the late nineteenth century, and whose estates were being looked after by managers, while the women themselves were living in purdah. But for most women this would not be an economically feasible option.

306 A field of one’s own This limited familiarity and mobility can also restrict women’s access to credit and agricultural inputs 1n both direct and indirect ways. For instance,

credit and input cooperatives situated in the urban centres are rendered relatively inaccessible to women who are unfamiliar with bus routes and forms of urban interaction, and are illiterate in addition. Several poor widows with whom I spoke in Kithoor village (Rajasthan), described a visit to the nearest town by themselves as a traumatic event. A few also said: ‘Ifa woman travels out of the village too often on her own, they say she roams around, that she is a loose woman.’ At the same time, many of them said they find it difficult to get loans within the village: “The moneylender often refuses to lend to us, but men can get credit more easily since they can find some wage work, if necessary by migrating, to repay the debt.’ Contacts that men develop socially and in the market place are critical to their ability to obtain production inputs and labour and to solicit reciprocal help from fellow farmers. It is through such contacts that arrangements are

made, bargains struck, and information exchanged. Women, restricted from speaking to male strangers either by direct strictures or by fear that aspersions will be cast on their character, and excluded from the market place, are strongly disadvantaged in seeking information on new agricultural practices, purchasing inputs, hiring labour, leasing in land from nonrelatives or leasing it out to them, selling their produce, and so on.?! All these factors can make male mediation imperative. The stricter the norms of purdah, the greater the need for this mediation. However, the mediators cannot be just any men, but only those with whom the woman’s interaction is deemed socially acceptable. A poor widow in Kithoor village

said to me: ‘If my brother-in-law helps me, people insinuate we have a sexual relationship.’ In Baluchistan, the NWFP, and Bangladesh, husbands and sons usually undertake all marketing activities.2? In Bangladesh,

even women’s transacting with house-to-house traders is disapproved socially: Harder (1981) found that women in only 13 per cent of her sample of 497 households interacted directly with tradesmen; the rest depended on

the mediation of children or household men. Abdullah and Zeidenstein (1982: 58-9) graphically summarize the constraints that strict purdah can impose on village women in Bangladesh: Maintenance of purdah, the behaviour society values and enforces, means that women cannot have access to the world that lies beyond the imposed physical boundaries of their mobility except through intermediaries — young children for small matters, husbands, fathers, brothers and grown sons for whatever they need that comes from outside. They do not go to the marketplace which is the center of 21 Also see Arens and Van Beurden (1977), Harder (1981), and Kabeer (1985) for Bangladesh. 22 For Bangladesh, see Abdullah and Zeidenstein (1982) and Cain et a/. (1979). For Pakistan, see Pastner (1978) for Baluchistan, and Ahmed (1980) and Lindholm (1982) for the NWFP.

Whose land? Who commands? 307 economic, social and political activity ... They do not go to the mosque, the center of religious and social activity. They do not go to the fields, the accepted center of agricultural activity. They do not go to school past puberty, even if they can afford it, if it involves being with males or walking beyond permissible boundaries. They do not have direct access to the products of their labor nor the chance to labor when in need. They do not go to the Union or Thana where medical and family planning services are available. They do not have access to the courts. They cannot see the families to whom they send their daughters in marriage ... Women without men simply cannot get their money’s worth or their rights.

This paints perhaps too dark a picture, since not all rural Bangladeshi women, at all times in their life cycles, are so constrained, but it does help to identify the ideological barriers against which women have to struggle, and the conflicts inherent in the imposition of such a restrictive set of norms in the context of a rapidly changing economic reality. At the same time, the variability of purdah norms between communities, classes, regions, and historical periods, suggests a degree of flexibility and a potential for the norms to be challenged and changed. To some extent, this

already appears to be happening in countries such as Bangladesh where extreme economic deprivation and the struggle for physical survival are forcing more and more women not only to seek off-bari work, but also, in

the process, to question the legitimacy of strict seclusion (as will be elaborated in chapter 9). The obstacles described are less acute in non-purdah contexts, or where purdah norms are more flexible. Women in Nepal, Sri Lanka, south India,

and among tribal communities in India would therefore have greater freedom to assert their claims and to control and self-manage land. But this freedom is still not equal to that enjoyed by the men of their classes and communities. Indeed, many aspects of the cultural construction of appropriate female behaviour are not confined to purdah-practising communi-

ties. Even Tibeto-Burman women of Nepal, who enjoy considerable freedom of movement and are significant and visible participants in all types of economic activity, including agriculture and trading, are not free

from the more subtle aspects of gendered behaviour patterns. These impinge, among other things, on women’s ability to assert their rights, including property claims within the family. As an illustration, March’s (1988: 19-20) description of the response of a Tamang woman, Nhanu, when her family property was being divided is revealing. She had left an expensive bronze drinking bowl, purchased from the profits of a trading expedition she’d made, in her parents’ house. After her father’s death, when the brothers were dividing the family property, she watched the fate of the bowl and described the event in the following words: I sat there quietly, without saying a word, just sitting and watching as they each took their separate shares of the family property.

308 A field of one’s own {Whispering] The bronze drinking bowl that I had bought that time in Kathmandu was given out in my younger brother’s — Busru’s Father’s — share. Well! While they were dividing the shares, I thought to myself, ‘Oh dear! My bronze

drinking bowl, the one I bought from the efforts of my trips to Kerong and Kathmandu, has been given out in Busru’s Father’s share!’ But I continued to sit there quietly. [Loudly] Then well! my second younger brother came up to get his share. He said, ‘That bronze drinking bowl must be given to Elder Sister! That’s the one she bought with the gallon measure of salt she was given after going to Kerong! The only thing that she bought from that salt was that bronze drinking bowl; that bowl’s hers! She

didn’t waste even one paisa on that trip...’

And then, right then!, he reached out and in a single sweep of his arm, Lo! he grabbed that bronze drinking bowl back and set it in a separate pile for me. Since he spoke up, they gave it to me and IJ took that bronze drinking bowl away with me — [laughing] ...?°

March (1988: 20) remarks: ‘Nhanu could have spoken up to claim her bowl, but instead she waited to see whether or not her rights would be remembered by her brothers themselves.’ March interprets Nhanu’s silence as a form of testing ‘the limits of [her] rights’ in the family. While such an interpretation adds a new and subtle dimension to the language of silence, silence is not necessarily a very effective way of affirming one’s rights: without her younger brother’s mediation, Nhanu may well have lost the

bowl. Her silence contrasts with the volubility of her brothers, and underlines accepted and expected differences in male and female behaviour

even in communities where women are not explicitly constrained from asserting themselves. The gendering of behaviour patterns similarly affects women’s interactions outside the home. For instance, in most south Indian Hindu commu-

nities, there is neither formal veiling nor any explicit rule mandating the

physical confinement of women. Yet behavioural norms can have a confining effect. The importance placed on female chastity is widespread, as

are cultural constructions of femininity which discourage women from engaging in the public bargaining and the assertive wheeling and dealing

that often mark lucrative marketplace transactions. Fisherwomen in Kerala, who find it necessary to sing or joke in order to attract customers in an increasingly competitive market, have been subjected to beatings from men in the community, who choose to interpret their behaviour as sexual

soliciting.2* More generally, the haggling, aggressiveness, and loudness associated with fish trading 1s looked down upon by the fisherwomen’s 73 In the above quotations, the insertions are as in the original. ?4 See the longer version of the documentary film Hidden Hands, Unheard Voices: Women in Indian Agriculture, directed by Rahul Roy and Saba Dewan (Roy and Dewan 1988).

Whose land? Who commands? 309 young educated daughters, who summarily reject such behaviour as ‘masculine’ (Ram 1989). Again, women retail traders in Madras city, seeking to keep within the

bounds of respectability, have adapted their mode of operation even though this involves an economic cost, as Lessinger’s (1989) study reveals.

Lessinger notes that the central wholesale market from which retailers procure their supplies is largely a ‘male space’ where few women, other than

prostitutes, typically go. As a result, urban poor women who depend on retail trading for a livelihood usually avoid going there altogether, instead buying their supplies from the larger male retailers of their own markets, at higher prices that cut into their slender profit margins. The few who venture into the wholesale markets do so in groups, avoiding the pre-dawn auction rush when the best bargains can be obtained. Also, women operate only as retailers of petty items, and within their specific retail area build up kin-like relations with other (especially male) traders. This provides them with a nominal chaperonage and shield in their interactions with ‘outsiders’ — male customers, market tax collectors, and moneylenders. But they are reluctant

to venture outside the immediate retail market area, such as to the wholesale markets, where no such social chaperonage is available. Nor do

they take advantage of contract-supplying in bulk, which is one of the routes to accumulating investible capital, since to obtain goods at concessional rates requires the building up of close contacts with the wholesaler. Any attempt to do so would leave the women open to aspersions of sexual immorality. For the same reason, if they have no male kin support, women do not hire male helpers who could enable them to function more efficiently and perhaps expand their businesses, since to do so would earn them a bad

name. Likewise, women’s dependence on social ties, protection, and chaperonage within their familiar trade markets makes it difficult for them

to relocate their trade if their market collapses, or to take advantage of expanding and more prosperous market locations. As a result, even in the retail market, women operating without adult male kin are amongst the poorest. It is also a telling point that nubile daughters are kept away from the retail shops for fear of tainting their reputations, even in families where the mother trades alone and critically needs an extra pair of hands. And in families that become prosperous on the basis of an initial equal participation in trading by both spouses, the wife often withdraws from active work for status considerations. This last response has also been noted in the rural context. Women in the villages of Karnataka (in south India), for instance, have been found to

withdraw from visible work in the fields with increasing agricultural prosperity just as they do in the purdah-practising north, since in the southern states also doing manual work outside the home is associated with

310 A field of one’s own lower social status.2° In other words, even in the absence of veiling or any explicit gender segregation of space, a preoccupation with the purity and chastity of women and the family’s social status tends to define appropriate female behaviour in ways which restrict women socially and, in indirect ways, also physically.?° The threat of male violence compounds the constraints already set by

social norms. During fieldwork in Janakpur village (Chitwan district, Nepal), for instance, Enslin (1990: 169-71) found that the women belonging to a local women’s organization were afraid to hold their meetings in certain public spaces such as the village panchayat, and the teashop and

bazaar areas where men drank and gambled. They felt that ‘bad men [would] come and make trouble’, and they described several incidents of sexual harassment when they moved about in the village, especially after dark. Indeed the threat of male violence has global resonance for women: it is often the primary way by which many public spaces, particularly at night, are appropriated by men even in western societies. The lurking shadows of

midnight keep women away from the streets as much in New York as in New Delhi! Constraints such as these which adversely affect their ability to function

as independent farmers are shared in greater or lesser degree by women across South Asia. Also shared across the region is another dimension of gendered social norms, namely the domestic division of labour, especially women’s primary (and usually sole) responsibility for childcare. This can

particularly restrict women in regions of high fertility, such as in the northern parts of the subcontinent. Some childcare responsibility could of course be delegated to other women, if there is a joint family, or to older siblings. But this still would not free the mother of central responsibility for 25 See e.g. Epstein’s (1973) study of two Mysore villages; also see Agarwal (1984). Underlying the social status associated with different types of work are not merely the economic returns from such work but also a complex set of attitudes which are not entirely coincident with thé economic, such as attitudes towards manual vs. mental labour, rural vs. urban location,

tasks done with machines vs. those done by hand, jobs requiring various levels of skills/ _ education vs. the unskilled (or so labelled), and so on. 20 It is revealing to compare similar contexts in Europe, where too the ideological division of public space into ‘male’ and ‘other’ has historically restricted women’s mobility and public interactions. For instance, Thomas Hardy’s description, in his novel Far from the Madding Crowd, of the stir caused among the male farmers by Bathsheba Everdene’s visit to the

cornmarket as an independent woman farmer, in nineteenth-century rural England, highlights the prevalence of a notion of ‘male’ space (that women were expected to avoid). This notion, in its essence, was not dissimilar to that discussed here in the context of village

India. Similarly, in 1844 when the House of Commons was built in England, it was with great difficulty that a Ladies’ Gallery was sanctioned. A compromise solution was finally reached: a grille was put up to screen the female occupants from the public gaze. This grille was removed only in 1918 (Altekar 1956: 178).

Whose land? Who commands? 311 the children’s care in the way that men are typically freed. If a woman’s farm is located at a distance, or in another village, this factor can constitute a serious constraint to her self-managing the farm.

(2) Post-marital residence: village exogamy and patrilocality In many Hindu communities of northern India, as we have seen, intravillage marriages are forbidden and village exogamy and long-distance marriages are the norm. But even among communities where village endogamy is allowed, as for instance among Muslims, tribals, and Hindus of south and northeast India, a certain proportion of marriages still take place outside the village; and the norm of patrilocality (or virilocality) ensures that it is the women who leave the village.27 When a woman inherits as a daughter while residing in a distant village, this poses difficulties for her

not only in claiming her legal rights (as noted), but also in managing the

farm. To begin with, there may be resentment from her natal kin. Hershman (1981: 76) anticipated that if a Punjabi woman, married in another village, returned with her husband to take over her father’s estate, she would face considerable opposition from her father’s collateral kin and ‘blood would no doubt be shed’. In addition, there are practical difficulties in managing land when residing at a distance. Even where veiling is not practised, women’s primary responsibility for childcare and housework would restrict their mobility between villages. The greater the distance, the more would be the difficulty. Where village exogamy coincides with purdah, the problems are compounded. This would be true for Muslim women married outside their parental villages and also for upper-caste Hindu women in northern India married at considerable distances from their natal homes. Indeed, Luschinsky (1963) notes that when the 1956 Hindu Succession Act was passed,

rural women 1n Madhya Pradesh found it hard to even conceive of a daughter as heir, for (they asked) how would she be able to manage the land

in her father’s village while living in her in-laws’ village? A woman in Rajasthan put it to me graphically: ‘where would I take the land, even if my brother parted with it?’

(3) Male control over labour and technology Typically village women tend to have less command over the labour of relatives than do men, since they usually cannot provide reciprocal labour or favours in the ways men can. Robinson (1968: 422) relates a telling case 27 For a detailed cross-community and cross-regional mapping, see chapter 8.

312 A field of one’s own of a Sinhalese widow, living in her natal village in Sri Lanka, who could not get help for cultivating her land either from her father or her half-brother, while the assistance rendered by her full brother was too little to make much difference: ‘It appears that there is no method of cooperative labour which can be used to help someone perpetually unable to pay.’ Although labour could in theory be hired, the noted difficulties women face when functioning in village markets place them at a considerable disadvantage in comparison with male farmers. Similarly, women’s higher illiteracy levels, their limited access to cash and to markets for purchasing inputs, and gender (along with class) biases in extension services, all become constraints to their self-managing land by limiting their access to production technology. These factors particularly restrict female heads of households who have no male relatives for market mediation. In general, the importance of this mediation has increased with

the shifts in crop technology from traditional to ‘modern’. Traditionally, women of farming households who participated in cultivation often had an extensive knowledge of indigenous seeds and farming techniques, and such seeds could be selected and stored for use by each household. In chapter 1 we noted that Garo women knew of some 300 indigenously cultivated rice varieties and the men always deferred to the women on this count. In Nepal women do the seed selection work among virtually all agricultural communities (Acharya and Bennett 1981). However, high-yielding variety seeds that are now extensively in use are developed on seed farms by specialized

agencies, and new ones have to be purchased every two-to-three years. These seeds, along with chemical fertilizers and an assured water supply that form the “Green Revolution’ technology ‘package’, require access to cash or credit, on which count women in general, and poor women in particular, are seriously disadvantaged. The Green Revolution has also amplified the role of agricultural extension agents in transferring the new technology and practices from the research stations to the cultivators. Such knowledge is typically transferred to male heads of households, in large part because the extension agents are usually men who do not see women as worthy targets for agricultural extension work, not only in purdah societies such as Bangladesh (Goetz 1990) but even in Sri Lanka (Kilkelly 1986). Additionally, ina purdah context male agents do not have easy access to the women farmers, and female agents are more difficult to recruit. But it is the taboo against women ploughing, found in most cultures, and,

to my knowledge, certainly in all communities of South Asia, which presents perhaps the biggest obstacle. Ploughing occupies a central place in intensive agriculture. Male monopoly over the plough is believed by some

scholars to date back to neolithic times, and to have been one of the significant factors that eroded the monopoly women historically enjoyed in

Whose land? Who commands? 313 cereal production among hunting/gathering societies (see e.g. Childe 1942).

Although it appears unlikely that there would have been a simple causal relationship between the advent of the plough per se and the decline in women’s role in agriculture, what appears undisputed is that while field preparation with the hoe has normally been done by women, ploughing has typically been done by men. According to Childe (1942), even in the oldest Sumerian and Egyptian documents, those who ploughed were always men, although the plough itself is said to have been developed from women’s digging sticks (Allaby 1977). Male control over both female labour and surplus production appears to have been facilitated by men’s prior control

over pastoralism and stock breeding (Childe 1942) and to have been entrenched subsequently by strong ideological control and by instituting punishments for transgressions. In India today, some communities (for example, the Oraon tribals of Bihar) believe that if a woman were to plough, there would be no rain, and calamity would follow (Dasgupta and Maiti 1986). Himachali men told U. Sharma (1980) that God had decreed that women should not plough. When women in dire circumstances have ploughed family land, they have often been severely punished. An illustrative case is that of a Bihari tribal woman, with a bed-ridden husband, who was unable to get help from neighbours for ploughing the family field and tried in desperation to do so herself. Within an hour or two of her starting, she was forcibly stopped by the villagers, and a village council was convened which decided to punish her by yoking her to

the plough along with one bullock, and forcing her to plough the village headman’s field for an hour in this way (Dasgupta and Maiti 1986). Ho women in Bihar, if seen to touch the plough even accidentally, are heavily fined by the tribal council and, in rare cases, even stoned to death (Kishwar 1987).

This taboo makes dependence on men unavoidable under settled cultivation, and severely constrains women’s ability to farm independently. Poor

female-headed households are placed in a particular quandary. As U. Sharma (1980: 114) notes: ‘It is at ploughing time that Durgi complained most bitterly of her widowhood; no-one was prepared to plough her fields for her without being paid, and even those who would do it for pay would

only do it after they had completed their own ploughing.’ I found that tractor owners in Kithoor village demanded advance or immediate cash payment for ploughing the fields of poor widows. One widow told me: ‘A man doesn’t face this problem because it is assumed that he will be able to work and repay.’ Delays in ploughing adversely affect crop yields, which are linked to timely field preparation. The justification often given for exclusive male control over the plough is

that ploughing is a heavy operation which women lack the strength to

314 A field of one’s own handle. Yet young boys, by no means always stronger than their mothers, are inducted into it at an early age: in Bangladesh, twelve-year-olds are taught to plough (Cain 1980), and in Uttar Pradesh (India), lower caste

boys learn to plough and thresh grain at eight to twelve years of age (Luschinsky 1962: 241). The real reason for women’s exclusion clearly lies elsewhere than in the ‘heaviness’ of the operation. I would like to suggest

that.a possible reason why men have sought to establish exclusive male

control over ploughing is that it serves to assert male claims over the agricultural surplus. Control over ploughing means control over an operation that is usually critical for good yields (and surplus production) under settled intensive cultivation; and it provides the ideological justifica-

tion for male right over that produce. The age-old analogy of sexual reproduction is often invoked in this regard, in which the woman is symbolized as the field, the man as the seed, and the produce (children, grain) is seen as belonging to the one who sows the seed.*® Here ‘sowing’ the

seed would be not the literal placing of the seed in the soil, which women often undertake (although some groups forbid even this), but preparing the ground for sowing by ploughing, which only men are allowed to do. It isa telling point that in many potter, weaver, and fishing communities in India,

women are barred from touching the very production technologies on which the livelihoods of these communities depend, namely the potter’s wheel, the loom, and the fishing net.?° Historically women are believed to have been the first potters, but once the potter’s wheel was developed this activity too became exclusively male (Childe 1942). The persistent nature of such taboos warrants further exploration.

In sum, so far we have seen that women’s ability to claim as well as control and self-manage land is likely to be a function of a number of factors, some of which appear to be uniform across all regions, such as the

taboo on female ploughing, while others vary cross-regionally, such as purdah practices, norms of post-marital residence, and female illiteracy rates. The next chapter will collate data on the geographic incidence of these

and other variables to provide pointers on regional variations in the difficulties women are likely to face in exercising their land rights and in functioning as independent farmers. In all regions, though, for women to enjoy rights in land comparable to those of men, many material and ideological changes will be necessary. But

it needs emphasis that the severity of the constraints women face in 28 For a discussion on this analogy, see, for instance, Dube (1986). 29 See e.g. Ram (1989) on taboos on women’s use of fishing nets. Among the Ho tribals of Bihar, women cannot touch bows and arrows either (Sachchidananda 1968).

Whose land? Who commands? 315 controlling and managing their land cannot justify depriving them of their claims. Rather (as will be elaborated in chapter 10) the situation calls for institutional support to increase women’s access to inputs and technology. It also calls for support systems to strengthen women’s ability to challenge the social norms that restrict their autonomous functioning, as indeed some gender-progressive organizations (which will be described in chapter 9) are today seeking to do.

8 Tracing cross-regional diversities

More than one student of India, confronted by the variety of its regional languages and cultures, has compared the subcontinent, in this respect, to

the whole of Europe. (Bhatt 1980: 43) A description can give but a generalized picture of a type of social conduct which is ever changing and it is necessary to understand the variety and mode of the changes which are found in each ... region... to understand

well the implications of a social structure. (Karve 1965: 378)

Woven through the discussion so far has been the argument that there are

marked geographic variations in the incidence and strength of factors which affect women’s ability to claim and control land. Here I will seek to systematically examine these variations, drawing upon a number of crossregional tables and maps that I have constructed. This is meant to provide a broad regional gradation of the degree of difficulty women are likely to face in realizing their inheritance claims in arable land and in exercising control over its management. In addition, the cross-regional presentation is meant to serve two purposes: one, to make the general point that there is a marked diversity in women’s situation across South Asia, and so to contradict the

excessive generalizations about the status of South Asian women that proliferate in the literature; and two, to share with other scholars my ethnographic information base which they could draw upon and perhaps use to answer questions not addressed in this book. However, what economic, sociological, and even ecological factors have operated historically to produce the noted geographic differences, especially in cultural practices, is a complex and contentious question which I do not attempt to address here. Indeed, in the absence of detailed cross-regional historical information on the range of variables examined, any answer about ‘origins’ must necessarily be highly speculative.

Below I will first summarize the arguments made earlier about why certain factors are significant in determining women’s ability to exercise their rights in land, and discuss the information sources used to trace the regional variations in these factors. I will then focus on each factor separately, offering in conclusion an overview identifying broad geographic zones that stand out when the variables are examined together. 316

Tracing cross-regional diversities 317 I. Some hypotheses The factors which appear important in determining women’s ability to claim and control land fall into two broad categories: —Social, economic, and demographic: such as post-marital residence, especially village exogamy/endogamy practices and distance from the natal village; close-kin, including cross-cousin, marriages; purdah practices; other forms of control over female

sexuality, as reflected in the extent of social tolerance for divorce, divorcee and widow remarriage, pre-marital sex and

adultery; female labour force participation rates; and total fertility rates;! and —Land-specific: such as land/person ratios, inequalities in the ownership of agricultural land, and the percentage of land under village commons and forests. Let us recall why each of these factors is likely to be significant. Post-marital residence. This variable has two aspects: (a) the type of residence: virilocal, uxorilocal, duolocal, and so on, and (b) the spatial location of the residence, that is, whether it is within the village or outside it, and if outside then at what distance from the woman’s natal home.

Both aspects were noted to be significant for two reasons. First, a daughter’s post-marital residence impinges directly on the degree of control

her natal family can exercise over the land she inherits. In chapter 3, we observed a close correspondence between a daughter’s customary inheri-

tance claims in land and her post-marital residence: in communities customarily practising matrilineal or bilateral inheritance, a daughter’s rights in land were associated with matrilocality/uxorilocality and occasionally with duolocality or neolocality, but not with patrilocality/virilocality. In matrilineal communities where the woman went to live with her husband and his matrilineal kin after marriage, as among the north Kerala Nayars, she did so for the duration of her marriage, returning to her natal home on divorce or widowhood; the land meanwhile was controlled and managed by

the karanavan, the seniormost male in her maternal home. Among the Kandyan Sinhalese, residence and inheritance were quite explicitly linked: it was the binna-married daughter (with an uxorilocally resident husband) who had the right to a share in the parents’ landed property, while the digamarried (virilocally resident) daughter forfeited that right. Even in traditionally patrilineal Hindu communities, the inheritance claim of a daughter in | The total fertility rate represents the number of live children on average that a woman would bear if she were to live to the end of her childbearing years and bear children at each age in accordance with prevailing age-specific fertility rates (see World Bank 1992: 297-8).

318 A field of one’s own a sonless family was linked to her remaining in the natal home with an inresident husband. Contemporary law gives women inheritance rights in land among most communities; but there is an inherent conflict between these laws and the prevailing norms of patrilocal post-marital residence followed by traditionally patrilineal communities. This conflict would be especially acute where patrilocality is linked with village exogamy, since a mere shift by the woman to the husband’s home within the same village is likely to be less problematic than a shift to a different and distant village. Hence among the Kandyan Sinhalese, even daughters living in the husband’s home sometimes inherited parental land as long as they were resident in the natal village and could look after their old parents, but their chances of inheritance were low in

cases of village exogamy. Moreover (as discussed in chapters 6 and 7), village exogamy presents a major practical constraint in claiming and controlling parental land, especially where the marital village is far from the natal home.

We can thus hypothesize that in regions where village exogamy and especially long-distance marriages are the norm, women as daughters are likely to face considerable hostility from their natal families as well as practical difficulties in claiming their shares in land and retaining and selfmanaging them. I will focus on village exogamy and residence distance, rather than patrilocality per se, since those variables (for the reasons stated) would have a greater predictive potential. Marriages within five miles of the

woman’s natal home are defined as ‘near’. This is about the maximum distance that a woman could reasonably cover on foot if she wished to visit

a farm in her natal village for purposes of supervision and return to her marital village on the same day. Marriage in the range of five to fifteen miles will be defined as ‘medium’-distance, and those over fifteen miles as ‘far’.

Within the medium-distance range, marriages closer to five miles will be termed medium-near and those closer to fifteen as medium-far. Since the village is taken as the unit, villages will be characterized according to the category (near, medium or far) in which over 50 per cent of village women’s marriages fall. Close-kin, including cross-cousin, marriages. In communities where

daughters are allowed to marry one or more category of close-kin, including cross-cousins, there is a greater likelihood of land remaining in the hands of the natal family. Hence in such communities we would expect less opposition to daughters inheriting land or being given use rights or gifts in land, than in those where all forms of close-kin marriages are forbidden. Similarly we would expect less opposition to widows inheriting where levirate is practised.

Tracing cross-regional diversities 319 Purdah practices. In regions where the physical seclusion of women is practised, women are likely to be particularly constrained in both claiming and controlling land. Seclusion practices would also adversely affect women’s labour force participation and literacy levels.

Control over female sexuality. Strictures on women’s sexual behaviour include constraints on pre-marital and post-marital sexual alliances, on freedom to initiate divorce, and on divorcee and widow remarriage. These strictures, unlike purdah, do not prescribe or necessitate

women’s physical seclusion. But they represent an important aspect of social control, constitute part of the cultural construction of gendered behaviour and, like purdah, restrict women’s interaction with men. The greater is this form of control, the greater would be the constraints on women’s ability to claim and effectively manage land.

Rural female labour force participation rate (RFLFPR). This can serve as a proxy for a number of factors, such as: —the degrees of physical and economic visibility of women’s work, which affect social perceptions about women’s productive contributions to the household and to the economy. The greater is this visibility, the greater will be the likelihood of a woman’s

needs being taken into account within the family and of her being able to claim social legitimacy in exercising her rights in land;

—the extent of women’s familiarity with their physical environment and the likelihood of their having some practical experience of farming operations; and —the extent to which women can be physically mobile, which impinges on their ability to assert their rights directly, or to join

women’s grassroots groups to fight for these rights along with other women in similar circumstances (of which more in chapter 9). Rural female literacy rate (RFLR). Illiteracy is likely to adversely affect women’s ability to claim as well as control land in many ways. For instance, it can limit women’s knowledge of laws and legal rights, their ability to deal with administrative and legal procedures in relation to land claims, their access to information on new agricultural technologies and practices, their physical mobility (facilitated, among other things, by the ability to read signboards), their overall self-confidence, and their intrahousehold bargaining power and autonomy in decision-making, including in fertility decisions. Regions of high literacy are therefore likely to be more

320 A field of one’s own conducive to women successfully claiming and controlling landed property than regions of low literacy. Literacy has been measured here for the age group fifteen years and above.

Total fertility rate (TFR). This gives us an indication of the average number of live births that women in different parts of the subcontinent tend to have during their reproductive lives. We would expect that the greater the number of births, the more time a woman will spend in pregnancy, lactation, and childcare (assuming the children survive), and

the greater will be the constraints on her physical mobility and overall ability to control and manage land. Land scarcity. Land/person ratios, inter-household inequalities in land ownership (including landlessness), and the availability of village commons and forests are all variables which, in different ways, serve to measure the extent to which land-dependent livelihoods are being squeezed in a region. Land-person ratios indicate the pressure of population on land in average terms. However, the more unequal is the distribution of private land, the less is available to the majority of households for subsistence. Also, the less the availability of non-private land (such as State forests and village commons), the greater the dependence on private land and the more the economic pressure on poor households. Taken together, we would expect that the lower the land-person ratio and availability of non-privatized land, the greater the landlessness, and the

higher the degree of inequality among those owning land (this last as measured by Gini coefficients), the greater would be the economic pressure on large sections of land-dependent populations. As subsistence possibilities get squeezed, this could lead to an increase in land conflicts not only between households but also between the genders, for at least three reasons.

First, traditional kinship support systems, including the support that brothers provide, would tend to get eroded, pushing women to opt for more direct ways of securing their future such as claiming their shares. Second,

husbands and sons would put greater pressure on women to assert their

rights in parental land. Third, there would be greater hostility from brothers towards sisters inheriting and from husbands’ relatives towards widows inheriting. * A Gini coefficient is a statistical measure of inequality in a given distribution, which ranges in value between 0 and I. Itis used here to measure land concentration in two distributions: (a) landowning households and (b) all rural households, including landed and landless, and

should be seen only as a broad indicator.

Tracing cross-regional diversities 321 By examining the regional variations in these factors we can identify, at least in broad terms, the regions in which women are likely to face most hostility from their relatives in exercising their rights. Consider now the available information base.

II. Information sources I have drawn primarily on two kinds of information sources: ethnographies and macro-surveys. Large-scale surveys — the census and others — have the

advantage of providing quantitative data on some variables for all the countries under study, such as on the land-use variables, female labour force participation rates, literacy rates, and total fertility rates. But macrosurveys give little or no information on many of the social variables; for these we thus have to depend on ethnographic evidence. Ethnographies have been used here for information on the norms and practices of postmarital residence,* close-kin marriages, control over female sexuality, and

purdah practices. For regions for which ethnographic information on purdah is not available, I have drawn upon my personal observations as well as the observations of people (especially anthropologists) familiar with the social practices of those areas.* In addition I looked at the percentage of

Muslim and tribal populations in the different Indian states. Although Islam prescribes female seclusion, the percentage of Muslim population ina

‘region in India is at best a partial indicator of the likelihood of purdah practice in that area, since in some states with sizable proportions of Muslims in their populations, such as Kashmir and Kerala, the rural Muslims practise purdah in very limited degree. The percentage of scheduled tribe population, however, is a more consistent indicator, in that tribal ‘populations in India do not practise purdah. For assessing purdah practices by state in India I have thus utilized all the above types of information.

On close-kin marriages, often communities permit them with some categories of kin while forbidding them with others. However, not all ethnographies give details of which categories are permitted and which forbidden: typically only the preferred categories are indicated. Where available, such information has been incorporated in the tables. But for our 3 For India alone it would have been possible to use census data on marriage migration and

migrants enumerated by their place of birth to obtain some estimates of territorial endogamy and marriage distance, as Libbee (1980) does from the 1961 census and the formulation of a mathematical model. But to my knowledge comparable data are not available for all five countries. I therefore decided to rely solely on the ethnographies. * It may be recalled that in chapter 7 we distinguished between purdah societies and other societies on the basis of whether or not there were explicit social strictures limiting women’s physical mobility and interaction with men.

322 A field of one’s own purposes here the main distinction is between communities which forbid all forms of close-kin alliances and those which permit at least some forms of them.

The ethnographies also have some general limitations (a few of which were briefly mentioned in chapter 1): ——There is a regional clustering of available studies: we find several

studies for some regions and none for others. Pakistan Punjab,

the states of Uttar Pradesh, Karnataka, and Kerala in India, and the middle-hills of Nepal are relatively well-studied, as are certain communities such as the Sinhalese in Sri Lanka or the Nayars 1n Kerala. However, there is a dearth of useful ethnogra-

phies on Sind in Pakistan and on some of the eastern, central, and western states of India. —There is no necessary relationship between the relative importance of a given community in the population of a region and the

availability of ethnographic material on that community. Hence, in some regions, the primary focus of ethnographies is on

the tribal groups even though such groups constitute only a small part of those regions’ populations.

—Not all ethnographies provide information on each of the variables mentioned. Information on levirate, divorce, and divorcee and widow remarriage is especially thin. —Most studies describe a community’s norms but not its actual

practice. For instance, they may say whether cross-cousin marriage or divorce are allowed, but not necessarily their actual

incidence within the village or community studied. (In the discussion therefore I have made a distinction between what is permitted, preferred, and actually practised.) Also they may say that marriages are arranged outside the village, but not always at what distances. Despite these limitations, an illustrative picture can be drawn for the social variables mentioned.*> And where the ethnographies do provide quantitative information (as many do for marriage distance and the incidence of close-kin marriages), this is summarized and presented in the tables. For rural women’s labour force participation and literacy rates across South Asia, I have used the 1981 census data for each of the countries. It is of course now well recognized that RFLFPRs based on data from national

censuses in South Asia underestimate women’s actual participation in * For India, some of the gaps mentioned above may be filled by the information being gathered (but not yet available) under the ‘People of India’ project, launched a few years ago by the Anthropological Survey of India (see the introductory volume by Singh 1992).

Tracing cross-regional diversities 323 economic activity (Agarwal 1985b, Sen and Sen 1985). There are several reasons for this: perceptions in many parts of South Asia (sometimes shared by women themselves) that women’s labour on the family fields or their doing farm-related work within the home compounds is ‘domestic’ rather than ‘productive’ work because it is unwaged and less physically ‘visible’; definitional biases in the census that tend to explicitly or implicitly associate

‘working’ with doing paid work; and cultural values which associate women’s involvement in work outside the home with low social status. Hence a good deal of women’s productive work is not reported as work by male respondents, and often not even by women respondents. However, since an important part of our concern is with the economic and physical visibility of women’s work, the census figures are still helpful, because they

do capture that component of a woman’s work which brings in some income and is done outside the home. The census figures also help, in a rough way, to indicate the physical mobility permitted to women and to measure one aspect of their fall-back position. They are, however, not a comprehensive measure of women’s familiarity with field-related work, since many women who work sporadically on the family fields get excluded.

For India the National Sample Surveys are less prone to these biases, but comparable surveys for other parts of South Asia are not available. The noted biases are minimal in regions where no negative connotation attaches to women doing field-related work as, for instance, among tribal communities in India and among most Nepalese communities. For Pakistan, Bangladesh, Nepal, and Sri Lanka, the 1981 census figures relate to female workers aged ten and above as a percentage of the female population of the same age group. For India, the 1981 census figures relate to female ‘main’ workers aged fifteen and over as a percentage of the female population of that age group, since disaggregated data for female workers between the ages of ten and fifteen are not available.°

Information on total fertility rates is obtained from two sources. For India they are taken from the Sample Registration Scheme of the Census: ° The 1981 Indian Census divides workers into ‘main’ and ‘marginal’ depending on whether or not they have worked for the major part (that is, for over 183 days) of the previous year. The figures for main workers have been used here (rather than for main plus marginal): these are more directly comparable with the estimates for other South Asian countries. Also our concern here, as noted earlier, is to capture the physical and economic visibility of women’s work and women’s physical mobility: these are better indicated by taking only the ‘main’ workers category. The ‘marginal’ workers would also include many women who are involved in work within the home compound, such as looking after family cattle and poultry. Although undeniably this is important to capture if our purpose were to measure women’s economic contribution, aggregating the main and marginal categories is less appropriate here in view of our present concern with women’s ability to claim and manage agricultural land. Of course, even the main workers category is only a very rough pointer for this.

324 A field of one’s own the latest available figures are for 1988 and are separated by rural and urban

areas. For countries other than India, figures for 1988 are taken from the World Bank’s World Development Report (WDR) 1990. These, unfortunately, are not disaggregated by rural and urban sectors. For comparative purposes, therefore, the map is based on the aggregate fertility figures, while the table also gives the rural estimates for India alone. The land-person ratios (population densities) are taken from the World Bank’s WDR 1992 for countries other than India, and from the 1991 census

for India. The information on rural landlessness and land distribution patterns by size class 1s drawn from several different sources, including agricultural censuses and large sample surveys; and the Gini coefficients for

land distribution among the rural landowning households and all rural households have been calculated from these land distribution data. The land ownership data for South Asia are poor in general, but worse in some

countries than others: those for India and Sri Lanka are relatively more reliable than those for Bangladesh, Pakistan, and Nepal. For Bangladesh the estimates made by Jannuzi and Peach (1980) based on a 1977 Land Occupancy Survey have been used, and for Pakistan I have drawn on Khan’s (1981) estimates for 1976, which do not include the province of Baluchistan. There appear to be no direct measures of landlessness for Pakistan, so I have used an indirect estimate cited in Singh (1990). For Nepal, recent information on land distribution appears to be available only for operational holdings, hence for land ownership I have had to rely on the rather-dated 1970 sample survey carried out by the FAO and the Govern-

ment of Nepal to evaluate the land reform programme. Assessing the availability of land under village commons and forests is also problematic. From the available data it is difficult to arrive at comparable assessments for village common land across the five countries. (For India, a state-wise breakdown of such land was given in table Al.1.) Hence the more readily available forest cover data alone have been used. Despite its noted lacunae, however, the available information suffices for a broad assessment of land scarcity. The information on the variables noted has been presented in tables 8.1

to 8.9, appendix tables A8.1 to A8.3, and in maps 8.1 to 8.10. Before discussing these below, three caveats need mention. First, in maps 8.1 to 8.3

(that is those relating to village endogamy, close-kin marriages and purdah), for some of the states, tracing the pattern has necessitated making ‘heroic’ generalizations on the basis of very few ethnographies. I have done so to enable a quick visual comparison of the variables. A reader interested in the detailed ethnographic evidence could, however, refer to the tables.

Regions for which I could locate no ethnographic or other evidence are

indicated by a question mark. Second, the hypotheses set out at the

Tracing cross-regional diversities 325 beginning of the chapter cannot be statistically tested, since there is very little information on what could be seen as the dependent variables, viz. land ownership and control by gender in different regions; or the degree of resistance from kin that women seeking to exercise their legal claims are likely to face. There is clearly a case here for macro-surveys to gather such information, especially gender-disaggregated data on land ownership. In India, for instance, land ownership surveys at the household level are periodically undertaken under the government’s National Sample Survey scheme, and a gender-wise disaggregation could be incorporated within that ongoing system. Third, by emphasizing the broad regional patterns in my discussion

below, I do not mean to suggest that there is uniformity within those regions. Indeed we would expect variations within a given state/province especially by caste and class, and (for some variables) also by districts. However, limitations of available information do not allow me to disaggre-

gate further by class/caste categories within each region, although the information that does exist on this count has been incorporated in the tables. Unless otherwise specified, the ethnographic information for India relates to non-Muslims, and mostly to Hindus, while that for Pakistan and Bangladesh relates to Muslims.

IIT. The cross-regional patterns’ (1) Marriage location and post-marital residence (see tables 8.1, 8.2, A8.1 and map 8.1)

In the northern part of the subcontinent, most women live patrilocally or

virilocally after marriage, while in south and northeast India and Sri Lanka, post-marital residence varies considerably, with cases of all forms of residence: patrilocal/virilocal, matrilocal/uxorilocal, ambilocal, neolocal, and avunculocal. However, as noted earlier, our concern here is more with village endogamy/exogamy and marriage distance.

Village endogamy is permitted among all Muslims in South Asia and among all communities in Sri Lanka and Nepal (with the exception of

upper-caste Brahmin and Chetri Hindus). Indeed there is a marked preference for village endogamy in Pakistan, Nepal, and Sri Lanka. However, Bangladeshis prefer to marry outside the natal village, although ’ As mentioned in chapter 6, the terms ‘northern India’ and ‘north India’ are used here only when a very broad comparison of the northern and southern (viz. the peninsular) parts of the country is intended, and would roughly include the northwestern, western, central, and eastern states, but exclude the northeastern (mainly tribal) states. Typically, however, the finer six-fold geographic division is used in the discussion.

326 A field of one’s own Table 8.1: Village endogamy norms in South Asia

Allowed and Strongly

Region accepted disapproved Not allowed INDIA Northwest

Haryana and — — Lewis (1958), Sharma

Delhi (1973), Freed and Freed (1976)

Himachal P. — — Newell (1970), Parry (1979), U.Sharma (1980) Kashmir — Madan (1989)! —

Punjab — — Hershman (1981), Leaf (1972), Nag (1960), Pettigrew (1975), U.Sharma (1980)

Rajasthan Carstairs (1954), Mandelbaum (1968) Chauhan (1967), Plunkett

Gupta (1974)3 (1973), Personal observation

Uttar P. Bhandari (1963)4 Berreman (1970)5 Gould (1960), Hu (1955), Luschinsky (1962), Majumdar (1954, 1955), Marriott (1955),

MacDorman (1987), Minturn and Hitchcock (1966), Sharma (1973), Singh (1970), Vatuk (1975), Wadley (1976) West and Central

Gujarat ~~ Chen (1990),° Fukutake e7 al. (1964) [all

(1963) castes], Pocock (1972), || Haekel Goody (1990)

Maharashtra Chapekar (1960)’ Laxminarayana — Malhotra (1980), (1968),

Rao and Chowdhury — Orenstein (1965) (1988)

Madhya P. Yadav (1970),® Haekel (1963), — Jay (1970)° Jacobson (1970), Mathur (1964), Mayer (1960) East

Bihar —— Sachchidananda Gallagher (1965), (1968) Standing (1987)

en

Tracing cross-regional diversities 327

Table 8.1: (cont.)

——— —eeeee——e—eeEe—————eeeeeeeee™e™==Qmmmmonaooeeooooeews

Allowed and Strongly

Region accepted disapproved Not allowed

Orissa Bailey (1957),° — — GOI (1965b,d,e),'° GOI (1967c,e)!°

W.Bengal Fukutake, ef al. Nicholas (1961), Fukutake et al. (1964)

(1964) {lower Klass (1966) [upper caste],

caste/tribe] Kolenda (1983)

Northeast

Arunachal P. Sarkar (1977)

Assam Cantlie (1984) Manipur Chaki-Sircar | (1984)

Meghalaya!! Agarwal (1990b), Majumdar (1978)

Mizoram GOI (1966a)!° Nagaland Furer-Haimendorf (1976)

Tripura GOI (1966b,c)!° South

Andhra P. Furer-Haimendorf (1979), Tyler (1970)

Karnataka Claus (1975), Dillon (1955), Epstein (1962), Harper (1971), Hill (1982), Ishwaran (1968), Laxminarayana (1968), Srinivas (1965)

Kerala Gough (196la, 1973), Mencher (1962, 1965)

Tamil Nadu Beck (1972) Good (1981) Kapadia (1990) Silvertsen (1963) BANGLADESH All regions and

communities!

328 A field of one’s own Table 8.1: (cont.)

Allowed and Strongly

Region accepted disapproved Not allowed NEPAL

All regions and Acharya (1981) Bennett (1983, and

communities [Maithili] personal communication)

except Chetri, {Chetri and Brahmin] Brahmin and Maithili

PAKISTAN

All regions and communities! 3

SRI LANKA All regions and communities Notes: ' This relates only to the Hindus; village endogamy is allowed among the Muslims. ? Relates to Bhils, a tribal group. 3 Relates to all castes in a village near Madhya Pradesh which is therefore likely to have been influenced culturally by the lesser insistence on village exogamy in that region. * Tribal group, Korwas. > Relates to the ‘pahari’ communities living in the lower foothills of the Himalayas. © Personal communication on the basis of her fieldwork in Gujarat in 1987. This is only true of lower castes; the upper castes practise stricter village exogamy. 7 Relates to a hill tribe. 8 Relates to the Gonds, a tribal community. ° Relates to the Kond community. '0 These are village surveys undertaken by the Government of India as a supplement to the 1961 census. In all such surveys mentioned in the table the noted communites are tribal, except in GOI (1965d) where the population surveyed was entirely Hindu. '! Primarily matrilineal tribal communities. '2 Hindus in Bangladesh can also marry within the village (Aziz 1979). '3] found no information on this for Hindus in Pakistan.

in a nearby one. Among [Indian Hindus, there is a marked regional variation, especially between the northwest, and the northeast and south.

The central and eastern states come in between the northwestern and southern. In northwest India, marriages among Hindus are almost always outside the natal village, village endogamy being forbidden among virtually all caste groups and especially the upper castes. The exceptions, as seen

from table 8.1, are few and relate primarily to tribal groups, some hill communities, and Rajasthani communities bordering central India which were probably influenced by the lesser emphasis on village exogamy norms

pone a | i Ee

|| Tracing cross-regional diversities 329

Ca, * ~ =n 7) th i :| 7oie (a. fa YoSF rtCO CT | & _. ~~» i ayySe ‘ =e

| oe } 7 eg ee: i

i Source: Table 8.1 eo Map 8.1 Village endogamy norms

in the latter region.®? Some groups, such as the Jats in villages near Delhi, forbid marriage into any village which shares even a border with the natal

one (Sharma 1973) or in which other clans of one’s village are well represented (Lewis 1958).

Often, the preferred direction in which the marital village should lie is also specified.? The ecology of the region, among other things, appears to have something to do with this. For instance, in Kangra district (Himachal Pradesh), villagers prefer to marry daughters westwards where the more fertile land and prosperous villages lie, rather than eastwards wherein lies increasingly inaccessible and barren hill country, although the villagers themselves justify this by arguing that people eastwards are less refined and 8 Also see Chauhan (1967) on this. ° Several studies for Uttar Pradesh and Himachal Pradesh note this: Lewis (1958), Kolenda (1983), Marriott (1955), Minturn and Hitchcock (1966), Newell (1970), and Parry (1979).

330 A field of one’s own civilized! (Parry 1979: 219-20). Similarly, in Chamba district, also in Himachal Pradesh, Newell (1970) finds a clear preference for marrying daughters down the valley into villages with less land hunger and a less physically arduous life, and taking (presumably hardier) daughters-in-law

from higher up (Newell 1970). | In northeast and south India, by contrast, there is a marked preference for

in-village marriage, and village endogamy is never forbidden. Indeed, the

Assamese have an appropriate proverb for the maximum appropriate marriage distance: ‘A girl within a day by road, a cow within shouting distance’ (Cantlie 1984: 57). In the western, central and eastern states, the picture is a mixed one: the lower castes often allow village endogamy, while

the upper castes often forbid it outright or allow it in rare cases, but disapprovingly. Information on the extent to which village endogamy is practised among communities which permit it is more limited but suffices to provide a broad

picture, as summarized in table A8.1. The table suggests that practice follows stated preferences: in Pakistan, Nepal, and Sri Lanka, the incidence of intra-village marriages is high, and in Bangladesh and much of India it is low. Gould (1960, 1961) identifies four factors underlying village exogamy in

north India: caste endogamy, territorial stabilization of kin groups, gotra exogamy, and the tendency to regard affinal and consanguineal kinship ties

as mutually exclusive in the patrilineal kinship system and so avoid conflicting claims.'!° This last factor, he argues, is what differentiates the northern Indian pattern from the southern one. Gould’s explanation helps illuminate the dominant north/south contrast, but Berreman (1962) points out that the four factors listed by Gould are not sufficient conditions and village endogamy may still occur in their presence, as it does among the Paharis of Garhwal in Uttar Pradesh whom Berreman studied. An explicit prohibition of village endogamy may therefore be necessary to entirely prevent it, as is in fact done by most northern Indian communities listed in table 8.1. Also factors such as the desire to create a geographically wide network of alliances for political or other reasons are likely to affect the decision to marry outside the village, as found by Parry (1979) in Himachal Pradesh, and noted below for Bangladesh. Regional variations in close-kin marriage preferences are also relevant to the geographic patterns of kin and

political networks; and their implications for women’s experience of marriage, and ability to control land, will be elaborated later in the chapter. 10 Spatial separation from the bride’s family, for instance, reduces interference from them on behalf of their daughter. The desire to avoid friction between the bride’s and groom’s families is often expressed as an important reason for endorsing village exogamy: see Mayer (1960) for Madhya Pradesh and MacDorman (1987) for Uttar Pradesh.

Tracing cross-regional diversities 331 On marriage distance, the contrasts between regions are less sharp, as seen from tables 8.2 and A8.1 (which give actual practice). In Nepal, Pakistan, and Sri Lanka, as we might expect, given the preference for village

endogamy, marriages outside the village tend to be within a five-mile | radius. Among the Tibeto-Burman communities of Nepal, the distances are small enough to be covered in an hour or two by foot. In fact, among the Limbus and Gurungs women normally do not move to the husband’s home immediately after marriage, but stay on with their parents usually till the first child is born and sometimes longer (Jones 1977; Andors 1976). In Bangladesh, however, although most marriages are within a five-mile

radius, there are also some at distances of over fifteen miles. Unlike Muslims in Pakistan, Bangladeshi Muslims prefer to spread their network

of kin and acquaintances through marriage alliances geographically: people argue that within their village they already know each other and therefore do not need to use marriages for this purpose (Aziz 1979). In India, although the south and northeast still contrast with the rest of the country, the differences in marriage distance are less marked than those

found for village exogamy/endogamy. In the south and northeast, marriages are almost always close to the natal home, within a five-mile radius or at most in the medium-near range. In the rest of the country the pattern is a mixed one. Certain caste groups strongly prefer long-distance marriages: the Rajasthani Rajputs, for instance, sometimes marry at distances of over

sixty miles. However, Jats and other middle castes in Uttar Pradesh and Rajasthan in India tend to marry closer to the natal village. In general, the bias in northwest India is towards distant marriage alliances, especially among the upper castes.

Libbee’s (1980) mapping of territorial endogamy and of marriage distance across rural India highlighted regional patterns similar to those described here. He found high village endogamy in south India and low endogamy in the north, as well as a ‘striking clustering of large marriage distances in the northwest, in Rajasthan, western Madhya Pradesh, Punjab, and western Uttar Pradesh’ (Libbee 1980: 93). The overall pattern for South Asia is therefore as follows. In Pakistan, Nepal, Sri Lanka, and south and northeast India, women are likely to be married either within or very close to their natal villages. In the rest of India women usually move to other villages, sometimes to nearby ones, but more

often (especially in the northwest) to distant ones. Bangladesh comes inbetween: village endogamy is allowed but marriage distances range from near to far. These post-marital residence patterns suggest that women in northwest India and Bangladesh are likely to have a particularly difficult time laying claim to and managing land inherited from parents.

_2 aSo ~ 3 os = 3 ~ > ps £s A S ~~, o 3S et wm OS -o=~ .eyare wee” —_

— ~~ m fo

n~— Ee

Ok wyant~anN 8~ mm ~ OS _N- Kea An= OS “sx m~ “nw wo ON“~~ CN ™ Cc oe ~OH OO oS ~~ a.@aé& =a DH t= 5 0 Qo e veel ~ —= -€— go © & Pi mn >seg eeasH#a EE 2 YY =a S& be — =f SG & Ot ©= .im>2>2SY 855 oS € | 20O0R 2zrz Ss a

< > tn pa = Oax3a —— e a ~ os oO So S So oe & gs Baa a fo = oe ~~ wo — = n faa) x = ow % ¢ oY z 2 os Sy a i!fan =o a7: se O>, == some Sang som oO 3 = nr

BS} Pang

A i se DSS % SS ° a so > A _ _—= % —_ ran) oS + ©pom 2) Sw- tn "TO oe

S12 Ss a =>> £742 ~2 “aio 2 Bes SER

= pitti, 8 = SSA = S Rw FAT S tN (~ =o ~~ Te —, a

8

a= RQ

S em

~ oF Wy)

8 Gg aS > 3 0.=Se3§ Nee” A. | neuen | woe”

SZe 3Om@ -SfsO 5an1| Z= a |=a 3smamse™ Svas) cS 2= >

Ss

3So ond

=

~.

—N ~ wnt c AQ. < 60 2 os a po is % 3 sles § So £26 on Ss > ¢ 2 3S =S = 8 oS 2 a > = Ss

a) mlIALre EUS s>bang 23 g cs eB) Se 8s C+ ome xs = 2s = Ss ~ 6) Lb y|/Ze2t AtTxwae =S>OO

-— ove) lo @)

ON wee”

g 3

mm)

a

ws

=}

w >a ae)

Ss af wel , ry

Ss

& oo

« OO

Oo 2 aS & _ S ¢ en —_~ A a S os NA ~ Hn ON ee RY a2 = :=&nNumD 77) QSoxa=o—e= 3Z| 422 a3 ”)ae J,YO fad xo en §) _-—_

2 00

rool rs) qQ O tn Oo a wn) + “1s§& Soe 3°= an e2 © eae Ser =m vay ASz Os .©@ Aso.aON om

= ‘oO . —™™O am SS swt ~OO No) —~ we &@a“4 ~~ +& So wn 3&

~s Cs acgr wnTt on) OAme .S emt rer ar~@ Ge=R= ~A oh z=TT = 0 crDR EOS 8aa™rOS& ask ew 5 92

& A SG =545 =x2 Oo BD nT SD 2 xs soeam 2ek &—~ +2ee35 om.~65 wta2 etwy 2een = 23 Oo = 2352 32 3 & com = Ss &oA 4 ossY9—— a =om OKre a xs S66¢s ao. nSSR? 3s ng@22r2 © oo oS O £= 65 B83 frst

~ Q2 2Ow oom Ss bs we} oom =—_ NN S 3 5 8 3 @ qs 3 5 ie

77)

Yeas = o=eS&S —_ O= |=oO a2 wa com fae eg = Q : ~ = & Ss} ZE = -ascoy 8 ~ S & o& rev A) a 7 = $B 8 VY a 3 MZ 8 oOEeSmo go aa i.sé5 z2¢qac8 Z. Aw

a ~~ Cn

vom fo)~~ =

~ STN E

as N «a Mo; tn “"~ wy

fa Pom

ee x EO — 3— oO o >Oooo=

=2 Z.

F meaten

©

e| SS ®a

ment fee yom,

vw mos

e| 38 2° 3| OH12c 12)

> saabe =>

ER ~- ~aeed —BEA‘S Om -s ty ‘mene=wn - Off>- . “5 5v¥6 .-4 5

S A2282EN¥5SBE SGES a wo ODD < & L w2 Seen Ses =

R RScIassK FHEesEse = =fon we ot OO) TU = wt FB \O ~~ yale ar ZQktAS g ~“fORM 37 7a

= s~~ &RO9oSee) aa Ao& Y9 Psa TT con aw} Of & sel&Sosan

4 >” 28> - oe oO CF Nee Oo 2

| | 32 ESS Rssu. wooE aos a = ol Oo Q “— 3Ss §6

E=

a:

Sm cond

E nn

wm

OD -= RZ =9 Se

3o .o

ax = ooOo” t “SS _ 3sve £28 358f $m SRT E —~ FS ad Oo Dus 40 Sele8 235 #4382 225 28] 238 €8 Y2Bo653t EE Eu.3h S& 6E 2ER 3 >, sgoad $28=SEES £ == Ell o oo t=

AAOMAH = 2S OC med fo me OarCome Ea Sowe 3edg 2s Q 6h ,.V ond nm NS 3 ag Oo wn — Pa= ® SES — ES as °on7i)Sea Aan, 5 &. Egome 20© =

os a. -Beds ELS s 23s & ae a Xs o 2>ane za2a 2 2H o's 3 ss t 5 Sie S La) [7e) 3 2 ae nO = 2-4 Qa sic38 Ec, DAE agocS — te Sw

~3s FF DS v

“Soe 3s os Eve

Ux SY © neRES oOo w zy 8 eoxesZezote

72) s Y ON = © a) < Ze 6g Sc < Zz. me oc oom Ym -E OTe, < Zz. - 8 2. re) — ral + ~~S — ‘s 2i © 3+ES3 08 ae3 S)

wa S&S aw

(Come meen,

>

nt = &os 72) > =— go > -— 2& —_ © va i To) Oo ~ LE By aS o ON o @ = Aon ‘a Sl ° —~ goo ean —_ Oo # = 8 im, pon Sa oa es cm «=Y oe ~ fan) nal ac oO © 5 ER DBD BO S — Be ESS =~ EY 3a =i — 23 ON =om eo) sé Eana~ Sos ~ w@ O re) O mm x aay 3 6 ~ “a a O "Ss om = oS CN an bs o ty a aw S&S Ce A & oO~~~~ S Sf 3)0aem 8 1. 8& is) "SS — O fn ~ co ~~ a=e==~oy Ze o>CD=SAK Sainnow cs Oo8 OD ss oO eet ONcon Oo o->s O= Pasom A~@6208 s:32S90a0R2 6 oma

O im‘=ms & mn EGESS O =a SToO on —~ TS aoteek AGRPSyHeFRAaQal Q SO —~e ~SAsaAnm ~~ SD s&Us=— GRRT TRS a hROTS Kalk hes eo ” ArTcSY%eaefooce ss Sees geevs 3 —_— & oS wo so o0wM — sS Ss cc ~o | ~aeveS SEs tSwe ee aes SSE BM SSE ra] Q S2SRmap EV oayw of Ss = SN Oo =a ZSEZSZESAGPM POS 's =“Asee Peeve Ss 8s -'o 5 a. 253 = i So & =Se2r ZB on | A AgxaoseLBSODELELSOO aDOD & A ae] = oo * 2 A & Nn wor & 82 ZC. * *# * | *S | osoom O& a oO om wut S o ov, ob _— 2 Secg w”Su ”n . Ss a we

n is! Egat Oo & -_ Sand fue Gon or WwW — » & 2 §s 83ao >5 — 6x » 2 vary Las o —~ 3 ‘= a x 2g SS = & Yy —< nee ay A rs§ o EgNLS &>5,08 5 Sgerte2ogzs £ © = UO Fr Ss si 8ess7s B88 SPs ase SAS eMac ete&etsSee, _ a Ow CS

~ o Oo S os : S rie

ae CRA eee Be 8 +m ee E S6e=55= * — Pp ibar: a eee st seks Phe ses eens esas .»«2peepee eS ESSER EEE SES 4x Seonie Gas c eee is eegesszES is

So - OoES 23>AG ~~ 62°” te BEV mS VSHtsS SsnS

ere2sUsed aV29

3oO Zzuntsooes EnS T= > 23s=%rrzl=cag sleegi-s Som eVvV oe

SED @ ei CE eel FSZg lB Se ovVS e253 ae moVEOUS ome o .- fF Y © SSsoSSenESs & BAMON= RH

a en A) ‘> = Bi aViyvesy - +F FF SS£ 272,222 /Ve Sf Flees ey2Soeruns

BoED SSE x c zee TU SHOE Oh osOCVvVasy: & ¢ && 3s0AD 9 — = .¥- ae WaVY oOo - ca &

2Es a5 Sr Qe & ES V| Gs+,= 8k OoOn Saez > Ot 6 a] & co o7rSse8e

@> Ss 2&man 89st 5OU§ ~~ BA OO G2 &~mure85 -sGel 7 SO .mGees Zs sz 8 ao Mel ck ageAef ful teaoh oa s oa 32 oO ie) Maa ae Mi Ao ® ywoa&8 SMe Zan tn OFOwD

i

ti| f|

372 A field of one’s own $

groaetavenrengcoreads BPO Re Oye BE OP PIEE EGO DEH ELER ES OTE RG USO DEO Aft SEO IES LTE EES OPE DED SOR STEIOC LG ENTE BEE OER OPE MIE DIO OPE Pl 4 BFA OY 0 AEs O98 DPE DPE REE EP ODP EEL O SEEDED ELEY PID BERBERS 2 BPS POMP AEDT ORS G OIROPE Be DYE RD HH 0 APS MGS ESt seneveneeneenechtg

Traditional forms of tand inheritance

;

5 ey , w 3

}

; > “Pinot senshi re ol : id b,laog ee ee

"y Rare if wpeen PIF Gy f Py, 2 OK i| CS he,bape, ’ ‘LAL an |al . ) i| 3 : : MEPS PIOLS y

et ef i t : "sebael gf i H =Ly “A § 2 yo ‘ f ad

;| *, r : ee a

i; 53

} | | | | 35

i | : z eadende oer t ; is aes } po} Patriineal so, “yet of i vy ee ; sity fo eorgh onfoeGy, Bilateral ’ Tu yp , Matriineal

_ COURP ERLE DEED FC DLE CIE RES BIRO BERL A BPERLE OPE EEREIEDIE OP ERPROPERECEPE EHS OF EG BEL OLE OLE GLE DLERPE OPT OFECIEOPERLECLEOP CELE DPSEIL EG LED EOE RTS EPEELEOIEOPECIREIE EFS OLY OIE EPERPE RPE ELEDP LOS R OPECLE OIE DLE DIE OIL CPE SIERPEDLD veucund

Pee eee ee ee ee ee ee ee ee ee ee ee eee ee eee ee eee et

Village endagamy norms

f| GE: GE

H LE MendGa ET i>, EE EEN in. if oncith

ee, “7 i CeEE YD Lh, dF f ye ty, yy a Lo Cr OC, ai i Se en BEE eee

i mae” Ly ee WE EES i‘ 7eee PESi ME eeas hg ESET tes lll ™“y le OE es ane fi)

‘| “,BIE ; 1 Mixed Se | Vk

‘i wren , & LEG PO We ES Pon GeBp es ee” va ee ‘f Rae SR hy GY eat Mae.

i ee LL ee eee i

gy {_ Mostly tarhidden

Source: Table 8.4 ~ 7 Mostly allowed |

*«

AP PEAT OD POD CPG CHE OLE ATE OPE HOE TLD APD TLD TPE RES OP OE FE OPE EE MLE OIE HE OPE AIT CLL OPE PEE APE OPE RLEEPIE BP CEG OEE EELEES RFE RLE RIE OLE REM CPL OP EAE BPERPE DLE APY PLE RPE SEY DIS OIE OPER C2 OPECEL CLE TID SLY DPD SPD SPE RFE OEE OPE Oe OEE CEE SHE CPOE tie eoene$

Maps 8.10 A comparative perspective

|

j44}i;i}i;ee|a32>‘;:i:ii ii;re: t i t;j$ zi;’ i”t

j{fj EEE teT > : t santtins j FE EEeea1 i;

* eecross-regional .ee rracing diversities 373

P70 NTE ee ore en ete oFe CFunte nee ote stn FY 800 ay O78 gn Ot8 tte ate nde OE SFO NdE RES Ot RE WIN DTS OTE aE NTE Cre noe sra ate ote O0e ltente dures 910 ty rte nte ete wie >E8 UE HOU CPE ILE HEE OPO HOE REALE MtE hte DIO PLO SLO AG 280 LEI SLE RDD ALO OLY OLO NTO 9G tt SLE BED EOO OG,

Close-kin marriage norms ¢

?

$5 CHGS. NL, KE M ‘7 3 PE OeEETRL Wame, sOAL:

iH Se Ma hs ; :f3ES IN ip, BO De : Wt a i EEE. ~~” ; », AE IN i te,EY ME oNae:: i EOD SE ee% piggies ; SE EEE 5 be cc ee eR H GE : ELeee me: a ee BY UA LIE fo ii ;4ji VEDI iUESee La | i PERE BS t IES Me GTEE ; GE hes Ss tl : BLE I UTE ? 9G PRE ER

TEES 7 ROPE REP IT REI RELL DP

H DEES ES IA a7ahengne. OE ieLAME LEILA eg pity NheZered LO tn ass Me F;:i :, LEYLA Ep te ying YEents Gee et Nie, oe : OL EGTA GsPANN FEN VIELE tlMLL, Eee LEE Z LILIES LE SIM IEGED OL COLtelecin, EE LSE IT eyWy NR EEE $ A TET LG Cee TEEELlelDOPE boy AM : mie = Se :

ELLE OPEL EE goa ee TEbing Tai Mee 5} §4HOPEL LL py pps Leleictias (ai heHi ODL ORR ET. Bee, SR hy af,Me i opie LY GI ee : ie 7 ; aes ee a ll } Ly On in % He ee Reishinciht om Ses ee OE H

iiiiEE Ce” 5 TE EE i TE TLES, EEE BEIGE : em ; MGI:F $;§ Wag Ga i£ gor”

ERR RG HLIORI LORE EE Ey 5 e Boe oie We,AERIAL EEE RNY MD ee tis neeSgtees‘age EEE GEie::i GRY a. AY SORE 2; RH iT IRS Bes ee?5z 3i SES ee ; RORY ws Bae SED Ree) aS 1 tan SO Be BS RY 0, OME ER ped FR dane A

i "4 ee LS Les Li LE pata EE ae ;

GOLESI f LE AID PUEDE © qz ; SET EE GE EY Se TYE Pf LOLI Ae Ba a iia igi ; SO CE i

iHLOGI ss, prrcrceoeeseny i GORE H ;; OME DE TLL .SOS EN Cafes Lesenacvoet Forbidden

a4EO EET | vererevene he, NfVite edie zF£ e¢ ORIOL ibe F §H . TE pa (tithe VAS i Source: Table 8.3’ Ge Ta ee Mostly allowed i UiCe: labie &. ~ =e fee, (ROSH anow é

bee OU RPEREE SOE AE HOT PEAR TIE SIU OIE EES APE OPE TP EID EMG OLE FO TPE IE SOY RPR ELD EPO OE HPD HPD OOD PPD OLE EPS OED SPE OLD EDR ODE HY BIUH ES OPE EOD BHO SOG OPE OHO OPE O08 TPE OSE OPO OIE OL0 O06 OPE OPE OEE SD OR OPE OLE OPT OPE OPE ete the ets ope cocare teste apy scesearonert

i £

Purdah practices i z

$ Mi AEE ‘ } ASE EE TE, H >

UL. REE, MP cage s Oe RR rs Wass EEE. ; gee oe eg j 4}iii CL Ne EE URE EE 5 4 EE GRR hes ee ; Fhe WEEE CE fis: Ny RY Sie POR i: +

CLL eee t Stn ee EEE ESS }Hit EGE Re RT eT a RIE i A ON ee Re, ~ mel H IE A Be EL OD ty nh iis fs “ae A 7 OEE IY SIE GL IS Pea ioscan eat ? :

i:itt :;*y, ;.Phe on i ; $ iifog ° ji Ee TET EE IE SASS, Ee ott DE *£, ’ vr, i

‘i UES lag IEE ESee ip ie al Ep Po dy ~h iH POE: ee Re a Gan eee i a Me ie Hs ES at i oR gs BE H er Kies PET ieee ets Lis SNES (SRR Oy LIER sie SSO i CMe Leg ge ee:j

H; PILE ELLIE NEE icing ONE Ee A Cea Te ae tN E IE SEE ELLE IELTS Eta pat 4 z

atin lates ERS MESSE (AGI i SEC: Eh ob oa A

i2;;z; ;q4pas bis ? i wh} het I i3 5 REE of 9 i yt EF fit gf { , fk ee 5 raed ie 4 7 i wed pa § 1 Pa ; H % yevsennnseen § r} be, ¢ ‘ f ‘HHi : von § : f y 4 Mu 8 Loman OW OF HORE H Pa ae aed ; f 4 a yy* » , f 4 tages Medium CTSai Eb EE BE YsRRO bigs Mite Be Ee GEESE IM, H I ‘paste pat SORES» RRR fe od

: wy Fie a a . PARR ph FO ie : ES TI, BRST O ALU AERO OE i OS RLS ¢ ERE IRL, Sl PERL SILO RTE LPT T RODS PRI ERR Or OR oT wiles '

, . Source: ‘ > LOLS .i Table 8.4 ISS we titi |Nigh i

i

ian apesocamu ase sovesnios CPERFLOPEOPURIA OPE CAEEIERPLOLED PEOPLE REEOPIOPE CPL OSERPT CIF OAA EAL OPER SEIT OEE SE GEESE TT O TEE ASE ede TSONGA THEE LO LTe tHe tHe Mee ode MS Osie thunder rea tle eee MheMie edu Ady the ase the tsetse tde ese ste ede sonsdurtectestasdested

Maps 8.10 (cont.)

374 A field of one’s) own PO Menem eRe EE OL I ALAA A ELE A EI A EAT

i bd bd LA H

é

j

Rural female labour force participation rates (1981)

|

; j Mins i Se i EE ME ; i4 Le Lue LE aE MM ag Ve Me IE ty REESE. 3 i Ce aye ;;

‘ rel tes i ag 3 wre ee ‘ ~ 4 , a oth io ; 8 ; Si = Pr Ie é et aaa , aa ee Pies WE geo 5 asalen ‘ate ““~ “Opp VF e & ae Sm, A wySeMY | Chey OF ce OS LN SON Sar J PES

nr cs NU ik aR A ST EN RRC & Lo { Mt act ctit ghRRO ee A aediedagt LETS : att Ge

ce” YS ee ‘ %

A EE RR antl

H a pen H |AVee) 4010 < 60% | ¢ 42WN Ff ISOS

.

eo tnt bess than au%

i ee eee ELL] 2010 < 40%

’ he Paes. : 3)

i«.

Source: Table 8.5 olen ick 60% and more

$

berrePrenennrn ramen sgn niatdebtsbeePTenydOd nt Ope AD) eb ant ane rey uATAPIAPIAATERGEPEURY > ALO DELAECEURA OLE LRIEIDOLGURLENT DAE LEEDDERDLERTLR OWI DTEAEEAT ONES AER RIEDEE IONE DORE HEE NEAR LEPLO TER RED CLO HEE SLD WPLENPEREEDLE DLO FIR SEN MERI SALE ood

| ant POG OPS OPE OEE RPL DPE PERIL OLE OPE OPE ROC ELE WOES IPED PE NTE OPE ONE RPS BPE ANE SpE eKegsapenerapdarinadardeptertontcodapiupdagtasdapdnedertoeteodartenteplaptordaproptaphipdortortuetinterte turdectentervathirterterdandog

f Population density (1981) (persons per sq. kr.)

§;

;3 FVII Ff *a ot Seg oem AltayUME og 4 , i epTL Tie EUR ee eS#> 5 3 ; f nigelh j . PR neti an a) Be, b i, ety, Pr " wy” for 5 rin }

4 be OLE A ila ayrae 4 Nae Nigay: oe Ge joi DY ii$ Dn aie font “Yrat ee ey aki eeeg nsaeon: IEeAll|

cP it5 Te ee Ls $i

we DO EES Peis 3aes ame 73i $$ , pens REnerdTS giea nae Kea . A weeod a oes ‘ “t J Y$ cote ere Bes GE EAS eT ES in Me EEO cen, F hsot tat eer, Chee, Bo Sei OI ee

;4 EeillREE SORE ghe Sites gel4 5j 4 |; Mohs we hy ;ene staan ti lth EOEES, res Ghee eet hemmed LOSS than 150

Pare ey, yy RR LS. Gontsoronteté 156 {6 < 300

Bo ee" BOG ge ene S$

MPU Ue ‘a

“|” Oo4Ue LipsHe, WigCire 300 to < 450 eee Ye, i Source: Table‘ *8.7 aleSees 5, tees «450 and more } 5 nreerenre PIA OPERIA ELA RIE PE OF ERS ECP ELE LV LOPLOPECPR ELECFERPUOIESETOPNCPLOOBCPIO PEELS PLO PTE LEL ILE IE | intendaovercereueenbvuctuptuntaptuntenienearsapeanesotsntsbesptrpernventinearecoespivstsporterserocsecierrectereocemenastesiron ned

Maps 8.10 (cont.)

Tracing cross-regional diversities 375 mee sei ate stertdatdaddntiadd obi atistartd etd edeatdatintestittentivtévthaldpiurdvtivtdsbtdvtivtdstintivedetinnistiabdiatiatinténtined rtd state eee ene ee AA ALLAN LAA

|| BE 4 3

Percent landiess to tatal rural households :

We | ine Uy LE ” :H$ Li, iy oy EG NE! CE ee LE ETE .ES eeyy LEO GAGE : | WS GelgLO YE tA eS ;| ty y uy ae ) Hag Ce tip “ ae "eed i Ds ee ee ;

!; J$ | .i

“ee es ; | SEE SMiew sugos j| BD ee CRE |esij OR Me iy o S Bs hire he é he £% ee i ok ae ner f

GP ee eee, ,

Teefechee boven A or heeet ls | Raa a 10 to < 16%1i

4| i /,. 1] : We — i| 2 20% and more

i Loney f

ercorronte, f Source: Table 8.7yi "*ae.an am, [+] No Info. i

Livan PPO RFARLE CFE CHE OPE BPE OLE OPE DPT BPE BPO CPE CPE ODE OIE OPE CEE CPG PEELE OPE OPO COT OPE CPG OPE OPE smveneaninsccecsersenspracecevcovenvencentsevesesevsoesoraseapropesoesrtenssmpererptresspsensensensersuneseqensapcassssessseespermnenstesnessed

GrereoewntannnaneenasneyAtn tee ete Mt eer ePdCeerOnn D0 bene e eA CLAN AN ACO IG LO A OT OOO ENO OE OO OO ON ALICE AE BFE EGS 196 OE Od APE RTE RPE DPE UPL RPE REEOTE DPE RIE OPE OPE OGY OPE =|

Percent geographic area under forests i

; ‘ 8 ie $ Otnk Same i; ;! Be EE aee ,seh, ante Og fee §; : \ Bo PSS ogee i pee a eg i ow,

PeeDY LE espare ae if BG © j 4 oeprveronatin te Rice LIE oy He j ie SE irpe:7kL EE SE PEED Bigs A an pe ELLEIG CEE EE i OES ai’ Cp: ae

oe ee eB eT Ea

TRIE ERIS” Laat cy BBO Si IS

i Le, CR oon ey f Gis UU i|

Ce ee eee ee H; eatin: WEN sere (ed Less than 10% ;

ae Sabin FE 10 to < 20%

an MUM, 2010 < 40% |

r ‘ Le % hy,8.7 COS. Source: Table ih, 40% and more Ps ys -

Hi

AG REO RIT ELE OPO DLE APE SIS OPEL OSHIP TOE OLS OPE OH AAT OLE PT OES OER PL ABE OPEL OLE APE OBL O PPO IS ORG CIEE PDD PE CPPEDG EDL © BPC LOERE® PPE REE PEE PIEDOER SS BOEEPE PPE RPE PFE RPE RP ERP EDP EDGE DEERE BP SOPL ELE RP LEFERPSOP EPA EEE OPE pcsovercarvenssccarerecaneunseed

Maps 8.10 (cont.)

376 A field of one’s own Attempts at the regional patterning of society and culture in the Indian subcontinent are by no means new, but most have been confined to postindependence India and usually do not cover the other four countries. Such attempts include those by geographers and social anthropologists such as

Sopher (1980), Miller (1981), Mandelbaum (1970), Libbee (1980), and Karve (1965). In the works of these writers, the geographic divisions vary from a two-fold north-south divide to a four-fold one or more, with some variation in the identification of areas included within these divisions. For instance, in Sopher’s (1980) proposed north-south separation of India’s historical-cultural regions, the north essentially relates to the northwestern Gangetic plains, and the south largely to the southeastern and far southern areas. He sees the ancient Vindhya—Narmada axis extending northeastwards asa significant cultural divider, as does Miller (1981). Mandelbaum’s

(1970) north-south division highlights differences in the institutions of marriage and family and the status of women, the south comprising the four southern Dravidian-language states (Kerala, Tamil Nadu, Karnataka, and Andhra Pradesh), and the north the Hindi-speaking belt centred around Uttar Pradesh. Libbee’s (1980) mapping of territorial endogamy and marriage distance again suggests a marked difference between the peninsular south and the northwest; while Karve (1965) discusses the agrarian organization of kinship within a four-zone linguistic-cultural division, with the four southern states comprising one zone and the rest three. In general, there is a fair degree of agreement among scholars in defining the south, and the differences among them relate largely to the rest

of the country, with several areas being left indeterminately characterized.*7

Clearly in any such geographic patterning, much depends on what variables constitute the basis of the divisions. My geographic divisions overlap in some degree with those of earlier scholars on some variables, such as Libbee’s on marriage residence, but not on others. In particular my analysis highlights four aspects. First it indicates the importance of a more

nuanced grading of India than is usually undertaken, by distinguishing within ‘northern’ India between the northwest, west, east, and central, and by taking account of northeast India. Second, it points to the usefulness of a South Asia-related analysis as opposed to only an India-based one, since some patterns which just begin to appear within India are revealed with greater starkness in the neighbouring countries. For instance, the associa-

tion between women’s post-marital residence and their traditional land 27 For additional attempts at defining regions in India, and discussions of the problems therein, see Crane, ed. (1966) and Sopher, ed. (1980). Also see Dyson and Moore’s (1983)

attempt to regionally map kinship structures, female autonomy, and demographic behaviour in India; and the discussion in Goody (1990).

Tracing cross-regional diversities 377 rights within matrilineal and bilateral communities, which is indicated by the Indian case studies, is affirmed by the Sri Lankan ones. Similarly the patterns of women’s disinheritance and the constraints on their autonomy noted in northwest India are found in their extreme form in the North West Frontier region of Pakistan. Again, Bangladesh brings out the contradictions between purdah ideology which confines women within the home, and economic necessity which pushes many out to search for work, in ways that

would not be revealed as starkly through a focus on north India alone. Including Bangladesh also helps to highlight women’s responses and contestations around the issue of purdah, as will be discussed in the next chapter. Third, the analysis brings out the strength of cultural commonalities over religious differences. The patterns in Hindu northwest India are closer to

those of Muslim Pakistan than to those of Hindus in Sri Lanka or south India. Similarly, the communities of Sri Lanka — Buddhist, Muslim, and Hindu — show greater similarity to one another in relation to the mapped variables than to the people of these religious persuasions in the northern part of the subcontinent. The identities sought to be constituted by religious and ethnic fundamentalists today thus often move in contradiction to these historically forged cultural links. Finally, the analysis suggests that favourable social and legal conditions are not sufficient in themselves to guarantee women their land rights. For instance, given that by most of the mentioned indicators, the conditions for

women inheriting and managing land in south India and Sri Lanka are relatively favourable, we would have expected more of them to be doing so

than appears to be the case.2® Although (as we saw in chapter 6) the situation in these regions is better than, say, in the northwest, in that many

women do inherit land in Sri Lanka and Kerala among the traditionally matrilineal and bilateral groups, even here not all who are legally eligible do

so. And in the traditionally patrilineal groups of south India, the gap between law and practice remains marked. In terms of land control, again the gender gap is considerable in all regions, including the south. What then are the missing ingredients? At least three could be suggested. One, there isa

wide gender gap in all the regions, including southern South Asia, in the

role that women play and the bargaining power they have in public decision-making bodies at every level (from village to national bodies):?? this, as noted earlier, both directly and indirectly and in complex ways 28 While this can be surmised from the ethnographies, systematic, region-wise informationgathering on the extent to which women in different regions do inherit and manage land is clearly warranted both in macro-surveys and in village-level ethnographies. 29 For instance, in 1987 only 5.5 per cent of Parliament members from the four south Indian states were women, and In no region was the percentage over [1 (GOI 1988b: 178).

378 A field of one’s own

ensures predominant male ownership and/or control over significant economic resources, including land. Two, women in all the regions are subject to restrictions imposed by the social structuring of appropriate female roles and behaviour and the gendering of public space, although admittedly less so in southern and northeastern South Asia than elsewhere. And three (in my view most importantly), favourable social conditions can complement but not substitute for women themselves taking the initiative in fighting for their land rights, both as individuals and in groups. Indeed

this appears to be the necessary condition for change as much in the southern parts of South Asia as in other parts, although the constraints against which this struggle has to be waged are significantly greater in northern South Asia than elsewhere. In the chapter that follows, we therefore turn to the ways in which women

have been resisting and contesting many of the constraints they face in making their legal rights in land effective.

Pan ~ ~ ~ _~ae *¢@ Ww

SS n +t S — ON w S x 2 ot S = mnt ct 4 = 9 2 sd S eS S 3 3 z 0 3 os a. = © yoo < a ) ae ew = O ) om,

Si) w Q

35 cord

po

% as}

~~ fm

8

av

o Sasa SS A ~~ ~ S as A xt x a QI » > S| 3 _ t 3 S a x7) vo% =}f2(3 o &R| 5S

i)

N na am 7) nm Fi = | fos Sein EDA ao IS EDAA o

Q. an

~

5

o. oD = 2B ‘3 Oo iz = so 7° = ob < 237 »~ a s 2 ao = @ vo || cles = 3 2,8 Ss 3556 = Clme-- & 5 23% & Same79 8

=Gs} mi ORS E SsEse Rese Fs @ etsMOBS Tsse SoS =O iD) pou

puns

rom

w

SS 50 Ya = 2 5 be 5% end 5 aS > © 2 263 S56 O el swe «= = VY Y = sas S22 cs bt | > > ~~ 32s £853 = zg - Ojelm — Sa iF ee enc ESees girs

5

a s >“ ns ofco-

[™ + ~_— io nN oo —_ ‘© oa S cn on ome on) os N 3 5 OS aA ci O mS ron} ba ™ Cy \O Ne 00 +: aaa On Y co to a .-in = ga com —hl ‘tend a lee ory “~ —

f¢ a osa > = Eg am #®SS af Fm

SNM A AS aZooooy Ss MmOenn > ra Tteo © wv) ArtSA°S8z B53 S24 > tent

cS aq A nw oO © So on as D —_ on a S SD ~ 4 + ron] ps res aen2eono, Ce aeeeececy, Seee yt O eococe ;N ~ ow 06 eed —= CaASZ= ore on SSeCI*S, EV SHlSRaSz tTTssez =

w 3) 2 a= Y aS Z 3 > 32 &fs)&2fo) & &~ 2 i+ 9)

7)

2FSe “gsc=8°24 BB Ho 2©3S22 2c= ® a¢VTosg Gs oe” § & Oo os =o) BE = =o Foe zwSA> S268 Oe. San an 5< 7£ SDAA EDA7SDA

yY “-_ oO CH D Co = a A ZL n 2 5 2 3.0 = < @ ~cs 2 2 So a 3S chk 6 oc = Ss o So 2s 3H PEL ac 2a =e
x = Ae) ~ = iS E $$ Md @ Ss Oo 3 ~ 2 9 2 =S 3 es 3 AY re) 3 re) Z os ER = 3 x @

Oo =z 4 2 & ==> 3 (T —

| | | | ce | | | e | | | | | S | | | Fe |

ca | : |

| | | | © | | sei La | | | | 8~| >HS | | £285 55|

¥, 3 ~=aTYB AG

2 OS 2 OU | oS US > | >

i? 9) ee i? 9) Nn

aset 7i= onw2.SY .2525 —ss— aA 1) S.2&So DH ay eg 8 22> y 5~no 2a asU eN = aor) oo =: = i OD 8B & S524 2 ws 5 8 BG So Y SF LU BS (aSEP SUT SSuSS pa TLE Sse stBE VG 83asepa SE SCT SSSE BL a6 z%90 3 a. 36438 © a, 3 =3 ¢ a 4, ~ & 88=SF & 2S535 ON mae Ss $32 Oo. S032 2 hm 2oO's 3 S54 EcBS 8 Ss sO SS Ha EE © Be 88223 3 2838 25 2882 2388 gy

~ ~ F A R ‘Oo ON Oo _— ~ ON ON os ond CN — fe ~~,

om)

wont _" Coane) ned _ ON fas

@a> eat a o& a 3 P< Ala ae =cs=ON ~o O

_ % So = 5.3 N = OE | | > & ar | | NZ. a) 8 m& ££ | | = oO NT oO

=Q ise}

=)

fu

o§ 1 & 1 & e's DH Boa as . So 3 o 2 ~in caer | | “ | | = 0 = 9 ) fe om Oo Oo ; c2if —_ 3 oe 3 s 6 font > v toa yy % 5) = 2 = nmsd om| ‘| om @ Yl « | | 9oL5 ¢“ S

Ee =.

Ca

,

oA

8

OQ

o

—,

wy 0) 2tom | | oo) c > | | ©S 42wa > |5) Oo

@

=~~ &

5=a © omg a)

on

~

we G s S ¢ Ke) SC x N G Ss nw Yr HE BLasB sss ora) 21 29 2g ws zeana xn}£92 = 232 —£P = we = oa Feet

< < = oO 30 TT UO se @ 4 Sc £ 22 sEl exe SF 8 gS §2 £2 S22 SHS

Ke) © quae «om Sos — “3 ro) SS S PEL ESQD S&Z§8 EFnN85200 SZEE

zc RS = mm,

© See OSs . 3 x a Zs & 3 = o) na| >4 a

n=

Yo © iP) .D)

= D n & x 6 8 2 7a) nm < a 5) co) | | y | v9 928 | | Os 9=x | fin, Nn

8}

-an ~ &s Oos »Fe c= oS v aS —,

5a > 22 exsgeB ase FHP 3 2 St

a proneam, SS ny ew ee oN pay = GQ) —~ VY F LL Oo =.Lsco wrYU f= 2S= a6s »Y QO BH O”~ eee 65 ~~ pLSoS 32S es U0 = 8Sea sO este “ 3SO sy,tse rrYS6Zsa f—

= nw ccm..5 OS“ws wow5) Qu |Vi>| nh >> >5) > o = O -©n> an2wo nae 9) ie | | 8e 2 | | | @& 61 ®

-a 3 Sed

& 3S 3 $= ™

= v, ~e” = NN i e BVSsg oe = S S pa w < 3fon 3 a“—, oo 3m O xs

“ ‘wan

5

ee 2f=3ete = Sap So s sew FSB os xOQ BSE|E $28 £2086 = 8 a v 2s = Ss 2 S & So EO 2 § 3 bo E 2 SZ io st a Se cg SS a> 9 oe ~ xe,o|x~ =e =¥Go =& 83h 0 =3

S 4 nN ~ oo ct = A ” & s | ee oe o~7SS x mr a00 0 A 5 RF = &&@ — on nA oO an ai o a) aoO ! = > = 9 oy io ee n S C Ss i.) E gs t — © A st ~ = G » an CN — Come 3 ‘fe o = c me “3 ws bent UO UO — — oS =~ {J ao) ° © mm on cs jaa) I fa in,

I

he 3) ™ wo t oom est (Pome aaa,

am,

a©

R66 OB : 20 22 Sr Ae v >666na&ownyG69”¥©>inY ~G SO mA 5)

ioe)

>

== S) 5 oo >=§ Bue x r) 3% S ie om ~~ on an oa wot Pe > 2 BR S 2 4% ze 55 Bs 3 Sage Oomo st x Soa oO Oo OoSs Gg5a 3& 2 «6 ss == fhe SSs

c >° SS ‘= Fes 2S i26=“ ¢2 Soe _me §& 5 7gepS ?52 PPS ESP Sheae 25% Og233 Se ERB SERGE SE ST ERS 28 O O's 2§ 8.SOO SE&as Stes ¢€ PS BRS ao 3S5yn% SERRE S8osm Q Fein SOY =BSE 3s Sse 6™ OESP ALN REBE!|! cB) ‘—_ a a one ee poi,

WW

Cy

oo ~

Nn ~~ fore 4 . on ote ° .

SUE at ON AN Ho I SEICZ Sa 8 3 Ze SZ ¢ u a £ ON Ne AN 1) jm Nw

per

1]

oo San

few

2. » .& i 2 = oe ZB nm HN

| oom rh Q wv coo) ” Oo 3 po S id==06 o.0 6a§& 5.90

-o beSSi 2 ar se wy 3B58 5 S Qo< oOO° woe £ a, wo

‘
> > Bg) 5 ho 25 gS S| Ss Lot n 2 2 “ee 2 o£ 2 8 or ie s

£| .

Cl c 928 225 225 > ra < i“ 2) 2)

: gg $ 25 6 2gOS 8 & 2 OS$S,éE< ae oO

5 aed

me

= Sb

Ln | | | | | | | | Ss) ~ on ~ = gy 2 Se a) >|2o>33 5Sc BS} g U2 s St ce a

fon

AS)

9

= 1)

~

iB)

00 mY yeOo ©P85 = Gy &sE SIPE.pPo 5s a 2 bm 8s6 nl nabd 3 3s §oO en ar & &§ 2&QO "Oo c= ses o 0 aS > D> s O-= cs eo &28|/2azr0Z00 Ss 7f£E SOS

© &© 66OW & ©> > >oOy ~ >©>A F men’

7)N

= ge = 53os *ZB ome7 some omanwo ~ omyey—“~ AU Qa.no ao —™ 3 .O =m 6'ssO 3sasMY —“ F£=£0 8S sag> S ‘S65 =5Ow 3p =& ?@ =& eae S STB x3Masa oF ‘= 0 fs =csSas SS besA

~3. u&8 3B 3 ho as! 258 Ss 5 3$=oes ESee s3-gtes 38

2235 £3 >ok BETES Shs eT se segs SAS & Sf6ts¥a gs 3Ss= on > aon =.od cfi(« . bdU's ~a SF OoLave FUFE REO.5 EE ZG P8eR

a+

") ” “a 00 ao. a5 S[5S Se. asFr30 S qe SES ar

fA SCRE BAS ES SC fs 73saepe eza2 2a 8& . hs a] CN my 753)NN>cos cry bc UK F fof O oO cr Se OO SA xs 8

7) < cm MM a cwwr _ cZoON UO

g = S Se

= oO ~ No} N Sos —_ Oo + a cn) . nwt eek . gt eat ogo ec |,

A. o = a. =& 2

2S OR ©

2=

Soil 25

a2 nm & ‘— 53 S 5 o

a $ € & Wee 2 2 : Bo 2 = A = . . ; EE & wn rf $>2“5 4;5 od” >

:ro) .— i) - mw o 38«28 ia “WW ©

2%=

~~ e6 As &. 2s

: A,.2 —~ 28 8 | aS bt HAE : 2° 6 oR S25 OO ~ sq

F QE . OG & TO nm um 2 9 RSES Avs

—_. 6 :-— 7) og $6% 824832 we OQ eb see esy. ges OOF ew ee

‘weno’ >3s8S =Y SSE ss cq QO. o . 8 S ae—~ ~~ _Pe SsStSe SEEESSS £3Og ¢€Hu 36 2EF age? 3 fora} S: =325 wn HY ae zZOKS.. aan °2ge| 886 Fs«@ 2a 8% ae $328 33 =S AaOon =m + mE! 6=~re los oO : QO mM Uz

$e o 2 E S as) = : : g| |3| gs §2 =| |Z] 32 EEo3 E Rr

rotE: 8 Sa Eas = aim, s= : -) Ae) ee — sete KR tr CS p : = i 213 22 : = ca »n om ~ ow Y w i S ‘me US 2 5 S fon 2 Sgr = < = 2 wa

Fy: oo = cas 2)Aas S== Ss) ma) 33Ax aS35 2K : eee” bee Fy: =©

: siti, Ss © 2-& 7sgi% bg 7: :S3: © ayoO 4 :) 8

3ms s| 52442 423% ic ileis SE & ES 53 r-% 2 j= 3 3 25¢5 :2

: s2 S222 found

a|| Z|23 3g\se: ra) reat


) =} x = fan Oo on el ¢ 4 ‘o) o1 328 Ss . &§ & &on) x o| ¢ = = eo= o) “ o) ° ° ° 8 QOlus fa) c 6 Ra) Re) O 3 feng

&

ry nl

a

os ced =< n n wn < Py a cs 3 wn < yn

es 6>£|gz. 28 28 c ~ @ 5 6 eS 38 a4 = ra) par) & © S & N o Ww . &N o| @

S

—fos ome —mle _ =CB) Pens se Fan«ont S jon) CF)PW =~ 7& Sew

3

ia feOooO of>$Faer: Pau Pr cn oO e

=) ra

Q1d)2S aas sS — a S a aos =e Tens ~~ ~~ oe ee 5 a e 2a a o) S SHSo ws Ena g; a

2©

— oN paeeey NS Nn ew’ ao OO WL Cc) CO im,

Site aa se S30 3° =-2 oN a Gt” Oo Sus SS Seo a = ec = 3 an % Sao

es ¢ 7 = ra oe g2 = 3s os x

=|oUGQoRraan oo E55 “z we 220 BM,

~~ OH -_ a 0 4 Vo Soe SHA a E oo 24 SS Yo ~~ o ao 2s ES ra io,

w poeed

S av © ao @ Ss .2 ZB re) KS) EE: 3 s= 2 ai2 [NN com 65; SEas26 6onl36607 6 6nn G oe56a oD cons >A Py

ot S

33te3& ro c 8 5 ° ° 3bn F°° fF & *& eEn < 2 BSS =“4 : RE : . na BR 33

Oo 0

am ae 84a ch ih a =& OO oa &

: SEPo a =2 mm —_

cy oO

cS ~~ =nr=Tes On ALYOa Oo ~~ ~~~ceagwo =A DAD

s 4 > a) - © V9 Da 2 oA s

o|