Social Approaches to an Industrial Past
Social Approaches to an Industrial Past
The Archaeology and Anthropology of ...
345 downloads
1412 Views
3MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
Social Approaches to an Industrial Past
Social Approaches to an Industrial Past
The Archaeology and Anthropology of Mining
Edited by A.Bernard Knapp, Vincent C.Pigott and Eugenia W.Herbert
London and New York
First published 1998 by Routledge 11 New Fetter Lane, London EC4P 4EE This edition published in the Taylor & Francis e-Library, 2002. Simultaneously published in the USA and Canada by Routledge 29 West 35th Street, New York, NY 10001 © 1998 A.Bernard Knapp, Vincent C.Pigott and Eugenia W.Herbert, selection and editorial matter; the contributors, the individual chapters All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data A catalogue record for this book has been applied for ISBN 0-415-18150-X (Print Edition) ISBN 0-203-06892-0 Master e-book ISBN ISBN 0-203-21435-8 (Glassbook Format)
To the memory of my coal-mining ancestors, especially my grandfather Harry Everett Davis. A.Bernard Knapp
To those nineteenth-century miners who left their ineradicable mark on the Rocky Mountains of Colorado and provided me from boyhood with a source of endless fascination. Vincent C.Pigott
To Alexander, Jesse, Bethany and Matthew. Eugenia W.Herbert
Contents
List of figures List of tables List of contributors Foreword 1
Social approaches to the archaeology and anthropology of mining A.BERNARD KNAPP
PART I Historical archaeology 2
3
1
25
The fabric and structure of Australian mining settlements PETER BELL
27
Gender and community structure on Australian colonial goldfields SUSAN LAWRENCE
39
4
Bedroom politics: ladies of the night and men of the day ALEXY SIMMONS
5
Power and the industrial mining community in the American West DONALD L.HARDESTY
6
x xii xiii xv
The mining camp as community WILLIAM A.DOUGLASS
59
81
97
viii 7
Contents Mining, colonialism and culture contact: European miners and the indigenous population in the sixteenth-century Arctic ROBERT M.EHRENREICH
PART II Anthropology and social history 8
9
109
121
‘Find the ekijunjumira’: iron mine discovery, ownership and power among the Toro of Uganda S.TERRY CHILDS
123
Mining as microcosm in precolonial sub-Saharan Africa: an overview EUGENIA W.HERBERT
138
10 A risky business: death, injury and religion in Cornish mining c. 1780–1870 JOHN RULE 11 Silver shackles and copper collars: race, class and labor in the Arizona mining industry from the eighteenth century until World War II THOMAS E.SHERIDAN
PART III Prehistory and protohistory
155
174
189
12 Producing copper in the eastern Alps during the second millennium BC STEPHEN SHENNAN
191
13 Prehistoric copper mining in the context of emerging community craft specialization in northeast Thailand VINCENT C.PIGOTT
205
14 Small-scale mining and smelting in ancient Cyprus VASILIKI KASSIANIDOU
226
Contents 15 Exploiting the desert frontier: the logistics and politics of ancient Egyptian mining expeditions IAN SHAW 16 Gold-miners and mining at Bir Umm Fawakhir CAROL MEYER
PART IV Overviews
ix
242
259
277
17 On the value of mixed methods in studying mining communities DAVID KILLICK
279
18 Mining communities, chaînes opératoires and sociotechnical systems BRYAN PFAFFENBERGER
291
Index
301
Figures
3.1 3.2 3.3 3.4 3.5 4.1 5.1 5.2 5.3 8.1 9.1 9.2 9.3 9.4 10.1 12.1
12.2 12.3 13.1 13.2 13.3 13.4 13.5
The location of Dolly’s Creek The study area at Dolly’s Creek Ages of the population at Dolly’s Creek in 1861 Site plan of House C75 at Dolly’s Creek Site plan of House C45 at Dolly’s Creek An early street scene in the Ahaura gold-mining settlement, Napoleon Hill, west coast, South Island, New Zealand, 1866 Map showing locations of sites mentioned in the text Photograph (c. 1910) of Reipetown, Nevada Site map, Reipetown, showing superimposed 1907 town plan. Map of the Toro and Nyoro kingdoms in Uganda during colonial times Mining malachite in Katanga (1924) Washing malachite ore in Katanga (1924) Iron mining in Lopanzo, Zaïre (1989) Iron mining in Banjeli, Togo (1985) Cornish humour illustrated in the cartoons of Oswald Pryor (also 10.2 and 10.3) Copper sources and known early Bronze Age/middle Bronze Age settlements in the Mitterberg copper production region of the Salzach valley The exchange region of the Mitterberg mining region, showing documented contacts View of the Klinglberg and its topographic location from the southwest Map of major prehistoric sites in Southeast Asia and Thailand Archaeometallurgy Project study areas Contour map of Phu Lon The ore body at Phu Lon View of the copper-mining area at Phu Lon River cobble mining maul from Phu Lon
44 45 46 51 53 62 83 91 93 126 143 144 147 148 164
194 195 196 206 208 209 210 211
List of figures 14.1 14.2 14.3 15.1 15.2 15.3
15.4 16.1 16.2 16.3
16.4 16.5
Map of Cyprus with sites mentioned in the text View of Almyras from a hill located to the north The rock chamber at Ayia Varvara Almyras Map of Egypt showing locations of sites mentioned in the text Scene from the painted wall-decoration of the tomb chapel of Khnumhotep III Bar charts showing the total number of expeditions sent to Wadi el-Hudi, Sinai and Wadi Hammamat during different reigns, and the number of expeditions per regnal year Annotated copy of the decoration along the bases of walls in the ‘silver-room’ at the temple of Hathor, Dendera Vicinity of Bir Umm Fawakhir Main settlement at Bir Umm Fawakhir—areas mapped in 1992, 1993 and 1996 View from the guard post over the northeastern end of the main settlement showing houses laid out on either side of the wadi bottom Three dimpled ore-crushing stones and the lower half of a rotary grinding stone Ancient mine shaft on the mountainside below the Roman tower
xi 228 230 233 245 249
252 254 261 262
264 266 268
Tables
3.1 3.2 4.1
4.2 10.1 15.1
Occupations of the residents of Dolly’s Creek in 1861 Distribution of artefacts from the houses excavated at Dolly’s Creek Hypothesis: percentages of functional artefact category expected in the archaeological record of contemporary frontier brothels, hotels and working-class households General hypothesis of types of non-Chinese prostitution and settlement pattern by economic phase and town type Mortality of Cornish tin-miners, 1849–1853, according to age cohort List of mining personnel given on an early twelfth-dynasty inscription at Wadi el-Hudi
48 52
72 78 159 247
Contributors
Peter Bell, Historical Research Pty Ltd., Adelaide, Australia S.Terry Childs, National Park Service/Smithsonian Institution, Washington, DC, USA William A.Douglass, Basque Studies Program, University of Nevada, Reno, Nevada, USA Robert M.Ehrenreich, National Research Council, Washington, DC, USA Donald L.Hardesty, Department of Anthropology, University of Nevada, Reno, Nevada, USA Eugenia W.Herbert, Department of History, Mount Holyoke College, South Hadley, Massachusetts, USA Vasiliki Kassianidou, Department of History and Archaeology, University of Cyprus, Nicosia, Cyprus David Killick, Department of Anthropology and Department of Materials Science and Engineering, University of Arizona, Tucson, Arizona, USA A.Bernard Knapp, Department of Archaeology, University of Glasgow, Scotland Susan Lawrence, Department of Archaeology, La Trobe University, Melbourne, Victoria, Australia Carol Meyer, The Oriental Institute, University of Chicago, Chicago, Illinois, USA Bryan Pfaffenberger, Division of Technology, Culture, and Communication, School of Engineering and Applied Science, University of Virginia, Charlottesville, Virginia, USA Vincent C.Pigott, Museum Applied Science Center for Archaeology (MASCA), University of Pennsylvania Museum, Philadelphia, Pennsylvania, USA
xiv
List of contributors
John Rule, Department of History, University of Southampton, Southampton, England Ian Shaw, Institute of Archaeology, University College London, England Stephen Shennan, Institute of Archaeology, University College London, England Thomas E.Sheridan, Arizona State Museum, University of Arizona, Tucson, Arizona, USA Alexy Simmons, New Zealand Historic Places Trust, Hamilton, New Zealand
Foreword
‘Mining is a community of occupation, not a community of place, like farming.’ Municipal Judge Neil V.Reynolds, a fifth-generation resident of Leadville, Colorado, was credited with that quotation in a recent newspaper article (International Herald Tribune, 7 November 1997, p. 24) describing the efforts of his home town to raise money from the Guggenheim Foundations in order to support Leadville’s National Mining Hall of Fame and Museum. The residents of Leadville turned to this source because the Guggenheim fortune, the basis for the four Guggenheim Foundations now located in New York City, began in Leadville. In the late nineteenth century, the Guggenheim family extracted an estimated $268 million worth of lead and silver from the mines in the vicinity of Leadville, but all this wealth was taken out of the region. Today the population is one-tenth what it was in the 1880s. As Judge Reynolds put it, ‘Nobody who made their money here did anything for this place.’ Today Leadville stands a bleak monument to the ‘get rich and get out’ mining ethic (as noted by James Brooke, author of the IHT article cited above). Having had the pleasure of participating in the Bellagio Conference on The Archaeology and Anthropology of Mining, a rewarding experience in every respect—intellectual, social and gustatorial—I believe that Judge Reynolds’s observation goes to the very heart of all the problems involved in any attempt to study the archaeology and anthropology of mining communities. Mining is indeed ‘a community of occupation, not a community of place.’ For this reason mining communities—the villages and encampments of the miners, their families and their female associates—are most likely to be rather ephemeral affairs, created by individuals who always saw their residence at the site as temporary and transitory, and therefore destined to leave little trace in the material record. The mines themselves, of course, do not move, and mining archaeology—to the extent that it has been devoted to the reconstruction of what went on in the mines themselves—has enjoyed phenomenal success in recent decades. The Roman mines of Spain, the medieval mines of France and the Neolithic and Bronze Age flint and copper mines of southeastern Europe and the British Isles have been studied and published in a series of highly successful research
xvi
Foreword
programs. The great impetus for the study of prehistoric mining archaeology has come from the introduction of radiometric dating techniques, making possible for the first time the accurate dating of archaeological contexts that in the past could only be described as ‘pre-Roman’ or ‘pre-Celtic.’ The archaeology of mining communities, on the other hand, provides quite a different story. The papers delivered at the Bellagio Conference, devoted to elucidating some ‘social approaches to an industrial past,’ all too clearly demonstrated the difficulties involved in trying to reconstruct the lives of the miners themselves, who they were, where they came from and how they interacted with the world around them. With the help of a written record, as is available for early modern mining towns (like Leadville, Colorado), answers to these questions can often be suggested. In a prehistoric context, however, answers have seldom even been proposed, in however tentative a fashion. What was attempted at Bellagio in July 1996 was to bring together scholars who work in ancient and modern, prehistoric and historic archaeological contexts, and to add the dynamic input of social historians and anthropologists, in order to compare and contrast, to look for difference and commonality, and to study mining communities as a social phenomenon, not one bound exclusively to the minutiae of the archaeological record. For a scholar like myself, attempting to make sense of excavations conducted in 1938–1939 at what remains the only investigated Late Bronze Age mining community (now utterly destroyed) on the copper-rich island of Cyprus in the Eastern Mediterranean, the Bellagio Conference was a sobering experience. It was obvious that, as a Cypriot archaeologist, I had much to learn. Any understanding of that Late Bronze Age site (Apliki Karamallos) in the mineralized zone of Cyprus’s Troodos Mountains will be forthcoming only if we ask the right questions. The chapters of this book formulate many of the right questions, and even suggest some possible answers. James D.Muhly, Director American School of Classical Studies at Athens
Chapter I
Social approaches to the archaeology and anthropology of mining A.Bernard Knapp
ABSTRACT The study of the mining of metals—the bailiwick of several different disciplines— reveals great diversity of approach. In the most general terms, social historians, archaeologists and archaeometallurgists tend to focus on the history and technology of mining, on metallurgical technology and on the mining process overall. Historians of mining and ethnographers tend to examine the socio-economic, spatial and ideological dimensions of past industrial cultures. In the attempt to reorient these divergent approaches in a more dynamic direction, contributors to this volume focus on the social context of the mining or metallurgical community as revealed in the material, ethnographic and ethnohistoric records of various cultures worldwide, from prehistory to the recent past. The mining community provides the focal point for information about human activity on an industrial frontier. Although mining communities are often socially and spatially remote, they are linked into broader social, communications, transport and economic networks by virtue of their ability to supply a raw material in demand to a regional or world system. This volume brings together the research of archaeologists, ethnohistorians and anthropologists to consider themes of common interest, and to explore relevant interdisciplinary issues. These studies are oriented around specific, interweaving themes: the social context of production; gender; power strategies and labour exploitation; imperialism and colonialism; and production and technology. Such themes, and the material or ethnohistoric features and human behaviour that permit us to discuss them, are well known to social historians and ethnographers; archaeology, however, has only begun to develop the kind of social theory essential for treating the material components of an industrial past.
INTRODUCTION The mining and manufacture of metals have formed a major aspect of production and social reproduction over the past six thousand years, and have employed millions of people throughout the Old and New Worlds. All industrial societies were erected on foundations of metal, and still require vast
2
A.Bernard Knapp
quantities of metals for their existence. Substantial areas of the earth’s surface have been transformed by mines or by the pollution that forms a common adjunct to mining and metallurgy. Mining may be regarded as an extension of people’s search for natural materials used in the production of tools, weapons and ornaments (Weisgerber and Pernicka 1995:159). The exploitation of resources, the selection of tools, the organization of labour and the sequence of acts that make up a technological process are in fact key components of social behaviour and human choice (Childs and Killick 1993:325). Once metal is obtained from ores, it is invested with multiple and changing social, economic and politico-ideological roles. Metals and metal artefacts often take on material and symbolic expressions of wealth/status, ethnicity, fertility and other ‘life changes’, as well as religious or ideological affiliation (Childs and Killick 1993: 330–332). In the preindustrial world, moreover, metals were valuable commodities, and Bronson (1992:65) argues that their value was enhanced when transported by sea. Despite the vast amount of financial and human capital devoted to mining and related activities, their scholarly study has until very recently focused on technological factors or on economic and labour histories. Only within the past decade have archaeology and cultural anthropology embarked upon more innovative studies of the social, spatial and ideological dimensions of technology and of past or present industrial communities (e.g. Childs and Killick 1993; Godoy 1985; Pfaffenberger 1992). Technology is not simply a scientific force but also a social process that involves human behaviour and negotiation, a complex ideology, and the extraction and manipulation of resources—human and material. Mining and metallurgy in preindustrial settings were often imbued with rich symbolic content, the study of which illuminates various aspects of culture that have little to do with metallurgy per se. Mining communities, like the emporia that distribute their products, cannot be understood in a cultural vacuum (Weigand 1982:3). Even though many cultural features associated with technology, mining and metallurgy are known from ethnohistoric documents or ethnographic fieldwork, archaeology has had to develop its own devices to assess the material components of an industrial past. The use of ethnohistoric and archaeological evidence as contrasting data sets, and the attempt to integrate archaeological, anthropological and historical approaches, fit squarely within current trends to develop a socially oriented, human-science approach to the study of the past. Recognizing this interdisciplinary dynamic, we (Bernard Knapp, Vincent Pigott, David Killick) organized an international conference which was held at the Rockefeller Foundation’s Bellagio Study and Conference Center on Lake Como, Italy (22–26 July 1996). The substantially revised papers from that conference which are published in this volume span the past 4,500 years and cover the areas of the Americas (including the Arctic), Australia and New Zealand, Egypt, sub-Saharan Africa, Europe, the Mediterranean and Southeast Asia.
Archaeology and anthropology of mining
3
This volume brings together prehistoric and historical archaeologists, social historians, archaeometallurgists and anthropologists, all of whom are widely recognized as innovative thinkers in their respective fields, to consider themes of common interest, and to explore certain social approaches to an industrial past. If the immediate goal of the conference was to acquaint scholars from different fields with the current status of research in other fields, our aim in this volume is to explore how these closely related but still diverse approaches may better be integrated; we believe this marks an important advance toward an interpretive, interdisciplinary approach to the archaeology and anthropology of mining. In his recent comprehensive study of early mining, Craddock (1995:1–2) confines his discussion to information that can be derived directly from material evidence. He believes that ‘the composition of materials can give far more unambiguous information on the technology of the process that produced them than on the society that made and used them’ (Craddock 1995:2). This belief in the superiority of ‘immutable chemical properties’ to ‘less perfectly formulated theories of workings of human society’ takes little account of current debates among scientists themselves about the value or worth of the ‘scientific’ record, the scientific community and the entire scientific endeavour (e.g. Ruse 1996; Shapin 1996). It is also a self-fulfilling prophecy, since the more that specialists who study mining and technology avoid theoretical and social approaches in favour of the empirical, the more difficult it will be to gain any new insights into the social aspects of those fields. Craddock’s opinion must be considered, and his work on early metal mining and production is essential to this field. But it is only one opinion and only one, quite narrow approach. In contrast, our main focus in this volume is to examine various social aspects of mining, especially mining communities (in addition to mining settlements or mining landscapes, ostensibly more accessible to archaeologists). To do so, several themes have been singled out for emphasis and exploration. In this introductory chapter I attempt first to consider the concept of the mining community as construed by fields ranging from sociology to the history of mining. I then present and expand upon the various themes as a means of introducing this volume, raising new questions, stimulating new ways of considering mining communities, and provoking contributors and readers alike to reconsider long-held notions in their respective fields. The intermingling of divergent scholarly interests should help to advance significantly the study of an industrial past and lead to a more dynamic, social synthesis.
THE MINING COMMUNITY The work of ethnographers, social historians and sociologists alike demonstrates a fascination with the distinctive character of mining communities (e.g. Bulmer 1975; Tenfelde 1992a; Thompson 1932), especially capital-
4
A.Bernard Knapp
intensive mining. Perhaps the most fascinating aspect is the collective behaviour resulting from the dynamic between working and living in settlements structured around a single commodity or, in the modern world, around a single industry or technology. The nature of minerals fixes the location of underground mines, which in turn influences the location and layout of a mining community. These same factors entail the movement of people, some degree of labour specialization, dependence on a single economic activity and— where there is capital investment—a large degree of reliance on the mining company (Bulmer 1975:61). In more recent mining communities, the fate and well-being of labourers are largely determined by economic forces external to the community; social change, moreover, is often instigated or mitigated as a result of large-scale economic factors beyond its control. Endogenous forces, both enabling and constraining, also impact the dynamics of a mining community; some degree of local autonomy and social solidarity, on the one hand, may be offset by rapid population turnover or by religious and ethnic boundaries, on the other (Bulmer 1975:69). Capital-intensive mining such as this must be distinguished from preindustrial, ‘carnival’ or ‘informal’ mining activities. The case studies in this volume help to clarify the distinction between large-scale, state-run mining endeavours, and smaller-scale, kin-based mining activities in precapitalist settings. Growing evidence exists for autonomous mining communities in prehistoric or preindustrial societies, in particular those that depend, apparently, on seasonal and non-specialist labour. The mining community—or camp, or village—represents the domestic space of people who often were or are heterogeneous in character, of diverse origins, and drawn together by the need to work (Roberts 1996:5, 10). Such communities are more often than not expedient and impermanent: when mines shut down or when ore bodies are exhausted, the mining community itself fragments or dies as miners scatter to new strikes and new opportunities. Geographic mobility, then, is common while social mobility is rare: the opportunity to rise to a higher occupational stratum was always extremely limited. Once a miner, always a miner…but what kind of miner? In a brief report on the Bellagio Conference, Ehrenreich (1996) suggested that the full-time, wage-earning miner was a phenomenon of the Industrial Age, in particular of the past two centuries. MacMillan (1995), writing on the Amazonia gold rush of the 1980s, used the term ‘informal mining’ to describe a system in which people engage in mining only as a part-time activity. Most of the time these ‘miners’ are in fact farmers, ranch hands or wage labourers. Godoy (1985) also makes the general point that, in highland Bolivia, periodic and part-time mining has long served as an alternative occupation for marginal peasants. Most early mining, from prehistoric to medieval, was carried out by individuals who lived in agrarian-based communities, and was conducted chiefly during periods of the annual cycle when people were unable to farm. Such individuals came together as a community—real or imagined—at specific
Archaeology and anthropology of mining
5
times to mine or smelt metallic ores whose products had utilitarian, economic, prestige or ideological value. Some of these communities mined or smelted only the ores needed for their own agricultural or everyday needs. During the eighteenth century in the American Southwest, mining camps ‘flared and faded’, but all served as temporary magnets of settlement for thousands of Native Americans, Spaniards and castas (people of mixed race), at least as long as the ore deposits continued to spawn metallic wealth (Sheridan 1992: 159). In an unpublished study of nineteenth-century mining in California and Nevada, Douglass (1979; this volume, Chapter 6) discussed the nature of social recruitment to the mining camps of that era. The typical California camp was more a collection of strangers than a community of friends or kinsmen. The camps exuded a sense of urgency and transience, and the typical ‘miner’ seldom remained in one. Indeed, by the end of the nineteenth century, there occurred an almost quasi-nomadic movement, primarily of men who were born and raised in mining or former mining towns in California and Nevada, and who followed the rich discoveries of gold and silver to the end of the rainbow. These ‘rushes’ from one field to another, in other words, resulted in a significant mobile population of miners in the American West, whom Douglass described as ‘a community without a locus’. This phrase cuts to the heart of many a mining community, throughout time, and in some respects accounts for the strikingly thin level of evidence, even of material variability, for mining communities in the archaeological and ethnohistoric records. Such communities came together briefly as mining camps centred on new discoveries. Douglass maintains that those involved in these communities without a locus came to know each other well, and developed their own traditions and values which transcended the experiences of any single camp. As a result, a skilled, self-aware community of miners persisted well beyond the use span of any specific rush site. These examples, and several of the case studies presented in this volume, suggest that the organization of a small mining community was quite fluid, often temporary and ephemeral, and certainly expedient as far as structures or other material remains are concerned. In fact, the material culture recovered through excavations at historical sites in the USA, Australia and New Zealand usually reveals limited evidence of ethnicity, class, gender or even wealth (see this volume, Chapters 3, 4 and 5, by Lawrence, Simmons and Hardesty, respectively). Beyond factors of impermanence and fluidity, the lack of variability in material culture would have resulted from the fact that many miners previously had little access to wealth and planned to work only a short time, strike it rich, and make their investments elsewhere. Moreover, the remoteness of mining camps, or the fact that the only store available was a company store, meant that there was little variety in material culture to choose from. Finally, the wealth amassed by the mining companies was only rarely invested or spent in the mining town or camp itself; more often it went to owners and stockholders in the major urban centres (Ehrenreich 1996:55).
6
A.Bernard Knapp
How then may we envision the social structure and material character of a traditional mining community? Bulmer (1975:85–88) suggested that such communities may exhibit the following characteristics: 1 2 3 4 5
physical isolation, and a dispersed settlement system; economic predominance of mining; exacting, dangerous and periodic work; occupational homogeneity and isolation; communal leisure activities (religious, sporting, drinking) where work remains the chief interest and topic of conversation; 6 sharply segregated family and gender roles; 7 economic and political conflict between miners and managers; 8 multiple and complex communal social relationships: solidarity, shared histories of work and living, inward focus. But how, if at all, do such factors, which typify more recent and capital-intensive mining communities or camps, relate to smaller-scale, temporary communities or smelting sites; that is, to archaeological realities and the study of an industrial past? How, if at all, may they help us to understand better the material or social record of mining communities? We cannot assume that all prehistoric metal producers operated on a local and therefore small scale; metals, especially when transported, were valuable and very tradable commodities, often in high demand over vast regions (e.g. throughout the ancient Mediterranean world for almost three thousand years, c. 2500 BC to AD 500). Bronson (1992:104–105) argues that the image of the village smithy mining and smelting only enough for local needs is emotionally appealing but culturally improbable: most small-time producers of metal who are known historically or ethnographically are or were commercial smelters who sell at least a portion of their output—whether ingots or finished products— to non-locals. The majority of documented preindustrial miners or metalworkers did not labour alone but in communities where there were multiple furnaces and a significant level of cooperation in mining, producing fuel, and building and maintaining equipment as well as lodgings. Bronson (1992) believes that these factors may have been typical of prehistoric mining communities, which would have contained up to twenty-five families or households producing between 5 and 15 tons of metal each year. Hardesty’s work in the historical archaeology of mining communities on the American frontier during the nineteenth and early twentieth centuries indicates that the most common household situation was a small group of unrelated males living under one roof and sharing some domestic chores, but otherwise unintegrated by other organizing principles of society (Hardesty 1992:180– 182). Hardesty (1992:186–191) suggests that the most common household types in the frontier mining communities of the Old West were organized around mutual aid or cooperation, around families, as occupational collectives
Archaeology and anthropology of mining
7
(e.g. miners, brothels, saloon-keepers or store-owners), and as work groups. The material evidence, however, differs from written accounts: instead of clearly defined households (based on census records in Hardesty’s case), the archaeological record represents the end product of a time series, a kaleidoscopic view of domestic cycles—or of a ‘household series’ (Smith 1992:30) —that changed throughout the occupancy of various dwellings within a village. The main material indicators of household variability in American mining communities were architecture, artefact assemblages and site layout, including the use of house floor area to estimate the size of the residential group. There is some general archaeological evidence to suggest that the most typical house size in a mining community was about 350 square feet (33 square metres), in which groups of two to five people would have lived. A study focused on the faunal and pottery samples from the historic townsite of Grantsville, Nevada, revealed that pottery was a more reliable indicator of variability and ethnic or socioeconomic status than were bones of animals consumed at the site (Schmitt and Zeier 1993). Images and perceptions of recent or modern mining communities obviously cannot be retrodicted directly into the prehistoric, classical or medieval past. Shennan (1993:66–68) nonetheless has noted that the individual mining communities of the central European Bronze Age had limited numbers of inhabitants and therefore did not sustain any significant degree of internal differentiation. Moreover, he has argued (Shennan 1991) that Bronze Age copper mining was mainly a winter activity, an observation that tallies well with the more recent and contemporary evidence for ‘informal’ mining. Similarly, mining expeditions during Egypt’s Middle Kingdom (c. 2040–1650 BC) seem to have been dispatched when the Nile River had flooded and farming was impossible (Shaw, this volume, Chapter 15). A hilltop mining village consisting of 125 stone-built structures and dated to the Egyptian Old Kingdom (c. 2650– 2180 BC) showed no evidence of seasonal activity but reflected very well the isolation and even vulnerability of a tightly clustered mining settlement expediently situated in close proximity to the turquoise and copper mines of the Sinai Peninsula (Shaw 1994: 114–115). In addition to seasonality, isolation, economic orientation and household make-up, several other facets of a traditional mining community can be seen in the material record. The themes highlighted here (see next section) represent an attempt to tease some of those facets out of the archaeological studies presented in this volume. The relative isolation of the traditional mining community is increasingly rare in the postmodern, consumerist, communication-oriented society of the late twentieth century (Tenfelde 1992b:1211–1212). After World War II, when the study of social history enjoyed a new prominence in the wake of the Annales’ impact on the field, the attention focused on mining communities sharpened. More interesting is the fact that the phenomenon of analysing the miner’s community occurred at the same time that these very communities, particularly those which belonged to the older, industrial nation-states (i.e. the UK,
8
A.Bernard Knapp
Germany, France, Belgium) were fading out of existence. The changing energy demands of the decades between 1945 and 1965 led to the exhaustion of many formerly productive mineral deposits, and to increasing costs for producing a resource such as coal in comparison with newer and cheaper sources of energy. The mining of iron ore, moreover, had been reduced to a level of activity more suited to a relatively small branch of the industrial world. However, it is likely that the more basic reasons for the intensified study of mining communities are to be found in the changes that occurred in mining technology and in the lifeways of the miners themselves (see, for example, Godoy 1991; Holloway 1996; Nash 1993). The modernization of mining technology and the political strategies and successes of capitalism over socialism fundamentally reversed the image as well as the reality of the mining way of life. In sum, the new awareness and concern with traditional mining communities arose at approximately the same time that this unique, uncommon way of working, earning and living was coming to an end. Most of the studies presented here seek to expand upon our knowledge of the traditional mining community and its social technology, even as this page of the industrial past is about to be turned for ever. Themes and issues The history of mining and metallurgical technology is a topic well represented in both academic and popular literature (Craddock 1995; Domergue 1989; Lavender 1962; Shepherd 1993). Archaeologists have considered the technology of mining, and the productive process overall (e.g. Dutton et al. 1994; Ehrenreich 1991; Hardesty 1988; Lechtman 1996; Weisgerber 1982, 1989; Weisgerber and Pernicka 1995), but have yet to integrate fully into their work the social, politico-economic, spatial and ideological dimensions of past industrial cultures (see Childs and Killick 1993). Anthropologists have considered economic, ideological and organizational aspects of mining (e.g. Godoy 1985, 1991; Nash 1993), but often from opaque Freudian or structuralist perspectives (e.g. Eliade 1978; Taussig 1980). Social historians of mining and some historical archaeologists, however, have written extensively about gender, class, age, spatial segregation and other issues directly relevant to the themes of this volume (e.g. Frazier with Brown 1992; Herbert 1993; Hosler 1995; Lankton 1992; Schmitt and Zeier 1993; Simmons 1989). In addition to the book’s focus on the social context of the mining or metallurgical community as a means of engaging the wide range of interests and expertise represented in this volume, a series of general, overlapping themes helped to orient our diverse approaches: gender; power strategies and labour exploitation; imperialism and colonialism on the mining frontier; production and social reproduction; and production and technology. Various chapters emphasize other themes that were developed in order to utilize the full, interdisciplinary range of evidence:
Archaeology and anthropology of mining
9
1 examining social structure and social interaction in mining towns and camps; 2 understanding earlier technical practices and technological knowledge; and 3 reconstructing the social organization of metal production from material remains. Such topics have relevance to any people whose culture and history are bound up with the mining and exploitation of natural resources; they also highlight the significance of archaeological or ethnographic fieldwork to cultural heritages, to ethnographic histories, and to understanding and negotiating difference across changing societies. The literature—or, more accurately, literatures—on past mining and metallurgy have yet to achieve a synthesis of the technological, economic and social aspects of these activities. Archaeologists and archaeometallurgists impressively reconstruct the chaînes opératoires of past miners and metalworkers from surviving material evidence, and on that basis assess the growth of technical knowledge. As yet, however, there exist few attempts to investigate the social reality of ancient or prehistoric miners, mining towns and camps. Cultural anthropologists and social historians, for their part, have produced rich accounts of the social life of miners, but still seem largely unaware of the wealth of evidence and insights to be gained from material culture. Material remains, whether from the field or in the laboratory, are the stock in trade of archaeologists and archaeometallurgists, who are often able to reconstruct mining processes and metallurgical facilities with an extraordinary degree of detail. The composition and microstructures of various metal artefacts may offer good evidence for long-distance trade and/or technology transfer. Discussion of these topics may be found in the chapters by Robert Ehrenreich (Canadian Arctic— Chapter 7), Vasiliki Kassianidou (Iron Age Cyprus— Chapter 14) and Vincent Pigott (Bronze Age Thailand— Chapter 13). Historians of mining rarely make adequate use of the available material evidence to reconstruct either the working or the home environments of past miners and metallurgists. However, a number of fields—anthropology, industrial archaeology, art history, folklore, rural history—reveal a growing concern with the use of material culture as historical documents. Of widespread current interest in the social sciences, this issue is treated in chapters by Terry Childs (Africa — Chapter 8), Susan Lawrence (Australia— Chapter 3) and Alexy Simmons (west/northwest USA, New Zealand— Chapter 4). Conversely, social and labour historians and cultural anthropologists have made issues of power and resistance, poverty, and risk central to their research on mining. Miners, both today and in the recent past, are prone to sudden death and traumatic injury, are generally poorly paid and housed, and are often ruthlessly exploited by powerful employers. In many precapitalist mining camps, they were slaves or convicts. These conditions led to frequent strikes or revolts, which constitute a recurrent theme in the literature on mining. Clearly the enormous benefits that metals have bestowed upon humankind in historic
10
A.Bernard Knapp
times have to be balanced against the toil and misery by which they were won. Near the base of the Cerro Rico mountain in Bolivia, the city of Potosí became rich on silver and copper mining, grew to be larger than Madrid, Rome and Paris, and was one of the most famous centres in colonial Latin America (Holloway 1996). The local Indians in this region have been exploited for centuries in mines that have always been unsafe; those who escape death by accidents succumb to disease, with the vast majority contracting tuberculosis by middle age (Nash 1993). Issues related to poverty, risk and injury are treated in chapters by Robert Ehrenreich (Canada/Arctic— Chapter 7), Peter Bell (Australia— Chapter 2), Donald Hardesty (western USA— Chapter 5), John Rule (Cornwall, England— Chapter 10) and Thomas Sheridan (southwest USA — Chapter 11). The social structure of mining towns and camps cannot be reduced to simple class divisions. In the colonial mining settlements of Australia and the Americas, linguistic and ethnic divisions within the working class also structured the communities, and it was not uncommon for miners from one particular ethnic group to constitute a ‘labour aristocracy’. And in all but the most rudimentary and most isolated mining camps, gender divisions were present, along with a wide variety of domestic arrangements. Studies dealing specifically with gender and/or ethnicity are presented in this volume by Peter Bell (Australia — Chapter 2), Terry Childs (Africa— Chapter 8), Eugenia Herbert (precolonial Africa— Chapter 9), Susan Lawrence (Australia— Chapter 3), Stephen Shennan (Europe, Bronze Age— Chapter 12), Thomas Sheridan (southwest USA— Chapter 11), and Alexy Simmons (west/northwest USA, New Zealand — Chapter 4). As Craddock (1995) maintains, archaeologists have limited access to knowledge concerning the social life of prehistoric miners and metallurgists; their attention is thus usually focused upon mines, slags and furnaces rather than on living areas. If nothing else, ethnohistoric studies can expose archaeologists to the vivid social history of mining communities, and stimulate them to balance impressive reconstructions of past technology with investigations of the social circumstances of the miners themselves. Studies by historical archaeologists who have worked in nineteenth- and twentieth-century mining and metallurgical communities in North America and Australia (Susan Lawrence, Robert Ehrenreich, Donald Hardesty, Alexy Simmons) help to provide a conceptual bridge between prehistorians, on the one hand, and social historians and ethnographers, on the other. Working with a mixture of documentary and archaeological evidence, several of the chapters in this volume explicitly tackle questions related to class, power and ethnicity, and whether these factors may be recognized in the spatial patterns and distribution of material culture in excavated mining: Carol Meyer (Byzantine Egypt— Chapter 16) and Ian Shaw (Bronze Age Egypt— Chapter 15) consider these themes in precapitalist contexts, while Vasiliki Kassianidou (Iron Age Cyprus— Chapter 14), Vincent Pigott (Bronze Age Thailand— Chapter 13) and Stephen
Archaeology and anthropology of mining
11
Shennan (Bronze Age Europe — Chapter 12) report on excavations of mining settlements in strictly prehistoric contexts. Studies of mining and metallurgy have the capacity to expand and alter our understanding of processes of colonialism and imperialism, and the challenge of frontiers. Deposits of metal ores are unevenly distributed over the surface of the earth, and empires have often established mining colonies at a great distance to extract metals for home consumption. Very rarely are mining expeditions sent to an area devoid of an indigenous population; the earliest camps set up in a given region therefore often function as a type of colonial outpost (Ehrenreich 1996:55). Several chapters specifically examine issues related to mining colonies, or colonies established to trade in metals by precapitalist empires: Peter Bell (Australia— Chapter 2), Eugenia Herbert (Africa— Chapter 9), Carol Meyer (Egypt— Chapter 16) and Ian Shaw (Egypt— Chapter 15). These studies offer instructive comparisons of mining in both ancient and modern polities, a comparison that should be of interest to anyone who deals with issues related to colonialism, postcolonialism and imperialism. How have such themes been considered within the disciplinary contexts of ethnography, ethnohistory and archaeology? How may they be linked to a more unified, social approach to an industrial past?
ETHNOHISTORY AND ETHNOGRAPHY In ethnohistorically attested and modern mining communities, certain ecological, demographic and politico-economic patterns tend to recur. Mining populations, for example, often have higher fertility than do agricultural populations, not least because of higher infant and adult mortality or debility, but also as a result of early marriage for women, low child-rearing costs, a minimum number of females in the labour force (at least, outside the home or brothel), and the existence of menstrual or sexual taboos (Godoy 1985:206; Herbert 1993:72–73). Historical archaeologists working in the American Northwest have demonstrated from material and documentary evidence alike that the frontier prostitute comprised a significant and ever-present component of this labour pool (Simmons 1989; this volume, Chapter 4). Mining is almost always a labour-intensive activity, and the successful mining operation necessitates a reliable, inexhaustible pool of labour. Most historically attested miners came from highly diverse ethnic and sociocultural backgrounds, but mining households were often co-residential, with the majority of men living under one roof. Some lived in bunkhouses or boarding groups, others with families, or in small, single-sex clusters (Hardesty 1988:15–16; this volume, Chapter 5). In the American West, mining communities were made up mainly of adult males from urban centres; they had high mobility and an incoherent sense of community—that is, they were highly individualistic. Most mining camps were also rife with class distinctions. The ‘hillside stratification’ in the
12
A.Bernard Knapp
mining town of Virginia City, Nevada, is a good example: elite households typified the upper streets, and were followed in order—descending the hill—by commercial and government residences, working-class homes, the red light district and Chinatown (Reps 1975:276–277). Precolonial African metallurgical communities, from large-scale undertakings exploiting a servile workforce to smaller-scale efforts by single family groups, were essentially temporary camps rather than permanent mining settlements. Such communities did not necessarily involve either specialist production or distinct social groups, but the organization of production was certainly non-egalitarian: the labour of women and younger men, even children, was controlled by senior males on the basis of their technical and ritual knowledge (Childs and Killick 1993:329–330; Herbert 1993:25–31). The politico-economic organization of most permanent mining communities is conditioned in part by locational factors such as isolation and inaccessibility, which, at least in historically recent periods, have given rise to self-contained communities such as unstructured, open-system camps (Bolivia, Colorado), paternalistic company states (Arizona, Chile, Peru), or the hermetic, totalitarian institutions of South Africa (Godoy 1985:205). Companies or their agents often sought to integrate the surrounding regions into a coherent economic sphere, in order to obtain not only capital and labour, but also food and other supplies for the physical operation of the mining camp. Although the production process needs to be examined in its entirety, many studies that focus on the ideology and ritual of mining seem somewhat idiosyncratic. Eliade’s (1978) Freudian analysis, for example, draws an analogy between obstetrics and mining that some might find questionable or even objectionable. On the other hand, while Eugenia Herbert’s (1993) work reveals clearly such an analogy between obstetrics and metallurgy, she has stressed the social construction of gender and reproduction rather than the biological analogy of Eliade. Taussig’s (1980) Marxist-inspired study of ideology and consciousness among Bolivian tin-miners, despite its literary qualities, tells us simply that belief systems—with their interrelated symbols and meanings—help to mediate between tensions or to fetishize evil. The material witness of this phenomenon is quite striking: each group of Bolivian miners sets up next to their work site a sculpted figure, often life-size, of a spirit called El Tío (‘uncle’). El Tío is reminiscent of the Christian devil, a sort of underground divinity thought to own and control the disbursement of riches within the mountain deposits it inhabits (Holloway 1996). In African metal production, the shape and decoration of furnaces may offer evidence of a conceptual association between smelting and ceremonial rites involving gestation, birth or marriage (Childs and Killick 1993:328). These furnaces could also provide material clues relevant to the study of prehistoric behaviour. Moreover, the spatial separation of village sites and smelting areas (to keep sorcerers and sexually active women at bay) might-be evident in archaeological surveys. The social organization of mining is partially conditioned by the physical and/or sociocultural isolation of mining communities, and partially by the harsh
Archaeology and anthropology of mining
13
working conditions and labour requirements of the extractive and productive phases of mining (Godoy 1985:205). Hardesty (1988:1) envisions the mining frontier as a series of ‘patches’ or ‘islands’ (where ore bodies were located) colonized by miners who, because they brought their own cultural traditions with them, found themselves implanted in a sociocultural ‘wilderness’. This does not deny, of course, the emergence or development of cultural and material traditions specific to the mining communities themselves, as exemplified by the widespread use of the El Tío figures in the Bolivian mines. To cope in their various, often remote, situations, miners adopted either a strategy of opportunity (to maximize their resource gains) or a strategy of resiliency (to survive under sudden environmental or politico-economic change). The maximizing strategy would tend to reduce variability in material culture, for example as miners would adopt the most effective, standardized tools and technology. The resiliency strategy, however, would tend to increase variability, for example as miners experimented with ways to cope under stress, or ways of changing the organization of labour to cut mining costs (Hardesty 1988:112–114). A conceptual framework to examine the variability and interrelations between mining communities and the surrounding rural and/or agricultural villages might involve consideration of transport factors, provision of subsistence and fuel, mechanisms for social exchange (marriage, seasonal labour, prostitution), ideological dimensions and ecological alterations. In the African case, mining was almost always a seasonal activity, and whole families of miners would move into bush camps to work at mining and smelting, following which they would return to their villages and traditional agricultural pursuits. This situation reveals an almost complete overlap between differing modes of production. In such a situation, where miners are agriculturalists, is there any overlap in the different residential occupations? Such considerations offer another possible means to bridge the gap between ethnohistoric or ethnographic and archaeological approaches to the study of mining communities and industrial society. How else might the information and evidence established by ethnographers and historians impact our understanding of prehistoric mining and material culture? How different are the approaches based on production and technology typically used by archaeologists?
ARCHAEOLOGY The material record of mining sites, rural/agricultural villages and production/ distribution centres is well suited to a study of archaeological variability from a broad comparative and chronological perspective (Knapp 1997). Within such a study, it is necessary to relate archaeologically visible mining ‘settlements’ to the more abstract concept of mining ‘communities’ (Hardesty 1988:101). Such communities will most often be situated in close proximity to ore bodies;
14
A.Bernard Knapp
secondary locational determinants include water, supporting resources (timber, agricultural land) and gravity components (access roads, transport system, towns). Since much primary production occurred in close proximity to the mines, a certain amount of variation through time may be expected. Such variation, including craft specialization, depends on factors such as access to ores, micro-environments for effective smelting, and the organization of the labour force with respect to various sequences of production, political expediency, social control and economic demand (e.g. Brumfiel and Earle 1987; Costin 199; Lechtman 1991:45; Stech 1985:103). Within a mining settlement, the social interactions that occurred, and the social and ethnic stratification that developed, will be evident to varying degrees in artefact morphology and settlement patterns; in the size, age and sex composition of the human population; in the extent and quality of prestige items and imports; in other industrial, ceremonial or habitational features preserved in abandoned or even buried mining ‘sites’; and in information recorded during surveys of mining districts. The social and economic organization of mining communities, of course, can only be inferred from these material remnants, their spatial arrangement and their contextual coherence, which in turn must be considered within a framework built upon the productive, technological and ideological dimensions of mining. Where documentary evidence is preserved, another arena of investigation may be opened. Within historical mining communities, demographic variability was often related not only to ecology, technology, ethnicity and social class, but also to individualism (Paul 1963; Hine 1980). The variability that results from differences in ethnicity, place of origin, status, gender and class may be visible in the material record, and thus should form one of the primary areas of interest in an archaeology of mining. The individualism typical of miners might also produce high variability in the assemblages of mining households, in contrast to the cooperative behaviour, and more limited variability, that might be expected to typify the material assemblages of the classic peasant or agricultural community (Hardesty 1988:103–104). The exception, of course, is the analogy of the more transient African mining community, where individualism versus cooperation would be relevant only on a symbolic level, where anyone with exceptional power risks the onus of isolation (as a ‘witch’) (Herbert 1993:15). Despite their social and spatial remoteness, and by virtue of their ability to supply a raw material in demand, mining communities—past or present—are inevitably linked into broader social, communications, transport and economic networks. Hardesty (1988:1–5) maintains that links between the mining frontier and the outside world involve three types of interaction sphere: the material (transport networks, including food and supplies), the social (labour pool, inmigration, gender ratio of population) and the informational (symbolic, financial, ideological). These interactions may also be visible in the archaeological record in the form of settlement layout, trash dumps, luxury goods and ‘mass-produced’ commodities, imports, skeletal remains, cult
Archaeology and anthropology of mining
15
paraphernalia, etc. Such interaction spheres necessarily would have overlapped: the transport of primary smelted copper from mines to production, refining or distribution centres, for example, may have necessitated travel through regions controlled by different communities. Such movements would have involved the negotiation of social relations beyond those required for the supply and distribution of a resource in demand (Lechtman 1991:46). Political ideologies, and the sanctions associated with them, helped to organize and develop economic strategies, and to establish and legitimize the human configurations that directed these strategies (Knapp 1988, 1996a). It is not uncommon for specific historical situations, such as political revolutions, technological innovations or geological events, to serve as ‘traumatic events’ (LeRoy Ladurie 1979) or ‘kickers’ (Deiner 1980) and eventually result in a new sociocultural trajectory. It is generally acknowledged that technological innovations in and of themselves do not lead directly to major social changes: rather, the social milieu into which such innovations are introduced forms the backdrop for change (e.g. Heskel and Lamberg-Karlovsky 1980:260–261; McNutt 1990:98). Because special interest groups like miners or managers often use material items to restructure relations of power (Gamble 1986:39), such transformations may be visible in the archaeological record, and thus should form another focus of the archaeology of mining.
FURTHER CHALLENGES The archaeology and anthropology of mining may be discussed in the context of several other issues, or themes, at least four of which feature in the chapters of this volume. Mining and agriculture Mining interests and activity may complement or compromise agricultural interests and activities (an example of MacMillan’s [1995] ‘informal’ mining). For example, in the African setting, there are overlapping modes of production, where mining is often a seasonal activity and agriculture the norm. In twentiethcentury Cyprus, to take a counter-example, agriculture has often suffered because mining companies discouraged drilling for water or any other activity that might distract potential workers from filling those companies’ insatiable need for cheap labour. As an economic activity, mining is very sensitive to changing commodity prices on world markets, and mining companies brook no interference from other modes of production. The traditional agricultural base of Cypriot villages diminished dramatically during the middle decades of the twentieth century, when mining activities reached their production peak. To cite a much earlier example in the Mediterranean: during the later stages of the Roman Empire in the Western Mediterranean (Edmondson 1989:98),
16
A.Bernard Knapp
state-run mining became divorced from the local economy inasmuch as the mines were administered independently and often from a distance, while labour and skilled personnel were often imported. On another level, however, mining was integrated into the local economy, because the miners and anyone associated with the mining community had to be fed and provisioned. Moreover, the small-scale mining practised in the later Roman Empire was closely associated with an agriculturally based economy. Landowning farmers, for example, were often the only people with the resources available to exploit mineral deposits or to lease contracts to operate mining shafts in imperially owned mining districts. Agricultural workers also provided the readiest source of labour, especially during slack seasons of the agricultural year. When the Roman state fell on hard economic times, it was logistically and financially more sensible not to tie up its own administrators or soldiers in mining activities. When mining finally became economically unfeasible, or when the wider politico-economic systems that created demand for copper broke down, mining villages were abandoned unless their inhabitants had other means of support, or unless the level of production was reorganized to suit the needs and resources of a rural, village society. The mining landscape Within the mining community, various factors dictate the appropriation of the local landscape: natural resources; agriculture; production and dwelling sites; communications; ideology or religion. Colourful gossans and sparkling ores, along with fluxes such as manganese oxide, haematite and silica (which were used to lower the melting point and viscosity of the gangue during smelting) were among the most obvious features of the ancient mining landscape. The deep brown layers of the umber and black nodules of the manganese concretions take on further significance in a mining landscape (Given and Knapp forthcoming). Mining galleries usually require pit props, and smelting needs fuel: the production of metal, in other words, requires forests, not just trees. Examination of timber preserved in mining galleries at Skouriotissa has shown that Pinus brutia, a species that grows particularly well on Cyprus’s pillow lavas, was used most commonly as pit props (Constantinou 1992:61). Olive, oak and hawthorn were also common hardwoods in the Cypriot forests of antiquity; because they produce a long, hot burn and can be coppiced easily, they are ideally suited as a fuel source for roasting and as charcoal for smelting. The denuded landscape of certain parts of modern-day Cyprus, not far distant from the pillow lavas of the Troodos Mountains, may have resulted at least in part from four thousand years of copper mining on the island. The relationship between mining, settlement and landscape varies according to the scale and organizational level of production (Raber 1987:301–302). Thus the intermittent small-scale, localized production of the medieval period on Cyprus contrasts significantly with the larger-scale industries and major labour
Archaeology and anthropology of mining
17
forces typical of the Imperial Roman period. Moreover, it is often assumed that the collapse of the Roman Empire resulted in the cessation of the mining of copper sulphides for metal; recent survey work in Cyprus suggests that this presumption may have to be reassessed (Clough, in Knapp and Given 1996:308–316). When copper was produced and exported on a large scale, as it was during the Bronze Age and the Roman era, a viable and fully functional system of communications was essential. Whenever the production of copper exceeded the local or regional scale, mining communities were perforce linked into interregional, national or global networks of communication and exchange. Invariably this impacted on the mining landscape in the form of imports, migrant labour, possibly even new settlement plans in the face of other sociostructural changes. Gender and the archaeology of mining The basic tenet of a gendered archaeology must be that, if we can discuss and theorize about ethnicity, class, status, ideology or other social constructs in the material record, then the activities of women and men are also accessible to archaeology. Having accepted this premise, one must ask how archaeologists might make firm associations between the material remnants of mining activity and sex or gender. Susan Lawrence points out (this volume, Chapter 3) that any meaningful and defensible inferences about gender require the social context of the material remains to be established and the conditions under which any associations might pertain to be determined. And, whereas it remains essential to ‘add women’ and remove sexist agendas from archaeological discourse, it is equally important that archaeologists—female and male —accept the need to theorize gender, and to rethink accordingly their traditional research priorities and writing styles, as well as the objects and subjects of their study. A gendered archaeology must involve both women and men; it is an archaeology informed by feminism, one that looks critically at theories of human action and allows archaeological data to challenge existing social theory (Knapp 1998). The archaeology of mining has for so long presented men as gender-neutral (and thus as universal, or ‘human’) that it is nearly impossible to separate mining terms from their androcentric essence or to break down approaches that privilege techno-environmental phenomena. Yet this is precisely what we must attempt to do, not just in studies of prostitutes on the mining frontier, but also in the study of the everyday material record of miners and the mining community. As we know from the publications and fieldwork of an entire generation of social historians, anthropologists and historical archaeologists, women were almost always present in mining communities, working in both industrial and domestic contexts. In the tin mines of Bolivia, for example, miners always emphasized how firmly the women were committed to the historic confrontations between the workers and the bosses (Nash 1993). In precolonial African iron production, women provided essential labour for the preparation of
18
A.Bernard Knapp
ores and fuels, even if they were usually excluded from the overall metallurgical process because they lacked access to supernatural assistance (Herbert, Chapter 9, and Childs, Chapter 8, this volume). Perhaps a more focused, less biased examination of written and material resources will help us to challenge the dominant androcentric portrayal of the mining experience, and to realize that both women and men were fully integrated into the sociocultural mainstream of the mining community. Susan Lawrence has attempted to investigate the ways in which material culture was used to negotiate gender, and to consider how women’s roles helped to structure relationships within the community. By so doing she has demonstrated the archaeological reality of both women and men in the mining experience. Technology, mining and social process Innovations in technology had deep-seated and long-lasting social and ecological effects, placing constraints as well as conferring benefits on miners and mining. But the technology of mining involves much more than extracting minerals or metals from rocks, or describing the tools involved in crushing and smelting ores. Technology as a social process cannot be separated from the people involved, their daily life and human abilities, their ideology and beliefs, and their capacity to negotiate complex social, economic and political relationships within the mining context (Childs and Killick 1993). There exists a rich body of symbolism, based in part on the conspicuous and prestigious nature of metals such as copper or gold, in part on a complex set of rituals. Such symbolism offered multiple historically situated images to rural cultivators who first became involved with the extraction of native ores, to others who were enslaved by colonial regimes, or to yet others who became industrial workers answering the call of supply and demand from a global economy. From Galen’s now well-known account of his visit to the mines near Soli (Lefka), Cyprus, in the second century AD, to the tales and rituals associated with the supernatural powers (Huari, Supay) who ‘owned’ the hills worked by the Bolivian tin miners (e.g. Godoy 1991; Holloway 1996; Nash 1993), to the rich ethnographic accounts of ‘ancestors’ who developed and passed along to their kin the knowledge of metal-smelting processes in precolonial Africa, there is no doubt that mining and extractive metallurgy are activities with dynamic social and intellectual meaning, meanings which impacted significantly on the social and natural order of things and gave social structure to human existence.
CONCLUSION Archaeologists who emphasize the active role played by material culture in people’s past social practices (an approach that should be termed ‘social
Archaeology and anthropology of mining
19
archaeology’ —Knapp 1996b) feel that there are other directions which might be taken in examining the nature of a mining community; for example: 1 How were social life and settlement interlocked? 2 How did the entire social entourage of the mining community create, exploit, idealize, modify, and ultimately—in most cases—abandon the mining landscape? 3 Can the accounts presented by social historians and historical archaeologists inform prehistorians about individual human agency and historical process; about gender, cognition, ideology and mentalité? 4 How do inherently male-gendered terms like ‘miner’ and ‘prospector’ fit into the context of twentieth-century Bolivian tin production, for example, where the women expressed their sense of involvement by referring to themselves as ‘we miners’ (Nash 1993)? 5 Would archaeologists all agree that our discipline is social in nature, that social practices are concerned with meanings, and that social interpretations deal more with understanding than with explanation? Wylie (1993:15) has observed that the archaeological record has an ‘infinite capacity to surprise’ and to shape or constrain the interpretation of the past. Archaeological interpretations, moreover, will continue to have multiple voices even as we aspire to find common ground among different archaeologies. The ‘meanings’ of material things in the past, as in the present, were not fixed but instead were ‘read’ into them from other experiences and expectations, and then acted upon. Social life, therefore, is enacted within material conditions which are constantly reworked according to different people’s interpretations of material reality. The archaeology of mining requires us to examine a range of factors, social and cultural, physical and technical. By considering together information on mining settlements, agricultural resources, communications routes and gendered activities, we may view technology in its social context, and sharpen our vision of the mining landscape. Finally, we should consider the ‘hidden histories’ of mining (Paul Sant Cassia, personal communication), and the symbolic construction of the mining community. It is also essential to examine the collective representations of miners about memory and the past, and to consider how miners anticipated the future. Mining as a politico-economic activity had the capacity to establish social relations between individuals that transcended community boundaries and transformed regional interactions. Mining villages often were built rapidly and carelessly to satisfy entrepreneurial or state needs; occasionally they evolved from agricultural villages in cycles of boom and bust, when villagers and farmers become urban workers or industrial pawns. The cycles of mining and agriculture in many mining regions suggest that we must concentrate not just on mining as an activity, but also on the history of the people who made up the mining community.
20
A.Bernard Knapp
ACKNOWLEDGEMENTS I wish to thank David Killick and Vincent Pigott for their help in organizing the Bellagio Conference and for input on various ideas discussed in this chapter. I am grateful to Suzanne Young and Eugenia Herbert for very helpful comments on a first draft of the chapter. I also wish to express my sincere thanks to everyone at the Rockefeller Foundation’s Bellagio Study and Conference Center, namely the Director, Pasquale Pesce, the Assistant Director, Gianna Celli, and the Conference Assistant, Linda Triangolo. Thanks also to Susan Garfield, Manager of the Rockefeller Foundation’s Bellagio Center Office in New York. The original draft of this study, and the task of co-organizing the conference, were undertaken while the author held an Australian Research Council Postdoctoral Fellowship at Macquarie University, Sydney: I am most grateful for all the support and infrastructure provided to me during that time.
REFERENCES Bronson, B. (1992) Patterns in the early southeast Asian metals trade. In I.Glover, P. Suchitta and J.Villiers (eds), Early Metallurgy, Trade and Urban Centers in Thailand and Southeast Asia, pp. 63–114. Bangkok: White Lotus. Brumfiel, E.M. and Earle, T. (eds) (1987) Specialization, Exchange and Complex Societies. Cambridge: Cambridge University Press. Bulmer, M.I.A. (1975) Sociological models of the mining community. Sociological Review 23: 61–92. Childs, S.T. and Killick, D. (1993) Indigenous African metallurgy: nature and culture. Annual Review of Anthropology 22: 317–337. Constantinou, G. (1992) Ancient copper mining in Cyprus. In A.Marangou and K. Psillides (eds), Cyprus, Copper and the Sea, pp. 43–74. Nicosia: Republic of Cyprus. Costin, C.L. (1991) Craft specialization: issues in defining, documenting, and explaining the organization of production. In M.B.Schiffer (ed.), Archaeological Method and Theory 3, pp. 1–56. Tucson: University of Arizona Press. Craddock, P. (1995) Early Metal Mining and Production. Edinburgh: Edinburgh University Press. Deiner, P. (1980) Quantum adjustment, macroevolution, and the social field: some comments on evolution and culture. Current Anthropology 21: 423–443. Domergue, C. (1989) Minería y metalurgia en las antiguas civilizaciones mediterráneas y europeas. Colloque Internacional Asociadio, Madrid, 24–28 October 1985. Madrid: Ministerio de Cultura, Dirección General de Bellas Artes y Archives, Instituto de Conservación y Restauración de Bienes Culturales. Douglass, W. (1979) The mining camp: a unique social phenomenon. Unpublished paper presented at the Southeastern Anthropological Association meeting, Santa Barbara, California. Dutton, A. and Fasham, P.J., with Jenkins, D.A., Caseldine, A.E. and Hamilton-Dyer, S. (1994) Prehistoric copper mining on the Great Orme, Llandudno, Gwynedd. Proceedings of the Prehistoric Society 60: 245–286. Edmondson, J.C. (1989) Mining in the late Roman Empire and beyond: continuity or disruption? Journal of Roman Studies 79: 84–102.
Archaeology and anthropology of mining
21
Ehrenreich, R.M. (ed.) (1991) Metals in society: theory beyond analysis. MASCA Research Papers in Science and Archaeology 8: 2. Philadelphia: University Museum, University of Pennsylvania. ——(1996) Mining communities in history: an industrial legacy. Journal of Metals 48(12): 54–56. Eliade, M. (1978) The Forge and the Crucible, 2nd edn. Chicago: University of Chicago Press. Frazier, C.A. (with Brown, R.K.) (1992) Miners and Medicine: West Virginia Memories. Norman: University of Oklahoma Press. Gamble, C. (1986) Hunter-gatherer studies and the origin of states. In J.A.Hall (ed.), States in History, pp. 22–47. Oxford: Blackwell. Given, M. and Knapp, A.B. (forthcoming) The Sydney-Cyprus Survey Project and the archaeology of mining. In G.C.Ioannides (ed.), Proceedings of the Third International Conference on Cypriot Archaeology (Nicosia, 16–20 April 1996). Nicosia: Society of Cypriot Studies. Godoy, R. (1985) Mining: anthropological perspectives. Annual Review of Anthropology 14: 199–217. ——(1991) Mining and Agriculture in Highland Bolivia: Ecology, History and Commerce among the Jukumanis. Tucson: University of Arizona Press. Hardesty, D.L. (1988) The Archaeology of Mining and Miners: A View from the Silver State. Society for Historical Archaeology, Special Publication Series no. 6. Pleasant Hill, CA: SHA. ——(1992) The miner’s domestic household: perspectives from the American West. In K.Tenfelde (ed.), Towards a Social History of Mining in the 19th and 20th Centuries (International Mining History Congress, Bochum, 3–7 September 1989), pp. 180– 196. Munich: Beck. Herbert, E.W. (1993) Iron, Gender, and Power: Rituals of Transformation in African Societies. Bloomington: University of Indiana Press. Heskel, D. and Lamberg-Karlovsky, C.C. (1980) An alternative sequence for the development of metallurgy: Tepe Yahya, Iran. In T.A.Wertime and J.D.Muhly (eds), The Coming of the Age of Iron, pp. 229–265. New Haven, CT: Yale University Press. Hine, R. (1980) Community on the American Frontier. Norman: University of Oklahoma Press. Holloway, M. (1996) I am rich Potosí, king of the mountains, envy of kings. Natural History 105(11): 36–43. Hosler, D. (1995) Sounds, color and meaning in the metallurgy of ancient Mexico . World Archaeology 27: 100–115. Knapp, A.B. (1988) Ideology, archaeology and polity. Man 23: 133–163. ——(1996a) Power and ideology on prehistoric Cyprus. In P.Hellström and B.Alroth (eds), Religion and Power in the Ancient Greek World, Boreas: Acta Universitatis Upsaliensis 24: 9–25. Uppsala: Department of Classical Archaeology and Ancient History. ——(1996b) Archaeology without gravity? Postmodernism and the past. Journal of Archaeological Method and Theory 3: 127–158. ——(1997) The Archaeology of Late Bronze Age Cypriot Society: The Study of Settlement, Survey and Landscape. Glasgow: University of Glasgow, Department of Archaeology, Occasional Paper 4. ——(1998) Who’s come a long way, baby? Gendering society, gendering archaeology. Archaelogical Dialogues 5. Knapp, A.B. and Given, M. (1996) A report on the Sydney-Cyprus Survey Project (SCSP) —third season (1995). Report of the Department of Antiquities, Cyprus: 295–366. Lankton, L. (1992) Cradle to Grave. Oxford: Oxford University Press. Lavender, D. (1962) The Story of the Cyprus Mines Corporation. San Marino, CA: Huntington Library. Lechtman, H. (1991) The production of copper-arsenic alloys in the Central Andes: highland ores and coastal smelters? Journal of Field Archaeology 18: 43–76.
22
A.Bernard Knapp
——(1996) Arsenic-bronze: dirty copper or chosen alloy? Journal of Field Archaeology 23: 477–514. LeRoy Ladurie, E. (1979) The Territory of the Historian. Chicago: University of Chicago Press. MacMillan, G. (1995) At the End of the Rainbow: Gold, Land, and People in the Brazilian Amazon. New York: Columbia University Press. McNutt, P.M. (1990) The Forging of Israel: Iron Technology, Symbolism and Tradition in Ancient Society. JSOT Supplement Series 108. Social World of Biblical Antiquity 8. Decatur, GA: Almond Press. Nash, J. (1993) We Eat the Mines and the Mines Eat Us: Dependency and Exploitation in Bolivian Tin Mines, 2nd edn. New York: Columbia University Press. Paul, R. (1963) Mining Frontiers of the Far West, 1848–1880. New York: Rinehart & Winston. Pfaffenberger, B. (1992) The social anthropology of technology. Annual Review of Anthropology 21: 491–516. Raber, P. (1987) Early copper production in the Polis region, western Cyprus. Journal of Field Archaeology 14: 297–312. Reps, J. (1975) Bonanza towns: urban planning on the western mining frontier. In R. E.Ehrenberg (ed.), Pattern and Process: Research in Historical Geography, pp. 271–289. Washington, DC: Howard University Press. Roberts, B.K. (1996) Landscapes of Settlement: Prehistory to the Present. London: Routledge. Ruse, M. (1996) Monad to Man: The Concept of Progress in Evolutionary Biology. Cambridge, MA: Harvard University Press. Schmitt, D.N. and Zeier, C.D. (1993) Not by bones alone: exploring household composition and socioeconomic status in an isolated historic mining community. Historical Archaeology 27: 20–38. Shapin, S. (1996) The Scientific Revolution. Chicago: University of Chicago Press. Shaw, I. (1994) Pharaonic quarrying and mining: settlement and procurement in Egypt’s marginal regions. Antiquity 68(258): 108–119. Shennan, S.J. (1991) The excavation of an early bronze age settlement at St Veit-Klinglberg, Land Salzburg, Austria: an interim report. Antiquaries Journal 69: 205–224. ——(1993) Commodities, transactions, and growth in the central-European Early Bronze Age. Journal of European Archaeology 1(2): 59–72. Shepherd, R. (1993) Ancient Mining. London: Chapman & Hall. Sheridan, T.E. (1992) The limits of power: the political ecology of the Spanish Empire in the greater southwest. Antiquity 66(250): 153–171. Simmons, A. (1989) Red Light Ladies: Settlement Patterns and Material Culture on the Mining Frontier. Anthropology Northwest, no. 4. Corvallis: Department of Anthropology, Oregon State University. Smith, M.E. (1992) Braudel’s temporal rhythms and chronology theory in archaeology. In A.B.Knapp (ed.), Archaeology, Annales, and Ethnohistory, pp. 23–34. Cambridge: Cambridge University Press. Stech, T. (1985) Copper and society in Late Bronze Age Cyprus. In A.B.Knapp and T.Stech (eds), Prehistoric Production and Exchange: The Aegean and East Mediterranean. UCLA Institute of Archaeology, Monograph 25: 100–105. Los Angeles. Taussig, M. (1980) The Devil and Commodity Fetishism in South America. Chapel Hill: University of North Carolina Press. Tenfelde, K. (ed.) (1992a) Towards a Social History of Mining in the 19th and 20th Centuries (International Mining History Congress, Bochum, 3–7 September 1989). Munich: Beck. ——(1992b) The mining community and the community of mining historians. In K. Tenfelde (ed.), Towards a Social History of Mining in the 19th and 20th Centuries (International Mining History Congress, Bochum, 3–7 September 1989), pp. 1201–1215. Munich: Beck. Thompson, E.T. (1932) Mines and plantations and the movement of peoples. American Journal of Sociology 37: 603–611.
Archaeology and anthropology of mining
23
Weigand, P.C. (1982) Introduction: Special Issue (P.C.Weigand and G.Gwynne, eds) — Mining and Mining Techniques in Ancient Mesoamerica, Anthropology (Stony Brook) 6(1–2): 1–6. Weisgerber, G. (1982) Towards a history of copper mining in Cyprus and the Near East: possibilities of mining archaeology. In J.D.Muhly, R.Maddin and V.Karageorghis (eds), Early Metallurgy in Cyprus, 4000–500 BC, pp. 25–32. Nicosia: Pierides Foundation. ——(1989) Montanarchäologie. Grundzüge einer systematischen Bergbaukunde für Vorund Frühgeschichte und Antike, Teil 1. In A.Hauptmann, E.Pernicka and G.A. Wagner (eds), Old World Archaeometallurgy, Der Anschnitt 7: 79–98. Bochum: Deutsches BergbauMuseum. Weisgerber, G. and Pernicka, E. (1995) Ore mining in prehistoric Europe: a short overview. In G.Morteani and J.P.Northover (eds), Prehistoric Gold in Europe, pp. 159–182. Dordrecht: Kluwer. Wylie, A. (1993) Invented lands/discovered pasts: the westward expansion of myth and history. Historical Archaeology 27(4): 1–19.
Part I
Historical archaeology
Chapter 2
The fabric and structure of Australian mining settlements Peter Bell
ABSTRACT There seem to be distinctive elements identifiable in the plans of towns and the forms of buildings that served the mining industry throughout the world in the nineteenth century. However, both the settlement plans and the built forms are also subject to considerable diversity. In some individual cases, the fabric and form of mining settlements can be attributed variously to economic, ethnic or cultural causes. This chapter begins to analyse the pattern of nineteenth-century mining settlement in Australia by identifying some circumstances in which each of these forces may be expected to operate more effectively than the others. It also looks briefly at the distortion of evidence that can arise from the processes of decay on settlement sites.
INTRODUCTION AND HISTORICAL BACKGROUND Why should mining create a distinctive cultural landscape? There are a number of reasons. First, mining industry obeys different geographical imperatives. Geographers have created a number of useful models which, given parameters of climate and terrain, can accurately predict the location of farms, agricultural settlements, towns, ports and roads. But one cannot model the location of mining industry. Mines are located where the minerals are; they cannot be anywhere else. And frequently the location of those minerals takes settlement into regions where farmers and pastoralists do not go: across Australia there have been mines and mining settlements in utterly waterless deserts, in tropical rainforest and on frozen mountain tops. Second, the economic impact of mining happens very quickly. Population and wealth can both boom remarkably in a very short time. In 1851, when gold was discovered, the population of the entire colony of Victoria was 97,000. In the following decade the population increased by more than five times, to 540,000 in 1860. In that same decade, Victoria produced something like 25 million ounces of gold, worth about £100,000,000 that simply had not existed previously: in every year a new £10,000,000 of wealth was being created (Serle 1963:382, 390).
28
Peter Bell
A third fundamental difference between mining and most other forms of economic activity is its ephemeral nature. Mining is non-sustainable, inevitably depleting and sometimes extinguishing the commercial resource. This means that all mines close, sometimes after more than a century like the silver—lead mines of Broken Hill, sometimes after a few weeks. To describe the mining industry truthfully, historians must keep fixed in mind the reality that most mines are failures. Probably about one in a hundred attempts at mining makes a profit; the other ninety-nine are abandoned as failures. But most mining history books concern themselves predominantly with the very rare cases that succeeded. Alluvial mining is the most ephemeral industry of all. One of the truly remarkable population movements in history was the series of great alluvial gold rushes that swept around the world in the second half of the nineteenth century. Hundreds of thousands of participants created a transient society that remained in constant movement for decades across the Americas, Australia, Africa and Oceania. Bill Douglass described it in the American West as ‘a community without a locus…diffused over a vast geographical region and crystallised in the form of a camp for brief periods of time around this or that discovery’ (Douglass 1979:13; this volume, Chapter 6). It flowed like a cloud of insects, apparently formless, but with a single purpose. People in places as remote as Coolgardie, Skagway and Johannesburg were linked into a single world community by their common motivation. No one organized the gold rushes. They had no leader. They arose from the collective free will of many thousands of people who saw an opportunity to improve their lot, and responded spontaneously. In the nineteenth century, such behaviour was very un-Australian. Australia had been settled by Britain largely in response to the geopolitical demands of the British Empire’s worldwide rivalry with that of the French. The colonies of Australia were short on free enterprise and democracy, and strong on law and order, with their civil governments backed up by the imperial army and navy. There were certainly no large numbers of people moving about the country exercising their free will in pursuit of wealth. In this quiet, orderly backwater of the empire, the onset of the gold rush was greeted with apprehension at times bordering on terror. The sudden influx of gold-diggers threatened anarchy and disorder, crime and drunkenness, and, perhaps worst of all, egalitarianism and Americaninspired notions of democracy. Most of the affected colonial administrations initially tried to prevent the gold rush occurring, and when that failed, fell back on ways of managing it by regulating, taxing and controlling whatever aspects of it they could. The first fifty years of gold mining in Australia were played out amid a constant tension between the free and unpredictable will of the mining community, and anxious government attempts to control what was happening.
Fabric and structure of Australian mining settlements
29
AUSTRALIAN MINING SETTLEMENTS There were thus a number of elements that characterized mining-related settlements and distinguished them from other colonial communities. Their location was determined by the mineral resource, so that they were often in difficult and inhospitable places. Frequently they were the first form of settlement in the region, so that no previous infrastructure was in existence. They generally came into being suddenly and with great excitement. There was a dramatic population rise and an influx of wealth, both of which inflated the local economy so that goods and services became very expensive. Mining settlements usually did not last long, and, more importantly, no one expected them to. All economic decisions made in a mining town took into consideration the likelihood of the mine closing in the near future, and their population was free, cosmopolitan and highly mobile. The fact that mining settlements were distinctive does not mean that they were uniform. There is very great diversity in the nature of what is loosely called the mining industry. The fundamental difference between alluvial and underground mining had major implications for the kinds of settlement they generated. The population of alluvial settlements was usually predominantly male, and their stay in one place was usually short. There was little confidence in the future, so there was not much incentive to build anything substantial. The industry rarely generated much commercial infrastructure, so the supply of material goods was limited, and everything was terribly expensive. Contrary to romantic folklore, alluvial mining fields were not prosperous places. On some fields, a few of the early arrivals achieved a healthy income, but even here the average income typically dropped to perhaps a third of a labourer’s wage within a few years (Serle 1963:391–392; Bell 1984:37–38). In underground mining settlements the population was likely to be more evenly gender-balanced, including a fair proportion of married couples with children. Most people worked for mining companies at fixed wages, so there was a more stable domestic economy. The stability of underground mining towns varied considerably, but they extended along a spectrum from the smaller base-metal towns, where the way of life was only marginally more secure than that of the alluvial camps, up to the grand gold-mining towns such as Bendigo, Charters Towers and Kalgoorlie, which were economically as prosperous and culturally as sophisticated as any cities in the country. These complexities were reflected in the diversity of miners’ housing in Australia, whose fabric ranged from stone masonry, through flattened oil drums, to holes in the ground. A recurring motif of the industry was that its ephemeral nature discouraged participants from investing either money or time in a substantial dwelling. Miners’ buildings had to accommodate a transient lifestyle. Perhaps surprisingly, there is relatively little evidence of Australian miners building in primitive materials; that is, material gathered in the local area and worked by hand. This was generally the first choice of farmers and
30
Peter Bell
agricultural workers, who improvised shelters of bark, split slabs and thatch, or adobe and other forms of earth construction. For a miner, such materials had the twin disadvantages of being time-consuming and non-portable. In the early stages of an alluvial rush, every minute was precious in the essential business of proving the ground and selecting a claim; there was no time for searching the bush for tree bark. If the choice of location proved unfortunate, a primitive habitation and the labour invested in it had to be abandoned. The solution almost universally adopted was the tent. Illustrations of alluvial diggings and the early stages of underground mining settlements, from Ballarat in the 1850s to the Golden Mile in the 1890s, invariably show a landscape of hundreds or thousands of small tents. These could be erected in a few minutes, and loaded back onto a horse in the same time. They were also suited to the many parts of Australia where the landscape does not provide even sticks and grass as building materials. Tent-makers became conspicuous among the tradespeople who followed the mining rushes. The early Cornish copper-miners of South Australia established another tradition, of excavating dwellings into hillsides and creek banks, and living in underground houses. In the mining town of Burra about 2,000 people, nearly half the town’s population, lived in burrows. A census in early 1851 reported that in one part of Burra, ‘[t]here are no houses, the dwellings being excavated in the banks of the Burra Creek’ (Auhl 1986:126). This development of course signalled that the size and wealth of the Burra copper lode had been proved, and there was an economic incentive to invest labour in a dwelling in anticipation of a long stay. The underground house tradition was very durable in South Australia; it was taken north to the copper mines of the Flinders Ranges, and as late as the 1960s opal miners in the towns of Andamooka and Coober Pedy were excavating their houses to escape the desert heat in an arid landscape that provided no building materials other than the ground itself. (Indeed, houses are still being excavated in Coober Pedy today, but principally for the admiration of tourists.) Despite this diversity, there is such a thing as a typical miner’s dwelling in Australia, more durable than a tent and less exotic than a hole in the ground. As we rise up the spectrum of prosperity in the mining industry, we also enter the economic sphere of industrialization. An underground mining company had an enormous demand for manufactured materials—engines and other machinery, bricks, nails, glass, sawn timber and sheet iron—in addition to the fuel, explosives and tools that it consumed daily. Many of these materials had to be transported great distances. Bricks were usually made locally, although there are specialized refractory bricks from Glasgow and Stourbridge scattered in unlikely places across the Australian outback. In the nineteenth century most of Australia’s nails, hinges, doorknobs and glass came from Europe, and every sheet of corrugated iron used in Australia before 1921 was imported. There are very few places in Australia where a company could saw its own building timber, let alone its pit props, in the local area. In the mines under Broken Hill
Fabric and structure of Australian mining settlements
31
there is a forest of Douglas fir, shipped as mine timber from Seattle and Portland (Australians still call it Oregon pine). Inevitably, the mining industry had to create an international supply infrastructure. Sea freights were very low in comparison to land carriage, so the cost differential between Australian-made products and imported goods was slight, and there was little incentive for local manufacturing. Within this global market, the costs of industrially processed building materials came within the reach of the wage earner, and the Australian miner working for a reasonably stable underground mining company could expect to live in a modest cottage of sawn timber with a corrugated iron roof. Its form and materials were established even before the gold rushes began, for in 1849 the Burra company was building cottages for its employees out of sawn timber imported seemingly impossible distances from Europe and the Atlantic seaboard of North America (Bell 1990). From the outset, the standard form of these houses was a two-roomed plan derived directly from the rural cottages of English farm labourers. So inextricably is this house form linked with mining in the Australian mind that today, 150 years after its adoption by the industry, it is known almost universally as a ‘miner’s cottage’, even in districts that never in their history saw a miner. It is found built in a variety of materials, but most commonly has a frame of sawn timber studs or light posts, clad at first with sawn weatherboards, and in later decades with corrugated iron. This light stud frame was also a product of rural England, used since the eighteenth century in the southeastern counties for small buildings made of Scandinavian timber. In Australia it has sometimes been postulated that the building technique is the ‘balloon frame’, imported from the USA during the gold rushes, but this theory is simply wrong. The light stud frame had been in use in Australia and New Zealand since the 1820s, ten years before the so-called balloon frame is said to have been invented in Chicago (Bell 1984, 1990). One sometimes reads about gold rush ‘boom-style’ architecture, and it is true that buildings in some of the great mining towns had a certain brashness, especially evident in the design of buildings such as town halls, hotels and theatres. But this very rarely extended to housing. When a mining town was experiencing great prosperity, demand for labour was high and the population was increasing rapidly, so large numbers of houses were being built. But the cost of materials and particularly of labour were also very high, so these houses were small, cheap and standardized. Ironically, the most characteristic architectural product of an economic boom in an Australian mining town was a suburb of identical timber-framed two-roomed cottages.
ETHNICITY AND SETTLEMENT The role of ethnicity in determining the forms of miners’ housing in Australia has never been explored satisfactorily. In the clamour of the gold rushes, specific
32
Peter Bell
ethnic groups found it difficult to assert their identity amid the great cosmopolitan throng that surrounded them, and in the face of powerful economic forces at work enforcing standardization. The result was that a German miner and a Welsh smelter in Mount Morgan would most likely have lived in identical houses, with little choice available to them. However, there are specific ethnic groups associated with the mining industry in Australia, such as the Cornish and the Chinese, who appear to have built settlements which have some consistent recurring characteristics despite being widely separated in space and time. The triggering circumstance enabling them to assert an ethnic identity appears to be that they were present in sufficient numbers to dominate the local population. In some early copper-mining towns such as Kapunda and especially Burra, which are known to have had a significant Cornish population, there are rows of attached stone cottages dating from the 1840s, identified as examples of traditional Cornish architecture. They do indeed resemble row housing in some mining towns in Cornwall and Devon. The problem with this ethnic determinist theory is that twenty years later a far more intensively Cornish enclave developed in Australia at the Moonta and Wallaroo mines; in 1873 Anthony Trollope wrote, ‘so many of the miners were Cornishmen as to give Moonta and Wallaroo the air of Cornish towns’ (Payton 1984:73). But Moonta and Wallaroo have no such row cottages. The most likely explanation is economic rather than ethnic. The rows of stone cottages in Burra and Kapunda were built by the mining companies, anxious to attract labour from Cornwall in the earliest years of overseas recruitment to the Australian mining industry. A generation later, the Moonta and Wallaroo companies had the easier task of recruiting from Cornish mining towns which had sunk into deep depression, and were far more hard-nosed about housing policy: they built no houses at all. The ‘Cornish’ houses of Kapunda and Burra are not really Cornish buildings, but rather company buildings. Australian mining companies were not the least bit interested in providing housing to employees in the second half of the nineteenth century. After these few experiments in South Australia in the 1840s (and even there the companies’ concern for their employees’ welfare fluctuated over time in inverse proportion to the supply of labour), company-built housing was virtually unknown for eighty years. The American phenomenon of the company town did not exist in Australia in the second half of the nineteenth century. Not until the 1920s did mining companies such as Mount Isa Mines and the Electrolytic Zinc Company in Hobart begin to provide housing to employees on any significant scale. Another ethnic group present in large numbers on the nineteenth-century Australian mining fields was the Chinese. From the 1850s, there were significant numbers of Chinese on the Victorian goldfields, typically 15 to 25% of the population on most fields. They suffered from discriminatory legislation; most tended to be poor, and they lived communally. Descriptions of their housing suggest they lived exactly as European miners did: at first in tents, then
Fabric and structure of Australian mining settlements
33
mostly in sawn timber huts with iron roofs. A few of the poorer Chinese used bark or scrap materials for shelter. But there are no descriptions of the Chinese in Victoria building culturally distinctive houses; only temples and clubhouses were built in the Chinese architectural tradition. However, this was to change in later decades. In the 1870s, a second phase of Chinese immigrants arrived on the mining fields of northern Australia, numerous and well organized. On the sparsely settled tropical goldfields they quickly came to dominate the population: at the 1876 census, the Palmer goldfield was 73% Chinese, and in the Northern Territory, the entire population was over 50% Chinese from 1879 to 1910 (Jones 1990:131). Here the Chinese established their own commercial infrastructure, supplied by steamers arriving from Chinese ports that were little further away than the cities of southern Australia. On these fields, conditions were right for the Chinese to import their material culture intact, and Chinese miners lived in huts of bush timber and thatch that were modelled on buildings from Guangdong province. We know this principally from written evidence and a few photographs, for the archaeological evidence left by thatch huts in a tropical climate is very slight. What we find on the ground today is the evidence of how the thatch was attached to the frame: a thin scatter of metal fasteners improvised from wire, nails, machinery parts and discarded horseshoes. The most monumental evidence is in the forms of hearths—knee-high structures of earth and stone, one to each hut site—and pig ovens, much rarer large stone cylinders used for roasting pigs on feast days. These Chinese settlements are the most culturally distinctive sites created by the mining industry in Australia. The lack of comparable sites on the older goldfields strongly suggests that there is a threshold proportion of the population below which a particular ethnic group is lost in the background, but above which the group can successfully preserve its material culture. The layout of Australian mining settlements Just as individual miner’s houses were extremely diverse in their design and materials, so were the overall forms of mining settlements. Here the diversity can generally be explained by the variety of design processes at work; some mining towns were planned by government surveyors, some by a mining company’s engineers, and some simply grew spontaneously with no plan at all. The Australian tradition of strong government control manifested itself in a preoccupation with orderly town planning. The square grid plan was the preferred model for the military surveyors who laid out towns like Adelaide with geometric precision. It was hardly an innovation: the grid plan was used by the ancient Greeks to lay out their colonial towns, and Vitruvius recommended geometrical town plans in the first century BC. Each Australian colony provided its surveyors with a manual with a name like ‘Handbook for Government Surveyors’, which gave instructions for laying out towns. So
34
Peter Bell
carried away was the Queensland Mines Department by the need for orderly planning that in 1892 a mining surveyor was sent around the Etheridge goldfield to lay out town plans ‘at any likely permanent camp’ in case there was ever a mineral discovery in the vicinity (Bell 1984:31). Thus, ideally, Australian mining settlements were laid out in a neat grid of wide, straight streets, and there are plenty of examples: Broken Hill, Kalgoorlie, Georgetown, Cloncurry, Moonta, Wallaroo. Of course there were many cases where topographical constraints made this impossible, and many towns, like Walhalla in Victoria and Irvinebank in Queensland, were forced into tortuous plans between hillsides and creek banks with allotments laid out wherever there was enough flat land for a building. There were many cases where the main determinant of the town plan was outright anarchy. Ravenswood in Queensland, seen from the air, looks like a wavering spider’s web radiating outwards from a centre. On the ground, the centre is the most convenient creek crossing, where a bridge was built and where all tracks converged. From there, the community made their own way to the diggings, and a network of tracks evolved. Buildings were built along them, and the whole plan was a fait accompli before the government surveyor arrived. Not uncommonly, there were effectively two towns a short distance apart. At Burra in the 1840s the mining company monopolized the land, and surveyed the neat town of Kooringa for the workforce. But it was undercut by rival entrepreneurs who surveyed their own land subdivisions just outside the company’s lease, creating a separate community with suburb names like Redruth, Aberdeen and Llwchwr calculated to attract tenants who saw themselves as ethnically distinct from their English employers (Auhl 1986). At Moonta in the 1860s there was a neat government grid plan laid out. But there was no reason for anyone to live in the government town; they could not be compelled to do so. A large part of the population preferred to live on the company’s lease, where they built their own cottages along their own streets, creating a spontaneous, unstructured tangle of settlement immediately alongside the government town. To this day it is called Moonta Mines to distinguish it from the official town. Even more complicated was the settlement that grew up at the Sons of Gwalia gold mine in the Western Australian desert at the end of the century. Here there were three towns, two created by the government and one by the miners. The government surveyor laid out two town plans, one called Leonora, which was to be the administrative and commercial centre for the district, and one called Gwalia, which was to be the dormitory suburb near the largest mine. The Sons of Gwalia became a very large mine which completely dominated the district, and its manager (who incidentally was Herbert Hoover, thirty years before he became president of the USA) inadvertently divided the community into ethnic groups by hiring an Italian and Croatian workforce. Leonora became the ‘British’ town for the mine officers and business community, while Gwalia became the Italian/Croatian town where the miners and their families lived. In an attempt to centralize the community, only leasehold land was made
Fabric and structure of Australian mining settlements
35
available in Gwalia, so there was no incentive to build substantially, and the houses tended to be smaller and simpler than those in Leonora, which were on freehold land. But even Gwalia was too far from the mine for many of the workers, especially single men, who built themselves cottages on the mining company’s lease, a few hundred metres from the headframe. These bachelor establishments tended to be even smaller and meaner than those in Gwalia, and some had no kitchens or bathrooms, as the men ate and washed at boarding houses in the town. Many of the houses were built of cast-off materials from the mine; filter cloth from Oliver filters was used to line most of the rooms. There were other advantages besides proximity to work, for the company allowed residents to help themselves to firewood from the mine stocks and water from the mine’s condensers. This untidy settlement naturally drew the indignation of the mining warden, who reported in 1901: They of course would prefer having their camps right at the mouth of the main shaft if they were allowed to, but for many reasons it has become necessary that their camps should be erected with some system and therefore they will have to remove into the residential area by degrees. (Bell et al. 1985:62) They never did, and the warden had no powers to compel them. When the mine finally closed in 1963, the community was still divided into three, despite sixty years of government threats and inducements. If a mining settlement has a composite plan, part chaotic and part orderly, it is tempting to assume that the untidy part is earlier, and the surveyor later tidied it up. This is not a safe assumption. Charters Towers in Queensland was laid out in early 1872 as a regular square grid of streets and mining leases, with some allowance for the topography. Then in 1888 a new reef, the Brilliant, was discovered outside the surveyed area. The town expanded to the east, and for some reason on this occasion the government surveyor arrived too late (Roderick 1980). Charters Towers now has a street plan derived partly from administrative tidiness and partly from the free will of the miners.
MINING SETTLEMENTS: DECLINE AND DECAY What we see in a nineteenth-century mining town today of course reflects not only the process of settlement creation, but the process of decay, which has usually proceeded for an even longer period. Ravenswood was an active mining town for fifty years, then the underground mines were abandoned for seventy years, and they have now been reopened as opencut mines for ten. The town is now as much a product of the seventy years of decay as of the fifty years of construction and use.
36
Peter Bell
While a mining operation is continuing, conservation of the mine and its settlement is undertaken by the company and the community in the course of its everyday operations. The normal processes of care and maintenance represent the most sympathetic regime that the structures on a historic mining site are ever likely to have. Of course, at the time no one regards the company’s activities as the conservation of a historic resource. When the mining operation ceases, the process of destruction commences. Normally a large proportion of buildings on the site are immediately removed for use elsewhere, or sold for scrap. Care and maintenance is discontinued, and decay begins immediately. When a site is first abandoned, natural deterioration usually proceeds very rapidly at the same time as the removal of portable items from the site is at its peak. The fastest processes of deterioration are those affecting organic materials: timber components succumb rapidly to fungal decay, fire and insect attack. Corrosion of metals follows next, although some, such as cast iron, are often very durable even in damp conditions. Masonry elements such as stone and brick survive longest. On many mining sites, most of the natural deterioration happens within the first five years. When a site has been abandoned for many decades, it usually reaches a state of near-equilibrium in which the rapid deterioration has already occurred, and the more durable parts that are left, typically consisting of masonry, cast iron and heavy timbers, are in a slow state of natural decay. Human intervention continues. On abandoned mines, people visit the site for many years to take away useful industrial items such as building materials, machinery parts, drill steel, ore trucks and wire rope. Some of these visitors are engaged in what might be thought of as industrial recycling; others are simply pilfering souvenirs. Most minesites are quickly stripped of distinctive and portable items such as mining tools and assay crucibles. Houses are stripped of useful parts such as doors, plumbing and light fittings. Iron is taken from roofs, and timber-framed buildings lend themselves to complete dismantling; that is one of the reasons why miners often chose to live in them. One recent and highly destructive practice is the theft of bottles and other small artefacts from historic sites. Besides the damage done to the site and the unsightly aftermath left by digging, irreplaceable historical information is removed. Even in the rare cases where the stolen material is deposited in a museum or other public collection, the archaeological information conveyed by its location and context is invariably lost for ever. Bottle-stealing has increased dramatically in the past two decades, and general minor vandalism has become much more frequent. Many sites which were still remarkably intact in the 1970s have been devastated since four-wheel drive vehicles became commonly available. The selective nature of the decay processes often distorts the historical evidence. The durability of materials determines their rate of survival in the archaeological record, and this can create a false impression of their original
Fabric and structure of Australian mining settlements
37
prevalence. We learn from documentary sources that Burra in 1850 was a town of new timber houses, with a few stone houses among them. Today only one early timber house is left, but all the stone ones are still standing; the visitor is given the impression that early Burra was a town built in stone. In the tropical north of Australia, where the lives of settlements tended to be short and most building materials were flimsy, there is often very little evidence to be found. Maytown, the principal settlement of the fabulous Palmer goldfield, is a clearing in the bush. The most monumental remains in the town site are the stone kerbs along the main street. Sometimes nothing survives of smaller towns but exotic vegetation; fig trees, mangoes and vivid flowering bougainvillea constitute the entire physical evidence of some nineteenth-century mining communities.
CONCLUSION Since the mid-nineteenth century, the mining industry has been one of the most powerful forces shaping Australia’s economy and society. The physical evidence of its effects is to be found in mining settlements, both extant and abandoned, throughout virtually every part of the Australian continent. However, despite the burgeoning of the discipline of historical archaeology in the past twenty years, and the large number of research projects which have investigated mining sites for cultural resource management or environmental impact purposes, there is as yet no systematic study of Australian mining settlements in existence (but cf. Lawrence, this volume, Chapter 3). There were some elements common to all mining communities which assisted in shaping their physical structure, but there are limits to the usefulness of these elements as predictors of the forms of settlement that will arise. It is evident that there are also more pervasive differences which act to create diversity. This chapter has described some of the resulting divergent forms of miners’ housing in Australia, and the larger context of the surrounding mining settlements, and investigated some of the origins of their diversity. Some tentative conclusions can be drawn about the roles of such forces as the chronic ephemerality of the industry, of the inflated economy created by mining, of the psychological state of its participants, and of the commercial infrastructure which supplied the mining community. Although I shall not descend into the geographical determinism which plagues much architectural history, it is evident that topography and climate played a part. It is also apparent that the tension between government and free enterprise was a crucial element in determining the forms of settlements. But given this state of background tension, it is not always clear why one or the other came to dominate in a particular case. Ethnicity suggests itself as a determinant in some circumstances, but needs more careful analysis; some of what seem to be the manifestations of
38
Peter Bell
‘Cornishness’, for example, are in fact more likely to be the product of nineteenth-century corporate culture. The miners who built the Gwalia camp were probably behaving less like Italians than like bachelors. The example of the Chinese also tells us that the role of ethnicity in determining settlement forms is a function of the proportion of the ethnic group present in the community. Identifying the potential determinants of a mining settlement’s form and fabric (e.g. economics, geography, ethnicity) is relatively simple. The difficult part is explaining what circumstances cause a particular influence to become the determining one. This chapter has examined some case studies and pointed the way to more systematic research.
REFERENCES Auhl, I. (1986) The Story of the ‘Monster Mine’: The Burra Burra Mine and Its Townships 1845– 1877. Hawthorndene: Investigator Press. Bell, P. (1984) Timber and Iron: Houses in North Queensland Mining Towns 1861–1920. St Lucia, Brisbane: University of Queensland Press. ——(1990) Continuity in Australian timber domestic building: an early cottage at Burra. Australian Journal of Historical Archaeology 8: 3–12. Bell, P., Connell, J. and McCarthy, J. (1985) Gwalia conservation study. Unpublished report to the Western Australian Tourism Corporation. Douglass, W. (1979) The mining camp: a unique social phenomenon. Unpublished paper presented at the Southeastern Anthropological Association meeting, Santa Barbara, California . Jones, T. (1990) The Chinese in the Northern Territory. Darwin: Northern Territory University Press. Payton, P. (1984) The Cornish Miner in Australia: Cousin Jack Down Under. Trewolta: Dyllansow Truran. Roderick, D. (1980) Development of the Charters Towers Urban Area. Unpublished paper, Townsville, Queensland. Serle G. (1963) The Golden Age: A History of the Colony of Victoria 1851–1861. Parkville: Melbourne University Press.
Chapter 3
Gender and community structure on Australian colonial goldfields Susan Lawrence
ABSTRACT The archaeology of mining is an explicitly gendered area of research in Australasian historical archaeology, notable for emphasizing machines and technologies at the expense of settlements and social structures. Dwellings associated with mining sites are assumed to be the huts of single male miners while settlements are described impressionistically. Such work both substantiates and perpetuates the popular image of mining as an exotic male domain and ignores the need for detailed and substantive examination of mining communities. A feminist perspective facilitates such an examination by challenging the dominant portrayal of the masculine experience of mining, and highlighting the role of women and families in shaping goldfields culture. The case study of Dolly’s Creek, Victoria, demonstrates insights gained from a multilayered analysis of material culture and textual sources and the multilayered relationships between settlement and community.
INTRODUCTION Mining has played a prominent role in Australian culture and history, and in the modern heritage industry. Of the various kinds of mining in Australia, the gold industry had a widespread impact on the country as a whole because of its geographic, economic and demographic extent. Large quantities of gold were discovered in New South Wales and Victoria in 1851, setting off a series of rushes that by the end of the century had reached every colony and region. The gold rush has received considerable attention from Australian historians (Bate 1978, 1988; Blainey 1963; Goodman 1994; Serle 1963, 1971) and there is general agreement that the 1850s—the main decade of the gold rush—was a watershed in the country’s development. Mining has been no less important to archaeologists and heritage managers because of the scale of its impact on the environment and the quantity of physical fabric remaining. Archaeologists have been extensively involved in recording mining sites throughout Australia and New Zealand for the past twenty-five years. With few
40
Susan Lawrence
exceptions these studies have been notable for privileging what Conkey and Williams (1991) have called technoenvironmental phenomena. There has been a tendency to emphasize machines and technology, cataloguing stamper batteries, boilers, steam engines and water races (e.g. Bannear 1993; Bell 1987; Birmingham et al. 1979; Comber 1992; Coroneos 1993; Davey 1986; Gaughwin 1992; Milner 1986; Ritchie 1981; Townrow 1986). Settlements and isolated dwellings associated with mining sites have received less attention. The few exceptions to this pattern have been studies of the Chinese on the goldfields Jack et al. 1984; McCarthy 1988; Ritchie 1986), McGowan’s (1992, 1994) study of settlements in New South Wales, and Holmes’s (1983, 1989) study of settlements in the Northern Territory. Because of the lack of attention to the social aspects of mining in Australia, much remains to be learned about the ways in which goldfields communities were structured: their physical layout, their archaeological signatures, their networks of social, economic and political interaction. This task is made more difficult by the ephemeral nature of most mining settlements, which has resulted in a scattered and poorly preserved documentary and archaeological record. It is argued here that the study of mining settlements can usefully be addressed by employing perspectives drawn from feminist archaeology and from community studies. Feminist archaeologists advocate ‘tak[ing] up critical perspectives on the scientistic reconstructions of the past’ (Conkey 1993:11) and investigating the dynamic nature of gender as a process that explicitly and implicitly shapes human life (Conkey and Gero 1991:9–10; Purser 1991:13; Wylie 1991: 34). A focus on women has been taken as the starting point for an expanded archaeological discourse that is people-centred and arises from a finegrained understanding of the details of daily life (Spector 1993; Tringham 1991; Wall 1994; Yentsch 1994). This kind of feminist analysis diverts attention from the mechanical aspects of mining towards a humanized emphasis on the people who lived and worked at the mines and on the social structures which sustained them. Further, by problematizing the role of gender in structuring community life, such an approach challenges the popular image of the gold rush as a male domain which has been perpetuated and substantiated by traditional historical and archaeological research. Terms like prospector and miner are so inherently gendered that it is usually unnecessary to add that the individuals concerned were men, but women were there as well, working in both domestic and industrial contexts. Community studies provide a methodological basis for understanding historical contexts in specific geographic locales. Community studies involve the description of the geography of a settlement, its subsistence base, its material culture, the demography of its population, the social structures that it supported, and the relation of that place to the world around it, what Demos (1982:221) has called the ‘overall shape of life’. As Cusick (1995:59) argues, this approach facilitates the ‘study of a town or other small settlement at the household level, comparing numerous household sites, with information on the occupants
Gender and community structure on Australian goldfields
41
compiled both from documents and from excavation’. In essence, community studies provide a geographically defined basis for comparative ethnographic research. This is readily applicable to mining communities which are both geographically circumscribed and amenable to comparative, household-level analysis. Indeed, in a recent review of community studies in archaeology, Cusick (1995:65) cites Hardesty’s 1988 study of Nevada mining communities as an ‘elegant demonstration of both the strengths and limitations of a community studies approach’. Further, the emphasis on the entire fabric of community, rather than just the subsistence base, shifts the focus of mining research away from a concentration on technological issues. Industry remains important, but it is reintegrated into its social context.
SETTLEMENT AND COMMUNITY The terms settlement and community are frequently used interchangeably but it is useful to distinguish between community, a mental construct of shared interaction and values that differentiates a group from the outside world, and the more archaeologically visible settlement, or physical features of a place. Settlement refers to the distribution of buildings and roads in the landscape and the pattern of how those elements relate to one another. Anthropologists consider community to be the basic unit for the organization and transmission of culture within a society (Arensberg and Kimball 1965). At its simplest it can be defined as ‘a group of people who live in the same place and interact daily with one another’ (Murdock 1949 in Hardesty 1988:13). More complex definitions of community include several elements. Community begins with the physical element of settlement and the residence of a group of people in the same location (Murdock 1949:29). The size of territory the settlement may occupy varies, but the size of the group is approximately that which can regularly engage in face-to-face interaction. The frequency and nature of meetings and exchanges between individuals is one measure of the strength of the community: the more frequent the interaction, the stronger the community (Geismar 1982:5). In small non-industrial communities, face-to-face meetings with the same individuals occur frequently, although the social roles of the individuals involved may shift as each person in a small population assumes multiple roles and responsibilities. Each person knows and is known in a variety of situations and contexts (Demos 1982:311). Such intensive and varied interaction marks a dense community structure with frequent opportunity for conflict as well as cooperation. Shared communal activities create the social centres of communities while their location indicates the physical centre of the community. Participation in those communal activities is one of the signs of community membership (Rutman 1973:71). The nature of the central activities that give community life meaning provides clues to the identity and shared goals of the group, while it is
42
Susan Lawrence
the possession of shared goals that creates the community. Recognition of common interests brings communities into existence, and when those interests cease to be of central importance the community fragments (Geismar 1982:5). Relative homogeneity of activities and state of mind within a community are part of the shared identity of the group and help to make it distinct from neighbouring groups (Redfield 1955:4). Knowledge of that distinctiveness is founded on knowledge of the world beyond the community, and the distinctiveness is maintained by an ongoing process of contrasting the community with that outside it (Tuan 1974:210; Glassie 1982:591). Settlement is the material manifestation of the mental construct of community (Hardesty 1988:13), but the two cannot be separated. The archaeological remains are primarily evidence of the physical structure of settlement, yet the social web of events and structures cannot be separated from the physical frame that supports and sustains it, a point made by Demos in his study A Little Commonwealth (1970). There are several points at which social and physical aspects come together most strongly. The first point is in the nature of the settlement pattern. The spatial organization of a settlement structures the kind of interaction possible between individuals. Regular contact has been identified as key to a community’s cohesion, but informal daily contact is inhibited in situations where settlement is dispersed over a wide area. To combat spatial separation, customs may arise that incorporate regular formalized occasions at which contact can take place and the community can be maintained. In contrast, where close, nucleated settlement patterns prevail, regular informal interaction between all members of a community is facilitated and a more inclusive ethos may result. Community and settlement also come together in the infrastructure supporting the population. Phases in the development of community consciousness can be identified in changes to the physical fabric of the settlement. This is particularly clear on goldfields, where the phases of mining development and decline have parallels in the establishment of other community facilities (McGowan 1992:45). During the rush phase of mining, individual and small-scale enterprise is most apparent, and the impetus is to provide for immediate needs with businesses such as hotels and shops. During the latter phases of mining, community awareness of corporate needs is manifest in the building of public institutions and facilities such as schools, churches, meeting-halls and cemeteries. The special nature of gold-rush communities cannot be separated from the distinctive set of physical and economic circumstances that created them. The location of the settlements and the form the settlements took were based on the location of the gold deposits and the geology of those deposits. Shallow alluvial deposits tended to support dispersed settlements while deep lead and quartz gold supported nucleated settlements (Hardesty 1988:102). Social relationships within the community were also related to the nature of the gold deposit and the means of working it. Shallow deposits and deposits that could be worked with a minimum of capital investment could be worked by independent miners and so
Gender and community structure on Australian goldfields
43
fostered a complex mix of individualism, cooperation and egalitarianism in community relationships (Philipp 1987; Hardesty 1988:102–104). Hogan (1990) describes these as carnival-type camps, characterized by dispersed economic and political power and a poorly developed class structure, and associated with a nonindustrial mining pattern of limited technology and capitalization. In the Australian vernacular they were called ‘poor man’s diggings’, where subsistence miners could secure a livelihood on their own terms, independent of industrial wage structures. The historian June Philipp (1987) has eloquently described the ethos that developed on such fields, where miners struggled consciously to maintain their ideal of rough egalitarianism and self-sufficiency. Deeper and more complex ore bodies required capital and technology to work and so became the domain of large companies. Shareholding and wage labour created a different kind of community altogether, with stratified social groups and more structured forms of interaction (Hardesty 1988:104). Hogan (1990) has described these as caucus-type camps, characterized by corporate interests which manipulated political and economic structures, and associated with a corporate industrial pattern of capital- and technology-intensive mining. As much as gold-rush communities were tied to settlement pattern, they were also divorced from it. The shared goals, ideals and group identity of miners extended community boundaries far beyond the single settlement. The mining community has been characterized as one without a locus, consisting of a large group of individuals, connected by ties of kin, friendship, business and previous experience, that crystallized periodically around the location of gold discoveries (Douglass 1979:13). It was the shared goals and shared experience, not the requirement of co-residence, that made the larger mining community, and in that respect the community existed wherever and whenever gold was discovered. The mining community was international also in the sense that many of the men living on the goldfields had left families in other parts of the world. Many intended to return and many continued to send earnings back to families elsewhere.
DOLLY’S CREEK Feminist archaeology and community studies provide the intellectual framework for a case study of community structures in one settlement on the goldfields, that of Dolly’s Creek, Australia (Figure 3.1). The settlement is named after a small seasonal watercourse west of Melbourne, Victoria. Gold was discovered there early in 1857, and the population grew to a peak of more than 600 people in 1861, when the settlement included several stores, hotels, a bakery, a post office and a school. The shallow alluvial workings were never particularly rich, but they sustained approximately 100 miners throughout the 1860s and 1870s. By the 1880s mining had ceased at Dolly’s and in 1888 the area was declared a Forest Reserve (Geelong Advertiser, Mining Surveyors Reports). The sluices, races and mine shafts of the alluvial workings still remain, as well as
Figure 3.1 The location of Dolly’s Creek
Gender and community structure on Australian goldfields
45
at least sixty fireplaces where huts once stood (Figure 3.2). A survey of the settlement was conducted in 1990 and sample excavations at four huts were carried out in 1991 and 1992 (Lawrence 1995). Status, class and gender all play significant roles in shaping community interaction, and of the three, gender was the most prominent at Dolly’s Creek. The principal ethnic divisions on the Australian goldfields were between the Europeans, primarily British, and the Chinese. Although there is considerable documentary evidence indicating the presence of a Chinese community at Dolly’s Creek, most of the population was of British extraction. Ethnic backgrounds are primarily identifiable from census records, which indicate that 9% of the population was from China and 89% was either from the United Kingdom or born in the Australian colonies to British parents. Archaeological
Figure 3.2 The study area at Dolly’s Creek
46
Susan Lawrence
evidence of Chinese neighbourhoods has not been located, and the sites excavated were associated with the non-Chinese portion of the community. The analysis will thus be confined to the non-Chinese. Marked status differences also appear to have been absent at Dolly’s Creek, which was described at the time as a poor man’s diggings. As is consistent with a carnival-type organization, there was minimal economic basis for status distinctions and there was little evidence in either the archaeological or documentary sources to suggest that status played a large role in the settlement. Demography In order to understand the role of gender on the diggings it is first necessary to determine the demographic composition of the settlement. The Victorian colonial census of 1861 provides the most detailed picture of the people living at Dolly’s Creek, and to supplement that information newspaper accounts, Mines Department records and birth, death and marriage records were utilized. With the information provided by these sources, files were compiled on individuals within the community, in order to chart kin relationships and to describe with some certainty particular aspects of life at Dolly’s Creek. In April 1861 the census-taker recorded that 619 people lived at Dolly’s Creek, 57 of them Chinese men. Over half the people recorded were either women or children below the age of 15 (Figure 3.3). Most of the adults were young, 82% of the men and 85% of the women being less than 40. Ninety-four of the 110 adult women stated that they were or had been married. Widows comprised only 2% of the married women in the whole of Grant County, of which Dolly’s Creek was a part, and it can be assumed that the number of widows at Dolly’s itself was similarly small. As women were less likely than men to be on the diggings without their spouses, it seems probable that the number of married women was close to the number of married men living with their wives on the goldfield. Thus most of the women and a good many of the men were living as part of a married couple, and most of the couples probably
Figure 3.3 Ages of the population at Dolly’s Creek in 1861, based on figures from the census of Victoria
Gender and community structure on Australian goldfields
47
had children, because most of the women were of child-bearing age. There were an average of 2.1 children for every married woman. While studies of comparable ‘poor man’s diggings’ have not been done, evidence from larger towns suggests that these demographic patterns are not unusual for the Victorian fields as a whole and that, in general, women and children were present in significant numbers (Fahey 1983; Grimshaw and Fahey 1982; Grimshaw and Willet 1980; Philipp 1987; Bate 1978). The number of women and children present at Dolly’s Creek had many implications, one of which was household structure. The census-taker in 1861 recorded 215 houses in the settlement. With a total population of 619, this suggests an average of 2.87 people per household. When the numbers of married women, men and children living in nuclear families are considered, however, this changes to approximately 4.1 people in each nuclear family group and 1.9 people in each of the remaining households. The latter were composed primarily of adult males, and many would have been groups of mates sharing accommodation. However, the family households may have been even larger, incorporating unmarried siblings, adult children and domestic help, and in consequence the number of all-male, single-person households smaller. While these figures are speculative, they do provide some indication of the kinds of households that might have been found in the settlement. Other documents provide evidence of the role of kin relationships within the community. Many historians and anthropologists have shown that in frontier situations links forged between individuals as the result of marriage are instrumental in both cementing and initiating politico-economic relationships (Derry 1992; Purser 1991; Russell 1994). In areas of recent settlement instant extended families of multiple brothers, sisters and cousins-in-law provided social and business contacts and support in times of distress. In such circumstances neither men nor women necessarily experienced goldfields as places of hostile or indifferent strangers. Both sexes frequently travelled in family groups and continued to experience the companionship and support that had been familiar to them before migrating. Derry (1992) has been able to link kin networks with settlement patterns and material culture in frontier Alabama, but the paucity of documentary records precludes that kind of analysis here. However, it is still possible to recapture something of the dense network of kin ties that underlay the otherwise disconnected community structure of Dolly’s Creek, and women’s lives in particular were shaped by a network of kin structures that they created and maintained. The McCormack family provides one example of the rapid creation of a large extended family. Mary McCormack’s husband Thomas Haines had come to Morrisons as a miner, although he later took up a selection and became a dairy farmer (Morrisons Cemetery Records). Mary’s father and her three brothers were also miners, one of whom, William, married Sarah Lavery, whose father John was a farmer at Morrisons (Morrisons Cemetery Records). The marriages of these two women made the Lavery, McCormack and Haines households into a clan, providing workmates and companionship for both men and women.
48
Susan Lawrence
Work Now that we have established that significant numbers of women were at Dolly’s Creek, it is possible to investigate ways in which male and female experiences of mining differed. One difference was in the area of work. The 1861 census recorded that most adults in the community fell into two occupational categories, miners and wives or widows. Of the men, 88% were miners while 92% of the women were wives or widows (Table 3.1). Of the remainder, men’s jobs included rural labouring, farming and squatting, the trades, and shopkeeping. Women’s jobs included one inn-keeper, one pastoral labourer and four domestic servants. This breakdown gives a deceptively simple impression of employment, however, and ignores the fact that both ‘miner’ and ‘wife’ were broad categories that concealed a number of activities. While men identified themselves primarily as miners, for many mining was only one part of a broader subsistence strategy. Periods on the diggings were frequently interspersed with timber cutting, labouring or a range of other odd jobs. Men moved between jobs as they moved between diggings, mining on one field, keeping store on the next, or turning to a previous trade. Farming provided another supplement to mining income, particularly after the initial rush period Table 3.1 Occupations of the residents of Dolly’s Creek in 1861
Source: 1861 census
Gender and community structure on Australian goldfields
49
when land was gradually opened for selection. McGowan (1992:46) reports that on subsistence goldfields in New South Wales the rhythms of the agricultural cycle began to influence the course of mining, and excavated washdirt was left piled on claims during the harvest period as miners worked their selections. Miners elsewhere were able to take up selections close to the diggings and could schedule activities at both places during each day, as for example did one miner in Victoria who spent eight hours a day mining and three hours a day erecting farm buildings and fences (Evans 1975:45). The pooled efforts of the entire family were required to make this combination of mining and farming successful. Sons assisted fathers on the mines but the day-to-day running of the farm was commonly the responsibility of mothers and daughters. The Australian evidence indicates the persistence of what MacMillan (1995) has called ‘informal mining’, or the pattern of combining mining with other forms of subsistence labour. It is particularly common in non-industrialized economies where mining is not sufficiently intensive to require full-time specialization. This pattern is described elsewhere in this volume, among traditional African ironworkers (Herbert, Childs), the Inuit (Ehrenreich), copper-miners in Bronze Age Cyprus (Kassianidou) and in Bronze Age Thailand (Pigott), and miners in Dynastic Egypt (Shaw). In these societies mining and metalworking are considered craft specialities but practitioners engage in mining activity on a sporadic and limited basis. Mining is restricted to short periods of several weeks or months which are separated by much longer periods, sometimes as long as several years, in which miners are employed in other forms of labour, primarily farming. When mining becomes industrial in scale, with accompanying complexities in technology and the organization of capital and labour, informal patterns are replaced by full-time specialization. While most western European mining had been industrial for centuries, the gold rush facilitated a brief revival of informal mining which persisted for decades alongside industrial models, and which extended into other branches of mining, such as tin mining (e.g. Jackman 1996), where small-scale deposits could be similarly worked. Archaeological evidence provides further clues about the interaction between subsistence mining and subsistence farming. The goldfields diet is popularly characterized as comprising little but mutton chops and damper (bread). However, faunal remains at Dolly’s Creek included eggshell and the bones of rabbits, chickens, pigs, cows and goats, as well as sheep. Keeping domestic animals and growing vegetables on the diggings was possible because one of the terms of the Miner’s Right, or licence to mine, was a 20 perche (500 square metre) allotment on which to build a hut. Women, and also children, provided the labour that made small-scale production on such plots feasible. Selling surplus eggs and butter was one way in which women supplemented the family income. The cooperation required to maintain such an operation successfully is just one example of the importance of the family or household group as the unit of production for subsistence miners, few of whom were able to survive on the efforts of one individual alone.
50
Susan Lawrence
The role of women in productive agricultural labour indicates the inadequacy of ‘wife’ as an occupational category for women. ‘Wife’ is not strictly speaking an occupation at all, but a social category, and many women who happened to be living with their husbands were thus unable to declare the full range of their activities. Other hints about what some of those activities might have been come from contemporary letters and diaries. In addition to the ubiquitous tasks of bearing and raising children, cleaning house, shopping, cooking, sewing, doing the laundry and, as has been argued, running subsistence farms, contemporary commentators recorded that women ran shops and sold alcohol, bought gold and worked family claims. Economic labour was only one component of women’s work, however, and an equally important component was that of social labour, both at the community level and in the home. Feminist historians (Lake 1988; McGaw 1989) have pointed to the role of women in initiating and maintaining community institutions such as churches and schools, and Hardesty (1994) in the USA and McGowan (1994) in New South Wales have cited the presence of such structures archaeologically as indications of the exertions of women in mining settlements. On the diggings around Dolly’s Creek, three churches, three schools and a continuous round of concerts, picnics and dances reported in the local newschapter are evidence of the work of women in that community. Another important social task which fell to goldfields women as much as it did to their sisters in cities was the maintenance of the family’s social identity. During the Victorian era the domestic environment was central to nineteenthcentury culture because it was there that the Victorian esteem for family life, or domesticity (Cott 1977), found material expression. Objects and spaces were used to symbolize and facilitate the unity and sanctity of the family (Ames 1976; Russell 1994; Wall 1994). Historians such as Russell (1994) have emphasized the role of women in facilitating that culture, and doing so constituted a major portion of women’s work. Women had the necessary skills and training to acquire and present the consumer goods that were a fundamental part of family identity and social position. Artefacts from Dolly’s Creek suggest that in some households at least, a version of the domestic ideal was being enacted by the subsistence miners. Wallpaper, pressed glass dishes and a decorative clock, all of which might be considered appropriate furnishings for a parlour, were found at two households where there is also firm evidence for the presence of women. This association further substantiates the identification of maintaining a proper household environment as women’s work. Gender and material culture Excavations at four hut sites at Dolly’s Creek provide more detailed evidence of the role of gender within the community and for the social work of women within the home. Two types of household were identified on the basis of the archaeological evidence. One type consisted of sites with a purely or primarily residential function, while the other type consisted of sites with a combined
Gender and community structure on Australian goldfields
51
residential and commercial function. The two residential huts (C75, B47) were simple timber-framed, canvas-covered structures with a stone fireplace on one gable end (Figure 3.4). The artefact assemblages included high proportions of tablewares, faunal remains, architectural items and small quantities of alcoholrelated items, i.e. bottle glass (Table 3.2). In contrast, the two mixed commercial/residential assemblages (A27, C45) contained smaller proportions of tablewares, faunal remains and architectural items, and large proportions of bottle glass. One of the latter structures was a canvas building with stone fireplace, similar to the residential huts, but the other (C45) was far more substantial. It had walls built partially of stone and partially of canvas covered with tin on the exterior and wallpaper on the interior (Figure 3.5). On the basis of the large quantities of alcohol bottle glass recovered, it is argued that these two buildings were either pubs, or sly-grog shops where alcohol was sold illegally. Further evidence suggests that the stone hut may also have been a store. The strong construction of the hut and the substantial padlock found there would have provided protection from theft, which was a constant worry for storekeepers in particular (Korzelinski 1979:137). The artefact assemblage also included multiple items of some kinds of goods, such as shovels, and evidence for bulk storage of liquids in casks, both of which are also consistent with this household running a store.
Figure 3.4 Site plan of House C75 at Dolly’s Creek
52
Susan Lawrence
Table 3.2 Distribution of artefacts from the houses excavated at Dolly’s Creek
Of particular interest here is the association between women’s goods and goods associated with Victorian domesticity. Census information from Dolly’s Creek suggests that approximately 50% of the households there included women, and two of the four assemblages included items of women’s clothing and jewellery. Personal artefacts such as buttons, spectacles and shoes accounted for between 2.5 and 5% of the total assemblage at three of the sites, while none was present at A27. However, items of clothing and jewellery worn by women were found at only two of the sites, one of the residential huts and the store. These items included a woman’s shoe, decorative buttons and clasps, a brooch, and a woman’s gold wedding-band. Men’s shoes and buttons and buckles from men’s work clothing were also included in these assemblages, which have for convenience been labelled family households. The two assemblages from the family households included a diverse range of decorative household items, while nothing similar was recovered from the two assemblages lacking women’s items. All the assemblages are meagre and the goods represented limited in number and poor in quality. However, particularly at the two family
Gender and community structure on Australian goldfields
53
Figure 3.5 Site plan of House C45 at Dolly’s Creek
households they reveal a diversity of form and function extending well beyond the basic goods necessary for subsistence in a short-term diggings camp. Fragments of several different pressed glass dishes were found, including the stem of a glass cake-plate. A clock case found at one house was equally decorative. Standing 18 cm high and cast in brass, it has an ornate moulded floral design and was probably gilded or ‘electro-copper bronzed’ for a French ormolu affect. The space left for the clock itself was for a standard 2 inch faced clock which could have been purchased separately from the frame. Similar clocks were mass-produced and readily available in the late nineteenth century (Cuffley 1984:139). Ceramics were present at all
54
Susan Lawrence
houses, representing 76 separate vessels in 43 different decorative patterns. Forms were primarily dinner plates, cups and saucers, indicating that crockery, not tin, was the normal form of tableware used throughout the diggings; thus decorative ceramics alone cannot be seen as indicative of differentiated domestic culture. However, at one of the family households more than half of the ceramic assemblage was teawares, and the prominence of this group may reflect the importance of the social ritual of taking tea. In addition to home furnishings, special care was also taken with decoration at these two houses. The fireplace in the residential hut was given a special coating of white pipeclay brought from the mine shaft itself. This would have cleaned and brightened the fireplace and hence the interior of the tent, and according to oral sources was a relatively common practice among goldfields women (Elsie Bayard, personal communication; Coroneos 1993:166). At the residence-cum-store a fragment of wallpaper was found, layered with newsprint and hessian and backed with a sheet of tin. The practice of covering tent walls in this way is also well documented. Mobility among subsistence miners dictated that houses be ephemeral and possessions few, but acceptable standards of decency were continuously negotiated in decisions made about the nature of the possessions. Residents of Dolly’s Creek chose to set their tables with decorative ceramics rather than with tin, and women wallpapered the interiors of their tents and whitewashed the rough fireplaces. On the mantelpieces they displayed pressed glass bowls and French-style clocks. Objects and actions like these were part of the work of maintaining genteel standards. The decision to use crockery rather than tin, to bear the expense of transporting fragile and heavy dishes, indicated awareness of and conformity to notions of what a respectable home was. The decision to divert time and labour from mining itself into building better fireplaces and papering the walls of the tents suggests a desire to improve the overall quality of life. Actions like these created refined homes in even the roughest conditions.
CONCLUSION A feminist approach to the study of community structures brings women of the goldfields into the foreground, but at the same time it extends far beyond simply finding women. Demographic structures, the nature of work, and activities within the home have been examined and several patterns have emerged, resulting in new insights into the operation of mining communities on a day-to-day basis. Women and children made up half the population of at least one diggings and there is no reason to expect that that was not the case elsewhere. Although identified only as miners and wives, both men and women worked at a range of activities and cooperatively pooled the results of their labours within productive household units. Such
Gender and community structure on Australian goldfields
55
cooperation and diversity were central components of a successful subsistence mining strategy typical of informal mining. Questions about women’s work revealed much about men’s as well, and about the place of mining in a wider subsistence strategy. Family farms, maintained by women, contributed meat and vegetables that broadened the diet and cash that supplemented and in many cases sustained mining activity. The material domain of daily life was also enlarged by women, who were bearers of genteel culture and who replaced tin dishes with decorative crockery and placed fancy clocks above whitewashed fireplaces. The archaeological evidence demonstrates that gender affected in tangible, material and archaeologically visible ways the experience of mining life for both men and women.
REFERENCES Ames, K. (1976) Meaning in artefacts: hall furnishings in Victorian America. In D.Upton and J.Vlach (eds), Common Places: Readings in American Vernacular Architecture, pp. 240– 260. Athens, GA: University of Georgia Press. Arensberg, C. and Kimball, S. (1965) Culture and Community. New York: Harcourt, Brace & World. Bannear, D. (1993) Historic mining sites in the Maldon mining division. Unpublished report for the Department of Conservation and Natural Resources, Bendigo. Bate, W. (1978) Lucky City: The First Generation at Ballarat. Melbourne: Melbourne University Press. ——(1988) Victorian Gold Rushes. Ringwood, Victoria: McPhee Gribble/Penguin. Bell, P. (1987) Gold, Iron and Steam: The Industrial Archaeology of the Palmer Goldfield. Townsville, Queensland: History Department, James Cook University. Birmingham, J., Jack, I. and Jeans, D. (1979) Australian Pioneer Technology: Sites and Relics. Richmond, Victoria: Heinemann Educational Australia. Blainey, G. (1963) The Rush That Never Ended. Melbourne: Melbourne University Press. Comber, G. (1992) Palmer Goldfield Heritage Sites Study (Stage 2). Unpublished report to the Queensland Department of Environment and Heritage, Cairns. Conkey, M. (1993) Making the connections: feminist theory and archaeologies of gender. In H.du Cros and L.Smith (eds), Women in Archaeology: A Feminist Critique. Occasional Papers in Prehistory, no. 23: 3–15. Canberra: Australian National University, Department of Prehistory, Research School of Pacific Studies. Conkey, M. and Gero, J. (1991) Tensions, pluralities and engendering archaeology: an introduction to women in prehistory. In J.Gero and M.Conkey (eds), Engendering Archaeology: Women and Prehistory, pp. 3–30. Oxford: Blackwell. Conkey, M. with Williams, S. (1991) Original narratives: the political economy of gender in archaeology. In M.di Leonardo (ed.), Gender at the Crossroads of Knowledge: Feminist Anthropology in the Postmodern Era, pp. 102–139. Berkeley: University of California Press. Coroneos, C. (1993) A poor man’s diggings: an archaeological survey of the Lisle Denison goldfields. Unpublished report prepared for the Forestry Commission, Hobart, and the Queen Victoria Museum and Art Gallery, Launceston, Tasmania. Cott, N. (1977) The Bonds of Womanhood: ‘Woman’s Sphere’ in New England 1780–1835. New Haven, CT: Yale University Press.
56
Susan Lawrence
Cuffley, P. (1984) Chandeliers and Billy Tea: A Catalogue of Australian Life. Fitzroy, Victoria: Five Mile Press. Cusick, J.G. (1995) The importance of the community study approach in historical archaeology, with an example from Late Colonial St. Augustine. Historical Archaeology 29(4): 59–83. Davey, C. (1986) The history and archaeology of the North British Mine Site, Maldon, Victoria. Australian Journal of Historical Archaeology 4: 51–57. Demos, J. (1970) A Little Commonwealth: Family Life in Plymouth Colony. New York: Oxford University Press. ——(1982) Entertaining Satan: Witchcraft and the Culture of Early New England. New York: Oxford University Press. Deny, L. (1992) Fathers and daughters: land ownership, kinship structure, and social space in old Cahawba. In A.Yentsch and M.Beaudry (eds), The Art and Mystery of Historical Archaeology: Essays in Honor of James Deetz, pp. 215–228. Ann Arbor: CRC Press. Douglass, W. (1979) The mining camp: a unique social phenomenon. Unpublished paper presented at the Southeastern Anthropological Association meeting, Santa Barbara, California. Evans, W. (1975) Diary of a Welsh Swagman 1869–1894. Melbourne: Macmillan. Fahey, C. (1983) Bendigo 1881–1901 : a demographic portrait of a Victorian provincial town. Australia 1988. Bulletin 11: 88–107. Gaughwin, D. (1992) Trade, capital and the development of the extractive industries of north east Tasmania. Australasian Historical Archaeology 10: 55–64. Geismar, J. (1982) The Archaeology of Social Disintegration in Skunk Hollow: A Nineteenth Century Black Community. New York: Academic Press. Glassie, H. (1982) Passing the Time in Ballymenone: Culture and History of an Ulster Community. Philadelphia: University of Pennsylvania Press. Goodman, D. (1994) Gold Seeking: Victoria and California in the 1850s. Sydney: Allen & Unwin. Grimshaw, P. and Fahey, C. (1982) Family and community in nineteenth century Castlemaine. Australia 1988. Bulletin 9: 88–125. Grimshaw, P. and Willet, G. (1980) Family structure in colonial Australia: an exploration in family history. Australia 1988. Bulletin 4: 5–27. Hardesty, D. (1988) The Archaeology of Mining and Miners: A View from the Silver State. Society for Historical Archaeology, Special Publication Series no. 6. Pleasant Hill, CA: SHA. ——(1994) Class, gender and material culture in the mining west. In E.Scott (ed.), Those of Little Note: Gender, Race and Class in Historical Archaeology, pp. 129–148. Tucson and London: University of Arizona Press. Hogan, R. (1990) Class and Community in Frontier Colorado. Lawrence: University Press of Kansas. Holmes, K. (1983) Excavations at Arltunga, Northern Territory. Australian Journal of Historical Archaeology 1: 78–87. ——(1989) Arltunga: a minor goldfield in arid central Australia. Australian Journal of Historical Archaeology 7: 43–49. Jack, I., Holmes, K. and Kerr, R. (1984) Ah Toy’s garden: a Chinese market garden on the Palmer River goldfield, north Queensland. Australian Journal of Historical Archaeology 2: 51–58. Jackman, G. (1996) An archaeological survey of the Blue Tier tin-field. Unpublished report prepared for the Forestry Commission, Hobart, Tasmania. Korzelinski, S. (1979) Memoirs of Gold Digging in Australia. Translated and edited by S. Robe. St Lucia, Brisbane: University of Queensland Press.
Gender and community structure on Australian goldfields
57
Lake, M. (1988) Women, gender and history. Australian Feminist Studies 7/8: 1–9. Lawrence, S. (1995) No abiding city: the archaeology and history of an ephemeral mining settlement. Unpublished PhD thesis, School of Archaeology, La Trobe University, Melbourne. McCarthy, J. (1988) The new gold mountain: Chinese trade networks in northern Australia. In Archaeology and Colonisation: Australia in the World Context, pp. 139–148. Sydney: Australian Society for Historical Archaeology. McGaw, J. (1989) No passive victims, no separate spheres: a feminist perspective on technology’s history. In S.Cutcliffe and R.Post (eds), In Context: History and the History of Technology. Essays in Honour of Melvin Kranzberg. Research in Technology Studies 1: 171–191. Bethlehem, PA: Lehigh University Press. McGowan, B. (1992) Aspects of gold mining and mining communities in the Shoalhaven area of New South Wales: an archaeological and historical study. Australasian Historical Archaeology 10: 43–55. ——(1994) Lost Mines: Historic Mining Sites in the Monaro-Southern Tablelands Districts of New South Wales. Canberra: Barry McGowan. MacMillan, G. (1995) At the End of the Rainbow? Gold, Land and People in the Brazilian Amazon. New York: Columbia University Press. Milner, P. (1986) Engineering at the Mount Wills Proprietary Mine. Historic Environment 5(2): 15–19. Murdock, G. (1949) Social Structure. New York: Macmillan. Philipp, J. (1987) A Poor Man’s Diggings: Mining and Community at Bethanga, Victoria, 1875– 1912. Melbourne: Hyland House. Purser, M. (1991) ‘Several paradise ladies are visiting in town’: gender strategies in the early industrial west. Historical Archaeology 25(4): 6–16. Redfield, R. (1955) The Little Community: Viewpoints for the Study of a Human Whole. Chicago: University of Chicago Press. Ritchie, N. (1981) Archaeological interpretation of alluvial gold tailing sites, central Otago, New Zealand. New Zealand Journal of Archaeology 3: 51–69. ——(1986) Archaeology and history of the Chinese in southern New Zealand during the nineteenth century: a study of acculturation, adaptation and change. Unpublished PhD thesis, Department of Anthropology, University of Otago: Dunedin. Russell, P. (1994) A Wish of Distinction: Colonial Gentility and Femininity. Melbourne: Melbourne University Press. Rutman, D. (1973) The social web: a prospectus for the study of the early American community. In W.O’Neill (ed.), Insights and Parallels: Problems and Issues of American Social History, pp. 57–124. Minneapolis: Burgess Publishing Co. Serle, G. (1963) The Gold Age: A History of the Colony of Victoria, 1851–1861. Melbourne: Melbourne University Press. ——(1971) The Rush to be Rich: A History of the Colony of Victoria 1883–1889. Melbourne: Melbourne University Press. Spector, J. (1993) What This Awl Means: Feminist Archaeology at a Wahpeton Dakota Village. St Paul: Minnesota Historical Society Press. Townrow, K. (1986) An archaeological investigation of early mining in Tasmania: a preliminary assessment of resources. Unpublished report for the National Parks and Wildlife Service, Hobart, Tasmania. Tringham, R. (1991) Households with faces: the challenge of gender in prehistoric architectural remains. In J.Gero and M.Conkey (eds), Engendering Archaeology: Women and Prehistory, pp. 93–131. Oxford: Blackwell. Tuan, Y. (1974) Topophilia: A Study of Environmental Perception, Attitudes, and Values. Englewood Cliffs, NJ: Prentice Hall.
58
Susan Lawrence
Wall, D.D. (1994) The Archaeology of Gender: Separating the Spheres in America. New York: Plenum Press. Wylie, A. (1991) Gender theory and the archaeological record: why is there no archaeology of gender? In J.Gero and M.Conkey (eds), Engendering Archaeology: Women and Prehistory, pp. 31–56. Oxford: Blackwell. Yentsch, A. (1994) A Chesapeake Family and Their Slaves: A Study in Historical Archaeology. Cambridge: Cambridge University Press.
Chapter 4
Bedroom politics Ladies of the night and men of the day Alexy Simmons
ABSTRACT In the historic mining towns of the USA and New Zealand, the workplaces of female prostitutes are identifiable in the historical and archaeological record. The political economy that emerged in the frontier mining communities was conducive to conflict and compromise, innovation, unprecedented political growth, and competition. Prostitutes entered this volatile milieu of rapid economic change as entrepreneurs answering a demand for sexual services and female companionship. The marketing of services varied greatly and was influenced by the political and economic structure of the community over time. The spatial arrangement of prostitutes was closely linked to sociopolitical structures. The status of women was lowest in communities with tightly differentiated public and private spaces. Zoning was used to control the access of prostitutes to public spaces and material goods. Restrictions were overcome through covert sociopolitical relationships including bribes, charitable aid and bedroom politics. Spatial controls on prostitution often accelerated the growth of a class system among the demimonde.
INTRODUCTION In the frontier mining towns and camps of the USA and New Zealand, prostitutes or ‘red light ladies’ constituted an identifiable socio-economic group. Prostitutes are identifiable in the historical and archaeological record as part of the entertainment sector of the community. Implicit information about the economic, social and political aspects of a mining community can be extrapolated from the lifeway of prostitutes within a mining settlement. Historically, prostitution was dominated by women and is culturally based. The criteria used to distinguish the female prostitute from other women is the exchange of a fee for a specific sexual act. Prostitutes see the sex—money transaction as the business transaction that constitutes their job (Heyl 1979:20). The mining frontier was a melting-pot of classes, with a common bond and conflict built on the resource being exploited. The spatial location of prostitutes in a settlement related to prevailing social and community
60
Alexy Simmons
structures (i.e. economics, demography and politics). The material culture associated with prostitutes is composed of the artefacts of the profession and the possessions of women.
GENDER: AN ARCHAEOLOGICAL TOOL The study of identifiable groups, in this case female prostitutes, increases our understanding of the processes that underlie human behaviour. Social and historical archaeology can be used to extrapolate information about social processes associated with this engendered group within the context of the frontier community and settlement. The study of prostitution on the frontier provides an opportunity to explore gender roles in terms of cultural constraints on behaviour, power and gender relationships, spatial dynamics, gender and race, and the cultural meanings assigned to an engendered group. Historical records disclose the community view on prostitution through both descriptive or factual data, and suppressed information. Historical literature documents prohibitions, both legal and social, that affected the lifeway of female prostitutes, and attests to the usefulness of this sociological group for deriving information about gender relationships and community attitude. The euphemisms to describe prostitution and prostitutes—e.g. harlot, courtesan, housekeeping, bawdyhouse, whore, demimonde, lady of easy virtue, house of ill fame, gay damsels, etc. —in the census, newspaper and other records are indicative of the attitudes the community held about prostitution. Data deficits and censorship evident in some newspapers and other records also have implications for understanding the community.
THE CONTEXT: COMMUNITY AND LOCATION The extractive frontier of the USA and New Zealand included people involved in whaling, trapping, sealing, mining and logging. The settlements of these people all shared the commonality of being instant and focused on a single resource in their initial stages of occupation. The frontier community was composed of opportunistic migrants who all shared an expectation of wealth accrued through exploitation of a resource, usually in an unsettled location. Social or political persecution or restrictions also stimulated migration to the frontiers. Migration is a key mechanism for spreading technology, language, customs and most other social behaviours (Morrill 1970:141). One result of the isolation of mining communities was that innovation, technology modification and social change occurred on the frontier. The western frontiers of the USA were populated by first-generation nationals and immigrants from the eastern and mid-western states, Europe, Latin America, South America and China. People, industries and technology were transferred directly to New Zealand
Bedroom politics
61
from the frontiers of colonial Australia and, to a lesser extent, from those of North America (Mackay 1992:11). The mining community, particularly in the early years, differed from other frontier communities because of a combination of factors linked to individualism. Factors that combined to set the mining frontier apart from other extractive frontiers included: (a) a high percentage of individual pursuit of the resource; (b) high potential for monetary gain paired with great resource unpredictability and substantial physical risk; (c) a high percentage of privatization or individual mineral titles/rights; and (d) secrecy about resource sites and exploitation. The result was often isolation and conflict within the context of a remote and demanding environment. The commonality of these factors created a tacit social bond that underlined the formation of a collective identity and organizational structure that became the mining community. Frontier mining settlements grew up virtually overnight. The settlement was new and unfinished and the community equally raw and transient. The mining community operated within the vicissitudes of boom and bust cycles typical of mineral exploitation. The frontier has been typified as containing ‘a strident ambivalence between gross barbarism on the one hand and an equally endemic yearning after refinement and respectability on the other’ (Davis 1967:19). These conflicting social traits found expression in the settlement pattern and material culture of the mining community.
DOING BUSINESS: THE ENTERTAINMENT TRADE The clustering of individual miners at a central locality created a market for various services. The usual mining population ratio of high adult male numbers to low numbers of adult females resulted in a demand for female companionship and sexuality. Entrepreneurs, including prostitutes, established businesses in the new settlements and mined the reliable bonanza of the miners’ needs and wants. Settlement segregation of prostitutes was not apparent in the mining camps that typified the initial settlement phase. The facilities of merchants, transport companies, saloon-keepers and prostitutes were usually interspersed in a random pattern. Brothels were established side by side with other businesses. Prostitution was generally practised openly and was an accepted and expected part of the entertainment sphere of a frontier mining camp or tent town (see Figure 4.1). The relationship between space and status is an expression of the operative social system. The properties of a social system express themselves through daily activities; at the same time these actions reproduce the structural properties of a social system and reinforce that system (Spain 1992; Bourdieu 1977; Giddens 1979). The mining camps matured in a brief period to become less transient settlements, i.e. towns and cities. The temporal nature of these changes and the fragility of archaeological evidence often result in a skewed interpretation of the operative social system.
Figure 4.1 An early street scene in the Ahaura gold-mining settlement, Napoleon Hill, west coast, South Island, New Zealand, 1866. Note the ‘ladies’ of the Casino de Venise on the right Source: Alexander Turnbull Library, Wellington, NZ: F-148337–1/2
Bedroom politics
63
As the mining camps grew into a town, business clustering occurred. Morrill (1970:71–78) suggests that the spatial patterns of businesses were located in a hierarchy, within mutually beneficial groups. Entertainment facilities clustered, as did other businesses. Hierarchical subclusters are apparent within these economically beneficial groupings. For example, higher-status prostitution facilities, such as parlour-houses, tended to be located near theatres or hotels, or at the edge of the retail district. The establishments of lower-status prostitutes— cribs and small houses—were close to the edge of town, next to the Chinese district, or outside town. The aggregation pattern of prostitutes, businesses and institutions was affected by the type of town that evolved and the economic patterns that supported it. The boom phase or initial mineral find phase was associated with transient tent towns that sprang up overnight. At Tuapeka, New Zealand, in 1861, within weeks of gold being found, sixty-four storekeepers, butchers and bakers had established themselves (Belich 1996). If economic stability of the find was established, a regional town or business centre might develop to service a district of small, transient camps or corporate investments in large hydraulic operations, and hard-rock mining might result. The regional centre was a service centre but not a consistent residential location for the miner. The corporate town was dominated by a stable workforce of miners, supervisors/managers and commercial entrepreneurs. A false economy based on outside investments and mineral stocks was common. Settlement patterns of the community within the regional centre and corporate town changed in response to the economic tide, i.e. bonanza periods, bonanza consolidated, recession and depression.
PROSTITUTION AND GENDER RELATIONS Prostitution is intrinsically linked to a demand for sexual services. This demand is born out of ‘the institutional control of sex, the unequal scale of attractiveness, and the presence of economic and social inequality between classes and between males and females’ (Davis 1966:371). Women enter the profession for many reasons, but the three main reasons are poverty, inclination and seduction. To understand prostitution on the frontier in the mid-nineteenth century it is essential to understand the differences in cultural values in the context of class, race/culture and economic opportunities. Being female paired with prostitution as a profession provided a behavioural script for both the prostitutes and members of the community. Behavioural expectations were modified by the context within which the prostitute operated. This resulted in the temporary integration of prostitutes not apparent in the established urban centres, where the Victorian code of gentility among the middle class imposed controls on sexual companionship. Sexuality was repressed but alluded to; it was to be reserved as a bargaining resource for the acquisition of male wealth and status through the marital contract (Collins 1972). Sex in exchange for a marital
64
Alexy Simmons
contract was an accepted social construct. The marriage contract provided economic security for middle-class women during a time when owing to social restrictions there were few economic options. Women were poorly educated; gender-based job restrictions existed; women’s work was poorly paid; and there was general societal opposition to women working. The economic necessity of marriage therefore dictated the control of a bridal virginity as a bridal dowry. Among the working class, economic survival dictated the acceptance of a modified view of purity and bridal virginity. During the nineteenth century it was possible for working-class women to leave prostitution, marry, and return to respectable society; situational factors were considered before judgement of vice and virtue (Pearsall 1976; Hammerton 1979; Seifert 1991; Perkins 1993). For the upper class, wealth allowed sexual latitude among women. At the same time, men were ‘conceptualised as having a strong, irresistible, and uncontrolled sex urge which justifies their recourse to whores’ (Robinson 1984:8). The definition of male sexuality not only served to increase the importance of purity but also provided an excuse for male fecundity. The repression of sexual intercourse outside marriage created an obvious market demand. For the prostitute this resulted in the marketing of virgins, exotic experiences and sexual partners, and sexual ambience. Prostitutes as a socio-economic group were stratified. They gained their stratum through the customers they associated with and the type of service they provided. Client preference was often limited by economics which served to reinforce community-based socio-economic class lines. Client preference also appears to have included a desire for cultural familiarity, particularly among non-English-speakers. Collins (1972) links stratification and changes in stratification in a market economy to the resources held by any one individual. The theory is useful for generalizing economic power and gender relationships from around 1840 to the early twentieth century, when men largely controlled income and property. Collins identifies female resources (assets) as personal attractiveness (not just physical); domestic service; and emotional support. To maximize the gains a woman could obtain from these assets it was necessary to appear both as attractive and as inaccessible as possible. Thus status among sex workers was generally a product of the percentage of companionship or entertaining involved in the actual prostitution service being provided. The more ‘ladylike’ a prostitute was—that is, the more entertainment and companionship she provided—the higher her status, whether she was European, Chinese or mulatto. The commercial markers or material indicator of status in the sex industry included location within the mining settlement, architectural type of establishment, hierarchy of management within the establishment, range of services provided including associated facilities/chattels/consumable goods, social links with respectable or disreputable members of the community, and clothing of the sex workers (both on and off the job). On the frontier, higher-class prostitutes provided companionship and filled some of the roles ‘virtuous middle-class’ women usually played in the
Bedroom politics
65
community, for example the provision of food for the poor and other charity work. The ladies of the night also built relationships with political leaders of the day and men in general. These relationships protected the business operations of the demimonde during the metamorphosis of the mining settlement into a developed town/city. The number of middle-class women in the mining towns of the frontier was limited. Where they did come in numbers, such as at Thames, New Zealand, there were fewer prostitutes in the population. In other mining towns like Hokitika, also in New Zealand, or Virginia City, Nevada, prostitutes made up a very evident part of the town’s population and entertainment sphere. In Hokitika the resident magistrate court judge responded to a charge involving abusive language filed by one prostitute against another, and noted, ‘he could not inflict any fine in this case. It would be a very dangerous precedent if he did. A third of the women in the district would be laying informations’ (West Coast Times 5 October 1867:41). Demand for female companionship and sex paired with weak gender role differentiation between the public and domestic sphere in the mining camps resulted in social integration of prostitutes and their political power. As gender ratios became more balanced, differentiation between the female-dominated domestic sphere and the male-dominated public sphere developed. This gave rise to conflicts between social attitudes about gender subordination and economic power. Madams and prostitutes, with the exception of Chinese prostitutes in the USA, held economic control over the service they provided. Female economic power is an essential precedent to political power and authority (O’Kelly 1980). Within a community that had weak institutional systems the economic power that many prostitutes held gave them a basis for political participation and mechanisms to protect their economic interest. Notwithstanding the fact that a substantial amount of this power was covert, some changes in sociocultural attitude occurred. Acceptance might also swing to discrimination and rejection in poor economic times as competition for resources increased. Such discrimination is documented in newspaper reports, legislation, ordinances and informal residential zoning.
STRATIFICATION AND ETHNICITY Status is the amount of honour a society chooses to confer on the individual because of his/her personal characteristics, skills and abilities, or because of ascribed parameters such as age and sex (Pyszczyk 1988:23). Prostitution was stratified according to the services provided, for example entertainment versus a brief sexual service. Mobility within the profession was apparent, although race did crosscut this mobility and could impose status barriers that resulted in location predetermination in some mining communities. Lillian Powers started out as a farm girl turned prostitute in a Cripple Creek (Colorado) crib; from there she managed a brothel in Salida, Colorado, and later opened a brothel of
66
Alexy Simmons
her own in Florence, Colorado (Lee 1958, 1968). Minority groups of a distinct religion, colour or natural origin have long been set apart from the rest of the populace (Morrill 1970:172). Ghettos form owing to both external discrimination and internal preference; they provide a community of people of similar cultural background (e.g. language, religion and custom) and aid migrants in their transition to a new society. Racial integration occurred in smaller mining towns and camps, probably owing to the minimal number of individuals from any one ethnic group. Chinese or black residents were exceptions to this rule and were often segregated, even in the smaller settlements. The ethnic composition of the prostitute population generally reflected the ethnic and racial mix apparent in the frontier mining community. Racial segregation was evident among prostitutes with the exception of ‘exotic’ prostitutes, who transgressed settlement pattern. Prostitutes termed ‘exotic’ often included octoroons, quadroons and mulattos, or Japanese. Their racial characteristics afforded them market opportunities in contrast to prohibitions associated with the racial characteristics of women from other ethnic groups. Chinese ghettos or ‘Chinatowns’ were particularly notable and are identifiable archaeologically because of the suite of artefacts associated with the Chinese. In Virginia City, black women occupied the same lodging houses as white women, but worked in segregated establishments. Black prostitutes in Silver City, Idaho, in the early twentieth century lived on the outskirts of town, beyond Chinatown, occupying an establishment called the Pink House. In the larger US mining towns, Hispanic prostitutes were often undervalued, since they spoke little English, and many were identifiable by genotype. Ironically, French prostitutes were desired, and many Euro-American prostitutes took French names and accents, and claimed French blood, as a marketing tool. In a similar way, Hispanic prostitutes and entertainers capitalized on their ‘hot Latin blood’. New Zealand did not have any Hispanic, Japanese or black prostitutes and very few (if any) miners of these racial backgrounds. New Zealand emigration was dominated by Caucasian British subjects. The lack of manuscript census for the goldfields and census data that linked ethnicity and employment makes it difficult to identify the ethnic background of prostitution in New Zealand. Newspaper reports and secondary sources identified Irish women with prostitution in New Zealand mining settlements. Many Irish prostitutes came from Australia with the wave of Irish gold-miners who travelled from Victoria to the Otago and West Coast fields in the 1870s (Brooking and Rabel 1995). Others came over as part of the New Zealand government immigration schemes of the 1850s and 1860s (Macdonald 1990). Irish prostitutes also made up a proportion of the US prostitute population as well. Literature on the historic mining towns in New Zealand suggests that Irish prostitutes were generally considered lower stratum. This stratum difference did not exist in the USA, where some of the better-known madams and high-status prostitutes were Irish. Strong anti-Irish sentiments existed in New Zealand that affected job and marriage prospects. Signs noting that ‘no Irish need apply’ were common.
Bedroom politics
67
Discrimination, paired with a higher percentage of female Irish immigrants to male Irish immigrants, resulted not only in Irish women being forced to consider prostitution but in stratum barriers within the sex trade. Silver City saloon owners made a trip to Germany and brought back a number of girls to dance in the saloons at 50 cents a dance (Chadwick 1960: 50). In New Zealand, Hokitika and Greymouth hotel-keepers imported ‘dancing girls’ from Melbourne and Sydney. One such contract required the girl to serve as a barmaid, and to dance and to otherwise make herself generally useful for the term of twelve months (May 1967:310). Mrs Alpenny’s Registry Office in Dunedin provided barmaids to the West Coast. Her advertisement in November 1865 notes, ‘Mrs Alpenny now has on her books several respectable girls wanting situation as barmaids, and willing to go to Hokitika at the current rate of wages’ (West Coast Times 7 November 1865:4). Openings for women were also advertised locally: ‘Wanted two respectable Girls for the Star Hotel Kanieri’ (West Coast Times 11 November 1865:3). Life as a dance-hall girl did not directly involve prostitution. Whether to engage in prostitution was up to the discretion of the women, but the work did involve physical contact with miners, making them even more accessible than brothel prostitutes. At Bannockburn, a small mining camp on the Otago fields, the local newspaper noted, ‘The dancing that followed was kept up for three or four hours by some 30 males and two females. Struggles for the possession of the fair demoiselles were the source of fun not included in the programme’ (Cromwell Argus April 18732 in Mackay 1992: 102). Whether this access to female companionship increased the demand for the resource or decreased the value of the resource is questionable. On the basis of Collins’s theory of sexual stratification the value should have decreased. In actuality, devaluation was nullified by the rarity of women on the mining frontier and the fluid nature of social attitudes.
INDIGENOUS WOMEN The involvement of Polynesian women, Hawaiian in the USA, or ‘Kanakas’ as they were called there, is recorded in newspaper accounts. In New Zealand, Kanakas were Melanesian, and no records of Melanesian women involved in mining settlement prostitution have been found. The involvement of New Zealand Maori women in prostitution has been recorded primarily in association with whaling settlements (Eldred-Grigg 1984; Brooks et al. 1992; Belich 1996). Goods were the currency of the transaction. Economics probably resulted in some Maori women’s involvement in prostitution during the mining period, but there is little documentation. The involvement of Amerindians in prostitution was noted in both early Virginia City and Jacksonville. Because of the demand and economic stress which southern Oregon tribes and the Paiutes of Virginia City were experiencing, prostitution by native women is not unexpected. Many native
68
Alexy Simmons
women in both the USA and New Zealand became common-law wives to miners, and as a result had even less recourse than a legal wife when the relationship ended. Stories of wives becoming prostitutes when widowed or deserted are common to most mining communities.
CHINESE EMIGRATION Chinese emigration was a result of strong political and economic pressures. Overpopulation in China coupled with the breakdown of the irrigation system as a result of the decline in the Qing Imperial Government led to frequent famines. Combined with the Opium Wars (1840–1842, 1853–1860) and internal rebellion (Taiping Rebellion 1851–1864), this led to the push to emigrate from southern China (Ip 1995). They came to (the old) Gold Hill (United States) and New Gold Hill (Australia and New Zealand) as sojourners with familial economic obligations in China and no mining experience (Ip 1995; Ng 1993). Chinese prostitution in the USA In the historic Chinese mining communities of the USA and New Zealand, the Chinese operated within their own community. Culturally based ideas about women differed from those of Europeans. In the U S Chinese mining community there was acceptance of prostitution as the act of a ‘good daughter’, not a moral degenerate or a criminal. Chinese prostitutes were ‘big-footed’ peasant women from southern China. The Chinese family was patrilocal in structure; thus a girl child would contribute to the economic well-being of the family only briefly as an adult. The economic return on raising a daughter could come only from her purchase or sale. Bride price, sale as a concubine and contracts that indentured young women as moy nui (‘little sisters’ —maids and companions) were commonly negotiated by the family. The contract system was open to abuse and often used as a channel for obtaining prostitutes. Chinese women were also procured directly from their family as indentured prostitutes. Others were sold for a bride/concubine price, or alternatively kidnapped or conned. Lee Yow Chun in testimony to the US House of Representatives typified the Chinese women procured for prostitution: Mother told me one day that a go-between had been to see her on behalf of a wealthy Chinese merchant living in San Francisco, who wanted a wife from China; that she wanted me to go out and take a walk, so that I could be viewed by the go-between, that she could make her report. I did what I was bid, though I knew that mother told me with some misgivings, as he had declined the offer once and would decline again but for her good nature and her poverty. The go-between having made her a present of
Bedroom politics
69
$380 as coming from my intended husband, mother said she consented to take so little because I could only marry that merchant as a concubine and also that in two or three years I could come back to pay her a visit. (US House of Representatives 1901:774) Chinese prostitutes could be purchased out of the profession through the economic satisfaction of their contract. The contract often stipulated the cost and other requirements relating to the resale of the prostitute. Chun Ho’s contract for prostitution allowed her to be purchased for $2,100 after fulfilment of two years of her four-year contract. The economic means of the average Chinese miner limited his ability to purchase a wife or concubine under contract. In the USA, some concubines were owned by several Chinese miners. Economics also played a key role in the integration, segregation, moral attitude and acceptance of prostitutes in the Chinese community. A class system existed in Chinese prostitution that followed traditional lines. High-stratum Chinese women were cloistered from contact with Euro-Americans. The Chinese male view held that it was degrading for Chinese women to have sexual intercourse with a white. The services of low-stratum Chinese prostitutes were sold to all. The majority of the Chinese prostitutes in the mining towns were considered low stratum within the sex industry. US trade in Chinese prostitutes The Hip Yee Tong dominated the US Chinese prostitution trade in the mid1870s. Between 1852 and 1873 the Hip Yee Tong imported 6,000 women, or 87% of the total female Chinese population immigrating between those years (Hirata 1979). Many of these women worked in the Chinese brothels of San Francisco; others were sold to service the mining camps. In Silver City, for example, the same Chinese prostitutes worked in that community for at least ten years. The average age of the Chinese prostitutes in 1870 was 25, in 1880 the average age was 32 and a number of the names listed in the 1870 census are similar to those in the 1880 census. There was also a shift from households composed of two women to individual households, suggesting that the Chinese women may have changed their status from indentured prostitutes to independent prostitutes. The Chinese and prostitution in New Zealand There were no Chinese prostitutes in New Zealand (Ritchie 1986; Ng 1993). The Reverend Don, who worked with the central Otago Chinese, noted, ‘There is not, as yet, the least appearance of the importation of Chinese women for immoral purposes’ (Ng 1993:119). Immigration controls have been credited with the lack of Chinese prostitutes. Economics and market control may be the real underlying reasons. The Chinese population size was relatively low in New
70
Alexy Simmons
Zealand in comparison to the USA, resulting in a lower demand. The Hip Yee Tong also appear to have had no trade interests in New Zealand, and historical evidence suggests that control on the supply of Chinese prostitution was exerted at the source. In describing the Chinese miner population in New Zealand, Ritchie (1986: 44) noted, ‘during the nineteenth century the Chinese population in New Zealand was virtually all male’. Thirteen years after the influx of Chinese miners in 1878, there were 4,427 male Chinese and 15 Chinese women in New Zealand (Ng 1993) (the census records 9 women). John Ah Tong, the only Chinese man to be interviewed by the New Zealand Select Committee on Chinese Immigration in 1871, [reported] that there was no law against women leaving China, but after the large exoduses to California the ‘head men’ would only allow wives to go and join their husbands or women who were sent for and had a definite promise of marriage upon arrival. (AJHR 1871, H5:53 in Ritchie 1986:45–46) Almost as soon as Chinese settlements were established in New Zealand, they attracted European women. These women were described as ‘of the lowest type’, or women who would find it difficult to induce any respectable European to marry them (AJHR 1871 H5; Butler 1977; Ritchie 1986). Irish women were said to have resided in Chinatown. Whether these women became wives, concubines or prostitutes is not known. Newspaper reports indicate there were some European women married to Chinese men. Cultural ties to China, as well as familial ties and wives in China, reduced the number who married. Mixed marriages were relatively more common in New Zealand than in the USA, probably because of the deficit of Chinese women in New Zealand and the high number of European women associated with some Chinese settlements. The Chinese settlement at Tuapeka, New Zealand, was located on 0.4 ha of wet ground 1.2 km north of Lawrence next to Tuapeka stream and the main road. Records suggest that in the 1880s eight to eleven Chinese men in the camp were married to European women, several of whom had been prostitutes and continued to be associated with the sex trade. Ellen Sing Lee and her husband were charged with permitting three prostitutes to meet together and remain in their house. Ellen was later imprisoned for one month for assaulting prostitute Jennie McKnight. In 1881 Elisabeth See Que’s husband Wong was beaten by Mr and Mrs Humphry when he tried to recover his wife from a house of vice in Lawrence, where she had eloped with a European man. The Humphrys’ house was described as ‘one of the worst if not the worst in Lawrence’ (Ng 1993:258). In nearby Cromwell, ‘Qui Sing, was charged with keeping a brothel in Melmore Terrace…and sentenced to two months hard labour’ (Cromwell Argus 20 June 1896; Ritchie 1986:45). Obviously Chinese brothels operated in New Zealand’s Chinese camps and towns, but the prostitutes were European, not Chinese.
Bedroom politics
71
MATERIAL CULTURE Historical data suggest the kinds of material culture associated with prostitutes of various strata, based on the type of service provided and the context or phase of community occupancy. The type of attributes of material goods used becomes the means by which information about the system is transmitted visibly and concretely (Pyszczyk 1988:66). An artefact’s quality or function may be less a result of personal need than of the social milieu within which the owner lived. Thus material culture is often a visible display of affiliation or status. The possessions of prostitutes include the usual possessions of women, plus the essentials of their profession and stratum markers (e.g. clothing, consumables such as food and alcohol, medicines, tobacco and pastimes). The most effective goods to communicate group affiliation or differentiation are ones that are highly visible to others (Pyszczyk 1988:55). In comparison to women factory workers, workingclass brothel prostitutes had a relatively large amount of clothing and relatively numerous personal artefacts, and consumed expensive food types. Women workers (outside the sex industry) often reported the difficulty of dressing in style and eating well on the wages they earned (Pearsall 1976; Hammerton 1979; Seifert 1991). Seifert (1991) made a comparative analysis of a brothel and several working-class black and white households and confirmed archaeologically the comparatively ‘good life’ of prostitutes. Material culture must be understood in context. Artefacts from contemporary households and businesses should be used as a comparative gauge with which to understand the artefactual remains of brothels and other establishments of prostitution. Deficits in material culture, such as a lack of men’s tools, must also be considered in identifying probable houses of prostitution. Parlour-houses and brothels Parlour-houses and brothels were larger than the other establishments that provided prostitution services. They catered for entertainment as well as sex. The brothel is a distinctive type of household, composed of boarding women who live together in their workplace (Seifert 1991). Often referred to as ‘female boarding’ (USA), unlike other boarders prostitutes did not go away from their boarding-house to work. In the higher-class brothels and parlour-houses, the provision of food as well as alcohol was part of the entertainment service. In some instances more might be spent by a customer on alcohol and other goods than on sexual services. Artefacts associated with men are those indicative of a brief visit, entertainment and sexual interaction. A general hypothesis of functional artefact category percentages expected in the archaeological record of contemporary brothels, hotels and working-class households in the same community is presented in Table 4.1. The comparative percentages are based on Seifert (1991) and on archaeological data on frontier hotels in New Zealand (i.e. Halfway House Hotel (Bedford 1986),
72
Alexy Simmons
Table 4.1 Hypothesis: percentages of functional artefact category expected in the archaeological record of contemporary frontier brothels, hotels and workingclass households
B, brothel; H, hotel; F, working-class household Note: Children’s artefacts include clothing, shoes, toys, school slates, etc.; prostitutes’ children generally boarded out of the brothel and often in other cities. Despite a high rate of sexual activity, prostitutes did not have a high birth rate, owing to disease, abortion, miscarriage, etc.
Rotomahana Hotel Catalogue (Simmons n.d.), Three Hotels in Cromwell (Simmons and Ritchie n.d.), the Blue Goose Brothel in Goodsprings, Nevada (Becker 1981), and historical records on prostitutes (Simmons 1989). Comparative analysis of places occupied by prostitutes with contemporary households and hotels alleviates problems that arise from variability such as isolation from goods, economic fluctuation, community size, demographic factors, community phase, etc. Hypotheses about comparative artefact frequency will differ for women operating out of a house/cottage, as did many mid-status and low-status prostitutes. The isolated residence of a low-status prostitute may be
Bedroom politics
73
difficult to distinguish from that of a widow; in many cases they were one and the same, because of economic pressure. The identification of women involved in occasional prostitution is problematic and reliant on several signature artefacts. Signature artefacts may include medicines used for treating venereal disease in association with a high percentage of men’s clothing fasteners and artefacts relating to women. Julia Bulette was a mid-status prostitute who worked out of her residence. Her probate will and newspaper articles relating to her death provide the following picture of her furnishings. The cottage parlour was furnished with a carved black walnut set including a sofa, two rocking-chairs and four matching cane chairs arranged around a stove. The cottage had no kitchen. The bedroom windows were covered with damask curtains and contained a small box stove and mahogany bedstead, two washbasins and two spittoons (Simmons 1989). At the time of her death Julie had an outstanding liquor bill that included ale, rum, bourbon, whisky, brandy, port and claret (McDonald 1980). The personal ornaments described in the notice of sale excluded minor articles like hairpins and combs. Among the items listed were a gold hunting watch, chain and charms, watch case, jet breast pin, earrings, cross and silver brick marked ‘Julia’ (Territorial Enterprise 17 March 1868:44). Differentiation based on status differences and associated provision of entertainment (alcohol and food) as well as sex will aid in discerning the mid-status establishment from those of lower-status prostitutes. As with most class lines, differentiation is not usually clear-cut. In the lower-class brothels there was little in the way of romantic trappings or service: Carter Patrick Coyle visited Louisa O’Brien, a Hokitika prostitute. He gave her two pounds and got into bed without taking his boots off. Her brothel-keeper later came into the room and stole 20 pound from the pocket of the sleeping client, and then tried to blame it on Miss O’Brien when the police investigated the theft. (Mackay 1992:89) The low-status cribs or adjacent one-room structures, usually in a line, may be identifiable by their group signature. The poverty of these structures individually may provide little evidence of their use, as a well-known evocative description of a crib suggests: A bed in one corner, in another a stove, a coal bin and a bundle of kindling. A small dresser with a wash basin was against the wall. Permeating every thing a mixed odour of disinfectant, hair oil and cheap perfume. On the walls a few art pictures, oddly enough, usually of some pastoral scene. (WPA 1943:180) Others involved in prostitution that are difficult to identify in the archaeological record are women associated with establishments such as hotels or saloons.
74
Alexy Simmons
Disposal patterns, occupational supposition and fire may result in substantial information deficits in the archaeological record of the initial mining settlement. Archaeologically the tent brothels, like tent residences and shops, may be almost invisible in the archaeological record. Yet this period of initial instant settlement is of the most significance for considering social changes on the frontier. The markers of Chinatown prostitution Chinese prostitutes should be identifiable in association with women’s artefacts such as hair ornaments, jewellery, clothing, and medicines for females, but owing to the wide variety of artefacts of Chinese manufacture used by both genders, gender variation may be indiscernible. Gender-related artefacts largely consist of ornaments because of the generic male/female composition of Chinese traditional clothing and use by both males and females of fans, the same body-care products and medicines, as well as pastimes (mah jong, gaming pieces) and indulgences (opium and rice wine). For example, Ritchie (1986) found French perfume bottles in isolated Chinese miners’ huts and towns where there were no recorded women. Upper-class Chinese prostitutes may be easier to identify in the archaeological record, but few of this class were evident in the mining towns. In New Zealand Chinatowns/camps, European prostitutes may be discernible because of gender and artefacts that can be linked to the profession.
ECONOMIC AND SOCIAL RELATIONSHIPS In the frontier mining towns of the western USA and New Zealand, prostitutes (other than Chinese ones) largely controlled their own income. Notwithstanding that portion of their income which was expended on operational expenses (such as bribes and purchase of the trappings and refreshments associated with their business), female prostitutes controlled their earnings during a period when women had few job opportunities and even less financial control. There were silent partners or investors who had little control over the daily operation of the sex trade but exerted an influence on political decisions that might have affected the business. Madame Josephine Airey (Joe Chicago) of Helena, Montana, mortgaged her brothel to Harriet Truett, wife of the probate judge (Petrik 1981). Joe entered into other transactions with the business community over the years, including loans from dry goods merchants Salig and Alex Larenberge, and Jacob Switzer, wholesale liquor merchant, who commonly lent money to the demimonde (Petrik 1981). Most merchants ran lines of credit for the madams and higher-status prostitutes. Business relationships commonly existed between saloon proprietors and prostitutes. Maria Clapper, a Silver City prostitute, held common investments with William Williams, a saloon keeper (as revealed in her probate will) (Simmons 1989).
Bedroom politics
75
The relationship between the demimonde and the business community had implications linked to their settlement pattern and resulting informal restrictions. Spain (1992:18) notes, ‘Most status differences are reinforced by subtle forms of spatial segregation. Instead of being visibly manifest in spatial barriers, status hierarchies often are determined in secret. Secrets, in turn, are preserved often through spatial boundaries.’ In Silver City, prostitutes were allowed to come in on the stagecoach and out on it, but not to wander in the central trade area of the city. There were no ordinances relating to mobility or settlement zoning, only informal agreement. European shopkeepers delivered goods to the prostitutes, although historical ledgers show that prostitutes did collect their orders from shops such as Andy Capp’s cigar shop, which was adjacent to a brothel; they could also move freely in adjacent Chinatown. Shop-keepers often forced the relocation of red light districts, as was the case in Cripple Creek, Colorado, after the fire of 1896. In 1883 in Helena, Montana, business transactions with the demimonde ceased and the regular commercial district moved away from the red light district. In 1886 prostitution was prohibited in the city, with limited success (Petrik 1981). The New Zealand police had an interactive relationship with the demimonde. Sergeant McMwyn was the 1866 owner of a Hokitika brothel. His ownership was not unusual, given the example set by policeman Martin Cash, owner of the well-known ‘Red House’ and ‘White House’ in Christchurch (Eldred-Grigg 1984). When Sergeant Thomas Ryan arrived in Fullerton (Mt Ida District) with his revolver drawn to the distaste of the populace, [he] made it clear authority resided in him. When he then repaired to an insubstantial brothel, however, an angry crowd gathered, the tent strings were cut…. Ryan emerged from under the canvas firing his revolver and the crowd fell upon him and administered a beating. (Hill 1989:603) The prostitute was not included in the beating. The integration of prostitutes and the community varied over time. Factors linked to social attitudes about prostitution include acceptance of prostitution as a profession and willingness to acknowledge that acceptance through overt actions (e.g. involvement in business partnerships, loans, public appearance in a prostitute’s funeral procession). Lea Perry, a Silver City madam of the early twentieth century who ran a brothel adjacent to the commercial district, was well liked and respected. Despite her acceptance she was still considered to be outside the regular community. Miner Ned Williams said that when Perry died, all her pallbearers were married men and she was buried as close to the Ruby City cemetery gate as they could place her (Simmons 1989). Others of the demimonde fared better in death. Madam Jessie Lester, who left her estate to the Sisters of Charity, was buried in consecrated grounds. Her probate will indicates that she owned a high-status brothel. Julia Bulette, a mid-stratum
76
Alexy Simmons
prostitute, was buried in the Flowery Hill Cemetery with a procession of sixteen carriages, the Engine House No. 1 company and the Metropolitan Band (McDonald 1980). Differences in integration existed along status lines. Many madams and prostitutes, like other business people, remained in a community for most of their working life. In the mid-1860s, Madam Cad Thompson of Virginia City operated a parlour-house that had a grand piano and rooms downstairs for the girls (Territorial Enterprise 20 November 1866). The 1880 census attests to her continued role as a resident of the community and head of a household of four women. Cad was described by newspaper man Alf Doten as ‘the Brick’. She was never mentioned in the newspapers in association with disorderly conduct or arrests, although she did suffer from a prank that involved the fire engine company. The integration of Cad Thompson contrasted with that of outcast prostitute Liz Hayes of the same city. Alf Doten noted in his journal (21 June 1869): ‘Liz Hayes buried this PM—no one at funeral, not even fancy women…was 28 years old and her real name was Maggie Glynn—Born in Boston—leaves a 12 year old boy at San Francisco’ (Clark 1973:1050). Barbara Weldon of New Zealand’s west coast was frequently in gaol for her loud, drunken escapades. A gaoler’s report described her as a perfect pest outside the gaol, and a continual source of trouble when in gaol (Dictionary of New Zealand Biography 1993:581). The Dunedin police had packed Barbara Weldon onto a coach to Hokitika in 1869 with a one-way ticket to rid their respectable town of her.
REGULATION OF PROSTITUTION US mining communities had greater political autonomy than the New Zealand goldfields. In New Zealand, authorities drew on overseas experience in governing the goldfields, especially the experience of Victoria, Australia. Legislation based on the Victorian model for governing the goldfields was introduced in 1858 and adapted in 1860 to the New Zealand situation. New Zealand laws on prostitution were based almost entirely on English precedents (Eldred-Grigg 1984). Parliament allowed people the right to sell sexual favours, but tried to keep the traffic orderly, and thus to control conduct and disease. The philosophy behind such a system maintains that prostitution cannot be eradicated and therefore must be controlled (Davidson 1984). The 1869 venereal disease ordinance saw little use in the frontier mining towns. The 1824 and 1866 vagrancy laws were generally applied for purposes of police arrest of prostitutes. The vagrancy law gave the courts the power to punish prostitutes found ‘wandering’ in public places and behaving in a ‘riotous or indecent manner’ (Eldred-Grigg 1984). Prostitutes tried under the vagrancy law often experienced little more than a verbal reprimand. Of the sixty-three prostitutes brought before the magistrate Charles Bowen between 1864 and 1871, almost half were released with a warning or required to quit their brothel.
Bedroom politics
77
The economic cycle of the mining community gave rise to the regulation of prostitution and informal restrictions within the settlement. With the growth of the camp into a more complex political unit, i.e. a town, the sex industry often faced restrictions that included delineated trade areas and employee segregation. Zoning was not commonly used to restrict the location of other business types on the mining frontier during the period from 1850 to the early twentieth century, but was used both formally and informally to restrict the sex trade. The effects of the economy underlay many of the actual changes in the aggregation of prostitutes. Changes in prostitute aggregation patterns can be used to identify economic and resulting social changes in the frontier mining community; there are of course exceptions such as areas that have a high population of wives and daughters. Table 4.2 provides a general hypothesis of mining community economic phases and prostitution by strata and community integration within this economic context and town type, based on historical records for Silver City, Virginia City, Jacksonville, Helena and Cripple Creek in the USA (Simmons 1989) and Thames, Lawrence, Cromwell and Hokitika in New Zealand.
SUMMARY The lifeway of prostitutes and female entertainers in the mining towns changed over time as a reflection of the context within which they worked. The economy of the frontier mining community was based on an exhaustible resource with variable geological composition. Other factors also affected the community, such as paper economies based on stock options, natural disasters such as flood or, more commonly, fire, or transportation systems and proximity to an established urban city. Female prostitutes as elements of the community wherein they operated leave a historical and archaeological record of daily activities associated with their business and their lifeway as women. The spatial dimension of their lifeway changed over time in relation to changes in the community. The aggregation pattern is indicative of their status in the community, community composition and economic changes. There was a hierarchy of status within the stratified demimonde. Increased consideration by archaeologists and anthropologists can and should be given to the composition of the mining community by examining spatial factors and their link to gender, profession and race. Female prostitutes as a group represent a specific gender and profession, and are often members of minority ethnic or racial groups. Gender, work and racial spaces must be examined more closely so that the relationship between social constructs and social relations can be exposed. Aggregation patterns are shaped by the dominant group’s ability to control and the subordinate group’s willingness to accept control. This control was relaxed on the mining frontier in the initial periods of community settlement and during times of rapid change.
78
Alexy Simmons
Table 4.2 General hypothesis of types of non-Chinese prostitution and settlement pattern by economic phase and town type
L, low-stratum prostitution: crib, 1–2 women operations, vagrants; M, mid-stratum: 1–2 women operations, brothels, dance-hall girls/hurdy-gurdies; H, high-stratum: brothels and parlourhouses, entertainers (actresses/singers); +, present to a greater degree, - present to a lesser degree. Note: The number of Chinese prostitutes, who were lower-stratum, generally increased and decreased parallel to economic phases apparent in the non-Chinese community.
Bedroom politics
79
NOTES 1 2 3 4
West Coast Times, Hokitika. National Archives, Wellington. Cromwell Argus, on file, Cromwell Museum, Cromwell, New Zealand. Appendices to the Journal of the House of Representatives, New Zealand. Territorial Enterprise, Virginia City, Nevada. Microfilm, Nevada State Library.
REFERENCES Becker, R. (1981) Mining communities: how to interpret a Blue Goose. Unpublished MA thesis, Department of Anthropology, University of Nevada, Las Vegas. Bedford, S. (1986) The Final Report on the Excavation of the Halfway House Hotel Site, Cromwell Gorge. Cromwell: New Zealand Historic Places Trust. Belich, J. (1996) Making People: A History of New Zealand. Auckland: Allen Lane/Penguin. Bourdieu, P. (1977) Outline of a Theory of Practice. Cambridge: Cambridge University Press. Brooking, T. and Rabel, R. (1995) Neither British nor Polynesian: a brief history of New Zealand’s other immigrants. In S.Greif (ed.), Immigration and National Identity in New Zealand: One People, Two People, Many People, pp. 23–49. Palmerston North, NZ: Dunmore Press. Brooks, B., Macdonald, C. and Tennant, M. (eds) (1992) Women in History, vol. 2: Essays on Women in New Zealand. Wellington: Bridget Williams Books. Butler, P. (1977) Opium and Gold. Martinborough, NZ: A. Taylor. Chadwick, A.G. (1960) Tales of Silver City. Boise, ID: Boise Printing Co. Clark, W. Van T. (1973) The Journals of Alfred Doten 1849–1903, vols 1–3 . Reno: University of Nevada Press. Collins, R. (1972) A conflict theory of sexual stratification. In H.P.Dreitzel (ed.), Recent Sociology no 4: Family, Marriage, and the Struggle of the Sexes, pp. 53–79. New York: Macmillan. Davidson, R. (1984) Dealing with the ‘social evil’. In So Much Hard Work: Women and Prostitution in Australian History. Sydney: Fontana/Collins. Davis, K. (1966) Sexual behavior. In R.Merton and R.Nisbet (eds), Contemporary Social Problems, 2nd edn, pp. 322–372. New York: Harcourt, Brace & World. Davis, R.L. (1967) Soiled doves and ornamental culture. The American West 4(4): 19–25, 69–70. Dictionary of New Zealand Biography (1993) The Dictionary of New Zealand Biography, vol. 2: 1870–1900. Department of Internal Affairs. Wellington: Bridget Williams Books. Eldred-Grigg, S. (1984) Pleasures of the Flesh: Sex and Drugs in Colonial New Zealand 1840– 1915. Wellington: A.H. & A.W.Reed. Giddens, A. (1979) Central Problems in Social Theory. Berkeley: University of California Press. Hammerton, J. (1979) Emigrant Gentlewomen: Genteel Poverty and Female Emigration, 1830– 1914. London: Croom Helm. Heyl, B.S. (1979) The Madam as Entrepreneur. New Brunswick, NJ: Transaction Books. Hill, R.S. (1989) The Colonial Frontier Tamed; New Zealand Police in Transition 1867–1886. Wellington: Dunmore Press. Hirata, L.C. (1979) Free, indentured, enslaved: Chinese prostitutes in nineteenth-century America. Signs 5(1): 3–29. Ip, M . (1995) Chinese New Zealanders. In S.Greif (ed.), Immigration and National Identity in New Zealand: One People, Two People, Many People, pp. 161–199. Palmerston North, NZ: Dunmore Press.
80
Alexy Simmons
Lee, M.B. (1958) Cripple Creek Days. Garden City, NY: Doubleday & Co. ——(1968) Back in Cripple Creek. Garden City, NY: Doubleday & Co. Macdonald, C. (1990) A Woman of Good Character. Wellington: Allen & Unwin New Zealand Ltd. McDonald, D. (1980) The Legend of Julia Bulette and the Red Light Ladies of Nevada, ed. Stanley Paner. Las Vegas, Nevada: Nevada Publications. Mackay, D. (1992) Frontier New Zealand: The Search for Eldorado 1800–1920. Auckland: HarperCollins. May, P.R. (1967) The West Coast Gold Rushes. Christchurch, NZ: Pegasus. Morrill, R. (1970) The Spatial Organization of Society. Belmont, CA: Duxbury Press/ Wadsworth Publishing Co. Ng, J . (1993) Window on the Chinese Past, vol. 1. Dunedin: Otago Heritage Books. O’Kelly, C. (1980) Women and Men in Society. New York: Van Nostrand. Pearsall, R. (1976) Public Purity, Private Shame: Victorian Sexual Hypocrisy Exposed. London: Weidenfeld & Nicolson. Perkins, J. (1993) Victorian Women. London: John Murray. Petrik, P. (1981) Capitalists with rooms: prostitution in Helena Montana. In Montana (Spring): 28–42. Montana Historical Society, Helena, Montana. Pyszczyk, H. (1988) Economic and Social Factors in the Consumption of Material Goods in the Fur Trade of Western Canada. Volumes in Historical Archaeology, vol. 6. South Carolina Institute of Archaeology and Anthropology, University of South Carolina, Columbia. Ritchie, N. (1986) Archaeology and history of the Chinese in southern New Zealand during the nineteenth century. Unpublished PhD thesis, Anthropology Department, University of Otago, Dunedin. Robinson, J. (1984) Canterbury’s rawdy women: whores, madonnas and female criminality. New Zealand Women’s Studies Journal 1(1): 6–25. Seifert, D. (1991) Within site of the White House: the archaeology of women. Historical Archaeology 25(4): 82–108. Simmons, A. (1989) Red light ladies: settlement patterns and material culture on the mining frontier. Anthropology Northwest, no. 4. Corvallis: Oregon State University. ——(n.d.) Rotomahana Hotel Collection, Te Wairoa. Unpublished catalogue. Simmons, A. and Ritchie, N. (n.d.) Three hotels in Cromwell. Unpublished manuscript. Spain, D. (1992) Gendered Spaces. Chapel Hill: University of North Carolina Press. US House of Representatives (1901) Reports of the Industrial Commission on Immigration, vol. 15. Fifty-seventh Congress of the USA, Government Printing Office, Washington, DC. WPA Writers Project (1943) Copper Camp. New York: Hastings House.
Chapter 5
Power and the industrial mining community in the American West Donald L.Hardesty
ABSTRACT This chapter focuses on archaeological approaches to the power structure of silver-, gold- and copper-mining communities in the American West in the social and cultural context of nineteenth- and twentieth-century industrial capitalism. Using examples from Nevada, California and Colorado, the chapter explores the material expression of power in the context of productive relationships mediated by class, gender and ethnicity, especially patterns and strategies of domination and resistance. The archaeological remains, architecture and landscapes of company towns, for example, ref lect domination strategies. In contrast, the material expression of ‘alternative’ towns growing up outside the limits of company control reflect strategies of resistance. Alternative towns such as Reipetown in eastern Nevada’s copper-mining region, for example, offered saloons, gambling-houses, brothels, labor unions’ cheap housing and more individualistic lifestyles.
THE INDUSTRIAL MINING COMMUNITY IN THE AMERICAN WEST The industrial metal-mining communities of the American West offer a unique opportunity to explore the anthropology and archaeology of power. These communities, first emerging on western Nevada’s Comstock Lode within a few years after its discovery in 1859, were an outgrowth of the expansion of industrialism, including industrial technology, highly mobile wageworkers, the labor movement, ethnic enclaves and corporations, throughout the USA (Robbins 1994; Schwantes 1987; Trachtenberg 1982). During the late nineteenth century, the American West in general was transformed from ‘a region dominated by preindustrial societies to a fully integrated segment of the modern world capitalist system’ (Robbins 1994:147). The building of transcontinental railroads quickly developed the industrial infrastructure of the American West and attracted global capital investment. By the turn of the century, large-scale corporations and monopoly capital controlled most
82
Donald L.Hardesty
industries in the American West and strongly integrated the region into global markets and the modern world-system. Most commodities harvested in the region—coal and metals, timber, fish and farm crops—were exported with prices set in the global marketplace (Robbins 1994:180; Schwantes 1987:41–42).
POWER AS AN ORGANIZING PRINCIPLE Power was the key organizing principle underlying western industrial mining communities. Social power is ‘a set of processes whereby one party gains the capacity to impose its will on another by its potential to contribute or to withhold critical resources and rewards or by threatening or inflicting punishment’ (Lipman-Blumen 1994:110). Mining-related critical resources, for example, might include capital, labor, water and access to ore bodies. However, power relationships among individuals and groups, based on controlled access to resources, are constantly being negotiated because of changing meanings or values of the resources. Power relationships are thus ‘situational.’ Power networks The western industrial mining community is essentially a network of power relationships. Sociologist Michael Mann (1986:1–2) thinks of social groups such as the mining community as overlapping power networks rather than as bounded systems. He identifies the four key power networks that connect individuals into social groups as ideological, economic, military and political. In the modern world-system, the four networks may be closely linked but historically are not necessarily linked. Military power, for example, may have operated independently in the past. Individuals and groups negotiate power relationships within these networks. Power networks in western mining communities sometimes approximate hierarchies, such as in company towns, but typically are much more flexible, situational and changing. Crumley (1994:186–187) refers to such power networks as heterarchies. Informal mining and power structures What Gordon MacMillan (1995), in a book on the 1980s Amazonia gold rush, calls an informal mining system is one component of a heterarchical power structure. Informal miners typically are smallholder farmers who engage in mining as a part-time activity during the farming off-season. Winthrop and Chambers (1988) develop such a model of informal gold mining for western Siskiyou County, California, for the period between 1900 and 1930 (see Figure 5.1). In the model, the mining community is marked by relatively small geographical scale and localization of labor, supplies, capitalization, profits and expenditures (Winthrop and Chambers 1988:19). Economically, the gold-
Power and the mining community in the American West
83
Figure 5.1 Map showing locations of sites mentioned in the text
miners are portrayed as mining seasonally or part-time and supplementing their income when necessary by engaging in other occupations such as wage labor or ranching. They are locally oriented and cash-poor, and have few ties to the outside world. The gold market itself often is the only real link to the outside world, which provides a small cash income that could be used to purchase manufactured goods. Most commodities were locally acquired by hunting, home gardening and bartering with neighbors. During the Depression era, the gold-miners were only minimally affected, since they already were locally oriented, diverse, cash-poor and self-sufficient, with little dependence upon the national economy. In addition, gold mining is economically counter-cyclical. Mobility and power structures Mobility is another component of heterarchical power structures in western mining communities. In a recent doctoral dissertation, for example, Allyson
84
Donald L.Hardesty
Brooks (1995) argues that the concept of anticipated mobility, first proposed by ethnoarchaeologist Susan Kent for interpreting the archaeological record of hunting gathering communities, often explains material variability in mining sites and landscapes better than class or ethnicity. The anticipation of a short stay levels or equalizes differences in the material culture of social formations within the community. Brooks found that an anticipated short stay encouraged a minimalist material culture such as dugout housing and tin tableware regardless of the social status of the individual. In contrast, the anticipation of a long stay increased differences in the material culture of class/ethnic/gender formations. Individuals brought with them a more complete repertoire of material things reflecting their social and cultural context. Mobility is also a key characteristic of a distinctive social class of wageworkers created by the industrial transformation of the American West. Carlos Schwantes (1987) developed the concept of a wageworkers’ frontier to describe the region, which David Emmons (1994:449) further describes as consisting of ‘thousands of “industrial cowboys” riding (the rails in this case) from one mining town or lumber camp to another.’ In general, geographical mobility, ethnic diversity and labor specialization worked against the development of a single working-class culture of wageworkers and toward social and cultural diversity (Dubofsky 1985:13; Trachtenberg 1982). Furthermore, Dubofsky (1985:16) contends that, in comparison to those in other regions, the wageworkers in the American West ‘proved to be the most radical, militant, and class conscious of working people.’ Western miners not only fit this description but also to a considerable extent were the archetype. In general, western mining communities ‘reflected the basic elements of industrial activity elsewhere in North America: the development of urban living patterns, the striking pervasiveness of wageworkers in those growing communities and a reliance on global markets for capital, prices, and labor’ (Robbins 1994:84).
CLASS AND POWER Sociologist Richard Hogan (1990) implicitly develops the concept of heterarchical power relationships in his discussion of the role of class in community formation and dynamics on the Colorado frontier. He focuses upon power networks involving social classes in the local community and their interplay with a national/global monopoly capitalist class for control of local industrial production. The power networks are based upon both relations of production and relations of exchange. Hogan portrays the economic and political institutions (e.g. municipal governments) evolving in local communities ‘as unstable coalitions representing the short-term interests of various classes that possess both the economic resources and the political organization required to defend their control of a local political economy’ (1990:208).
Power and the mining community in the American West
85
In this view, Hogan attempts to combine consensus and conflict interpretations of the development of frontier communities in the American West. The consensus perspective focuses upon the self-interested role of local boosters such as merchants and land speculators (e.g. Elkins and McKitrick 1954:321–353, 565–602). In contrast, conflict theorists argue that conflicts among local and regional groups play the pivotal role in explaining the evolution of frontier communities. Robert Dykstra (1968), for example, in his studies of cattle towns, followed Allan Bogue in pointing to the conflict that arose as the result of a lack of predetermined community structure in the political and social arenas. He argued that the lack of a stable social order intensified competition, which often resulted in conflict and instability. The lack of leadership, instead of forcing collective action, created factions or individuals who competed for power. In many cases, however, self-interested classes such as laborers, merchants and real estate speculators worked together in the formation of local governments to control industrial production. Hogan (1990) shows that in Colorado the struggle for local control of industrial production created two distinctive types of frontier communities. ‘Carnival’ communities formed local governments that placed emphasis upon the rights of persons where laboring classes were more or less economically independent and politically organized, such as in non-industrial mining towns. In such communities, laboring classes fully participated in public government. In contrast, ‘caucus’ communities with private governments that stressed property rights formed where local capitalist classes had unchallenged control of local industry and where laboring classes were dependent upon wages. Both types of local communities, however, competed with national/global monopoly capitalists and states for control of local industrial production. They differed in how the conflict was expressed. In carnival communities, for example, where labor defended its economic independence in the public arena, conflict took the form of struggle between labor and capital, such as mining strikes. The economic and political domination of caucus communities by local capitalist classes, in contrast, led to conflict between local capitalists and national/global capitalists, such as the railroad wars.
ETHNICITY AND POWER Another key player in the heterarchical power networks of western mining communities is ethnicity, which often integrates the community and overrides conflicts arising from other power networks or relations, such as those based upon class or gender. The role of ethnicity in the formation of the western mining community, however, is complex. David Emmons’s (1989) study of the evolution of the Irish community in Butte, Montana (see Figure 5.1), a coppermining town, for example, found that ethnicity played the dominant role in the early formation of the community. Despite some class differences, ‘those with
86
Donald L.Hardesty
power and those without were more alike in values than was ordinarily the case where disparities of wealth and status were involved’ (Emmons 1989: 180). Following Eric Hobsbawm, Emmons argues that such common values gave the Butte Irish the ‘power to exclude’ others from the workplace. Thus, they controlled the hiring process in the copper mines by dominating the local labor union. Furthermore, they achieved a high level of social respectability within the town of Butte. No stigma was attached to being an Irish miner. But common ethnicity could not hold the community together during economic recessions, such as in the 1890s, when unemployment was high. Radicalized out-of-work Irish miners and the threat of unemployment broke the ethnic bridge that once connected the working-class and middle-class Irish and dissolved the community. In contrast, Brian Frehner (1996) found that common ethnicity never bridged the gap between working-class and middle-class Italians in the silversmelting town of Eureka, Nevada (see Figure 5.1), and, consequently, an Italian community failed to develop. He attributes the schism to the contrast between urban and isolated rural life. Working-class Italians in Eureka were mostly charcoal-burners or carbonari who lived in isolated pockets in the remote mountains around the town, a lifestyle that encouraged the maintenance of traditional village cultures and provincialism. In contrast, middle-class Italians, many of whom were store-owners and charcoal merchants, lived in town and were class conscious, upwardly mobile and acculturated. Ethnic loyalty took second place to the protection of their social and economic investment in the local community. The so-called Charcoal Burners War of 1879 illustrates not only the dominance of class over ethnicity in the Eureka Italian community but also the role of violence in negotiating power relations. Town-dwelling Italian charcoal merchants interacted with the rural carbonari in a version of the padrone system. The charcoal merchants contracted with the smelting companies in Eureka to provide charcoal fuel at a negotiated price. They purchased the charcoal from the carbonari, who found that the smelters would not buy directly from them at the same price. Furthermore, the charcoal merchants paid with script redeemable only at their stores, which offered commodities at inflated prices. The carbonari soon organized the Charcoal Burners’ Protective Association, demanded a price of 31 cents a bushel for charcoal, threatened to cut off the charcoal supply to the smelters and forcefully prevented teamsters working for non-members of the association from loading and hauling their charcoal. In response, a deputy sheriff and other townspeople organized a posse, confronted the striking carbonari, and killed five of them. The Italian charcoal merchants joined non-Italian merchants and townspeople in opposing the association and supporting the posse. Giuseppe Tognini, one of the merchants, went so far as to refer to his fellow countrymen, the carbonari, as ‘foreigners’ in response to a newspaper editorial (Frehner 1996:58).
Power and the mining community in the American West
87
TOWARD AN ARCHAEOLOGY OF POWER The archaeological study of industrial mining communities in the American West, then, must focus on power, and necessarily begins with the supposition that the material world both reflects and creates a network of power relations. Shanks and Tilley (1987:107) argue that ‘material culture plays an active symbolic role in social relations. Interacting groups manipulate and negotiate, consciously and unconsciously, material symbols according to their strategies and intentions.’ In his book Uncommon Ground, for example, Leland Ferguson (1992) shows how slaves on antebellum plantations in the American South actively manipulated material things associated with architecture, foodways and ritual to create a distinctive ethnic identity. In the same sense, ‘class consciousness’ is a key concept in the industrial community that argues for the use of practices, material things and symbols in everyday lives that set off one class-defined group from another. Wageworkers in an industrial community, for example, should have practices and material goods that symbolize working-class identity in opposition to other social classes. Shackel (1993:18.2) further notes that ‘[m]aterial culture can construct power relations, define social boundaries and create hierarchical identities.’ Pierre Bourdieu gives an example in his book Distinction, A Social Critique of the Judgement of Taste (1984). He argues that ‘tastes’ in consumer goods helped to create and maintain the power structure among social classes in twentieth-century France. Comparing the tableware used by working-class and bourgeois families while entertaining, Bourdieu found significant differences. Working-class families focused upon the ‘substance’ of the food served at the meal and used ‘common’ tableware. The meal symbolized the contrast between the home and the public worlds. They entertained only friends who could be treated like ‘family’ and with whom they felt ‘at home.’ Bourgeois families, on the other hand, focused upon the ‘form’ of the service used for the meal and used ‘special-purpose’ tableware. The meal symbolized the continuity between the home and the public worlds. They often entertained professional and business acquaintances toward whom they felt social obligations and who required an appropriate ritual performance. Coping with power; cultures of domination and resistance Paynter and McGuire (1991) argue that the key to the archaeology of power is the way in which social dominance and subordination structure power relations among individuals and groups. Deagan (1996:366) observes that a key assumption of such an approach is that ‘the most important individual decisions and acts in shaping culture are those that reflect, mitigate, or enforce the neverending cycle of attempts by both individuals and groups to assert social dominance in some cases and resist it in others.’ Orser (1988) gives an example
88
Donald L.Hardesty
from antebellum plantations, which included the interplay between planters and slaves as social groups who worked out strategies of domination and resistance. The domination strategies of planters involved the use of force, withholding material goods and prestige jobs to discourage resistance, and the giving of valued goods and prestige jobs to reward good work. Slaves could resist such strategies, in turn, by ‘malingering, feigning ignorance, sabotaging machinery or tools, running away, or outright rebellion’ (Orser 1988:741). In the industrial mining communities of the American West, the key social groups playing such power games are mine owners or managers and wageworkers. Mine owners typically devised a domination strategy that attempted to control labor through mechanization to make workers produce more through specialization and repetition of tasks (see Paynter 1988). Workers may resist such a strategy by forming labor unions (e.g. the Western Federation of Miners) and with labor strikes. The strategies of domination and resistance often evolve into cultures that have a distinctive material expression visible in the archaeological record.
MODEL COMPANY TOWNS Model company towns, settlements maintained by a single mining company, best illustrate the culture of dominance that emerged in the industrial mining communities of the American West. Of recent origin, they first appeared in the mining camps and smelters of Appalachia and the Monongahela Valley in the late nineteenth century but rapidly spread into the mining industry of the American West (Allen 1966; Roth 1992). Mining companies actively attempted to control acts and beliefs within the towns, creating a social order ‘derived from labor routine, isolation, and company-imposed rules or policies’ (Gardner 1992:4) and an environment that reflected and reinforced company culture. The culture of dominance is expressed in the layout and landscaping of model company towns. The layout of the model company town of McGill in eastern Nevada (see Figure 5.1), for example, reflects a fairly rigid class structure organized around ethnicity, race, class, wealth and occupation. High company officials lived in an exclusive area called The Circle; middle management and skilled workers lived in Middle Town; and unskilled workers lived in the neighborhoods of Lower Town (Elliott 1990). Settlement patterns also reflect the culture of dominance. Consider, for example, the Robinson mining district in eastern Nevada (Hardesty et al. 1994). The district was organized in 1868 after the discovery of gold and silver deposits the year before. Copper deposits, however, proved to be the most promising. Early efforts in the 1870s to mill the copper ore failed, and mining virtually disappeared in the district for the next twenty-five years. The emergence of the Electrical Age by the end of the nineteenth century created a new demand for copper and renewed interest in the Robinson district. In 1904, the Guggenheim family entered the district and, after working out stock ownership and financial
Power and the mining community in the American West
89
agreements with competing companies, heavily capitalized and quickly developed the Robinson copper industry into a large-scale operation. Within a couple of years, the newly formed Nevada Consolidated Copper Company and other companies developed the copper mines, built a large smelter and company towns at McGill, Ruth and Kimberly, and constructed the Nevada Northern Railroad to transport copper from the mines to the smelter and then to the Central Pacific railhead for shipment to the refineries in the Salt Lake City area. In order to meet the demand for cheap labor created by the copper boom in the Robinson district, the mining companies recruited newly arrived immigrants with the help of labor bosses working with different ethnic groups. The immigrants mostly came from eastern and southern Europe, especially Greeks, Hungarians, Serbs, Croats, other Slavs and Italians, and from Japan. A classic industrial mining community with well-defined occupational, ethnic and social classes emerged in the Robinson district and was reflected in architecture, wealth differences, power and prestige. Settlement patterning especially well reflects power relations within the Robinson community. Company towns (e.g. Ruth, Kimberly and McGill) and the railroad/ redistributive town of Ely formed the economic and political power centers of the regional community, which also included several other towns, hamlets and isolated households along with the mines and mills. The copper mines and mills make up the economic center of the community. Radiating out from the mines are outlying neighborhood settlements such as the company towns, the satellite settlements, dairy farms, hay farms and isolated households. Road networks link together the center and the outliers. As a social interaction network, the regional community is a network of social nodes where people living in the outlying neighborhood settlements congregate, not only the mines and mills but also the medical facility, the schools, the churches, the company stores and the union halls. The regional community can also be viewed as a diverse functional network with specialized functions at different geographical nodes: McGill as a milling node; the copper mines as extractive nodes; Ely as a redistribution center for people and services and supplies made possible by warehouses and railroad facilities (for example, Ely’s White Pine High School serviced students from the entire Robinson district by means of the railroad); the company towns as residential nodes for most of the workers and managers, and the like.
SATELLITE TOWNS: A CASE STUDY OF REIPETOWN, NEVADA In addition to model company towns, industrial mining communities in the American West included satellite towns with alternative lifestyles and subcultures that reflect resistance strategies. In the Robinson district the best known of these was Reipetown (see Figure 5.1). The paper trail of Reipetown goes back to Richard Reipe, a German immigrant and prominent resident of the
90
Donald L.Hardesty
area, who developed a building stone quarry and small workers’ settlement at the locality in the 1890s. When the quarry failed shortly thereafter, Reipe sold the land, the new owners filing a plat (proposed plan of town/settlement) on the property in 1907. The town boomed as a wide-open mining camp with cheap housing and less than respectable entertainment not permitted in the tightly regulated new company towns (Figure 5.2). From the beginning, immigrants from southern and eastern Europe dominated Reipetown’s population. In 1908 the town boasted eleven saloons, two groceries, two restaurants, a hotel and several other ‘places of business’ (Ely Daily Mining Expositor). During the next ten years, the town’s bad reputation grew, fueled not only by the rapid expansion of saloons, gambling-houses and brothels but also by Reipetown’s emergence as a hotbed of union radicalism. C.E.Mahoney, a vice-president of the Western Federation of Miners, lived in the town for several years as a union organizer and became the town’s first and only mayor when it was incorporated in 1918. The state of Nevada legislature, however, disenfranchised Reipetown, along with several other small towns, the next year, citing their total disregard for good taste. After World War I, the reduced demand for copper forced the Robinson district mines and mills to curtail production, ushering in a period of depression that lasted several years. Not so, however, for Reipetown, which benefited shortly thereafter from the passage of the Volstead Act and the Prohibition era. The town soon enjoyed a reputation as a center of the ‘best bootleg liquor’ in the area. Among the leaders of the new industry were the Data brothers, two Italian immigrants, who operated a store and saloon in the town before becoming liquor dealers and who ultimately became the most prominent citizens of the town. Al Capone reportedly sent an intimate lady friend to Reipetown to operate a brothel during the same period, and brothels continued to be a featured attraction of the town. With the end of prohibition, however, and the continued decline of copper production in the district, Reipetown finally sank into depression and much of its population left, abandoning residential and commercial buildings. After the late 1930s, the copper industry revived and World War II started another boom in the Robinson district. The War Production Board declared copper to be vital to the war effort and established production quotas. Reipetown once again overflowed with miners unable to find housing in, or unwilling to live in, the nearby company towns of Ruth and Kimberly. The town’s population reoccupied abandoned residential and commercial buildings, and in 1943 and 1944 the Federal Housing Authority built barracks and houses just outside the town, filling in the road between Reipetown and Kimberly. Life in the town retained its wide-open and rough character, even though many more families lived in Reipetown and continued to do so into the 1950s and 1960s. The end of the war, however, brought the town to a close, reflecting, in part, another period of declining copper production in the Robinson district. Only a handful of residents remained in the 1960s and the last one left a decade
Source: Courtesy of the Special Collections Archive, University of Nevada-Reno Library
Figure 5.2 Photograph (c. 1910) of Reipetown, Nevada
92
Donald L.Hardesty
later, leaving the buildings to fall into decay and the townsite to become a trash dump (see Figure 5.3). In the 1990s, Magma Copper Company acquired the townsite and developed a plan of operations for a milling there. Archaeological investigations took place at the site in 1992 and 1993 as part of an environmental assessment study by Woodward-Clyde and Western Cultural Resources Management (Hardesty et al. 1994). In contrast to the company towns in the Robinson district, there is no evidence of a class structure in Reipetown, but documentary data and oral testimony show that the settlement occupied a low social status in the social structure of the larger Robinson community. Visible power and prestige differences were minimal within the Reipetown settlement. The Data brothers are exceptional, both within the settlement and within the Robinson community. No archaeological or other evidence shows significant differences in wealth among the households, nor do architectural differences among the houses reflect significant wealth differences. Interestingly enough, despite the documentary evidence and oral testimony of a population dominated by ethnic Greeks, Slavs, Italians, Mexicans and Japanese, the archaeological record of Reipetown spoke not at all about ethnicity and ethnic relationships. Certainly site formation processes obfuscated archaeological evidence of any tendency of ethnic groups to cluster in geographical neighborhoods; more remarkable, however, is the virtual absence, with the exception of a couple of pieces of Japanese ceramic tableware, of artefact markers of ethnicity in the Reipetown archaeological record. The culture of resistance in Reipetown appears to have minimized the role of ethnicity in the organization of the community. Housing, living conditions and lifestyles in Reipetown differed significantly from the company towns in the Robinson mining district. The wide variety of house forms and household organizations in Reipetown, for example, is more typical of the mid-nineteenth-century mining camps in their early stages than of twentieth-century company towns and mature mining camps in the American West. Dietary patterns also appear to be more diverse than those in company towns. Bea Mathis, a Reipetown resident in the 1900s and 1910s, remembers that Reipetown had a dairy, documented in the archaeological record, that provided fresh eggs and butter to the residents. Others remember a Greek bakery and a number of ‘ethnic shops’ or corner stores in Reipetown. Reipetown restaurants provided a wide range of ethnic foods, including Italian, Greek, Mexican, Slavic, etc. Reipetown bakeries and restaurants provided ‘fast food’ lunches to the miners; oral testimony suggests that the Data brothers began their career in this fashion. To judge from the animal bone refuse recovered from the excavation, the typical meat diet of the Reipetown residents consisted mostly of beef, pork, mutton or goat, and chicken, supplemented by smaller amounts of domestic rabbit and turkey, fish and wild deer or antelope. In addition, flotation samples from several features turned up evidence of fig (Ficus domestica), raspberry (Rubus cf. idaeus) and domestic grape (Vitis sp.).
Figure 5.3 Site map, Reipetown, showing superimposed 1907 town plan
94
Donald L.Hardesty
Power and the archaeology of violence The culture of resistance in Reipetown also included a strong dose of violence. Many years ago, philosopher Michel Foucault (1978) in his book Discipline and Punish pointed to the importance of violence as a way to acquire power and maintain social inequalities, an idea further developed by Giddens (1985) and Mann (1986) in sociology. The culture of violence permeated some mining communities as part of a strategy of resistance. Reipetown offers a classic example with its saloon culture. Documentary accounts, oral testimony and archaeological data also suggest that a saloon culture existed in Reipetown that provided a recreation outlet for visitors from the company towns as well as for the Reipetown residents. The saloon culture, which was associated with gaming, prostitution and violence, continued more or less unchanged during the history of the town until at least World War II. Shootings, murders, beatings and general mayhem appear to have been much more common in Reipetown than in the company towns. Unstable or rapidly changing power relations in the community led to a culture of violence as well. The ‘bad man from Bodie’ (California) is typical and also illustrates the myth or ideology of violence. Violence is often expressed in landscapes. Landscapes of violence, for example, are illustrated by the earth and rock fortifications built by the striking miners of Cripple Creek, Colorado, at the town of Independence (see Figure 5.1). The miners, after blowing up the train station at Independence, retreated to the hillside above the town in the face of an advancing military/police force organized by the mining companies.
CONCLUSION In conclusion, then, the mining communities of the American West are best viewed as overlapping power networks of economic, political, military and ideological relations that range between hierarchical and heterarchical structures. Individuals and groups negotiate power within the networks, using methods that vary from violence to ideology to politics as a way of controlling mining resources such as capital or ore deposits, water, or labor. The archaeology of power begins with the material expression of the economic, political, military or cultural control of mining-related resources. Hierarchical power structures such as company towns have the highest archaeological visibility, with landscapes that reflect strategies of domination and resistance, cultures of violence and so forth. In contrast, heterarchical power structures leave behind ‘palimpsests’ of materially expressed power relations that reflect negotiations of individuals and groups that were extremely ‘situational’ and changing. Mobility and informal mining systems played key roles in heterarchical structures.
Power and the mining community in the American West
95
REFERENCES Allen, J.B. (1966) The Company Town in the American West. Norman: University of Oklahoma Press. Brooks, A. (1995) Anticipating mobility: how cognitive processes influenced the historic mining landscapes in White Pine, Nevada, and the Black Hills, South Dakota. Unpublished PhD dissertation, University of Nevada, Reno. Bourdieu, P. (1984) Distinction, A Social Critique of the Judgement of Taste. Cambridge, MA: Harvard University Press. Crumley, C.L. (1994) The ecology of conquest: contrasting agropastoral and agricultural societies’ adaptation to climatic change. In C.L.Crumley (ed.), Historical Ecology: Cultural Knowledge and Changing Landscapes, pp. 183–201. Santa Fe, NM: School of American Research Press. Deagan, K.A. (1996) Environmental archaeology and historical archaeology. In E.J. Reitz, L.A.Newsom and S.J.Scudder (eds), Case Studies in Environmental Archaeology, pp. 359–376. New York: Plenum Press. Dubofsky, M. (1985) Industrialism and the American Worker, 1865–1920, 2nd edn. Arlington Heights, IL: Harlan Davidson. Dykstra, R.R. (1968) The Cattle Town. New York: Alfred A.Knopf, 2nd edn, 1972, New York: Athenaeum. Elkins, S. and McKitrick, E. (1954) A meaning for Turner’s frontier, parts I and II . Political Science Quarterly 49(3–4): 321–353, 565–602. Elliott, R. (1990) Growing Up in a Company Town. Reno: Nevada Historical Society. Emmons, D.M. (1989) The Butte Irish: Class and Ethnicity in an American Mining Town, 1875–1925. Urbana: University of Illinois Press. ——(1994) Constructed province: history and the making of the last American West. Western Historical Quarterly 25: 437–459. Ferguson, L. (1992) Uncommon Ground. Washington, DC: Smithsonian Institution Press. Foucault, M. (1978) Discipline and Punish, trans. A.Sheridan. New York: Vintage Press. Frehner, B. (1996) Ethnicity and class: the Italian charcoal burners’ war, 1875–1885 . Nevada Historical Society Quarterly 39(1): 43–62. Gardner, J.S. (ed.) (1992) The Company Town: Architecture and Society in the Early Industrial Age. New York: Oxford University Press. Giddens, A. (1985) The Nation-State and Violence. Berkeley: University of California Press. Hardesty, D.L., Mehls, S., Stoner, E. and Kimball, M. (1994) Reipetown: A Data Recovery Report for the Historic Townsite of Reipetown, White Pine County, Nevada. Report prepared for Bureau of Land Management, Ely District, Ely, Nevada. Hogan, R. (1990) Class and Community in Frontier Colorado. Lawrence: University Press of Kansas. Lipman-Blumen, J. (1994) The existential basis of power relationships: the gender role case. In H.L.Radtke and H.J.Stam (eds), Power/Gender: Social Relations in Theory and Practice, pp. 108–135. London: Sage. MacMillan, G. (1995) At The End of the Rainbow? Gold, Land and People in the Brazilian Amazon. New York: Columbia University Press. Mann, M. (1986) The Sources of Social Power, vol. 1: A History of Power from the Beginning to AD 1760. Cambridge: Cambridge University Press. Orser, C.E., Jr (1988) The archaeological analysis of plantation society: replacing status and caste with economics and power. American Antiquity 53: 735–751. Paynter, R. (1988) Steps to an archaeology of capitalism: material change and class analysis. In M.P.Leone and P.B.Potter, Jr (eds), The Recovery of Meaning: Historical Archaeology in the Eastern United States, pp. 407–433. Washington, DC: Smithsonian Institution Press.
96
Donald L.Hardesty
Paynter, R. and McGuire, R.H. (1991) The archaeology of inequality: material culture, domination, and resistance. In R.H.McGuire and R.Paynter (eds), The Archaeology of Inequality, pp. 1–27. Oxford: Blackwell. Robbins, W.G. (1994) Colony and Empire: The Capitalist Transformation of the American West. Lawrence: University Press of Kansas. Roth, L.M. (1992) Company towns in the western United States. In J.S.Garner (ed.), The Company Town: Architecture and Society in the Early Industrial Age, pp. 173–205. New York: Oxford University Press. Schwantes, C.A. (1987) The concept of the wageworkers’ frontier: a framework for future research. Western Historical Quarterly 18: 39–55. Shackel, P.A. (1993) Prospects for an archeology of the people without history. In P.A.Shackel (ed.), Interdisciplinary Investigations of Domestic Life in Government Block B: Perspectives on Harper’s Ferry’s Armory and Commercial District, Chapter 18. US Department of the Interior, National Park Service, Harper’s Ferry National Historical Park. Shanks, M. and Tilley, C. (1987) Social Theory and Archaeology. Cambridge: Cambridge University Press. Trachtenberg, A. (1982) The Incorporation of America. New York: Hill & Wang. Winthrop, K. and Chambers, A. (1988) Poor But Not So Poor: The Depression Era in Western Siskiyou County, California. Report prepared for the Klamath National Forest, Yreka, California.
Chapter 6
The mining camp as community William A.Douglass
ABSTRACT The social anthropological literature on mining is surprisingly sparse. In part this stems from the mining camp’s dubious status as a community with any degree of permanence. Drawing upon f ield and archival research, the present chapter examines social recruitment to the mining boom triggered by the discovery in 1900 of a rich silver lode in Tonopah, Nevada. It is argued that this discovery and ancillary ones represented the end of an era initiated by the discovery of gold in California in 1848. Half a century later there was an established mining community in the American West disposed to relocate whenever there was a new discovery—in effect a focus without a locus—yet stitched together with kinship, friendship, and business and ethnic ties. Anthropological methodologies—including genealogical reconstruction, network analysis, ethnic stratification and the articulation of part societies with wider social wholes— proved to be particularly effective ways to study the mining camp phenomenon. The approach holds considerable promise for a better understanding of mining history in other periods and world venues.
INTRODUCTION AND BACKGROUND At first blush, from an anthropological perspective the mining community presents us with a unique scholarly anomaly. Most of the literature regarding it has been generated by historians, both professional and amateur, whereas there are few anthropological treatments. This is so despite anthropology’s notorious fixation upon community studies, a legacy that the discipline has transcended (and then not entirely) only in recent years. Upon closer examination, however, there are logical explanations for anthropology’s relative neglect of the mining community. First, despite the plethora of anthropological subdisciplines, there has yet to emerge something called ‘industrial anthropology.’ In the modern world, mining is a form of industrial activity conducted in First World settings and by First World economic interests in Third World venues. To the extent that there is an anthropological literature of it at all, there is the tendency to examine
98
William A.Douglass
mining’s impacts (usually adverse) upon ‘native’ populations (e.g. Nash 1979; Cleary 1990; Kirsch 1996; Polier 1996). However, given that the discipline has increasingly concerned itself with contemporary issues in contemporary world settings (whether First or Third), the paucity of anthropological studies of mining camps remains remarkable. There is a second explanation, one which accesses the realm of perceived authenticity. For there is a sense in which the mining camp fails to qualify at all as a ‘community’ within anthropological conventional wisdom (i.e. the discipline’s unstated premises). The very term ‘camp’ invokes, embodies and projects a notion of impermanence. In speaking of the creation of the mining camp there is the imagery of the ‘rush’ (Holliday 1981), ‘stampede’ (Adney 1900) or ‘commotion’ (Beebe 1954). In the Stereotypic view someone (the lone prospector) discovers a gold or silver deposit and the rush is on. Those comprising it are recruited from the four corners of the globe and from every social sector. As a disparate collection of human beings they fail to pass the implicit anthropological means tests for community. That is, there is seemingly a dearth of the kinship, friendship, religious and neighborhood ties that form the substantive definition of human community. Anthony Cohen (1985:115– 118) treats community formation in oppositional terms, i.e. we versus them, as in our community versus theirs. However, again most western mining camps would fail to qualify since they are generally instances of frontier settlement. That is, they tend to be discovered at the fringes of existing human settlement and therefore at best constitute a kind of humans-versus-nature symbolic opposition. Nor is there any sort of continuity that might permit ties of intimacy and oppositions to develop. The life cycle of many a mining camp is measured in months rather than years, let alone decades or centuries. Many succumbed without even being constituted politically as a municipality. Again, the language of such demise is illustrative. It employs the vocabulary of ‘bust’ and ‘borasca.’ The latter is from the Spanish term borrasca or ‘squall,’ the sudden windstorm that carries all before it while leaving stark ruins (and ghosts—as in towns) in its wake. Mention should be made of the growing interest of historical archaeology (an emerging subdiscipline) in mining camps. Historical archaeologists combine the results of standard excavation with archival, journalistic, literary and oral historical accounts to reconstruct the lifeways of recently defunct human settlements. In this regard the western (North American) mining camp is an ideal candidate, approximating a controlled laboratory experiment, for the historical archaeologist. Many western mining camps had their own newspaper within months of their ore discovery, attracted considerable journalistic and literary attention, generated a body of official documentation (mining district legislation, mining claims, township incorporations, etc.) and experienced a boom-to-bust cycle over a short time frame, thereby creating ample ‘sites’ for archaeological excavation (Hardesty 1988).
The mining camp as community
99
The foregoing description of the intellectual climate is derived mainly from my reading of the literature regarding mining in the American West. It provided the backdrop for anthropological field research that I instituted in 1975 in the central Nevada community of Tonopah, site of one of the most important American mineral discoveries. The district, owing to its significance within the annals of mining history, was scarcely terra incognita. While it was less well documented than the California gold camps or Nevada’s earlier Comstock lode, it had been the subject of literally hundreds of popular and technical articles, book-length personal narratives and even several scholarly monographs. Their focus, however, was almost exclusively upon the discovery of the ore bodies and the halcyon period of the boom years. Tonopah, unlike so many of the western mining camps, had become neither a ghost town nor a tourist curiosity. Rather, it remained a regional servicing center of about 3,000 inhabitants. The thrust of my research, then, was to determine the causes of Tonopah’s survival as a community while analyzing the social and economic structure of a post-bust mining camp society. Obviously such a study required historical treatment of the town’s earlier boom years and I felt certain that such could be reconstructed from the plethora of existing secondary sources. However, delving into the available county, church and municipal records and reading the town’s early newspapers, I became increasingly uncomfortable with both the published accounts of life in Tonopah and the literature on mining camps in general. Glass differences within the town were fairly well documented, since many of the descriptions dwelled upon the economic successes and social activities of the entrepreneurial elite, on the one hand, and labor unrest between the mine owners and miners’ unions, on the other. Moreover, despite the wealth of information regarding the individual lives of the town’s original settlers, no one had attempted to trace their kinship patterns and other social and economic networks. Similarly, there was little sensitivity to the fact that Tonopah’s population was both stratified and segmented in terms of a rich variety of ethnic groups and self-aware subgroups defined by geographic origin in the American West prior to creation of the community. Finally, Tonopah’s impact upon surrounding areas was described almost solely in terms of its catalytic effect within the mining industry. Tonopah’s relationships with the wider society, and particularly with the non-mining sectors, on the other hand had received little attention. Obviously, then, there was considerable room for reanalysis of Tonopah’s early years from the typically anthropological perspectives of genealogical reconstruction, network theory, ethnic group stratification and the articulation of one part of a society with its wider social whole. Space does not permit any of these approaches to be developed thoroughly; nevertheless, we might consider some examples that illustrate the ways in which an anthropological perspective can trigger reexamination of many of the conclusions regarding the mining community of the American West. Here I wish to examine the nature of social recruitment to the mining camps in the
100
William A.Douglass
broadest sense. Regarding this point the literature may be characterized as California-centric. The discovery of gold there in 1848 clearly created the historical baseline for a mining craze that over the next half-century was to sweep eastward across the Sierra Nevadas, to the Rocky Mountains and Black Hills of South Dakota, as well as northward to British Columbia and the Klondike. With respect to significance, however, the California discoveries remain in the forefront both chronologically and in terms of total mineral wealth. Consequently, much of what has been written about mining camps in general is most applicable to the California experience and there are several ways in which California was unique.
THE MINING EXPERIENCE IN THE AMERICAN WEST Prior to the California discoveries, the American West was practically devoid of a non-Native American population. The architects of the first mining boom were the famed argonauts, fortune-seekers from the four corners of the globe recruited from all walks of life who flocked to the area in chaotic fashion. It is out of their experience that the image of the mining camp populace as a disparate collection of rootless, and often ruthless, individuals emerges. It was in the California camps that men (for there were few women), speaking a plethora of tongues, sojourners by predilection (for they had little commitment to a California future), engaged in a unique social experiment. They were faced with forging some sort of collective modus vivendi within a context in which individual human greed was the prime moving force. In the words of one historian, ‘The task was to protect the property and claims of the miners, to prevent theft and murder and…to prevent chaos’ (Mucibabich 1977:43). The populace of the typical California mining camp was, then, a collection of strangers rather than a community of kinsmen, friends and neighbors. Furthermore, camp life was imbued with a sense of urgency. Finding one’s personal bonanza was the object, and the rumors of new and bigger strikes elsewhere proved an irresistible lure. The average fortune-seeker never remained long in one place. Another aspect of the California mining scene deserves mention, namely the lack of mining expertise on hand even after thousands of supposed ‘miners’ had fanned out over the Sierra Nevada foothills. There is a sense in which the western mining boom could only have begun in California, since it was there that on the well-watered western slopes of the Sierras free gold in placer (or alluvial) deposits was present in sufficient quantity for the panning or the sluicing. California gold, therefore, required a minimum of mining skill and technology to find, and presented virtually no processing problems. As the exploration for minerals expanded throughout the American West, it became clear that the California deposits were an aberration. Rather, most of the region’s interior arid lands’
The mining camp as community
101
mineral wealth had to be extracted from veins, requiring elaborate tunneling and shafting. Some of the requisite technology had not even been invented as yet. It was in meeting this new challenge that mining became an elaborate enterprise, involving close cooperation between trained engineers and skilled miners. In sum, then, many of the stereotypes in the narratives regarding the miner and mining camp derive from the first years of the California boom. It was a period in which the miner was likely to be an amateur, working a small claim of his own, a stranger to the area, little committed to its future, and predisposed to pulling up stakes at the first hint of new strikes.
THE MINING CAMP Nevertheless, observers and students of the subsequent mining booms in the interior arid lands of the American West, Canada and Alaska continued to emphasize the fluid and cosmopolitan nature of the mining camp population. Trimble (1972:275), for example, notes: One fact stands out prominently, and that is that the population was very heterogeneous. In addition to an original basis of French half-breeds and of mountain-men, representatives from all parts of the United States and from every corner of the globe were to be found—Americans, Canadians, Englishmen, Germans, Frenchmen, Italians, Spanish, Chinese, Mexicans, Chilenos, Australians, Hawaiians. Similarly, in discussing the Rocky Mountain camps Smith (1974:11) notes, ‘It was a heterogeneous population composed of people from all parts of the world.’ Such statements, accurate as far as they go, are derived by examining the US census to determine the birthplace of the mining camp populace. When it is viewed in such statistical fashion, one could scarcely conceive of a more variegated and disparate collection of human beings. There is no question that each new discovery held its own particular fascination and fired the imaginations of persons in such distant places as New York and London. Each camp therefore had its collection of neophytes, as well as its contingent of drifters and desperadoes. However, what is too frequently ignored is the fact that with the passing years the mining frontier was itself aging, as were its earlier pioneers. The majority of men who first panned for gold in California subsequently abandoned mining and settled into some other pursuit. Nevertheless, a large number of the early California miners stayed with the profession and constituted the phalanx of the extension of the mining frontier throughout the American West. Again to quote Trimble (1972:277), ‘whatever elements of population prevailed in one or the other place, there was one everywhere present, everywhere respected, everywhere vital—the Californian.’
102
William A.Douglass
Nor was the Californian likely to remain a transient bachelor forever. As more and more miners established families, a new element was introduced into the camps. No longer did recruitment depend solely upon immigration. Rather, persons began to be born into the mining way of life. Though the individual camp lacked stability and longevity, there emerged in the American West a skilled and self-aware mining community, increasingly interlaced with kinship and business ties. The aging of the mining frontier also effected changes in the character of immigration into the region. Easterners and foreigners with experience in the early California camps began to send for friends and relatives. Such chain migration meant that the seemingly disparate and cosmopolitan make-up of the mining camp populace masked the fact that each of the regional and ethnic subgroups was becoming increasingly more integrated internally. Fewer and fewer persons arrived in the region without at least one friend or kinsman to rely upon. In some cases whole extended families were reunited in the mining districts of the American West.
TONOPAH, NEVADA: CASE STUDY OF A MINING CAMP Viewed in such terms, the Tonopah-Goldfield strike is particularly informative. If the western frontier mining boom began in California (1848), there is a sense in which the Tonopah discovery (1900), and its spin-offs, constitute the other bookend. The Tonopah strike triggered new exploration throughout much of central and southern Nevada. A nearby discovery in 1902 led to creation of Goldfield (Zanjani 1992), an even more impressive camp, and Rhyolite to the south. There were a number of lesser strikes as well that demonstrated little staying power. The presence of rich silver deposits in a remote corner of the Nevada desert acquires added significance when considered against the backdrop of the economic climate within the mining industry at the turn of the twentieth century. Many observers of the day were convinced that the mineral wealth of the American West had been effectively depleted. To be sure, there were pockets of activity in places like Cripple Creek, Colorado, and Butte, Montana. Delamar, Nevada, was still producing respectable amounts of silver. However, the major thrust of the search for precious metals had essentially passed out of the American West to British Columbia, the Klondike, and Sonora, Mexico. The former California gold camps, as well as those in Nevada, such as Virginia City, Austin and Eureka, were practically somnolent. While still centers of a degree of activity, these major foci of earlier boom periods were existing largely upon memories of past glory. The day-to-day reality was that the most accessible ore bodies had long since been extracted, while recovery of the remaining ores posed formidable technical problems. Furthermore, it had become difficult to finance the increasingly costly and complex operations
The mining camp as community
103
because of a history of mining scams gone awry. Possibly the most eloquent testimony of the mining depression was the State of Nevada’s perilous position. Born out of the excitement of the Comstock years and sustained by a series of smaller mining strikes during the 1870s and 1880s, by the turn of the century the state’s mining-based economy was so moribund that between 1890 and 1900 the area’s population shrank by 11 per cent, triggering a serious movement in the US Congress to strip Nevada of its statehood. However, throughout much of the American West skilled prospectors continued the search for new El Dorados. Many of these men were second-and third-generation descendants of the early California miners. One such man was Jim Butler, rancher and resident of the town of Belmont, seat of Nye County, Nevada. Butler was born in a gold camp of El Dorado County, California, and spent much of his life in several silver camps of central Nevada. In the spring of 1900 he traveled south from Belmont to the fledgling mining camp of Klondike, site of what ultimately turned out to be an insignificant silver strike. Nonetheless, the deposits were being worked largely by Butler’s Belmont friends, and he wanted to check out the rumors that the Klondike discovery was promising. While returning to Belmont, accompanied only by his burro, he decided to prospect along the way. It was on 19 May 1900 that he made his momentous find in a high canyon of the Nevada desert. The first area to feel the effects of the impending mining boom was Butler’s home town of Belmont. Since he was practically impoverished, Butler offered a percentage of his claims to his friend Tasker Oddie with the condition that Oddie pay for the assays of the ore samples (Douglass and Nylen 1992:225). Equally strapped for cash, Oddie then gave the assayer Walter Gayhart, his former business partner in Austin, a percentage of his percentage in return for doing the assays. Once it was clear that the ore values were significant, the Belmont merchant Wilse Brougher was offered a share in return for equipment and a grubstake to begin work on the claims. Oddie’s letters detail the discovery and first two years of Tonopah’s existence, providing a fascinating prospective rather than the usual retrospective account of a mining camp’s development. And so it began. Tonopah was situated about 50 miles (80 km) to the south of Belmont; to the west at a similar distance was the town of Silver Peak. To the northwest about 60 miles (100 km) distant were the towns of Candelaria, Douglass Camp and Sodaville. Both Candelaria and Sodaville were on the railroad line that served the southeastern California mining camp of Bodie. All of these towns were formerly sites of mining booms, so they housed experienced miners and prospectors, and yet their combined population was fewer than 2,000. When news of the developments at Tonopah began to spread, the first people to rush to the new camp were drawn almost exclusively from these immediate hinterland communities. The Belmont group of original locators lacked the capital to work all of their claims themselves, so they instituted a leasing system. A newcomer to the camp was assigned an area. He provided his own tools and labor and had to do an agreed-upon amount of exploratory work
104
William A.Douglass
within a specified period of time. Twenty-five per cent of his gross proceeds went to the Belmont owners. For the first year of Tonopah’s existence development of the camp proceeded very slowly and its future was in doubt. High assay values were obtained at or near the surface of many of the leases, but the leasers themselves were without the capital to work the properties properly. Shipment of the ore was an extremely costly venture, since 60 miles of trackless desert separated Tonopah from the nearest railroad. Only the highest-grade ore provided sufficient return to justify the high costs of transport in twenty-mule-team wagons to the railhead in Sodaville. However, by year’s end the camp’s reputation was growing and its population had swollen to about 1,000. The area’s future was secured when, in 1901, a group of Philadelphia investors purchased the claims of the original Belmont group and began large-scale deep-shaft mining operations. Other outside investors quickly followed suit and within two years there were half a dozen mining companies in the area employing hundreds of miners. By 1904 the town was linked to the outside by its own railroad. Meanwhile, prospectors were fanning out in every direction in hopes of finding the next Tonopah. In the winter of 1902 rich silver and gold discoveries were made 30 miles (50 km) to the south and the camp of Goldfield was born. Within two years it had eclipsed Tonopah in size and importance. Both camps had peaked and were in decline by the time of the 1910 census, so it is impossible to provide a wholly accurate figure for their populations at the height of the boom. Still, it is likely that Tonopah at its zenith had about 8,000 inhabitants, while Goldfield may have had as many as 15,000. Furthermore, the Tonopah-Goldfield discoveries triggered activity throughout Nevada and in surrounding states that, over the next two decades, led to the creation of scores of new (if lesser) mining camps. On balance, then, Tonopah-Goldfield was one of the three or four most important mining booms in the history of the American West. While it may not be denied that outside capital as well as people from throughout the USA and abroad flocked to the Tonopah-Goldfield area, it is equally true that its original social and economic infrastructure was created by a core population of ‘old-timer’ Nevada and California miners and mining-camp merchants. The first year the majority of the camp’s inhabitants were drawn from the immediate Belmont-Silver Peak-Candelaria-Sodaville hinterland. Within a short time Tonopah co-opted almost all of this population. Given the scarcity of construction materials these communities were physically cannibalized, as buildings were dismantled, transported by wagon and then reassembled in Tonopah. By the end of the first year there was scarcely a business in the new camp that was not owned by a former resident of one of the aforementioned areas. In some cases existing businesses in Belmont and Candelaria opened a Tonopah branch for a few months, before succumbing totally to the attraction of the new camp.
The mining camp as community
105
This core population, then, was interlaced with pre-existing business ties and, in many cases, with kinship and affinal ties as well. Furthermore, by the second year of Tonopah’s existence the Belmont, Silver Peak, Candelaria and Sodaville factions were becoming increasingly intertwined in the new context. As an example, the discoverers of Goldfield, Harry Stimler and William Marsh, originally of Belmont, were affinal kinsmen. They were grubstaked by one former Belmont resident and two former residents of Douglass Camp (Douglass and Douglass 1996). As Tonopah’s fame spread, there was a noticeable ripple effect. By the end of the second year the town’s growth came at the expense of a somewhat further distant concentric ring of moribund mining camps. Nevada supplied Tonopah with a clearly discernible contingent of residents from Austin, Eureka and Tuscarora to the north, and Caliente and Delamar to the east. Randsburg and Bodie, both in California to the west, lost many of their inhabitants to Tonopah. There was also a growing contingent of former Virginia City residents, as well as a group from the Grass Valley-Nevada City area in the California motherlode country. Finally, when Tonopah really hit its full stride, about 1904, population growth in the region was being fueled by the arrival of hundreds of miners from Cripple Creek (Colorado), Rossland (British Columbia) and Nome (Alaska). They arrived with their own collective histories elsewhere, but many had previously resided in one or another Nevada or California mining camp and therefore had prior acquaintances in Tonopah. Nor were the newcomers drawn exclusively from mining backgrounds. By 1900 there was a well-established agricultural sector within the economy of eastern California and Nevada. It included settled cattle and sheep ranchers with significant acreage and labor force, family farms in the hands of homesteaders who had availed themselves of federal homestead laws designed to reward Civil War veterans with land while extending the nation’s agricultural frontier, and itinerant sheep outfits (mainly Basque-owned) that ‘cruised’ the public lands throughout the year in search of pasturage. Indeed, prior to Jim Butler’s centennial date with fate in May 1900, a Basque sheepman, Isadore Sara, took ore samples at the Tonopah site. He abandoned them along the way when they wore through the saddlebags of his pack animal (Keeler 1913: 965–966). The interpenetration of the mining and agricultural frontiers is considerably more complex than is implied by the standard statement that the camps stimulated agricultural development by providing a local market for agricultural products. The discovery of mineral wealth redefined the agricultural districts within a several hundred mile radius of Tonopah as buckaroos found more lucrative employment in the camps and homesteaders left marginal family farms to work for a steady miner’s wage. Conversely, the established ranches expanded by absorbing the abandoned smallholdings while, in many cases, coming under the control of successful mining entrepreneurs anxious to validate their new fortunes with land ownership and the more socially prestigious ranching lifestyle.
106
William A.Douglass
Tonopah, like practically every mining camp, had a great deal of ethnic diversity as well. Ethnic groups within the camp with sufficient self-awareness to form voluntary associations or stage ethnic events included the Irish, Italians, Slavs, Germans, Finns, Chinese, Mexicans, blacks and Native Americans. However, this ethnic diversity was not the result of an argonaut experience comparable to that of the earlier California camps. To look at but a few of the groups, Tonopah’s Italian contingent came mainly from Delamar, a community in which a large number of former residents of Eureka, with an established Italian colony of the 1870s, had resettled after Eureka’s mining boom waned. Many of the Italians who came into Tonopah had therefore been born in a Nevada mining camp. Of the Old World-born Italians who came directly to Tonopah, not surprisingly, many had kinsmen in this earlier Nevada Italian colony. The Chinese in Tonopah came primarily from Candelaria and Bodie, two communities in which several hundred Asians established residence after constructing the Carson to Bodie railroad in the early 1880s. The Slavs, like the Italians, were a well-defined ethnic element in Delamar’s population prior to coming to Tonopah. Candelaria and Douglass Camp were dominated by the Irish, whereas Austin and Belmont had significant German contingents.
CONCLUSION To conclude, I would note that even this very limited description of social recruitment to the Tonopah-Goldfield mining boom suggests the power of anthropological analysis for understanding the mining camps of the American West. It also underscores the need for a historical perspective. The type of mining community that I have been discussing is no longer being replicated within the region. Today’s uranium, petroleum or ‘invisible’ gold boomtown, for instance, is of a different nature, if for no other reason than that its wider social and economic setting is so different from the frontier conditions that obtained in the American West during the latter half of the nineteenth century. Nevertheless, regarding any of the earlier major mining strikes there is a wealth of historical documentation available that easily lends itself to anthropological network analysis. Once this is effected, the image of the mining community begins to shift. The individual mining camp, with its short-lived boom-to-bust cycle, becomes almost epiphenomenal when we consider that, by the latter decades of the nineteenth century, we can discern a mining community here in the American West that was unique in one important respect, namely it was a community without a locus. Rather it was diffused over a vast geographical region and crystallized in the form of a camp for brief periods of time around this or that discovery. But the community as a whole persisted beyond the life span of any of its short-lived manifestations. It evolved its own traditions and mores which transcended the experience of any single camp.
The mining camp as community
107
Finally, whether undertaking holistic studies of individual mining camps or framing network analyses within the larger context of the mining community of the American West, anthropologists are well situated to contribute new insights to our understanding of one of the major and most distinctive lifestyles on the American and other world frontiers. I would situate the foregoing analysis within the annals of world mining history in two ways. First, it is equally possible to carry out similar anthropological analyses of mining booms within colonial contexts such as the Spanish Empire (e.g. Bakewell 1971, 1984). If anything, recruitment to them was often realized in ways that configured an even more narrowly circumscribed community of neighbors, friends, kinsfolk and business associates than the one that I have described. Second, the developments in the American West were paralleled both temporally and structurally by the late nineteenth- and early twentieth-century expansions of the mining frontier throughout Canada, Australia, New Zealand, South Africa, Siberia and several Latin American countries. This ‘modern’ era of global mining is characterized by transnational capitalization and expertise producing mineral wealth for the world market. In this regard, mining provides one of the earliest and most elaborated examples of a globalized industrial activity.
REFERENCES Adney, T. (1900) The Klondike Stampede of 1897–1898. New York and London: Harper & Brothers. Bakewell, P. (1971) Silver Mining and Society in Colonial Mexico, Zacatecas 1546–1700. Cambridge: Cambridge University Press. ——(1984) Miners of the Red Mountain: Indian Labor in Postosi, 1545–1650. Albuquerque: University of New Mexico Press. Beebe, L. (1954) Comstock Commotion. Palo Alto: Stanford University Press. Cleary, D. (1990) Anatomy of the Amazon Gold Rush. Iowa City: University of Iowa Press. Cohen, A. (1985) The Symbolic Construction of Community. Chichester: Ellis Horwood; London: Tavistock. Douglass, J. and Douglass, W.A. (1996) Tap Dancing on Ice: The Life and Times of a Nevada Gaming Pioneer. Reno: Oral History Program, University of Nevada, Reno. Douglass, W.A. and Nylen, R.P. (eds) (1992) Letters from the Nevada Frontier: Correspondence of Tasker L.Oddie, 1898–1902. Norman and London: University of Oklahoma Press. Hardesty, D.L. (1988) The Archaeology of Mining and Miners: A View from the Silver State. Society for Historical Archaeology, Special Publications Series 6. Pleasant Hill, CA: SHA. Holliday, J.S. (1981) The World Rushed In: The California Gold Rush Experience. New York: Simon & Schuster. Keeler, E.P. (1913) Nye County. In S.P.Davis (ed.), The History of Nevada, pp. 965–966. Reno and Los Angeles: Elms. Kirsch, S. (1996) Anthropologists and global alliances. Anthropology Today 12(4): 14–16. Mucibabich, D. (1977) Life in Western Mining Camps: Social and Legal Aspects 1848–1872. Hicksville, NY: Exposition Press. Nash, J. (1979) We Eat the Mines and the Mines Eat Us: Dependency and Exploitation in Bolivian Tin Mines. New York: Columbia University Press.
108
William A.Douglass
Polier, N. (1996) Of mines and min: modernity and its malcontents in Papua New Guinea. Ethnology 35: 1–16. Smith, D.A. (1974) Rocky Mountain Mining Camps: The Urban Frontier. Lincoln: University of Nebraska Press. Trimble, W.J. (1972) The Mining Advance in the Inland Empire. New York and London: Johnson Reprint Corporation. Zanjani, S. (1992) Goldfield: The Last Gold Rush on the Western Frontier. Athens, OH: Swallow Press.
Chapter 7
Mining, colonialism and culture contact European miners and the indigenous population in the sixteenth-century Arctic Robert M.Ehrenreich
ABSTRACT Mining expeditions usually go to regions far beyond the control of the societies that sent them. Rarely are the regions so remote that there are no indigenous populations with which to interact, however. Mining expeditions can thus mark the first encounter between cultures and, based on the perceived threat, develop many of the militaristic and societal characteristics of colonial outposts. This study examines the cultural dynamics of the sixteenth-century English mining expeditions to Baffin Island, Northwest Territories, Canada, and the effects of the interaction with the indigenous population.
HISTORICAL BACKGROUND Martin Frobisher was an Elizabethan explorer who traveled to Baffin Island in 1576 in search of the Northwest Passage. English merchants were at a serious disadvantage during this period. Spain was receiving gold and other luxury goods from its colonies in the New World. Portugal controlled the colonies and shipping lanes along the west and east coasts of Africa and thus most of the trade between western Europe and China. In the other direction, the journey around Cape Horn was long and dangerous. Thus, England needed a new route to China. The primary potential short-cut pursued by the English was the Northwest Passage, which supposedly circumvented the New World via the north. Frobisher initially sought funding for the expedition from the merchant class of Britain since they had the most to gain from its discovery, but since they ‘never regard virtue without sure, certain, and present gains, he repaired to the Court (from whence, as from the fountain of our common wealth, all good causes have their chief increase and maintenance)’ (Best 1578/1:47). The Earl of Warwick primarily funded the initial expedition, which consisted of 2 ships (Gabriel and Michael), a pinnace (ship’s boat) and 36 men. Frobisher set sail on 15 June 1576. The ships traveled up the east coast of Britain and then to Iceland and Greenland. Disaster struck approximately halfway through the voyage. The pinnace sank during a storm and the Michael abandoned the expedition.
110
Robert M.Ehrenreich
Frobisher continued the voyage in the Gabriel, however, eventually arriving at what is now called Frobisher Bay, but which he christened Frobisher Strait, thinking that he had discovered the new route to China. Although Frobisher could not prove that he had found the Northwest Passage, he returned to England with a number of ‘tokens of possession’ from the New World. According to the journals of George Best,1 these tokens included a piece of black stone, much like to a sea-coal in color, which by the weight seemed to be some kind of metal or mineral. This was a thing of no account, in the judgment of the Captain at the first sight. And yet for novelty it was kept, in respect of the place from whence it came. (Best 1578/1:51) Record has it that the spouse of an adventurer who received a piece of the black rock as recompense for his service justifiably threw it into the fire over the pitifulness of the prize. When the fire was later cleaned, however, the rock had supposedly changed color and sparkled ‘with a bright marquisette of gold’ (Best 1578/1:51). Michael Lok, who was Frobisher’s principal commercial backer and organized the private financing for the expeditions, had the rock assayed five times in England (Fitzhugh et al. 1993). The first four assays were negative. The last assay supposedly determined it to be a high-grade gold ore, containing almost 4 ounces of gold per hundredweight (approximately 2,200 parts per million; Hogarth 1993). When the assayer, John Baptista Agnello, was asked how he was able to obtain a positive reading when so many could not, he supposedly replied, ‘bisogna sapere adulare la natura’ (‘it is necessary to know how to coax nature’ —Stefansson and McCaskill 1938/2:83). On the basis of this assay, Frobisher and Lok formed the Cathay Company and secured investments from the merchant and royal classes of Britain, including Queen Elizabeth. Mining expeditions to Baffin Island were launched in both 1577 and 1578. The 1577 expedition consisted of 3 ships and 145 men. The crew contained 8 miners and 3 assayers (Hogarth 1993). Two blacksmiths were also present to maintain the miners’ equipment. All of the roughly 158 tons of ore collected during this voyage came from an island that the Inuit have called Kodlunarn since Frobisher established his base camp there over four hundred years ago, qallanat being the Inuktitut word for Europeans (Rowley 1993). The ore was collected by simple open-cast mining. The efforts of the miners were aided by some of the crew, as Best wrote in his journal: ‘only with five poor miners and the help of a few gentlemen [officers] and soldiers, brought aboard almost 200 tons of gold ore in the space of twenty days’ (Best 1578/1:75). The ore collected during the second voyage was to be officially assayed immediately on return to England, but the testing was strangely delayed (Fitzhugh et al. 1993). Thus, it was not discovered to be barren rock until after the third expedition had departed.
European miners and the indigenous population in the Arctic
111
The third and final expedition consisted of a staggering 15 ships and over 400 men. The crew contained 147 miners, 4 blacksmiths and 5 assayers (Hogarth 1993). The ore from the third expedition came from seven sites in the region. Mining camps were established at each of the different locations, but little information exists about their structure. For example, Best makes only one statement concerning the camps: The first of August every Captain, by order, from the General and his council, was commanded to bring ashore unto the Countess Island, all such gentlemen, soldiers, and miners as were under their charge, with such provision as they had of victuals, tents, and things necessary for the speedy getting together of Mine…the Muster of the men being taken and the victuals with all other things viewed and considered, every man was set to his charge, as his place and office required. The Miners were appointed where to work and the Mariners discharged their ships. (Best 1578/1:103) Frobisher returned to England laden with what he thought was 1,136 tons of gold ore but what further assaying showed to be worthless rock as well. The Cathay Company went bankrupt and Lok was discredited and ruined. Although financially devastated, Frobisher soon regained his stature in Queen Elizabeth’s court and was knighted in 1588 for his part in the fleet action against the Spanish Armada. The cause of the false assays is still a mystery. Recent analyses of the black rock from Kodlunarn Island reveal its gold content to be 130 ± 40 parts per billion and its silver content to be 11 ± 3 parts per billion, which is four orders of magnitude lower than Lok’s gold assays and five orders of magnitude lower than his silver assays (Hogarth 1993). These analyses raise the question of how the assayers were actually testing the ore during the expeditions. As stated above, the journals confirm that assayers were present and that they were performing tests. For instance, Best states for the second voyage: our General rowed thither with…his goldfiners to search for ore…and the first small island which we landed upon, here all the sands and cliffs did so glitter and had so bright a marquisette that it seemed all to be gold; but upon trial made, it proved no better than black lead and verified the proverb: all is not gold that shineth. (Best 1578/1:61) The ore was probably assayed using the cupellation method, which was the main pyrotechnological assaying technique of the period and relied on the use of lead to separate silver from ore (Biringuccio 1556). It has been postulated that the inaccurate assays were caused by the unwitting use of lead that already contained silver (Hogarth 1993). There are three problems with this
112
Robert M.Ehrenreich
hypothesis, however. First, it would explain the false silver readings but not the false gold readings. Second, if the lead were contaminated, the assayers would have obtained positive results on the first use of the lead but negative results thereafter. Third, contaminated lead might have accounted for incorrect assays during one voyage, but it seems inconceivable that six assayers would not have suspected that something was amiss over the course of two years and two additional voyages. Were the assayers really that incompetent? Could they have been using a less sophisticated method that relied on an indirect indication of the presence of silver and gold as opposed to determining the actual contents? Could the entire endeavor have been an innocent case of wishful thinking or a ploy to continue the search for the Northwest Passage or to establish a British colony in the New World? The chief assayer for the second voyage, an accomplished German metallurgist named Jonas Shutz, quit before the departure of the third voyage because of difficulties with Frobisher. He was replaced by his assistant, a London goldsmith named Robert Denham (Hogarth 1993). There is even speculation that a Portuguese informant was present on the third voyage. Thus, the history and development of the Frobisher expeditions may have been influenced by a range of political, economic and societal factors that are no longer discernible.
CULTURAL CONTACT AND COLONIALISM The Frobisher expeditions marked the first contact between Inuit and European on Baffin Island, with definite effects. The logs and memoirs of the European explorers reveal a fascination with and fear of the indigenous inhabitants as well as an acceleration in the adoption of many of the militaristic and societal characteristics of a colonial outpost based on the interaction. Inuit culture perpetuated folklore about the contact and place names like Kodlunarn Island for over four hundred years. The remainder of this chapter explores the dynamics and effects of the interaction between the two cultures, based on the explorers’ journals and Inuit folklore. Mining and colonialism Since Frobisher’s first voyage to the Arctic was to search for the Northwest Passage, the expedition was initially fairly open in its encounters with the Inuit. For example, Best (1578, 1:49) states that Frobisher had sundry conferences with them and they came aboard his ship and brought him salmon and raw flesh and fish…they exchanged coats of seal and bearskins and such like with our men and received bells, looking glasses, and other toys in recompense thereof.
European miners and the indigenous population in the Arctic
113
The situation changed dramatically, however, when ‘after great courtesy and many meetings, our mariners, contrary to their Captain’s direction, began… to trust them and five of our men going ashore, were by them intercepted …and were never since heard of to this day again’ (Best 1578/1:49). The situation was complicated by the fact that the men had taken the only shore-going boat that remained after the sinking of the pinnace. Thus Frobisher could not even mount a landing party and search for his lost men. The expedition became increasingly militaristic thereafter, with subsequent contact being limited to ploys for luring the Inuit onto the European vessels and then using them as hostages to negotiate an exchange. Frobisher captured one Inuk toward the conclusion of the first expedition: to make them more greedy of the matter, he rang a louder bell, so that in the end one of them came near the ship side to receive the bell, which when he thought to take at the Captain’s hand, he was thereby taken himself. For the Captain being readily provided, let the bell fall and caught the man fast and plucked him with main force, boat [kayak] and all, into his bark, out of the sea; whereupon when he found himself in captivity, for very choler and disdain he bit his tongue in twain within his mouth. Notwithstanding, he died not thereof, but lived until he came in England, and then he died of cold which he had taken at sea. (Best 1578/1:50) The militaristic nature of the expeditions further increased during the second voyage. As stated above, all of the ore mined during this voyage came from Kodlunarn Island, on which Frobisher also established his base of operations. Five turf and stone structures were built on the island, four of which were related to mining: a smithy, an assayer’s office, a structure that probably served as a store for the blacksmiths’ and assayers’ charcoal brought from England, and a long-house that probably served as a storage facility for miners’ tools (Fitzhugh 1993). The fifth structure, known as Best’s Bulwark, was purely defensive. George Best supervised the construction of a modest fort of rock and sod walls on a promontory at the southeast end of the island. The small size of the fort, even prior to four hundred years of erosion, would suggest that it was not meant to house a mining camp or garrison but was probably intended to serve as a refuge in the event of an attack (Fitzhugh 1993). Two major skirmishes occurred during the second voyage, with five Inuit men killed during one of them. A man, woman and child were also abducted and Frobisher’s men would have seized more, if possible: [Frobisher] presently upon a watchword given, with his Master suddenly laid hold upon the two savages. But the ground underfoot being slippery …their handfast failed and their prey, escaping, ran away and lightly recovered their bow and arrows. And being only two savages in sight, they
114
Robert M.Ehrenreich so fiercely, desperately and with such fury assaulted and pursued our General and his Master, being altogether unarmed and not mistrusting their subtlety, that they chased them to their boats and hurt the General in the buttock with an arrow. (Best 1578/1:59)
It was also during this voyage that the journals began to refer to the Inuit as being devious, murderous and probably cannibalistic. For example, Best states that during an interrogation of an Inuk he gave us plainly to understand by signs that he had knowledge of the taking of our five men the last year…and when we made him signs that they were slain and eaten, he earnestly denied and made signs to the contrary. (Best 1578/1:65–66) Yet on finding the tunic of one of the lost men containing what appeared to be arrow holes, Best states again, ‘considering their ravenous and bloody disposition in eating any kind of raw flesh or carrion, howsoever stinking, it is to be thought that they slain [sic] and devoured our men’ (Best 1578/1:68–69). Paradoxically, the Europeans continued to present gifts to the Inuit in order to try to coerce them to return the men. The gifts were left in the deserted settlements after the Inuit fled, however. As Best (1578, 1:66–67) states, Captain Yorke…discovered certain tents of the country people…but found the people departed, as it should seem for fear of their coming… and so without taking anything away in their tents, leaving there also looking glasses, points, and other of our toys (the better to allure them by such friendly means), departed aboard his bark. The militaristic and colonial aspects culminated during the final voyage. Prior to departure, Frobisher established a set of strict guidelines for all ships (Best 1578/1:103–104) that foreshadowed the British Navy’s Articles of War by roughly seventh-five years. All contact with the local population was strictly forbidden. All captains, mariners and miners were restricted from digging, transporting or loading any ore without Frobisher’s permission and any mariner or miner caught smuggling ore was to be fined three times its supposed value. There were also rules against quarreling, drawing weapons, swearing, brawling or cursing, upon pain of imprisonment. An outpost of 30 miners, 40 mariners, 30 soldiers and 3 ships was also to be established at Kodlunarn Island in order for mining to continue through the winter (Best 1578/1:81). Luckily for the men to be left behind, however, the ship carrying the material for their ‘strong fort or house of timber, artificially framed and cunningly devised by a notable learned man here at home’ (Best 1578/1:81) sank with no loss of life during the voyage over (Best 1578/1:89). Peas, corn and
European miners and the indigenous population in the Arctic
115
grains were sown and a small stone and mortar house was constructed on Kodlunarn Island, however, to determine whether an outpost could be established in the future. No captives were taken during the third voyage, although attempts were made. As Best states, ‘the people are now…so wary and so circumspect, by reason of their former losses, that by no means we can apprehend any of them, although we attempted often in this last voyage’ (Best 1578/1:129). However, Best devoted significant portions of his journal to detailed descriptions of the local population. He obviously spent a great deal of time with the Inuit captives and included what he learned in a special section of his journal entitled A General and Brief Description of the Country and Condition of the People Which are Found in Meta Incognita. Some of his comments are surprisingly perceptive: These people are in nature very subtle and sharp witted, ready to conceive our meaning by signs and to make answer, well to be understood again …they will teach us the names of each thing in their language which we desire to learn and are apt to learn anything of us. They delight in music above measure…. They are exceedingly friendly and kind-hearted, one to the other and mourn greatly at the loss or harm of their fellows and express their grief of mind, when they part one from another, with a mournful song and dirges. They are very shamefaced in betraying the secrets of nature and very chaste in…manner of their living: for when the man which we brought from thence into England should put off his coat or uncover his whole body for change, he would not suffer the woman to be present but put her fourth [sic] of his cabin. (Best 1578/1:124–125) Thus, the societal organization of the expeditions went through a metamorphosis over the course of the three voyages. The changes in the objectives of the voyages probably accounted for some of the adoption of militaristic and societal characteristics. As opposed to pure exploration, the Frobisher expeditions increasingly became full-fledged mining operations that required systems for protecting the miners, efficiently collecting the gold ore and ensuring its safe passage back to England. The effect of the interaction with the indigenous community should not be discounted, however. The perceived threat engendered during the first voyage and the dynamics of the subsequent contact accelerated the development of military and colonial aspects within the expeditions.
IMPACT ON THE INDIGENOUS POPULATION The Inuit folklore is unfortunately not as extensive as the journals. Most of the tales were also collected by Charles Hall in the nineteenth century and reveal
116
Robert M.Ehrenreich
many of his biases. Thus the actual extent of the effect of the appearance of the European explorers is difficult to discern. Hall traveled to the Arctic on the whaling ship George Henry in 1860 to search for the ill-fated Franklin expedition, which disappeared in 1848. He firmly believed that members of the crew were still alive and living with the Inuit. He had to abandon his quest for the Franklin expedition, however, when a storm destroyed the small exploration vessel that he was to use to undertake the search. Hall was thus forced to winter on Baffin Island with the whaling expedition. He used the time to augment his understanding of Inuit culture by interviewing local elders about previous explorations to the region. Although he was doubtful at first, the level of detail convinced him that some of the tales he was hearing concerned the Frobisher voyages. For example, Hall heard tales about the arrival of qallunats over three successive summers in an increasing number of ships; the killing of five Inuits in a skirmish with the foreigners; and the capture of four Inuits by the explorers. The method by which the tales were passed down the generations was eloquently described by one Inuit as When our baby boy gets old enough, we tell him all about you and about all those qallunats who brought brick, iron, and coal to where you have been, and of the qallunats who built a ship on Kodlunarn Island; when boy gets to be an old Inuit he tell it to other Inuits, and so all Inuits will know what we now know. (Hall 1865:445) Although the tales were corroborated by Western accounts, Hall’s limited interest in some of them restricted the data accumulated and, thus, our understanding of the actual extent of the effects of the contact on the Inuit. Since Hall’s original objective was to show that the lost Franklin explorers could have survived by living among the native population and that their fate could still be determined sixteen years after their disappearance, he devoted most of his energy to recording the Inuit folklore about the five lost Elizabethan sailors. Inuit folklore states that the five mariners wintered on Kodlunarn Island and then three of the men built a ship in one of the mining trenches on Kodlunarn the following spring while the other two watched, per one Inuit, ‘all same as captains’ (Hall 1865:473). The Inuit supposedly did not assist with the building of the ship but did help put it to sea. The masts were erected at a cliff near Kodlunarn Island, which is still called by the Inuktitut term for a place ‘to set up a mast’ (Nepouetiesupbing; Rowley 1993:37). Before departing, the Inuit sang a song wishing the qallunats quick passage and much joy. Unfortunately, the five men departed too early in the season and were forced back by storms and frozen hands. Lore states that the Inuit built igloos and brought them food, but they all perished. The significance of the voyages to the indigenous population can be glimpsed in the history of a ‘heavy stone’ left behind by the explorers (Ehrenreich and
European miners and the indigenous population in the Arctic
117
Wayman 1993). The Inuit spoke of a very large, smooth object that was used as a test of strength by younger Inuit men and as a seat by visiting or important Inuit. Hall (1865) concluded that the object was an anvil. It transpired that the Inuit tumbled the anvil into the sea in the early part of the nineteenth century, believing that it was adversely affecting hunting. The significance acquired by the anvil within Inuit culture is displayed by the fact that it was never exploited for its metal content, that it remained within Inuit culture for roughly two hundred and fifty years as an integral part of Inuit folklore and that it was ultimately rolled into the sea because of a belief that it was adversely affecting hunting. The Inuit possessed a technology that could have allowed the removal of metal flakes from the anvil and thus its exploitation as a resource (Ehrenreich and Wayman 1993). Although not smelters of metal, the Arctic people coldworked iron as early as 800 BC and had a well-established exchange system for metal (McCartney 1992). One of the main sources of iron was the meteorite fall of the Cape York region of northwest Greenland, which provided at least 58 tons of metal. Meteoritic iron artifacts have been found in archaeological contexts all along the eastern Arctic from Greenland to Hudson Bay. The metal was exploited by hammering small flakes of iron into the shape of blades while still on the meteorites and then detaching these flakes by further laboriously pounding the extraterrestrial metal with hammerstones (Wayman 1988). Although the anvil was smoother and softer than a meteorite, it is still possible that the same technique could have been used to exploit the metal in the anvil. Thus, the cultural interaction with the European mining community must have been sufficiently notable for the iron-work to have acquired value within Inuit society.
CONCLUSIONS A dramatic change occurred in the nature of the communities that comprised the three Frobisher expeditions. The initial voyage was the least militaristic and the most open to contact with the indigenous population, as clearly shown by two facts: (a) that the Michael could desert the voyage yet still participate in the second and third expeditions; and (b) that the five missing sailors could have even considered disregarding Frobisher’s orders and engaging in personal interaction with the local inhabitants. The voyage was probably less militaristic because it was essentially a trade mission. Frobisher’s objectives were to discover a new passageway to Cathay and establish new trade agreements. Thus, the expedition had to be open to individual opportunism and trade with local populations. The objectives and nature of the second and third expeditions were remarkably different, however. Although the search for the Northwest Passage was still within Frobisher’s purview, the inaccurate assaying of the black rock transformed Frobisher’s primary concern to ensuring the safe passage of the
118
Robert M.Ehrenreich
ships and men to Baffin Island, the collection of as much ore as possible as quickly as possible and the safe return of the ore back to Britain. As opposed to the expeditions’ being open to meeting with the indigenous population, the disappearance of the five sailors caused the final two expeditions to be organized along more militaristic lines, with the establishment of very clear regulations for ship’s conduct and a tenet that the local population was exceedingly dangerous and needed to be either subdued or expunged. This chapter has attempted to show that studies of mining societies cannot neglect the influence of the indigenous population. Even if contact is limited to sporadic gift-giving, hostage-taking or skirmishes, both societies will be affected. For the Frobisher voyages, the perceived threat engendered by the interaction during the first voyage caused them to accelerate their adoption of the militaristic and societal characteristics of colonial outposts. For the Inuit, the fact that the place names and folklore persisted for over four hundred years and that the anvil acquired value within the society shows that the interaction was sufficiently significant to be worth commemorating. Therefore, mining communities must be studied within the context of their environment, which includes the indigenous population.
ACKNOWLEDGMENTS I would like to thank Bernard Knapp, Vincent C.Pigott and David Killick for inviting me to the conference ‘Social Approaches to an Industrial Past: The Archaeology and Anthropology of Mining.’ I would also like to express my gratitude to the RTZ-CRA Group and Homestake Mining Company for the travel funds that allowed me to participate. Finally, I would like to thank Ray Beebe and Carmel McGill for helping me locate and secure this funding.
NOTE 1
Although biased and inaccurate in places, the journals of George Best are the only comprehensive first-hand publications of all three voyages and thus are regarded as the official record of the Frobisher expeditions. The pay-lists record him as a lieutenant on the second voyage and the captain of the Anne Francis on the third. There is no official record of his accompanying the first voyage, but his memoirs imply that he did.
REFERENCES Best, G. (1938 [1578]) A true discourse of the late voyages of discoverie, for the finding of a passage to Cathaya, by the northweast, under the conduct of Martin Frobisher generall: devided into three bookes. In B.Stefansson and E.McCaskill (eds), The Three Voyages of Martin Frobisher. vol. 1. London: Argonaut Press.
European miners and the indigenous population in the Arctic
119
Biringuccio, V. (1942 [1556]) The Pirotechnia, trans. by C.S.Smith and M.Teach. New York: American Institute of Mining and Metallurgical Engineers. Ehrenreich, R.M. and Wayman, M.L. (1993) Acculturation in the Arctic: the Inuit meet Martin Frobisher. In S.Alsford (ed.), The Meta Incognita Project: Contributions to Field Studies. Mercury Series no. 6: 139–147. Ottawa: Canadian Museum of Civilization. Fitzhugh, W.W. (1993) Archeology of Kodlunarn Island. In W.W.Fitzhugh and J.S. Olin (eds), Archeology of the Frobisher Voyages, pp. 59–97. Washington, DC: Smithsonian Institution Press. Fitzhugh, W.W., Laeyendecker, D. and Hogarth, D.D. (1993) Exploration history of Frobisher Bay. In W.W.Fitzhugh and J.S.Olin (eds), Archeology of the Frobisher Voyages, pp. 11–25. Washington, DC: Smithsonian Institution Press. Hall, C.F. (1865) Arctic Researches and Life among the Esquimaux. New York: Harper & Brothers. Hogarth, D.D. (1993) Mining and metallurgy of the Frobisher ores. In W.W.Fitzhugh and J.S.Olin (eds), Archeology of the Frobisher Voyages, pp. 137–145. Washington, DC: Smithsonian Institution Press. McCartney, A.P. (1992) Canadian Arctic trade metal: reflections of prehistoric to historic social networks. In R.M.Ehrenreich (ed.), Metals in Society: Theory beyond Analysis, pp. 26–43. Philadelphia: Museum Applied Science Center for Archaeology (MASCA). Rowley, S. (1993) Frobisher Miksanut: Inuit accounts of the Frobisher voyages. In W. W.Fitzhugh and J.S.Olin (eds), Archeology of the Frobisher Voyages, pp. 27–40. Washington, DC: Smithsonian Institution Press. Stefansson, B. and McCaskill, E. (1938) The Three Voyages of Martin Frobisher, vols 1 and 2. London: Argonaut Press. Wayman, M.L. (1988) The early use of iron in Arctic North America. Journal of Metals 40(9): 44–45.
Part II
Anthropology and social history
Chapter 8
‘Find the ekijunjumira’ Iron mine discovery, ownership and power among the Toro of Uganda S.Terry Childs
ABSTRACT In many instances of precolonial African iron-working, it is difficult to examine the people and activities associated with mining in isolation from smelting and smithing. Some of the same people are involved in all three processes as a result of social, politicoeconomic and ideological factors. This chapter focuses on the significance of discovering a new source of iron ore among the Toro of Uganda and its impact on subsequent smelting and smithing. Recent ethnographic interviews with elderly iron-workers reveal the social and politico-economic consequences of a discovery. A person was lucky when he found the glittery stones dug up by an ekijunjumira beetle because the discovery could bring him wealth. It also initiated new investments and social responsibilities for the discoverer and new owner, including obligations to his clan spirit, recruiting and employing iron-working specialists, protecting his source of wealth, and fulfilling socioeconomic demands on him as a man of power. The importance of mining to the Toro economy and society was also evident when it was banned by colonialists in the early twentieth century. Ethnographies of mining can provide archaeologists with important tools to interpret past metal-producing societies. Unfortunately, we learn that informative sources of insight, such as songs, will rarely be available.
INTRODUCTION Precolonial miners of the Toro kingdom in western Uganda left behind numerous open mine shafts that are hidden today under dense woods, tangled shrubs or high grasses. This legacy of pock-marked hillsides now testifies to the regular exploitation of iron ore up to the late 1920s and early 1930s, when Toro miners were ordered to stop working by their king (omukama) and the British. Who were the people who created these shafts? What were their relationships to the small communities of people who lived nearby? What other material and non-material record have they left behind, besides deep holes, that document their craft and the social and physical context in which it occurred?
124
S.Terry Childs
In our days of industrialization, permanent mining communities are commonly characterized as isolated and self-contained (e.g. Godoy 1985), although their constant need for supplies, information and transportation requires outside interaction (Hardesty 1988). These communities are seen to contain a large and highly specialized workforce of men who work full-time and for wages on ore extraction. Various processes to win metal from the ore are usually conducted elsewhere by other specialists. Women are often present in the communities but are rarely involved in mining or seen as instrumental to its success. Finally, the miners are recruited labor paid by investors far removed from their immediate communities (see also the chapters by Sheridan and Hardesty, this volume, Chapters 11 and 5 respectively). Miners from precolonial and early colonial Toro communities in western Uganda, like most pre-nineteenth-century miners (Ehrenreich 1996), do not fit this characterization. They were not isolated but participated fully in the sociopolitical and economic spheres of their society. The miners did not work fulltime but seasonally. They also participated in other production activities, such as farming and cattle-raising, as well as other metallurgical operations subsequent to mining, such as smelting the ore to raw metal or bloom and forging the raw metal into usable objects. Although Toro women did not participate in the technical aspects of iron production, they played crucial roles in its success. Finally, the benefits of the iron technology—the ore mined, the objects produced, the jobs created, the social status achieved—were largely fed back into the local communities and Toro society rather than siphoned off to distant sponsors. Recent interviews with elderly iron-workers who remember the precolonial technology reveal that the discovery of a hill of iron ore had major social and economic impacts within a community. The discoverer and new owner of ore was confronted with the possibility of considerable wealth, but also new investments, religious responsibilities and social obligations. Given that it was precapitalist and integrated with other economic pursuits, MacMillan (1995) and others (Knapp and Pigott 1997) might label the socioeconomics of Toro mining and iron-working communities as ‘informal.’ It was also small-scale, served a relatively small population, and lacked mechanized extraction methods. I find the term ‘informal’ misleading, however, for two reasons. First, although Toro miners, smelters and smiths participated in other economic pursuits, they characterized themselves as iron-workers first and foremost. Their participation in iron production was in no way casual or unofficial within the society (cf. MacMillan 1995:135). Toro agriculture, hunting and defense relied on iron, and its producers were highly regarded members of society. Second, the work relations involved in Toro mining were not ‘totally informal’ (MacMillan 1995:3). Social rules affecting the dynamics of work and related activities abounded in the Toro kingdom. The following case study focuses on the social and technical dynamics of Toro iron ore mining, situated within the context of local communities involved in the full iron-making process, as well as other activities. The first section
Iron mine discovery, ownership and power among the Toro
125
provides a brief overview of the sociopolitical and economic setting of the technology. The subsequent focus on Toro mining begins with the question ‘what did it mean to nearby Toro communities when a new source of ore was found?’ and then examines its relationships to other community activities. I conclude by exploring the material and non-material record of Toro mining and iron production that may provide helpful clues to archaeologists working on similarly oriented communities of the past.
THE SOCIOPOLITICAL AND ECONOMIC CONTEXT OF TORO MINING AND IRON PRODUCTION Precolonial Toro society was hierarchical and patriarchal. These social principles were constantly enforced and reaffirmed in the interactions between men and women, young and old, and the living and the dead. Women had fewer rights than men, and married women were dominated by their husbands. The longer a person lived, the more social stature he or she accrued. Spirits and ancestors had considerable influence over the living owing to experiences and transformations they had undergone during life and death. These principles were manifest in Toro politics, as well as during most daily activities such as planting, hunting, iron-working and cattle-keeping. Toro was a kingdom that broke away from the long-standing and nearby Nyoro state (Figure 8.1) (Ingham 1975; Beattie 1971). Kaboyo, the first omukama or king of the Toro, rebelled against his father, Omukama Kyebambe III Nyamutukura of Nyoro. Once the chief and tax collector for the southwestern portion of the Nyoro state, Kaboyo claimed the area occupied by Nyoro, Toro and Konjo peoples around 1830. He fought wars during his long reign to acquire new territories for Toro that were rich in pasture lands for his cattle and in tradable goods, such as salt and iron, for his treasury. Upon Kaboyo’s death, the Toro kingdom underwent a series of power struggles. The Nyoro sought to recover the polity on a number of occasions and succeeded in the late 1800s. They lost Toro again when the British intervened in the 1890s. When the borders between Nyoro and Toro were reset, largely through British protection, some border districts that previously had been a part of the Nyoro state, such as Mwenge, were incorporated into Toro. Throughout their political history, the Toro people were organized into three social strata, despite some social mobility. The social elite included the omukama and other members of the royal Babito clan, who were often appointed territorial chiefs. Through ritual, chiefs became local manifestations of the omukama’s ultimate power and authority. Wealthy cattle-keepers, called bahuma, were also members of this elite. The lowest social stratum consisted of slaves. The majority of people made up the middle stratum and were subsistence farmers or bairu. This grouping also included part-time craft specialists, such as potters, salt workers, iron producers and bark-cloth makers. Another important
126
S.Terry Childs
Figure 8.1 Map of the Toro and Nyoro kingdoms in Uganda during colonial times
group of specialists, who could be either bahuma or bairu, were religious personnel. Spirit mediums (nyakatagara) communicated with a family or lineage deity (omuchwezi) and performed rituals to influence the successful outcome of activities, such as childbirth, planting and hunting. The nyakatagara were also important to all stages of Toro iron-working. Interestingly, whereas most of the jobs involved in iron production were filled by men, Toro spirit mediums were usually women (Taylor 1962; Beattie 1968a). Women seem to have been preferred for this role because of their powers related to fertility, procreation and abundance. The settlement pattern across the hilly Toro landscape mostly consisted of dispersed villages. Toro population was relatively sparse, and lineage and clan affiliation provided the social, economic and religious structure in everyday
Iron mine discovery, ownership and power among the Toro
127
village affairs. Trade between local villages and between villages and the omukama’s capital was constant, particularly to distribute specialized goods such as iron, salt and ivory. This was because some raw materials and specialists who used or worked them were not evenly dispersed across the Toro kingdom. Family or clan groups of iron-workers, in particular, lived and worked where essential raw materials, such as iron ore and charcoal, were plentiful. The Mwenge district of the Toro kingdom was especially well known for iron production. Within this context, the family-oriented groups of iron-workers were headed by a male elder, most commonly called the chief smith (omuhesi omukuru). He was both powerful and wealthy in his village and the surrounding area, because of his skills and the importance of his craft to the society. The omuhesi omukuru knew the fundamentals of all aspects of iron production. He assessed risk, handled problems and assigned appropriate jobs during mining, smelting or forging. He carefully controlled both the technical and the esoteric knowledge related to the technology as he took on apprentices, usually his sons or nephews, enforced rules and taboos during production activities, and performed required rituals to enhance success and minimize risk. Finally, he always recognized and took advantage of cultural principles and values during ironworking that could enhance his own and his family’s social and economic position in relation to his clan, his community and wider Toro society.
THE SOCIAL AND TECHNICAL DYNAMICS OF TORO MINING AND IRON PRODUCTION Mining for iron ore and smelting were performed primarily during the dry months of January-February and July-August. Agricultural fields lay largely dormant during these seasons. Possibly more importantly, very little rain fell to destroy the mines through collapse or to jeopardize smelting by flooding the furnace or ruining the required raw materials. The miners and smelters labored hard during the dry months to make enough iron to work later at the forge. Smiths forged frequently prior to the planting season when many hoes were needed, but interspersed their work with many daily or seasonal activities during the year. Mining The discovery of a new hill of ore (obutale), a matter of luck, was not a frequent event. More commonly, miners obtained permission to open a new mine on someone else’s hill. The search for ore was done exclusively by Toro men, primarily married, older men who were experienced in iron production. Smelters and smiths, including the chief smith of a family group, might be involved. They limited
128
S.Terry Childs
participation of younger men to their sons and nephews, whom they could control through rules of social hierarchy and patriarchy. They did not have to discourage women from joining these ventures because of the danger involved and the amount of time taken away from domestic responsibilities. This search was very seriously pursued because to find a new hill of iron ore meant wealth and prestige within one’s family, clan, village, and beyond. Ore was not easily found on Toro hillsides covered with forest, tall grasses, grazing cows and wild animals, so men cooperated in parties of six or more. The teams looked for specific clues to find ore, such as significant changes in vegetation. Members of the Abachwamba clan sought a dung beetle or ekijunjumira called Kahinda whose hind legs dug through cow and elephant dung into the soil below. Sometimes the white-necked beetle dug up glittery black stones that might be ore. If a search party member found the beetle with possible ore, he called his friends. The other members of the search team then sought to verify the presence of ore by digging a pit and testing the stones retrieved. If the stones broke easily, they were not ore. If they were hard and emitted sparks when hammered, the Muchwamba man who first found the beetle and its leavings claimed he was the discoverer (omujumbuzi) and owner of the hill of ore. This announcement initiated or maintained a complex network of social relationships and rules of social behavior familiar in daily life that were played out during the subsequent stages of mining, smelting and smithing. The omujumbuzi immediately called the spirit medium (nyakatagara) of his family deity (omuchwezi) to come to the hill. The word of the discovery traveled rapidly and many people quickly gathered at the site with the nyakatagara. The social gathering was significant in a number of ways. First, it publicly established who was omujumbuzi for the new source of iron ore and accepted its responsibilities. Second, it provided the context for a crucial ritual dealing with all future mining on the hill and the first of several rituals for iron production using that ore. Third, it was the setting for celebrating the future wealth that the discoverer would reap and his family and friends might share. The nyakatagara conducted the ritual by first making contact with the discoverer’s family spirit and then slaughtering a white sheep provided by the discoverer. She dripped the sheep’s blood into the test pit and spat on the people around her using available water or beer. This was done to thank the family deity for the discovery of the mine, to ‘bewitch any evil that would spoil the mine’ (Ndunga, 7/941), such as collapse of a shaft or debilitating rain, and to ensure that the ore would not spoil during smelting. The act also extended luck, success and fortune from the deity to the people gathered. When the sheep was roasted, everyone drank, danced, sang and feasted until nothing remained, not even the bones. The discoverer compensated the spirit medium with a small piece of iron. The sheep and its color were important. Not only did sacrificed sheep skins cover the bellows pots used to stimulate combustion during smelting and forging, but iron-workers often wore them for their protective properties
Iron mine discovery, ownership and power among the Toro
129
(Ndunga, 7/95). Also, white meant purity, luck and bounty among the Toro and neighboring Nyoro (Needham 1967; Beattie 1968b), and its significance was evident at other occasions related to iron-working. For example, the iron produced in a successful smelt was called ‘white.’ A Toro song, sung at this and other events, refers to the smith’s hammer as ‘white’ and commemorates the prosperity it brings to iron-working families and to Toro society in general: When we were dead Say the hammer widens, We were saved by white. When we were finished Say the hammer widens, We were saved by white. In the days of eating clay Say the hammer widens, We were saved by white. The men and women who came to this event at the new hill of ore were primarily members of the discoverer’s family and clan. Men who had undergone the blood brotherhood ritual with the discoverer, often iron-workers from other clans and villages, also attended. Women, particularly clan members and the wives of the search team, were allowed, but no women who were menstruating. That condition would ‘spoil’ the ore when it was being smelted. These people took advantage of an opportunity that could offer them social and economic benefits, but acted within the limits of known social rules. Some hoped to reap from the good fortune of the discoverer through the potential jobs he had to fill and the wealth he might distribute. Some also wanted the blessing of the family spirit through the nyakatagara. Finally, they welcomed the feast of roasted meat at the celebration. Serious mining began the day after the ritual and celebration, after all nonminers had left the hill. The discoverer rarely mined himself, but supervised all activity on his hill. Given the hard work involved and the dangers associated with digging deep shafts, he negotiated with two to four young men who had come to the event to be the miners (abahaige). They were often his sons, nephews, other relations, or friends who had previous mining experience. He also negotiated with several older men, called fundi, who knew the properties of good ore to sort it out at the edge of the mine. The task of this group was to retrieve enough ore immediately required by the discoverer; no ore was stockpiled. The miners used a hoe to dig out the top layers of soil and create small channels around the mouth of the mine to catch and divert any rain that fell. They used an iron pick (omusuma, literally translated as thief) to mine the ore as they created narrow shafts approximately 1 m in diameter and 5–8 m deep. The miners sent up the mined earth and stone in baskets at the end of a rope. The
130
S.Terry Childs
men who sat at the top of the shaft used the omwangatta, a small hammer shaped much like a railroad spike and also used by smiths for finishing work, to separate out the ore from rocks. They used an open-weave basket to shake out the attached mud and dirt. Often many basket-loads of material came out of the mine before they yielded a basket of sorted ore. On rare occasions the rope and basket down the mine shaft with a miner went limp and a mine had collapsed. The miner’s body was usually left in place because the work-related death could be explained in ‘natural’ terms and no other lives were to be jeopardized (see Herbert, this volume, Chapter 9). All involved in mining knew that, as part of the job, they had to abstain from sexual relations to ensure that the ore transformed into iron during smelting. They also knew that they should negotiate a payment from the discoverer before they began work. The abahaige, those who dug the mines and risked death, could negotiate a higher payment than the fundi. These payments, usually pieces of raw iron in various sizes, were retrieved after the ore was smelted. Additional rules and activities that were neither technical nor located at the mine were initiated upon discovery of ore. The omujumbuzi instructed his most trusted wife, ‘Do not untie your belt’ (Ndunga and Kamara, 7/95). This meant that she must be faithful to her husband until he returned to his compound, sometimes not for many days, and be careful about who greeted her. No one could touch her, particularly on her shoulder, which meant bad luck, or around her stomach (a euphemism for her private parts), which was disrespectful and suggested infidelity. Since the discoverer owned the hill, family members, clansmen and neighboring iron-workers often asked permission to open other mine shafts on the hill. By this method the latter groups minimized the risk of finding ore, a risk already borne by the discoverer and his search party. Visiting work groups of four to eight men were assigned a place to work on the hill by the omujumbuzi. These groups did not have a designated leader; each man got an equal share of the ore they found by the end of the project. Since no one could be assured of finding ore, sometimes the omujumbuzi favored his relatives and assigned them a place near a known iron vein. He had to be careful to distribute the mining groups spatially, however, ‘because they were scared the world would turn upside down’ (Ndunga, 7/94). Before these groups began mining, they negotiated with the discoverer for appropriate compensation. This was usually based on the amount of ore they expected to recover. The payment would be made either in ore or in pieces of iron that the group later smelted. When they eventually presented the discoverer with his payment (omukinda), he brought it to his family’s nyakatagara. She blessed the iron or ore and verified that it was good to be used; in other words, that it was not harmful to him or his family, or bewitched. A second kind of iron ore was also collected before a smelt could be initiated. This was a red, dense material called entabo. There was no known ritual associated with the discovery of entabo, but it was valuable. Iron-workers paid to
Iron mine discovery, ownership and power among the Toro
131
obtain it from those who mined it from only a few sources. It was combined with the black, glittery ore as they dried them prior to smelting. This was done to ‘befriend’ or ‘embrace’ the black ore and make it stick together during the smelt, terms that may have sexual overtones. Notably, the iron-workers of the neighboring Nyoro and Ganda states referred to the hard, black ore they used as male and the soft, red ore as female (Roscoe 1911, 1923). When the discoverer brought the two types of ores to dry in his compound prior to smelting, other rules directly applicable to his family commenced. First, his most trusted wife now had to have sexual relations with him on a regular basis: ‘She had to sleep facing her husband until the smelting of the ore’ (Ndunga, 7/94). Perhaps this was done to reinforce and reproduce an activity that would soon occur in the furnace. Another rule was that the husband and wife must be faithful to each other during this time. The directive against touching this wife also continued. Finally, the only people who could touch the drying ore were the discoverer, his trusted wife, and small children, who were considered ‘pure.’ This rule is noteworthy because the ore had to be set out to dry each day and quickly brought in when rain threatened, but only a few people could do these jobs. In sum, the social relationships that were most important during this period of protecting the ore were those that stressed fertility and marital fidelity. When the ore was ready, the discoverer prepared to reduce it to raw iron in an iron-smelting furnace. Occasionally, if he was not experienced in smelting, he had to negotiate with fellow clansmen or others who were more knowledgeable to do the work. Only men smelted, including two or three bellowers (abajugusi) who spelled each other pumping the bellows. None of them could sleep with a woman the night before they smelted. Subsequent iron-working activities Toro smelting involved digging a furnace pit approximately 1 m in diameter and up to 1m deep, lining it with ash, and building a protective shelter over it. Toro smelters placed one pair of bellows at the side of the pit and aligned the flared mouth of the one tuyère or blowpipe with the two bellows nozzles. They set the tuyère into the pit so that it lay above layers of dried reeds and charcoal and below layers of charcoal and ore. No walls or superstructure were built above the pit as was done by neighboring groups (Roscoe 1911, 1923; Schmidt and Avery 1978). The owner of the ore summoned the nyakatagara to come and make contact with his family spirit when the furnace was prepared and first lit. A white sheep, supplied by the owner of the ore or by someone who wanted a sizable piece of iron from the smelt, was slaughtered. The nyakatagara sprinkled its blood into the furnace (ejugutiro), on the bellows, and on the iron-workers. This was done to ensure bountiful success and prevent any harm to the ore, process and workers.
132
S.Terry Childs
During these furnace preparations and the smelt, certain people were not permitted nearby. A smelter’s wife had to call from a distance to announce the food she brought. Neither she nor any other woman was allowed at the site; those who were menstruating were particularly dangerous to the furnace. No one, male or female, who had been unfaithful to their spouse could look at the furnace and ore. Finally, no one who might sneeze or pass gas by the furnace was tolerated, since the abrupt noises and gas, like the presence of the other individuals, could cause the ore to turn black in the furnace and die. The smelters set up a blockade (enkomo) of branches across the path to the furnace to enforce these rules and warn people of their activities. This blockade was identical to that used when the nyakatagara performed various rituals for a family. Fidelity, as well as the removal of any potential barriers to furnace productivity, was emphasized in the social interactions allowed at a smelt (Herbert 1993). Such social values and rules were also symbolized and reinforced by ironworking tools, particularly the two clay bellows pots integral to smelting and forging. One bellows was female, called the mother of twins (nyinabarongo). It was identified by a roulette design around the pot bowl or a small mark depicting a navel (omukundi), although the female sexual organs were obviously displayed by neighboring groups (Lanning 1954). The other bellows was male, called the father of twins (esebarongo), and was identified by a molded penis located above the air shaft. The birth of twins was and still is a mixed blessing among the Toro and the neighboring Nyoro (Beattie 1962) because it connotes abundance and prosperity but also involves serious ritual danger. Parents of twins must perform expensive and extended rituals to counter the danger to their family. They are held in high esteem once their ritual obligations are met (Beattie 1962:1). The two anthropomorphized bellows were tied together under one sheepskin with the male on the left side and the female on the right. This is the customary arrangement of Toro men and women in many social situations, including sleeping in one bed. A stick was tied into the sheepskin above each bellows chamber so that an alternating pumping rhythm efficiently pushed air into the furnace during smelting. The sexual pairing of the male and female bellows and the efforts of the sexually potent and faithful iron-workers, then, yielded great wealth—an iron bloom that was referred to as abarongo or twins. The iron-workers watched for specific clues to determine the progress of the smelt. When they saw slag dripping through the tuyère, the smelt leader retrieved a small pot filled with red earth, sprinkled it into the furnace, and asked the bellower to pump hard. This earth was used ‘to guard the ore’ (Ndunga, 7/94) and to help the obutale (black ore) and entabo (red ore) join. When the growing iron was very red, the leader might call the nyakatagara to the smelt. She filled her mouth with water, spurted it into the furnace, and said, ‘Let it be white, let it be pure white’ (Ndunga, 7/94). The whiteness of the iron bloom relates to purity, bounty and goodness, as mentioned above. The ironworkers knew that the smelt was near an end when they heard hissing sounds
Iron mine discovery, ownership and power among the Toro
133
and saw sparks flying from the white iron. These signs were a cause for celebration, and they sang: Ntogota, come and accompany me; it is almost hissing but there is no tuyère; Woman and child, come and accompany me; it is almost hissing but there is no tuyère; Cow and sheep, come and accompany me; it is almost hissing but there is no tuyère. Ntogota, the deity of the Abachwamba clan, helped the ore form into hissing iron. The woman, child, cow and sheep referred to here were other highly valued products obtained from the wealth brought by a successful smelt. When the iron hissed, the tuyère finished working. The end of the smelt was a time of celebration called the ‘serving,’ attended by old men and women, sexually active yet faithful adults, and women who were not menstruating. This was when the bloom was pulled out of the furnace by two or four servers (abahereza) who used long sticks (emihano). The splitter (omwasi w’ebyoma) then roughly hammered out and cut the bloom into pairs of pieces with an axe. A clear relationship exists between the cut iron pairs and the successes of the parents of twins during the smelt. All the individuals involved in the various tasks leading up to the smelt, such as the miners, ore separators, charcoal makers, bellows makers and bellowers, attended the serving. This was when they were compensated for their services with pieces of iron sized according to the job they performed. It was incumbent on each individual to be present on this occasion to ensure that he received the share he had negotiated. There was singing and dancing at the serving. Another important and commonly sung iron-working song about the bellow and tuyère (‘short-legged’) was: Clansmen, thank you. Clansmen, thank you. Chorus: I slept when the bellows were nice/good/wealthy. Short-legged, where did you sleep? Short-legged, where did you sleep? Chorus: I slept when the bellows were nice/good/wealthy. Abwooli, who is sweeping, where did you sleep? Abwooli, who is sweeping, where did you sleep? Chorus: I slept when the bellows were nice/good/wealthy. These songs were sung to the family spirit and were always accompanied by someone hitting or ‘playing’ the smith’s hammers.
134
S.Terry Childs
The year-round, final phase of iron production was forging. The forge was where objects were fashioned from bloom (now truck and Land-Rover springs) and repaired, and worn items were reforged into other object types. It is and was a small, roofed hut open at the sides that contained a forge pit, a pair of decorated bellows, a pit of water and sprinkler to dowse the flames, at least one anvil, and various hammers and other tools. Notably, the hammers had particular names and shapes. The male omurubiko was the massive round stone hammer used to roughly pound out the raw iron during the serving and forging. The female enyondo was the large iron hammer used to begin shaping an object. The smaller, but similarly shaped hammer called the child was used for shaping small objects or finishing work. The names and relationships of these hammers reinforced yet again Toro social values and rules. The forge was, and still is, a social place. Smiths built them in visible locations, such as by a main path, near their compounds to enhance their economic success. Anyone could come by the forge to buy, trade or pass the time; all passers-by, including women and the king of Toro, were encouraged to pump the bellows. The master smith, however, also took precautions to protect the forge. An elaborate ritual involving the nyakatagara and a celebration with family and friends occurred when a forge was first established. The smith sacrificed a sheep for the forge at least once a year, which also involved the nyakatagara and others. In the morning before the smith worked, he healed or cleansed (kutamba) the forge by sprinkling water around it and on the tools with the sprinkler (isiza) used to stifle flames during forging. The smith was also careful to avoid having intercourse with a woman the night before he planned to forge.
DISCUSSION Many scholars who study ancient and precolonial African iron technology have focused on the social context in which it occurred and the dynamic factors involved in its social construction for several decades (for recent references see Childs and Killick 1993; Herbert 1993; Schmidt 1996). This perspective has benefited from the depth and variety of the available ethnographic record, early observers’ fascination with the associated ritual and symbolism, and the ability of current researchers skilled in both anthropological archaeology and metallurgy to work with craftsmen who still remember the precolonial technologies. The Toro case study attempts to document a technology that was once integral to its society, yet will soon be extinct. It also provides some further interpretive insights for archaeologists working on the prehistory of this region, metallurgical communities in general, and mining in particular. First, iron-working was in no way an ‘informal’ sector of Toro socio-economics even though its components were conducted part-time and seasonally. Its practitioners were often wealthy, highly respected members of society as a result of their integrated skills in and knowledge of mining, smelting and smithing. They greatly valued their tools both
Iron mine discovery, ownership and power among the Toro
135
for functional purposes and as symbols of their success and prosperity. Toro ironworking tools were handed down over generations and carefully stored for use. In a similar prehistoric context, excavating the compound of a family elder might yield a variety of tools, such as mining picks, forging hammers, and bellows used for both smelting and smithing. Second, social rules and values of daily life were often used during ironworking to help ensure its success. This was done by controlling who participated in its activities, using specially named and shaped tools, and through songs sung during iron-working events. It will be very difficult for archaeologists to determine who was allowed at ancient mines, smelts or forges. Unfortunately, too, significant songs that often highlighted the relationships between people, meaningful objects and rules are usually lost to the archaeological record. The use of tools to construct and reinforce social rules might be traced archaeologically, however. The importance of male and female relations in Toro society, for example, was clearly evident in the vividly decorated bellows. The male and female hammers were also differentiated by material and shape, although these characteristics would be impossible to interpret in the archaeological record. Archaeologists should never assume that tools lack social and ideological information just because their primary and most obvious function was technical. Third, African mining, smelting and smithing often involved rituals. The Toro case, however, illustrates that the critical importance of the nyakatagara and her connection to a family deity would be hard to determine in an archaeological context. This is because the material culture involved was generally perishable, such as blood, beer and a sheep that was fully consumed, including its bones. A hearth used to cook the sheep might be the only evidence of ritual events on a mined hill, but this find might be difficult to interpret. The lessons learned through ethnoarchaeology, however, can offer critical clues on where and what to investigate (Schmidt and Mapunda 1997). Finally, an accurate archaeological reconstruction of the complex social context in which mining occurred will be elusive. Numerous mine shafts and pits offer the best evidence of mining operations. Study of their number and distribution might indicate past scales of production, while mine shafts might contain lost tools and skeletons of dead miners for productive interpretation. Many of the tools used, however, were both perishable and similar to domestic objects, such as baskets for hauling up mined materials, sifting ore and dirt, and transporting ore. Other tools were multipurpose, such as pickaxes and hammers. As mentioned above, discovery of durable material culture would be needed to detail any related ritual performances, but archaeologists should know to look for such clues. An iron-worker’s compound where ore was dried and stored might be found archaeologically, although probably with little insight into any associated social rules. The methodological and interpretive relationships between ethnographic and archaeological studies are still being worked out around the world. Some
136
S.Terry Childs
fruitful information is being gained in Africa where the observations, opinions and feelings of elders are heard. The importance of a technology such as ironworking to its practitioners and society is especially insightful as models of ‘informal’ technologies are offered in our present-day context.
ACKNOWLEDGMENTS This chapter is based on a series of interviews I held in 1994 and 1995 with Toro elders from the Mwenge district of the Toro kingdom, particularly N.Ndunga, M.Bamulimbya, A.Kamara, Z.Kirigwajo, P.Tinkasimere, A.Mpeerabusa and P.Bamanyisa. I dedicate it to the memory of N.Ndunga, who with his wife Abwooli, provided great hospitality and patience. Peter Robertshaw contributed financial (NSF grant no. SBR-9320392), logistical and collegial support for which I am most grateful. John Sutton and Justin Willis of the British Institute of East Africa generously provided use of a Land-Rover in both 1994 and 1995. Charlotte Karungi, my translator and friend during both field seasons, did a tremendous job. I also wish to thank the Uganda National Council for Science and Technology for research clearance, the Ugandan Department of Antiquities and Museums, particularly Dr Ephraim Kamuhangire, for their support, and the District Executive Secretary of Kabarole and his assistants for their assistance. Thanks also go to Robert Ehrenreich for comments on an earlier version of this chapter and to Mark Oviatt for creating the map. Finally, I want to thank the organizers of the mining conference held at the Rockefeller Center in Bellagio for a most stimulating and fruitful gathering.
NOTE 1
In-text references, such as ‘(Ndunga, 7/94),’ relate to the results of interviews held with one or more Toro elders during specific months in 1994 and 1995.
REFERENCES Beattie, J. (1962) Twin ceremonies in Bunyoro. Journal of the Royal Anthropological Institute of Great Britain and Ireland 92: 1–12. ——(1968a) Spirit mediumship in Bunyoro. In J.Beattie and J.Middleton (eds), Spirit Mediumship and Society in Africa, pp. 159–170. New York: Africana. ——(1968b) Aspects of Nyoro symbolism. Africa 38: 413–442. ——(1971) The Nyoro State. Oxford: Clarendon Press. Childs, S.T. and Killick, D. (1993) Indigenous African metallurgy: nature and culture. Annual Review of Anthropology 22: 317–337. Ehrenreich, R. (1996) Mining communities in history: an industrial legacy. JOM 48(12): 54–56.
Iron mine discovery, ownership and power among the Toro
137
Godoy, R. (1985) Mining: anthropological perspectives. Annual Review of Anthropology 14: 199–217. Hardesty, D.L. (1988) The Archaeology of Mining and Miners: A View from the Silver State. Society for Historical Archaeology, Special Publication Series no. 6. Pleasant Hill, CA: SHA. Herbert, E.W. (1993) Iron, Gender and Power: Rituals of Transformation in African Societies. Bloomington: Indiana University Press. Ingham, K. (1975) The Kingdom of Toro in Uganda. London: Methuen. Knapp, B.A. and Pigott, V. (1997) The archaeology and anthropology of mining: social approaches to an industrial past. Current Anthropology 38: 300–304. Lanning, E. (1954) Genital symbols on smiths’ bellows in Uganda. Man 54(262): 167– 169. MacMillan, G. (1995) At the End of the Rainbow? Gold, Land and People in the Brazilian Amazon. New York: Columbia University Press. Needham, R. (1967) Left and right in Nyoro symbolic classification. Africa 37: 425–452. Roscoe, J. (1911) The Baganda. Cambridge: Cambridge University Press. ——(1923) The Bakitara. Cambridge: Cambridge University Press. Schmidt, P.R. (ed.) (1996) Culture and Technology of African Iron Production. Gainesville: University of Florida Press. Schmidt, P.R. and Avery, D. (1978) Complex iron smelting and prehistoric culture in Tanzania. Science 201: 1085–1089. Schmidt, P.R. and Mapunda, B. (1997) Ideology and the archaeological record in Africa: interpreting symbolism in iron smelting technology. Journal of Anthropological Archaeology 16: 73–102. Taylor, B.K. (1962) The Western Lacustrine Bantu. London: International African Institute.
Chapter 9
Mining as microcosm in precolonial sub-Saharan Africa An overview Eugenia W.Herbert
ABSTRACT Iron, gold and copper were the three primary metals worked in sub-Saharan Africa before the colonial period. This chapter examines the ‘culture of mining’ in Africa; its aim is to illuminate not only technologies and social relations but also the patterns of belief which underlie them. The available evidence suggests a great deal of variation both in organization of labor and in ritual. In most regions, mining was a dry-season pursuit, complementary to farming and involving the same labor force, although in some areas it was the work of specialists and of distinct social groups. The labor of women, children and younger men in mining seems to have been almost uniformly under the control of senior males who possessed both the technical and the ritual knowledge believed to be indispensable to a successful outcome. Where women collected, pulverized and winnowed ores, the analogy with agricultural pursuits might be explicit; and where ores and fuel had to be hauled to metallurgical sites, it was primarily women who carried them. The degree of ritualization also varied: the use of medicines, sacrifices and invocations to ancestors and deities, the imposition of sexual and menstrual taboos, and so forth. There is some evidence that deep mining and mining by specialists involved greater ritualization than panning and shallow mining. Although women might provide key labor components in the mining and preparation of ores and fuels, they were rarely allowed to take part in the more highly skilled smelting activities, which demanded access to supernatural assistance; just as rarely did they share in the wealth and status that accrued to metalworkers.
INTRODUCTION Metals have been worked in sub-Saharan Africa for more than two millennia. Those who produced these metals and the objects they fashioned have played dominant roles in the history of the subcontinent. In many cultures, myths of origin refer to the revolutionary impact of metalworking. Objects of iron, copper and gold have long been central to political, economic, social and religious life. So abundantly was Africa endowed with deposits of gold that it
Mining as microcosm in precolonial sub-Saharan Africa
139
was for centuries synonymous with the glittering metal, first to the Arab and Indian Ocean worlds, then to Europeans. However, in Africa itself iron has been the primary metal in all spheres of life, from utilitarian to ritual. Furthermore, in many societies copper was traditionally more highly esteemed than gold. In this chapter I will focus on the organization of mining labor in precolonial subSaharan Africa as a whole, with particular attention to the ways in which it may or may not have differed from the organization of agricultural and domestic labor, and to the ways in which it reveals underlying patterns of belief.
RESOURCES AND HISTORY Much of Africa south of the Sahara is exposed lateritic crust; that is, low-grade iron. Metalworkers exploited primarily the richer deposits of oxides such as hematite, magnetite and limonite. Gold, in both alluvial and reef form, also occurs much more widely throughout the continent than is generally realized. Copper deposits, in contrast, are relatively scarce in West Africa, with the exception of areas of the southern Sahara in Mauritania and Niger, although they are more plentiful in regions of Central and Southern Africa. Lead occurs occasionally, especially in association with copper. Tin was also exploited in a few areas, such as southern Africa and the Bauchi plateau in Nigeria. Despite tenacious Portuguese beliefs to the contrary, there is little silver to be found in Africa, and use of the metal is recent except in the Sahara and Sahel. During the period from about 600 BC to AD 500, the working of iron and copper spread over the entire subcontinent, apparently in tandem. There is no evidence of a sequence from copper to arsenical copper or bronze and finally to iron as in the Middle East and Europe. The lack of clear routes of diffusion from North Africa, the Arabian Peninsula or the Indian Ocean has led some scholars to propose that metallurgy may have developed independently in Africa, but this remains unproven (Herbert 1984:8–10; Schmidt 1996a:8–10). If metalworking was not invented independently, it may have reached Africa from several sources rather than a single one; this hypothesis fits the scatter of early dates at great distances from each other better than theories of a single route of diffusion. The spread of metalworking throughout the Congo Basin and into eastern and southern Africa was long held to be associated with the migration of Bantu agriculturists. More recent archaeological and linguistic reconstructions have undermined this theory except in southern Africa. The technology seems to have spread in complex patterns, and stone-using hunter-gatherers coexisted with iron-using farmers in many areas. Indeed, iron-workers themselves frequently continued to use stone tools for rough forging until recently. The effects of metallurgy may have been gradual rather than revolutionary, but they were undoubtedly profound. Iron hoes, axes, knives, hooks and spears increased the efficiency of agriculture, hunting, fishing and warfare. They contributed to the evolution of complex societies characterized by the
140
Eugenia W.Herbert
production of surpluses, craft specialization and social stratification. Iron, copper and gold all entered circuits of trade. Raw iron and copper, as well as forged objects, were traded locally to peoples lacking resources or specialists. With the development of the trans-Saharan trade during the first and early second millennium AD, copper and brass wares became major imports into West Africa, often exchanged directly for gold. The opening up of maritime routes between Europe and West Africa in the second half of the fifteenth century led to an exponential increase in these imports and, for a relatively brief period, in exports of gold. Eventually, too, iron bars also became a staple import from Europe. Africa’s mineral wealth, actual and mythical, was a major catalyst for European discovery and eventually for conquest. Archaeology provides a time frame for African metalworking and some indication of the technologies used in mining and smelting, but it tells us too little about social organization and beliefs. For these we need other kinds of sources, and they are sparse indeed before the nineteenth century. Indigenous cultures were mostly non-literate, and travelers who have left written accounts rarely visited metalworking sites. To some extent their exclusion from mining areas, especially gold-mining areas, seems to have been a deliberate strategy on the part of local peoples to keep unwelcome and possibly predatory strangers away from indigenous sources of wealth (see, for example, Garrard 1980:130). A rare exception is Ibn Battuta’s all too brief eyewitness account of copper mining at Takedda in the southern Sahara in the mid-fourteenth century. By early in the twentieth century, indigenous metalworking had died out in most areas, so that much of the documentation comes from interviews with elderly informants rather than direct observation, and tends to be weighted much more heavily toward smelting and smithing than toward mining. From the evidence that does exist, both oral and written, it is hard to detect a distinct ‘culture of mining’ in sub-Saharan Africa. That is to say, mining involves a great variety of patterns, but individually these mirror the societies of which they are part rather than being unique to the mining activity itself. In general, mining involved the creation not of permanent settlements that took on a character of their own but of temporary camps established by villagers in a particular region at appropriate seasons of the year—an important factor to which I shall return. But among the many ethnic and subethnic groups for which there is information, one can identify two main areas where differences show up. The first relates to the social and gender categories of the miners: were they free or slave; members of smithing lineages or the population at large; men, women, children or all three? The second area of difference concerns the degree of ritualization associated with mining and the elements of the rituals themselves. At the same time, we must resist the temptation to assume a static model of precolonial mining that changed only with the intrusion of imperial mining technology and capital. Although it is difficult to document change in social organization and technology over the two millennia or more before European
Mining as microcosm in precolonial sub-Saharan Africa
141
rule, change there undoubtedly was, especially in response to increasing demand for metals for both local consumption and trade.
MINES AND MINERS Within Africa, iron and copper mineralization often occurs close to the surface so that ores were dug out easily in trenches or opencast mines with simple picks and hoes (Cline 1937:26, passim). Some of these opencast mines were huge: ancient copper workings at Kansanshi in modern Zambia covered an area some 7 1/2 km in diameter, while workings at Bwana M’Kubwa, also in Zambia, extended to a depth of over 50m (Herbert 1984:25). In the Sahelian region of Takedda, it appears that native copper was literally gleaned from the ground with virtually no digging at all (Herbert 1984:67–68). Both alluvial gold and forms of iron (magnetite and limonite) were gathered from river sands. In a few areas, especially in central and southern Africa, however, deep mining of gold and copper involved shafts that went down 30m or more (Summers 1969: 163– 165). Miners used gads and hammers as tools and were familiar with techniques such as fire-setting and timbering, but encountered major problems of ventilation, illumination and flooding as well as cave-ins (Summers 1969). The inability to solve the problem of flooding meant, first, that pit mining was limited to the dry season and, second, that mines could not be sunk below the water table. In the agricultural societies that predominated in sub-Saharan Africa, those who provided the labor for mines were most commonly the same people who provided the labor for other activities; there was nothing that set miners apart. They could be men, women or children or, often, a team of all three. There is some suggestion that men tended to do deep mining, while women and children did surface mining and prepared ores. However, this runs up against finds of skeletons of both women and children in underground shafts in some areas (e.g. Herbert 1984:44–45; Schofield 1925:6–7). Miners sometimes included slaves just as the larger labor force included slaves, but the role of slaves in mining has probably been exaggerated—and, indeed, we cannot always be sure what the term may have meant in a given society. Before we examine exceptions to these generalizations, let me offer several examples that support them. For much of the year, Kwanyama metalworkers in southern Angola were farmers and cattle raisers. During the dry season, however, whole families decamped from their regular villages to a region rich in iron ore which was considered !Kung Bushmen territory. There they built huts, turned cattle loose and, together with Bushmen ‘friends’ (probably clients), they settled in for some weeks, mining, smelting and forging iron. Mining in the sandy soils was relatively easy. Often ore could be found close to the surface of the ground. Women hoed away the light covering of organic material to expose it, then men pried loose blocks with a sort of crowbar made of a large iron point attached to
142
Eugenia W.Herbert
a wooden handle. Next, women crushed the blocks using two stones as rough hammers and anvils. It was their job, also, to transport the ore to the smelting furnace using the same basket-traps used for fishing. In the copper-rich region of Katanga to the northeast, whole villages also set up temporary settlements near copper mines during the dry season. Monsignor de Hemptinne, Apostolic Bishop of Katanga, has left a particularly graphic account of such a displacement early in the twentieth century, although we must be aware that it was a command performance, initiated by him. He describes (Hemptinne 1926) how BaYeke men, women and children all left for the mines under their elderly chief. The actual mining in this case was done by men only, in open pits or underground shafts, using picks similar to their everyday axes, except that the handle and blade were heavier and the cutting blade angled differently. As shafts were sunk deeper, ladders made of fiber were lowered. At the bottom, three miners worked to dislodge the ore, filling cylindrical baskets made of fiber from local trees (Figure 9.1). Light but strong, these baskets could hold about 25 kg of ore. They were passed to the surface by men and women stationed on the ladders. If blocks were too heavy to fit into the baskets, they were raised by fiber ropes. Once the blocks were on the surface, women washed, sorted and pulverized chunks of ore until they could be winnowed to separate the malachite from the rocky matrix (Figure 9.2). When the miners struck rock walls too hard for their picks, they used firesetting to break them up. The shaft was filled with wood to the desired height and left to burn until morning. By that time the heat had burst the obstructing stone and mining could resume. Mining in this region might reach a depth of 35 m, but generally went no deeper than 10 or 15m, proceeding at a rate of about 1 m per day. Even this pace could not be sustained for long, so that a shaft of 30 m might represent three months’ labor. The only source of light for the shafts was a burning bunch of straw—no other type of lamp seems to have been used. Timbering was rarely necessary because of the circular shape of the shafts and the solid nature of the terrain. Furthermore, underground galleries were not long: the longest were about 20 m. The only mineral sought was malachite, which would have been found in the weathered deposits above the water table (Hemptinne 1926, summarized in Herbert 1984:51–52). In the Akan regions of modern Ghana, there were several methods for obtaining gold: panning, shallow pit mining and deep shaft mining. Panning for alluvial gold was mainly the task of women and young boys. They would stand in shallow streams or pools at the base of waterfalls or river estuaries after a heavy rain to scoop up gold-laden sands in shallow bowls or calabashes. To separate the grains of gold from the sands was extremely time-consuming; about 5 lb of sand or soil was considered the maximum that could be washed in a day and this yielded only a modest amount of pure gold (Dumett 1979). Consequently, the lion’s share of the precious metal produced in the Akan states and exported both north and south over many centuries came from shallow pits dug to expose the sedimentary deposits of fine gold and nuggets
Figure 9.1 Mining malachite in Katanga (1924)
Figure 9.2 Washing malachite ore in Katanga (1924)
Mining as microcosm in precolonial sub-Saharan Africa
145
that lay beneath the strata of soil, laterite, clay and gravel. These pits were usually no more than 1–3 m deep. Ideally, they bordered rivers and streams or dried-up river beds. Areas exploited over a long period of time were honeycombed with holes averaging less than a meter in diameter, creating a hazard for the unwary traveler (Dumett 1979). It was seasonal work, concentrated in the months at the very end of the dry season and beginning of the rainy season. This was the time when the extraction of alluvial or sedimentary gold was easiest: after the ground was dampened and the deposits in streams loosened, but before shafts became flooded and shallow pits turned into quagmires. It also complemented the agricultural calendar, falling between the harvesting and planting seasons for most of the local staple crops, so that there was no competition for labor (Dumett 1979:44–46; Curtin 1975:204–206). Unlike panning but typical of other forms of mining, shallow pit mining for gold was organized primarily around entire family groups among the Akan. At the height of the mining season, whole villages would relocate to the richest mining districts. They would take provisions and tools and set up camps near the goldfields for several weeks at a time. A single labor unit, mobilized by a male head of household, would normally include several wives and numerous children, including nieces and nephews under the Akan matrilineal system. It might also include domestic slaves incorporated as part of the kin group. It is important to note that these slaves did the same work as other members of the family, so that, as Dumett argues, ‘if there was any division of labor in mining, it was based on age and sex, not on slave versus “free.”’ For the most part, men and older boys dug and broke up ores, while women washed and separated gold from earth and gravel (Dumett 1979). Manuel Barreto, writing in 1667, describes a very similar type of mining by family units in the Kingdom of Baroe in what is now Zimbabwe (Theal 1898–1903:3/487–488). To a much lesser extent, Akan mining also involved ‘hard rock’ or digging for reef gold by means of deep shafts. Formations of vein gold were probably recognized less by geological markers than by distinctive forms of surface vegetation; or they might be located by following an outcropping underground. The depth and size of shafts depended on a number of variables such as the nature of the rock and the levels of groundwater. A common type of pit was bottle-shaped—with a long, narrow top and neck, opening up at the center and bottom. Probably most of these shafts were 3–7 m deep, but a number were much deeper, 25–35 m and more. The very deep shafts were accessible by bamboo ladders and/or foot- and hand-holds dug into the sides; otherwise miners simply used pressure of the legs and back against the walls to go up or down. At the bottom, galleries were enlarged to accommodate thirty or more workers. Nevertheless, profitable as such lodes might be, deep reef mining was probably confined primarily to the richest districts since it required so much hard labor, presented so many problems of ventilation and water, and was even more hazardous than the shallow pits. Even this type of mining relied principally on family units or sometimes cooperation between families from a
146
Eugenia W.Herbert
village (contrary to claims that it was the monopoly of the state and the domain par excellence of royal slaves [see Terray 1974]). But it was not always necessary to establish temporary settlements at the mines. Some Akan communities were close enough to gold sources to come and go on a daily basis (Dumett 1979:45). At Lopanzo in equatorial Zaire, miners had access to rich deposits of iron almost at their doorstep (Figure 9.3). Indeed, local traditions explained that the village had been moved to its present site precisely to exploit them. Here men could dig the ore in shallow open pits and carry it to the nearby smelting site for roasting (Herbert 1993). Similarly, the rich iron ores of the Bassar region in Togo were readily available to metalworkers in Banjeli and surrounding villages. At Banjeli, mining was the work of women and children, including, apparently, slave women acquired through trade in iron blooms and finished goods (Figure 9.4) (Goucher and Herbert 1996). However, it appears that in nearby areas men also mined (Martinelli 1982:39, 41). The main exception to this model of mining as a complement to agriculture concerns specialist communities that devoted themselves almost entirely to metal-lurgy. Bellamy has left an excellent description of such a community in western Nigeria near the Yoruba capital of Oyo at the turn of the nineteenth century. It consisted of about 100–120 individuals and ‘beyond the cultivation of a few acres of provision-grounds for their daily requirements, the whole of them— men, women, and children down to the ages of five or six—are occupied in the mining, smelting, &c., of iron’ (Bellamy 1904:101–102). They had migrated to their present location a few years earlier because it provided betterquality ore. Bellamy compared their methods of mining with those found in ancient stream-workings in Devon and Cornwall: the shale lying below about 6 ft of gravel was excavated by means of a pick and broken into pieces weighing 3–5 lb. It was roasted at night, then ‘pulverized in any ordinary wooden mortar, such as the natives use for pounding yams, &c., with the aid of a pestle of the same material, both shaped in the rudest manner.’ This work was done by women and children, ‘who seem to take to this work as soon as they are strong enough to wield the pounder.’ The ore is then passed through a basket-sieve before being carried by the women to a nearby river for washing. Bellamy describes the washing process in some detail; it seems very close to panning for alluvial gold with its tedious repeated sluicings to separate finer material from dross. Finally the ore, by then a rich dark red, was ready for the smelting furnace and could even be poured in while still damp (Bellamy 1904:103). A similar specialization seems to have characterized the Lemba, specialist copper-workers in the Messina area of South Africa. More commonly, however, even where metalworkers formed a distinct, hereditary group, they also farmed part-time. Or, while one had to belong to a smithing lineage to work metals, not everyone in the lineage would choose to exercise the craft. As Hemptinne put it somewhat parochially, ‘Every BaYeke is not a miner any more than is every inhabitant of the Borinage’ (Herbert 1984:45).
Figure 9.3 Iron mining in Lopanzo, Zaïre (1989)
148
Eugenia W.Herbert
Figure 9.4 Iron mining in Banjeli, Togo (1985)
Mining as microcosm in precolonial sub-Saharan Africa
149
To summarize, then: since mining was of necessity a dry-season activity in most areas of sub-Saharan Africa,1 it did not conflict with agriculture. Indeed, it was only because it could be done at a time when agricultural demands on women’s labor were diminished that it was feasible for women to engage in inordinately labor-intensive activities such as panning and washing gold. A very recent estimate suggests that ‘female labor accounted for at least two-thirds of all labor inputs in Akan precolonial gold mining’ (Dumett n.d.: 79). Unlike smelting and smithing, which were exclusively male activities, there seems to have been much more fluidity in the allocation of gender and age roles in mining, perhaps because mining itself meant very different things in different areas. True, in some areas such as Futa Jallon and Hausaland, mining was not only a male monopoly but also a monopoly of men from smithing families (Appia 1965; Krieger 1963). Where ores were relatively near the surface or found in streams and rivers, however, it was almost a form of gathering and often fell into the domain of women and children; deep mining, in contrast, required greater skill and endurance, although even here we cannot be sure that this always marked it as men’s work. Transporting ores and preparing them for smelting, on the other hand, were much more commonly women’s work. This is hardly surprising, since when there is carrying to be done in Africa, women almost always do it, or did so in the past. The use of ordinary mortars and pestles to pulverize ores, remarked on by Bellamy and others, underscores the analogy of ore preparation with food preparation, which was universally women’s work in black Africa. Dumett (1987: 217) describes how Akan women would ‘roll out’ handfuls of pounded gold ore with an oblong stone on a slab until it became powder in the same way that they prepared ‘kenkey’ by crushing maize between two stones. Mungo Park (1969 [1799]:231), who has left one of the earliest and most extensive descriptions of Mandingo gold mining in the western Sudan, notes that while the pits were dug by men, they always passed the calabashes of sand to women to wash since they ‘are accustomed from their infancy to a similar operation, in separating the husks of corn from the meal.’
RITUALS AND BELIEF Most activities of basic importance in Africa are embedded to some degree in ritual and behavioral prescriptions and proscriptions. Is this true of mining? If so, what does it tell us about beliefs? Let me propose a tentative hypothesis suggested by the admittedly very incomplete evidence: namely, that there was more ritualization in the case of deeper mining than of surface collecting and therefore more ritualization where miners were predominantly male. There may also have been more ritualization in societies where there were restrictions on who could mine (especially where those who did the smelting were themselves miners), but here the available evidence is even scantier (but see Schmidt 1996b:88).
150
Eugenia W.Herbert
Thus, where mining was simply a matter of collecting river sands, there seems to have been little ritualization of the process and no restrictions on who could do it. Among the southern Mbeere in the Kenya Highlands, for example, anyone could gather iron-rich sands and take them to a specialist who smelted along with the charcoal necessary for reduction (Brown 1971). Similarly, Akan women and children panning for gold seem to have done so with no preliminaries as long as they avoided certain days sacred to the spirits of the rivers. In addition, women did not pan during their periods (Dumett 1987:214; Garrard 1980: 129, 131, 138). Garrard (1980:138) asserts that sacrifices were also made to rivers where gold was panned, but it is not clear how widespread this was or how often it had to be done. Nor does there seem to have been any ritual accompaniment to Kwanyama surface mining (Herbert forthcoming). Pit and shaft mining, however, were quite a different matter. Here one often finds the full panoply of sacrifices, invocations, sexual and menstrual taboos, and exclusion of whole categories of people. While Mandingo prospectors in the Bambuk-Bure region of western Sudan depended on practical indicators such as particular types of vegetation to identify prospective mines, they also used a range of divinatory techniques: dreams, verses from the Koran, kola nuts. Once a site was located, a priest offered prayers, recited the Koran, and distributed red kola nuts before sacrificing a cock, goat or bull. If the mine proved rich, there would be further sacrifices to the spirit of the gold (Garrard 1980:128; M.Keita 1945). Akan miners, too, believed that it was necessary to propitiate the spirits of the ancestors or of a mine before it was dug. Priests offered sacrifices, libations and prayers, repeating them if a mine was desecrated or if it proved unproductive (Garrard 1980:138; see also Herbert 1984:34, 39). Some traditions even mention human sacrifices in the past (Krieger 1963:319). Sexual abstinence was also required of male miners in many areas (sources do not mention any corresponding requirement for women) (see, for example, Kita 1985:87; Martinelli 1982:39; Evans-Pritchard 1967:31; Yandia, pers. comm.). In some cases, too, men whose wives were pregnant or menstruating were banned from mines (e.g. Kita 1985:87). In a rare switch, there is a report from a region in northern Nigeria where women did all the mining and men watching them were seized and fined (Meek 1925:1/149–151). Sometimes male miners had to work naked or in leather shorts (Krieger 1963:319; Herbert 1993:93); sometimes whistling was forbidden and dogs had to be kept away (M.Keita 1945; F.Keita 1945). Certain days were propitious, others to be avoided (Garrard 1980:138). A functionalist interpretation of these rituals and practices—which are certainly not unique to Africa—would simply underscore the fact that mining is dangerous. Even shallow pits often caved in; deep shafts must have been even more perilous, where light was dim and air foul. Mines in which people were killed were often abandoned, no matter how rich. Mining was also capricious, yielding fantastic wealth one day, sterile sand or rock another. The rituals were intended, like all rituals, to protect and to encourage a favorable outcome by
Mining as microcosm in precolonial sub-Saharan Africa
151
appeasing the djinns, spirits, ancestors, etc., all of whom were part of the world of the miner. But such an interpretation does not do justice to the vividness or detail of beliefs found in particular areas, beliefs especially well documented for gold mining. We will probably never know, for example, why, in the region of Siguiri, in modern Guinea, it was believed that if a miner brought balls of black soap or sticks of wood to the work-site, the gold would disappear, or why bird songs, a column of ants, or a chameleon were considered auspicious signs (M.Keita 1945). Gold was seen not as inert but as a living substance (Curtin 1975:204). In some legends, it was linked to a ram with a fleece of gold (F.Keita 1945), in others to a serpent, in still others to a plant that grew in the ground nourished by the rain. The tenth-century Arab writer Ibn al-Fakih claimed that in the land of Ghana, on the southern fringes of the Sahara, gold was planted in the sand like carrots and gathered at daybreak (Cuoq 1985:54). Other Arab sources, later retailed by Europeans, referred to strange, dog-faced people who worked the fabled mines of the West African interior. Shy cannibals, they were afflicted with terrible sores that could be healed only with salt—hence their eagerness to trade gold for salt, an exchange which dominated the transSaharan trade for centuries but which seemed incomprehensible to goldobsessed outsiders.
CONCLUSION If the legends about gold and gold mining were most avidly collected by Arab and European travellers, the association of metals with the earth and mining with agriculture appears to have been general. It is hardly surprising, then, that agriculture provided the model for both the social organization and what we may call the cosmology of mining. Mining was interdigitated into the agricultural calendar, governed by similar divisions of labor by gender and age, and permeated with the same understanding of how the world works. Just as the rains preceded the preparation of ground for planting, so the rains seemed to replenish streams with grains of gold and magnetite. Both the winnowing of river sands and the pounding of denser hematite ores with mortar and pestle were women’s work, closely analogous to their role in food production and preparation. Even the miner’s pick was ‘very probably adapted and developed from the agricultural hoe’ (Phimister 1976:11). Here, Phimister is referring to picks used by the Shona in gold mining, but his comment seems to have larger validity. Like agriculture, the outcome of mining was unpredictable: cave-ins were frequent and success elusive. While technical skills—the ability, for example, to recognize tell-tale signs of mineralization and to dig and reinforce shafts— counted for a great deal, manipulation of what we would call the supernatural world was equally necessary. This demanded the services either of a metallurgical specialist or of a priest who supervised ritual actions. But these
152
Eugenia W.Herbert
rituals were similar in form to those applied to other major activities in an attempt to control outcomes; as far as I can ascertain, there were no rituals unique to mining. Their aim was to placate local spirits, deities, djinns and lineage ancestors, all of whom determined human fortunes, whatever the sphere. In theory, if hardly in practice, the world was entirely rational, knowable and controllable. In his celebrated Forgerons et alchimistes (1956), Mircea Eliade claimed that male miners extract minerals from the womb of the Earth, the ‘Terre-Mère,’ in a reenactment or usurpation of the birth ritual, but one must be cautious about applying such a sweeping, universal model to African mining. In cases where mining was restricted to males and accompanied by sexual taboos, the exclusion of women or specifically of menstruating women—that is, where the operation was sexualized—Eliade’s model may have relevance. The real difficulty is that, unlike smelting or forging, mining was not a single operation; it came in many different forms in Africa. Not only must we reckon with a multiplicity of cultures, constantly in the process of change, but also with an activity that ranged from panning in rain-swollen streams to shallow trenches to pits and deep shafts. In addition, the metals that were the object of mining—iron, copper and gold—played quite different roles in African economies and social systems. It is tempting to speculate that the situation is more confused precisely because women and children were so commonly involved at least in some aspects of mining, while smelting and smithing were exclusively male prerogatives. In fact, preindustrial mining involved a much larger labor force in general. Where miners labored collectively in trenches or pits, the evidence points to control by senior men. But where they panned or gathered as individuals, they may have escaped the kind of control, technical and ritual, that empowered male smelters and smiths and gave them their often legendary status. Even here there are echoes of agricultural systems that combined both collective labor and individual plots, labor for the family unit and labor on one’s own account.
NOTE 1
Manyika, in present-day Zimbabwe, seems to have been an exception. Here panning was possible year round because many rivers were both gold-carrying and perennial (Phimister 1976:25).
REFERENCES Appia, B. (1965) Les forgerons du Fouta-Djallon. Journal de la Société des Africanistes 5: 317– 352. Bellamy, C.V. (1904) A West African smelting house. Journal of the Iron and Steel Institute 56: 99–126.
Mining as microcosm in precolonial sub-Saharan Africa
153
Brown, J. (1971) Ironworking in southern Mbere. Mila (Institute of African Studies, University of Nairobi) 2: 37–50. Cline, W. (1937) Mining and Metallurgy in Negro Africa. Menasha, W I: George Banta. Cuoq, J.M. (1985) Recueil des sources arabes concernant l’Afrique Occidentale du VIIIe au XVIe siècle (Bilad al-Sudan). Paris: Éditions du Centre National de la Recherche Scientifique. Curtin, P.D. (1975) Economic Change in Precolonial Africa: Senegambia in the Era of the Slave Trade. Madison: University of Wisconsin Press. Dumett, R.E. (1979) Precolonial gold mining and the state in the Akan region: with a critique of the Terray hypothesis. Research in Economic Anthropology 2: 37–68. ——(1987) Precolonial gold mining in Wassa: innovation, specialization, linkages to the economy and to the state. In E.Schildkrout (ed.), The Golden Stool: Studies of the Asante Center and Periphery, vol. 65, pt. I, pp. 209–224. Anthropological Papers of the American Museum of Natural History, New York . ——(1998) Eldorado in West Africa: the Mining Frontier, African Labor and Colonial Capitalism in the Gold Coast, 1875–1900. Athens, OH: Ohio University Press. Eliade, M. (1956) Forgerons et alchimistes. Paris: Flammarion. Evans-Pritchard, E.E. (1967) Zande iron-working. Paideuma 13: 26–31. Garrard, T.F. (1980) Akan Weights and the Gold Trade. London: Longman. Goucher, C.L. and Herbert, E.W. (1996) The Blooms of Banjeli: technology and gender in west African iron making. In P.Schmidt (ed.), The Culture and Technology of African Iron Production, pp. 40–57. Gainesville: University of Florida Press. Hemptinne, Msgr de (1926) Les ‘mangeurs de cuivre’ du Katanga. Congo 7: 371–403. Herbert, E.W. (1984) Red Gold of Africa: Copper in Precolonial History and Culture. Madison, WI: University of Wisconsin Press. ——(1993) Iron, Gender, and Power: Rituals of Transformation in African Societies. Bloomington: Indiana University Press. ——(forthcoming) Kwanyama Iron-working. Iowa Studies in African Art. Keita, F. (1945) Autour des placers du Cercle de Siguiri. Notes Africaines 27: 16–18. Keita, M. (1945) A-propos des placers du Cercle de Siguiri. MS, Institut Fondamental d’Afrique Noire, Dakar. Kita, K.M. (1985) La technique traditionnelle de la métallurgie du fer chez les Balega de Pangi (Zaïre). Muntu 3: 85–100. Krieger, K. (1963) Notizen zur Eisengewinnung der Hausa. Zeitschrift für Ethnologie 88: 318–331. Martinelli, B. (1982) Métallurgistes Bassari: techniques et formation sociale. Lomé, Togo. Meek, C.K. (1925) The Northern Tribes of Nigeria, 2 vols . London: Oxford University Press. Park, M. (1969 [1799]) Travels in Africa. New York: Dent. Phimister, I.R. (1976) Pre-colonial gold mining in southern Zambezia: a reassessment. African Social Research 21: 1–30. Schmidt, P.R. (1996a) Cultural representations of African iron production. In P.R. Schmidt (ed.), The Culture and Technology of African Iron Production, pp. 1–28. Gainesville: University of Florida Press. ——(1996b) Reconfiguring the Barongo: reproductive symbolism and reproduction among a work association of iron smelters. In P.R.Schmidt (ed.), The Culture and Technology of African Iron Production, pp. 74–127. Gainesville: University of Florida Press. Schofield, J.F. (1925) The ancient workings of south-east Africa. Native Affairs Department Annual (Salisbury) 3: 1–7. Summers, R. (1969) Ancient Mining in Rhodesia and Adjacent Areas. Salisbury: National Museums of Rhodesia.
154
Eugenia W.Herbert
Terray, E. (1974) Long-distance exchange and the formation of the state. Economy and Society 3: 315–345. Theal, G.M. (ed.) (1898–1903) Records of South Eastern Africa, 9 vols. Cape Town: C.Struik.
Chapter 10
A risky business Death, injury and religion in Cornish mining c. 1780–1870 John Rule
ABSTRACT By the early nineteenth century, the tin- and copper-miners of Cornwall made up a settled occupational community. Although rapid expansion of the workforce continued until a peak in the mid-1850s, there was by then little of the ‘frontier’ about the longsettled mining villages. On the contrary, their way of life was the outcome of generations of tradition-building and, in addition to the shared experience of mining as a distinctive activity, included other cultural characteristics, the most important of which was an attachment to the chapel culture of Methodist revivalism. In general the living standards of the miners and their families were not set by the working-class standards of the time, which were especially poor and depressed. Wages were generally good but individually subject to fluctuation which meant that occasional experience of material hardship was widespread. Even more common was hardship of another kind: Cornish metal mining was probably the most dangerous occupation of any pursued by significant numbers in nineteenth-century Britain. The toll of men, both young men through accident and older men through lung disease, was extreme. This chapter discusses some of the ways in which the mining community responded to this situation, in particular the role of religion in providing consolation and acceptance, while at the same time reinforcing its own claims to be the only means of understanding risk and tragedy. Analysis is made of the broadsheets which were produced after mining accidents, and which provided a ‘narrative’ of suffering, redemption and, sometimes, the working of providence, over a message of the need for perpetual spiritual readiness for death or bereavement. The role of funeral ceremonies and of fatalism as expressed in ‘black humour’ is also considered.
INTRODUCTION Around the tin and copper mines of Cornwall there had developed by the middle decades of the nineteenth century what was probably the largest longsettled metal-mining community in the world. It is also one of the bestdocumented mining communities of that era, with written sources of all kinds. The national population census provides basic demographic data from 1801, which were augmented from 1841 by additional statistics, including data on
156
John Rule
migration and on the distribution of occupations by age and gender. Several local weekly newspapers were in existence by the 1830s, and the interest they showed in the morbidity and mortality of the working miners was shared by a number of local doctors who published detailed investigations at the parish level. There were several major parliamentary inquiries between 1825 and 1864 which assembled hundreds of pages of testimony on working and living conditions in the mining community from a large and varied population of local witnesses. These written sources provide the main evidence for this study. They have been supplemented by others, including manuscript letters and autobiographical or other author-printed works, especially those deriving from Methodism, whose role is a central concern of this chapter. Aspects of the material culture of the mining community are frequently discussed in detail in the parliamentary inquiries, but little archaeological research has been done on the area’s later history. Visible reminders are hardly rare, from engine houses at the now disused mines to village chapels, now often converted to other uses, but a boom in industrial archaeology in the 1960s and 1970s concentrated almost exclusively on technical and engineering aspects of the mining industry and only marginally on social conditions. Several contributions to this volume stress the importance of women as a presence even in mushrooming ‘frontier’ mining communities. In the longsettled mining districts of Cornwall, women outnumbered men as a consequence of the high death rate among miners. Young women worked at the mines, but only as surface labour, breaking and sorting ores. In the Cornish mines there was no use of female labour underground. Occupational data sorted by gender and by age in the census tables of the mid-nineteenth century indicate that women usually ceased mine work on marriage (Rule 1971). They married young and were frequently pregnant at marriage. In the mining township of Camborne, calculation from the parish registers for the last quarter of the eighteenth century reveals that 45% of traceable first baptisms took place within eight and a half months of marriage (Rule 1986:197). Although the central purpose of this chapter is to examine the high mortality and injury risks which daily faced the underground male workforce, it is clear that women too must be its subject. Young marriage and large families aggravated the material impact of the death or disablement of ‘breadwinners’. If the folk religion, which Methodism had become, provided for the miners a quietist acceptance, just as importantly it provided for the bereaved a culture of consolation.
CORNISH MINING: BACKGROUND Cornish tin production has a very long history, dating from pre-Roman times. The industry was protected and privileged by the Crown through the Middle Ages, although only from the sixteenth century did mining replace streaming as
Death, injury and religion in Cornish mining
157
the predominant technology (Hamilton Jenkin 1962:38–39; Lewis 1908). First arousing interest towards the end of the sixteenth century, copper came to exceed tin in importance during the county’s boom years from 1740 to 1860; it was produced only from deep mining. By 1800 together the two metals already employed around 10,000 people. The underground labour force was male, but copper in particular also employed numbers of surface workers breaking and sorting ores: these were mostly young women and children. In 1851 total employment was 36,287. Decline set in by 1861 and was rapid after 1880 (Rule 1971). Large numbers emigrated and took their by then famous mining skills around the world (Rowe 1974; Payton 1984; Burke 1985). In Cornwall by the mid-nineteenth century, settled mining communities had existed through several generations. Although the stereotyping of miners as ‘a race apart’ has been overdone (Levine and Wrightson 1991:275–278), these communities had particular structural characteristics, such as a very high male death rate, and distinctive cultural ones, such as the chapel culture deriving from a strong attachment to Methodism. In the extensive literature on the middle decades of the nineteenth century, words like ‘risk’, ‘chance’, ‘speculation’ or even ‘lottery’ appear frequently. Shares in Cornish mines—proven ones, promising ones and prospective ones— attracted great interest. Cornish mining involved a multiplicity of shareholders who, along with those who historically chanced their capital on the high seas, were known as ‘adventurers’. Mining booms do not occur simply because there is a shortage of other investment opportunities. The element of excitement is usually present, for metal mining is an industry which must at one and the same time produce and prospect. South Caradon commenced in 1835. Its original issue of £1 shares attracted little interest, but within a year they had reached £2,000 (Barton 1964: 32–39). Such returns, however, were not usual. Even adventuring in an established mine with a previous good history was not safe. All business ventures have a history, but in mining it is essentially a natural one. Higher lodes might prove profitable, but lower ones disappointing. Alternatively, the best-paying levels might prove to be deep down. Rewarding as a mine might ultimately become, it could prove too expensive for some to stay in through the financial calls of a long, hard and deep sinking. West Caradon, an enterprise stimulated by the success of South Caradon, produced good ore in its higher levels in the 1850s and early 1860s, and its shares peaked at £90; by 1868, however, they were down to £6. Cook’s Kitchen, near Redruth, was one of the largest single enterprises in the country, already employing 1,000 people by the 1790s. In its early years it had raised ore to the amount of an incredible £2,000,000, but from 1810 it did not manage to pay a dividend for fifty-one years. In 1837 the copper mines of Cornwall employed 27,028 persons across 160 mines. However, less than 12% worked in mines employing fewer than 100 people; more than 60% worked in mines employing more than 250, and 31% in
158
John Rule
the five largest concerns, which each employed more than 1,000 (Lemon 1838). Cornish mining had by then been a highly capitalized industry for more than a century. But corporate mining forms still retain some prospector aspects (Hardesty 1988:102): they do not make, but discover and remove.
RISKS OF MINING An unusual feature of the mining economy of Cornwall was the sharing of prospecting risk by capital and labour. The miners who raised the ore were paid under the tribute system. Tributers formed small groups, perhaps only two men with a boy; they undertook measured portions of the mine (known as ‘pitches’) for an agreed share of the price fetched by the ore they raised. Contracts usually ran for two months and were expressed in terms of a rate in the pound, which ranged from a shilling or less (5p) to as much as 15s. This variation was due to the fact that pitches contained ores of different quality and in different amounts. Tributers had to judge the rate at which good wages could be made from a particular pitch on offer. If the ore was large in quantity or high in quality, this could be done at a lower rate in the pound. The prospective and speculative elements are evident. The selling price of the copper or tin was subject to fluctuation in a market beyond the miner’s influence. Luck played a part in whether lodes turned out better or worse than they had promised. The skill and judgement of the tributer were critical, for the pitches were not allocated by lot, or distributed by management. Instead they were bid for at a form of auction, and were taken by the tributers offering to work at the lowest rate (Burt 1969). The system was acclaimed by those nurtured in the axioms of nineteenthcentury political economy. It promised a high degree of task application because the miners were paid by results; and it associated labour with capital in the risks of enterprise. It naturally regulated the price of labour by supply and demand, for when work was scarce, competition was fierce, and miners in effect bid down their own wages. Certainly contemporaries noted that involving a miner in periodic competition with his comrades was hardly propitious for collective action, and contributed to the weak development of trade unionism in the mines (Rule 1992). The skilled miner was offered not constant wages but the chance of a windfall and the risk of poverty. However experienced the miners, the unexpected could happen in one direction or the other: [W]ithin the last six or seven days two poor men were working in a part of the mine where the lode was very hard and poor. They suddenly cut into a bunch of rich copper ore, very soft and consequently easy to break. The poor fellows having by their contract 12s [60p] out of the pound… Their time expires at the end of the month and if the lode continues so good…I
Death, injury and religion in Cornish mining
159
expect they will get £100 each for themselves for two weeks labour, and that same piece of ground will be taken next month for less than one shilling in the pound. (Jenkin MS 24 November 1804) Such good fortune might happen only once or twice in a working life, but it helps explain the attachment of the miners to a system which often denied them a living wage and sometimes left them with nothing. The adventurers did not mind the occasional paying out of large sums, for when that did happen, it served to ‘animate all the others who increase their exertion in the hopes of similar discovery’ (Burt 1969:25). They also encouraged competition at the next bidding and helped to bring neglected parts of the mine into operation. In 1805 after a spectacular gain, a pitch was taken for only 5d (2p) in the pound, when the manager’s own estimation was that at least a shilling (5p) was needed to make adequate earnings, but ‘there was no stopping them in the survey so eager were they to have that pitch’ (Jenkin MS 25 February 1805). The tribute system meant that from time to time some miners and their families would suffer deprivation. Such reversals of fortune could, in a religious community, find themselves attributed to a greater power. When a Methodist miner had a good run in 1853, he knew who to thank: ‘the Lord has greatly prospered me at the mine lately’, while for another it was ‘Providence’ which had blessed his labours (Tyack 1866:54–55; Edwards 1955:38). When times were otherwise, hunger for the mining family linked a matter of risk to one of health. But working underground had more direct effects on the health of the miners. Official figures reveal Cornish mining to have been the unhealthiest of employments. Mortality was more than two and a half times that of coal miners and half as much again as that for seamen (Hunt 1981:43). Just as revealing is a comparison between miners and non-mining males in Cornwall itself. From a sample of six mining parishes in 1849–1853, if the latter group (non-miners) is indexed at 100, then the respective age cohort figures for miners’ mortality are as given in Table 10.1. Two significant things emerge (a) mortality among the youngest age group was significantly higher for miners, and, (b) after parity in the age range 25–35, miners’ mortality began increasingly to exceed that of non-miners, especially after age 45. The death rate for miners in middle age was more than double that for non-miners. The heavier mortality between Table 10.1 Mortality of Cornish tin-miners, 1849–1853, according to age cohort
Source: British Parliamentary Papers 1864:x-xi
160
John Rule
15 and 25 is largely explained by the risks of mining, i.e. by the accident rate, which also crippled or blinded even larger numbers. Deaths and sickness in middle age are for the most part differently explained. Older miners were killed or maimed often enough, but there was a general consensus that their greater caution, patience and experience made them less vulnerable. What explains their heavier mortality is not risk, but inevitability: the impact of years of working in oxygen-deficient underground levels, breathing dust, the exhaustion of hard work in wet and hot places, and in particular the long ladder climbs back to the surface at the end of each shift. When miners answered the questioning of a parliamentary inquiry in 1864 by declaring, ‘You cannot expect miners to live as long as other men’, and when another said of the ‘old miners’, ‘They dwindle down, they waste away in time’, they were in effect stating the simple facts of life and death as they knew and experienced them. The conclusion of the report was, in its official way, as clear: At a comparatively early age the miners almost invariably exhibit in their features and persons the unmistakable signs of debilitated constitutions. Their faces are sallow, they have an anxious expression of countenance, and their bodies are thin. At the border of middle-age, or soon after their health begins to fail, the maturity and confirmed strength of that time of life seems to be denied to them. (British Parliamentary Papers 1864:viii, 6, 18) The Poor Law Relieving Officer for the mining parish of Gwennap reported in 1847 that of 240 families receiving payments, 200 were miners’ families. Of the fathers, 15 had been killed outright in the mines; 40 had been blinded or otherwise maimed to the point of being unable to earn a living; 63 had died after lingering for various periods after serious injury; and 15 others who had between them 80 children were too ill from progressing lung disease to continue working. A local doctor estimated that in that parish one miner in five would be killed at work (Burt 1984:186).
MINING DISASTERS AND DISEASE Large-scale disasters were less common than in coal mines. Metal mines did not suffer the dreaded ‘firedamp’ produced by the explosive gases naturally occurring in the coal itself. There were exceptions like the flooding of West Wheal Rose in 1846, which drowned more than forty men, but the toll was generally the aggregation of accidents to individuals or, at times, a few men working together. These resulted from premature explosions of gunpowder; plunging falls from ladders or faulty platforms; rockfalls; or incidents involving the lifting or pumping machinery. Deaths from accidents accounted for 25% of miners dying between the ages of 20 and 30 in the four central mining poor law
Death, injury and religion in Cornish mining
161
districts between 1837 and 1847; the contrasting figure for deaths of non-miners was only 6% (Burt 1984:186). Newspaper reports suggest a link between the accident rate in the mines and youthful inexperience or carelessness. For example, in a single issue of the weekly West Briton in 1853, reports are included of three inquests on young miners. One was killed when he went to the underground smith’s to have his pick sharpened, and a spark ignited two sticks of gunpowder he was keeping dry beneath his shirt. Another fell to his death when climbing over the shaft casing to save time, while the third was killed when he put his head over the shaft just as the engine bob was coming up. Some died very young: William Lawry was just 13 when he was blasted to death in 1848. He had been working with an older brother preparing a hole for blasting and had a can of gunpowder under one arm and a lighted candle in the other: the lid of the can fell off when he bent to pick up a cartridge. Henry Turner was only 17 when he fell to his death from a ladder when a spark fell on his hand from a lighted length of rope he was carrying (West Briton: 1 April 1853; 18 February 1848; 2 March 1855). But it was not always a matter of youth. Familiarity can breed contempt. As a mine agent remarked in 1801, ‘Men always in the habit of working [with gunpowder] seem to lose in a great measure a sense of their danger—so that many lives are lost, more than, I believe, would have been had due care always been taken’ (Jenkin MS 27 February 1801). Ingrained work practices contributed too. A mine captain told the accident inquiry of 1835 how difficult it had been to persuade miners to reduce the chance of striking a spark by substituting copper needles for the iron ones they customarily used to tamp-in explosive charges. The copper needles were offered at the same price as the iron ones, but it was only when the mine ruled that for those who were involved in accidents using the latter, no payment from the accident fund would be made, that they were more widely adopted (British Parliamentary Papers 1835:13–14). Of the greater toll taken by lung disease (phthisis), much could have been prevented. While accepting that conditions had improved with assisted ventilation over the last thirty years, a major investigation in 1864 still found that only 17 of 142 samples contained the normal proportion of oxygen, while 87 were condemned as extremely bad (British Parliamentary Papers 1864:xvi). A mine captain in 1835 had defined an ill-ventilated place as one in which a candle would not burn (British Parliamentary Papers 1835:12). In fact candles could burn where the air was already harmful. Nor was the test always applied. The miners were paid according to the value of the ore they raised, and the best ore did not necessarily coincide with the best air. Miners working on good ore at the level end were not inclined to give it up: ‘I have worked on tribute where I had the candle perhaps eight feet behind me, as it would not burn where I was working, and we were doing very well, and you do not think about the injury then’ (British Parliamentary Papers 1864:35).
162
John Rule
The annual average death rate of Cornish miners (per thousand miners) between 1849 and 1853 exceeded that of northern coal-miners in every age cohort, but was most noticeable in the age ranges 45–55 and 55–65 years. In the former group the Cornish rate was 33.51 per thousand, while that of northern coal-miners was 16.81 per thousand. In the second group Cornish miners suffered a mortality rate of 63.17 per thousand against a northern one of 24.43. Not surprisingly, the proportion of widows to total female population in Cornwall in 1851 was higher than in any other county (Burt 1984:184).
THE MINING COMMUNITY AND METHODISM How far can these unhappy statistics be linked to forms of behaviour and attitudes among the mining community, other than through reference to careless habits or some generalized concept of fatalism? Popular religion may be one link. Statistics of membership and attendance reveal Cornwall to have been the most Methodist county in England, and within it membership and influence were strongly concentrated in the mining communities. A report written in 1842 concluded that in these communities adherents of Methodism were ‘so numerous, that its qualities become prominent features of the whole body, when it is compared with other communities’ (British Parliamentary Papers 1842:760) Wesleyan Methodism in west Cornwall did not follow the national tendency of becoming a respectable ‘middle-class’ Victorian church, losing proletarian adherents to more democratic and revivalist splinter sects. In the religious census of 1851, the parent Wesleyan body in Cornwall had an attendance index equal to 20.5% of the county’s population out of a figure of 32% for all Methodist denominations. Analysis of baptism lists reveals no marked occupational pattern sharply distinguishing the parent body from those generally considered to be more proletarian offshoots, such as the Primitive Methodists. While it is true that middle-class Cornish Methodists were more likely to be Wesleyan, it is just as true that Wesleyan miners massively outnumbered those from other divisions. The recurrent revivalism found among the Cornish miners was a feature which happened across Methodism; it was not a function of any particular segment. Jabez Bunting, the conservative head of the Wesleyan movement in the early Victorian era, reputedly described the Cornish as ‘the mob of Methodism’ (Shaw 1967:81–82). It was noisy, enthusiastic revivalism, not political radicalism, which he had in mind. The independence of the Methodist communities of west Cornwall had long been notorious. John Wesley noted in 1788 that some at Redruth ‘were apt to despise and very willing to govern the preachers’ (Shaw 1967:78). It was a continuing characteristic. A minister remarked in 1908, ‘In the town…the minister represents the Connexion and takes over the chapel as his own…. In the village the chapel belongs to the people…and the people take over the minister as a newcomer attached to them’ (Currie 1968:30–31).
Death, injury and religion in Cornish mining
163
If Cornish Methodism was controlled by local laymen, preachers and class leaders to an extent unparalleled elsewhere, this is explained partly by the county’s geographical and cultural remoteness and partly by the special affinity of many aspects of Methodism to a traditional culture. The relative poverty of the members had the effect of limiting the number of official itinerant ministers who could be supported. In 1830, for example, Redruth had a ratio of members per minister of 658 and St Agnes of 929, against a national circuit average of 333. Whatever the pretensions of the London-based governing Conference, we can, following Ward (1972:245–246), represent Cornish Methodism as indigenized to the extent that it was a true Volkskirche (Luker 1986:605–610). Revivals were intense periods of mass conversions when large numbers were added to the Methodist societies. Phenomenologically they were mass displays of the conversion characteristics of sighs, groans, weepings and fits and, Methodism’s adherents hoped, the eventual joy of salvation. In Cornwall widespread revivals have been noted for 1764, 1782, 1798, 1814, 1824, 1831– 1833, 1841–1842, 1848–1849 and a period between 1862 and 1865. Up to 1814, the year of the so-called ‘Great Revival’, such meetings were separated by roughly sixteen years; thereafter they became somewhat more frequent. The usual features of Cornish revivals were remarkably consistent over time. Scenes of mass hysteria were common. Two descriptions from the revival of 1831– 1832 can illustrate this: The loud and piercing cries of the broken hearted penitents drowned the voice of prayer, and all that could be done at this stage of the meeting was to stand still and see the salvation of God. At length the penitents were conducted and upheld, each of them by two persons, into one part of the chapel. And now when their cries and groans were concentrated, one of the most affecting scenes appeared before the people. Their humble wailings pierced the skies. Sometimes a burst of praise from the pardoned penitents mingled with the loud cries of the broken hearted; and this greatly encouraged those that were in distress. The chapel was crowded almost to suffocation. The steam ran down the walls; the gallery stairs were flooded with it; and had not all the windows been opened, every light would have gone out. The people were yet determined to enter into this kingdom; yet everyone suppressed his feelings as long as he could. When any have cried out, it has been the involuntary burst of a soul overwhelmed with a sense of guilt and a sense of danger. The seats in many places were literally covered with tears. This night the preacher concluded the first service about eight o’clock; but scarcely any left the chapel. From that hour until nearly eleven the people were crying out for mercy in every part of the chapel, both in the gallery and below. Nearly fifty found peace that night; and more than double that number returned home weary and heavy laden. Many of these continued in prayer all night; and not a few during the night found peace in their own houses. (Treffry 1837:84–85, 91)
164
John Rule
Figure 10.1 Cornish humour illustrated in the cartoons of Oswald Pryor Source: P.Payton (1984) The Cornish Miner in Australia: Cousin Jack Down Under. Redruth: Dyllansow Truran. Figures 10.1, 10.2 and 10.3 reproduced by kind permission of Lindsay Pryor
Death, injury and religion in Cornish mining
Figure 10.2 Cornish humour illustrated in the cartoons of Oswald Pryor
165
166
John Rule
Figure 10.3 Cornish humour illustrated in the cartoons of Oswald Pryor
Death, injury and religion in Cornish mining
167
It was not only in the chapels that miners were overcome: ‘in the bowels of the earth…the spirit strives with them. They come up from their work and explain to…the captain…we cannot work, we must cry for mercy!’ (FletcherTooth MSS 30 March 1831). In the ‘Great Revival’ of 1814, many were said to have believed that the revival was preparatory to the end of the world (Anon. 1814:5–6; Smith 1888:551): [I]t continued for a few weeks, when the great noise subsided, but the fire still existed. The same extraordinary agitation extended to divers neighbouring parishes. But the great current seemed to run westward. It was not so very rapid or noisy about Truro and its vicinity, as it was in the parish of Gwennap. At Illogan and Camborne it was more violent. But at Hayle Copper works and the adjoining villages (taking in Wheal Alfred mine) the agitation was extremely so—at the latter mine are about 800 labourers at the surface (male and female), chiefly young people, where the torrent bore down everything that stood in its way. Were I to attempt to describe it I could not find words sufficient to draw in colours strong enough. All labour for some days was suspended, and the underground labourers (equally as numerous as those at the surface) seemed to be struck with the same power —but being more advanced in years I think they appeared to have a greater mastery over their passions than the others had. (Hamilton Jenkin 1951:179) Explaining Cornish revivalism in terms of ‘frontier’ mentalities derived from the ‘burned-over district’ experiences of North America does not seem applicable. For although elsewhere newly upheaved industrial and mining communities characterized by social dislocation may well have been susceptible to emotional and enthusiastic forms of religion, the mining villages of west Cornwall were by the nineteenth century long-standing industrial communities, characterized by established traditions and with marked cultural and social cohesion. So far as the coal-miners of northeastern England are concerned, Colls has suggested that some outbreaks of revivalism can be linked to the psychological impact of colliery disasters (Colls 1987:163). In Cornwall, however, major disasters were rare while revivals were regular. I have argued elsewhere (Rule 1971) that Cornish revivalism was an internally driven phenomenon deriving from social imperatives within the local religious community itself. The evidence is overwhelming that youth was the most likely point in the life cycle for an individual to experience conversion. The periodicity of revivalism meant that a major revival would occur during this period of everyone’s life. A revival could therefore function as a means of joining young people to the Church, and to a lesser extent of rejoining older ones. This does not necessarily imply that revivals were deliberately staged with this end in view, but rather that they would perform this role, deliberately or not. Once a revival had broken out, the neighbouring societies were naturally put into a state of expectation that they
168
John Rule
would soon experience it for themselves, or else a state of anxiety that they would not. Either way the necessary emotional state was attained. A miner later recalled of the revival of 1856, ‘hearing of revivals around us, and feeling the want of one for ourselves and the people, we agreed three weeks ago, to have special services’ (Tyack 1866:81). Revival effects were in part produced by a style of preaching which could amount to mental terrorism: ‘Stout hearted men from the mines come to the prayer meetings and with tears streaming over their faces and in great agony they fall down on their knees and wrestle till they have found mercy’ (FletcherTooth: 30 March 1831). Contemporaries made no bones over this. Methodism’s official historian, a Cornishman, wrote that the awakened sinner ‘trembling in agony’ beneath God’s frown could hardly be expected to control his emotions. During the 1814 revival some were said to have been ‘convinced that they are sinful, perishing, helpless creatures’ and to have been ‘driven to the throne of mercy by terror’ (Smith 1888:604, 607; Anon. 1814:12). If widespread revivals were exciting and spaced events, some of their features were a usual feature of worship in Cornish mining communities. A minister recalled, ‘In their meetings shouting and ecstasies were rather the rule than the exception…they let themselves go; they shouted, they wept, they groaned, they not seldom laughed aloud with a laugh of intense excitement’ (Rigg 1904:67–68).
DISCUSSION Religious excitement was endemic and periodically epidemic in the mining communities, but can it be linked to the dangers of the mine and the constant need for consolation? The special need for the miners to be prepared for death is a recurrent theme of many forms of contemporary Methodist literature. Acceptance of their lot brought the miners much approval, from those who did not have to share it: ‘Nothing can indeed be more admirable than the cheerful confidence with which, in trust of a future life, the miner contemplates that termination, often an early one, of his labours’ (British Parliamentary Papers 1842:760). There was also a political dimension: No rankling ill-will to those whom Providence had placed in easier circumstances arose in their minds. No hatred to their employers, or rebellious thoughts against the government of the country, for a moment found a place in their hearts. They laboured, they sorrowed, they suffered, but they patiently endured. (Cornish Banner 1846:22) More pertinent here, perhaps, is another article in the same evangelical periodical, which drew lessons from the death of forty miners in the flooding of East Wheal Rose in 1846:
Death, injury and religion in Cornish mining
169
We have mentioned the terror and dismay of those who were not prepared to die. But there were other and very different scenes which we rejoice to chronicle… [among the victims] were not a few, belonging to various Christian denominations, who had realised the power of godliness, and some of whom had been useful labourers in the cause of Christ. Although taken by surprise, they were not unprepared. Two…met in one of the levels, and conversed a moment on their awful situation. As they bid each other farewell, one of them exclaimed: ‘I hope to meet you soon in heaven’. ‘Glory be to God!’ was the reply…One [of those who died] was heard to exclaim as the waters rose about him, and just before he was drowned: ‘All is well! all is well!’ The article concluded that among the 1,260 men employed at the mine were ‘many of a very wicked and abandoned character; and by many persons on the spot, the visitation is looked upon in the light of a judgement’. Certainly there is an appeal to the whole community, and especially to all who are similarly employed. ‘Be ye also ready; for in such an hour as ye think not, The Son of man cometh’ (Cornish Banner 1846:55–56). Working miners and their families did not generally read such homilies in middle-class periodicals. They did, however, receive them in another form: in verse from one of the broadsheets which were commonly produced after mine accidents. These follow a typical narrative and didactic form. The scene is set and the incident and its aftermath are described in great detail. The victims are named, and the painful consequences for their families described: Young Peter Eddy’s head was gone, Upon the skip he lay The sollar struck him as they passed, And took his head away Yet further down were others found, Thomas and Richard Wall, Father and son—an awful sight— Lay in that dreadful hole Thomas Nankervis lay to death An awful sight to view And Michael Nicolas by his side, Both bruised and mangled too. Each one was injur’d fearfully, Bruis’d, broken, smashed and dead; A sickening spectacle! for some Had lost part of the head.
170
John Rule *********** Near twenty children now are left Without a father’s care; Three widows for their husbands mourn In sorrow and despair The widow’s mother too is left, Her only son is gone; His earnings did her wants provide, Nought else to live upon
Typically, the closing verses carry the message: An accident, or sudden death, Above or under ground, May call thee, reader, but how blest If waiting thou art found. Then mourn not, weeping friends, for those That suddenly are gone; In Heaven they from their labours rest, Before the great white throne. Prepare! prepare! to follow them, While yet it’s called today! Lest unprepar’d, grim Death should come, And summon you away. (Noall and Payton 1989:186–187) This particular broadsheet followed a tragedy at Botallack Mine in 1863, but the form is general. Another, describing an accident through a cave-in in 1859, concludes: May miners one and all prepare Lest by some accident, They may be hurried from this life, With no time to repent. After seven lives were lost in 1858, another concludes (original broadsheets in Cornwall County Record Office): Great God, in mercy stay thine hand, May this a warning be, May sinners need no other call, To bring them unto thee.
Death, injury and religion in Cornish mining
171
May they but think of those who lie, Deep in Porkellis mine, So many fathoms down below Cut off just in their prime. To call such narratives of suffering a ‘celebration of death’ is perhaps unhelpful, but the mining communities certainly celebrated funerals: Then borne unto the silent tomb, What thousands gather’d there, To mingle with the relatives, Sorrow’s lingering tear. That such numbers attended funerals is well supported from other evidence. A correspondent to the Mining Journal described the funeral of one miner killed in a rock fall in the late 1850s. A ‘black mass’ of ‘several thousands’ of people processed several miles to the church. It was headed by the Methodist minister, who gave out the verses of a hymn, the singing of which was then accomplished by the chapel choirs of the district, about eighty persons in number, after whom came the representatives of the mine, the captains and agents. Then came the coffin, borne in the traditional underarm way by relays of bearers. Behind the body followed the widow with ‘a long string of relations’ and behind them thousands of miners. At the church, matters were handed over to the clergyman. While the service was conducted, the crowd waited bare-headed until the moment of burial, before dispersing—many, it must be said, to the neighbouring pubs (Burt 1972:12–14). Cornish mining funerals have been described many times by historians of their communities abroad as well as at home. As well as the more sombre, heavy hymns of the Wesleys, there seem to have been some peculiarly local songs, the best known of which was known as the ‘burying tune’: ‘Sing from the chamber to the grave,’ I hear the dying miner say; ‘A sound of melody I crave Upon my burial day. ‘Sing sweetly whilst you travel on And keep the funeral slow; The angels sing where I am gone And you should sing below.’ (Payton 1984:7–8) Attending funerals was part of life in the mining villages. Unrelated people and those who were not especially close to the deceased presumably went for
172
John Rule
a variety of reasons, among which were socializing and entertainment. When the dead was the victim of a mine accident, there was a special sense of the need for the community to show its respect and its sympathy. There is a humorous saying in Cornwall, ‘If you don’t go to other folk’s funerals, then they won’t come to yours.’ Perhaps this links to another way in which oral traditions accommodate death and injury. Not all mining families were Methodists, and not all Methodists were too ‘serious’ to joke. There is a parallel tradition of black humour, which travelled with emigrating miners. At copper mines in South Australia, worked by expatriate Cornish from about 1840 to the 1920s, one became known locally as a cartoonist. Many of Oswald Pryor’s drawings illustrate this traditional humour (Pryor 1976; see Figures 10.1 – 10.3). Things did not improve for the smaller numbers still working in Cornish mines at the end of the nineteenth and the beginning of the twentieth century. The introduction of powered rock drills abridged labour, but only at the expense of more dust. Around 1900, rock-drillers were able to work for only about eight years before contracting phthisis. As Gillian Burke (1985) has remarked, ‘To work a rock drill in Cornwall was to face almost certain death.’ Over its long history, Cornish mining certainly created much wealth; at the same time, and just as certainly, it destroyed many a miner’s health. Exploitation is never an easy concept to define. Although their earnings were subject to fluctuation, Cornish miners were at most times among the relatively well-paid sections of the British working class, but the death and injury rates among men and the consequences these also had for women and children, including desperate poverty, surely amount to a form of exploitation. Methodism helped the community to come to terms with this situation, although alongside consolation it should be noted that its diversion of suffering from a cause of resentment and protest into a vehicle for personal redemption contributed to the reputation of Cornish miners by the midnineteenth century for industrial and political passivity (Rule 1992).
REFERENCES Anon. (1814) An Account of the Remarkable Revival of the Work of God in Cornwall. Dublin. Barton, D.B. (1964) A Historical Survey of the Mines and Mineral Railways of East Cornwall and West Devon. Truro: Bradford Barton. British Parliamentary Papers (1835) V Report of the Select Committee on Accidents in Mines. ——(1842) XVI Report of the Royal Commission on Child Employment. ——(1864) XXIV Report of the Commissioners Appointed to Inquire into the Condition of All Mines in Great Britain etc. Burke, G. (1985) Disease, labour migration and technological change: the case of the Cornish miners. In P.Weindling (ed.), The Social History of Occupational Health, pp. 78– 88. London: Croom Helm. Burt, R. (ed.) (1969) Cornish Mining: Essays on the Organisation of Cornish Mines and the Cornish Mining Economy. Newton Abbot: David & Charles.
Death, injury and religion in Cornish mining
173
——(1972) Cornwall’s Mines and Miners. Truro: Bradford Barton. ——(1984) The British Lead Mining Industry. Redruth: Dyllansow Truran. Colls, R. (1987) The Pitmen of the Northern Coalfield: Work, Culture and Protest 1790–1850. Manchester: Manchester University Press. Currie, R. (1968) Methodism Divided. London: Faber. Edwards, M.E. (1955) John Wesley and the Eighteenth Century. London: Epworth Press. Fletcher-Tooth Correspondence. Methodist Church Archives, John Rylands Library, Manchester. Hamilton Jenkin, A.K. (1951) Letters from Cornwall. Bath: Westaway. ——(1962) The Cornish Miner, 3rd edn. London: Allen & Unwin. Hardesty, D.L. (1988) The Archaeology of Mining and Miners: A View from the Silver State. Society for Historical Archaeology, Special Publication Series no. 6. Pleasant Hill, CA: SHA. Hunt, E.H. (1981) British Labour History 1815–1914. London: Weidenfeld & Nicolson. Jenkin, W.William Jenkin Manuscript. Letterbooks in Courtney Library, Cornwall County Museum, Truro. Lemon, C. (1838) The statistics of the copper mines of Cornwall. Journal of the Statistical Society of London 1: 65–84. Levine, D. and Wrightson, K. (1991) The Making of an Industrial Society: Whickham 1560– 1765. Oxford: Clarendon Press. Lewis, G.R. (1908) The Stannaries: A Study of the Medieval Tin Miners of Cornwall and Devon (reprint, 1965). Truro: Bradford Barton. Luker, D. (1986) Revivalism in theory and practice: the case of Cornish Methodism. Journal of Eccesiastical History 37(4): 603–619. Noall, C. and Payton, P. (1989) Cornish Mine Disasters. (Introduction by P.Payton.) Redruth: Dyllansow Truran. Payton, P. (1984) The Cornish Miner in Australia: Cousin Jack Down Under. Redruth: Dyllansow Truran. Pryor, O. (1976) Cornish Pasty: A Selection of Cartoons. Adelaide: Rigby. Rigg, J.H. (1904) Wesleyan Methodist Reminiscences Sixty Years Ago. London: Robert Culley. Rowe, J. (1974) The Hard-Rock Men: Cornish Immigrants and the North American Mining Frontier. Liverpool: Liverpool University Press. Rule, J.G. (1971) The labouring miner in Cornwall c. 1740 to 1 870 : a study in social history. Unpublished PhD thesis, University of Warwick. ——(1986) The Labouring Classes in Early Industrial England 1750–1850. London: Longman. ——(1992) A configuration of quietism? Attitudes towards trade unionism and Chartism among the Cornish miners. Tijdschrift voor Sociale Geschiedenis 18(2/3): 248–262. Shaw, T. (1967) History of Cornish Methodism. Truro: Bradford Barton. Smith, G. (1888) History of Wesleyan Methodism, vol. 3. London. Treffry, R. (1837) Memoir of Mr John Edwards Tresize. London. Tyack, W.D. (1866) The Miner of Perranzabuloe. London. Ward, W.R. (1972) The religion of the people and the problem of control. Studies in Church History 8: 237–257.
Chapter 11
Silver shackles and copper collars Race, class and labor In the Arizona mining industry from the eighteenth century until World War II Thomas E.Sheridan
ABSTRACT The lure of silver and gold made Arizona shimmer in the dreams of prospectors and speculators during the Spanish colonial and early Anglo-American territorial periods, but it was copper, not precious metals, that drove the Arizona mining industry, beginning in the 1880s. From then until World War II, Arizona was dominated by extractive industries—the Three Cs of Copper, Cattle, and Cotton—but Copper was clearly king. Because copper was an industrial commodity, enormous quantities of ore had to be dug and enormous capital investments in mines, smelters and railroad transport had to be made. Moreover, copper mining required a large and stable workforce. Most of the early workers came from northern Mexico, but soon the immigrant surge that was reshaping eastern industrial America brought Cornish, Irish, German, Spanish, Italian, Croatian, Serbian and Finnish miners to Arizona’s isolated copper towns. Those towns often became ethnic battlegrounds as Anglo-Irish miners and the labor unions that organized them strove to exclude Mexican and southern European workers from the best-paying jobs. Some mining communities were segregated into ‘white men’s’ or ‘Mexican’ camps. Others instituted a differential wage scale for Anglo-Irish and Mexican miners, paying Mexicans less for the same jobs. Copper companies pitted ethnic groups against one another to cripple organized labor and depress wages. In Arizona’s copper towns, idioms of race undercut idioms of class to weaken working-class solidarity and forge the ‘copper collar’ that enabled the companies to break the unions during World War I, when xenophobic fears of sabotage helped brand all strikers as Wobblies (members of the Industrial Workers of the World). Unions did not regain power until after World War II, when Mexican-American veterans spearheaded successful strikes in the copper towns.
INTRODUCTION To most casual observers, Arizona conjures up images of saguaro-studded deserts, Monument Valley or the Grand Canyon; of cowboys and Indians; and of open spaces. From the Spanish colonial period until World War II, however, mining dominated the political ecology of the region, even when
Race, class and labor In the Arizona mining industry
175
the lure of mineral wealth proved chimerical. This chapter focuses primarily upon the ways in which industrial copper mining structured relations of ethnicity and class in the isolated mining communities that developed on the Arizona frontier. It therefore articulates very well with the work of archaeologists and social historians who have investigated the material correlates and ideological manifestations of those idioms of identity and power in other mining communities across the globe. This study also addresses the changing political and economic nexuses between geographically isolated mining settlements and outside sources of capital and power such as corporations, labor unions and the state. International immigration streams and transcontinental railroad networks are just two of the most important and most visible ways in which the global intersected the local in Arizona’s mining towns. Highly capitalized and technologically sophisticated companies from the eastern U SA and the British Isles appropriated the Arizona political as well as physical landscape in response to an explosive international demand for copper. Those same companies then tapped an international labor market to mine and smelt their ore. The interplay among technology, capital and labor was unique to the southwestern U SA in some respects, particularly regarding the importance of Mexican labor. But in many other respects, it paralleled the trajectories of industrial mining communities in Australia, Canada and other parts of the USA where a colonial or neocolonial extractive order took shape, using divisions of race or ethnicity to undercut solidarities of class. And even though this chapter does not address directly the impact of changing technology, technological change had an immense effect upon relations of race and class in Arizona mining communities as well as upon Arizona’s physical and political landscapes. Those changes, of course, were global rather than local, affecting copper mining communities throughout the world. One final introductory note: even though this chapter concentrates upon industrial mining, other themes addressed by the volume resonate throughout Arizona mining history as well. Many of the earliest prospectors were Mexicans and Indians from the state of Sonora just south of Arizona. They, like some African miners, were farmers and small-scale stockraisers who searched for gold during the winter months after the summer planting season. As alluvial placer deposits played out, however, hard-rock gold and silver mining required greater capital investment and technological expertise. Preindustrial patterns became industrial. Prospectors became proletarians. And, as Arizona mining shifted from precious metals such as gold and silver to industrial metals such as copper because of technological advances, rising worldwide demand, and denser and more extensive railroad networks, Arizona’s isolated mining communities became bound by copper collars and iron rails to global commodity production and capital flows.
176
Thomas E.Sheridan
ARIZONA MINING: RACE AND CLASS During the extractive period in Arizona’s history—from the 1880s until World War II—boosters liked to brag that Arizona’s economy was dominated by the Three Cs: Copper, Cattle and Cotton. Of the three, Copper was clearly king. Arizona’s state house has a copper dome, and copper companies like Phelps Dodge exercised enormous political and economic power, breaking unions, manipulating territorial and state legislatures, writing state tax and county codes to their own advantage. In order to keep their labor costs down, the copper companies also pitted ethnic groups against one another to cripple organized labor and depress wages. In Arizona’s copper towns, idioms of race undercut idioms of class to weaken working-class solidarity. Later, idioms of ‘foreignness’ were employed to exacerbate xenophobic fears of sabotage during World War I, but those fears had strong racial, and particularly anti-Mexican, undertones. Because of Arizona’s proximity to the border and its reliance upon Mexican labor, ethnic conflict in Arizona’s copper communities usually revolved around management’s exploitation of, and organized labor’s profoundly ambivalent relationship to, Mexican workers (Sheridan 1986, 1995). The asymmetrical balance of power between management and labor in the copper industry was tipped even further by ethnic divisions among the miners and smelter workers themselves. Earlier mining activity both contrasted with and foreshadowed what historian James Byrkit calls ‘the forging of the copper collar’ (Byrkit 1982). Before transcontinental railroads and technological innovations allowed the extraction of low-grade copper ore, silver and gold made Arizona shimmer in the dreams of prospectors and speculators for 150 years. In a world where most of the good things in life were scarce, the promise of precious metals exercised a peculiarly powerful hold on the imagination. Europeans in the seventeenth and eighteenth centuries came to the Americas from a continent where periodic famines still stalked the peasantry and where necessary resources like land, water, firewood and wild game were often controlled by aristocratic elites. Prospecting, on the other hand, offered a glimmer of the great escape. Silver drew the Spaniards northward from Mesoamerica, and they leapfrogged from strike to strike, creating great mining centers like Zacatecas, Guanajuato, Parral and Alamos. And when a Yaqui Indian discovered the planchas de plata (slabs of silver) at Arizonac southwest of modern Nogales in 1736, silver gave Arizona both its name and its legend for fabulous mineral wealth (Officer 1987, 1991). The Spaniards gave Arizona miners not only a body of lore but also a repertoire of techniques perfected over four centuries in the deserts of northern Mexico. Placer mining (mining surface deposits rather than veins), dry washing (winnowing gold from heavier sand and gravel by blowing on it or tossing it in the air), the patio process (separating or ‘reducing’ silver ore from other minerals by mixing it with mercury in round arenas where animals or people trampled the sludge) —all were skills the Spaniards brought to the Arizona frontier.
Race, class and labor In the Arizona mining industry
177
Mining in the region, like stock raising, developed from a colonial base (Officer 1987, 1991). In many respects, that base consisted of preindustrial rather than industrial technologies. During the early territorial period, these simple and flexible methods satisfied the needs of most miners. Many of these early placeros (placer miners) were Sonorans who had left their homes for the California goldfields a decade earlier. Placer mining was a part of their heritage, a set of skills they absorbed along with farming and ranching as they grew up on the northern frontier. By the 1870s, however, silver had replaced gold as Arizona’s most valuable mineral. It was harder to reduce than gold and only one-sixteenth as valuable, but there was more of it. So most of the great strikes of the 1870s and 1880s were silver strikes. And even though prospectors could pick away at outcrops, most silver deposits required more capital than individual miners could muster. Hundreds of Anglo-American, British and Irish miners, many of them from gold and silver strikes in Nevada, started out searching for their own bonanzas and ended up working for wages in Arizona mines and mills. The days of the placero with his blanket (for winnowing) and batea (wooden bowl for panning) were gone. The transition from preindustrial to industrial technologies and from mercantile to industrial capitalism in the mining industry had come. This had a profound effect on the social organization of mining and metallurgy in Arizona. From the very beginning of the placer boom, Anglo prospectors in central Arizona—and I use the term ‘Anglo’ as an unsatisfying but convenient shorthand for English-speaking workers from the USA or the British Isles —tried to turn their diggings into ‘white men’s camps.’ In the Walker District along Lynx Creek south of Prescott, a miners’ meeting passed a resolution to exclude ‘Asiatics and Sonoranians…from working this district’ as early as 12 July 1863 (quoted in Sheridan 1995:151). But even though silver mines in places like Tombstone gave the Arizona territory its first spectacular mining fortunes and its Wild West notoriety, copper became the mineral that drove the Arizona mining industry. And because copper was an industrial commodity, not a precious metal, enormous quantities of ore had to be dug and enormous capital investments had to be made in mines, smelters and railroad transportation networks. The copper industry also required large and stable workforces—workforces that crowded into the tight, hilly quarters of mining communities that were surrounded, paradoxically, by hundreds of miles of desert or mountain. The copper towns were also ethnic battlegrounds. During the early years, most copper mines—like most other businesses in Arizona—depended on Mexican labor. By the late 1870s, for example, most of the several hundred men working in Clifton’s mines and smelters were Mexican. Because of Apache hostilities, however, it was difficult to attract laborers to the district, so mineowner Henry Lesinsky finally took his cue from the Southern Pacific Railroad and imported Chinese laborers. By 1883, 100 of the 400 miners in Clifton were Chinese. According to mine supervisor James Colquhoun, ‘if occasionally a few
178
Thomas E.Sheridan
were killed no questions were asked, and the work went on as usual’ (quoted in Sheridan 1995:169). Such callousness was common among employers of the Chinese. To Mexican and Anglo workers, on the other hand, the Chinese were intolerable rather than expendable. L.C.Hughes, editor of the Arizona Weekly Star, expressed the sentiments of many Anglos and Mexicans on 24 July 1879 when he called the Chinese ‘an ignorant, filthy, leprous horde.’ With racism fueling the already bitter competition for jobs, violence was inevitable. In Clifton there was so much terrorism and sabotage that Colquhoun and other supervisors were forced to dismiss the Chinese and hire Mexicans and Anglos in their place. During the early 1880s, Anglos and Mexicans joined together to drive the Chinese off the railroads and out of the mines. In Bisbee, Chinese were not even allowed to own businesses or remain overnight (Vaughan 1992). Anglo and Mexican miners then spent the next fifty years fighting one another. With the conquest of the Apaches and the construction of transcontinental railroads during the 1880s, copper-mining districts like Bisbee, Globe-Miami and Clifton-Morenci flourished as corporations from the eastern USA and the British Isles invested millions and the demand for labor skyrocketed. Experienced miners with hard-rock skills came to Arizona’s young copper communities from established mining areas in Germany, Ireland and Cornwall as well as the USA. Then, as technological innovations such as the pneumatic drill reduced the need for traditional hard-rock mining skills, Spanish, Italian, Croatian, Serbian and Finnish immigrants flocked to towns like Bisbee and Jerome. Copper towns became miniature models of industrial America, tiny colonies where the immigrant surge from Europe swirled into the Arizona territory. The rest of Arizona might speak English, Spanish or Apache, but in Bisbee or Globe the Babel that was reshaping cities like Cleveland and Chicago could often be heard (Byrkit 1982; Mellinger 1995; Sheridan 1995). Mexican immigration also swelled during the brutal reign of Porfirio Díaz and the carnage of the Mexican Revolution (García 1981). Companies frequently employed Mexicans to break strikes, including the bitter struggle in the Colorado coalfields organized by the Western Federation of Miners in 1903 and 1904. The presence of Mexican scabs infuriated Anglo workers, hardening their attitudes toward people they already mistrusted and misunderstood. Many organizers therefore ignored Mexican miners or turned them into scapegoats to win the support of Anglo workers. By the late nineteenth century, two other responses to Mexican labor had evolved as well. One was the segregation of mining communities into ‘white men’s’ and ‘Mexican’ camps. There were relatively few Mexican miners in the Globe-Miami mining district. Clifton-Morenci, in contrast, retained its predominantly Mexican labor force. What was going on was a complex and shifting process of both ethnic compression and ethnic differentiation. Cultural and religious differences among English-speaking Anglo-American, Irish, Cornish and Scottish miners were submerged, at least for political purposes, as
Race, class and labor In the Arizona mining industry
179
opposition to Mexican labor and, to a lesser extent, southern and eastern European labor hardened. Because ‘whiteness’ and a command of the English language translated into better-paying jobs, Irish Catholics found more common political cause with the traditionally anti-Irish Cornishmen, or ‘Cousin Jacks,’ than they did with their fellow Catholics from Mexico (Mellinger 1995). This segregation undoubtedly was reflected in the material culture of the mining communities. Unfortunately, so little historical archaeology has been carried out in Arizona’s copper towns that we do not know in a systematic fashion what the material correlates of those ethnic boundaries were. Differences of diet, house construction, household spatial organization and housewares may have reinforced those boundaries. Until historical archaeologists have the opportunity to excavate in established communities like Jerome, Clifton-Morenci or Bisbee, however, we have only the photographic and documentary record to consult (Schwantes 1992).
ARIZONA MINING: LABOR The other response to Mexican labor was a differential wage scale for Anglo and Mexican miners. During the 1890s, the lowest wage for Anglos in the Old Dominion and other Globe-Miami mines was $3.00 a day. In Clifton-Morenci, Mexicans received $1.75–$2.00 for a ten-hour shift. The difference in wages only increased after the turn of the century. By 1910, Anglo miners were making $4.00 a day or more in most Arizona copper towns, but Mexican wages leveled out at $2.00 a day and stayed there for several decades. No other figure reveals more dramatically the systematic subordination of Mexican labor (Park 1961; Sheridan 1986; Mellinger 1995). An example of such racial ‘exclusionism’ occurred in 1896, when the owners of the Old Dominion mine in Globe tried to weaken racial boundaries by lowering Anglo wages to Mexican levels. First, mine supervisor S.A.Parnell and his foreman Alexander McClain dropped wages for shovelers and car men from $3.00 a day to $2.25. Then they began replacing Anglo workers with Mexican nationals. Soon afterward, 200 miners held a meeting in the Globe courthouse and vowed to keep the community a ‘white man’s town.’ They marched to the home of Parnell and hauled him out into the street at gunpoint, throwing a rope over his head. Parnell fired McClain and his Mexican workers the next day. He himself resigned shortly thereafter (Park 1961). But Mexican workers were never passive pawns in this process of exclusion or exploitation. Mexican, Italian and Spanish miners carried out the next major strike in Arizona despite being ignored by the Western Federation of Miners (WFM) and other unions. In 1903 the twenty-second territorial legislature passed an act reducing the workday for underground miners from ten to eight hours with no cut in pay. The law went into effect on 1 June. The next day, mine-owners in Clifton-Morenci complied by shortening shifts but agreed to pay
180
Thomas E.Sheridan
their miners only for nine hours of work. The result was a 10% drop in the daily wage (Mellinger 1995). Supervisors in the district must have reasoned that since most of their workers were Mexicans or recent Italian immigrants, they would docilely accept the pay cut just as they had accepted the differential wage scale. But to the surprise of nearly everyone, several thousand Mexican, Italian and Spanish miners immediately went on strike. Because the WFM and other labor organizations refused to admit them, they coordinated the strike through their mutualistas, or mutual aid societies, and brought one of Arizona’s largest mining districts to a complete halt. By 3 June, Clifton-Morenci’s mines and smelters had shut down, throwing 3,500 men out of work. Contrary to the stereotype, Mexican miners, not the Western Federation of Miners, carried out the most significant strike in the Arizona territory before World War I (Mellinger 1995; Sheridan 1995). But nature and Arizona’s power structure conspired against them. On 6 June, Governor Alexander Brodie called out the Arizona Rangers, a paramilitary organization that had been created in 1901 to patrol the border and prevent cattle-rustling. Three days later, 2,000 miners defied the rangers and paraded through Morenci in a torrential downpour, many of them armed with rifles, pistols or knives. The determination of the strikers terrified state and county authorities. Brodie asked President Theodore Roosevelt for federal troops, so six companies of the National Guard marched into town. It was the largest military force assembled in Arizona since the Apache campaigns of the 1880s. By 12 June, Clifton-Morenci had become occupied territory under martial law. After summer rains came early, the town was also nearly under water as floodwaters surged through it. As many as fifty people died. With their relatives missing and their town buried in mud, the defiant euphoria of the strikers quickly suffocated as well. When the federal troops arrived, the miners surrendered their weapons and watched sullenly as the National Guard arrested their leaders. Ten Mexicans and Italians were later convicted of rioting and sentenced to the Yuma territorial prison even though no riot took place. At the height of the strike, the WFM president, Charles Moyer, proclaimed, ‘The men of Morenci have the full support of the Western Federation of Miners,’ but those words were the only assistance the strikers received (quoted in Sheridan 1995). The grudging admiration of the WFM soon fell victim to internal conflicts within the union itself. ‘Inclusionists,’ who favored organizing Mexican and southern European miners, quarreled with ‘exclusionists,’ who wanted to halt the recruitment of Mexican, Spanish, Italian and Slavic immigrants. And since copper companies imported Mexican scabs to break Anglo strikes and Anglo scabs to break Mexican strikes, ethnic hostility escalated. By the end of the decade, exclusionists dominated WFM locals in Jerome and Globe, where they fought to end the employment of Mexicans and southern Europeans. Such divisiveness played into the hands of the copper companies. Strike after strike sputtered and died (Sheridan 1995; Mellinger 1995).
Race, class and labor In the Arizona mining industry
181
Organized labor had more luck in the political arena. During most of the territorial period, large corporations dominated Arizona politics, and bribery was a common practice. The foremost goal of the corporations was to keep their taxes down—a goal they pursued with great success. At the turn of the century, mines worth an estimated $100 million were officially valued at only $2 million. This rankled many Arizonans, especially the small businessmen and property owners who paid the lion’s share of the territory’s taxes. The result was a strong but evanescent coalition of populists, progressives and labor leaders who first coalesced in the crusade to win Arizona statehood. This coalition called for a progressive charter that would place as much power as possible in the hands of the voters themselves in order to wrest Arizona’s government away from the mining companies and the railroads. The coalition therefore championed constitutional provisions guaranteeing the initiative, the referendum and the recall of all public officials, including judges, who were often controlled by the corporations. All but the recall of judges passed. Labor, the strongest element in the coalition, also successfully incorporated into the constitution a whole series of pro-labor provisions including an eight-hour day, employer’s liability, workmen’s compensation and the abolition of child labor. Labor’s greatest failure was the collapse of its crusade to restrict ‘alien’ labor in Arizona. Most union leaders considered Mexican workers the Achilles’ heel of the organized labor movement in the Southwest. In their opinion, Mexicans broke strikes, worked for lower wages and endangered other workers because they did not speak English—a racist assumption with no basis in the accident records of the mines that employed them. And there were always thousands more Mexicans just across the border, ready to pour into the copper towns whenever the corporations needed to crush the unions. Despite the CliftonMorenci strike of 1903 and the bloody strike in Cananea, Sonora, three years later, labor leaders insisted that Mexicans could not be organized, so they tried to write a constitution that banned most Mexican labor in Arizona (Park 1961; Sheridan 1995). The heart of organized labor’s anti-Mexican agenda was Proposition 91, which prohibited anyone who could not ‘speak the English language’ from working in ‘underground or other hazardous occupations.’ It also forbade any ‘individual, firm, corporation, or association’ from employing ‘alien labor’ as more than 20% of its workforce. In other words, Mexicans, Italians and other non-English-speaking immigrants could not work in the mines or as brakemen or engineers on the railroads. Moreover, Mexican nationals could not constitute more than 20% of any ranch’s cowboys, any lumber company’s mill workers, or any farm’s field hands. According to a US Immigration Commission study in 1911, 60% of Arizona’s smelter workers were Mexicans, half of whom had been in the USA less than five years. There are no reliable figures for the agricultural sector, but the percentage of Mexicans on farms and ranches was undoubtedly higher. Proposition 91 therefore threatened not just Mexican labor but the entire foundation of Arizona’s extractive economy (Park 1961; Sheridan 1995).
182
Thomas E.Sheridan
Not surprisingly, the coalition between labor and its allies quickly weakened. Realizing that their own access to cheap labor would be restricted, farmers and ranchers backed away from the 20% clause, while other delegates attacked the ‘English language’ stipulation. When the roll was finally called, Proposition 91 died by a margin of 26 votes to 19 (McGinnis 1930). The defeat of Proposition 91 revealed the weakness of what historian James Byrkit has called the ‘laborbourgeois coalition’ (Byrkit 1982). Demanding more taxes from corporations or supporting direct-democracy measures such as the initiative and referendum did not affect the running of a small firm, ranch or farm. But choking off Arizona’s supply of cheap labor threatened most businesses regardless of their size. Delegates voted their progressive ideals until their own livelihoods were affected. They also backed away from any provision that would have seriously damaged the copper industry. Even though industry representatives railed against ‘socialists’ and ‘radicals’ at the convention, the Arizona Constitution was a progressive, not a radical, document. Nonetheless, the fight to limit Mexican labor continued for the next few years. After statehood, the unions took their fight against Mexican labor to the voters. In 1914 they used the initiative to resuscitate the provision that 80% of the employees of any individual or firm had to be ‘qualified electors or native born citizens of the United States.’ The copper companies and railroads opposed the initiative, as did hundreds of Mexican miners across the state. But most citizens of the state agreed with the Arizona Labor Journal of 3 July 1914, which warned that Arizona ‘cannot assimilate untold hordes of aliens.’ The initiative passed by an overwhelming margin of 10,694 votes. What labor could not win at the convention it won in the ballot boxes (Sheridan 1986). Or so it thought. Although the primary goal of the law was to restrict Mexican labor, it affected other non-citizens as well. Both the British and the Italian ambassadors to the USA protested its passage, and in December 1914 an Austrian cook in Bisbee named Mike Raich sued to keep from being fired under the new law. Raich was represented by corporation attorneys from Bisbee and Tucson, who took the case to the US District Court in San Francisco. The court ruled that the law violated the Fourteenth Amendment, and the US Supreme Court upheld the decision on 1 November 1915 (Byrkit 1982). Between 1912 and 1916, however, the unions and their allies controlled state politics. The WFM and other unions also organized a round of strikes in 1915 —strikes in which, for the first time, Mexican and Anglo workers tentatively joined together to achieve common goals. The largest strike erupted in CliftonMorenci, where miners were tired of making the lowest wages in the state: $1.92 for muckers, $2.39 for Mexican miners and $2.89 for Anglo miners, compared with $3.50 per shift or more in the other major districts. Mexican mutual aid societies coordinated the early resistance, and then both Anglo and Mexican miners asked the WFM to organize the district. Flush with its successes in other Arizona mining districts, the WFM agreed, even though about two-thirds of the workforce were Mexican. At one meeting, Mexican miners interrupted an
Race, class and labor In the Arizona mining industry
183
Anglo organizer, shouting, ‘We can’t understand you! Speak in Spanish!’ Intentionally or unintentionally, it was a sardonic commentary on the WFM’s long ‘English only’ fight in the mines. The strike itself started on the night of 11 September. Most of the 5,000-man workforce walked off the job, shutting down operations in Clifton, Morenci and Metcalf. In contrast to earlier confrontations, however, this one had the sympathy of key county and state officials. When the strike was finally settled after five months in January 1916, labor across the nation hailed it as a major victory. On 26 January, the mines’ 5,000 workers went back to work with a sliding scale of wage increases pegged to the price of copper. Unskilled laborers who had made $1.92 a shift in September 1915 were making $3.08 by March. During the same period, the wages of skilled miners rose from $2.89 to $4.08. Moreover, the differential pay scale for Anglo and Mexican workers was abolished (Kluger 1970; Byrkit 1982; Sheridan 1995; Mellinger 1995). Nevertheless, the strike was not a complete success. Even though wages rose by 39%, the price of copper soared by 56% because of World War I, so profits outstripped wage increases by a considerable margin. The miners also agreed to ban the WFM from the Clifton-Morenci district in return for the right to join any other union. At the height of labor’s power, its most successful confrontation with the copper companies was little better than a draw. The strike also marked the beginning of a counteroffensive by the copper companies. On 23 September 1915, soon after the strike broke out, two men arrived in Clifton on the same train. One was Charles Moyer, president of the WFM. The other was Walter Douglas, who was soon to become vice-president and then president of Phelps Dodge. By the end of the strike, Moyer and the WFM had been rejected. Douglas, on the other hand, was carefully mapping out a campaign to destroy the power of organized labor in Arizona (Byrkit 1982). For years the copper companies’ deep distrust of one another had kept them from developing a united front. The taxation policies of the first Arizona state legislature, however, rapidly brought them together. Progressives and labor leaders abolished the exemption of ‘productive’ mines from property taxes. The legislators also tripled the assessment of Arizona mines by declaring that assessments were to be based on a mine’s ‘full cash value.’ In 1913 the mining companies claimed that the value of their mines was $31 million. The new state tax commission assessed them at more than $108 million. To turn back this tide, the copper companies joined together and mounted a multipronged assault on labor and its progressive allies. As World War I drove up the price of copper and created an atmosphere of patriotic hysteria, Walter Douglas and the managers of other copper companies picked apart the ‘laborbourgeois coalition.’ They targeted the Arizona senate and helped elect politicians sympathetic to the corporations. They purchased newspapers or bribed editors to turn public opinion against anticompany forces. By 1915, Phelps Dodge alone controlled Clifton’s Copper Era, Bisbee’s Daily Review,
184
Thomas E.Sheridan
Douglas’s International Gazette, Phoenix’s Arizona Gazette and Tucson’s Arizona Daily Star. They influenced public-school curriculums, manipulated lawyers and doctors, and even intervened in church politics to eliminate liberal ministers who stood up to the copper companies (Byrkit 1982). But the decisive battles were fought in the copper towns themselves. The USA and its allies needed copper for shell jackets, cable, wire—all the industrial demands of the war. So the value of Arizona copper production soared, from $40 million in 1910 to $200 million in 1917. Copper, which was selling for 13.4 cents a pound in 1914, was fetching 26.5 cents a pound two years later. And since it cost a company like Phelps Dodge only 9.5 cents a pound to produce the copper, profits were enormous—a staggering $24,030,905 for Phelps Dodge in 1916. These profits gave the corporations tremendous clout. Even more important was the patriotic fervor sweeping over Arizona and the rest of the nation. When the USA plunged into World War I in the spring of 1917, a fear of industrial sabotage spread across the country. In the mining communities, much of this fear focused upon the Industrial Workers of the World (IWW), a loose and anarchic amalgam of miners, loggers and unskilled laborers better known as the Wobblies. Influenced by socialist and syndicalist doctrines, the IWW opposed the war and advocated direct action, including violence and sabotage, in the struggle between labor and management. Wartime profits for the copper companies also inspired wartime demands from the miners. The International Union of Mine, Mill, and Smelter Workers (IUMMSW, the Western Federation of Miners under a new name) called the first strike in Jerome on 25 May 1917. Walkouts followed in Globe, Morenci and Bisbee in June and July. Company-controlled newspapers attacked the strikes and branded all strikers as seditious Wobblies. Yet card-carrying Wobblies constituted only a small minority of the workers. In Jerome, for example, there were no more than 125 members of the IWW among 4,000 miners. Wobblies were little more than specters invoked to inflame old fears. But the copper companies did not stop at sensational journalism. On 11 July, Walter Douglas gave a speech in Globe, where several thousand IUMMSW strikers had shut down the mines. ‘There will be no compromise because you cannot compromise with a rattlesnake,’ Douglas vowed. ‘That goes for both the International Union and the IWW’s…. I believe the government will be able to show that there is German influence behind this movement…. It is up to the individual communities to drive these agitators out as has been done in other communities in the past’ (quoted in Sheridan 1995). Douglas engineered just that in Bisbee, the center of Phelps Dodge’s most important mining district. Phelps Dodge persuaded Sheriff Henry Wheeler of Cochise County to deputize the Bisbee Citizens Protective League, which was formed to defend the community against the Wobbly menace. Some 1,600 miners also joined the Workmen’s Loyalty League, which vehemently opposed both the IWW and the strike.
Race, class and labor In the Arizona mining industry
185
Then, at 6:30 a.m. on 12 July, two thousand ‘deputies’ under the command of Sheriff Wheeler fanned out across town, breaking down doors and pulling strikers from their beds at gunpoint. An hour later a huge travesty of a parade filed from the Bisbee post office to the ballpark in nearby Warren. Armed guards surrounded the prisoners, who eventually numbered two thousand. Once inside the ballpark, mine managers promised the prisoners that they could go back to work if they renounced the strike. More than eight hundred did so, and they were released. Of the rest, twenty-three boxcars from the El Paso and Southwestern Railroad pulled into Warren station, and 1,186 men were jammed into the cars. Only about a third of the deportees actually belonged to the IWW. Nearly two hundred vigilantes guarded them from on top of the train. The miners were abandoned by their captors at 3 a.m. the next morning in the southwestern New Mexican desert. They had no food or water. It was the largest mass kidnapping in Arizona history. The Bisbee deportation appalled New Mexico and federal officials. President Woodrow Wilson sent the deportees food from El Paso and ordered the army commander to set up a camp for them and provide them with rations. For the next two months, most of the deportees remained at Columbus, New Mexico, under federal care, where a census revealed that 804 of the 1,003 men in Camp Furlong were foreigners from twenty different nations. Mexicans (268) comprised the largest group, followed by Austro-Hungarians (179) and British (149). That must have made it easier for the vigilantes to convince themselves that they were rounding up alien saboteurs (Byrkit 1982). There was undoubtedly a strong anti-Mexican flavor to this anti-foreign crusade, however. German agitation was strong along the Mexican border; the German foreign secretary, Arthur Zimmermann, even sent Mexican president Venustiano Carranza a telegram offering to return all the territory the USA had seized from Mexico in 1848 if Mexico allied itself with Germany. In 1915, a group of Mexican revolutionaries even promulgated the Plan de San Diego, which called for a race war to take back Texas, New Mexico, Arizona, Colorado and California and to execute all Anglo males over 16. Such episodes, along with Pancho Villa’s raid on Columbus, New Mexico, must have reinforced longstanding prejudices against Mexican labor in Arizona’s strategic copper mines. For a time, however, it looked as though the organizers of the Bisbee Deportation might actually be punished for their actions. The Arizona attorney general and then a federal commission investigated the deportation, which the commission condemned. On 15 May 1918, the US Department of Justice even indicted twenty-one Bisbee luminaries, including Walter Douglas and Sheriff Wheeler, on charges of conspiracy and kidnapping. In response, defense attorney E.E.Ellinwood argued that no federal laws had been broken, so the case could only be heard in state courts. The state of Arizona tried two hundred Bisbee residents for kidnapping in 1920, but after three months of testimony, all the defendants were acquitted. ‘The verdict of the jury is a vindication of the deportation,’ jury foreman J.O.Calhoun proudly proclaimed (Byrkit 1982).
186
Thomas E.Sheridan
A year later, Phelps Dodge created a company union whose constitution could be abrogated by the corporation’s board of directors. Other companies quickly followed suit. With a Republican administration back in office in Washington and postwar anti-bolshevik witch-hunts sweeping the country, the labor-progressive coalition shattered and independent unionism withered and died. It did not resurrect itself until the end of World War II, when MexicanAmerican veterans in Clifton-Morenci carried out the first successful major strike against Phelps Dodge since the forging of the copper collar during World War I. By then, growing demand for civil rights and economic equality on the part of Mexicans in Arizona weakened the idiom of race and strengthened the idiom of class, at least in the copper towns.
DISCUSSION AND CONCLUSION Between 1880 and World War II, copper mining was the most important extractive industry in Arizona. Copper companies manipulated territorial and state legislatures, pitted Anglo-Irish against Mexican workers to weaken the power of labor unions and keep wages down, and instituted what was in many respects a neocolonial social order in Arizona mining communities. Although particulars may have varied, similar processes characterized mining communities in other areas of the world that depended upon immigrant labor and large quantities of outside capital and technological expertise. Wherever mining depended upon high capital investment and a permanent, relatively stable workforce that was not locally available, landscapes of social stratification and ethnic segregation developed, often in geographically restricted spaces. Many of those same landscapes manifested a specialized form of gender segregation as prostitution districts thrived. Unfortunately, the material correlates of either stratification or segregation are not well understood in Arizona mining communities because so little historical archaeology has been done. One promising area of research would be to investigate the extent to which patterns of household consumption re-created patterns of consumption in donor areas. Did the isolation of Arizona’s mining settlements homogenize consumption patterns, or was consumption ethnically marked, especially as transportation networks became more sophisticated? An even more fascinating avenue of archaeological exploration would be to examine processes of ethnic compression and ethnic differentiation in the material record. Did Irish, Cornish and Anglo-American dietary preferences, house construction and consumption of luxury items converge as Anglo-Irish miners attempted to differentiate themselves from ‘Mexican’ or southern European miners? Or were private ethnic spaces preserved within hearth and home? This chapter has examined issues of race, class and social control as they were played out in the public arena of mining communities. In order to understand how those same issues played out in more intimate domestic spaces,
Race, class and labor In the Arizona mining industry
187
historical archaeology—complemented by oral history and different forms of documentary research—must become a research priority.
REFERENCES Byrkit, J.W. (1982) Forging the Copper Collar: Arizona’s Labor-Management War, 1901–1921. Tucson: University of Arizona Press. García, M. (1981) Desert Immigrants: The Mexicans of El Paso, 1880–1920. New Haven: Yale University Press. Kluger, J.R. (1970) The Clifton-Morenci Strike. Tucson: University of Arizona Press. McGinnis, T.A. (1930) The influence of organized labor on the making of the Arizona constitution. Master’s thesis, University of Arizona, Tucson, AZ. Mellinger, P.J. (1995) Race and Labor in Western Copper: The Fight for Equality, 1896–1918. Tucson: University of Arizona Press. Officer, J.E. (1987) Hispanic Arizona, 1536–1856. Tucson: University of Arizona Press. ——(1991) Mining in Hispanic Arizona: myth and reality. In M.Canty and M.N. Greely (eds), History of Mining in Arizona, vol. 2, pp. 1–26. Tucson: Mining Club of the Southwest Foundation and the American Institute of Mining Engineers, Tucson Section. Park, J. (1961) The history of Mexican labor during the territorial period. Master’s thesis, Department of History, University of Arizona , Tucson, AZ . Schwantes, C.A. (ed.) (1992) Bisbee: Urban Outpost on the Frontier. Tucson: University of Arizona Press. Sheridan, T.E. (1986) Los Tucsonenses: The Mexican Community of Tucson, 1954–1941. Tucson: University of Arizona Press. ——(1995) Arizona: A History. Tucson: University of Arizona Press. Vaughan, T. (1992) Everyday life in a copper camp. In C.Schwantes (ed.), Bisbee: Urban Outpost on the Frontier, pp. 57–84. Tucson: University of Arizona Press.
Part III
Prehistory and protohistory
Chapter 12
Producing copper in the eastern Alps during the second millennium BC Stephen Shennan
ABSTRACT Despite a long history of debate over the role of metal production and exchange in defining the character of European Bronze Age societies, until recently remarkably little work had been done on the social organization of copper production. This is somewhat curious inasmuch as the copper mines of the period have been known in some parts of Europe for a hundred years and a great deal of work has been done on mining and production techniques. This chapter describes the results of a recent fieldwork project in one of the main Bronze Age copper production centres, the Mitterberg region of the eastern Alps, designed to provide information on this subject. The excavated site, St Veit-Klinglberg, was a base settlement for a small community which mined and smelted copper in the vicinity and then exchanged it downstream to agricultural communities in the hills and plains outside the mountains. Finds indicating links between these areas and the copper producers include amber, flint, tin-bronze and imported pottery. The site, like its contemporaries in the region, was fortified and in a defensible position, which may indicate the threat of raiding. It is suggested that these communities were autonomous and that production and exchange were not monopolized by any single group. The copper entered patterns of circulation where it functioned as a protocurrency. Eventually the small autonomous settlements were abandoned and the copper production region may have come under centralized control.
INTRODUCTION Discussions of the organization of Bronze Age metallurgy have played an important role in studies of European prehistory for much of the twentieth century. From the 1920s onwards the Bronze Age was seen by Gordon Childe as the period in history when the contrast between the ‘dynamism’ of Europe and the ‘stagnation’ of the East, which had long been a staple of Enlightenment social philosophy, first began to develop. Subsequently, the focus of European Bronze Age studies has moved away from ‘the irradiation of European barbarism by [decadent] oriental civilisation’ to a concern with the emergence of social ranking and stratification, which has
192
Stephen Shennan
come to be regarded as the key development of this period. In some accounts (e.g. Kristiansen 1984; Shennan 1982, 1986), control over the production and exchange of copper and bronze has been seen as playing a central role in this process; in others (e.g. Gilman 1981), the inequality and power differentials are seen as arising from the control of subsistence resources, but here too the use of copper and bronze objects is seen as an important aspect of the expression of these social differences. More recently (Shennan 1993a) it has been suggested that the emphasis on ranking and stratification has been excessive, driven on the one hand by a desire to see Bronze Age Europe as a participant in the ‘rise of civilization’ and on the other by acceptance of the ‘myth of control’ which pervades social evolutionary approaches and which, in effect, subscribes to the ideology of early elites that everything happens in society as a result of their efforts. In this alternative view the different aspects of Bronze Age economic and social life are more loosely coupled than is normally assumed, and different sources of social power may be in a number of different hands and in a variety of spatial locations. From the perspective of the subjects which have been central to Bronze Age studies, one of the most important kinds of spatial location concerns those places where metal was actually produced. Of these, the regions where copper was mined and smelted are certainly among the most significant, because they can provide information about the organization of the production and distribution of the single most important resource for Bronze Age metallurgy. Investigations of early copper-mining sites have now been under way for over one hundred years, but recently there has been increasing interest in looking at the social organization of production and not merely the technology. O’Brien (1994) has discussed the social organization of the production and distribution of copper at the Mount Gabriel mine in Ireland, although he was handicapped by the lack of contemporary settlement evidence. Montero-Ruiz (1993) has argued that the settlements of the Argaric Bronze Age in southeast Spain obtained their copper from adjacent sources and that the control of copper cannot therefore have been an important basis for social inequality and power differentials. However, until recently no such examination had taken place of the social organization of copper production in any of the major mining areas of Central Europe, where the scale of production not only was much larger than in either Spain or Ireland but also had an enormous impact over a wide area. Many if not most of the major copper-mining areas of Central Europe are still poorly known, but the Mitterberg mining region in the Alpine zone of the valley of the river Salzach south of Salzburg has been the object of study over a long period. The location of the major Bronze Age mining areas has been established and smelting places and settlements are also known. The region therefore provides an excellent focus for an investigation of the nature of Bronze Age mining communities and of the organization of copper production in a
Producing copper in the eastern Alps
193
region which has been estimated to have produced 10 tons of copper a year from the later early Bronze Age onwards (Eibner 1993). The project, part of whose results are summarized below, focused on the excavation of a single settlement site, St Veit-Klinglberg (Shennan 1995), but as always a regional perspective is essential to its understanding. In this case there are two relevant regional scales: a local one, the copper production region itself, with its mines, ore processing areas, smelting places and settlements, and patterns of movement between them (Figure 12.1); and a larger one which includes the areas downstream from the mining region and north of the Alpine zone. These areas were the recipients of the copper produced, and they maintained exchange links with the producers (Figure 12.2).
MINING COMMUNITIES The Klinglberg is typical of the earlier Bronze Age settlements of the Mitterberg mining region in its small size, ca. 4 ha, and in its defensible position on the edge of a steep drop to the bottom of the Salzach valley (Figure 12.3), in this case, and in most if not all of the others, supplemented by a wall on the more accessible side. However, only the Klinglberg has extensive surviving Bronze Age deposits which have been excavated on a large scale. The distribution of finds and features in the main excavated area suggests that activities on the site were predominantly domestic. Estimation of the number of houses on the site is difficult, but on the basis of those already excavated, and allowing for the fact that much of the site area was too steep for houses to be built, a total of about fifteen houses seems an absolute maximum and ten or fewer in contemporary occupation seems more likely. Assuming a modular house size of approximately 7 m×3 m on the basis of the known structures, a range of 4–7 people per house might be assumed. For the site as a whole this would give an estimated population of about 30–40 people at the bottom end of the range, up to 100–110 at the top, with a figure towards the lower end perhaps more likely. As the map (Figure 12.1) makes clear, the Klinglberg, like the other contemporary settlements, is not situated adjacent to the copper sources but rather is at some distance from them and closer to the valley bottom. Thus the assertion—based on a variety of different lines of evidence—that it was connected to copper production clearly needs justifying. First, there is the presence of small pieces of copper ‘casting cake’, lumps of raw cast metal. Second, the diagnostic feature of the local pottery, here as elsewhere in the Alpine copper-producing area at this time, is the use of copper slag as temper. Its addition was clearly intentional since the slag added to the finer fabrics was more finely ground than for the coarser ones (Knowles and Quinlan 1995). Third, a small proportion of the sherds had been refired to high temperatures, while two burnt pine needles from the site had been vitrified. These indications that very high temperatures
194
Stephen Shennan
Figure 12.1 Copper sources and known early Bronze Age/middle Bronze Age settlements in the Mitterberg copper production region of the Salzach valley
Figure 12.2 The exchange region of the Mitterberg mining region, showing documented contacts
Figure 12.3 View of the Klinglberg and its topographic location from the southwest. The arrow indicates the position of the site. The river Salzach is running eastwards, away from the camera. The Burgstein, Burgschwaig and Einoeden copper lodes are on the higher part of the slopes rising to the left of the river in the middle ground. The Mitterberg lodes are on the nearest of the background ridges on the left of the picture. The mines marked A-G on Figure 12.1 are high on the ridge in the foreground, but off the picture to the left.
Producing copper in the eastern Alps
197
were sometimes produced at the Klinglberg settlement may indicate that smelting occasionally went on there even though no definite ovens, as opposed to hearths, were found. However, the vast bulk of smelting activities must have taken place elsewhere, and indeed there is evidence for this. Finally, in addition to smelting, involvement in ore preparation is also suggested by the presence on the site of hammerstones like those found adjacent to the mines themselves, where they were used in the final stages of ore crushing. The connection this evidence establishes with copper production still leaves open the question of how far the occupants of this settlement, and presumably of the others like it, specialized in these activities, and indeed whether the site was occupied permanently or merely seasonally, as is often the case with mining activities in non-industrial contexts (see, for example, Herbert 1984). Seasonality is always difficult to establish, and in the case of the Klinglberg the bone assemblage, which is generally the basis of such inferences, was very small and badly preserved. However, there was one indication that the community was fairly sedentary, or at least not engaged in regular seasonal long-distance movements from the plains to the mining area in the mountains. This was the presence of pig skull bones and teeth. These indicate that pigs were being kept locally, and pigs are not suitable for long-distance migratory movements. The subsistence evidence is also relevant to the degree of specialization of the Klinglberg community. It appears from the absence of charred cereal rachis fragments and other waste in the plant remains from the site (Green 1995) that only highly processed cereals were being used. This may suggest that the Klinglberg was associated with cereal consumption rather than production, since there is no evidence of crop processing in the form of sieving, threshing and winnowing and the charring of the resulting waste products. It may be that the Klinglberg community limited food production to give more time to copper production. Some crops could have been cultivated locally, such as peas, for which there is evidence; these have a short growing season and require less input of effort. No doubt locally available wild plants were also exploited. However, already processed cereals may have been obtained in exchange for smelted copper. If the Klinglberg was occupied on a relatively permanent basis as suggested, then one might expect it to represent a community of men, women and children rather than, for example, young adult males only, who might be the occupants of a temporary camp. In any event, ethnographic and historical evidence (e.g. Herbert 1984; Rule 1971) indicates that women and children often had an important role in the metal production process, as well as men, particularly in the key labour-intensive process of ore preparation: smashing up the ore obtained from the mine into ever finer grains and separating it from the waste so that the ore fraction which was finally smelted would be as rich as possible. However, unless there is a contemporary adjacent cemetery, actually identifying the presence of men, women and children at a settlement site, as opposed to merely assuming it, is not easy in a situation where there is no ethnographic or
198
Stephen Shennan
historical evidence to suggest the gendering of activities. In the case of the Klinglberg the only indication comes from two artefacts found on the site, an amber bead and a bronze dagger. Contemporary burial evidence from adjacent areas suggests that amber beads tend to be more associated with females while daggers are usually associated with males.
THE PRODUCTION REGION It has already been suggested that the Klinglberg and the other communities like it represent part of a larger network of activities associated with copper production which involved a variety of different locations. The mines are well known and were certainly being exploited during the period of occupation of the Klinglberg, in the range 1800–1400 BC (Eibner 1993; Gstrein 1988). There is extensive evidence of ore preparation close to the mines, as well as smelting (e.g. Eibner 1993). However, smelting was also carried out some distance from the mines (Herdits, personal communication). Dispersing the smelting sites in this way made sense in light of the considerable demands on timber of the smelting process and was possible because the volume and weight of the ore had been drastically reduced by the painstaking preparation process. Given the evidence outlined above that inhabitants of the Klinglberg were involved in copper production, it appears likely that at least some members of the community spent greater or lesser periods of time working elsewhere, at mining and smelting sites. The indication that they may have been importing cereals from elsewhere also points in this direction, rather than in favour of the alternative possibility that they were exchanging agricultural products with other communities which permanently occupied sites near the mines. That such local movements may have been quite complex is suggested by information derived from lead isotope and trace element analysis of some of the pieces of copper ‘casting cake’ from the Klinglberg (Romanov 1995). Three different metal types could be distinguished, suggesting that the pieces came from different ore zones within the Mitterberg region. One of the types probably came from the Mitterberg main lode (see Figure 12.1), on the other side of the Mühlbach valley from the Klinglberg, too distant for regular access on a daily basis. One of the other two types probably comes from the Mitterberg southern zone, south of the Mühlbach valley and much closer to the Klinglberg, but perhaps still rather far for daily travel given the steep nature of the terrain. The indications of access to a range of sources, including some at a distance and actually nearer to other settlements, point to two conclusions: that there was at least some freedom to move through the landscape, and that there was no exclusive ownership and defence of local sources by specific groups. Indeed, there is no evidence whatsoever of physical defence of the mines or of the smelting places, although particular groups no doubt had rights to the shafts
Producing copper in the eastern Alps
199
and tunnels they had created. This situation contrasts markedly with that implied by the settlements, with their inaccessible positions and additional walls or ditches, which do point to a defensive motive although it is not known whether contemporary settlements in different situations remain to be discovered. This brings us back to the role of the settlements in the local regional system. One possibility might be that they represent the strong-points of an upper social stratum controlling the surrounding area and its people, like the medieval castles which are found in the same area, and in some cases on the same sites, in a later period. It is very hard to refute such an idea conclusively on the basis of settlement evidence, but on balance it does not seem convincing. Currently there is no evidence for the lower-level settlements which this would imply. There is little evidence within the site or the region as a whole for artefacts indicating an elite subcultural style, in very marked contrast to the wealth evidenced a millennium later at the nearby salt-mining site of the Dürrnberg. Within the wider region of the Alpine foreland defended and defensible sites are very frequent in this period (Biel 1987), but again little indication exists that they represent higher levels in a settlement hierarchy, in contrast to regions further east (see, for example, Shennan 1993b). A more likely scenario is that the settlements represent small autonomous communities and that there was a widespread climate of hostile relations between neighbouring groups, who were at least occasionally raiding one another. In most areas the object of such raiding was no doubt stock. In copper production regions it would have been raw smelted copper. Since the valley bottom settlements were the bases of the mining and smelting communities, this is where the metal produced over the weeks and months higher up the mountains would have been concentrated prior to sending it downstream; the broken bits of casting cake found scattered around the Klinglberg site would be the residue of this. Given the complexity of the copper production process and its demands on time, labour and expertise, it would have made much more sense to attempt to seize, and equally to defend, the product rather than the raw materials and facilities involved in producing it.
THE EXCHANGE REGION We have seen that the Klinglberg and other settlements like it appear to represent fairly permanent small-scale autonomous communities engaged in the production of copper in the source region in the mountains. The potential consumers of the metal, on the other hand, occupied the hills and plains of the valley of the Danube and its tributaries outside the Alps to the north. This situation obviously presupposes the existence of exchange links down the Salzach valley, along which the smelted metal no doubt went downstream by canoe. Copper and bronze still remain problematical to characterize, so the final
200
Stephen Shennan
destinations of the Mitterberg copper still cannot be identified. On the other hand, the Klinglberg project revealed extensive evidence of the kinds of things that were reaching the copper producers from the wider world outside the mountains (see Figure 12.2). Perhaps the most striking find in this respect is the amber bead (Shennan 1995). At the time of the occupation of the Klinglberg, amber from the Baltic region was circulating widely in central Europe (Shennan 1982), so there is no reason to assume any direct long-distance connection between the Baltic region and the Alpine copper producers. Nevertheless, the find confirms for the first time the long-assumed connection between amber and metal exchange and shows that the producers were connected to the major interregional exchange networks of this period and through them were able to get access to valuable materials. Given the likely low loss rates of a valuable material like amber on a settlement and the problems of finding it when present (the bead was recovered from a flotation sample), it seems likely that the bead should be considered the tip of an iceberg. The metal artefacts from the Klinglberg also indicate ultimately long-distance contacts since they were made of tin-bronze rather than the unalloyed copper of which the pieces of ‘casting cake’ are made. Whether or not the copper used was local is at present unknown, but the finds show that tin, or already combined tin-bronze, or the artefacts themselves were reaching the copper producers via the interregional exchange networks. Shorter-distance links in the exchange chain are also evidenced. There are very few chipped stone artefacts from the Klinglberg but a number of them, projectile points in particular, are made of Plattensilex, which must have come from the area around Salzburg (Moosleitner, personal communication), either in the form of raw material or, given the skills involved and the lack of lithic debitage at the Klinglberg, perhaps more likely as finished artefacts. Finally, it appears that about 1% of the pottery from the Klinglberg is nonlocal (Knowles and Quinlan 1995). Perhaps the most interesting of these imported sherds are the small number containing graphite in their fabric, which almost certainly come from the area around Passau (see Figure 12.2). As the map shows, this is at the point where the Danube is joined by the river Inn, of which the Salzach, which runs through the mining area, is a tributary. Here we have direct evidence of precisely the kind of exchange link to be expected if the Klinglberg community were sending their metal downstream to supply the agricultural communities of the Danube. All these different lines of evidence appear to suggest that the individual small-scale metal-producing communities in the mining region had their own exchange links with the outside world, rather than being linked to it only indirectly via some controlling superordinate centre; thus we have yet another argument in favour of their autonomy. As noted earlier, it has long been suggested that from the later early Bronze Age (c. 1800–1700 BC) onwards the Mitterberg region was producing large quantities of copper, perhaps up to 10 tons per year (Eibner 1993), which would
Producing copper in the eastern Alps
201
have had a major impact on the circulation of metal over a wide area. Liversage (1994) has recently argued that such a large scale of production would have required centrally organized coerced production and could not have been achieved by small-scale autonomous communities. However, not only does all the evidence reviewed here contest such a claim, but estimates of copper production rates based on ethnographic and historical evidence (Shennan 1995: 300–302) indicate that a population in the low hundreds for the Mitterberg region as a whole, which seems entirely plausible, could easily have produced at least several tons of copper per year.
THE SOCIAL CONTROL OF METAL EXCHANGE We are now in a position to return to some of the questions concerning metal production and its control in the earlier Bronze Age which were raised in the introduction to this chapter. It has already been argued that within the mining region itself autonomous small communities were producing copper and obtaining various things in return for it, thus realizing an exchange value, but how did the terms of exchange operate on the larger spatial scale of the exchange region? Immediately outside the production zone in the Alpine foreland, there prevails a similar impression of a lack of overarching control. One of the few areas close to the Alpine sources where monopoly and control might in principle be possible is the vicinity of present-day Salzburg, since the vast majority of the copper produced in the Mitterberg region must have come through here, down the river Salzach. Yet although there is extensive evidence of earlier Bronze Age settlement (summarized in Shennan 1995), there is no evidence of rich burials representing concentrations of wealth and power and symbolizing social differences. In fact, traces of copper-working are found in both the open river terrace sites and the local hilltop settlements. Furthermore, even if one supposes that the distinction between the two settlement types is indicative of a settlement and social hierarchy, which is very much open to question, there is no single dominant hilltop settlement in the area. It appears that the picture in this key exchange location is very similar to that in the mining region itself; metal did not automatically provide a basis for social control and the growth of hierarchies. This depended on whether its control was monopolized: in the earlier Bronze Age of the Mitterberg production region and its immediate exchange hinterland, this appears not to have been the case. Nevertheless, outside the mining zone, in the Alpine foreland, there does seem to have been a change in the nature of the circulation of copper, since it was transformed into ingots of standardized weights which are widely found in this region and elsewhere (see most recently Lenerz-De Wilde 1995). It used to be thought that these ingots represented batches of raw copper in the form in which it was originally smelted in the production zone. However, it is now clear
202
Stephen Shennan
that the situation is more complex than this. No ingots have been found in the mining regions. Moreover, while in any given hoard the majority of ingots usually belong to a single copper type, there are often small proportions of ingots of different types, suggesting the possibility of different origins. Occasionally, ingots contain small quantities of tin (Menke 1982), indicating an element of recycling. However, it is unlikely that recycled metal played a significant part in these hoards. The vast majority of ingots were of pure copper whereas the standard artefact metal was tin-bronze, so one would expect ingots containing tin to be much more frequent if recycled metal was significant. More convincingly, the ingot hoards have been regarded as evidence that by the later early Bronze Age, metal had come to acquire some of the functions of a ‘primitive money’, as store of value, standard of value and means of exchange (see Lenerz-De Wilde 1995). Since ingots are not known from the mining regions, it would appear that this transformation to a standardized exchange commodity took place outside the mining community rather than within it. This occurred perhaps at places like Obereching, north of Salzburg on a terrace of the river Salzach, where four hoards of rib ingots of standardized weights were found in what were clearly pits beneath the floors of four different houses in a settlement (Moosleitner 1988). This may suggest not only a greater intensity of exchange transactions but also a more acute sense of the value of the material among the intermediaries along the exchange chains than among the producers themselves. However, as we have already seen, these intermediaries were unable to achieve monopolies. Indeed, it could be argued that the apparent existence of an extensive network of exchanges involving quantitative equivalences in itself indicates a significant separation between the ‘economic’ and the ‘political’ in which the former has the potential significantly to undermine existing power relationships based on other foundations. It was not something anyone could easily control. The miners of the Mitterberg and elsewhere just kept on producing copper because it was in their local interest to do so. The situation changed only when it became possible to exercise political power on a much larger spatial scale, at the beginning of the later Bronze Age. Around 1400 BC the autonomous hilltop settlements in the Mitterberg mining region (and many other areas—see Biel 1987) were abandoned and large, open areas of occupation and metallurgical activity are characteristic, most likely as a result of the eventual establishment of centralized control of production and producers which previously seems to have been so conspicuously lacking.
DISCUSSION AND CONCLUSION It seems that even in fully prehistoric situations it is possible to make at least some inferences about the nature and organization of mining communities using archaeological evidence, particularly when relevant questions are
Producing copper in the eastern Alps
203
formulated at a local regional scale. In this case it has been suggested that small communities, probably permanent and consisting of women and children as well as men, organized copper production at a variety of different spatial locations within their vicinity. It seems likely that copper production was not just incidental to these people’s lives but that they specialized in copper production to the extent that they were unable to carry out a full range of ‘normal’ subsistence activities. The small size of the communities, their apparent concern with protection and the lack of evidence for social differentiation all point to the conclusion that they were autonomous and relatively egalitarian, but potentially hostile to one another. The threat of raiding for accumulated copper has been suggested. This argument for autonomy has been supported by pointing to the lack of contemporary centres immediately outside the mining regions which might have exercised control within them. This last point leads us to a conclusion which is apparent in all the case studies in this volume. It is impossible to understand the nature of mining communities without a knowledge and understanding of the wider social, economic and political networks of which they form a part, because the whole basis of such communities is that they provide for their own needs and desires by producing for and exchanging with others. The larger socio-economic system (in recent centuries a global one) conditions if it does not determine the nature of the mining communities, whether in defining the nature and scale of the labour pool, affecting the scale of investment, or defining the existence and terms of domination and resistance and rates of exchange. In the eastern Alps in the earlier second millennium BC, where small-scale autonomy seems to have been the rule over a wide area, relations of domination and resistance seem unlikely to have played a major role in defining the character of mining communities, as they have in other situations. Correspondingly, the lack of a dominant monopoly exchange partner suggests that at least adequate terms of exchange could always be maintained, despite the fact that copper seems to have been more consciously commodified once it moved outside the mining regions. Unsurprisingly, though, once the larger political and economic system changed with the onset of the later Bronze Age, the mining communities inevitably felt the effect.
REFERENCES Biel, J. (1987) Vorgeschichtliche Höhensiedlungen in Südwürttemberg-Hohenzollern. Forschungen und Berichte zur Vor- und Frühgeschichte in Baden-Württemberg 24. Stuttgart: Konrad Theiss. Eibner, C. (1993) Die Pongauer Siedlungskammer und der Kupferbergbau in der Urzeit. In W.Günther, C.Eibner, A.Lippert and W.Paar, 5000 Jahre Kupferbergbau Mühlbach am Hochkönig-Bischofshofen, pp. 10–26. Mühlbach and Bischofshofen: Gemeindeamt. Gilman, A. (1981) The development of social stratification in Bronze Age Europe. Current Anthropology 22: 1–23.
204
Stephen Shennan
Green, F. (1995) The plant remains from the Klinglberg. In S.J.Shennan (ed.), Bronze Age Copper Producers of the Eastern Alps. Universitätsforschungen zur prähistorischen Archäologie 27: 223–231. Bonn: Habelt. Gstrein, P. (1988) Neuaufnahme eines vorgeschichtlichen Abbaues im Arthurstollen (Bergbau Mitterberg). Mitteilungen der Gesellschaft für Salzburger Landeskunde 128: 425– 438. Herbert, E.W. (1984) Red Gold of Africa: Copper in Precolonial History and Culture. Madison: University of Wisconsin Press. Knowles, K. and Quinlan, A. (1995) The pottery. In S.J.Shennan (ed.), Bronze Age Copper Producers of the Eastern Alps. Universitätsforschungen zur prähistorischen Archäologie 27: 147–196. Bonn: Habelt. Kristiansen, K. (1984) Ideology and material culture: an archaeological perspective. In M.Spriggs (ed.), Marxist Perspectives in Archaeology, pp. 72–100. Cambridge: Cambridge University Press. Lenerz-De Wilde, M. (1995) Prämonetäre Zahlungsmittel in der Kupfer- und Bronzezeit Mitteleuropas. Fundberichte aus Baden-Württemberg 20: 229–327. Liversage, D. (1994) Interpreting composition patterns in ancient bronze: the Carpathian Basin. Acta Archaeologica 65: 57–134. Menke, M. (1982) Studien zu den frühbronzezeitlichen Metalldepots Bayerns. Jahresbericht der Bayerischen Bodendenkmalpflege 19/20: 5–305. Montero-Ruiz, I. (1993) Bronze Age metallurgy in south-east Spain. Antiquity 67: 46–57. Moosleitner, F. (1988) Vier Spangenbarrendepots aus Obereching, Land Salzburg. Germania 66: 29–67. O’Brien, W. (1994) Mount Gabriel: Bronze Age Mining in Ireland. Galway: Galway University Press. Romanov, P. (1995) Archaeometallurgical investigations of ‘casting cake’ and a copper ore sample from the Klinglberg excavations. In S.J.Shennan (ed.), Bronze Age Copper Producers of the Eastern Alps. Universitätsforschungen zur prähistorischen Archäologie 27: 263–273. Bonn: Habelt. Rule, J.G. (1971) The labouring miner in Cornwall c. 1740–1870 : a study in social history. Unpublished PhD thesis, University of Warwick. Shennan, S.J. (1982) Exchange and ranking: the role of amber in earlier Bronze Age Europe. In C.Renfrew and S.J.Shennan (eds), Ranking, Resource and Exchange, pp. 33– 45. Cambridge: Cambridge University Press. ——(1986) Central Europe in the third millennium BC: an evolutionary trajectory for the beginning of the Bronze Age. Journal of Anthropological Archaeology 5: 115–146. ——(1993a) Commodities, transactions and growth in the Central European Early Bronze Age. Journal of European Archaeology 1(2): 59–72. ——(1993b) Settlement and social change in Central Europe, 3500–1500 BC. Journal of World Prehistory 7: 121–161. ——(ed.) (1995) Bronze Age Copper Producers of the Eastern Alps. Universitätsforschungen zur prähistorischen Archäologie 27. Bonn: Habelt.
Chapter 13
Prehistoric copper mining in the context of emerging community craft specialization in northeast Thailand Vincent C.Pigott
ABSTRACT Early in the second millennium BC the production of copper and bronze began in Southeast Asia. Excavations conducted by the Thailand Archaeometallurgy Project (TAP) revealed a major prehistoric copper mining complex located at Phu Lon on the Mekong River in northeast Thailand. The first portion of the discussion treats resource procurement and copper/bronze production at the site. It is proposed that by means of ‘mining expeditions’ —possibly akin to the model of recent stone procurement strategies for ax manufacturing by the Tungei of highland New Guinea—prehistoric peoples extracted sufficient amounts of copper ore and/or smelted ore to ingots and transported these products downstream or overland to consumer villages located in the Sakon Nakhon Basin on the Khorat Plateau. Next, a series of questions are posed to ascertain whether the considerable effort expended to extract such a resource reflected the presence of a central authority which controlled access to and distribution of mined resources. The discussion concludes with the application to the Phu Lon-Sakon Nakhon Basin interaction sphere of Andrew Sherratt’s multizoned model for the structure of material production and distribution channels.
INTRODUCTION Finds of metal artifacts in copper and bronze from archaeological sites in Southeast Asia have been a commonplace as long as sites have been excavated. However, the revelations in the mid-1970s of bronze in the prehistoric village of Ban Chiang in northeast Thailand put Southeast Asia on the map as an Old World culture area with its own unique metalworking traditions (Gorman and Charoenwongsa 1976; White 1982, 1986, 1988, 1997). The ensuing discussion concerning the origins and development of copper/bronze metallurgy in Southeast Asia has resulted in important advances since the initial recognition that this area was relevant to the topic of Old World metallurgical development, not least White’s (1988:179) characterization of a ‘Southeast Asian metallurgical province’ (on this term,
206
Vincent C.Pigott
Figure 13.1 Map of major prehistoric sites in Southeast Asia and Thailand Archaeometallurgy Project study areas
see Chernykh 1980). Distinct from other Old World metallurgical centers, this province is characterized by a technological ‘internal coherence’ of adaptive metalworking traditions. A second important advance, stimulated, in part, by the Ban Chiang Project, was the initiation in 1984 of the Thailand Archaeometallurgy Project (TAP). 1 Over the next decade TAP focused exclusively on the excavation of prehistoric mining and metal-producing
Prehistoric copper mining in northeast Thailand
207
communities in northeast and central Thailand (Figure 13.1). This work has helped to develop and refine our understanding of the Southeast Asian metallurgical province, at least for the core zone of Thailand. While metal artifacts from excavated contexts are increasing in number (Higham 1996), only in Thailand has problem-oriented archaeology begun to collect the range of data necessary to place copper/bronze metallurgy in its proper socioeconomic context. Therefore, it is primarily from the Thailand evidence that we can begin to make critical assessments in anthropological terms concerning the mining and metal-producing communities which were active in Southeast Asia during the last two millennia BC. The sociopolitical organization of prehistoric Thailand during these two millennia (up until c. 500 BC) has been characterized as decentralized, autonomous and heterarchical (White 1995a; White and Pigott 1996). How does a copper/bronze ‘industry’ fit into such a context? What does the presence of widespread metallurgy and exchange networks suggest about societal organization? Were major prehistoric industrial activities and works undertaken with or without an organizing authority? To begin to answer such questions we must first look at the evidence.
RESOURCE PROCUREMENT The prehistoric copper-mining complex at Phu Lon TAP conducted an extensive survey in 1984 in the ore-rich Loei province along the northeastern edge of the Khorat Plateau in northeast Thailand. This survey intended not only to determine the location of ore sources, but also to identify the nature and degree of exploitation by prehistoric peoples. The only welldocumented prehistoric culture area in proximity to Loei province is known primarily from excavated sites on the ore-poor Khorat Plateau, especially the sites of the bronze-using Ban Chiang Cultural Tradition (White 1982) in the Sakon Nakhon Basin. The most significant result of the TAP survey in northeast Thailand was the documentation of the site of Phu Lon—a large prehistoric copper-mining complex located on the banks of the Mekong River upstream from the Ban Chiang Cultural Tradition sites (Figure 13.1). The uniqueness of Phu Lon and its importance as a major early source of copper in northeast Thailand are underscored by the fact that more than twenty years of economic geological reconnaissance and several archaeological surveys never revealed another copper-mining location of comparable size.2 Other enormous copper ore bodies are known in the Loei region, but there is no evidence that any of them was exploited in antiquity. The TAP excavations at Phu Lon which took place in 1985 revealed the nature and duration of mining activity at Phu Lon (Figure 13.2).
208
Vincent C.Pigott
Figure 13.2 Contour map of Phu Lon Mapping by Udom Theetiparivatra
The ore deposit at Phu Lon is a typical weathered copper sulfide ore body with an indigenous oxide zone, or gossan cap, exposed at the surface (Figure 13.3). Here mining took place in an extensively weathered skarn deposit cut by numerous quartz veins (Pigott and Weisgerber 1998). Malachite, the principal ore mined, occurred in fractures in the skarn rock as well as in association with cavernous quartz veins. Later mining did not destroy all the evidence for prehistoric mining techniques. Located in the upper reaches of the hill are galleries with rounded configurations which suggest mining with stone hammers. In the lower reaches of the mined area, where, it is presumed, the most recent mining operations took place, only deep shafts with relatively
Prehistoric copper mining in northeast Thailand
209
Figure 13.3 The ore body at Phu Lon Diagram by William W.Vernon
narrow (1 m diameter), winding configurations remain. These shafts often bear evidence of the use of pointed, probably metal, picks, in some later period.3 Much of the West Face of Phu Lon or ‘Bald Mountain’ has been cut away and the resulting mining debris is poorly vegetated (Figure 13.4). On the Lower Flat, along the west slope of the hill, mining rubble (mostly broken host rock) is over 10 m in depth and several hundred square meters in area. The substantial tailings which accumulated as a result of mining at Phu Lon suggest that the site was exploited over a considerable period. This rubble is the product of periodic visitations, most probably seasonal, over the centuries. This seasonal activity involved days and months of manual work in which heavy igneous river cobble mauls were used to mine ore from the deposit (Figure 13.5). Hundreds upon hundreds of mining and crushing tools of various types are associated with the Phu Lon deposits. At two separate locations—the Pottery Flat and near the river at Ban Noi— smaller ore pounders and anvil stones were used to crush the ore (and perhaps slag as
f igure 13.4 View of the copper-mining area at Phu Lon Source: Courtesy of the Thailand Archaeometallurgy Project
Figure 13.5 River cobble mining maul from Phu Lon (note flake scarring from heavy use of both ends of the maul) Source: Photo by S.Natapiutu. Courtesy of the Thailand Archaeometallurgy Project
212
Vincent C.Pigott
well) to fines, i.e. fragments only a few millimeters in size. We learned from TAP excavations at the prehistoric copper production settlement of Nil Kham Haeng in central Thailand that finely crushed slag can occur in great quantity mixed in the crushed ore fines (Pigott et al. in press). At another Phu Lon location, Bunker Hill, substantial amounts of fractured quartz fragments indicate mining of malachite in quartz veins close to the surface. Such indications of phased exploitation, together with the considerable depth of mine tailings on the Lower Flat, and the sizeable deposit of crushed ore gravel on the Pottery Flat and elsewhere on the site, combine to support the carbon-14 dates (for the carbon-14 dates from Phu Lon, see Natapintu 1988). Phu Lon was probably exploited over a period of at least two millennia or more, beginning as early as the early second millennium BC. While most of the activity occurred during the first millennium BC, further activity continued in the Historic Period. Excavations on the Pottery Flat revealed a gently sloping, open activity area of more than 200 m2 on the east side of the hill. The most prominent feature of the Pottery Flat was a single stratum of crushed ore fines about 15–50 cm thick. Crushing stations on the Pottery Flat were identified by the presence of small ore pounders and anvil stones in association with small pieces of malachite. In the process of beneficiation, the mined copper ore was being dressed by hand to free the malachite from the host rock. Hand sorting must have occurred all during the process, but most of the ore and host rock was crushed to a rather uniform size a few millimeters in diameter. It is difficult to imagine that such fines were easy to sort by hand, thus it is possible that they were sieved and winnowed in the final stage of beneficiation, prior to smelting or bagging for transportation. O’Brien (1994:173–177), at prehistoric Mount Gabriel in Ireland, discusses the process of ore beneficiation and explains the process of formation of spoil heaps made up of fines. An analogous situation may have prevailed at Phu Lon. Domestic artifacts, including a high density of cordmarked potsherds, occurred in the stratum of crushed material as well. Mixed in the fines were polished, chipped-stone adzes which were being prepared at Phu Lon from shale brought in from elsewhere. Ground stone bracelets were also being manufactured on the spot. Interestingly, no animal bone was found.
PRODUCTION Crucibles on the Pottery Flat Evidence for metal processing on the Pottery Flat includes charcoal, which occurred in amounts suggesting use beyond the household level. In addition to the charcoal, excavations revealed two socketed-tool mold fragments (one ceramic and the other stone) and almost one hundred ceramic crucible
Prehistoric copper mining in northeast Thailand
213
fragments with slag adhering to them. The fragments indicate small crucibles (no intact crucibles were found) whose size suggested, initially, that they were used only for melting copper. However, analysis revealed that the adhering slag contained copper and, in some instances, contained both copper and tin. The presence of both metals suggests they were used in the production of tin-bronze. It is possible, therefore, that ingots of copper and tin-bronze were being produced at Phu Lon. Tin could have been imported from Laotian tin fields (the closest source) in ingot form or as cassiterite (SnO ), the primary ore of tin 2 which is 80% tin by weight. Thus, despite their small size, the crucibles appear to have been used for both smelting copper and alloying it with tin. A rich, low-impurity ore like malachite, a primary ore at Phu Lon, lends itself to crucible smelting. This carbonate ore, if properly dressed, will reduce at relatively low temperatures and yield little or no slag. Tin metal or cassiterite can be easily processed with copper to make bronze and would leave little slag (for a useful discussion of three techniques for making tin-bronze, see Wayman et al. 1988). A small amount of slag has been identified from the Pottery Flat excavations, though much of it may have been crushed up among the ore fines and gangue which compose the layer on the Pottery Flat. By excavation’s end we had a splendid example of a prehistoric copper mine, but one which served what populations and satisfied what demand? William O’Brien (1994:236), in his theoretical spatial analysis of metal production in the zone of production around the Bronze Age copper mine at Mount Gabriel, Ireland, asks a question relevant to the Phu Lon context: Is a model of sequential production within the context of a regional distribution system more appropriate than a local mine-centered one? In this model of sequential production at Mount Gabriel, ore mining and processing progressed to the stage of a crushed malachite concentrate and/or easily transported ingots. At Phu Lon we believe a quite similar mode of production prevailed. While we have no direct evidence that ore was being exported from Phu Lon, we hypothesize that copper and/or bronze ingots (an intermediate product) produced at the mine were transported, as in O’Brien’s model, to a base settlement in the immediate vicinity or further away, to be…refined …and finally cast into prepared moulds to make (items in metal). Intermediate or finished products may also have been transported to settlements located some distance away, with further redistribution taking place from these centres. (O’Brien 1994:236–237) Although we have indications that bronze production at Phu Lon included the casting of socketed tools, we believe that ingot production was the primary objective. We base this observation on the substantial deposit of crushed material (which may contain slag) and the quantity of crucible fragments
214
Vincent C.Pigott
excavated on the Pottery Flat. These hypothesized ingots, produced on the Pottery Flat, would have been transported downstream to consumer villages such as those in the Sakon Nakhon Basin. Tin could have been coming from Laotian sources both to Phu Lon and to these Ban Chiang Culture Tradition villages where alloying could have been accomplished. In O’Brien’s terms, we would see Phu Lon neither as a factory site nor as one among many producing for a local consumption area, but rather as a raw material and intermediate product source within a regional production system. If sufficient amounts of excess metal were available to feed into exchange networks, redistribution would have spread outward from the primary consumer villages. In fact, these villages may have dispatched the mining expeditions to Phu Lon. At the completion of the Phu Lon excavations only a few weak links could be established for a connection to the copper/bronze-using settlements on the northern Khorat Plateau, the only known settlement area in proximity to the mine (south of the Mekong River). However, it was not until we undertook the analysis of the crucibles, which in their fragmentary condition appeared to resemble closely those from the Khorat sites, that a strong direct technological link to Ban Chiang Cultural Tradition sites was established. Their study sheds direct light on the nature of the exploitation of the Phu Lon deposit and may allow some insights into how metal products were being distributed. Analysis of the ceramic crucibles Petrographic analysis of the Pottery Flat crucible fragments by TAP geologist William Vernon showed that the dominant fabric type was clay tempered with crushed rock, rice chaff and crushed slag. His analysis indicated that the crucibles had been manufactured locally (Vernon in press). Though smaller in size, the Phu Lon crucibles resembled closely those from the excavations at Ban Chiang. The Ban Chiang crucibles were subsequently analyzed by Vernon, who discovered a remarkable structural similarity between the two groups of crucibles (White et al. 1991; Vernon 1997). Not only were there examples of slag tempering in the fabric of both groups of crucibles, but both groups had examples with lagged interiors; that is, the crucible interiors had been lined with an insulating layer composed of finely ground quartz-rich sand mixed with clay. This layer was apparently formed by careful washing of the very same clay from which the crucibles were manufactured in order to concentrate the quartz-rich component—perhaps as a slurry. Such a layer, which was not fired prior to crucible use, would have been highly refractory and may have facilitated the removal of the crucible product after smelting. It is difficult to imagine that the presence of this layer in both groups of crucibles is coincidental. Rather, it reveals a quite specific technological approach in a metalworking tradition common to inhabitants at these sites and as such may well serve as a marker of the ‘technological style’ peculiar to northeast Thailand.
Prehistoric copper mining in northeast Thailand
215
While there is technological evidence available on the crucibles from Ban Na Di (Higham and Kijngam 1984:84–92), which resemble closely those from Phu Lon and Ban Chiang, detailed technological studies of the crucibles found at other Ban Chiang Cultural Tradition sites have not been published. The crucibles from Non Nok Tha (Ng 1979: Plate I, h), which remain unanalyzed, also resemble the crucibles under discussion. Thus, the information available suggests that crucibles from the northern Khorat Plateau are typologically similar to those from Ban Chiang and Phu Lon and may well share the same technological style. This comparability of crucible technology on and off the Khorat Plateau strongly supports the argument that Khorat villagers were traveling some distance to Phu Lon to exploit the substantial and readily accessible copper reserves found there. This considerable distance between Phu Lon and settlement sites raises an important question, posed in O’Brien’s terms (1994:239).
CONSUMER INTERACTION What forces, social and otherwise, played a role in shaping the procurement strategies and the production and supply systems associated with the mine at Phu Lon? O’Brien (1994:239) observes that The organization of metal production around Mount Gabriel was to a great extent determined by the way society was organized…. The underlying social fabric influenced many different facets of this production, while control over the latter may also have had long-term social consequences [emphasis added]. O’Brien makes two points here. The first is that correctly modeling the organization of production is fundamental to the proper archaeological interpretation of the nature of procurement strategies. Suffice it to say that with increasing social complexity the potential for complex procurement systems is enhanced but is not a given. In the attempt to delve into this issue, a review of the literature related to ore mining and smelting (production) and mining expeditions (supply systems) yielded a few excellent examples; however, more often than not they hailed from the context of complex social systems (e.g. de Barros 1988; Yener and Vandiver 1993; Shaw 1994), situations not comparable to the social context of northeast Thailand c. 2000–500 BC. However, the literature related to quarrying of stone for tool manufacturing was found to be far more appropriate to the Phu Lon example. Exemplary among these was John Burton’s (1984) examination of quarrying in a tribal society in which mining expeditions were dispatched by the Tungei of the Papua New Guinea highlands to a recently used, twentieth-century stone ax factory. We have here not only an example of what O’Brien argues for above, namely that the organization of society will in turn influence the organization of production, but also a convincing example of how
216
Vincent C.Pigott
one might view the context of resource procurement at Phu Lon. However, before exploring his example further, O’Brien’s second point made above, concerning resource control, merits brief mention. At Phu Lon there is no evidence to suggest that the deposit was the exclusive provenance of a particular group over time; in fact the evidence suggests quite the opposite. At Phu Lon nothing resembling the Mount Gabriel context of restricted access has been observed. This might include, for example, displays by local populations of ‘marked territoriality’ and/or a ‘developed sense of land ownership’ (O’Brien 1994:238). Hence the Phu Lon example would not seem to support arguments for controlled access by socially complex entities (e.g. chiefdoms). The Burton example provides a different conceptual framework for resource ‘control.’ Access to the stone quarries is ‘restricted’ in that they are owned by members of the clan on whose territory the quarries sit. Ownership is facilitated by the proximity between village and quarry just a few kilometers apart. The Tungei live in what Burton (1984:234) termed ‘factory areas’; that is, what is termed a ‘production zone’ in a particularly relevant model of the distribution channel for metal products for exchange designed by Andrew Sherratt (1976). Burton undertook the examination of the Tungei ‘with respect to the concept of chiefdoms and the emergence of centralised societies in the archaeological record’ (Burton 1984:247). While, on the basis of current evidence, the task of reconstructing prehistoric mining expeditions and their social milieu is problematic at best, Burton’s approach to this issue has a particular relevance to how we might view the development of resource procurement at Phu Lon over the centuries. The Tungei example is relevant to this volume in that it is the only example discussed which illustrates a non-resident mining community and the economic mechanisms of mining expeditions. Tungei mining expeditions were responsible for the removal of thousands of tons of rock from the quarries. The rock was shaped into axes, which were the principal focus of an exchange network ranging over a 250 km radius from the quarries. The quarries extend as a line of pits over 2 km, and run along the river Tun. Each quarry site is owned by members of the clan which owns the territory. As at Phu Lon over the centuries, substantial human effort was involved in the process of extraction of stone from these Tungei quarries. Quarrying expeditions were not directed by a central authority, but rather were ‘cooperative ventures in so-called egalitarian, tribal societies’ (Burton 1984: 234). All the men of the seven quarry-owning clans participated in an expedition. Burton stresses that Economic demands ultimately fueled the axe industry but, in the short run, social forces controlled the timing of new expeditions. Social forces also disciplined the workforce into making a constructively cooperative effort. …In this the Tungei belief system played a powerful role. Successful quarrying was attributed to ritual purity and correct axe making magic. (Burton 1984:239)
Prehistoric copper mining in northeast Thailand
217
Conducting an expedition While the quarries were only a few kilometers from the villages, the men moved to quarry camps for the duration of the mining expedition. Work gangs grouped along the same sub-clan lines as those for daily gardening, and the individual gangs worked side by side in the same pit (Burton 1984:240). Four parties of about twenty-five men and boys each worked in the large main pit and two parties at the other two pits. Quarrying was performed with basic tools including hammerstones (as heavy as 2 kg). The quarryman’s basket, easily made and durable, was used in efficient basket chains, facilitating removal of quantities of spoil which was thrown down a talus slope (Burton 1984:241). The basket chains employed large numbers of unskilled laborers as only a few skilled quarrymen could work the quarry face at one time. This process involved all males, thereby increasing output and giving a stake in the final production to every man. The dry months were preferable for expeditions, which lasted three to five months. The first month was devoted to preparing the quarry face area and erecting the camp fence. Quarrying began and, after suitable stone was mined, the shaping of roughouts began. Roughouts were stockpiled until each man’s full share was acquired. Men were not entirely cut off from their home villages as women brought the food to the camps. The end of an expedition was marked by a ritual fight in which women participated. Then the stone axes were collected and everyone went home (Burton 1984:242). Organizing the quarrying The Tungei went to the quarries ‘because the big men said so’ —by force only of their persuasive oratory and acute understanding of the situation (Burton 1984:243). The decision when to quarry was consensus-driven among all clans. Equality between clansmen, a guiding principle of Tungei society, played an important role as men worked cooperatively down to the final ‘shareout’ of the quarried stone. One major reason to join a quarrying expedition stemmed from the ‘desire to compete successfully in the exchange system’ operating with neighboring clans and tribes (Burton 1984:243). Quarrying was communal and there was no room for discord. Expeditions were not sponsored by one big man and no one man owned the production. A respected man divided up the day’s production among each of the clansmen. The point stressed here is that ‘great cooperative works are by no means the sole province of centrally organised societies…. A mass effort by individuals …was perfectly in line with the decentralised, inter-personal nature of trade over the highlands’ (Hughes 1977, cited in Burton 1984:244). Burton demonstrates that ‘acephalous tribal societies can mobilize a large workforce for co-operative, economic ventures…in which leadership involved direction rather than command’ (Burton 1984:244).
218
Vincent C.Pigott
Mining expeditions to Phu Lon could have been mounted from the Khorat villages by walking along the Mekong or paddling upstream. Bamboo rafts would have served to move bags of ingots or malachite concentrate downstream— certainly easier than returning home on foot. However, overland traffic cannot be ruled out. The dramatic alteration of the hill at Phu Lon probably resulted from repeated, seasonal visits of expeditions consisting of, for example, large kinbased groups. Thus, on the basis of the preceding arguments, one need not read into the significant scale of the workings at Phu Lon the organized and controlled efforts of chiefdom-level societies. A second relevant example can be briefly reviewed here. Sagona and Webb (1994) have demonstrated what Aboriginal tribal ochre miners in Australia were able to achieve in terms of substantial workings over the centuries and the complex exchange networks that evolved around the demand for the precious ochre. In Tasmania, ochre mines were owned by tribal groups and access was restricted, but socially complex overarching hierarchical entities were not in control. Mining expeditions were mounted, and they ‘were not necessarily motivated by economics.’ As observed among the Tungei as well, ‘the expeditions themselves and objects exchanged were ritually significant, associated with ceremonies fundamental to Aboriginal culture. Journeys were viewed as continuing the ancestral tradition of epic adventures, and, in turn, the goods acquired became symbols of the expeditions and were attributed great social value’ (Sagona and Webb 1994:139). The relevant point here is that the seeking of ochre was driven by the desire for a precious and sacred commodity. In Thailand, burials in association with prehistoric sites are commonly characterized by the presence of red ochre. For example, from one of the large copper-smelting sites in central Thailand, Non Pa Wai, a burial c. 1500 BC yielded an early copper-base socketed ax and a carved disk of red ochre, which might be a crayon of sorts for body or ceramic painting (Pigott et al. in press). Within a kilometer of Non Pa Wai is a large iron ore body, Khao Tab Kwai, which happens to be a major source of ochre. It is a site which could have served as a regional source of this rare commodity and a central point from which a broad-based exchange network fanned out across the prehistoric landscape. Here too mining expeditions mounted from afar could have traveled great distances to exploit this unique deposit. To judge from the Tungei example it is important to make the point that, at least in the prehistoric context, multiple motivations (e.g. social forces) may have existed behind large-scale efforts and prompted such arduous undertakings. The need for metal to make efficient implements and decorative ornaments may well have been the least powerful of motivations, while belief systems and the symbolic role of metal and its shapes and associated needs constituted equally powerful motivations. Moreover, it is clear that significant amounts of resources were capable of being mined and exchanged through the cooperative efforts—perhaps ritual-driven to varying degrees—of societies of, at best, modest social complexity.
Prehistoric copper mining in northeast Thailand
219
Phu Lon in the context of northeast Thailand The discussion now turns to the broader social and cultural context of northeast Thailand, the context in which Phu Lon played an integral role as a regional source of copper. Khorat prehistoric societies were decentralized and modestly complex (arguably pre-chiefdom until the late first millennium BC). However, they were engaged in extensive exchange involving natural resources (metal ores included) and manufactured products on a substantial basis. O’Brien (1994:237) observes that the organization of metal production varies according to resource procurement strategy, which is in turn linked to both territoriality and resource ownership. One model of procurement which he presents offers the possibility of unrestricted, i.e. direct, access to resources. This type of procurement was not only irregular but also ‘inefficient, unsystematic and only undertaken for short periods by non-specialist labourers,’ and was most probably seasonal and undertaken as an ad hoc enterprise (O’Brien 1994: 237–238). Today Phu Lon is riddled randomly with galleries and shafts. An assessment of the methods of ore prospection and winning indicates that they were unsystematic. Mining methods used at Phu Lon can be likened to those known from the fourth millennium BC copper mine at Rudna Glava in former Yugoslavia, perhaps Europe’s earliest copper mine (Jovanovic 1980). Here geology dictated the path of the mining as miners focused on easily accessible ore outcrops and then followed the sinuous path of copper veins. It was not ‘organized’ mining activity by any means, and the same held true at Phu Lon. O’Brien argues that metallurgy is technologically more complex than lithic procurement, involving more skills, more complex means of obtaining raw materials, and requiring individual specialization within the community. However, in my view, direct access procurement on a seasonal basis by a variety of groups over time (and/or at the same time) is exactly what occurred at Phu Lon as a precursor to low-tech, rather unsophisticated metallurgical production. Further arguments will be presented below on why access to Phu Lon copper ores was not restricted. The mining activity appears to have been opportunistic and inefficient, following surface-exposed outcrops and veins using the most basic mining maul technology. The mauls do not have hafting grooves that would facilitate swinging them with more force and no obvious indications of the use of fire-setting in extant and excavated galleries and shafts were encountered. The model of direct access procurement suits the Phu Lon example well, but what of the alternative to direct access procurement: that is, restricted access to its copper riches? Was the extensive working at Phu Lon the product of a controlled labor force mining for a central authority who restricted access to the deposit as well as to the metal produced? Asking the above question is critical not only to how organizational developments in the metalworking traditions are viewed, but also for an understanding of the larger picture of how prehistoric society itself was organized in this region. At the early
220
Vincent C.Pigott
stages of Phu Lon research, hypotheses focused on issues of access to and control of this unique resource and on reconstructing the role such access plays in the evolution of sociopolitical complexity and, in particular, in the rise of chiefdoms in the Old World. For example, as noted earlier, O’Brien (1994: 238) has argued cogently for control of and restricted access by local population groups to the substantial Mount Gabriel copper deposit during the Irish Bronze Age. As Phu Lon research progressed, it became increasingly clear that at different periods a variety of groups were exploiting several locales at the mine. This was evident in the composition of the archaeological assemblages of the Pottery Flat, Bunker Hill and Ban Noi. The presence of multiple exploiting groups is hardly indicative of a controlled context. Moreover, there certainly was no homogeneous archaeological assemblage across the site (the Pottery Flat orecrushing area is distinct from that at Ban Noi, for example), nor were there indications of divisions of labor or single-task activity areas. Furthermore, there was not much evidence of long-term occupation: we found no burials, animal bone, nor permanent living areas with occupation stratigraphy including living floors, hearths or pits. In addition, in Loei province surveys conducted have not revealed the presence of substantial occupation in the region during the prehistoric period. The lack of substantive evidence to suggest that Phu Lon was controlled for an extended period by a single entity, and the absence of any trace of occupation near this copper source raised the possibility that mining expeditions were mounted from the copper-consuming villages on the Khorat Plateau. Thus, we are exploring the nature of a mining/metallurgical community which was, in effect, non-resident and, given this, its archaeological remains are apt to be more ephemeral than those of a mining community in situ. Research into prehistoric exploitation of natural resources has helped construct a model which is applicable to a study of the structure of the distribution system involving Phu Lon’s products. In his Mount Gabriel example, O’Brien (1994: 238) cites Andrew Sherratt’s (1976) multizoned model for the structure of material production and distribution channels. The Phu Lon example fits well into this model, which describes resource modification and use from procurement through consumption. The four main zones are the source zone, the production zone, the direct contact zone and the indirect supply zone. The source zone of the material, often not located in the zone of agricultural production, would have been accessible by mining expeditions sent to extract the ores and accomplish a ‘preliminary reduction in bulk’ (Sherratt 1976:558); that is, beneficiating ores and/or smelting to produce ingots. Beneficiation was fundamental to the procurement process. Concentrated resources eased the task of the necessary long-distance transportation. At Phu Lon mining evidence and beneficiating debris are abundant while evidence of agricultural settlement lies to the east on the Khorat Plateau. In the production zone the primary exploiters would have been metalworking craft specialists in the northern Khorat Plateau settlements. Evidence suggests that
Prehistoric copper mining in northeast Thailand
221
these were independent rather than attached specialists (White and Pigott 1996:157–158). For example, the shape, size and technology of the ornaments and implements produced do not suggest items appropriate to display on the communal scale. The bangles are modestly executed and the implements serve utilitarian functions which in Costin’s (1991) terms suggest independent specialization (White and Pigott 1996:157). The craft specialists would have dispatched mining expeditions, and would have actively melted and cast ingots into finished products at the settlements. Such products included mostly personal ornaments like bracelets and anklets, and implements including socketed adzes, and also fishhooks and spear points (White and Pigott 1996:155). The processing debris at the sites on the Plateau (e.g. Ban Chiang, Ban Na Di, Non Nok Tha) suggests only the melting and casting of metal, not the smelting of transported ores. This evidence argues for small-scale, kin-based enterprises utilizing small crucibles and molds to cast simple artifacts. These activities were most probably conducted during the dry season as melting and casting under conditions of monsoonal high humidity are not only difficult but can result in poor-quality copper. Thus the modest labor input and investment in ‘equipment’ support the contention that the metal specialists worked independently. These craftsmen produced an array of unsophisticated artifact types easily distributed via exchange networks (White and Pigott 1996:158) in both indirect and direct contact zones. In the direct contact zone production is linked directly by face-to-face contact between the producers and the settlement-based consumers who originate the demand. Supply is maintained through close kinship links among producers and consumers. Exchange may proceed on the basis of services rendered in exchange for products (see Brookfield and Hart 1971:315, cited in Sherratt 1976:558). Costin has shown that demand directly influences production in determining ‘cost parameters, levels of output, appropriate technology and the exclusivity of distribution’ (Costin 1991:13, cited in White and Pigott 1996:157). We have yet to see an elite level of demand reflected in the evidence for metal objects and associated metal production remains in prehistoric Thailand prior to c. 500 BC. In sum, archaeological investigations in northeast Thailand have revealed seasonal copper/bronze production on a modest scale which yielded simple cast utilitarian artifacts for consumers in roughly egalitarian settlements which were minimally differentiated in social terms. Evidence to date does not support the presence of ‘chiefdom’ levels of social complexity for most of the third, second and first millennia BC (White and Pigott 1996). The indirect supply zone comprises those settlements linked by exchange networks along which the desired goods travel, but without access to the production zone. Middlemen transact the trade by moving a variety of material goods (not services) to and from trading partners. This zone is less likely to influence demand than the direct contact zone (Rapport 1968:106, cited in Sherratt 1976:558). Riverine distribution may well characterize the indirect contact zone. White (1995b) (see also Higham and Kijngam 1984) has suggested that the Mekong
222
Vincent C.Pigott
River, flowing right past Phu Lon, and the finely feathered river system which drains the Sakon Nakhon Basin into the Mekong, were critical factors in the uneven distribution of bronze-using settlements in Thailand, including those of the Ban Chiang Cultural Tradition (White and Pigott 1996:158). This pattern of distribution may well be yet another defining characteristic of the Southeast Asian metallurgical province. As an example, we argue that the strength of connections indicated by the use of identical crucible shape and technology suggests a direct link for source-to-consumer interaction between Phu Lon and at least some of the Ban Chiang Cultural Tradition sites. Moreover, the very existence of this technological link between zones may well be due to the ease with which peoples could move via the rivers between these two zones.
CONCLUSION We have argued elsewhere that from the prehistoric period onward, ‘craft specialization—in particular that of copper—has largely developed within a community-based mode of production’ (White and Pigott 1996:151). This is what Cathy Costin (1991:8) has defined as ‘autonomous individual or household-based production units, aggregated within a single community producing for unrestricted regional consumption.’ We maintain that the context of northeast Thailand which has been discussed above may serve as an example of a less intense production system that may have characterized the earlier stages of metal production in the region beginning sometime after c. 2000 BC. Furthermore, we (White and Pigott 1996:157) argue that ‘independent producers and suppliers and a relatively broad and flexible group of consumers are being “served by a production system that minimized production and transaction costs” (Costin 1991:11).’ This model for a community-based mode of craft production ‘has implications for the region’s patterns and structures of accumulation of wealth and the centralization of political power’ (White and Pigott 1996:151). White has argued elsewhere that an economy founded on communitybased craft production helped to foster sociopolitical developments in Southeast Asia which, in turn, promoted heterarchical societal systems during the prehistoric period and later (Crumley 1987; White 1995a). These social systems are ‘often characterized more by flexible hierarchy and lateral differentiation’ (White and Pigott 1996:151). The prevalence of heterarchical contexts in prehistoric Thailand would have served in part to condition the nature of metal production and production organization outlined above.
ACKNOWLEDGEMENTS I would like to acknowledge support of TAP research by the National Geographic Society, the National Science Foundation, the American
Prehistoric copper mining in northeast Thailand
223
Philosophical Society and the University of Pennsylvania Museum. The generous support from individual benefactors Betty Starr Cummin and Maude de Schauensee is also gratefully acknowledged. The comments and criticisms of Julie Pearce Edens and Joyce C.White were of great value in the preparation of this chapter. I alone remain responsible for the ultimate outcome.
NOTES 1
2
3
TAP is co-directed by the author and Surapol Natapintu of the Faculty of Archaeology at Silpakorn University in Bangkok. TAP is a joint collaboration of the Thai Fine Arts Department and the University of Pennsylvania Museum. For publications of the project, see Pigott and Natapintu (1988; in press), Natapintu (1988) and Pigott et al. (1992). The geological reconnaissance of Loei Province was conducted by Udom Theetiparivatra of the Thai Department of Mineral Resources. It was he who first located the mine at Phu Lon. Archaeological site surveys in Loei province have been conducted by teams led by Donn Bayard of the University of Otago (Bayard 1980), by James Penny, then of the University of Pennsylvania (Penny 1982, 1986), by Samsuda Rutnin, then of Australian National University (Rutnin 1988), and by a TAP team composed of Udom, Surapol and myself (Pigott 1984). A study of mining techniques at Phu Lon was conducted by mining archaeologist Gerd Weisgerber of the Deutsches Bergbau-Museum in Bochum, Germany. The report is in the TAP files at MASCA, University of Pennsylvania Museum (see Pigott and Weisgerber 1998).
REFERENCES Bayard, D.T. (1980) The Pa Mong Archaeological Survey Programme, 1973–75. University of Otago Studies in Prehistoric Anthropology 13. Dunedin: University of Otago. Brookfield, H.C. and Hart, D. (1971) Melanesia: A Geographical Interpretation of an Island World. London: Methuen. Burton, J. (1984) Quarrying in a tribal society. World Archaeology 16: 234–247. Chernykh, E.N. (1980) Metallurgical provinces of the 5th-2nd millennia in Eastern Europe in relation to the process of Indo-Europeanization. Journal of Indo-European Studies 8: 317–336. Costin, L.C. (1991) Craft specialization: issues in defining, documenting, and explaining the organization of production. In M.B.Schiffer (ed.), Archaeological Method and Theory 3: 1–56. Tucson: University of Arizona Press. Crumley, C.L. (1987) A dialectical critique of hierarchy. In T.C.Patterson and C.W. Gailey (eds), Power Relations and State Formation, pp. 155–169. Washington, DC: American Anthropological Association, de Barros, P. (1988) Societal repercussions of the rise of large-scale traditional iron production: a West African example. African Archaeological Review 6: 91–113. Gorman, C. and Charoenwongsa, P. (1976) Ban Chiang: a mosaic of impressions from the first two years. Expedition 18(4): 14–26. Higham, C. (1996) The Bronze Age of Southeast Asia. Cambridge: Cambridge University Press.
224
Vincent C.Pigott
Higham, C. and Kijngam, A. (1984) Prehistoric Investigations in Northeastern Thailand, vol. 1. British Archaeological Reports, International Series 231(i). Oxford: BAR. Hughes, I. (1977) New Guinea Stone Age Trade: The Geography of Traffic in the Interior. Terra Australis 3. Canberra: Research School of Pacific Studies, Department of Prehistory, Australian National University. Jovanovic, B. (1980) The origins of copper mining in Europe. Scientific American 242(5): 152–167. Natapintu, S. (1988) Current research on ancient copper-base metallurgy in Thailand. In P.Charoenwongsa and B.Bronson (eds), Prehistoric Studies: The Stone and Metal Ages in Thailand. Papers in Thai Antiquity 1: 107–124. Bangkok: Thai Antiquity Working Group. Ng, R .C.Y. (1979) The geographical habitat of historical settlement in mainland South East Asia. In R.B.Smith and W.Watson (eds), Early South East Asia: Essays in Archaeology, History, and Historical Geography, pp. 262–272. New York and Kuala Lumpur: Oxford University Press. O’Brien, W. (1994) Mount Gabriel: Bronze Age Mining in Ireland. Galway: Galway University Press. Penny, J.S., Jr (1982) Petchabun Piedmont Survey: an initial archaeological investigation of the western margins of the Khorat Plateau. Expedition 24(4): 65–72. ——(1986) The Petchabun Piedmont Survey: an initial archaeological investigation into the prehistory of the western borders of the Khorat Plateau, northeast Thailand. Unpublished PhD dissertation, Department of Anthropology, University of Pennsylvania. Pigott, V.C. (1984) The Thailand Archaeometallurgy Project 1 984 : survey of basemetal resource exploitation in Loei Province, northeastern Thailand. South-East Asian Studies Newsletter 17: 1–5. Pigott, V.C. and Natapintu, S. (1988) Archaeological investigations into prehistoric copper production: the Thailand Archaeometallurgy Project 1984–1986 . In R.Maddin (ed.), The Beginnings of the Use of Metals and Alloys, pp. 156–162. Cambridge, MA: MIT Press. ——(in press) Investigating the origins of metal use in prehistoric Thailand. To be published in D.Bulbeck (ed.), Proceedings of the ‘Conference on Ancient Chinese and Southeast Asian Bronze Age Cultures’ (Kioloa, New South Wales, 8–15February 1988). Canberra: Australian National University. Pigott, V.C., Natapintu, S. and Theetiparivatra, U. (19 92) The Thailand Archaeometallurgy Project 1984–5 : research in the development of prehistoric metal use in northeast Thailand. In I.Glover, P.S.Suchitta and J.Villiers (eds), Early Metallurgy, Trade and Urban Centres in Thailand and Southeast Asia, pp. 47–62. Bangkok: White Lotus. Pigott, V.C. and Weisgerber, G. (forthcoming) Mining archaeology in geological context. The prehistoric copper mining complex at Phu Lon, Nong Khai Province, northeast Thailand. In T.Rehren and A.Hauptman (eds) Metallurgia Antigua. In Honour of HansGest Bachmann and Robert Maddin. Der Auschmitt Beiheft (1998). Pigott, V.C., Weiss, A. and Natapintu, S. (in press) The archaeology of copper production: excavations in the Khao Wong Prachan Valley, central Thailand. To be published in R.Ciarla and F.Rispoli (eds), Proceedings of the ‘Fifth Conference of the European Association of Southeast Asian Archaeologists’ (Rome, Italy, 1993). Rome: L’Istituto Italiano per l’Africa e l’Oriente. Rapport, R.A. (1968) Pigs for the Ancestors. New Haven, CT: Yale University Press. Rutnin, S. (1988) The prehistory of western Udon Thani and Loei provinces, northeast Thailand. Unpublished PhD dissertation, Australian National University, Canberra. Sagona, A.G. and Webb, J.A. (1994) Toolumbunner in perspective. In A.G.Sagona (ed.), Bruising the Red Earth: Ochre Mining and Ritual in Aboriginal Tasmania, pp. 133–151. Melbourne: Melbourne University Press.
Prehistoric copper mining in northeast Thailand
225
Shaw, I. (1994) Pharaonic quarrying and mining: settlement and procurement in Egypt’s marginal regions. Antiquity 68: 108–119. Sherratt, A. (1976) Resources, technology and trade: an essay in early European metallurgy. In G.de G.Sieveking, I.H.Longworth and K.E.Wilson (eds), Problems in Economic and Social Archaeology, pp. 557–581. London: Duckworth. Vernon, W. (1997) Chronological variation in crucible technology at Ban Chiang: a preliminary assessment. Indo-Pacific Prehistory Association Bulletin 16: 107–110. ——(in press) The crucible in copper/bronze production at prehistoric Phu Lon, northeast Thailand: analyses and interpretation. To be published in D.Bulbeck (ed.), Proceedings of the ‘Conference on Ancient Chinese and Southeast Asian Bronze Age Cultures’ (Kioloa, New South Wales, 8–15February 1988). Canberra: Australian National University. Wayman, M., Gualtieri, M. and Konzuk, R.A. (1988) Bronze metallurgy at Roccagloriosa. In R.M.Farquhar, R.G.V.Hancock and L.A.Pavlish (eds), Proceedings of the 26th International Archaeometry Symposium (Toronto, Canada, 16–20 May 1988), pp. 128–132. Toronto: Archaeometry Laboratory, Department of Physics, University of Toronto. White, J.C. (ed.) (1982) Ban Chiang: Discovery of a Lost Bronze Age. Philadelphia: The University Museum, University of Pennsylvania; Washington, DC: Smithsonian Institution Traveling Exhibition Service. ——(1986) A revision of the chronology of Ban Chiang and its implications for the prehistory of northeast Thailand. Unpublished PhD dissertation, Department of Anthropology, University of Pennsylvania. ——(1988) Early east Asian metallurgy: the southern tradition. In R.Maddin (ed.), The Beginning of the Use of Metals and Alloys, pp. 175–181. Cambridge, MA: MIT Press. ——(1995a) Incorporating heterarchy into theory on sociopolitical development: the case from southeast Asia. In R.M.Ehrenreich, C.M.Crumley and J.E.Levy (eds), Heterarchy and the Analysis of Complex Societies, pp. 101–123. Arlington, VA: American Anthropological Association. ——(1995b) Modeling the development of early rice agriculture: ethnoecological perspectives from northeast Thailand. Asian Perspectives 34(1): 37–68. ——(1997) A brief note on new dates for the Ban Chiang cultural tradition. Bulletin of the Indo-Pacific Prehistory Association 16: 103–106. White, J.C. and Pigott, V.C. (1996) From community craft to regional specialization: intensification of copper production in pre-state Thailand. In B.Wailes (ed.), Craft Specialization and Social Evolution: In Memory of V.Gordon Childe, pp. 151–175. University Museum Monograph 93. Philadelphia: University Museum Publications. White, J.C., Vernon, W., Fleming, S., Glanzman, W., Hancock, R. and Pelcin, A. (1991) Preliminary cultural implications from initial studies of the ceramic technology at Ban Chiang. In Indo-Pacific Prehistory 1990, Proceedings of the 14th Congress of the IndoPacific Prehistory Association, Yogyakarta, Indonesia, vol. 2. Bulletin of the Indo-Pacific Prehistory Association 11: 188–203. Yener, K.A. and Vandiver, P.B. (1993) Tin processing at Göltepe, an Early Bronze Age site in Anatolia. American Journal of Archaeology 97: 207–238.
Chapter 14
Small-scale mining and smelting in ancient Cyprus Vasiliki Kassianidou
ABSTRACT Impressive slag heaps located in the mining areas of Cyprus’ Troodos Mountains are relics of the intensive exploitation of the island’s mineral resources and of a large-scale copper industry in antiquity. Yet only one primary smelting site has been excavated: this is a small copper workshop at Ayia Varvara Almyras. Traces of opencast mining and a small mine were located 30 m west of the smelting site. Its proximity to the ancient city of Idalion has raised the possibility that Almyras may have been one of the workshops supplying the kingdom of Idalion with copper. The archaeometallurgical finds include a roasting furnace and a variety of smelting furnaces, but slag is by far the most common find. The archaeological finds are limited to a small quantity of ceramic vessels, mainly of domestic and utilitarian nature, oil lamps, figurines and a spindle whorl. On the basis of the ceramic finds and radiocarbon dating of charcoal samples, the metallurgical activities at Almyras may be placed in three chronological phases, the earliest being the Cypro-Archaic (c. 600 BC) and the most recent the Hellenistic period (c. 150 BC). Before the discovery of Almyras, what we knew of the copper industry suggested that it was characterized by large-scale, centralized, state-run production areas, an impression confirmed in part by the enormous slag heaps. With the discovery of Almyras, however, we are now aware of another type of production centre which seems to have been active in parallel with the larger ones, but for a shorter period of time. This chapter presents the site of Almyras and discusses some of the ideas that have been formulated about the organization of an ancient small-scale copper-producing workshop.
INTRODUCTION Although the earliest references to Cypriot copper (Muhly 1972:204) and the earliest archaeological evidence for mining in Cyprus (Merrillees 1984) date to the beginning of the second millennium BC, it is in the Late Bronze Age (LBA), in other words after the mid-second millennium BC, that large-scale production and long-distance trade of copper begin. From the LBA until the end of the first millennium AD, perhaps even until the Ottoman conquest in AD 1570 (Kassianidou in press), the copper industry was one of the most
Small-scale mining and smelting in ancient Cyprus
227
important sources of revenue for the island, as well as one of the reasons behind internal conflicts and foreign invasions. The abandoned mines were rediscovered in 1912 by C.G.Gunther, an American prospector, who came to the island after finding numerous references in the ancient sources to the rich Cypriot deposits (Rickard 1930:287). The modern industry shipped the ore abroad and did not produce metallic copper by pyrometallurgical techniques. This is an important point, as it means that any slag found on the island is the product of past metallurgical activities. It has been estimated that there are 4 million tons of copper slag in forty different locations spread over the periphery of the Troodos Mountains (Constantinou 1982:19). These impressive slag heaps are remnants of the ancient copper industry and clearly testify to the scale of production in the past. The production and trade of Cypriot copper have been the subject of numerous publications and research projects. Some have focused on the ore deposits and the slags in an effort to reconstruct the smelting technology and to estimate the scale of production in antiquity (e.g. Constantinou 1982; Koucky and Steinberg 1982). Others have tried to understand the organization of the ancient copper industry, mainly that of the Bronze Age, through the study of archaeometallurgical and archaeological finds from settlements and sanctuaries (e.g. Keswani 1993; Knapp 1988; Muhly 1989, 1991; Stech 1985; Tylecote 1982). Finally, a great number have dealt with the trade of Cypriot copper in prehistoric and protohistoric times (e.g. Budd et al. 1995; Gale 1991; Gale and Stos-Gale 1995; Knapp and Cherry 1994; Muhly 1982; Muhly et al. 1988). Most of these studies have considered either technological aspects of the copper industry or the political and cultural organization of the industry as part of a complex society. One aspect which has hardly been touched upon is the study of the mining community. This is due not to a lack of interest as much as to the lack of sources, both archaeological and documentary, for both prehistoric and historical periods. Most fieldwork has focused on urban and religious centres located either on the coast or in the fertile plains of the Mesaoria, but not in the metalliferous mountain region of the Troodos. Some exceptions exist, the most notable being the LBA miners’ camp at Apliki Karamallos, discovered by chance in 1938 during modern mining operations and subsequently excavated in two short seasons before it was destroyed (Du Plat Taylor 1952). Recently archaeologists and archaeometallurgists have started to search actively for information regarding the mining and smelting industry and community in the mining areas (e.g. Raber 1987). Two projects are actively doing fieldwork in the Troodos foothills. The first is the Sydney-Cyprus Survey Project (SCSP), which is conducting a fully interdisciplinary intensive survey in the area between the modern villages of Mitsero and Politiko, in an effort to examine the area’s resource potential and to determine the location of sites in relation to these resources (Knapp et al. 1992, 1994; Knapp and Given 1996).
Figure 14.1 Map of Cyprus with sites mentioned in the text
Small-scale mining and smelting in ancient Cyprus
229
The second is the excavation of Ayia Varvara Almyras, a small mining and smelting site which will be presented in more detail here and will be used to investigate some of the relevant themes of this volume. AYIA VARVARA ALMYRAS The archaeological site consists of a small mine and a smelting workshop (Figure 14.2). It is located 20 km south of Nicosia close to the large ore deposits of Sha and Mathiatis, both of which have been extensively exploited since antiquity. The workshop is on top of a low hill and takes its name, Almyras, from a small creek which runs nearby. On the basis of pottery typology and radiocarbon dating, one may distinguish three phases of occupation at Almyras. These phases correspond to the Cypro-Archaic II period (c. 600 BC), the Classical period (c. 400 BC) and the Hellenistic period (c. 150 BC) (Fasnacht and Kassianidou 1992:79). The mine A geological study of the area located the closest cupriferous deposit just 30 m from the hill where the smelting furnaces are located. It is important that this deposit constitutes the only significant occurrence of chalcopyrite in the area (Fasnacht et al. 1989:75). The copper mineralization, mainly chalcopyrite and pyrite, occurs disseminated or in the form of veins in the highly silicified Basal Group rock. Whole-rock analysis of some samples detected up to 15 wt % of copper. Test trenches led to the discovery of traces of opencast mining (Fasnacht et al. 1990:131) and later to the discovery of a small ancient mine. The area around the mine had been completely filled with soil, eroding from the hill, which covered the ore deposit and thus protected it from mining enterprises of later periods. Geophysical prospection of the area did not locate any other shafts. The mine is not large in size; in fact, it is an adit large enough for only one person. The total height is 2 m and its horizontal extent into the mountain is slightly less (Fasnacht et al. 1996:114). Even so, the depth of the adit created a small practical problem in the spring season of the excavation. After the winter rains, the water table had risen and the mine would flood. This casts some interesting light on ancient mining activities, as presumably ancient miners would have been faced with the same problem. Granted, it would not have been a major difficulty, because the deposit is practically on the surface and draining would not have been difficult. Nevertheless, this may suggest that mining and smelting took place during the summer or even in the autumn, which tend to be dry. The adit had been entirely filled with large rock lumps representing mining waste. The fact that no artefacts or charcoal were found hinders the dating of
Figure 14.2 View of Almyras from a hill located to the north. The mine is on the right of the olive trees (west) and the workshop is to the left (east). In the distance are the mining areas of Sha and Mathiatis
Small-scale mining and smelting in ancient Cyprus
231
the mining activities. No tool marks have been preserved on the rock face. This type of mineralization, however, is extremely hard, and according to Fasnacht et al. (1996:98) would have been accessible only to ancient miners with iron; bronze tools would have been too soft to break this rock. Nevertheless, one should consider the possibility that fire-setting may have been used, even though at the moment we have no solid evidence to support this idea. Fire-setting is a technique in which the rock face is heated by an intense fire and subsequently cooled abruptly. The thermal shock leads to the cracking of the rock and thus greatly facilitates mining (Craddock 1995:33–37). Fire-setting would be unsuitable and even dangerous for underground mining in Cyprus, where modern miners have noted that freshly mined ore oxidizes with unusual rapidity, depleting the atmosphere of oxygen, a phenomenon which would sometimes lead to explosions (Bruce 1948:212). In the case of Almyras, however, where the deposit is located practically on the surface, this would not be a severe problem. Beneficiation The nature of the mining debris suggests that once the chunks containing chalcopyrite veins were broken away from the host rock, they were reduced in size, and after selection were taken for crushing and grinding to the hill where the smelting operations took place (Fasnacht et al. 1992:71–72). During the excavation of the site, numerous lenses and fine layers of oxidized pyrites with a thin upper and lower crust of malachite were found in association with crushing and grinding tools (Fasnacht et al. 1996:117). The latter consist of hammerstones, pounders, mortars and anvils, as well as various types of grinding stones (Fasnacht et al. 1992:71–72). All the stone tools were made of basal group rock locally available on top of Almyras hill (Fasnacht et al. 1989:73–76). Roasting It is well known that sulphide ores must be roasted before they can be smelted. Two roasting furnaces have been discovered at Almyras. The first consists of a rectangular pit cut into the bedrock and lined with coarsely tempered clay (Fasnacht et al. 1989:63). The furnace, dated by carbon-14 to the Late Hellenistic period, is very similar to roasting installations depicted by Agricola (Hoover and Hoover 1950:274). The second roasting-pit is not cut into bedrock but rather is stone built and has a stone slab floor (Fasnacht et al. 1996:97). Smelting Four different smelting furnaces have been unearthed at Almyras. They were all built on bedrock, after the removal of any soil cover. Around the furnaces, horizontal working platforms were cut into the rock. One of the types found
232
Vasiliki Kassianidou
was a double furnace which has two separate chambers approximately 40 cm in diameter. As indicated by the different layers of the clay wall, the furnace was rebuilt at least twice. The other furnaces are free-standing, with an approximate height of 80 cm and a diameter of 40–60 cm (Fasnacht and Kassianidou 1992:83). Scale of production On the basis of available data—in other words, the size of the mine, the copper content of the ores, the amount of slag found on the site, and the number and size of the excavated furnaces—it has been possible to estimate roughly the amount of copper produced in all three phases of occupation at Almyras. This amounts to only 1 ton of metal (Fasnacht et al. 1996:98). This small-scale production stands in contrast to the industrial-scale production witnessed by the large slag heaps, some of which are, according to carbon-14 dating, at least partly contemporary with Almyras (for carbon-14 dates of samples from the slag heaps, see Zwicker 1986). MINERS AND MINING AT ALMYRAS Having presented the archaeometallurgical finds, I now turn to the archaeological material and what this reveals about the people who once worked at Almyras. No permanent architectural remains were found on the site, but this is not surprising. Had the operations been on a larger scale, one would expect some buildings, as indeed were found at Apliki Karamallos (Du Plat Taylor 1952) or at Timna, where a miner’s camp complete with temple has been excavated (Rothenberg 1972, 1988). Almyras is less than two hours on foot from the site of the ancient kingdom of Idalion. The small size of the deposit and its proximity to the city of Idalion may explain the absence of a mining settlement. On the hill was found a chamber, measuring 3m×3m×2m, cut entirely out of the havarra rock (Figure 14.3). The roof had collapsed in antiquity and is only partly preserved at one point to a maximum height of 1.80m. Two post-holes parallel to the western wall may at one stage have offered extra support to the roof or may have supported a roof of some other material (Fasnacht et al. 1996:99). The chamber was practically devoid of finds and its function was, therefore, far from clear. Although initially we investigated the possibility that this was a tomb, it soon became obvious that the feature bore no relation to any known Cypriot tomb types (Fasnacht et al. 1996:101). The discovery of a small figurine in the shape of a ram in the entrance of the chamber and the presence of a large concentration of olive wood in the vicinity led to the hypothesis that this feature may be related to religious practices (Fasnacht et al. 1992:62). A second terracotta figurine was found on the site but, unfortunately, out of context, in an area disturbed by a bulldozer.
Figure 14.3 The rock chamber at Ayia Varvara Almyras
234
Vasiliki Kassianidou
As the excavation of the site continues, it is becoming apparent that there are more rock chambers on the eastern side of the hill (Fasnacht et al. 1996: 98). In the 1995 season a second rock chamber was excavated and in the centre of this feature a small hearth was discovered. Within the hearth a concentration of charred olive stones as well as burnt fragments of pottery were found (Fasnacht et al. 1996:102). This may suggest that rock chambers cut into the soft rock acted as living and storage spaces. From the stratigraphy of the site and the carbon-14 dating we can reconstruct the different stages of occupation and other transformations the site went through. The first visitors to Almyras constructed the rock-cut chamber, as well as a large horizontal platform at the same level as the chamber’s entrance. On this platform, a smelting installation was built (furnace no. 1), the first furnace to be discovered on the site. A second platform, also constructed in the Archaic period, was located in the northern part of the site but was severely damaged by a bulldozer. Unfortunately, we are not able to say how much more of the site has been destroyed by modern landscaping of the area (Fasnacht et al. 1992:68). In the following phase, the Classical period, there was massive earth-moving. Large parts of the Archaic remains were either destroyed or covered up. The roof of the rock chamber collapsed at this time (Fasnacht et al. 1992:67). Working platforms were formed by levelling the ground using the destruction debris. A furnace dating to this period was also built on bedrock, as was the earlier example (Fasnacht et al. 1992:68). The only features dating to the third phase, in other words the Hellenistic period, are metallurgical installations such as the roasting furnace and the double furnace. This is interesting inasmuch as a double furnace recently discovered by the Almyras team at South Mathiatis mine is contemporary with the earliest phase of Almyras and not with this later phase (Fasnacht et al. 1996:106). Preliminary analysis of the pottery assemblage from Almyras reveals that nearly half of the collected sherds come from closed vessels, the most common forms being cooking pots, jars and jugs (Fasnacht et al. 1989:65). The second largest group consists of Plain White ware medium bowls, jugs, lamps, amphorae and jars. There is a relatively small number of sherds from coarse wares, but these total almost a third of the weight of pottery collected, because they come from large vessels such as pithoi and storage amphorae. A small number of fine wares have been found, including a small sherd of Attic Black Glaze ware. It is not surprising that most of the pottery consists of either utilitarian wares or storage vessels. Storage vessels would have been needed to store food and water or other beverages for the workmen, while cooking pots would have been needed to prepare their meals. A few oil lamps suggest that the miners may have worked at night, or at least beyond the daylight hours. A rather intriguing find is a spindle whorl, the discovery of which may indicate other, non-metallurgical activities, perhaps even the presence of a woman at the site. As pointed out by other contributors to this volume, women
Small-scale mining and smelting in ancient Cyprus
235
often formed a significant part of the labour force in mining communities. Female slaves crushing ore appear in Agatharchides’ description of Ptolemaic gold mines, which is mentioned by Diodorus Siculus (Book III. 13). ALMYRAS IN ITS WIDER CULTURAL AND HISTORICAL CONTEXT As pointed out above, the site lies on a low hill not too far from the river near to Almyras and only 30 m from an ore source. Clearly the rich copper deposit, albeit small and uneconomical by today’s standards, was the reason why the site was located at this particular position. The river gave an added advantage inasmuch as water is needed not only for people working at the site but also for the manufacture of refractories, and perhaps also for the beneficiation of the ore. Finally, water would also have been useful if fire-setting had been practised. Access to the site is very easy as it is located at the lower part of the Troodos foothills. Both the archaeological and archaeometallurgical evidence indicate that all periods of occupation at Almyras were fairly short, but we have no way of knowing how long each lasted and whether there was seasonal occupation of the site during each phase. Today, in the general area around the site, there are orchards of citrus trees, olive groves and wheat fields. It is not unreasonable to imagine a similar landscape in antiquity, although there is a strong possibility that the area was much more wooded, if one considers the following statement in Strabo (Geography 14.6.5): ‘Eratosthenes says that in ancient times the plains were thickly overgrown with forests, and therefore were covered with woods and not cultivated.’ The presence of Pinus brutia (pine) and the olive tree is attested from charcoal samples collected during the excavation. These two species represent 95% of the wood identified so far (Fasnacht et al. 1996:110). The point here is that subsistence goods would have been easily accessible to the miners. Although there has been no archaeobotanical study of the site, charred olive stones were found in one of the non-metallurgical hearths excavated on the site. Moreover, apart from a single sheep’s tooth, no faunal remains have been found on the site (Fasnacht et al. 1996:105). We can only estimate roughly the minimum time it would have taken to mine, beneficiate, roast and smelt the ore. Because the rock is so hard, mining and beneficiation of the ore would have required significant time and effort. The enriched ore would then have been roasted, which in itself does not require much physical effort: the reactions are exothermic and once a slow fire had been started in the roasting heap, the ore would have been left to roast by itself. The metalworkers would have had to turn the heap from time to time in order to ensure that all the ore was exposed to the oxidizing atmosphere. Roasting is, however, time-consuming: depending on the size of the heap it may have lasted for days (Hoover and Hoover 1950:273–274; Doonan 1994:88). Roasting was
236
Vasiliki Kassianidou
followed by smelting, which is both time- and energy-consuming. A smelting run may last up to seven hours if not more (Merkel 1990:87), during which time a constant draught of air is necessary to keep the high temperatures needed to drive the reactions and to melt the charge. We do not know how many tuyères were incorporated in each smelter, but even if a single set of bellows was operating for each one there must have been more than one workman operating the bellows. Finally, time and effort were required to collect the necessary fuel. While wood may have been used for the roasting, smelting required charcoal, which would have had to be prepared as well. Although it would be an exaggeration to picture an entire mining community at Almyras, it is clear that even a small-scale operation like this may have employed a fair number of people, who would have stayed there for a minimum of one week. It is, unfortunately, impossible to tell at what time of year the site would have been used, although it seems more probable that it would have been either in the summer or in the autumn: dry weather is necessary not only for the metallurgical processes but also, as argued earlier, for the mining of the ore. What is puzzling is the fact that the miners took the time to construct the rock chambers. If they were staying at the mine for a short period during a dry season, surely a shelter of this type would not have been essential. On the other hand, if we assume that these were skilled miners accustomed to extracting copper ore from hard rock, then the construction of a rock chamber out of the soft havarra would have been easier and quicker than building a shelter. Although there is no other solid evidence apart from its geographical location, we may assume that the workshop would have been situated within the kingdom of Idalion. The lives of the miners, therefore, would have been directly affected by any political and historical developments in this wider community. Cyprus’s geographical position, along with its mineral and timber resources, rendered the island a much-desired territory, and in antiquity as indeed today all major powers of the Eastern Mediterranean occupied the island. In general, however, during the Iron Age and until the Hellenistic period, the conquerors, whether they were Assyrians, Egyptians or Persians, recognized the royal status of the local rulers (Maier 1989:16). The kings were absolute monarchs, not only heads of state but also the high priests and commanders of the army. It is likely that they also owned the minerals and forests of their kingdoms, both of which would have provided them with a substantial revenue (Spyridakis 1972: 119). Copper smelting was one of the main industries, and the trade of copper still played a leading role in the island’s economy (Gjerstad 1948:459). The incorporation of copper workshops within temple complexes at the sanctuary of Aphrodite-Astarte and Cybele at Tamassos (Tatton-Brown 1985:71) suggests that, just as in the Late Bronze Age, so too in the historical periods there was a connection between religion and copper production. Internal strife was not uncommon, and the copper mines were often the centre of discord between warring kingdoms. Idalion too suffered the
Small-scale mining and smelting in ancient Cyprus
237
consequences of war in the fifth century BC. According to a bronze tablet which contains the longest text written in the Cypriot syllabary, Idalion was attacked by a joint force of Persians and Phoenicians from Kition. The date of this attack, which was unsuccessful, is disputed, but it is placed sometime between 475 and 450 BC (Spyridakis 1972:73). Later on, the Phoenicians finally conquered Idalion and then continued their westward expansion. By the third quarter of the fourth century BC, they also controlled Tamassos, another kingdom with abundant copper mines, located some 15 km northwest of Idalion. It has been suggested that the reason behind the Phoenician conquest of these areas was nothing more than a desire to control the copper mines and profit from the production and export of this metal. Tamassos, Idalion and Kition marked the main route the copper metal would have taken from the mines of the Troodos to the port of Kition (Cross 1974:78). A similar pattern emerges in later periods. For example, when Alexander decided to reward King Pnytagoras of Salamis for his assistance in the siege of Tyre, the latter asked to be given Tamassos. Presumably the reason behind Pnytagoras’s request was once again control over the copper mines (Hill 1940: 150). After the death of Alexander, Cyprus became part of the Ptolemaic kingdom of Egypt, losing all vestiges of independence (Maier and Karageorghis 1984:222). The city kingdoms were abolished, and the island was governed as a single entity by a strategos appointed by the king in Egypt (Hill 1940:173). In the Hellenistic kingdoms, and especially during the time of the Ptolemaic kingdom of Egypt, mines belonged to the state (Healy 1978:112); the copper mines of Cyprus would also have belonged to the Ptolemaic ruler. An inscription referring to the antistrategos Potamos, son of Aegyptus, administrator of the mines, may indicate that the strategos was also in charge of the mines (Michaelidou-Nicolaou 1976:101). This has been contested by Hill, who argues that the state’s financial officer would have been the one responsible for the mines (1940:178).
CONCLUSION In all three periods during which the deposit at Almyras was exploited, copper production and trade were central to the island’s economy, and the operation of the mines was centralized and state-controlled. The copper industry was very active, and in fact, up to the discovery of Almyras, it was seen as being characterized by large-scale production areas, represented by enormous slag heaps. These industrial-scale workshops fit well with a state-run industry, glimpses of which we get in the ancient sources (not specifically of the Cypriot industry but, for example, of the Ptolemaic gold mines in Egypt mentioned earlier). Since the discovery of Almyras, however, there is suddenly another type of production centre which seems to have been active parallel to the larger ones. This is a small-scale operation lasting for a brief period of time.
238
Vasiliki Kassianidou
Where does a workshop such as Almyras, however, fit in a model of stateowned and state-operated mines, where the production and the trade of copper are centrally controlled? How much control would the state have over small production centres? Was the copper produced at Almyras turned over to a central authority or was it intended to meet the demands of a local group, or village? If this was the case, did the private entrepreneurs working at Almyras pay some sort of lease to the state, the rightful owner of all mineral resources? Did small workshops like Almyras fulfil local demands for copper, while larger workshops like Kalavasos produced copper in bulk for export? At the moment one can only hypothesize. Current and future fieldwork in the mining districts as well as in the urban centres should shed more light on these matters. An important archaeological discovery from the ancient site of Idalion is one of the largest Phoenician archives found to date. Excavated by Dr Maria Hadjicosti of the Department of Antiquities of Cyprus, the archive consists of economic and administrative texts (Hermary and Karageorghis 1995: 76), some of which may well refer to the exploitation of the mines of the area and the production and trade of copper.
ACKNOWLEDGEMENTS The chapter is based on data collected and published by the Almyras Excavation team directed by Walter Fasnacht. A large number of people have participated in the project, but the following have been in charge of publishing various aspects and finds from the project used in the present chapter: A.Connoly, C.Deslex, W.Fasnacht, D.Gerboth, V.Kassianidou, J.Kunz, R.Morris, M.Schenk and K.Zubler. I would like to thank all of them. Ideas and interpretations given in the chapter, when not otherwise stated, are the author’s and in no way are binding upon any of the other team members.
REFERENCES Bruce, J.L. (1948) Cyprus mines copper again. American Institute of Mining Engineers, Technical Publication 2459. Mining Technology 12: 212–214. Budd, P., Pollard, A.M., Scaife, A. and Thomas, R.G. (1995) Oxhide ingots, recycling and the Mediterranean metals trade. Journal of Mediterranean Archaeology 8: 1–32. Constantinou, G. (1982) Geological features and ancient exploitation of the cupriferous sulphide orebodies of Cyprus. In J.D.Muhly, R.Maddin and V.Karageorghis (eds), Early Metallurgy in Cyprus 4000–500 BC, pp. 13–23. Nicosia: Pierides Foundation. Craddock, P.T. (1995) Early Metal Mining and Production. Edinburgh: Edinburgh University Press. Cross, F.M. (1974) Inscriptions from Idalion in Greek, Cypriot syllabic and Phoenician scripts. In L.E.Stager, A.Walker and G.E.Wright (eds), American Expedition to Idalion, Cyprus. First Preliminary Report: Seasons of 1971 and 1972 . Bulletin of the American Schools of Oriental Research, Supplement 18: 77–81. Boston: ASOR.
Small-scale mining and smelting in ancient Cyprus
239
Diodorus Siculus (1979 [about 50 BC]), vol. 2. Translated by C.H.Oldfather. The Loeb Classical Library 303. Cambridge, MA: Harvard University Press. (First edition 1925 .) Doonan, R.C.P. (1994) Sweat, fire and brimstone: pre-treatment of copper ore and the effects on smelting techniques. Journal of the Historical Metallurgy Society 28(2): 84–98. Du Plat Taylor, J. (1952) A Late Bronze Age settlement at Apliki, Cyprus. Antiquaries Journal 32: 133–167. Fasnacht, W. and Kassianidou, V. (1992) Copper at Almyras: a mining and smelting site in Cyprus. In A.Marangou and K.Psillides (eds), Cyprus, Copper and the Sea, pp. 78–90. Nicosia: Government of Cyprus. Fasnacht, W., Morris, R.S., Schenk, M. and Zubler, K. (1989) Excavations at Ayia Varvara—Almyras. Report of the Department of Antiquities, Cyprus: 59–76. Fasnacht, W., Schenk, M. and Zubler, K. (1990) Excavations at Ayia Varvara—Almyras. Second preliminary report. Report of the Department of Antiquities, Cyprus: 127–139. Fasnacht, W., Connolly, A., Gerbothe, D., Kassianidou, V., Morris, R.S., Rehren, Th. and Zubler, K. (1992) Excavations at Ayia Varvara—Almyras. Fourth preliminary report. Report of the Department of Antiquities, Cyprus: 59–74. Fasnacht, W., with the collaboration of Kunz, J., Deslex, C., Zubler, K., Boll, P., Kassianidou, V., Connolly, A. and Maradi, T. (1996) Excavations at Agia Varvara— Almyras. Fifth preliminary report. Report of the Department of Antiquities, Cyprus: 95– 125. Gale, N.H. (1991) Copper ox-hide ingots: their origin and their place in the Bronze Age metals trade. In N.H.Gale (ed.), Bronze Age Trade in the Mediterranean. Studies in Mediterranean Archaeology 90: 197–239. Jonsered: Paul Åström’s Förlag. Gale, N.H. and Stos-Gale, Z.A. (1995) Comments on ‘Oxhide ingots, recycling and the Mediterranean metals trade’. Journal of Mediterranean Archaeology 8: 33–41. Gjerstad, E. (1948) The Cypro-Geometric, Cypro-Archaic and Cypro-Classical periods. The Swedish Cyprus Expedition 4: 2. Stockholm: SCE. Healy, J.F. (1978) Mining and Metallurgy in the Greek and Roman World. London: Thames & Hudson. Hermary, A. and Karageorghis, V. (1995) Les Phéniciens à Chypre. Chypre au Cœur des Civilisations Méditerranéennes. Dossier d’Archéologie 205: 72–79. Hill, G. (1940) A History of Cyprus, vol. 1: To the Conquest by Richard the Lionheart. Cambridge: Cambridge University Press. Hoover, H.C. and Hoover, L.H. (1950) Agricola: De Re Metallica. New York: Dover Publications. Kassianidou, V. (in press) Hellenistic and Roman mining in Cyprus. To be published in Proceedings of the Third International Congress of Cypriot Studies (Nicosia, 16–20April 1996). Keswani, P.S. (1993) Models of local exchange in Late Bronze Age Cyprus. Bulletin of the American Schools of Oriental Research 292: 73–83. Knapp, A.B. (1988) Copper production and Eastern Mediterranean trade: the rise of complex society on Cyprus. In J.Gledhill, B.Bender and M.T.Larsen (eds), State and Society: Emergence and Development of Social Hierarchy and Political Centralisation, pp. 149–169. London: Allen & Unwin. Knapp, A.B. and Cherry, J.F. (1994) Provenience Studies and Bronze Age Cyprus: Production, Exchange and Politico-economic Change. Monographs in World Archaeology 21. Madison: Prehistory Press. Knapp, A.B. and Given, M. (with contributions by R.Clough, V.Kassianidou, H. Wright, L.Wells and E.Gabet) (1996) The Sydney Cyprus Survey Project (SCSP) — third season ( 1995 ). Report of the Department of Antiquities, Cyprus: 296–337.
240
Vasiliki Kassianidou
Knapp, A.B., Johnson, I., Held, S.O. and Zangger, E. (1992) The Sydney Cyprus Survey Project—first preliminary season (1992). Report of the Department of Antiquities, Cyprus: 319–336. Knapp, A.B., Held, S.O., Johnson, I. and Keswani, P. (1994) The Sydney Cyprus Survey Project (SCSP) —second preliminary season (1993). Report of the Department of Antiquities, Cyprus: 329–343. Koucky, F.L. and Steinberg, A. (1982) The ancient slags of Cyprus. In J.D.Muhly, R.Maddin and V.Karageorghis (eds), Early Metallurgy in Cyprus 4000–500 BC, pp. 117–137. Nicosia: Pierides Foundation. Maier, F.G. (1989) Palaces of Cypriot kings. In V.Tatton Brown (ed.), Cyprus and the East Mediterranean in the Iron Age, pp. 16–19. London: British Museum Publications. Maier, F.G. and Karageorghis, V. (1984) Paphos: History and Archaeology. Nicosia: Leventis Foundation. Merkel, J.F. (1990) Experimental reconstruction of Bronze Age copper smelting based on archaeological evidence from Timna. In B.Rothenberg (ed.), The Ancient Metallurgy of Copper: Archaeology—Experiment—Theory, pp. 78–122. London: Institute of Archaeometallurgical Studies, Institute of Archaeology, University College London. Merrillees, R.S. (1984) Ambelikou Aletri: A preliminary report. Report of the Department of Antiquities, Cyprus: 1–13. Michaelidou-Nicolaou, I. (1976) Prosopography of Ptolemaic Cyprus. Studies in Mediterranean Archaeology 44. Göteborg: Paul Åström’s Förlag. Muhly, J.D. (1972) The land of Alashiya: references to Alashiya in the texts of the second millennium BC and the history of Cyprus in the Late Bronze Age. In V.Karageorghis (ed.), Proceedings of the First International Congress on Cypriot Studies (Nicosia, 14–19 April, 1969), vol. 1: 201–219. Nicosia: Department of Antiquities. ——(1982) The nature of trade in the Late Bronze Age Eastern Mediterranean: the organization of the metals trade and the role of Cyprus. In J.D.Muhly, R.Maddin and V.Karageorghis (eds), Early Metallurgy in Cyprus 4000–500 BC, pp. 251–269. Nicosia: Pierides Foundation. ——(1989) The organisation of the copper industry in Late Bronze Age Cyprus. In E. Peltenburg (ed.), Early Society in Cyprus, pp. 298–314. Edinburgh: Edinburgh University Press. ——(1991) The development of copper metallurgy in Late Bronze Age Cyprus. In N. H.Gale (ed.), Bronze Age Trade in the Mediterranean. Studies in Mediterranean Archaeology 90: 180–196. Jonsered: Paul Åström’s Förlag. Muhly, J.D., Maddin, R. and Stech, T. (1988) Cyprus, Crete and Sardinia: copper oxhide ingots and the Bronze Age metals trade. Report of the Department of Antiquities, Cyprus: 281–298. Raber, P.A. (1987) Early copper production in the Polis region, western Cyprus. Journal of Field Archaeology 14: 297–312. Rickard, T.A. (1930) Copper mining in Cyprus. Transactions of the Institution of Mining and Metallurgy 39: 285–301. Rothenberg, B. (1972) Timna, Valley of the Biblical Copper Mines. London: Thames & Hudson. ——(1988) The Egyptian Mining Temple at Timna, Researches in the Arabah, vol. 1. London: IAMS. Spyridakis, C. (1972) H oikonimiki politiki ton basileon tis Kiprou. Meletai, Dialekseis, Logoi, Arthra, vol. 1: 112–122. Nicosia: Istorikophilologikai Meletai kai Dialekseis. Stech T. (1982) Urban metallurgy in LBA Cyprus. In J.D.Muhly, R.Maddin and V. Karageorghis (eds), Early Metallurgy in Cyprus 4000–500 BC, pp. 105–115. Nicosia: Pierides Foundation.
Small-scale mining and smelting in ancient Cyprus
241
——(1985) Copper and society in Late Bronze Age Cyprus. In A.B.Knapp and T. Stech (eds), Prehistoric Production and Exchange: The Aegean and the Eastern Mediterranean. UCLA Institute of Archaeology Monograph 25: 100–105. Los Angeles. Strabo (1989) The Geography of Strabo, ed. H.L.Jones. The Loeb Classical Library 223. Cambridge, MA: Harvard University Press. (Fir st edition 1929.) Tatton-Brown, V. (1985) Archaeology in Cyprus: 1960–1985 . Classical to Roman periods. In V.Karageorghis (ed.), Archaeology in Cyprus 1960–1985, pp. 60–72. Nicosia: Pierides Foundation. Tylecote, R.F. (1982) The Late Bronze Age: copper and bronze metallurgy at Enkomi and Kition. In J.D.Muhly, R.Maddin and V.Karageorghis (eds), Early Metallurgy in Cyprus 4000–500 BC, pp. 81–100. Nicosia: Pierides Foundation. Zwicker, U. (1986) Ancient metallurgical methods for copper production in Cyprus, Part 2 : Sulphide ore and copper-arsenic-alloy production. Bulletin of the Cyprus Association of Geologists and Mining Engineers 3: 92–111.
Chapter 15
Exploiting the desert frontier The logistics and politics of ancient Egyptian mining expeditions Ian Shaw
ABSTRACT Attempts to study the nature and composition of Egyptian mining communities in the Middle Kingdom (c. 2040–1650 BC) are not common in Egyptology. Although a small number of epigraphic and historical studies have described and discussed briefly the personnel of Egyptian mining expeditions to remote areas, the wider question of the relationships between these mining encampments and pharaonic society as a whole has not been addressed. The overall aim is to understand the significance of mining expeditions in terms of the administration, organization and ideology of pharaonic society. (I) How important was the act of dispatching expeditions for such materials as gold, copper, galena, turquoise or amethyst, and how much of an economic risk did such tasks entail for the overall economy? In this context, the regularity and frequency of the expeditions to particular mines is an important factor, as is the impact of a royal monopoly on the procurement and distribution of certain materials. (2) To what extent were mining expeditions also military operations? The chapter will also discuss the strong links between the string of Egyptian twelfth-dynasty fortresses in Lower Nubia and the exploitation of the Nubian copper and gold resources. It is also necessary to consider the ways in which the desire for precious metals may have fuelled Egyptian ‘imperialism’ at this date. (3) How well adapted to the task were the individual members co-opted into such expeditions? Large numbers of workers were taken on in an ad hoc manner, i.e. in roughly the same manner as those employed on large-scale agricultural, hydraulic or constructional projects. There was, however, probably a nucleus of professional prospectors and miners, who would presumably have been employed by the Egyptian state on a more permanent basis. Archaeological sources provide glimpses of their likely origins and way of life.
INTRODUCTION A small number of epigraphic and historical studies have described and analysed the personnel of the ancient Egyptian mining expeditions dispatched to remote areas in Nubia, Sinai and the Eastern and Western deserts, but the question of the fundamental characteristics and motives of such expeditions has
Logistics and politics of ancient Egyptian mining expeditions
243
not been seriously addressed. In particular, little research has been devoted to the exploration of the political, sociological and ideological repercussions of the Egyptians’ procurement of metals and gemstones. This chapter is an attempt to remedy that situation, concentrating particularly on the gold, copper, turquoise and amethyst mining expeditions of the Old and Middle Kingdoms. The first section summarizes the archaeological and textual data concerning mining in Egypt and Nubia during the early pharaonic period. The second section examines the practicalities associated with assembling a crew of miners: what was the ethnic and sociological composition of the workforce and was there any sense of group identity or ‘community’? The third section is concerned with the political importance of the acquisition of precious materials in terms of the power of the king, the provincial governors and the bureaucracy of central and local government. The final section of the chapter deals with the religious and cosmogonic beliefs that link the world of mining and raw materials with the Egyptians’ ideas concerning the nature of the universe.
MINING IN THE OLD AND MIDDLE KINGDOMS: THE MAJOR SOURCES OF EVIDENCE There is a growing amount of evidence for Egyptian mining and processing of metals and gemstones in prehistoric times. El Gayar and Rothenberg (1995), for instance, have analysed slag from predynastic copper mines at Bir Nasib in the central southern Sinai peninsula, while Beit Arieh (1980) has presented evidence for the presence of Chalcolithic turquoise miners near Serabit elKhadim in the same region. These prehistoric workings, however, are on a comparatively small scale compared with the huge state-sponsored expeditions which began in the Early Dynastic period, and comparatively little is known about the social composition, or even the geographical origins, of the prehistoric miners. From the Old Kingdom (c. 2650–2180 BC) onwards, a great deal of archaeological and textual information has survived concerning mining expeditions in pursuit of copper, gold, turquoise, malachite and a variety of other gemstones. The inscriptions and graffiti associated with mining and processing sites of the Old and Middle Kingdoms provide information on the dates of the expeditions, as well as occasional lists of personnel, and—in rarer instances—detailed narrative accounts of specific expeditions. Texts and paintings from other contexts (particularly funerary ones), such as the sixthdynasty (2345–2181 BC) ‘funerary autobiography’ of Weni at Abydos and the twelfth-dynasty (1985–1795 BC) depiction of the transportation of the colossal statue of Thuthotep at Deir el-Bersheh, are also important in terms of understanding the ancient Egyptians’ own perceptions of quarrying and mining expeditions. Archaeological remains relevant to the period in question have been excavated at such sites as Wadi Dara (Old Kingdom copper mines), Buhen
244
Ian Shaw
(Old Kingdom copper-smelting area), Qubban (Middle Kingdom copper processing), Askut (Middle Kingdom gold processing), Wadi Maghara and Serabit el-Khadim (Old and Middle Kingdom turquoise mines), Wadi el-Hudi (Middle Kingdom amethyst mines) and Gebel el-Zeit (Second Intermediate Period galena mines). In addition, numerous pharaonic gold-mining sites in the Eastern Desert of both Egypt and Nubia have been identified and surveyed during the 1980s and 1990s (see Klemm and Klemm 1994). In the Old Kingdom, the Egyptians were still obtaining much of their gold and copper (and probably also tin, in the form of cassiterite: see Rothe et al. 1996) from within their own borders. The work of Georges Castel at Wadi Dara, to the east of Dendera, has revealed a copper-mining site of the Early Dynastic Period and Old Kingdom. The excavated remains include a workmen’s camp in which the miners appear to have been divided into five or six separate gangs, each with its own installations for ore processing and copper reduction (Castel and Mathieu 1992; Grimal 1993, 1994). Klemm and Klemm (1994) have located at least nineteen gold mines of the Old and Middle Kingdoms, including the important site of Bakari, located a short distance to the north of the Edfu-Marsa Alam road, where there is evidence for almost continuous exploitation of the gold from the predynastic period through to medieval times (Klemm and Klemm 1994, 217–219, fig. 11, plates 31, 35). Lower Nubia was already beginning to be exploited for metals by the Old Kingdom or even earlier, to judge from the Old Kingdom copper-smelting works at Buhen (Emery 1963) and two sixth-dynasty graffiti in the vicinity of the gold mines of Wadi Allaqi (Piotrovsky 1966). The Buhen copper-processing area has been dated by seal impressions to the fourth to sixth dynasties, and it is possible that there may also have been some activity at the site in the first or second dynasty (3100–2686 BC), although the radiocarbon dating of the site is extremely ambiguous (see Shaw 1985:302–303, table 3). In the Middle Kingdom, the primary functions of the string of twelfth-dynasty Egyptian fortresses in the Second Cataract region of Nubia must have been military and commercial, but at least a few of them probably served as bases for the procurement and processing of metals, to judge from the copper slag heaps near Qubban, the limited evidence for gold processing at Askut, the proximity of the Serra fortress to the gold-mining region of Wadi Hagar Shams, and the presence of riverside gold mines at Saras and Duweishat (the latter probably dating to the twelfth dynasty).
WERE EGYPTIAN MINERS A DISTINCT SOCIAL GROUP? It seems clear that a great deal of mining in the pharaonic period was essentially performed by ad hoc seasonal crews rather than full-time professionals. The so-
Logistics and politics of ancient Egyptian mining expeditions
Figure 15.1 Map of Egypt showing locations of sites mentioned in the text
245
246
Ian Shaw
called Satire of the Trades (see Lichtheim 1973:184–192), which presents a jaundiced view of such occupations as soldier, farmer and leather-worker, does not list ‘miner’ as a specific occupation, although ‘copper-worker’ (i.e. smith) does appear, and some scenes in private funerary chapels of the Old Kingdom include detailed depictions of the processes of melting and hammering copper (e.g. the fourth-dynasty mastaba-tomb of Queen Meresankh at Giza (c. 2550 BC—see Dunham and Simpson 1974: fig. 5). Mining seems rarely to have been the main livelihood of any individual, although arguably there is a certain amount of overlap with soldiery, in that most expeditions to remote mines appear to have been accompanied by a military escort. Indeed, there may be something of a definitional problem here, in that the small core of mining professionals were probably more usually defined by the Egyptians themselves as a subset of military or trading officials. The overall leaders of such expeditions in the Old and Middle Kingdoms tended to hold either naval titles or the specific office of ‘god’s treasurer’ (sedjawty netjer), a title which was also held by the leaders of major trading ventures. It might therefore be argued that mining—which was also never depicted by the Egyptians in their tombs or temples—was perhaps regarded not as an independent activity in itself but as a specialized form of a military or commercial expedition. Given this picture of a mining workforce in the Old and Middle Kingdoms which was (a) usually seasonally employed, (b) apparently lacking in explicit professional distinctiveness and (c) frequently dependent on military or naval organization, it is perhaps not surprising that—in contrast to more modern mining—the concept of the long-term mining ‘community’ was somewhat overshadowed by the less tangible and more short-term idea of the ‘expedition’ (see Seyfried 1981). In other words, Egyptian miners were not, at that date, static, specialized groups, but simply a more fluid series of motley crews pulled together for specific endeavours. There was, however, a certain consistency in the composition of such ad hoc forces. It is difficult to assess the degree to which the various different types of individual members of the expedition were trained for their tasks. Large numbers of workers were clearly taken on in roughly the same way as those employed on large-scale, state-organized agricultural, hydraulic or architectural projects. In all these cases, a system of corvée seems to have operated, whereby young men (either Egyptians or local Nubians/Asiatics) could simply be pressganged into such tasks by the provincial or national government. Thus the more detailed lists of mining personnel usually specify the geographical origins of various groups of recruits, who would almost certainly have been used for such non-skilled tasks as digging and transportation (as well as such ancillary work as building and guarding). Each expedition, however, seems to have included at least a small core of professional workers, including ‘prospectors’ (sementyw, represented by a hieroglyphic symbol consisting of the figure of a man holding a bag or stick) and professional miners/quarriers (ikyw), who might well have been employed by the state on a more permanent basis. One prospector, a man
Logistics and politics of ancient Egyptian mining expeditions
247
called Thuthotep, described as a ‘reckoner of gold’, left a graffito recording his presence in the vicinity of the Abrak well, in the Nubian Eastern Desert about 240 km southeast of Aswan (see de Bruyn 1955). The inscriptions at both the Wadi el-Hudi amethyst mines and the Sinai copper/turquoise/malachite mines provide a good opportunity to study the composition of the mining workforces (Fakhry 1952; Seyfried 1981; Shaw and Jameson 1993). The stele of Intef at Wadi el-Hudi, from the reign of the late eleventh-dynasty ruler Mentuhotep IV, describes the expedition members as ‘thousand after thousand’ of local Nubians, but one inscription at Wadi el-Hudi (text WH6, dating to the early twelfth dynasty) gives a genuinely detailed sense of the amethyst-mining personnel (see Table 15.1). Because the lower ranks were recruited mainly from local towns, there were effectively about 1,400 locals being supervised by about 100 officials from the north. Somewhat smaller numbers of workers are recorded at the Serabit el-Khadim turquoise mines, where the workforce during the Middle Kingdom varied from 168 (inscription 124) to 734 (inscription 23), usually accompanied by between 200 and 600 donkeys. The average expedition thus would seem to have comprised about 300 men and 400 donkeys. The manual labourers are often described as ‘stonecutters’, and two inscriptions record the presence of 200 stone-cutters. There appear to be two major differences between Old and Middle Kingdom (2055–1650 BC) mining inscriptions in the Sinai. The Old Kingdom inscriptions frequently include naval titles and mention the presence of specific military detachments, whereas both of these elements are absent in the Middle Kingdom lists. Thus, the mining of copper, malachite and turquoise in the Sinai peninsula was at first, in the Early Dynastic Period and Old Kingdom, a series of militarized forays into dangerous territory, whereas the lack of naval titles or specific mention of military detachments in Middle Kingdom mining inscriptions implies that by then the work was taking place in a context of more peaceful cooperation with the non-Egyptian inhabitants of the central and southern Sinai (see Pons Mellado 1995). The same situation seems to have been true of copper and gold mining in Lower Nubia from at least the Old Kingdom Table 15.1 List of mining personnel given on an early twelfth-dynasty inscription at Wadi elHudi (text WH6)
248
Ian Shaw
onwards. Thus Gratien (1995) discusses the archaeological evidence from various Lower Nubian sites showing that indigenous people (probably late AGroup) were present at the Buhen Old Kingdom settlement, presumably participating in the mining and smelting of copper. Notwithstanding this increasing degree of cooperation with foreigners inhabiting the regions adjacent to mineral sources, the mines exploited by the Egyptians were nevertheless invariably remote from the Nile Valley. Inscriptions sometimes mention that routes through the desert were ‘opened up’ for the workers, and many surviving traces of specially constructed roads have been found in the surrounding areas of mines, quarries and major structures. Henry Fischer (1991) has identified various instances of the Old Kingdom titles ‘master of the roads’ and ‘official of the masters of the roads’, both in the Memphite necropolis and in the mining areas of the Wadi Hammamat and Wadi Abbad (in the Eastern Desert), suggesting that the coordination and maintenance of land routes through the desert was a high priority for the Egyptian administration. In the case of the more extensive mineral sources, which were revisited year after year, considerable time and energy were clearly expended on road construction, the nature of each road being dictated mainly by the bulk and quantities of the minerals, the character of the topography and the materials locally available for road-building.
POLITICAL MOTIVATIONS There are two important pieces of evidence that help to place the launching of mining expeditions in a broader political context. The first is a scene from the tomb of Khnumhotep III (c. 1875 BC) at Beni Hasan, showing a group of Asiatics (probably bedouin of the Eastern Desert) bringing a consignment of lead sulphide (galena, used for eye pigment) from a set of mines on the Red Sea coast into the presence of Khnumhotep, the provincial governor of the Oryx nome in Middle Egypt (see Figure 15.2). Since Khnumhotep’s titles also include the important role of ‘Overseer of the Eastern Deserts’, we are given a strong indication that the local governors controlled and exploited not only the agricultural resources of the Nile valley but also the mineral resources of the adjacent desert regions. It is not clear, however, whether such provincial administrators were generally obtaining metals or stones for their own use and enrichment, or whether they were always acting as representatives of the king, who would presumably have regarded all mineral resources as his own property. If Khnumhotep’s tomb painting gives some sense of the place of metal procurement in provincial administration, a recently discovered text from the city of Memphis (a few kilometres southwest of modern Cairo) indicates the political importance of the royal acquisition of metal and gemstones (specifically copper and turquoise). This very unusual text is a tantalizing fragment of the royal
Figure 15.2 Scene from the painted wall-decoration of the tomb chapel of Khnumhotep III (tomb 3 at Beni Hasan), showing the arrival of a Bedouin tribe bringing a consignment of eye-paint in the form of galena (lead sulphide), probably deriving from mines on the Red Sea coast
250
Ian Shaw
annals of the Middle Kingdom ruler Amenemhat II (Malek and Quirke 1992). It describes the important events of parts of two years in this king’s reign, especially in terms of the acquisition of various commodities from abroad, as a result of trade, war, mining or the collection of tribute. It is therefore effectively a detailed balance sheet of the economic achievements in these two particular years, and several columns are devoted to the sending of an expedition to the ‘turquoise terraces’ (presumably a reference to the turquoise-mining region in southern central Sinai), and a description of the acquisition of large amounts of precious materials such as copper. These are listed alongside the description of spoils of war, and it is noticeable that the quantities of copper acquired by the mining expedition are cited in thousands, whereas those acquired by various military forays are comparatively small amounts. Even the numbers of workers on quarrying and mining expeditions appear to have been at least comparable with those sent on military campaigns. Thus an inscription in the Wadi Hammamat indicates that an expedition consisting of 17,000 workers was sent to quarry siltstone in the thirty-eighth year of the reign of Senusret I (1927 BC). The large scale of such an expedition may be appreciated by comparing it with the army of 20,000 Egyptian soldiers who are said to have fought at the Battle of Qadesh in the reign of Rameses II (1274 BC). It is also possible to study the changes among the higher echelons of the personnel by looking at various figures whose names recur over certain periods of time. We can do this to good effect in the reign of Amenemhat III (1855– 1808 BC) in Sinai, where it appears that most of the repetitions of names are restricted to periods of less than a decade, and it is possible to observe that when a name reappears, the title has usually risen by a rank or two with promotion. Only about 10% of the names recur over the years (see Murnane 1975), and those with lower ranks (e.g. doctors, foremen and workmen) seem to have been sent on a larger number of successive expeditions than their superiors. The lower-ranked officials were also less likely to be promoted, whereas higher officials, such as treasury representatives and scribes, were less likely to return, but more likely to have been promoted if they did so. The overall impression is of a steady change in personnel, especially in the higher-ranking jobs. There are at least two possible reasons for this: first, it was an unpopular job, and second, the king may have deliberately encouraged a steady turnover in key positions, so as to avoid any accumulation of power in the hands of a few individuals. The control over precious raw materials could be as powerful a tool in the hands of officials as it was in the king’s own hands. It is not clear precisely how much importance, whether symbolic or practical, was attached to the act of dispatching a mining or quarrying expedition in relation to the pharaonic economy as a whole. Nor is it known how much of an economic risk such enterprises entailed. An important factor in attempting to answer these two questions is the evidence concerning the regularity and frequency of the expeditions to particular mines. In the Middle Kingdom at
Logistics and politics of ancient Egyptian mining expeditions
251
least 39 expeditions were sent to the turquoise mines in Sinai, 65 into the Wadi Hammamat (for siltstone) and 15 into the Wadi el-Hudi (for amethyst). The average interval between expeditions was 31 years in Sinai, 35 in the Wadi Hammamat and 17 in Wadi el-Hudi. Figures 15.3a and 15.3b present the numbers of expeditions dispatched per reign and per regnal year respectively. One of the most interesting aspects of Figure 15.3a is the close similarity between the numbers of expeditions sent to the three destinations in each reign, particularly during the early and middle twelfth dynasty, i.e. from Amenemhat I to Senusret III. Apart from an unusual surge of turquoise expeditions to Sinai in the reign of Amenemhat III, there is a surprising consistency both within and between reigns. The procurement of siltstone from Wadi Hammamat appears to be the steadiest and most consistent of the three activities. This is even more apparent if we look at Figure 15.3b, the bar chart for expeditions per regnal year, which shows us that for most of the Middle Kingdom there was no more than one expedition every five years sent to each of the three destinations. We can see that in most reigns during this period there was a fairly constant ratio between reign length and number of expeditions. It is interesting to note that the only two real exceptions are Mentuhotep IV and Amenemhat IV, both of whom undertook an amount of mining that was much higher than we would expect for the lengths of their reigns. Is it a coincidence that both ruled at the end of a dynasty, at a time when there may have been greater political instability? It is likely that mining and quarrying expeditions, like military campaigns, were an essential means of bolstering central authority, both because this was a way of demonstrating the king’s ability to dominate even the most remote regions of his kingdom, and because the products of mining and quarrying played an important role in the king’s relationship with his courtiers. Thus we know from funerary inscriptions that the loyalty of high officials was regularly rewarded with gifts of funerary equipment such as sarcophagi or jewellery, the raw materials for which were invariably obtained through royal mining expeditions. It is not clear whether the king actually exercised a monopoly on the largescale mining of most metals and gemstones, but in practice few other individuals would have ever commanded the resources or manpower to do so. In addition, it might be argued that mining expeditions would have served as virility symbols, particularly for kings who were experiencing political or socio-economic pressures. In the case of the last eleventh-dynasty ruler, Mentuhotep IV, they are virtually the only written records that his reign took place at all, and it is perhaps significant that his vizier Amenemhat—the man whom many Egyptologists identify as the man who was to become the first ruler of the twelfth dynasty, Amenemhat I—is encountered in a rock inscription along the Wadi Hammamat, identifying him as the leader of a quarrying expedition said to consist of 20,000 workers (Breasted 1906:436– 450). The control of such expeditions must have been one of the more effective routes to political power.
252
Ian Shaw
Figure 15.3 Bar charts showing (a) the total number of expeditions sent to Wadi el-Hudi, Sinai and Wadi Hammamat during different reigns (statistics based on the number of surviving dated texts at the mines and quarries in question); and (b) the numbers of expeditions per regnal year
Logistics and politics of ancient Egyptian mining expeditions
253
MINING AS A RELIGIOUS DUTY The discussion has so far centred on the practical socio-economic and political aspects of Egyptian mining expeditions. Although the physical act of mining metals or gemstones is rarely either depicted or described in Egyptian temples or tomb chapels, a number of texts indicate that precious metals and stones were an important part of the Egyptians’ view of the universe. As in other parts of precolonial Africa, gold was regarded by the Egyptians not as an inert material but as a living substance; the Pyramid Texts (Faulkner 1969) therefore include the statement of the belief that the flesh and bones of Egyptian deities were composed of gold and silver respectively. This connection between metals and divine beings meant that gold and silver were often employed—by those who could afford it—in the production of funerary equipment, as exemplified by the golden mask and coffins of Tutankhamun (1335–1323 BC). In the fifth-dynasty tomb of Iy-Mery (c. 2400 BC) at Giza (G6020), an inscription points out that the shape of the nebw sign was being imitated by pairs of dancers in the funerary dance known as the tcheref. There are numerous other instances of the paramount role played by gold, particularly in the royal funerary cult. For example, the sarcophagus chamber in the New Kingdom royal tomb was known as the ‘house of gold’, while the goddesses Isis and Nephthys were often shown at the ends of sarcophagi or coffins kneeling on the hieroglyphic sign for gold (nebw). The deities Horus, Hathor and Sopdu were particularly closely associated with the protection of mining expeditions in remote regions. They were regarded as manifestations of divine goodwill in foreign lands, sometimes effectively comprising Egyptian versions of foreign gods or goddesses, as in the case of Hathor, Lady of Byblos, who was actually the local goddess Baalath Gebal. In the Middle Kingdom, Hathor, Lady of Nekhent, was the patroness of the Gebel el-Asr gneiss quarries in the desert west of Tushka; Hathor, Lady of Mefkat, presided over the acquisition of copper, turquoise and malachite in the Sinai; while the twelfth-dynasty fortresses at Buhen, Aniba and Qubban were the cult-centres of Hathor, Lady of Iken, Hathor, Lady of Abshek, and Horus, Lord of Meha, respectively. In the New Kingdom a temple dedicated to a local form of Hathor was established in the vicinity of the copper mines at Timna in the Wadi Arabah (Rothenberg 1972). In view of this close association between Hathor and mining, it is perhaps not surprising that our most detailed information concerning the Egyptians’ cosmological views on mining has been found on the walls of the Ptolemaic and Roman temple of Hathor at Dendera. In several of the rooms of the temple, rulers are shown presenting precious metals and stones to the goddess, but the bases of the walls of the so-called ‘silver-room’ (or side-room XI) are decorated with a series of kneeling figures, each representing regions that produced particular metals or precious stones (see Figure 15.4, based on Mariette 1870: pl. 12). The accompanying hieroglyphic texts make it clear that the purpose of this depiction was not
Figure 15.4 Annotated copy of the decoration along the bases of walls in the ‘silver-room’ at the temple of Hathor, Dendera, showing anthropomorphic representations of the regions from which particular metals or precious stones derived Source: After Mariette (1871: pl. 12)
Logistics and politics of ancient Egyptian mining expeditions
255
simply to stress that precious gifts were being offered to the goddess, but that the entire mineral universe—gold, turquoise, emeralds, jasper, calcite, q c mineral (black schist?), copper, carnelian, green feldspar, galena, lapis lazuli, silver—was being reassembled in microcosm as a means of enabling the temple to function as a reflection of the cosmos (see Aufrère 1991). The significance of this situation in terms of Egyptian mining was the perception that the procurement of minerals was not merely an act of economic or political exploitation but also—perhaps most importantly—an act of religious devotion through which the ruler was able to fulfil his obligation to maintain the harmony of the universe and to attempt to re-enact the process of creation itself.
DISCUSSION AND CONCLUSION The above discussion has dealt primarily with the interaction between the temporarily desert-based miners and the Nile Valley-based agricultural communities from which they were drawn. This interaction essentially corresponds to the three interaction spheres of a mining frontier identified by Hardesty (1988: 1–5): material, social and informational. The material sphere is dominated not only by the archaeological remains of the miners’ settlements but also by the specially built roads or traditional caravan routes along which the expeditions (and their essential supply lines) travel en route to and from the mining areas. These desert remains—the physical records of the prospection, extraction, processing and transporting of metals and gemstones—are supplemented by the survival of artefacts incorporating the precious materials themselves within the settlements and cemeteries of the Nile Valley and beyond. Hardesty’s social sphere is characterized principally by the textual and archaeological evidence for the ethnic and social mixture of the Egyptian miners, the vast majority of whom were neither professionally trained nor motivated by any likelihood of personal gain: the core of the workforce was made up of a number of discrete groups of young men taken from agricultural communities. The ancient Egyptian manner of mining seems to have been characteristically ‘African’ in that it was primarily a seasonal occupation of agriculturalists rather than purely the preserve of professionals (see Herbert 1993; this volume, Chapter 9). However, pharaonic mining settlements also presage more modern mining communities in North America and Australia in terms of their ethnic mix. Thus, Canaanites worked alongside Egyptians in the turquoise and copper mines of central Sinai, while groups of Midianites and the local people of the Negev participated in the copper mining at Timna. Similarly, in Lower Nubia, many ‘C-Group’ Nubians were almost certainly involved in the Egyptians’ procurement of gold, copper and amethysts. The mining expeditions of the pharaonic period were therefore scenarios in which Egyptians came into direct—often cooperative—contact with foreigners.
256
Ian Shaw
Concerning Hardesty’s informational sphere, it may be defined in the Egyptian context as a combination of religious, ideological and political links which—like much of pharaonic society—hinged on traditional perceptions of the king’s duties and obligations. On a religious level, the king’s sponsorship of mining was part of his role as mediator between the worlds of gods and men; as the reliefs in the ‘silver-room’ at Dendera indicate, it was part of the king’s fundamental role to seek out the precious constituents of the universe and, in a sense, ‘restore’ them to the gods. On a level which, from a modern perspective, might be regarded as political pragmatism, the king also relied on mining expeditions both as a demonstration of his control over the most far-flung frontier regions of Egypt and as a means of obtaining the material resources with which to reward his loyal supporters and therefore maintain administrative stability. In this sense, the Egyptians’ attitude to mining was very closely related to their views on warfare, foreign relations and international commerce. The sending of an expedition—whether for mining, trade or conquest—was concerned not only with material results (i.e. the acquisition of such commodities as gold, copper, incense, slaves or territory) but also with the reaffirmation of the Egyptian king’s absolute control over the universe, which he wielded on behalf of the gods. The religious aspects of mining also derive from geographical factors. Old and Middle Kingdom Egyptian mining expeditions were virtually always situated physically in the remote desert, which was a quintessentially liminal and ‘frontier’ location. The mining expedition therefore was ritually charged not only because of the religious and political resonances of the minerals but also because of the nature and associations of the desert itself, which was symbolic of chaos and the unknown. Above all, the desert was a place where strange and extraordinary events might occur, particularly in connection with the provision of water or the quality of the materials being mined. In a turquoise-mining inscription of the sixth regnal year of the twelfth-dynasty ruler Amenemhat III (1849 BC), the ‘god’s seal-bearer’ Horurre describes how, although he arrived at Serabit elKhadim in the summer ‘when it was not the proper season for coming to this mining region’, he was nevertheless rewarded with reserves of good-quality turquoise through the ‘power of the king’ and the benevolence of the goddess Hathor (see Parkinson 1991:97–99). In a more elaborate example of divine intervention, a group of rock inscriptions relating to the Wadi Hammamat expedition of the vizier Amenemhat, in the second year of the eleventh-dynasty ruler Mentuhotep IV (1991 BC), include an episode in which a gazelle is said to have given birth to its young on the quarried stone itself, thus supposedly ensuring the arrival of a rainstorm which filled the local well (Breasted 1906:436). The mining expedition in the desert could be a stark backdrop for dramatic events, and it is this sense of theatre in the Egyptian mining inscriptions that foreshadows much later ideas of the miner as pioneer or hero (see Moser 1996).
Logistics and politics of ancient Egyptian mining expeditions
257
Very similar literary tropes, involving divine assistance or epiphany, are also found in Egyptian narratives of military conquest, thus helping to reinforce the idea that mining expeditions were part of the same distinctive culture as military or commercial ones. The various types of expedition not only shared similar personnel, organization and administrative systems (including the extensive use of naval terminology), but also had similar aims and motives: on an ideological level, the fulfilment of the king’s pact with the gods, and on a socio-economic level, the procurement of those exotic resources which formed the basic currency of contracts and transactions between the members of the Egyptian administrative hierarchy.
REFERENCES Aufrère, S. (1991) L’Univers minéral dans la pensée égyptienne, 2 vols. Cairo: Institut Français d’Archéologie Orientale du Caire. Beit Arieh, I. (1980) A Chalcolithic site near Serabit el-Khadim. Tel Aviv 7: 45–64. Breasted, J.H. (1906) Ancient Records of Egypt I. Chicago: University of Chicago Press. Castel, G. and Mathieu, B. (1992) Les mines de cuivre du Ouadi Dara: rapport préliminaire sur les travaux de la saison 1991. Bulletin de l’Institut Française d’Archéologie Orientale 92: 51–65. de Bruyn, P. (1955) A graffito of the scribe Dhuthotpe, reckoner of gold, in the southeastern desert. Journal of Egyptian Archaeology 41: 121–122. Dunham, D. and Simpson, W.K. (1974) Giza Mastabas I: The Mastaba of Queen Meresankh III. Boston: Museum of Fine Arts. El Gayar, El Sayed and Rothenberg, B. (1995) Investigation of predynastic and Old Kingdom ore and slag samples from Sinai. Abstracts: International Conference on Ancient Egyptian Mining and Metallurgy and Conservation of Metallic Artifacts, Cairo Egypt. Cairo: SCA Press. Emery, W.B. (1963) Egypt Exploration Society: preliminary report on the excavations at Buhen, 1962. Kush 11: 116–20. Fakhry, A. (1952) The Inscriptions of the Amethyst Quarries at Wadi el-Hudi. Cairo: Government Press. Faulkner, R.O. (1969) The Ancient Egyptian Pyramid Texts. Oxford: Oxford University Press. Fischer, H. (1991) Sur les routes de l’Ancien Empire. Cahiers de Recherches de l’Institut de Papyrologie et Égyptologie de Lille 13: 59–64. Gratien, B. (1995) La basse Nubie à l’Ancien Empire: égyptiens et autochtones. Journal of Egyptian Archaeology 81: 43–56. Grimal, N. (1993) Travaux de l’IFAO en 1992–1993 . Bulletin de l’Institut Française d’Archéologie Orientale 93: 482–488. ——(1994) Travaux de l’IFAO en 1993–1994. Bulletin de l’Institut Française d’Archéologie Orientale 94: 385–480. Hardesty, D.L. (1988) The Archaeology of Mining and Miners: A View from the Silver State. Society for Historical Archaeology Special Publication Series no. 6. Pleasant Hill, CA: SHA. Herbert, E.W. (1993) Iron, Gender, and Power: Rituals of Transformation in African Societies. Bloomington: University of Indiana Press. Klemm, R. and Klemm, D.D. (1994) Chronologischer Abriß der antiken Goldgewinnung in der Ostwüste Ägyptens. Mitteilungen der Deutschen Archäologischen Instituts Abteilung Kairo 50: 189–222.
258
Ian Shaw
Lichtheim, M. (1973) Ancient Egyptian Literature I: The Old and Middle Kingdoms. Berkeley and Los Angeles: University of California Press. Malek, J. and Quirke, S. (1992) Memphis 1 991 : epigraphy. Journal of Egyptian Archaeology 78: 13–18. Mariette, A. (1871) Denderah II. Paris: Librairie A.Franck. Moser, S. (1996) Mining as metaphor: the culture of archaeology. Paper presented at the conference: Social Approaches to an Industrial Past: The Archaeology and Anthropology of Mining (Rockefeller Foundation, Bellagio Center, Bellagio, Italy, 22–26July 1996). Murnane, W.J. (1975) A note on the personnel of the Sinai expeditions under Amenemmes III . Göttinger Miszellen 15: 27–33. Parkinson, R. (1991) Voices from Ancient Egypt: An Anthology of Middle Kingdom Writings. London: British Museum Press. Piotrovsky, B.B. (1966) Two Egyptian inscriptions of the Sixth Dynasty in Wadi Allaki. Vestnik Drevnei Istorii 95: 80–82. Pons Mellado, E. (1995) Egyptian mining in Sinai during the Middle Kingdom. In Abstracts of Papers presented at the Seventh International Congress of Egyptologists, Cambridge, September 1995, ed. C.Eyre, pp. 142–143. Oxford: Oxbow Books. Rothe, R.D., Rapp, G. and Miller, W.K. (1996) New hieroglyphic evidence for pharaonic activity in the Eastern Desert of Egypt. Journal of the American Research Center in Egypt 33: 77–104. Seyfried, K.J. (1981) Beiträge zu den Expeditionen des Mittleren Reiches in die Ostwüste. Hildesheimer Ägyptologische Beiträge: Hildesheim. Shaw, I. (1985) Egyptian chronology and the Irish Oak calibration. Journal of Near Eastern Studies 44: 295–317. Shaw, I. and Jameson, R. (1993) Amethyst mining in the Eastern Desert: a preliminary survey at Wadi el Hudi. Journal of Egyptian Archaeology 79: 81–97.
Chapter 16
Gold-miners and mining at Bir Umm Fawakhir Carol Meyer
ABSTRACT Three seasons of archaeological survey at Bir Umm Fawakhir in the Eastern Desert of Egypt have identified the site as a fifth- to sixth-century Byzantine/Coptic gold-mining town (not a Roman caravan station), have mapped in detail 152 out of approximately 216 buildings, and have identified peripheral features such as wells, guard posts, ancient paths and roads, outlying clusters of ruins, cemeteries, quarries and mines. Gold was extracted from quartz veins in Precambrian granite, so mining, ore reduction and washing required a large labor force; the population of ancient Fawakhir is estimated at over 1,000. Contrary to the first century BC account of Diodorus, there is no compelling reason to believe the miners were criminals, war prisoners or slaves. The lack of town planning, formal defenses or town walls, the uniformity of finds within the site, and the houses’ individual grain silos do not suggest control of prisoners. The town, in a hyper-arid desert, must nonetheless have been supplied entirely from the Nile. Given the government’s critical need for gold at that time, the Fawakhir gold must have been monitored by imperial authorities in the Thebaid. The labor force was probably also supplied by the government, perhaps on the pattern of military recruitment.
INTRODUCTION AND BACKGROUND The Bir Umm Fawakhir Survey Project of the Oriental Institute of the University of Chicago completed its third field season in the central Eastern Desert of Egypt in January 1996. This chapter, however, is not about what was found in the past three seasons of archaeological survey so much as what the site can tell us about gold mines and mining in Coptic/Byzantine Egypt, and more specifically about the miners themselves. As this is a relatively new site, however, it is necessary first to present a brief overview of some of the empirical results and then to discuss mining, miners and their milieu. Bir Umm Fawakhir, long called a Roman caravan station, can now be identified as a fifth- to sixth-century AD Coptic/Byzantine gold-mining town. It
260
Carol Meyer
lies in the Precambrian mountains of the central Eastern Desert of Egypt, almost exactly halfway between the Nile at Quft (ancient Coptos) and the Red Sea at Quseir, about 65 km or two and a half to three days by camel. No agriculture has ever succeeded in this hyper-arid desert, though the rare flash floods do fill pools and cisterns from time to time and do produce a sudden growth of desert plants such as camel thorn. The site of Bir Umm Fawakhir itself is about 5 km northeast of the Wadi Hammamat, just past a natural gate in the mountains (Figure 16.1). The main settlement of Bir Umm Fawakhir lies in a long, narrow wadi not visible from the modern road, but having turned a spur in the hills, one can see the ancient mining town spread out over approximately one-half kilometer. The layout of the town along the sandy wadi bottom that serves as a main street and the individual nature of the house units can clearly be seen. As virtually nothing apart from natural decay has happened to the site since it was abandoned, some of the houses stand a meter or more in height, and many preserve doors, stone benches, wall niches for storage, and other built-in features. In all, 152 out of an estimated 216 buildings in the main settlement have now been mapped in detail. The expectation was that an accurate map would help clarify the layout of the town, and this has indeed proven to be the case (Figure 16.2). The basic pattern of the houses is a two- or three-room structure, but these are often joined into larger agglomerated buildings (Meyer 1993a, b, 1995a, b, 1996; Meyer et al. in preparation). The largest structure mapped to date, Building 106, has twenty-two rooms. Scattered around the houses is a series of one-room outbuildings, both a smaller rounded type and a larger rectangular type. Without excavation it cannot be determined whether they were used for storage, as workshops, as kitchens, for animals, as latrines, or something else altogether. Ancient trash heaps lie adjacent to many of their houses or upslope on the skirts of the cliffs. During the 1993 season we plotted an area called the ‘Plaza’: this is a grouping of houses opening into a central sandy space rather than onto the main wadi street. With this as a clue, three other groups of houses functioning together could be picked out (Meyer et al. in preparation). I shall not go into details of the two methods used (square meters of roofed house space per person and estimates of number of persons per household), but a population estimate of a little over a thousand people in the main settlement has been derived, assuming that nobody lived in any of the eight outlying clusters found so far (Meyer 1993b; Meyer et al. in preparation). This is a far larger population than any modern one on the Hammamat road. Four cemeteries have been detected thus far on the ridges around the site. The graves are simply natural clefts in the granite or small cists built of granite slabs, roofed over with granite cobbles. All the burials have been thoroughly looted, but the considerable amount of pottery scattered around them indicates that the graves are of the same date as the main settlement. The pottery may pertain to funeral or memorial feasts; the graves are barely large enough for a body, and if Christian are less likely to have burial goods anyway.
Gold-miners and mining at Bir Umm Fawakhir
Figure 16.1 Vicinity of Bir Umm Fawakhir Source: Based on Klemm and Klemm (1993)
261
Figure 16.2 Main settlement at Bir Umm Fawakhir—areas mapped in 1992, 1993 and 1996
Gold-miners and mining at Bir Umm Fawakhir
263
So far all the buildings appear to be domestic. (One possible exception may be Building 91, which has a large internal circular feature that just might be a community bake oven. Building 65, which may enclose a similar feature, has a considerable amount of ash.) To date no churches, fortifications or large central buildings or warehouses or workshops have been found. Crosses stamped on the pottery and Chi Rhos (X/R) on the wine jars point to a Christian or Christianized population, but there is as yet no church. There is, however, some reason to think that this structure, along with other official buildings, may have been located near the modern settlement and road where the wadi wash is heaviest. The lack of defenses is surprising, particularly in the desert, where security is often a concern, and especially considering the product of the community, gold. The survey has, however, identified a couple of guard posts high on the ridges overlooking the main settlement. The larger post has no formal defenses, only a few rough walls for shelter from the sun and wind, but it is marked as ancient by graffiti and it does command an excellent view of all three roads leading to the wells, many of the mines, and much of the main settlement (Figure 16.3). Eight outlying clusters of ruins of the same date as the main settlement have been identified; the houses have the same type of dry-stone masonry and the pottery is the same if less abundant. One of the clusters, Outlier 2 on the Roman road and between the wells and Quarry 1, was investigated in more detail in 1996 because of its exceptional preservation. One house appears to be preserved to its original height, about 2 m, not tall by modern standards. Also, two cylindrical grain silos were noted next to two of the houses. The silos are made of granite cobbles set in heavy mud plaster and are of interest because such features have not been detected elsewhere at the site and because the association with individual houses does not suggest day-to-day central control over grain rations.
BIR UMM FAWAKIR: SITE FUNCTION Bir Umm Fawakhir means ‘Well of the Mother of Pots,’ and it is literally impossible to walk over the site without crunching sherds. Although pottery studies are well advanced, here it will only be noted that in the absence of any useful written evidence, the sherds are the basis of the fifth- to sixth-century AD date for the site (for detailed descriptions and analyses, see Heidorn in Meyer 1995a, b; Meyer et al. in preparation). Epigraphic material is not wholly lacking, but it is so far limited to the dipinti, dockets painted in red ink on the shoulders of wine jars. The jars are of a common type called Late Roman A, and the inscriptions give some indication of their place of manufacture, contents and a date. All the dipinti recovered to date are highly cursive and mostly fragmentary (Wilfong in Meyer 1995a, b; Meyer et al. in preparation). Apart from sherds, and crushing and grinding stones, which will be discussed below, surface finds
Figure 16.3 View from the guard post over the northeastern end of the main settlement showing houses laid out on either side of the wadi bottom. The ‘Plaza’ is the circle of houses towards the middle of the left edge of the photograph. Granite Quarry 2 lies at the foot of the sunlit knob at the right Photograph by Carol Meyer
Gold-miners and mining at Bir Umm Fawakhir
265
from Bir Umm Fawakhir are meager: a few pieces of Byzantine glass, figurine fragments, or faience. What was Bir Umm Fawakhir? It had always been called a Roman caravan station of the first or second century AD, but its sprawling layout looks nothing like the orthogonal, fortified hydreumata (orthogonal, fortified, watering stations) elsewhere on the Wadi Hammamat route, and the pottery is nothing like the Roman material from Quseir al-Qadim on the Red Sea (Whitcomb and Johnson 1979, 1982). Compounding the problem was the long-held belief that Byzantine rulers were too weak to control the desert and so abandoned it bit by bit to nomadic tribesmen (Maspero 1912). The project therefore considered four possibilities for site function: a caravan station, a military post, a granite quarry and a gold mine. The caravan station hypothesis was rejected for the reasons listed above, and also because it would require only a small resident population, far fewer than a thousand people. The military installation hypothesis was rejected because of the lack of any formal defenses at all. The granite quarry hypothesis could not be dismissed so easily because quarrying does require a large labor force and because there are in fact two quarries at the site, plus two even smaller workings. They are, however, considered to be minor Roman activities, possibly only exploratory. The quarry marks are identical to marks at a known major quarry at Mons Claudianus to the north, and the Romans had an unparalleled passion for elegant stones. There is a little Roman pottery in Quarry 2, and some of the partly quarried blocks have been built into later houses. Unlike the Mons Claudianus to Qena (Cainopolis) route, however, Fawakhir and the Hammamat route preserve no traces of animal shelters for the large number of oxen needed to haul multi-ton blocks of granite to the Nile. As for gold extraction, the evidence comes from the dozens of mines and trenches cut into the surrounding mountainsides, hundreds of dimpled orecrushing blocks, scores of grinding stones of at least two different types (Figure 16.4), and the geological evidence. Hard-rock mining does require a large labor force, which has to be recruited, organized and reliably supplied. In the fifth and sixth centuries all three tasks were probably best carried out by the government, which in turn had compelling reasons to make the effort, as I will argue below.
MINING AND MINERS AT BIR UMM FAWAKHIR How productive were the Fawakhir mines? Here it is necessary to consider a little of the region’s geology. The Precambrian mountains of the Eastern Desert represent a complicated geological zone, but the economically important feature at Bir Umm Fawakhir is the granite stock, called the Fawakhir granite, intruded into older Precambrian rocks. The granite is naturally jointed and in places exfoliated, and it has been quarried for
266
Carol Meyer
Figure 16.4 Three dimpled ore-crushing stones and the lower half of a rotary grinding stone Photograph by Henry Cowherd
building stone to no great extent. It is also the aquifer, carrying water in tiny cracks until it is stopped by the dense ultramafics to the west. Here the wells, all-important in the desert, are sunk and probably always have been. Most of all, however, the quartz veins injected into the granite are auriferous, and the richest veins occur towards the edge of the stock. The mines in the Wadi elSid about 4 km southeast of Bir Umm Fawakhir have been called the richest in all Egypt. The main vein was worked out only in the 1950s, by modern deep mining, and a secondary vein is believed to have been exhausted then also (Kochin 1968). The estimates of yields of grams of gold per ton vary
Gold-miners and mining at Bir Umm Fawakhir
267
widely, apparently according to the methods of evaluation, but what is consistent is that the estimates for Bir Umm Fawakhir are far lower than for the Wadi el-Sid mines, far too low to be economically viable under normal circumstances (Barron and Hume 1902:261; Kochin 1968; Sommerlatte 1994:137–138). This may explain why the Fawakhir veins were exploited only in late antiquity, after the richer lodes had been worked out by ancient techniques, and why little mining has been attempted since then. The ancient miners used two techniques at Bir Umm Fawakhir: open-cast trenches following a quartz vein from the surface, and underground mines cut horizontally or diagonally into the mountainsides. Some of the latter still have stone-built revetments or platforms at the mouths, presumably to aid in the raising and lowering of men and boys, baskets, ropes and the like (Figure 16.5). Diodorus Siculus and mining Here I turn, with some reluctance, to the first-century BC account of Diodorus Siculus on Egyptian gold mines, based in part on Agatharchides’ secondcentury BC work. The reluctance comes from the fact that Diodorus is repeatedly cited as the sole written source on Egyptian gold mining, all the way from pharaonic through Roman times and later. Diodorus did use the best sources he could find and he says he tried to check them, but it is not at all clear whether he or Agatharchides ever saw a gold mine, and there is some question as to whether Agatharchides was basing his account wholly or in part on Spanish mines. This chapter therefore treats Diodorus as a source to be tested rather than accepted uncritically. Diodorus states: For the kings of Egypt gather together and condemn to the mining of the gold such as have been found guilty of some crime and captives of war, as well as those who have been accused unjustly and thrown into prison because of their anger, and not only such persons but occasionally all their relatives as well…. For no lenience or respite of any kind is given to any man who is sick, or maimed, or aged, or in the case of a woman for her weakness, but all without exception are compelled by blows to persevere in their labours, until through ill-treatment they die in the midst of their tortures. (Diodorus III, 12, 13) We have no idea what a Byzantine prison establishment looked like. Also, the kings of Egypt were long gone by the fifth century, replaced by distant rulers in Rome and Constantinople. Half a millennium after Diodorus, the economic situation and the entire political system had changed around the mines. Also, Diodorus’s account does not agree very well with the seemingly unstructured layout of Bir Umm Fawakhir. There are no signs of central planning, much less
268
Carol Meyer
Figure 16.5 Ancient mine shaft on the mountainside below the Roman tower. Pieces of a revetment wall still survive. The wadi at the upper left-hand corner is the Roman road; the cluster of huts that is Outlier 2 can be seen jutting out between two spurs of rock. The wadi opening at the upper right-hand corner of the photograph encloses the main settlement. The modern settlement and road are out of sight to the right Photograph by Henry Cowherd
Gold-miners and mining at Bir Umm Fawakhir
269
the regimentation of a military fort or camp. House construction is quite idiosyncratic and includes small built-in comforts such as benches and niches. The dwellings are similar from one end of the site to the other, and one would expect at the least a differentiation of the keepers and the kept. The large number of wine jars may have been used only once for wine, but that still represents a considerable amount of an imported luxury in the desert. The silos in Outlier 2 do not look like central distribution of daily rations. There are no defensive works to speak of, either to keep people in or keep them out. The workers were not slaves, and it is necessary to emphasize this because it is a common assumption that they were. Egypt has never been a slave society on the order of, say, ancient Greece or Rome; even Diodorus does not call the miners slaves. In short, the civil status of the workers at Bir Umm Fawakhir is less certain than Diodorus’s account might indicate. He continues: The gold-bearing earth which is hardest they first burn with a hot fire, and when they have crumbled it in this way they continue the working of it by hand…. And the entire operations are in charge of a skilled worker who distinguishes the stone and points it out to the labourers; and of those who are assigned to this unfortunate task the physically strongest break the quartz-rock with iron hammers, applying no skill to the task, but only force. (Diodorus III, 12) In 1996 four of the ancient mine shafts were inspected for evidence of firesetting. The longest of the mines runs horizontally for about 100 m, is tall enough for a short person to walk upright, has two short side passages and an air shaft, but no signs of fire, no charcoal or ash and particularly none of the characteristic fire-spalled niches. Rather, the working faces have oblong holes pounded in them, apparently for hammering out the next piece of rock. The Fawakhir granite is jointed and fractured, and in places downright rotten. In short, there would have been no need for fire-setting here. The account continues: The boys there who have not yet come to maturity…laboriously gather up the rock as it is cast down piece by piece and carry it out into the open to the place outside the entrance. Then those who are above thirty years of age take this quarried stone from them and with iron pestles pound a specified amount of it in stone mortars, until they have worked it down to the size of a vetch. (Diodorus III, 13) Here the material evidence from Bir Umm Fawakhir is more in line with the documentary record. Mortars, in the sense of deep basins for pounding, are not
270
Carol Meyer
common at the site, and the ones recovered are limestone, highly unsuitable for crushing quartz. On the other hand, there are hundreds of small (c. 20 cm×20 cm) crushing blocks with pecked depressions in the center (see Figure 16.4). Most of them appear to be in secondary contexts, but one was found in situ at the mouth of the fourth mine inspected. Fist-sized and smaller chunks of quartz are scattered around the crushing stone where the workman walked away from it nearly 1,500 years ago. This makes sense. Ore mined in virtual darkness would be immediately reduced at the mouth of the mine to pick out the pieces worth the considerable effort of further reduction, and the rest tossed downslope. Diodorus continues: Thereupon the women and older men receive from them the rock of this size and cast it into mills of which a number stand there in a row, and taking their places in groups of two or three at the spoke or handle of each mill they grind it until they have worked down the amount given them to the consistency of the finest flour. (Diodorus III, 13) Here again some of the evidence from Bir Umm Fawakhir would seem to correspond with Diodorus’s description, for there are scores of rotary mills or querns at the site, both the upper and lower stones (see Figure 16.4). They, like the crushing stones, are made of granite, porphyritic granite or basalt. What is not so clear is the function of the grinding stones, oblong slabs of granite (c. 80 cm long) with a shallow dished depression on top, presumably ground down by an upper hand stone like a mano. It is not clear whether the grinding slabs represent an earlier stage in ore reduction, coarse versus fine grinding, or whether they were used for something else altogether such as flour; they may even have been earlier in date (Klemm and Klemm 1991). Another aspect of Diodorus’s account that seems to hold up in the face of the Bir Umm Fawakir evidence is the presence of complete or nearly complete household groups: women, children and old people in addition to the miners themselves. None of the bones from the cemeteries have been analyzed yet, so the only corroborating evidence comes from earlier and later periods. The Coptos tariff (AD 90) on animals crossing the desert, ships’ spars and the like includes rates for soldiers’ wives, as well as prostitutes (Bernard 1984). Excavations at Quseir al-Qadim, ancient Myos Hormos, also suggest the presence of women at the Red Sea port end in the first and second centuries (perfume flasks, a girl’s burial), and by the Mamluk period (thirteenth and fourteenth centuries AD) the evidence becomes overwhelming: henna, bracelets, combs, kohl jars, a veil, and children’s caps and shoes (Whitcomb and Johnson 1979, 1982). Until the bones have been analyzed, we cannot determine whether the miners were of a different ethnic type from the Nile villagers. Diodorus (III, 12)
Gold-miners and mining at Bir Umm Fawakhir
271
does mention that the guards are foreigners lest the prisoners be able to plead with them, but again we have as yet no evidence of enclosures, differentiation in housing, quality of pottery, distribution of wine jars, small finds, or anything else. There is just as much imported North African Red Ware as possible Nubian pottery (none of which has been analyzed for provenience), which suggests only that the Mediterranean coast and Nubia were the two countries closest to Egypt. Furthermore, what slender evidence is available suggests that the miners did in fact originate from the Nile villages and towns. Diodorus (III, 14) also describes gold washing and refining. Nineteenthcentury travellers’ accounts do mention gold-washing tables at Bir Umm Fawakhir, but they have been destroyed by modern mining activity. It is far from certain, however, that the final refining was carried out on site; it would make much more sense to carry the washed gold to the Nile Valley, where fuel is less precious, for final purification.
BIR UMM FAWAKHIR IN ITS WIDER CONTEXT If we turn from Bir Umm Fawakhir to consider its position in regard to the rest of Egypt, it is becoming increasingly clear that the Eastern Desert was not abandoned to nomads, as had previously been thought. In addition to the gold mines at Bir Umm Fawakhir, recent archaeological work has demonstrated the Byzantine/Coptic occupation of the Roman fort at Abu Sha’ar (Sidebotham 1994); the continued exploitation of the imperial porphyry quarries at Mons Porphyrites (Kraus et al. 1967); heavy utilization of the port at Berenice and its desert road (Sidebotham and Wendrich 1996); an attempt to mine gold at the Old Kingdom copper mines at Wadi Dara (Grimal 1994:423); and activity at Wadi Nakheil (Prickett 1979) and other smaller sites. There is also mention of an alabarch, a customs official. Most alabarchs were concerned with trade into and out of Alexandria, but there is a late sixth-century reference to an alabarch’s office in Antinoopolis (Johnson and West 1949:106), which was the terminus for the desert road across to the Red Sea and down the Via Hadriana to Myos Hormos and Berenice. A reading of Coptic/Byzantine documents from Egypt gives the impression that the ruling polity was concerned with taxes, heresy, taxes, property rights, taxes, collection of taxes, taxes, and enforcement of tax collection. There is virtually nothing directly concerned with mines or quarries. What is clear, however, is that these were hard times in Egypt. Repeatedly one hears of peasants fleeing their land because they could not pay their taxes, or losing their land to great landowners who could pay taxes even in bad years, or becoming monks who had to support themselves but at least were not taxed. Unlike in pharaonic times, there is no mention of adjusting taxes to mitigate the damage of low Niles. Egypt was expected to feed Constantinople just as it had earlier fed Rome, regardless of the effect on the country.
272
Carol Meyer
Egypt suffered several divisions and redivisions of the duchies in the fifth and sixth centuries, but these mainly affected the Delta region. Upper Egypt, which has always included the Hammamat route, remained an administrative unit, the Thebaid, under the duke of Thebes, who was appointed by the emperor himself. Although the Thebaid was treated as a limes (border fortification) on an extreme edge of the empire and the dukes had both military and civil powers, the civil authority was actually the more important. Threats to the borders of Upper Egypt simply were not as serious or common as, say, in Persia. Another high official was the comes largitionum sacrarum, who among other responsibilities had charge of mints, mines and a number of state factories (King 1980). Although there are many obscure features of their bureaucratic offices, it seems likely that the comes largitionum sacrarum or his officers were concerned with the product of the Bir Umm Fawakhir mine, specifically with collecting the state’s share of the gold, and perhaps with recruiting workers and provisioning them. It is difficult to see who apart from the government could have routinely supported so large a population in the central desert. It is a two-and-a-half to three-day camel trek from the Nile to Fawakhir and everything except water had to be carried in: baskets, cloaks, grain, lamps, lettuce, mats, oil, pottery dishes, rope, tools, wine, wood, etc. (The modern Bedouin do maintain herds of goats and some camels, but they are constantly moving to new sources of water and vegetation, which, once eaten, will not be replenished until the next flash flood, perhaps ten years hence.) There is mention of a rhabduchus who escorted the camels or donkeys to the mines being paid a wage (Johnson and West 1949:331). A camel can carry 227 kg and requires neither food nor water for weeks, though there are at least two reliable wells on the route to Bir Umm Fawakhir. If the population were over 1,000, then something like a biweekly camel train of 30 camels would have been required or 60 camel loads per month, not to mention assembling the supplies and requests in the valley. One reference to imperial mines in Egypt comes from a document of AD 349, earlier than Bir Umm Fawakhir, recording a tax levied at Hermopolis for the imperial at the rate of 12,250 drachma per aroura (land measure). This is believed to have been in lieu of providing requisitions of labor, but it is not known whether the tax was universal or a special assessment (Johnson and West 1949:255). Criminals may still have been sent to the mines, but workmen were requisitioned from the villages as a liturgy, a service obligation. Commuting this liturgy by a tax on land seems to have become general (Johnson and West 1949: 331). This practice has some similarities to the recruitment of soldiers, and the paying of money to avoid service suggests that poorer people, those who could not buy their way out or who had no other prospects, would have been the most likely to end up in the mines. Considering Egypt, specifically Egyptian gold, in relation to the Byzantine Empire brings up the Byzantine gold question (my thanks to Dr Walter Kaegi, Department of History, University of Chicago, for the following insights). Imperial expenses in the fifth and sixth centuries were staggering. Gold was
Gold-miners and mining at Bir Umm Fawakhir
273
required for armies and wars, navies, hostage ransoms, improving the loyalty of dubious allies, and major building programs such as Hagia Sophia and far-flung churches and monasteries at places such as Ravenna in Italy and St Catherine’s on Sinai. One gold source was the confiscated accoutrements of the pagan temples, but this was finite. Gold supplies from East Africa, sub-Saharan Africa (at a later date), Turkey itself, Armenia and other parts of the Caucasus, the Urals and the Balkans have been suggested, but never Egypt (Meyer et al. in preparation), which is a little surprising inasmuch as it was famous throughout antiquity as a gold source. The quest for gold may explain why the low-grade ores at Bir Umm Fawakhir were worked in the fifth and sixth centuries and not thereafter. The workmen, then, would have at least made a living for themselves and their families in hard times. Unlike placer mining, the hard-rock miners at Bir Umm Fawakhir had no hope of striking it rich individually. Rather, the gold was carried away from the desert mines to the centers of power in the Thebaid, Alexandria and Constantinople. In other words, the state could use real wealth —grain—to support the mines and miners, and the gold in turn helped to prop up the imperial regime.
CONCLUSION Bir Umm Fawakhir is a relatively ‘new’ site. No legal excavation has ever been undertaken there, and analyses of survey findings are still in progress. Although we can draw some preliminary conclusions about the site, many questions remain. It seems clear, however, that the miners were recruited as labor to meet liturgical obligations, or perhaps as prisoners; their immediate dependants probably accompanied them. Wives and sons may have been employed in ore collection and reduction; certainly there was no agriculture and little if any animal husbandry to occupy them. The house ruins appear to be family units, and to judge from the number of houses sharing party walls or an enclosed space, the house-builders coordinated their activities, at least initially. The Coptic crosses stamped on dishes point to a Christian or Christianized population with all that entails in terms of work-weeks, holidays, ethics and hope of glory. As yet, however, there is no direct evidence of formal or informal religious practices. If the miners were recruited from villages on the Nile, we would expect Bir Umm Fawakhir to have had the style of town leadership typical of the period, which is to say that responsibility for meeting government laws and exactions rested on the heads of important families rather than on outsiders appointed by the imperial government (Bagnall 1993:136–137). In addition to the vital supply lines that fed ancient Fawakhir, we suspect that the inhabitants maintained personal contacts with the Nile Valley. Ostraca and letters from earlier and later sites (e.g. Guéraud 1942) contain greetings to family members and requests for supplies.
274
Carol Meyer
The site of Bir Umm Fawakhir belongs basically to one period and is estimated to have survived for about 150 years in the fifth and sixth centuries AD. Almost all mining towns are ephemeral in the sense that the ores will play out, but Bir Umm Fawakhir was especially vulnerable in this regard. It could exist only when the politico-economic organization in the Nile Valley and the demand for gold were strong enough to support such a remote mining operation. We do not know whether the mines were worked continuously for 150-odd years or whether they were abandoned for years or even decades and then reopened. Nor do we know whether the mines were worked year-round or only in the winter; excavation might help to answer both questions. Seasonal or interrupted occupation would have had a profound effect on the mining community in that its internal organization could have been much more ad hoc. Although we have scores of houses attributed to the workers and their families, we have none that can be clearly identified as belonging to a middle level; that is, to specialists such as Diodorus Siculus’s ‘skilled workers who distinguish the stone and point it out’ (mining engineers in effect), or to priests and scribes, soldiers and others. As concerns other questions such as the relations, if any, between the people of Bir Umm Fawakhir and the nomads who presumably inhabited the surrounding Hammamat region, we have not a clue. Future work at the site will, if possible, be directed to answering some of these questions. At the moment, however, we look to the Nile Valley for the home town of the miners, for the source of their supplies and the destination of their product (gold), and for the pattern of their lives and town organization.
REFERENCES Bagnall, R.S. (1993) Egypt in Late Antiquity. Princeton, NJ: Princeton University Press. Barron, T. and Hume, W.F. (1902) Topography and Geology of the Eastern Desert of Egypt. Cairo: National Printing Department. Bernard, A. (1984) Les portes du désert. Paris: Éditions du Centre National de la Recherche Scientifique. Diodorus Siculus (1967 [about 50 BC]) Translated by C.H.Oldfather. Cambridge, MA: Harvard University Press. Grimal, N. (1994) Travaux de l’IFAO en 1993–1994 . Bulletin de l’Institut Français d’Archéologie Orientale 94: 420–434. Guéraud, O. (1942) Ostraca grecs et latins de l’Wadi Fawakhir. Bulletin de l’Institut Français d’Archéologie Orientale 41: 141–196. Johnson, A.C. and West, L.C. (1949) Byzantine Egypt: Economic Studies. Princeton, NJ: Princeton University Press. King, C.E. (1980) The Sacrae Largitiones: revenues, expenditure and the production of coin. In C.E.King (ed.), Imperial Revenue, Expenditure and Monetary Policy in the Fourth Century A.D. British Archaeological Reports International Series 76: 141–173. Oxford: BAR. Klemm, R. and Klemm, D. (1993) The development of gold ore milling technique from Predynastic to Arabic time in Egypt. In Archaeological Stone Abstracts: Conference at the British Museum (14–16 November 1991). Abstract.
Gold-miners and mining at Bir Umm Fawakhir
275
Kochin, A. (1968) Unpublished report to the Egyptian Geological Survey and Mining Authority on Eastern Desert gold occurrences. Kraus, T., Röder, J. and Müller-Wiener, W. (1967) Mons Claudianus—Mons Porphyrites. Mitteilungen des Deutschen Archäologischen Institute Abteilung Kairo 22: 108–205. Maspero, J. (1912) Organisation militaire de l’Égypte byzantine. Paris: H.Champion. Meyer, C. (1993a) The Bir Umm Fawakhir Survey Project. The Oriental Institute 1991– 1992 Annual Report. Chicago: The Oriental Institute. ——(1993b) The Bir Umm Fawakhir Survey Project. The Oriental Institute Annual Report 1992–1993. Chicago: The Oriental Institute. ——(1995a) Gold, granite, and water: the Bir Umm Fawakhir Survey Project 1992. Annual of the American Schools of Oriental Research 52: 37–92. ——(1995b) A Byzantine gold-mining town in the eastern desert of Egypt: Bir Umm Fawakhir, 1992–1993. Journal of Roman Archaeology 8: 192–224. ——(1996) The Bir Umm Fawakhir Survey Project. The Oriental Institute 1995–1996 Annual Report. Chicago: The Oriental Institute. Meyer, C., Heidorn, L., Kaegi, W. and Wilfong, T. (in preparation) A Byzantine Goldmining Town: The Bir Umm Fawakhir Survey Project 1993. Oriental Institute Communications. Chicago: The Oriental Institute. Prickett, M. (1979) Quseir Regional Survey. In D.S.Whitcomb and J.H.Johnson (eds), Quseir al-Qadim 1978: Preliminary Report, pp. 257–359. Cairo: American Research Center in Egypt. Sidebotham, S.E. (1994) Preliminary report on the 1990–1991 seasons of fieldwork at ‘Abu Sha’ar (Red Sea coast). Journal of the American Research Center in Egypt 31: 133– 158. Sidebotham, S.E. and Wendrich, W. (eds) (1996) Berenike ’95. Leiden: Research School Centre for Non-Western Studies (CNWS). Sommerlatte, H.W.A. (1994) Das Gold der Ägypter—Wahrheit und Legende? Betrachtungen eines Bergingenieurs zu einem altem Thema. Nubica 3: 111–140. Whitcomb, D.S. and Johnson, J.H. (eds) (1979) Quseir al-Qadim 1978: Preliminary Report. Cairo: American Research Center in Egypt. ——(1982) Quseir al-Qadim 1980: Preliminary Report. Malibu, CA: Undena Publications.
Part IV
Overviews
Chapter 17
On the value of mixed methods in studying mining communities David Killick
WHY BOTHER WITH A SOCIAL HISTORY OF MINING? AND WHY NOW? As Bernard Knapp notes in his introduction to this volume, mining should feature prominently in historical studies because of the central role that metals and fossil fuels have played in the development of industrial societies. Over the past five thousand years literally millions of people have lived, labored (often involuntarily) and died in mining settlements, while the quest for new sources of metals and minerals has been a major motive for exploration, conquest, colonialism and imperialism since at least the Egyptian Middle Kingdom (Shaw, this volume, Chapter 15). Yet the lives of miners have attracted far less attention from social historians, archaeologists and anthropologists in the Englishspeaking countries than have the lives of urban factory workers, fisherfolk or farmers. (There is, however, a substantial literature in German, much of it emanating from the Deutsches Bergbau-Museum in Bochum.) There is certainly much written in English on the technological, economic and business history of mining, but the only aspect of social history that has been researched in real depth is the conflict between capital and labor. It is not clear to me why historical and anthropological studies of mining communities should have lagged behind studies of many other working people. Perhaps it has something to do with the ephemeral existence of many mining communities—a theme that runs through most of the chapters in this book. Mining towns tend not to die lingering deaths; usually the end is sudden, and the dispersal of the population abrupt. These circumstances work against the preservation of the archival documents and collective memories that the historian needs. There is also a distinct geological bias in the existing literature. The disproportionate attention to coal- and tin-mining communities reflects the fact that geological reserves of these materials are often large enough to support longterm settlement, with all that this implies for the preservation of written records. This volume goes some way toward restoring the balance, because most chapters deal with small, ephemeral mining settlements. In writing about these, the authors
280
David Killick
explore a number of new approaches, including the use of archaeology and material culture to complement (or substitute for) written records. This is a particularly appropriate time for scholars from western Europe and North America to explore the social history of mining, particularly since, during the past twenty years, these regions have experienced a precipitous decline of employment in mining. To our parents’ generation, the miner was as familiar a figure as the postman, while miners’ unions used to be a potent political force in almost all industrialized nations. Most of our children are about as likely to know a miner as they are to encounter a knight in armour. The political clout of miners’ unions has diminished sharply as pit faces have fallen silent, whole mining communities have disappeared, and union coffers have shrunk dramatically. (How many Britons under 25 recognize the name of Joe Gormley, former president of the National Union of Miners, who in the 1970s brought down a British government?) Even in Arizona, which still presents itself as the ‘Copper State’ (Sheridan, this volume, Chapter 11), mining now accounts for only 10% of economic activity and 5% of employment. There are many reasons for the rapid decline of employment in mining in the Western world, but the most important are competition from mines elsewhere (e.g. Chile, Peru, South Africa, Bougainville), the withdrawal of state subsidies from European mines (especially coal mines), and technological change, especially the substitution of highly mechanized open-pit and strip mining for underground tunneling, and of heap leaching and electro winning for smelting of non-ferrous metals. Stricter controls on pollution of air and water in North America and Europe have also contributed to this decline. Rather than pay the high costs of compliance at home, major Western mining houses have chosen to invest in countries where environmental protection is less stringent. Overall, therefore, it is an appropriate time to stimulate research on past mining and miners, and to develop method and theory for that purpose. Many regions of Europe and North America are trying to preserve some record of former mining and metallurgical industry, but inevitably even the best ‘living museums,’ like that at Coalbrookdale/Ironbridge Gorge in England, present a greatly sanitized version of the original (Clark 1987; Gordon and Malone 1994). The green and pleasant land around the Ironbridge Gorge today, for example, is difficult to reconcile with the fiery coke heaps, sulfurous fumes and industrial clutter depicted in Phillippe de Loutherbourg’s famous painting of Coalbrookdale in 1801. The task of reconstructing past mining communities in all their sweat, smoke and sorrow can only be accomplished on paper (or film, or CD-ROM) by the historian and the archaeologist. This has never been better done than in John Rule’s vivid portrait in this volume of the physical and spiritual torment of Cornish tin-miners. I should also draw attention to Larry Lankton’s superb study of copper mining in the Upper Peninsula of Michigan (Lankton 1992). This is, to my knowledge, the only published study that integrates all of the relevant historical approaches to the study of a mining district—geological, technological, economic, labor and social—within a single volume.
On the value of mixed methods in studying mining communities
281
Before rolling up their sleeves and getting down to work, students of past mining communities need to pause and to consider carefully what issues are important and what methods are appropriate for their investigation. This was the aim of the 1996 Rockefeller conference at Bellagio, of which this book is the ultimate product.
APPLES AND ORANGES? One of the questions posed by the organizers of the conference was whether historians, archaeologists and anthropologists could find common ground on the major issues in the study of mining communities. Class, gender and ethnicity are clearly the major issues for historian Tom Sheridan and historical archaeologist Alexy Simmons, while the anthropologist Terry Childs is most interested in shared beliefs and their symbolic expression. Yet there is clearly much crossover between the historical and anthropological approaches. John Rule and Eugenia Herbert are historians, but both their chapters explore the role of shared beliefs in accommodating people to the risks, disasters and losses that have always stalked mining communities. Conversely, anthropologists William Douglass and Robert Ehrenreich adopt the techniques of the historian in their studies of peripatetic miners in the American West and of the clash of cultures around early mining in the Arctic. What are the issues that animate the archaeologists? Historical archaeologists Donald Hardesty and Susan Lawrence are interested in the same issues as the historians (class, gender, ethnicity) but ask whether these issues can be approached by studying the spatial distribution of material culture. This seems to me a particularly interesting question, and I shall return to it below. The last set of chapters in this volume are those of prehistorians working entirely, or almost entirely, with archaeological evidence (Pigott, Shennan, Meyer, Shaw, Kassianidou). These scholars have a distinctly different focus, which is the organization of production in past mining. They discuss economic specialization (Pigott, Shaw, Shennan, Kassianidou), reconstructing the labor process or chaîne opératoire (Pigott, Kassianidou) and regional exchange of the products of mining (Shennan, Pigott, Shaw). A linking theme between this set of chapters and the others offered at the conference is that of ethnicity; both Shaw and Meyer draw upon written records to discuss the ethnic composition of the workforce on mines in Egypt and its hinterland.
RECONCILING ARCHAEOLOGY AND HISTORY: THE ROLE OF MATERIAL CULTURE STUDIES Must we have written evidence to investigate gender, class, ethnicity or belief in past mining communities? I don’t think so; rather, I agree with Bernard Knapp
282
David Killick
(this volume, Chapter 1) that the lack of discussion on these topics reflects the choice of previous students of prehistoric mining and metallurgy to focus on technology, trade and the organization of production rather than on the social existence of past miners. (The excavation of the Egyptian miners’ temple at Timna [Rothenberg 1988] is a rare and fascinating exception to this generalization.) The excavations of Steve Shennan at Mondsee and the survey of Carol Meyer at Bir Umm Fawakhir are among the first to have an explicit focus on the lives of prehistoric miners. The current Russian/Spanish project at the Early Bronze Age mining district of Kargaly is another that is devoting as much attention to the settlements of miners as to mines and metallurgical processing sites (Chernykh 1994). The ongoing Sydney-Cyprus Survey Project, directed by Bernard Knapp (Knapp and Given 1996), adds an additional dimension in examining mining as part of a regional pattern of settlement—an approach also advocated in industrial archaeology by Gordon and Malone (1994). But can credible evidence of gender, class, ethnicity and ideology be obtained from material culture? Bruce Trigger is quoted (in the publisher’s blurb for the new Journal of Material Culture) as saying that the study of material culture is ‘the last serious gap in the structure of the social sciences.’ This is not a gap that, on the evidence of its first two volumes, the Journal of Material Culture is likely to close very rapidly, but more substantial demonstrations of the potential of material culture for social history are available elsewhere (e.g. Weiner and Schneider 1989; Kingery and Lubar 1993; Kingery 1996). The studies that have been most successful so far are those related to the social division of space, whether at the scale of house plans (e.g. Glassie 1975) or at community and regional scales (e.g. Hillier and Hanson 1984). Prehistoric and historic archaeologists alike also pay close attention to spatial variation in the distribution of ‘sumptuary goods’ in the expectation that such variation, whether in households, cemeteries or settlements, reflects variation in the social status of the occupants of those localities. The ‘sumptuary goods’ that have received most attention are exotic materials (e.g. turquoise, gold, silver, amber) and manufactured items that represent (in Marx’s terminology) large amounts of crystallized labor, such as the Gundestrap cauldron (Taylor 1992), or to which are ascribed special properties that endow the holder with political legitimacy, such as Chinese ritual bronzes (Murowchick n.d.). Food and beverages are obviously another marker of class and ethnic differences in modern times. Zooarchaeologists and archaeobotanists are aware of this, and actively search for spatial patterning in animal bones and crop residues in archaeological sites.
MATERIAL CULTURE AND CLASS How well does this experience serve us when we turn to the study of mining communities? In some cases, quite well. As far as class is concerned, Don Hardesty (this volume, Chapter 5) notes distinct differences in town layout
On the value of mixed methods in studying mining communities
283
between the ‘company towns’ of Colorado (those wholly owned by the mining companies) and the ‘free’ town of Reipetown, constructed by the miners themselves in opposition to the ‘company towns.’ Carol Meyer also makes a convincing argument from the spatial layout of the Roman mining settlement of Bir Umm Fawakhir that mining at the site cannot have been based upon slave labor; there is no evidence at the site of any system of constraint, of a military presence, or of centralized facilities for storing rations. On the other hand, neither house plans nor food residues have proven very informative in the short-lived Australian mining camps discussed by Susan Lawrence, or in the equally ephemeral mining towns of the American West (discussed here by William Douglass and Don Hardesty in Chapters 6 and 5 respectively). In these cases miners and their families seem to have lived in shanties patched together with recycled timber, tin, paper and canvas. Hardesty usefully invokes Susan Kent’s concept of ‘anticipated mobility,’ developed in the context of Kalahari hunter-gatherers, to explain the makeshift character of housing in these settlements; an obvious parallel can be drawn to shanty towns in the Third World today, where the constant threat of demolition by civic authority also makes for ‘anticipated mobility.’ As far as food was concerned, miners in remote frontier camps seem to have had little freedom of choice, and therefore little opportunity for expression of ethnic preferences. In settlements of this kind it seems that discarded personal possessions are likely to be more revealing of class than are house plans or food wastes.
MATERIAL CULTURE AND GENDER In the Australian mining settlement of Dolly’s Creek (Lawrence, this volume, Chapter 3) the census records for 1861 show that more than half of the households included women. There is ample confirmation of this in the material record, in the form of ornaments and fittings from women’s clothing. Lawrence also assumes, probably correctly, that the presence of refinements such as wallpaper, good tableware and ormolu clocks indicates the civilizing hand of woman in those households (though as a semi-civilized male I would like to see a robust statistical correlation between the occurrences of these artifacts and those more certainly indicative of the presence of women!). The question of women’s economic roles in mining camps is not as easily addressed. Herbert’s survey of the historical and ethnographic literature from precolonial Africa shows that women were often responsible for the transport of ore and charcoal, the crushing of ore and the panning of gold, while Lawrence suggests that women engaged in farming and stock-raising at Dolly’s Creek. It seems highly unlikely that an archaeologist could deduce from the material remains alone which gender was responsible for these activities. Nor am I convinced by Alexy Simmons’s argument, in an otherwise exemplary chapter, that houses of prostitution can be recognized by their material culture alone.
284
David Killick
Her assertion that working-class prostitutes were better clothed and better fed than working women outside the brothel is one that needs much more supporting evidence than she provides; nor is her negative evidence (the absence of men’s artifacts) particularly diagnostic of brothels. The only indisputable material evidence of women’s work cited is the physical remains of African women killed at the mine face by collapsing shafts (Herbert, this volume, Chapter 9). These send a warning that we cannot assume that all prehistoric miners and metalworkers were male, but the specifically gendered signatures of work around mining settlements have yet to be identified, if indeed they can be.
MATERIAL CULTURE AND ETHNICITY Before we begin to look for material evidence of ethnicity in archaeological sites we need to be sure that we understand what we mean by the term. Alonso (1994:391) distinguishes two distinct uses in academic writing: What is called race in much of the literature is that variant of ethnicity that privileges somatic indexes of status distinctions such as skin color, hair quality, shape of features, or height. What is called ethnicity is the variant that privileges style-of-life indexes of status distinctions, such as dress, language, religion, food, music or occupation. As this excerpt makes clear, race is a social product, not a biological fact; racial distinctions are assertions of relative status in stratified societies. When Carol Meyer (this volume, p. 270) writes of the fifth- to sixth-century mining settlement of Bir Umm Fawakhir that ‘[U]ntil the bones have been analyzed, we cannot determine whether the miners were of a different ethnic type from the Nile villagers,’ she is confusing socially defined race with biological variation. A study of the human remains from this site might indeed reveal that workers came from regions as far distant as the Aegean, lower Egypt and Nubia; indeed, Brace et al. (1996) have already shown that the ancient populations of these regions can be distinguished by multivariate statistical analysis of cranial measurements. But this would reveal nothing about ‘ethnic type’ (i.e. race), since modern physical anthropology firmly rejects the concept of races in favor of ‘clines’ — contours showing the frequency distribution of genotypical or phenotypical characters across space (Brace et al. 1996; Shanklin 1994). Ethnic groups (as opposed to ethnic categories within a particular ideology) form around shared interests and values, and sometimes (but not always) around shared language, shared class position or a putative common origin. Class interests may divide an immigrant group with a common language, as is illustrated by the split between Italian merchants and Italian carbonari in the Nevada Charcoal Burners War of 1879 (Frehner 1996, cited by Hardesty, this
On the value of mixed methods in studying mining communities
285
volume, Chapter 5). Ethnic consciousness is often sharpened by economic competition. In North America a large frontier mining town typically housed, among others, Cornish, Irish, Scandinavian, Slovakian, Chinese, Mexican and Native American inhabitants (Lankton 1992; Schwantes 1992; Hardesty 1988; Sheridan, this volume, Chapter 11). Times were often hard, and competition for a dollar (or for company scrip) was fierce. As Sheridan’s chapter demonstrates so well, ethnic divisions undercut common class interests, and thereby weakened labor solidarity against the mining companies, which deliberately promoted ethnic consciousness and racism in order to divide labor and contain (or even drive down) wages. Paradoxically, ethnicity also served as the basis for resistance to oppression, as in the fight of miners of Mexican origin for equality with Anglo miners in Arizona. What evidence of ethnicity can we find in the archaeological record? In the case of nineteenth-century American and Australian mining towns, the major material manifestation of ethnicity outside the workplace was the de facto division of the towns into ethnic neighborhoods. Unfortunately this may not be evident from street plans or architecture in company towns, nor (for reasons outlined above) in the short-lived and undercapitalized mining camps. In longlived mining towns, and particularly in those where residents held title to their plots, one might perhaps find more expression of ethnicity in house plans. Small keepsakes, brought by immigrants from their regions of origin, might also be more widely useful indicators of ethnicity; for example, at the Bellagio Conference Patrick Martin (Michigan Technological University) noted that Scandinavian coins have been recovered from the sites of known Scandinavian quarters in former mining towns of northern Michigan. Food, and facilities for preparing food, are other markers of ethnicity in mining towns. (Cornish pasties, for example, can still be purchased in the former copper-mining towns of Houghton, Michigan, and Bisbee, Arizona, but are essentially unknown elsewhere in the USA.) Unfortunately, one cannot assume that food was prepared at the spot where the archaeologist finds its imperishable residues; perhaps the distribution of special tools for food preparation (such as the metates formerly used to prepare corn dough for tortillas in Mexican and Mexican-American cuisine) would be more reliable indicators of ethnicity. Detection of significant spatial variation, whether in house plans, in cuisine or in ‘small things forgotten,’ is, however, going to require excavation of much larger areas of former mining settlements than has been undertaken so far. Shrines and other immobile places of worship are particularly useful markers of ethnicity. For example, icons of the divinity Hathor at the Timna mining temple in southern Israel (fourteenth to twelfth centuries BC) offer certain proof of the presence of Egyptian personnel at that site (Rothenberg 1988). While there is certainly no one-to-one correspondence between religious affiliation and ethnicity, places of worship have tended to be the foci of different ethnic communities in urban settlements. John Rule’s extraordinary chapter in this
286
David Killick
volume on Methodism and miners in Cornwall (Chapter 10), for example, immediately raises questions about the role of ‘chapel’ among expatriate Cornish miners in the Americas. The wider issue of the role of religion in dividing mining communities is clearly one that deserves much more attention than it has received to date. The study of cemeteries could prove to be another valuable window on ethnicity in mining communities.
MINING, COLONIALISM AND IMPERIALISM Miners have been the shock troops of imperialism since at least the Egyptian Middle Kingdom (Shaw, this volume, Chapter 15), so the documentary and material records of mining ought to be prime sources for the study of colonialism and imperialism. This is not news to historians, and particularly not to historians of Africa or of the Americas, where the quest for metals played so large a role in the conquest and subordination of indigenous peoples, but many archaeologists seem to be unaware of this. I find it truly remarkable, for instance, that a recent major review of the literature on the archaeology of imperialism (Sinopoli 1994) fails to include a single reference to mining! Few studies better illustrate the value of these records than Robert Ehrenreich’s chapter in this volume (Chapter 7). His discussion of the clash in the Arctic between the English state and the Inuit is a dark tale, woven from strands of greed, ambition, delusion, betrayal, kidnapping, revenge and ultimately farce; it would have made an extraordinary movie in the hands of Werner Herzog! More importantly, it breaks new ground in reminding us of the indigenous populations affected by mining expeditions from powerful states, whether in the context of Egyptian imperialism in the fourteenth century BC, English and Spanish incursions into the Americas in the sixteenth century AD, or the clashes between Europeans and Native American, African and Australian peoples over mining in the nineteenth and twentieth centuries. Except in Spanish America (Craig and West 1994), historians of mining have not paid nearly enough attention to the fate of the indigenous people among whom colonizing miners settled. Among the issues that need fuller study are those of discovery and possession of ore deposits and their products in precapitalist societies (a theme well explored by Terry Childs and Ian Shaw in their respective chapters in this book— Chapters 8 and 15), the beliefs of the dispossessed towards the earth and its mineral products (e.g. Reichel-Dolmatoff 1981, Herbert 1984), and how (if at all) these may have influenced resistance to mining by indigenous peoples, particularly within the capitalist world system. In this connection, there is a striking contrast between the capitalist world-system and the earlier Islamic world-system (c. 850–1550). As is well known, the Islamic monetary system was based in large part upon African gold, which was obtained from two regions: from sub-Saharan West Africa,
On the value of mixed methods in studying mining communities
287
via the West African caravan trade; and from East African sources in what is now the nation of Zimbabwe (Austen 1987). Yet at no point did the various Islamic powers attempt to seize the gold-producing regions, nor did hordes of prospectors from the circum-Mediterranean or Arabia descend upon subSaharan Africa to stake claims to mineral deposits. Instead, Islamic traders bartered for gold at entrepôts like the cities of Jenne, Gao and Timbuktu in West Africa (see the contemporary accounts in Levtzion and Hopkins 1981), and at ‘fairs’ along the Zambezi river in East Africa (Axelson 1973). Islamic demand for the metal certainly stimulated gold production—there is almost no evidence of the mining or use of gold in either region before the Islamic era—but there is no indication that Islamic traders had any direct role in its production. The contrast between the Islamic and the capitalist world-systems in this respect is an intriguing problem deserving of much closer examination.
THE SOCIAL ORGANIZATION OF WORK IN MINING Few technologies, even in the modern world, are so constrained by natural laws that they can be done in only one way. Usually there are many ways of getting to the same end, and the path actually chosen (the chaîne opératoire) often reflects ideological and social choices as much as (or more than) the search for ‘efficiency’ (for two examples, see Gordon and Killick 1993). This is, in a nutshell, the credo behind the ‘social construction of technology,’ currently the dominant approach within the history of technology (Bijker et al. 1987), and increasingly influential within anthropology and archaeology (Pfaffenberger 1992; Dobres and Hoffman 1994). As Bernard Knapp notes in his introduction, the extensive literature on ancient mining and metallurgy has been almost exclusively concerned with the reconstruction of techniques, and has paid little attention to the people who devised and practiced them. The chapters written by archaeologists in this volume go a considerable way towards restoring the balance. Those by Pigott and Shennan both display two of the great strengths of modern archaeology, which are (a) the sophisticated use of material remains; and (b) the systematic use of a regional frame of reference. Both of these scholars effectively use these tools to challenge entrenched positions. Shennan argues that evidence for copper processing, and the absence of evidence for cereal threshing at Mondsee, together with a comparative analysis of sites in the wider region, indicate a greater degree of economic specialization than is accepted for that time. Pigott uses the material evidence to reconstruct in great detail the chaîne opératoire from ore to copper. From this evidence, and a consideration of the relationship of the mining sites to habitation sites and to potential consumers of the smelted copper, he argues that the impressive quantities of mine waste and smelting debris were
288
David Killick
the product of many visits by autonomous producers, rather than from production organized by a central authority. Meyer also makes effective use of material evidence and spatial analysis to suggest that it is unwise to rely wholly upon historical documents, especially when these are not directly related to the site under investigation. I hope that historians of mining will add these approaches to their bag of tools; if this makes them into historical archaeologists, so much the better. Historians can also look more closely into the social construction of work in mining, taking as an example John Rule’s analysis of the Cornish practice whereby groups of miners would bid for the right to work the next ‘pitch’ of rock. So entrenched was this practice that when Cornish miners were imported to work the native copper mines of Michigan, they demanded and won the right to apply the ‘pitch’ system there (Lankton 1992). This concession privileged the Cornish miners as a ‘labor aristocracy’ and heightened ethnic divisions within the mining community. Gordon and Malone (1994) provide an excellent demonstration of the value of material evidence in reconstructing the history of work, and introduce a number of methods that any historian of mining would find useful.
SUMMARY The major aim of the Bellagio Conference was to undermine the disciplinary walls that have separated historical, anthropological and archaeological studies of mining and miners. In this respect, I believe, it was successful, and I look forward to productive mingling of methods in future research. This is not to say that all problems raised were successfully answered. To give but one example, it is evident that the scale of all archaeological excavations on mining settlements so far is too small to address the kinds of issues that are of most interest to the historians—those of class, ethnicity, gender and power. Archaeological excavation is hugely expensive, and opportunities for very extensive excavation are therefore rare. The challenge ahead for archaeologists of mining is to identify the attainable. For historians and anthropologists of mining communities the challenge presented by the conference is to move beyond traditional categories of analysis and traditional forms of evidence.
REFERENCES Alonso, A.M. (1994) The politics of space, time and substance: state formation, nationalism and ethnicity. Annual Review of Anthropology 23: 379–405. Austen, R. (1987) African Economic History. London: James Currey. Axelson, E. (1973) Portuguese in South-East Africa, 1488–1600. Cape Town: Struik.
On the value of mixed methods in studying mining communities
289
Bijker, W.E., Hughes, T.P. and Pinch, T. (eds) (1987) The Social Construction of Technological Systems. Cambridge, MA: MIT Press. Brace, C.L., Tracer, D.P., Yaroch, L.A., Robb, J., Brandt, K. and Nelson, A.R. (1996) Clines and clusters versus ‘race’: a test case in ancient Egypt and the case of a death on the Nile. In M.R.Lefkowitz and G.M.Rogers (eds), Black Athena Revisited, pp. 129–166. Chapel Hill: University of North Carolina Press. Chernykh, E.N. (1994) Kargalí: origenes de la metalurgia en Eurasia Central. Revista de Arquelogía 153: 12–19. Clark, C.M. (1987) Trouble at t’mill: industrial archaeology in the 1980s. Antiquity 61: 169–179. Craig, A.K. and West, R.C. (eds) (1994) In Quest of Mineral Wealth: Aboriginal and Colonial Mining and Metallurgy in Spanish America. Baton Rouge, LA: Geoscience Publications. Dobres, M.-A. and Hoffman, C. (1994) Social agency and the dynamics of prehistoric technology. Journal of Archaeological Method and Theory 3: 211–258. Frehner, B. (1996) Ethnicity and class: the Italian charcoal burners’ war, 1875–1885 . Nevada Historical Society Quarterly 39: 43–62. Glassie, H. (1975) Folk Housing in Middle Virginia. Knoxville: University of Tennessee Press. Gordon, R.B. and Killick, D.J. (1993) Adaptation of technology to culture and environment: bloomery iron smelting in America and Africa. Technology and Culture 34: 243–270. Gordon, R.B. and Malone, P.M. (1994) The Texture of Industry: An Archaeological View of the Industrialization of North America. Oxford: Oxford University Press. Hardesty, D.L. (1988) The Archaeology of Mining and Miners: A View from the Silver State. Society for Historical Archaeology, Special Publication Series no. 6. Pleasant Hill, CA: SHA. Herbert, E.W. (1984) Red Gold of Africa: Copper in Precolonial History and Culture. Madison, WI: University of Wisconsin Press. Hillier, B. and Hanson, J. (1984) The Social Logic of Space. Cambridge: Cambridge University Press. Kingery, W.D. (ed.) (1996) Learning from Things: Method and Theory of Material Culture Studies. Washington, DC: Smithsonian Institution Press. Kingery, W.D. and Lubar, S. (eds) (1993) History from Things: Essays on Material Culture. Washington, DC: Smithsonian Institution Press. Knapp, A.B. and Given, M. (1996) A report on the Sydney-Cyprus Survey Project (SCSP) —third season (1995). Report of the Department of Antiquities, Cyprus: 295–366. Lankton, L. (1992) Cradle to Grave: Life, Work and Death in the Copper Mines of Northern Michigan. Oxford: Oxford University Press. Levtzion, N. and Hopkins, J.F.P. (eds) (1981) Corpus of Early Arabic Sources for West African History. Cambridge: Cambridge University Press. Murowchick, R. (n.d.) The emergence of bronze metallurgy in ancient China: the interplay of technology and power. Paper presented to the Rockefeller Foundation Conference ‘Social Approaches to an Industrial Past: The Archaeology and Anthropology of Mining’, Bellagio, Italy, July 1996. Unpublished transcript. Pfaffenberger, B. (1992) Social anthropology of technology. Annual Review of Anthropology 21: 491–516. Reichel-Dolmatoff, G. (1981) Things of beauty replete with meaning: metals and crystals in Colombian Indian cosmology. In Sweat of the Sun, Tears of the Moon, pp. 17–33. Natural History Museum of Los Angeles County, Los Angeles. Rothenberg, B. (1988) The Egyptian Mining Temple at Timna. London: Institute for Archaeometallurgical Studies. Schwantes, C. (1992) Bisbee: Urban Outpost on the Frontier. Tucson, AZ: University of Arizona Press.
290
David Killick
Shanklin, E. (1994) Anthropology and Race. Belmont, CA: Wadsworth Press. Sinopoli, C. (1994) The archaeology of imperialism. Annual Review of Anthropology 23: 159–180. Taylor, T. (1992) The Gundestrup cauldron. Scientific American 266: 84–90. Weiner, A.B. and Schneider, J. (1989) Cloth and Human Experience. Washington, DC: Smithsonian Institution Press.
Chapter 18
Mining communities, chaînes opératoires and sociotechnical systems Bryan Pfaffenberger
Is the term ‘mining community’ an oxymoron? Any meaningful discussion must begin by challenging the myth of the isolated ‘mining camp,’ in which are found—putatively—the rude cabins and tents of temporary residents, exclusively male. In place of community, or so goes this myth, there is only an aggregate of adventurous, violent men, united only by greed and the pursuit of some valuable metal. Their society, if it can be so dignified by the use of that term, is controlled and shaped by its extreme dependence on resource extraction; if the resource disappears, what passes for community also disappears. To one rather unsympathetic observer, mining and its social relations more closely resemble an ‘addictive disorder’ than a normal human community (Freudenberg 1991). I hope the chapters in this volume put this stereotype to rest, once and for all. As Lawrence (this volume, Chapter 3) convincingly argues, this pejorative view of mining communities is partly a product of the fetishized scholarly fixation on ‘technology and machines,’ to the exclusion of an interest in the social dimensions of mining communities. Lawrence instead emphasizes the community studies model, which ignores the ‘hard’ dimensions of technology and economics in favor of a sensitive exploration of what were face-to-face communities of human beings. What is emerging, thanks to the work of scholars working in this tradition, is a richer view of mining communities, in which women and complex patterns of ethnic cooperation and competition play key roles. In this commentary, which focuses on the social rather than the archaeological dimensions of mining communities, I should like to go a bit further and ask just how this very valuable contribution could be reintegrated with a renewed focus on technology, one that is enriched by some very important advances that have been made in the scholarly understanding of technology. I shall argue that just as a fixation on technology and machines obscures the human and social dimensions of mining communities, so too does a sharp distinction between technology and society obscure some of the factors that give mining communities their distinctive dynamics.
292
Bryan Pfaffenberger
‘AN ENDEMIC YEARNING FOR RESPECTABILITY’ Let us begin by looking at the considerable gains that come from looking at people rather than ‘cataloguing stamper batteries, boilers, steam engines and water races’ (Lawrence, this volume, p. 40). For one thing, surprising numbers of women turn out to have been present at all but the most rudimentary camps. What is more, it is becoming increasingly clear that no adequate account of the politics of mining communities can be constructed without acknowledging the often pivotal role of women. Time and again, it is found that women provided support and networking that proved crucial to the improvement of working conditions. In nineteenth-century Andalusia, for example, networks of women played a key role in extending political activism into the mining industry, which was then dominated by foreign (especially British) ownership (Kaplan 1981). Thanks to the community studies focus, a picture of rich ethnic diversity continues to emerge. In the American West, for instance, the Anglo miner now appears in a more complex drama, involving Chinese and Mexican laborers, Sonoran placeros. Native Americans, Irish women and Italian entrepreneurs, such as the fascinating Data Brothers of Reipetown, described by Hardesty (this volume, Chapter 5). Clearly, ethnic differentiation appears to form part of the process by which people adapted to mining communities. In southwestern Pennsylvania (Bauman 1979), for example, ethnicity turns out to have played a vital role: ethnic ties provided solidarity and mutual aid, and formed the basis for employment in the mines. The mining community’s heterogeneity may have been marred by violence and racism, but it is also clear that prosperity could bring a degree of economic integration that would be missed by those focusing merely on riots and lynchings: in the Comstock mining district of Nevada, white miners voiced fears of exotic diseases and had very few kind things to say about Chinese launderers in their camps—but they took their clothes to them (James et al. 1994). Although conflict occurred and divisions persisted, it is also clear that mining communities are more complex, and doubtless offer more cohesion, than the previous myth would have it. And as Simmons notes (this volume, Chapter 4), the harsh conditions of miners’ lives would naturally lend themselves to the formation of a tacit social bond. Recent work in rural sociology and social psychology bears out the picture of cohesion in the face of impermanence and isolation, for the most part. In Australian mining towns, Cotterell (1984) finds that women living in mining communities were under somewhat more psychological stress, had more fragile social networks, and needed greater social support, in comparison with women living in country service towns. But a reply comes from Robinson and Wilkinson (1995), who administered J.C.Buckner’s Neighborhood Cohesion Index (NCI) in Elliot Lake, a remote Canadian single-industry mining town. Robinson and Wilkinson (1995) found that the mining community could not be distinguished from the NCI’s reference neighborhoods in a North American city. One could argue that this is damning with faint praise, considering the
Mining communities, chaînes opératoires and sociotechnical systems
293
anomic state of neighborhoods in the urban USA, but it is reasonably well established that the mining communities of frontier America and Australia do not necessarily contrast in any fundamental way with other social aggregations. It is striking to note that Australian and American miners themselves seem very sensitive to the question of how their communities should be evaluated qua communities, suggesting that Anglo-American scholars and miners alike are working under the sway of a larger and deeper cultural issue that has to do with the question of valid community as against the stereotype of barbarism and impermanence. Davis (1967:19, cited in Hardesty, this volume, Chapter 5) speaks of an ‘endemic yearning after refinement and respectability,’ which leads to a surprisingly high incidence in Australian and American frontier mining communities of refined artifacts such as decorated china and clocks. A lack of community feeling is likewise a problem to be explained away: recalling the company-built town of Centralia, Pennsylvania, the remaining members of the community wryly lamented the lack of community-wide bonds during the town’s heyday, blaming this deficiency on the mining company’s decision to mine in ways that compromised the community’s health and future (KrollSmith and Couch 1991). Here I am suggesting that the community studies model is not only a useful analytical tool; in some sense, it is also a normative construct—and frankly, that makes me nervous. Undeniably, the community studies model has enriched our understanding, but it is equally possible that its biases have deflected attention away from matters that are just as deserving of attention. In particular, community studies scholars sometimes hold up their scholarship as something morally superior to the work of those who do not see the people along with the machines. Considering the theoretically uninteresting way in which most scholars have looked at mining technology, that is hardly surprising, but it does not exactly fill communities studies scholars with enthusiasm for hearing the latest news from those horrible people who are obsessed with technology. As the next section illustrates, though, there are new ways of looking at the grim world of tools and machines, ways that bring the social and the human back into the picture, and these provide the explanatory power that is sometimes missing when technology is left out. FROM ‘TECHNOLOGY’ TO THE CHAÎNE OPERATOIRE The gains that come from focusing on the human dimension of mining communities are considerable—so much so, in fact, that it is clear that no adequate account of a mining community’s vicissitudes can omit the stories of people who in previous generations of scholarship might very well have been viewed as marginal, unimportant or nonexistent. And the question of the extent of community cohesion is important too, not least because it is important to miners themselves. If the goal of the sociological project is Weber’s Verstehen
294
Bryan Pfaffenberger
(subjective understanding), scholars working in the community studies tradition are doing a fine job. What I should like to explore, though, is the potential that a somewhat different view of mining communities might offer additional explanatory value. This view brings ‘technology and machines’ back into view, but in a new and, I think, much more fruitful way. My basic point here can be summed up very simply: in the literature on mining communities, the distinction between technology and community is often too sharply drawn. It is important to realize that technology is no longer regarded as a merely physical or material dimension of social life, divorced from the ‘softer’ aspects of human life such as community and meaning. The use of even the most rudimentary tools, after all, requires some degree of socialization and knowledge, and is thus a social as well as a material phenomenon. Increasingly, technology is understood to be not so much a matter of things, but of activities, in which the technical and the social are very difficult to distinguish. For this reason, French anthropologists and archaeologists prefer to speak not of the vague concept of technology, but rather of a chaîne opératoire, a concept that stems from the seminal work of André Leroi-Gourhan (1943), which is unfortunately not very well known in English-speaking countries. In what follows, I should like to show that this concept is very useful for understanding the nature and dynamics of mining communities. In brief, the chaîne opératoire (Gosselain 1992; Lemonnier 1983, 1986, 1989; Schlanger 1994) is a conventionalized, learned sequence of technical operations, tightly imbricated with patterned social relations, that must be carefully and empirically described as an initial step in grasping the nature and implications of technological activities. A chaîne opératoire is not dictated by the ‘one best way’ to accomplish a task technologically; rather, it represents a choice among what generally turn out to be dozens, hundreds, or even thousands of options, a point not generally appreciated by those unfamiliar with technology. In the choice of technique, people express meaning and political intention—or, in the case of an imposed chaîne opératoire, have it chosen for them. If we draw the distinction between technology and community too sharply, we fail to see chaînes opératoires, which are simultaneously technical, social and cultural. As a result, they are not often described. Happily, there is a wonderful description of a chaîne opératoire to be found right in this volume: Rule’s description of the tribute system in Cornish mining, in which small groups—as few as a couple of men and a boy—would contract to work a measured portion of the mine (called a pitch) in exchange for a share of the price fetched by the ore. There is a great deal to be gained by focusing on chaînes opératoires. From Rule’s description of the tribute system, we learn that the tribute system simultaneously encouraged exploration and speculation while, at the same time, it elegantly and efficiently exploited the varying skill levels of available tributers; for mining companies, it served the further benefit of placing the miners in competition with one another, placing barriers in the path of any
Mining communities, chaînes opératoires and sociotechnical systems
295
movement to foster solidarity among the mine-workers. Fortune played a role as well: tributers who happened upon a spectacular gain provided motivation for others. Combined with the very real dangers of mining, the tribute system drove home a social experience that sounds like a page out of Ecclesiastes (9:11), such that time and chance happened to them all—and it is hardly surprising, as Rule’s analysis suggests, that this experience would find a profound expression in a Methodism that stressed acquiescence to the vagaries of fate. (Here, one is reminded forcefully of the Bolivian miner’s esteem not for Christ, but rather for the devil [Nash 1979; Taussig 1980], who is seen to control the supply of ore and to monitor underground safety.) Taken together, all these attributes help to explain the notorious political apathy of Cornish mine-workers, and the relative infrequency of industrial action in Cornish mining communities. As Rule’s analysis shows, a focus on the chaîne opératoire provides a means for addressing some of the questions that a community focus cannot fully answer. In addition, it works just as well for preindustrial mining, as shown in Childs’s equally exemplary analysis (this volume, Chapter 8) of Toro mining, which reveals how Toro communities forestalled the very real potential for conflict over mining discoveries. Competition for ore was intense, for it provided a pathway to considerable wealth and status. Childs’s description of the public festival that served to announce the discoverer’s ownership of the vein, as well as the careful means by which the vein owner distributed spatially separated portions of the area to relatives, neighbors and clients, shows the concern which would be quite naturally felt, namely that the discovery could divide the community. Here, too, one sees the imbrication of the religious, social and technical in chaînes opératoires; the various parties working the hill were given widely separate plots, because ‘they were scared the world would turn upside down.’ The men involved in mining did not miss the chance to drive home the patriarchalism of Toro society, in which women are seen as a threat to the community’s spiritual welfare (see Herbert, this volume, Chapter 9); men had to abstain from sexual relations while mining lest they ‘spoil’ the iron. Rule’s and Childs’s chapters powerfully demonstrate another point I should like to make about chaînes opératoires: that is, their potential to create and sustain fields of intersubjective meaning, a job supposedly fulfilled by face-to-face interaction in the community studies model. Although mining community interaction is a powerful generator of community culture, it is also quite clear that chaînes opératoires appear equally powerful in this respect. To understand the culture of mining communities fully, it is important to look at chaînes opératoires as well as face-to-face interaction outside the mines. As the next section argues, this is particularly important when the chaîne opératoire is imposed, as is the case in company mining towns and what Godoy (1985:207) calls the ‘hermetic, total institutions’ of the South African mines (Crush 1992).
296
Bryan Pfaffenberger
DIVIDE AND CONQUER In industrial mining, the workers’ communities are situated within a broader sociotechnical system (Hughes 1983) of enormous scope and complexity. Hughes developed the sociotechnical system concept to explain the history and dynamics of the electrical power industry in the USA. As Hughes found, those who seek to develop new industrial ventures must not only concern themselves with Lawrence’s tools and machines, but also engineer the social, economic, legal, political and cultural dimensions of the system they are trying to build. A successful sociotechnical system reflects the deliberate modification of the system’s various components so that they work together harmoniously; in this sense, all of the system’s components—laws, political relationships, social arrangements of manufacture and use—are rightfully seen as artifacts. Little understanding of any one component can be gained by looking at it in isolation, for as Hughes remarked, ‘the social is indissolubly linked with the technological and the economic’ (1983:112). What I would like to argue here is that, as against the community studies model, modern mining communities can be understood only by situating them within the sociotechnical system of capitalist mining, and by understanding this system’s peculiar dynamics. Like any sociotechnical system, industrial mining is fabulously complex. Its components include vast hinterland supply regions, transport systems such as roads and railways, manufacturers and distributors of mining equipment, political arrangements (bribes and donations), laws and regulations, labor unions, management styles, and much more. But a dominant factor is the difficulty of securing capital (Godoy 1985). Mining is inherently an extremely risky investment, and for a multitude of reasons: discoveries might not pan out, scams are common, ore prices fluctuate wildly on world markets and, until the IMF started throwing its weight around, Third World nations were quite eager to nationalize mines once investors had sunk their money in (just as the mines were becoming productive). The capitalization of mining has often been left to regional investors who are better positioned to understand, manipulate and profit from mining, such as the merchant monopolists of Cornwall and Devon tin mining (Lewis 1915), who both advanced risk capital and transported the ore to coinages. Starved for capital, industrial mining faces a worsening picture as time goes on: the richest deposits have already been exploited, and the newly discovered ones contain poorer-quality ore, which requires all the more capital to mine profitably. Since mining is inherently a labor-intensive business, and capital is increasingly hard to come by, pressure mounts to reduce costs by holding down the cost of labor. Here, the age-old strategy is ‘divide and conquer.’ Mining companies have a great deal to gain by manipulating and exploiting potential ethnic divisions in mining communities, and exploit they did. This point is sensitively described in Sheridan’s brilliant chapter on the history of technology, capital, labor and ethnicity in Arizona mining (this volume, Chapter 11). Mining companies cynically manipulated ethnic and racial divisions to hold
Mining communities, chaînes opératoires and sociotechnical systems
297
down the cost of labor. Many Anglo mine-workers wanted white-only mining communities, but companies did not share much enthusiasm for this; for example, Sheridan shows how companies at first (unsuccessfully) attempted to bring in Chinese workers, and then used cheaper Mexican labor to forestall worker solidarity. The ‘divide and conquer’ mentality is seen everywhere in the company town: Hardesty (this volume, Chapter 5) shows how the company town’s elaborate spatial plan was designed both to impose and to maintain a class structure based on ‘ethnicity, race, class, wealth, and occupation,’ which could only serve to deepen ethnic and racial antagonisms and further divide the community (for an example drawn from New Guinea, see Polier and Lindenbaum 1995). The more the various constituent groups are at each other’s throats, the less the threat of unionization. The company’s actions would seem to encourage ethnic and racial solidarity as a mode of adaptation, and that is doubtless the case, but a thriving mining community introduces yet another mechanism of cleavage: class. Hardesty (this volume, citing Frehner 1996), recounts how working-class and middle-class Italians came to blows over the price of the charcoal that the working-class Italians (carbonari) supplied to their store-owning compatriots. The store owners, headed for acculturation and upwardly mobile, joined with other townspeople in opposing the carbonari when the working-class Italians organized an association and withheld their charcoal. Ethnic divisions would rise to prominence as companies exploit them in order to hold down the cost of labor—but, paradoxically, the community’s economic success begins to undercut ethnicity’s value as a principle of association and solidarity, resulting in a crosscutting fabric of divided loyalties that plays right into the companies’ hands.
RESISTANCE AND UTOPIANISM The companies’ drive for control is met with a mixture of acquiescence and resistance. Within the town, secure employment and advance require conformity to the company’s many rules and regulations, which often extend far beyond work regulations in an attempt to control virtually every aspect of miners’ lives. In a contemporary Western company mining town, for example, company residents state that they are willing to conform to the regulations, despite disliking them, in the hope of material gain and advancement (Graham 1980). The more companies drive for control, though, the more chance that they will encounter covert forms of resistance, such as ore theft, desertions, loafing and deliberate wastefulness (Godoy 1985:207), not to mention a good deal of what Vaught (1978) calls ‘horseplay.’ In this, we see a social strategy that Moore (1975) calls situational adjustment, in which people react to invidious, domineering structures by engaging in redressive actions, which are meant to ward off the negative status implications that dominating system imposes. At the extreme, the companies’ efforts to divide workers fail, and unions succeed in uniting the mine-working community.
298
Bryan Pfaffenberger
Sabotage, resistance and unionization are not the only possibilities; several of the chapters in this volume highlight a different strategy altogether, one in which miners opt out of the company framework in order to create their own, freewheeling mining communities. Several of the authors of this volume would ascribe this to the relatively loose (‘heterarchical’) control that mining companies are able to exert over the wider mining region, and this is surely a factor, but it is worth stressing that this redressive process is yet another instance of the age-old human response to unwanted domination, namely, utopianism, millenarianism, communitarianism, counter-cultures, and other forms of intentionally rebellious community formation. Lawrence (this volume, Chapter 3) cites Phillips’s (1987) study of the ethos of egalitarianism found in the ‘poor man’s diggings’ of nineteenth-century Victoria, as well as Hogan’s interesting distinction (1990) between caucus communities—towns controlled by company officials and local capital—and the more independently minded carnival communities in Colorado, which were self-governed and far more egalitarian. In contemporary Nevada, the citizens of self-governed Red Butte do not have the economic opportunities of miners living in a nearby company town, but they place greater value on the freedom of their lifestyle (Graham 1980). The utopian impulse is seen most tellingly in Hardesty’s interesting portrait of Reipetown, a wide-open, turn-of-the-century mining camp in Nevada. An 1890s version of Haight-Ashbury, Reipetown fostered union radicalism as well as ‘alternative lifestyle’ emphasizing saloons, gambling, brothels and a tendency to take the law into one’s own hands. Ethnic differentiation and conflict, so common where companies exploited it for their own purposes, seem to have been absent or nearly so, as were class divisions. Reipetown (and analogous manifestations of miners’ counter-cultural egalitarianism) suggests that the bitter ethnic and racial conflicts of frontier mining communities are best ascribed not so much to endemic racism, but rather to the actions of the mining companies in deliberately fanning the flames. One could imagine that the mining arrangements in Reipetown more closely resembled the egalitarian chaînes opératoires of a preindustrial mining community, which would take enormous care to make sure—as do Toro miners—that open conflict is avoided, rather than expressly invited. In the end, Reipetown is best seen as the last act in a technological drama, one that is set into motion by a sociotechnical system prone to accentuate ethnic differentiation and, in the extreme, as in South Africa, to approximate a total, despotic institution.
SOME POTENTIAL LESSONS FOR ARCHAEOLOGY For archaeologists, what is to be learned from this chapter’s attempt to reconcile the community focus with a renewed emphasis on technology and, in particular, chaînes opératoires and sociotechnical systems? As we look closely at a mining community, it is clear that a great deal can be gained by piecing
Mining communities, chaînes opératoires and sociotechnical systems
299
together the chaîne opératoire—just how mining was organized as a step-by-step sociotechnical activity, and whether the details of this activity were developed with some degree of independence, or imposed by some external agency. Stepping back a bit, it is equally important to consider just how the specifics of productive organization were adapted so that they could operate smoothly in a broader field of constraints, those imposed by the wider sociotechnical system of which the community was a part. Piecing together this picture can provide a great deal of insight into the nature and dynamics of mining communities, even if it is not possible to achieve the subjective understanding that a study of face-to-face relations makes possible. And most importantly, I would hope this chapter will convince the skeptical that looking at technology by no means rules out an appreciation of the human and social dimensions of community life, but, on the contrary, makes a fuller appreciation more possible.
REFERENCES Bauman, J. (1979) Ethnic adaptation in a southwestern Pennsylvania coal patch, 1910– 1940 . Journal of Ethnic Studies 7: 1–23. Cotterell, J. (1984) Social networks of mining town women. Australian Journal of Social Issues 19: 101–112. Crush, J. (1992) Power and surveillance on the South African gold mines. Journal of Southern African Studies: 825–844. Davis, R. (1967) Soiled doves and ornamental culture. The American West 4: 19–25, 69– 70. Frehner, B. (1996) Ethnicity and class: the Italian charcoal burners’ war, 1875–1885 . Nevada Historical Society Quarterly 39(1): 43–62. Freudenberg, W. (1991) Addictive economies: extractive industries and vulnerable localities in a changing world economy. Unpublished paper presented at the annual meeting of the Rural Sociological Society, 1991. Godoy, R. (1985) Mining: anthropological perspectives. Annual Review of Anthropology 14: 199–217. Gosselain, O. (1992) Technology and style: potters and pottery among the Bafia of Cameroon. Man 27: 559–586. Graham, S. (1980) Community, conformity, and career: patterns of social interaction in two Arizona mining towns. Urban Anthropology 9: 1–20. Hogan, R. (1990) Class and Community in Frontier Colorado. Lawrence: University Press of Kansas. Hughes, T.P. (1983) Networks of Power: Electrification in Western Society. Baltimore: Johns Hopkins University Press. James, R., Adkins, R. and Hartigan, R. (1994) Competition and coexistence in the laundry: a view of the Comstock. Western Historical Quarterly 25: 164–184. Kaplan, T. (1981) Class consciousness and community in nineteenth-century Andalusia. Political Power and Social Theory: 21–57. Kroll-Smith, J.S. and Couch, S. (1991) The Real Disaster is above Ground: A Mine Fire and Social Conflict. Lexington, KY: University Press of Kentucky. Lemonnier, P. (1983) L’Étude des systèmes techniques, une urgence en technologie culturelle. Techniques et Culture 1: 11–34.
300
Bryan Pfaffenberger
——(1986) The study of material culture today: toward an anthropology of technical systems. Journal of Anthropological Archaeology 5: 147–186. ——(1989) Bark capes, arrowheads, and the Concorde: on social representations of technology. In I.Hodder (ed.), The Meaning of Things: Material Culture and Symbolic Expression, pp. 156–171. London: Unwin Hyman. Leroi-Gourhan, A. (1943) Évolution et techniques: l’homme et la matière. Paris: Albin Michel. Lewis, G. (1915) The Stanneries. Cambridge, MA: Harvard University Press. Moore, S.F. (1975) ‘Epilogue’. In S.F.Moore and B.G.Myerhoff (eds), Symbol and Politics in Communal Ideology: Cases and Questions, pp. 210–245. Ithaca, NY: Cornell University Press. Nash, J. (1979) We Eat the Mines and the Mines Eat Us: Dependency and Exploitation in Bolivian Tin Mines. New York: Columbia University Press. Phillip, J. (1987) A Poor Man’s Diggings: Mining and Community at Bethanga, Victoria, 1875– 1912. Melbourne: Hyland House. Polier, N. and Lindenbaum, S. (1995) A view from the ‘cyanide room’: politics and culture in a mining township in Papua New Guinea. Identities: Global Studies in Culture and Power 1: 63–84. Robinson, D. and Wilkinson, D. (1995) Sense of community in a remote mining town: validating a neighborhood cohesion scale. American Journal of Community Psychology 23: 137–148. Schlanger, N. (1994) Mindful technology: unleashing the chaîne opératoire for an archaeology of mind. In C.Renfrew and E.Zubrow (eds), The Ancient Mind: Elements of Cognitive Archaeology, pp. 143–151. Cambridge: Cambridge University Press. Taussig, M. (1980) The Devil and Commodity Fetishism in South America. Chapel Hill: University of North Carolina Press. Vaught, C. (1978) Hanging, hairing, greasing, and coal dust queens: deviant games in a hostile work setting. Unpublished paper presented at the annual meeting of the Mid-South Sociological Association.
Index
Agricola 231 agriculture and mining 15–16, 19, 48, 49, 149, 151 Akan regions, Ghana 142–6 alabarch (customs official) 271 Almyras; see Ayia Varvara Almyras Alonso, A.M. 284 Amenemhat II, royal annals of 248–50 Amenemhat III 250 anthropology 2–19, 97–8, 99, 106, 207 anvil as Inuit ritual object 117, 118 archaeology: gendered 17–18, 40, 60; lack of in Arizona 186; of mining community 13–15, 87–8, 94, 205–7; social 18–19 Arizona: corporate domination of 174–5, 181; mining industry, race and class in 174–87; 1917 strikes in 184–5; racial conflict in 175, 176, 178, 179; statehood campaign in 181; taxation policies in 183 Australia: Chinese in 32–3, 45–6; Cornish in 32; gender and community structure in 39–55; gold rush in 28; mining settlements in 27–38; tribal ochre miners in 218; women in mining camps in 27–38, 39–55 Ayia Varvara Almyras 226, 229–32; artefacts from 232–4; beneficiation at 231, 235; cultural and historical context of 235–8; miners in 232–5; mining at 229–31, 235–6; production, scale of at 232; roasting ore 231, 235; rock chambers at 232–3, 236; smelting 231–32, 236; women in 234–5 Baffin Island 109, 111 Ban Chiang Cultural Tradition, bronze in 205, 207, 222
basket chains, use of 217 Beit Arieh, I. 243 Bellagio Conference on the Archaeology and Anthropology of Mining xiii-xiv, 2–3, 281, 288 Bellamy, C.V. 146, 149 Best, George 110–15 Bir Umm Fawakhir, Egypt: artefacts at 263–5; context of 271–4; description of 259–62; Diodorus Siculus and 267–71; mining at 259–74; population of 260 Bir Umm Fawakhir Survey Project 259 Bisbee Deportation 185 Bogue, Allan 85 boom aspect of mining settlements 27, 29, 61, 63 Botallack Mine, Cornwall 169–70 Bourdieu, Pierre 87 Brodie, Alexander 180 Bronson, B. 6 Bronze Age, studies of copper mining in 191–3 Brooks, Allyson 83–4 brothel, as household type 71–4 Buckner, J.C. 292–3 Bunting, Jabez 162 burials, use of red ochre in 218 Burke, Gillian 172 Burra, Australia 30, 31, 32 Burton, John 215–17 business, hierarchichal spatial patterns of in mining towns 63 Butler, Jim 103 Butte, Montana, Irish in 85–6 Byrkit, James 176–82 Byzantine empire 271
302
Index
California, mining stereotypes from 100–1 carnival community 43, 85, 298 Castel, Georges 244 casting cake, analysis of 198, 199 Cathay Company 110, 111 caucus community 43, 85, 298 chaîne opératoire 281, 287, 294–5, 298–9 Chambers, A. 82 chapel culture; see Methodism in Cornwall Childe, Gordon 191 Chinese: in Australia 32–3, 45–6; exclusion of in Arizona 177–8; prostitution 68–70, 74 Christianized population, Bir Umm Fawakhir 263, 273 class: and community power 84–5; and layout of company towns 88; and material culture studies 282–3; and race in Arizona 174–7 Clifton-Morenci strike 179–80 Coalbrookdale/Ironbridge Gorge 280 Cohen, Anthony 98 Colls, R. 167 colonialism: and mining 112–15; understanding of 11 Colquhoun, James 177–8 comes largitionum sacrum, role of 272 communal activities, shared 41–2 communities: autonomous 201–2, 203; development of 85, 298; as generators of culture 295; see also mining communities community studies perspective 40–1, 293 company towns 88–9 competition, economic and ethnic consciousness 284–5 Comstock Lode 81 Conkey, M. 40 copper communities, ethnic conflict in 176, 177–9, 186 copper companies, anti-labor tactics of in Arizona 183–5 copper: in Bronze Age 201–2; in Cornwall 157–8; corporate investment in 175, 178; economic and political role of in Cyprus 236–7; effect of fluctuations in 90; in Katanga 142; religion and 236 Cornwall, mining in 155–72; accident rate in 160–1; copper in 157–8; disasters 160–2; documentation of 156; gender differentiation in 157; history of 156–8; lung disease in 161; mortality rates in
159–60, 162; risks of 158–60; tribute system in 158–9, 288, 294–5; weak unionization in 158 corvée, system of 246 Costin, Cathy 221, 222 Craddock, P. 3, 10 craft specialists, independence of 221 Crumley, C.L. 82 Cyprus, ancient, mining and smelting in 226–38; copper industry in 236–7; in Late Bronze Age 226–7; slag heaps as evidence of 227; study of 227–9; see also Ayia Varvara Almyras Data brothers 90, 92, 292 Davis, R. 293 Deagan, K.A. 87 decay process, selective nature of 36–7 Demos, J. 42 Deutsches Bergbau-Museum, Bochum 279 Diaz, Porfirio 178 Diodorus Siculus 235, 267–71 dipinti 263 Discipline and Punish (Foucault) 94 Distinction, a Social Critique of the Judgment of Taste (Bourdieu) 87 distribution, riverine 199, 201, 221 Dolly’s Creek, Australia 43–54; demography of 46–7; gender and material culture in 50–4; work in 48–50 dominance, culture of in company towns 88 Douglas, Walter 183–4, 185 Douglass, Bill, 28 Dubofsky, M. 84 Dumett, R.E. 149 Dürrnberg, salt-mining site in 199 Dykstra, Robert 85 East Wheal Rose, flooding of 160, 168 Egypt; see Ayia Varvara Almyras; Bir Umm Fawakhir; Middle Kingdom; Old Kingdom ekijunjumira beetle 123, 128 El Gayar, El Sayed 243 El Tío figure 12, 13 Eliade, Mercia 12, 152 Emmons, David 84, 85–6 ethnic: compression and differentiation, process of 186; consciousness and economic competition 284–5; divisions, manipulation of 85–6, 296–7; exclusion 177
Index ethnicity: archaeological markers of 285– 6; and community power 85–6; and material culture studies 284–6; and prostitution 65–8; of workforce 281 ethnohistory and ethnography 11–13 Eureka, Nevada, Italians in 85–6 exchange networks, interregional 200–1 family: at Bir Umm Fawakhir 270; as unit of production 49, 127; Victorian esteem for 50 Ferguson, Leland 87 Fischer, Henry 248 folklore, Inuit 116 Forgeron et alchimistes (Eliade) 152 Foucault, Michel 94 Frehner, Brian 86 Frobisher, Martin, expeditions of 109, 110–12 frontier: understanding challenge of 11; wageworkers 84 Galen 18 gender: and archaeology of mining 17–18, 40, 60; and material culture 50–4, 283– 4; and mining community structure 39–55; in mining occupations 135, 140, 157; roles in Australian mining community 45, 46–7 General and Brief Description of the Country and Condition of the People Which are Found in Meta Incognita (Best) 115 Globe, Arizona, race conflict in 179 Godoy, R. 295 gold mining: in Akan region, Ghana 142– 6; informal 82–3; in Old Kingdom 244; symbolic elements of 151, 253 gold rush communities 41–3; extended boundaries of 43; spontaneity of 28 Goldfield, Nevada, establishment of 104 Gratien, B. 248 Great Revival 163, 167 Gunther, C.G. 227 Hadjicosti, Maria 238 Hall, Charles 115–17 Hamilton-Jenkin, A.K. 167 hammerstones, Tungei 197, 217 Hathor and mining in Egypt 253–5 Hemptinne, Monsignor de 142, 146 heterarchies 82 historical events, impact of 15
303
Hogan, Richard 43, 84–5, 298 household types in mining communities 7, 47, 50–4, 71 housing, company built 31, 32 Hughes, L.C. 178 immigration: international 89, 175, 178; Mexican 178–9 imperialism, understanding of 11 individualism in mining communities 61 industrial mining: class-consciousness in 87; fear of sabotage in 184; labor and management in 296–7; local control of production in 85 industrial mining community, power in 81–94; in American West 81–2; archaeology of 87–8, 94; and class 84– 5; ethnicity and 85–6; informal mining and 82–3; mobility and 83–4; networks 82–4 Industrial Workers of the World (Wobblies) 174, 184 Intef, stele of 247 interaction of miners and indigenous population 109–18 interdisciplinary approaches, integration of 3, 8–9, 281–2 International Union of Mine, Mill and Smelter Workers (IUMMSW) 184 Inuit, impact of Europeans on 115–16; miners’ relationships with 112–15 Irish in Butte, Montana 85–6 iron ore, types of 130–1, 132 ironworking: gender attributes of tools for 132, 134; socio-economic role of among Toro, 134–5 Italians in Eureka, Nevada 85–6 Journal of Material Culture 282 Kent, Susan 84, 283 Khao Tab Kwai, as ochre source 218 Khnumhotep III, tomb of 248–9 kin structure, network of 47 Klemm, D.D. 244 Klemm, R. 244 Klinglberg, the 193–8; amber bead, as evidence of trade in 200; exchange region of 195, 199–201; metal exchange, social control of 201–3; production region of 194, 196, 198–9 Kwanyama metalworkers, Angola 141–2
304
Index
labor organization, Arizona 179–86; and prehistoric, 191–8, 282; professional Arizona constitution 181; and Egyptian 246–7; sexual abstinence of CliftonMorenci strike 182–3; 130, 150; social and gender categories differential wage scale, impact of 179, of, sub-Saharan Africa 140 183; Mexican labor, fight to limit 181– mining: and agriculture 15–16, 19, 48, 49, 2; see also Western Federation of Miners 151; archaeology and anthropology of, Lankton, Larry 280 2–19; boom aspect of 27, 28, 61, 63; Leroi-Gourhan, André 294 capitalism and 88, 107, 296; centralized Lesinsky, Henry 177 control of 202; colonialism in 112–15; Little Commonwealth, A (Demos) 42 decline of 8, 102–3; division of labor in Liversage, D. 200–1 124–5, 129–30, 135, 145, 295; food Lok, Michael 110, 111 production analogies in 149, 151, 152; Loutherbourg, Phillippe de 280 frontier, interaction spheres of 105, lung disease 161, 172 255–6; functions of 2; informal 4–5, 7, 49, 60–1, 82–4, 124; and Islamic world system 286–7; resistance to 286; McGill, Nevada 88, 89 ritualization of 22, 130, 131, 140, 149– McGuire, R.H. 87 51, 246; seasonal 127, 145, 218, 219, Macmillan, Gordon 82 220, 244–6, 255; technological change Mandingo, of Sudan 150 in 18, 280, 281; types of, 28, 141, 142, Mann, Michael, 82 145–6, 265, 267, 268 manufactured goods, demand for 30–1 mining camps: anthropological analysis of marriage contract, importance of 63–4 99, 106; archaeology of 98; as Martin, Patrick 285 communities 97–107; heterogeneity of material culture: as display of affiliation or 101–2; impermanence of 98; see also status 71; functions of 9, 87; gender mining communities; mining and 50–4, 283–4; lack of variability in settlements 5–6; minimal in informal mining, 84; mining communities: characteristics of 6, spatial distribution of 281 60–1; cohesion in 292–3; conditioning Mbeere, of Kenya 150 of by socioeconomic system 13, 14–15, Mentuhotep IV, written records of 251 39, 203; daily operation of 54–5; decline metal: symbolic role of 218; as primitive of 35–7, 280; demographic variability of money 202 14; dynamics of 4; ephemerality of 279; metallurgy: effects of 139–40; functions of gender and structure of 39–55; 2; Hellenic installations 2, 34; link of household types in 7, 47, 50–4, 71; with forests 16; symbolic content of 2, individualism in 61; locational factors of Methodism in Cornwall 162–72; 12, 14; pejorative view of 291; broadsheets, religious 169; death, prehistoric 6, 7; residential patterns in preparation for 168–9; funerals 171–2; 11–12, 178; social and sociotechnical local control in 162–3; revivals in 163– systems in 10–11, 106, 291–9; see also 8; role of 156, 162–8; Wesleyan 162 mining camps; mining settlements Mexican labor: exploitation of 175, 176; mining communities, study of 39–41, 279– opposition to 178–9, 181–2; as scabs 88, 293; anthropological neglect of 97– 178; strike by 179–80 98; colonialism and imperialism as Mexican Revolution 178 focus for, 286–7; as conference theme Middle Kingdom, mines in 244, 247 2–4; historical approaches to 281; miner’s cottage 31 material culture studies, role of in 281– Miner’s Right 49 6; work, social organization of as focus minerals: distribution of 4, 139; location of for 287–8; themes and issues in 8–18 27, 28, 42–3 mining expeditions, ancient Egyptian 242– miners: mass kidnapping of in Arizona 57; evidence for 243–5; frequency of 185; mobility of 4–5, 29–30, 83–4; 250–2; miners in 244–7; political mortality rates of 159–60, 162;
Index
305
motivations for 248–52, 256–7; ritual significance of 218, 253–5; seasonal nature of 244–6; see also Tungei mining expeditions Mining Journal 171 mining settlements: decline of 35–7, 280; ethnicity and 31–3, 255; history of 27– 8; housing in 29–31, 35; layout of 33– 5, 42, 88, 126–7; population of 29; in regional system 199; study of 39–41; undefended 198–9, 263; see also mining camps, mining communities mobility, anticipated 283 Montero-Ruiz, I. 192 Moore, S.F. 297 Morrill, R. 63 Mount Gabriel mine, Ireland 192 Moyer, Charles 180, 183 mutualistas (mutual aid societies) 180, 182
phthisis; see lung disease Phu Lon, Thailand 207–14, 218–22; crucibles, use of 212–15; domestic artefacts at 212; evidence of mining techniques at 208–12; extent of mining at 212; ingot production at 213–14; in regional production system 214, 219; seasonal expeditions to 218, 219, 220; see also Northeast Thailand pitch system; see tribute system placeros (placer miners) 177 Powers, Lillian 65–6 preindustrial mining technologies, transformation of 177, 186 processed cereals, Bronze Age use of 197, 198 procurement: of siltstone 250, 251; seasonal direct access 219 prospecting: divinatory techniques for 150; among Toro 127–8; for silver in Arizona 175, 176–7; unsystematic, in Neighborhood Cohesion Index (NCI) 292–3 Phu Lon 219 neocolonialism in Arizona 175 prostitutes: economic power of 74; political Nevada Charcoal Burners War 86, 284 power of 64–5; as socioeconomic group New Zealand, prostitution in 66–7, 76 59–60 newspapers, corporate control of 183–4 prostitution 59–78; attitudes about 75–6; Northeast Thailand, prehistoric copper economic relations in 74–5; mining in 205–22; anthropological entertainment trade and 61–3; and approach to 207; archaeological sites of gender relations 63–5; indigenous 205–7; consumer interaction 12, 215– women in 67–8; local distribution of 22; controlled access, lack of 216, 220; 77–8; material culture of 71–4; in New production and distribution channels, Zealand 66–7, 76; reasons for entering multizonal model for 220–2; see also 63; regulation of 76–7; social Phu Lon; Tungei mining expeditions relationships in 75–6; stratification and Northwest Passage, search for 109, 112 ethnicity in 64–8 nyakatagara; see spirit mediums Pyramid Texts 352 O’Brien, William 192, 212, 213, 215–16, 219, 220 Old Kingdom, gold mines of 244–7 omujumbuzi (Toro ore discoverer) 128, 129, 130, 131 ore, Toro prospecting for 127–8 ore sources, access to in Klinglberg 198 ore-crushing blocks at Bir Umm Kawakhir 265, 266, 270 Orser, C.E., Jr 87–8 Park, Mungo 149 parlour houses 71–4 Paynter, R. 87 Phelps Dodge 176, 183–6 Philipp, June 43
qallunats (Europeans) 116 querns 270 quota law, anti-Mexican 182 race, as social product 284 Raich, Mike 182 railroads, building of 81 Reipetown, Nevada 89–94, 298 revivals: descriptions of 163–7; periodicity of 163, 167; mental terrorism in 168 Reynolds, Neil V. xiii rituals; see mining, ritualization of river cobble mauls 209, 211 Robinson mining district, Nevada 88–9 Roman empire, state-run mining in 16
306
Index
Roosevelt, Theodore 180 Rothenberg, B. 243 Rudna Glava, Yugoslavia 219 Sagona, A.G. 218 StVeit-Klinglberg; see Klinglberg, the Salzach valley, exchange links in 199–200 satellite towns 89–94 Satire of the Trades 246 Schwantes, Carlos 84 Shakel, P.A. 87 Shanks, M. 87 Sherratt, Andrew 216, 220 Siguiri region, Guinea 151 siltstone, procurement of 251 silver: prospecting for 176–7; technology of mining 177 situational adjustment 297–8 slavery: power games in 87–8; role of 141 smelting: in classical Cyprus 234; dispersal of in Klinglberg 198; preparation of ore for in sub-Saharan Africa 149; Toro technology and rituals of 131–3 space, social division of 75, 282 Spain, copper mining in Bronze Age 192 speculation, in Cornish mines 157 spirit mediums, Toro (nyakatagara) 126, 128, 130, 131, 132, 134 stratification: emergence of in Bronze Age 192; in prostitution 64–8; among Toro 125–6 sub-Saharan Africa, mining in 138–52; history of 139–41; miners in 141–9; rituals in 149–51 sumptuary goods 282 supply infrastructure, international 31 supply zone, indirect 221 Sydney-Cyprus Survey Project (SCSP) 227, 282 Taussig, M. 12–13 technology: dimensions of 2; effect of change in 18, 280, 281; social context of 2, 19, 287; socialization to 294; and social process 18 Thailand Archaeometallurgy Project (TAP) 205, 207
Tilley, C. 87 Tonopah, Nevada 99–100, 102–6 Toro of Uganda, precolonial iron mining among 123–36, 295; characteristics of 124–5; dynamics of 127–31; economic context of 125–7; forging 134; smelting 131–3; sociopolitical context of 125–7 trade, evidence of in Klinglberg 140, 199– 200 Treffry, R. 163 tribute system 158–9, 288, 294–5 Trigger, Bruce 282 Trimble, W.J. 101 Tungei mining expeditions, 215–17; as cooperative ventures 216, 217; conduct of 217; organization of quarrying 217 Uncommon Ground (Ferguson) 87 Vernon, William 214 Wadi el-Sid, mines in 266–7 Wadi Hammamat, inscription in 250–1 wage workers: frontier 84; power games of 88 Ward, W.R. 163 Webb, J.A. 218 Wesley, John 162 Western Federation of Miners 90, 178, 179, 180, 182–3, 184 Wheeler, Henry 184–5 White, J.C. 221–2 Williams, S. 40 Wilson, Woodrow 185 Winthrop, K. 82 Wobblies; see Industrial Workers of the World women: importance of 156, 283, 292; in mining 146, 149, 150; prohibition of at smelt 132; see also gender World War I, impact of on copper industry 184 Wylie, A. 19 zoning to regulate prostitution 77