Craft Specialization: Operational Sequences and Beyond: Papers from the EAA Third Annual Meeting at Ravenna 1997. Volume IV 9781407350370, 9780860548973

Citation preview

BAR S720 1998 MILLIKEN & VIDALE (Eds): CRAFT SECIALIZATION EAA 3 RAVENNA 1997 VOL IV

B A R

01/07/2010 10:23:54 red cover template.indd 1

UniversitadegliStudi di Bologna Dipartimento di Archeologia Facoltadi Conservazione deiBeniCulturali.Sededi Ravenna

Craft Specialization: Operational Sequences and Beyond Papers from the EAA Third Annual Meeting at Ravenna 1997 Volume IV Edited by

Sarah Milliken & Massimo Vidale European Association

of

Archaeologists

Third Annual Meeting Ravenna, September 24-28 1997

BAR International Series 720 1998

Published in 2016 by BAR Publishing, Oxford BAR International Series 720 Craft Specialization: Operational Sequences and Beyond

© The editors and contributors severally and the Publisher 1998 Volume editor: David Davison The authors' moral rights under the 1988 UK Copyright, Designs and Patents Act are hereby expressly asserted. All rights reserved. No part of this work may be copied, reproduced, stored, sold, distributed, scanned, saved in any form of digital format or transmitted in any form digitally, without the written permission of the Publisher.

ISBN 9780860548973 paperback ISBN 9781407350370 e-format DOI https://doi.org/10.30861/9780860548973 A catalogue record for this book is available from the British Library BAR Publishing is the trading name of British Archaeological Reports (Oxford) Ltd. British Archaeological Reports was first incorporated in 197 4 to publish the BAR Series, International and British. In 1992 Hadrian Books Ltd became part of the BAR group. This volume was originally published by Archaeopress in conjunction with British Archaeological Reports (Oxford) Ltd/ Hadrian Books Ltd, the Series principal publisher, in 1998. This present volume is published by BAR Publishing, 2016.

BAR

PUBLISHING BAR titles are available from:

EMAIL

PHONE FAX

BAR Publishing 122 Banbury Rd, Oxford, OX2 7BP, UK [email protected] +44 (0)1865 310431 +44 (0)1865 316916 www.barpublishing.com

UniversitadegliStudidi Bologna Dipartimento di Archeologia Facoltadi Conservazione deiBeniCulturali- Sededi Ravenna

Papers from the EAA Third Annual Meeting at Ravenna 1997 Volume I

Pre- and Protohistory Edited by Mark Pearce and Maurizio Tosi with Anna Maria Bietti Sestieri, Serge Cleuziou, Alessandro Guidi, Ludmila Koryakova, Pietro Laureano, Mike Rowlands, Nataliya Shishlina, Simon Stoddart, Andrea Zifferero BAR International Series 717

Volume II

ISBN 0 86054 894 5

Classical and Medieval Edited by Mark Pearce and Maurizio Tosi with Andrea Augenti, Hugo Blake, Paolo Carafa, Cristina Tonghini, Guido Vannini BAR International Series 718

Volume III

ISBN 0 86054 895 3

Sardinia Edited by Alberto Moravetti with Mark Pearce and Maurizio Tosi BAR International Series 719

Volume IV

ISBN 0 86054 896 1

Craft Specialization: Operational Sequences and Beyond Edited by Sarah Milliken and Massimo Vidale BAR International Series 720

ISBN 0 86054 897 X

Credits Conference Editors:

Mark Pearce, Maurizio Tosi

Archaeopress Editor:

David Davison

Scientific Secretariat:

Maurizio Cattani, Alessandro Naso, Mark Pearce, Maurizio Tosi

Editorial Service:

A.B.A.C.O. / M.A.C. srl, Forli, Italy

European Association

of

Archaeologists

Third Annual Meeting Ravenna, September 24-28 1997

Contents Contributors

iii

Introduction

V

The ghost of Childe and the question of craft specialization in the Palaeolithic Sarah Milliken

Technical behaviour and the identification of social patterning: a preliminary discussion of some new evidence from the Late Neolithic ofNorthern Greece Katerina Skourtopoulou

9

Ornament production centres on the French Atlantic coast during the Late Neolithic Luc Laporte

17

Signs of specialization in a settlement group of the Lengyel culture Judit Regenye

25

Clay, twigs, threads and ritual: ceramic "industrial" decoration in the East European Neolithic and Chalcolithic Dragos Gheorghiu

35

"Sequential slab construction" and other problems concerning hand-building techniques in Chalcolithic Iran: experimenting with mammographic x-ray images Sergio Dipilato & Nicola Laneri

59

Biface standardization accompanying organized chert quarrying efforts: an argument for intensifying lithic production AnneS. Dowd

69

Ceramic production among the Maros villagers of Bronze Age Hungary Kostalena Michelaki

77

The question of levels of specialization levels in pottery production between the end of the Bronze Age and the Early Iron Age in Daunia Paolo Boccuccia

89

Defining social and symbolic changes from the Bronze Age to the Iron Age through the operational sequences in NW Iberian pottery Isabel Cobas Fernandez & Pilar Prieto Martinez

95

Technological study as a means of identifying bronze production forms: the archaeological record of Etruria in the Early Iron Age period Anne Le Fevre-Lehoerff

107

Iron production and power: a story of early large scale production of iron in mid-Norway Kristin Prestvold

117

The reconstruction of manufacturing sequences on the basis of iconography: the case of the Foundry Cup at Berlin Gabriella Prisco & Massimo Vidale

125

The organization of production in the artisan quarter at Rocca d'Evandro Emilia Chiosi

133

Glass beads as an archaeological source Anna Mastykova

139

Pre-industrial mining techniques in the mountains ofCampiglia Marittima (Livorno) Giovanna Cascone & Alessandra Casini

149

The production of metals for coinage in Mediaeval Tuscany: the technological context Silvia Guideri

153

The art of making bread by the charcoal burners of the Calabrian mountains Francesca Lugli

159

Towards a theory of social production and social practices Pedro V. Castro, Sylvia Gil, Vicente Lull, Rafael Mico, Cristina Rihuete, Roberto Risch & M_tlEncama Sanahuja Yll

173

Operational sequences beyond linearity Massimo Vidale

179

11

Contributors

Paolo Boccuccia

Universita degli Studi di Roma "La Sapienza", Dipartimento di Scienze Storiche, Archeologiche e Antropologiche dell'Antichita, Via Palestro 63, 00185 Rome, Italy

Alessandra Casini

Parco Archeologico Minerario di San Silvestro, 57021 Campiglia Marittima (LI), Italy

Giovanna Cascone

Parco Archeologico Minerario di San Silvestro, 57021 Campiglia Marittima (LI), Italy

Pedro V. Castro

Departament d'Antropologia Social i Prehist6ria, Edifici B, Universitat Aut6noma de Barcelona, 08193 Bellaterra, Spain

Emilia Chiosi

Via Cuma-Licola 229, 80072 Pozzuoli (NA), Italy

Isabel Cobas Fernandez

Universidad de Santiago de Compostela, Laboratorio de Arqueologia, Edificio Monte da Condesa, USC - Campus Sur, 15706 Santiago de Compostela, Spain

Sergio Dipilato

Centro Diagnostico di Radiologia Medica, Ospedale Nuovo Regina Margherita, Via E. Morosini 30, Rome, Italy

AnneS.Dowd

Brown University, Department of Anthropology, P.O. Box 1921, 02912 Providence, Rhode Island, USA

Dragos Gheorghiu

Academy of Art, Str. G. Ral Bodisteanu 19, 70744 Bucharest, Romania

Sylvia Gili

Departament d'Antropologia Social i Prehist6ria, Edifici B, Universitat Aut6noma de Barcelona, 08193 Bellaterra, Spain

Silvia Guideri

Parco Archeologico Minerario di San Silvestro, 57021 Campiglia Marittima (LI), Italy

Nicola Laneri

Istituto Universitario Orientale di Napoli, Piazza San Domenico, 80134 Napoli, Italy

Luc Laporte

CNRS UMR 6566, Laboratoire d'Anthropologie, Universite de Rennes I, Campus Beaulieu, 35042 Rennes, France

Anne Le Fevre-Lehoerff

Ecole Fran~ise de Rome, Piazza Farnese 67, 00186 Rome, Italy

Francesca Lugli

Via Ouchi di Castro 1, 00194 Rome, Italy

Vicente Lull

Departament d'Antropologia Social i Prehist6ria, Edifici B, Universitat Aut6noma de Barcelona, 08193 Bellaterra, Spain

Anna Mastykova

The Rescue Archaeology Department, Institute of Archaeology, Russian Academy of Science, Dmitria Ulianova str. 19, 117036 Moscow, Russia

Kostalena Michelaki

University of Michigan, Ann Arbor, Department of Anthropology, 1109 Geddes, MI-48109-1079 Ann Arbor, USA

iii

Rafael Mic6

Departament d'Antropologia Social i Prehist6ria, Edifici B, Universitat Aut6noma de Barcelona, 08193 Bellaterra, Spain

Sarah Milliken

Dipartimento di Scienze Geologiche e Paleontologiche, Universita di Ferrara, Corso Ercole I d'Este 32, 44100 Ferrara, Italy

Kristin Prestvold

South-Trnndelag County Council, Fylkeskommunen, Postuttak, 7004 Trondheim, Norway

Pilar Prieto Martinez

Universidad de Santiago de Compostela, Laboratorio de Arqueologia, Edificio Monte da Condesa, USC - Campus Sur, 15706 Santiago de Compostela, Spain

Gabriella Prisco

Istituto Centrale peril Restauro, Piazza San Francesco 9, 00184 Rome, Italy

Judit Regenye

Laczk:6 Dezso Muzeum, Erzsebet setany 1, 8200 Veszprem, Hungary

Cristina Rihuete

Departament d'Antropologia Social i Prehist6ria, Edifici B, Universitat Aut6noma de Barcelona, 08193 Bellaterra, Spain

Roberto Risch

Departament d'Antropologia Social i Prehist6ria, Edifici B, Universitat Aut6noma de Barcelona, 08193 Bellaterra, Spain

~

Departament d'Antropologia Social i Prehist6ria, Edifici B, Universitat Aut6noma de Barcelona, 08193 Bellaterra, Spain

Encarna Sanahuja Yll

Katerina Skourtopoulou

Department of Archaeology, University of Cambridge, Downing Street, CB2 3DZ, Cambridge, United Kingdom

Massimo Vidale

Istituto Centrale per ii Restauro, Piazza San Francesco 9, 00184 Rome, Italy

IV

INTRODUCTION

This articles in this volume were presented at the Third Annual Meeting of the European Association of Archaeologists in Ravenna, Italy, in September 1997. The papers span from the Palaeolithic period to the present day, and address various aspects of material culture including pottery, stone tools, beads, metals and architecture. Although many of the papers are based on research carried out in Italy, which is not surprising considering the location of the conference, the volume also includes papers on research in Spain, France, Greece, Hungary, Romania, Russia and Norway, as well as Iran and North America.

Laporte's reconstruction of the technical and operational sequences involved in the manufacture of shell beads along the French Atlantic coast during the Late Neolithic is principally based on data from the site of Ponthezieres on Oleron Island. He reveals a standardized operational sequence leading to a very normative production which may have resulted from the symbolic charge of the shell beads or their role in a long-distance exchange network. The technical specialization of the people involved in this activity is revealed by the standardization of the gestures and the way they are linked together, and Laporte suggests that they may have been part-time specialists belonging to a rural community which, at least occasionally, received some compensation for this product, perhaps copper.

Despite this apparent chronological, material and geographic heterogeneity, a unifying thread is provided by the fact that the papers clearly reflect the widespread interest which has arisen in the last decade in the social agency of ancient material culture production. Because all technologies depend on social relations of production, technology can be defined as a materially grounded arena for dynamic social interaction involved in the planning, production, use, repair and discard of material culture (see Dobres & Hoffman 1994; Ingold 1988, 1990; Lemonnier 1986, 1989, 1990, 1993; van der Leeuw 1993). The papers in this volume make a valuable contribution to the development of the theoretical foundations of the study of ancient technology.

Regenye approaches the question of specialization in a settlement group of the Late Neolithic and Early Copper Age Lengyel culture in western Hungary by focusing on flint mining activities. She describes the Szentgal-Tuzkoveshegy flint mine supplied a ring of eight specialized settlements (a workshop zone) which were built for the exploitation and protection of the mine, and suggests that workshop production for the whole of a community operated at a regional level, while individual specialization is indicated by the graves of craftsmen. Gheorghiu presents a well illustrated study of the Eastern European incised ceramics from the Neolithic to the Chalcolithic which reveal the existence of complex and repetitive designs. His experiments permit a reconstruction of lost gestures, showing how the designs were applied by using templates and splines, and he suggests that the repetitive character of this 'industrial' tradition may imply a ritual behaviour.

Following a critical discussion of the various contrasting definitions of 'craft specialization' found in the archaeological literature, Milliken illustrates some examples of craft specialization in the Palaeolithic: the degree of standardization and labour investment in the manufacture of Aurignacian ivory beads and Solutrean foliate points, the presence of specialist flintknappers at the Perigordian quarry at Corbiac in the Perigord region and at the Magdalenian site of Etiolles in the Paris basin, and the manufacture of Middle Magdalenian spear throwers. She suggests that the appearance of craft specialization in the Palaeolithic should be seen as an adaptive measure in relation to specific contexts and socio-political conditions, such as seasonal aggregations, reduced mobility, and the need for alliance networks.

Dipilato and Laneri present the results of a series of tests carried out on Chalcolithic potsherds from the Iranian site of Cheshmeh Ali using x-ray mammographic analysis. The images thus obtained consent a gestural reconstruction of the building sequences of the vessels, which is of direct relevance in the question of the use of sequential slab construction techniques and the evolution of coiling and hand-building techniques.

Skourtopoulou' s examination of some Late and Final Neolithic assemblages from northern Greece takes an alternative view of the patterns of production and movement of chipped stone and questions the generally accepted hypothesis of the existence of intrasite specialized production and the development of local and regional exchange networks. She shows how the correlation between the raw material procurement evidence and the evidence available on the location of tool manufacture seems to point to the co-existence of different patterns of craft production.

Dowd attempts to define and measure the degrees of craft specialization among lithic tool makers by assessing variability in manufacturing technique, location and scale of production for ca. 4150 BP hunter-fisher-gatherers in the Hudson valley region, USA. She demonstrates that the degree of biface standardization was a function of workshop size, and that distinct stages of tool production are found to be spatially segregated at different sites. Certain stages of the biface reduction sequence appear to have been the result of part-time, short-term seasonal specialist producers.

V

In her article on ceramic production among the Maros villagers of Bronze Age Hungary, Michelaki focuses on the extent to which this production involved specialists. Using a combination of published data from Maros cemeteries and unpublished data from two Maros settlements, she argues that in the Early Bronze Age the scale and intensity of production was limited, while by 2000 BC and during the Late Bronze Age production was intensified and there is strong evidence for some degree of specialization with independent artisans producing in their own households for a broader community of consumers. This specialization was reflected by the relative degree of labour investment measured by the percentage of fine and burnished wares, the relative complexity of the shapes of the vessels, and the relative skill involved, but not by dimensional standardization.

Prestvold presents a study of iron production in Mid-Norway during the early Iron Age in the context of social relations, based on evidence from the production site at Fjergen. Fluctuation in early iron production is shown to have been conditioned by social stability and social change which is reflected in the nature of the burials and the presence of hillforts in the landscape, and peaksin production coincided with peaks in the struggle for power. Iron production therefore seems to have been based on a society with a cohesive social organization, and possibly with a monopoly of both the iron production process as well as the distribution of the product. Prisco and Vidale present an eloquent reconstruction of the manufacturing techniques of bronze statues based on the iconography of an Attic kylix. They suggest that in this and in other representations of craftspersons, the illustration of the technical events was not aimed at documenting the production process, but rather that it provided the occasion for re-asserting the essence of the relationships of production among apprentices, senior craftsmen and affluent citizens. They conclude with the cautionary note that attempts to reconstruct ancient techniques on the basis of similar iconographies should make a careful evaluation of the ideological and functional context within which the images were produced and exhibited.

Boccuccia bases his reconstruction of the pottery manufacturing techniques used in the Late Bronze Age and Early Iron Age in southern Italy on the results of a xeroradiographic study which focused on different production techniques used for the manufacture of figulina ware. He shows that wheel-made pottery did not develop because of its inner technical advantages, which improved production standards and favoured a hypothetical mass production, but rather as a function of the need for obtaining more regular and standardized forms using simpler operational sequences, and suggests that this might be related to the need for balancing and lowering the overall production cost of this class of fine ware in the framework of a set of processes of social interaction and confrontation.

Chiosi' s paper is concerned with the social organization of production at the late Republican workshop site of Rocca d'Evandro in west-central Italy, where amphorae, tiles and pantiles were made using slave labour. On the basis of her analysis of the production units, the kilns, she reveals the relative equality among the various people who worked there and of the objects they produced, while on the basis of the products themselves, and in particular the seals on the amphorae, she is able to reconstruct the juridical relationship between the dominus and the officinator and the vertical type of social organization which characterised the late Republican period.

Cobas Fernandez and Prieto Martinez set out to define the patterns of formal regularity and their continuities and changes in the ceramic material culture from the Bronze Age and Iron Age in north-west Iberia, in association with the social processes of the appropriation and semantization of space. They demonstrate a marked change in both the landscape and the ceramics, from a clear separation between domestic and non-domestic space and decorated and undecorated ceramics in the Bronze Age, to a non-existent separation of the landscape and decorated and undecorated ceramics in the Iron Age. These differences are thought to be related to a conscious intent, and differentiated cultural standards.

Using various examples from her work in the Crimea and the northern Caucasus, Mastykova illustrates the value of glass beads as a chronological indicator and as a source for reconstructing the history of material culture in ancient societies, the history of glassware technology and the history of economic and cultural contacts.

Le Fevre-Lehoerff argues in favour of the use of metallographic microstructural analyses for understanding bronze production in central Italy in the Early Iron Age, and in particular for determining the techniques used for making the decorative designs. The results of her analysis of two metal girdles of the "Tarquinia type" show that in both cases chiselling was used rather than engraving, although the techniques and tools used were different, and she questions whether the craftsmen at the beginning of the Iron Age in fact had the necessary tools and techniques to undertake engraving.

Cascone and Casini present a short account of ancient mining techniques and the organization of labour in the mountains of Campiglia Marittima, Tuscany, from the Etruscan period through to the 16th century. Based on research carried out in the same area, Guideri presents the results of an archaeological and experimental study of the operational sequences used for the production of coins in Mediaeval Tuscany, and highlights the relationship between this production and the socio-economic systems. She effectively demonstrates how the inspiration for the development of the multi-stage process for the reduction of

vi

polymetallic sulphides was the need to obtain silver, a commodity which had a strategic importance for defining power relationships.

Bibliography

Dobres, M.-A. & C. Hoffman 1994 Social agency and the dynamics of prehistoric technology. Journal of Archaeological Method and Theory I (3): 211-258. Ingold; T. 1988 Tools, minds and machines: an excursion in the philosophy of technology. Techniques et Culture 12: 151-176. 1990 Society, nature, and the concept of technology. Archaeological Review from Cambridge 9 (1): 517. Lemonnier, P. 1986 The study of material culture today: towards an anthropology of technical systems. Journal of Anthropological Archaeology 5: 147-186. 1989 Bark capes, arrowheads and concords: on social representations of technology. In I. Hodder (ed.) The Meaning of Things: Material Culture and Social Expression. London: Unwin Hyman, pp. 156-171. 1990 Topsy turvy techniques: remarks on the social representation of techniques. Archaeological Review from Cambridge 9 (1): 27-37. 1993 Technological Choices: Transformation in Material Cultures since the Neolithic. London: Routledge. van der Leeuw, S. 1993 Giving the potter a choice: conceptual aspects of pottery techniques. In P. Lemonnier (ed.) Technological Choices: Transformation in Material Cultures since the Neolithic. London: Routledge, pp. 238-288.

Lugli' s ethnoarchaeological study of the communities of charcoal burners in the Calabrian mountains provides a fascinating glimpse of social life in contemporary southern Italy. A characteristic feature of the charcoal burners camps is the bread oven, and by describing the operational sequences used in their construction and use, and the taphonomic processes which take place after their abandon, she shows how the technological choices in cycles of camp and bread oven building are a function of both the individual history of the families and of the wider-scale patterns of adaptation by the craft communities. Castro et al. introduce their Marxist inspired 'theory of production of social life'. The aim of this theory is to identify and explain the objective conditions on which the production of social life is based, and to determine whether the social relations established in and between societies exploit, hide or alienate the social subject. In the last paper, Vidale argues that the representation of operational sequences in the form of a linear progression of steps sometimes limits the understanding of the wider implications of technology, and instead we should focus on the various forms of deviation from this ideal linear path. He illustrates his five forms of deviation from linearity with various examples including the manufacture of 3rd millennium BC talc beads in the Indus valley, Bronze Age faience beads and sandstone bangles, 5th century BC bronze statues, contemporary pottery in Nepal and agate beads in India. By defining and exploring the various forms of deviation from linearity and their multiple implications, he suggests that we are able to avoid the reconstruction of ancient manufacturing sequences in terms of social norms (labour) and instead approach a dimension of work.

Sarah Milliken July 1998

vii

The Ghost of Childe and the Question of Craft Specialization in the Palaeolithic Sarah Milliken

INTRODUCTION

The theoretical foundations for an anthropological study of Palaeolithic technology have not been developed as far as the methodological ones. Typically, Palaeolithic technology is interpreted as the material means of making artifacts which was shaped by practical demands, rather than as a dynamic cultural phenomenon embedded in social action and social reproduction (Dobres & Hoffman 1994: 211). However, technological acts are a fundamental medium through which social relations, power structures, worldviews, and social production and reproduction are expressed and defined (cf. Ingold 1990; Lemonnier 1986; Pfaffenberger 1992), and the question of craft specialization is central to these issues. After a brief discussion of the theoretical background, I will illustrate some examples of craft specialization in the Palaeolithic.

'CRAFT SPECIALIZATION'

Generally speaking the term 'craft specialization' has been used to refer to occupational specialization associated with the development of stratified social organization in pre-state and early state-level societies (eg. Tosi 1984). More than any other person, it was Gordon Childe who fostered the current interest in craft specialization with his emphasis on the Urban Revolution and the place of specialized craft production in it. Childe's basic idea was that craft specialization occurred in history once humankind had mastered subsistence techniques and could produce a surplus, and therefore it arose as part of the Urban Revolution and was not present in preceding, self-sufficient societies (Childe 1936). Craft specialists were withdrawn from food production and supported from the common surplus of food. The key concepts are therefore: (1) differentiation of productive tasks within a community and (2) mutual dependency among "withdrawn" specialists sponsored by surplus. These notions are still a part of many current definitions of craft specialization (Clark 1995). Prejudice of praxis, or 'the ghost of Childe', therefore means that craft specialization is commonly conceived as a state of being that is achieved in the course of technological and social evolution and that consequently it was absent from prehistoric hunting and gathering societies (Kenoyer et al. 1991). This concept is however inappropriate, since in all societies the need for tools or objects made from specific raw materials requires the development of new technologies and individuals capable of producing such objects efficiently. However, not all humans have the same degree

of dexterity or skill and available raw materials are themselves quite variable. Consequently the ability to produce specific objects using specialized technologies is the result of a wide range of processes that are closely tied to the available raw materials, the current technologies, and the overall economic and social adaptations. From this perspective craft specialization should be defined as an adaptive process (Rice 1984: 46-47). As an adaptive process specialization in one form or another has been practised in all societies and is not limited to highly ranked or stratified societies. The ethnographic record in fact indicates that community-based specialization is an important component of many contemporary tribal societies, and this may also be the case in the archaeological record of non-state societies. Therefore for archaeologists working in non-state societies, identifying productive specialization in the material record may not be synonymous with identifying social complexity (see Cross 1990, 1993; Cobb 1993, 1996). Material correlates of craft specialization may represent a host of organizational structures and specialization often reflects community-based craft production in regional exchange systems rather than administratively controlled distribution systems (Stark 1991). Individuals who have progressively increased their competence on a limited range of produced goods, and have developed appropriate motorhabits for manufacturing specific objects, are in fact craft specialists, regardless of the chronological time frame or cultural context (Kenoyer et al. 1991). The question then does not centre on the presence or absence of craft specialists, but rather on the role of craft specialization in the overall social and economic structure of a particular society. While in some societies individual specialists may produce items for personal use or for the extended kingroup, in other societies specialists may produce items primarily for trade or as a service to other individuals or communities. Most would generally agree that craft specialization is the production of durable goods for the purpose of exchange; that is,.it involves the transfer of goods from the producer to someone outside his/her immediate household. Like most socio-cultural phenomena, craft specialization is not a discrete but rather a continuous variable: specialization is a continuum along which any economy may be gauged. At one end is the domestic mode of production, in which the division oflabour is limited to age and sex differences in the family. At the other end of the economy is the modem industrial economy in which the division of labour is

Sarah Milliken

enormously complex (Brumfiel & Earle 1987: 5). Many traditional societies are characterised by part-time specialists, and such a mode of production is not unknown in the ethnographic literature. For example, various forms of part-time specialization have beendocumented among such diverse groups as the Yir Yiront (Sharo 1934-35), the Wintu (Dubois 1935), the Kwakiutl (Boas 1921) and the Patwin (McKern 1922). At the lowest level of specialization, the production of some items is used to augment the basic economic activity. At the highest level, individuals work full-time at very specialized tasks. The degree of specialization depends on the amount of time devoted to the craft and to the quantity of production.

production. However, the presence of large numbers of unfinished beads allows a detailed reconstruction of the production sequence. The beads were created from pencil like rods of ivory. Since no mammoth bones have been recovered at the site, the large quantities of ivory used in bead manufacture seem to have been imported, either as tusk segments or in the form of the cylindrical rods found in the collections studied by White. Rough-outs for rods, seldom more than 10 centimetres long and between 0.45 and 1.40 centimetres in diameter, were split from desiccated laminae before being scraped and smoothed into long cylinders. The rods were then circumincised and snapped into cylindrical blanks from one to two centimetres in length and these were bilaterally thinned at one end to form a sort of stem. A perforation was then created at the junction of the stem and the unaltered end by gouging from each side. These rough-outs were then ground and polished into their final basket-shaped form using coarse and then fine abrasives with water. Experiments performed by White suggest that from one to three hours of labour per bead are required by this process.

In the archaeological literature the terms 'specialist' and 'specialization' are used to refer to a vast variety of concepts, people or things (see Evans 1978; Rice 1981; Tosi 1984; Arnold 1984, 1987; Brumfiel & Earle 1987; Clark & Parry 1990; Peregrine 1991; Shafer & Hester 1991; Costin 1991; Durden 1995; Clark 1995). In this paper the notion of specialization is used on a number of different levels:

•

•

•

The second example comes from the Aurignacian site of Sungir in Russia, where the intact burials of one man and two children were found. Each individual was lavishly decorated with thousands of ivory beads: the man was adorned with 2936 beads, the small boy with 4903 and the girl with 5274. The technology of the dominant form of bead production was systematic and redundant. The process began with an ivory baton elliptical in section which was circumincised and snapped into blanks. The blanks were scored across the width of each face creating a grooved blank, and bilateral drilling through this groove was then employed to perforate the piece. The final form of the bead was embedded in the production process from the very beginning. In other words the reduction sequence was designed with a particular standardised final product in mind, in this case a particular aesthetic effect when conventional modes of ornamental attachment were applied. Experiments by Semenov demonstrated that each of the small ivory beads at Sungir took about 45 minutes to fabricate, an estimate which White believes to be a severe underestimate, on the basis of his own experiments. Therefore the man's beadwork took more than 2000 hours, while that of each child took more than 3500.

firstly, if artifacts are found in a restricted set of contexts they may have had a restricted set of roles or a certain exclusivity. It may be the case that the contexts of use were different from those of final deposition, but their regular occurrence in a restricted set of contexts suggests that there may have been an awareness of the significance of certain artifacts even during a possibly mundane use-life. In this sense the artefacts may be called specialized; secondly, if the mode of production of such artifacts is very different from everyday production in that it is far more accomplished and relatively uncommon, then this technology may also be termed specialized; and thirdly, those people who have the ability to produce such artifacts using a technology that requires a level of skill above that required for mundane production activities may be termed specialists, as they are performing a task which only a limited number of people would have been able to carry out (cf Olausson 1992).

The following are a few examples of craft specialization in the European Upper Palaeolithic.

Both production sequences produced a remarkable degree of metric and morphological standardisation. This degree of standardisation and the amount of time invested in their production has important implications for craft specialization, quantitative production standards and the reproduction of complex systems of information.

AURIGNACIAN NORY BEADS

Detailed studies of the operational sequences used in the manufacture of Aurignacian ivory beads have been made by Randall White (1989, 1995), two of which I will summarise here. The first type is the so-called basket-shaped bead, such as those from the French Aurignacian site of Abri Blanchard. At this site the beads were not associated with a burial context, but rather were excavated from living surfaces. Unfortunately the collection, which was excavated fifty years ago, was never studied adequately before it was dispersed and thus it is difficult to judge the scale of

SOLUTREAN FOLIATE POINTS

Foliate points are among the most complex of Palaeolithic artifacts, exhibiting a diversity in production techniques,

2

The ghost of Chi/de and the question of craft specialization in the Palaeolithic

finishing retouch, shape and surface form, and raw materials. Fragments of foliate points usually comprise between 10 and 50% of all retouched tools in Solutrean assemblages, and occur alongside a range of other types or so-called substrate tools which were made on either flake or blade blanks, and which were more simply and sparsely retouched. There are four recognised types of foliate point which fall readily into two classes of production technology: those made on blades and finished primarily through pressure retouch (shouldered points and willow leaves), and those made on flakes and thinned with invasive percussion retouch (unifacial points and laurel leaves). Few exceptions to these rules occur, although some laurel leaves may have been finished with pressure retouch, and laurel leaves were very infrequently made on blades.

platform or possible ridges. The time taken to manufacture the leaf-shaped points depends on the size of the piece as measured by its long axis. According to Jaques Pelegrin, the smaller leaf-shaped points, which are between 4-5 centimetres in length, may take about 20 to 30 minutes to retouch. The largest pieces, such as those found at the site of Volgu which are up to 30 centimetres in length, may take up to eleven hours to retouch. By contrast, the more sparsely retouched 'substrate' tools can be manufactured in less than 5 minutes (Sinclair 1995). In the light of this great disparity in manufacturing time,

one might be tempted to think that they were not in fact utilitarian items at all. The effort expended in their manufacture, or their 'cost', does not make sense in the light of the likely return in hunting success, in other words their 'benefit', since they easily fracture when used as projectile points. However, with the exception of the very largest of the leaf points, the majority of them do show evidence of their practical use, either in the form of microscopic usewear or evidence of use damage.

Each class has its raw material prerequisites. The flake points required massive chert rocks to produce the appropriate flake blank. For laurel leaves the chert had to be durable enough to endure the repeated shocks of bifacial thinning, since the Solutrean technique consisted of striking down directly on the edge of the biface to remove the flakes that left the typically thin, wide, and flat invasive scars on the surfaces. Only a few chert types available in the Perigord region of south western France met these requirements, and among them Cenozoic chalcedonies and Senonian nodules stand out as the prime materials for laurel leaves. Blade points had different production requirements, and on the whole were made from different raw materials. Maastrichtian nodules were particularly sought for their large size, homogeneous texture, great resiliency and general flawlessness. Many brightly coloured and finetextured rare Senonian cherts and Liassic jaspers were used for manufacturing blade points, which suggests that aesthetically pleasing attributes were also important.

Foliate points circulated from production sites to their sites of discard. It was not necessarily completed foliate points that circulated, although the lack of attention paid to debitage in Solutrean assemblages means that it is only possible to make hypotheses. Laurel leaves would probably have been at least roughed out while still close to the quarry since they require a large blank to produce a relatively small tool, and it would have been impractical to carry blanks very far before the initial shaping. In addition, the massive blanks would tend to hide imperfections that might later prevent completion. Since breaks are most likely to occur as the biface attains its thinnest cross-section, the most sure strategy would have involved the laurel leaf being finished where production had begun. Laurel leaves are therefore likely to have travelled in their completed form. For blade points, on the other hand, the critical act of production comes when the blade is struck from the blank, and finishing could have taken place away from the quarry site without incurring risks.

Bifacial thinning retouch entails a sequence of applications of retouch blows that requires great order and precision. There are sequences of roughing out blows, and as the sequence develops each blow requires progressively greater precision in its application and the final blows are especially difficult. The overriding characteristics of this production process are care, order and thinking ahead as the risks of failure gradually increase (Sinclair 1995). Evidence that Solutrean points were heat-treated before the finishing retouch was applied also suggests a sophisticated theoretical and practical knowledge of heat and its effects on chert on the part of the Solutrean flintknappers.

Foliate point technology was therefore complex and sophisticated and might well be investigated in terms of craft specialization (Larick 1987). The mode of production of such artifacts is very different from everyday flintknapping in that it is far more accomplished and relatively uncommon, and thus this technology may be termed specialized. The production of these points may indeed have been limited to a small group of individuals who controlled access to the appropriate knowledge, skills and materials. Those people who have the ability to produce such artifacts using a technology that requires a level of skill above that required for mundane knapping activities may be termed specialists, as they are performing a task which only a limited number of people would have been able to carry out. Adopting the terminology introduced by Clark and Parry (1990) we can refer to the manufacture of Solutrean

When these bifacial implements have been replicated it is clear that their manufacture takes significant lengths of time. The leaf-shaped points require the most time, since the retouch not only covers the surface of the piece but is also the means by which the blank is successively thinned. As work progresses each retouch blow becomes successively more difficult, with the end result that the platform for each retouch blow needs careful preparation, often to the extent that pressure flaking will be required to remove any pseudo-

3

Sarah Milliken

points as ad hoc specialization, in other words the sporadic, informal production of goods for exchange. These goods circulated widely, but their production is considered to be ad hoc because the items appear to have been originally made for personal use. The higher labour expenditure involved in the production of foliate points may have "envalued" (sensu Pollock 1983 in Clark & Parry 1990) the object with a special social message indicative of privileged consumption.

highly specific, and according to Pigeot two connected aspects may have promoted a full integration between social organization and mode of production. Firstly, being scarce the raw material was not relinquished to disorganised or private exploitation. The better and more productive blocks were reserved and restricted to the specialists, while less experienced novices were training on useless or exhausted cores and the more competent knappers used cores which could potentially provide adequate blades. Secondly, being difficult to work the raw material could not be exploited by anyone. The difficulties inherent in the continuous production of series of blades up to 50 centimetres long from voluminous cores with variable constraints would undoubtedly require a vast sum of knowledge and practical know-how, which could only be acquired through a long period of technical initiation and apprenticeship.

PERIGORDIAN FLINTKNAPPERS AT CORBIAC

At the Upper Perigordian quarry site of Corbiac in the central Perigord, south western France, Gibson (1980, 1984) has suggested that the characteristics of the lithic assemblage indicate the presence of socially related, parttime flintknapping specialists. These characteristics are: the very high degree of skill utilized; the low incidence of errors; the use of a technology which minimizes the input of time, effort and raw material; the use of standardized techniques of manufacture; and the high degree of consistence in the size and shape of both the products and the waste material. At Corbiac the critical resource was the copious amounts of high-quality Bergerac flint. This flint occurs in long nodules which often exceed 50 centimetres in length and which contain few impurities. In fact the presence of artifacts made of Bergerac flint at Abri Pataud, which is located 25 kilometres from Corbiac, suggests that this raw material was valued by prehistoric flintknappers. The specialists at Corbiac may therefore have been making blades for the needs of a larger local population.

It therefore seems legitimate to envisage the presence of knapping specialists at Etiolles. Though they would probably not have been truly full time professionals, the more complex and elaborate flintworking cannot simply be accredited to the experience of an average adult. Were all 25 of the elaborate debitage clusters in habitation U5 to represent familial production, a more widespread range of technical practices would be expected. On the contrary, they are striking in their stylistic, technical and economic consistency. The hypothesis put forward by Pigeot is therefore that specialization and increased competence in certain technical activities, such as the production of sizeable blades for communal purposes, could well be a necessary and almost imperative response to the Etiolles society striving for rational management of those scarce and difficult to work resources.

MAGDALENIAN FLINTKNAPPERS AT ETIOLLES

MIDDLE MAGDALENIAN SPEAR THROWERS

At the Magdalenian site of Etiolles in the Paris Basin there was an artisanal production of lithic artifacts which was undoubtedly much larger than was necessary for local needs (Bodu et al. 1990; Roux 1990). The site consists of a number of workshops, and through the reconstructed sequence of lithic production Nicole Pigeot showed that there was a spatial order in the lithic products which might have been caused by a varying technical rank of flintknappers who were working according to precise rules (Pigeot 1986, 1987, 1990; Pigeot et al. 1991; Olive & Pigeot 1992). Three kinds of flintknappers were identified: best technicians, or specialists, whose superior technique resulted in the manufacture of large quantities of well-made blades; less talented technicians, perhaps occasional flintknappers, who were probably satisfied with simple debitage; and apprentice-debutantes, for whom the goal of flintknapping seemed to have been in the gesture itself.

Middle Magdalenian spear throwers such as those from the sites of Mas d'Azil and Bedeilhac in the French Pyrenees clearly indicate not just adherence to manufacturing guidelines or a design target, but also some kind of craft specialization on the part of one or more makers (Conkey 1985). Not only do they attest to a formalised system for their manufacture, but they also derive from regional aggregation sites that may have been the context in which ritual communication was efficacious. These sites also yield other types of artifact that most strikingly appear to be the products of a kind of craft specialization or standardization in concept and execution. In addition to the spear throwers, there are similarly sculpted and engraved horse heads, identically decorated horse incisors, decorated polishers and the famous incised spiral rods from Isturitz which exhibit decoration that is, as Saint-Perier wrote in 1920, "so particular that the series may well be the work of a single craftsman" (in Bahn 1982: 257).

Spatial analysis of flaking debris was used to identify "training areas", and these have been employed to infer a system of apprenticeship. The lithic sources at Etiolles are

4

The ghost of Chi/de and the question of craft specialization in the Palaeolithic The fact that these artifacts are found in a restricted set of contexts suggests that they may have had a restricted set of roles or a certain exclusivity. It may be the case that the contexts of use were different from those of final deposition, but their regular occurrence in a restricted set of contexts suggests that there may have been an awareness of the significance of certain artifacts even during a possibly mundane use-life. In this sense these artefacts may be called specialized.

BIBLIOGRAPHY Arnold, J. 1984 Economic specialization in prehistory: methods of documenting the rise of lithic craft specialization. In S. Vehik (ed.) Lithic Resource Procurement. Proceedings from the Second Conference on Prehistoric Chert Exploitation. Carbondale: Center for Archaeological Investigations, Southern Illinois University, Occasional Paper 4: 37-58. 1987 Craft Specialization in the Prehistoric Channel Islands, California. Berkeley: University of California Press.

________ A

2

g ..

.

.

.

.-610

c:=::::, 20

c::>

21

22

~

~

23

24

..:::::::::.25

29

~

26

30 0

1

2

3cm

Fig. 4: Szentgal-Fuzi-kut, lithic material from the 1986 excavation (after Biro 1993-94)

32

~ 27

31

Signs of specialization in a settlement group of the Lengyel culture

3

~

~

6

5

•,

7

•.

10

.c=::.12

0

1

2

13

3cm

Fig.5: Szentgal-Fiizi-kut, lithic material from the 1986 excavation (after Biro 1993-94)

33

Clay, Twigs, Threads and Ritual: Ceramic "Industrial" Decoration in the East European Neolithic and Chalcolithic Dragos Gheorghiu

Some traits of this tradition are also to be found in the Neolithic Hamangia culture (c.5000 BC), in the shape of large curves (Figs. la; lb) that delimit some incised patterns (see Hasotti 1997: fig.30). The method is also found in the Boian culture, where it was used to delimit incised or excised patterns.

INTRODUCTION

The potter's work is a complex activity that begins with the identification of the sources of material and finishes with the decoration and firing of the vases. Not all the stages of the process have been studied with the same intensity. Although the formation of ceramic vases has been studied by experiment (see Coles 1973) or by analytical techniques (see van der Leeuw 1997: 136), analysis of the decoration did not rise above the level of description.

All the curvilinear designs applied in the above-mentioned cultures consist of simple curves, parallel curves, complex "natural" curves resulting from the gradual bending of vegetal stems or twigs, generally labeled "Bezier curves" in this text, and double spirals, part of which were awkwardly drawn by free hand and part by remarkable techniques of plotting. In the Boian or Gumelnita cultures, this difference of quality is easily perceived even on a single vase, where a "perfect style" is continued by an "awkward style". Bezier curves, which are also known as Bezier-Casteljau curves, are "calculated mathematically to connect separate points in a smooth, free-form curves and surfaces" (Bezier Curve, Microsoft Encarta Encyclopedia) and are defined by at least three points. "The two end points of the curve are called anchor points. The other points which define the shape of the curve are called handles, tangent points or nodes. Attached to each handle are two control points. By moving the handles themselves, or the control points, you can modify the shape of the curve" (Ralph Martin, Bezier curve, PC Webopaedia, Definition and Links, 1994, p.l.)

The study of the Eastern European incised ceramics from the Neolithic to the Chalcolithic reveals the existence of some complex repetitive designs, made with a skilful control of the incised curved lines that are impossible to be made by free hand. The difficulties raised in plotting an accurate double spiral, parallel curves or a complex curve without the help of instruments are insurmountable. There is no scientific explanation for this assertion which, unfortunately, is specific for the empmc1sm of connossoirship. For a debate on this concept as an analytical tool, see Schmitt et al. (1988: 122-123). Due to the complexity of curves, one can assume that they are the result of an elaborate process of "industrial" design, in other words the result of an "industrial" process of plotting the decorative design with instruments that produce a repetitive pattern. The technology of the decoration of Neolithic and Chalcolithic ceramic vases as the result of a repetitive technological process with an "industrial" character has not been analysed until today. In this respect the existence of a technological tradition in the plotting of the decoration would presuppose the existence of a ritual tradition, in the form of a "cultural continuum" (see also Hodges 1989: 82) from the Neolithic to the Chalcolithic.

EXPERIMENT WITH CURVILINEAR DESIGN

Experimental archeology seems to be, at present, one of the major methods that can be utilized to identify the technology used to decorate ceramic vases in prehistory. With respect to the curvilinear design referred to above, experiments with ceramic designers and potters confirmed that all the above mentioned curves and double spirals can not be accurately plotted by free hand. Also, from the same experiments one can discover the difficulties of positioning a pattern of curves or spirals on a curved surface without the use of an instrument to partition the vase's surface. This process of division of a curved surface is a time consuming action that, to maintain a character of efficiency specific for a design action, needs specialized instruments, i.e., templates (Gheorghiu 1997).

From all the complex decorations of prehistoric ceramic vases, the present study deals with the family of incised curves characterizing a long span of time in Carpathian and Balkan prehistory. Thus, from the Neolithic to the Chalcolithic there seems to be a cultural tradition of incised complex curves that lasts from the Linear Pottery culture (c.5300-4900 BC), to the Boian culture (c.5000-4500 BC) and Precucuteni culture (c.4700-4300 BC), and ends in the Gumelnita culture (c.4700-4200 BC). This tradition of complex curvilinear patterns was continued in the painted ceramics of the Chalcolithic, in the Cucuteni culture (c.4500-3500 BC) (Gheorghiu 1997).

The use of simple templates for incision of simple patterns in the Neolithic and Chalcolithic has already been demonstrated by the archeological finds of small instruments made from notched bones (see Marinescu-Bilcu

35

Dragos Gheorghiu

1993: 217, figs. 26-28), and consequently the principle of repetitive action may have been applied as a design principle also to plot complex curves.

approach to the vases' decoration is a "general reading" of the entire pattern. This formal analysis can highlight the following:

If a template can be defined as "a pattern or a gauge, used as a guide" (The Oxford Encyclopedic English Dictionary) in plotting by incision a certain ornament, thus the simplest template for the decoration with curves of ceramic vases would be a vegetal spline ("a flexible wood or rubber strip used especially in drawing large curves", The Oxford Encyclopedic English Dictionary). What would be the advantages of using a template? From an economic perspective, the employment of such an instrument would correlate low effort with maximum efficiency in the replication of a pattern. The question is what kind of templates would have beeen used in the Neolithic and Chalcolithic ceramics of Eastern Europe? The materials at hand that could produce complex curves would probably be vegetal stems and vegetal threads. For example, some stems or willow twigs could generate Bezier curves which, from a fitosemiotic (Deely 1997) perspective, display a "natural" design. A Bezier curve, in computer graphics, is a curve that is calculated mathematically to connect separate points in smooth, free-form curves and surfaces of the type needed for illustration programs and CAD/CAM models. Bezier curves need only a few points to define a large number of shapes, hence their usefulness over mathematical methods for approximating a given shape ("Bezier Curves", Microsoft Encarta 98 Encyclopedia). Experiments with splines made of twigs produced analogous Bezier curves as those incised on ceramic vases or produced by a computer programme (Figs. 2a; 2b.)

1. the presence of Bezier curves which cannot be accurately drawn by freehand; 2. the simultaneous presence of Bezier curves drawn in an awkward manner and in a "perfect" style; 3. the awkwardly incised curves generally having a very small radius cannot be produced by bending a vegetal twig without breaking it (Fig. 4); 4. parallel curves whose incision is unrealizable by freehand (Fig. 5); 5. parallel curves which slowly diverge (Fig. 6); 6. "incomplete" spiralled patterns, cut off at the upper part and at the bottom by two parallel lines (Figs. 7; 8; 9), which indicates that the spiralled template was fixed on the vase by two horizontal strips. A second approach to the decoration of vases is "close reading" (see Marshack 1972), a method regarded as inoperative (see Elkins 1996: 185-226) at a certain enlargement of the detail which deals with the examination of the pattern details. This method is applicable to ceramic fragments because of the holographic character of Bezier curves and double spirals which permits a quick (mental) reconstruction of the whole image. Such an approach would reveal the following aspects, which support the argument for the use of splines for ceramic decoration:

1. the existence in all the cultures discussed of some interruptions in the incised curves, which cannot be explained in a freehand incision; • in the Precucuteni (Figs. IOa; IOb; IOc.) and Gumelnita (Figs. Ila; llb; 12.) cultures, the presence of an interruption in every anchor point of Bezier curves, the place where the spline was fixed with fingers on the surface of the vase. While in the Linear Pottery notenkopf phase these points were made by nail incision, in the Precucuteni and Gumelnita cultures they were replaced with simple interruptions of the incised lines. The method to fix the splines with fingers on the vase's surface reappeared later in the painted ceramics of the Cucuteni culture phase B (c.3900-3700 BC), where dots of color produced by the fingers' impressions were positioned on the anchor points of the curves. • in the Boian culture, the presence of interruptions between the straight lines and the curved ones in some large curves, due to the difficulty of connecting the curved lines of the spline to the straight lines (Figs. 13a; 13b); • in the Gumelnita culture the frequent location of these interruptions at the beginning of the segment of the curve with the smallest radius which cannot be reproduced with a vegetal spline (Figs. 14a; 14b);

Other experiments with twisted templates made of threads lead to images of spirals or double spirals. Textile threads soaked in clay with water or grain powder with water, a method used in Romanian ancient peasant society in the weaving process (Iordache 1985: 227), and positioned on the wall of the vase, can, by a simple twist, produce in a few minutes a rigid curved template that will permit the plotting of a perfect spiral. (Figs. 3a; 3b.) The fixing of a rigid spiral's beginning needs only one point if its contour is fixed with two parallel threads. A first inference would be that splines or textile templates had a double role: in a first stage to position the pattern on a certain surface (vase's wall or lid), and in a second stage to permit the plotting of this pattern on the freshly made clay object.

READING THE DECORATIVE TEXT

The evidence of the utilization of splines to plot Bezier curves can also be derived from the study of shards by means of the hermeneutic method, namely the interpretation of the whole from the detail and of the detail from the whole (Gadamer 1975: 258; see also Hodder 1992: 150). A first

2. in the Linear Pottery culture (Figs. 15; 16), the Boian and sometimes in the Gumelnita culture (Fig. 17), the presence

36

Ceramic "industrial" decoration in the East European Neolithic and Cha/eolithic

of an incised "starting point", in the shape of a simple line or a triangle, as a generating source for two divergent incised lines;

bit consisting of body motion and environmental effort" (Marvin, H. 1964: 42). From experiments made "when the clay is still plastic [and] design based on curved lines are possible" (Hodges 1989: 82), one can infer that the main actones of the "industrial" incised decoration with Bezier curves on lids and vases walls would be the following:

3. in the Boian culture, the presence of traces of plotting on both sides of the broken line mentioned above, the result of a change in the direction of plotting due to the angle of the broken vegetal stem (Figs. 18a; 18b);

•

4. in the Gumelnita culture, the presence of two directions of incision in the places of interruptions, which is unnatural for a freehand plotting but usual for the plotting with a spline, together with the incision made to hold the spline (Figs. 14a; 14b); •

5. in the Gumelnita culture, the presence at the interruptions of shifts of the incised lines, as if the plotting was done on the two sides of the spline (Figs. 19a; 19b);

• •

6. the presence in the Gumelnita culture of parallel spirals, the first one incised and perfectly plotted, and a second one overlapping the first spiral, painted by free-hand;

• •

7. in the Boian culture, the presence of some "broken lines" existing in the group of parallel curves, as if the vegetal stem or twig of the spline was curved too much and collapsed (Figs. 20a; 20b; 18a; 18b).

• • •

The analysis of the above pattern regularities led to the following preliminary conclusions concerning the usage of splines utilized to plot complex curves: •

•

•

the fastening of the template (a vegetal stem or twig, or a textile thread) with two fingers at its anchor points situated at the extremities (these locations are visible as interruptions of the plotted line and also as slight hollows on the vase's surface, because of the finger's pressure on the spline, an "accident" which cannot be corrected easily); the curving of the vegetal splines, in order to generate the "natural" Bezier curves; the twisting of the threads in order to generate spirals; the fastening of the Bezier curve or of the spiral at the main anchor points with fingers; a double fastening for simple or double spirals with fingers and threads; a first plotting with a pointed instrument from the marginal anchor point to the anchor point situated in the centre of the Bezier curve; a second plotting from the centre of the curve to the second marginal anchor point of the spline; the untying of the spiralled template; the reposition on other object of the spline or the spiraled template.

It should be noted that double spirals use the same methods of plotting their centre but have a different technique of fixing their periphery. Because double spirals need several points to be fixed on the vase's surface, the textile templates resulting from the twisting of a thread on the vase's surface were generally tied with two horizontal threads at their upper and lower extremities. If such a template was soaked in a mixture of water and flour, the resulting spiral would preserve its shape and a certain rigidity when repositioned on another vase, and would permit its fastening on the vase's wall.

in the Linear Pottery notenkopf (see Ursulescu 1991: 222, figs. 10d, 10:C 10g) phase, in the Boian (see Pandrea 1994: 23, fig. 6.4) and in the Precucuteni cultures, Bezier curves were fixed in two and three anchor points; in the Hamangia and Boian cultures splines were used to plot curves with a large radius to delimit impressed patterns; in the Boian and Precucuteni cultures vegetal stems were used to plot straight lines and curves.

The analysis of technological actones would reveal a change in the importance of the positioning of splines between the Neolithic and the Chalcolithic, as the methods of generating anchor points changed from the simple nail incision in the Linear Pottery culture to marked incised "decorative" points in its late phase and in the Boian culture, and then losing importance in the Gumelnita culture. From this remark one can presume that a temporal change occurred in the technological sequences of plotting.

By studying these "industrial" decoration methods with splines, one can notice that all the above mentioned cultures display an analogous technique in the utilization of splines to incise complex curves, which can be explained by their genetically cultural relationship (see Berciu 1937: 105; Marinescu-Bilcu 1993: 194; 1972; 1976).

THE TECHNOLOGY OF PLOTTING WITH SPLINES AND TEXTILE TEMPLATES THE "VEGETAL" STYLE OF SPLINES

In addition to the readings of the decorative text in the manner discussed above, the incised patterns can be fractioned in small visible units of (technological) significance, called actones and representing "a behavioural

As one can observe from experiments of plotting with

vegetal stems or twigs, their curvatures gradually produce

37

Dragos Gheorghiu

several types ofBezier curves up to a double spiral. It is well known that "decorative" patterns with complex curves were frequently produced in basketry in the process of coiling and "by differently colored sewing elements." The presence of analogous curves as in basketry incised on ceramic vases could be explained from the skeuomorphic character of ceramics which sometimes copies the objects of value. The "vegetal" curves incised on ceramic vases would be in this interpretation the result of a copying process of images expressing the "natural" properties of materials, i.e., the curvature of stems and twigs.

a template would produce a more standardized gesture and determine the potter to perform in series the same ritualistic action. An empirical reconstruction of the technological actones would reveal the following levels of decoding in the text of a meta-ritual:

For the use of spirals as "decorative" patterns, the argument seems easier, because the spiral is the most common structure in nature, from vegetals to the positioning of human organs (Cook 1979), and in prehistoric cultures in the process of preparing vegetal or natural threads or ropes or basketry. Thus, in analyzing the sources of the "vegetal" style one can infer that, because the curvings and twisting exist in nature as well as in culture, Bezier curves could be an index of the "vegetal", namely, of wild or domestic plants and the double spiral of the wild and of the process of transformation from wild into domestic.

• •

• • •

• •

the ritual of positioning of the splines or templates on the vase; the ritual of curving or twisting; the ritual of fixing the splines or templates and of visualizing these points with the nails or with instruments; the ritual of incision; the ritual of correction, polishing and of hiding the traces of the splines; the ritual of burning the vase in the kilns; the ritual of filling the incisions with color.

It can be inferred that, in spite the fuct that the emic significance of the ritual remains unknown to us, from an ethic perspective, the experiments with splines and textile templates when explored more fully could permit a reconstruction of lost gestures and, subsequently, of the invisible images of some of the ritual.

TEMPLATES AND RITUAL Note This paper represents a developed version of a paper delivered at the Third Annual Meeting of the European Association of Archaeologists in Ravenna, 1997.

Prehistoric rituals as well as cult are one of the invisible cultural aspects in the archeological record, because they were made of repetitive actions (Barrett 1988: 31) which may not leave a tangible trace. However, the repetitive character of the industrial tradition may also imply a ritual behaviour. Since an "industrial" technology to decorate ceramic vases by splines or textile templates can produce the same unaltered design on an unlimited number of vases, a cultural and encultural action at the same time that, by its characteristics, implies a ritualistic behaviour due to its repetitiveness.

Acknowledgments My gratitude goes to Professor Radu Florescu who guided my first research on the use of templates in Chalcolithic ceramics and to Dr Sarah Milliken for her patience and comments during the genesis of the present text. Many thanks to my colleague Ernest Budes, for the help given to certify some of my technological intuitions. The present text would not be understandable without the constant help of Linda Miller and Cornelia Catuna.

According to the anthropologist Gilbert Lewis (1991: 122) "ritual involves a performance but a special form of performance, that is acted out according to explicit and communally recognized rules of behaviour, which are artificial, and which therefore require teaching and learning. These rules provide clear guidance concerning the practical aspects of the performance: when and where to hold it, for example, and how and with whom to do it. A ritual thus becomes an arena for rigidly prescribed and highly formalized action."

REFERENCES

Barrett, J. 1988 The Living, the Dead and the Ancestors: Neolithic and Early Bronze Age Mortuary Practices. In J. Barrett, & I. Kinnes (eds). The Archeology of Context in the Neolithic and Bronze Age. Recent Trends. Sheffield: University of Sheffield, pp. 3041. Berciu, D. 1937 Primele consideratiuni asupra neoliticului din valea Dunarii inferioare in legatura cu descoperirile din jud. Vlasca. In Buletinul Muzeului Jud Vlasca. Bucharest.

In the present text I used the term "decoration" to symbolize

an ensemble of patterns applicable to the vase's surface, thus the significance of the term is purely conventional, and in fuct, defines a ritual text. The decoration is a rigid repetitive (c£ Gediga 1989: 51) action. Seen from this perspective, the prehistoric decorated vases by use of templates could be perceived not only as a mode of more efficient production but also as texts of the act of ritual. In this sense a spline or

38

Ceramic "industrial" decoration in the East European Neolithic and Cha/eolithic

Coles, J. 1973 Archaeology by Experiment. New York: Charles Scribner's Sons. Cook, T. A. 1979 The Curves of Life. New York: Dover Publications [1914]. Deely, J. 1997 Basic of Semiotics. Romanian translation, Bucharest: All. Elkins, J. 1996 On the imposibility of close reading. The case of Alexander Marshack. Current Anthropology 37 (2): 185-226. Hodder, I. 1992 Reading the Past. Current Approaches to Interpretation in Archaeology. Cambridge: Cambridge University Press. Hodges, H. 1989 Artifacts. An Introduction to Early Materials and Technology. London: Duckworth. Gadamer, H. G. 1975 Truth and Method. New York: Seabury Press. Gediga, B. 1989 Remarks on Studies of the Religion of the Communities of the Lusatian Culture. In M.L.S. Sorensen & R Thomas (eds). The Bronze Age Iron Age Transition in Europe. Oxford: British Archaeological Series International Series 483, pp. 430-439. Gheorghiu, D. 1997 The "Industrial" Decoration of Eneolithic Ceramics. Abstracts, Third Annual Meeting of the

van der Leeuw, S. E. 1997 Cognitive aspects of 'technique'. In C. Renfrew & E.B.W. Zubrow (eds) The Ancient Mind Elements of Cognitive Archaeology. Cambridge: Cambridge University Press, pp. 135-142. Lewis, G. 1991 Caves, Cult and Children in Neolithic Abruzzo, Central Italy. In P. Garwoord, R Skeates & J. Toms (eds.) Sacred and Profane, Proceedings of a

Conference on Archaeology, Ritual and Religion. Oxford, 1989. Oxford: Oxford University, Committee for Archaeology, Monograph No.32. Marinescu-Bilcu, S. 1972 A propos des influences de la culture Precucuteni sur la culture de Hamangia a la lumiere de quelques decouvertes inedites de Dobrogea. Dacia NS 16: 53-73. 1976 Relatiile dintre culturile Precucuteni si BoianGumelnita. SC/VA 37 (3): 347-354. 1993 Les Carpathes Orientales et la Moldavie. In Atlas du Neolithique Europeen. L'Europe Orientale. Liege: ERAUL, pp. 191-241. Marshack, A. 1972 The Roots of Civilization: The Cognitive Beginnings of Man's First Art, Symbol, and Notation. New York: McGraw Hill. Marvin, H. 1964 The Nature of Cultural Things. New York: Random House. Pandrea, S. 1994 lnceputul culturii Boian in campia Brailei. lstros VII: 7-26. Schmitt, M., E. Bakewell, W. Beeman & C. McMichael Reese 1988 Object, Image, Inquiry. Santa Monica: The Getty Art History Information Program. Ursulescu, N. 1991 La civilisation de la ceramique rubanee dans les regions orientales de la Roumanie. In V. Chirica & D. Monah (eds). Le Paleolithique et le Neolithique de la Roumanie en Contexte Europeen. Iasi: Bibliotheca Archaeologica Iassiensis, IV, pp. 188224.

European Association of Archaeologists, Ravenna, September 1997. Forli: A.B.A.C.O. s.r.l., p. 77. Hasotti, P. 1997 Epoca Neolitica in Dobrogea. Constanta: Muzeul de Istorie Nationala si Aheologie. Iordache, G. 1985 Ocupatii Traditionale pe Teritoriul Romaniei. Craiova: Scrisul Romanesc.

39

Dragos Gheorghiu

Fig. la. Hamangia culture, Baspunar Valley, Litnanu, Dobroudja.

Fig. lb. Detail ofFig. la.

40

Ceramic "industrial" decoration in the East European Neolithic and Cha/eolithic

Fig 2a. Stages of twisting of a Bezier curve.

41

Dragos Gheorghiu

Fig. 2b. Stages of twisting ofa Bezier curve.

42

Ceramic "industrial" decoration in the East European Neolithic and Cha/eolithic

Fig. 3a. The repositioning on a vase of a textile template soaked in water with flour.

Fig. 3b. The fastening of the textile template and the plotting of the double spiral.

43

Dragos Gheorghiu

Fig. 4. Gumelnita culture, Baspunar island, Limanu.

Fig. 5. Boian culture, Ciorogarla, Bucharest.

44

Ceramic "industrial" decoration in the East European Neolithic and Cha/eolithic

Fig. 6. Boian culture, Ciorogarla, Bucharest.

45

Dragos Gheorghiu

. lture with Precucuteni influence, wherea fig. 7. Hamang1a cu

46

bo ts unknown Dobrouja. u '

Ceramic "industrial" decoration in the East European Neolithic and Cha/eolithic

Fig. 8. Precucuteni culture, Tarpesti (after Marinescu-Bilcu, 1974: 245, Fig.8).

47

Dragos Gheorghiu

Fig. 9. Gumelnita culture, Vladiceasca (after Dumitrescu 1985: 30).

48

Ceramic "industrial" decoration in the East European Neolithic and Cha/eolithic

Fig. 1Oa.Precucuteni culture, whereabouts unknown, Moldova.

Fig. 1Ob.Detail of fig. 1Oa.

49

Dragos Gheorghiu

Fig. 1Oc.Detail of fig. 1Oa.

Fig. I la. Gumelnita culture, Baspunar island, Limanu.

50

Ceramic "industrial" decoration in the East European Neolithic and Cha/eolithic

Fig. 11b. Detail of fig. 11a.

Fig. 12. Gumelnita culture, Baspunar island, Limanu.

51

Dragos G"heorghiu

Fig. 13a. Boian culture, Ciorogarla, Bucharest.

Fig. 13b. Detail of fig. 13a.

52

Ceramic "industrial" decoration in the East European Neolithic and Cha/eolithic

Fig. 14a. Gumelnita culture, Baspunar island, Limanu.

Fig. 14b. Detail of fig. 14a.

53

Dragos Gheorghiu

Fig. 15. Linear Pottery and Boian culture, Ciorogarla, Bucuresti.

Fig. 16. Linear Pottery culture, detail of fig. 15.

54

Ceramic "industrial" decoration in the East European Neolithic and Cha/eolithic

Fig. 17. Gumelnita culture, Baspunar island, Limanu.

Fig. 18a. Boian culture, Ciorogarla, Bucharest.

55

Dragos Glleorghiu

Fig. 18b.Detail of fig. 18a.

Fig. 19a. Gumelnita culture, Baspunar island, Limanu.

56

Ceramic "industrial" decoration in the East European Neolithic and Cha/eolithic

Fig. 19b.Detail of fig. 19a.

Fig. 20a. Boian culture, Ciorogarla, Bucharest.

57

Dragos Gheorghiu

20b. Detail of20a.

58

"Sequential Slab Construction" and other Problems concerning Hand-Building Techniques in Chalcolithic Iran: Experimenting with Mammographic X-Ray Images Sergio Dipilato & Nicola Laneri

INTRODUCTION

specialization and improvement. During the Vlth millennium BC, the agricultural revolution which followed trends already established in the Mesolithic, dramatically changed the economy of the sedentary communities, from hunting and gathering to farming, across a large area of South Asia from the Mediterranean basin to the Indus Valley. A transformation in the general socio-economic conditions is visible during the Vth and the lVth millennia BC, with a surprising increase in the size of the settlements (through the lVth millennium BC small towns such as Susa, Malyan and Uruk became large urban sites). In most cases, these towns rose in the middle of river basins, engaged in long-distance trade, and developed specialized craft industries (pottery manufacture, flint and semi-precious stone industries, metalworking) (Goren & Gopher 1995; Moore 1995) .

Medical mammography is a very sensitive radiographic technique which allows, under specific conditions, a very fine enhancement of the inner features of ceramic potsherds such as organic and inorganic inclusions, voids and other phase discontinuities, because it gives one the possibility of using a higher time of exposure (from 20 to 500 mAs) and a lower peak kilovoltage in to obtain brighter radiophotographic images (Carr 1990; Carr & Riddick 1990). In this article we present a series of tests carried out on early Chalcolithic potsherds from the Iranian site of Cheshmeh Ali (late Vth-early lVth millennium BC). The images thus obtained consent a gestural reconstruction of the building sequences of the vessels, which is of direct relevance in the question of the use of sequential slab construction techniques, and the evolution of coiling and hand-building techniques in the earliest ceramics of the Iranian plateau.

SEQUENTIAL SLAB CONSTRUCTION THE MOUND OF CHESHMEH ALI

With the help of various traditional methods of scientific analysis (mainly Scanning Electron Microscopy, X-Ray radiography and Xero-radiography), some scholars in the past have studied large collections of vessels and potsherds coming from early Iranian sites. These studies came to the conclusion that a peculiar hand-building technique, the socalled "Sequential Slab Construction" technique (see Fig. 1) was used for a long time before the development of the first coiling techniques (Vandiver 1986, 1988).

The mound ofCheshmeh Ali (''The Spring of Ali", from the name of the Prophet's son-in-law) lies not far from the ancient site of Rhages (Ray), 10 kilometres south of Teheran. The site was initially excavated by J.P. De Morgan during his exploration of Susa; the famous specialist in Iranian archaeology Erich Schmidt, sponsored by the Boston Museum of Fine Arts, then excavated the site between 1934 and 1936 (Schmidt 1935; Matney 1995). During the excavations two major prehistoric phases were discovered by Schmidt's team: Cheshmeh Ali I A (ca. 5500-5000 BC, roughly contemporary with the so-called Halaf cultural period in Mesopotamia), and Cheshmeh Ali I B (ca. 50003500 BC, comparable to the Obeid period in the Mesopotamian region). These excavations were useful for an early reconstruction of the archaeological sequence of prehistoric occupations across the Iranian plateau (McCown 1957; Vanden Berghen 1959; see Matney 1995). The potsherds we analyzed, consisting of about 200 fragments, are related to these two major phases. The pottery was sampled during a visit to the old trenches in the early seventies, and is presently kept in IsIAO deposits in Rome. Stylistic comparisons suggest that the majority of the sample belongs to the second phase of the occupation (Cheshmeh Ali I B).

When compared with coil-building, Sequential Slab Construction, which is rarely observed in modem ethnographic contexts (and therefore even more difficult to reconstruct), seems to have been used by ancient potters for making various types of vessels. Small clay balls or lumps were pre-formed and then flattened to form large and irregular slabs. According to an evolutionary interpretation, coil-building would have appeared in the archaeological record of South Asia only in a second and later phase (end of the Vth-beginning oflVth millennium BC). The base of the vessel, which is the first piece of the vessel made by the potter, is usually constructed by flattening and soldering two or three different slabs one on top of the other. Then, other smaller slabs were added in the form of overlying bands, first along the base and then upwards, thus forming the whole body of the vessel. In a later phase a turntable would have been used to help in the finishing processes (outer shaving and decoration) carried out after a stage of partial or complete drying (Gibson & Woods 1990).

Chesmeh Ali was an ancient agricultural village, probably affected by incipient social stratification, and its craft industries are expected to reflect processes of craft

59

Sergio Dipilato & Nicola Laneri

With the help of Xero-radiography, which enhances even slight phase discontinuities by means of a toner-like effect integrated with a careful scrutiny of fracture surfaces, the interface among the superimposed slabs can be detected, and the whole process reconstructed, step by step. We began our research with the purpose of observing in detail the inner structure of the potsherds of the Cheshmeh Ali collection, and testing the hypothesis of manufacturing techniques based upon the techniques of sequential slab construction.

given the specific texture and composition of the clay of the fragments in our sample. The XR images were obtained with 500 mAs and 23\28 Kv. These settings combined a really high time of exposure (which gave us the opportunity of increasing the contrast of the particles within the inner structure), with a very high contrast film (3M Mammo Trimax HM2), and provided us with bright and easily readable XR plates. A selection of XR images was drafted, evidencing inclusions, voids and possible constructive interfaces, following the graphic conventions used by the authors in previous articles. The graphics interpreting the XR images are monitored to the reconstruction of the manufacturing sequence.

THE SAMPLE The majority of the potsherds we studied are walls of hemispherical bowls of various sizes, and medium-sized jars. In the case of the larger bowls, the fracture surfaces of the sherds usually show a rather coarse ware, with chaff impressions and vegetal inclusions. Some of the smaller bowls are a medium to fine ware, with a homogeneous texture and lesser amounts of rocky and vegetal inclusions. Through visual analysis of the fractures, it is also possible to identify discontinuities which probably correspond with the interfaces between different lumps of clay. The depressions on the inner and outer surfaces are the fingerprints left by the potters (or by other persons involved in some stages of the manufacture of the vessels) while joining the different slabs during the construction of the vessel. In most cases the rim appears to have been added in a later moment, using elongated slabs ultimately resembling coil preforms. On the other hand, the high degree of finishing of the surface does not allow a direct observation of the assembled slabs, nor a simple assessment of the manufacture techniques.

The XR images seem to confirm the hypothesis of sequential slab construction for the ceramics of Cheshmeh Ali. The paleotechnological indicators of this peculiar hand-building technique are the following: 1. Radio-opaque macro-areas (probably corresponding to the contour of the original slabs of clay) with a variable orientation inside the wall of the vessel. Inside a large number of specimens, the contours of lumps or slabs of variable size are visible; the interface lines between the lumps may not always be visible, but they are clearly recognizable in some XR images of the larger-sized specimens. 2. Interior voids running along the interfaces between incorrectly joined slabs. 3. Random orientation and distribution of chaff inclusions and other vegetal particles (visible as voids) inside the radioopaque macro-areas representing the possible slabs.

The outer and inner surfaces, which are often covered by a thick (1-2 millimetre) layer of red slip, show irregular, shallow grooves with an oblique orientation which were left by spatulas or other tools used by the potter to thin the wall of the vessel by shaving. The walls are often extensively burnished outside and inside with a bone or a wooden stick. Such burnishing is a recurrent feature of the ceramic industries of the Neolithic and Early Chalcolithic periods in Iran.

4. Rounded radio-translucent areas, with the outline of vegetal and straw particles characterized by a concentric structure, along some interface lines, corresponding to the fingerprints left by the potter while pressing and joining the slabs onto one another.

RECONSTRUCTING THE PRODUCTION SEQUENCE INTERPRETATION BY MEANS OF MEDICAL MAMMOGRAPHY (X-RAY) ANALYSIS

Our paleotechnological reconstruction of the production techniques of the Cheshmeh Ali ceramics is based on the visual and radiographic analyses, as well as on the knowledge of manufacturing techniques as observed in other archaeological and ethnographic contexts. The following examples will show various hypotheses of paleotechnological reconstructions. Each Figure shows a picture of the potsherd, the X-ray image, a graphic interpretation of the X-ray image, and a conjectural reconstruction of the original assembling/joining pattern of the slabs, obtained by expanding the interfaces visible in the X-ray plates; the section of the potsherd, whenever visible,

All the fragments from both phases (Cheshmeh Ali IA and IB) were investigated by medical X-ray radiography and by medical Mammography. We also tested the application of Computerized Axial Tomography on some sherds. The methodological background and the relevant indicators have been discussed at length in previous publications (Rye 1981; Rice 1987; Carr 1990; Carr & Riddick 1990; Vandiver et al. 1987; Marano et al. 1992; Vidale & Tosi 1996). At the end of our experiments, we decided to continue our study using a Medical Mammography device (Metaltronica Compact Mammo HF\A, with a focal spot of 0.lx0.3 centimetres),

60

"Sequential slab construction" and other problems concerning hand-building techniques in Cha/eolithic Iran

shows possible joining interfaces recognized on the fracture surfaces.

E) Fragment of a bowl, fine ware (Fig. 6). Mouth diameter: 28cm.

A) Fragment of a small bowl, fine ware (Fig. 2). Mouth diameter: 15 cm.

Decorated with a thick layer of red slip and with a waveshaped painted design. The XR mammographic image and the graphic reconstruction show vegetal and straw inclusions with a sub-horizontal orientation. Some irregular interfaces are visible along the body of the bowl due to joint lines between slabs. These slabs would have been larger than in the previous example. The rim was probably made with a sequence of smaller lumps of clay.

Decorated with a red slip and with metope-like patterns alternating with horizontal simple tracts and vertical curvilinear motifs. The XR plate provides us with a clear image of the inner structure of the vessel. Starting from the rim, one can observe two radio-translucent bands with a subhorizontal shape, each band having a width of 2-3 centimetres. The rim was made using several small lumps of clay applied one onto the other to form a kind of continuous horizontal band. In this upper band, the orientation of the particles and voids is particularly regular, and the vegetal and straw inclusions follow a parallel and doubtless oblique orientation. The second band from the rim is characterized by straw and vegetal inclusions and voids oriented in a subhorizontal fashion. In the lowermost preserved part of the fragments other finger impressions are visible ( darker circular areas) and some lines of discontinuity ( or interfaces) between sub-circular slabs (translucent areas). The bowl was probably constructed in two different phases: the body, by using and joining subcircular slabs; and the rim, for a maximum height of about 3-5 centimetres, formed by applying preformed coil-shaped slabs, joined in two larger bands. The oblique orientation of the voids and included particles within the upper part of the rim suggests that the vessel was partly worked on a revolving table or on an early type of potter's wheel.

C) Fragment of a large bowl or basin, coarse ware (Fig. 4). Mouth diameter: 38 cm.

CONCLUSIONS

Decorated with a thick red slip and with horizontal spirals alternating with vertical spirals. The XR mammographic image shows four radio-translucent areas along the rim. These areas are not clearly defined by interfaces, but are characterized by vegetal and straw inclusions with a random orientation (running in sub-horizontal and/or vertical and/or oblique directions) and chaotic concentrations, clues to a slab-building technique. The slabs should have been small lumps of clay, perhaps with a diameter of 3-4 centimetres, joined with noticeable care and regularity. B) Fragment ofa bowl, medium-coarse ware (Fig. 3). Mouth diameter: 30 cm.

XR mammographic analysis is an effective technique for enhancing important indicators of the production process of these ancient ceramics, in spite of their variable texture. In the early agricultural village of Cheshmeh Ali, between the end of Vlth and the middle of IVth millennia BC, vessels appear to have been made using preformed lumps, flattened into slabs joined together to form the wall of the vessels. In some cases, after having been applied these slabs retained a sub-circular or egg-shaped contour. In other cases the slabs were flattened and joined forming short sub-horizontal bands, which resemble the structure ofhorizontal coils. This is the case, for example, of the rims of the vessels in Figs. 5 and 6.

Decorated with a thick red slip and with vertical and oblique painted bands. A hole along the section of the fragments was made in ancient times to "restore" and reuse the broken vessel. The XR image shows four overlying sub-horizontal oriented bands and one fingerprint (along the interface of these four bands) with a cluster of concentric vegetal and straw inclusions. In this case the slabs had probably been preformed in regular lengthened clay strips, and were later joined by the potters in the form of irregular, sub-horizontal bands. The rim shows a similar pattern, but with smallersized inner elements. D) Fragment of a large bowl or basin, medium-coarse ware (Fig. 5). Mouth diameter: 40 cm ca.

One is tempted to view these vessels as evidence for a transition from a pure "Sequential Slab Construction" technique to a coil-building technique. The forming of the mouth of the vessel would have been an obvious occasion for applying elongated strips of clay ·instead of flattened rounded sheets. Moreover, at least in one case (Fig. 6) we come across evidence for the use of a turntable or (more probably) of a potter's wheel used in forming, thinning, and decorating the vessel. If we have to judge on the basis of the inner structures commonly observed in later wheel-thrown ceramics, in order to orient the inner inclusions and voids as

Decorated with a thick red slip and with a wide, horizontalrunning band of broken lines. The XR mammographic image clearly shows a translucent egg-shaped area overlying another two translucent areas and some darker areas corresponding to ancient finger-impressions, with inner clusters of concentric oriented vegetal inclusions. The eggshaped slab, which would have had a diameter of 3-4 centimetres, was probably added in a second moment, in order to mend a defective joint between two slabs. In this case too the rim seems to have been made by applying smaller, elongated slabs.

61

Sergio Dipilato & Nicola Laneri

observed in Fig. 6, the rim of the bowl must have been rotating at a relatively high speed.

Marano, P., A. Meduri, T. Pirronti, D. Calicchio, F. Rispoli &M. Vidale 1992 Xeroradiografia e Radiografia Digitale a Luminescenza: Risultati Preliminari di Indagini su due Campioni di Ceramica Pre-protostorica Orientale. Atti de/ Convegno dell 'Associazione Nazionale Prove Non Distruttive, Viterbo, pp. 393414. Matney, T. 1995 Re-excavating Cheshmeh Ali. Expedition 37/2: 2632. McCown,D.E. 1957 The Comparative Stratigraphy of Early Iran. Chicago University Press. Moore, A.M.T. 1995 The Inception of Potting in Western Asia and Its Impact on Economy and Society. In W.K. Barnett & J.W. Hoopes (eds.) The Emergeru:e of Pottery:

Thus, if our interpretations at this preliminary stage of the research are correct, the ceramics collected on the surfuce of the Iranian mound would give us a different perception of Sequential Slab Construction: this technique, instead of appearing as a static tradition, unchanged for millennia, might have represented an earlier evolutionary stage of coiling, and would have allowed the experimentation of efficient rotating devices. The development of the research will confirm or disprove this new view. Acknowledgments Our thanks are due: to M. Vidale for his precious and

essential collaboration in this study; to M. Tosi for allowing us to study the Cheshmeh Ali collections; to Francesca Fazzini and Timothy Matney for their friendly help.

Technology and Innovation in Ancient Societies.

Washington D.C., pp. 39-54. Rice, P.M. 1987 Pottery Analysis: A Sourcebook. Chicago. Rye, O.S. 1981 Pottery Technology. Manuals on Archaeology 4, Taraxacum, Washington. Schmidt, E. R 1935 The Persian Expedition. University Museum Bullettin 5 (5): 41-49. Vanden Berghen, L. 1959 Archeologie de I 'Iran Ancien. Leiden. Vandiver, P.B. 1987 Sequential Slab Construction: A Conservative Southwest Asiatic Ceramic Tradition ca. 7000-3000 BC. Paleorient 13 (2): 9-37. Vandiver, P., W.A. Ellingson & T.K. Robinson 1991 New Applications of X-rays Imaging Technologies for Archaeological Ceramics. Archeomaterials 5: 185-207. Vidale, M. & M. Tosi 1996 The Development of Wheel Throwing at Shahr-i Sokhta. Slow and Fast Revolutions Towards Statehood. East and West 46, 3-4: 251- 269.

REFERENCES

Carr, C. 1990 Advances in Ceramic Radiography and Analysis: Applications and Potentials. Journal of Archaeological Scieru:e 17: 13-34. Carr, C. & E.B. Riddick 1990 Advances in Ceramic Radiography and Analysis: Laboratory Methods. Journal of Archaeological Scieru:e 17: 35-66. Courty, M.A. & V. Roux 1995 Identification of Wheel Throwing on the basis of Ceramic Surfuce Features and Microfubrics. Journal of Archaeological Science 22: 17-50. Gibson, A. & A. Woods 1990 Prehistoric Pottery for the Archaeologist. Leicester. Goren, Y. & A. Gopher 1995 The Beginnings of Pottery Production in the Southern Levant: A Model. In P. Vincenzini (ed.) The Ceramics Cultural Heritage. Faenza, pp. 2128.

62

"Sequential slab construction" and other problems concerning hand-building techniques in Cha/eolithic Iran

without kinetic rotational energy

dlscontlnuos pressures on a unique clay mass

with lclnetlc rotational energy

discontlnuos pressures on an assembled clay mass

continuos pressure on a unique clay mass

transf11nnatlonl and aCdltlue

sequences

pressure on a assembled clay mass

transformational

SEMIFINISHED BESSEL

without lclnetic rotational energy

sequences

with kinetic rotational energy

reductlue sequences

SEMIFINISHED BESSEL without kinetic rotational energy

with kinetic rotational energy

Fig. 1: Flow chart showing a classification of hand-building and wheel-throwing pottery manufacture, based upon the distinction between techniques involving or not kinetic rotational energy, and between discontinuous versus continuous pressure movements (M. Vidale, modified after Courty & Roux 1995)

63

Sergio Dipilato & Nicola Laneri

1 2

10

11

12

13

"

~ l

I

/

4 1' I

/

:

.:_ \I I I

I \ \'

\

'

5 '· I

I

/

I/

).

I

I

l -1,

I.... ,, J I .,._.,.

• • ...

.'

I

,I

Fig. 2: Cheshmeh Ali, fragment ofa bowl: 1. photograph, 2. XR mammographic image, 3. graphic interpretation of the XR mammograpic image, 4. archaeological drawing, 5. conjectural graphic reconstruction of the production technique

64

''Sequential slab construction" and other problems concerning hand-building techniques in Cha/eolithic Iran

1

I

I

4i

I

I

I

I

5i

I I

I

Fig. 3: Cheshmeh Ali, fragment of a bowl: 1. photograph, 2. XR mammographic image, 3. graphic interpretation of the XR mammograpic image, 4. archaeological drawing, 5. conjectural graphic reconstruction of the production technique

65

Sergio Dipilato & Nicola Laneri

1

10

11

12

13

1,

4

5

Fig. 4: Cheshmeh Ali, fragment of a bowl or basin: 1. photograph, 2. XR mammographic image, 3. graphic interpretation of the XRmammograpic image, 4. archaeological drawing, 5. conjectural graphic reconstruction of the production technique

66

"Sequential slab construction" and other problems concerning hand-building techniques in Cha/eolithic Iran

1

2

10

11

12

_n

3

~

\\J

I

:'

I'

\'

I '_

'I

)

,, ,, I

:

'\ \

'

I I

'

\

I

41

/

\ ·--~~---( ~

I

I

.-

5

Fig. 5: Cheshmeh Ali, fragment of a bowl or basin: I. photograph, 2. XR mammographic image, 3. graphic interpretation of the XR mammograpic image, 4. archaeological drawing, 5. conjectural graphic reconstruction of the production technique

67

Sergio Dipilato & Nicola Laneri

1

2

4

)~----===--:;::::...__-::::----.....r

-

_

5

\

..-.---;:.--·,

...·

),

I

/~

___

.,.,

' •.

·--------~---··

Fig. 6: Cheshmeh Ali, fragment ofa bowl: 1. photograph, 2. XRmam.mographic image, 3. graphic interpretation of the XR mammograpic image, 4. archaeological drawing, 5. conjectural graphic reconstruction of the production technique

68

Biface Standardization Accompanying Organized Chert Quarrying Efforts: An Argument for Intensifying Lithic Production AnneS. Dowd

58-59). For quite awhile, it has been assumed that these broad-bladed points heralded a migration of newcomers into the region. Among their other accomplishments, they engaged in some kind or degree of lithic craft specialization (Cross 1990, 1993). New research on the River phase (ca. 4150-3289 BP) of the Piedmont (or Narrow Point) tradition challenges simple diffusion or migration explanations for culture change, by addressing the complex origins of intensifying lithic procurement and production (Figure 1).

INTRODUCTION

The question of whether standardized manufacturing techniques indicate craft specialization, and if so under what circumstances, begs consideration by lithic specialists. Possible examples of lithic standardization have recently been explored using biface reduction sequences. The results challenge researchers to examine their underlying assumptions about degrees of variability among production technologies. Flintknappers may have chosen among a variety of technological options based on seasonal shifts in social organization and settlement. For example, if craft specialization originated as a part-time seasonal activity, production solutions would be visible archaeologically that contrasted with efforts by nonspecialist members of the same culture.

Data on lithic procurement patterning suggests that larger, more organized quarrying efforts took place when populations expanded, making organized labor teams more feasible. This implies that part-time, short-term, occupational specialists may have been equally pragmatic, timing their efforts to periods of seasonal sedentism and increased economic cooperation. Late stage River phase bifaces have a distinctive character, not unlike the Meadowood blades. Perhaps variable degrees of biface standardization indicated differing intensities of tool production linked with ritual, for example, rather than subsistence.

To define and measure degrees of craft specialization among lithic tool makers, variability in manufacturing technique, location and scale of production was assessed for ca. 4150 BP North American hunter-fisher-gatherers. Assuming that tools and techniques, as well as the tools to make tools, can all be standardized (and frequently are, in situations where craft specialists control production), analysis of reduction sequences (or chaines operatoires) helps to identify the distribution and extent of lithic production events.

Unlike the Susquehanna Broadspear-using people, River phase groups using Normanskill spearpoints produced more limited quantities ofbifaces. A cache reported from the Bent site on the Mohawk River contained eight blades (Ritchie & Funk 1973: 56, 58). Also unlike the Susquehanna, Normanskill-type projectile points do not exhibit tremendous formal variation (except as a result of repeated resharpening or lateral recycling). Like the Susquehanna, specialized tools to make tools existed. Also like the Susquehanna, small but well organized task groups systematically mined chert quarries and produced bifaces at workshops nearby.

Several key trends have been identified so far: 1) the degree ofbiface standardization was a :functionof workshop size; 2) quarry events were organized for small task groups during periods of aggregation; 3) distinct stages of tool production were spatially segregated at different sites; and, 4) certain stages appear to have been the result of specialist producers. Regional scale spatial distributions of more standardized bifaces correlate with situations of seasonal aggregation (base camps) rather than dispersal (upland camps) within the hunter-fisher-gatherer populations studied.

DATA

By three thousand years ago, caches ofbiface blades became more common in the Eastern Woodlands ofNorth America. Caches, such as the North Hadley cache from Massachusetts which contained 319 Meadowood phase blades, have been dated to ca. 3250-2450 BP or 1300-500 BC (Gregg in Krech 1994: 145, 160). Contemporaneous caches with as many as 1500 bifaces have been reported (Ritchie 1969 [1965]: 185). Earlier, ca. 3800-2600 BP or 1850-650 BC, groups also cached Susquehanna tradition preforms for future use. Forty-two were found in Riverside, Rhode Island .. Susquehanna Broadspear points were widely distributed along the coast and major river drainages (Dincauze 1972:

Ten assemblages from a variety of site types were inspected to evaluate the possibility that at ca. 4150-3289 BP or 22001339 BC flintknappers produced relatively standardized bifaces, and that these bifaces showed some variation that corresponded with seasonal changes in settlement size. The site types sampled included three rockshelters, two inland camps, two workshops, two fishing stations on the banks of the Hudson, and one central base camp. Together these represent upland hunting camps, flintknapping sites and base camps in between the uplands and the lowlands; as well as lowland fishing and habitation zones. Biface

69

AnneS. Dowd

(Cesarski 1996: 62). One notable example was between a Stage 4 and a Stage 5 biface, as only one notch forming the tang had been removed. Five late-stage bifaces were measured from the Black Birch site, which was in a similar setting downstream on the west bank of the Hudson.

measurements were obtained to provide data useful for answering questions about the degree to which bifaces (n=51) were standardized. If bifaces were standardized, whether or not standardization peaked at certain site types was addressed, for example, at base camps inhabited by larger groups who aggregated during the summer and fall.

Rockshelters Eight bifaces were excavated at the Moonshine Rockshelter, a temporary hunting camp (Weinman & Weinman 1969). Three were whole examples. Bifaces from two other rockshelters were also measured. These included the Boehm Rockshelter (n=3) and the Rimmer Rockshelter (n=l) (Weinman & Weinman 1970). Whether biface production became more idiosyncratic at short-term occupations like rock shelters was considered by comparing them with larger, longer term settlements.

Standardized bifaces may be identified by consistent size and shape. Biface dimensions and ranks for shape conformity were collected from different site types and compared to determine relative standardization. The degree of standardization, or lack thereof: was estimated using measures of central tendency. The examples under discussion are Stage 4 bifaces, which are elongated and carefully thinned, with a square base and a tapered point. They have undergone secondary thinning using the percussion method only. Stage 4 bifaces lack only the notching and pressure flaking along the edges. By contrast, following Callahan's (1979) biface reduction categories, Stage 5 is equivalent to a finished projectile point. Stage 3 has undergone primary biface thinning, with flakes taken off both sides of the piece and a more tapered shape. Stage 2 has undergone rough-shaping or initial-edging on one or both sides of the implement and has a lanceolate shape. Stage 1 is a core or blank, amorphous or tabular, either the form in which the chert was quarried or an exhausted remnant.

ANALYSIS According to Trubitt (1996: 43), "craft specialization can be contrasted with nonspecialized craft production by the amount of time spent on craft working, the technological expertise required for production, and the production of craft goods for exchange." In a hypothetical archaeological situation where specialists as well as nonspecialists were producing stone tools, differences in time (quantity of artifacts produced), expertise (quality of artifacts produced) and distribution (location of artifacts produced) should be observed.

Workshops At the Breezy Knoll lithic workshop, six Stage 4 bifaces or preforms were identified (Dowd 1993, 1995 [1992], 1997a, 1997b, 1998). At this site, spent Stage 5's, and broken stage 4, 3 and 2 bifaces predominated. Exhausted Stage 1 cores were also present. The site's assemblage also included tools to make other tools, such as hammerstones and abraders. Abraders are distinctive for the oblique grooves on either side of the stone. Battering and extensive wear suggests that these tools were curated and reused as part of the flintknapper's tool kit. A similar artifact in the Susquehanna assemblage is the beveled-cobble abrader, "a type of multipurpose tool exhibiting battering, pecking and polishing on its edges and surfaces" (Cross 1990: 121). Pickle Hill II (n=l) appears also to have functioned as a workshop (Weinman & Weinman 1977).

Neither large quantities of artifacts, nor their far-flung distribution, indicated that surplus production of goods characterized the River phase lithic economy. Nevertheless, one variable, artifact quality, was evaluated from the standpoint of greater or lesser degrees of production standardization. Greater standardization of late-stage bifaces is taken as an indication that technological expertise was relatively high. Intersite variability among diagnostic latestage bifaces, used as Normanskill projectile point preforms, is expected to show different levels of standardization based on the degree to which flintknappers coordinated and organized production. The data needed to answer questions about the degree of standardized production include measurements on whole bifaces (Table 1). Mean and standard deviations were calculated for site size rank (1-10), biface length (mm), biface width (mm), biface thickness (mm), biface weight (g), biface shape rank (1-10).

Inland Camps Other biface examples included those from the A-B-C site, a central base camp covering about an acre and a half Seventeen Stage 4 bifaces or preforms were measured of the large number present in the assemblage. The majority of these were from a single feature, 4(6-A), possibly a burial pit, although no bones were present. The Fred Young site (n=3) and Pickle Hill (n=l) were smaller inland camps (Weinman & Weinman 1967, 1968).

Cross (1990: 68) has suggested that "the lower the width/thickness ratios for bifaces, the greater the probability that form is constrained by production techniques and/or raw material rather than a high degree of control over fracture properties." In other words, lower width/thickness ratios may indicate intentional standardization of the biface form. To investigate this possibility, mean width/thickness

River Camps At the Becker site, a fishing station on the banks of the Hudson, six Stage 4 bifaces or preforms have been measured

70

Biface standardization accompanying organized chert quarrying efforts

ratios were compared by site size. Mean biface width/thickness ratios were compared by two size categories, large (50% of the sites [n=5]; 69% of the artifacts [n=35]) or small (50% of the sites [n=5]; 31% of the artifacts [n=16]).

Certain stages but not others appear to have been the result of a degree of specialization among producers. Stage 4 bifaces appear to have been produced by specialists, resulting in standard size and shape. Future research will be directed toward evaluating the Normanskill spearpoints themselves to determine relative degrees of standard or idiosyncratic production.

The result was a mean biface width/thickness ratio of3.02:1 and a standard deviation of 0.58 for the larger site rankings (>1) and a mean of3.26:1 and a standard deviation of0.73 for the smaller site rankings (> 1). Higher mean widththickness ratios are evident at the smaller sites, and the standard deviations are larger. Lower mean width-thickness ratios are evident at the larger sites, and the standard deviations are smaller. When the standard deviations of the width/thickness ratios for the bifaces at each site were compared, the observed biface width/thickness indices confirmed the expectation that in comparison with the smaller sites, the larger sites had thinner and less variable biface width/thickness ratios.

CONCLUSIONS

Because available evidence points to different degrees of social organizational complexity from one drainage basin (for example, the Delaware) to another (for example, the Hudson), and one season (winter/spring) to another (summer/fall), biface standardization may not have been uniformly practiced across the area where people used Normanskill points. A greater degree of standardized production appears to have been correlated with the small, but well-organized task groups who mined large quarries in the Hudson drainage during the summer or fall. These ca. 4150 BP River phase economic patterns may have been tied to larger settlements and concomitant increases in ceremonialism or ritual, contemporary with or prefiguring the developments of the Susquehanna, starting around ca. 3800 BP, and the Meadowood, beginning at ca. 3250 BP. Variability in social organization within, as well as among cultures, is important for understanding sociocultural change. Efforts of full-time specialists engaged in mass production are only one of the many kinds of crafting behaviors worthy of archaeological investigation.

RESULTS

Preliminary results suggest that the degree of biface standardization was a function of workshop size. Analysis of a limited sample shows a correlation between site size and biface standardization, with the greater degrees of standardized production occurring at larger sites, or at workshops, where one or more task groups worked to produced late-stage bifaces exclusively. More data are required to augment the sample and to evaluate these themes in greater detail. Analysis of regional chert procurement patterns indicate that larger quarries may have been mined by small but well organized teams during periods of population aggregation (Dowd 1997b). This appears to be part of a larger pattern of organized forays from a base camp to resource extraction and processing areas for the purpose of formal biface manufacture (Table 2). Logistical planning helped move bulk resources to consumers (contrasting with consumers moving themselves to resources) (Stevens 1995). Other complementary patterns existed, for example expedient (amorphous core) technology for flake tool production was documented at the Hollister Lake Site (used during the Sylvan Lake phase, which predates and is coeval with the River phase) (Cobb & Webb 1994: 216).

Acknowledgements My research has been supported in part by Brown University's 1997 Graduate School Fellowship, a 1996 Manning Fellowship, and a 1996 Dr. Frances E. Price Harnish Fellowship, as well as a Geochron Laboratories (a division of Krueger Enterprises) 1996 Research Award for radiocarbon age determinations.

BIBLIOGRAPHY

Beardsley, R K., P. Holder, A. D. Krieger, B. J. Meggars, J. B. Rinaldo & P. Kutsche 1956 Functional and evolutionary implications of community patterning. Seminars in archaeology: 1955. Society for American Archaeology, Memoir 11: 129-155. Binford, L.R 1980 Willow Smoke and Dog's Tails: Hunter-Gatherer Settlement Systems and Archaeological Site Formation. American Antiquity 45: 1-17. Callahan, E. 1979 The Basics of Biface Knapping in the Eastern Fluted Point Tradition: A Manual for Flintknappers and Lithic Analysts. Archaeology of Eastern North America 7: 1-180.

Distinct stages of tool production were spatially segregated at different sites. For example, Stage 1 cores were produced at the quarry location for transport to a separate workshop area where a group could work more comfortably (on a level area rather than a steep slope). Stage 2-4 rough shaping, initial edging and bifacial thinning occurred in workshops relatively close to major quarries; unless the chert production was opportunistic, for example, expediently quarrying chert from the rockshelter walls of an upland hunting camp. Stage 5 were produced at habitation or processing sites, such as at the Becker site.

71

AnneS. Dowd

Cesarski, E. 1996 Archaeology Survey of the Town of Schaghticoke, Rensselaer County, New York. On file at the Department of Anthropology, State University of New York at Albany. Cleland, C.E. 1976 The Focal-Diffuse Model: An Evolutionary Perspective on Prehistoric Cultural Adaptation of the Eastern United States. Midcontinental Journal of Archaeology I: 59-76. Cobb, C. & P. Webb 1994 A Source Area Perspective on Expedient and Core Technologies. North American Archaeologist 15 (3): 197-219. Cross, J. 1990 Craft Specialization in a Nonstratified Society: A Late Archaic Example from the Northeast. Unpublished Ph.D. Dissertation. Department of Anthropology, University of Massachusetts, Amherst. 1993 Craft Specialization in Nonstratified Societies. Research in Economic Anthropology 14: 61-84. Dincauze, D.F. 1972 The Atlantic Phase: A Late Archaic Culture in Massachusetts. Man in the Northeast 4: 40-61. Dowd,A.S. 1993 The Normanskill Reduction Sequence from the Breezy Knoll Site. Poster presented at the Society for American Archaeology, St. Louis. On file at the Department of Anthropology, Brown University, Providence, Rhode Island. 1995(1992] Site 199-9-1 [Breezy Knoll Site]. Iroquois Gas Transmission System Phase III Archaeological Data Recovery Report, Volume III: The Hudson Valley Region. Garrow & Associates, Inc., Atlanta. 1997a Understanding Hunter-Fisher-Gatherer Complexity via Chert Resource Acquisition and Allocation. Paper presented at the Annual Meeting of the Society for American Archaeology, Nashville, Tennessee. 1997b Lithic Procurement and Social Complexity in New York's Hudson River Valley. Unpublished Ph.D. Dissertation. Department of Anthropology, Brown University, Providence, Rhode Island. 1998 Operationalizing an Anthropology of Technology: Lithic Procurement and Tool Production among North American Hunter-Gatherers. Proceedings of

the XIII Congress of the International Union of Prehistoric and Protohistoric Sciences. Forli, Italy. Funk, RE. 1976 Recent Contributions to Hudson Valley Prehistory. New York Museum and Science Service Memoir 22. Albany, New York. Krech III, S. 1994 Passionate Hobby. Studies in Anthropology and Material Culture, Volume VI. Haffenreffer Museum of Anthropology, Bristol, Rhode Island. Pagoulatos, P. 1992 Native American Land Use Patterns in New Jersey. Journal of Middle Atlantic Archaeology 8: 57-78. Ritchie, W.A. 1958 An Introduction to Hudson Prehistory. New York State Museum and Science Service, Bulletin Number 367. Albany, New York. 1969(1965] The Archaeology of New York State. Natural History Press, Garden City. Ritchie, W.A. & RE. Funk 1973 Aboriginal Settlement Patterns in the Northeast. New York State Museum and Science Service Memoir 20. Albany, New York. Stevens, J.S. 1995 A Comparison of Technological and Adaptive Strategies between Normanskill Occupations in the Delaware and Hudson Valleys. North American Archaeologist 16 (3): 239-279. Trubitt, M.B.D. 1996 Household Status, Marine Shell Bead Production, and the Development of Cahokia in the Mississippian Period Unpublished Ph.D. Dissertation. Department of Anthropology, Northwestern University, Evanston, Illinois. Weinman, P.L. & T.P. Weinman 1967 The Pickle Hill Site. New York State Archaeological Association Bulletin 39: 18-22. 1968 The Fred Young Site - A River Phase Component. New York State Archaeological Association Bulletin 43: 1-6. 1969 The Moonshine Rockshelter. New York State Archaeological Association Bulletin 46: 11-15. 1970 The John Rimmer Rockshelter. New York State Archaeological Association Bulletin 48: 19-23. 1977 The Pickle Hill II Site. New York State Archaeological Association Bulletin 70: 35-36.

72

Biface standardization accompa,rying organized chert quarrying efforts

Table l: Rive Phase late stage biface dimensions

Site

Type

Rank 1*

Lemrth

Width

Thick

Wff Ratio

Weight

Rank2**

A-B-C

base camp

6.07

57.5

26.9

8.1

3.32:1

18.0

4

A-B-C

base camp

6.07

57.5

27.9

9.2

3.03:1

22.0

6

A-B-C

base camp

6.07

84.2

29.5

9.2

3.21:1

30.0

10

A-B-C

base camp

6.07

71.9

21.9

6.8

3.22:1

13.0

8

A-B-C

base camp

6.07

62.4

20.0

7.9

2.53:1

12.0

7

A-B-C

base camp

6.07

60.1

19.6

5.7

3.44:1

11.0

A-B-C

base camp

6.07

40.8

23.9

8.2

2.91:1

10.0

3 4

A-B-C

base camp

6.07

46.1

24.4

7.5

3.25:1

8.0

4

A-B-C

base camp

6.07

74.4

19.9

8.0

2.49:1

11.0

6

A-B-C

base camp

6.07

69.6

33.6

8.5

3.95:1

29.0

9

A-B-C

base camp

6.07

50.3

24.6

6.7

3.67:1

11.0

9

A-B-C

base camp

6.07

58.6

34.0

4.10:1

29.0

3

A-B-C

base camp

6.07

NA

27.8

8.3 9.8

2.84:1

18.0

7

A-B-C

base camp

6.07

NA

26.9

3.16:1

11.0

8

A-B-C

base camp

6.07

68.5

35.8

2.91:1

38.0

6

A-B-C

base camp

6.07

49.8

29.1

4

A-B-C

basecamo

6.07

74.7

23.4

8.5 12.3 8.0 8.0

0

12.1

4.9

6.07

61.8

stand-dev mean

3.64:1

15.0

2.93:1

14.0

8

1.4

0.45:1

8.8

2

26.4

8.3

3.21:1

17.6

6

2.82:1

NA

4

Becker

river camp

1.80

NA

24.0

Becker

river camp

1.80

42.3

27.6

8.5 11.9

2.32:1

NA

8

Becker

river camp

1.80

40.6

26.8

7.9

3.39:1

NA

6

Becker

river camp

1.80

49.2

20.9

8.3

2.52:1

NA

9

Becker

river camp

1.80

39.2

24.4

7.5

3.25:1

5.0

6

Becker

rivercamo

1.80

44.3

20.2

6.6

3.06:1

6.0

7

0

3.9

3.0

1.8

0.42:1

0.7

2

1.80

43.1

24.0

8.4

2.89:1

5.5

7

stand-dev mean Black Birch

river camp

6.00

80.5

30.1

7.6

3.96:1

29.0

6

Black Birch

river camp

6.00

50.5

19.6

8.4

river camp

6.00

NA

26.9

9.1

6.0 12.0

7

Black Birch

2.33:1 2.96:1

Black Birch

river camp

6.00

NA

25.1

7.8

3.22:1

12.0

3

Black Birch

river camp

6.00

NA

31.6

10.5

3.01:1

21.0

5

stand-dev mean

8

0

21.2

4.7

1.2

0.59:1

9.0

2

6.00

65.5

26.7

8.7

3.10:1

16.0

6

Boehm

rock shelter

0.10

59.2

32.5

7.4

4.39:1

20.0

5

Boehm

rockshelter

0.10

87.5

34.6

7.0

4.94:1

31.0

Boehm

rockshelter

0.10

59.1

33.4

8.3

4.02:1

25.0

3 2

stan-dev mean

0

16.4

1.1

0.7

0.46:1

5.5

2

0.10

68.6

33.5

7.6

4.45:1

25.3

3

Breezy Knoll

workshop

5.40

45.9

25.1

9.3

2.70:1

21.0

6

Breezy Knoll

workshop

5.40

43.1

15.6

12.8

1.22:1

33.0

5

Breezy Knoll

workshop

87.2

33.0

12.0

2.75:1

22.4

Breezy Knoll

workshop

5.40 5.40

80.3

27.9

8.4

3.32:1

32.8

8 10

Breezy Knoll

workshop

5.40

76.8

33.7

12.1

2.79:1

32.2

8

Breezv Knoll

workshop

5.40

79.6

32.1

9.3

3.45:1

15.1

9

stan-dev mean

0

19.2

6.9

1.9

0.79:1

7.6

2

5.40

68.8

27.9

10.7

2.71:1

26.1

8 3

Fred Young

inland camp

0.31

60.0

24.9

10.1

2.47:1

12.0

Fred Young

inland camp

0.31

40.0

23.6

8.4

2.81:1

8.0

1

Fred Young

inland camp

0.31

41.9

22.0

8.8

2.50:1

9.0

2

stan-dev mean Himmer

rockshelter

0

11.0

1.5

0.9

0.19:1

2.1

1

0.31

47.3

23.5

9.1

2.59:1

9.7

2

0.04

42.1

17.7

6.1

2.90:1

2.5

4

73

AnneS. Dowd

stan-dev mean Moonshine Moonshine Moonshine Moonshine Moonshine Moonshine Moonshine Moonshine

rockshelter rockshelter rockshelter rockshelter rockshelter rockshelter rockshelter rockshelter

stan-dev mean Pickle Hill

inland camp

stan-dev mean Pickle Hill 2

workshop

stan-dev mean

0

0

0

0

0

0

0

0.04

42.1

17.7

6.1

2.90:1

2.5

4

0.18 0.18 0.18 0.18 0.18 0.18 0.18 0.18

52.6 73.6 64.6 NA NA NA NA NA

21.8 29.4 29.2 20.9 21.1 22.9 23.4 29.3

5.7 8.7 8.5 9.1 6.4

12.0 19.0 24.0 10.0 5.0 10.0 7.0 21.0

9 8 6 3 4 4

7.2 9.7

3.82:1 3.38:1 3.44:1 2.30:1 3.30:1 2.60:1 3.25:1 3.02:1

8.8

4 5

0

10.5

3.9

1.4

0.49:1

6.9

2

0.18

63.6

25.8

8.0

3.14:1

13.5

5

2.02

53.6

29.5

6.8

4.34:1

10.0

2

0

0

0

0

0

0

0

2.02

53.6

29.5

6.8

4.34:1

10.0

2

0.50

NA

36.3

12.2

2.98:1

30.0

7

0

NA

0

0

0

0

0

0.50

NA

36.3

12.2

2.98:1

30.0

7

Bettinger 1991

Pagoulatos 1992

* site size ranking (1-10) ** biface shape ranking (1-10) measurements in milliemetres, weights in grams

Table 2: River Phase site size, type and function Site

Type

A-B-C Black Birch Breezy Knoll Pickle Hill Becker Pickle Hill 2 Fred Young Moonshine Boehm Himmer

base camp river camp workshop inland camp river camp workshop inland camp rockshelter rockshelter rockshelter

Rank 2*

6.07 6.00 5.40 2.02 1.80 0.50 0.31 0.18 0.10 0.04

Beardsleyey al. 1956 tethered tethered tethered tethered tethered tethered free wandering free wandering free wandering free wandering

Cleland 1976 focal focal focal focal focal focal diffuse diffuse diffuse diffuse

Funk 1976 inland open camp lowlyingsite on Hudson inland open camp backcountryopen camp lowlyingsite on Hudson backcountryopen camp inland open camp backcountryrockshelter backcountryrockshelter backcountryrockshelter

* site size ranking (1-10)

74

Binford 1980 collector collector collector collector collector collector forager forager forager forager

processor processor processor processor processor processor traveller traveller traveller traveller

base I base II target I base II base II target I target II target II target II target II

Corrected C14 Years B.C. w

w

8

§

u. I..audez Farrell A-B-C Five Mile Dam Banana Belt (36Pl36) Farrell Bent

~

~..,

-.., 0

Cl

-...J VI

]0 ..., ~ Q..

~r

I

Jil ~

River Bend £live Mile Dam Kuhr No. 2

Fortin No. 1 A-B-C Five Mile Dam Five Mile Dam

A-8-C Miller Field (28WA15) FomnNo. l Peters-Albrecht (361'122) A-B-C Five Mile Dam ForunNo. l A-D-C Brodhead-Heller (36PI 17) Pickle Hill Beut River Bend Brodhead-Heller (36PI 17) Wickers Creek

N

N

c5

§

u. ,.....

I

-8 §-

u.

V. ,.....

0

I I I I I I

I I

0:,

~

~

I

~

I I I I I I I

"' is;::i

§~

I I

N' ~

o· ;::i

I

g

I

0

~s::i

I I

I I I I I

~

~-

I

0

~ §

I

~-

I

s::i...

I

I

(",

[

I I I

~

I I I

I I I I

~

~

~~

~

I I I I I

a

I I I I I

I I I

Ceramic Production Among the Maros Villagers of Bronze Age Hungary Kostalena Michelaki

islands of dry land in the middle of a flood prone, marshy environment.

INTRODUCTION

Maros ceramics have received a lot of attention from the initial appearance of the Maros group (also known as the Perjamos Culture, or the Szoreg Group) in the archaeological literature of Bronze Age Hungary. In the tradition of the times, they were initially used to define an archeological culture (Banner 1931; Childe 1929) and later on to devise its internal chronology (Bona 1975; Foltiny 1941). However, little is known about the economic and social role of Maros ceramics, as it is manifested through evidence for their production, use and distribution. My goal here is to focus specifically on understanding the organization and technology of ceramic production. Did it remain constant throughout the Maros Phases? Was it specialized and, if so, to what extent?

Test excavations in Popin Paor in Serbia and Kiszombor, Klarafalva, and 6szentivan in Hungary (Fig. 1) allow us a glimpse into the layout of tell and open settlements (Banner 1928, 1929; Foltiny 1942; Giric 1987; Horvath 1985; O'Shea n.d.). In both settlement types the houses are small and rectangular, with prepared clay floors, wattle and daub walls and roofs of matted or woven reeds. Some preserve evidence that has been interpreted as prepared interior ovens and hearths, interior room dividers, and subterranean storage pits. Single nuclear, or small extended families are suggested as the occupants of these dwellings. Besides houses, large bell-shaped storage pits, exterior ovens, kilns and prepared clay living surfaces have also been recognized (O'Shea 1996).

Archaeologists make inferences about the organization of pottery production using either direct evidence (such as the presence of kilns, wasters, wheels), or indirect information coming from the vessels themselves (such as the technology of their production; whether the products are standardized or diverse). In this paper I address the issue of production indirectly, by examining both published information from Maros cemeteries and unpublished ceramic materials from two Maros settlements, Kiszombor and Klarafalva. In particular, I focus on the extent to which Maros pottery production involved specialists. I argu~ that in the Early Phase the scale and intensity of production was limited, but the data on specialization are insufficient. By 2000 BC and during the Late Phase, even though the scale remained relatively limited, production was intensified and there is strong evidence for some degree of specialization.

The people who lived in these settlements exploited a diverse array of both domestic and wild plants and animals (Bokonyi 1972). A wide array of ceramic vessels, loom weights and spindle whorls, bronze and copper ornaments, daggers and axes as well as crucibles, tuyeres and metallic slag testifies that they were also involved in the manufacture of ceramics and textiles, as well as copper and bronze goods. No study to date has focused on the organization of ceramic production.

PRODUCTION AND SPECIALIZATION

Unlike many studies of ceramic production and specialization that have focused on administratively complex societies (Brumfiel & Earle 1987; Clark & Parry 1990; Costin & Hagstrom 1995; Costin 1991; Feinman et al. 1984; Peacock 1981; Rice 1981; Sinopoli 1988; Van der Leeuw & Pritchard 1984) studies of less complex societies, like those of the East European Early Bronze Age, are not as common (see though Day et al. 1997; Kaiser 1984; Kiriatzi et al. 1997; Mills & Crown 1995; Stark 1991; Underhill 1991; Whitelaw et al. 1997).

Since published data from other Maros sites are limited, and the analysis is in a preliminary stage, this paper cannot provide final conclusions about the organization of Maros ceramic production. It provides preliminary findings along with questions and hypotheses to guide future research, and joins the recent literature that tries to understand changes in the organization of production in the absence of centralized administrative control.

Here I shall rely on Cathy Costin's definition of four important parameters according to which the organization of production can be described: context, concentration, constitution of the production unit, and intensity (Costin & Hagstrom 1995: 620; Costin 1991: 8-18). Even though she refers to specialized production, these parameters are useful when thinking about the organization of production in general.

THE MAROS: GENERAL BACKGROUND

The Maros villagers lived in the region of the Tisza-Maros angle, in southeastern Hungary, northern Yugoslavia and western Romania (Fig. 1), between 2500 and 1650 cal. BC (O'Shea 1991), during the Early and Middle Bronze Age. In the southern Great Hungarian Plain they occupied low

77

Kostalena Michelaki

1) The context of production refers to the nature of demand for craft goods. A distinction is made between specialists who are attached to elite sponsors and patrons, at one extreme, and those who are independent, at the other. Attached specialists are expected to produce goods that encode social and political information that serves to reenforce social distinctions and maintain political power. Independent specialists are expected to produce utilitarian goods that serve household needs.

1981; Sinopoli 1988). Specialized systems are assumed to be competitive and cost conscious. Specialists, having to repeat the same task many times, not only become very skillful and able to replicate their products, but are also motivated to increase their efficiency by simplifying and standardizing their products as well (Allen 1984; Hagstrum 1985). At the same time, standardization is also usually interpreted as the product of a limited number of production units because, as Barbara Stark has shown, the lower the ratio of producers to products, the higher the standardization (Stark 1995). In other words, archaeologically, homogeneity in material culture is often associated with large scale, specialized production, while variability is associated with small-scale production.

Archaeologists focus on labor investment as a variable to assess the amount of time required to produce goods and, in this way, the context of production. Utility wares, with their simple forms and lack of decoration, are usually considered to be products manufactured for general consumption. They may also reflect a competitive economic situation, where the producers try to be as efficient as possible and produce more in less time. On the other hand, elites are guaranteed to consume the products of the specialists attached to them, and thus the producers are relieved from the economic constraints of a competitive market and can devote more time to each of their products. At the same time, these goods carry more social and political information, and are more visible than utilitarian goods. It is expected that more energy is required for their production.

Recent ethnoarchaeological work however has pointed out that the relationship between standardization and specialization is not that simple. Stark (Stark 1995) compared the products of specialists and non-specialists from Papua New Guinea, Guatemala and the Amazon Basin. On average the products of specialists were more standardized than those of non-specialists. However, the degree of standardization among specialists varied considerably and in some cases non-specialist products were more standardized than those of specialists (Stark 1995: 254).

In other words, archaeologically, vessels that show high labor investment are expected to be associated with producers attached to elite sponsors, while low labor investment in ceramic production is associated with the economic competition that underlies independent production (Costin & Hagstrum 1995; Costin 1991).

Most importantly, researchers have also pointed out that a variety of factors not related to the organization of production may contribute to standardization, such as the existence of cultural limits to the amount of variability tolerated for a particular kind of vessel or function (Arnold 1991: 364; Costin & Hagstrum 1995: 622; Costin 1991: 3334).

2) The concentration of production refers to the distribution in space of producers and consumers and requires a regional perspective. At one extreme the place where ceramics are produced is the same as the place where they are used, while at the other production is concentrated on one locality and the goods and/or the consumers (or distributors) have to be moved. Given our current knowledge about the Maros pottery, it is not possible to speculate about this parameter, and thus it shall not be discussed in further detail here.

It is evident that standardization is not an absolute variable, measurable in a presence/absence fashion. It is a relative measure that must be assessed within specific industries and technologies and, therefore, even though it is easy to measure it remains hard to assess. It is also a variable hard to interpret and one must always be aware that it may be influenced by factors other than the organization of production.

3) The constitution of the production unit describes the size of the productive groups and the social relations of their members. Household and workshop production are seen as the two extremes of the continuum. In a household context the tools and facilities are modest and the producer is an individual or a small number of usually related individuals. In a workshop the tools and facilities suggest considerable economic investment and the producers are unrelated.

4) Finally, intensity of production refers to the amount of time producers devote to their craft, as opposed to other economic activities. At one extreme are part-time producers and at the other full-time ones. If production is specialized, then the repetition of the task and the experience acquired is expected to produce goods that are technologically consistent, with fewer errors and accidents, and production sequences that are technologically and artistically complex.

Standardization is a variable commonly considered by archaeologists in order to assess the constitution of the production unit. It is often assumed that as production becomes more specialized, the products become more standardized. In many cases, research has supported this relationship (Hagstrum 1985; Longacre et al. 1988; Rice

Therefore, by examining skill, one can assess the experience, proficiency and talent of a craftsperson, and ultimately, the intensity of production. Since skill involves a measure of subjectivity, however, it is harder to quantify.

78

Ceramic production among the Maros villagers of Bronze Age Hungary

which regional social and political integration is achieved by means of broadly shared ideological structures and sodalities" (O'Shea 1996: 376).

MAROS ORGANIZATION OF PRODUCTION

In the case of Maros, my data come primarily from two sources: a) John O'Shea's r~tly published mortuary analysis of all the currently known Maros cemeteries (O'Shea 1996) and b) my analysis of the ceramic collections from the Maros settlements ofKiszombor and Klara:falva.

There is no direct evidence for elite control of ceramic production. Thus, it can be hypothesized that the Maros potters were independent artisans, producing in their own households for their own needs (as non-specialists), or in a broader community of consumers (as specialists). I will examine labor investment to assess this indirectly.

Kiszombor is an open level site, dated to 2500-2000 cal. BC (O'Shea 1991), belonging to the Early Maros Phase (Early Bronze Age). Klara:falva is a tell, dated to 2000-1650 cal. BC (O'Shea 1991), belonging to the Late Maros Phase (Middle Bronze Age). The samples studied included 1500 rims, bases, handles and decorated body sherds from Kiszombor and 2880 from Klara:falva. The material was excavated as part of a collaborative project between the University of Michigan, Museum of Anthropology (UMMA) and the Mora Ferenc Muzeum, Szeged, Hungary, under the joint direction of Dr. Ferenc Horvath and Dr. John O'Shea, during the summers of 1987 through 1989. With the permission of the Mora Ferenc Muzeum the material was loaned for study to the Kipke laboratory :facility of the University of Michigan Museum of Anthropology.

In both Kiszombor and Klara:falva coiling and scraping was the main vessel forming technique. The coils were not applied directly on top of the previously built wall, but on the inside or the outside of it. Clay was then smeared up and down to seal the juncture and eliminate air pockets that might expand during firing and cause the pot to break. Additional clay was then pinched and pulled up to continue building the pot. In many cases, subsequent finishing eliminated the thin-thick areas at the coil joints. The contact point between the coils remained however a weaker point in the vessel and in some cases vessels break along the coil junctures, in straight lines, with tails protruding on either side of the clay which was smeared over the joint (Vitelli 1984: 119). Slab building was also used in very large vessels, while a combination of pinching and coiling was used in some small restricted vessels. There are no wheel made ceramics or any extensive use of molds.

The most widely used typology of Maros ceramics is that defined by Istvan Bona (Bona 1975), based on the ceramics from the Szoreg cemetery (Fig. 2). His purpose was to define the group's internal chronology. His phases 1 and 2 represent the Early Maros. They include one and two handled vessels, biconical bowls and beakers. A transitional Phase 3 shows the beginning of highly arched handles called kantharos. Phases 4 and 5 represent the Late Maros, including the so called ''baroque" vessels, with rhomboidal rims, pedestal bases, ansa lunata (half-moon shaped handles) and kantharos handles.

To determine whether the variation from orange to black in vessel color was affected by differences in clays as well as variability in firing, I refired 60 sherds from Kiszombor and 91 from Klara:falva for half an hour in an oxidizing atmosphere in an electrical muffle kiln at 950 degrees Celsius. I chose 950 degrees, not only because it is expected to be higher than the temperature in which the sherds were initially fired, but so that it would allow the iron oxides to oxidize into ferric oxides (Morariu et al. 1991: 161). All the samples, from both sites, became the same color of dusky red (2.5 YR 3/4 in the Munsell code). This suggests that variation in color is primarily due to firing, rather than different kinds of clay used.

In the present analysis the first challenge was that the sample was so fragmentary that measurements providing information on shape or size could not be accurately made on the majority of sherds. Accordingly, I could not follow Bona's typology which was based on complete vessels. Furthermore, it is not clear to what extent that typology was appropriate to settlement ceramics. Instead, I focused on clay inclusions and surface treatment. Rim diameter was the only morphological variable measurable on the majority of the sample and thus it was included to provide a general idea of sizes. Vessel forming techniques were observed as well. The following is a discussion of the mortuary and ceramic analysis data as they relate to the parameters of production context, scale and intensity.

Preliminary petrographic analysis of six potsherds from each site revealed that the only temper used was ground potsherds. In general, the samples looked relatively homogeneous in almost every way. However, in Klara:falva, the later site, preliminary results suggest that coarse wares may have been made from different clays than fine wares (Stoltman n.d.). A considerably larger sample is currently under analysis. Based on the temper size and sorting, the samples were divided into coarse, medium, and fine inclusion categories. Within each category the surface treatment was noted and in this way nine wares were defined for Kiszombor and ten for Klara:falva (Table 1). The percentage of Plain and Rough sherds decreases in Klara:falva, while the percentage of

CONTEXT OF PRODUCTION/LABOR INVESTMENT

O'Shea describes the Maros as "a clear example of an autonomous village type of social and regional organization. They exhibit a segmentary form of organization that emphasize the autonomy of local communities and one in

79

Kostalena Michelaki

burnished sherds increases. In particular, the percentage of highly burnished sherds rises.

pebbles and ones that included shells (O'Shea 1996: 229230), suggesting the possibility that certain wares were produced by some potters and not by others (burnished vs. roughened vessels), or that labor was divided in some way whereby scraping was performed by some members of the group, perhaps apprentices, and surface treatment was executed by others.

Even though the analysis is preliminary, a pattern emerges that suggests higher labor investment in the production of . the Late Maros Phase ceramics. Their ''baroque" shapes, as known from the cemeteries, are far more complicated to build, taking more time, and their surfaces are treated in ways that also would take more time to execute. At the same time, a higher percentage of vessels is decorated and the decoration occupies a larger area of the vessel body, suggesting again higher labor input in the production of Late Phase ceramics. Further analysis, following the production step measure proposed by Fienman et al. (1981), will be very useful in quantifying the increase in labor input suggested here.

During the Late Maros Phase tools are rarer in grave assemblages and we have to tum to the ceramic vessels themselves. Evidence from the funerary vessels is somewhat surpnsmg. O'Shea's measurements show a strong correlation between rim diameter and vessel height, which suggests an emphasis on relative proportions. As vessels get bigger in rim diameter they also get taller. The proportionality is maintained throughout both the Early and the Late Phases (O'Shea 1996: 95-96). Relative dimensional standardization seems to characterize the mortuary assemblage.

At the same time, the main building technique did not change. No molds or wheels were introduced, nor were the features interpreted as kilns more efficient in any way during this Phase.

Since the variables of rim diameter and vessel height may be highly influenced by the intended function of the vessels, the relative standardization of morphological and technological attributes measured on the sherds from Kiszombor and Klarafalva was assessed as well. Rim diameter (included as an anchor between the mortuary and the household ceramics), rim thickness, and body thickness were measured for each ware. Histograms were initially evaluated for modality in their distribution. In the case of rim and body thickness, distributions were unimodal with a few outliers. Extreme outliers were eliminated before summary statistics were calculated. In the case of rim diameter certain histograms showed multi.modality, which was interpreted as indicating discrete size classes. The wares were divided into the respective size classes and then summary statistics were calculated.

SCALE OF PRODUCTION/STANDARDIZATION

There does not appear to be any evidence for the presence of large workshops, storage rooms, wheels, real kilns, or anything that would point to considerable investment in pottery production. Mercury porosimetry analysis on potsherds from the Bronze Age strata of the Pecica settlement in Romania (Morariu et al. 1991), and scanning electron microscope analysis of sherds from the Early Bronze Age material of the T6szeg tell in Hungary (Maniatis & Tite 1981) agree that the samples were fired at temperatures below 700 degrees Celsius. Thus, although features interpreted as kilns are often found in Maros sites, even when they were used to fire pottery they did not reach high temperatures. One can expect that the scale of production, whether specialized or not, would be limited, probably to household or interhousehold groups.

The coefficient of variation was calculated for each variable, since it describes the distribution of a sample without being influenced by size (c.v. = standard deviation x 100/mean). As table 2 shows, coefficients of variation range between 10.22 percent and 47.21 percent. Since standardization is a relative measure, finding an appropriate example to compare it against becomes crucial. Many standardization studies compare their results against ethnographic examples of known production organization. When compared to coefficients of variation of ethnographic samples, the Maros ones are much larger than even the ones of non-specialist products (e.g. the Kalinga pots in Longacre et al. 1988). Is this to be interpreted as an example of extreme variation, suggesting individual household unspecialized production?

Towards the end of the Early Maros, around 2000 BC, however, more evidence exists from the Maros cemeteries. When the graves that contained tools in the Mokrin cemetery were plotted, the ones with presumed pottery production tools (burnishing pebbles or shells for scraping) clustered to the western half of the cemetery, while the ones with textile production tools (spindle whorls or loom weights) clustered to the eastern half (O'Shea 1996: 229). This suggests that, at least as early as 2000 BC, ceramic production in some Maros communities was specialized (O'Shea 1996: 270, 310). The status of "potter" was recognized as separate from that of other craftspeople (textile producers in this case) and it was not accompanying everyone in the cemetery.

Longacre et al. studied an archaeological ceramic sample from the Grasshopper Pueblo and noticed that: "the archaeological situation creates a major analytical stumbling block because statistics that describe variation are totally dependent on how the ceramic classes or types happen to be defined .... Inadvertent lumping of multiple size classes by

Furthermore, graves with the presumed pottery production tools in Mokrin were also divided into ones that included

80

Ceramic production among the Maros villagers of Brorize Age Hungary

the archaeologist has the effect of inflating the apparent variation of the measurements" ( 1988: 106-107).

and accurate timing for the building and assembling of the various vessel parts. The appearance of slipped surfaces in the Late Phase further supports this argument.

They conclude that, however an archaeological ceramic class is defined, it is most likely to include one or more emic classes. Variation statistics calculated on an archaeological class will probably be too high (Longacre et al. 1988: 109). In the particular Maros sample various emic classes are bound to be present within at least some ware categories, since shape information is currently lacking. Even if we could be sure, however, that we have captured emic classes it would still be expected that a sample produced over a millennium, as in the case ofMaros, would be more variable that a sample produced over a year, as in the case of the Kalinga.

Furthermore, in the early site of Kiszombor various vessels showed forming mistakes. Areas on the interior of some vessel walls were hatched with a pointed tool while the clay was still quite wet so that more clay could adhere to them, possibly because they had been unevenly scraped. In some cases the piece of clay that was used to patch the spot is also found, with the negative of the hatching marks. No such mistakes were found in Klarafalva. In sum, there is a qualitative difference between the Early and Late Phase Maros ceramics. The Late Phase Maros ceramics suggest higher skill.

Compared to other archaeological examples of hand-made ceramic assemblages the coefficients of variation do not seem that large. Coefficients of variation calculated for morphological variables of Prehispanic Wanka cookwares and storage jars and lnka jars from Peru ranged between 29.46 percent and 49.09 percent (Costin & Hagstrum 1995: 632). Still, one should not expect some kind of universal scale against which to measure standardization in archaeological cases either (Costin 1991: 36).

DISCUSSION AND CONCLUSIONS

In summary, the results of the indirect ceramic analysis show that the relative levels of dimensional standardization remain constant throughout the Early and Late Maros Phases. On the other hand, the levels of labor investment and skill are higher during the Late Phase. The next step is to consider how well these variables reflect the organization of pottery production in the case ofMaros.

Instead of comparing the coefficients of variation of my samples to those of other examples, either ethnographic or archaeological, I think it is more important to compare them to each other. It is most important, for the time being, that there is no important difference between the two samples. They are both equally standardized (or variable). Even though it is not appropriate to compare the household with the mortuary coefficients of variation, since the first are based on ware classifications and the latter on shape categories it is, nevertheless, interesting to note that they are similar. In other words, the degree of relative standardization in both the mortuary and the household ceramics does not reflect the suggested appearance of specialized producers during the Late Maros Phase in cemeteries.

Current settlement and mortuary evidence suggests that there is no hierarchy of settlements or communities and an absence of any effective overarching central control. Yet, the linkages between communities were sufficiently strong to promote a potent and functional regional identity. In this light the standardization of Maros ceramics could be due to its social role as a marker of the Maros group affiliation and identity. Low tolerance for variability, and a strong idea of what Maros pottery was supposed to look like, rather than an economically competitive market, seems a more appropriate explanation for the "standardization" evident throughout the Early and Late Maros Phases.

As mentioned earlier, the information from settlements is limited. However, the available evidence shows that households participated in a variety of economic activities, producing generalized, rather than specialized inventories. It would seem likely, therefore, that pottery producers were part-time producers, who did not have to depend completely on others for their subsistence goods.

In the Late Phase, however, higher labor investment and skill are evident in the ''baroque" styles that became prevalent. At the same time, individual status that was formerly expressed through the mortuary program now leaves the funerary context (O'Shea 1996: 252-253). A change in the role of ceramics is suggested. Besides group affiliation, individual status may also now be displayed through the more labor intensive skillfully made ceramics used for feasting and display during life. The northern and southern trends that appear in the pottery may communicate participation of their owner in a northern and/or southern network of relations.

Indirectly, the complex morphology of the "baroque" vessels of the Late Phase mortuary and household contexts, with the ansa lunata and kantharos handles, the pedestal bases and the rhomboidal or flower-like rims required a very competent understanding of the drying properties of clay,

To conclude, I think that currently we lack sufficient conclusive evidence to assess whether production was specialized or not in the beginning of the Early Maros Phase. It is evident though that both the scale and the intensity of production were limited.

INTENSITY OF PRODUCTION/SKILL

81

Kostalena Michelaki

Banner, J. 1928 Az Oszentivani Asatasok. Dolgozatok Szeged IV: 148-243. 1929 Az Oszentivani Bronzkori Telep es Temeto. Dolgozatok SzegedV: 52-81. 1931 A Marosvidek Bronzkori Telep es Temeto.

By 2000 cal. BC however, towards the end of the Early Phase, there is evidence from the Mokrin cemetery that potters are recognized by their communities as having that status. By the beginning of the Late Phase the vessels themselves show higher labor investment and skill. In this case, I believe that the evidence is sufficient to suggest some degree of specialization. The scale and intensity remain limited, but are intensified in relation to the Early Phase. Further, the division in the Mokringraves between the ones including pebbles and the ones including shells is paralleled by the initial finds of petrographic analysis in the Late Maros site of Klarafalva, where the fine ceramics seem to be made from clays different than the ones for coarse ceramics. Some degree oflabor differentiation in craft specialization is suggested, along with the possibility that the production of some wares remains unspecialized, while that of others becomes specialized.

Dolgozatok a Szegedi Tudomar,yegyetem Regisegtani Intezeteb6l V (1-2): 52-78. Barraclough, A. 1992 Quaternary Sediment Analysis: a Deductive Approach at A-Level. Teaching Geography 17: 1518. Bokonyi, s. 1972 Animal Remains from the Graves of the Bronze Age Cemetery at Mokrin. In N. Tasic (ed.) Mokrin II: The Early Bronze Age Necropolis, pp. 91-97. Dissertationes et Monografiae. Narodni Muzej, Kikinda. Bona, I. 1975 Die Mitt/ere Bronzezeit Ungarns und Ihre Siidostlichen Beziehungen. Akademiai Kiado, Budapest. Brumfiel, E. & T. Earle (editors) 1987 Specialization, Exchange, and Complex Societies. Cambridge University Press, Cambridge. Childe, V. G. 1929 The Danube in Prehistory. Clarendon Press, Oxford. Clark, J. & W. Parry 1990 Craft Specialization and Cultural Complexity. Research in Economic Anthropology 12: 289-346. Costin, C. L. 1991 Craft Specialization: Issues in Defining, Documenting, and Explaining the Organization of Production. In M. Schiffer (ed.) Advances in Archaeological Method and Theory 3, pp. 1-56. University of Arizona Press, Tucson. Costin, C. & M. Hagstrum 1995 Standardization, Labor Investment, Skill, and the Organization of Ceramic Production in Late Prehispanic Highland Peru. American Antiquity 60 (4): 619-639. Day, P., D. Wilson & E. Kiriatzi 1997 Reassessing Specialization m Prepalatial Cretan Ceramic Production. In R Laffineur & P. Betancourt (eds.) TECNH: Craftsmen, Craftswomen and Craftsmanship in the Aegean Bronze Age, pp. 275-289. Aegaeum 16: Annales d'Archeologie Egeenne de l'Universite de Liege et UT-PASP. Universite de Liege, University of Texas, Austin, Texas. Feinman, G., S. A. Kowalewski & R E. Blanton 1984 Modelling Ceramic Production and Organizational Change in the Pre-Hispanic Valley of Oaxaca, Mexico. In S.E. Van Der Leeuw & A.C. Pritchard (eds.) The Mar,y Dimensions of Pottery: Ceramics in Archaeology and Anthropology, pp. 295-337. Universiteit van Amsterdam, Amsterdam.

Even though the picture of ceramic production presented here is sketchy and far from definitive, I hope that this paper has contributed to an understanding of changes in Maros ceramic production through time. I also hope that the need to consider many dimensions, from various contexts, when assessing production systems became evident. Finally, I hope that further analyses of ceramic production, use and distribution will take place in the context of Eastern Central Europe contributing to a better understanding of its economic and social prehistory.

Acknowledgments I am grateful to my advisors, John O'Shea, who kindly allowed me access to his Maros ceramic collections and has always supported my efforts, and Carla Sinopoli who commented extensively on the various drafts of this paper. Many thanks to Dr. Otto Trogmayer, Director of the Mora Ferenc Muzeum, for the opportunity to study Maros collections under his care, and to Professor William Farrand, who provided access to the University of Michigan Geology Lab muffle kiln for the re-firing tests. Evangelia Kiriatzi has supplied me with articles, support and stimulating ideas on pottery production during the last ten years and I thank her.

BIBLIOGRAPHY Allen, J. 1984 Pots and Poor Princes: A Multidimensional Approach to the Role of Pottery Trading in Coastal Papua. In S.E. Van Der Leeuw & A.C. Pritchard (eds.) The Mar,y Dimensions of Pottery: Ceramics in Archaeology and Anthropology, pp. 407-473. Universiteit van Amsterdam, Amsterdam. Arnold, P., III 1991 Domestic Ceramic Production and Spatial

Organization: A Mexican Case Study in Ethnoarchaeology. New Studies in Archaeology. Cambridge University Press, Cambridge.

82

Ceramic production among the Maros villagers of Bronze Age Hungary

Feinman, G., S. Upham & K. Lightfoot 1981 The Production Step Measure: An Ordinal Index of Labor Input in Ceramic Manufacture. American Antiquity 46 (4): 871-884. Foltiny, I. 1941 A Szoregi Bronzkori Temeto. Dolgozatok a Szegedi Tudomanyegyetem Regisegtani Intezetebol 17: 189. 1942 Bronzkori Leletek Klarafalvarol es Kiszomborr61. Dolgozatok Szeged 18: 99-103. Giric, M. 1987 Naselja Moriske Kulture. Rad Vojvodancki Myseja 30: 71-83. Hagstrum, M. 1985 Measuring Prehistoric Ceramic Craft Specialization: a Test Case in the American Southwest. Journal of Field Archaeology 12 (4): 65-75. Horvath, L. A. 1985 Kora-Bronzkori Telepiiles Kiszomborban. A Mora Ference Muzeum Evkonyve 1982-83 (1): 72-93. Kaiser, T. 1984 Vinca Ceramics: Economic and Technological Aspects of Late Neolithic Pottery Production in Southeast Europe. Unpublished Ph.D. thesis, University of California. Kiriatzi, E., S. Andreou, S. Dimitriadis & K. Kotsakis 1997 Co-Existing Traditions: Handmade and Wheelmade Pottery in Late Bronze Age Central Macedonia. In R Laffineur & P. Betancourt (eds.) TECNH: Craftsmen, Craftswomen and Craftsmanship in the Aegean Bronze Age, pp. 361-367. Aegaeum 16: Annales d'Archeologie Egeenne de l'Universite de Liege et UT-PASP. Longacre, W., K. Kvamme & M. Kobayashi 1988 Southwestern Pottery Standardization: an Ethnoarchaeological View from the Philippines. The Kiva 53 (2): 101-112. Maniatis, Y. & M. Tite 1981 Technological Examination of Neolithic-Bronze Age Pottery from Central and Southeast Europe and from the Near East. Journal of Archaeological Science 8: 59-76. Mills, B. & P. Crown (editors) I 995 Ceramic Production in the American Southwest. The University of Arizona Press, Tucson. Morariu, V., M. Bogdan & I. Ardelean 1991 A Physical Investigation of Bronze Age Ware from Pecica. In T. Soroceanu (ed.) Studien zur MuresKultur, pp. 161-168. Internationale Archaologie. vol. 7. Buch am Erlbach, Milnchen. O'Shea, J.M. 1991 A Radiocarbon-Based Chronology for the Maros Group of Southeast Hungary. Antiquity 65: 97-102. I996 Villagers of the Maros. Plenum Press, New York. n.d. Report on Excavations at Klarafalva-Hajdova and Kiszombor Uj Elet. University of Michigan Museum of Anthropology.

Orton, C., P. Tyers & A. Vince 1993 Pottery in Archaeology. Cambridge University Press, Cambridge. Peacock, D. P. S. 1981 Archaeology, Ethnology and Ceramic Production. In H. Howard & E.L. Morris (eds.) Production and Distribution: a Ceramic Viewpoint, pp. 187-194. Oxford: British Archaeological Reports International Series 120. Rice, P. 1981 Evolution of Specialized Pottery Production: A Trial Model. Current Anthropology 22 (3): 219240. 1987 Pottery Analysis: A Sourcebook. The University of Chicago Press, Chicago. Sinopoli, C. 1988 The Organization of Production at Vijayanagara, South India. American Anthropologist 90 (3): 580597. Stark, B. 1995 Problems in Analysis of Standardization and Specialization in Pottery. In B. Mills & P. Crown (eds.) Ceramic Production in the American Southwest, pp. 231-267. University of Arizona Press, Tucson. Stark, M. 1991 Ceramic Production and Community Specialization: a Kalinga Ethnoarchaeological Study. WorldArchaeology23 (1): 64-77. Stoltman, J. n.d. Report on the Petrographic Analysis of Potsherds from the Bronze Age Settlements of Kiszombor and Klarafalva, Hungary. University of Wisconsin. Underhill, A. 1991 Pottery Production in Chiefdoms: the Longshan Period in Northern China. World Archaeology 23 (1): 12-27. Van der Leeuw, S. & A. C. Pritchard (editors) 1984 The Many Dimensions of Pottery: Ceramics in Archaeology and Anthropology. Universiteit van Amsterdam, Amsterdam. Vitelli, K. 1984 Greek Neolithic Pottery by Experiment. In P. Rice (ed.) Pots and Potters: Current Approaches in Ceramic Archaeology, pp. 113-131. Institute of Archaeology Monographs, vol. XXIV, Institute of Archaeology, University of California, Los Angeles. Whitelaw, T., P. Day, E. Kiriatzi, V. Kilikoglou & D. Wilson 1997 Ceramic Traditions at EM IIIB Myrtos, Foumou Korifi. In R Laffineur & P. Betancourt (eds.) TECNH: Craftsmen, Craftswomen and Craftsmanship in the Aegean Bronze Age, pp. 265274. Aegaeum 16: Annales d'Archeologie Egeenne de l' Universite de Liege et UT-PASP.

83

Kostalena Michelaki

Table l: Ware definitions for the Kiszombor and Klarafalva ceramic assemblages

Sorting An inclusion sorting chart was used, based on the one published in Orton et al. (1983), after Barraclough (1992), with a few modifications.

1 = very poor 1.5 2 =poor 2.5 3 = fair 3.5 4 = good 4.5 5 = very good 5.5 6 = excellent Coarse= 1, 1.5, 2, 2.5 Medium = 3, 3.5, 4, 4.5 Fine= 5, 5.5, 6

Exterior surface treatment Smoothing, wiping, burnishing and brushing were observed, as defined by Rice (1987). Plain = smoothed, highly wiped Burnished = lightly burnished, burnished, highly burnished Rough = lightly wiped, brushed Burnished and rough = top is burnished and bottom is rough Burnished and Plain = top is burnished and bottom is plain

84

Ceramic production among the Maros villagers of Bronze Age Hungary

Table 2: Summary statistics of rim diameter, rim thickness and body thickness of the various Kiszombor and Klarafalva wares, divided by site (*ware appearing only in Klarafalva, ** sample size too small for meaningful calculation of c.v.)

Rim Diameter/Settlements Ware Kiszombor/Klarafalva Coarse/Plain Coarse/Burnished-small Coarse/Burnished-medium Coarse/Burnished-large Coarse/Burnished-v.large Coarse/Rough Coarse/Burnished+Plain Coarse/Burnished+Rough *Coarse/Slipped Medium/Plain Medium/Burnished *Medium/Slipped Fine/Plain Fine/Burnished

N. Kiszombor/Klarafalva 29 47 276 35 26 130 25 72 11 16 32 1 1 -

-

20 117

9 219

Mean Kiszombor/Klarafalva 17.41 22.04 12.91 14.16 23.73 24.86 35.90 35.72 51.81 49.81 18.96 17.00 42.00 -

St.D. Kiszombor/Klarafalva 5.82 10.40 4.25 3.31 2.61 2.42 3.80 3.34 3.97 2.67 7.17 -

-

-

-

-

-

-

17.95 15.44

14.44 14.34

6.65 6.49

3.46 7.80

37.06 42.06

24.00 54.38

-

-

-

-

-

-

-

-

-

-

3 5

-

12.00 13.40

-

6.92 5.02

8

c.v. Kiszombor/Klarafalva 33.46 47.21 25.60 30.05 10.22 10.52 10.59 9.36 5.16 7.97 37.81 ** ** -

7.87

2.10

-

-

** **

**

Rim Thickness/Settlements Ware Kiszombor/Klarafalva Coarse/Plain Coarse/Burnished Coarse/Rough Coarse/Burnished+Plain Coarse/Burnished+Rough *Coarse/Slipped Medium/Plain Medium/Burnished *Medium/Slipped Fine/Plain Fine/Burnished

N. Kiszombor/Klarafalva 46 38 96 560 1 39 1 -

-

-

20 128

10 259

Mean Kiszombor/Klarafalva 8.18 9.29 7.79 8.25 8.07 8.10 6.75 -

St.D. Kiszombor/Klarafalva 2.12 2.28 2.06 1.78 1.64 -

-

-

-

7.72 6.12

7.40 6.33

1.52 1.36

-

1.04 1.63

c.v. Kiszombor/Klarafalva 25.96 24.63 22.91 25.01 20.34 ** ** -

-

-

-

-

19.71 22.25

14.14 25.79

-

-

-

-

-

-

-

-

2 5

-

10.25 4.51

-

0.49 2.23

-

** **

**

9

3.93

1.01

-

Body Thickness/Settlements Ware Kiszombor/Klarafalva Coarse/Plain Coarse/Burnished Coarse/Rough Coarse/Burnished+Plain Coarse/Burnished+Rough *Coarse/Slipped Medium/Plain Medium/Burnished *Medium/Slipped Fine/Plain Fine/Burnished

N. Kiszombor/Klarafalva 25 28 296 669 23 187 4 28 14 5 1 2 30 464 63 3 -

Mean Kiszombor/Klarafalva 8.92 9.31 7.16 7.90 9.12 8.90 8.67 7.71 8.95 8.30 10.50 4.15 7.97 5.70 6.15 10.53 -

St.D. Kiszombor/Klarafalva 2.61 2.92 2.02 2.01 2.26 2.67 1.59 1.80 1.41 1.33

0.21 1.65

-

-

-

-

-

-

4

-

3.97

-

85

2.38 1.65 0.90

0.62

c.v. Kiszombor/Klarafalva 29.28 31.41 28.12 25.55 24.83 30.05 20.85 ** 14.94 ** ** ** 29.98 28.98 26.94 ** ** ** **

Kostalena Michelaki

LEGEND 1 Kiszombor 2 Klarafalva 3 Oszentivan 4 Sz!!reg 5DeszkA,F 6Mokrin 7Perjamos 8 Pecska

10km

Principal Maros Sites •

Cemeteries

.A.Settlements Other Sites

A

Settlements

Figure I: Some of the main settlements and cemeteries of the Maros (after O'Shea 1996: 28)

86

Ceramic production among the Maros villagers of Bronze Age Hungary

Istvan B6na's Typology

Szoreg 1

Early Maros Phase Szoreg 2

Szoreg

[) g t3

6

@

3

t

Szoreg 4

Late Maros Phase Szoreg 5

cJ ~

Figure 2: Maros ceramic forms characteristic of the five phases recognized by Bona (1975) (after O'Shea 1996: 30)

87

The Question of Specialization Levels in Pottery Production between the End of the Bronze Age and the Early Iron Age in Daunia (Southern Italy) Paolo Boccuccia

developments in southern Italy between the end of the 2nd and the beginning of the 1st millennium BC, in particular reference to the manufacturing techniques used.

INTRODUCTION

Manufacturing techniques of Southern Italian figulina ware have frequently been the subject of discussion for several scholars as a key argument towards understanding cultural developments at the end of the 2nd and the beginning of the 1st millennium BC. Recently, a xeroradiographic study has been carried out on Daunian figulina sherds focusing on how different production techniques (handmade, wheelmade in a single piece or by joining different parts) can be contemporary in a single pottery class. According to many scholars the use of a wheel, being a technical innovation of great importance for its impact on the organization of production and manufacturing processes, would imply a precise technical know-how and a higher level of social organization of production, and it therefore might represent an important indicator in the global interpretation of social structure. However, the patterns regarding the different types of social organization proposed by various scholars are too strict, when the use of wheel has ·been mechanically related to specialized forms of production organization defined as "workshops". The great stylistic and technical homogeneity of the pottery analysed, and the probable impossibility for the user to distinguish between wheel-made from hand-made vessels, points out how the actual importance of such vessels was principally related to the macroscopic characteristics of this class (such as the kind of clay, the light colour of the surfaces and the decorative patterns), features that probably determined the use made of them and/or the purpose of their possession.

THE RECONSTRUCTION OF MANUFACTURING TECHNIQUES

The reconstruction of manufacturing techniques is one of the aspects that can be effectively isolated in the study of production technology, and in this case it is particularly important: during the period considered, in fact, there are more and more frequent signs of the use of the potter's wheel in Southern Italy, following the local production at different sites ofltalo-Mycenean ware. By wheel-making, I mean the pottery forming technique performed by the means of a rotating support, in which the deformation of plastic clay is produced by centrifugal force combined with hand pressure. In order to make a pot on the wheel it is necessary to obtain enough speed to "raise" the clay, and I do not accept the problematic idea of a "slow wheel" frequently proposed for the production of this kind of ware (Yntema 1990; Gorgoglione et al. 1994). Hulten (1974: 69) states that it would be necessary to have a minimum tangential speed of about 0.5 m per second. The use of a rotating support is generally thought to be a technological innovation of great importance for the influence on the organization of production, manufacturing process and the morphology of the finished products. As frequently proposed by several authors, it would also have involved an advanced level of technical knowledge and social organization of production; thus it might represent an important indicator in the interpretation of the overall social structure.

FIGULINA WARE

The painted figulina ware (a distinctive type of fine ware datable between the end of the Bronze Age and the Early Iron Age in Daunia, southern Italy) recently found at the site of Coppa Nevigata (Boccuccia 1997) and attributed to the Southern Italian Protogeometric, Southern Italian Early Geometric and Daunian Middle Geometric (Figure 1), is characterised by a recurrent geometrical decorative pattern that finds comparisons in all those Apulian and Lucanian sites that have pottery datable between the end of the Bronze Age and the Early Iron Age. This would seem to indicate that during this period there was a wide diffusion of similar decorative patterns (especially in the Southern Italian Protogeometric) of figulina ware which represented an uncommon and sought after product. These ceramic products made of tempered clay have often been a key argument for the comprehension of the cultural

Starting with Taylour, for whom "nearly all the pottery shows wheelmarks but some of it was made by hand" (Taylour 1958: 159), scholars have often debated the problem of the use of the potter's wheel in relation to the manufacturing techniques of the Late Bronze Age and Early Iron Age painted figulina ware. Lo Porto (Lo Porto 1964: 210 and 212) states that both were hand-made, while Peroni (1967: 119) affirms that the presence of these wheel-made and painted wares joined the production of impasto (coarse ware) pottery, and that this was the most significant evidence of the social and economic development which occurred in Apulia during the Late Bronze Age. Like Lo Porto, De Juliis considers these wares to have been completely handmade (De Juliis 1977: 23, note 1). For

89

Paolo Boccuccia

Yntema (1990: 19 and 3I)figulina ware pottery produced during these phases was all hand-made, even if in some cases the use of a slow wheel can be hypothesized. Pots were formed by joining together different portions made separately, as shown by the joints clearly visible inside closed vessels. Yntema also states that the tools used for the final work of the pots frequently left horizontal traces that can be confused with wheel-marks. Ruby (1988), in a technological study carried out exclusively on a specific 8th century BC class of pottery called ceramica a tenda (literally "tent-like ceramics", after a class of distinctive painted motifs) confirmed Yntema's hypothesis, and stated that the pottery he analysed was not wheel-made.

clay used (the sercalledfigulina), but also by the occurrence of the use of the potter's wheel since the Southern Italian Protogeometric. Wheel-making again became the main (but not the only) method used for Daunian Middle Geometric ware.

LEVELS OF SPECIALIZATION IN POTTERY PRODUCTION As previously stressed, the use of a rotating support is generally considered to have been a technological innovation of great importance for its occurrence in the organization of production, manufacturing process and the morphology of the finished product. The use of such a tool has frequently beenfocused on as one of the main indicators of specialized production. Many and diverse patterns have been proposed by different authors about the implications of this simple machine for the overall social organization of production, but generally the use of rotating supports is considered to be a reliable indicator of a form of production often referred to as the "workshop" level (van der Leeuw 1984).

At the end of the 1980s Peroni (1989: 306) stressed again how the Protogeometric period was characterised by the production of a wheel-made, painted fine ware, very homogeneous in style and technology (una produzione di ceramicafigulina tomita e dipinta fortemente unitaria nello stile, nelle tecniche) and how this is an important indicator of a proterurban society (Peroni 1989: 307). Recently Gorgoglione stated that Protogeometric ware from Torre Castelluccia was made on a kind of slow wheel, but using a colombino (coil-building) technique (Gorgoglione et al. 1994: 76). Finally, manufacture on the wheel has been hypothesized for some Protogeometric sherds from Lipari, Termitito and Madonna del Petto (Bernabo Brea & Cavalier 1980: 566-567, 608; Bianco & De Siena 1982: 75; Laviano et al. 1995).

Problems regarding the manufacturing techniques of these ceramics are hence a central matter for discussion. A recent archaeometric study carried out on the figulina ware from Coppa Nevigata (Boccuccia et al. 1998) tried to verify, through the application of adequate analytical techniques, i.e. by means of xeroradiography, the rate of occurrence of the evidence of the use of the potter's wheel in the production process. The research confirmed the use of the potter's wheel in Southern Italy for this kind of pottery beginning in the Protogeometric period, and stressed how different manufacturing techniques (hand-building, wheelmaking out of a single clay piece, or by joining different parts together) can be combined together within the same production framework.

The existence of "workshops" for the production of wheelmade figulina ware has been proposed, for example, for Aegean type wares produced in Italy during the phases preceding the Bronze Age (Jones & Vagnetti 1991; Vagnetti 1994). Therefore, also as far as the Southern Italian Protogeometric is concerned, the presence of wheel-made ware could confirm this hypothesis. Due to the cer occurrence of hand-made products, one may also argue that the potter's wheel in itself did not play, in this phase, a central role in improving production, and that there is no specific relationship between the morphological standardization of the products and the evolution of manufacturing techniques. The great homogeneity of these wares, both from a stylistic and from a technical point of view, also leaves open the possibility that for the users it was actually impossible, or irrelevant, to make a clear distinction between wheel-made and non wheel-made pots. This could support another view: the importance given to this class of ware, and the possible status implications of its ownership, depended upon its outer appearance, i.e. upon the type of clay and ceramic texture, upon the light colour of the surface and the decorative patterns, and not upon its commonly applied manufacturing techniques.

At Coppa Nevigata one may observe a diachronic difference in the manufacturing techniques offigulina ware: from the constant use of the potter's wheel for Aegean type wares, we come to a clear decrease (but not to a total disappearance) of its use for the manufacturing of Southern Italian Protogeometric and Southern Italian Early Geometric ware. The continuity with ItalerMycenean ware production traditions in Southern Italy during the phases preceding the ones examined here is therefore evidenced not only by the technical-stylistic features of the decoration and the kind of

A transformation in the organization of production towards a higher level of specialization seems to emerge more clearly during the Daunian Middle Geometric, where there is a quantitative increase in figulina wares of which the percentage, in relation to that of impasto pottery, increases, and hand-made products are mostly the thicker ones, thus expressing an explicit technical choice. In the case of a vessel dated to the Daunian Middle Geometric, different manufacturing techniques have been recorded for different parts of the same vessel. Two sherds which fit together and

FIGULINA PRODUCTION AT COPPA NEVIGATA

90

Pottery production between the end of the Bronze Age and the Early Iron Age in Daunia

production times than impasto pottery, even without taking into consideration the times, the technical processes and the labour required for the production of the pigments used for these decorations. It could therefore be hypothesised that all these aspects could have lengthened the apprentices' training phase, generating a greater social and economic effort for all the community. It could thus be suggested that all the figulina ware production was made by specialized craftsmen, not all of them provided with a wheel-making know-how, but all of them specialized enough to allow them to perform all of the other stages of the production process. It should not be ignored that some pottery classes in Southern Italy have all the aspects illustrated here, and I do not think that in these cases there was a specialized production as has been hypothesized by some scholars (see Malone 1985), considering the general context in which they develop. During the end of the Bronze Age and the Early Iron Age, one witnesses the emergence of either clearly specialized production (i.e. metallurgy) or the general organization of sites and ecological adaptations which were more complex and articulated compared with the Neolithic ones.

which belong to adjoining parallel superimposed sections, in met turned out in radiography to have been made by two different techniques: one by hand, and the other on the wheel (Boccuccia et al. 1998: fig. 3A). The question is therefore whether a single potter mastered two different techniques or, on the contrary, if two potters belonging to the same production unit collaborated with two different techniques in the manumcture of a single production unit. However, it seems difficult to imagine an undifferentiated production of both wheel-made and hand-made pottery within the same "workshop". Recent chemical analyses (Boccuccia et al. 1995) indicate that pottery production, even if not strictly local, seems to have occurred in the same regional context. There seems to have been no clear-cut correlation between chemical composition and the manufu.cturing technique, at least in the cases that allow both a compositional and a radiographic analysis of the same sample, nor between manumcturing technique (wheel-made versus hand-made) and the different styles for the different periods. It also seems that during the Southern Italian Protogeometric and the Southern Italian Early Geometric there were no major patterns of variation regarding the correlation between manufu.cturing techniques and form. Thus, the technological and social context of this important innovation between the end of the Bronze Age and the Early Iron Age is still rather obscure.

CONCLUSIONS

In this light it could be hypothesised that the development of wheel-making, in this case, improved not because of its inner technical advantages, which improved production standards and mvoured a hypothetical mass production, but rather as a function of the need for obtaining more regular and standardized forms using simpler operational sequences. Thus, wheel-making in this case study might be related to the need for balancing and lowering the overall production cost of a class of fine ware, in the framework of a set of processes of social interaction and confrontation whose precise meaning, for the time being, is still elusive.

Could we hypothesise that figulina hand-made pots were produced in Daunia at a lower productive social organization level? Can the use of the potter's wheel alone differentiate between two specialization levels within such a unitary class of products? One may imagine, for example, that hand-made products were a sort of "imitation" of wheel-made products, made by craftspersons that still did not fully master a wheel-making technical know-how, and therefore were distinguished by a lower level of specialization when compared with the craftspersons who were able to use this technique efficiently. If this was not the case, the only possible explanation could be that the use of the wheel alone, as I suspect, might not be a sufficient indicator in itself to distinguish between different levels of specialization.

Ackowledgements I wish to thank A. Cazzella for having supported this work; many thanks also to I. Baroni, M Moscoloni, G. Recchia and M. Tafuri.

Figulina ware production goes together with a long tradition of impasto pottery production: in these phases the latter was still the predominant pottery class (with a presence of more than 90% during the end of the Bronze Age). Figulina wares differentiate themselves from the impasto ceramics, not only for the use of rotating supports during the manumcturing process, but also for the type of clay used, the firing techniques and the presence of painted decorations: all these aspects influence the manumcturing techniques and somehow may be viewed as complicating the production process. In met, a more tempered clay and a higher firing temperature with a highly controlled atmosphere are required. Moreover, these wares show a painted decoration which, although simple, necessarily involves longer

REFERENCES Bernabo Brea, L. & M. Cavalier 1980 Meligunis Lipara IV. Palermo: Flaccovio Editore. Bianco, S. & A. De Siena 1982 Termitito (Montalbano Ionico, Matera). In L. Vagnetti (ed.) Magna Grecia e Mondo Miceneo. Nuovi Documenti. Taranto: Istituto per la Storia e l'Archeologia della Magna Grecia, pp. 69-96. Boccuccia, P. 1997 Nuovi dati sulla frequentazione protostorica di Coppa Nevigata. Atti del 15° Convegno Nazionale sulla Preistoria, Protostoria e Storia della Daunia

91

Paolo Boccuccia

(San Severo, 27-28 novembre 1993), San Severo, 117-144. Boccuccia, P., P. Desogus, F. Fratini, S.T. Levi & E. Pecchioni 1995 Manufacturing Techniques, Raw Materials and Provenance of Italo-Mycenean, Protogeometric and Early Geometric of Southern Italy and Daunian Middle Geometric Pottery at Coppa Nevigata (Foggia province, Italy), XIII-VIII Century B.C. In B. Fabbri (ed.) Fourth Euro Ceramics, vol.14. The Cultural Ceramic Heritage. Faenza: Gruppo Editoriale Faenza Editrice, pp. 77-88. Boccuccia, P., P. Desogus & S.T. Levi 1998 11problem.a dell'uso del tomio tra la fine dell'eta del Bronzo e la prima eta del Ferro: ceramica figulina da Coppa Nevigata (FG), Protovillanoviani e/o Protoetruschi. Ricerche e scavi, Preistoria e Protostoria in Etruria. Atti del Terzo Incontro di Studi, Manciano - Farnese, 12/14 maggio 1995, pp. 249-259. De Juliis, E.M. 1977 La Ceramica Geometrica de/la Daunia. Firenze: Sansoni Editore. Gorgoglione, M., G. Fiorentino, C. Corridi, L. Sadori & P. Panetta 1994 La capanna 7 di Torre Castelluccia (Pulsano, Taranto), dalle ultime fasi dell'eta del Bronzo alla prima eta del Ferro. Taras XIII, 1-2 (1993): 25114. Hulthen, B. 1974 On Documentation of Pottery. Lund. Jones, RE. & L. Vagnetti 1991 Traders and Craftsmen in the Central Mediterranean: Archaeological Evidence and Archaeometric Research. Proceedings of the Oxford Meeting, Oxford, pp. 127-147. Laviano, R, I. Muntoni & F. Radina 1995 Technological and Compositional Characteristics of the Matt-Painted Production in the South-Italian Late Bronze Age (Madonna del Petto, BarlettaBA). Estudis sobre Ceramica Antiga. Actes del simposio sobre ceramica antiga, Barcelona 18-21 novembre 1993, pp. 29-34.

Lo Porto, F.G. 1964 Satyrion (Taranto). Scavi e ricerche nel luogo del piu antico insediamento laconico in Puglia. Notizie degli Scavi di Antichita serie VIII, vol.XVIII: 177279. Malone, C. 1985 Pots, prestige and ritual in Neolithic southern Italy. In C. Malone & S. Stoddart (eds.) Papers in Italian Archaeology IV. Oxford: British Archaeological Reports International Series 244, pp. 118-151. Peroni, R 1967 Archeologia de/la Puglia Preistorica. Roma. 1989 Protostoria nell1talia Continentale. La Penisola Italiana nelle Eta de/ Bronzo e de/ Ferro. Rome: Popoli e Civilta dell'Italia Antica 9. Ruby, P. 1988 Les questions sous la tente: pour une approche technologique de la ceramique "a tenda". Melanges de !'Ecole Fran~aise de Rome 100: pp. 649-686. Taylour, W. 1958 Mycenean Pottery in Italy. Cambridge: Cambridge University Press. Vagnetti, L. 1982 Quindici anni di studi e ricerche sulle relazioni tra il mondo egeo e l'Italia protostorica. In L. Vagnetti (ed.) Magna Grecia e Mondo Miceneo. Nuovi Documenti. Taranto: Istituto per la Storia e l'Archeologia della Magna Grecia, pp. 9-40. 1994 Ceramiche · protostoriche del Mediterraneo: il contributo dell'archeometria alla definizione dei circuiti di scambio fra l'Egeo e l'Italia. 1st

European Workshop on Archaeological Ceramics (Roma, 1991), pp. 43-53. van der Leeuw, S.E. 1984 Dust to Dust: a Trasformational View of the Ceramic Cycle. In S.E. Van der Leeuw & A.C. Pritchard (eds.) The Many Dimensions of Pottery. Amsterdam, pp. 707-774. Yntema, D. 1990 The Matt-Painted Pottery of Southern Italy. Scienze dell'Antichita 4, Galatina.

92

Pottery production between the end of the Brome Age and the Early Iron Age in Daunia

6

GW I

I

8

7

'

) '

10

9

13

14

12

11

15

Fig.l: Coppa Nevigata (Foggia). Some sherds ofpainted.figulina ware from the south eastern area of the site. 1-3: Southern Italian Protogeometric; 4-6: Southern Italian Early Geometric; 7-15: Daunian Middle Geometric.

93

Defining Social and Symbolic Changes from the Bronze Age to the Iron Age through Operational Sequences in NW Iberian Pottery Isabel Cobas Fernandez & Pilar Prieto Martinez

INTRODUCTION

spatial entities, produced through social action, related to a socio-cultural context, and comprehensible within this :framework. Material culture is therefore considered as the product of a society combining both practical and symbolic aspects as, consciously or unconsciously, the concept of the material indicates a meaning.

The aim of this paper is to study the ceramic material culture :fromthe Bronze Age and the Iron Age in NW Iberia. We will try to define the patterns of formal regularity and their continuities and changes between the pottery styles of both periods. To do so, we will apply principles based on landscape archaeology, and comprehend the archaeological record as embedded within the social processes of the appropriation and semantization of space. In this case, all of the social dimensions (both material and ideal) are directly connected with the social material products which objectify these dimensions. We will examine how the stylistic trends revealed through the definition of the operational technological sequence of each period reflect these social conditions.

As social products created within a community are closely related to all of the material and imaginary areas of its reality, the characteristics and elements of a particular society are reflected in all of the facets of its material production, causing patterns of similar regularity and complementary relationships between codes, as proposed by Levi-Strauss (1987) for studying myths, and as applied by Criado {1993: 41) for the study of the archaeological landscape. Based on these principles, we defend relying on a structural line. Despite not working with modern societies and not having access to their language, meaning that in archaeology it is not possible to carry out a structural analysis as proposed by Levi-Strauss, this does not prevent the existence of hypotheses and procedures that may prove very useful :froma methodological viewpoint.

The main concepts used in this paper will be: style and operational technological sequence. This is implemented either as a methodological or interpretative tool: methodological as it allow us to deconstruct :from the final material product the technological processes related to the production of ceramic material culture, and interpretative as it gives us access to the social and symbolic dimensions which determine the decision-making process of this production. Style will be used in a broad sense, understood as a mechanism ~f the power discourse of a culture reflected in the formal products of its society. In the light of certain dominant opinions, we propose to understand style as a formalisation of power according to Foucault. The patterns of formal regularity are better displayed comparing material products belonging to two very different cultures :from the same geographical area.

Style The concept of style will be used in a broad sense, understood as a mechanism of the power discourse of a culture reflected in the formal products of its society. In the light of certain dominant opinions, we propose to understand style as a formalisation of power according to Foucault (1980). We therefore consider it important to assess how the patterns of formal regularity which exist between different codes of material culture may be defined within a given social group, making them coherent between themselves, and with a definite pattern of rationality: in short, identifying what can be referred to as a style (see Figure 2). However, it is not possible to arrive directly at the underlying conceptual schemes only by means of the identification of cultural choices. Accordingly, accepting the limits of interpretative practice in archaeology, our interest centres on identifying the matrix of theoretical possibilities, compiling all of the wide array of choices produced within an equally viable series of options, and extrapolating :from them elements of regularity. The method used to arrive at a definition of the matrix of possibilities is formal analysis, using the concept of the technical-operative sequence as a basis.

PRESENTATION The :framework of this study is an examination of ceramic cultural material :from the north-west of the Iberian peninsula according to the perspectives of Landscape Archaeology. It has a basically methodological value as the intention is to outline a new line of investigation for the study of different registers of material culture, based on the study of ceramics and landscape :from the Bronze and Iron Ages in this part of Spain (Figure 1).

Theoretical-methodologicalprinciples Formal analysis

The theoretical basis for this study is Landscape Archaeology (see Figure 2), according to which archaeological entities do not stand alone but instead are

Considering that ceramics, like any human product, are the formalisation of a particular rationality, we propose that

95

Isabel Cobas Fernandez& Pilar Prieto Martinez formal analysis be used for the study of material culture (see Figure 2). This consists of studying shapes and the relationship between them, understood not just as morphology and finished products, but as a unit of activities, ideas, premises and mental schemes brought together in their fabrication. Investigation takes place reconstructing the technical sequence used to make ceramic products.

• •

visible megalithic monuments in the landscape: dolmens and tumuli; invisible sites in the landscape: underground cists and pits.

Having carried out the technical sequence of the ceramics, we offer a summary of the most important formal features. The most important characteristic of ceramics from the Bronze Age in Galicia is of decorated material, bell beaker pottery and undecorated ceramics. Two types of morphologies are documented: simple and composite (see Figure 5).

The Technological-Operative Sequence

The concept of the technical-operative sequence (see Figure 3) involves three interwoven aspects: the strictly technical aspects, considered in what is effectively the technical sequence; those aspects which refer to social instances, which may be denominated the conceptual sequence, considering the economical, social, territorial and imaginary elements which merge together and indicate all of the object's process of fabrication; and finally the result of these two processes, defined as the final product: a group of formal characteristics which, determined by social demands, is registered within a stipulated pattern of formal regularity, in line with other codes, and as a reflection of the pattern of rationality to which it belongs.

• •

simple open or closed profiles and closed composite profiles are associated with undecorated ceramics; open composite profiles are, however, associated with bell beaker pottery.

Accordingly, we have a strong morphological opposition between plain and decorated beaker pottery. The characteristics of the decoration are as follows: • • •

By using the concept of a technical-operational sequence, the technological aspect is incorporated as a new style parameter, meaning that by following the technical processes we may have access to criteria which are less susceptible to changes, as indicated by Gosselain (1992). With this, we may both define the 'points of continuity' which allow us to talk of a style, and the breaking points which form categories. We therefore underline the active role of the object in ancient societies, and the existence of a variable cultural choice for each society (which Lemonnier, 1986, denominated a socially suitable technological choice), over and above external constraints.

•

the use of incision and impression techniques; decoration instruments are combs, punches and cords; the decorative elements are always geometrical and horizontal; body decoration on the object may be divided into segments, outlined, integral or in strips.

We will now characterise ceramics within each context (see Figures 6 and 7). In a domestic context the morphological opposition is maintained between plain and decorated pottery, although this opposition is reflected in other aspects of the phases of production: the finish, colour, size of the gritting, the accessories and, above all, the final product. Nondecorated ceramics are roughly finished, with large gritting, and light, matt colours which are practically invisible. Bell beaker pottery, however, has a better quality finish, with smaller gritting, and lively, visible colours in various tones. The decoration displays a certain variety enabling us to group it into four substyles.

INTERPRETATION

We continue with a summary of the landscape and ceramics of the Bronze Age, then refer to the characteristics of the Iron Age, and finally examine the relationship which exists between both periods.

In a funerary context, differences may be seen according to the type of burial. In visible burial sites, or megaliths, the pottery is similar to that found in domestic contexts, a formal opposition between undecorated and decorated ceramics, which are always beaker shaped. Differences are only found in two phases of the technical sequence: in the modelling and application of the decoration. The other production stages use the same technology: a good quality finish, small gritting sizes, bright, light colours and a lack of accessories complementing the profile. This leads to both types of ceramic having a good visual appearance, with a visible character.

The Bronze Age

We shall firstly refer to the contexts in which Bronze Age ceramics appear, and then discuss the general characteristics of these ceramics. Three types of contexts are known in Galicia: ritual, with petroglyphs (see Santos 1996 for a summary), and domestic and funerary (see Mendez 1994 for a summary). Ceramics are only registered in the two latter contexts. The domestic context (see Figure 4) is characterised by settlements which are invisible in the landscape and have: perishable constructions (holes for posts, foundation trenches for huts), divided arable land nearby, and defined wetlands suitable for grazing. The funerary context (see Figure 4) is formed by two main groups according to their visibility in the landscape:

Ceramics from invisible burial sites differs between cists and pits. In cists only undecorated items appear, with two morphologically differentiated groups: one with simple,

96

Operational sequences in NW Iberian Pottery

open morphologies, and another with closed compos~te morphologies, well finished and with bright, clear, visible colours. The differences are only found in one production stage: the modelling phase.

relation to the Bronze Age is the disappearance of the separation (at least visual) between the domestic and nondomestic landscape. This is due to the application of a strategy whereby the symbolic space was completely hidden (both funerary and ritual), resulting in a complete lack of examples in the archaeological record. On the contrary, an apparent strategy of exhibiting the domestic landscape characterised by the monumentalisation of domestic space is used, represented by the concept of the hill-fort not merely as a focus of population, or even as a fortified area, but as a major achievement. Instead of merely being situated in the landscape, the hill-fort organises it and gives it a conscious and ostensible meaning, and converts the landscape into territory by imposing a cultural order upon it (see Parcero 1995 for a review).

In pits, undecorated or decorated pottery may be found, although these are not beaker shaped. There are two well defined morphological groups: one with simple, open morphologies, and another with complex, closed morphologies. Both use matt, dark tones, with rough, invisible finishes. The differences are only found in one production stage: the decoration phase. If we compare the ceramic record of both contexts, we can underline the following aspects: (1) decorated ceramics in a domestic context are always beaker shaped, while in funerary contexts they may be beaker shaped or of another style; (2) the strong opposition between plain and decorated ceramics in a domestic context is diluted in a funerary context as more homogenous phases exist: finish, colour and gritting: (3) the existence of an association between the type of undecorated utensils and the type of funerary architecture which does not exist in a domestic context; (4) the decoration is more varied in a domestic context, giving rise to substyles, whereas in a funerary context it is more standarised. Differences may be seen in the following features: •

•

With regard to Iron Age ceramics from this region, and focusing on the domestic context, we should point out that once the ideal technical-operative process is reconstructed, separate from the varieties and internal categories which we will not consider here, a series of regularities may be identified in their formal configuration (see Figure 10). The most significant characteristic is that unlike Bronze Age ceramics, the opposition between decorated and plain ceramics lies only in the presence or lack of decoration, as it does not imply a different treatment during the technicaloperative process ( see Figure 11). We may therefore say that in general they share: modelling (as all of the recipients are hand made using a potter's wheel, and a duality is maintained in the object's morphology between simple profiles, which may be open or closed, and complex profiles, which are always closed, although divided into flexed and edged shapes; finishing techniques (as they use burnishing, combing, spatula designs and burnishing); type of firing (with a predominance of the use of an oxidising atmosphere) and surface colouring (with dark colours, either dark brown or black).

in a domestic context, instruments are used for decoration, which do not appear in a funerary context (shells, nails, fingers); in a domestic context, the decoration is horizontal, integral and flowing, and the most significant decorative element is the horizontal line. In a funerary context, the decoration is also horizontal but may be integral and in stripes for bell beaker pottery, and zonal and flowing for ceramics that are not beaker shaped. The most outstanding decorative elements are zigzags, reticulated shapes and oblique lines.

With reference to decoration (see Figure 12), it is possible to isolate a series of common features throughout all of the Iron Age. For thematic issues, the constants refer to decorative elements which are always geometric, and to perfectly independent decorative motifs in the same item, without the existence of an element which acts as an axis of what is represented. It may be said that more emphasis is given to the individual motif than in the general theme. For morphological issues, various aspects may be differentiated:

In summary, the rupture points within the technological process vary according to the contexts (see Figure 8): whereas in a domestic context the differences between plain and decorated items may be seen in five phases of production, in the funerary context these are reduced to two phases of production (with the exception of the technical process for ceramics in graves, which is totally different to the other ideal technical processes). Two types of relations exist in the visibility strategies between the landscape and the ceramics which in some cases are represented by way of an opposition between both codes, and in other cases by complementing each other.

•

• The Iron Age For this period, we found different characteristics from the Bronze Age both in its landscape and ceramics. For the landscape (see Figure 9), and following the scheme proposed by Parcero (1995), the greatest difference found with

97

'verticality' of orientation, as the combination between orientation, disposition, distribution and the decorative scheme is executed so that the decoration is always vertical, without needing to turn the object; divisory distribution which instead of uniting the profile of the object divides it, and makes its different parts independent from one another. This may be either by giving special attention to a single area on an otherwise undecorated piece, or individually detailing each section using different decorations;

Isabel Cobas Fernandez & Pilar Prieto Martinez

•

fields, the difference between types of finish and decorative styles, etc.), and 'territorial character' through the divided and differentiating nature of the decoration with regard to the piece as a whole. This is seen in the predominantly successive decorative scheme, in the distribution and composition of the motifs which is different in each of its parts, and in the perfect definition given to the majority of the decorative motifs using straight horizontal lines.

location in a visible area, as even though examples exist of relative invisibility, for example in the case of decoration which appears in the upper part of the rim, no invisible decoration as such exists, as neither the interior nor the bottom part of the piece ever appear with decoration.

However, within this regularity it may be seen that these ceramics respond to a system of categorisation based on ternary series. These take shape in a series of complex relations and combinations of choices among a wide range of varieties in each step of the decorative technical process. Three groups may therefore be identified (see Figure 13): undecorated items and items with simple decoration, which share the same technical operative process; undecorated items and items with complex decoration sharing the same technical operative process, and a group of items with complex decoration which belong to a different technical operative process.

CONSEQUENCES A difference in styles may be seen between these two periods, despite the existence of common cultural choices in both (see Figure 15). These differences are related to a conscious intent, and differentiated cultural standards in each of them. The changes from one period to another are evident in the two codes of material culture which were studied. For the landscape, a change occurred from a clear separation between the domestic and non-domestic environment in the Bronze Age, to a less apparent or even inexistent separation of these in the Iron Age. For ceramics, although a type of decoration exists in both periods which received a differential treatment (bell beaker pottery decoration in the Bronze Age and stamped decoration in the Iron Age), there is a change between the Bronze Age, with a clear definition between decorated and undecorated pieces, and the Iron Age, where the technological treatment given to decorated and undecorated ceramics is the same.

Accordingly, we may propose a hypothetical model which includes three differentiated groups. Firstly, ceramics with scarcely specialised shapes including undecorated and decorated pieces with only moderate differences (as they share their morphology and technological treatment, although there are exceptions), where the decoration is ambiguous due to being very simple, barely visible, either unemphasised or with very little emphasis, or zonal. Secondly, we have a more specialised group of ceramics which we may refer to as non-domestic, with a probably more limited and specific nature than the previous. It is characterised by having more decoration, which is more complex, clearly visible and defined, and even integral in some cases. Within this second group we may differentiate between pieces which have a common technical operative process with both decorated and undecorated ceramics, and pieces which are always decorated and belong to a specific technical operative process (ceramics with an 'S' shaped profile with mainly stamped decoration and a varied decorative pattern).

BIBLIOGRAPHY Cobas Fernandez, M' I. 1995 Bases metodo/6gicas para la descipci6n y el

estudio formal de la ceramica de/ yacimiento de Alto do Castro (Cuntis, Pontevedra). Unpublished third year investigation. Santiago de Compostela: Facultad de Xeografia e Historia. 1997 Estudio de la ceramica castrena de/ yacimiento de Alto do Castro (Cuntis, Pontevedra). Unpublished thesis. Santiago de Compostela: Facultad de Xeografia e Historia. Cobas Fernandez, M' I., C. Gonzalez Perez & M' P. Prieto Martinez 1995 The Pots Database: an example of sistematisation on pottery research. Archaeological Computing Newsletter 4: 1-5. Criado Boado, F. 1993 Visibilidad e interpretaci6n del registro arqueol6gico. Trabajos de Prehistoria 50: 39-56. Foucault, M. 1980 Poderes y estrategias. In La microfisica del poder: 163-74. Madrid: La Piqueta. Gosselain, 0. P. 1992 Technology and style: potters and pottery among Bafia of Cameroon. MAN (New Studies) 27: 55986.

It: finally, we relate the ideal model of the landscape with the ideal model of the ceramics, a coherent formal regularity is observed between both, which could be understood as an inverted structural relationship (see Figure 14). In the landscape a strategy of exhibition of domestic space exists, by means of the territorialisation of the surroundings and the monumentalisation of the hill-fort. This is transformed into an authentic landmark which organises and gives meaning to the landscape, combined with a strategy of concealment of the ritual space. On the contrary, a strategy of inhibition or even concealment is used for domestic ceramics, and a strategy of exhibition for non-domestic ceramics, in which the elements of the occupied space are reproduced: its monumental and territorial character: 'monumental character' through the existence of a precise decorative style, with a clear intention of visibility (the greater expansion of decoration in the piece, outlining of decorative

98

Operational sequences in NW Iberian Pottery

Lemonnier, P. 1986 The study of material culture today: toward an anthropology of technical systems. Journal of Anthropological Archaeology 5: 147-86. Levi-Strauss, C. 1987 Antropo/ogia estructura/. Barcelona: Ed. Paid6s (Primera edici6n, 1974: Anthropo/ogie structurale, Buenos Aires). Mendez Fernandez, F. 1994 La domesticaci6n del paisaje durante la Edad del Bronce gallego. Trabajos de Prehistoria 51 n° 1: 77-94. Parcero Oubifia, C. 1995 Elementos para el estudio de los paisajes castrefios del noroeste peninsular. Trabajos de Prehistoria 52 n° 1: 127-44. Prieto Martinez, M'.P. 1993 Aproximaci6n al ana/isis formal de la ceramica de la Edad de/ Bronce en Galicia. Trabajo de lnvestigaci6n, inedito. Santiago de Compostela:

Departamento de Historia I, Facultade de Xeografia e Historia, Universidade de Santiago de Compostela. 1995 Definici6n de un sistema metodol6gico para el estudio de la ceramica de la Edad del Bronce en Galicia: La tradici6n campaniforme del yacimiento de A Lagoa (Toques, A Corufia). In Actas de/ XXII Congreso Nacional de Arqueologia (Vigo, 1993); vol. II: 17-24. Vigo: Concello de Vigo, Conselleria de Cultura e Xuventude, Xunta de Galicia. 1996 Estudio de la ceramica de/ segundo milenio a. C. de la Sierra de/ Boce/o y el Occidente gal/ego. Unpublished thesis. Santiago de Compostela: Facultade de Xeografia e Historia, Departamento de Historia I, Universidade de Santiago de Compostela. Santos Estevez, M. 1996 Los grabados rupestres de Tour6n y RedondelaPazos de Borben como ejemplos de un paisaje de petroglifos. Minius V: 13-40.

()

•

200 Km

Fig. 1: Galicia in the Iberian peninsula

99

Isabel Cobas Fernandez & Pilar Prieto Martinez

LANDSCAPE ARCHAEOLOGY THEORETICAL BASIS

and ceramic material culture studies

Methodology

Sludylng the archaeological entitles as the forms produced by social action In relation 1D their soclocultural context

Identifying: theoreticalmatrlxes formal regularttles

FORMAL ANALYSIS

STRUCTURAL ANALYSIS

STYLE

Definingthe relationships among codes of material culture

TECHNOLOGICAL SEQUENCE OF OPERATIONS

Fig. 2: Theoretical-methodological basis to the material culture studies

-

Raw materialpreparation 1-----~

Technical sequence context

,.

..

Non-rigid materfal

To obtain the raw

Material culture

Clay

preparaUon

1

I

Product manufacture

Final Product

AlternaUngprocce-

I

,--D..1.rle_d_,,-Fi_nl_.,sh_e_d.,,-D-ec_oratl.._on__, Function

material

[]

Individual

Social group

Context:space and time

Social context

Cultural patterns

Social patterns

Imaginaryworld

Territorial nature

-

Fig. 3: Technological sequence of operations

100

Economy

Operational sequences in NW Iberian Pottery

...... L....\~:;, ♦

,----,.

...•

.••••

: ••••

Q•

. . . . ... . . . . . . . . .. ·....

.......... . . ·o.·.

.

1,1,;; . . . .,; • . ., : ..

't

.

. ...

. . . . . . .. .... . . ...

Funer-, CGlllut

r,_

~

Cillo

... Fig. 4: Brone Age.contexts in Galicia

Simple morphologies OPEN SHAPE

Complex morphologies

CLOSEDSHAPE

OPENSHAPE

CLOSED SHAPE

OJmrn mm mrn QJ mw \[/~

CD

CI]~

D]~ GJOJ ID GJ

WCJJ UJ

ffi~OJ OJ~ wmw

m(D Ww~

U)ID

ID(l)·~

\,:.-\

-.. .......

.

!..

·.

!

'•

...

\_

Figure 1: Map ofNorway with an inset map showing the area in the easternmost part of the two counties ofTrnndelag in Mid-Norway where the iron production sites are located (Inset map after Stenvil< 1991c).

121

Kristin Prestvold

Figure 2: Aerial view of the three furnaces and slag heaps comprising the production site at Fjergen in Nord-Trnndelag. The shallow pits surrounding the furnaces give the impression of rosette shaped patterns. One of the slag heaps was thoroughly excavated as shown in the photo (Photo: K. Prestvold)

(---------

',.__). @(-~

([J (/\ ,

.

,

/

•

0

0

.

'

\.,.__,/

,

' '

Figure 3: Plan of the iron production site at Fjergen. The furnaces or slag pits are in the middle, surrounded by the shallow pits that make these production sites so characteristic. Notice the postholes (marked in black) associated with the surrounding pits (Drawing: K. Prestvold & S. del Bjerck)

122

A story of early large scale production of iron in mid-Norway

Figure 4 and 5: The remains of the furnaces today; the slag pit from one of the furnaces at Fjergen in Nord-Troodelag. The

first photo shows the slag pit after excavation but before the layers of mixed clay and stones covering the original wall surface have been removed. The second photo shows the original wall. The repairs of the slag pit, among other factors, attest to these being permanent constructions and the pits give the impression of solid and professional craftsmanship (Photo: K Prestvold)

123

Kristin Prestvold

Figure 6: One of the :furnaces at Fjergen in Nord-Trnndelag before excavation. Burnt clay fragments and charred stones from the last furnace shaft spread over and cover the immediate vicinity and top part of the slag pit (Photo: K. Prestvold)

Figure 7: The slag heap below one of the :furnaces at Fjergen in Nord-Trnndelag before excavation. These heaps spill out over the terrace and the sloping ground and usually consist of between 10 and 20 tons of slag each (Photo: K. Prestvold)

124

The Reconstruction of Manufacturing Sequences on the Basis of Iconography: The Case of the Foundry Cup at Berlin Gabriella Prisco & Massimo Vidale

Achilles. The goddess leans with her right arm against a long spear, while on her left arm she bears a shield, and she is looking at the helmet offered by Hephaistos. The divine smith has just finished using the hammer he is carrying in his right hand. Between the two gods, suspended in the background, is a couple of schinieri, and nearby a second hammer hangs upside down. Hephaistos is offering his finished product to Tethis exactly like the craftsmen, on the opposite side, who are delivering their statue to people of a superior social status. We are doubtless at the end of the operational sequence. The images therefore establish a precise analogy between Hephasitos and the bronzesmiths, an analogy whose function is to nobilize the metal craft.

INTRODUCTION

The study of the vessel nowadays known as the "Foundry cup" begun with a complete misunderstanding of the manufacturing cycle it represented. The exterior of the vessel (Fig. 1) in fact showed a horizontal human figure with a severed head lying near a tall furnace crowned by a cauldron; in the first half of the last century, immediately after its discovery, this was interpreted as a scene of butchering and cooking of human flesh. The vessel was therefore called the Coppa dell 'Antropofago (literally "The Cannibal's Cup") and was considered to represent reliable proof of the fact that the ''Etruscans" were cannibals (in those days the Greek ceramics found in Tuscany were commonly ascribed to Etruscan craftsmen) (Hamilton Gray 1840). As we shall see these errors, compared with others, were easy to correct. The vessel is an Attic kylix datable to around 490-480 BC. It was found in one of the graveyards at Vulci during clandestine excavations carried out with antiquarian goals, and thus no information is available on the actual context of the find (Buranelli 1992). We may presume that a vessel depicting a metallurgical workshop could have been appreciated by the rich families of Vulci, a town that was intensively involved in the production and trade of metal commodities. After its recovery, the cup was sold and transported to Berlin. It is the only iconographic document we have which shows the manufacture of large bronze statues. For this reason it has been frequently discussed or mentioned by art historians, archaeologists and paleotechnologists (among others, see Gerhard 1840; Furtwangler 1885; Bliimner 1886; Furtwangler & Reichhold 1904; Thompson 1964; Burford 1972; Ziomecki 1975; Mattusch 1975, 1980, 1982, 1988, 1992, 1996; Griefenhagen 1978; Zimmer 1982; Oddly & Swaddling 1985; Beazley 1989; Haynes 1992; Heilmeyer 1993). Nonetheless, there have been many doubts and major mistakes made on basic issues: which part of the manufacturing sequence is actually depicted on the vessel? And is there any reliable information on the forms of labour organization adopted by the craftsmen portrayed on the vessel?

ASSIMILATING CRAFTSMEN TO GODS

The least controversial part of the painted image is the inner tondo. It shows the god Hephaistos in the act of giving Tethis the wonderful weapons he manufactured for her son,

SENIOR CRAFTSMEN AND APPRENTICES

The exterior of the kylix (Fig. 1) has been traditionally subdivided into two halves. On the first halt: which is the longer one because its figures protrude below one of the two horizontal handles of the vessel, we observe two senior craftsmen, distinguished by their beards, and two young apprentices, naked and with an ephebic look. The scene begins near a high vertical furnace, topped by a globular cauldron with a series of lids with progressively smaller diameters. Below the furnace one of the apprentices, a boy, is busy with the bellows. In front of the furnace a senior craftsman, bearing the typical cap of the smith, is stoking the mouth of the furnace with a long bar which ends in a hook. On the right, the other apprentice turns to observe his master, as if distracted from his work; he is leaning on the long handle of a hammer. Near the boy, in the background hang a foot and a hand. The second senior craftsman is busy with his hammer on the headless, naked statue of a young man, lying in a horizontal position on some kind of irregular basement or support. The head of the statue pops out between the legs of the standing craftsman.

A FURNACE-FORGE, SOME PINAKES AND OTHER OBJECTS

The interpretations so far advanced on the actors and the elements of this scene are varied and often contradictory. Some scholars thought that the exterior of the cups represented a large part of the manufacturing sequence of bronze life-size statues, from the melting of the bronze to the finishing of the surface (on the opposite side of the exterior) within a single, large workshop employing five or more craftsmen (for example Seki 1985: 69). Others thought

125

Gabriella Prisco & Massimo Vidale

that the vertical :furnacewas charged with layers of charcoal and metal, ready to be melted and poured into a buried mould (Kluge 1929; Charbonneaux 1958; Beazley 1989; Mattusch 1982, although a different interpretation by the same author appears in Mattusch 1996; Rolley 1994). Pictures of bronze workshops dating to the beginning of the century suggest that there was no lack of ethnoarchaeological evidence for supporting this hypothesis (Kluge 1929: Abb. 2-4).

The hand and the foot which appear in the background have been considered by some authors as proof of the use of finished models to be reproduced by means of negative moulds, according to the method of indirect lost-wax casting (see Formigli 1984). These objects might also be parts of finished statues, or · they might allude to a metrological systems used by the craftsmen.

WORKING AT THE HORIZONTAL STATUE

It was thought that the upper cauldron represented some kind of chimney, but it is probably a closed container made of terracotta or bronze. It may be slightly uplifted from the top of the :furnace, so as to allow an escape for the gases. Other scholars saw it as a crucible, where metals such as tin or lead could be melted; others suggested that it contained acid solutions for removing oxidized patinas, or oils for quenching iron or, again, a vessel for melting wax. Today, most scholars agree that the tall :furnacerepresents a kind of forge; in fact, in most of the other vessels depicting the same facility, it accompanies ironsmiths or helmet-makers (see, for the various opinions, Oddly & Swaddling 1985; Schwander & Zimmer 1983, including a contribution by U. Zwicker). In 1991 E. Formigli built an experimental replica of this forge and successfully used it for melting small amounts of bronze within crucibles and for heating iron objects to be shaped on the anvil (Formigli 1983: 109-122). The upper vessel was made of terracotta, but it was of no practical use. Formigli's ideas was that the cauldron could have contained water, and that by regulating the lids the craftsmen would have been able to perform some kind of foresight and control on the times of some operations.

The naked, headless statue of a young man with raised arms has so far defied every attempt at finding useful iconographic comparisons. In the past some scholars thought that this figure was part of a single group together with the colossal statue of an attacking hero on the opposite halt: The theme of a young boy attacked by an armed warrior might recall some scenes of Ilioupersis (the sack and destruction of Troy) with the killing of Astyanax or the ambush against Troilos. But the iconographies of these two famous episodes are definitely quite different. Some scholars have viewed the horizontal headless figure as the portrait of a runner at the very moment of the start (Thompson 1964: 326; Burford 1972: 95), an athlete (Harrison 1985: 41), a jumper (Richter 1930: 70), or a wrestler (Mattusch 1980: 442-443); we have also tested, always on iconographical grounds, the hypothesis of a horse raider and, again, possible identification but Daedalus and Prometheus, but each time with no positive match. Who was the personage portrayed by this imaginary bronze? The problem is still unresolved. What seems to be certain is that the tall :furnace-forgein the picture was used for melting the alloy prepared for soldering the head to the torso. From a technical viewpoint, the casting of bronze statues in separate pieces to be soldered together is well known in the Vth century BC. The casting of heads with an oblique joint at the base of the neck is attested by the two famous bronzes from Riace and other contemporary specimens. The subtle line which runs through the wrist of the right hand of the horizontal figure is also coherent with paleotechnological evidence. The bearded craftsman seems to be hitting part of the statue with his hammer. The latter rests on a irregular support, whose edge projects itself outward, a feature contrasting with the prevalent interpretation of this element, i.e. a heap of sand or clay meant to hold the statue during the assemblage of the various parts. In contrast, we think that this support closely resembles the graphic conventions applied by the Foundry Painter when, on other vessels, he draws rocks. But against this suggestion there is the evidence that a rock would fit a finished work, and it is clear that in this case the painter wanted to represent an unfinished statue. On the other hand, the head on the ground clearly shows the eye and the iris, ideally inserted before mounting and soldering the head to the torso: a lack of technical congruence, because no craftsman would have run the danger of spoiling the eyes,

Behind the :furnace-forge we see a series of pinakes, i.e. small tablets or pictures clustering around two heads or masks; this group of objects is crowned by a couple of goat's horns, and embellished by some vegetal garlands. These objects are probably votive offerings, and may be inspired by dyonisiac believes or rituals. On the other hand, they could have an apothropaic :function:to avoid evil events during the delicate phases of heating or melting metals (see Gerhard 1840: 13; Thompson 1964: 326; Oddly & Swaddling 1985: 13; Mattusch 1980: 436, note 13; Mattusch 1986: 182; Burford 1972: 122-123). The two heads, one male, bearded, and the other female, might be allusions to Hephaistos and Athena, the two major divinities of craftspersons. But the heads could also be parts of statues, and the painted tablets could be models or sketches of bronze statues designed or fabricated by the artisans (Beazley 1989); the apothropaic and votive meaning of these objects could perhaps go together with the more practical will of illustrating the high technical standards of the bronzesmiths. In this light, it is interesting to observe that the designs on the tablets refer to images which can be easily opposed with categories such as animal-human, male-female, standing-sitting-running, as if the craftpersons were expressing their capability of creating every possible kind of bronze statue.

126

The reconstruction of manufacturing sequences on the basis of iconography

whose application required time and care, with the rougher handling involved in soldering.

craftsmen are scraping its surface with the goal of removing the most apparent defects, irregular casting spills or crests, and the casting skin, the outer dark peel accompanying cast bronzes. But it is impossible that such a finishing stage could take place after having mounted and applied parts such as the eye, the hair locks, the helmet and the shield. This is doubtless another technical inconsistency, to be explained only by a limited understanding by the painter of the technical details of the process he depicted, or alternatively it may have had a precise explanatory purpose, to make it clear to the onlookers that this second statue is finished and ready to be delivered, while the former was.not.

SENIOR CRAFfSMEN AND AFFLUENT CITIZENS The opposite half shows the same couple of bearded senior craftsmen, clearly identified by the same cap and loincloth, while they finish a colossal heroic statue of a young oplite which occupies the centre of the panel, framed within a large wooden structure. The frieze-like continuous structure of the first half is now replaced by a concentric, metope like graphic arrangement, focusing on the finished statue. In this half we no longer see the young apprentices, but at the two ends, in a symmetric position, there are two bearded mature men, elegantly dressed, wearing comfortable shoes and fillets and walking sticks; beside them, hanging in the background, one sees two strygils with two aryballoi, the oil flasks used for cleaning the body after sport and physical exercise. One of them is extending his arm, perhaps in a gesture of command, and speaks to the bronzesmiths.

In spite of its colossal size, the statue of the warrior does not represent the god Ares, but the strength and the value of a Homeric-like hero (Stewart 1990: 134). It is hard to deny the precise will by the painter of depicting two completely different, if not opposed, bronze statues: one life-sized, in a dynamic posture, horizontal, naked, which somehow reminds us of the statuettes of athletes which were to become more common during the Vth century BC; the other double life-size, in a stiff posture, vertical, furnished with clothes and weapons, closely reminding us of the conventional representations of warriors by the painters of the Archaic period (Mattusch 1996). Perhaps this choice was aimed at expressing the capability of the craftsmen of creating statues in quite different styles.

In the past, these two figures have been interpreted as the owners which, after cleaning themselves with oil, were giving the last instructions to their craftsmen before leaving the workshop (Thompson 1964; Burford 1972). A development of such an interpretation saw the two figures as the sculptor and the caster (Beazley 1989; Stewart 1990: 33; Mattusch 1990: 441 ), a dichotomy echoing famous associations such as that of the couples of Rhoikos and Theodoros, or Kritios and Nesiotes, recalled by the ancient writers. More recently, another interpretation has gained some support, according to which the two figures symbolize two of the Eponymous Heroes of the original Athenian tribes (Mattusch 1996: 84). A simpler possibility is that they represent the affluent citizens of Athens, perhaps in the role of the persons or institutions that ordered the statue or paid for it, or perhaps as epistates, i.e. the appointed superintendents or managers responsible for a correct administration and execution of the work. The oil flask and the strygil stress, in our view, their total extraneity to the manual labour and the workshop. There might be a conscious, formal opposition between the world of the palaistra, the place reserved for the performance of sports so dear to the higher classes, and the workshop: the respectable citizens sweat because of the pleasant effort of physical exercise, while the craftsmen sweat under the hot blow of the forge.

The wooden infrastructure surrounding the colossal statue, consistent with the idea of the "large workshop", has often been interpreted as the "door" of the room (for example Mattusch 1980: 443). In this period, however, doors were as a rule represented with quite different graphic conventions. According to a second interpretation, it represents a system of scaffolds supporting the statue and allowing the craftsmen to access its upper parts by means of ladders (Harrison 1985; Mattusch 1996). This second interpretation seems more likely, in particular in the light of the inscription with the report of the expense accounts for the Hephaisteion of the agora. This inscription, dealing with the expenses for the manufacture of the two larger than life-size cult statues of Athena and Hephaistos, clearly mentions the wood used for protecting the basement and the statues, the interior of the temple, and the wooden scaffoldings erected around the statues (Thompson 1969: 114).

ORGANIZATION OF LABOUR AND RELATIONSHIPS OF PRODUCTION In the past, some scholars counted between six and eight different craftsmen on this vessel, and proposed that it showed a "large industrial workshop" where various statues could be manufactured at the same time. It was also thought that the scenes referred to the main stages in the manufacturing sequences, namely melting and downpouring of the alloy, assembling the various parts and finishing the surface. The anonymous "Foundry Painter" was considered

THE CENTRAL FINISHED STATUE The two craftsmen in the central "metope" appear to be almost overwhelmed by the size of the wonderful statue they are finishing. They are scraping its surface with two rasps whose shape curiously reminds us of the strygils of the two "citizens". If we accept the idea of a finished statue, just released from the outer mould, we might think that the two

127

Gabriella Prisco & Massimo Vidale

to be a "Greek realist" because of his supposed realistic representation of the technical details. In contrast, in our view there is no doubt that there are only two senior craftsmen on the exterior, portrayed twice, and two young apprentices. Therefore the two technical episodes depicted on the two halves of the cup had been conceived by the painter as not being contemporaneous: we are not dealing with a "large workshop" but rather with only two craftsmen who were capable of making different types of statues. The first half: in fact, shows an unfinished statue in the context of the collaboration between senior craftsmen and young apprentices. There was probably no intention of illustrating the flow, or the various stages, of a craft process; the main concern of the painter was to attribute a well defined role to each actor, and to express the hierarchy of these roles. The whole vessel is permeated with a hierarchical ordering of four couples of actors: at the lowermost level, the two young apprentices; the two senior craftspersons; the two "affluent citizens"; and finally, on the interior of the vessel, the two divinities, which grant a superior assimilation of the craft. The two technical episodes depicted on the exterior are completely independent from each other, and do not exemplify any manufacturing sequence: these scenes were selected only on the basis of the opposition between ''unfinished" (not prestigious and ideally associated with the work of the apprentices) and "finished (prestigious, associated with the idea of divine protection and with the affluent citizens). All the attention of the painter was centred on expressing this hierarchy efficiently, and perhaps this is why we find "technical errors" in his (or her) designs. The senior craftsmen are the only ones to be represented twice, and appear to be the main characters on the vessel.

the unfired vessel in a leather-like state of hardness. In this case we observe another ''technical error": the turning of a vessel like a kylix, made by assembling together parts which had to be thrown separately, had to take place on the individual parts, before assemblage, and not after, and certainly before the application of the handles. As we have seen in the Foundry cup, in this second part of the vessel the apprentices are absent: instead a finished object is presented to a fully dressed figure representing an affluent citizen. He is a young man, coated with a heavy cloak with a red border. His gesture is identical to that seen in the first vessel: his mouth is open, and his arm is raised in a gesture of command. It thus appears that the match between the iconographic structure of the two vessels is very close. Even if only two cases are known, we may state that we are dealing with a recurrent scheme. In these and other representations of craftspersons, it is clear that the illustration of technical events simply provided the occasion for re-asserting the essence of the relationships of production among apprentices, senior craftsmen and affluent citizens, under the seal of divine authority. When we attempt to reconstruct ancient techniques on the basis of similar iconographies, the first necessary step is a careful evaluation of the ideological and functional context in which the images were produced and exhibited; and our experience suggests that such reconstructions will often be difficult and uncertain.

REFERENCES Beazley, J.D. 1989 A Greek Realist. In D.C. Kurt (ed.) Greek Vases. Lectures by J.D. Beazley. Oxford, pp. 78-83. Blfunner, H. 1886 Technologie und Terminologie der Gewerbe und Jndustrie Griechen und Romern. Leipzig. Buranelli, F. 1992 Gli scavi di Vulci della Societa Vincenzo Campanari - Governo pontificio (1835-1837). Roma. Burford, A. 1972 Craftsmen in Greek and Roman Society. London. Charbonneaux, J. 1958 Les Bronzes Grecs. Paris. Formigli, E. 1984 La tecnica di costruzione delle statue di Riace. In Due Bronzi da Riace. Bollettino d'Arte 3, Vol. I. 1993 Antiche Officine del Bronzo. Siena. Furtwangler, A. 1885 Beschreibung der Vasensammlung in Antiquarium. Berlin. Furtwangler, A. & K. Reichhold 1904 Griechische Vasenmalerei Vol. I. Munich. 1910 Griechische Vasenmalerei Vol. II. Munich. 1931 Griechische Vasenmalerei Vol. Ill. Munich.

A RECURRENT SCHEME This interpretation is supported by a second vessel, as famous as the Berlin cup: the cup at Karlsruhe (Metzler 1969; Ziomecki 1975: Figs. 2, 3; Schleiber 1983: Abb 68). This black-figure vessel shows two small, almost minimalist scenes on the exterior. On one side of the exterior (Fig. 2), a naked senior craftsman, distinguished by his beard, is busy at a huge potter's wheel. He has centred a large cylinder of clay, and is raising the walls of a (presumably) large vessel with some effort. In front of him is a young apprentice, also naked and sitting on a low base, pushing the wheel with his right hand in a clock-wise direction. His eyes seem to be fixed on his master, in an attempt at co-ordinating his own work with the throwing movements of the potter. On the opposite side (Fig. 3) another craftsman, this time dressed, has completed a kylix. The vessel sits on a cylindrical support centred on the potter's wheel. The craftsman is holding a tool in his left hand which is painted in white, a colour that often indicates the shining of metal. This scene could therefore refer to the scraping or turning of

128

The recomtruction of manufacturing sequences on the basis of iconography

Gerhard, E. 1840 Trinkschalen und Ge/assen des koniglichen Museums zu Berlin. Beschreibung der Vasensammlung in Antiquarium I, Berlin. Griefenhagen, A. 1978 Griechischen Gotter. Bilderhefte der Staatliche Museen Preussicher Kulturbesitz Heft 10, Berlin. Hamilton Gray, Mrs 1840 Tour to the Sepulchres of Etruria in 1839. London. Harrison, E.B. 1985 Early classical sculpture: the bold style. In C.J. Boulter (ed.) Greek Art, Archaic into Classical. Leiden, pp. 40-65. Haynes, D. 1992 The Techniques of Bronze Statuary. Mainz am Rhein. Heilmeyer, W.-D. 1993 Progresso tecnico nella fusione dei bronzi di eta classica? In E. Formigli (ed.) Antiche Officine de/ Bronzo. Siena, pp. 13-28. Kluge,K. 1929 Die Gestaltung des Erzes in der ArchaischGriechische Kunst. Jarbuch des Deutschen Archiiologische Imtituts 44: 1-15. Mattusch, C.C. 1975 Casting Techniques of Greek Bronze Sculpture: Foundries and Foundry Remaim from the Athenian Agora with Reference to Ancient Sources. Ph.D. Dissertation, University of North Carolina, Chapel Hill. 1980 The Berlin Foundry Cup. American Journal of Archaeology 84, 435-444. 1982 Bronzeworkers in the Athenian Agora. American School of Classical Studies at Athens, Athens. 1988 Greek Bronze Statuary from the Beginning through the Fifth Century BC. Ithaca and London. 1992 The Production of Bronze Statuary in the Greek World. In G. Hellenkemper Salies, H.-H. von Pritziw & G. Bauchness (eds.) Das Wrack. Der antike Schifftfund von Mahdia. Koln, Band II, pp. 789-800. 1996 Foundry Cup. In C.C. Mattusch (ed.) The Fire of Hephaistos. Large Classical Bronzes from North American Collectioru. Cambridge (Mass.), pp. 182184.

129

Metzler, D. 1969 Eine Attische Kleinmeisterschalen mit Topferszenen. Archiiologische Anzeiger LXXXIV: 138-149. Oddly, W.A. & J. Swaddling 1985 Illustrations of metalworking on Greek vases. In P.T. Craddock & M.J. Hughes (eds.) Furnaces and Smelting Technology in Antiquity. British Museum Occasional Paper no. 48, pp. 43-57. Richter, G.M.A. 1930 The Sculpture and Sculptors of the Greeks. New Haven. Ridgeway, B.S. 1977 The Archaic Style in Greek Sculpture. Princeton. Rolley, C. 1994 La Sculture Grecque. 1, Des Origines au Milieu du Ve siecle. Paris. Schleiber, I. 1983 Griechische Topferkurut. Munchen. Schwander, E.-L. & G. Zimmer (with a contribution by U. Zwicker) 1983 Zurn Problem der Ofen Griechischer Bronzegiesser. Archaologische Azeiger pp. 57-80. Seki, T. 1985 Untersuchungen zum Verhaltnis von Gefassform und Malerei Attische Schalen. Berlin. Stewart, A. 1990 Greek Sculpture. New Haven and London. Thompson, H.A. 1964 A note on the Berlin Foundry Cup. In Essays in Memory of Karl Lehmann (Marsyas supplement I), pp. 323-328. Thompson, W.E. 1969 The inscriptions of the Hephaisteion. Hesperia XXXVIII: 114-118. Zimmer, G. 1990 Griechische Werkstattbilder. Bilderhefte der Staatliche Museen Preussicher Kulturbesitz, Heft 42, Berlin. Ziomecki, J. 1975 Les Representatiom d'Artisam sur /es Vases Attiques. Wroclaw.

. & Massimo TT."da/e rl GabriellaPrisco

. 1· The exterior ofthe Berm l. Foundry cup (after Bliimner 1886) Fig.

130

The reconstruction of manufacturing sequences on the basis of iconography

Fig. 2: The exterior of the Karlsruhe cup: a potter and his apprentice turning the wheel (drawing M. Vidale)

Fig. 3: The exterior of the Karlsruhe cup: a potter presents a finished kylix to a well-to-do citizen (drawing M. Vidale)

131

The Organization of Production in the Artisan Quarter at Rocca d'Evandro Emilia Chiosi

INTRODUCTION

The production of pottery in Roman rustic villas is seen to be a ''particular" case in the world of pottery production (Peacock 1982), since although the presence of elementary factors such as the location, the technology, the type and quantity of distribution are common in workshops of this type, the historical variables in the late Republican period make the production of these workshops particular. The location of the site of Porto di Rocca d'Evandro (Caserta) is emblematic and responded to the intrinsic needs of the production cycle. The site is located in an area with raw material sources, sources of energy, areas for the chemical-physical transformation of the material and easy and economical transport to the market area. The model followed was one which was widespread in central and southern Italy and it depended on the exploitation of the slave mode of production. The activity found its strength and, in a certain sense, its limitations in the fact that it used slave labour as a source of capital. It was therefore only profitable when the price of a slave was low and the supply of labour was continuous and regular.

PORTO DI ROCCA D'EV ANDRO

The site of Porto di Rocca d'Evandro, where the Soprintendenza Archeologica of the province of Naples and Caserta carried out two seasons of excavation between 1992 and 1995, is located at the junction of the Sacco Liri valley, along which the via Latina runs, and the Garigliano valley, between the Aurunci mountains to the north and the slopes of the Roccamonfina mountains to the south (fig. 1) (for a preliminary report on the excavations see Chiosi & Gasperetti 1991 which describes the problems related to the discoveries made at Porto). Located next to the river Garigliano, in one of the loops of its ancient channel, and delimited on two sides by streams, the site had access to raw material sources (sand and clay), water, an agricultural hinterland, and good transport routes. Apart from river transport, two roads paved with dark lava have been found, one of which is a byway between the via Appia and the via Latina, and another road paved with white limestone which runs perpendicular to the other two roads towards the river, where there must have been a landing place for embarking the products made at the site.

same type as those produced at Porto. In all these workshops documented along the lower course of the river Garigliano, Greek-Italic type amphorae have been found, of the Dressel 1 and Dressel 2-4 types. A kiln was found a few hundred metres upstream of Minturnae, on the right bank of the Garigliano, and a second workshop is located a dozen kilometres from the sea, on the edge of the SS. 430 road, on the left bank of the river (Hesnard et al. 1989: 24-26, figs. 710). This is further proof of the well-known convenience of this location. "The privileged geographical position chosen, the vicinity of the river in addition to the road network of which the tracts found are part, facilitated the transport by river of wine which, arriving from the interior of the region, could have been decanted into the amphorae made in the artisan quarter and eventually embarked in the direction of the coast, and in particular to the port of Minturnae (Chiosi & Gasperetti 1991: 122-123). It is very likely that the activities which were developed in the area also involved the exploitation of the woodland for the procurement of wood which was indispensable for the kilns, as well as the exploitation of the many local quarries in the area for paving the roads which connected the various centres in northern Campania and the two main communication routes: the via Appia and the via Latina. In his historical topographic study of the road network (see Wightman 1981) of the lower Garigliano based on historical cartography, Di Biasio (1994) mentions a place called Porto delle Mole where the via Appia crossed the lower course of the river, and it would be interesting to study the relationship between identical place names in such a small area since the stream to the north of the site is called Fosso Porto di Mole. Although only a small part of the site at Porto has been excavated, it is the first example of a production site in the Liri-Garigliano valley which has been the object of an archaeological investigation. Towards the end of the 2nd century BC, a quarter was established which had a mainly artisan function with kilns for the production of amphorae, tiles and pantiles (which are documented by seals and manufacturing waste), as well as small forms of common pottery (documented by manufacturing waste) and perhaps of antefixes (in dried but unfired clay).

THE KILNS

The excavated kilns are of the vertical type with a rectangular shaped combustion chamber (fig. 2), separated from the firing chamber by a perforated floor of which

Along the course of the river Garigliano, the majority of which was certainly navigable in antiquity, there are other pottery workshops for the production of amphorae of the

133

Emilia Chiosi

A) The kilns have analogous dimensions and analogous spaces allotted, which is the sign of a relative equality among the various people who worked there and of the objects produced.

various fragments have been found although none of them are in situ. Kilns were found in two different areas of this artisan quarter, but only the combustion chamber is preserved, filled with layers of material which is not useful for determining the specific type of production. In the first area, which is located close to the river, the excavated kiln has a rectangular plan oriented east-west, delimited by walls made of limestone blocks and layers of clay earth mixed with fragments of amphorae and tile which served to isolate the heat produced during the firing. A rectangular ditch filled with sand was found next to this kiln, which may have been used as a degreaser for the clay, and a channel which ran towards the river. In the second area in the quarter the kilns were built inside walls made in opus incertum, and may have constituted real blocks which enclosed various work spaces or rooms destined to the various phases in the production process. Kiln number 2 (fig. 3), which also has a rectangular plan and has the praefurnium oriented westwards towards the river, is delimited by perimeter walls made of tiles and tufa held together by mortar. To the south of it, in the same work space (room), two column drums made of local sandstone were found, as well as an antefix made of unfired clay. This work space can be interpreted as a sort of work bench which rested on column drums, and which was used for working on and drying the products before they were fired. Kiln number 3, which has not been completely excavated, also has a rectangular plan and the praefurnium is oriented towards the south. This kiln was also part of a larger room, located at the western edge of the complex towards the river.

B) Some fragments of amphora, many of which appertain to the Dressel 1B type, have seals which were stamped before firing. The seals, which are still being studied, all adopt the same formula consisting of a name in the nominative case, followed by an aristocratic name, by a prenomen in the genitive case and finally the initial S, the abbreviation of servus. The last part of the formula adopted in the seals thus highlights the juridical relationship between dominus and officinator, indicating one of the internal articulations of the production process. As is well known the function of the seals was not that of providing a guarantee of the product for the buyer, but rather of indicating the type of relationship that permitted its production for internal control purposes. Eleven seals on the amphorae refer to MOSCA LUCCEI L(uci) S(ervus), and these are the only seals with a circular scroll ornament (fig. 4); of those with a rectangular scroll ornament, four refer to MELIT( o), one to MANES and four to ANTIOCUS. On the tiles, on the other hand, the seals are again rectangular in form and they refer to PAMP(hilus ), which is documented four times on four legible inscriptions, while the only pantile with a seal mentions SOPH(ilos). From the evidence at the site of Porto di Rocca d'Evandro it seems clear that:

The individual production units at the site therefore had similar characteristics (form, orientation) which were related to the general systemisation and orography of the site. The evidence seem to be characterised by a substantial homogeneity: the structures are aligned and well organised from the point of view of the space allotted to the different activities, and from a structural point of view in particular, it seems that the kilns at Porto were the fruit of a single operation of territorial planning.

1) the dominus who was proprietor of the figlinae of the artisan quarter, and perhaps of the praedia which they served, was a member of the gens Lucceia. In fact all of them were servants of Lucceius L(ucius); 2) of the different officinatores (or tegularii), the servi had the same type of relationship with the dominus and perhaps controlled specific sectors of the production (amphorae, tiles, pantiles). Leaving aside the inscription on the pantiles, of which there is only one example, it seems that some people controlled the production of the amphorae (Mosca, Melito, Manes and Antiocus) and others the tiles (Pamphilus). Therefore the name of the same servant is never found on different products, such as amphorae and tiles.

The amphorae produced there were mainly of the Dressel IB type and on the basis of the chronology of these the floruit of the site can be fixed just after the end of the first quarter of the 1st century BC. Therefore we are in the period of the creation of an original Italic model of amphora, the Dressel 1, which had such a large expansion that it has been rightly connected with the development in Italy of the slave mode of production which has been defined as ''the unity of the economic and productive world of late Republican Italy'' (Manacorda 1981).

The identity of Lucius Lucceius is unclear (De Caro 1979; Chiosi & Gasperetti 1991). One Lucius Lucceius has been documented by means of one of his slaves at Delo, a certain Nikephoros, in 94 BC, and at the same site some seals have been found on amphora lids dating to a later period which also document an L. Lucceius (Hesnard & Gianfrotta 1989: 404-405, 418-419). Two years later a certain L. Lucceius Mf, a Roman legate, was initiated in the mysteries of Samothrace (Munzer 1927, nr. 4). The best known Luci Luccei are those mentioned by Cicero: a son of Quinto, who was an unsuccessful candidate for the consulate in 59 BC

THE SOCIAL ORGANIZATION OF PRODUCTION As regards the social organization of the production, the available elements are those deriving from the production unit (A) and the product itself (B).

134

The organization of production in the artisan quarter at Rocca d'Evandro

(Munzer 1927, nr. 6), and in 50 BC Lucius Lucceius son of Marco and perhaps nephew of the legate of 92 BC (Munzer 1927, nr. 5). The chronology of the site at Porto seems to indicate that the dominus who was the owner of these figlinae was one of the two L. Luccei known and mentioned by Cicero, even though this does not exclude the possibility that their forefathers hadn't set up an artisan quarter previously. The presence of the dei Luccei slaves at Delo documents that the distribution of the products was personally controlled by the gens. The hypothesis that the gens Lucceia owned buildings in the area of the Foro Boario in Rome is well known, as they gave their name to a road divided into two parts by the Porta Flumentana, the two vici Luccei (Palmer 1980, pp. 136-140, fig. 1). For historical and topographic reasons these vici Luccei (Cicero Ad Att. VII 3, 6, 9) have been dated to the mid 3rd century BC, when a member of this gens, a Lucceius aedile, may have given his name to the road he paved. It therefore seems to be more than just a hypothesis that the Luccei owned property within the same quarter, well before the eel/a Lucceiana is documented; this wine cella which was built or rebuilt in Trajan times is recalled in the 161 AD dedicatory inscription at the religious complex of Agathopo (Panciera 1970-1971). Until the end of the late Imperial period the gens Lucceia continued to have a large and well documented interest in the production and distribution of wine, with the obvious corollary of the choice of the sites to be used for this, such as the bend in the river Tiber towards which the vici Luccei ran. If the hypothesis about the chronology of the vici Luccei mentioned above (Palmer 1980, pp. 137-140) is correct, the other important point in the story of this gens is that the Foro Boario could have served as a model for the workshops at Porto di Rocca d'Evandro.

transport to the distribution of the finished product. On the other hand, on the basis of the data which have emerged so far, one cannot exclude the possibility of a diversity of ownership (a hiatus) between the management of the agricultural production and the other activities (pottery production, commerce). And it is also possible that the site at Porto was not tied to only one praedia. CONCLUSIONS

In the context of the complex relationship between ownership and management of the activities which existed in late Republican and proto Imperial Italy, a situation of notable entrepreneurial vivacity which may have been speculative seems to have emerged at the site of Porto di Rocca d'Evandro, run by families related to the emerging Italic ranks which were inserted in the Roman entrepreneurial class. However, at the site at Porto the system of slave production was very important, and with the peace brought temporarily by Trajan and then definitively by Hadrian the sources of capital and of slaves slowly started to dry up (Lo Cascio 1995), thus throwing the large slave businesses into crisis. It is probable that the workshops at Rocca d'Evandro, as well as the other workshops along the course of the river Garigliano, fell into decay for this very reason. Acknowledgements I would like to thank dott. Stefano De Caro, Archaeological Superintendent for the provinces of Naples and Caserta, for having continuously and constantly encouraged the study of this site. I would also like to thank my friend dott.ssa G. Gasperetti, who is in charge of this area and who followed the various phases of the research.

The widespread presence of slaves of the gens Lucceia with names of eastern origin also confirms that the import of slaves (Musti 1981) resulting from conquests made by Rome in those years in the east was fundamental for the production and wide diffusion of the Dressel 1 amphorae and their wine content, the raw material of the production cycle. An analogous diffusion of eastern names is also documented Minturnae, which was certainly the port used by the artisan quarter at Porto.

BIBLIOGRAPHY

Chiosi, E. & G. Gasperetti 1991 Rocca d'Evandro (Caserta). Localita Porto. Un quartiere produttivo romano sulla riva sinistra del flume. Bollettino di Archeologia 11-12: 121-125 and 194-196. De Caro, S. 1981 Le lucerne dell' officina L VC. Rendiconti de// 'Accademia de/le Scienze Fisiche e Matematiche di Napoli XLIX: 107-134. Di Biasio, A. 1994 II passo de/ Garigliano nella storia d'Italia. II ponte di Luigi Giura. Mintumo. Hesnard, A. et alii 1989 Aires de production des greco-italiques et des Dr. 1. In Amphores romaines et histoire economique. Dix ans de recherche. Atti del colloquio di Siena (22-24 maggio 1986), pp. 21-65. Hesnard, A. & P. Gianfrotta 1989 Le bouchons d'amphore en Pouzzolane. In Amphores romaines et histoire economique. Dix

It remains to be resolved whether the various workshops at Porto were all property of the dei Luccei, as would seem to be indicated by the high percentage of seals bearing this name. However, only a small area of the quarter has been excavated, and therefore this can not be demonstrated with certainty. What has been demonstrated is the social organization of a ''vertical" type, centralized in the hands of a single gens, of which some members, perhaps of south Italian origin like many entrepreneurs of this period, were already part of the Roman entrepreneurial world. In this relationship the dominus maintained an active role in the production process, controlling all the various different phases in the economic activities connected with the agricultural exploitation of the land, from storage to

135

Emilia Chiosi L 'Italia: Insediamenti e Forme Economiche. Bari,

ans de recherche. Atti del colloquio di Siena (22-24 maggio 1986), pp. 393-441. Lo Cascio, E. 1995 L'impero patrimoniale e la morte lenta del capitalismo antico: l'interpretazione weberiana del passaggio dalla repubblica all'impero. In

L 'incidenza dell 'antico. Studi in memoria di Ettore Lepore, Napoli, pp. 261-279. Manacorda, D. 1981 Produzione agricola, produzione ceramica e proprietari nell'ager Cosanus nel I sec.a.C. In A. Giardina & A. Schiavone (eds.) Societa Romana e

Produzione Schiavistica, IL Merci Mercati e Scambi nel Mediterraneo. Bari, 1981, pp. 3-54. Munzer 1927 s.v. Lucceius, in RE. 13, coll 1553-1562. Musti,D. 1981 Modi di Produzione e reperimento di manodopera schiavile: sui rapporti tra l' oriente ellenistico e la Campania. In A. Giardina & A. Schiavone (eds.)

1981,p.3 Palmer, RE.A. 1980 The Vici Luccei in the Forum Boarium and some Lucceii in Rome. Bul/ettino del/a Commissione Archeologica Comunale di Roma 85 (1976-1977): 135-161. Panciera, S. 1970-1971 Nuovi documenti epigrafici per la topografia di Roma antica. Rend Pont. Acc. XLIII: 110-117, fig.

I. Peacock, D.P.S. 1982 Pottery

in the Roman World: EthnoarchaeologicalApproach. London.

Wightman, E.M. 1981 The Lower Liri Valley: Problems, Trends and Peculiarities. In G. Barker & R Hodges (eds.)

Archaeology and Italian Sociery. Prehistoric, Roman and Medieval Studies. Oxford: British Archaeological Reports International Series 102, pp. 275-287.

Societa Romana e Produzione Schiavistica, 1

Fig. I

an

View of one of the blocks in the artisan quarter at the site of Rocca d'Evandro

136

The organization of production in the artisan quarter at Rocca d'Evandro

Fig. 2

Fig. 3

Kiln 2 and its work space

Detail of the combustion chamber in kiln 2, seen from the praefumium

137

Emilia Chiosi

Fig. 4

Example of one of the most widespread seals at Porto, that on the Dressel 1B amphorae in a circular scroll ornament, which refers to MOSCA LVCCEI L S.

138

Glass Beads as an Archaeological Source Anna Mastykova

production methods, the origin of the beads and the type of workshop where they were made; 3. comparison of the data obtained with all the available information about the site and its historical and cultural context.

INTRODUCTION

Glass beads are mass archaeological finds which contain a large amount of information on the ethno-cultural and social processes in ancient societies (Von Freeden 1997). Recently beads have become an object of specialized research (see for example Andrae 1973; Alekseeva 1975, 1978, 1982; Callmer 1977; Guido 1978; Glasperlen 1983, 1987, 1995; Templemann-Maczynska 1985; Theune-Vogt 1990; Glass Beads 1995; Perlen 1997). In these studies the detailed morphological method was used as well as technological and chemical analyses; classifications of the types of glassmaking workshops and their production is conducted in parallel with research on the chronology and geographical distribution ofbeads.

In modem research, a formalised description of the beads according to a strictly hierarchical set of attributes is often used (Sasse & Theune 1995, 1996, 1997; Sasse & Vach 1997). The structure of the descriptive tree provides accurate determination of the beads, which consents computer processing and the application of statistical analyses. A clear indication set is necessary for the creation of the structure of classification on the one hand, and for its generalised character to make it possible to include any new material on the other. Modem approaches to the investigation of beads are not associated with any new, previously unknown methods, by which I mean that a combination of various methods have been applied in archaeology. The possibilities of such an approach may be illustrated by the following examples.

This article aims to demonstrate the potential of beads as an archaeological source. A systematic approach to the investigation of beads should be applied in order to obtain the most complete information, as this method reflects not only technological peculiarities but also the archaeological context. In this case one can obtain information not only on the development of ancient glassmaking and bead production in particular, but also on the cultural and historical processes which took place in ancient societies.

BEADS AS A CHRONOLOGICAL INDICATOR

The unique character of beads is based on their mass presence in so-called closed assemblages, such that different types of statistical analyses can be applied to them (for more details see Theune 1995). Examples of this are the works by Sasse, Theune and Siegmund on the chronology of Merovingian beads (Theune-Vogt 1990; Sasse & Theune 1996, 1997; Siegmund 1995), or Hoilund Nielsen's investigation of early Medieval Scandinavian beads (Hoilund Nielsen 1997).

Modem methods of archaeological research on beads are based on the following general principles: 1. beads are not considered as separate artefacts, but in connection with each other and according to the context of the find; 2. the results of field observations and laboratory studies are taken into account; 3. classification and the creation of a database is necessary in order to highlight the regularities in the material under investigation.

Presently I am carrying out a project on the relative chronology of beads recovered from the early Medieval sites of the Kislovodsk basin (Northern Caucasus). These investigations resulted in the identification of some regular combinations of certain types of stone and glass beads. Each type of bead exists for a long period of time, while their combinations have a relatively short duration. As a mass material, beads are suitable objects for seriation analyses (analysis of the type combinations within assemblages) (see Sasse & Theune 1996; Hoilund Nielsen 1997; Sasse & Vach 1997). By this method groups or combinations of beads are outlined. The chronology of the groups is established by examining the occurrence of beads in closed assemblages with reliably dated pottery or metal artefacts. The latter, in their tum, also undergo seriation (for the Kislovodsk basin see Gavritukhin & Malashev 1998). The correlation of beads

The method used for such a complex investigation of archaeological beads includes three main stages: 1. field work: this is one of the main stages, as detailed records of the horizontal and stratigraphic distribution of the material and its relationship to the other finds provides the possibility of preserving very important information on ethnic, cultural and other characteristics of the population studied, and forms the basis for further research on the beads; 2. investigation of the beads themselves: this includes morphological, technological and, if necessary, chemical analyses. In this way one obtains information on their

139

Anna Mastykova

with pottery and metal objects indicates the short period of their existence. Beads constitute good material for chronological studies for two reasons. Firstly, the seriation of beads can and must be compared with pottery and metal artefact seriations. Possible matches or differences of the seriation results represent a problem for special studies as well as for the creation of a common chronological sequence for the sites or cultures under investigation on the basis of such a procedure. The direct introduction of beads (or their sets) into correlation tables is also possible with other finds, such as metal or pottery ones, as a chronological indicator (see Perin 1980; Roth & Theune 1988). Secondly, the incorporation ofbeads into local chronological scales will evidently let one compare variation in their circulation and peculiarities of their spreading within different populations (for example, Tempelmann-Maczynska 1985). In different local groups of sites seriations display essential differences, and this fact certainly represents an important historical source.

BEADS AS A SOURCE FOR THE HISTORY OF MATERIAL CULTURE IN ANCIENT SOCIETIES The most abundant information concerning beads is available in cases when the latter are not separated from the context of the other archaeological sources. Accurate records of the discovery of the beads within the excavation trench and their spatial distribution, as well as joint studies with other specialists, provide interesting results for the history of dress. It is known that dress is one of the most important indicators of the evolution of material culture in ancient societies. Thus, the spread of female dress with numerous beads in the early Medieval period demonstrates Mediterranean and, above all, Byzantine cultural influence on European barbarian peoples (Koch 1974; Schulze 1976; Hinz 1978; Vierck 1981; Schellhas 1997) (Figs. l, 2 and 3).

ornament large polychrome eyed beads of individual production were used, combined with metal pendants. Beads of that type were probably manufactured in a workshop which practised the Egyptian glassmaking tradition. By "Syrian", "Egyptian" or some other production, I do not mean the location of the workshop, but rather particular manufacturing traditions. These conventional terms are used for certain morphological, technological and chemical groups of glassware implying the region of their origin. It is well-known that numerous tribes practised embroidery decoration on the edges of female dress, and endowed this ornamentation with a magical meaning. It served as an amulet protecting them from demonic forces and the evil eye, and it is highly possible that bordering female dress with beads of certain combinations and colours had the same magical and symbolic function. The way beads were used, which can be traced from their location on the skeleton, is without any doubt an important cultural attribute. Bordering dress edges with beads was a wide-spread tradition among Iranian-speaking tribes of the steppe region, such as the Scythians and the Sarmatians. In the Sarmatian period (the 3rd century BC to the 4th century AD), this custom was widely represented in the Crimea and in the central part of the Northern Caucasus (Abramova 1993: 92). It is evident that the finds from Kulchuk reflect the Sarmatian influence on the local population which the cemetery belonged to. It should be pointed out that in the same burial of the Kulchuk cemetery, beads were also discovered on the wrist of the dead which evidently constituted a bracelet. Similar finds of the Sarmatian period have also been reported among the tribes of the Northern Caucasus (Abramova 1993: 92).

In the 5th and 6th centuries AD in Europe, large beads of glass or other materials (such as stone and amber) were popular in female dress as a separate pendants worn on the side, on the waist-band, or near the knee (Martin 1994, Abb. 139: 2-3). Such cases are known from the Alamanian cemetery at Eschborn which is dated to the middle and second half of the 5th century (burials nos. 12, 18, 24, 29) (Ament 1992, Abb. 17.6; 19.1,4; 20.2; S. 60, 62, 64-66). Again, in the Burgundian burial No. 304 of the Beaune necropolis dated to the second half of the 5th and the early 6th centuries, beads were found above the knees on female skeletons (Gaillard de Semainville, Sapin & Maranski 1995, Fig. 7.304). The magical function of bead-pendants in dress is undoubted, and the wide area of their occurrence in Europe displays some common features in the cultural development.

During the excavations of the classical cemetery at Kulchuk in the Crimea, which dates from the 1st century BC to the 1st century AD, a family chamel-house was discovered containing several burials (Golentsov 1990). By the accurate recording of the distribution of the beads on the skeleton and their further analysis in co-operation with the physical anthropologists, a large part of the female dress of that period was reconstructed. Its characteristic features were glass beads of serial production which were sewn onto the edges of sharovary-type trousers and the sleeves of chemisedresses. The beads were manufactured in workshops of the Syrian glassmaking tradition, and small beads of Syrian production were sewn onto the dress collar. The beads sewn onto the trouser edges were of the same form, size and decoration. It is noteworthy that children's clothes were not ornamented with beads, apart from the head-dress (cap) which was bordered by minute beads. For the child's neck-

Research on beads is also valuable for the reconstruction of the social structure of ancient societies. For this purpose it is important to reveal ''prestige" types of beads linked to the costume of the social elite. Different territories had ''prestige" types of their own and in Scandinavia large glass beads with mosaic decoration consisting of squares placed in

140

Glass beads as an archaeologicalsource a chess-board order were widespread from the late Roman period onward (Olldag 1994: 237). However in Eastern Europe beads of that type were found mostly in the graves of military leaders of the Great migration period, where they were used as sword-pendants, such as in the Shestaya Shakhta burial in the Northern Caucasus (Abramova 1997, Fig. 39,8) and two burials of the Koshibeevo cemetery on the Oka river in Central Russia (Spitsyn 1901, 11, Table 13,6) (Fig. 4). They are also recorded in rich female graves in the Baytal-Chapkan and Pashkovsky-1 cemeteries in the Northern Caucasus, Shapka-Abghydzrakhu in Abkhazia and Chersones in the Crimea. Prestige types are also present in the stone and amber beads of the Great migration period, such as large faceted crystal beads, large date-shaped chalcedony ones, and large amber well-made beads of a rounded shape and with an incised decoration.

4. workshops with an incomplete production manufacturing only semi-finished products.

cycle

In its tum, the information obtained on the types of

workshops and the distribution of beads according to glassmaking traditions practised by the population of Khazar Chaganate formed the ground for certain suppositions concerning the level of development of Medieval glassmaking in the Mediterranean during the early Middle Ages. While the workshops of the "Byzantine" tradition apparently practised multi-purpose production, the workshops of the "Syrian" or "Egyptian" traditions reached the stage of the specialized production of beads, which was part of a diversified production and economic system.

BEADS AS A SOURCE FOR THE HISTORY OF ECONOMIC AND CULTURAL CONTACTS BEADS AS A SOURCE FOR THE HISTORY OF GLASSWARE TECHNOLOGY

Each artefact preserves the traces of the technological operations used for its manufacture. Some characteristic peculiarities of various processes of production are expressed in the finished products. By studying the latter and the character of its structural elements, one can define the processes used for manufacturing certain ancient artefacts (Lvova 1979). The collection of glass beads recovered from the archaeological sites of the Don basin which date to the second half of the 8th and the early 10th centuries and ascribed to the population of Khazar Chaganate was investigated (Mastykova 1991, 1993). In Eastern European archaeology these sites are known as the Saltovo-Mayatskoe archaeological culture. The results of morphological, chemical, and technological studies of glassware made it possible to link the beads with certain glassmaking workshops. It became clear that the set of glassware of the Saltovo-Mayatskoe culture had been imported. Beads worn by the Khazar Chaganate population were imported from several centres, which belonged to different manufacturing traditions: "Egyptian", "Syrian", and "Byzantine" ones. According to the results of the chemical and technological analyses of beads, the following types of workshops were determined based on the character of glassmaking (Mastykova 1997, 1997a): 1. workshops with a totally continuous production cycle; 2. workshops with a total production cycle and with the division of labour; among the workers there were skilled melters capable of glassmaking, blowers manufacturing objects out of liquid glass, moulders who further treated the non-cooled products, blowers who dealt with halffinished products and polishers/decorators who treated the cooled glass; 3. workshops with an incomplete production cycle treating only semi-finished products;

As mass archaeological material, beads contain valuable information on trade, exchange centres and routes of contacts. I should point out the classical works by Callmer concerning these problems devoted to Scandinavian and Eastern European beads of the Viking period (Callmer 1977, 1991, 1995). Research on the glass beads recovered from sites of the Saltovo-Mayatskoe culture made it possible to define their origin, and thus to identify the regions linked by exchange contacts. The trade of finished products could be carried out directly by the country of manufacture, or by the mediator countries located along the transit trade routes. The museum collections of antiquities from the Kislovodsk basin in the North Caucasus were studied in order to reconstruct one of the supposed trade routes by which glass beads were transported to Khazar Chaganate, and the materials yielded by the sites chronologically close to Saltovo-Mayatskoye culture were considered. Subsequently sites with morphologically similar bead sets were chosen, which then underwent further investigation aimed at the identification of the beads and their geographical distribution. Morphological and technological analyses of the abundant material made it possible to identify glass beads analogous to those of the Saltovo-Mayatskoye culture. By mapping the glass beads it was possible to identify the direction of trade routes, which coincided with two roads known from ancient written sources, such as the Misimian and Dara roads, which were the Northern Caucasian branches of the Great Silk Route. According to Menander the Byzantine, these two roads acted simultaneously in the 6th century and were used by the merchants depending on the situation on the roads (Fig. 5). While the mapping of the Northern Caucasian beads has made it possible to trace these two routes, while the detailed study of the chronology of the beads has furnished indications concerning the dynamics of the function of the

141

Anna Mastykova

Northern Caucasian part of the Great Silk Route. It turned out that the earliest beads sets of the 8th and 9th centuries were found along the Misimian road, while ones of the later 9th century were concentrated along the Dara part of the route (Mastykova 1997a, 1997b).

1995 The influx of oriental beads into Europe during the 8th century A.D. In Glass Beads. Cultural History, Technology, Experiment and Analogy. Lejre, pp. 49-54. Gaillard de Semainville, H., C. Sapin & D. Maranski 1995 Les decovertes de Beaune (Cote d'Or). Des Burgondes en Bourgogne? In Les Burgondes, apports de l'archeologie. Dijon, pp. 143-165. Gavrituhin, I. & V. Malashev 1998 Perspektivy izucheniya khronologii

Thus, a modem complex approach to the investigation of beads provides an opportunity not only to determine the origin of glass artefacts and to link them with known glassmaking centres, but also to discover new ones. It also makes it possible to reconstruct trade routes, to distinguish cultural links between the regions, and to clarify the meaning of those artefacts within the framework of the culture under consideration and the period of their function, and to determine and verify the dates of the burial assemblages and the artefacts.

rannesrednevekovykh drevnostei Kislovodskoi kotloviny, in: Kultury evropeiskikh stepei vtoroi poloviny I tysyacheletiya n.e. (voprosy khronologii) (I'he perspectives of chronological studies of the early Medieval antiquities from Kislovodsk basin, in: The cultures of European steppes of the second half of the 1st millennium A.D. (Problems of chronology). Samara, pp. 28-86.

BIBLIOGRAPHY

Glasperlen der Vorromischen Eisenzeit 1983 Marburger Studien zur Vor - und Fruhgeschichte I. Marburg, Mainz. 1987 Marburger Studien zur Vor - und Fruhgeschichte II. Marburg, Mainz. 1995 Marburger Studien zur Vor - und Fruhgeschichte III. Marburg, Mainz. Golentsov, A.S., 1990 Otchet Kulchukskoy ekspedicii. Arkhiv Instituta

Abramova, M.P. 1993 Tsentralnoe Predkavkazie v sarmatskoe vremya (Ill

v. do n.e. - IV v. n.e.) (Central Caucasus in the Sarmatian Period, 3rd century B.C. - 4th Century A.D.). Moscow. 1997 Rannie a/any Sevemogo Kavkaza, III-V vv. n.e. (I'he early Alanians of the North Caucasus, the 3rd - 5th centuries A.D.). Moscow.

Archeologii NAN (Report of the Kulchuk expedition. Archive of the Institute of Archaeology, National Academy of Sciences of Ukraine). Kiev.

Alekseeva, E.M. 1975 Antichnye

busy Severnogo Prichernomorya (Classical beads of the North Pontic area. Code of archaeological sources Al-12). Svod Archeologicheskich lstochnikov Al-12. Vol. 1

Guido, M. 1978 The Glass Beads of the Prehistoric and Roman Periods Britain and Ireland. London. Hinz,H. 1978 Zurn Frauentracht der Volkerwanderungs und Vendelzeit im Norden. Bonner Jahrbucher 178: 347-365. Hoilund Nielsen, K. 1997 Die fruhmittelalterlichen Perlen Skandinaviens. Chronologische Untersuchungen. In U. Von Freeden & A. Wieczorek (eds.) Per/en. Archaologie, Techniken, Analysen. Bonn, pp. 187196. Koch, U. 1974 Mediterrane und frankische Glasperlen des 6. und 7. Jahrhunderts aus Finnland. In Studien zur Vorund Fruhgeschichtlichen Archaologie. J. WemerFestschrift. II. Munchen, pp. 495-520. Lvova,Z.A. 1979 Technologicheskaya klassifikatsiya izdeliy iz stekla (po materialam rannesrednevekovykh steklyannykh ukrasheniy) (Technological classification of glass articles, based on early medieval glass decoration).

Moscow.

1978 Antichnye

busy Severnogo Prichernomorya (Classical beads of the North Pontic area. Code of archaeological sources Al-12). Svod Archeologicheskich Istochnikov Al-12. Vol. 2

Moscow.

1982 Antichnye

busy Severnogo Prichernomorya (Classical beads of the North Pontic area. Code of archaeological sources Al-12). Svod

Archeologicheskich Istochnikov Al-12. Moscow. Vol. 3. Ament, H. 1992 Das alamannische Grabeifeld von Eschborn (Main-Taunus Kreis). Wiesbaden. Andrae, R 1973 Mosaikaugenperlen. Untersuchungen zur Verbreitung und Datierung karolingerzeitlicher Millefioriglasperlen in Europa. Acta Praehistorica et Archaeologica 4: 101-198. Callmer, J. 1977 Trade Beads and Bead Trade in Scandinavia ca. 800-1000 A. D. Lund. 1991 Beads as a criterion of shifting trade and exchange connections. Studien zur Sachsenforschung 7: 2538.

Arheologichesky Sbornik Ermitazha 20: 90-103.

Gosudarstvennogo

Martin, M. 1994 Fibel und Fibeltracht. K. Spate Volkerwanderungszeit und Merowingerzeit auf dem

142

Glass beads as an archaeological source Kontinent. Reallexikon der Altertumskunde Band 8: 541-582.

Germanischen

Mastykova, A. V. 1991 Tipologiya bus iz pogrebenii Mayatskogo selishcha, (Typology of beads from the burials of the Mayatskoye settlement). In A.Z. Vinnikov & G.E. Afanasiev (eds.) Kultovye kompleksy Mayatskogo selishcha (The cu/tic assemblages of the Mayatskoye settlement). Voronezh, pp. 170-182. 1993 Ste/do yugo-vostochnoi Evropy vtoroi poloviny I

tys. n.e. (po materialam saltovo-mayatskoi kultury) (South-Eastern European Glass of the second half of the /st mill. A.D., based on the SaltovoMayatskoe culture). Ph.D. Thesis, Moscow University, Moscow. 1997a Study of the trade links of south-eastern Europe during the second half of the first millenium A.D. on the basis of the Saltovo-Mayatskoye archaeological culture. In Exchange and Trade in Medieval Europe - Papers of the "Medieval Europe Brugge 1997" Conference. Vol. 3. Zellik, pp. 8996. 1997b K izucheniyu roli Kavkaza v sisteme vostochnoevropeiskikh torgovykh svyazei vtoroi poloviny pervogo tysyacheletiya n.e. (po materialam saltovo-mayatskoi kultury) (On the study of the Caucasus' role in the system of East-European trade links of the second half of the 1st millennium A.D. (according to the materials of SaltovoMayatskoye culture). Istoriko-Arheologitchesky Almanakh 3: 80-88. Olldag, J.E. 1994 Glasperler i danske fund fra romersk jemalder. Aarboger for Nordisk Oldkyndighed og Historie (1992): 193-280. Perin, P. 1980 La datation des tombes merovmgzennes. Historique-Methodes-Application.Geneve. Rassmussen, M., U. Lund Hansen & U. Nasman (eds) 1995 Glass Beads. Cultural History, Technology,

1997 Das Programm ProPer. Klassifizierung und Anwendung. In U. Von Freeden & A. Wieczorek (eds.) Per/en. Archaologie, Techniken, Analysen. Bonn, pp. 169-176. Sasse, B. & W. Yach 1997 Das Programm ProPer. Uberlegungen zur Weiterverarbeitung der Daten. In U. Von Freeden & A. Wieczorek (eds.) Per/en. Archaologie, Techniken,Analysen. Bonn, pp. 177-186. Schellhas, U. 1997 Perlen als Fibelanhanger - die merowingische Interpretation eines mediterranen Vorbildes. in: Von Freeden, U. & A. Wieczorek (eds.) Per/en. Archaologie, Techniken, Analysen. Bonn, 339-348. Schulze, M. 1976 Einflusse byzantinischer Prunkgewander auf die frankische Frauentracht Archaologisches Korrespondenzblatt 6: 149-161. Siegmund, F. 1995 Merovingian Beads on the Lower Rhine. Beads. Journal of the Society of Bead Researchers 7: 3753. Spitsyn, A.A. 1901 Drevnosti basseinov rek Oki i Kamy, (The

Experiment and Analogy. Proceedings of the Nordic glass bead seminar. Studies in Technology and Culture vol.2. Lejre. Roth, H. & C. Theune 1988 Zur Chronologie

merowingerzeitlicher Frauengraber in Sudwestdeutschland. Stuttgart.

Sasse, B. & C. Theune 1995 Merovingian Glass Beads - A Classificational Model. In Glass Beads. Cultural History, Technology, Experiment and Analogy. Lejre, pp. 75-82. 1996 Perlen als Leittypen der Merowingerzeit. Sonderdruck aus Germania 74 (1): 187-231.

143

antiquities of the Oka and Kama rivers' basins). St.-Petersburg. Tempelmann-Maczynska, M. 1985 Die Per/en der romischen Kaiserzeit und der

fruhen Phase der Volkerwanderungszeit im mitteleuropaischen Barbaricum. Mainz. Theune, C. 1995 Moglichkeiten und Grenzen der Seriation. Ein Diskussionsbeitrag. Etnographish-Archaologische 'Zeitschrift36: 323-341. Theune-Vogt, C. 1990 Chronologische Ergebnisse zu den Per/en aus dem alamannischen Grabeifeld von Weingarten, Kr. Ravensburg. Eine Datenanalyse. Kleine Schriften aus dem Vorgeschichtlichen Seminar Marburg. Heft 33. Marburg. Vierck, H. 1981 Imitatio imperii und interpretatio Germanica vor der Wikingerzeit. in: Les pays du Nord et Byzance (Scandinavie et Byzance). Uppsala, pp. 64-113. Von Freeden, U. 1997 Perlen - Kulturbegleiter der Menschheit. In U. Von Freeden & A. Wieczorek (eds.) Per/en. Archaologie, Techniken, Analysen. Bonn, pp. 1-12. Von Freeden, U. & A. Wieczorek (eds.) 1997 Per/en. Archaologie, Techniken, Analysen. Akten des Internationalen Perlensymposiums in Mannheim. Bonn.

Anna Mastykova

a

b

Fig.I. The examples of rich Mediterranean dress decorated with beadsaccording to iconographic data a. the empress Theodora's image from San Vitale church in Ravenna (after H. Hinz 1978) b. the image from St. Agnes church in Rome (after H. Hinz 1978)

144

Glass beads as an archaeological source

a

0

C

d

Fig.2. The imitation of Mediterranean dress decorated with beads in Merovingian female dress (after M. Schulze 1976) a. Selzen, burial 11 (after W. andL. Lindenschmidt, s. Anm. 25. Ta£ 11) b. Koln-Mungersdor:t: burial 89 (after F. Fremersdor:t: s. Anm. 18. Ta£ 78) c. Wittersheim, burial 4 (after Anm. 27, Ta£ XVII, 4) d. Koln-Mungersdor:t: burial 142 (after F. Fremersdor:t: s. Anm. 18, Ta£ 81, 142

145

Anna Mastykova

0 () a

b

Fig.3. The example of Scandinavian and Merovingian imitations of Mediterranean dress decorated with beads (after R Vierck, 1981) a. Orby (Denmark); b. Pareds (France)

146

Glass beads as an archaeologicalsource

h a

u..u

Fig.4. The finds of mosaic beads- sword-pendants a Shestaya Shakhta (after M.P. Abramova 1997) b. Koshibeevo (after A.A. Spitsyn 1901)

147

Anna Mastykova

Black Sea

0

80

\

Fig.5. The map of the distribution of beads dated to the 8th-9th centuries (after A. Mastykova 1997)

148

Pre-industrial Mining Techniques in the Mountains of Campiglia Marittima (Livorno) Giovanna Cascone & Alessandra Casini

INTRODUCTION

This paper presents the results of research that developed out of a major field project carried out by the Mediaevalists of the University of Siena on the archaeology of minerary landscapes in Tuscany. The information provided here is due to a joint effort of the University of Siena and the Gruppo Speleologico Livomese, which collaborated in the study of pre-industrial mining in the mountains of Campiglia Marittima (Livomo), with the goal of discovering and documenting ancient mining techniques and their evolution from the Etruscan period (from the Vllth century BC) to the XVIth century AD.

GEOGRAPHICAL AND GEOLOGICAL SETTING

The mountains of Campiglia Marittima are the northwestern extremity of the Colline Metallifere (literally "Metal-bearing Hills") of Tuscany. These hills are famous for their skam mineralizations, including mixed sulphides such as calchopyrite (a copper-iron sulphide), galena (lead sulphide, sometimes with silver) and blenda (zinc sulphide) (Tanelli 1992, 1993). Surface surveys carried out over an area of approximately 10 square kilometres, revealed the location of the pits of more than 200 mines datable to preindustrial times. These pits, which are located below the surface, are mostly vertical shafts and follow the local geostructural features. In fact, skam mineralizations are usually included within sub-vertical tectonic fault lines; the karstformed cavities so far identified in this area have a similar sub-vertical setting too. The mining necessarily developed from the surface downward, strictly following these natural features and without lateral expansions. The presence of rock subject to karst-like erosion also played an important role in the formation of mixed subterranean features (Cascone 1991, 1993). Only in rare cases, in fact, is an ancient mine, in its evolution, entirely artificial. We frequently entered natural caves through the shafts dug by ancient miners and, conversely, found evidence of ancient mining during the exploration of natural caves. For this reason, minerary activities in these hills have never stopped. Metal-bearing deposits were located at the base of the surface altered outcrop, or of the altered skam; underground prospecting could start from a karst-formed entrance or by enlarging the natural faults nearby. (G.C.)

SETTLEMENT AND MINING ACTMTIES

Four major historical periods of intensive mining, alternating with phases distinguished by an apparently total lack of interest for these metal sources, have been recognized. These periods may be so defined: from the Etruscan to the Roman Republican period (about Vllth-Ist century BC); Mediaeval period (from the end of the Xth to the XNth century AD); Modern period (XVIth century AD); Contemporary (XIXth-XXth century AD) (Casini 1993; Cascone & Casini 1997). Furthermore, a protohistoric mining phase, datable to approximately the Xllth century BC, has also been hypothesized (Fedeli 1994-1995). The Etruscan settlement network was made up of hilltop villages, of which the residual evidence is represented by faint remains of stone-built houses. Every settlement probably exploited only the mine of its small territory. While such minerary villages were certainly settled until the IVth century BC, some survided until the Roman Republican age (llnd-Ist century BC) (Casini 1993). A major settlement on the top of Poggio Le Strette (which is probably related to a graveyard found at the site of San Dazio, dated from the midth VIIth-Vth century BC), is located on the edge of the minerary area and may be considered to have been the leading and coordinating center of these sites (Romualdi, Settesoldi & Pacciani 1994-1995). Mediaeval settelments specifically centered on mining are less frequent than the Etruscan ones; these consist of castles built following a south-east/north-west orientation, i.e. the exact orientation of the metal-bearing mineralization. The foundation of these castles was aimed exclusively at the exploitation of the silver-bearing deposits. This metal was used for coinage. The interest of the Mediaeval lords for silver production probably had a strong impact on the organization of mining labour, but it is still unclear how these lords actually exerted their power of control. At any rate in this feudal system, which was distinguished by the absence of conflict between labour and political power (Francovich & Wickham 1994), it was the lord who was the only mining entrepeneur. Under Cosimo I of the Medici family, the whole of Tuscany witnessed a phase of intesive exploitation of its minerary sources, and also the mountains of Campiglia Marittima were involved in the search of copper, lead and silver. In spite of the fact that in our district such phases of intensive mining activities lasted only for 10 years, from 1549 to 1559, it left substantial traces in the territory.

149

Giovanna Cascone & Alessandra Casini

On the basis of written documents, it is possible to reconstruct various aspects of the organization of mining under the Medici family. We also know that the specialized miners from Germany, called Lanzi, were called in. They gave their name to a valley where the richest deposits of silver were found and dug. (A.C.)

THE MINES AND THEIR MORPHOTYPES

The mouths of the mines may appear as elongated openings, following natural faults, or as round or elliptical holes. In both cases, they immediately give access to vertical shafts. While elongated cavities may have a height of some metres, following the local natural stratigraphy, round or elliptical holes, partially or entirely artificial, never exceed diameters of 1-1.2 metres. Sometimes both types of entrance are found in the same mine, perhaps as a result of specific technical and organizational needs. The size of the shafts depends upon the importance of the deposit and the interaction with pre-existing karst caves, frequently found at the interface between the bedrock and skarn. When the deposits were shallow and laterally limited, pits have a diameter of 1.20 centimetres or less, and sometimes stop at a depth of 30-50 centimetres. When the deposits were more extensive, the pits could develop into larger pits or mining rooms having a maximum axis of 6 metres. Sometimes, as already stated, larger sizes were due to the intersection of karst caves. In many cases, miners dug series of sub-parallel pits, communicating at regular intervals with small opening or round ''windows", with a maximum diameter of30-40 centimetres. In our survey, we studied systems of interconnected mines which reached depths of 90-120 metres from the opening, to be explored only by the means of speleological techniques involving the use of ropes for descending and climbing the pits. The few horizontal galleries were short tunnels dug for the purpose of searching and connecting nearby pits. Such galleries have a variable height of between 0.8 and 2 metres, and a width rarely greater than 1.5 metres. They are often filled with layers of rock, piled to form true walls, called ripiene in Italian; in order to ascertain the actual extension of these galleries, we should carry out regular archaeological excavations. The actual total depth of these mines is also probably greater. The bottom of the end pits, in fact, is obstructed by huge amounts of debris formed by mining residues and, to a lesser extent, by the natural decay of the walls of the pits.

Both horizontal and vertical shafts are often covered by deposits of calcium carbonate (flows, stalactites, stalagmites, excentric formations, and so on). Radiometric analyses of such deposits might provide in future reliable datings for the progressive growth of the mines. (G.C.)

MINING TECHNIQUES AND ORGANIZATION OF LABOUR

The most ancient mining techniques were simple: the veins with metallic minerals were thoroughly followed by excavation. This technique was used without change for more than 2000 years, and is attested both in the Etruscan and the Mediaeval periods. Even the traces left by the miners' picks on the walls of the shafts are difficult to interpret, because marks which appear identical are visible in mines which doubtless belong to different periods. One cannot rule out the possibility that mining was carried out by re-opening or enlarging pre-existing shafts, superimposing the marks of the excavation tools of different periods, further complicating a precise recognition of the various periods of minerary exploitation. The study of the marks left by digging tools revealed the presence of four types of points: there are two types of conical points and two types of pyramid-shaped points, ascribable to punches hit with a hammer-like tool and to hafted mining picks respectively. Sometimes small niches for holding lamps are visible on the walls. Where mineralized veins were substantial, a mining yard was formed, and when these yards were particularly large, the vaults were supported by leaving some unexcavated portions in the form of pillars. The traces visible on the walls are, as a rule, not parallel, and lack a systematic orientation. The presence on the walls of square holes (10xl0x5 centimetres) suggest the existence, in ancient times, of wooden structures for supporting stairs, or of wooden scaffolding for speeding up the excavation of the galleries. Along the edge of some of the pits, within the mines, one may easily recognize the holes left by the wooden supports of the cranes used for lifting the excavated mineral. The presence ofunexcavated steps of mineralized rock at the edge of the pits, within the galleries, may probably be explained by the need for firmly anchoring ropes and scaffolding to the rock wall for safety purposes. These steps are still used today for entering and exploring the mining shafts. In order to avoid the burden of lifting the first selected

material, free from metal bearing ores, the miners used to fill up the previously excavated and exhausted shafts, building structural fillings or ripiene. In the absence of cavities to be filled, they used to build retaining walls with wooden planks sunk into the rock, whose traces are still recognizable along the walls and at the rear of the galleries, and to fill the spaces thus obtained. In the calcite deposits which coat the floor of some oblique galleries we also identified the imprints of wooden elements; these imprints demonstrate that the miners used to build true wooden staircases for moving up and down these passages (Cascone & Casini 1997).

150

Pre-industrial mining techniques in the mountains ofCampiglia Marittima

Immediately outside the entrance to the mine the ore was crushed (pesta de/ minerale ). The selected rocks were broken into small fragments and crumbles in order to separate the useless mother rock (ganga) from the sulphides (Casini & Francovich 1992). Then, the selected fractions were washed and stocked near the mine, waiting for transport to the the sites where the ores underwent further processing.

Casini, A. 1993 Archeologia di un territorio minerario: i Monti di Campiglia. In R Mazzanti (ed.) La scienza de/la terra nell'area de/la provincia di Livorno a sud de/ flume Cecina. Quaderni del Museo di Storia Naturale di Livomo 13, 2, pp. 303-314. Casini, A. & R Francovich 1992 Problemi di archeologia mineraria nella Toscana Mediaevale: il caso di Rocca S.Silvestro (LI). In Les Techniques Minieres de l 'Antiquite au XVIII Siecle. Paris: Editions du CTHS, pp. 249-265. Fedeli, F. 1994-1995 Ricerche pre-protostoriche nel territorio di S.Carlo (S.Vincenzo-LI). 11materiale di Vallin del Mandorlo. Rassegna di Archeologia 12, pp. 149234. Francovich, R 1991 Rocca San Silvestro. Roma: Ilva. Francovich, R & C. Whickham 1994 Uno scavo archeologico ed il problema dello sviluppo della signoria territoriale: Rocca San Silvestro e i rapporti di produzione minerari. Archeologia Mediaevale XXI, pp. 7-30. Pierre, F. 1993 Etude de !'apparition de la poudre noire dans !'evolution des techniques minieres de percement. In R Francovich (ed.) Archeologia de/le Attivita Estrattive e Metallurgiche. V Ciclo di Lezioni sulla Ricerca Applicata in Archeologia, Pontignano (SI) e Campiglia Marittima (LI), Firenze, pp. 413-423. Romualdi A., R Settesoldi & E. Pacciani 1994-1995 La necropoli orientalizzante del podere San Dazio nel distretto minerario di Populonia. Rassegna di Archeologia 12, pp. 271-312. Tanelli, G. 1992 Mineralizzazioni ed aree geotermiche de/la Toscana meridionale. Convegno annuale SIMP-76a riunione estiva SGI, Firenze. 1993 I minerali e le miniere del Campigliese. In R Mazzanti (ed.) La scienza de/la terra nell'area de/la provincia di Livomo a sud de/ flume Cecina. Quaderni del Museo di Storia Naturale di Livomo 13, 2, pp. 165-182.

The galleries and mining yards datable to the XVIth century AD are larger than the older ones. The excavation technique continued to be based on the use of punches hit with hammer-like tools; but the progress into the mass of mineralized ore was more systematic. While in the past such punches were used chaotically, in every direction, starting with the XVIth century AD there is evidence of regular excavations, with steps approximately 60 centimetres high and wide, cut into ogive-shaped niches opening in the rock wall. This technique may be compared with mining techniques already reported from mines in the Alps (Pierre 1993). These innovations, affecting both mining and excavation methods, may be referred to the activities of the Lanzi, the professional miners which came from Germany on behalf of the Medici family. (A.C.)

BIBLIOGRAPHY Cascone, G. 1991 Ricerche speleologiche nei Monti di Campiglia Marittima (LI): peculiarita e problematiche di questa area carsica. In Atti de/ VI Congresso de/la Federazione Speleologica Toscana, Stazzema (LU), 1-3/11/1991, pp. 97-137. 1993 La zona speleologica del massiccio del Monte Calvi. Primo contributo alla sua conoscenza. In R Mazzanti (ed.) La scienza de/la terra nell'area de/la provincia di Livomo a sud de/ flume Cecina. Quaderni del Museo di Storia Naturale di Livomo, 13, 2, pp. 183-212. Cascone, G. & A. Casini 1997 Le miniere antiche di Campiglia M.ma (LI). In Atti de/ IV Convegno Nazionale sulle Cavita Artificiali, Osoppo, 30/05/1997 to 1/06/1997, Trieste, 29-50.

151

The Production of Metals for Coinage in Mediaeval Tuscany: The Technological Context Silvia Guideri

This paper presents the results of an archaeological and experimental research which was the subject of the author's doctoral thesis (Guideri 1996) and which is a part of a major field project carried out for several years now by the University of Siena on the minerary landscapes of central Tuscany. The study area is represented by the Colline Metallifere, in central-southern Tuscany, an ideal context for defining the relationships between production and the related socio-economic systems, as well as for gathering all the evidence we need for a careful reconstruction of ancient metalworking processes. In particular, we analysed two archaeological sites which exemplify two different models for the organization and control of metal production. The first may be defined as the "lordship model"; here it is represented by the minerary village of Rocca San Silvestro, in the territory of Campiglia (LI), for which a substantial amount of information was available. This site, which has been excavated since 1984, now belongs to the Parco Archeologico Minerario (see Francovich 1994). The second may be defined as the "commune model", which was later in age and different, and which is represented by the ancient communal workshop of Marsiliana, near Massa (GR), presently still under study.

The process we studied in detail is the transformation of mixed sulphides for the production of copper and lead/silver. In the Mediaeval period sulphides represented the only available source in central Europe for obtaining the metals needed for coinage. For this reason, sulphides played a fundamental role in the development of Mediaeval metallurgy (Bachmann 1982: 26). In fact, in the middle centuries of the Mediaeval period, the economy of this territory was based to a great extent on the exploitation of the local deposits with mixed sulphides (mainly silverbearing galena and chalcopyrite). In these same centuries the area was subdivided into districts ruled by a network of castles, which formed the basis of the productive organization of the local lords (Francovich & Whickham 1994). Fortified villages such as Rocca San Sivestro, which grew up near important minerary basins, were founded and developed in order to control the various steps of mining and processing the metals needed for coinage. These settlements represent the pivots of a productive system which was strictly dependent upon the capability of the ruling elites of managing metal production. Nonetheless, the efficiency of this system was limited by the entirely political need for concentrating in a single site all

the infrastructures demanded by the production cycle. This limit became impossible to overcome when the new hydraulic-powered technologies required the shift of the processing sites to new, more functional locations. This same period, which was characterised by a general development of the exchange economy and a consequent growth of mining activities, witnessed the rise of a different model of managing the mineral resources, the "commune model"; as has recently been proposed, this model may actually be considered a different version of the previous "lordship model" (Francovich & Whickham 1994). Besides the lordly elites which used to rule over the castles and the relative districts, new urban ruling classes grew up in the framework of a more general transformation of the socioeconomic system. These classes often sprung from the same families which, until that moment, had exerted a monopolistic control over the resources of their land. In the Xlllth century the growing demand for metals, with the need of new investments for optimizing production, caused the decadence of the old forms of organization of minerary production, forms which, although initially well suited to the original needs of the old minerary enterprises, could not compete any longer with the tradesmen and entrepreneurs' skill, with the availability of resources and the degree of technological development reached by the new urban classes. Such a process may be closely monitored in the territory of Massa, where an extreme degree of productive specialization provided the push which allowed a minerary settlement originally founded under lordly control to assume the role of free Common.

The minerary code of Massa "Ordinamenta super arte fossarum rameriae et argenteriae civitatis Massae" is the fourth section (distinzione) of the Comon of Massa. It was written in 1325, while its core may be dated back to the middle of the 13th century AD. The Ordinamenta were anastatically resprinted in 1983, on the basis of the edition prepared by Rodolico in 1938. These law codes of the common of Massa are clear and very detailed when they deal with the various stages of the processing of copper and silver. The workshops for the processing of copper ores were scattered in the territory, and traded for managing to private entrepreneurs, while the Common concentrated in its hands the production of silver. From an archaeological viewpoint, the most meaningful feature is the growth of specialized production centres along the main watercourses, where the new technologies applying

153

Silvia Guideri

hydraulic power to the metallurgical plants were first exploited (in Marsiliana: Francovich & Guideri in press). For iron working, this innovation involved a substantial technological jump, causing a radical change in the reduction process of this metal (see Cortese & Francovich 1995). In contrast, for the metallurgy of the metal needed in coinage the technological evolution was more gradual. In fact, when we compare the analyses of metallurgical slags from workshops belonging to the "lordship model" with those produced by plants controlled by the communes, we see no evidence for a greater efficiency. In both cases, good ores are matched by well formed tapping slags, with small amounts of residual metal (Tab. 1). Major changes, instead, affected the organization of the production cycle and the overall amount of production, because plants endowed with water-powered bellows enjoyed improved manufacturing conditions. In the Marsiliana territory, where a waterpowered plant was at work, according to an estimate of 1939 the overall amount of metallurgical residue ranged around 20,000 tons (Baiocco et al. 1990: 91-92).

decreasing density of these materials favoured their segregation. Lead, being the heaviest, was gathered at the bottom of the furnace, from where it could be spilled as the first product; second came the molten slag; and finally the copper matte that, from this point onward, could be treated with the well-known copper processing cycle. One is struck by the similarity between the process we have described and the information provided by the Mediaeval treatise by Theophilus (Diversarum Artium Schedula) when he deals with copper-processing techniques:

Cuprum in terra nascitur; cuius uena cum inuenitur, summo labore fodiendo et frangendo acquiritur. Est enim lapis co/ore uiridis ac durissimus et plumbo naturaliter mixtus. Qui lapis abundanter efossus imponitur rogo et comburitur in modum calcis, nee tamen mutata colorem, sed duritiam amittit, ut confrangi possit. Deinde minutatim confractus imponiturfornaci, et follibus atque carbonibus adhibitis incessanter die ac nocte conflatur. Quod ipsum diligenter et caute fieri debet; id est, ut primo carbones imponantur,. deinde lapidis minutiae superfundantur, rursumque carbones e denuo lapis. Sicque fiat, donec ad capacitatem fornacis sufficiat; cumque lapis coeperit liquefieri, per cauernulas quasdam plumbum effluit et cuprum intro rimanet. Quod cum diutissime coriflatum fuerit, refrigeratur et eicitur, rursumque aliud imponitur eodem ordine" (emphasis by the

In order to observe the effects of a growing empirical knowhow in the Mediaeval metallurgy of mixed sulphides, these effects were searched for in the evolution of the processing of polymetallic ores (or multi-stage process), in the framework of the "lordship model". The treatment of mixed sulphides required high levels of skill because, apart from the need to eliminate sulphur from the ores, specialized craftsmen had to tackle with problems due to the complex nature of the ore itself. In fact, the various metallic sulphides (chalcopyrite, sphalerite, galena, etc.) could not always be efficiently separated by means of simple mechanical crushing. This task is difficult even by applying contemporary techniques (see Meneghini 1947).

author; the Latin text was quoted from Dodwell 1961). For the so-called multi-stage process, we have very important written evidence which has so far been generally overlooked. This description, which is considered by the editors of an English version (Hawthorne & Smith 1979: note 1) to be a misinterpretation by Theophilus, who must have confused such a technique as a process of separation of silver from copper (or liquation), may however be true, and could refer to the processing of polymetallic sulphides. One notices, in fact, the mention of mixed ores in the unprocessed state, and the description of the spilling of lead from the furnace before copper. This process, caused by the need to optimize the available materials, could have provided the example and the prototype for the liquation technique discovered, perhaps not incidentally, during the Middle Ages.

The analysis of slag samples from castle-like settlements, and a detailed study of manufacturing waste from the craft quarter of Rocca San Severo (Guideri 1997; Benvenuti, Mascaro & Tanelli 1995), reveal a process which was subdivided into various stages that, starting from a mixed ore, allowed the extraction first of lead and later, in a subsequent stage, of copper (Fig. 1). A similar technique has been hypothesised for the silver-bearing ores of Monte Calisio (TN) (see Cucini in press). The first stage was the roasting of the ore, a task that could be accomplished in an open pit-like structure. Already in this stage, as one had to reach temperatures which were high enough to transform copper sulphide (600-800°C), metallic lead and slag with a high metal content could be obtained. The process was then continued in a furnace with high vertical walls, or within a pit, where the material obtained with the first stage (desulphurized copper mineral and slag rich in lead) was heated to still higher temperatures (1100-1200°C) using charcoal and an efficient air system. Lead, matte and slag were deposited within the furnace in different times (Fig. 2), and it is curious to note the analogy between this process just described and the contemporary metallurgical techniques used for the same task (Meneghini 1947). The

As a matter of fact, one could suppose that during the reduction of polymetallic sulphides, the silver amount originally included in the copper ore was absorbed by lead, thus performing a self-induced process ofliquation, and that only in subsequent times did the craftsmen realize the function of lead as a collector of precious metals, knowledge that was to be fully exploited and encoded as a scientific principle only in the XVIth century AD.

154

Theproduction of metalsfor coinage in Mediaeval Tuscany

Most probably, the inspiration for the multi-stage process we have just described was the need to obtain silver, while copper matte was simply a by-product (Bachmann 1993: 490). All the available evidence suggests that the process we have outlined is the result of the attempt to exploit to the full the available resources and raw materials, as well as the exploitation potential of ores that were difficult to reduce but which were relatively rich in a commodity which had a strategic importance for defining power relationships: silver. This process may be best understood in the context of a limited market economy in which the capability of exploiting the local resources became a basic issue; in time the inner complexity itself of this process turned out to be a stimulus for technological evolution.

Francovich, R 1991 Rocca San Silvestro. Uva, Roma. 1993 Archeologia de/le attivita estrattive e metallurgiche. Firenze. 1994 Le Ragioni di un Parco a/le Radici dell 'Archeologia Mineraria. Le Miniere di

Campiglia Marittima nelle Pagine dei Naturalisti e dei Geologi del/'800. Regione toscana, Venezia. Francovich, R & R Farinelli 1994 Potere e attivita minerarie nella Toscana altoMediaevale. In R Francovich & G. Noye (eds.),

La Storia dell'Altomedioevo Italiano (VI-X secolo) al/a Luce dell 'Archeologia. Atti del Convegno Internazionale Siena, 2-6/12/1992, Firenze, pp. 443-465. Francovich, R & C. Wickham 1994 Uno scavo archeologico e il problema dello sviluppo della signoria territoriale: Rocca San Silvestro e i rapporti di produzione minerari. Archeologia Mediaevale XXI, 7-30. Francovich, R & S. Guideri in press La produzione de/ rame dai solfuri misti nella Toscana Mediaevale [Paper presented at the seminar on the Metallurgy of Copper, Bourg en Bresse, 17-19/10/1997] Guideri, S. 1996 fl contributo dell'archeometallurgia per lo studio

In the following periods, which were characterised by longer-distance trade relationships, the local production of silver lost part of its importance. Nevertheless, the multistage process remained as a component of the technical know-how that was then optimized and thoroughly described by the technical manuals of the XVIth century: thus confirming that " .... the concentration and rationalization of production distinguishing the lordly phase represented the stimulus towards the technical changes of the ensuing centuries" (Francovich & Farinelli 1994: 463).

di un territorio a vocazione mineraria: le Colline Metallifere nella Toscana Mediaevale.

BIBLIOGRAPHY Bachmann, H.G. 1982 The Identification of Slags from Archaeological Sites. Occasional Publications of the Institute of Archaeology, London, 6. 1993 The archaeometallurgy of silver. In R Francovich (ed.) Archeologia delle Attivita Estrattive e Metallurgiche, Firenze, pp. 487-496. Benvenuti, M., S. Guideri & I. Mascaro 1991 Inventario de/ Patrimonio Minerario e

Unpublished doctoral thesis, University of Pisa. in press I paesaggi minerari nella Toscana Mediaevale: ii progetto Massa [Paper present at the seminar organized by the Rotary Club of Massa Marittima and Follonica, January 1996]. Hawthorne, J.G. & S.C. Smith 1979 Theophilus - On divers arts. The foremost

Mineralogico in Toscana. Aspetti Nauralistici e Storico-Archeologici. Firenze. Benvenuti, M., I. Mascaro & G. Tanelli

Meneghini, D. 1947 Chimica Applicata ed Industriale [5th edition]. Milano. Moesta, H & G. Schlick 1989 The furnace of Mitterberg. An Oxidizing Bronze Age Copper Process. Bullettin of the Metals Museum XIV, pp. 5-16. Rodolico, N. 1938 Ordinamenta super arte fossarum rameriae et argenteriae civitatis Massae [Anastatic reproduction, Grosseto 1983]. Rostoker, W. & M. Sadowski 1980 The carbon reduction of fully oxidized chalcopyrite (copper) ores. Journal of the Historical Metallurgical Society XIV-1, pp. 38-42.

Mediaeval Treatise on Painting, Glassmaking and Metalworking. New York.

1995 Mineralogy applied to Archaeometallurgy: an investigation of Mediaeval slags from Rocca San Silvestro (Campiglia M.ma, Tuscany). Science and

Technologyfor Cultural Heritage. Cucini, C. in press Analisi di scorie di riduzione di minerale argentifero de/ Calisio, Civezzano- Fornace (TN) Dodwell, C.R 1961 Theophilus. The VariousArts - De Diversis Artibus, Oxford University Press, New York. Forbes, RJ. 1971 Studies in Ancient Technology vol. VIII. Leiden 1972 Studies in Ancient Technology vol. IX. Leiden.

155

Silvia Guideri

MIXED ORE-----, CRUSHING

... ROASTING

LEAD

DESULPHURIZED COPPERORE sbg ricli in Pb FLUX

f + ft=

FUEL(~----f

REDUCTION FUSION SLAG..--.J

+

►

LEAD

COPPER METALLINE AIR

FUEL (wood)----f

f + r----

S02

I L..

ROASTING OF THE METALLINE

FUEL(ch~

----f

f ++

LEAD OXIDIZATION CUPELLATION

FLUX

*

SILVER

REDUCTION FUSION SLAG

..--.J

L_.

SECOND METALLINE

BLACK COPPER (95-97"/eCa)

AIR

FUEL (green wood)

+

REFINEMENT OXIDIZATION

PURE COPPER

Figure 1: Scheme of working ''polymetallic" ores (multiphase process)

156

LEAD OXIDE

The production of metals for coinage in Mediaeval Tuscany

B

A

Figure 2: Section through the interior of a furnace. A) Loading the furnace at the beginning of the process; B) Sequence of fused material at the end of the process

SiO2 S/SO3 CaO FeO Cu ZnO Pb Sb2O3 Na2O MgO Al2O3 P2O5 K2O TiO2 MnO NiO Sn

RSS 2289 39.95 1.1

14.1 34.6 0.15 3.5 0.3 0 0 0 2.1 0 0.7 0 0.6 0 0.1

RSS 2291 39.66 1.3 14.2 34.9 0.5 3.4 0.3 0 0 0 2.1 0 0.7 0 6.6 0 0.01

RSS 2458 43.5 0.8 14.2 29.2 0.08 2 0.15 0 0 0 3 0 1.2 0 6.7 0 0.1

RSS 2459 35.3 0.8 12.8 41.3 0.09 0.5 1 0 0 0 1.8 0 0.5 0 6.9 0 0.01

PdG 102(1) 34.4 1.9 7.9 33 0.3 1 0 0 4.1 3.1 8.7 0.4 2.1 0.2 0.2 0 0

PdG 102(2) 39.6 0.4 13.1 38.7 0.1 0.4 0.01 0 0 0 4.3 0 1.7 0 0.5 0 0.01

MARS 200A 40.6 0.7 11.6 38.3 0.1 0.5 0.01 0 0 0 5.8 0 1.9 0 0.3 0 0.1

Table 1: Percentage of the chemical elements present in the slag at Rocca San Silvestro (RSS) and Marsiliana (MARS, PdG)

157

MARS 200F 34.2 1.6 14.3 41.4 0.4 0.8 0.07 0 0 0 5.3 0 2.3 0 0.4 0 0.01

The Art of Making Bread by the Charcoal Burners of the Calabrian Mountains Francesca Lugli

INTRODUCTION

Since 1991 the author, in collaboration with the Azienda di Stato Foreste Demaniali (the Italian military corps for the management of state-owned forestry resources), has been carrying out ethnoarchaeological research focused on the communities of charcoal burners in the territories of Mongiana and Serra San Bruno in the Serre (literally "mountain ranges") region (Vibo Valentia, Calabria: Fig. 1). The Serre are an uninhabited mountain region whose peaks reach an elevation of 1400 metres above sea-level; their woods, which are sometimes quite inaccessible, still today represent an ideal escape route and shelter for outlaws.

the patterns of the exploitation of the forest and to the problem of wood cutting (Lugli 1993, 1994; Lugli & Pracchia 1995: 440-442). In the Serre region, following a widespread pattern, metallurgical activities followed the production of charcoal and, therefore, the slashing of the wood, and not the contrary. For example, the Ferriere (iron extraction plants) of the Stilaro and Allaro rivers were moved to Mongiana, in order to exploit a still productive woodland (Caracciolo 1995: 12). Presently, charcoal burners carry out only the production of charcoal, and only on very rare occasions actually perform wood cutting. In central-southern Italy there are three basic ways of working as a charcoal burner: 1. he or she may be a part-time craftsperson, burning

IDSTORICAL BACKGROUND

For centuries, the activity of the charcoal burners has been associated with the extraction and transformation of metallic ores. The earliest historical reports of metallurgical activities in the Serre region seem to date back to 1094 AD: in this year the monks of Santo Stefano del Bosco (Serra San Bruno) were allowed by king Rodger the Norseman to cavar ferro (literally "to dig out iron") and, as a consequence, to produce charcoal (Franco & Riggio 1992: 10; Zadra 1995: 39-40). Since then, metallurgical activities were gradually intensified, so that in the 18th century the Borboni (the kings of Naples and Sicily) installed their Ferriere (iron foundries), which closed only around the middle of the 19th century. After the political unification of Italy (1870), charcoal burning continued to be performed in order to supply the northern regions, both for industrial and domestic consumption. Today in the Serre region the production of charcoal survives thanks to the massive demand for charcoal by the many restaurants which apply themselves to barbecue cooking.

THE CONTEMPORARY ORGANIZATION OF CHARCOAL BURNING

The charcoal burners of the Serre region, almost always natives of the region, have handed down their trade from father to son for many generations. Even if they usually own a house in their home town, for the majority of the year their life takes place in the woods, with the exception of the month of August and the most important celebrations (Christmas, Easter and the festival of the patron saint of the home town). In the past, the charcoal burners shifted from place to place after having slashed the wood, and therefore the mobility of the production of charcoal was closely tied to

charcoal only on a few occasions, in the summer near his or her house; 2. he or she may be a semi-nomadic specialist, cutting wood and burning charcoal in the spring on the plains, and in the summer up in the mountains; or 3. he or she may be a full-time, permanent craftsperson who has given up wood cutting and periodically receives the wood to be burnt into charcoal near his or her camp; in this case the camp is periodically shifted according to its particular needs (Lugli 1993, 1994; Lugli & Pracchia 1995: 440-441). Thanks to the availability of mechanical devices, the charcoal burners get huge amounts of wood near their camps, and this allows for a lesser mobility in the territory than was possible previously. Nonetheless, the camps are periodically abandoned, either for management requirements by the forest authority or because of the need for new wooded areas. A camp may be inhabited from a minimum of a single season to a maximum of approximately fifteen years; sometimes the camps are re-settled. Such a shifting pattern provides ethnoarchaeologists with a precious opportunity for studying the local settlement and abandonment phases.

THE STUDY AREA: CAMPS AND BREAD OVENS In 1997, the writer carried out a preliminary survey of

charcoal burning camps in the Serre region, with the purpose of mapping the settlements, both inhabited and abandoned, to be selected for future research. The camps of the charcoal burners of the Serre region may include from a minimum of one to a maximum of 20-25 nuclear families, usually with five or six members each. Each family owns a

159

Francesca Lugli

wooden hut, with a roof covered by straw, tarred paper and plastic, a shelter for a car, a covered structure used as a toilet, a fence sheltering some domestic animals (on average each family owns three sheep, three pigs and seven to eight chickens), plus a kitchen using wood or charcoal in front of the main hut, and the oven for baking bread (Fig. 2). Such ovens have been observed in all the charcoal burners camps visited by the writer, and also in other regions of southern and central Italy (Lugli 1993, 1994). We studied the ovens of the camps at Inticco, Boscoreggio, Pizzenti and Centofontane; their different state of conservation satisfactorily represented the various phases of use and abandonment of these peculiar features.

one of the families used to periodically visit the nearest village in order to buy its bread. When two or more nuclear families are related by a first degree relationship, the bread oven and the animals may be shared (Fig. 3). This peculiar spatial and functional organization has been observed in the camps at Pizzente, Centofontane, and in the camp at Croceferrata, not far from Serra S. Bruno. The camp at Croceferrata has been inhabited for 15 years. The charcoal burners have bought their own land at the site and built, alongside the traditional wooden structures, huts made with metal sheet and, in one case, a house built in masonry. In this camp immigrant workers also live who are hired by the charcoal burners. Although Croceferrata, like the other camps, has its own bread ovens, due to its anomalous social context and relationships of production it was not included in this preliminary study. In terms of archaeological analogy, we learn from this simple example that the sharing of such an important domestic facility among various huts might imply the existence of close kinship relationships among their inhabitants.

The camp at Boscoreggio (Fig. 2), located between Mongiana and the Speranza track, was suddenly abandoned in the winter of 1997 because the local charcoal burner had been arrested, having been charged with the accusation of hosting fugitives belonging to the 'ndrangheta (the Calabrian mafia). It was a one-family settlement, with a family composed of four people, and had one bread oven. Inticco (Fig. 2), near Mongiana, is a camp where charcoal burners have alternated for 15 years. In 1997, there were two families, one composed of five individuals, namely the father, the mother and three sons. Another two sons and one daughter lived elsewhere: the girl was married and lived in the town of Serra S. Bruno, one of the sons lived with his sister, while the other earns his living by cutting wood near Siena in Tuscany (central Italy). The other family was composed of three individuals. One bread oven was visible in this camp. The two nuclear families were kin-related, but not through a first degree relationship. In 1998, during a subsequent visit carried out with A. Stoppiello, we learnt that one of the families had been shifted to another area by the contractor firm.

BUILDING AND USING A BREAD OVEN: OPERATIONAL SEQUENCES

Within the communities of charcoal burners, the production of charcoal is shared by men and women, while building is a traditionally male activity. Therefore, the building and maintenance of the bread oven is a male activity, while the production of bread is entrusted to the women. The bread oven is built beside the family's huts in a single day's work. The actual oven is formed by a hollow cone-shaped infrastructure built of bricks, usually old hand-made bricks recovered from ruined buildings, and coated with an outer layer of clay. According to the elder craftspersons of the Serre, in the past the oven used to be built with stones.

The camp at Pizzente, near Serra S. Bruno, has been inhabited for ten years by three families, amounting to a total number of ten people. Two nuclear families were kinrelated through a first degree relationship. In the camp we saw three bread ovens, one of which was in a state of abandonment and partial collapse. Prateria is a one-family camp inhabited since the winter of 1997, and had a bread oven built against a slope. Finally, Centofontane, beside the road to Lamezia Terme, is located at an altitude of 600 metres above sea-level and, isolated from other settlements, was abandoned 15 years ago. Although no standing parts were visible, and the ground was covered by vegetation, the architectural features were still easily recognizable. We saw two huts, a bread oven and several charcoal burning grounds.

AN ETHNOARCHAEOLOGICAL MODEL

According to our data, in these camps each family usually owns its bread oven. In the camp at Inticco, inhabited by two families, we saw only one baking facility: as a matter of fact,

The oven is built over a platform made of earth and chaff: sometimes contained in or supported by a substructure made of logs. The floor on which the oven sits is therefore suspended, and craftspersons say that this grants an effective insulation from ground moisture. At the same time, the ground is not affected by the heat of the oven: in fact, Fig. 9 shows a dog peacefully sleeping during the baking of bread, although the oven has been alight for hours. All around the oven, charcoal burners often excavate a small drainage trench while in the centre, between the oven, the outer kitchen area and the huts, one may see a floor made with stones, bricks or broken tiles. The oven is always oriented against the prevailing winds, and sheltered by an upper roof with two (Fig. 4c, Fig. 6) or more eaves (Fig. 4a, Fig. 5) or protected by building the structure against a slope. The most common type has double eaves. Among the camps we visited, two ovens had more than two eaves (Inticco and one of the ovens at Pizzente), while the other facilities had double eaves. This type of oven is used exclusively for the production of bread, once a week or once every two weeks, according to the size of the families. A single oven may be

160

The art of making bread by the charcoal burners of the Calabrian mountains

used for several years; the oldest infrastructure we documented was a double-eaved oven in the camp at Pizzente which was nine years old. Figures 7 and S show the various stages of the chaine operatoire of the construction of an oven: first, the building material is acquired and collected on the spot (Fig. 7a); • the ground is levelled and cleaned from wild grass (Fig.

•

7b);

•

• • • • •

•

• •

four parallel logs are laid on the ground within a square space delimited at the corners by four vertical poles (Fig. 7c); or, as an alternative, two horizontal trunks are placed on top of these logs (Fig. Sa); above this structure a floor is laid made of parallel logs (Figs. 7d, Sc, Sd); two lateral logs are placed for containing the platform (Figs. 7e, 7f, Se); the surface is filled with earth and chaff: for a thickness of 10-15 cm (Figs. 7g, Sf); a floor ofbricks is laid on top of the earth (Figs. 7h, Sg); the cone-shaped dome of the oven is built. The base diameter of the cone measures 120-130 centimetres and the height is about 130 centimetres. The total number of bricks used is 400-430. The charcoal burners insert a stone inside the brick structure, which has the function of a ''pyrometer": from the colour of the stone they can assess the temperature of the inner chamber. A horizontal steel supporting bar is inserted in the upper part of the oven's mouth which functions as a lintel. The mouth is closed with an iron sheet or a wooden door called timpagnu in the local dialect (Figs. 7i, Sh); the cone-shaped structure is coated with a layer of clay (at Inticco this outer layer is further coated with an additional layer of cement) (Figs. 7j, Si); the logs for supporting the roof are set in place (Figs. 7k, Sj); finally, the roof structure is covered with tarred paper and plastic sheets (in former times, with bark and branches) (Figs. 71, Sk).

The manufacture of bread starts early in the morning, with the preparation of the yeast. The latter is left to rest for three hours; after this it is kneaded, and after lunch it is mixed with flour and water to form the paste. Bread is prepared by the women indoors, using a special wooden tool called a majed(h)a, a rectangular board with a central cavity and two lateral flat supporting projections. Women use two different majed(h)a boards, one for kneading bread and the other for pork sausages. If this tool is not left out under the sun which tends to crack it, it may last for about 15 years. When the paste is ready, the first product to be put into the oven is pitta (a local type of thin bread, not to be confused with pizza), ready in about 10 minutes; this kind of bread is consumed on the same day. Usually, before the bread loaves are baked, pitta is eaten with olive oil and chillies, or with boiled pork parts such as ears or kidneys conserved in lard

(called frittole ), or with tomatoes, olives and anchovies. Next comes the bread, which requires a baking time of about three hours. On both the pitta and on the bread loaves one may see incised signs, and different families seem to use slightly different signs. Bread making for the charcoal burners plays the role of a festivity ritual, often corresponding to the holy Sunday, which is charged with shades of magic and superstition. In order to bake a loaf of bread, people employ not less than 40 kilograms of wood; whenever possible, heather wood is preferred. The oven is heated for about an hour and a half before the bread is inserted, and the actual baking takes approximately three hours. Usually, the task of lighting the oven is performed by the male members of the family. When the inner brick wall reaches a white-heat temperature, the oven is judged to be ready for baking bread, although the stone is deemed to be a more reliable indicator. At this point, the ash is cleaned from the oven floor. For this purpose, the charcoal burners use a stick with, at one end, a bundle of wet cloth: this tool is called cadipu. A handful of flour is thrown into the chamber. If it burns, the floor is cooled, always by means of the cadipu. Before putting the bread into the oven, a white cross is drawn with flour on the oven's mouth for good luck.

ECONOMY OF BREAD PRODUCTION

Far from the influence of the Catholic creed, and sometimes in close contact with outlaws, the charcoal burners of the Serre region have always been viewed with suspicion by people living in permanent villages and towns; social relationships between these two types of communities have traditionally been sporadic and difficult and, still today, as a rule charcoal burners get married with members of families with the same profession. The purchase of food represents an obvious and recurrent occasion for social interaction. In fact, the procurement of flour (today possible at the grocery shop of the nearest village) is one of the most important needs of the charcoal burners. The daily need of flour per person is about one kilogram, calculated on a rate of about 600-700 grams for bread, and 300 grams for pasta. On the occasion of religious celebrations, this amount is increased since it has to include the flour used for the production of cakes and biscuits called zippole. Every week, an average family of four to five people therefore consumes not less than ten to fifteen loaves of bread, and five to six loaves of pitta, for a total amount of about 40 kilograms of flour and 0.5 kilograms of salt.

ABANDONMENT AND TAPHONOMIC PROCESSES

In some cases, the oven is abandoned while the camp is still inhabited; it thus becomes a collecting place for various kinds of refuse. When a camp is abandoned, in most cases

161

Francesca Lugli

the charcoal burners take away the steel bar which had been placed as a lintel on the oven's mouth, with the purpose of re-using it for the next oven. As a consequence, the brick cone-shaped structure collapses onto itself Our informants reported the use of dismantling the ovens for the purpose of also recycling the stones (originally used in place of the bricks). The old, hand-made bricks are sometimes deemed to be a valuable possession, and it is not surprising to see the whole structure dismantled, and the bricks transported to a new camp to be re-used for a new construction. In contrast, the wooden roof covering is almost always abandoned on the spot, and reportedly this happened in the past as well.

huts may suggest the intervention of important kinship relationships among the dwellings. Bread making, the related technical processes, as well as the consequent taphonomic processes, involve a certain range of variability. As far as the cycles of camp and bread oven building and bread production are concerned, technological choices are a function ofboth the individual history of the families and of wider-scale patterns of adaptation by the craft communities; the observation of these variations within the chaine operatoire may reveal important, often unexpected glimpses of social life in contemporary southern Italy.

In summary, when a camp is abandoned the oven may be spoiled of its bricks, but it is not uncommon to find that it was abandoned intact (Fig. 10), as in the case of Centofontane (Fig. 11); in the case of the camp at Boscoreggio (Fig. 12), as we have seen, the sudden desertion of the site was due to an unexpected event. After the abandonment, the camp underwent a gradual pilfering by local inhabitants, as well as by the charcoal burners which came back to the same spot. First they removed the most valuable or useful items while later, in the course of the following months, almost every structural element was removed. This process of pilfering and dismantling is more effective when the site is near a village or farmland. In the case of Centofontane, which is isolated from other settlements and which was not re-settled, the original buildings which were partially undisturbed and collapsed in situ were still easily recognizable.

REFERENCES Caracciolo, V. 1995 Gestione del bosco e conservazione dell'ambiente nella legislazione del Regno delle due Sicilie, con particolare riferimento al bosco di Stilo ed il funzionamento degli stabilimenti metallurgici di Mongiana e Ferdinandea. In Associazione Culturale Ritomo Emigrati (ACRE), Atti de/ Convegno sulla Fabbrica d'Armi, Mongiana (VV), pp. 30-38. Lugli, F. 1993 La technique de la preparation du charbon de bois dans une communaute de charbonniers de la province de Viterbo. Paper prepared for the D.E.A. course on 'Technologie Comparative en Archeologie', Universite de Paris I, SorbonnePantheon. 1994 II carbone di legna - osservazioni presso una comunita di carbonai de/ casertano. Paper for the course on 'Paletnologia', Specialisation School in Archaeology, University of Rome "La Sapienza". Lugli, F. & S. Pracchia 1995 Modelli e finalita nello studio del carbone di legna in archeologia. Origini XVIII: 425-479. Franco, D. & S. Riggio 1992 Memorie industriali in Calabria - Siderurgia e Metallurgia nelle Serre Calabre. Quademi A.C.A.L Bivongi (RC). Zadra, E. 1995 Le ferriere: cenni storici. In Associazione Culturale Ritorno Emigrati (ACRE). Atti de/ Convegno sulla Fabbrica d'Armi, Mongiana (VV), pp. 39-52.

Nonetheless, as a rule it is rather difficult to find in situ structural features in the abandoned camps, and the bread oven may be exceedingly difficult to identify. At a macroscopic scale of analysis, one may recognize rare charcoal bits scattered on the original location of the oven, or in the space immediately in front of it. Furthermore the ground, as stated above, does not show any alteration caused by heating.

CONCLUDING REMARKS These data suggest that without any previous knowledge of such a peculiar ethnographic model, it would be virtually impossible to reconstruct the original form and function of the oven on the basis of what survives. On the other hand, as we have seen, the location of the oven among the various

162

The art of making bread by the charcoal burners of the Calabrian mountains

Fig. 1: A map of Calabria, in southern Italy. The Serre region is comprised between the town ofVibo Valentia and the eastern coast

163

FrancescaLugli

Bosc:oreg-9io

t::'i

~~\

--

Inticco

Fig. 2: Maps of the camps at Inticco (below) and Boscoreggio (above) with the main architectural features and infrastructures. At Inticco the charcoal burning area is placed near the huts, while at Boscoreggio it is further away. In both cases, the animal pen is also located at a distance. The camp at Inticco, at the moment of the visit, was still inhabited, while Boscoreggio was abandoned in the winter of 1997

164

The art of making bread by the charcoal burners of the Calabrian mountains

~I.

Fig. 3: When two families which occupy the same camp have a close kinship relationship, they often share the bread oven and the kitchen area

165

FrancescaLugli

A

PLAN

FRONTVIEW

,«t - ~.½'

B

_:f/};;;1,0.·

·•·.•...:1,z r .'/;,:; I, ~ ,/.-

..

I

I

l

J FRONT VIEW

PLAN

C

FRONTVIEW

Fig. 4: A: the multiple-eave roofed oven at Inticco; B: the oven at Prateria, without a roof; built against the slope; C: the double-eave roofed oven at Pizzente

166

The art of making bread by the charcoal burners of the Calabrian mountains

Fig. 5: The double-eave roofed oven at Pizzente

Fig. 6: The multiple-eave roofed oven at Inticco

167

Francesca Lugli

q--=.?' __ ..;-__· ---u_j) {...;....-_; :::..z= ...3=.::._-.,00

a

b ::: .. 6l

C

-

--

.

-~

~

•I ' • 1 . ' . t t \ i .,' ' I I 1'• , . . . - . ·- - I

~

I

I

I

I

'

(

I .,

{ ~

~

•d

e

d

~ g

~

:.=s:J ~ f

~ h

k

Fig. 7: The operational sequence of the building of a double-eave roofed oven placed on a base oflogs. The four corner poles support an upper wooden platform

168

The art of making bread by the charcoal burners of the Calabrian mountains

,..

-

-·..... .

Q

ID

=

a

--

•• ----:10o

l

--

l

- --

J

b

d

-C

P4◄iq

e

g

f

h

k

Fig. 8: The operational sequence of the building ofa multiple-eave roofed oven suspended on four comer logs

169

FrancescaLugli

Fig. 9: A dog sleeping under the oven while the bread is being baked

Fig. IO: An abandoned intact oven on the road to Vibo Valentia. This oven had no roo£ One may observe how the brick infrastructure collapses onto itself

170

The art of making bread by the charcoal burners of the Calabrian mountains

Fig. 11: An oven abandoned 15 years ago: the camp at Centofontane.

Fig. 12: The oven of the camp at Boscoreggio, abandoned in the winter of 1997

171

Towards a Theory of Social Production and Social Practice Pedro V Castro, Sylvia Gili, Vicente Lull, Rafael Mic6, Cristina Rihuete, Roberto Risch & Ma Encarna Sanahuja nl

Western Mediterranean, namely south-east Iberia and the Balearic Islands (e.g. Gasull et al. 1984; Chapman et al. 1987; Castro et al. in press b, in press c).

INTRODUCTION

Seldom has archaeology been so overwhelmed by empirical data, by increasing specialization and fragmentation of research, and by private management of the public heritage, as in the 90s. Despite an apparently more open world, never does archaeology in Europe seem to have been less critical, less communicative, and further from becoming a social tool for a better understanding of the past and, ultimately, changing the present. In our view, the increasing lack of historical content in the archaeological discourse, which all over Europe seems to have become lethargic, demands new and different ways of thinking and organ1zmg archaeological theory, as well as practice. Changes can only take place if technical and disciplinary fragmentation is reintegrated into a general archaeological discussion, if individual academic protagonism is substituted by team work and collaboration, and when private competition over the greater part of the archaeological practice, i.e. rescue excavation, is brought to an end.

The theory of production of social life forms the main part of this theoretical framework, as it establishes the conceptual structure, which provides historical explanation of the archaeological phenomenology we observe (for a more detailed discussion of the development and application of this theory see Castro et al. in press a, in press c). The aim of this theory is, first, to identify and to explain the objective conditions on which the production of social life is based, and, secondly, to determine whether the social relations established in and between societies exploit, hide or alienate the social subject.

THE STRUCTURE OF SOCIAL PRODUCTION

The initial premise of the theory of production of social life is that any type of social life requires the existence of three objective conditions: men, women and the material objects, which are used by them and which imply a socialization of certain parts of the physical world. The material expression of these three objective conditions forms the social matter. All social matter has to be produced, which implies that men, women, and material goods are social subjects, as well as .social objects. They operate as subjects in the moment they allow and decide, as social agents, the reproduction of society. They operate as objects in terms of their own production and in the relationships they establish between themselves. Thus men, women and material objects participate in an ambivalent way in the transformation of the physical world. This lies at the basis of the opposition between vulgar materialist and idealist forms of understanding society, but can probably be overcome through an approach that integrates into the historical analysis the social object, as well as the social subject, the products as well the agents, and which ultimately considers matter as well as energy.

Yet, the development of archaeological science, as of any other discipline, also demands a continued and dialectical discussion about the general ontology and epistemology that shall structure our research on past and present societies. Most theoretical developments, such as processualism, postprocessualism, Marxism or others, seem to have stagnated after a rather dynamic phase during the 70s and 80s, though possibly one of the few exceptions in this situation is the development of feminist archaeology. These conflicting discussions have been substituted by a whole range of mutually ignoring or independent research lines on particular aspects of social organization (ethnicity, gender, ritual, social complexity, operational sequences, etc.). Their results inevitably remain fragmentary, if not wrong, as the material evidence on which any archaeological interpretation or description is based cannot be disarticulated into different meanings without a general sociological theory that establishes the relationship between the different forms of phenomenological expression of social reality. In an attempt to proceed in another direction, different Marxist and feminist approaches have been integrated into a general theoretical framework. This consists of a set of interrelated theories, that range from the explanatory level of social organization down to the registration and description of their phenomenology. These have been developed and applied through a series of archaeological and ecological projects concerning the later prehistory of the

173

All social transformation of the world implies the combination of a series physical elements with a given expenditure of energy. On these factors the successive paradigms developed by western economic thought have been constructed. They can be articulated in what we have defined as the basic economicscheme:

LO+LF+MP~P where:

Pedro V. Castro et al.

LO = labour object; it represents the material basis or physical support of the social production. In the form of land it has been considered from Aristotle to the physiocrats as the main factor in the reproduction of society, as it is the main natural element that in terms of matter and energy provides a direct social benefit. LF = labour force needed for the economic activities; this refers to the effort men and women as social subjects undertake during production. Individual skill, information and experience increase the efficiency of the labour force, and therefore represent objective elements of the technical and social division of labour. The relevance of human labour and, therefore, the idea of the subject in social production only became fully acknowledged with Adam Smith and the development of Political Economy. MP = technical means of production; these are all the technical elements which become implemented in economic activities, mediating between LO and LF. Possibly one of Marx's most important contributions to economic theory was to explain the central role of the means of production not in generating, but in transforming energy in more socially efficient ways. P = the final product; this is the target and necessary condition of every economic practice, forming a good which is needed for, desired by or imposed on social reproduction. The final aim of all production is consumption. Production is a moment of consumption, but consumption is also a moment of production. Each element is at the same time its contrary (Marx 1973: 81-100). Therefore, any analysis of social production also has to deal with the individual consumption of what is produced, and vice versa. Yet, this unity does not imply identity, given that between production and consumption a spatial and temporal movement always takes place, which we call distribution. Given that production forms the general/social aspect, and consumption the singular/individual aspect of social production, distribution acts as the bridge between the social and the individual. It is this sphere of the production-consumption transitivity where the social relations of production are objectivelyestablished, and where the social matter becomes the subjective expression of the social experience, generating the communicative and symbolic structures and the aesthetic values necessary in any form of social organization. Finally, the movement from production to distribution and consumption results in and allows the reproductive cycle of society to take place. In social and economic theory, production has normally been a concept used in the singular, referring only to the generation of goods, considered as the bases of all economic activity. In this way a whole set of social practices, which fulfil the conditions of the basic economic scheme, have been omitted from the social analysis. In order to overcome this unequal and partial evaluation of social labour, it is considered that all societies reproduce themselves through

three types of production: basic production, in which women guarantee the biological reproduction of society, the production of material objects, responsible for the generation of utilities, and maintenance production, which is intended to keep in a satisfactory state the properties and capabilities of social objects and subjects. To place all three types of production at the same level of necessity in the social reproduction inevitably requires their joint evaluation in any kind of research. Basic production refers to the generation of new subjects, which provide the labour force necessary for social reproduction. To recognise this form of production means to consider biological reproduction as a specific and socially necessary labour process, and avoids explicitly all forms of naturalisation or occultation of it. It implies for women an activity, which only they can undertake, but which renders a collective product, as the new subjects become part of social life. Therefore, basic production has to be materially compensated if sexual exploitation on the basis of biological reproduction is to be avoided. Object production refers to the generation of food supplies and all other types of products designed to be used or consumed. Basically one can distinguish between subsistence goods, means of production and artifacts for direct consumption. In each case the distribution and individual consumption of these goods present specific characteristics. This form of production has normally been over-emphasisedby modern economic theories.

Finally, maintenance production allows the increase of the social value of things without changing their use value, either artificially or through the improvement of the physical, chemical, affective and aesthetic characteristics of social subjects and objects. This production is a key element in order to increase the production force and to avoid the exhaustion of subjects and objects. The recognition of maintenance production allows us to overcome the hierarchical differentiation between production and services, which has led to an unequal economic evaluation of producers, mainly male, against servants, mainly female. Nevertheless, the dependency of maintenance production in relation to the other two types of production lays the basis for a social dependency of the agents, who specialise in activities of maintenance or services. For this reason the labour invested in maintenance has tended to be considered oflow social value, despite the fact that most social products would be of short existence or used without this type of implementation. When these three forms of production are analysed in terms of the basic economic scheme, it becomes clear that men, women and objects can or must participate in very different ways in the production of social life. Natural resources (LO), labour force coming from men or women (LF) and the products which are used as means of production (MP),

174

Towards a theory of social production and social practice

provide the necessary conditions for object production. The obtained product (P), is normally individually consumed.

material consequence of this development is that some part of P becomes surplus value. This is that share of production which does not revert in any form to the group or individual that has generated it. How the surplus is produced relates to the economic analysis, yet its forms of appropriation and consumption concerns the social organization of the community. Therefore, the two questions which should direct any social analysis are:

In basic production the body (LO) and the energy (LF) of women fulfil all the factors of the scheme. The product is incorporated into social life in the form of new men and women. This supposes a fundamental sexual and social difference in society, and necessarily affects, by means of compensating or exploiting strategies, the organization of all other forms of production. The direct relation LO-LF has been used to legitimate sexual division of labour, naturalizing reproduction by reducing women to a labour object (LO) and omitting the physical effort (LF) necessary in the generation of new social subjects. Furthermore, the fact that in basic production no specific instruments (MP) are needed, implies that an increase in production is not possible through technical improvement, as in the other forms of production, but only through a greater physical effort of women. Women are confronted with the social relations of production with only one means to avoid exploitation: to remain in control of their bodies and to obtain material compensation for the products they generate.

1. who undertakes the social practices of the production of men, women, objects and the maintenance of all of these, and how do they do so? and 2. who benefits (consumption/use/enjoyment) from the resulting products? Depending on the answers we can approach social asymmetry, when exploitation is established between classes, or sexual differentiation when exploitation is exercised on the whole or parts of the female community, or a combination of both. In a general sense, the triad exploitation-property-surplus value is the expression of a society divided in terms of the participation in and access to forms of social production. The degree to which these differences become imposed, maintained and institutionalised defines the level of social differentiation and ultimately results in the formation of social classes and their political correlate, the state (Lull & Risch 1996).

In maintenance production the initial LO is already a product (men, women or objects), the LF can be provided by men or women, while the MP can be other products, or nonexistent. The frequent absence of a specialized technology has supposed that these activities have often been considered as natural aspects of social life, rather than as economic practices, and have often been ascribed to the domain of the woman. In this form of production the final product is individually consumed although .seldom, or only partially, by its producer.

The question about the allocation or appropriation of the products generated by the three types of production, makes it clear that the historical approach to social organization cannot restrict itself to the analysis of the transformation of energy into social matter (e.g. Caracciolo & Morelli 1996). In order to move from the description of the physical world and its dynamics, to an explanation of its social transformations, it is necessary to consider the problem of the distribution and consumption of the produced matter. This inevitably leads us to take into account the differentiated participation of the social agents in the transformation and appropriation of the physical world or, put in another way, to the distribution of the material and energy costs and benefits within society (Risch 1995).

Any, some, or all of the factors established by the basic economic scheme in the three forms of social production are liable to be appropriated by individuals or groups and transformed into property. Therefore the analysis of surplus value results ultimately in an historical analysis of the dialectical relationship existing between production and property. In the case of object production, the understanding of the appropriation of any of the economic factors involved lies at the basis of the definition of the modes of production in classical Marxist theory: feudalism is constructed on the control of the main LO (land), the slave mode of production is based on the direct appropriation of LF, capitalism implies the property of the MP, while the Asiatic organizations are centred on the unequal access to the factor P. Yet, if we also consider basic production and maintenance production, the analysis of social exploitation and its historical origins becomes much more complete, sexually differentiated and the social subject is not eliminated from production, as has occurred in many Marxist as well as functionalist approaches. The combination of the relations of production established in the three forms of production defines the modes of social reproduction.

The asymmetry which can arise in these terms between social production and individual consumption determines the level of economic exploitation and social distance reached by society. Economic exploitation implies an unequal individual appropriation of social production, and necessarily results in the existence of property. The direct

The difficulty for archaeology in defining surplus value through archaeological objects lies in the fact that artifacts are at the same time products and utilities. The existence of surplus value in a society is not determined by differences in the consumption of products, but by the differences in the social value of what is being individually consumed.

SOCIAL DIFFERENTIATION AND SURPLUS VALUE

175

Pedro V. Castro et al.

Furthermore, the social value of things is not absolute, but depends on the material costs and benefits of their production and consumption. This means that surplus value can only be determined:

record in terms of a limited set of techno-economic factors, whose relation to social organization has not been formulated.

1. through a global analysis of the types of social production, along the lines just described, and 2. through the definition of the function the artifact plays in the basic economic scheme of each form of production.

ARCHAEOLOGICAL IMPLICATIONS

The procedure proposed consists in the analysis of the factors of the basic economic scheme in all three types of social production in each of its moments, i.e. production, distribution and consumption/use of the social matter (e.g. Castro et al. in press c). Apart from its psychic, symbolic and aesthetic implications, any strategy of surplus value production also has direct physical consequences on the social matter, as it depends on the potential to force surplus labour and to alter or vary the social value of products. Thus, a strategy of aiming towards absolute surplus value implies an intensification of the labour force (LF) under constant technical conditions (MP). On the other hand, the production of relative surplus value reduces the social value of products through a development of the means of production (MP) and/or an improvement of the labour object (LO), leading to an increase in the productivity of the labour force (LF) (Marx 1959). It is the responsibility of archaeological theory to find the ways of recognising these material changes. The question regarding social distance and the notion of surplus value articulates and gives sense to an empirical analysis of production processes, social division of labour and the social mechanisms of appropriation of surplus production. Standardisation of the artifacts, technical exclusiveness of the spaces of production, simplification of the production processes and production volume are the axis along which the economic strategies of surplus production can be defined. The spatial distribution of labour objects and means of production, the relationship between natural resources, production areas and consumption areas, and nutritional and health patterns as identified by the human remains, are the main archaeological features which allow us to identify the social organization of surplus production and consumption. This also implies that archaeological remains cannot be understood as isolated objects with a unique meaning, but should be recognised in the multiple forms in which they express the appropriation of nature and society. The theoretical weakness of a concept such as an 'operational sequence' consists in its marginal relation to a general sociological theory. Independently of the cognitive aspects one wishes to attach to this concept, it implies a reduction of production to technology, an identification of technological processes with social relations to production and an inductive research procedure. Its application necessarily leads to an understanding of the archaeological

In order to put the theory of production of social life into practice in terms of the archaeological record, a series of interrelated descriptive theories and methods have been developed.

The aim of the theory of archaeological sets is to obtain, through archaeological excavation, groups of synchronic objects with a socio-historical meaning (Castro et al. in press c). The archaeological set is not a certainty nor an empirical fact, but an explanatory hypothesis about the observed sediments, archaeological objects and their meaningful associations. In this sense it represents an attempt to overcome the difficulties of the inductive and unit orientated Harris-system. The theory of archaeological objects articulates and defines the material sphere in which social practices and natural conditions interact, forming what we have defined as artifacts, arteuses and circundata (Lull 1988). The materiality of the archaeological object informs us about the environment (circundata), about its dimension as a socially appropriated part ofit (arteuso), and about its dimension as an artificially transformed material (artifact). Archaeological objects and social spaces do not constitute units of meaning, but present multiple spheres in which the dialectical relations between nature and society, and between at least two sectors of society, male and female, become expressed. The theory of social space articulates and defines the spatial sphere in which social practices and natural conditions interact (Gili 1995). Social space represents the integration of the environment into te"itories and landscapes. While in the territories the appropriation of natural resources and products takes place, the symbolic universe through which society perceives the environment and the territories configures the social landscapes. The theory of social practices operates at the highest descriptive level of the empirical world we observe (Castro et al. 1996). The social practices constitute the relational contexts between men and women, and between them and the material world they use, generate and represent. The socio-parental, socio-economic and socio-political practices result in experiences and consciences and therefore form the phenomenological expression and the factual universe of social existence. This brief description of this sociology, and the theories necessary for its application, might have put forward the range of problems and considerations archaeological theory

176

Towards a theory of social production and social practice

needs to approach in order to arrive at a better understanding of the development and organization of society. It might also have shown that such an aim can hardly be achieved through an academic proliferation of unconnected interpretative :frameworks, nor through an arbitrary and superficial recovery of the archaeological record. Quantity of interpretations and of data cannot be mistaken for a situation in which plurality of and conflict between ideas generates an advance towards the explanation of society. The theory of production of social life attempts to propose an alternative conceptual structure for such an explanation in terms of social matter.

south-east of the Iberian Peninsula. Brussels: DGXII of the EC in press c Proyecto Gatas 2. La dinamica arqueoecol6gica de la ocupaci6n prehist6rica. Sevilla: Junta de Andalucia. Chapman, R W., V. Lull, M. Picazo & M.E. Sanahuja Yll 1987 El Proyecto Gatas. 1. La prospecci6n Arqueoecol6gica. Oxford: British Archaeological Reports International Series 348. Gasull, P., V. Lull & WE. Sanahuja 1984 Son Fornes I: La Fase Talay6tica. Ensayo de reconstrucci6n socio-econ6mica de una comunida,d prehist6rica de la is/a de Mallorca. Oxford: British .Archaeological Reports International Series 209. Gili, S. 1995 Tellitorialidades de la Prehistoria Reciente Mallorquina, Publicacions de la Universitat Aut6noma de Barcelona, Bellaterra (ed. microfilm). Lull, V. 1988 Hacia una teoria de la representaci6n en arqueologia. Revista de Occidente 81: 92-76. Lull, V. & R Risch. 1996 El Estado Argarico. Verdolay 7: 97-109.

BIBLIOGRAPHY Caracciolo, A. & R Morrelli 1996 La cattura dell'energia. Rome: La Nuova Italia Scientifica. Castro, P., R Chapman, S. Gili, V. Lull. R Mico, C. Rihuete, R Risch & M.E. Sanahuja 1996 Teoria de las pnicticas sociales. Complutum 6: Homenaje a M Fernandez-Miranda: 35-48. Castro, P., R Chapman, S. Gili, V. Lull. R Mico, C. Rihuete, R Risch & M.E. Sanahuja in press a La producci6n de la vida social. Mecanismos de explotaci6n en el sudeste peninsular (c. 3000-1550 cal BNE). In P. Nocete & 0. Arteaga (eds.) I Congreso de Arqueolog'a Social. Huelva: La Rabita. in press b Aguas Project: Palaeoclimatic reconstruction and the dynamics of human settlement and land-use in the area of the middle Aguas (Alemria) of the

Marx, K. 1959 El Capital. Critica de la economia politica I. Mexico: Fondo de Cultura Econ6mica. 1973 Grundrisse: Introduction to the critique of Political Economy. Harmondsworth: Penguin. Risch, R 1995 Recursos naturales y sistemas de producci6n en el Sudeste de la Peninsula Iberica entre 3000 y 1000 ANE. Bellaterra: Publicacions de la Universitat Aut6noma de Barcelona (ed. microfilm).

177

Operational Sequences Beyond Linearity Massimo Vidale

archaeological excavation needs to be adjusted to the degree of detail required by this specific type of organizational research.

INTRODUCTION

Paleotechnology,the reconstruction of ancient technology, is based upon various fields of research, such as archaeological stratigraphy, archaeometry and ethnoarchaeology (Vidale 1992). It is sometimes presented as a highly sophisticated field requiring complex knowledge and depending upon esoteric specializations and this, at least in some cases, may be certainly true; however, on the other hand, it is important to stress that paleotechnology is probably the easiest conceivable form of archaeological research. The reason for this paradox is that technical systems, in other words organized strategies of human behaviour aimed at the transformation of nature, are among the simplest components of cultural systems, and are by far the most consistently represented in the archaeological record. Technical systems are relatively simple to study because they may be described as linear trajectories in which, given the relative position of some steps, the position of the ''missing links" is predetermined. Furthermore, within most manufacturing cycles the internal technical events, both reductive and pyrotechnical, leave a rich body of decayresistant material evidence in the archaeological record in the form of archaeological craft indicators: layers, interfaces, manufacturing waste and powders, tools, and fragments of infrastructures which can be traced back to their source operations (Tosi 1984; Vidale 1992). In the archaeological record, as a rule, manufacturing techniques are easier to understand and reconstruct than communication or rituals might ever be. Manufacturing techniques, therefore, represent a realistic target for reliable archaeological reconstructions.

Once operational sequences have been reconstructed, the second context to be explored on the basis of the evidence thus provided is technology. Technology, besides its internal technical processes and their material components, includes other cultural aspects such as inherited technical traditions, technical know-how, mental templates, symbolic behaviour, manufacturing styles, and a wide range of social relationships among traders, producers, users of the given goods and their social contexts (a great part of these relationships would fall under the Marxist label of "relationships of production"). We may look at ancient operational sequences from two other, complementary points of view. First, the representation of operational sequences in the form of a linear progression of steps, which is reflected in the French term chaine operatoire, appears to be a useful tool for describing manufacturing techniques although sometimes, particularly when research is restricted to the lowermost level of palaeotechnological reconstruction, it may limit the understanding of the wider implications of technology (Vidale, Kenoyer & Bhan 1992). Using the terms originally proposed by M.B. Schiffer (1972, 1987), I suggest that linearity is useful for understanding the archaeological context, but might hinder our capability of understanding the systemic context of the living culture.

Once a manufacturing technique has been archaeologically reconstructed and pieced back together into a formalized hypothesis of an operational sequence, the ideally linear flow of this latter may be used as a guiding thread for exploring a set of much wider implications. The first context to be explored is represented by the formation processes of the archaeological record where the original indicators have been found. Once all the archaeological craft indicators have been associated with well defined internal manufacturing operations or steps, their variable distribution within the stratigraphical context of a site may reveal important aspects of the overall organizational strategies of ancient labour. In fact, many activity areas are distinguished by means of complex pluri-stratified deposits, and the differential distribution of archaeological craft indicators among these depositional units may reveal where, when and with which time intervals some critical operations were performed. Obviously, in such cases the scale of resolution of

Secondly, it is by focusing on various forms of deviation from the ideal, linear path of manufacturing techniques that we may investigate technology, by enhancing some basic aspects of the social relationships involved in craft production and, ultimately, important parts of the overall organization oflabour in an ancient society. These two aspects are closely dependent upon one another. In fact, I believe that when we describe technical behaviour

in the form of a linear progression of discrete operations or "steps" leading to the finished product, we produce a highly ideal, abstract model that should be considered as a refined analytical tool rather than, as often happens, the final aim of our research. Its utility rests mainly in the possibility of testing our understanding of the archaeological record on functional and organizational grounds: for example, once we know that indicators A, B and C are functionally and chronologically interconnected, then the stratigraphic units 4, 5 and 6 which contain them must also be connected; if indicator D is absent from stratigraphic unit 7, it must have

179

Massimo Vidale

been discarded somewhere else; and so on. The historical expression of a manufacturing cycle, on the other hand, consists of deviations of various orders of magnitude from the ideally "linear" norm of our reconstructions. These deviations are ascribable, to a great extent, to social interaction. In order to describe, measure and quantify such deviations, it may be useful to briefly outline in which ways technical behaviour may deviate from such ideal linearity.

are quite unpredictable but cultural systems, particularly in conditions of conflict, may be very prompt and efficient in envisaging and exploiting the new emerging opportunities. The substitution of one technique for another may have farranging implications for the whole society, and can play an important role in major political conflicts. For example, in a Saharan oasis society studied by G. Bedoucha (1993), water could be distributed either by irrigation time (the traditional system, based upon the daily schedule of praying) or by volume. The shift to the second technique, which might appear a "more rational" choice to our eyes, accelerated the individualization of rights on water, then the selling of water shares, and finally promoted a general absorption of the oasis traditional activities within a wider monetary economy. Obviously, the current debate on the use of nuclear power versus other forms of energy in our own society is another, large-scale example of the same order of issues. By observing how and when similar forms of technical options have beenat play in a given archaeological record, we may be able to approach important social processes.

FORMS OF DEVIATION FROM LINEARITY

Deviation from technical linearity in manufacturing sequences may be described in five basic types of technical contexts, namely:

I. Parallel operational sequences, when different techniques were available to a given society to be used for the same technical or productive goal. 2. Branching operational sequences, when at critical points of an operational sequence, different techniques could be freely selected for the execution of the same technical steps. 3. Constrained operational sequences, when technical branching, i.e. the selection of a different technique at a point of the sequence is constrained by an error occurring at a specific point, instead of being a free choice. As an alternative, the error may result in type 5 (see below). 4. Intersecting operational sequences, when at some critical points manufacturing residues, semifinished products, tools or other items may be shifted from one operational sequence to another. A more common term used in some of these cases is recycling. 5. Suspended operational sequences, when at some critical points semifinished products can be extracted from the production flow and put in storage for variable times, to be eventually re-inserted in different modes into the original manufacturing cycles.

Particular forms of parallel operational sequences are parallel production cycles. These are encountered when the same type of good, usually an ornament, may be produced with costly materials but may also be imitated with artificial base materials and completely different production cycles. For example, in the early states of the Bronze Age, the invention of faience provided cheap replicas of status symbols in semiprecious stones, making these ornaments available to a wider part of the protourban societies and in particular to a growing population of ritual specialists and bureaucrats. A similar evolution may be observed in Mesopotamia, in Egypt, in the Mycenean early states as well as in the Indus Valley Civilization since the 6th millennium BC (Vidale in press; Vidale & Miller in press).

Some examples, taken from different cultural contexts, will help to illustrate these five basic types. Parallel operational sequences In some cases, the same commodity can be obtained, or the same technical task can be performed, by means of different, sometimes contrasting techniques. When a set of different techniques are available to a society, we deal with a potential basin of technical choices. This concept has been brilliantly developed and applied in different cultural contexts by Pierre Lemonnier (1993). One technique may involve types of costs, social relationships and symbolic implications which are quite different from those of another technique. In other words, a technique is not a neutral option, because it carries with it important implications of the technological environments where it was originally developed. Sometimes a technical choice may be a matter of a conscious political strategy; in other cases its implications

180

Branching operational sequences Within the operational sequence of a given production cycle, the end-product may sometimes be obtained with manufacturing techniques which diverge at some critical point. Technical divergence may depend upon the involvement of more individuals in the same manufacturing cycle. Such forms of discontinuity are to be observed in repetitive, lot-processing crafts such as shell or stone bead making, but also in quite unexpected manufacturing sequences focusing on unique, highly individual products. For example, while we were micro-excavating the interior casting core of the two Riace warriors, two life-sized Greek bronze statues still filled with the original clay core which date to the 5th century BC, it was found that the inner core had been hand modelled with flat pre-formed earthen slabs (Micheli & Vidale in press). The lower left leg of Statue Bis made with flat, parallel slabs enveloped by a perimeter slab, without a central support, while the core of the right leg was made with concentric slabs rolled around iron bars. The modelling style is so different that we may hypothesize that it was a different person that made the lower left leg. Thus,

Operational sequences beyond linearity

it appears that at least part of the model was a collective creation, and this would imply that the chief craftsperson who made such statues, as we know him from the written sources, was more similar to a superintendent than to a modern artist modelling his figures with his or her own hands. A diverging technique, in this case, suggests a hierarchy among the craftspersons.

tools such as, for example, drill heads or hammerstones. Another possibility is that the nature and technical output of a given operation may turn out to cause :frequent errors in a following stage, but the whole process may be perceived as a standard. This seems to have beenthe case of the sandstone bangles produced at Mohenjo-daro, in Sindh (Pakistan) (Vidale 1993). These ornaments were fired inside small cylindrical saggars in a reducing atmosphere, close to sintering point; since two or three bangles were fired together in a pile in each box, the rings partially melted at the contact and, after firing, had to be separated by means of a strong blow. This left a circular scar on the bangle's surface, and the potters had to carry out a careful, timedemanding grinding and polishing process in order to make the surface smooth. As such abrasion-polishing marks are rather common, this seems to be an example of a systematic error purposefully included in a manufacturing sequence; however, the bangles thus obtained may have been deemed to be second-choice products.

A skilled craftsperson may be able to skip some manufacturing steps, while less experienced personnel may be obliged to go through longer, more painstaking technical processes. By monitoring such forms of technical variability one may be able to define different degrees of standardization in technical behaviour by different groups, an aspect ideally related to the degree of economic specialization of the producers. Furthermore, the same approach may be used for establishing the presence of unskilled personnel in a craft group. For example, at the Magdalenian factory site of Etiolles, the chipping loci of apprentices were singled out on the basis of the deviation from the expected standard technical behaviour: a low level of standardization of raw material, a high rate of technical errors, small sized blades, a very low percentage of finished tools, and a lower percentage of tools used directly in the chipping locus (Olive 1988; Pigeot 1988). In the study of a talc bead-making area of Mehrgarh, Pakistan, dated to the 5th millennium BC, a typometric analysis revealed that unskilled bead-makers were unable to slice sufficiently thin blanks. Therefore they were compelled to resort to an additional edge-grinding operation, whereas skilled craftsperson saved time and raw material through a simpler manufacturing sequence which involved cutting and rounding substantial sets of disk beads which had been previously strung together (Vanzetti & Vidale 1995).

As a matter of fact, besides forming and patterning the archaeological record, technical errors obviously have major effects on the sequences themselves. They interrupt the cycle at undesired points, opening new paths for recycling, i.e. forcing some :fractions of raw material to re-enter the cycle at another point, or to join a different manufacturing cycle (see below). In other cases, an error may modify some features of the designed finished product, without changing its position and path within the manufacturing cycle. We observed a similar case while we studied the contemporary pottery manufacturing techniques at a village of the centraleastern Terai, in Nepal {Lugli & Vidale 1998). Here, a very common type of carinated vessel is shaped by male potters as a pre- form on the wheel, and set to dry until it reaches a leather-like state of hardness. Next, manufacturing is resumed by female potters, who place the pieces onto a concave mould and squeeze the thick base of the pre-form with a pottery anvil, thus thinning and enlarging the lower body. The pots thus obtained have a distinctive carination at the base of the shoulder. Sometimes the lower wall is exceedingly thin, and the vessel folds and collapses down. In these cases, the women may insert a new slab to support the wall, and fix it with the paddle-and-anvil technique: but in this operation the carination is lost, and the pots assume a simpler, sub-globular profile. We had no opportunity to observe how these anomalous, modified containers fared in the village market, where some kind of selective pressure might be at work against them, but apparently they were fired and traded with all the others, maintaining the same name. Thus, what archaeologists would have considered to be a different ceramic type on the basis of the shape, is in reality a defective version of the same vessel. Only by Xraying a large number of unbroken carinated and globular pots one might have noted a recurrent correlation between globular pots and pre-firing repairs in the form of small slabs added and soldered to the wall from the interior.

Constrained operational sequences Because of such a generalized emphasis on the normative evaluation of manufacturing sequences, technical errors have rarely been the subject of specific research. Nonetheless, it is because of technical errors that the archaeological record so often includes precious and meaningful craft indicators: broken beads, casting spills poured onto the ground, overfired and warped ceramics. Archaeologists normally use this kind of evidence for palaeotechnological recontsruction, but rarely concentrate their attention on the failures which generated these objects, for example by attempting to define the rates of failures of each single operation and comparing such rates in terms of :frequencyand ergonomic cost.

Manufacturing errors may be due to the lack of specific skill, to the employment of untrained personnel, to the use of defective raw materials, tools and craft facilities or, more often, to the emergence of particular conditions of stress which affect one or more technological aspect at a given time. They might also have beenconsidered an unavoidable component of the technical sequence, for example when failures are due to the "'normal" process of consumption of

181

Massimo Vidale

Intersecting operational sequences It is well known that recycling, both in pre-industrial and industrial manufacturing sequences, may assume different and complex forms. Recycling has been defined as primary when technical intersection takes place within the same manufacturing cycle and an operational sequence intersects with itself;, and secondary when it intersects with another, different cycle. The modem amber industry is based to a large extent upon the principle of primary recycling, as the majority of the raw material discarded by reductive sequences is melted at a high temperature and pressure in order to create re-formed raw amber (Rice 1987). One of commonest forms of primary recycling in various cultures is the crushing of earthenware potsherds which are then used as an artificial temper for making other ceramics. Similar, but belonging to the field of secondary recycling, is the use of crushed iron slag on the interior surface of ceramic basins for making ceramic graters in the Iron Age in northern Italy (Maioli 1976). Some other examples of secondary recycling are parchments, sheep bowels and horn from butchering facilities which were used in the late Renaissance for beating gold foil (Martini 1965: 235). In the 16th century the ash from carbonized goat bones was used for detaching the moulds for casting large bells from the pig fat used for the inner model. Similarly, in the 3rd millennium BC in the Indus Valley, powder and chips resulting from the manufacture of talc beads were systematically used for making not only steatite-faience beads (a form of primary recycling) but also refractory containers and furnaces used for firing high temperature ceramics (secondary recycling) (Vidale 1985). Systematic technical intersecting is expected to increase with the development of urban societies and professional specialization.

Technical intersecting may involve economic collaboration or, when important raw materials become scarce, conflict. Economic interest in the same raw material may also stimulate the growth of control institutions for mediating among different craft groups. Similar relationships may be independently observed in the archaeological record, for example detecting specific spatial patterns in the distribution of workshops and dumping areas. On the other hand, they may also involve the expression and formalization of hierarchical social relationships: for example, in the contemporary agate bead making centre of Khambhat in India, agate flakes derived from the manufacture of large, expensive beads is given to lower caste craftspersons for the production of cheap beads; outcast, untouchable craftspersons may also mine the communal dumping area in order to recover even less valuable flakes for making other types of beads. In India, garbage and its handling are ritually polluting and socially dangerous. To accept somebody's else refuse means to formally accept a subordinate position, as well as the long-established reciprocal obligations that come with it.

Suspended operational sequences A feature shared by protohistoric and contemporary industrial systems is the capability of suspending or freezing the material flow of manufacturing at critical points. While the manufacturers of the Industrial Revolution attempted to concentrate large parts of the operational sequences in the same place and at the same time, mainly for political reasons and for the need to contrast with other traditional forms of production, in contemporary industry manufacturing cycles are suspended at critical points, strategically segmented, spatially distributed and recombined on a global scale. Cheaper base materials, cheaper labour, low transport costs and, above all, the same old political constraint, the need for an easy and efficient control oflabour, seem to be at play in this strategy.

What was the meaning of technical suspension in prehistoric and protohistoric times? Hoards of objects such as lumps of unworked raw materials, semifinished products and stocked unworn commodities are not uncommon in prehistoric and protohistoric sites. Their presence is usually left unexplained or, in many cases, simply referred to ritual behaviour, but in many cases such an interpretation is doubtful. In the late Bronze Age in Europe, large hoards of bronze objects contain finished objects including weapons, tools and ingots, both complete and fragmentary. Large bronze objects may have a critical economic importance: they may have a use value (they are valuable because they have a destined function) but also a potential exchange value (they are valuable as shapeless metal to be remelted and cast). An ingot found in a hoard, a settlement or a grave represents a suspension of the metalworking cycle, but it has a substantial exchange value. Plano-convex ingots can be easily broken into wedge-shaped standard parts, but they are also potential pre-forms for making vessels, helmets or other objects with the technique of rising. Observing how metal manufacturing cycles were suspended or, in other words, in which forms metal was strategically extracted and reinserted in the processing cycle, may provide us with important hints on ancient social evolution. The following two examples deal with bronze hoards in the Bronze Age in Europe and South Asia. In the Italian peninsula, the composition of bronze hoards shifted from groups of standardized assemblages made of a few metallic types such as axes and heavy personal ornaments in the early Bronze Age, to deposits containing ornaments, weapons, craft and agricultural tools, vessels, and every kind of metal scraps in the latest phases of the Bronze Age. This change has been interpreted as the expression of a shift from a pre-monetary use of standardized bronze types monopolized by the uppermost layers of a society ruled by warriors to a much more dynamic society where metal traders, merchants, craftspeople and farmers could more freely concentrate, exchange and distribute metal goods in various forms and stages of processing (Peroni 1989). In the Indus Valley, large merchants' houses dating to the 3rd millennium BC

182

Operational sequences beyond linearity

have been excavated. Large hoards containing bronze vessels, ingots and a large amount of probable woodworking tools were buried under the floor. This suggests that bronzes were buried both for their exchange and use value, as well as for the control of a craft of strategic importance in building, one of the leading economic sectors in every society. By freezing the flow at the ingot stage and in the critical form of usable tools, metalsmiths could suspend, together with manufacturing, sets of important social and economic relationships; and hoarding may be seen for what it must have often represented in most early states, a simple but powerful form ofbanking.

complete military equipment of the defeated "army'' (this hypothesis was first suggested to me by R Peroni). Alternatively, the weapons may have been prepared for a small scale military operation, but suddenly had to be buried and for some reason were never recovered. Another important aspects of this hoard is that on the basis of the technical style of some features (direction of the spiral-like closure of the haft, shape of the central ridge, shape of the rear of the blade of the spearhead, technique of piercing the haft for fixing it to the wood) it is possible to argue that more than one smith had made the weapons: evidently, various ''houses" had collaborated in this enterprise.

In other cases, the suspension of a manufacturing cycle may be a hint to other complex forms of economic behaviour. In the contemporary agate workshops of Khambhat, hoards of semifinished beads may result from the irregular fluctuation in demand, causing unexpected changes in the normal relationships of production and frequent episodes of overproduction. For example, when craftspeople in such cases are paid by means of daily salaries, and not by counting each bead (the traditional system), they become careless and, given the low inner value of the base material, frequently manufacture beads in excessive amounts. Such lots of beads are hardly ever re-inserted in the normal manufacturing cycle. In other cases, under the pressure of a fast-rising demand, new craftpersons may try to enter the agate bead trade, but immediately fail. The deposits of semifinished, unprocessed beads one sometimes sees in the workshop of a smith or carpenter are the result of their lack of specific technical know-how, unreliable trade connections and dependence upon very unstable markets (Kenoyer, Vidale & Bhan 1990).

OF LABOUR AND WORK It is interesting to observe how the Marxist dichotomy between labour (having an abstract meaning) and work (being the material aspect of the transformation of nature: see Eaton 1971: from p. 36) has a direct relevance to the present discussion. Labour, in the classical Marxist political economy, indicates a part of the social labour-force, the overall display of productive activities by a given society, whose general recognition is the basis of exchange value and trade; work, on the contrary, indicates a specific type of production cycle (i.e. pottery making or metalworking) and its physical expression. Furthermore, according to this theory, the exchange value of a commodity is not due to the amount of labour invested in the single product by a single operator, with her or his idiosyncratic behaviour, but rather to the "socially necessary labour" required, on average, for manufacturing that commodity. It is easy to see that, in this light, by concentrating and limiting the research to the linear flow of the average manufacturing sequence, we limit our research to the reconstruction of ancient manufacturing sequences in terms of"social norms", i.e. to labour.

The suspension of the normal manufacturing sequence may also have more direct political implications. This is suggested by a hoard of iron weapons datable to the 5th century BC which was found by chance near Aquileia (Vitri et al. 1992). It includes about twenty axes, forty spear heads, a single horse fillet and an elegant knife, in other words the weapons needed by a small unit of warrior footmen led by a horse-mounted chieftain. During conservation a careful scrutiny of the surfaces revealed that all of the weapons apart from one axe were new and completely unworn. They were buried in a hurry, as some hammer scales from the forges still adhered to their surface. Four of the spear heads were left unfinished, the haft still full of iron slag. The point had clearly been forged by a master smith, and the haft had to be finished by somebody else, perhaps a less skilled smith. Discontinuity in the manufacturing of the weapons might point to hierarchical relationships within craft communities; but it might also suggest exceptional manufacturing conditions. Perhaps the hoard is a ritual offering, and the weapons had been collected on the battlefield from the escaped or deceased enemies. In this case, the unfinished spear-heads with the non-functional haft might have been just a symbolic replacement of the missing weapons, manufactured when it was necessary to ritually represent the

This latter perspective might be useful for understanding ancient technology from the point of view of the users or consumers of a given craft commodity, but probably not as useful for observing the social behaviour and strains within ancient craft communities and between craftspeople and the rest of the society. I think that by defining and exploring the various forms of deviation from linearity and their multiple implications we may approach a dimension of work and, ultimately, move towards a dimension of micro-history which can be more fruitfully controlled with scientific means.

BIBLIOGRAPHY Bedoucha, G. 1993 The watch and the waterclock. In P. Lemonnier (ed.) Technological Choices. Transformations in Material Cultures since the Neolithic. London and New York: Routledge, pp. 77-107.

183

Massimo Vidale

Eaton, J. 1971 Economia Politica. Introduzione al/a Teoria Economica Marxista (Political Economy. A Marxist textbook, 1950). Torino. Kenoyer, J.M., M. Vidale & K.K. Bhan 1990 Contemporary stone bead making in Khambhat, India: patterns of craft specialization as reflected in the archaeological record. World Archaeology 23, 1: 44-63. Lemonnier, P. (ed.) 1993 Technological Choices. Transformations in Material Cultures since the Neolithic. London and New York: Routledge. Lugli, F. & M. Vidale 1998 Making and Using Ceramics: On the Role of Technical Events in the Generation of Functional Types. Origini XX (1996): 353-382. Maioli, M.G. 1976 Bacchiglione. In Padova Preromana, Padova, Tav. 5, Fig. 48. Martini, A. 1965 Arti Mestieri e Fede nella Roma dei Papi. Roma. Micheli, M. & M. Vidale in press The evolution of Greek bronze statuary from the Archaic to the Hellenistic period: new evidence on ancient manufacturing techniques. Proceedings of the International Co,iference on Ancient Greek Technology, Thessaloniki. Olive, M. 1988 Une forme particuliere d'economie de debitage a Etiolles. In J. Tixier (ed.) Journee d'Etudes Technologiques en Prehistoire. Paris: CNRS, pp. 27-36. Peroni, R 1989 Protostoria dell'Italia Continentale. La penisola italiana nelle eta del Bronzo e del Ferro. Popoli e Civilta Vol. IX, Roma. Pigeot, N. 1988 Apprendre a debiter des lames: un cas d'education techniques chez des Magdaleniens d'Etiolles. In J. Tixier (ed.) Journee d'Etudes Techologiques en Prehistoire. Paris: CNRS, pp. 64- 70. Rice, P.C. 1987 Amber. The Golden Gem of the Ages. New York. Schiffer, M.B.

1972 Archaeological context and systemic context. American Antiquity 37: 157-165. 1987 Formation Processes of the Archaeological Record. Albuquerque. Tosi, M. 1984 The notion of craft specialization and its representation in the archaeological record of early states in the Turanian basin. In M. Spriggs (ed.) Marxist Perspectives in Archaeology. Cambridge: Cambridge University Press, pp. 22-52. Vanzetti, A. & M. Vidale 1995 Formation processes of beads: defining different levels of skill among the early beadmakers of Mehrgarh. In A. Parpola & P. Koskikallio (eds.) South Asian Archaeology 1993, II: 763- 766. Vidale, M. 1985 Some observations and conjectures on a group of steatite debitage concentrations on the surface of Mohenjodaro. Annali dell 'Istituto Universitario Orientale di Napoli 47, 2: 113-129. 1992 Produzione Artigianale Protostorica. Saltuarie dal Laboratorio del Piovego. Padova. 1993 Bracciali in gres dell'eta del Bronzo. Le Scienze 300: 46-57. in press Indus craftspeople and why we study them. Rome: Istituto Italiano per l' Africa e l'Oriente. Vidale, M., J.M. Kenoyer & K.K. Bhan 1992 A discussion of the concept of 'Chaine Operatoire' in the study of stratified societies: evidence from ethnoarchaeology and archaeology. In A. Gallay (ed.) Ethnoarcheologie. Justification, Problemes, Limit es. XII Rencontres International es d'Archeologie e d'Histoire d'Antibes. Juan-les-Pins: Editions APDCA, pp. 181-194. Vidale, M. & H.M.-L. Miller in press On the development of Indus technical virtuosity and its relationship to social structure. South Asian Archaeology 1997. Vitri S., N. Proen~ de Almeida, P. Coghi, M. Marabelli & M. Micheli 1992 II ripostiglio di armi in ferro di Porpetto (UD): un progetto di conservazione analitica. 3rd Co,iferenza Internazionale sulle Prove Non Distruttive, Viterbo, pp. 837-856.

184