Kevin Greene - Archaeology. An Introduction

An Introduction

The History, Principles and Methods ofModern Archaeology

Third Edition Fully Revised

Kevin Greene

ARCHAEOLOGY

First published 1983 by B.T.Batsford Ltd Routledge is an imprint of the Taylor & Francis Group This edition published in the Taylor & Francis e-Library, 2003. © Kevin Greene, 1983, 1990, 1995

All rights reserved. No part of this book may be reprinted or reproduced orutilized in any form or by any electronic, mechanical, or other means, nowknown or hereafter invented, including photocopying and recording, or in anyinformation storage or retrieval system, without permission in writing fromthe publishers. ISBN 0-203-44720-4 Master e-book ISBN ISBN 0-203-75544-8 (Adobe eReader Format)ISBN 0-415-16607-1 (Print Edition) A CIP catalogue record for this book isavailable from the British Library

3

ContentsList of illustrations 4Acknowledgements 6Note on references 6Preface 7

1 THE IDEA OF THE PAST 81 Human origins 92 Avenues of investigation 153 Archaeology and the Enlightenment 184 The recognition of human artefacts 245 The discovery of civilizations 306 Achievements of early antiquarians 35

2 DISCOVERY, FIELDWORK ANDRECORDING 37

1 The discovery of new archaeologicalsites 37

2 Fieldwalking and site recording 383 Aerial photography 414 Geophysical surveying 465 Archaeology and the landscape 516 A Mediterranean field survey project 537 GIS 568 Conclusions 57

3 EXCAVATION 581 The development of excavation

techniques 582 An example of the development of

excavation techniques: Corbridge,Northumberland 69

3 Excavation procedure 774 Excavation: special cases 855 The excavation of structures 89

4 DATING THE PAST 1011 Background 1012 Historical dating 1013 Typology 1044 Sequence dating and seriation 1075 The advent of scientific dating

techniques 1076 Environmental methods 1097 Absolute techniques 1148 Radioactive effects on crystal

structure 1239 Derivative techniques 125

10 The authenticity of artefacts 12911 Conclusions 129

5 SCIENCE AND ARCHAEOLOGY 1301 Is archaeology a science? 1302 The examination of objects and raw

materials 1303 Conservation 1374 The environment 1395 Climate 1416 Rocks and soils 1417 Plant remains 1438 Animal remains 1469 Human remains 15210 Statistics 15511 Experimental archaeology 15612 Conclusion 158

6 MAKING SENSE OFTHE PAST 159

1 Where is archaeology at the end of thetwentieth century? 159

2 Archaeological theory 1603 Social evolution 1614 Diffusionism 1625 Nationalism and racism 1676 Towards processualist archaeology 1687 Ethnoarchaeology 1718 Post-processualism 1729 Reconstructing archaeology 17310 Current issues in archaeology 17411 Managing our heritage 17512 Ethical issues 18013 Conclusion 184

GUIDE TO FURTHER READING 185General works 185The idea of the past 185Discovery, fieldwork and recording 186Excavation 187Dating the past 188Science and archaeology 190Making sense of the past 193

BIBLIOGRAPHY 195

INDEX 206

4

1.1 Jacques Boucher de Perthes 111.2 Flint implements from the Somme valley 121.3 Joseph Prestwich 121.4 John Evans 131.5 Section drawing by Boucher de Perthes 141.6 William Camden 201.7 John Aubrey 201.8 William Stukeley 221.9 Ole Worm’s collection 251.10 Stone implements from Britain and the

New World 261.11 C J Thomsen in the Oldnordisk Museum,

Copenhagen 271.12 Pitt Rivers’ typology of Australian

weapons 291.13 Engraving of the Parthenon, Athens 311.14 Schliemann’s excavations at Troy 341.15 Section drawing from excavations at

Knossos 352.1 Graph of archaeological finds in Denmark 382.2 Field survey in action 392.3 Standard recording form for sites 402.4 Features visible on aerial photographs 422.5 Aerial photograph of Roman forts at Chew

Green, Northumberland 422.6 Plan of Chew Green earthworks 432.7 Aerial photograph of Neolithic site at La

Panetteria, Italy 442.8 Aerial photograph of prehistoric fort at

Chatton, Northumberland 452.9 Computerized enhancement of aerial

photographs 452.10 Buried features detected by geophysical

prospecting 472.11 Geophysical surveying equipment in use 482.12 Magnetic survey of Groundwell Farm 492.13 Location map of Dalmatian survey area 542.14 Settlement densities from Dalmatian

survey 552.15 Part of the Upper Tisza Project’s survey

area, Hungary 563.1 Mound 2 at Sutton Hoo, Suffolk 603.2 General Pitt Rivers 613.3 Pitt Rivers’ excavation at Wor Barrow,

Dorset 623.4 Sir Mortimer Wheeler and Tessa Wheeler 633.5 Box trenches at Stanwick, Yorkshire 65

3.6 Open area excavation at Stanwick 653.7 New plan of Stanwick 653.8 Magnetic survey of The Tofts, Stanwick 653.9 Drawing of a section at Segontium Roman

fort, Caernarvon 663.10 Open area excavation at Nørre Fjand,

Denmark 663.11 Neolithic settlement at Elsloo, Netherlands 663.12 Diagram of the excavation and dating of a

post-hole 673.13 Recording a stratigraphic sequence with

the Harris Matrix 683.14 Plan of area excavated at Corbridge 713.15 Excavation techniques at Corbridge, 1909 723.16 Sections of relationships between structures

at Corbridge 733.17 Location map showing Corbridge and the

Red House site 753.18 Excavations at the Red House site,

Corbridge 753.19 Photograph of new site museum at

Corbridge 773.20 Excavations in London 783.21 Recording form for excavated features 823.22 Stages from excavation to publication 843.23 Timber structures at Deer Park Farms,

Antrim 863.24 Recording techniques on an underwater

excavation 883.25 Anglo-Saxon burial at Sutton Hoo, Suffolk 893.26 Construction methods for timber and

stone buildings 913.27 Stone facade excavated at Meonstoke,

Hants 933.28 Excavation within the church of St Peter,

Barton-on-Humber 943.29 Elevation of a church at Rivenhall, Essex 953.30 Timber building excavated at Cowdery’s

Down 973.31 Details of building at Cowdery’s Down 973.32 Reconstruction drawing of building at

Cowdery’s Down 973.33 Mud bricks in a section at Tepe Ali Kosh 993.34 Reconstruction of Roman fort gate at

South Shields 1004.1 Inscription dating the construction of

Hadrian’s Wall 103

List of illustrations

List of illustrations

5

4.2 The typology of Bronze Age axes(Montelius) 104

4.3 The typology of Bronze Age axes(Pitt Rivers) 105

4.4 Minoan pottery from Crete and Egypt 1064.5 Typology of harpoons 1064.6 The applicability of scientific dating

methods 1084.7 Materials examined by scientific dating

techniques 1084.8 Apparatus for measuring tree-ring

thicknesses 1104.9 Dating by dendrochronology 1104.10 Timbers preserved at Pueblo Bonito, New

Mexico 1114.11 Cores from sedimentary lake deposits in

Sweden 1134.12 Volcanic eruption revealed by acidity in ice

cores 1144.13 The basis of radiocarbon dating 1154.14 Tree-ring calibration curve for

radiocarbon dates 1174.15 Statistical margins of error for radiocarbon

dates 1184.16 Triple date from a radiocarbon sample

from Galgenberg, Germany 1204.17 AMS radiocarbon date for the Turin

shroud 1214.18 Graph of light released during

thermoluminescence dating 1244.19 Section through the hydration rim of an

obsidian artefact 1264.20 Changes in the direction of magnetic

North 127

5.1 Microscopic structure of metals 1335.2 Obsidian flakes, core and arrowheads 1345.3 Sources of obsidian revealed by analysis 1355.4 Metal content of Egyptian copper alloy

axes 1365.5 Anglo-Saxon buckle from Broadstairs,

Kent 1385.6 X-ray of Anglo-Saxon buckle from

Broadstairs 1385.7 Anglo-Saxon buckle from Broadstairs after

conservation 1385.8 Rendsina soil profiles 1425.9 Pollen grains 1445.10 Pollen diagram of Jutland, Denmark 1455.11 Neolithic animal bones from Troldebjerg,

Denmark 1475.12 Cattle-raising in early medieval Ireland 1485.13 Photomicrograph of textile 1495.14 Oxygen isotope ratios in a limpet shell 1515.15 Scan of a mummified Egyptian head 1535.16 Experimental farm at Butser, Hants 1576.1 The spread of influences from the

civilizations of the Near East 1626.2 Megalithic tomb at Kits Coty House, Kent 1636.3 Distribution map of megalithic tombs 1656.4 Typology and dating of megalithic tombs

in Brittany 1676.5 Ethnoarchaeological observation of a

smith, Kenya 1716.6 The Archaeological Resource Centre,

York 1786.7 Druids at Stonehenge 1806.8 Model of Iron Age family group 1836.9 Model of Anglo-Saxon family group 183

6

The content of this book includes aspects of the sub-ject that I have always found interesting, supplementedby issues raised by Adult Education students when Ifirst began teaching in higher education. I remain grate-ful for the intellectual stimulation that I gained fromthese students on a variety of courses in Hexham,Ponteland, Rothbury, Wallsend and Newcastle uponTyne. The book also reflects the content of a first-yearlecture course, ‘Introduction to Archaeology’, that Ihave taught in the University of Newcastle upon Tynesince 1978. It is taken by many students who wish toinclude archaeology in a broader Arts and Science de-gree, in addition to those specializing in Single Hon-ours Archaeology. Other students follow the course asa subsidiary part of degrees such as History, Geogra-phy, Psychology or Music. The book has been influ-enced by what I felt this mix of students might usefullylearn about archaeology. I appreciate their many com-ments (positive and negative) over the years.

Several individual archaeologists have influenced thisbook. Among my formative influences were RichardAtkinson and other members of his Department of Ar-chaeology at University College, Cardiff, notably BillManning and Leslie Alcock. My debt to the writings ofGlyn Daniel and Stuart Piggott on the history of archae-ology will be apparent throughout this book. GrahamWebster provided a ‘role-model’ both as archaeology’sleading ‘adult educator’ and as a writer of high-qualitybooks for general readers. I also gained much from work-ing in adult education with Colin Burgess during my earlyyears in Newcastle. Sadly, the most influential of thecolleagues I joined in the Department of Archaeologyin 1977—George Jobey, Martin Harrison and JohnGillam—have all died in recent years. Fortunately, mycurrent colleagues Chris Tolan-Smith and John Chapmancontinually enhance my awareness of developments inarchaeological method and theory. I am also indebtedto Newcastle University’s library facilities, in both themain Robinson Library and the departmental CowenLibrary, where Pat Southern provides outstanding re-sources and the empathy of a fellow Batsford author.

Finally, I thank my wife Julia and my children Danny,Liam and Eva for help, support and forbearance at allstages of the Great Rewriting that took place in 1993and early 1994. I also thank my parents, John and BettyGreene, for stimulating and nurturing my early child-hood passion for archaeology, and maintaining theirinterest ever since. My mother’s scientific training in bi-ology opened my eyes to the natural setting of archaeo-logical sites, and she has acted as a useful proofreader

and scrutineer of my written style recently. My father’scareer in the Health Service acted as a useful reminderthat compassion for the living should not be overlookedin the pursuit of the distant past, while his research andpublications since retirement show that there is noth-ing indispensable about secondary and higher education.I have also learnt a lot from my brother, Patrick, whofollowed a very different path through archaeology thateventually led to the directorship of the Greater Man-chester Museum of Science and Industry.

Advice and illustrationsI have received help from a large number of individualsin Britain and abroad, some contributing ideas aboutthis new edition, many in providing illustrations. Mostof the latter are mentioned in the captions, but othersinclude: Martin Aitken; Lindsay Allason-Jones; MartinCarver; Andy Chopping; Mike Corbishley; J G B Haigh;Heinrich Härke; Robert Hedges; Dr S Ipson; CatherineJohns; Andrew Jones; Mick Jones; Catherine Lavier;David Mattingly; Samantha Middleton; Martin Millett;Barbara Ottaway; Ian Oxley; Georgina Plowright;Julian Richards; Bob Rutland; Chris Saunders; EleanorScott; Celia Sterne; Liz Watson; Humphrey Welfare;Felicity Wild; John Peter Wild.

Kevin Greene, May, 1994

Note on referencesReferences to publications made in the text are given inauthor/year form (e.g. Evans 1860) and are listed in thealphabetical order of their authors’ names in the con-solidated bibliography at the end of the book. Wher-ever possible, books (rather than articles in periodicals)have been cited in section headings within chapters.These may either be the principal publication on thesubject or a particularly good example of the subjectthat is being explained. References made within the textalso include articles in journals (notably Antiquity). Theextensive ‘Guide to further reading’ that begins on p.185 includes many further works related to the themesof the individual chapters; details of these are not re-peated in the consolidated bibliography.

Acknowledgements

7

PrefaceWho is this book aimed at?I have tried to provide a readable and informative booksuitable for just about any interested reader from mid-teens upwards. I envisage it as a first point of contactwith the subject, and hope that it may lead on to widerreading. It is certainly not designed as a textbook, forI lack the comprehensive range of knowledge that thisterm implies. The book reflects my interests, and it doesnot claim to venture very far beyond my own exper-tise in the archaeology of Roman and early medievalwestern Europe. Since introductions to archaeologytend to be written by prehistorians, readers will prob-ably find more examples drawn from historical peri-ods in this book than in most others.

The back cover of the first edition carried an ap-parently warm recommendation—‘this book scores a‘bull’s-eye’—that had actually concluded a rather un-enthusiastic review:

Kevin Greene’s book is written for the dominantmarket of the 1980s, that is the first year under-graduate aiming to become a professional archae-ologist. Indeed any general reader picking up thisnew book would be overawed by the massive scaleof everything, and the complexity of all the opera-tions, to say nothing of the jargon of the ‘new’ ar-chaeology. But though one might regret thatuniversity teaching is so biassed in favour of large-scale, rather than small scale operations, that’s theway it is, and for the market that actually exists,this book scores a ‘bull’s-eye’. (Current Archaeol-ogy January 1984, 221)

In reality, a book aimed solely at archaeology undergradu-ates in Britain would sell very few copies. The NorthAmerican tradition of ‘blockbuster’ textbooks has neverfound favour in Britain; despite its British origins, Ren-frew and Bahn’s Archaeology: Theories, methods, andpractice (1991) clearly targets an American market. Myown experience, reinforced by reports from colleaguesin other universities, is that students read all or parts ofmy book to gain a general understanding of a topic, andthen turn to Renfrew and Bahn (or specialist books andarticles) for further details and examples.

‘Blurbs’ dreamed up by copy-writers for book-clubbrochures captured my intentions more perceptively(and concisely) than most reviewers. My favourite

examples of this genre remain ‘No other survey saysso much that’s new about so much that’s old’ and‘Shows you how fieldwork is done; explains how tointerpret remains to get a vivid picture of the past’. Ialso appreciated the unintentional irony of ‘The per-fect initiation, clearly written for students and armchairarchaeologists’ and ‘Even if you never take a trowel inhand, this is a totally absorbing look at the practicesand methods of the people who spend a lifetime dig-ging up the past’.

Why is a new edition needed,and how does it dif fer from itspredecessor?Archaeology has moved forward since 1983. Techni-cal advances are most visible in chapters 2, 4 and 5(fieldwork, dating and science). Existing methods haveimproved in range and accuracy, and new methodshave been developed. Relatively minor changes inemphasis and interpretation appear in chapters 1 (‘ori-gins and growth’) and 3 (excavation), but chapter 6(theoretical archaeology) is substantially new, and re-veals how much a subject may change in ten years. Itnow contains discussions of matters that were only justgetting on to the archaeological agenda in the early1980s, such; as ‘heritage’, gender, or the rights of in-digenous peoples. Even where the content has changedrelatively little I hope that the style has improved, forthis edition has been prepared with the help of a mi-crocomputer; I do not think that a single sentence hassurvived intact. There are now entire books devotedto topics that in 1983 appeared only in specialist jour-nals. However, the proliferation of books on individualarchaeological techniques or scientific methods has notbeen matched by new general books (with a few hon-ourable exceptions, such as Aitken’s Science-basedDating in Archaeology, 1990). Thus, guidance on fur-ther reading remains an important part of my book,but I have gathered it into a single section this time toallow it to be updated more easily whenever the bookis reprinted.

8

One cannot see archaeology emerging as arecognizable discipline until the later nine-teenth century; before this we have an amor-phous antiquarianism. (Piggott 1989, 8) The history of archaeology is the history ofthe ideas that have prevailed, the ideas thathave been right in the long term, and it isproper to see the past from the present. Thenecessary qualification is some real under-standing of context, some real respect for thepotentials and possibilities that were avail-able, and a decent humility about our ances-tors’ accomplishments. (Chippindale 1989,33)

Our knowledge of archaeology is founded uponmany basic assumptions that are far from secure andunchanging. Theories produced by scholars may bemodified by discussion and criticism, only to be over-thrown by a single new idea or piece of evidence;equally, they may simply go out of fashion for no par-ticular reason. The vision of a subject suddenly re-versing its ideas can be disconcerting to people whoare not directly involved, especially if they have re-ceived a traditional education based upon the learn-ing of ‘facts’ that lead to ‘correct’ answers. Peopletend either to reject new ideas or to embrace thementhusiastically, for reasons that may be emotionalas much as rational. It is easy to understand why theo-ries from outside the academic world (frequently in-volving extra-terrestrial influences) are so popular,particularly when they dismiss academic ‘experts’ asrea-ctionary snobs unwilling to consider ideas pro-posed by anyone outside their ivory towers.

The most sensible approach to archaeologicalevidence in the closing years of the twentieth cen-tury is not to attempt to keep up with every shift inopinions, but to understand how they were arrivedat in the first place. It takes a certain amount of cour-age to accept that there are no ‘right’ answers, and

that all interpretations are based upon assumptionsof varying probability, never certainty. It is verystimulating to approach archaeology without nec-essarily deciding whether a new interpretation is‘right’, but by analysing it in a straightforward man-ner. How did the archaeologist reach that conclu-sion? What kind of evidence was involved? Whatare the assumptions upon which it is based? Sev-eral chapters of this book attempt to explain thesources of information on which the assumptionsand interpretations are based, and they also aim toshow how the basic evidence is actually recoveredand studied.

The aim of this first chapter is to give an impres-sion of the slow emergence of archaeology by look-ing at how some of the principal concerns of modernarchaeologists developed in the past. Four issues willbe examined: the idea of early human origins; therecognition and interpretation of human tools; theobservation and recording of ancient sites; and, fi-nally, the investigation (with the help of excavation)of early civilizations. These issues are not entirelyseparate, and they will not be looked at in strictchronological order. They reflect my personal judge-ment of the importance of the concepts involved; Iwill discuss some aspects further in later chapters.

At first sight, the assumptions early antiquariesmade about the past, and the concepts within whichthey framed their writings, may seem naive today.However, it is important that the benefit of hind-sight does not make us forget the constraints of thesocial and intellectual context in which antiquar-ians lived and worked. For example, in the earlynineteenth century, Danish scholars first organizedprehistoric objects into three successive ‘Ages’(Stone, Bronze and Iron), but they were content toset these Ages into a time-span that began in 4004BC, the date of the creation of the Earth calculatedfrom the Book of Genesis by seventeenth-centurytheologians. The much longer time scale demandedby geology and evolution did not finally displace

1 The Idea of the Past

The Idea of the Past

9

the biblical scheme until the 1860s. However, suchmomentous reorientations did not end in the nine-teenth century; the dating of prehistory has under-gone two significant revisions since the SecondWorld War, first as the radiocarbon dating techniquewas introduced and accepted, and later as furtherresearch demanded a revision and recalibration ofits results. Some of today’s most convincing inter-pretations of the past will undoubtedly require com-parable re-examination as more sites are excavated,dating techniques are improved and new ideas aredevised.

The lesson to be learned from the history of ar-chaeology is clear. We work within the limitationsof the best available assumptions, but these are sub-ject to unpredictable changes. Little is gained fromtreating older research as simply redundant andsometimes humorous; we may learn a great deal ofvalue by examining how early antiquaries and ar-chaeologists tackled the formidable problem ofmaking sense of the human past, without most ofthe libraries, museums, travel and technical facili-ties available today. From this point of view, a studyof their work will result in admiration, not amuse-ment. At the same time, it is important to avoid thetendency to reduce early scholars to ‘textbook card-board’ by giving them credit for inklings of ideasthat we consider important today, without under-standing the setting in which those ideas were origi-nally formulated (Gould 1987, 5).

1 Human origins

An interest in origins may be a very early aspect ofhuman consciousness. The possibility that the phe-nomenon of death led to reflections on afterlife orrebirth amongst early prehistoric peoples is sug-gested by burial rites. For example, around 23,000years ago two young boys and a man were buriedat Sunghir (200 km north-east of Moscow) withstone, bone and ivory objects including spears, pen-dants, beads and animal carvings (Gamble 1986,188). Many societies have developed sophisticatedmythologies to explain their origins, and, in asso-ciation with religion, the whole environment maybe fitted into an orderly system in which major natu-ral features may be the work of gods (Trigger 1989,29). Artificial mounds, abandoned occupation sitesand ancient objects were often associated with dei-

ties, fairies, ancestors or other denizens of the worldof mythology, and explanations of this kind aboundin surviving folklore; an awareness of the physicalremains of the past, and attempts to explain them,lie behind the modern disciplines of archaeology andhistory. However, the development of strictly ar-chaeological (as opposed to antiquarian) thinkingsimilar to our own belongs to very recent centuries.

Collections of antique objects and works of artwere not uncommon in the Ancient World, fromBabylon in the sixth century BC to the civilizationsof Greece and Rome, and more sporadically in themedieval period (Trigger 1989, 29–35). Romanphilosophical poetry of the first century BC con-tained ideas about the successive importance ofstone, bronze and iron as materials for the manu-facture of implements. Although this ‘Three-AgeSystem’ was widely accepted by AD 1800, it wasnot applied in a practical way to ancient objects until1816 (below, p. 26). Greek and Roman philoso-phers and travellers wrote accounts of peoples (realand mythological) whose lifestyles were more primi-tive than their own, partly out of curiosity but pri-marily in an attempt to judge their own societies(Blundell 1986, 187–201). It is difficult now for usto appreciate the basic problem that confronted his-torians or philosophers in literate societies right upto the eighteenth century AD. They were able topursue their origins through surviving historicalrecords, but beyond the earliest documents lay acomplete void, containing unverifiable traditionsthat merged into a mythological and religious worldof ancestors and gods. Most conspicuously lackingwas a concept of the depth of time (Rossi 1984).Until the mid-nineteenth century, most historiansfound no difficulty in compressing the period be-fore written records into a biblical time scale involv-ing only six to seven thousand years. Interpretationsof geology or evolution that demanded an unimagi-nable length of time were fiercely contested. Eventoday, the depth of archaeological and geologicaltime is still grossly underestimated in the contem-porary mythology of cartoons, in which no prehis-toric man is complete without a stone axe orwooden club, a simple one-piece animal-skin gar-ment, and perhaps a vehicle with stone wheels.

The fundamental problem of chronology didnot change significantly between the Greek andRoman period and the eighteenth century AD.The only possible basis for the study of sites and


10

artefacts was to link them to peoples and eventsknown from documents, or otherwise dismissthem into a vacuum. Samuel Johnson stated: ‘Allthat is really known of the ancient state of Britainis contained in a few pages…. We can know nomore than what old writers have told us’ (Daniel1963, 35). Historians of the Middle Ages (such asGeoffrey of Monmouth, who died c. 1155) filledout early periods with fantastic tales of mytho-logical and real figures like Brutus the Trojan,King Arthur and Julius Caesar. The only real dif-ference between Geoffrey and the ‘enlightened’historians of the eighteenth century was in thequantity of mythology that they accepted; laterwriters tended to associate earthworks with Ro-mans or Danes rather than Trojans or Druids,but a concept of prehistoric time was still miss-ing. One positive development was that a greaterunderstanding of the way of life of ancient peo-ples (such as the Britons described by Julius Cae-sar) could be gained from reports of ‘savages’encountered by European traders and colonistsin Africa or the Americas; prehistoric artefactsfound in Europe could also be compared withthose still in use in primitive societies (Trigger1989, 53; below, p. 25–6).

1.1 Prehistory and history (Daniel & Renfrew 1988)

Archaeologists today still tend to be divided intotwo categories—prehistorians and historical ar-chaeologists. This division is not particularlyhelpful, but it does distinguish between the latter,who study people or places within periods duringwhich written records were made, from theformer, who are concerned with any period be-fore the use of documents. Historical archaeolo-gists usually possess a basic framework of datesand a general idea of the society of a particularperiod into which to fit their findings, while pre-historians have to create some kind of frameworkfor themselves from artefacts and sites alone. Themethods used by both kinds of archaeologist arevery much the same, and there is considerableoverlap between their ideas and interests; themost effective of today’s scholars happily crossthe boundary between prehistory and history.There was a distinct difference in the past, how-ever. Ancient historians or biblical scholars could

set out to locate physical traces on the ground ofevents and civilizations described in literature,unlike local historians, natural scientists or col-lectors who tried to make sense of artefacts orgraves surviving from times before the earliestsurviving written records. Because of earlyprogress in classical archaeology, many historicalarchaeologists assumed that the written wordhad an innate primacy over physical evidence,and ‘…continued to regard prehistoric archaeol-ogy as greatly inferior to the archaeological studyof periods that can be illuminated by writtentexts’ (Trigger 1989, 40).

If early ‘prehistorians’ (the term only becamecurrent after 1850: Chippindale 1988) believedin a biblical Creation in 4004 BC (or some otherdate calculated from the Old Testament), at leastthere was an upper limit to the age of any of theitems that they studied. If they did not, the po-tential questions were almost infinite, and at firstsight insoluble. Which sites and objects were inuse at the same time, and how many years elapsedbetween those that looked primitive and thosethat seemed more advanced? Did technical im-provements represent a gradual series of inven-tions made by a single people, or did innovationsmark the arrival of successive waves of conquer-ors with superior skills? The first step essentialto any progress was a recognition of the amountof time occupied by human development in pre-historic times, and this advance took place in thefirst half of the nineteenth century. In the view ofBruce Trigger (1989, 70–3), the liberation of ar-chaeologists from this ‘impasse of antiquarian-ism’ had two distinct consequences. The first wasthe invention of new dating methods in Scandi-navia, and the second was the study of humanorigins in France and England, which ‘added avast, and hitherto unimagined time depth to hu-man history’. I will examine dating methods be-low, after exploring the more fundamental anddramatic issue of human origins.

1.2 Human antiquity (Grayson 1983)

The Society of Antiquaries of London wasfounded in 1717, and in 1770 it began to publishanimpressive periodical, Archaeologia, whosestyle and format have changed little up to the


11

present day. Volume 13 (published in 1800) in-cluded a minor item, whose full significance didnot become apparent for sixty years. Amongst anassortment of papers on subjects ranging from aRoman fort in Germany to historical documentsassociated with British royalty was a short letterfrom John Frere, drawing attention to some obser-vations he had made in a clay pit at Hoxne in Suf-folk. He reported flint weapons found at a depthof twelve feet in a layer of gravel, overlain by a bedof sand containing bones of extinct animals and,remarkably, shells and remains of marine crea-tures, ‘…which may be conjectured to have beenonce the bottom, or at least the shore, of the sea’.Frere was evidently conscious of the problematicimplications: ‘It may be conjectured that the differ-ent strata were formed by inundations happeningat distant periods…. The situation in which theseweapons were found may tempt us to refer themto a very remote period indeed; even beyond thatof the present world…’ (1800, 205). Frere madeno reference to the biblical Creation and Flood,and he died before an accumulation of similarfinds began to suggest an alternative view of hu-man origins.

Finds of human bones and artefacts associ-ated with remains of extinct animals were notedwith growing frequency in Europe in the earlynineteenth century—and were as often ex-plained away by theologians as accidental. Bythe time of Frere’s death in 1807 a key figurewas already becoming interested in archaeologyin France: Boucher de Perthes (fig. 1.1) spentmany years studying the gravel quarries ofnorthern France and was a strong advocate oftwo ideas. First, he was impressed by the greatdepth and variety of the deposits of sedimentsand he felt that they were far too complex toresult from the biblical flood, although he didnot reject the authority of the Old Testament.Second, he fought hard to convince contempo-raries that the flint tools that he had collectedfrom the gravels were made by humans, andthat they could be recognized by their artificialshaping (fig. 1.2). It was an uphill struggle: hecommented that ‘at the very mention of thewords “axe” and “diluvium”, I observe a smileon the face of those to whom I speak. It is theworkmen who help me, not the geologists’(Daniel 1981, 52). Because he was able to prove

that these tools came from within ancient gravelbeds, he concluded that humans had existed be-fore ‘the cataclysm that gave our country itspresent configuration’, and that these humanswere therefore also contemporary with a widerange of extinct animals. Because Boucher dePerthes did not abandon the idea of floods, hesuggested that Adam and Eve resulted from alater and separate Creation, long after the Floodwhose results he observed had wiped out earlierhumans. ‘Let us not bargain over the durationof ages; let us believe that the days of the crea-

1.1 Jacques Boucher de Perthes, portrayed byGrèvedon in 1831. He published many ideas aboutartefacts and their stratif ication derived fromCasimir Picard without acknowledgement. Hisbombastic manner diminished the credibility of hisown beliefs: ‘Few people took his book seriously;of those who had met him, none’ (Evans 1956,282). Despite this, Boucher de Perthes’ centralidea—that human artefacts of great age were to befound in the gravels of northern France—wasconfirmed by Evans and Prestwich in 1859. Societyof Antiquaries, London


12

1.2 These flint implements from the Somme valleywere published by John Evans soon after his visit tothe sites where they were found. ‘That they really areimplements fashioned by the hand of man, a singleglance at a collection of them placed side byside…would, I think, be sufficient to convince eventhe most sceptical. There is a uniformity of shape, acorrectness of outline, and a sharpness about thecutting edges and points, which cannot be due toanything but design’ (Evans 1860, 288). The ‘handaxe’ (centre) is approx. 17 cm long. Evans I860, pl. 15

1.3 Geologist Joseph Prestwich (left) examining flintimplements of the kind found in gravel pits innorthern France, where Boucher de Perthes claimedto have found them in deep deposits together withbones of extinct animals. Prestwich and John Evanswere convinced of the authenticity of these finds,and published reports that led rapidly to acceptanceby most scientists and archaeologists. Prestwich 1899,facing p. 126


13

tion, those days that began before our sun, werethe days of God, the interminable days of theworld. Let us remember, finally, that for thiseternal God a thousand centuries are no morethan one second…’ (Daniel 1967, 62). Whetherpeople accepted this modified view of the Crea-tion or not, not only was the Earth becomingincreasingly ancient, but humans were also be-ing drawn back into a void of seemingly im-measurable depth.

Not all geologists treated Boucher de Perthes’work with disbelief and amusement. An Englishgeologist, Joseph Prestwich (fig. 1.3), together withan authority on ancient implements, John Evans(fig. 1.4), organized a visit to France to meet himand to see the celebrated gravel pits. On the firstof May, 1859, they were rewarded with the op-portunity of seeing a flint axe still firmly embed-ded in an ancient gravel deposit, and any remainingdoubts were removed (fig. 1.5). Prestwich read anaccount of their observations to the Royal Societyin London before the end of May, and a summary

of his paper appeared in print in 1860. He referredto John Frere’s letter published in 1800, andpointed out that Frere’s observations conformedwith the new findings from France.

In 1869 an important pioneer of many aspectsof archaeology, Pitt Rivers, successfully sought andfound flint implements in association with bones(elephant, hippopotamus, extinct deer, etc.) atActon, near London. They occurred in a gravel ter-race 25–30 metres above the River Thames; how-ever, Neolithic and Bronze Age finds from the riveritself demonstrated that its present course was over

1.4 John Evans (1823–1908) combined a busy lifein the paper industry with interests in geology andthe collection and study of ancient coins andartefacts. This photograph demonstrates hiscontinuing involvement in archaeology thirty yearsafter his visit to the Somme valley in 1859; he isthe f igure in shirtsleeves, excavating a Romanburial at Youngsbury, Ware, in Hertfordshire.Society of Antiquaries, London


14

2000 years old: ‘…this gives us some idea of thegreat length of time it must have taken to erode thewhole valley…’ (Bowden 1991, 74). He designed aparticularly elegant method of proving the antiq-uity of early flint artefacts in Egypt, by looking forthem in the walls of tombs constructed around 1500BC near Thebes. The tombs had been dug into hardgravel that included (along with other artefacts) aflint flake cut by the builders; the geologist whoaccompanied Pitt Rivers commented: ‘It belongs tothe geological delta formation, and beyond ques-tion it is older beyond calculation than the tombwhich was cut into the gravel, and cut through theend of this particular flint flake’ (Bowden 1991, 91).

1.3 Catastrophists and fluvialists (Gould 1987)

In some ways, the recognition of authentic associa-tions between flint axes and the bones of extinctanimals increased the problems of dating faced bygeologists and historians: how long ago did thesehumans and animals live? The predicament wasexpressed particularly well by Joseph Prestwich:

The author does not, however, consider that thefacts, as they at present stand, of necessity carryback Man in past time more than they bringforward the great extinct Mammals towards ourown time, the evidence having reference only torelative and not to absolute time; and he is of theopinion that many of the later geological changesmay have been sudden or of shorter durationthan generally considered. In fact, from the evi-dence here exhibited …the author sees no rea-son against the conclusion that this period ofMan and the extinct Mammals…was brought toa sudden end by a temporary inundation of theland. (Prestwich 1860, 58)

1.5 Section drawing published by Boucher dePerthes in his Antiquités celtiques et antediluviennes(1847) showing the geological strata in which hehad found f lint implements (labelled couteau/haches en silex) in the Somme valley gravels. Thecarefully numbered and delineated layers andartefacts, with a vertical scale in metres, illustratehow geologists used this method of recording,decades before it was adopted by archaeologicalexcavators (compare f igs 1.15 and 3.9 below).


15

Prestwich reveals that the idea of sudden catas-trophes survived as late as the 1850s, butCharles Lyell (1797–1875) had published a se-ries of books in the 1830s (entitled Principles ofGeology) that asserted that gravel, sand andclay deposits were formed by the same proc-esses of erosion and deposition by weather andwater observable today. Lyell, and subsequenthistorians of geology, have expressed the debatein terms of catastrophists and uniformitarians.In fact, after AD 1800 few geologists still be-lieved that layers of gravel and sedimentaryrocks were formed simply by the catastrophicfloods described in the Book of Genesis, andfew were constrained by the very short timespan for the Earth derived from the Old Testa-ment (Gould 1987, 112). ‘Fluvialists’ and‘catastrophists’ both studied and interpreted se-quences of rocks and fossils, and their methodsoffered a solution to the problem of early hu-man tools and weapons. If the levels observedby Frere and Boucher de Perthes really had beenlaid down by slow erosion by wind and water,and gradual deposition by rivers and oceans, animmense length of time must be involved. Itcould not yet be measured, but if these proc-esses were assumed to have operated uniformlyin the present and the past, their duration couldperhaps be sensed and visualized rather moreeasily than mysterious catastrophic floods.

During this period, human bones as well asstone tools were found in early geological depos-its in many parts of Europe. In 1863, Lyell fi-nally combined the ‘new’ geology with thisarchaeological evidence in a book entitled TheGeological Evidences of the Antiquity of Man.This book also took full account of the evolu-tionary theory, perhaps the most significant sci-entific advance of the nineteenth century. CharlesDarwin’s theory, proposed in The Origin of Spe-cies by means of Natural Selection (1859), didnot just provide an appreciation of the depth oftime demanded by geology. It also provided alinear, almost historical, notion of progress thatcould be adopted easily by archaeology. Sciencein the nineteenth century was not divided intosmall specialized compartments in the way it istoday, and Darwin was well aware of the impli-cations of recent geological thinking. Darwinand the geologists both demanded the accept-

ance of the same concept: the present surface ofthe Earth, and the plants and animals (includinghumans) that inhabited it, resulted from an im-mense period of change.

The slow development and acceptance of aclear concept of human antiquity have been de-scribed here to illustrate how archaeology pro-gressed by changing the explanations thatprevailed at the time, and to remind us that newideas normally meet resistance. Individuals wereforced to look at fresh evidence in a scientificmanner and work out its implications, rather thansimply fitting new information into an existingframework of ideas in the least offensive way.This attitude is equally necessary today; we mustnever forget that we are trapped in the outlook ofour own time. The study of changing attitudes tothe past should act as a constant reminder of this,and we must avoid a smug satisfaction that wemay distinguish the ‘right’ ideas of the past simplybecause they accord with the prevailing judge-ments of the present. Gould’s thoughtful exami-nation of the different concepts of time held byearly geologists (including Lyell) contains manyimplications about our attitude to the work ofsuch pioneers (1987).

2 Avenues of investigation(Casson 1939)

Although human antiquity was the most impor-tant single idea that had to be established beforearchaeology could develop into its modern form,it was far from being the first field of study. Ar-chaeology is still fundamentally concerned withsites and objects, and the growth of research intothe material remains of past human activitiesthrows further light on the basic principles thatunderlie the work of modern archaeologists. Con-siderable progress in setting these tangible re-mains into meaningful contexts had been made bythe sixteenth century, with the help of travel, thecollecting of artefacts, and anthropological infor-mation that began to reach Europe from the NewWorld after AD 1500.

Speculation about the past was not uncom-mon amongst the Greeks and Romans. Some-thing akin to modern anthropology (rather thanarchaeology) was prominent in ancient Greece,


16

where intense interest in his own society andpolitical system probably led Herodotus to travelamongst ‘barbarian’ (i.e. non-Greek) peoplessuch as the heavy-drinking, head-hunting Celtsin the sixth century BC. His observations of in-stitutions, customs and ways of life earned himthe respect of one twentieth-century historian ofanthropology, Stanley Casson (1939, 42–53).Greek and Roman travellers may have felt thatan understanding of other peoples would givegreater insight into their own society, but on amore practical level, their observations were use-ful to other travellers, for Greek and Roman cul-ture and commerce embraced the wholeMediterranean as well as parts of its barbarianhinterland. These studies were paralleled by ad-vances in anatomy, astronomy and biology, tothe extent that ideas of biological and social evo-lution were emerging (Blundell 1986, 73–97).These ideas were taken up again during the Ren-aissance and advanced to a stage where true ar-chaeological research could begin.

With few exceptions, Roman writers were con-tent either to ignore barbarian cultures, or simplyto rehash observations made by earlier Greek in-vestigators. A notable exception was Tacitus, whowrote an interesting account of the Germans inthe late first century AD. However, it was not sim-ply scientific curiosity that motivated his descrip-tion of the simple life and virtues of thesebarbarians; he wished to make a political point bycontrasting them with the corruption of Romansociety. His Germania is an early example of thecreation of a ‘Noble Savage’ myth, a literary formthat has remained popular ever since. Unlike hisGreek predecessors, Tacitus made no attempt togather first-hand information by travellingamongst the Germans. He embellished and up-dated Greek writings through conversations withsenior army officers and civil servants from hisown social circle who had held appointments inthe Rhine provinces on the frontiers of the RomanEmpire. His book could have formed the basis ofa comparative study of the origins of Roman soci-ety, but in practice it provided an opportunity tomake a political and philosophical point.

Some features of modern archaeology did existin the Roman world. Collections of Greek sculp-ture and vases were popular, various stages of 16architectural development were appreciated and

tourist visits to ancient monuments had alreadybecome common, not only in Italy and Greece butalso in Egypt. The Emperor Hadrian (AD 117–38) is a good example of a traveller and collector:during tours of the Empire, he visited ancientGreek shrines and restored or completed Greekbuildings. He designed a country villa inland fromRome at Tivoli that housed a library and a collec-tion of Greek sculpture, and incorporated gardensand lakes reminiscent of places he had visited inEgypt and Greece. He changed a century-old fash-ion by wearing curly hair and a beard in theGreek manner, in contrast to the severe clean-shaven and short-haired appearance of his pred-ecessors.

This kind of self-conscious antiquarianism wasunlikely to lead to any scientific inquiry into mat-ters relating to archaeology. It was in any caseswept away by the political and economic chaosof the third and fourth centuries AD, after whichthe western half of the Roman Empire graduallydisintegrated and passed under the control of thedescendants of Tacitus’ Germans—tribes fromnorthern and eastern Europe who invaded andsettled in the old Imperial lands in the fifth andsixth centuries. Because the remaining Romanculture of these areas and the surviving easternRoman Empire had become largely Christian,philosophical speculation was replaced by rigidbiblical doctrine.

2.1 Medieval attitudes to antiquity

The attitudes of Christian theologians help to ex-plain the lack of significant progress in archaeologi-cal thinking before the nineteenth century. It tookrevolutionary developments in geology and biologyto force a new scientific view of human origins uponthe Christian world. For most of its history, Chris-tianity has been founded on total belief in the Bi-ble; to doubt its word offended not only God, butalso the political organization of Church and Statethat enforced its acceptance. Thus, independentthinking was discouraged by both intellectual andsocial circumstances, and new ideas were likely tobe treated as heresy. In particular, archaeologicalspeculation was hampered by the account of theCreation given in the Old Testament, together witha description of the subsequent settlement of knownlands by descendants of Adam and Eve. The cred-


17

ibility of the Bible was enhanced by the fact that itslater books contained episodes set in contexts withindependent historical records such as Egypt or theRoman Empire.

Some of the aspects of antiquarianism found inthe medieval Church are superficially similar tothose associated with Romans like Hadrian, but oncloser inspection are usually found to be motivatedby religion. ‘Tourism’ was common, in the form ofpilgrimages to ancient shrines, as was the collect-ing of manuscripts and relics. Many travellers com-bined both purposes; collections of relics enhancedthe status of churches as centres for pilgrimage, andgood libraries improved the reputation of monas-tic centres of learning. Since these libraries mightalso contain the works of some of the more accept-able pagan Latin and Greek authors, educated ec-clesiastics could gain some knowledge of theClassical world and its culture.

A medieval bishop of Winchester made a purelyaesthetic collection of Roman antiquities in thetwelfth century, including at least one ship-load ofmarble sculptures from Rome itself; his interest hadpresumably resulted from visits to Italy and the read-ing of Roman authors such as Pliny and Vitruviuson art and architecture. A growing number of an-cient Greek authors became known in western Eu-rope in the twelfth century AD, thanks to thetranslation into Latin of important Greek manu-scripts.

2.2 Archaeology from the Renaissance to the ‘Age of Reason’ (Piggott 1989)

In theory, when the west broke up into ‘barbarian’kingdoms in the fifth century, classical cultureshould have survived in the Byzantine Empire,where ‘Roman’ rule lasted until the capture of Con-stantinople by the Turks in 1453. However, the civi-lization that emerged from the ruins of the formereastern Roman Empire was very much a GreekChristian culture. Although much of Greece wasruled by Italian states in the final years before theTurkish conquest, they took little interest in its an-cient monuments. The heritage of ancient Rome wasof more immediate interest to Italians involved inthe Renaissance, an intellectual resurgence thatbegan in Italy during the fourteenth century.

Scholars, artists and architects turned to pre-

Christian Roman sources for largely forgotten in-formation, ideas and new inspiration. The rapidlydisappearing monuments of the city of Rome itselfwere studied by Poggio Bracciolini and FlavioBiondo in the fifteenth century, using every possi-ble source of written evidence to elucidate the physi-cal remains (Weiss 1969, 59–72). In some ways theRenaissance attitude to the examination of the pastresembled that of the Romans, for it involved travel,the study of buildings and the collection of worksof art and manuscripts. One scholar with this out-look who looked beyond Italy to Greece was Cyriacof Ancona, who was born in 1391, well before thefall of Constantinople (Weiss 1969, 131–44). Hespent twenty-five years of the early fifteenth centuryin Greece, visiting sites and libraries for himself andpublishing commentaries on his observations.

Cyriac embodied some of the principal compo-nents of a modern archaeologist, notably the activerecording and study of physical remains of the past,whether sites or objects, through extensive field-work. In addition, as a historical archaeologist,Cyriac carried out his researches with the help ofthe literary background of the culture that he in-vestigated. On the negative side, Cyriac displayeda typically selective attitude to what he recorded,and failed to record or comment upon changes thathad affected the condition of Athenian monuments:‘A true child of the Renaissance, he was the first ina long line of travellers eager to overlook the bar-barous rubbish of a barbarous people’ (McNeal1991, 52). By the sixteenth century, ‘The testimonywhich could be extracted from ancient ruins, fromstatues and mosaics, but particularly from coins andinscriptions, was now fully appreciated, and anyserious historian who dispensed with it, obviouslydid so at his personal risk’ (Weiss 1969, 206). Oneimportant element—the idea of using systematicexcavation of buried remains to supplement infor-mation gained from the surface inspection of sites—still lay far in the future.

The Renaissance atmosphere of discovery andspeculation gradually spread to the rest of Europe,including areas in the north whose connection withthe Classical world had been either brief (like Brit-ain) or non-existent (like much of Germany andScandinavia). In these countries the same spirit ofenquiry was also directed towards the non-Clas-sical past, and the first steps began to be taken to-wards the methods of prehistoric archaeology.


18

Some of this research was undertaken by individu-als whose means did not permit them to travelwidely in southern Europe. Thus, most of the ad-vances towards scientific archaeology occurred innorthern Europe, and the methods and ideas fos-tered on the fringes of the Classical world wereonly applied to sites in Greece and the Near Eastmuch later.

The Renaissance also coincided with many voy-ages of discovery, which began shortly before AD1500 and continued into the sixteenth century.Many societies at different levels of savagery or civi-lization were encountered in the Americas, and theworld was finally proved to be a sphere; both dis-coveries conflicted with the authority of the Bible.The significance of these discoveries was potentiallydramatic, but few dared to state the logical conclu-sions. Peoples unknown to the Bible in other partsof the globe could not have spread out from theGarden of Eden over a single flat continent, and theycould not be related easily to the creation storysurrounding Adam.

If this were indeed true, perhaps human racessimilarly unrelated to Adam could have existed inthe Old World before the Creation described in theBible. One writer, Isaac de la Peyrère, voiced thesedoubts; he was a French protestant who proposedin a theological book published in 1655 that Adamwas simply the ‘father of the Jews, not of all men’(Casson 1939, 115). His views were founded uponknowledge of the ancient civilizations of the NearEast and the newly discovered inhabitants of vari-ous parts of the world. Peyrère was forced to recantby the Inquisition and his book was publicly burnedin Paris. Many must have sympathized with hisviews, but they were quite beyond proof until de-velopments in geology and biology in the nineteenthcentury (outlined above) enabled archaeologists likeBoucher de Perthes to prove the existence of ‘ante-diluvian’ tool-using humans by observation andfieldwork.

The Renaissance interest in pagan classical lit-erature, combined with New World discoveries,created an favourable atmosphere for archaeo-logical work. Herodotus and Tacitus had writ-ten about primitive peoples who lived on thefringes of their (Greek and Roman) world, andthis area included Germany and Britain, whichwere both now involved in the new scholarship.The precedent of these ancient authors made it

respectable to investigate the primitive state ofEurope; for example, John Shefferius (a Swed-ish professor of law) published a study ofLapland in the 1670s, inspired by Tacitus’Germania. Since primitive peoples like the Lappswere not readily available for study in the restof Europe it is not surprising to find that thealternative in other countries was the examina-tion and description of archaeological remains.Rising sentiments of nationalism enhanced theinterest of searching for the origins of peoplessuch as the Celts, Germans or Slavs (Sklenar?1983, 24–8). As we have already seen, the ex-ploration of this kind of material was more com-plicated in northern Europe than inMediterranean countries, where research wasdominated by classical sites recorded in docu-mentary sources. Prehistoric earthworks, tombsand artefacts offered a greater challenge of clas-sification and explanation precisely because theylacked direct historical evidence.

3 Archaeology and theEnlightenment

The age which set the ‘light of reason’ againstthe darkness of medieval and baroque super-stition and prejudice based its philosophy pri-marily on the results achieved in the naturalsciences (Sklenar? 1983, 47)

The most important change in attitude to the pastcaused by the Enlightenment of the seventeenthto eighteenth centuries AD was the abandonmentof a view that humans had degenerated since thetime of the Creation. The economic and techno-logical development of Europe encouraged an al-ternative idea involving progress in humanmaterial, intellectual and spiritual culture (Evans1982, 17; Trigger 1989, 57–8). This shift in out-look was reflected in the work of many Frenchand Scottish philosophers—rather than antiquar-ies—who used reports of ‘primitive’ cultures inan attempt to define stages of social evolution:‘The ancient state of mankind was the properstudy of the philosophers untrammelled by thepetty facts of the antiquary’ (Piggott 1989, 151).Nevertheless the adoption of an evolutionary


19

frame of mind clearly favoured the acceptance ofthe implications of scientific investigations intogeology, biology and artefacts. As we have seenabove, by the early nineteenth century Europeanscholars finally possessed a range of essential con-cepts suitable for confronting the problem of hu-man antiquity.

3.1 Antiquarian fieldwork in Britain(Piggott 1989)

The aims and concepts of research into the pastthat followed the diffusion of Renaissance think-ing into northern Europe may be illustrated by thework of a series of antiquarians who engaged inactive field archaeology in Britain between theearly sixteenth and mid-eighteenth centuries:Leland, Camden, Aubrey and Stukeley. Before thesixteenth century, historical writers occasionallyreferred to monuments, but with little purposeother than to display sheer wonder, or to add cir-cumstantial detail to some actual or invented epi-sode in their works. For example, a recognizableillustration showing Stonehenge being built by themagician Merlin appeared in a fourteenth-centuryBritish manuscript (Trigger 1989, 32, fig. 2). TheTudor dynasty coincided with an increase in na-tional consciousness, which produced rationalattempts to examine the continuity of Britain fromRoman times. The founding of Britain was nolonger attributed to unlikely or imaginary indi-viduals and tribes (such as Brutus the Trojan orPhoenicians); instead, greater reliance was placedon references contained in Classical sources, andanalogies from the New World. This was the at-mosphere in which systematic attention was firstpaid to field monuments in Britain, by individualswho, from the sixteenth century, were describedby their contemporaries as ‘antiquaries’.

John Leland (1503–52)Leland was educated in London, Cambridge, Ox-ford and Paris and held the post of Keeper of King’sLibraries under Henry VIII. Like Cyriac of Anconaa century earlier, Leland travelled extensively, pe-rusing the libraries of monasteries and colleges;unlike Cyriac, his fieldwork did not involve therediscovery of sites belonging to a well-known anddocumented culture, but ill-understood remains ofunknown age, or at best fleetingly documented

monuments like Hadrian’s Wall or Offa’s Dyke. Hisinterests were primarily directed towards historicaldocuments and genealogy. His records of items thatattracted his attention (whether ancient or contem-porary) in the landscape through which he travelledmay have been intended to accompany a map oran annotated gazetteer. Leland’s significance lies inhis general idea of recording non-literary evidenceas part of wider researches. Although never pub-lished, his work anticipated the many county his-tories of the eighteenth century, and illustrated abreadth of interest in the landscape that was char-acteristic of the new, wider horizons of Renaissancescholarship.

William Camden (1551–1623)Like Leland, Camden (fig. 1.6) progressed througha sound education and eventually held a state ap-pointment that gave him ample opportunity to fur-ther his antiquarian researches. After teaching andholding the headship at West-minster School, hebecame Clarenceux King of Arms in the Collegeof Heralds. He was less devoted to the study of her-aldry and genealogy than Leland, but it neverthe-less formed an essential part of his historicalwriting. Antiquarianism was Camden’s passionfrom childhood, and he consciously acquired thenecessary skills for his purposes. Any educatedperson would possess Latin, but Camden learnedAnglo-Saxon and Welsh to study place-names. Hetravelled extensively and was already an author-ity of European standing in his twenties; when only35, his major work was published: Britannia, the‘first general guide to the antiquities of Britain’(Daniel 1967, 36).

Camden’s intention was to set Britain into arespectable position in European culture,largely by emphasizing the importance of theRoman occupation, which linked Britain to theContinental centres of the Renaissance. Thebook proceeds through the Saxon and medievalperiods, stressing the relationship of the Romanprovince to the recent historical past; no traceof the wild foundation myths of many Tudorwriters is found in the reasoned prose ofCamden. His descriptions of antiquities arethorough and detailed, and sections on coinsand language were also included. As withLeland, descriptions of the present configura-tion of the places he visited form an insepara-


20

ble part of his account of Britain; indeed,Camden made extensive use of Leland’ s unpub-lished manuscript. Camden is noted for a con-cise account of crop-marks, the visible effectsof buried structures on growing plants that areso important in the detection of sites by aerialphotography today. He also identified pre-Ro-man coins minted by the late Iron Age tribes ofsouth-eastern England—a notion rejected byLeland (Piggott 1989, 133–4).

Part of Camden’s achievement was to organizean enormous collection of information into pub-lished form, a feat Leland never managed. His in-terest in material culture, and the recognition ofthe part it could play in elucidating the past, wasfundamentally important. His Britannia enjoyedgreat popularity, and its careful organization al-lowed additions not only in his lifetime, but fornearly two hundred years after Camden’s death.The first edition of 1585 was in Latin, with onlyone illustration, but further illustrations wereadded to subsequent editions. The first Englishedition appeared in 1610, and a notable expandedversion was produced by Edmund Gibson in 1695.‘An annotated copy of the 1695 or subsequenteditions of Camden formed the almost inevitablenucleus around which comment and additionswould grow as local investigations were carried bythese amateurs who were beginning to build up the

tradition which crystallised into the great nine-teenth-century county histories and the foundationof regional archaeological societies’ (Piggott 1985,18).

John Aubrey (1626–97)John Aubrey (fig. 1.7) lacked the education ofLeland or Camden, and was described by his biog-rapher Anthony Powell as a ‘Wiltshire Squire fallenon evil days’ (Daniel 1967, 37). Nevertheless,Aubrey participated in a new kind of scholarshipthat came to prominence in the seventeenth century,centred upon the Royal Society. It was character-ized by a scientific outlook, and a wish to approachany subject from a sound basis of classification andcomparison, whether astronomy, medicine, botany,or antiquities. This new attitude was expressedclearly in 1686 by a contemporary of Aubrey, DrRobert Plot, Keeper of the Ashmolean Museum inOxford: ‘I intend not to meddle with the pedigreesor descents either of families or lands, it being in-deed my Designe…to omit, as much as may be, bothpersons and actions and chiefly apply myself tothings; and amongst those too, only of such as arevery remote from the present Age’ (Piggott 1989,26). This is the approach of an archaeologist, ratherthan a historian who happened also to be interestedin monuments, and it contrasts with that of Lelandand Camden.

1.6 Engraving of William Camden used as afrontispiece to Gough’s translation of Camden’sBritannia in 1789

1.7 Engraving of John Aubrey by William Fairthorne,dated 1666. Bodleian Library, Oxford


21

One attitude that did remain firmly establishedfrom the time of Leland and Camden was that an-tiquities formed only one part of a complete studyof the landscape. Aubrey included all kinds ofnatural and artificial phenomena in accounts ofhis beloved Wiltshire. Sadly, his great work of apurely archaeological nature, MonumentaBritannica, was never published, except for someparts included in the 1695 edition of Camden’sBritannia . The contents of MonumentaBritannica reveal Aubrey’s interests and ap-proach. The first part was focused on the greatprehistoric monuments of Wessex, includingStonehenge, Silbury and Avebury; he was one ofthe earliest writers to assign these sites to the pre-Roman Celts and their priesthood, the Druids,known from the writings of Tacitus and other Ro-man authors. Part two concerned pre-Roman andRoman camps and forts, and Roman civil settle-ments, again predominantly those of Wessex. Partthree was more diverse, including barrows, pot-tery, burials, linear earthworks, roads andtrackways, mosaic floors and coins. The fourthpart (never finished) contained assorted materialof medieval date, including a study of architec-tural features from which he deduced (correctly)that windows were particularly diagnostic fordating purposes: ‘the windows ye mostremarqueable, hence one may give a guess aboutwhat Time ye Building was’ (Piggott 1976, 17).Aubrey was right, but because his deductionswere not published they had to be worked outagain when interest in church architecture wasrenewed in the nineteenth century.

The most significant feature of Aubrey’s workwas the idea that information was worth collect-ing and classifying for its own sake, rather thansimply to illustrate a particular theory. The sameapproach is found in the work of his contempo-raries in other fields of learning such as botanyand the study of fossils. Aubrey’s observationsand interpretations also reveal the effects of an-other important feature of his age—the dissemi-nation in Europe of accounts of AmericanIndians. He did not share the ‘Noble Savage’view of native Americans that might have re-sulted from reading the Germania of Tacitus;this is clear from analogies he drew betweenancient Britons and Indians: ‘…the inhabitants(of northern Wiltshire) almost as savage as the

21 Beasts whose skins were their only rayment…They were 2 or 3 degrees I suppose less savagethan the Americans…The Romans subdued andcivilized them’ (Piggott 1989, 62). Clearly,Aubrey shared Camden’s view that the Romanperiod made Britain acceptable in the eyes ofEuropean post-Renaissance scholarship. Rem-nants of this concept survive today; the archae-ology of Roman Britain is normally studied inthe wider context of the Roman Empire, butrarely with sufficient reference to the periodsthat preceded or followed it.

Aubrey was not able to escape from the conun-drum of dating ancient monuments. Although hewas right to place Stonehenge and Avebury in aritual context of pre-Roman date, he attributed IronAge hillforts to Britons, Romans or Danes with wildinconsistency (Piggott 1989, 118–20). Sadly, deathcut short the activity of Edward Lhwuyd (1660–1708), a younger scholar and friend of Aubrey. Hehad participated in preparing Gibson’s edition ofCamden’s Britannia, but died before the publica-tion of his own account of the monuments of Brit-ain that should have followed his book of‘staggering brilliance’ on Britain’s Celtic languages(Piggott 1989, 30). Aubrey’s style of scholarship hada final successor in the eighteenth century—WilliamStukeley.

William Stukeley (1687–1765) (Piggott 1986)The early eighteenth century saw a decline in thespirit of objective scientific inquiry that had char-acterized the Royal Society, and a shift in tasteaway from Classical art and architecture towardsfanciful ‘Gothic’ buildings incorporating medi-eval features; the same period also saw the devel-opment of ‘Romantic’ art and literature. Stukeley(fig. 1.8) reflected these changes in the spirit ofthe age; his early researches belonged to a pass-ing tradition, but his later interpretations of thembelonged to the new era. Stuart Piggott (Stukeley’sprincipal biographer) pointed out the contradic-tions contained in his work by suggesting that‘…perhaps we should see William Stukeley’sgreatness as a field archaeologist not in his beingan innovator, but in his being provincial, old-fash-ioned and out-of-date, continuing the high tradi-tion of Restoration antiquarianism unaware ofthe changed intellectual mood of the metropolis’(1982, 24).


22

Stukeley was trained in medicine at Cambridgeand became a Fellow of the College of Physiciansin 1720, having practised as a doctor. He had alsostudied botany, and this experience probably ledhim to an appreciation of the ancient monumentsin the countryside that captured his imagination,especially after reading the manuscript of Aubrey’sMonumenta Britannica in 1718. Extensive field-work in Wessex followed in the 1720s, includingaccurate and thorough surveys of Avebury, Stone-henge and Silbury. He went on to travel extensivelythroughout Britain, from the south coast to Hadri-an’s Wall, making surveys and excellent sketches.So far, Stukeley was a typical scholar of the ‘Age ofReason’; but signs of ‘unreason’ are evident in histaste for dramatic landscapes such as the Lake Dis-trict, and for Gothic architecture (to the extent ofdesigning mock-ruins or ‘follies’). His life changeddirection from medicine to religion in the 1720s, andhe was ordained in 1729.

From this point, Stukeley attempted to use theresults of his collected fieldwork from Wessex toestablish a theological connection between the Dru-ids and Christianity. Aubrey had made observations,

sorted them into a sensible order and drawn lim-ited conclusions from common sense and histori-cal information; for him, Stonehenge and its relatedmonuments did not fit into the Roman period, sohe attributed them to the preRoman Britons. Sincethe sites were obviously ritual, not functional,Aubrey had assigned them to the only known cultand priesthood attested by Classical authors, theDruids. Stukeley accepted this attribution (whilefailing to acknowledge his debt to Aubrey), but wenton to invent a vast theological system for the Dru-ids, supported by quite unwarranted connectionswith features of the monuments: ‘The form of thatstupendous work [Avebury] is the picture of theDeity, more particularly of the Trinity’. He pub-lished two major books—Stonehenge (1740) andAvebury (1743)—that he claimed to be part of alarger enterprise entitled Patriarchal Christianity ora Chronological History of the Origin and Progressof true Religion, and of Idolatry.

Stukeley’s reputation as a field worker rests noton these publications, but on the unpublishedmanuscript of 1723 from which he selected his in-formation. One early historian of archaeologysummarized the curious dualism of Stukeley par-ticularly clearly: ‘Just as Dr Stukeley may be saidto be the patron saint of fleldwork in archaeology,so can the Rev. William be held to be the evil gen-ius who presides over all crack-brained amateurswhose excess of enthusiasm is only balanced bytheir ignorance of method’ (Casson 1939, 150).

1.8 A drawing by William Stukeley (1687–1765)showing him engaged in fieldwork with friends. Evenin this light-hearted sketch a number of antiquitiesand features of the landscape are drawn andlabelled; his observations and plans remain animportant source of information.Bodleian Library, Oxford (Ms Eng Misc b 65 fol 43r)


23

Nevertheless, Stukeley’s basic evidence still formsan invaluable record of monuments that have suf-fered severely since his day. He recorded an avenueof stones leading from Stonehenge to the riverAvon that was subsequently destroyed; it was onlyrelocated by aerial photography in 1920 (Piggott1985, 92). His observations were careful and logi-cal, and included the ability to relate groups ofseparate earthworks in an area into a coherentpattern. Stukeley also made analytical observa-tions, such as deducing that a Roman road onOakley Down, Dorset, must have been constructedlater than some ‘Druid’ burial mounds, because itcut across the ditch of one barrow (Piggott 1989,pl. 27). Skills like these are still essential today. Itis noteworthy that Stukeley was already aware ofthe role of fieldwork as part of rescue archaeol-ogy: as it ‘perpetuates the vestiges of this celebratedwonder & of the barrows avenues cursus &c forI forsee that it will in a few years be universallyplowed over and consequently defaced’ (Piggott1989, 127).

3.2 Fieldwork elsewhere(Malina & Vas?íc?ek 1990)

Advances in antiquarian research and fieldworkwere not restricted to Britain, of course. Historiansof ideas, science or archaeology can all point to simi-lar phenomena taking place elsewhere in Europe atthis time. In Scandinavia, Johan Bure and Ole Wormundertook antiquarian research—with royal pa-tronage—in the early seventeenth century (Trigger1989, 49), and similar efforts were devoted to Ro-man and earlier antiquities in central Europe(Sklenar? 1983, 6–43). A German pioneer of thesystematic investigation of Roman art and architec-ture in Italy, Johann Winckelmann, was a near con-temporary of William Stukeley. An indigenousarchaeological tradition had also emerged inAmerica by the nineteenth century (Trigger 1989,104–9). Inevitably it began with ethnographic ac-counts of the native Americans, but gradually ex-tended to sites and artefacts. The literatecivilizations of Central and South America attractedcomment as early as the sixteenth century, for theirarchitecture, sculpture and inscriptions offered thesame kind of possibilities for study as those ofGreece or Italy. By the end of the eighteenth cen-tury, it was generally accepted that the native popu-

lation of North America had migrated from Asiaby way of the Bering Straits. Nevertheless, specula-tion about the origins of Indians was still influencedby a desire amongst European colonists to justifytheir conquests by proving that the natives were insome way inferior to themselves. This was fre-quently achieved by claiming a glorious past forareas that they now occupied. Some archaeologi-cal fieldworkers who recorded ritual earthworks,such as burial mounds reminiscent of those foundin northern Europe, attempted to explain them interms of ‘lost races’, identified with Israelites, Danesor even Welshmen. Willey and Sabloff aptly catego-rize the phase of archaeological study in Americafrom 1492–1840 as ‘the speculative period’ (1980,12–27).

From a methodological point of view, field ar-chaeology in Britain could develop no further af-ter Stukeley until some new element wasintroduced (‘For prehistoric earthworks the utterconfusion and ignorance of 1695 was to persistinto the present century’—Piggott 1989, 120).Accurate recording was continued and extended,and many county histories maintaining the styleof their work appeared into the nineteenth century;however, the interpretation of recorded monu-ments remained static, because historical evidencescarcely stretched back beyond the Roman period.Historical events could be shuffled into a differ-ent order, or fanciful theories could be constructedto expand them, but no new source of evidence wasavailable until the idea of excavation was adoptedon a large scale in the nineteenth century, and re-fined in the twentieth. This development will befollowed in the appropriate chapter below, but afurther stage in the unravelling of the past requirescomment. The collection and study of objects,rather than sites, ran parallel to the developmentof archaeological fieldwork but triumphed in thenineteenth century when excavations began toprovide growing quantities of pottery, metal andstone artefacts for study.

3.3 Touring and collecting(Impey & MacGregor 1985)

The Renaissance revived the Roman penchant forvisiting monuments and collecting works of artfor aesthetic reasons, in contrast to the medievalChurch’s concentration upon shrines and relics.


24

The concept spread to northern Europe, and edu-cated people of sufficient financial means beganto visit the Mediterranean centres of classical civi-lization in Italy, Greece, Turkey and the NearEast. Naturally, travellers purchased ‘souvenirs’to adorn their northern residences (constructedand decorated, of course, in a classical manner),and the process was accelerated by agents sent toseek out further items and to arrange for theirshipment to the new owners’ homes. An earlyexample of an English ‘Grand Tour’ aristocratwas Thomas Howard, Earl of Arundel (1585–1646), who first travelled with a large entourageto Italy in 1612; there, he bought, and even dugfor, antiquities. His agent, William Petty, extendedthe search to Greece with the help of the localambassador, who described him as ‘all things toall men that he may obtain his ends’ (Casson1939, 128). Petty built up a collection (at a bar-gain price compared with buying in Italy) thatbecame a centre of great learned interest, knownthroughout Europe after its publication in 1628.Although Arundel’s collection suffered neglectand dispersal after the Civil War, it had alreadygenerated similar desires amongst other noblemenand even royalty. Indeed, King Charles I statedthat ‘The study of antiquities is by good experi-ence said to be very serviceable and useful to thegeneral good of the State and Commonwealth’(Daniel 1975, 19).

Tours had other effects too; learned societies,such as the Society of Dilettanti (an organiza-tion of British antiquaries) sponsored expedi-tions to record Classical sites, rather than simplyto loot them. Individuals of lower social statusand wealth also began to form collections thatincluded a wider variety of items (fig. 1.9). Forexample, John ‘Gardener’ Tradescant’s collec-tion was created in the first half of the seven-teenth century, and a catalogue of its contentsappeared in 1656. Although largely made up ofbotanical specimens, i t also comprised‘Mechanick artificial works in carvings, turn-ings, sowings and paintings’ and ‘warlike instru-ments’, mainly from Polynesia, Africa andAmerica (Casson 1939, 136). After his death, thematerial passed to the University of Oxfordthrough Tradescant’s friend, Elias Ashmole. Anew museum was opened in Oxford in 1683 bythe future King James II, and it moved in the

nineteenth century to the building knownthroughout the world as the Ashmolean Mu-seum; the original building, now a museum ofthe history of Science, still exists. Thus, the Ren-aissance fashion for collecting contributed to theestablishment of public museums attached tocentres of learning or to cities. Museums havebecome the first point of contact with archaeol-ogy for many members of the public. The essen-tial features of the early Ashmolean (collecting,scholarship and public display) have becomeaccepted as integral parts of the cultural life ofalmost every modern country. The interest ofantiquaries like Aubrey and Stukeley in prehis-toric sites and objects was connected to the samephenomenon; indeed, many travelled in Britainbecause they could not afford to go abroad.However, early field archaeologists naturallyconcentrated on sites, because the potential forusing objects to distinguish between stages of de-velopment in prehistory remained extremely lim-ited until a meaningful concept of time becameaccepted. We have already seen that this accept-ance came only in the second half of the nine-teenth century. The history of the study ofobjects, like that of fieldwork, provides a usefulillustration of some basic principles that stillunderlie the subject to this day.

4 The recognition of humanartefacts

There was not normally any doubt that Romanand later artefacts were the work of humans, butordinary artefacts from historical periods attractedlittle interest unless they were considered to haveaesthetic qualities. Unlike works of art, they wereonly recovered by accident until excavation be-came an essential part of archaeology during thenineteenth century. Sophisticated prehistoric ob-jects made of cast bronze were commonly assignedto the Romans or Danes, because antiquarianslacked a clear idea of what to expect from prehis-toric material culture. For these reasons the sys-tematic study of objects began with simple stonetools from very early periods. Casual finds of finelyworked flint arrowheads or polished axes mustalways have suggested human manufacture to any-


25

one who actually thought about them, and it wouldnot have been difficult to reach the idea that theymight have been used before metals were known.Concepts of successive ages of stone, bronze andiron, suggested by actual finds, are known fromChinese literature as early as the first century BC,and Shen Kua made remarkable studies of artefactsin the eleventh century AD (Evans 1982, 13–14).

A number of Italian collectors made accurateidentifications of human artefacts in the sixteenthcentury (Piggott 1989, 89–90), but the conceptspread slowly, so that it still seemed a novelty whenthe seventeenth-century French theologian, Isaac dela Peyrère, proposed that stone implements were not‘thunderbolts’, but the tools and weapons of peo-

ples who had preceded the creation of Adam(Piggott 1989, 45–7). The matter was soon placedbeyond doubt when similar items became availablefor study in ethnological collections from the SouthSeas and the Americas, where they could still beobserved in use (fig. 1.10). In 1800, John Frere’scelebrated letter published in Archaeologia (above,p. 11) included drawings of typical flint hand-axesof the Old Stone Age, ‘…evidently weapons of war,fabricated and used by a people who had not theuse of metals…’. However, fifty years later, Boucherde Perthes was still fighting for the acceptance ofsimilar artefacts as the work of early humans. In-terestingly, bronze artefacts caused more problemsthan those made of stone or iron for, while earlytravellers could observe Stone Age communities inAmerica and Australia, and Iron Age societies inmany parts of Africa, no Bronze Age peoples hadbeen encountered. Bronze artefacts were normallyassigned to the Romans, because they seemed too

1.9 Ole Worm’s collection of natural and archaeologi-cal curiosities, illustrated in 1655, illustrates thebreadth of such collections, which embraced natural,geological, ethnographic and archaeological speci-mens. Bodleian Library, Oxford


26

complex to have been made by ‘savages’, but sug-gestions of an earlier date by Camden were builtupon by Edward Lhwuyd and others, and foundsome support by the eighteenth century (Piggott1989, 95–100).

4.1 Scandinavia and the Three-Age System(Graslund 1987)

Why has Scandinavian archaeology, generallyspeaking, an advantage over foreign archaeol-ogy, if not because Scandinavian archaeologistshave had an opportunity to study in their mu-seums not isolated specimens but whole seriesand their development? (Hans Hildebrand,1873, quoted in Graslund 1987, 16)

The archaeology of Scandinavia is particularlyrich in finely made artefacts dating from the pre-

historic to Viking periods, and many of them arefound in good condition in graves. Hildebrandwas right to stress these factors, for increasedbuilding, agriculture and excavation in the nine-teenth century had provided a plentiful supplyof discoveries. Fortunately, Scandinavia also hadmuseums where objects could be preserved, stud-ied and displayed. An Antiquities Commissionwas set up by the Danish government in 1807to protect sites, promote public awareness of an-tiquities, and to establish a museum. The firstcurator of the resulting National Museum in Co-penhagen was Christian Thomsen (fig. 1.11),who held the post from 1816 to his death in1865.

Thomsen would have been well aware of theconcept of successive ages of stone, bronze and iron;it was particularly well expressed by Simonsen in1816:

1.10 Stone artefacts from the Old and NewWorlds. Recognition of prehistoric implements inEurope was helped by observations of similarobjects, still in use, in other parts of the world. In1699, Edward Lhwuyd, keeper of the AshmoleanMuseum and an authority on fossils and antiqui-ties, wrote: ‘I doubt not but you have often seenof these Arrowheads they ascribe to elfs orfairies: they are just the same chip’d f lints the

natives of New England head their arrows with atthis day; and there are also several stone hatchetsfound in this kingdom, not unlike those of theAmericans’ (Piggott 1989, 86). The artefacts onthe left come from North and South America;those on the right are from northern Britain, anddate to the later Stone Age and the Bronze Age.Hancock Museum and Museum of Antiquities,Newcastle upon Tyne


27

At first the tools and weapons…were made ofstone or wood. Then the Scandinavians learntto work copper and then to smelt it and hardenit…and then latterly to work iron. From thispoint of view the development of their culturecan be divided into a Stone Age, a Copper Ageand an Iron Age. (Daniel 1967, 90–1)

Thomsen was the first to demonstrate the validityof these hypothetical ages by examining ‘closedfinds’ (graves, hoards, etc.) in which artefacts hadbeen discovered. He restricted his central definitionof the Three Ages to cutting weapons and tools, andestablished their relative order. Some finds con-tained only stone tools, plus a few stone withbronze, but never iron. Bronze weapons were alsofound without iron, and bronze continued to beused for other kinds of objects during the Iron Age,which was observably the most recent age becauseof associations with Roman and medieval coins.Once this analysis had established the order ofweapons and tools, Thomsen was able to observe

their association in closed finds with artefacts madefrom other metals and materials, as well as specificburial practices and grave forms. The basic objec-tivity of these observations meant that later findselaborated and refined his system, rather than re-placed it. Effective classification was indispensableto the advance of the study of prehistory, and theThree-Age System provided a framework that sur-vives (with modifications) today.

Thomsen presented the evidence for his chrono-logical deductions in his museum by displayingtogether groups of objects that had been found inassociation. He was keen to show them to visitingarchaeologists, and also to peasants, who werelikely to discover objects that could be added to thecollections. The displays were described in a guideprinted in 1836, and received wider attention afterit was translated into English in 1848. The phenom-enon of collecting antiquities was initially a hobbyof a social elite, typified by the Earl of Arundel, butby the nineteenth century it had been transformedin a remarkably democratic fashion. Happily, thepopularizing approach of Thomsen was reinforcedby other archaeologists, such as Pitt Rivers, and itremains characteristic of most museum curatorstoday. However, unlike Pitt Rivers, Thomsen did notattempt either to study the development of the formsof individual artefacts (‘typology’) or to explain thereasons for the changes that he had observed(Graslund 1987, 26–8).

Thomsen’s successor as Director of the Dan-ish National Museum was another remarkableman, Jens Worsaae (1821–85). His recommenda-tions for the use of systematic excavation wereinspired by the need to recover still more artefactsfrom specific contexts that would allow Thom-sen’s broad classifications to be refined. GlynDaniel (1967, 99–100) noted a key phrase inWorsaae’s writings, published in 1843: ‘As soonas it was pointed out that the whole of these an-tiquities could by no means be referred to one andthe same period, people began to see more clearlythe difference between them.’ In 1861 Worsaaesubdivided the Stone Age into three periods ac-cording to the nature of stone artefacts. The ear-liest period was characterized by hand axes andlarge flakes, found in the gravels and caves ofwestern Europe; these were followed by finertools found in Denmark in ‘kitchen middens’(mounds of shells and bones left by hunter-gath-

1.11 C J Thomsen in the Oldnordisk Museum inCopenhagen thirty years after its foundation in 1816.His enthusiasm for increasing public awareness ofantiquities is well illustrated by this drawing by MagnusPedersen. National Museum, Copenhagen


28

erers). Finally, polished stone tools were associ-ated with elaborate tombs that occasionally alsocontained the earliest metal objects. These divi-sions of the Stone Age were soon namedPalaeolithic, Mesolithic and Neolithic (old, mid-dle and new) by Sir John Lubbock in his bookPrehistoric Times (1865). Worsaae used a differ-ent method to divide the Bronze Age. He identi-fied a series of different burial practices and graveforms, and was able to place them into chrono-logical order either by reference to artefacts foundin them, or by observation of excavated siteswhere examples of different forms had been foundin a stratigraphic sequence. Like Thomsen,Worsaae relied primarily on the contexts of arte-facts, rather than a typological study of the arte-facts themselves.

4.2 Typology

Typology differs fundamentally from mere clas-sification. It studies classes of artefacts from thepoint of view of developments and changes thatmay allow them to be placed into a hypotheticalchronological order (below, p. 104). Until re-cently, the development of typology has beenlinked to the Three-Age system and the influenceof Darwin’s theory of evolution. However, BoGraslund (1987) has demonstrated that this wasnot the case by means of a thorough study of theoriginal writings of Thomsen, Worsaae,H.Hildebrand, Montelius and many otherScandinavian scholars (few were ever translatedfrom Danish or Swedish). Studies of artefactswere based primarily on the contexts in whichthey had been discovered, and in Scandinaviathese were sufficiently plentiful for virtually allclasses of artefacts to be placed into chronologi-cal order without the use of typology. Once thishad been done, typological studies could beginon a secure basis. Evolution provided a conven-ient, if rather misleading, metaphor, and it un-doubtedly stimulated the development oftypology from the 1870s onwards.

The influence of classical archaeology on ty-pology has been underestimated because mosthistories of archaeology have been written byprehistorians. Systematic studies of Greek andRoman architectural and artistic styles beganduring the Renaissance, and were formalized by

Johann Winckelmann in the eighteenth century.Ancient coins were even more significant:Petrarch studied inscriptions and portraits in thefourteenth century, and classifications of largecoin collections were published from the six-teenth century (Berghaus 1983, 19–23). Josephvon Eckhel’s Doctrina numorum veterum(1782–98) and similar works by other authorsprovided comprehensive geographical andchronological classifications that must have beenuseful reference tools for C.J.Thomsen and hissuccessors. Graslund has rightly stressed the im-portance of the numismatic knowledge ofThomsen, Hans Hildebrand and Montelius, whoall appreciated the importance of coins as dat-ing evidence that could be used to subdivide theScandinavian Iron Age (1987, 66). It is alsoimportant to recognize that coins are artefacts,and that their study by means of stylistic se-quences of portraits or other ornamentation,combined with changes in size and weight, bearsmany similarities to typology. The styles of clas-sical sculptures and Greek painted vases werealso studied primarily from the objects them-selves, largely because their contexts were rarelyrecorded.

Augustus Lane-Fox (1827–1900, a lsoknown as Pitt Rivers, after taking this nameunder the terms of an inheritance in 1880) col-lected artefacts from all over the world from theearly 1850s while serving in the GrenadierGuards. He was involved in replacing musketsby rifles in the British army, and in testing vari-ous models and modifications for reliability andefficiency. This probably contributed to the wayhe sought to illustrate progressive improve-ments and developments of various classes ofartefacts in his personal research. He liked tocollect examples of the principal stages in-volved, and, in contrast to earlier collectors likeJohn Tradescant, he assembled artefacts ‘…notfor the purpose of surprising anyone, either bythe value or beauty of the objects exhibited, butsolely with a view to instruction. For this pur-pose ordinary and typical specimens rather thanrare objects have been selected and arranged insequence’ (Daniel 1981, 140). Pitt Rivers’ con-cept of typology was very different from thatof Montelius, for he invoked analogies withDarwinian evolution as early as the 1860s


29

(Bowden 1991, 54). His scheme for the deriva-tion of Australian weapons placed a variety ofclubs, boomerangs, throwing sticks, shields andspears into sequences from simplicity to complex-ity (fig. 1.12), but a little critical thought soon un-dermines this idea. A shield is only a shield whenit is broad and flat, and a boomerang is not a boo-merang if it does not fly; Pitt Rivers’ model failedto take account of invention. Nevertheless, theanalogy derived from Darwin’s concept of linearevolution still underlies many modern studies oftypology (Basalla 1989).

As soon as Scandinavian prehistory had beensubdivided according to studies of groups of arte-facts found together in graves and other contexts,attention was turned to the artefacts themselves.The work of Oscar Montelius encompassed thewhole of Europe from the 1880s, and he used hisbroad knowledge to fix dates for the Bronze andIron Ages by cross-referencing north Europeanfinds to datable objects exported from the civili-zations of Egypt and the east Mediterranean (see

p. 104). Fellow Swedes Salin and Aberg continuedtypological research in the twentieth century bystudying objects and ornamental styles associatedwith Germanic tribes of the Roman and ‘Dark Age’periods. Like Montelius, they used dated findsfrom southern Europe to provide fixed points inthe archaeological sequences of Scandinavia. Theintroduction of radiocarbon dating in the 1960srevealed major errors in the dating of European

1.12 Pitt Rivers justif ied his idea of typology byderiving a wide range of Australian aboriginalweapons from a plain cylindrical stick. ‘…I only askyou to glance at the sequence shown in thisdiagram…in order to convince you of the truth ofthe statement…that although, owing to thecomplexity of modern contrivances and the largersteps by which we mount the ladder of progress inthe material arts, their continuity may be lost sightof, when we come to classify the arts of savagesand prehistoric men, the term growth is fully asapplicable to them as to the development of theforms of speech…’ Lane-Fox 1875, pl. III and p. 514


30

prehistory, and cast typology in a bad light. Nev-ertheless, with appropriate caution, the techniqueremains fundamental to the classification andstudy of artefacts of virtually any kind or datefound anywhere in the world. Unfortunately, theclassification of objects and cultural stages pre-ceded the general development of improved exca-vation techniques by several decades.

Modern archaeologists should not allow theyears that have elapsed since Thomsen, Worsaae orPitt Rivers to provide an excuse for ignoring theintellectual context of their work, or forget that theyunderpinned the political attitudes of their day. AliceKehoe expresses strong scepticism of their portrayalin ‘Whiggish consensus histories’ by Glyn Danieland others:

The three-age system was a variant of a myth ina standard, familiar format, the three offers ortests to which the culture hero must respond.Stone, Bronze, Iron—like the Three Little Pigs,he who persevered in working the least tracta-ble material in the end won out…. When thegreat historical change from landed estates toindustrial capitalism and bourgeois democracieshad been achieved, the format for a sanctioningmyth was already given. Archaeology furnishedthe hero—mankind—and the substantive de-tails—stone, bronze, iron—to lend verisimilitudeto the myth…. To ignore this context is to missthe forest for the trees. (Kehoe 1989, 106)

5 The Discovery ofCivilizations(Daniel 1981)

Interest in material remains, and in particular theconcept of excavating sites for information ratherthan in search of treasures, developed long after thegreat period of descriptive study characterized byantiquarians like Camden or Aubrey. Archaeologi-cal exploration usually began for one of two rea-sons. Some structures, such as Hissarlik(Schliemann’s Troy), were investigated because theywere thought to relate to historical people, periodsor events. Conversely, mysterious monuments suchas the pre-Columbian North American moundswere dug into in the hope of revealing their nature

and date (Daniel 1981, 90–2). A third factor existedalmost universally: treasure hunting, either forpurely financial gain, or, on a more intellectualplane, in search of curiosities or objets d’art forcollectors.

The Mediterranean civilizations of Greece andRome formed an important background to Euro-pean culture, and they received special attentionduring the Renaissance and Enlightenment peri-ods. This degree of familiarity reduced classical ar-chaeology’s potential for introducing newtechniques and concepts, in contrast to the stimu-lus given by difficult questions such as HumanAntiquity, or the exploration of Egypt and Meso-potamia. In the years around 1800, Europeans firstbegan to turn to more systematic forms of exca-vation. By a happy coincidence, the declining Turk-ish Empire was allowing easier access to theseregions, resulting in the presence of diplomats andsoldiers from France and Britain (and later Ger-many) around the Red Sea and Arabian Gulf—thestrategic routes that connected the Mediterraneanto the Indian Ocean. Many of these individualscame from the same educational and social back-ground as antiquarians, who had studied the clas-sics and travelled to historic sites on the GrandTour, and who, increasingly, turned to the investi-gation of ancient sites on their estates. It is there-fore not surprising to find that Claudius JamesRich (agent of the East India Company at Bagh-dad from 1807) or Paul Emile Botta (French con-sul from 1842 at Mosul, the ancient Nineveh)investigated the remains of Babylon and Ninevehand other sites in Iraq near the towns where theywere based (Lloyd 1980). National prestige be-came embroiled in the pursuit of antiquities, andas a result vast sculptures and even large portionsof buildings were transported to the museums ofLondon, Paris and Berlin. Napoleon’s invasion ofEgypt in 1798 was even more striking; althoughNelson ensured that it was not a military successit certainly was an academic triumph. The 200scholars who accompanied Napoleon’s army es-tablished the foundations for decades of subse-quent research into Egypt’s civilization andprehistory (Trigger 1989, 39).

The failing grip of the Ottoman Empire in thenineteenth century also stimulated the explorationof Greek civilization. Greece gained its independ-ence in 1831 and foreign excavators rapidly


31

cleared the Acropolis of Athens, releasing the re-mains of such buildings as the Erechtheum andParthenon from the encumbrances of a harem andmosque respectively (fig. 1.13). This action alsoillustrates that archaeological research is highlyselective when combined with nationalism. Socie-ties tend to select the past that they wish to em-phasize; the removal of physical reminders ofTurkish rule and its religion, Islam, allowed thenew Greek nation to underline its connections withthe European roots of classical culture.

5.1 Egypt and Mesopotamia(Maisels 1993)

Interest in Egypt and Mesopotamia was not en-tirely separate from the investigation of classicalGreece and Rome. Both areas had fallen withinthe power of Alexander the Great in the fourthcentury BC, and both were absorbed into theRoman empire in the first century BC. Thus someindications of the early history and antiquities ofEgypt and Mesopotamia could be gleaned fromclassical writers, while even earlier referencesabounded in the Old Testament. A further impor-tant aspect was that a wide public could take asafe interest in news of discoveries made in theNear East, for they promised to enrich and con-

firm two major roots of European culture—theClassics and the Bible. In contrast, unsettlingclaims were being made about unspectacularstone tools of dubious human manufacture foundin France and elsewhere in the early nineteenthcentury (above, p. 11). Many geologists relatedthese finds to the ideas of Charles Darwin, andhad the effrontery to deny both the date and na-ture of the Creation recorded in the Book of Gen-esis: ‘To the side of scholarship and literaryinterest there gravitated, imperceptibly, the bulkof those religious-minded traditionalists whowere alarmed at the tendency of the times’(Casson 1939, 207).

The methods developed since the Renaissance

1.13 Stuart and Revett’s engraving of the Parthenon,Athens, published in 1787, shows Turkish houses anda mosque that were removed when Greece becameindependent in 1831. They published five volumes ofarchitectural studies and views of buildings between1762 and 1830, and placed great emphasis uponaccurate recording, for these books were intendedfor use by architects building in the neo-classicalstyle. Fortunately for modern researchers with awider interest in these sites, they began by sketchingthe actual condition of each monument. Stuart &Revett 1787, pl. 1


32

for the study of classical Greece and Rome, basedupon a co-ordinated investigation of literature, artand architecture, provided a model for the studyof Egypt and the Near East (Trigger 1989, 35–40).Literary interest was soon given a tremendousboost, for the written languages of both regionswere deciphered by the middle of the nineteenthcentury. An inscription on the Rosetta Stone (dis-covered in Egypt by a French officer in Napole-on’s army in 1799) turned out to have been writtenin two different Egyptian scripts and also in Greek.The stone was taken to Britain after Napoleon’sdefeat, but attempts to use the Greek text as a keyfor understanding the Egyptian scripts culminatedin success by a French scholar, Jean FrançoisChampollion, who published a grammar and dic-tionary of Egyptian hieroglyphics in the 1830s(Andrew 1992). The cuneiform script of Mesopo-tamia was first translated around the same time,and the early Babylonian language of the regionwas deciphered with the help of a gigantic inscrip-tion carved on a high cliff at Behistun in Persia,recorded by Henry Rawlinson, a soldier and dip-lomat in the region who eventually became cura-tor of the British Museum in 1876. It includedidentical texts written in Persian, Babylonian andElamite to proclaim the authority of the Persianking Darius over his conquests, and the study wascompleted by 1857.

The implications of these translations were tre-mendous: ‘The development of Egyptology andAssyriology in the course of the nineteenth centuryadded 3000 years of history to two areas of theworld that were of particular interest in terms ofbiblical studies, but for which no direct documen-tation had been available’ (Trigger 1989, 40).Countless Egyptian hieroglyphic inscriptions werealready known (their use had continued underGreek and Roman rule until at least the end of thefourth century AD), and buildings could now bedated according to the names of Pharaohs in-scribed on them. The decipherment of ‘cuneiform’writing allowed the translation of thousands ofclay tablets found on excavations throughout thearea; these tablets frequently provided details ofpalace stores and accounts, as well as historicalinformation. Egypt and Mesopotamia thus joinedGreece and Rome in having a detailed historicalframework for the study of their culture and physi-cal remains.

The increasing interest in Near Eastern civili-zations was not entirely beneficial, for it led to in-tensive plundering of sites for carvings andinscriptions to satisfy greater demands from mu-seums and collectors. In Mesopotamia, even pal-aces and temples were largely built out of sundriedmud brick (see fig. 3.33)—unlike their stone coun-terparts in Egypt. Fragile structures and perishableor unimpressive artefacts were neglected for mostof the rest of the nineteenth century, along withany earlier prehistoric levels underlying historicalsites. Casson pinpointed the problem: ‘Scientificmethod existed. But for the archaeologists of thevarious phases of civilized man there were no sci-entific collaborators…. This divorce of sciencefrom archaeology, in so far as the later phases ofcivilization were concerned, was largely due to thefact that historical sites fell automatically under thecontrol of literary men’ (1939, 215). Frere,Worsaae and Boucher de Perthes observed andrecorded the stratigraphical contexts of prehistoricartefacts because it was the only possible sourceof chronological evidence; but with historicalrecords written in hieroglyphs or cuneiform, whoneeded strata?

So far, then, planned fieldwork was almost ab-sent from the process of discovery, and it was in-hibited by the nature of the study of literatecivilizations. As a result of the publicity surround-ing Schliemann’s major discoveries at Troy in the1870s, a request was made in England to obtainfinancial support from the Treasury for work onburial mounds in the same area of Turkey, on thegrounds that they were of as much potential inter-est as the Temple of Diana at Ephesus, which wasalready receiving financial support. The officialreply was mortifying: the work at Ephesus wasundertaken ‘…not for the purpose of ascertainingthe site or the form of the Temple, objects quitebeyond the scope of the Trustees [of the BritishMuseum], but for the sake of such relics of ancientart as might be found buried among the ruins. Theascertainment of the site was a mere incident….The question then is: are excavations undertakenfor the purpose of illustrating the Iliad a propersubject for the expenditure of public money? I amsorry to say that in my judgement they are not’ (Rt.Hon. Robert Lowe, 1873). Mention of Schliemanndoes hint that clearer objectives were finally com-ing into the study of early civilizations. The late


33

nineteenth century also witnessed a more system-atic approach to the recording of surface remainsof monuments, using improved surveying tech-niques, combined with the rapidly developing tech-nique of photography.

5.2 Schliemann and Troy(McDonald & Thomas 1990)

Heinrich Schliemann was born in Germany in1822. His commercial skills and gift for lan-guages allowed him to close down his businessinterests in 1863 to devote himself to travellingand studying the ancient Greek world until hisdeath in 1890. Part of the enduring appeal ofSchliemann’s life-story lies in his rather dubiousrole as an outsider who took on the academicestablishment and outwitted the Greek andTurkish authorities in the relentless and success-ful pursuit of his theories. How far this view iscorrect may be debated, but the persistence, dis-cipline and intelligence that brought him com-mercial success and a rapid rise from shopassistant to Californian banker would have beenhelpful in approaching excavation: ‘The grocerwho unpacks crates is better equipped to unpackthe middens of antiquity than the polite scholarwho has never seen the inside of his own dust-bin’ (Casson 1939, 224). However, Schliemannwas not the only archaeologist in Greece orTurkey to pay attention to the recognition andrecording of stratigraphy and finds during anexcavation. In the 1870s an Austrian, Alexan-der Conze, working at Samothrace and a Ger-man, Ernst Curtius, at Olympia both appliedrigorous methods of excavation inspired by therecent work of Guiseppe Fiorelli at Pompeii inItaly (Trigger 1989, 196–7).

Nineteenth-century German literary scholarsconsidered that the Iliad (Homer’s epic poem re-counting stories of the Trojan Wars) was not basedon a historical reality, but involved miscellaneousaccounts of mythical heroes. Schliemann held theopposite view and, having combined study of theHomeric text with fieldwork in Greece and Tur-key, he published observations about Mycenae andthe location of Troy in 1869—two years before hebegan to excavate the latter site (fig. 1.14). Hedrew wide attention to his findings through therapid publication of his work at Troy and related

sites, as well as popular reports to newspapers suchas The Times. His results have undergone consid-erable reinterpretation, initially by his co-workerDörpfeld, who only three years after Schliemann’sdeath, redefined the occupation level at Troy thatwas considered to have belonged to the Homericperiod.

Although Schliemann’s excavations and re-search around the Aegean were initially moti-vated by the desire to elucidate a specific literarytext, they brought the Greek Bronze Age and itsantecedents to light for the first time. He con-ducted his work as a conscious problem-orientedexercise, rather than simply to recover attractivefinds from a known historical site; he also paidattention to the whole stratigraphic sequence atTroy, not just a single period. His approach wasin stark contrast to Mariette’s discovery of theSerapeum, at Memphis in Egypt, in 1851.Mariette knew about the site from an ancientGreek traveller’s account and from references inEgyptian papyri, but only discovered it thanks toa good memory and the chance observation of thehead of a sphinx sticking out of the sand; fouryears of excavation followed (Daniel, 1967, 229).Happy accidents of this kind were the rule ratherthan the exception. Many sites mentioned in his-torical sources or the Bible were only identifiedbecause their names appeared on building inscrip-tions or clay tablets found during plunder for mu-seum exhibits. One example of this kind was thesite of Sippar in southern Mesopotamia (the bib-lical Sepharvaim) where Rassam excavated for theBritish Museum in 1881. Ironically, one of the cu-neiform inscriptions that he found recorded anexcavation carried out by the Babylonian kingNabonidus in the sixth century BC. Nabonidusdug beneath the foundations of a temple dedi-cated to the Sun-God Shamash to find out whohad built it, and discovered an inscription thatanswered his question (Lloyd 1980, 156).Nabonidus was evidently a rather more problem-oriented excavator than Rassam.

5.3 Evans and Knossos(McDonald & Thomas 1990)

One of the final stages in revealing the early civili-zations of Europe and the Near East took placewhen Sir Arthur Evans investigated the origins of


34

the Mycenaean civilization revealed by Schliemannin Greece. Soon after the independence of Crete in1898 Evans excavated the Minoan palace atKnossos, where a ‘literate’ civilization had devel-oped around from around 2000 BC. Evans, likeSchliemann, was testing a hypothesis suggested byprior research. He was aware that engraved seal-stones bearing a pictographic script had been foundin Crete, and that their script (now known as Lin-ear A) was independent of those of Egypt or Tur-

key. It indicated that a system of writing had beendeveloped well before the adoption of an early formof Greek by the Mycenaeans. Unlike Schliemann,Arthur Evans did not suffer opposition or ridicule;he had an impeccable academic background, andworked in the Ashmolean Museum, Oxford. Hehad even accompanied his father John on the fa-mous visit to Boucher de Perthes at Abbeville in1859 (above, p. 13); eight-year-old Arthur actuallyfound a flint implement.

Unlike Egypt, Mesopotamia, or even HomericGreece, the Minoan world was almost entirely un-known; the notion of a civilization preceding thatof Classical Greece was a real revelation. As at Troy,earlier levels were found below the palace atKnossos; they extended back into the prehistoricperiod and emphasized the depth of time that pre-ceded the literate stages of these early civilizations.Thus, archaeology alone had provided almost eve-rything that was known about Minoan civilization,and this achievement paralleled the contributionmade by prehistorians tothe understanding of hu-man antiquity. The excavations at Knossos weredirected at the solution of a specific cultural prob-

1.14 Schliemann’s excavations at Troy (Hissarlik,Turkey) were not a good model of archaeologicaltechnique. Only solid structures were noticed andrecorded, and they were rapidly demolished toreveal earlier features. Schliemann’s awareness thata succession of cities had occupied the site, and hisdetermination to find the Homeric level, did at leastforce him to take note of the occurrence of artefactsin different levels. His motivation for digging is ofparticular interest; it was the culmination of a longprogramme of literary research, fieldwork andexcavations on other sites, all aimed at identifyingthe geographical setting and physical remains of anearly Greek culture known only from literature.Schliemann 1880, facing p. 265


35

lem, using a variety of evidence, including somesmall previous excavations on the site: the resultswere spectacularly successful (fig. 1.15). Evans washelped by the fact that the Minoan palace was notoverlain by extensive remains of subsequent occu-pation. He was able to make really detailed inter-pretations because it had been destroyed by volcanicactivity, and contained the remains of most of itsartefacts and furnishings.

5.4 Beyond Europe and the Near East(Fagan 1977)

After the discovery of Minoan Crete, the onlyearly European or Near Eastern civilization toremain unknown until the twentieth century wasthat of the Hittites in Turkey. Like the Mesopo-tamian civilizations, it was known from the Bi-

ble; it was illuminated in 1906–8 by the discov-ery of large numbers of inscribed tablets at thelarge fortified city of Hattusha (now Bogazköy).Further East, fieldwork and excavation in thetwentieth century in India and China producedevidence of urban civilizations, dating back tobefore 2000 and 1000 BC respectively.

In the New World, Spanish colonists andchurchmen had reported the existence of sophis-ticated urban civilizations since the fifteenth cen-tury, but the literate civilization of the Maya thatflourished in Yucatan was first described by JohnStephens and Frederick Catherwood in the1840s. Fortunately, the objectivity and accuracyof their fieldwork set an example for work else-where in Central and South America. High-qual-ity excavation of Maya sites, organized by thePeabody Museum of Harvard University, fol-lowed in the 1890s. Further south, fieldwork andexcavation took place from the 1850s onwards,notably by Max Uhle in Peru and elsewhere. Allof this exploratory work falls within Willey andSabloff’s ‘classificatory-descriptive period(1840–1914)’ of American archaeology, and itwas of course influenced by European work bothon human antiquity and early civilizations(1980, 34–76). In contrast to the languages ofEgypt and Mesopotamia, the exciting break-through of deciphering Mayan script did nottake place until the 1960s (Coe 1992).

6 Achievements of earlyantiquarians

The purpose of this chapter has been to demon-strate that the basic principles of contemporary ar-chaeology are illuminated by studying how theydeveloped in the first place. I hope to have doneso without conflicting with Piggott’s demand toavoid both hindsight and period charm: ‘What weare left with, and surely should be our aim, is anattempt to see the early development of archaeol-ogy as far as possible in terms of its own past, withits individual exponents thinking and acting aschildren of their age (as we are of our own), andassessing what now seem for us their praisewor-thy or unfortunate contributions to the subject inrelation to their contemporary social and intellec-tual context, and their individual world-picture’(1982, 19).

1.15 A section drawing from Arthur Evans’ report onexcavations at Knossos, Crete. In an unusually carefulmanner at such an early date, Evans illustrated theprecise location of a vital piece of dating evidence (animported Egyptian lid) beneath the wall and floor ofpart of the palace. Perhaps his father, John Evans, hadmade him familiar with geological sections such as fig.1.5 above. Evans 1899–1900, p. 64


36

The discovery of the ‘lost’ civilizations (otherthan Greece and Rome), the appearance of scien-tific excavation techniques, and the increasinglysophisticated interpretation of past societies, allbelong to a phase of archaeology that had scarce-ly begun before the nineteenth century. However,the rapid developments of the nineteenth and twen-tieth centuries incorporated several of key factorsestablished during the Renaissance and the Age ofReason. Pursuits that were considered respectablein intellectual circles happened to include the study,recording and collecting of ancient sites and arte-facts, as part of a wider scientific interest in naturalhistory. The efforts of individuals, usually amateursand often eccentrics, established the methods offieldwork, and led to the opening of museums thathad to be staffed, displayed and catalogued. Oth-ers extended the existence of humans on Earth froma mere 6000 years back into an immeasurable pe-riod. As a result of all these achievements, greaterefforts were made to collect human artefacts, andto organize them in more sophisticated ways toprovide a new source for the documentation ofhuman technical and social progress.

Unfortunately, the discovery of the great civili-zations did little to advance archaeological tech-niques until conscious problem-solvinginvestigations began in the late nineteenth century.Luck, accident and careless plunder were common-place; historians and art-historians were interestedin documents and works of art, and these could begained without much attention to stratification orplanned research. These awaited the growth of asense of responsibility about antiquities, and the

systematic approach of people like Pitt Rivers orUhle. New methods had to be developed to investi-gate problems such as the prehistoric backgroundof Egypt, or the Neolithic communities accidentallyrevealed beneath the Minoan palace at Knossos andat the bottom of many Near Eastern tells. Becauseradiocarbon dating only began to provide independ-ent dates for prehistoric sites in the 1950s, theseproblems demanded the kind of study that had beendeveloped in northern Europe, involving carefulexcavation (taking account of stratification), andtypological study of pottery and other artefacts toprovide relative dating.

Co-operation between specialists in separatedisciplines was much easier in the restricted circlesof learning that existed before the twentieth century.Fortunately, the multi-disciplinary approach thatbrought such beneficial results to the study of earlyhumans in the mid-nineteenth century has grownever since, with the result that archaeology remainsone of the few subjects available in the educationalworld that forms a genuine bridge between scienceand the humanities. Two less welcome traditionswere also established early. One was the inabilityof some scholars to complete their work for publi-cation; another was the distortion of evidence tosupport fanciful theories. Although characteristic ofearly antiquaries such as Aubrey and Stukeley re-spectively, both phenomena are still very much apart of archaeology today, and they will be exam-ined further in later chapters.

Note: a guide to further reading that includes top-ics covered in this chapter begins on p. 185.

37

This chapter aims to introduce a variety of tradi-tional and modern techniques used in discover-ing and examining archaeological sites. I hope toemphasize that there are many non-destructiveways of looking at sites in their local and regionalsettings; excavation should be a last resort, sinceit involves irreversible physical intervention. It isremarkable how much can be revealed about asite using surface observations alone; however,these depend on a reasonably good state of pres-ervation of visible structural remains orearthworks. In densely populated, intensivelycultivated countries well-preserved sites are verymuch the exception. In much of western Europemany prehistoric sites must already have beenploughed flat by the end of the Roman period. Inthese circumstances, anything that can ‘see be-neath the soil’ has an important role, from broad-scanning techniques such as aerial photographyto geophysical devices used on individual sites.Twentieth-century technology cannot claim allthe credit, however, for Camden and Stukeleymade sensible observations and interpretations ofburied features or structures revealed by varia-tions in growing crops in the sixteenth and eight-eenth centuries (Daniel 1967, 37; 45).

1 The discovery of newarchaeological sites

Chapter 1 explained how antiquaries like Cyriacof Ancona and William Stukeley, or the telldiggersof the Near East, relied on straightforward visualinspection to find ancient sites. Some travelled toinvestigate unknown areas, others made system-atic attempts to increase knowledge about regionsthat had already proved productive. Limitationsof transport continued to impose severe restric-tions on fieldworkers until at least the mid-twen-tieth century. Although many technical devicesare used today, the human eye remains an ex-

tremely important and sensitive instrument.Chance discoveries of artefacts or structures dur-ing farming or building work have been particu-larly important in adding to the basic corpus ofarchaeological knowledge in the nineteenth andtwentieth centuries, and finds made in this waymay be followed up by more informed examina-tion. The pattern of discovery of artefacts andsettlements in Denmark illustrates this well (fig.2.1; Hedeager 1992, 14–21).

Discovery is pointless without recording, butobservers (ancient or modern) only record whatthey see, and what they see is determined by whatthey consider to be significant. Early antiquariesdid of course make invaluable observations aboutsites which in some cases have since disappeared,but they normally left frustratingly inadequate ac-counts. There was slow progress from terse de-scriptions to schematic illustrations, and thenfrom picturesque drawings to accurate surveys(Piggott 1979). The scientific attitudes that ac-companied the Enlightenment involved an in-creased interest in classification, which naturallyrequired more careful observation. Recordingwas revolutionized in the 1840s and 1850s by therapid development of photography (Feyler 1987).British and French expeditions carried out exten-sive photography in Syria and Egypt; when theCrimean War began in 1854, the Society of Anti-quaries of London requested the English Army toinstruct its photographer ‘to take and transmitphotographic views of any antiquities which hemay observe’ (Evans 1956, 291).

Perhaps the greatest contrast with the past is thatfieldwork today is rarely directed at a single site. Itusually forms part of a comprehensive study of anarea selected either because it is threatened withdestruction, or because it offers potential answersto questions generated by wider archaeological re-search. One of the first major fieldwork projectscarried out in central Italy began in the 1950s be-

2 Discovery, Fieldworkand Recording

Discovery, Fieldwork and Recording

38

cause redistribution of agricultural estates to small-scale farmers led to a sudden increase in ploughing.Fortunately, archaeologists from the British Schoolof Archaeology in Rome realized the implicationsof this change, and walked over large areas of farm-land recording finds and traces of sites or structures.The result was a view of the distribution of ruralsites over a long period, from later prehistory to theearly Middle Ages. Interesting variations in the lo-cation and intensity of sites in different periods re-quired careful interpretation to explain the changesin terms of settlement patterns, population andmethods of farming. Partly as a result of this earlysuccess, studies in Greece or the Roman Mediter-ranean provinces shifted from the investigation ofindividual sites to the study of regions.

Formerly, significant sites (most commonlytowns) were surveyed and excavated because theirplans, buildings and inscriptions were expected toreflect the political and military history of their area.Now, studies are more likely to involve extensiveanalysis on a regional scale, designed to elucidatethe broad agricultural, economic and social devel-opments within which individual sites operated(Greene 1986, 98–143). Prehistoric archaeology hasalso undergone a similar shift in focus away fromindividual events to the consideration of longer-termprocesses (see chapter 6). An example of a recentmulti-period Mediterranean field survey project inDalmatia (Croatia) is described in a case-study be-low (p. 54).

Fieldwork underwater is not unlike that onland, for it relies on visual scanning of areas ofthe seabed by divers. Photography, recording andsurveying practices on wreck sites and cargo scat-ters use the same general principles employed ondry land, but they are of course more time-con-suming and cumbersome. Echo-sounders andsonar scanners replace aerial photography fordetecting anomalies on the seabed, except in clearshallow water, and magnetic location devices arevery effective in surveying wreck sites (Dean1992, 128–46).

2 Fieldwalking and siterecording(Brown 1987)

Although fieldwork projects are often conducted ona very large scale today, they continue to employtechniques developed in the past for the study ofindividual sites or small areas. The simplest (andoldest) procedure is fieldwalking, which relies uponthe observation of minor fluctuations in the char-acter of the ground surface, and, where possible, therecognition of ancient artefacts lying upon it. Iffieldwalking is conducted systematically the resultscan be analysed to reveal significant patterns offinds. The area selected for examination is markedout with a regular grid for the guidance of teams ofwalkers, to ensure that the ground is inspectedevenly. Any finds, whether artefacts or surface fea-tures, must be recorded accurately in relation to thisbasic grid, and plotted on to a master-plan to give

2.1 Sources of archaeological finds in Denmarkchanged markedly in the nineteenth and twentiethcenturies. The expansion of agriculture led to manyaccidental discoveries in fields and bogs, but thesewere overtaken by finds from graves as archaeologi-cal work increased later in the nineteenth century.However, it was only after 1930 that the changingfocus of archaeological excavation made settlementsthe predominant source of new finds. It can be seenthat there was a plentiful supply of finds to nine-teenth-century antiquarians who, like C J Thomsen,were primarily interested in typology. Hedeager 1992,14–20


39

an overall distribution of the results (figs 2.2–3). Ifan area is too large to be explored in its entirety, itshould be sampled in such a way that the resultsmay be assessed statistically, and extrapolated to thewhole area with some confidence; computer pro-grams now allow this to be done with ease. Clus-ters of potsherds, flint flakes or building debris lyingon the surface may suggest likely centres of occu-pation or other human activity in the past. A lackof obvious concentrations does not mean that nooccupation existed, of course; agricultural exploi-tation or geological weathering of the land mayaffect the results in different ways, even in a smallarea.

Traces of structures, or earthworks such as

ditches or field-boundaries, should be surveyedstraight away with simple equipment, but a moredetailed follow-up may be desirable if greater ac-curacy is required for the interpretation of a com-plex site (fig. 2.2). A grid of measured spot-heightsprovides contour plans that may reveal subtle sur-face variations not immediately apparent to the eye.If the results are logged by a portable computer thereadings may be processed to enhance variationsbefore plotting the contours. Large sites are moreeasily understood when the features recorded in asurvey are drawn out at a small scale; inconsisten-cies requiring further investigation will be moreobvious. Relationships between the various com-ponents (enclosures, trackways, building plots, etc.)are more likely to be evident on a plan than on theground. The analysis and interpretation of site-plansare an important aspect of archaeology, and a com-bination of field observations and the examinationof plans may elucidate the sequence in which over-lapping earthworks were created, altered or super-seded. It must not be forgotten that Stukeley wasperfectly capable of making deductions of this kindin the eighteenth century, however (above, p. 23).

2.1 Sites and Monuments Records(Larsen 1992)

The more elaborate scientific techniques of ar-chaeological prospecting outlined below are nor-

2.2 A team from the University of Bradford recordingsurface traces of a medieval monastic settlement andvillage on the Isle of May in the Firth of Forth,Scotland. It is important that all field recording,whether of sites or artefacts found on the surface, isconducted on a carefully planned basis. The surveyingequipment used on this site included a site included aGRID-pad write-top, held by an operator in theforeground. This specialized portable computer has ascreen that can display a general plan of the area, andis connected to an EDM (electronic distance meter).The exact location of features indicated by thearchaeologist in the middle distance can be loggedaccurately and added to the site plan. Steve Dockrill


41

mally applied to individual sites to maximize theunderstanding of their nature and extent before ex-cavation. Broader methods, such as aerial photog-raphy or field survey, are more likely to beemployed in a regional research project, with theaim of collecting information about all periodsthoroughly. Ideally, a database should be createdto serve a number of different purposes, includingfurther research by individuals who may not havebeen involved in its collection. ‘Rescue’ archaeol-ogy has provided a stimulus to this kind of record-ing, and many countries now have a policy ofmaintaining Sites and Monuments Records on aregional basis. A good knowledge of the numberand distribution of known sites makes it mucheasier to assess the implications of plans for build-ing construction or other development threats. Ifplanning authorities and contractors are able toconsult these records at an early stage, they maybe able to adjust developments to avoid sites, or,at least, to take account of them so that excava-tion can be arranged well in advance without costlydisturbances to construction schedules.

Sites and Monuments Records are normallystored on computers for rapid access, and the de-tails of each site include cross-references to maps,publications and aerial photographs. They are veryuseful for research, and provide an obvious start-ing point for anyone who wishes to study sites orfinds of a particular type or period. Research of thiskind increases our understanding of regions andtheir sites, and makes it easier to draw up prioritieswhen difficult choices have to be made about theirpreservation or destruction. Thus, fieldworkershave many responsibilities besides the discovery andrecording of sites; they have an obligation to inter-pret the results in the light of the latest research, andto present their conclusions in a convenient andcomprehensible form for consultation by non-spe-cialists.

3 Aerial photography(Wilson 1982; Riley 1987)

The greatest single contribution to fieldwork andrecording has undoubtedly been made by aerialphotography. Besides giving a bird’s-eye view ofsurviving sites, aerial photography can, in favour-able circumstances, record details of buried sitesrevealed by discolorations in the overlying soil orvegetation (fig. 2.4). The visual effectiveness ofaerial photographs had been appreciated since the1850s; occasional archaeological shots were takenfrom balloons, but the First World War stimulatedthe practice of taking high-altitude reconnaissancephotographs from aeroplanes both for mapping andstrategic purposes. Many pioneers of archaeologi-cal aerial photography gained their experience inthis way, including O G S Crawford, who publisheda manual on the subject in 1929. The use of thetechnique has expanded exponentially since then,and the understanding of the conditions for its op-timum application has also increased. Cameras andfilm-types used in conventional photography arenow very versatile, and remote-sensing from satel-lites is becoming a valuable additional source ofhigh-altitude images (Shennan and Donoghue1992).

3.1 Visible sites

Aerial photography provides a useful supplementto observations made during fieldwork on visibleearthworks (figs 2.5–6). Isolated features may be-come more coherent when seen in an overall view,and new features not easily noticed on the groundmay be revealed; a common example is the relationof outlying field boundaries and trackways to afarming settlement. The best conditions are pro-vided by low light, because it emphasizes irregulari-ties by highlighting bumps and filling hollows withdeep shadow (fig. 2.5). Clear sunlight cannot beguaranteed at exactly the right time of morning orevening in temperate countries, of course, and thebest shadow effects do not last long. This form ofaerial photography is most suitable for adding de-tails to known sites, rather than for prospecting fornew sites. Frost or light snow may enhance bothshadow sites and buried sites, either because of theirreflective qualities or because of variations caused

2.3 A systematic approach to discoveries isillustrated by this standard form for recordingsites during a survey project around Leptiminus, aRoman port in Tunisia. Fieldworkers can notemost traces that they encounter simply by tickingboxes, and the information from the completedforms is ready for entry into a computer data-base. Lazreg & Mattingly 1992, f ig. 3


42

by differential thawing. Such conditions are ofcourse rare, and are more likely to be recorded byluck rather than planning.

3.2 Invisible sites(Whimster 1989)

Many of the most impressive photographs taken bypioneers like Crawford and Keiller were of knownsites, but aerial photography has been much moreimportant in providing information and discover-ies about sites that have been levelled, and are there-fore unlikely to be spotted during normal fieldwork.Even when sites are discovered by fieldwalking,after ploughing has scattered finds on the surface,their form and extent are rarely evident. Such sitesare commonest in areas of heavy agricultural ex-ploitation, where different settlement patterns and

2.4 Circumstances leading to the formation ofcharacteristics visible on aerial photographs. A: Slightsurface variations, which might not be apparent toan observer on the ground, are enhanced byhighlights and shadows produced by low sunlight inthe early morning or evening. B: Even if thesesurface features have been removed, subsequentdisturbance of the soil by ploughing may still revealvariations in colour or texture. C: Irregularities inthe depth and moisture content of the soil may leadto marked variations in the height and colour of acereal crop during its growth or ripening. Severedrought conditions exaggerate these effects and mayalso reveal much smaller features, not only in cerealcrops but also in permanent pasture like A. Notethat the post-hole and hearth at the left and centreof these diagrams would not be likely to be re-vealed, even in extreme conditions. Fig. 2.10illustrates the results of geophysical prospecting overthe same buried features. Audio Visual Centre,University of Newcastle

2.5 The ditches of Roman military earthworks atChew Green, Northumberland, are enhanced byshadows cast by low sunlight. This extensive sitelies on sloping rough moorland in the CheviotHills, and is very diff icult to understand on theground because of its complexity and size: themost distinctive earthwork in the centre of thephotograph is approximately 150 metres square.Tim Gates, National Monuments Record


43

field systems may have come and gone several timesduring the evolution of the modern landscape. Theirdetection relies on a number of phenomena that in-fluence vegetation or the soil. Crop-marks are cre-ated when buried features either enhance or reducethe growth of plants (figs 2.7–8). Since a key factoris the availability of moisture to their roots, abnor-

2.6 A new plan of Chew Green earthworks compiledby surveyors and investigators from the RoyalCommission on Ancient and Historical Monuments.There are four overlapping square or rectangularmilitary camps and forts, besides other features thatprobably result from more recent farming activities. ©Crown copyright: RCHME


44

mal conditions emphasize the height and colour ofcrops (particularly cereals) during growth and rip-ening. These effects are not consistent or easily pre-dictable, and a complex site should bephotographed over many years, under different con-ditions, to compile a cumulative record of its fea-tures. As with shadow sites, the optimum conditionsdo not last long; root crops and pasture are veryinsensitive, and only reveal marks during extremedrought conditions.

Soil-marks may be observed when land has beenploughed; human activity in the past may lead tovariations in the character and colour of the top-soil, and like crop-marks, subtle variations are bestseen from above to reveal a coherent plan. The mostdramatic sites discovered in this way are Romanvillas photographed by Roger Agache in north-east-ern France; ploughing has brought fragments oftheir chalk foundations to the surface and revealeddetailed plans of buildings, clearly visible as whitelines against dark brown soil (Greene 1986, 116–18). Unfortunately a site that shows up as a soil-mark is probably being severely eroded, and maysoon disappear altogether if regular ploughing con-tinues. Some soil-marks are merely ‘ghost sites’,

made up of soils of differing consistencies and col-ours derived from pits, ditches and other featuresthat have been destroyed by deep ploughing (Clark1990, 110–12).

Aerial photography is not to be undertakenlightly; like fieldwork conducted on the ground, itmust be well planned and systematic. Photographyshould be carried out with a good knowledge of thesites, crops and geology of an area, so that it is timedto coincide with the best conditions. Expertise isrequired in selecting the optimum cameras andfilms, for some colour film emulsions are speciallysensitive to particular colour ranges. The use ofinfra-red photography, or multispectral scannersthat record the reflective properties of the ground,may clarify the results (Shennan and Donoghue1992). Variations in colour or contrast can be en-hanced by special developing or printing processes,or by computer scanning and filtering (fig. 2.9). Sitesmust of course be mapped to be used by archaeolo-gists on the ground. Oblique views of sites requirecomplicated adjustments to be plotted on to a hori-zontal plan, and even the best cameras suffer somedistortion away from the centre of their lenses.Computer-based techniques have been designed toovercome these problems, but they are restricted tospecialized mapping centres. It is important thataerial photographs include several fixed referencepoints that can be identified on large-scale maps,to enable archaeologists to create working plans byhand, using geometry—and a considerable amountof patience.

Experience is also required not only to photo-graph meaningful things from the air, but to recog-nize and interpret the results. It is not always easyto distinguish between archaeological features andnatural geological phenomena, and a deep knowl-edge of archaeology is required for any attempt to

2.7 At La Panetteria, Foggia, Italy, a neolithicsettlement is revealed by particularly clearcropmarks that mark enclosure ditches and circularfoundation trenches for timber roundhouses. Manyfurther dark lines and patches appear within andoutside the enclosures; excavation would benecessary to determine whether these are naturalgeological effects, prehistoric features, or the resultof modern farming. Sherratt 1980, p. 147


45

classify and date sites by their form. This problemis reduced by using an objective method of describ-ing the traces recorded on photographs, and byusing a computer database system to enforce stand-ard methods of description (Edis, MacLeod andBewley 1989). It is also important to recognize ar-eas where sites are absent, and to decide whetherthey really did not exist there, or if modern condi-tions simply do not provide circumstances that willbring them to light. One of the first effects of aerialphotography in Britain and elsewhere was to revealcrop-mark and soil-mark sites on heavy valley soilspreviously considered unsuitable for cultivation; ithad been assumed that they were covered by denseforest until comparatively recent historical times.The discovery of these sites caused an upward revi-sion of estimates of early populations and theiragricultural technology. However, areas whereearthwork sites had already been recorded haveperhaps been neglected as a result; photographers

understandably concentrate on the most promisingregions. Proper research programmes are less ex-citing and more expensive than unplanned explo-ration, but their results allow much firmerconclusions to be reached about site distributions,settlement patterns and other features of ancientlandscapes.

A different problem is that land that was onceploughed may subsequently have remained underpasture for a considerable period; in Britain, manyfields used for grazing today show the tell-tale tracesof medieval ridge-and-furrow ploughed strips.Medieval ploughing will not only have destroyedany earthworks that preceded it, but today’s ma-ture pasture will prevent the detection of earlierburied sites except under extreme drought condi-tions. The use of land for modern arable or pasturedepends on factors such as the surface geology,drainage, climate and altitude. Thus, a map of ar-chaeological sites in any area is incomplete unlessit shows how the sites were discovered (shadow,

2.8 Dry conditions have revealed the remains ofthree concentric ramparts that once belonged to afortified late prehistoric settlement at Chatton,Northumberland. The crop has ripened most wherethe subsoil is shallowest, producing a visual effectexactly opposite to that observable at La Panetteria.A further set of irregular enclosures may be seen inanother part of the field. Tim Gates, National Monu-ments Record

2.9 Aerial photographs may be digitized for computeranalysis, and then processed by programs that filterout irrelevant background variations and enhancesignificant details. Oblique photographs can also beadjusted to compensate for distortions of angle, andmatched precisely to maps. This illustration shows aninexpensive PC-based imaging system in which a videocamera is examining a negative. The graphs on the leftrepresent scales of tones found in the photograph;they can be altered to produce modified images onthe other screen. Booth et al., 1992, f ig. 27.1


46

crop- or soil-mark), and records the conditionslikely to reveal visible traces.

A final comment on the value of aerial reconnais-sance was provided after an exceptional number ofsites was photographed in the dry summer of 1989in Britain. Flying funded by state organizations andprivate individuals covered much of England andWales, at a total cost well below that of a season’sexcavation on an average urban site (Griffith 1990).

4 Geophysical surveying(Clark 1990; Scollar 1990;Spoerry 1992)

Some sites never show up on aerial photographsat all, and some show different features at differ-ent times; others are revealed only in part becausethey extend into adjacent land where conditionsare unsuitable for photographs. When excavationis planned on a site of this kind, further details maybe required about areas that do not appear clearlyon photographs. It may even be difficult to estab-lish the precise location of a site known only fromoblique photographs that lack clear referencepoints that can be found on a map. In these cir-cumstances a selection of geophysical prospectingdevices is available for detecting buried features.It must be stressed that these instruments are onlysuitable for use on sites whose location is alreadyknown or suspected, for their operation is muchtoo time-consuming to be applied ‘blind’ to largeareas. As with aerial photography, their main pur-pose is to distinguish anomalies, hopefully of hu-man origin, from the natural subsoil (figs2.10–11).

Geophysical prospecting devices are useful wher-ever details need to be checked, or where trial ex-cavation would be an inefficient method of locatingburied features. For example, at Usk, Gwent, theposition of the southern defences of a Roman le-gionary fortress was thought likely to be in fieldsunder pasture, insensitive to crop-marks. It was acomparatively simple exercise to take measurementsalong a line that began within the known easternand western ramparts; two major anomalies werefound, consistent with the presence of buried ditcheslying outside the former rampart (Manning 1981,86–8). Geophysical surveying is also used within

known sites to suggest areas where excavationmight be most profitable. Anomalies suggestingoccupation debris or structures are a very helpfulguide in ‘rescue archaeology’ situations where alarge exploratory excavation would not be feasible;indeed, the first significant uses of geophysical in-struments took place in exactly these circumstances.At Dorchester, near Oxford, prehistoric ditches(known from aerial photographs) were located in1946 by using a resistivity meter, and they wereexcavated before their destruction by gravel extrac-tion. Magnetometers were first used in Northamp-tonshire in 1958 to find pottery kilns of Roman date(known from the writings of a local early nineteenth-century antiquary) before road construction (Clark1990, 12–17).

A great advance in geophysical surveying oc-curred in the late 1960s, when portable instrumentsbecame available that could supply continuousreadings, rather than readings taken with the instru-ment positioned at a series of individual points.Geophysical instruments normally require a secondoperator to record readings in relation to the sur-veying grid, but an automatic logging system isfaster (fig. 2.11). Early examples produced imme-diate visual output by means of graphs plotted onpaper, but these were superseded in the 1980s byportable computers that could compile and displaylarger site plans on their screens (Clark 1990, 19–26). Results are normally presented in the form oflinear graphs, whose peaks and troughs indicatepoints of high and low resistance that indicate theposition of buried anomalies (fig. 2.12). Alterna-tively, if a series of linear surveys has been compiledover an area, it is possible to convert the readingsinto a contour plan. The data can be ‘filtered’ bycomputer programs to eliminate natural variations,so that archaeological features are emphasized whenplotted on a plan. The availability of laser printershas enhanced the subtlety of filtered plans in a waydescribed by Clark as ‘nothing short of a revolu-tion’ (1990, 147).

It is dangerous to place too much reliance on theresults of geophysical surveys, however. At SouthCadbury Castle, Somerset (an Iron Age hillfort withadditional ‘Dark Age’ occupation) an area of theinterior was excavated because anomalies identifiedby a survey suggested the presence of a rectangularbuilding—possibly, the excavator hoped, a timberhall of early medieval date. On excavation it was


47

found to be a fortuitous combination of entirelyunrelated features of different dates. However, atimber hall was indeed found in an adjacent area,but its construction was based on post-holes toosmall to have shown up in this kind of survey(Alcock 1972, 71, fig. 8).

Two main classes of instruments are used in geo-physical prospecting. Resistivity meters detect theresistance of subsoil to the passage of an electricalcurrent, and magnetometers measure variations inthe subsoil’s magnetic characteristics. Related in-struments obtain useful information by measuringthe magnetic susceptibility of the superficial soils,rather than that of features dug into the subsoil.

4.1 Resistivity surveying

When an electric current is passed through theground between electrodes, the resistance to its flowmay be measured. A current will pass relativelyeasily through damp soil, but drier compact mate-rial such as a buried wall or a cobbled road surfacecreate higher resistance. Resistivity surveying is

rather cumbersome, because it normally requires anumber of electrode probes (usually four) to bepushed into the ground at precise intervals for eachreading. A variety of configura—tions has been triedto speed up the surveying procedure; for instance,instead of forming a line. probes have been fixed in

2.10 Diagram showing the ability of the twoprincipal methods of geophysical prospecting todetect buried features.‘+’ indicates a positive readingfrom a strong magnetic field, or high resistance tothe passage of an electrical current, and ‘-’ shows anegative response (lower magnetism or resistance)in comparison with the site’s average backgroundreading (‘0’). The size of the symbols gives a roughimpression of the relative strengths of these results.The different kinds of instruments do not react toanomalies in the same way, and neither can beexpected to detect minor features such as a post-hole. The nature of the site and the operator’sexperience will determine which method is em-ployed. The same soil profile is also used in fig. 2.4to illustrate how these features would show up onaerial photographs. Audio Visual Centre, University ofNewcastle


48

a square table-like arrangement with the instrumentmounted on top for single-handed surveying (Clark1990, 45–7, figs 35–8). An even, well-drained sub-soil with features buried at a fairly constant depthis best for resistivity; otherwise, natural disturbancesand variations may confuse the readings. In tem-perate climates, the best conditions for the moisturecontent of the soil occur in the months from July toSeptember.

Because of the laborious procedure involved,resistivity is best suited to the detection of linearfeatures such as roads, walls or ditches by takingmeasurements along a straight line at 90° to theirsuspected position. Some electromagnetic instru-ments measure soil conductivity without probes,and although less sensitive they have proved par-ticularly successful on arid sites in desert areas.

4.2 Magnetic surveying

Magnetometers also detect deviations from the gen-eral background of the subsoil, in this case indicatedby variations in its magnetic field. Several aspectsof past human occupation cause suitable anoma-lies. Heating at approx. 700° C or above by hearths,kilns, furnaces, etc., causes the randomly alignedmagnetic particles present in most soils and claysto realign along the prevailing magnetic field of theEarth, and to retain this new alignment on cooling.The alignment of magnetic particles is also affectedby digging and refilling ditches and pits; solid fea-tures, such as walls or road surfaces, contain fewermagnetic minerals and therefore provide lowerreadings than their surroundings.

Magnetic surveying does not require probes in theground; the instrument can be carried along a lineor a grid by its operator, and it is generally preferredto resistivity for this reason when conditions are fa-vourable. One form of instrument, the protonmagnetometer, takes readings of the absolute mag-netic field at given points on a grid. The protongradiometer is less sensitive, but in many ways easierto operate; it measures the difference between twoseparate detector bottles at either end of a pole heldvertically by the operator. Buried anomalies affect thelower bottle more than the higher, and the difference

2.11 Unlike aerial photography, geophysical surveyingis never used on its own as a method of discoveringsites. It is best used for checking details that are notclear from surface remains. Here, a resistivity meter isbeing used to record a traverse across a suspectedmedieval fortified farm at Easter Bradley in Northum-berland. The operator is following a precise line(marked by tapes) which can be related to a precisesite-plan by the EDM (electronic distance meter)visible in the foreground. Jim Crow


49

is recorded, rather than the absolute field. This per-mits rapid surveying of areas threatened by develop-ment, especially if a ‘bleeper’ mechanism producesan audible sound when an anomaly is detected. Evenmore rapid is the fluxgate gradiometer, which takescontinuous, rather than spot, readings (‘…the work-horse—and the race-horse—of British archaeologi-cal prospecting’ (Clark 1990, 69)). Gradiometerresults are easier to interpret than magnetometerreadings, which are more easily disturbed by natu-ral variations in the subsoil, or by the effects of wirefences, electrical storms or railways. Magnetometersare also an effective aid in surveying shipwreck siteson the seabed, where metal rivets or larger objectssuch as cannons may have been dispersed over a largearea and covered by sediments (Green 1990, 81–5).

Some additional types of magnetometers and soilconductivity meters are employed in the surveyingof archaeological sites. Unlike normalmagnetometers, they inject a signal into shallowsurface deposits and record the way the response isaltered by magnetic susceptibility. Since susceptibil-ity is influenced by human occupation and otheractivities, particularly when burning has been in-

volved, it is particularly useful when it is necessaryto determine the full extent of a site, for only a fewvery widely spaced measurements need be taken. Ona smaller scale, susceptibility is a good indicator ofintensively utilized domestic or industrial areaswithin sites; it provides a useful complement to theresults of phosphate analysis, which reflect differ-ent kinds of activities such as waste-disposal oranimal husbandry (Clark 1990, 107–9).

4.3 Metal detectors

Metal detectors are not only popular with mem-bers of the public who regard their use as an in-nocent hobby, but also with professionaltreasure-hunters who plunder sites for profit. Thefine dividing line between these types of users ac-counts for the bad press metal detectors have re-ceived from archaeologists. Most types penetratethe soil only to a very limited extent, but they havebeen used by archaeologists to locate dispersedmetal artefacts—for example, a hoard of Romancoins scattered by ploughing. A more sophisti-cated device (the pulse induction meter) gives awarning of metal objects in graves that are aboutto be excavated. This is not normally necessaryif the site has already been surveyed bymagnetometer, for when these instruments en-counter iron objects they produce readings thatare distinguishable from archaeological features.Some archaeologists and museum curators workin partnership with metal detector enthusiasts tomake full records of any artefacts that they dis-

2.12 Magnetic survey of a site at GroundwellFarm, Wiltshire. As at La Panetteria (f ig. 2.7)the site is a prehistoric settlement consisting ofbanks and ditches surrounding timber housesconstructed in circular foundation trenches. Theexcavated central portion of the site confirmsthat the survey has been successful in detectingditches. Each line represents a continuousreading along a measured traverse, and is ineffect a graph of magnetism. Because thereadings are higher over ditches, the visual effectmakes them look more like raised banks; theirpositions are very clear, however. The surveytechnique would not have been suff icientlysensitive to detect the complexities of theoverlapping foundation trenches of houses in theinterior, however. Clark 1990, f ig. 99; Gingell 1982


50

cover, and to provide expert assistance whenimportant antiquities are found. At Snettisham(Norfolk) several hoards of gold and silver coinsand ornaments dating from the first century BChad been found by accident during ploughing,and by metal detector users. After the most re-cent find, archaeologists from the British Museumremoved the topsoil mechanically, and locatedfurther undisturbed hoards buried at a depth thatwas beyond the range of metal detectors used onthe original surface. The topsoil that had beenremoved was also thoroughly checked both by eyeand by metal detector. The law of treasure troveensured that the original finder received the mar-ket value of his discovery, and this provided animportant incentive to report the finds immedi-ately. Thus, the thrill of discovery and financialreward were combined with full archaeologicaldetails about the nature of the site and its buriedmetal objects (Stead 1991).

4.4 Radar and sonar location devices

A recent development with considerable poten-tial for the examination of buried sites in futureis the use of radar, but a disadvantage is that itperforms best on very dry deposits. It works inthe same way as sonar scanning, but electronic(rather than sonic) signals are transmitted intothe soil, and bounce back into a receiver. Thesignals are altered by the density and positionof whatever they encounter, and the patternsreceived from the ground are plotted as a dia-gram. When colour is used to enhance the vari-ations, it is possible to see the shapes not onlyof solid features, such as buried walls, but alsothe profiles of pits or ditches. Radar images havebeen used in England at the Anglo-Saxon burialground at Sutton Hoo to look for signs of pos-sible ship-burials below mounds and at York toexamine sites before excavation (Stove andAddyman 1989). Sonar scanning is a routinetechnique used in sea-bed surveys and it is ableto detect archaeological anomalies such as ship-wrecks as well as natural rocks and sand banks.The Mary Rose, a Tudor warship subsequentlyexcavated and lifted for display in a museum atPortsmouth, was first located with the help ofthis method by a nautical archaeologist whoknew, from documentary sources, the approxi-

mate position where it sank in 1545 (Green1990, 50–1).

4.5 Soil analysis(Courty 1989)

Scientific location instruments have not replaced alltraditional pre-excavation techniques for examin-ing sites. For example, if the ground is struck witha mallet it should produce a light resonance over aburied wall or thin topsoil, but a dull thud over ahumus-filled ditch. Probing or augering are alsouseful for testing the depth of soil, or to removesamples to gain some idea of buried stratification.The latter is dangerous on a small complex site, fora regular series of bore-holes could easily damageslight traces of structures or fragile artefacts.Augering is more commonly used to provide soilsamples for pollen analysis, or to measure variationsin their phosphate content to detect areas of a siteused for habitation or related activities (Bethell &Maté 1989). An early medieval site at Vallhagar, onthe island of Gotland in the Baltic, was thoroughlysampled for phosphate analysis. The site consistedof scattered buildings and enclosures representinga farming community; concentrations of phosphatesderived from animal urine and dung were foundnext to some buildings, as well as in open areas thatmay have been used for milking and byres.

4.6 Dowsing

An unpredictable prospecting technique is dows-ing, traditionally employed to locate undergroundwater sources. Exponents use a Y-shaped twig (orwhatever else they find suitable) and carry itabove the ground, noting places where it twistsdownwards. Apparent successes in finding ar-chaeological features, such as the plans of demol-ished Anglo-Saxon churches, are difficult toexplain scientifically, for they rely heavily on theintuition of the dowser (Bailey 1988). I have wit-nessed the survey of a buried rock-cut ditch thatenclosed a Romano-British farmstead in SouthDevon; the farmer had observed it as a crop-mark,and located its position by dowsing with a forkedhazel twig, cut unceremoniously from the near-est hedge. A small trial excavation confirmed itsposition and recovered dating evidence (Greeneand Greene 1970).


51

5 Archaeology and thelandscape(Wagstaff 1987)

Landscape archaeology places humans into abroad context where they are seen to have been ina state of continuous interaction with the environ-ment; sometimes their actions helped to shape thephysical form of the natural world, normally theirlives were shaped by it to a large extent. The studyof the environment had already come to form animportant part of Stone Age studies by the 1870s,for Edouard Lartet had divided the upperPalaeolithic into phases, such as the ‘cave bearperiod’ or the ‘woolly mammoth and rhinocerosperiod’, according to the dominant species repre-sented by animal bones found with human arte-facts in France (Trigger 1989, 95). By the 1920s,studies of plant remains and microscopic pollenhad revealed a very elaborate sequence of climaticphases since the last Ice Age, marked by changingpatterns of vegetation (below, p. 145). Aerial pho-tographs contributed to a stronger appreciation ofthe essential integration of finds, sites and theirnatural setting, and the potential significance ofthese relationships.

Environmental approaches also formed part ofthe ‘new’ archaeology of the 1960s, when at-tempts to understand ancient societies were basedupon an analysis of their place in an ecological‘system’. The position of settlements in relationto each other, and to their agricultural and mate-rial resources, was an important research goal inprocessual archaeology (below, p. 170). It mightinvolve spatial archaeology, which adopted math-ematical methods and statistical techniques de-veloped by geographers to elucidate modernsettlement patterns. Mathematical approaches areuseful for examining many kinds of information,from broad regional studies down to the positionsof individual artefacts recovered from the surfaceof a single field. One basic element is to determinewhether scatters of sites or artefacts contain sig-nificant clusters or a regularly spaced pattern, orwhether their disposition is purely random. Whenthe distributions of more than one kind of arte-fact or settlement have been recorded, compari-sons may be made to establish connections(correlations) between them. The results are likely

to be expressed in levels of statistical probabil-ity, rather than a simple ‘yes’ or ‘no’; reliabilitywill depend upon the quality of the informationbeing used and the statistical techniques em-ployed. Two early exponents of spatial analysisin archaeology presented these problems in a verypositive light:

It is important that most of the techniques…demand good data…it is to be hoped that archae-ologists will be stimulated by the possibilitiesoffered by the techniques to collect in the futuremore data of high standard. (Hodder & Orton,1976, 238).

5.1 Landscape archaeology(Aston 1985; Rackham 1987)

The mathematical techniques associated with spa-tial archaeology are most commonly used in thecontext of prehistoric archaeology. Much moreinformation survives from recent historical peri-ods, not only in the form of evidence in the field,but also in documentary archives (Hoskins 1988).Traditions of field archaeology that stretch backto Aubrey and Stukeley were revived andstrengthened in the twentieth century by a grow-ing interest in local history, in particular its so-cial and economic dimensions. It has incorporatedresearch into all kinds of documents, includingcharters recording the extent and ownership ofestates, tax registers, parish registers of births anddeaths, and analyses of place-names. This kindof research leads not just to studies of the physi-cal remains of sites and buildings mentioned indocuments, but to the idea that archaeologicalobservations about the landscape and settlementsmay be used to fill in gaps in the documentaryrecord. Thus, landscape archaeology now re-quires a thoroughly integrated approach, involv-ing both written and material evidence. Onemanual summarized its objectives with particu-lar eloquence:

The landscape is a palimpsest on to which eachgeneration inscribes its own impressions andremoves some of the marks of earlier genera-tions. Constructions of one age are oftenoverlain, modified or erased by the work of an-other. The present patchwork nature of settle-


52

ment and patterns of agriculture has evolvedas a result of thousands of years of human en-deavour, producing a landscape which pos-sesses not only a beauty associated with longand slow development, but an inexhaustiblestore of information about many kinds of hu-man activities in the past. The landscape ar-chaeologist needs to develop an eye and afeeling for patterns in town and country and,even more important, to recognize anomaliesin, for instance, the large isolated medievalchurch of the deserted medieval village; thestraight stretch of stream channelled by monksin the thirteenth century; the regular eighteenthcentury Parliamentary enclosure hedge lyingacross medieval ridge and furrow; the lumpyground next to the church, marking the site ofan old settlement; and even a fine Jacobeanbuilding in an otherwise apparently poor area,indicating a former prosperity linked to a longforgotten trade or industry. Ideally it shouldbe possible to look at any feature in the land-scape, know why it is there in that form, andunderstand its relation to other features (Aston& Rowley 1974, 14–15).

This process may sound rather intuitive, but it isof course based on informed fieldwork and docu-mentary research. It is also an essential accom-paniment to the interpretation of aerialphotographs; changes in land-use in the past mayhave determined whether surface features of an-cient sites survived, or whether they will onlyappear as crop-marks. The term retrogressiveanalysis is sometimes employed to describe howlandscape archaeologists work back from mod-ern features to the fragmentary remains of ear-lier landscapes. In a fascinating study of a hillfarm in Derbyshire, the medieval and modernfield systems associated with Roystone Grangewere subjected to aerial photography, detailedsurveying and selective excavation. Features ofknown date were gradually eliminated, and bymeans of identifying changes in the design ofenclosure walls, the outlines of Roman and pre-historic settlements were revealed:

The typology of walls permits us to strip off—layer by layer—the succession of changes madeto Roystone over the past six millennia. By

plotting the gates, tracks, stiles and sheep-creeps too we can begin to build up a pictureof how the modern landscape has been shaped.Walls were employed in different ways at dif-ferent stages in the history of Roystone. Tounderstand these stages we have investigatedthe archaeology of the succession of farmingcommunities themselves. Their farms, yards,pens and even their rubbish enable us to putflesh upon the skeletal picture formed by map-ping the walls. With this mixture of informa-tion we can begin to reconstruct the history ofthe farm and its place in the making of theWhite Peak. (Hodges 1991, 43)

Landscape archaeology is also inseparably con-nected with environmental archaeology, notablyin the context of the soils and surface deposits thathave influenced agriculture and the exploitationof other resources for food-production or craftsand industries (Simmons 1989). If a ‘systemstheory’ approach is adopted, it will be assumedthat, given a free choice, a settlement is likely tobe located wherever its inhabitants have optimumaccess to all the necessary parts of their economicsystem; this concept gave rise to site catchmentanalysis. A line is drawn around a site to deter-mine the potential resources that lie within a day’swalking distance; an arbitrary circle is less in-formative than a ‘territory’ constructed to takeaccount of the actual terrain, however. Althoughrather mechanical, this technique does at leastallow a clear perception of a site’s potential, andit also allows finds from an excavation to be re-lated to its resources. A standardized unit, suchas an arbitrary 10 km circle, allows statisticalcomparisons to be made between the territoriesof different sites. In a ‘classic’ application of thistechnique at San José Mogoté, Mexico, KentFlannery looked first at finds recovered from anexcavated site dated to around 1000 BC, and thenconsidered how large a catchment area wouldhave been required to supply them. He deter-mined that all essential agricultural needs couldbe obtained within 2.5 km, and all but the mostexotic materials obtained within 5 km. When thedistribution of other local sites was studied, it wasfound that a 2.5 km circle could be drawn aroundeach one to define a ‘territory’, without overlap-ping (Flannery 1976).


53

Catchment analysis is open to the same criti-cisms that are levelled at systems theory andprocessual archaeology (below, p. 170). It assumesa purely physical interaction between people andtheir environment, in which every detail of theirlife was ruled by the location of material resources.Settlements and the landscape are perceived in verydifferent ways in many non-European cultures,where aspects may be related to social or spiritualfactors rather than mere subsistence. A visit toWessex today emphasizes this alternative view, forthe landscape is still dominated by survivingNeolithic and Bronze Age ritual monuments suchas Stonehenge, Avebury and Silbury Hill, whilehundreds of burial mounds are dotted along thecrests of nearby hills and ridges (Richards 1990).It is very unlikely that people who invested timeand energy in the construction of conspicuousmonuments to death and ritual would make a cleardistinction between the world of deities and an-cestors and that of their everyday lives. It wouldbe dangerous to analyse the economy of a farm-ing settlement of this period solely by examiningthe potential of soils within its hypothetical terri-tory through catchment analysis. A settlementmight have been constructed in a particular loca-tion according to spiritual factors, and its economycould have involved gifts and exchanges of foodand raw materials between relatives scattered overa wider region.

Anthropological studies and local history bothunderline the view that human settlement is a verycomplex subject with an indisputable social com-ponent (Arensberg 1968). The growth of large-scale field survey projects in the 1970s and 1980swent some way towards integrating fieldwork,environmental science, and the social and eco-nomic development of regions over extendedperiods of time. These projects also focused at-tention on some very basic aspects of archaeol-ogy, notably the concept of a ‘site’. Fewarchaeologists would find any difficulty in defin-ing an artefact found on the ground, or a featuresuch as a stone wall or a burial mound. But whatexactly is a site, what are its limits, and howshould it be recognized from artefacts or featuressurviving on the surface? Ethnoarchaeologicalstudies of hunter-gatherer societies suggest thatindividual ‘sites’ are meaningless if they are notviewed in terms of the shifting patterns of activi-

ties that make up the overall manner of subsist-ence (Smith 1992, 11–26). The term ‘site’ is anartificial concept invented in the present, with nomeaning in the past; it is only with the growth oftwentieth-century fieldwork that functionalwords like ‘monument’, ‘camp’, ‘village’ or ‘fort’began to be replaced by the objective term ‘site’(Dunnell 1992, 22). Some purists would like toabandon it:

In the last analysis, site, as an archaeologicalconcept, has no role to play in the discipline. Itsuses are not warranted by its properties, it ob-scures crucial theoretical and methodological de-ficiencies, and it imparts a serious andunredeemable systematic error in recovery andmanagement programs. In spite of the technicalproblems its abandonment will cause, the con-cept of archaeological sites should be discarded,(ibid., 36–7)

A more practical solution for archaeologists in-volved in survey projects is to continue to use theterm as a descriptive label for a place where a par-ticular concentration of artefacts and/or featuresoccurs, while remembering the many distortingfactors of human and natural origin that deter-mine whether a ‘site’ survives in a form that maybe recognized today. ‘Most of the time we haveone or a very few sites that must guide our ap-preciation for the types of systems that once ex-isted. If we are to use site information we mustlearn to see a past system from a site perspective.’(Binford 1992, 56)

6 A Mediterranean fieldsurvey project(Barker & Lloyd 1991)

It was stressed at the beginning of this chapter thatfieldwork today is rarely directed at individual sites.It is normally part of a regional study, devoted toanswering questions generated by wider archaeo-logical research. The most successful projects havebeen aimed at the analysis of long-term changes insettlement patterns, seen in the perspective of envi-ronmental factors that influenced, or were affectedby, human exploitation.


54

2.13 Location map of the Neothermal DalmatiaProject’s survey area. It consisted of a survey blockand linear transects across a peninsular on theAdriatic coast of Croatia, near the modern city ofZadar. Chapman & Shiel 1993, f ig. 4

6.1 The Neothermal Dalmatia Project(Chapman & Shiel 1991, 1993)

The Neothermal Dalmatia Project was designed toinvestigate the development of the landscape andhuman settlement over the 12,000 years since theend of the last Ice Age. It was centred upon an areanear the modern city of Zadar in Dalmatia, a prov-ince of Croatia (fig. 2.13). Dalmatia is character-ized by an indented coast and limestone mountains,with little in the way of level plains. The survey wasconducted by John Chapman and Robert Shiel (anarchaeologist and a soil scientist) from the Univer-sity of Newcastle upon Tyne, and Sime Batovic fromthe University of Zadar. John Chapman had writ-ten a doctoral thesis on neolithic agriculture andsettlement in Yugoslavia, and Sime Batovic had al-ready made a detailed study of prehistoric sites inDalmatia. The aim of the survey was to conduct anintegrated programme to investigate the origins offood production, and the relationships betweenagriculture, society and environmental changes.Robert Shiel’s role in studying the agricultural po-tential of soils was fundamental to the project. Inaddition, it was hoped that the emergence of socialhierarchies in later prehistory could be investigated,

along with the relationships between the nativepopulation and the Roman empire.

6.2 Research design and methods

Zadar lies on the coast of a short peninsula ap-proximately 35 km long and 25 km wide. Thesurvey selected three strips of land (‘transects’), 1km wide, running across the grain of the geologi-cal structure of the peninsula with the purpose ofsampling different types of soil from the edge ofthe sea to the highest ground. Field-walkers weredeployed in a uniform manner to cover the groundevenly, and when archaeological sites were encoun-tered, samples of finds were collected from unitsof a standard size. Physical traces of sites, such ashillforts, buildings, ditches, or boundary walls,were surveyed, and trial excavations were carriedout on some examples to recover dating evidenceand other finds that would help to explain theirfunction.

Besides fieldwalking for archaeological finds,soils were recorded in detail, particularly wheredeep sequences could be studied in sections cut bywells or streams. Soils reveal the history of erosionthat normally follows the clearance of forest forfarming. On hilly ground exposed to extreme peri-ods of heat and rainfall (typical of the Mediterra-nean region), soil erosion is dramatic, buryinglow-lying sites and removing others from hillsides.This distorts the pattern of discovery, and has to betaken into account when the distribution of sites isanalysed (Greene 1986, 85–6; 140). Other factorsthat might distort patterns of finds (known as ‘trans-forms’) were also noted. For example, there was alack of uniformity in finds of flint, because of localvariations in its availability, while well-made pot-tery of the Roman period survived much better thanits less durable prehistoric and medieval counter-parts.

6.3 Interpretation

The benefits of a carefully designed sampling strat-egy for sites, dating evidence and soil types areapparent when the results are interpreted in humanterms. A fundamental factor is that sites were dis-covered and finds collected on a uniform basis, sothat valid comparisons could be made not onlybetween the three linear transects, but also within


55

a larger rectangular area (5 km x 11 km) around aprehistoric hillfort that eventually developed intoan important Roman town. Furthermore, sites ofall periods could be related to the agricultural po-tential of the soil type on which they were located.Fig. 2.14 summarizes the results of the Dalmatiansurvey very succinctly. It shows at a glance that, forexample, in the Roman period, settlement of thelimestone uplands (Kras) with the lowest agricul-tural potential was as intense as neolithic settlementof fertile low-lying arable land.

No trace of occupation by late Stone Age(mesolithic) hunter-gatherers was encountered,possibly because coastal sites disappeared when sealevels rose after the Ice Age, while inland sites wereburied by later soil erosion. The first settled farm-ers of the Neolithic period lived in small sites dis-persed over the highest quality soils. In the BronzeAge, settlement was wider (but still dispersed) andincluded monuments such as stone cairns andboundaries situated on stonier ground. In the IronAge all available types of soil were used for the firsttime, and large defended centres appeared amongstthe scattered settlements. This pattern was intensi-fied in the Roman period, with evidence for higherpopulation and the use of some low-potential soilsfor specialized crops such as fruit and nuts. Medi-eval sites were fewer, but they still occupied a widerange of soils; many survive to the present, but theirchronological evolution is difficult to study becausevery uniform pottery has been used from the earlymedieval period to the twentieth century.

6.4 Analysis

It is one thing to collect and quantify survey data,but quite another to interpret it in human terms,when the only information available is whateverhappened to show up on the surface during a sea-son of fieldwork. The finds will also have been af-fected by numerous ‘transforms’ over the years, ofcourse. However, archaeologists only make progressby testing ideas against data, for without the de-mands made by the process of interpretation therewould be no reason to improve methods of datacollection.

One theme that emerged from Chapman’s dis-cussion was the increase in investment of capital (interms of labour) required by the intensification ofagriculture. This phenomenon was observed from

the Bronze Age onwards in the form of stone-walledenclosures, boundaries and cairn-fields, coincidingwith defensive structures built to protect these in-vestments. Intensification reached a peak in theRoman period, when the participation of Dalma-tia in an empire resulted in a move from local sub-sistence to agriculture on an ‘industrial’ scale.Chapman proposed that ‘…there are strong theo-retical grounds for supposing that intensification isstrongly correlated with increases in social hierar-chies’ (1987, 142). In a different publication, he usedthe survey data to test rival theories about the na-tive population’s adaptation to life under Romanrule: ‘The strategy which we shall follow is a de-ductive test of two hypotheses which make almostdiametrically opposed assumptions about theseprocesses’ (Chapman & Shiel 1991, 64). He con-cluded that one view (substantial immigration ofsettlers from Italy) was true in large towns, but thatthere was a considerable degree of continuity in thecountryside and smaller towns.

The final publication of the Dalmatian surveywill contain many more hypotheses and tests, par-

2.14 Settlement densities in the Dalmatian survey areaare summarized by this chart. The larger the filledsquare, the greater the number of finds per squarekilometre. Clearly arable land was popular throughoutthe history of the area, and exploitation of the lessfavourable land expanded through time to reach apeak in Roman times. Medieval finds are moredifficult to recognize, but Iska pottery shows that allfive categories of land were utilized to some extent.Chapman et al., 1987 f ig. 10


56

ticularly where the interaction of agriculture andthe environment is concerned. I hope that this shortsummary of an individual project makes it clearthat data collected in a systematic manner may beused for many purposes by scholars interested indifferent themes or periods. Furthermore, it is verymuch more efficient (and less destructive) to col-lect survey data from a large area than to excavatea single site. Hard economics as well as changingresearch objectives led to the rapid growth of fieldsurvey programmes in the 1980s (Greene 1986,98–101).

7 GIS(Allen 1990)

The extraordinary pace of development in compu-ter technology has transformed the potential of fieldsurvey and spatial archaeology by borrowing againfrom techniques employed in the study of geogra-phy. Computers that combine large data-storagecapacity with fast mathematical processors and ahigh quality of graphic display, typified in the 1990sby Sun workstations, are very suitable for runningprograms known as Geographical Information Sys-

tems. In the Upper Tisza Project, data collected froma river valley in Hungary during a fieldwork projectdirected by John Chapman has been processed andanalysed by Mark Gillings (fig. 2.15). It involvedan immensely time-consuming preliminary stage inwhich large-scale maps of the survey area were digi-tized; the contours, natural features, geology andsoil types were converted into mathematical form.All details recovered during archaeological field-work (notably site categories, locations and sizes,and the find-spot, type and date of artefacts) werealso digitized.

One example of the potential of GIS was pro-duced by calculating the effects on the river systemof the thawing of mountain snows each spring. By

2.15 Greyscale image of part of the Upper TiszaProject’s survey area in Hungary. The contours havebeen digitized from maps, and processed by a GISprogram to give an oblique view with artificialshading; dark meandering lines show rivers. Whenintegrated into maps like this, details of sites or theresults of fieldwork give much deeper insights into thelandscape than can be gained from two-dimensionalmaps. The results are normally displayed in contrast-ing colours that intensify the effects. Mark Gillings &John Chapman


57

fixing a specific depth of water, the map of the land-scape could be transformed on the computer’sscreen by flooding the river and its tributaries upto a particular contour. Sites of a specific periodcould also be added to test ideas about their rela-tionship to this phenomenon. Early farming com-munities in the Neolithic period were found to havebeen located near the water’s edge or on temporaryislands formed by the floods, with the implicationthat fishing may have provided an important sup-plement to the farmers’ diet in spring, before cropswere ready for harvesting. This idea is supportedby studies of fish bones from sites where suitablesamples have been recovered.

This exercise could have been performed manu-ally by making tracings from contour maps, shad-ing in the flooded areas, and adding neolithic sitesrecorded in files of survey data. The advantage ofa GIS approach is that questions can be asked re-peatedly in slightly different ways. What if thefloods were deeper? Do burial sites of the sameperiod have the same relation to water as the set-tlements? Do early Bronze Age sites share the samedistribution? The answers can be displayed veryrapidly, and results that look significant are meas-urable in statistical terms. Once maps and datahave been entered into the system, they can berecalled in any combination or permutation. Someexercises could only be performed with the helpof a computer; for example, if the investigatorselects a specific site (such as a farmstead), it ispossible to instruct the computer to display a viewof the landscape from that location as it wouldhave appeared to someone standing there, look-ing in any specified direction. By adding furtherinformation, perhaps by assuming that certaintypes of soil were still covered by mature treeswhile others had been cleared for fields, the viewcould be modified accordingly. Could a neolithicfarmer see other farmsteads, and were ‘ritual’

monuments and burials positioned so that theycould be seen from settlements?

Thus, GIS provides an important new dimensionto field archaeology that is very widely applicableto complex problems of location and distribution.As with other forms of computer-based data han-dling, an initial investment of time and energy inrecording data in an appropriate form is repaid bythe flexibility that it allows at the analytical stage.Similarly, publication is facilitated by printing outselected graphic displays rather than drawing themby hand. In the future, the actual data could bepublished on disks to make it accessible to otherresearchers, along with moving graphics to presenta dynamic view of the director’s own interpretationof developments over time, rather than traditionalstatic printed maps of separate periods or phases.

8 Conclusions

Archaeological fieldworkers now require manyskills, some of them traditional and subjective, oth-ers based on new scientific techniques. They alsoneed to be experts in the use of documentary evi-dence and aerial photographs, besides understand-ing geology and geography. These skills haveimportant implications for the process of excava-tion, which is examined in the next chapter. Ourability to investigate ancient landscapes and envi-ronments, without resorting to the destructive proc-ess of digging into sites, means that no excavationwork should ever be carried out until a programmeof fieldwork and documentary research has beencompleted. It is impossible to ask valid questionsabout an individual site without understanding itsplace in the historical and natural environment.


58

Archaeological evidence for past human ac-tivity is an irreplaceable part of our commonheritage. Activities aimed at recovering thisinformation can destroy all traces of the evi-dence unless it is preserved by adequate re-cording. Those of us who are involved inarchaeology, whether as amateurs or profes-sionals, have a duty to the rest of society totreat this heritage responsibly. Investigationsshould be conducted in such a way that reli-able information is acquired, and the resultsthen made available for others to study. Theseobjectives should apply to all archaeologicalwork, whether on land or underwater. (Dean1992, 300)

Of all the archaeological activities described in thisbook, excavation has the highest profile in publicperception, to the extent that anyone could be for-given for gaining the incorrect impression that ex-cavation is the primary activity of mostarchaeologists. This chapter aims not only to ex-plain how excavation is conducted, but also toemphasize the background work that precedes it,and the time-consuming processing of site recordsand finds for publication that should follow. I hopethat it will also explain the thinking that lies behindthe ethical standpoint outlined above in the Nauti-cal Archaeological Society’s principles of conduct.

Along with fieldwork, excavation is the mostimportant source of new information in archaeol-ogy. If reliable data are to be recovered, excava-tion techniques must be sound. Although majormodern excavations involve complex technicalequipment under the control of virtuoso directors,the basic principles are comparatively simple andhave changed little since their importance was firstrecognized. The process of excavation incorporatestwo approaches that are frequently in conflict: theexposure of vertical sequences of layers, and thedefinition of horizontal plans of occupation levels

or individual structures. In both cases full recordsmust be kept of the contexts in which all artefactsand other finds were discovered. The most impor-tant development of the twentieth century has beenan improved understanding of the vertical aspect,and the design of excavation techniques to revealit with the utmost care. Recent large excavationshave gone further, and have attempted to maximizethe horizontal aspect while still recording the ver-tical sequence in necessary detail (Barker 1993;1986).

1 The development ofexcavation techniques

Whatever their motives, very few excavations be-fore the late nineteenth century were much betterthan treasure hunts. A few caves with prehistoricoccupation, such as Brixham in Devon, were ex-plored by removing the filling in layers to confirmthat tools really were associated with bones of ex-tinct animals (Grayson 1983). Most excavatorssought to recover objects of commercial or aes-thetic value by random digging on known sites, orby systematically plundering monuments such asburial mounds. Luck, accident and careless plun-der were commonplace; historians and art-histo-rians were interested in documents and works ofart, and these could be gained without much at-tention to stratification or planned research. Othersites were uncovered in a more academic mannerto reveal structures as well as finds. Frequently, thetwo objectives were combined; many ancient sitesin Egypt and the Near East were extensivelycleared, and the best sculptures and objects re-moved and transported to museums in Europe.Impressive foundations or ruins were left behind,but much important evidence about the sites hadbeen shovelled away for ever; this phase has been

3 Excavation

Excavation

59

described aptly as ‘digging before excavation’(Maisels 1993, 30). Finely engraved architecturalplans and elevations are no substitute for the in-formation that might have been gleaned fromunspectacular layers of accumulated soil and de-bris, and the modest artefacts they once contained.The careful work carried out at Pompeii from 1860by Fiorelli was very much an exception (Daniel1981, 85–7). Most familiar views of ancient Greekor Egyptian sites frequented by tourists today arethe product not of abandonment and natural de-cay, but of eager clearance by archaeologists in thenineteenth century.

The discovery of civilizations did little to ad-vance archaeological techniques until the latenineteenth century, when excavators such asSchliemann and other German archaeologists be-gan to ask more sophisticated historical questionsabout classical sites around the eastern Mediter-ranean (Trigger 1989, 196–7). At about the sametime advances in excavation and the study of findsalso began to emerge from work on prehistoryand ethnography, but only after a depressinglylong period of careless barrow-digging that se-verely damaged the majority of prehistoric burialmounds in Europe (Piggott 1989, 153–9). Indi-viduals such as Pitt Rivers and Flinders Petriedeveloped a sense of responsibility about antiq-uities, and began to formulate more sophisticatedquestions about sites. Many problems could onlybe studied by archaeological methods (includingexcavation), for example the prehistoric back-ground of Egyptian and Mesopotamian civiliza-tions or the origins of the prehistoric farmingcommunities accidentally revealed beneath manyNear Eastern tells. Without historical texts orradiocarbon dating, these problems required theuse of techniques developed in northern Europe,notably excavation conducted with reference tostratification, combined with classification andtypological study of pottery and other artefactsto provide relative dates.

Two positive products of the depth of tell depos-its were that awareness of stratigraphy increased,and sequences of various kinds of artefacts fromsuccessive levels were recognized. One of the firstand most famous tell excavations was bySchliemann at Troy (Hissarlik, in Turkey). Manyof his assistants went on to apply high standardson classical sites in Greece, while Koldewey and

Andrae, who investigated Babylon and Ashur inMesopotamia, came from the same background(Daniel 1981, 122–3). These excavators were wellaware of the horizontal and vertical aspects of theirwork. They were interested both in exploring largeareas of buildings on individual levels of tells, andin excavating complete stratigraphical sequencesfrom the top to the bottom. Glyn Daniel summa-rized the context of their work:

In Egypt every monument was built of the stoneor cut in the solid rock, and the arid climatepermitted the preservation in a remarkablemanner of objects…which would have been de-stroyed elsewhere. It was, therefore, in Meso-potamia that the classical techniques werereshaped and that new techniques ofstratigraphical excavation, and of the excava-tion of perished and semi-perished materials,were developed. The architecture of Mesopo-tamia is executed in sun-dried bricks; the tech-niques of tracing these were quite unknown toearlier excavators. Koldewey and Andrae firstsuccessfully traced the walls of sun-dried brick,and this work reached its highest technicalachievements in the work of Delougaz, atKhafajah, where every single brick was articu-lated, the chips being blown away by com-pressed air. The excavations at Ur were anoteworthy model of the whole modern tech-nique of archaeology, from extraction and pres-ervation to interpretation and publication.(Daniel 1975, 290–1)

One of the most prominent excavators of Ur in the1920s was Sir Leonard Woolley, who made his firstfumbling beginnings in archaeology at Corbridgein Northumberland in 1906, while Koldewey andAndrae were already at work in Mesopotamia (be-low, p. 70).

Arthur Evans’ excavations at Knossos (above, p.33) were directed at the solution of a specific cul-tural problem, using a variety of evidence, includ-ing some small previous excavations on the site; hisresults were spectacularly successful. He was helpedby the fact that, unlike many sites, the Minoanpalace was not overlain by extensive remains ofsubsequent periods of occupation. He was able tomake really detailed interpretations because it hadbeen destroyed by fire (possibly caused by an earth-

Excavation

60

quake or volcanic activity), and still contained theremains of most of its artefacts and furnishings. Incontrast to most Near-Eastern excavators, Evansmade a conscious effort to preserve the crumblinggypsum masonry of the palace at Knossos while theexcavation proceeded. His earliest photographsshow a meticulously cleaned site, and the text dem-onstrates close attention to the stratigraphic posi-tions of finds (see fig. 1.15), both as dating evidenceand as a means of interpreting the destruction of

the palace (Evans, 1899–1900). In Britain, Pitt Riv-ers was paying similar attention to detail in the clos-ing years of the nineteenth century (below, p. 61–2),and his work must have been familiar to Evans.However, these high standards were far from uni-versal, and were certainly not reached in the exca-vations at Corbridge between 1906 and 1914.Unlike Pitt Rivers, Evans did not publish full exca-vation reports; fortunately, his detailed commentar-ies and the notebooks kept by his assistant, DonaldMackenzie, have allowed more recent archaeolo-gists to review the evidence in detail. For example,it was possible in the 1960s for Popham to carryout a review of pottery from deposits associatedwith the destruction of the Minoan palace, thanksto careful labelling during the original excavations(Popham 1970).

1.1 The destruction of evidence(Mytum & Waugh 1987)

It is instructive to consider the circumstances of oneof England’s richest discoveries to realize just howmuch information has been lost on other sites (fig.3.1). The Anglo-Saxon barrow cemetery at SuttonHoo in Suffolk was excavated in an exploratorymanner in 1938 and 1939, and an undisturbedburial was found in one mound in 1939 (Evans1986, 19). A deposit of extraordinary richness wasexcavated, including armour, weapons, silver table-ware and other items from an Anglo-Saxon royalhousehold. Years of laboratory work in the BritishMuseum have been devoted to the conservation andreconstruction of corroded iron and bronze itemssuch as a magnificently decorated iron helmet. His-torical documents, together with dates derived fromgold coins found in a purse in the burial, suggestthat it was the grave of King Redwald, who died inc. AD 625.

The burial deposit had been placed in a cham-ber in a 24-metre ship that had been hauled froma nearby river, placed in a deep trench and cov-ered by a large mound. Robbers dug a hole intothe centre of the barrow in the sixteenth or earlyseventeenth century, but, fortunately, earlier ero-sion of the mound meant that the burial cham-ber no longer lay directly below its centre. Sincethe hole was not taken below the surroundingground level, the excavators missed all signs ofthe ship as well. If this early ‘excavation’ had been

3.1 This Anglo-Saxon ship burial (Mound 2) at SuttonHoo, Suffolk, had been disturbed by nineteenth-century barrow-diggers, who had cut steps down thefar side of the burial chamber. The process had beenrepeated in 1938 by an archaeologist, but fullexamination had to wait for the campaign of excava-tions mounted in the 1980s. Many significant detailswere deduced by fragments of artefacts left behind bythe earlier excavators, but most information aboutthe burial itself had been irretrievably destroyed.However, most of the structural details of the mound,which had covered an upturned ship placed over awooden chamber, could still be discerned. © SuttonHoo Research Trust; photograph by Nigel Macbeth

Excavation

61

successful, very little would now be known aboutthe find. Some objects would have been melteddown or sold off immediately; others might havesurvived as fragments, but the majority wouldhave been discarded immediately as worthlesscorroded metal. We would know nothing aboutthe ship itself, which would have survived onlyas discoloured sand and iron nails, or the layoutof the grave goods in the burial chamber. Thedate, identity and cultural connections of the deadking would remain unknown, together with thelight that the boat burial has shed on Beowulf,an important early Anglo-Saxon poem.

Although these hypothetical effects of early ex-cavations at Sutton Hoo represent an extreme case,the analogy is applicable to most of the ‘great’ ex-cavations or discoveries made before the twentiethcentury. Even in comparatively well protected coun-tries like Britain, the proliferation of metal detec-tors has revived the plundering of sites. The problemis much greater in areas of the world where sitesregularly produce items desired by museums andcollectors. What suffers from uncontrolled treasurehunting is the quality of information: coins, preciousobjects, or carvings have little meaning when di-vorced from their stratigraphic level or structuralcontext.

1.2 The concept of stratification(Harris 1989)

John Frere’s observations published in 1800(above, p. 11) clearly embodied the common sensenotion that in a series of layers, those at the bot-tom will be older than those at the top. Geologistshad already used the same principle to arrangefossils from different strata into developmentalsequences, and it was well established among stu-dents of human origins, such as Boucher de Perthesand McEnery, by the early nineteenth century.

Unfortunately, other archaeologists whose studieswere focused on later periods paid little or no at-tention to stratification. A modern excavationpeels away these layers in reverse order, andrecords the finds and structural evidence that theycontain.

Good examples of stratigraphic observations onexcavations took place at least as early as the sev-enteenth century in Scandinavia (Trigger 1989,49), and J J Worsaae wrote with lucidity and in-telligence on the requirements of controlled exca-vation in the 1840s (Daniel 1967, 103–6). Therecovery of closed groups of artefacts from strati-fied contexts had been an important element inThomsen’s development of the Three-Age Systemsome years earlier, but these tended to be self-de-fining contexts such as graves, rather than indi-vidual layers found in a sequence (Graslund 1987;above, p. 26–7). However, the majority of nine-teenth-century excavators observed stratificationpassively, rather than using it actively to guide theirexcavation strategy.

1.3 Pitt Rivers (Bowden 1991)

Modern excavation practice does not have a singlefounder, but the kind of progress made in the latenineteenth century is exemplified by Pitt Rivers(1827–1900; fig. 3.2). His wide range of activities

3.2 General Pitt Rivers (1827–1900), portrayed byFrank Holl R.A. in 1882, shortly after inheriting alarge estate in Hampshire and Dorset containing manyarchaeological sites. The prehistoric shield and apickaxe relate to his interest in ancient weapons,which led him to conduct excavations, while theobservant stance and notebook draw attention to hisrole in recording and protecting ancient monuments.Pitt Rivers Museum, University of Oxford

62

provides fascinating insights into the cultural atmos-phere of late Victorian England.He would be remembered today even if he had notturned to excavation late in his life, for he had al-ready played an important part in the developmentof the typological study of artefacts, besides con-ducting fleldwork and surveys in England and Ire-land. Through an unlikely course of inheritance, PittRivers came into possession of a large tract ofCranborne Chase in Wessex, one of the richest ar-chaeological areas in England. He devoted the restof his life (1880–1900) to the study of this area,while indulging in many other interests of an edu-cational and ‘improving’ nature for the benefit ofhis estate workers and their families.

The wealth derived from his estates allowed PittRivers to employ full-time archaeological assistants,and to use his own labourers to carry out excava-tions on a variety of monuments from a neolithicbarrow to a standing medieval building (fig. 3.3).Pitt Rivers had conducted his first excavations inIreland in the 1860s, and from the very beginninghe devoted detailed attention to artefacts as well asto sites and their structures. The importance he at-tached to all finds, however trivial, led him to recordnot just the artefacts themselves but also the con-texts in which they were found. Pitt Rivers’ exca-vation technique was far from revolutionary,although it was neat, methodical and accurately sur-veyed, in accordance with his military training. Themost remarkable novelty was that his excavations

and finds were published in great detail in a seriesof massive volumes, containing copious illustrationsand ‘relic tables’ that are still informative today.

1.4 Developments in thetwentieth century(Barker 1993)

The requirements of ‘scientific’ excavation werefinally met when Pitt Rivers’ approach to record-ing and publication was combined with a clearperception of the significance of stratification, and

3.3 Pitt Rivers’ excavation of Wor Barrow, aneolithic burial mound in Dorset, in 1893. Despitethe quality of the recording, the actual excavationtechnique was crude (Bowden 1991, 131–4). Thelabourers simply moved forward against a verticalface on horizontal levels; sloping layers of barrowmaterial are clearly visible on the sections, but noattempt was made to remove them individually.Worse still, the excavation was taken down intonatural chalk well below the old ground surface,whose buried soil shows as a dark layer at wheel-barrow height in the left section. Fortunately, PittRivers eventually realised his mistake when the largefoundation trench of a timber mortuary structurewas found underneath the barrow, but many minorfeatures must have been dug away unnoticed. Ascale model of the barrow before excavation can beseen propped up at the front of excavation. Salisburyand South Wiltshire Museum

Excavation

63

the ability to recognize and excavate it layer by layer.Contemporaries of Pitt Rivers, such as Schliemannor Petrie, who were working in Turkey, Egypt andPalestine, possessed a more advanced understand-ing of stratigraphy and the use of artefacts for dat-ing the observed layers.

It is difficult to overestimate the contributionmade to archaeological method in the last quar-ter of the nineteenth century by Schliemann, PittRivers and Petrie. It would be no exaggerationto say that, with the experience of the Danes andthe Swiss behind them, they forged the essentialtechnique of archaeology. (Daniel 1975, 177)

Thus, three indispensable elements of excavationhad emerged by the beginning of the twentieth cen-tury. Horizontal observations had improved con-siderably, and were combined with accuraterecording, notably on German excavations of clas-sical cities in Turkey (Trigger 1989, 196–7). Verti-cal sequences were increasingly important,particularly on deeply stratified tell sites in the NearEast such as Ashur, excavated by Andrae (Daniel1981, 123). Systematic attention to all classes offinds was the newest and most important element,

and it led to a growing number of authoritativepublications of sites and catalogues of artefacts thatallowed archaeologists to make critical evaluationsof the work of other excavators.

1.5 Mortimer Wheeler(Hawkes 1982)

All three elements were combined in work conductedduring the 1920s and 1930s by Mortimer Wheeler,a man who (like Pitt Rivers) developed an outlookand methods that reflected his military background;unlike Pitt Rivers, he trained many younger archae-ologists who went on to have distinguished careersin archaeology. During a long career interrupted bytwo world wars Wheeler conducted excavations inBritain, France and India on sites selected for the

3.4 Sir Mortimer Wheeler (1890–1976) and TessaWheeler during their excavations at Verulamium (StAlbans) in the 1930s. Mortimer Wheeler firstentered professional archaeology in 1913, and hisperception of sites and the landscape was soonincreased by artillery experience during the FirstWorld War. He remained a vigorous and colourfulfigure active in archaeology until his death.Verulamium Museum

Excavation

64

investigation of specific research questions. For ex-ample, he explored the relationship between the pre-Roman hill-forts of southern England and northernFrance by excavating the defensive structures of sev-eral sites and comparing the styles of pottery foundon them. In India, he used imported Roman coinsand pottery (that could be dated by reference to sitesin Europe) as a fixed point in a sequence of phases inIndian sites and finds. Wheeler also organized andpublished museum collections, wrote books and ar-ticles for the general public, and, in his later life,became a remarkably successful radio and televisionpersonality. His first wife, Tessa, played an under-rated part in the development of Wheeler’s methods,and also instigated the creation of the Institute ofArchaeology at the University of London. She met atragically early death in 1936 (fig. 3.4; Hawkes 1982,122–43).

Wheeler is chiefly remembered for perfecting the‘box system’ of excavation, whereby a site wasuncovered by means of a grid of square trenches,with baulks left standing between them as a perma-nent record of the stratification of all four sides ofeach trench (fig. 3.5). Wheeler did not always ex-cavate in this fashion, however, and his work at theRoman city at Verulamium (St Albans) or the IronAge hill-fort at Maiden Castle (Dorset) involveduncovering large open areas in a manner similar tomodern practice (Sharpies 1991). His techniquesemphasize the conflicting requirements of lateraland vertical excavation. Stratification is essential forunderstanding the sequence of a site, but an arbi-trary grid of vertical baulks imposed on a site maymask important horizontal features (figs 3.6–8).Wheeler was of course aware of this basic conflict,and summarized the optimum compromise afterdiscussing it at length (1954, 126–9):

With the proviso, then, that all horizontal dig-ging must proceed from clear and comprehensi-ble vertical sections, the question of priority isfundamentally not in doubt. Careful horizontaldigging can alone, in the long run, give us thefull information that we ideally want (ibid., 129).

This compromise cannot always be reached eventoday, especially in urban rescue excavations con-strained by limitations on time and access (see fig.3.20). Wheeler was undoubtedly well aware of thecomparable problems posed by tells in the Near East,

where it had become standard practice to sink deepshafts (‘sondages’) to sample their successive occu-pations and artefacts, but where horizontal excava-tion of the lower levels was physically impossible.Wheeler’s approach to stratigraphy (developed fur-ther by his student Kathleen Kenyon) included closeattention to layers and features such as pits or walls,and the interfaces between these stratigraphic units.He continually stressed the need to excavate theseunits, rather than to follow the common practice ofremoving arbitrary horizontal levels (‘spits’). BecauseWheeler’s whole approach was combined with sys-tematic numbering of layers and the finds that theycontained, Harris has compared its significance withthat of the ideas about strata and fossils introducedinto geology by Hutton and Smith in the late eight-eenth century (1989, 11).

1.6 From keyholes to areas(Barker 1993)

An opposite extreme to the open-area approach wasthe procedure of digging many very small widelyspaced trenches on large shallow sites. A more spe-cialized form of this approach (sometimes describedas ‘keyhole’ excavation) was used extensively inRoman military archaeology (for example, atCorbridge: below, p. 72). Ian Richmond producedoverall plans of many Roman forts and fortressesin northern Britain by the judicious excavation ofsmall narrow trenches, carefully placed to checkcritical details of the fairly predictable layout of theirinternal structures. Richmond was also an astuteobserver of stratification; his pioneering drawingsof occupation layers and their associated dated coinswere included in a report on Wheeler’s excavationsin the Roman fort at Caernarvon in Wales publishedin 1923 (fig. 3.9).

The interwar decades saw the development ofopen-area excavations on sites lying on flat alluvialland in southern Scandinavia, northern Germanyand the Netherlands, a region that contains manyrural sites, ranging from neolithic farming settle-ments through to villages of the early medieval pe-riod (fig. 3.10). Because their buildings wereconstructed entirely from wood, there was littlebuild-up of construction and demolition layers ofthe kind found on sites where stone or mud brickshave been used. Many sites were located on subsoilsthat facilitated the recognition of pits, post-holes

Excavation

65

and other structural remains of buildings. UnlikeRoman forts, the plans of these settlements couldnot be predicted, and complete uncovering was es-sential to the understanding of their form and de-velopment (fig. 3.11). A German archaeologist,Gerhard Bersu, was invited to England in 1938 tocarry out an area excavation of this kind on an IronAge farmstead at Little Woodbury, Wiltshire (Evans1989). He demonstrated that pre-Roman Britonslived in substantial round timber houses, not, aspreviously thought, in minute pit dwellings. Theplans of their houses could only be recovered bystripping large areas, and by analysing the plans ofpost-holes and other features that had been care-fully excavated and recorded (Bersu 1940). The sitewas excavated in strips rather than as a continuousopen area, but the contrast with ‘keyhole’ trenchesis nevertheless dramatic.

3.5–6 Box trenches and open area excavation in TheTofts, part of an Iron Age fortification at Stanwick,Yorkshire, excavated by Wheeler in 1951–2. Fig. 3.5shows a typical grid of unexcavated baulks thatprovided four permanent sections in each squaretrench; wooden pegs at their intersections arereference points for surveying and plotting importantfinds. In fig. 3.6, the removal of the baulks hasrevealed the foundation of a circular timber building.The shallow ploughsoil did not justify the numeroussections visible in fig. 3.5, and the baulks concealedimportant features. Wheeler 1954a, pl. VIII, facing p. 8

3.7–8 As part of a re-examination of the Stanwickfortifications since 1983, the area known as The Toftshas been investigated further by Colin Haselgrove.Wheeler’s area excavation is marked A on fig. 3.8,which shows that he sampled only a very small partof the interior. A geophysical survey of most of the

rest of the site demonstrated that it had been denselyoccupied, and placed Wheeler’s excavation, andlarger open-area investigation by Haselgrove (see B),into context. Haselgrove 1990, f ig. 11

Excavation

66

3.9 Drawing of a section through the filling of thestrongroom (sacellum) beneath the headquartersbuilding of Segontium Roman fort, Caernarvon,Wales. Because the phases of occupation of thismilitary site were assumed to reflect the history ofRoman Wales, particular emphasis was placed upondating evidence in this drawing. Thus, for example,the layer of rubble cannot have fallen there before c.AD 290, for a hoard of that date was found in the soil(‘black mould’) beneath it. Wheeler 1923, 55, f ig. 17

3.10 The Iron Age settlement at Nørre Fjand inDenmark, explored by Hatt from 1938–40, is anearly example of open area excavation. The fainttraces of timber buildings, fragmentary stoneworkand superimposed floor levels could only have beenunderstood by uncovering large areas, and examin-ing each feature and layer. Hatt 1957, 75, f ig. 50;National Museum, Copenhagen

3.11 The neolithic settlement at Elsloo in theNetherlands illustrates a common problem on open-area sites where no occupation levels survive toprovide stratigraphical relationships betweenseparate structures. The only clues to the sequenceare given by chance intersections between featuresin the few places where building plans overlap. It isalso difficult to tell whether the site was crowdedwith buildings for a comparatively short time, orwhether a few buildings existed at any one time,which were gradually replaced in slightly differentpositions over a long period. It may be helpful toexamine the distribution of datable artefacts, such aspottery, to see whether they cluster in separateareas; this approach has proved fruitful at Mucking,in Essex (Hamerow 1993). Modderman 1975, f ig. 86

Excavation

67

The history of the development of modern ex-cavation techniques is completed by examining theextension of the concept of open-area examinationto sites that possess a significant depth of accumu-lated layers. It should be repeated that good exca-vation should aim to satisfy the demands of boththe vertical and horizontal aspects of a site. Evenon sites excavated in the past by a grid system, orin the case of mounds excavated in quadrants, thefinal stages usually consisted of removing thebaulks. Open-area excavations rarely use baulks atany stage. On sites of limited depth, this causes noproblems, for the disturbed ploughsoil can bestripped off to reveal various features for individualstudy. This process can be extended to sites withdeep stratigraphy if the extent, contours, depth andconsistency of each layer are carefully recordedbefore it is removed. If this is done properly, a sec-tion of the sequence of deposits could be drawn onany given line from the site records alone. Alterna-tively, lines are fixed before excavation at selectedpositions where a cross-section of each layer willbe drawn, so that a cumulative drawing of the se-quence is built up in stages as the layers are removed(Barker 1993, 113–17). These techniques both suf-fer from an unavoidable drawback, for unlike sec-tions preserved in a baulk at the sides of a trench,they are never actually seen by the excavator all atonce, and they can only ever be checked on paper.However, this is a legitimate sacrifice when it is veryimportant to obtain the maximum exposure of eachhorizontal level on a site.

1.7 The interpretation of stratification(Harris 1993)

Stratification can be defined, therefore, as anynumber of relatable deposits of archaeologi-cal strata (from a stake-hole to the floor of acathedral) which are the result of ‘successiveoperations of either nature or mankind’.Stratigraphy, on the other hand ‘is the studyof archaeological strata…with a view to ar-ranging them in a chronological sequence’.(Barker 1993, 21, quoting Harris)

All excavation is based on the fundamental princi-ple of the succession of levels, which assumes thatlayers of soil (or any other material) were depos-ited in chronological order, with the oldest at the

bottom. Today, even the simplest archaeologicalsituation is still approached from the same point ofview, whether it is a long sequence of layers recorded

3.12 Dating by means of stratif ication: theexcavation of a post-hole. Left: a post stands in ahole with packing stones added for stability. Bychance, the rim of a pot datable to c. AD 1300was included when the hole was back-filled with amixture of topsoil and subsoil. A coin of AD 1520found amongst the packing stones could not havegot there offer the post was erected and the holeref illed. Thus, the date of the potsherd is irrel-evant, for it is ‘residual’ from earlier occupationof the site, but the coin provides a terminus postquem showing that the post was erected after1520. Right: a layer has accumulated after thedemolition or decay of the structure, f illing thecavity left by the post, and a new topsoil hasformed. The coin of AD 1600 could only havereached its position offer the decay or removal ofthe post, but 1600 is only a terminus post quem forthe formation of this layer. Thus, the life of thetimber structure began after 1520, and endedbefore the new layer was formed, sometime after1600. However, one or both of the coins mighthave been residual or old when lost. Greaterprecision could only be achieved by consideringother evidence from the site. Without the coin ofAD 1 520, the residual potsherd would have givena t.p.q. at least two centuries too early. If theprecise f indspot of the second coin in the laterlayer of the post-hole had not been recorded, afalse t.p.q. of AD 1600 would have been given tothe structure. Audio Visual Centre, University ofNewcastle

Excavation

68

in section, or a direct relationship between two in-tersecting features in an area excavation. The analy-sis of dating by means of stratification involves theclumsy but apposite phrases terminus post quemand terminus ante quem (fig. 3.12). The interpreta-tion of sections is not strictly objective, and it is animportant reminder that excavation is not a science.Excavators should prepare detailed reports andmake them available to other archaeologists as soonas possible, so that they can decide whether theconclusions are based on reasonable interpretationsof the evidence.

The first stage in writing an excavation reportinvolves the integration of layers and structures withdating evidence recovered from them, whether it ispottery, coins, radiocarbon samples or anything else

of chronological significance. It is of course essen-tial that all excavated material was carefully re-corded in relation to the deposits where it wasfound. The Harris Matrix is a useful developmentthat has brought considerable benefits to the studyof stratigraphy; it was devised in the 1970s as aresponse to the over-whelming complexity of deeplystratified urban excavations (Harris 1989, 34–9; seefig. 3.13). Harris emphasizes the fundamental va-lidity of Wheeler’s approach to stratigraphy, butdemands a more systematic approach. His term‘unit of stratification’, rather than ‘stratum’ or‘layer’, conveys the importance of interfaces be-tween layers. Analysis of the stratification of com-plex excavations is simplified by the reduction ofsections to diagrams, and the recognition of theseinterfaces.

Thus, although the sides of a silted-up ditchhave no existence other than as the boundary, orinterface, between its filling and the materialthrough which the ditch was dug, they neverthe-less represent a human action that should be as-signed to its correct place in the sequence of layers.In the case of working surfaces, such as floors, itis not always fully appreciated that the interfacesrepresent an important period of human activitythat may have lasted for a very long time, whereasa visible layer that overlies it may have been de-posited very quickly. Harris stressed that the onlyway to make sense of a site is to study the wholeextent of each interface to understand the activitythat took place:

No amount of sectional drawing is of the slight-est help in such a composition of these periodplans, as it becomes clear that the horizontalrecord of stratification is far more importantthan the vertical…. What is needed is not agrain-by-grain plan, but a record of scaleddrawings with each stratigraphic unit on a sepa-rate sheet, and which shows, at the very least,the area of the stratum and spot heights of itssurface—as recorded prior to excavation. Withsuch an archive any desired configuration ofstratigraphic units can be made at any time….(Harris 1977, 94)

Harris Matrices have not replaced traditional sec-tion drawings in archaeological reports; site di-rectors will still wish to illustrate relationships

3.13 The recording of a stratigraphic sequence usingthe Harris Matrix. The simple section (top left) andthe exploded three-dimensional diagram (right) showthe same stratigraphic units and the order in whichthey were formed. It should be noted that (2) and(6) are interfaces, rather than actual layers. Bottomleft is a Harris Matrix showing all stratigraphicalrelationships; in the centre, these superfluous orduplicated lines have been removed to give a clearsummary of the original section. This method ofsummarizing stratigraphy can be carried out during anexcavation, and is useful in published excavationreports. Audio Visual Centre, University of Newcastle,after Harris 1989, f ig. 12

Excavation

69

that are crucial to dating or interpretation. Sec-tion drawings remain a convenient method ofsummarizing individual features, for example thecomposition of silted-up ditches or rubbish pits;colour photographs rarely bring out sufficientdetail, and they are expensive to publish. A draw-ing of a section is by definition an interpretation,and it can be expressed in a variety of differentstyles, from baldly diagrammatic to naturalistic(Harris 1989, 69–81). Section drawings can sup-ply an impression of the actual appearance of astratigraphic sequence that cannot be included ina matrix. Some excavators go much further, andattempt reconstruction drawings to explain howstructures have decayed to result in a particularsection. This valuable intellectual discipline cangenerate useful ideas that improve the quality ofinterpretation.

2 An example of thedevelopment of excavationtechniques: Corbridge,

Northumberland

The large Roman site commonly referred to asCorstopitum lies just west of Corbridge, a smallmedieval town in the Tyne Valley, 24 miles (38km) from the mouth of the river. It lay at the junc-tion of two important Roman roads: Dere Streetled from York into Scotland, while Stanegate con-tinued west to the Solway, and marked the north-ern frontier of the Roman Empire for aroundthirty years before Hadrian’s Wall was built 4 kmto the north. The history of exploration atCorstopitum illustrates the changing aims andtechniques of archaeology, from treasure huntingto investigation, from research to rescue, andfrom excavation to publication and display. Asearly as the seventh century AD, Corstopitum wasused as a convenient source of stone for localAnglo-Saxon churches, which were the first ma-jor stone buildings to be constructed in the areasince the end of the Roman period. This kind ofquarrying continued for more than 1000 yearsbefore any serious archaeological excavationbegan. The site was probably still impressivewhen it attracted the attention of King John in1201; the terse contemporary record was elabo-

rated in an eighteenth-century edition ofCamden’s Britannia:

No inconsiderable remains of antiquity, however,are to be found here, among which King Johndug for treasure supposed to be buried by theantients; but fortune mocked his vain pursuit, ashe had formerly done Nero when searching forDido’s treasure at Carthage. He found nothingbut stone with marks of brass, iron, and lead.Whoever views the adjoining heap of ruins calledColecester will pronounce it a station of Romansoldiers. (Gough 1789, III, 235)

The name Corstopitum came from an inaccuratemedieval copy of a Roman document, and is cer-tainly incorrect; modern interpretations vary, but aplausible view is that the place was called Coria bythe local native Celtic population, and that it wasadapted to Corioritum (‘hosting-place by a ford’)to describe the Roman military base at this impor-tant river crossing (Hind 1980).

2.1 Antiquarian observations

Many travellers commented on the visible ruins wellinto the eighteenth century, but besides stone robbing,the expansion of agriculture was already having se-vere effects. A notable local antiquarian, JohnHorsley (1684–1732), gathered information in thefirst quarter of the eighteenth century, and observedin his Britannia Romana that ‘It is now almost intirelylevelled… Pieces of Roman bricks and pots were lyingeverywhere on the surface of the ground in tillage,when I was on the spot’ (1732, 397). Stukeley com-mented on both stone robbing and agriculture in IterBoreale, an account of a tour made in 1729: ‘Theytell us with some sort of wonder, that it is the richestand best thereabouts for ploughing: they discern notthat it is owing to the animal salts left in a place thathad been long inhabited. Corbridge is built out of itsruins, which are scattered about there in every house’(1776, 63). A large area of the site was cleared foragricultural improvement in the first decade of thenineteenth century.

2.2 Excavation

Archaeological excavations were mounted in 1861:‘…a labourer had been placed by Mr. Cuthbert of

Excavation

70

Beaufront at the service of Mr. Coulson,…for thepurpose of making investigations at Corbridge.’Dere Street and the Roman bridge across the Tynewere examined along with some internal structures(Bruce, 1865, 18–19). Woolley did not benefit fromthis earlier work when he began a major investiga-tion in 1906: ‘the results obtained were apparentlyof great interest. Unfortunately the plans, reports,and drawings made by Mr. Coulson have entirelydisappeared, and the only record of his work is aninadequate résumé given at second hand…’ (1907,162). The excavations of 1906–14 were stimulatednot only by past accounts of buildings and richfinds, but also by ignorance about the real signifi-cance of the site. The best insight is provided not inthe excavation reports, but in an autobiographypublished in 1953 by the first director, LeonardWoolley (1880–1960):

My first experience of digging was at Corbridgein Northumberland, and I know only too wellthat the work there would have scandalized,and rightly scandalized, any British archaeolo-gist of today. It was however typical of whatwas done forty-five years ago, when field ar-chaeology was, comparatively speaking, in itsinfancy and few diggers in this country thoughtit necessary to follow the example of the greatpioneer, Pitt Rivers. The NorthumberlandCounty History was being written and the writ-ers wanted to know more about the Romanstation at Corbridge, so proposed a small scaledig to settle the character of the site. The com-mittee naturally appealed to ProfessorHaverfield as the leading authority on RomanBritain, and he, as he had intended to take aholiday on the Roman Wall, agreed to super-vise the excavations. Somebody, of course, hadto be put in charge of the work and because Iwas an Assistant Keeper in the AshmoleanMuseum my qualifications were, in the eyes ofan Oxford professor, ipso facto satisfactory.Haverfield arranged with Sir Arthur Evans, theKeeper, that I should go to Corbridge. In pointof fact I had never so much as seen an excava-tion, I had never studied archaeological meth-ods even from books (there were none at thetime dealing with the subject), and I had not anyidea of how to make a survey or a ground plan;apart from being used to handling antiquities

in a museum, and that only for a few months,I had no qualifications at all. I was very anx-ious to learn, and it was a disappointment tome that Haverfield only looked in at the exca-vation one day in the week and then was con-cerned only to know what had been found—Idon’t think that he ever criticized or correctedanything. (Woolley 1953, 14–15)

The excavations of 1906–14 explored enormousareas of the site (fig. 3.14), and resulted in a seriesof prompt annual excavation reports that totalled676 pages and were as detailed as the general stand-ards of their day expected. In addition, Volume Xof The Northumberland County History (1914)contained a 48-page synthesis of the results, writ-ten by Prof. Haverfield:

It seems plain that we have here something thatwas neither an ordinary fortress nor an ordinarytown. The castramentation of the one and thestreet-planning of the other are alike, wanting.We may rather guess that Corstopitum was nowa store-base for armies operating further northand possibly even for the eastern garrisons ofHadrian’s Wall, with a half-military, half-civilpopulation which would gather round such abase (op. cit. 479). As the [fourth] century ad-vanced, Corstopitum declined. It survived…onlyas an ill-built town of hucksters and mechanics,whose mean shops intruded on the nobler ruinsof the Antonine Age; a town of vanishing wealth,on the extreme border of a sinking empire (op.cit. 11).

Thus coins, pottery and dated inscriptions indicatedoccupation throughout the Roman period, from thelate first century AD until the late fourth or earlyfifth century. Most structural information relatedto stone buildings of second-century date; whatpreceded or succeeded them was largely unknown,because excavation had consisted of stripping soiluntil stone structures were found, and either uncov-ering them completely or simply following the walls(fig. 3.15). Some digging was carried out beneathfoundations, but earlier structures were rarely re-vealed, even though ‘disturbed ground’ containingrecognizably earlier Roman finds was frequentlyencountered. The excavation was certainly superiorto treasure hunting, for the horizontal aspect of the

3.14 A large area was excavated at Corbridge beforethe First World War. Good plans were obtained ofmajor stone structures in the centre of the site (thegranaries and ‘forum’, labelled X, VII, and XI, andwalled compounds south of Stanegate). To the northand west, many buildings had been at least partiallyconstructed in timber, or had been disturbed bymodern agriculture. The positions of shops openingonto Stanegate were indicated by stone gutters andby floors with gravel surfaces or stone paving. Earliertimber forts that lay beneath the stone structureswere not discovered until excavations began again inthe 1930s. Craster 1914, facing p. 481

Excavation

72

site—plans of buildings in particular—was beingexplored with the clear purpose of finding out moreabout the site. Attention had still not been devotedto the examination of vertical sections or sequencesby the time that the Great War ended excavation in1914.

2.3 Excavations between the wars

Durham University Excavation Committee beganwork at Corbridge in 1934, under the joint direc-tion of Ian Richmond and Eric Birley, because HisMajesty’s Office of Works was in the process ofpreparing the central portion of the remains foundin 1906–14 for public display. Birley had beenasked to write an official guide to the site, and, asa result of his own recent excavations on Hadri-an’s Wall, he had hypothesized four firmly datedperiods of construction, occupation and destruc-tion for the frontier. Stratigraphic excavation had

become familiar by the later 1930s (Richmond hadworked with Wheeler in the 1920s), and artefactstudies, notably the dating of pottery, had im-proved considerably. The new excavations weretherefore carried out with strict reference to indi-vidual layers and deposits, which could be datedby associated artefacts, and by the examination ofstratigraphic relationships between structures (fig.3.16). Furthermore, the directors had a distinctworking hypothesis: ‘It seemed axiomatic that thefour main Wall periods should be reflected in thestructural history of Corstopitum too’ (Birley1959, 3). Part of their work consisted ofresurveying and analysing the stone structures thatwere to be conserved and displayed to the public,but this was combined with the excavation of newtrenches to examine the relationships between thestructures and to assign them to their appropriatehistorical period.

Richmond and Birley also wanted to explorebelow the stone buildings to investigate the layersthat had been recognized, but not understood, since1909. They knew how to recognize remains of tim-ber structures, and were well informed about thenature and planning of Roman forts. Thus, a late-first-century fort gateway, together with its ditchsystem and parts of internal buildings, was identi-fied by means of small trenches. Further layers wererecognized as belonging to later forts constructedbetween this first phase and the overlying stonestructures. The 1936–8 excavation report speakswith an air of confidence entirely lacking from the1906–14 series:

…it is impossible…to correlate with history allthe periods of occupation so far discovered…The deep sections, however, by relating to thesubsoil and to one another buildings of whichthe planning has long been known, mark a newdeparture in our understanding of the site. It isnot possible both to describe and to date thework of later periods; while the fact that theproblems concerned with early periods can bedefined, may be regarded as the first step towardstheir solution. (Birley & Richmond 1938, 260)

After a second break caused by world war, exca-vations took place every year from 1947 to 1973on a comparatively small scale, heavily con-strained by the consolidated remains of the ma-

3.15 This illustration shows excavation techniques atCorbridge in 1909. The columns of porticoesbelonging to two stone granaries (X and VII on fig.3.14) had been partly buried by rises in the level ofthe Roman main street (Stanegate); its stone guttercan be seen resting against them. The excavators haveremoved the vital deposits between the road to thegranaries. These stratified layers might have revealedimportant details about the date and nature ofchanges to the buildings and their entrances. How-ever, everything except solid masonry was simply dugout, destroying evidence that later excavators wouldhave liked to re-examine. Knowles & Forster 1910

Excavation

73

jor stone structures in the central area. The sitewas used for training students as well as for re-search; new details were added, and old ideaswere confirmed, modified or rejected. Four suc-cessive superimposed forts (two with additionalphases of internal modification) were graduallydisentangled by means of the detailedstratigraphic observation of intercutting features.In addition, a round timber house was discovered,belonging to a native farmstead that had existedbefore the first Roman fort was built. The phasesof military occupation were correlated with dat-ing evidence drawn from historical events, datedbuilding inscriptions, coins and stratified pottery,while comparisons with other forts suggested thesize and nature of the army units that occupiedeach phase.

The stone structures on the site, and in par-ticular their histories after the second centuryAD, were more difficult to assess than the ear-lier forts because crucial overlying strata hadbeen removed either by earlier excavators orduring clearance in advance of public display.Debate has raged, complicated by the fact thatit has had to rely on observations recorded inthe early reports from the 1906–14 excavationsthat cannot be checked, for the evidence has

gone for ever. The area surrounding the centralmilitary and official core lies outside the limitsof the remains open to public display underfarmland, and it has not been touched by archae-ologists since 1914. However, regular plough-ing has clearly done further damage to theuppermost levels, for the cultivated areas arenow up to a metre lower than those under pas-ture (Bishop & Dore 1989, 12). It is known thatopen-fronted shop buildings flanked a mainstreet leading west, but we know nothing abouttheir construction or subsequent history. Many

3.16 Sections drawn by Ian Richmond in trenches dugat Corbridge in the 1930s to explore relationshipsbetween structures discovered before the First WorldWar. The excavators related these buildings to roadlevels that they had assigned to specific historicalperiods (Antonine, Severan, etc.). Several layersbelonging to timber forts of the late first and secondcenturies AD can be seen beneath the lowest roadlevel, together with traces of timber structures cutinto the subsoil. With hindsight, the interpretation ofthe site in terms of a rigid historical sequence seemsover-simplified and arbitrary, but it was a considerableadvance over the treatment of the same road a fewmetres away illustrated in fig. 3.15. Birley & Richmond1938, facing p. 254

Excavation

74

were only recorded when traces of stone gutters,hearths or gravel levelling and floors were found;the actual buildings were probably constructedmainly of timber. Other structures may well havebeen missed completely, for the detection of tim-ber buildings without stone drains or floors layoutside the technical competence of the early ex-cavators. To the north, the plan is even less sat-isfactory. Fragments of buildings, minor roads,ditches and furnaces indicate intense activity inwhat was probably an industrial area, overly-ing the northern parts of the early timber forts.A few stratigraphic sequences were recorded onthe plans in places where major features over-lapped each other, and, occasionally, structureswere dated by associated finds. Little else can besaid about the overall development and declineof the whole Roman town of Corstopitum with-out further fieldwork and excavation.

2.4 Area excavations

The application of new excavation techniques atCorbridge did not end in the 1960s; it was broughtup to date by two large open-area excavations inthe 1970s and 1980s. The first of these, caused byroad building, was a classic example of plannedresearch, chance discovery, and systematic field-work that revealed a whole new phase inCorstopitum’s early military history. During thefinal season of training excavations at Corbridgein 1973, several fragments of pottery were foundin the foundations of the headquarters building be-longing to the very first fort on the site, providinga textbook example of a terminus post quem. Thepottery was a well-known form of samian warewith moulded decoration, imported from south-ern France. It could be dated by historical evidencefrom other sites, including Pompeii; experts agreedthat it was unlikely to be earlier than c. AD 90,and certainly later than c. 85 (Bishop & Dore1989, 219). This created an immense problem, forit had been assumed since the 1930s that the firstfort was built by Agricola, the governor of Britainwho garrisoned the region in the early eighties ADduring his campaigns in northern England andScotland.

A solution to this problem was offered by aRoman bath-house that had been discovered byaccident in 1955, one kilometre west of the main

site (fig. 3.17). Pottery dating to the time ofAgricola had been found there, and it seemed al-most certain that a military site must lie nearby,for civilian baths would be unthinkable in this areaat such an early date. Since the early fort was defi-nitely not situated on the main site, a level riverterrace immediately to the east of the bath-housenow seemed to be a possible location. By a remark-able chance, a new dual carriageway by-pass wasabout to be constructed, which had been designedto avoid the northern limits of Corstopitum butwould cut right across the site of the newly sus-pected Agricolan fort. Rescue excavations were or-ganized, and, thanks to cooperation from the roadcontractors, the fort’s existence was confirmed; acontinuous open strip over 150 metres long wasexcavated, and the remains of fifteen timber build-ings were found (fig. 3.18). Pottery and coins foundon the site could be dated to the time of Agricola,and were contemporary with the nearby bath-house. They proved that the first military occupa-tion was on the Red House site, and that it shiftedto the main site soon after AD 85/90, the date pro-vided by those critical pottery sherds found therein 1973.

Publication standards had risen dramatically bythe 1970s. The excavation of a small percentage ofthis single-period site, without occupation levels orrubbish deposits, generated a report of 98 pages inthe 1979 issue of Archaeologia Aeliana. The sameperiodical had carried reports on the 1906–14 ex-cavations, but although these earlier seasons in-volved incomparably larger and more complexareas, with abundant structures and other finds, theresults were presented in an average of only 85 pagesper season.

One very small research excavation was carriedout on the Corbridge main site in 1976 to checkan outstanding detail of the interpretation of thesecond-century forts. Then plans were drawn upto construct a new purpose-built site museum. This

3.17 Location map showing the relative positions ofCorstopitum (‘main site’) and the Red House siteexcavated in 1974 before the construction of a by-pass. The structures shown in black on the main siteare those found in 1906–14 (see fig. 3.14), and therectangular outline indicates the position of theunderlying timber forts revealed between the 1930sand 1970s. Hanson et al., 1973, f ig. 1

3.18 Open-area rescue excavations at the Red Housesite, Corbridge. A strip was cleared along the line ofthe new road, although space had to be left forcontractors’ vehicles working further along the route.

The outlines of foundation trenches for timberbuildings can be seen in the foreground. CharlesDaniels

Excavation

76

required a rescue excavation because the museum’sbasement would intrude into archaeological lev-els. As a result, the whole area was excavated torecord any information that had survived distur-bance by earlier trenches. A deep trench cut belowstone buildings on this part of the site in 1908 hadonly found four to five feet of ‘disturbed soil’. Thistrench was emptied again in 1980, and revealedthe remains of successive west ramparts of the earlyforts; a useful cross-section of their structure wasrecorded by the excavators, John Dore and JohnGillam.

2.5 Publication and cataloguing

Besides revealing a sequence of forts and civilianbuildings, excavation conducted over 75 years hadgenerated an enormous collection of artefacts.Many individual papers, notably on classes of Ro-man pottery, had appeared in the periodicalArchaeologia Aeliana from the 1930s through tothe 1970s. Ian Richmond published a particularlyinteresting paper in 1943, ‘Roman legionaries atCorbridge, their supply-base, temples and religiouscults’, that related the rich collection of sculptures,inscriptions and excavated structures to the Romanarmy’s calendar of religious ceremonies. Anotherlandmark was John Gillam’s ‘Types of Romancoarse pottery vessels in northern Britain’ (1957,subsequently reprinted in book form), a fundamen-tal reference work that used finds and dating evi-dence from Corbridge and other sites in the northernfrontier zone to provide dated typological classifi-cations of the principal forms of pottery found inthe region.

In common with other spheres of British archae-ology in the 1980s, further work at Corbridgeveered away from excavation towards publication.The Department of the Environment (in the formof English Heritage from 1984) put more of its re-sources into publishing a backlog of reports andfinds generated by the boom in rescue excavationin the 1960s. This policy was extended to impor-tant sites in State care, such as Corbridge, and re-sulted in two major publications. The first,Excavations at Roman Corbridge: The hoard(Allason-Jones & Bishop 1988) was a detailed ac-count of the discovery, in 1964, of the remains of alarge wooden box that contained the most exten-sive remains of Roman legionary armour ever

found, along with assorted weapons, tools andmiscellaneous other items. The authors concludedthat the contents had been gathered together whena workshop was cleared out at the end of a phaseof occupation of one of the site’s series of forts. Theitems had been packed up ready for transport be-fore a decision was made to bury them instead.Hundreds of detailed scale drawings of the armourand other artefacts were accompanied by a carefulanalysis of their form and function. Re-examina-tion of the excavation records and the stratigraphiccontext dated the whole find to c. AD 122–38—around 30–40 years later than had been thought in1964.

The second book published by English Heritage,Corbridge: Excavations of the Roman fort andtown (Bishop & Dore 1989), drew together all sur-viving information from the long series of excava-tions conducted between 1947 and 1980. Of its 323pages, 140 were devoted to a detailed analysis ofthe myriad of small trenches (rather than open ar-eas) excavated all over the site. Since it was verydifficult to correlate finds and layers from so manydifferent deposits, the results were published inelaborate tables and supplemented by four micro-fiches of further text and drawings. The remainderof the volume comprised reports by specialists oncoins, pottery, glass and ‘small finds’ (miscellane-ous objects made of metal and other materials). Thesequence of dates for the forts built on the site wasnot significantly altered by this research, but theidentification of exactly which contexts (and finds)were associated with each phase provided impor-tant supporting evidence for dating them. Thanksto this work, the report is not only a scholarly dis-cussion of the structural history of a complex mili-tary site, but also a valuable reference work fordating finds from the late first to second centuriesAD.

2.6 A new museum

The 1980s also witnessed a change in the manage-ment of ancient monuments in State care, particu-larly in the quality of their presentation to thepublic (below, p. 175). The construction of a newsite museum at Corbridge allowed finds to be dis-played in an aesthetically pleasing, as well as in-formative, manner, and it also provided anopportunity to improve the storage of the thou-

Excavation

77

sands of other items that are not on public display.The painstaking process of sorting out old recordsfor publishing a report on the excavations wasinseparable from improving the catalogues of themuseum store. The resulting report and museumarchives now offer a collection of accessible, well-documented material ready for future research. Atthe same time, an attractive modern museum en-hances the appeal of the site to ordinary visitors(fig. 3.19).

3 Excavation procedure(Barker 1993)

Every archaeological site is itself a document.It can be read by a skilled excavator, but it isdestroyed by the very process which enablesus to read it…when the site has been de-stroyed all that is left are the site records, thefinds and some unreliable memories. (Barker1993, 13; 1986, 108)

The account of excavation techniques at Corbridgegives an insight into the development of modernmethods; most of the remainder of this chapter willattempt to explain how excavations are carried outtoday. Excavation can destroy a site just as thor-oughly as ploughing, building construction or natu-ral erosion; the only difference is that destructionby excavation is initiated by people who should (intheory) understand the value of ancient sites. Thisawareness brings responsibilities; no site should betouched without consideration of the need to carryout the excavation with the utmost skill, after mak-ing careful plans for the preservation and publica-tion of the results. All excavators face the sameethical question: on what grounds can their actionbe justified? Some may claim that examination of asite will advance knowledge, others that they arerescuing a site from destruction. We should notforget that different motives may be involved, suchas economics, through the enhancement of tourism,or politics, by promoting nationalism.

It must be remembered that fieldwork techniqueshave advanced to a high level of sophistication,making use of remote sensing from the air or geo-physical surveying devices on the ground (Chapter2). Field survey programmes are conducted by non-invasive observation of the ground and limited col-lection of artefacts from its surface. In view of these

alternatives, excavation should be a last resort. Ifexcavation does take place, fieldwork techniquesshould have been exploited fully in advance tomaximize the amount of information availableabout the site, and to assist in the preparation ofthe excavation strategy.

3.1 Selection of a site (fig. 3.20)

New sites are discovered with growing frequency,but resources remain limited; it is therefore impor-tant that sites for excavation should be chosen withcare. Since the number of sites is finite, every exca-vation should form part of a wider programme ofresearch, whether the site is threatened with destruc-tion or not. One site might attract attention becauseit is particularly well preserved, another because itis going to be destroyed; are these good reasons forexcavation? The answer depends on fundamentalattitudes to archaeology. Should a picture of the pastbe built up from a gradual accumulation of inde-pendent observations (like clues in a detective story),or should there be overall strategies? The detective-story approach characterized most work conductedup to the early twentieth century, but our knowl-

3.19 Corstopitum, or Corbridge ‘main site’, today.The purpose-built museum is unobtrusive, and thefacade echoes the colonnaded porticoes of theRoman granaries that lie immediately in front of it;their columns, seen in fig. 3.15, may be distinguishedjust left of centre. Neal Askew, English Heritage

Excavation

78

edge is now so detailed in most parts of the worldthat we have sufficient information for posing mean-ingful questions; it is now indefensible, both ethi-cally and intellectually, to excavate a site simply tofind out what is there.

The kinds of decisions facing archaeologists in-volved in excavation today may be illustrated by ahypothetical problem. Suppose that an area hasbeen extensively examined by fieldwork of variouskinds, and that twenty Roman sites of a similar formhave been recorded. Ten are situated on marginalland, well protected from past or present destruc-tion; of the other ten, nine lie on arable farmland,and are suffering from the effects of ploughing,while the remaining one is about to be destroyedby building work. Sufficient funds are available forthe complete excavation of one site. What shouldbe done?

Some possible choices may be suggested: a) the site threatened with complete destruction should

be totally excavated to save a record of its infor-mation for posterity;

b) one of the well-preserved sites should be fully ex-cavated, because it is in better condition. It can beapproached at a more leisurely pace, since it is notthreatened; the quality of information should there-fore be higher than (a);

c) several sites, including the one threatened withdestruction, should be examined partially in orderto provide some basis for making comparisonsbetween them;

d) none of this class of site should be excavated at all.Resources should be redirected to non-destructivefieldwork aimed at the investigation of a differentperiod, or an area about which less is known. Thedestruction of one out of twenty similar sites istolerable.

There are good points about all of these choices.It would require much confidence to take choice(b), for no two sites are ever identical and surfaceindications may be misleading. There is no guar-antee that a well-preserved site will produce infor-mation of better quality than that found on adamaged site.

If (a) or (b) are selected, how safely can the re-sults from one site be generalized to the other nine-teen? Few scientists would accept results gainedfrom only 5% cent of a sample. Choice (c) may seemto be a good compromise, but partial excavationsalways leave unanswered questions, and completeknowledge of one site might be more informativethan one quarter of each of four; opinions wouldvary over this solution. Whether (d) is better thanany of the others depends on the interests of thedecision-makers; a specialist on the Bronze Agewould obviously prefer to increase aerial survey and

3.20 Richard Rogers’ remarkable Lloyds building inthe City of London overlooks an excavation of theRoman basilica; this was a large hall attached to theforum of Londinium, capital of Roman Britain. Apartfrom a period of near-abandonment in early Anglo-Saxon times, London has been an administrative andcommercial centre linking the British Isles with Europeand the rest of the world for nearly two thousandyears. This site also illustrates the pressures of rescueexcavation in an urban setting. Procedures forexcavation and recording must be adjusted to takeaccount of the time scale and nature of building work.Large modern buildings require such deep founda-tions and basements that, unlike most constructions inprevious centuries, they remove all trace of earlieroccupation. Museum of London

Excavation

79

fieldwork on prehistoric sites rather than excavatea Roman site.

English Heritage has added a variation on thistheme of conflicting priorities (Exploring… 1991, 34):

The management of the archaeological resourceis based on a series of stages which form theframework for decision-making and the formu-lation of strategies. Three such stages can be pro-posed in the context of archaeological resourcemanagement and together these stages may becalled the management cycle:

Stage 1 Identification, recording and the understanding ofthe monument or historic landscape.

Stage 2 Option 1—curatorial management where the mainaim is to arrest the natural and man-induced proc-esses of decay through protection and manage-ment.

Stage 2 Option 2—exploitative management where thearchaeological resource can be used for publicenjoyment through interpretation and display, orfor academic interest through investigation andexcavation.

Stage 3 Recording—in exceptional circumstances, whenpreservation is no longer possible, because thevalue of the archaeological resource is out-weighedby some other factor, a site may be excavated torecord as much as possible of its structure and formand thus in effect preserve it as a record.

Stage 1 Identification, recording and the understanding ofthe monument or historic landscape.

Stage 2 Option 1—curatorial management where the mainaim is to arrest the natural and man-induced proc-esses of decay through protection and manage-ment.

Stage 2 Option 2—exploitative management where thearchaeological resource can be used for publicenjoyment through interpretation and display, orfor academic interest through investigation andexcavation.

Stage 3 Recording—in exceptional circumstances, whenpreservation is no longer possible, because thevalue of the archaeological resource is out-weighedby some other factor, a site may be excavated torecord as much as possible of its structure and formand thus in effect preserve it as a record.

Central to this cycle is the understanding of theresource which requires research and evaluation.For that reason the stages of the managementcycle are preceded by the proposed frameworkof the academic objectives which English Herit-age considers necessary in order to provide aframework for the future.

Although a cynic might be suspicious of the refer-ence to ‘public enjoyment’, and question how ‘the

value of the archaeological resource’ is to be meas-ured, the emphasis on placing excavation within a‘framework of the academic objectives’ is impor-tant. If resources are scarce, it is better that theyshould be distributed according to a strategic plan—even if no two archaeologists agree on the detailsof such a plan! The most dangerous phrase in thisextract lies in Stage 3, however, where elegant phras-ing makes it sound acceptable to ‘preserve’ a site‘as a record’; few archaeologists are sufficientlyconfident about the process of excavation to regardthe results as nearly equivalent to the site itself.

A further question may be added at this point:what should be done about ‘unique’ sites? Stone-henge, Avebury and Silbury Hill are all ‘unique’prehistoric sites in Wessex, if only because they areso much bigger than other contemporary stone cir-cles, henge monuments or mounds. Should effortsbe concentrated on more ‘typical’ sites, or can thesemonuments be expected to give clearer insights intotheir times because of their individual character? Itis instructive to frame the same question in a his-torical period: which will tell us more about medi-eval Kent—an architectural and archaeologicalanalysis of Canterbury Cathedral, or the study of aselection of churches from surrounding urban andrural parishes? Again, there is no clear answer with-out a statement of research objectives. In practice,the intellectual climate of the 1990s would not lookfavourably upon any further excavation of a majormonument such as Stonehenge unless it was facedby a dire threat that could not be met by protectionand management, rather than intervention. Perhapsthe only conclusion to emerge from this discussionis the need for conscious decisions and policiesabout all ancient sites, based on a good knowledgeof their significance in a particular region or coun-try, and their relevance to current research ques-tions.

3.2 Planning an excavation

The role of site director varies in scale and complex-ity. An individual engaged in a small research projectmight dig a minor site with a few volunteers,whereas the director of a permanent excavation unitinvolved in a major research or rescue project wouldemploy a full-time team of special assistants and apaid labour force. An average excavation will prob-ably be planned by a single director who takes re-

Excavation

80

sponsibility for preliminary research, setting up andconducting the actual excavation, and writing upthe results for publication. Even when a site has beenselected as a result of a sensible programme offleldwork and there is a clear reason for the exca-vation, an enormous amount of research and or-ganization still remains to be done, irrespective ofthe period or geographical location of the site.

3.3 Background research

No excavation should take place without a care-ful study of previous work on the site, or with-out an appreciation of how it fits into a widerlandscape. The amount of data that has beencollected in advance will determine how well thedirector is able to draw up detailed plans for theexcavation, and to respond to new informationthat is revealed as excavation proceeds. Much ofthis preparation must be conducted in librariesand archives where earlier accounts and illustra-tions of the site may be found. Records of previ-ous excavations, or stray finds recovered from thesite, may lead to material preserved in museumcollections, which must be examined before a newexcavation takes place, and, if unpublished, in-cluded in the final report. In countries wheredocumentation extends far enough back into thepast, archive offices may hold informative mapsand plans, such as those drawn up to record theownership of land. These might show features ofthe site itself or physical surroundings, along withforgotten but informative place-names and field-names. In England, a site or its area may evenhave been mentioned in an Anglo-Saxon charter,a form of land document that survives from asearly as the seventh century AD. Ideally, thisdocumentary research will form part of a pro-gramme of fieldwork to investigate the local land-scape (above, p. 51).

The most important modern source of infor-mation that must be explored is aerial photogra-phy. Ideally, a number of photographs will havebeen taken over a long period under varied con-ditions, and may reveal slight surface earthworksor the effects of buried features on growing crops(see fig. 2.4). Maps of soils and local geologyshould also give insights into the natural subsoilconditions that will be encountered. The siteshould be visited to check knowledge gleaned

from the above sources against its present stateon the ground. If it is not visible, detailed surveysmust be made to relate it to modern referencepoints, or, if necessary, to fix its exact positionwith the help of geophysical devices (above, p.46). All visible traces must be recorded, especiallyon a research excavation where the original formof the monument must be restored from an accu-rate contour survey. If sufficient resources areavailable, new aerial photographs may be com-missioned, along with more extensive explorationwith resistivity equipment or magnetometers, andprofessional surveyors may be employed. With allthis information, the director will be ready todecide what the priorities of the excavationshould be, and to devise the optimum layout oftrenches to explore them.

3.4 Staff and equipment

The number and nature of staff employed on anexcavation are directly related to its size, re-sources and complexity; on small sites, manytasks will be performed by the director. Althoughit is impossible to generalize, certain basic re-quirements always exist. Because excavation de-stroys a site, recording is the most important partof the process. Site records come in three forms—written, drawn and photographic—and everyexcavation must have staff with appropriate ex-pertise, both archaeological and technical, tocompile and maintain them. A large excavationthat produces large numbers of finds will alsorequire a full-time finds assistant, with sufficienthelpers to sort and wash pottery and other mate-rials. Lists of finds from each excavated contextmust be compiled, and every artefact must belabelled, bagged or boxed for storage, in such away that it is accessible for further study whenrequired. An on-site conservation laboratoryshould be provided on sites where delicate findsare likely to be abundant; remains of wood,leather or textiles recovered from waterloggeddeposits require immediate treatment, for exam-ple. The site should also employ someone with agood knowledge of environmental evidence,which involves the careful selection of soil sam-ples, and the use of sieving or flotation equipmentto increase the recovery of small bones, seeds, etc.Since all of these categories of informa-tion about

Excavation

81

the site and its finds are suitable for recording incomputer databases, an excavation should alsohave a competent computer operator as well asappropriate computing equipment. The ultimatequality of the interpretation and publication of asite depends not only on the skill of the excava-tor, but also on the comprehensive nature of therecording and preservation of finds; these factorsshould guide the director’s selection of staff.

3.5 Excavation strategy

The director’s experience, combined with all thebackground research that has been carried out inadvance, will determine how best to approach asite. Aerial photographs and the results of geo-physical prospecting may suggest the most in-formative parts; this is particularly important ifthe site is being sampled, rather than totally ex-cavated. Exploratory ‘sondages’, larger trenchesor extensive open areas are employed in variousconfigurations, according to the availability ofresources and the nature of the site. Since eventhe best-prepared director rarely finds that a siteconforms exactly to expectations, the strategyshould allow for modifications. Flexibility isimportant, and it depends to a large extent on thequality of the excavation’s specialist staff. Inter-pretation of excavated structures, combined withprovisional dating and other observations aboutthe finds, should provide continuous feedbackthat can help to reinforce, modify or reject thedirector’s working hypotheses. It is now possibleto compile up-to-date computer files of site infor-mation for this purpose during an excavation, andto carry out sophisticated analyses while the workprogresses.

Stratigraphic excavation is sometimes comparedto the dissection of a biological specimen, becauselayers or other features are removed in strict order,and detailed records are kept of each stage. Theanalogy is misleading in two major ways. First, ifdissection goes wrong another specimen can beobtained, but no two sites are ever identical. Sec-ond, the structure of tissues, bones and organs ismore predictable and easier to recognize than asequence of archaeological features; layers of soilmay merge into each other without clear interfaces,leaving the excavator to decide what divisionsshould be made for the purposes of recording.

However scientific a site’s recording system mayseem, it is inevitable that major subjective elementsremain. Individuals working with trowels makecontinuous observations and judgements about thetexture, colour and significance of soils, deposits orfeatures (some samples should be taken for exami-nation by soil micromorphology: Courty 1989). Inaddition, they must be able to recognize all kindsof finds, from solid stone or pottery and fragilecorroded metal to the faint discolorations left byorganic materials that have decayed away com-pletely. The alarming early history of excavationshows how easily this information was lost.

3.6 Recording

The large size of many modern excavations meansthat the role of director resembles that of a gen-eral manager who balances objectives against re-sources, while site supervisors control the detailsof the excavation. In the past, directors kept anoverall check on the entire process by writing de-scriptions of layers and features in site notebooks,and jotting down subjective insights that mighteventually help in the interpretation of the basicplans and drawings. An increasing demand forobjectivity and accuracy, combined with the com-plexity of, for example, a large open-area multi-period urban excavation, led to the design ofpre-printed recording forms (fig. 3.21). These al-low the director to impose standardized recordingmethods that considerably reduced the scope forerrors or omissions. Individual record forms areused to describe the position, size and character-istics of each separate excavated context. Formsshould be designed with future analysis of the sitefor publication in mind, so that the description ofa feature is cross-referenced to all relevant photo-graphs, plans and significant finds.

Further files are required for recording catego-ries of excavated material such as pottery, bones orscientific samples, as well as for supplementary in-formation like section drawings and photographs;all of these must also be cross-referenced back totheir relevant excavated contexts. Ideally, site recordforms and details of finds should be coded in sucha way that the information can be transferred rap-idly into data files on a computer during the exca-vation. It may even be possible to type informationstraight into a portable computer, while the plan-

3.21 A standard form for recording excavated featuresat South Shields Roman fort, Tyne and Wear. Thislayer has been given a unique context code (5885),and the excavator has assigned it to a phase (‘con-struction 7’) in the Roman sequence. Relationshipswith earlier or later contexts are identified by ‘over’or ‘under’, ‘cuts‘ or ’cut by‘; these could be added toa Harris Matrix (see fig. 3.13). A series of recordeddimensions follows, together with references to plans,sections or photographic records; more generalobservations and interpretations are entered into‘description’ and ‘event’. ‘Finds’ provides space for

mentioning important discoveries (in numberedtriangles) and indicating other finds, which will befully catalogued elsewhere. Finally, the date of therecord and the initials of the recorder are added sothat any discrepancies may be checked. When theform has been entered into a computer database, thecontext code will link it directly to other files that listphotographs, plans and finds, or to other featuresexcavated nearby. Paul Bidwell

Excavation

83

ning of excavated features is carried out with sur-veying equipment that logs its readings in electronicform for processing by computer. High-qualityvideo cameras offer an additional photographicrecord of all aspects of excavation (Hanson & Rahtz1988).

3.7 Computerized processing of siterecords

Information needs to be recorded with a clear ideaof the questions that are likely to be asked aboutit at a later stage, and the archaeologist must beaware of these factors while recording features orartefacts. The presence or absence of a long list ofattributes must be recorded for every layer of soil,post-hole, wall or whatever else is encountered,including dimensions, location, soil colour andtexture, any finds, stratigraphic relationship toadjacent features, etc. This stage is tedious andtime-consuming, but with forethought it is simpli-fied and speeded up by abbreviated codes and well-designed computer programs that check the dataas it is being entered.

Time devoted to computerized recording is re-paid when the director comes to write the excava-tion report, for a comprehensive site database canbe indexed in many different ways to produce spe-cific information sorted into any desired order. Forexample, it may be suspected that post-holes filledwith brown, stony soil have some particular sig-nificance. The computer can be instructed to searchfor all such features, and to select those examplesthat contained pottery or similar material suitablefor dating, to see if they all belong to the sameperiod. If it is suspected that these post-holes rep-resent buildings whose plans are difficult to dis-tinguish from other features on a plan by eye, acomputer with suitable graphic software can plotthem along with all other features that containedmaterial of the same date, such as rubbish pits,fences, hearths, etc. If the dimensions of excavatedfeatures have been digitized, GIS systems allowdata and plans to be combined in all sorts of com-binations.

The advantages of computerized recording be-come particularly obvious if the alternative is con-sidered: multiple card-indexes, masses of site plans,photographs, notebooks and perhaps only thememory of the excavator for some details. An enor-

mous amount of time would be involved in siftingthrough all of these records manually to provide theinformation demanded by the straightforward en-quiry outlined above, and the boredom it couldgenerate might seriously affect the accuracy andcompleteness of the process. The cost and availabil-ity of computers are now well within the reach ofsmall sites; no director of a large, complex site canpossibly afford to ignore the value of comprehen-sive computer-based recording of all aspects of theexcavation.

3.8 Publication(Cunliffe 1982)

Publication was assisted by the invention of pho-togravure in 1879, for photographs could be re-produced without having to be printedindividually or redrawn as engravings (Feyler1987, 1045). For the next century it was gener-ally assumed that a site should be published inbook form, and this concept was boosted by themassive printed volumes produced by Pitt Riversat the very end of the nineteenth century. By the1960s it was taken for granted that an excava-tion report would consist of an ‘objective’ accountof the excavated features and structures, followedby descriptive catalogues of each category offinds, along with scientific reports on bones, en-vironmental samples, etc. (fig. 3.22). However,the unparalleled size and number of rescue exca-vations in the 1960s and 1970s, followed byspending cuts in the 1970s and 1980s, made it im-possible to produce reports of this kind. A numberof committees proposed new standards of publi-cation, with varying emphases, but with a con-sistent recommendation that much materialshould be stored in archives, rather than pub-lished. Thus, the phrase ‘preservation by record’was coined to minimize the reality of the destruc-tion of a site at the same time that ‘preparationof an archive’ was substituted for compiling andpublishing a traditional excavation report.

English Heritage recently set out a definition ofminimum requirements for the contents of a pub-lished report (MAP 1991, 39): i. the research objectives as expressed in the

project designii. circumstances and organisation of the work

Excavation

84

and the date at which it was undertakeniii. identity of the individual/organisation by

whom the work was undertakeniv. summary account of the results of the projectv. summary of the content of the project archive,

where it is housed and how it may be consulted. The same document also defined the specificationof a research archive (37):

The research archive will be derived fromthe work done during the analysis phaseand will comprise: stratigraphical/struc-tural, artefact, environmental, and othercatalogues and all other records as well asdetails of the methods and selection strate-gies used in each case. Each separate datagroup should be cross-referenced to relateddata groups, to the final publication, andif necessary to a general context concord-ance. These should be supplemented by in-d ices to a l low the user maximumaccessibility to the contents.

The move towards unpublished archives coin-cided with a growing use of standardized ex-cavation records held on computers, and thishas interesting implications for how publica-tion might take place in the future. Why issuesite records in an indigestible and expensivelyprinted book, when they can be left in a com-puter archive to which specialists are allowedfree access? Why not publish a short summaryof the principal structures and finds, in a formthat will be more interesting to general readersand simply inform specialists about the exist-ence of archives? This attractive solution con-ceals the danger that excavations publishedonly in summary form do not allow readers tocheck the details upon which the excavator’sgeneral interpretations are based. A solutionmay lie in the growing availability of videodisksand CD-ROMs, together with indexing andsearching programs similar to the Hypertextsystems that began to reach an advanced stateof development in the late 1980s. In the twenty-first century, excavation reports may becomemultimedia experiences, with site records, pho-tographs, drawings and video sequences storedon a single disk.

3.22 There are many important stages betweenexcavation and publication. Perhaps the mostimportant is the amalgamation of plans, sectiondrawings and recording sheets into a stratigraphicsequence, and its interpretation in terms of periods ofoccupation and abandonment, and phases character-ised by particular activities. The finds should beprocessed at an early stage, for artefacts such aspottery and coins assist in dating and understandingthe nature of occupation, while bones or samples ofsoil, plants, etc., help to elucidate environmentalconditions and the economic functions of each phase.All finds, together with any information that will notappear in a published report, must be documented infull, ready for storage in a museum or archive. SandraHooper, after Harris 1989, f ig. 57

Excavation

85

4 Excavation: special cases(Schiffer 1987)

It has already been stressed at the beginning of thischapter that archaeology cannot be classed as ascience because one of its primary sources of infor-mation, excavation, involves the destruction ofunique specimens. The remainder of this chapterwill look at various forms of excavation, withoutmaking any attempt to provide a comprehensive‘manual’ of methods. My personal experience islimited to sites in Britain and north-western Europe,ranging in date from neolithic to medieval, involv-ing everything from almost invisible traces of tim-ber structures to elaborate Roman masonry. I haveworked on urban and rural sites situated on a rangeof subsoils, from the relatively easy chalks andsandstones of southern England to the more diffi-cult valley silts of south Wales, and the intractableglacial clays and dolerite of Northumberland. Be-fore tackling the interesting problems involved inexcavating and understanding various forms ofbuildings, it is important to look briefly at othercategories of sites that are important in the archae-ology of earlier periods or other parts of the world.

4.1 Camps and caves(Smith 1992)

A sedentary lifestyle is a very recent innovation inhuman history. It began at different times in differ-ent parts of the world, normally when settled farm-ing came to replace hunting and gathering as thepredominant means of subsistence. Before this,human occupation revolved around numerous tem-porary and/or seasonal sites whose locations weredetermined by hunting or gathering activities. Ar-chaeological traces of such sites should reflect theirfunction, and the nature and length of occupation.Thus, the home base of a hunter/gatherer groupmight contain semi-permanent houses and otherstructures, while a field camp associated with hunt-ing expeditions would be very different (Smith1992, 29, fig. 3.1). From the point of view of exca-vators, the principal problem of open sites is thatseasonal occupation and insubstantial buildings didnot lead to the formation of an appreciable depthof stratified deposits. They are particularly suscep-tible to erosion and other forms of natural distur-

bance; exceptionally, sites occupied at the edge ofwater were covered by silt, and coastal sites weresometimes buried by blown sand. The priority ofan excavator is to make exact records of the posi-tion of every fragment of bone, flint or other arte-facts, for an analysis of their distributions maysuggest the functions of different parts of a site andimply relationships between structures and activi-ties (Smith 1992, 30, fig. 3.2):

Many repetitive tasks such as fine flint knappingor bone carving are often most comfortably car-ried out in a seated position and, requiring littlespace, may take place inside. Butchery requiresspace so that the operative can move around thecarcass and, because it is associated with unpleas-ant smells, it is usually considered to be an out-side activity. Stationary tasks may requirewarmth and need to be carried out by the firewhile detailed tasks may require light and, if theyare undertaken inside, will need to be locatednear an opening. It is inconvenient, and even dan-gerous, to walk on flint waste or fractured ani-mal bone, and at a site occupied for a prolongedperiod such rubbish is likely to be cleared away,(ibid. 31–2)

The interpretation of sites as field camps or homebases, and the identification of the activities thattook place on them, is based largely uponethnoarchaeological observations of modernhunter-gatherers (below, p. 171).

Caves occupied from time to time offer muchbetter possibilities for finding stratified deposits,but little in the way of structures; furthermore,their stratigraphy is likely to result from naturalweathering and sedimentation as much as fromhuman activities. Cave excavations were very im-portant in the nineteenth century because assem-blages of tools and other artefacts could be foundin association with bones of animals and, occasion-ally, human burials or even cave-paintings. Thebones and tools that Boucher de Perthes retrievedfrom gravel beds in the Somme valley were onlyassociated in the sense that they had ended up inthe same geological deposit. However, animalbones found in caves are very misleading if thosethat were brought there by human hunters are notdistinguished from those introduced by carnivo-rous animals when humans were absent. This

Excavation

86

problem applies to most sites with signs of earlyhuman occupation, including those associated withfossil hominid remains in East Africa, with theresult that opinions are divided over whether earlyhumans were hunters at all, or mere scavengers.Donald Johanson’s informative and entertainingaccount of interpretations of Site FLK in OlduvaiGorge provides an excellent historical outline ofchanging perceptions of sites and their finds sincethe 1940s (1991, 213–44).

4.2 Waterlogged sites(Coles 1984)

Information about the indigenous peoples of theAmericas and Australia gave early prehistoriansin Europe valuable perspectives on the lifestylesof people known only through discoveries ofstone tools. An important additional source ofknowledge resulted from a drought in Switzer-land in 1853–5, when low water levels revealedseveral hundred ‘lake-dwellings’ (probably settle-ments on the shore, rather than dwellings con-structed on piles in the water itself, as originallythought). Ferdinand Keller published an influen-tial book about these sites in the 1860s that in-cluded illustrations of a wide range of organicstructures and artefacts, as well as bones and

plant remains, that complemented the stone andmetal artefacts being studied by Scandinavian ar-chaeologists (above, p. 26). In Trigger’s judge-ment, ‘…the continuing study of these prehistoricremains attracted wide interest. It played a ma-jor role in convincing Western Europeans of thereality of cultural evolution and that ancient timescould be studied using archaeological evidencealone’ (1989, 84).

The growth of ecological perspectives on archae-ology in the twentieth century enhanced interest inthe range of information that could be retrieved

3.23 An early medieval timber round-house pre-served in wet conditions at Deer Park Farms, Antrim,Northern Ireland. The walls were constructed frominterwoven branches, whose inherent strength andlight weight demanded little in the way of foundationsor supporting posts; a length of wall that collapsedinwards can be seen in the foreground. The exteriorwould presumably have been made weatherproofwith clay. Remains like these remind excavators ofsites where wood does not survive that they must notunderestimate buildings that lack impressive founda-tions. One of the excavation workers kneels beside astone hearth, providing a useful indication of scale; alarge number of people could be accommodated in astructure of this kind. C J Lynn

Excavation

87

from waterlogged sites. It led to the ‘classic’ inves-tigation of a mesolithic site at Star Carr in York-shire by Grahame Clark (Smith 1992, 110–21). Thistradition continues in Britain, and includes majorwork prompted by ‘rescue’ archaeology, such as theSomerset Levels project (Coles & Coles 1986) orFlag Fen, near Peterborough (Pryor 1993); wetlandarchaeology also thrives in many other parts of theworld (Coles & Lawson 1987; Coles 1992). Wetsites also enhance periods of historical archaeology(fig. 3.23). In Ireland, for example, the early medi-eval period is marked by the poverty of artefacts andabsence of pottery from settlement sites. Fortu-nately, some were constructed on artificial islands(crannogs) in lakes, and finds from these illustratewhat is missing on land—notably a range of bar-rel-shaped containers and lathe-turned woodenvessels that made pottery unnecessary for most stor-age and table purposes (Edwards 1990, 77, fig. 30).Crannogs also produce many finds of metal arte-facts, for it was difficult to recover objects that weredropped into water or fell between gaps in timberfloors.

John Coles gave the name sensitivity analysis tohis outline of the most important components of theassessment that should take place before anywetland site is examined (1984, 36–8): 1. site identification (e.g. settlement, burial, etc.)

and chronological position: the date and natureof any kind of site are fundamental to any re-search that is carried out.

2. site content: awareness of the types of evidencelikely to be encountered will deter mine the ex-cavation strategy.

3. condition of the site: the extent of waterloggingand the nature of the deposits may vary fromsoft peat to hard mud.

4. recovery techniques to be used: an excavatormust make a hard-headed estimate of aproject’s resources and effectiveness in planningthe tactics to be used.

5. requirements for post-excavation work: ‘Thisis probably the most crucial element in any pro-gramme, and unless it is fully debated andagreed beforehand, and arrangements made,the excavation should not proceed. Samplingand extraction of materials for analyses, spe-cialists and their specific requirements, andrelevance of such work to the project’s aims,

must be laid down and fully understood by all’(ibid. 37).

6. conservation needs: a very careful excavationstrategy is required to limit the exposure ofthese finds to the air during excavation, and toensure that adequate facilities are available forimmediate treatment and long-term preserva-tion of organic finds.

Ideally, these six aspects should be applied to theexcavation of any kind of site; the implications ofinsufficient planning are more immediately appar-ent on wet sites, however. The principal differencebetween ‘dry’ and ‘wet’ sites lies in the range oforganic finds that will be encountered. Excavationtrenches will be more difficult to manage, for it maybe necessary to erect platforms suspended above theworking surfaces to avoid walking on fragile re-mains. The greater complexity of structures com-plicates recording by planning and photography,while additional categories of finds make increaseddemands on the cataloguing system. Publicationcosts will also be affected by the scale and qualityof the discoveries.

Perhaps the single most important developmentin wetland archaeology has been the extension ofdating by dendrochronology into prehistory; it isnow possible to provide accurate dates for timberstructures back to about 9000 BC in Europe. Tree-rings can also be used to analyse periods of construc-tion on complex sites (below, p. 110). This degreeof precision allows a ‘historical’ perception of theconstruction and development of suitable prehis-toric sites that is discussed in relation to datingmethods.

4.2 Underwater archaeology(Dean 1992)

As with wetland archaeology, the guiding principlesand methodology of underwater archaeology areidentical to those that should be employed on drysites (fig. 3.24). However, the additional complex-ity of the tasks of discovery, excavation, recordingand conservation forces directors of underwaterprojects to take a much more stringent approachto their objectives and ethics. The Nautical Archaeo-logical Society, based in Britain but with one thirdof its members abroad, has clear principles of con-duct (Dean 1992, 300):

Excavation

88

i. NAS members shall adhere to the highest stand-ards of ethical and responsible behaviour in theconduct of archaeological affairs.

ii. NAS members have a responsibility for the con-servation of the archaeological heritage.

iii. NAS members will conduct their archaeologi-cal work in such a way that reliable informa-tion about the past may be acquired, andshall ensure that the results are properly re-corded.

iv. NAS members have a responsibility for mak-ing available the results of archaeological workwith reasonable dispatch.

Maritime archaeology is particularly troubled bythe indiscriminate damage to shipwreck sites bylooters and souvenir hunters. Much of this activityis not malicious in intention, for diving is a popu-lar leisure activity whose attraction is enhanced byvisiting wrecks. However, commercial salvage onmodern wrecks shades imperceptibly into the de-struction of historical evidence from older ships by

unscrupulous treasure hunters. Shipwrecks offerrare insights into the technology, warfare and com-merce of the past, and individual ships provide achance to study a range of artefacts that were all inuse at a specific date. For this reason they are fre-quently described as ‘time capsules’. On most dry-land sites, structures are found in the state that theyreached after their useful life had ended, and arte-facts often accumulated in rubbish dumps over anextended period. In contrast, finds from the MaryRose, a Tudor warship that sank in 1545, illustrateweapons, clothing and personal items that belongedto members of the crew at the moment of the ship’sloss. When sufficiently large numbers of ancientships and their cargoes have been recorded andpublished carefully, each find contributes to a widerpicture of ancient trade. We now have a good ideaof the changing patterns of shipping and trade inthe Mediterranean, thanks to a database compiledby Parker (1992) of 1259 wrecks dating from pre-historic times to AD 1500.

4.3 Graves(Roberts 1989; Pader 1981)

Occasionally the concept of ‘time capsules’ is ex-tended to burials, for the majority were placed inthe ground at a single moment in the past. For anarchaeologist, the most informative kind wouldconsist of the skeleton of a dead individual who hadbeen placed into a grave fully clothed, accompaniedby grave-goods—a selection of personal items, orgifts to take into an afterlife—that might indicatethe deceased’s sex, social status and religion, andhelp to date the burial (fig. 3.25). Unfortunately,burials are normally more complicated; bodies werecremated in many periods, grave goods were notalways included, and large numbers of individualswere sometimes buried together in collective tombs.Worse still, acidic soil conditions can remove virtu-ally all traces of human bone, and cause severedamage to artefacts. Ethnoarchaeological studiesprovide many warnings about the interpretation ofburials, which may just be the end-result of a com-plicated series of ritual practices (Humphreys andKing 1981).

Burial practices also vary in terms of cemeteriesand structures. Some graves were isolated, ratherthan placed in neat cemeteries with clear markersto indicate their position. Structures may range from

3.24 The principles of excavation and recording on anunderwater site differ little from those employed ondry land, but many special techniques have beendevised to maximize the effectiveness of divers who,for reasons of safety, can only spend limited amountsof time working. Here, a planning-frame with ameasuring grid is being used to support rigid plasticsheets onto which features may be drawn accuratelyby eye. Dean 1992, 168

Excavation

89

simple wooden memorials to massive stonemausolea, such as those of the emperors Hadrianand Augustus that can still be seen beside the Tiberin Rome, or the pyramids of Egypt. In many partsof the world prehistoric burials were incorporatedinto earthen barrows or stone megalithic tombswhose monumental character attracted the atten-tion of antiquarians, who ‘excavated’ them withlittle attention to anything other than the hope offinding grave-goods.

Because the process of burial was frequently ac-companied by ceremonies, modern excavationsshould not just examine a burial, but also its sur-roundings, where remains of ritual feasts or reli-gious structures might be found. In Christiancontexts, there may be signs of cemetery churchesor facilities for pilgrims visiting notable graves(Rodwell 1989).

The excavation of burials and cemeteries re-quires scrupulous recording for many reasons. Ifsoil conditions do not favour the survival of bonesit may nevertheless be possible to detect the formerposition of a skeleton from subtle changes in thetexture of the earth, or by chemical analysis of soilsamples from the grave. The exact positions of theremains of dead individuals must be recorded, forthey may be related to factors revealed by pathol-ogy, such as signs of injuries on bones. If gravegoods are present, their position may also be sig-nificant, particularly where items of jewellery orother clothing accessories are concerned, for thesemay suggest the form of costume worn by the de-ceased individual. Recent excavations at the royalAnglo-Saxon cemetery at Sutton Hoo in Suffolkillustrated the complexities that may occur on asingle site. The burials include both cremations andunburned bodies, humans and animals, some plaingraves and others under large barrows, graveswithout goods, and others amongst the richest everfound in Britain. The most famous barrow origi-nally covered a wooden chamber, filled with grave-goods, in a complete ship that had been draggedup a slope from the nearby River Deben and low-ered into a deep trench beneath the mound (Carver1992).

5 The excavation ofstructures(Barker 1993)

Early excavators cleared stone-built sites withoutany regard to traces of timber structures, and, inthe Near East, only solid stone or fired clay struc-tures were detected on sites whose buildings wereconstructed mainly from sun-dried mud bricks(see fig. 3.33). There is no substitute for experi-ence, based on an understanding of the manyprocesses involved in the formation of an ar-chaeological site—how ruined buildings decay,how occupation layers accumulate, how ditches

3.25 Photograph of a recently discovered Anglo-Saxon burial at Sutton Hoo, Suffolk (see also fig.3.1). Mound 17 covered an undisturbed gravecontaining a man in a coffin, placed in a large oval pit,along with assorted grave goods, including weaponsand horse-gear. Close by was a separate burial pitcontaining the skeleton of a horse. The excavation ofsuch complex and delicate remains requires consider-able expertise, and a comprehensive understanding ofthe problems of lifting and conserving finds for futurestudy and display. © Sutton Hoo Research Trust;photograph by Nigel Macbeth

Excavation

90

silt up, etc. Excavators should also have sometechnical understanding of the variety of con-structional techniques that may be encounteredon a site (fig. 3.26).

The study of buildings—whether fortifica-tions, temples, palaces, houses or workshops—is fundamental to archaeology. Antiquariansand travellers from the Renaissance to the nine-teenth century did produce many excellent pub-lications of buildings, but the possibilities forrecording them have been expanded by photog-raphy and film. At the same time, archaeologi-cal techniques have increased the quantity andrange of information provided by structuresfrom any area or period. Information gainedfrom the excavation or analysis of buildings isof enormous value, whether it comes from earlypalaeolithic structures such as the oval woodenhuts that date back to 380,000 BC at TerraAmata, Nice, in France, or from recent millsand factories associated with the IndustrialRevolution.

5.1 The excavation of stone structures

Second only to burial mounds and other majorearthworks, stone structures were the easiestform of site for early archaeologists to recognizeand excavate. In many cases quite reasonableconclusions were drawn from them, for manymore nineteenth-century excavators had an ar-chitectural training than those of today,paticularly where Greek and Roman archaeol-ogy were concerned. J T Wood, who excavatedat Ephesus in Turkey in the 1860s, was able toexplore Roman theatres and the Temple of Dianaknowing roughly what to expect, and with anunderstanding of how fallen columns or frag-ments of architectural sculpture might have fit-ted into the original structures (Wood 1877). Ifless was known about a site it was common fornarrow trenches to be dug to locate buried walls,and for digging to continue along the sides ofwalls looking for off-shoots or junctions in anattempt to recover complete plans. ‘Islands’ ofunexcavated soil left in rooms by this proceduremight then be cleared in the hope of recoveringobjects to indicate the date and function of vari-ous parts of the building. Vital stratigraphic re-lat ionships between walls , f loors, and

occupation or destruction levels were entirelyremoved, making it virtually impossible formodern archaeologists to make sense of the re-mains by re-excavation.

The use of stone for construction does notnecessarily have any significance for technicalability, wealth or social status, except in obvi-ous cases of monumental architecture such asStonehenge or the Parthenon. Stone was used inareas where it was conveniently available; whenit was not, timber, mud brick, or other usefullocal resources provided alternative buildingmaterials. The remains of stone or brick build-ings are much easier to recognize than mud brickor timber; as a result, problems of differentialsurvival distort evidence for human settlementin many parts of the world.

A great variety of construction techniques isemployed when building stone is readily avail-able. Dry stone walling is the simplest; uncutnatural boulders or roughly quarried blocks arestacked up with care or laid in regular courses,relying on gravity and friction for their stabil-ity. Outcrops or boulders that might otherwiseobstruct farmland would have provided veryconvenient sources of stone. Dry stone con-struction is not unskilled, but tools or other ma-terials are not essential. The stones are bondedtogether with tough clay or mortar to increasestrength and to improve the comfort of build-ings by reducing draughts and water penetra-tion. Many Roman buildings relied on thestrength of the concrete cores of their wallsrather than on the visible stone surface, whichformed a mould for the concrete while it set.Some major buildings of the second century AD

3.26 An understanding of construction techniques isvery helpful in the interpretation of excavated remainsof structures. This diagram presents eight imaginaryexamples of wood, stone and clay used in a variety ofcombinations (many other possibilities exist). Sinceonly four of these structures have holes or trenchesdug into the ground, the remainder might leave notrace on a site that had suffered from ploughing orsurface erosion. The location of structures builtdirectly on the surface may be detected on undis-turbed sites by means of scrupulously careful openarea excavations, such as those carried out atWroxeter by Philip Barker (1990, 1993). Audio VisualCentre, University of Newcastle

Excavation

92

and later were made of thick concrete with oc-casional courses of stone or tile, and thin or-namental wall claddings made of marble ormosaics. They can appear unsightly today if theconcrete cores have been stripped of their fac-ing by later stone robbers, especially wheregood building stone is not easily available lo-cally.

Much of the final work on Greek, Roman andmedieval buildings made of finer stone was car-ried out on site. Blocks were roughly shaped atquarries to minimize their weight during transport,and trimmed to the exact requirements duringbuilding; decorative carving was usually completedat this stage, too. Small chippings of stone anddiscarded tools can often be detected during exca-vations, and they are useful in relating thestratigraphic record to the building. Debris left bymasons cutting blocks of fine imported stone intogeometrical shapes for decorative wall claddingwas recognized at the Roman palace at Fishbourne,Sussex; the excavators also found spreads of build-ers’ mortar, and splashes of paint from the walls(Cunliffe 1971, pl. 11).

Structure (Adam 1994)Remains of stone walls uncovered on an excava-tion do not necessarily result from a stone build-ing, for low stone walls were frequently used toprovide firm dry footings for clay or timber build-ings. Their effectiveness is shown by the largenumbers of picturesque cob and half-timberedcottages that have survived for many centuries inthe south of England; dry footings combined withan overhanging roof that throws rainwater wellclear of the wall face are most effective. Mud orclay walls are ideal in drier climates, and quitecomplex structures were already being constructedon early agricultural sites of the eighth to seventhmillennium BC in the Near East, such as Çayönü,in Turkey or Jarmo, in Iraq. Stone footings wereused as a base for packed mud walls, while stonefoundations in curious ‘grill plan’ formations sup-ported floors made from bundles of reeds coveredin clay (Gates 1976, 83–4).

Another variable is the presence or absenceof foundations set in a trench dug into the sub-soil. Substantial foundations would be superflu-ous for a structure of modest height, or if soundbedrock lay immediately below the surface. On

farmsteads in Northumberland occupied duringthe Roman period, traditional round timberhouses of pre-Roman origin were often rebuiltin stone. Their thick stone walls were erected onthe surface of the ground without creating dis-tinctive subsoil features such as the post-holesof their wooden predecessors. This can create areal problem for excavators, for agriculturalclearance easily remove all trace of stone housesfrom a site. The only traces of stone buildingson a farmstead at Apperley Dene were largeboulders in a surrounding ditch, mixed with lateRoman pottery that also occurred in miscella-neous features elsewhere on the site (Greene1978). At the other end of the architectural spec-trum, the original specification for Hadrian’sWall was so broad that it did not require foun-dations for stability. When the plan was modi-fied during construction, probably to save stoneand mortar, it was reduced from c. 3 to c. 2.5metres thick, and proper foundations were thenconsidered necessary.

When a timber roof resting on stone walls de-cays through age it is possible to remove and re-place it without disturbing the walls; it istherefore difficult to trace the full history of asimple stone building, even when the walls re-main substantially intact. A long chronologicalspan for artefacts from such a structure may bethe only indication of its length of occupation.In contrast, the walls and roof supports of tim-ber houses require frequent repairs that leaveobservable traces in the subsoil. A glance at anystanding medieval building usually reveals in-serted and blocked windows, changes in roofline, extensions and rebuildings stretching overseveral centuries. If nothing but the foundationsand a few of the lower courses survive, this in-formation will have disappeared, leaving onlyfeatures such as blocked doorways or extendedfoundations to hint at modifications. If the ex-cavator is fortunate, further details may be re-vealed by fragments of distinctive masonryshowing architectural characteristics or decora-tive styles that are datable; stonework demol-ished at the end of one period is often reused inthe next. Documents containing building ac-counts, or illustrations made before the build-ing was completely ruined, may be found inarchives. Indeed, these records should have been

Excavation

93

revealed during preparatory research beforeexcavation began and used in the design of anexcavation strategy. This was the case at NortonPriory, in Cheshire, where a medieval monasticcomplex and the secular houses that replaced itwere analysed through an integrated programmeof excavation and documentary research(Greene 1989).

Internal features of stone buildings, such aspartitions, screens or flooring, were frequentlyconstructed from wood, and they are difficult todetect unless parts of their structure extendedbelow ground level. Floors of stone or beaten clayare more helpful because they were frequently laidon a foundation of rubble or other material con-taining architectural fragments from earlierphases, masons’ chippings, and even domesticrubbish. Datable sherds of pottery or coins dis-covered in these layers provide a useful terminuspost quem for the construction of the floor thatcovers them.

Survival of stone structures (fig. 3.27)Much of what has been said so far assumes thatstone structures have survived reasonably well.In reality, buildings that went out of use werenormally treated as quarries, especially if theycontained squared facing blocks. In areas wherestone is particularly scarce even the foundationsmay have been carefully removed, leaving onlyunusable fragments and scattered mortar. Theholes left behind are known as ‘robber trenches’,and they destroy the stratigraphic relationshipsbetween walls, floors and other surroundinglevels in a similar manner to trenches dug by‘wall chasing’ archaeologists. However, sincestone robbers did not shift more than the veryminimum amount of soil in their quest for stone,they normally followed the foundations veryclosely. An excavator can recover a ‘negative’plan of the building by removing the filling ofrobber trenches. Mortimer Wheeler recognized‘ghost walls’ at Verulamium (St Albans, Herts.)and excavated them in this way to reveal one ofthe city’s monumental Roman gateways, ofwhich not a single stone survived (Wheeler &Wheeler 1936, pl. 88a).

Standing structures (Rodwell 1989)In the seventeenth century, John Aubrey was aware

of the value of different forms of window styles fordating medieval buildings, while classical archae-ologists deduced Greek and Roman architecturalstyles from ancient authors and surviving buildingsin Greece and Italy. From the Renaissance to theeighteenth century architects examined and re-corded classical buildings to incorporate details ordesigns into their own work. The Gothic Revivalof the nineteenth century brought medieval build-ings back into favour, and many architects reshapedexisting buildings (especially churches) in a meticu-lously academic fashion—but destroyed genuinemedieval features in the process. The notion thatrepresentative examples of old buildings should bepreserved intact has only reallygained hold in thetwentieth century; ironically, it coincided with aphase of urban renewal that destroyed large areasof historic cities on an unprecedented scale. At the

3.27 Part of a building façade excavated atMeonstoke, Hampshire, where the end wall of part ofa Roman villa complex had collapsed outwards. Thisphotograph shows a set of three windows (ordecorative arcades) constructed from earthenwaretiles that would have made a visual contrast with thesurrounding stonework. Since the walls of mostRomano-British villas rarely survive above theirfoundations, this was the first opportunity to gain anyimpression of the appearance of their upper levels. Italso demonstrates the importance of careful cleaningand observation of stonework in open area excava-tions. If parts of the fallen wall had been encounteredin small trenches, they would probably have beeninterpreted as flooring or a paved surface. BritishMuseum; King & Potter 1990, 198

Excavation

94

same time many rural churches have become redun-dant, while major European cathedrals such as Yorkand Trier have undergone massive repairs. All ofthese factors have encouraged traditional architec-tural methods to be supplemented by archaeologi-cal activities.

The procedure of analysing a standing build-ing relies on principles similar to stratigraphy(Harris 1989, 56–61). The objectives are to es-tablish its original form, to work out the se-quence of later alterations, and to relate themto any available dating evidence. There may bevisible indications of changes such as the block-ing of redundant doorways or the insertion ofnew windows. It is surprising how often exist-ing walls were adapted to bring a building upto date with changes in function or new fashions,rather than being demolished and replaced. PittRivers pioneered modern archaeological meth-ods of examining buildings in his study of amedieval hunting lodge at Tollard Royal on hisestate in 1889, doing such things as strippingwall plaster to reveal earlier decoration andoriginal wall faces, and excavating both withinand outside the building (Bowden 1991, 122–6). Additional techniques are available today.

Samples of mortar compared in the laboratorymay reveal different phases of construction;earthenware tiles or bricks are datable bythermoluminescence (below, p. 123), and theexact age of structural timbers may be deter-mined by counting tree-rings (below, p. 109).

Medieval parish churches in many parts of Eu-rope have provided some of the best examples ofthis kind of analysis, because they are smallenough for comprehensive study (figs 3.28–29).They remind excavators who normally deal onlywith buried structures that many complexchanges can take place in the history of a build-ing without affecting the foundations. The com-bination of excavation with the study of standingremains can also tell architectural historians thatsignificant phases may have come and gone with-out leaving visible traces on the surviving struc-ture (Rodwell 1989). Meanwhile, new techniquesof spatial and functional analysis have brought adeeper understanding of the significance ofchanges made to the plans of buildings that wereused for many centuries. A study of a medievalcastle at Edlingham in Northumberland identifiedchanges in society, economics and the need fordefence that were reflected very clearly by the waythat people of different status gained access tovarious parts of the building (Fairclough 1992).Since this approach has been developed and vali-dated in the context of documented historicalbuildings, it can be extended to the explanationof prehistoric buildings, although the results willalways remain hypothetical.

5.2 Timber structures

Timber only survives where extremely wet or dryconditions have remained constant over a long pe-riod. Otherwise, it decays completely, leaving dif-ferences in the colour and texture of soil that areonly detected by careful excavation. Three exam-ples of archaeological discoveries of boats illus-trate these conditions. One side of the hull of theMary Rose survived in excellent condition nearPortsmouth because the ship sank on its side intodeep silt that protected the timber from erosionby currents and from marine worms that eat ex-posed hulls (Rule 1982). The world’s oldest sur-viving ship was excavated in Egypt in 1954,beside the Great Pyramid of Cheops (c. 2590 BC);

3.28 Churches present some of the most complexproblems of archaeological and architectural analysis,but offer rewarding results (Rodwell 1989). Thisphotograph shows the excavation of structures andburials stretching back over 1000 years in thechurch of St Peter at Barton-on-Humber, Lincoln-shire. Unfortunately, the earliest remains are likelyto have suffered from considerable later disturbance,and they may also be masked by later masonry thatcannot be removed. Dr W Rodwell

Excavation

95

it had been dismantled and placed in a speciallyconstructed stone chamber where arid conditionspreserved it from decay (Jenkins 1980). In con-trast, the Anglo-Saxon ship containing a royalgrave at Sutton Hoo in Suffolk was buried in along trench and covered by a large mound. Al-ternating wet and dry conditions in the sandysubsoil caused complete decay of all but one frag-mentary plank from the bottom of the hull. For-tunately, layers of stained soil marked thepositions of decayed timber, and the outline of the

ship was revealed by meticulous excavation andconfirmed by the positions of corroded iron riv-ets that had originally held the planks together(Evans 1986, frontispiece).

Construction of timber buildings(Brunskill 1985)A knowledge of construction methods is importantif the remains of timber buildings are to be exca-vated and interpreted correctly (see fig. 3.26). Al-though most forms require foundations, it ispossible for substantial structures to be built on thesurface without leaving any traces in the subsoil.Timber-framed buildings and ‘log cabins’ madefrom interlocking timbers rely for their stability onjoinery rather than earth-fast upright posts. Fortu-nately, problems with uneven ground and dampnessusually made it wise to erect these structures on aspread of gravel or rubble make-up, or on low stonewalls; even then, only indirect evidence for the ac-tual timber building may remain. One of the mostimpressive achievements of modern open-area ex-cavation has been the discovery of a phase of lateRoman timber structures built at Wroxeter in Shrop-shire after the Roman baths were demolished(Barker 1993, 207–20).

3.29 A detailed record of one external wall of achurch at Rivenhall, Essex, drawn to emphasizedifferences in the masonry. The remains of distinctivesmall round-headed windows (marked in black)provide evidence that this wall was initially con-structed in the Anglo-Saxon period. It was thenretained throughout the life of the church, and largerwindows were inserted later by knocking holesthrough the original Saxon structure. Because thesesignificant details had been concealed behind wallplaster, the church at Rivenhall was thought to havebeen constructed much later in the medieval period,on the evidence of visible windows (Rodwell 1985).Drawing: Dr W Rodwell

Excavation

96

Earlier excavators had used narrow trenches totrace the plans of Roman masonry buildings, dig-ging through a layer of rubble lying near the sur-face that they dismissed as debris left by thedemolition and decay of the Roman town, dis-turbed by later agriculture. However, open-area(and open-minded) excavation employed by PhilipBarker revealed a hitherto unsuspected final phasein the occupation of the town. The implicationsof his work are universal: wherever timber build-ings might have been erected on rubble levelling,leaving no detectable floor levels, an excavatorcannot claim with complete certainty that build-ings were absent during a phase when a site wasapparently deserted. Unfortunately, this kind ofexcavation is virtually impossible on other Romantown sites, because, unlike Wroxeter, most havebeen disturbed by medieval and modern occupa-tion, and assorted foundations, cellars, wells andpits reduce the area available for excavation. Deepploughing also destroys traces of this kind on manyopen sites (e.g. Corbridge—Bishop & Dore 1989,12); sadly, few places survive where the full ben-efits of the techniques learned at Wroxeter areapplicable.

Barker’s excavations also emphasize the im-portance of recording. The location of a build-ing may be indicated by nothing more than slightdifferences in the size and character of one areaof rubble compared to its surroundings. Arte-facts also require the same kind of scrutiny; frag-ments of flint, pottery or small coins may befound outside the limits of such structures, wherethey were broken or dropped and trampled intothe ground surface. It is possible to study theirexact distribution only if each find has beenaccurately surveyed, and detailed plans mayprovide valuable guidance for interpreting thesite. The significance of some of this informa-tion may only emerge after the excavation hasfinished, thus underlining the principle (firstproclaimed by Pitt Rivers) that everything foundshould be recorded, even if its meaning is notapparent at the time of discovery.

Timber structures that did possess below-ground foundations are not without problems forexcavators. One of the most recurrent archaeo-logical features encountered on sites is the post-hole. The simplest method of erecting a firmupright is to dig a hole, stand a post in the bot-

tom, and to pack the upcast from the hole firmlyback around it, perhaps with the addition of somepacking stones. The subsoil is rarely suitable forlarge posts to be rammed directly into the groundwithout some pile-driving equipment, but thistechnique was certainly in use as early as theRoman period.

Although separate post-holes are commonly dugfor each upright in a large complex structure, acontinuous foundation trench may be more efficientif a building contains lines of regular posts. It is alsoa useful technique if large posts need to be set closetogether; individual holes would tend to intercut,and would be inconvenient during the manoeuvringof heavy timbers. Roman military granaries areamongst the best examples of structures making useof linear foundation trenches (Manning 1981), butthe technique occurs widely. Gaps between indi-vidual posts of houses or fences were frequentlyfilled in with planks, which sometimes extendedbelow ground level, but perhaps only just penetratedinto the foundation trench. Particular care is re-quired to detect these traces during excavation ofthe top layers of the fill. Some walls or fences, andeven large ramparts, consisted of a continuous pali-sade of posts or planks set into a trench. An alter-native to setting vertical posts into the ground is touse sleeper beams. Large timbers are laid horizon-tally, and upright posts are mortised into them toprovide the structure of a building; time and effortsaved in digging post-holes is offset by additionalcarpentry, of course. The horizontal beams mightsometimes be set into trenches or raised on low stonesleeper walls, but they frequently stood on a simplerubble platform; some of the buildings recognizedat Wroxeter were probably of this form. Discussionof the sleeper beam technique merges into stoneconstruction, for stone footings were commonlyused together with timber or half-timbered super-structures.

These various methods of timber constructionmay coexist on different parts of a site, or even in asingle building. Neolithic longhouses of the LinearPottery culture in Europe (see fig. 3.11), Iron Ageround houses in Britain, and early medieval aisledhalls in the Netherlands and north Germany fre-quently combine two or more of these features.Many closely comparable (but unconnected) build-ings have been erected over wide chronological andgeographical spans. A large timber hall excavated

Excavation

97

at Balbridie in Scotland was initially assumed to beof early medieval date because it closely resembledone found on an Anglo-Saxon site in Northumber-land; however, radiocarbon dates revealed that ithad been occupied in the Neolithic period and was

more than 5000 years old (Fairweather & Ralston1993).

Interpreting timber structures (Dixon 1988)(figs 3.30–32)As with stone buildings, interpretation is helped by

3.30 Foundation trench of an Anglo-Saxon timber hall(Structure C12) excavated at Cowdery’s Down,Hampshire, before the construction of a new housingestate (Millett & James 1983). The magnificence ofthis building can be judged partly from its size, andpartly from the lavish use of large timbers. RoyalArchaeological Institute

3.31 Careful excavation within the foundations ofStructure C12 at Cowdery’s Down revealed alternat-ing vertical planks set either side of the wallingmaterial, which was probably woven hurdles. Largerposts stood at the corners, and the side walls wereflanked by large rectangular but-tresses set at an angleto receive the thrust of the weight of the roof. RoyalArchaeological Institute

3.32 Reconstruction drawing of the Anglo-Saxontimber hall (Structure C12) excavated at Cowdery’sDown, Hampshire, based on observations of struc-tural features illustrated in figs 3.30–1. The details ofthe roof are of course conjectural, but its overall formhas been worked out from observations of post-holesand other indications of elements that reached theground. It is likely that a high-status building wouldhave been decorated with elaborate paintings andcarvings similar to those found on surviving Anglo-Saxon metalwork. Royal Archaeological Institute

Excavation

98

studying recent structures that are still standing, orwell-preserved remains of ancient structures pre-served by waterlogging or desiccation. Excavatedremains of burnt buildings are also informative, foronce timber has been converted into charcoal itresists further decay.

The dimensions of post-holes or foundationtrenches are the best guide to the sizes of structuraltimbers and the loads that they were intended tobear; it is very important for excavators to recordthe exact location and measurements of every postby careful observation of the filling (see fig. 3.31).Features that did not support much weight, suchas internal partitions, did not have to penetrateinto the ground, and their traces are unlikely tosurvive on a site that has been damaged by activi-ties such as modern ploughing. This problem caneven affect the walls of a building if the mainweight of its roof was supported on internal pil-lars. A further complication is caused by the per-ishable nature of wood, particularly in temperateclimates. Decay affected timbers at different rates,and wooden buildings required running repairs.Posts were rarely replaced in exactly the same holesas their predecessors, and precise excavation andrecording are necessary if the correct sequence ofintercutting holes is to be sorted out and relatedto any significant dating evidence. When an entirebuilding had to be replaced, a very similar struc-ture was often constructed on almost exactly thesame spot; this could possibly happen several timesin succession. Careful open-area excavation isneeded to recover the separate plans of successivestructures, and to detect crucial points where post-holes or other features overlap, allowing the ex-cavator to establish a stratigraphic sequence (seefig. 3.11).

If large numbers of apparently unrelated post-holes are found on sites that were occupied for longperiods by structures that have left no complete orobvious plans, it may be possible to hypothesizenumerous rectangular or circular structures. Sets offeatures that belonged together may be revealed ifprecise records have been kept of differences in theshape, depth, soil filling, etc., of post-holes and otherstructural evidence. A computer database of theserecords would be particularly suitable for analysisin this way, and possible structures could be plot-ted directly onto a digitized site plan (Bradley &Small 1985).

5.3 Other building materials

Along with stone and timber, one of the most im-portant ancient building materials was clay, whetherapplied directly to wooden walls (wattle and daub),shaped into blocks and dried in the sun (mud bricks;fig. 3.33), or made into bricks or tiles fired in kilns.Building techniques are of course dependent notonly on the availability of suitable raw materials,but also on climatic conditions. The use of handmade bricks of sundried clay began around 8000BC in western Asia, where a move away from hunt-ing and gathering towards food production andfarming had led to a sedentary lifestyle, and theconstruction of permanent buildings. Hand-madebricks were used in circular houses at Jericho (Is-rael) in the ‘prepottery Neolithic A’ phase; rectan-gular mud bricks, shaped in a wooden mould, wereparticularly suitable for rectilinear buildings, andwent on to become the standard building materialof the first urban civilizations of Mesopotamia,Egypt and India. Mud bricks remain the dominantconstruction medium of the region today. Mudwalls were sometimes provided with stone footings,and, where a more temperate climate made theunfired mud building technique less suitable, bun-dles of reeds or timber might be combined with claywalls to increase their strength; Nea Nikomedia inGreece and Karonovo (Bulgaria) in the Balkansprovide good examples of mud and timber construc-tion dated to between 7000 and 6000 BC (Piggott1965, pl. 2a).

Mud-brick buildings have a limited life but areeasily demolished and replaced, so that the build-up of levels over many centuries leads to deeplystratified deposits, such as the characteristic tells ofthe Near East. They are very difficult to excavate,because demolished structures compacted by over-lying levels have a similar consistency to that of anysurviving structural remains. Mud bricks decayrapidly once they have been exposed; many pioneer-ing excavators in the Near East found that monu-mental structures simply crumbled to dust. Moredurable fired bricks form part of the archaeologyof many parts of the world. They were in regularuse in Mesopotamia by 3000 BC, but were mainlyrestricted to the ornate or exposed parts of ceremo-nial buildings. Roman brick buildings make a strongimpression around the west Mediterranean and inEurope.

Excavation

99

Although the emperor Augustus (27 BC–AD 14)claimed to have found Rome a city of brick, andleft it a city of marble, surface appearance is decep-tive: we have already seen that most marble-cladRoman buildings depended for their grandeur onbrick arches and vaulting, backed by the strengthof concrete.

The use of sun-dried clay is not restricted to theOld World; under the name ‘adobe’, it is in com-mon use in Mexico, and is often found in combina-tion with other materials. The vast preColumbiancity of Teotihuacan (Mexico) is famous for anumber of pyramids, made from adobe with a pro-tective facing of stone or plaster—a combinationreminiscent of Mesopotamia. The importance oflocal resources has been stressed throughout this

discussion of stone, timber and other materials; atTeotihuacan, the same phenomenon is found in theprivate houses:

The basic building materials of Teotihuacan wereof local origin. Outcrops of porous volcanic rockin the valley were quarried and the stone wascrushed and mixed with lime and earth to pro-vide a kind of moisture-resistant concrete thatwas used as the foundation for floors and walls.The same material was used for roofing; woodenposts spaced at intervals bore much of the weightof the roof. Walls were made of stone and mor-tar or of sunbaked adobe brick. Floors and wallsurfaces were then usually finished with highlypolished plaster. (Millon 1967, 43)

5.4 Reconstruction(Drury 1982)

An excellent way of increasing understanding of anexcavated building is to attempt to reconstruct thepossible structure in a drawing or a scale model (fig.

3.33 Unfired mudbricks bonded together with wetclay are notoriously difficult to locate and excavate.This scrupulously cleaned section at Tepe Ali Kosh,Iran, makes it perfectly clear why early excavators oftells in the Near East failed to find any structuresother than those finished in fired brick or stone.Hole et al., 1969, pl. 9b

Excavation

100

3.34; see 3.32). Some information may be particu-larly helpful, notably the size and strength of foun-dations, pillars and walls. Fragments ofarchitectural stonework such as window or doorframes, voussoirs from arches or vaulting, or roof-ing slates and tiles, may help to date the building aswell as to reconstruct it. Comparisons should bemade with contemporary structures and relevantdocumentary evidence, but specialist help from anarchitect or engineer may be needed to estimate loadfactors on walls, roof structures or vaulting. Al-though the results are merely informed guesswork,the display of an ancient site for the public is en-hanced considerably by a high-quality scale modelthat includes human figures and activities relevant

to its function. Excavators also benefit, for the scru-tiny of alternative reconstructions will require verydetailed analysis of the excavated remains; new in-terpretations may be suggested, and attention isdrawn to parts of the site that need further investi-gation. If several plausible reconstructions are de-duced from a single plan it is best to offer more thanone possible interpretation in the excavation report.Shakespeare’s Globe Theatre in London providesan excellent example of the difficulties involved inrelating excavated foundations to an above-groundtimber structure, even when its form is known fromcontemporary illustrations (Blatherwick&Gurr1992).

The elaborate nature of excavated remains ofbuildings is easily underestimated, especially if noth-ing more than post-holes remains. However, thecomplexity of joinery and the elaboration of purelydecorative detail that survive on many medieval andlater timber buildings must have existed in earliertimes. Anthropological analogies point in the samedirection. The timber buildings, boats, textiles andeveryday objects of the native Americans of thePacific coast of Canada still bear elaborate paintedand carved designs derived from animals, birds andfish. Norwegian stave churches provide a goodEuropean parallel; if none had survived, who wouldhave believed an archaeologist’s reconstruction ofthe ornate superstructure and intricate carved or-namentation of Urnes church from its simple groundplan?


3.34 Excavators rarely have an opportunity to testtheir interpretation of an excavated structure in a full-size reconstruction, rather than a drawing. The gateof Arbeia Roman fort (South Shields, Tyne and Wear)is based on extensive research into military architec-ture throughout the Roman Empire, but it is not clearwhether the towers would have had tiled roofs orbattlements. Although many significant aspects maynever be resolved, the value of the structure fortourism and education is very important, for therooms provide exhibition space, and the rest of thesite can be understood much more easily when seenfrom the parapet. Because the building had toconform to modern planning and building regulationsto allow public access, it was not possible to create anauthentic interior. The cost of materials was too greatto allow solid stone and mortar walls in the Romanmanner; cheaper concrete breeze-blocks lie behindthe sandstone facing. Paul Bidwcll

101

Dating is the key to organizing all archaeologicalevidence. Furthermore, the development of datingmethods, whether ‘traditional’ or scientific, illus-trates the ingenuity and lateral thinking that makearchaeological problem-solving such a fascinatingexercise. This chapter will examine the use of his-torical evidence and some methods of relative dat-ing based upon artefacts before looking at scientifictechniques. As in earlier chapters, the historicaldevelopment of the subject will be stressed at vari-ous points. It is interesting to see how tree-rings,varves and pollen analysis were used to constructrelative and absolute dating for prehistory in the firsthalf of the twentieth century. The intellectual ap-peal of this pioneering work is just as attractive asthat of more recent methods.

1 Background

Chapter 1 described how the biblical accounts ofthe Creation, the Flood and the peopling of theworld were gradually eroded, and how, by the1860s, scientists had undermined Bishop Usher’sdate of 4004 BC for the Creation. An awarenessof geological time scales, combined with Darwin’sconcept of evolution, emphasized the slow andgradual nature of developments in human socie-ties and artefacts. Prehistoric time could be subdi-vided with growing confidence and artefacts couldbe subjected to more detailed classification. Whileobservations of geological and archaeologicalstratification and contexts provided evidence forsequences of fossils and artefacts, they only placedthem into a correct relative order. Absolute datingremained firmly in the hands of archaeologistsworking on the literate civilizations such as Greeceor Rome. The scope of historical dating was ex-tended to Egypt and the Near East when theirscripts were deciphered in the early nineteenthcentury. By then, Thomsen had already used ar-chaeological finds in Scandinavia to validate the

concept of three successive ages of stone, bronzeand iron, but these remained essentially undatablebefore Roman imports appeared in Iron Agephases.

At the beginning of the twentieth century it muststill have been inconceivable that reliable datescould ever be established for European prehistory,other than those that depended on tenuous connec-tions between Egypt and the Aegean in the secondmillennium BC. Dating began later in most otherparts of the world; apart from South America, In-dia, China and other parts of the Far East whereliterate civilizations existed, dating began with thefirst contacts between native peoples and Europeanexplorers and colonizers. Not until 1950 did abso-lute dates become a reality for prehistoric archae-ology in areas outside Scandinavia and thesouth-west of the United States, where varves andtree-rings had begun to provided a locally applica-ble dating method some decades earlier.

The dating of sites by stratigraphy was examinedin Chapter 3 and the concepts of the terminus postquem and terminus ante quem were explored (p.67). Many of the dating techniques surveyed inchapter 4 are independent of stratification, but it isimportant to stress that they are most valuable whenobjects or samples to be dated come from properlyrecorded, stratified contexts on excavations.

2 Historical dating(South 1977)

Scientific dating techniques have received consid-erable attention since 1950, but their most spectacu-lar successes have tended to affect prehistory. It isimpossible for archaeologists working in historicalperiods to cause such dramatic changes as the de-struction of the accepted framework for datingNeolithic and Bronze Age Europe, or the additionof several million years to the estimated age of tool-making hominids from the Olduvai Gorge in East

4 Dating the past

Dating the past

102

Africa. Prehistorians had already constructed aframework from archaeological sources by the endof the nineteenth century, long before scientific datesbecame available; the introduction of independentdates caused adjustments to the framework, ratherthan complete rebuilding. Archaeologists workingin a historical setting are in a very different posi-tion, for documentary sources have already beenused to establish a framework of dates and culturesinto which they are expected to incorporate ar-chaeological evidence.

However, prehistorians sometimes overestimatethe accuracy and detail of frameworks based onhistorical evidence; in practice, early writtensources may provide little more information thana scatter of radiocarbon dates. The extent of docu-mentation varied considerably in ‘historical’ cul-tures and the information that survives today isdetermined by a variety of factors. People wroteabout a restricted range of subjects in the past; theirsuccessors only preserved what was still of inter-est to them, and they frequently rewrote it fromtheir own point of view, introducing errors andmisunderstandings. Historical writing has onlyrecently attempted to aim at objectivity. It wasnormally written with a clear purpose, either torepresent an individual or regime in a good or badlight (depending on the writer’s point of view), orto convey a particular philosophical or religiouspoint. A number of considerations have to beweighed up before a piece of information con-tained in a historical or biographical account isaccepted: the date and quality of the survivingmanuscripts; the distance (in time and place) of theauthor from the events described; the author’srecord of accuracy on items that may be checkedindependently; the quality of the writer’s sources;and any personal biases or motives for distortingthe truth.

Some documents were written with a clear his-torical purpose, but the value of others is a resultof attention from modern historians and archaeolo-gists. The first category includes narrative histori-cal works or biographies such as those written byTacitus or Bede, as well as the chronicles maintainedin many monasteries in medieval times. Documentswithout a historical purpose include laws, land-charters, wills, accounts, miscellaneous letters andanything else written for use rather than posterity.This kind of material is often preserved today in

archive offices, and it becomes more abundant as itdecreases in age. Post-medieval and industrial ar-chaeologists may find precise dates for sites andstructures in company accounts, building designsand detailed maps.

Historical documents may be discovered in ar-chaeological excavations; thousands of clay tab-lets with cuneiform inscriptions were found inMesopotamia before Rawlinson deciphered theirscript, while everything from the lost works ofGreek poets to gossipy letters, written on frag-ments of papyrus, have been recovered from thedesiccated rubbish tips of Graeco-Roman cities inEgypt. Inscriptions carved on stone were particu-larly important in Egypt and the Greek and Ro-man world, and their content ranges from tersebuilding dedications giving the date and builder’sname, to lengthy historical, religious or legal ma-terial (fig. 4.1). Coins are historical documents ofa kind, and besides dates they sometimes bear shortinscriptions about rulers and events that may notappear in surviving documents. A datable coinprovides an excellent terminus post quem when itis found in a significant stratigraphic position onan excavation (see figs 3.9; 3.12). The uniqueimportance of these kinds of historical evidence isthat they are primary documents that have notbeen copied out many times over the centuries byscribes who might introduce fresh errors at everystage.

Dates derived from historical informationshould be related to sites with care. Sometimes adirect association is established, perhaps by a coinin a stratified sequence, or an inscription from aspecific building. Otherwise, there tends to be atleast one remove between the evidence and thearchaeology, whether it is the use of cross-datingby dated finds, or the identification of placesnamed in texts with remains of sites found by field-work. Cross-dating is used extensively in the studyof artefacts in historical periods. Roman Germanyprovides a good sequence of military sites estab-lished between the late first century BC and thelater second century AD, resulting from advances,retreats and modifications along the Rhine-Dan-ube frontier. Sites of the first century AD are par-ticularly useful, for many new forts were foundedand they may be dated very closely, thanks to theHistories and Annals written by Tacitus towardsthe end of the century. By the early twentieth cen-

Dating the past

103

tury German archaeologists had worked out de-tailed typologies for pottery and other artefacts bycomparing finds from sites of different dates, andthese dates could then be applied to undated sitesin other areas where the same artefacts were dis-covered. In India, Mortimer Wheeler’s cross-dat-ing of sites near Pondicherry in 1945 involvedItalian tableware (Arretine ware) that had beenclassified and dated thanks to its occurrence onearly military sites in Germany (Wheeler, 1954,119–25).

A danger of historical archaeology is that datesmay have a rather mesmerizing effect. If a layercontaining burnt debris and broken artefacts isexcavated, there is an inevitable tendency tosearch the local historical framework for a refer-ence to an invasion or warfare in the region, andto date the excavated context accordingly. Unfor-tunately, historical information is patchy even inthe Roman period, and there may have been manyunrecorded episodes that could equally well ac-count for the remains found. In any case, build-ings and even whole forts or towns could burndown accidentally; it happened to London in1666. If an excavated context and the artefactsthat it contains are dated incorrectly in this way,there is a real danger that cross-dating will applyinaccurate dates to similar artefacts found onother sites.

One of the most precise examples of histori-cal dating is provided by Pliny the Younger’seye-witness account of the burial of Pompeiiand Herculaneum by the eruption of Vesuviusin AD 79. The volcanic deposits that sealedthese cities are a rare example of a terminusante quem, for everything found beneath themmust be earlier than AD 79. Objects found incircumstances that show that they were in useat the time of the eruption (such as potteryvessels left on a table) are particularly welldated, but finds from uncertain contexts couldbe several hundred years old. Thera, a BronzeAge city on the Greek island of Santorini in theAegean, has been compared to Pompeii becauseit was buried by an even more cataclysmic vol-canic eruption. The excavator dated the de-struction to around 1500 BC, ultimately on thebasis of cross-dating to historical records inEgypt, and the same eruption was thought tohave destroyed Minoan palaces on Crete, pro-

viding a valuable dating horizon for the AegeanBronze Age.

The analogy between Thera and Pompeiiproved to be misleading, for scientific tech-niques have produced a series of conflictingdates that still cause argument. On the whole,scientific evidence now favours an earlier datefor the eruption and fails to support any con-nection with events in Crete (Hardy & Renfrew1990). Many archaeologists now relate thedestruction of Thera to a major volcanic epi-sode that had been noted in Greenland ice-coresand has subsequently been dated by tree-ringsto 1628 BC (below, p. 114). Considering thewarning (issued above) about the danger ofusing historical dates, it is interesting to note

4.1 This stone slab, which is just over one metre long,is a primary source for dating the construction ofHadrian’s Wall. It was found in the 1750s at the site ofa milecastle that formed part of the original plan forthe Wall, and probably once adorned its gateway. Itwas common for this kind of dedication slab to becarved to mark the completion of a Roman building.The inscription reads IMP CAES TRAIAN HADRIANIAUG LEG II AUG A PLATORIO NEPOTE LEG PRPR, which can be translated as ‘This work of theEmperor Caesar Trajan Hadrian Augustus [was builtby] the Second Legion Augusta under Aulus PlatoriusNepos, propraetorian legate’ (RIB 1638; Collingwood& Wright 1965, 520). It associates the Wall not justwith Hadrian himself, but with Nepos, who wasgovernor of Britain from AD 122–6, and shows thatthe first phase of the frontier structure had alreadybeen completed early in Hadrian’s reign (AD 117–38). Museum of Antiquities, University of Newcastle uponTyne

Dating the past

104

that scientific dating may have now the sameeffect: ‘Any sloppily dated archaeological event,within a century or so, tends to be “sucked in”to the precisely dated tree-ring events. We allhave to be on our guard against circular argu-ments’ (Baillie 1989, 313). Ultimately archae-ologists and historians share the same generalobjectives; the principal contrasts lie in thekinds of evidence that they explore, and thedifferent aspects of the human past that they areable to address most successfully with the ma-terial or documentary information available tothem.

3 Typology(Graslund 1987)

Although Pitt Rivers was an early exponent oftypology, his ideas of its universal validity weretoo abstract to have any chronological promise(see fig. 1.12). In Sweden, Oscar Montelius ad-vanced typology into the realms of firmer dat-ing by producing comprehensive publications ofEuropean artefacts from the 1880s. He soughtassociations between artefacts of different formsburied together, such as grave-goods in indi-vidual burials, or collections of objects buriedin ritual deposits. Each form of artefact was clas-sified in a type-series, and the sequence of findcontexts normally confirmed progress from sim-plicity towards greater elaboration or efficiency(figs 4.2–3). These procedures are perfectly ac-ceptable today.

A third technique used by Montel ius ,cross-dating (or synchronism), was entirelylogical in theory, but, in retrospect, has beenvery misleading. In its strongest form, cross-dating takes account of artefacts made in his-tor i ca l ly da ted areas , such as Egypt orMesopotamia, found in associat ion withother artefacts made in undated areas. For ex-ample, in 1891 Flinders Petrie found potteryfrom Crete on Egyptian sites, in contextsdating to around 1900 BC (fig. 4.4). He sub-sequently ident i f ied Egyptian exports atMycenae in mainland Greece that could bedated to c. 1500 BC (Drower 1985, 182–5).Thus, dates derived from Egyptian historicalrecords were extended to sites and cultures inCrete and Greece that lacked internal dating

evidence. An obvious limitation was that no his-torical dates extended beyond 3000 BC, so thatthe age of earlier artefacts could only be guessed.Whereas Petrie’s links were based on direct asso-ciations with Egyptian material, Montelius ex-tended cross-dating right across Europe intoBritain and Scandinavia by noting associationsbetween artefacts found far from their area ofmanufacture and local types. These fixed pointsallowed type-series of different areas to be inter-locked, but unfortunately every step away fromEgypt increased the possibility of a weak link inthe chain (see fig. 6.1).

Two further criticisms are apparent: objects im-ported from distant sources may have been treas-ured for long periods before being lost or buried in

4.2 In an explanation of his methods for studyingtypology, Montelius illustrated the transition of theaxe head from stone to metal. The first copper axes(top row) were very similar to their stone counter-parts (extreme left), but it was soon realized thatmetal could be saved by making them thinner, whileincreasing their effectiveness by hammering out awider cutting edge (bottom row). Montelius 1903, 22

Dating the past

105

association with local items; and superficially simi-lar artefacts found in different areas may be uncon-nected and not contemporary at all. However,confidence in Montelius’ cross-dating was increasedby the assumption that all cultural advances inEurope were inspired by the civilizations of the

Aegean and Near East. The effects of this‘diffusionist’ view survived into the 1960s, whenradiocarbon dates suddenly snapped the chain ofconnections into many unrelated links.

Typology has not been superseded, but radio-carbon dates have reduced the burden of prehis-toric chronology that it was once made to carry

4.3 Some technical reasons for further modificationsof Bronze Age ‘celts’ (axes) were explained by PittRivers in 1875: ‘…the celt of the neolithic period,chipped only at first and subsequently polished…gaverise to the copper celt of the same form havingconvex sides, which grew into the bronze celt withflat sides. Then the bronze celt was furnished with astop to prevent its being pressed too far into thehandle by the blow. Others were furnished withprojecting flanges to prevent them from swerving bythe blow when hafted on a bent stick. Others hadboth stops and flanges. By degrees the flanges werebent over the stops and over the handle, and thenthe central portion above the stops, being no longerrequired, became thinner, and ultimately disappeared,the flanges closed on each other, and by this means

the weapon grew into the socket celt. On this socketcelt you will see that there is sometimes a semicircularornamentation on each side. This…is a vestige of theoverlapping flange of the earlier forms out of which itgrew, which, like the rings on our brass cannon, aresurvivals of parts formerly serving for special uses.’(Lane-Fox 1875, 507) The development of copperalloy axes ended at this point, for the introduction ofiron from c. 1000 BC provided a superior metal foredge tools, with radically different manufacturingtechniques. Audio Visual Centre, University of Newcastle,after Smith 1920

Dating the past

106

(fig. 4.5). Type-series remain an extremely usefulmeans of describing and classifying artefacts ofany period, and for understanding their technol-ogy and function. Radiocarbon dating (below, p.115) now allows the typology of prehistoricbronze objects to be checked independently, forremarkably precise results can be derived usingthe AMS technique from small fragments ofwooden handles or shafts that occasionally sur-vive in the sockets of spears or axes (Needham1986). Association and cross-dating are still im-portant in the historical period; Roman metal-work, pottery, glass and coins were traded toScandinavia, Central Europe and even India,where they provide valuable dates when foundwith local artefacts. In the early medieval Migra-tion period (fourth to sixth centuries AD), typo-logical analyses of brooches and buckles linkedto Germanic peoples (such as the Goths, Huns orFranks) are still important in the study of ‘bar-barian’ settlements within the former Roman

4.4 Cross-dating by pottery: Arthur Evans usedimported Egyptian artefacts to date his excavation ofthe Palace of Knossos in Crete (fig. 1.15). LocalCretan pottery found on this site could also be datedbecause similar sherds had been found in Egypt. A, Band D are from Crete and bear decoration of Evans’Latest Middle Minoan II Phase, while C was found atKahun in Egypt. Evans 1921, f ig. 198

4.5 Unlike Bronze Age axes, the shapes of harpoonpoints used by hunters in Britain after the end of thelast Ice Age show no clear typological development.However, small samples of the bone from which theywere made can now be dated by the AMS radiocar-bon technique, and these dates may be used to placethem into chronological order. Smith 1992, f ig. 1.2

Dating the past

107

Empire (Greene 1987). It is important to realizethat cross-datings and associations obey the sameprinciple as a terminus post quem in an excava-tion (p. 67); dated finds only establish fixed pointsafter which the contexts that they were discov-ered in must be dated.

4 Sequence dating andseriation

These dating techniques rely on careful exca-vation and recording, for they both place as-semblages of artefacts into relative order. Petrieused sequence dating to work back from theear l ies t h is tor ica l phases of Egypt intoPreDynastic Neolithic times, using grave-groups that could be assumed to consist of con-temporary artefacts deposited together at asingle time (Petrie 1899; Drower 1985, 251–4).Decisions were made about ‘early’ and ‘late’ ar-tefacts in graves by typological judgementsabout their form. Grave groups were then ar-ranged in a sequence according to their combi-nations of artefacts of early or late character,in a kind of ‘simultaneous typology’ thatweighed up the development of every itemfound in each grave. Petrie drew graphs of pot-tery types that occurred in his sequence of fiftypre-dynastic phases, and showed that types didnot appear and disappear abruptly, but becamepopular gradually before declining equallygradually (1920, pl. L). This phenomenon isconfirmed by the ways that modern clothingand jewellery go in and out of fashion at dif-ferent rates.

Seriation is based on the same principle, andit has been applied to finds from grave groups,strata or other kinds of assemblages, whetherfound on individual sites or over a wider area. Itworks best on assemblages that contain a rangeof definable characteristics, such as types of pot-tery or flints, especially those that are subject tochange rather than continuity. The numbers ofselected artefact types found in each assemblageare converted into percentages to make themcomparable. The figures are then arranged intothe best possible sequence on the assumption thatthe percentages of artefacts will have increased

and declined in an orderly manner. This processmay be carried out by eye if the percentages aremarked on individual strips of graph paper torepresent each assemblage and shuffled to find thebest fit. Random statistical variations and possi-ble differences in the character of the assemblagesthat are being compared make it very unlikelythat the results will form perfect ‘battleshipcurves’. Seriation is only a relative dating method,but it remains useful in the study of finds that donot occur on stratified sites where the sequenceis revealed by excavation; like artefact typologies,it is now used within an absolutely dated frame-work. Petrie’s desperate, but inconclusive, at-tempts to establish an absolute date for thebeginning of his prehistoric Egyptian sequenceunderline the enviable position of modern archae-ologists (Petrie 1899, 4–6).

5 The advent of scientificdating techniques(Zeuner 1946)

For Prehistory, no calendars are available. Upto not many years ago, the time-scales sug-gested for the evolution of early man and hiscultures were pure guesses, not to say imagi-nation. From a scientific point of view theywere worthless. (Zeuner 1946, 1)

This quotation underlines the complete transfor-mation of archaeological dating that began around1950 and continues to this day (figs 4.6–7). How-ever, archaeologists tend to forget that geology hadalready undergone a revolution in scientific dat-ing during the first half of the twentieth century.Seen in the context of the development of datingmethods over the previous century, radiocarbondoes not seem quite as dramatic as it is sometimesportrayed.

Frederick Zeuner’s book Dating the Past: An in-troduction to geochronology (first published in 1946)integrated geological dating with archaeology in anexemplary manner. The text was updated and ex-panded several times up to 1958, by when Zeunerwas able to document the introduction of new tech-niques such as radiocarbon and potassium-argondating. Because it gives such a vivid impression of

Dating the past

108

the difficulties and triumphs of archaeological dat-ing as it emerged from the nineteenth century,Zeuner’s book provides an excellent companion—and contrast—to Aitken’s ‘state of the art’ survey,Science-based Dating in Archaeology (1990). WhileAitken organized his book according to the scientificbasis of each dating technique, Zeuner had adopteda very different approach that began with techniquesapplicable to the recent past and worked back to-wards measurement of the age of the Earth. My ac-count will group methods together according to theirscientific basis, but it will begin with some of the ear-liest methods to emerge into general use in the hopeof retaining some of the atmosphere of discovery thatcharacterizes Zeuner’s writing.

5.1 Geological time scales

Nineteenth-century geologists were preoccu-pied with the age of the Earth, and Darwin’sdemand for gradual evolution underlined thelength of the time scales involved. Glimpses of‘deep time’ could be gained by estimating therate of erosion of geological formations; Dar-win suggested 300 million years to produce the

4.6 The leading scientific dating methods areapplicable to widely differing periods of the past. Eachhorizontal bar indicates the range of an individualmethod; interrupted bars show periods where thepotential is less good. Techniques with the greatesttime-span are not necessarily the most useful: see fig.4.7. Sandra Hooper, after Aitken 1990, f ig. 1.2

4.7 Summary chart of materials that can be examinedby different scientific dating techniques; the bestresults will be obtained from the techniques andsamples with the darkest shading. Thus, wood andother plants respond well to dendrochronology andradiocarbon, but no other techniques are applicable;

conversely, volcanic materials are unsuitable for eitherof these methods, but offer many other possibilities.Archaeologists need an understanding of figs 4.7–8to take the right kinds of samples for dating methods,appropriate to the period with which they areconcerned. Sandra Hooper, after Aitken 1990, f ig. 1.1

Dating the past

109

modern form of the South Downs. Lyell usedthe rate of evolution of certain shells to calcu-late the age of the Earth and arrived at around240 million years for the time that had elapsedsince the appearance of life (Zeuner 1946, 307–8). However, Lord Kelvin’s estimate of as littleas 20 million years, based on the rate of cool-ing of the planet, was widely accepted (ibid.315–16). The problem was solved by a grow-ing understanding of radioactive decay andmeasurement of the rate that uranium decayedto produce lead. From around AD 1900 ArthurHolmes and other scientists used the radioac-t iv i ty method to extend the date ofpreCambrian rocks back to an age of nearly2000 million years (Zeuner 1946, 333).

Thus, estimates of geological time underwenta transition from informed guesswork to scien-tific precision in the fifty years that followed thepublication of Darwin’s Origin of Species in1859. Accurate knowledge of the age of theEarth was of little direct help to archaeologists,but it emphasized the potential of scientific dat-ing techniques. The first half of the twentiethcentury witnessed a similar transition that be-gan with the dating of recent geological periodswhen early humans first lived, and ended withthe introduction of radiocarbon dating. As aresult, by 1950 absolute dates were available forimportant stages of recent prehistory, such as theinception of farming and the first use of metals.

While some geologists concentrated on theage of the Earth, others studied distinctive sur-face traces left behind by changes in the extentof the polar ice cap. They established the exist-ence of a succession of Ice Ages and worked outa sequence of climatic phases based on evidencefor alternations between glaciations and moretemperate conditions. If these could be dated,much could be learned about the emergence ofmodern humans and their interaction with dif-ferent environments. Some calculations wereattempted by measuring the depths of depositsand comparing them with the formation rate ofsimilar deposits in modern times, but there wasno way that these estimates could be checked(Zeuner 1946, 134–5). However, an attractiveidea that had been developed with increasingprecision since the 1780s was that glaciationscoincided with changes in solar radiation and

that these changes were caused by regular andmeasurable variations in the Earth’s orbit. How-ever, the correlation between periods of glacia-tion and periods of low solar radiation remainedhypothetical until independent dating wasachieved with the help of ocean-bed deposits andpotassium-argon dating between the 1950s and1970s (Aitken 1990, 17–19).

6 Environmental methods(Aitken 1990)

6.1 Tree-ring dating (dendrochronology)(Schweingruber 1988)

It had been recognized since at least the fifteenthcentury that trees produce annual growth rings,and their physiology was well understood by theeighteenth century (Schweingruber 1988, 256–7). Well-documented examples of their use fordating begin in North America in the late eight-eenth century; for example, the Reverend Cut-ler counted 463 rings in a tree that had grownon a native American burial mound at Mariettain Ohio and deduced (correctly) that the moundmust antedate Columbus (Daniel 1981, 40–2).Because annual growth rings are subject to sea-sonal factors that affect their thickness, distinc-tive patterns recognized in different samples oftimber may be compared and used to establishtheir contemporaneity (figs 4.8–9). In 1901 A EDouglass had begun to study fluctuations in so-lar radiation and their effects on climate by look-ing at patterns of varying ring thickness in treesin Arizona and his work became inseparablylinked to archaeological dating in the 1920s.Many timbers preserved in pueblos (prehistoricnative American sites in arid areas of Arizonaand New Mexico) could be dated by cross-ref-erencing them to his series (fig. 4.10). An over-lapping series of rings was gradually built upfrom many timbers found on sites in the south-west of the United States, and the sequence wasextended back to the fourth century BC. In 1954,bristlecone pines still growing in California werefound to be as much as 4000 years old, and acombination of specimens from living trees andold trunks preserved in the White Mountains

Dating the past

110

now provides a continuous record back to 6700BC that is of vital importance for checking radio-carbon dates (below, p. 116). An even more im-pressive achievement is the establishment of atree-ring sequence that extends beyond 5000 BC,based on a large number of trees from north-western Europe. Many of the early samples havebeen taken from ancient tree-trunks preserved inpeat bogs. Some earlier ‘floating’ sequences thatcould not yet be linked to absolutely dated tim-bers were dated approximately by ‘wiggle match-ing’ with the radiocarbon curve (p. 117) andextended the range of dendrochronology back toaround 9000 BC. By 1993, the sequence in Ger-many had reached 9494 BC by following pinetrees back beyond a period that was too cold foroaks to grow (Becker 1993).

Tree-rings may also be used for relative datingon waterlogged sites where successive timber struc-tures have been excavated. At Charavines (Isère,France), a large scatter of wooden stakes wasfound preserved on a neolithic village site sub-merged in a lake, but no coherent plan was appar-ent. However, when posts made from trees felledin the same year were plotted, they revealed theplans of two rectangular structures built in succes-sive years (Bocquet 1981). This technique becameeven more valuable when the tree-rings were datedabsolutely. Many studies have been conducted inmedieval buildings—such as the cathedrals at Trierin Germany or Chartres in France—to identify ordate periods of construction that were not fullydocumented in surviving historical records. Ro-man forts and bridges in Germany and the Neth-erlands have also been investigated in the sameway; the precision of tree-ring dating is impossi-ble to achieve by any other means. Once dated, thesites can be integrated into historical accounts ofthe area.

4.9 Dating by dendrochronology: A, B and C aresections from three different trees showing annualgrowth rings that cover a period of 83 years, from theinnermost ring at the left of timber A to the outer-most of C. The overlapping (contemporary) portionsof the timbers can be matched by observing similari-ties in the pattern of their rings, especially whenunusually wide or narrow rings reflect particularlygood or bad growing seasons for the trees. The graphrecords the average annual ring thickness for eachyear, allowing for the fact that the outer rings arealways narrower than the inner because their volumeof wood is spread thinly around a large trunk. Longoverlapping sequences from dated timbers provide areference graph against which individual undatedsamples can be compared. Thus, if this graph began inAD 1000, timber B was felled in AD 1060, and this isa terminus post quem for any structure into which itwas incorporated. Audio Visual Centre, University ofNewcastle

4.8 Apparatus for measuring tree-ring thicknesses.The screen in the centre of the photograph shows aseries of rings from a sample mounted beneath amicroscope and video camera on the left. Individualrings can be measured precisely and recorded by amicrocomputer, while the more sophisticatedcomputer on the right runs programs to match theseries of measurements with sequences of knowndate. Laboratoire de ChronoEcologie de Besançon;photograph by Olivier Girardclos

Dating the past

111

Unfortunately there are many problems in thedirect application of dendrochronological dating.Not all tree species are sufficiently sensitive to dis-play distinctive variations in their ring character-istics, particularly when growing in temperateclimates. Wood only survives under exceptionallywet or dry conditions, and large timbers must berecovered to provide sufficient rings for valid com-parisons, because they rely on patterns that accu-mulated over several decades. Timbers used inbuildings were normally trimmed into regularshapes, removing the evidence for the exact dateof their felling, and they may have been stored formany years before use. Worse still, timbers werefrequently reused several times in repairs or recon-structions of wooden buildings, whose founda-tions rot long before their roof. Re-use is aparticular problem on arid sites, where timbers donot decay easily. Despite these problems, tree-ringsare perhaps the only source of truly absolute dates,in terms of a single year. Unfortunately, they willnever be universally applicable, partly because ofregional and environmental variations in thegrowth of trees, but principally through the rarityof suitably wet or arid conditions that ensure theirpreservation.

The provision of samples of known age for test-ing the accuracy of radiocarbon dates is not the onlyindirect use of tree-rings. Variations in ring widthsreflect climatic conditions, and there are severalinstances of extreme disturbances to normal growth.A series of exceptionally narrow rings indicates anepisode of cold, wet weather from 1159 BC that wasalmost certainly the result of a volcanic eruptionmarked in ice-cores at 1100 ± 50 BC. Ash in theupper atmosphere reduced solar radiation to suchan extent that human settlement patterns and farm-ing practices were disrupted for sufficiently long tocause an abandonment of upland areas of north-ern Britain. Thus, tree-rings provide not only datesfor sites, but also for environmental catastrophesthat influenced changes in human behaviour (Baillie1989).

At a more intimate level, the precision of tree-ring dates adds an exciting dimension to other findsassociated with dated timbers. In the late 1980s tim-bers from the Somerset Levels trackways were tiedin to the dated European series; suddenly, theirconstruction ceased to be ‘somewhere in the earlyfourth millennium’, and became an event that took

4.10 Pueblo Bonito, in Chaco Canyon, New Mexico,is an extensive native American site built mainlyfrom stone but with timber beams, lintels, roofs, etc.Since the arid desert environment had ensured thepreservation of wood for hundreds of years, it wasthe first site to be studied systematically with thehelp of tree-ring dating. A combination ofdendrochronology and architectural analysis revealeda detailed picture of the development of PuebloBonito from c. AD 900–1100 (Judd 1964). The endof a large beam and a horizontal lintel visible in thisphotograph provide useful sources for dating thispart of the structure. Although the latest (outer)tree-ring will provide a terminus post quem (the dateafter which the structure must have been built),there is always a possibility that old timbers were re-used from an earlier phase of the building. Theobvious blocking of the doorway suggests that thefunction of these rooms changed during the life ofthe building. Neil Judd, 1926; copyright: NationalGeographic Society

Dating the past

112

place in 3807/6 BC. This precise date extended toother finds, such as tools and pottery, found in thesame context (Hillam 1990). The impact is similarin historical periods; it is now known that a woodengrave chamber erected within the famous Vikingship burial at Oseberg in Norway was constructedfrom trees felled in AD 834 (Bonde and Christiansen1993).

6.2 Varves

Every summer, the melting of glaciers causes ero-sion by streams and rivers, and the resultingsediments are eventually deposited on lake beds(fig. 4.11). The sediments become sparser andfiner as the year progresses, for the flow of wa-ter is reduced when temperatures begin to fall;winter freezing then stops erosion until the nextsummer. Sections cut through lake beds in gla-cial regions reveal a regular annual pattern ofcoarse and fine layers, known as varves. Varia-tions in climate produced observable differencesin the thickness of sediments, and, like the pat-terns of variation in tree-rings, this allows com-parisons to be made between deposits in separatelake beds. Varves had been recognized and un-derstood as early as the 1870s in Sweden. From1905 onwards, Baron de Geer carried out exten-sive fleldwork with the aim of establishing a con-tinuous sequence from overlapping depositspreserved in beds of the hundreds of lakes thatformed during the retreat of glaciers after the lastIce Age. Whereas tree-rings can be counted backfrom a tree felled today, de Geer lacked a securefixed point at the end of his sequence. However,a lake known to have been drained in AD 1796gave an approximate pointer, and he publisheda sequence covering around 12,000 years in1910.

Varves allowed the end of the last Ice Age tobe dated with confidence to around 6800 BC andprovided the first extension of ‘calendar’ datesinto European prehistory. They also made it pos-sible to date individual sites if their positionscould be related to former lakes or seashores.Work on varves continues, particularly in NorthAmerica, and it may one day be possible to tiethe sequence of Scandinavian varves to some ar-eas of the New World. Varves also contributeinformation to archaeomagnetic dating, for they

contain a record of the Earth’s magnetic field intheir iron-rich clay particles (below, p. 121). Evenmore important, until radiocarbon dating wasintroduced after 1950, varves provided the onlymethod that could be used to date the climaticsequence revealed by changes in vegetationknown from pollen analysis.

6.3 Pollen analysis(Dimbleby 1985)

Microscopic wind-blown pollen grains survivewell in many soil conditions, and the ease of dis-tinguishing different plant species is of consider-able value in the study of past environments (p.143). Pollen that has accumulated in deep depos-its such as peat-bogs supplies a sequential recordof changes in vegetation since the last Ice Age, forvariations in temperature and rainfall resulted inperiods of markedly different plant and treepopulations in the past. Work on pollen began inScandinavia in the 1920s and it confirmed thegeneral pattern of climatic change that had beenproposed from visible plant remains. Fortunately,since these changes were also reflected by varves,each distinctive climatic phase could be dated.

The value of this technique for archaeology layin the fact that climatic phases were likely to havebeen fairly uniform throughout northern Europe.Thus, plant species found in a sample of pollenfrom an archaeological site could be fitted into theclimatic sequence. Correlations could be estab-lished between sites belonging to similar climaticphases in different countries, and this form ofcross-dating did not have to rely on dubious linksbetween artefacts. However, even individual arte-facts could be dated if they were found in peat-bogs, or if they had sufficient soil attached to themfor the identification of pollen. For example, amesolithic bone harpoon dredged from the bottomof the North Sea was placed into the period whenpine was declining in favour of trees that preferredwarmer conditions around 7000 BC (Zeuner 1946,91–2).

A further benefit of dating sites and artefacts toclimatic phases was that new insights could begained into their environmental context. The sig-nificance of the location of settlements was in-creased by understanding the state of contemporaryvegetation and the landscape, and the functions of

Dating the past

113

tools also took on more significance. Since climaticzones established from pollen have been dated ab-solutely by radiocarbon, they are no longer requiredas chronological indicators; nevertheless, pollenanalysis continues to supply important evidence forthe interpretation of the ancient environment (be-low, p. 143).

6.4 Sea-bed deposits

An approach to geological dating analogous tovarves was developed during the 1950s. Deepsediments exist on the sea-bed, representing a slowaccumulation of shells and skeletal material fromdead marine creatures. Cores (typically 10 metresin depth) extracted from these deposits reveal vari-ations in oxygen isotopes in the shelly material,caused by fluctuations in the volume of the oceanthat reflect global temperatures and ice ages. Apattern of climatic variation is derived from tem-perature-sensitive species of marine fauna and frommeasurements of oxygen isotopes. It correlates withgeological evidence for cold and warm periods thatare dated according to deviations in the Earth’s orbitaround the sun (above, p. 109). Sea-bed sedimentsalso contain material derived from the erosion ofland containing iron particles; their magnetic align-ment has been measured to produce a dated se-quence of changes in the Earth’s magnetic field,which undergoes complete North-South reversalsfrom time to time. As a result of studies of deep-seacores, geologists and archaeologists interested in theearliest stages of human development now possessan integrated dated record of global temperatureand magnetism. Thus, if bones or tools associatedwith early hominids, such as the famous series fromEast Africa, are found in geological deposits relatedto periods of extreme temperature or magnetic re-versals, they are now datable (below, p. 128).

6.5 Ice-sheet cores

Yet another form of dating based on a cumulativenatural phenomenon has been developed by clima-tologists who have extracted cores from the icesheets of Greenland. Each winter’s snowfall createsa distinct layer, and the annual layers have beencounted back almost 6000 years in a core more than2 kilometres in depth, with an excelent level of re-liability within around 50 years (standard deviation

±10). The thickness of each layer varies, as do theproportions of different oxygen isotopes whoseformation is known to reflect temperature; thus,long-term patterns of variation reflect changes inclimatic conditions. A further factor of value fordating is the recognition that even when the indi-vidual layers are no longer distinguishable by eye,they contain annual fluctuations in dust and acid-ity that have extended the annual record back al-most 10,000 years, at which point the layers becometoo thin for counting (Aitken 1990, 23).

Some layers of ice contain high levels of dustand acidity caused by volcanic eruptions (fig.4.12). Volcanoes known from historical records,such as Krakatoa (1883) or Vesuvius (AD 79), canbe correlated with ice-cores; further undocu-mented eruptions in prehistoric times may alsobe detected. Ideally, prehistoric eruptions datedby ice-cores would provide precise dates for sites,especially when calibrated with tree-rings, whichmay show abnormal growth patterns caused byvolcanic disturbance of the climate. The massiveeruption that destroyed much of the island of

4.11 These cores bored from sedimentary lakedeposits in Sweden show distinctive varves; each bandof light to dark silt marks a single year’s deposition ofwater-borne sediment. Varves vary in thickness froma few millimetres to several metres; these averageapprox. 2.5 cm. Prof. D Tarling, University of Plymouth

Dating the past

114

Santorini in the Aegean should probably be linkedto signs visible in ice-core and tree-ring data from1628 BC, but many interesting conflicts betweenarchaeological and scientific dating remain unre-solved (Hardy & Renfrew 1990). Akrotiri, aBronze Age town on the island, was buried un-der deep volcanic ash, and the same ash also fellon Crete and Turkey, offering the potential fordating sites over a wide area. Volcanic eruptionsmay also give indirect dates for wider changes insettlement patterns, for ash in the upper atmos-phere may cause severe disturbances to theweather, and if these circumstances were pro-longed for many years they could lead to theabandonment of adversely affected sites (below,p. 183).

Finally, ice-cores and varves provide an addi-tional way of checking the reliability of radiocar-bon dating in periods beyond the range of samplesfrom precisely dated tree-rings. Abrupt signs of cli-matic change dated by ice-cores and varves to

around 8750 BC, are underestimated by approxi-mately 700 years by uncalibrated radiocarbon dat-ing.

7 Absolute techniques (Aitken 1990)

The proper meaning of absolute dating isthat it is independent of any other chronol-ogy or dating technique, that it is based onlyon currently measurable quantities. (Aitken1990, 2)

We have seen that, by 1950, a number of datingtechniques had emerged that could offer chrono-logical frameworks for the study of prehistory atleast as reliable as those used by historical archae-ologists. Unfortunately, all required special cir-cumstances, such as the survival of timber fortree-rings, the proximity of glacial lakes forvarves, or the existence of soil conditions thatfavoured the preservation of pollen. However, thesuccessful development of dating methods forgeological periods, whether they relied upon ra-dioactive decay or variations in the Earth’s orbit,offered the possibility that a similar, generallyapplicable, technique might one day be found thatwould give absolute dates for prehistoric archae-ology.

7.1 Radioactive decay

Several scientific dating techniques exploit thephenomenon of radioactive decay, including thosefirst used to date the age of the Earth in the earlyyears of the twentieth century (above, p. 109).Many elements occur in different forms, and someare unstable; these isotopes have extra neutronsbesides their standard number of protons andthey are designated by a number representingtheir atomic weight (carbon-14 or 14c). Unstableisotopes are radioactive and emit rays of parti-cles at a known rate. Some isotopes become sta-ble after emitting these particles, while others(such as uranium) go through a protracted seriesof ‘daughter’ elements before reaching a stableform (e.g. uranium to lead: p. 123 below). Thespeed of decay is expressed as the half-life, thetime taken for half of the total radioactivity todecay; this may vary from seconds to millions ofyears.

4.12 Major volcanic eruptions affect the atmosphereby emitting large quantities of acidic ash; it may berevealed by abnormal acidity in layers within corestaken from deep ice-sheets in Greenland. Even whenthe annual layers are not clearly visible, the pattern ofyearly temperature variations is indicated by changesin oxygen isotope levels. Here, an eruption that leftits mark around 1644±20 BC was almost certainly thesame event that caused damage to trees in ringsdated to 1628 BC; it is usually assumed to have beenthe explosion of Thera in the Aegean. Sandra Hooper,after Aitken 1990, f ig. 2.10

Dating the past

115

7.2 Radiocarbon dating|(Bowman 1990)

Amongst the numerous peaceful by-products ofaccelerated wartime research into atomic physicsand radioactivity in the 1940s was radiocarbondating. The rate of decay of carbon-14, which hasa half-life of 5730 years, is long enough to allowsamples of carbon as old as 70,000 years to con-tain detectable levels of radioactive emissions, butshort enough for samples from periods since thelate Stone Age to be measured with reasonable pre-cision. However, the feature of carbon-14 thatmakes it exceptionally important is that it is ab-sorbed naturally by all living organisms, but ceasesto enter them when they die (fig. 4.13). In theory,all that needs to be done is to measure the radio-activity of a sample from a dead animal or plantand to calculate from the level that remains thetime that has elapsed since its death. The practi-calities of age estimation are rather more compli-cated, and the discussion that follows will attemptto highlight the principal advantages and disad-vantages of carbon-14 rather than to provide acomprehensive account.

This simplified description does not do justiceto the inspired formation and testing of hypoth-eses carried out by Willard F.Libby in Chicago inthe 1940s for which he received a Nobel Prize in1961. However, the publication of his preliminaryresults in 1949 was only a beginning. By a happychance, the period of the past where it promisedto be most effective from the outset was one ofparticular significance to prehistoric archaeolo-gists, for it encompassed the transition from hunt-ing and gathering to farming, and the emergenceof the first civilizations. There are now more than80 radiocarbon laboratories all over the world andupwards of 30,000 archaeological dates have beencalculated. Accuracy and precision are improving,and the introduction of Accelerator MassSpectrometry laboratories in the 1980s allowedvery small samples, one hundredth of the size re-quired in the 1950s, to be dated. AMS allows di-rect dating of actual artefacts and bones, ratherthan just the contexts where they have been found.It therefore offers particularly exciting prospectsin early prehistory, for example in dating fragmentsof fossil bones associated with the disappearanceof Neanderthals and the appearance of modern

humans in Europe and Asia between 50,000 and30,000 years ago (Stringer 1986).

Radiocarbon dating has grown exponentially,and many problems and inaccuracies have beenisolated and examined, some leading to major ad-justments of the results. Despite many problems,radiocarbon dates now provide a framework for theprehistory of the world; for the first time, its studyhas become more like that of historical periods, andemphasis has shifted away from pure chronologytowards more fundamental social and economicfactors.

Positive factors• radiocarbon dating is universal, because the ra-

dioactive isotope carbon-14 is formed continu-ously throughout the Earth’s atmosphere by theeffects of cosmic radiation.

4.13 This drawing by Robert Hedges illustrates thebasis of radiocarbon dating with unusual clarity. Thearrows follow the formation of the carbon-14 isotopein the atmosphere by cosmic radiation, and itsincorporation into a tree through photosynthesis ofcarbon dioxide. It then passes to a deer that haseaten the foliage, but this animal ceases to take infresh carbon-14 when it dies. Thus, a bone is placedat the top of a graph that shows the steady decline ofthe radioactive isotope over time. Research Laboratoryfor Archaeology, Oxford University

Dating the past

116

• carbon-14 has a known half-life, and decays ata constant rate.

• the rates of formation and decay are in balance;cosmic radiation in the past should have main-tained carbon-14 and the other isotopes of car-bon in the atmosphere at constant levels.

• all life-forms contain carbon, and living organ-isms absorb carbon from the atmosphere,mainly in the form of carbon dioxide photo-synthesis by plants is one common mechanism.Animals and plants therefore maintain the sameproportion of newly formed carbon-14 as theatmosphere until their death, when it begins todecay.

• dendrochronology provides an independent‘benchmark’ of dated samples of wood fromannual tree-rings stretching back nearly 10,000years. The initial source was bristlecone pinetrees found in the south-west of the USA; someare still growing after more than 4000 years.Older samples come from dead trunks that hadresisted decay in this semiarid habitat, and fromtrunks of oak trees preserved in bogs or riversediments in Europe.

Thus, if a sample of ancient wood, charcoal orother organic matter is processed in a laboratoryso that carbon is isolated, the amount of radioac-tivity that remains can be measured; the older itis, the fewer radioactive emissions of beta-parti-cles will occur in a period of observation. Tengrams of modern carbon-14 produce 150disintegrations per minute; the age of an ancientsample of the same weight that produced only 75counts should therefore be equal to the half-life ofthe isotope, around 5730 years.

The measurement of radiocarbon requireshighly accurate laboratory equipment to keepthe margins of error within reasonable limits. Bythe 1950s the technique had moved rapidly fromusing solid carbon to gases such as carbon di-oxide, and in the 1960s liquid scintillationcounting joined gas counting. All of these radio-metric methods require some means of detect-ing natural cosmic radiation that may penetratethe apparatus, to ensure that only radioactiveemissions derived from the sample itself are re-corded. A new technique, accelerator massspectrometry, was developed in the late 1970s;AMS is fundamentally different because it meas-

ures the concentration of carbon-14 in a sample(relative to carbon-12) rather than its radioac-tivity.

Negative factorsSeveral aspects of radiocarbon dating require care-ful examination to achieve a correct understand-ing of the interpretation of its results. Some ofLibby’s original assumptions were incorrect, andthe method of calculating dates has been revisedseveral times since the technique began to be em-ployed.

• the half-life has been shown by more accuratemeasurement to be too low by around 3%; itis now judged to be around 5730 years, ratherthan 5568.

• different isotopes of carbon are taken into or-ganisms at different rates (fractionation); theproportions of carbon-13 and carbon 14 mustbe checked and an adjustment made to the es-timated date.

• he level of cosmic radiation has fluctuated overtime, perhaps in relation to sunspot activity andthe Earth’s magnetic intensity. This means thatthe formation of carbon 14 in the atmospherehas varied; thus, samples from organisms thatabsorbed abnormally larger or smalleramounts of carbon 14 will give misleadinglyyounger or older dates.

• a calibration curve must be used to convert‘radiocarbon years’ into calendar years (fig.4.14). Tree-rings have not only revealedshort-term fluctuations in carbon-14 levels,but also a divergence between carbon 14dates and ‘real’ calendar years that becomesincreasingly serious before c. 1000 BC. Sam-ples with a radiocarbon age of 5000–7000years require upward adjustment of as muchas 500–1000 years, and this trend increasesas dates extend further back in time; at apoint when Uranium-Thorium dating meas-ures coral as being around 30,000 years old,its age in radiocarbon years is only around26,000.

• calibration reveals that dates from thesouthern hemisphere are around 30 yearstoo old compared with those from theNorth; this is probably because the greater

Dating the past

117

area of oceans has affected the distributionof carbon 14 in the atmosphere.

• a statistical estimation of error, expressed as astandard deviation, is attached to laboratorycounts of radioactivity. Since isotope decaysoccur at random, a reasonably long countingperiod is needed to reduce this inherent error.Several counting sessions of the same fixedlength are normally carried out and the rangeof differences between the separate results isconveyed by a figure that follows the date,preceded by ‘±’. Fig. 4.15 shows how the reli-ability of a date should be envisaged.

4.14 Pearson’s tree-ring calibration curve forradiocarbon dates. The straight line shows what therelationship would have been had the amount ofcarbon-14 in the atmosphere had remained constantover the last 10,000 years: i.e. 4000 radiocarbonyears would be equivalent to c. 2000 BC. However,beyond 500 BC there is an increasing divergence, sothat a radiocarbon age of 8000 years has to beincreased by almost 1000 years, from c. 6000 to c.7000 BC. The process of calibration looks decep-tively simple at this scale, but the ‘wiggles’, com-bined with other statistical uncertainties, makecalculations much more complicated. After Pearson1987

Dating the past

118

Summary table

4.15 Every radiocarbon measurement has a statisticalmargin of error, which is quoted in terms of themean and one standard deviation (e.g. 2000±100 bp).A normal distribution curve shows how it should beinterpreted: one standard deviation either side of themean will give a 68% probability of the true date lyingwithin a 200-year bracket (and consequently a 32%chance of it not doing so), whilst two standarddeviations increase the probability of accuracy toaround 95%. Audio Visual Centre, University of Newcastle

Dating the past

119

7.3 Presenting and interpreting aradiocarbon date(Stuiver 1993)

Health warning! Proper calibration is not easyfor the non-mathematician, but doing it incor-rectly, wrongly interpreting the result, or evennot understanding the potential of calibrationmay seriously damage your archaeology. Takeadvice from the experts know whatcalendrical band-width is necessary for cor-rect interpretation and discuss this with thedating laboratory, preferably before takingand certainly before submitting samples.Think first, not after you get the radiocarbondate. (Pearson 1987, 103)

Because interpretation is so complex, all radiocar-bon dates included in an archaeological publicationmust be presented in a standard manner. For exam-ple, a series of charcoal samples obtained from alate neolithic site at Galgenberg (Bavaria) is shownin Table A below (Aitchison 1991, 113).

The first column contains the code for theGironingen radiocarbon laboratory (GrN) to-gether with a unique serial number for this par-ticular sample, so that it could be checked withlaboratory records if any problem arose. The ar-chaeological number refers to an excavated con-text at the Galgenberg site, and its nature isexplained in the final column. The determined ageof this sample is expressed in ‘raw’ uncalibratedform in years BP (before the ‘present’, standard-

ized to AD 1950), complete with an unavoidablecounting error estimated by the laboratory (±35).The ‘raw date’ has already been adjusted to com-pensate for fractionation, but it is calculated ac-cording to Libby’s half-life of 5568 years ratherthan the more recently determined estimate of5730 years; this practice is maintained to avoidconfusion in comparisons with older results. Thestandard counting error of 35 years means thatthe (uncalibrated) date has a 68% chance of ly-ing between 4350 and 4420 BP, and a 95%chance that it lies between 4315 and 4455 BP.This emphasizes the importance of regarding ra-diocarbon ages as ranges of possibilities, ratherthan ‘dates’.

This ‘date’ has not yet been calibrated. Referenceis normally made to the calibration curve, derivedfrom dated tree-ring samples, published by Pearsonin volume 28 of the periodical Radiocarbon in 1986(supplemented in 1993 by volume 35.1). A rapidinspection of the curve suggests that the radiocar-bon estimation will be transformed into a calendardate with a range falling roughly between 2900 and3100 BC. However, closer inspection of this particu-lar age determination reveals a common problem:a ‘wiggle’ in the calibration curve at around 4400BP means that it could represent three different ‘his-torical’ dates (fig. 4.16; Aitchison 1991, 113); seeTable B below. The tree-ring calibration curve isitself subject to statistical variations; for this reasonthe standard deviation should be considered as onlya minimum estimate of unreliability. Furthermore,precision varies according to which part of the curve

Dating the past

120

is being consulted; if the line is steep, the prospectsare good, but if it is flatter, the date range will bevery wide. Thus the ‘date’ of 3025 has the lowestof the three estimated levels of error. When all thir-teen samples from Galgenberg were examined to-gether, the main period of the whole site’soccupation was estimated to lie between 2810 and3100 BC.

Thus, Galgenberg illustrates some of the prob-lems that lie between the receipt of an age estima-tion from a laboratory and its interpretation inmeaningful chronological terms for a site or anartefact. This is why Pearson advised archaeolo-gists to consider the ‘calendrical band-width nec-essary for correct interpretation’ before submittingsamples. In the context of later prehistoric Britain,a sample from the British late Bronze Age and earlyIron Age that was expected to give calibrated re-sults between 1100 and 800 BC would be veryworthwhile, for it would coincide with a steepslope on the calibration curve. In contrast, sam-ples from the period between 800 and 400 BC are

almost useless, for this part of the curve is muchflatter, and does not permit refinement within arange of around four centuries; traditional formsof dating would be more accurate; see fig. 4.14;Bowman 1990, 55–7.

It should be noted that in many British publi-cations, uncalibrated dates are indicated with alower-case ‘ad’ or ‘bc’, while calibrated dates arecited as ‘1000 BC’ or ‘1000 AD’. This practicehas not been followed elsewhere; an InternationalRadiocarbon Convention in 1985 recommendedthat uncalibrated age determinations should al-ways be quoted in the form ‘1000 BP’ (BeforePresent; for this purpose, the ‘present’ is stand-ardized as AD 1950). If dates are calibrated ac-cording to ‘an agreed curve’, they should be citedin the form ‘1000 cal BP’. In areas of the worldwhere the AD/BC division is useful, calibrateddates can be converted to ‘1000 cal BC’ or ‘1000cal AD’ (Gillespie and Gowlett 1986, 160). ‘Per-haps with the benefit of hindsight it might havebeen preferable if radiocarbon measurements hadnever been expressed as “ages” or “dates”; thenthere could be no misunderstanding’ (Bowman1990, 49).

Radiocarbon samplesMost materials containing carbon are suitable fordating; the lower the carbon content, the larger thesample needs to be. Charcoal derived from theburning of wood is a common find on archaeo-logical sites and samples of around 10–20 gramsdry weight are adequate for conventional count-ing, compared with around 50–100 grams of peat,or 100–500 grams of bone (Aitken 1990, 91).‘Mini-counting’ methods cope with samples lessthan a tenth of this size (e.g. 0.1–0.5 grams ofcharcoal), while AMS requires only around onehundredth (e.g. 0.01–0.1 grams). Many othermaterials may be tested, including cloth, flesh,pollen, shell, soil and even iron, which usuallycontains some carbon impurities. The collectionof samples needs to be scrupulous and their stor-age and handling must avoid contamination, eventhough they are subjected to a chemical ‘laundry’process before being tested.

Archaeologists must know exactly what isbeing dated and, in the case of samples from ex-cavations, their precise stratigraphic relation-ship to the site. The nature of charcoal and

4.16 This diagram shows how a single radiocarbon ageestimation (from the Galgenberg, Germany) mayproduce three different calendar dates of varyingreliability if it happens to coincide with a difficult‘wiggle’ in the calibration curve. For the purposes ofdating a neolithic sample, it would normally besufficient to know that the calibrated date laysomewhere between 2800 and 3100 BC, but amargin of error of this size would be too great forhistorical periods. Sandra Hooper, after Aitchison et al.,1991. f ig, 4

Dating the past

121

wood samples is very important—twigs or nutsare ideal, because they only contain carbon-14taken in during a short growing season,whereas the central portion of a large tree willobviously give a date decades (or even centu-ries) earlier than its use for fuel or construction.Thought must also be given to the extent thatsamples are related to the objects or contextsthat they are intended to date. Thus, the samesignificance must not be attached to fragmentsof charcoal from a general occupation level andto a sample taken from part of a wooden arte-fact or a human body. One of the most widelypublicized examples of direct dating was the ex-amination of the Turin shroud; since only verysmall samples of linen could be provided fromthis unique artefact, AMS was an ideal method.By good luck the result (691 ±31 BP) matcheda favourable part of the calibration curve andgave a date of 1275–90 cal AD at the 68%confidence level (fig. 4.17); whatever the nature

and date of the strange image painted (?) on theshroud, the linen from which it was woven grewno earlier than the thirteenth century AD, mak-ing it impossible that it was ever associated withJesus.

Even in prehistory radiocarbon raises ques-tions of a ‘historical’ nature. For example, evi-dence of very early human settlement linked witha hunter-gatherer economy was recently found onthe island of Cyprus, whereas it had previouslybeen thought that farming communities had set-tled there in the Neolithic period. However, sincethe relevant radiocarbon dates were too early forthe conventional calibration curve, it was diffi-cult to provide a calendar date for the earliestoccupation. Evidence from varves, floating tree-rings, Uranium-Thorium dates from coral, andvarious other forms of dating suggest a datearound 10,000–11,500 BC in ‘calendar’ years—but this may be changed or refined in future(Manning 1991). Technical limitations upon ra-diocarbon dates are just as significant in the caseof relatively recent (in European terms, histori-cal) periods. The question of the date of coloni-zation of New Zealand is a good example:estimates ranged up to 2000 years ago, with amajority favouring around 1000 years ago. Alarge number of radiocarbon estimations nowdemonstrates that it took place as recently as thefourteenth century AD, but many samples derivedfrom shell, bone and old wood had given mislead-ing earlier dates (Anderson 1991).

4.17 The precision of AMS radiocarbon dating gavean unambiguous answer to questions about the age ofthe Turin shroud, a mysterious piece of cloth thatbears such a striking image of Christ that it wasbelieved possible that it really had been preservedfrom the Crucifixion in the first century AD. Accel-erator dating was able to test very small samples ofthe cloth, and the age estimation matched a particu-larly favourable part of the calibration curve. Resultsfrom three different laboratories combined to give aremarkably precise indication that there was a 68%chance that the cloth dated from AD 1275–90; thesafer 95% significance level increased this to 1275–90and 1355–85. Thus, whatever explanation for theimage on the shroud is proposed, it must start fromthe knowledge that the cloth is of medieval date.Sandra Hooper, after Aitken 1990, f ig. 4.10

Dating the past

122

First-order radiocarbon datingEmphasis on accuracy and precision in radiocarbondating is not always appropriate. Occasionally, re-sults from a much quicker and simpler method maybe quite sufficient for some needs—at only one four-hundredth of the cost. First-order radiocarbon dat-ing is applicable to carbon dioxide gas derived fromshells using the liquid scintillation counting tech-nique. It has proved very effective on shell middensup to at least 20,000 years old, and it can be checkedagainst samples of charcoal from the same depos-its. It revealed exploitation of marine resources inVictoria (Australia) three to five thousand yearsearlier than had been thought, because it could beemployed on older, less well preserved shell middensthat had not been considered good enough to betested by more expensive conventional methods(Frankel 1991). The advantage was summarizedwell in a recent article: ‘We believe that in manyarchaeological situations it is better to have 10 dateswith standard deviations around the mean of 200years, than one date with a 40-year standard error’(Glover 1990, 566).

The impact of radiocarbon dating(Taylor 1987; 1992)Without doubt, radiocarbon dating has made thegreatest single contribution to the development ofarchaeology since geologists and prehistorians lib-erated themselves, a century earlier, from the con-straints of historical chronology by rejecting thebiblical Creation. The major stages of human de-velopment from hunting through to urbanizationare now well dated over most of the world. How-ever, so few radioactive carbon-14 isotopes remainin a sample more than 40,000 years old that it isdifficult to measure the small number of particleemissions. The technique is therefore unsuitable forstudying most of the Palaeolithic period; fortunately,a related method based on an isotope of potassiumallows the examination of early hominid develop-ments beyond the range of radiocarbon.

7.4 Potassium-argon dating

Potassium is abundant throughout the Earth’s crust.Among its isotopes is a small percentage of K-40that decays into calcium-40 and a gas, argon-40.This gas escapes while new volcanic rocks are be-ing formed, but as minerals crystallize they begin

to trap Ar-40, which can be released from samplesin the laboratory and measured. At 1250 millionyears, the half-life of K-40 is staggeringly long incomparison with that of carbon-14. Its potential forgeological dating had been realized by 1940, butarchaeological applications began in the 1950s.Dates are arrived at by measuring the amount ofAr-40 trapped in potassium-rich minerals in com-parison to K-40; the less there is, the more recentwas the formation of the material involved. Theinaccuracies inherent in measuring minute quanti-ties of Ar-40 make it difficult to use in periods lessthan 100,000 years old. However, improvements inthe measurement of comparatively recent sampleswill allow checks to be made onthermoluminescence dating in the period beyond theeffective range of radiocarbon dating (at best 50,000years).

Potassium-argon is ideal for dating early homi-nid fossils in East Africa, for they occur in an areathat was volcanically active when the fossils weredeposited between one and four million years ago.Layers containing bones and artefacts may be found‘sandwiched’ between volcanic deposits of ash orlava that provide excellent samples of newly formedminerals for measurement. Very occasionally theassociation between human remains and volcanicdeposits may be much more intimate, as in the caseof human footprints around 3.6 million years oldfound on a layer of freshly deposited ash at Laetoli,Kenya (Leakey and Lewin 1992, 103).

Independent radioactive techniques, includinguranium series and fission-track dating, give simi-lar results that support dates derived from potas-sium-argon. Margins of error measured inthousands of years are unimportant in periods ofsuch long duration, but they are useless in laterprehistory, when, for example, the entire EuropeanBronze Age lasted for only around 1000 years. Amore serious problem involves the nature of thematerial required for sampling; few areas in theworld provide archaeological remains that arestratigraphically related to sequences of freshlyformed but undisturbed volcanic material, contain-ing crystallized minerals of the kinds best suited tomeasurement. Potassium-argon has also been veryimportant to geologists for checking the dates ofsome major reversals of the Earth’s magnetic field(below, p. 128) and climatic patterns revealed bysea-bed cores (above, p. 113); both sources of in-

Dating the past

123

formation are valuable for dating purpose in placeswhere volcanic minerals suitable for the potassium-argon method are not available.

7.5 Uranium series dating

The dating of geological periods followed the dis-covery of radioactivity, and the age of rocks backto the Pre-Cambrian was assessed by mesuring theproportions of uranium and lead or uranium andhelium. Uranium could be used in this way becauseit remains radioactive for very long periods; ele-ments with shorter decay periods are of course morehelpful for recent geological and archaeologicaldating. Thorium-230 is a useful isotope because ithas a half-life of 75,400 years. Although coral is theideal sample material, calcite crystals contained instalagmite may also be sampled, and this makes itsuitable for dating early human activity in caves,anywhere between 5000 and 350,000 years ago;calcium carbonate deposits from spring also pro-vide suitable material. Large samples of around 50grams are required unless mass spectrometry isavailable. In any case, the precise relationship be-tween the sample and an archaeological event oractivity must always be established. Uranium seriesdating is frequently used in conjunction with ESR(electron spin resonance—below, p. 124), for thelatter may also be carried out on the kinds of sam-ples typical of cave finds, such as teeth, shells andstalagmite calcite. Uranium-Thorium dating of coralhas also allowed radiocarbon dates to be calibratedback to around 20,000 BC, because coral, a livingorganism, also contains carbon.

8 Radioactive ef fects oncrystal structure(Aitken 1990)

The following absolute techniques do not simplymeasure radioactive emissions or the products ofradioactive decay; instead, they examine the effectsof radioactive impurities on the crystal structure ofminerals.

8.1 Thermoluminescence dating(Aitken 1985)

TL dating is most effectively appiled to fired clay,which normally contains crystalline impurities, or

burnt flint. In addition to being subjected to con-tinuous cosmic radiation, these materials are alsoaffected by radioactivity from uranium, thoriumand potassium, contained in the artefacts them-selves or in the soil where they have been burieduntil excavation. Crystals have defects in theirstructure that ‘trap’ electrons produced by thisradiation; ‘deep traps’ do not begin to release theseelectrons unless they are heated above 300°C.When electrons are released, some recombine im-mediately with a luminescence centre (another typeof defect) and emit light in proportion to theirnumber. As soon as heating is over, electrons be-gin to accumulate again, until reheating to the sametemperature occurs.

The first stage in calculating a date is to meas-ure the amount of light released by a suitably pre-pared sample from an appropriate material andto plot its ‘glow-curve’ on a graph as the sampleis heated up to 500°—the ‘natural’ glow-curve.This is then compared with another, ‘artificial’,glow-curve derived from an identical sample sub-jected to a known amount of radiation in thelaboratory (fig. 4.18). The relationship betweenthe two curves gives information about the reli-ability of the sample, as well as revealing theamount of energy that had accumulated since itwas last heated (the palaeodose). The palaeodosedoes not reveal the age straightaway. It is neces-sary to measure the annual dose derived not onlyfrom radioactive impurities in the sample itself,but also from the radioactivity of the soil whereit had been buried. When this has been done, anda number of additional factors have been allowedfor, the age is equivalent to the palaeodose dividedby the annual dose. Thus, a palaeodose of 8.5 Gydivided by an annual dose of 5.18 Gy (Grays, astandard measurement of absorbed radiation)would give an age of 1640 years—around AD 350(Aitken 1990, 151).

The most important material for TL is firedclay; hearths, kilns and especially pottery form animportant part of the archaeological record in mostparts of the world. Since pots are fired at a tem-perature well above that required to release all theelectrons that have been trapped in their crystallattices, the energy released in the laboratory to-day will have built up from the date of their fir-ing. The older the pots, the more energy thatshould have accumulated. Unfortunately, there are

Dating the past

124

many problems connected with measurement.Clays contain a variety of mixed minerals of dif-fering particle size that vary in their ability to trapelectrons. Samples require elaborate preparationto separate particles of the optimum size, and ex-tremely sensitive equipment must be used to recordthe very slight amounts of thermoluminescenceemitted. Furthermore, measurement of the radia-tion absorbed from the soil that surrounded aburied clay artefact is difficult, but crucial for anaccurate calculation of the age; this is even morecritical in the case of flint. Obviously, it is notpossible to measure this form of radiation in ob-jects from museum collections whose precise findspot is unknown; TL will still detect modern for-geries that lack trapped electrons, however.

The TL dating technique is particularly valuablefor dating in situations where no suitable materialsfor radiocarbon dating have been found, or if theage exceeds 40,000 years, when radiocarbon is ofrapidly diminishing usefulness. Fortunately, sitefinds from early periods of prehistory usually in-clude stones and flint implements burnt in fires incaves or camp sites at a sufficiently high tempera-ture to release their trapped electrons. Flints foundin deposits with relatively low radioactivity mayhave a potential datable range of up to 500,000 oreven a million years. In areas where volcanic mate-rials suitable for potassium-argon dating are absent,this is of great significance. Stalagmite may also bedated, and so too can volcanic materials or the soilover which molten lava has flowed (Aitken 1990,172). It is even possible to date deposits of soil orsediment that were subjected to intense sunlight andsubsequently buried. It has been established that thiskind of exposure to heat and light is sufficient toremove trapped electrons (‘bleaching’), but theybegin to accumulate again as soon as it is covered(ibid. 173–5).

8.2 Electron spin resonance (ESR)

The basis of ESR has much in common withthermoluminescence, for both measure electronsthat have become trapped in the crystal lattice ofminerals. It differs from TL in the nature of suit-able samples, which include teeth, shells and sta-lagmite calcite. The method of measurement is alsounlike TL, for ESR does not release the electrons,but subjects them to electromagnetic radiation in amagnetic field. At certain points of interaction be-tween the magnetic frequencies and the magneticfield, the electrons resonate and absorb electromag-netic power. The strength of resonance reflects thenumber of trapped electrons, and their quantity isrelated to the time that has elapsed since the crys-tals were formed. The radioactive content of sam-ples, combined with the external radiation that theyhave received, must always be measured so that theirage can be calculated by dividing the palaeodose bythe annual dose—in exactly the same manner as TL(see above).

ESR sample materials favour the study of thePalaeolithic period, for stalagmites may be relatedto cave occupation, and fossil teeth from largemammals such as mammoths may provide effective

4.18 Thermoluminescence apparatus provides a graphof light released by a sample prepared from anancient artefact as it is heated (a). A second measure-ment of the same sample provides a different graphfor the same material without its ancient energy (b);the bulge in curve (a) between 300° and 400°Cresulted from the electrons trapped in the sample.Curves c1–3 are further measurements taken tostudy the luminescence produced after the sample hasbeen exposed to known levels of modern radioactiv-ity in order study its sensitivity. When further factorsabout the context in which the artefact was foundhave been taken into account, a date may becalculated. Audio Visual Centre, University of Newcastle

Dating the past

125

dating. Tooth enamel is required, for the dentine ofthe core—like bone—is porous, and new mineralscontinue to form long after the death of the animal,giving an underestimate of the true age. Aitken citessome convincing examples of ESR dates derivedfrom samples of mammoth and rhinoceros teethfrom Canada and Germany; the resulting dates ofaround 100,000 and 350,000 years old correlatedwell with the climatic stages to which the finds couldbe assigned, and with uranium-series dating (Aitken1990, 198–9).

8.3 Fission-track dating

This method involves counting microscopic dam-age trails in minerals such as zircon, and glass,whether volcanic (e.g. obsidian) or of humanmanufacture. The trails are caused by fission frag-ments when the nucleus of uranium-238 splitsduring radioactive decay. In practice the most use-ful samples come from zircon, and from obsidian,a material used extensively for making tools. How-ever, the sample must have been subjected to heat-ing at the time of the archaeological context orevent that is to be dated. Obsidian tools or wasteflakes from tool production that have beendropped into a hearth would be ideal, for heatremoves the fission-tracks that have accumulatedsince the obsidian first solidified after its volcanicformation. Glass, or a vitreous glaze on pottery,should not have been re-heated in this way if thedate of its manufacture (rather than its last heat-ing) is to be discovered. Fission-track dating, alongwith potassium-argon, has also assisted in check-ing the age of volcanic deposits associated withearly hominid remains in East Africa (Aitken 1990,135).

9 Derivative techniques(Aitken 1990)

Aitken draws a clear distinction between abso-lute and derivative dating methods (1990, 2); thelatter may only be used for dating by compar-ing their results with a time scale or referencecurve that has been established by other datingmethods. Thus, the level of thorium-230 foundin a sample of stalagmite is a product of its

uraniurn content, and the sample’s age is calcu-lated from the known radioactive half-life ofthorium-230, which is not affected in any wayby its environment; this method can therefore bedescribed as absolute. In contrast, measurementof one form of amino acid changing to another(outlined below) is a derivative method, for therate of alteration is entirely dependent on thetemperature of the context where the sample hasbeen buried.

Fluorine, uranium and nitrogen testing wasone of the first scientific dating methods used inthe examination of bone. It did not attempt toprovide an estimate of age, but addressed a morefundamental problem that affects bones or ar-tefacts of any kind: are the finds excavated froma single level—for example, a layer containingartefacts and bones in a cave—really contempo-rary? Does the stratum contain older items thathave eroded out of earlier contexts, or items dugup accidentally during a later phase of occupa-tion? Amino acid racemization and obsidian hy-dration dating may also be used to detect strayitems, for bones and artefacts buried in uniformconditions over the same length of time wouldproduce identical results; if they do not, somedisturbance must have taken place and the ex-cavated material is of limited value for any dat-ing method.

9.1 Fluorine, uranium and nitrogen tests(Price 1989)

Buried bone absorbs water containing elementsthat react chemically with the bone, adding fluo-rine and uranium, while nitrogen decreasesthrough the decay of bone protein (collagen).Bones found in a single context should have beensubjected to the conditions that cause these changesin a uniform manner, and their levels of these threeelements should therefore be very similar. Oldersurvivals and recent intrusions should therefore bedistinguishable because of unusually high or lowlevels. The technique remains useful for checkingbone samples that are to be submitted for radio-carbon dating, but other methods (such as ESR)are now more informative for dating purposes.Fluorine/ uranium/nitrogen testing holds a specialplace in archaeological history because of its rolein proving that the skull and jaw of ‘Piltdown

Dating the past

126

Man’, excavated in Sussex in 1912 and claimedas evidence of an early hominid, was a forgery(Spencer 1990). Analysis showed that the find hadbeen assembled from bones of several differentages and origins.

9.2 Amino acid racemization

Samples taken from bone, teeth or shells containdetectable amino acids that undergo gradual change(racemization) from L-form to D-form over time;the ratio of the two is measured to indicate age. Sincethe rate of change is highly dependent on tempera-ture it is necessary to use an independent method,such as radiocarbon, to date a sample from the sameburial context. Once this has been done, the speedof racemization may be determined and other sam-ples may be dated by this means alone. The upperlimit for successful dating may range from 100,000to several million years, according to the kind ofsample material available and the amino acids se-lected for study.

9.3 Obsidian hydration dating

Like amino acid racemization, this dating tech-nique relies on a transformation that takes place

over time and, likewise, it is highly dependenton the temperature of the context where the sam-ple of obsidian has been buried. Obsidian is anatural volcanic glass that was a popular alter-native to flint for making flaked tools in manyparts of the world (see fig. 5.2). As soon as afresh surface of obsidian is exposed, for exam-ple during the process of making it into a tool,a microscopically thin ‘hydration rim’ begins toform as a result of the absorption of water (fig.4.19). Furthermore, obsidian from different geo-logical sources may weather at different rates.However, in regions where supporting dates aresupplied by radiocarbon (notably Japan andSouth America), large numbers of measurementscan be compiled to provide a calibration curvethat may be used for checking the rim thick-nesses of individual artefacts or assemblagesfound on similar sites.

4.19 This photomicrograph shows a section throughthe hydration rim of an obsidian artefact. The interiorof the specimen is on the left; the diagonal band is alayer of weathering on the surface, and its depth isdemarcated by a diffusion front that shows up as apaler line. This can be measured quite accurately,even though it is only three microns thick in thissample. Prof. J Michels

Dating the past

127

This technique may also be used to check thecontemporaneity of material from a single depositand to detect modern forgeries. The thickness ofthe hydration rims of large numbers of artefactsfrom individual sites are plotted on a graph; if thesite was occupied continuously they should dis-play a reasonably smooth progression with age,while discontinuities may represent periods ofabandonment. In warm climates where hydrationis rapid, obsidian dating is useful for quite recentcenturies, but, generally a margin of error of atleast ±1% must be expected—in other words, abracket 100 years either side of an age of 1000years.

9.4 Archaeomagnetic dating(Tarling 1983)

Fine grains of iron oxide are present in most clayand soil, and they take on a new magnetic align-ment in two main ways. Thermoremanent magnet-ism is acquired when they realign according to theEarth’s magnetic field after having been disorientedby heating above 650°C; some grains may retainthe new field for hundreds of thousands of years.Magnetism is also acquired by means of the depo-sition of sediments, for instance in lake beds, whereparticles may settle into alignment with the prevail-ing magnetic field. Magnetic North wanders at ran-dom around the North Pole (and indeed reversescompletely to the South Pole for long periods). Fromany reference point its position is measurable interms of two components: movement up or down(inclination or ‘dip’) and from side to side (declina-tion).

Magnetic dating measures the alignment in anancient sample and attempts to relate it to arecord of past changes in the magnetic field (fig.4.20). However, although records of magneticalignment have been made by scientists in Brit-ain since before AD 1600, they began much morerecently elsewhere in the world. Thus, informa-tion about the pattern of past variations has tobe derived from suitable samples from archaeo-logical sites that have been dated independentlyby some other means, such as historical evidenceor radiocarbon dating. The reference curve ofinclination and declination has been extendedback to 1000 BC in Britain in this way. Unfortu-nately, the Earth’s magnetic field varies from re-

gion to region, so that results from Britain are noteven applicable in France. Thus, magnetic datingclearly illustrates Aitken’s definition of a deriva-tive method, for it is necessary to establish a sepa-rate independently dated series of measurementsfor every region where the technique is required.When multiple dates result from over-laps in therecord of the magnetic field, one particular datemay be selected on archaeological or historicalgrounds. Fortifications that were possibly erectedby Charles the Bald at Pont-de1’Arche on theSeine in France produced dates around 360 BC,

4.20 The movement of Magnetic North, measuredfrom Britain. The graph shows declination, in degreeseast or west of true North, and inclination, in degreesbelow the horizontal. These wandering lines arecompiled from contemporary observations as farback as records allow, but samples from dateddeposits or structures on archaeological sites must befound to project them further back into the past.Samples from undated sites can be measured in thelaboratory, and dated according to where theirmagnetic alignments coincide with the curve estab-lished for the relevant geographical area. Difficultiesdo exist, however: identical readings occur whereverthe curve crosses itself, for example between AD1600, when the curve also matches late-Saxonmeasurements. Prof. D Tarling, University of Plymouth

Dating the past

128

AD 580, AD 860 and AD 1580; obviously onlyAD 860±20 was appropriate for a historical ref-erence to a Viking attack in AD 865 (Dearden andClark 1990).

This technique may only be used on archaeo-logical sites where solid clay structures that havenot moved since becoming magnetized are found;kilns, hearths and burnt clay walls or floors areideal. Small samples are selected, and their posi-tions are carefully recorded in relation to thepresent magnetic field. This allows their align-ment to be duplicated in a laboratory, so that dif-ferences between the ancient and presentalignment can be measured. It is also possible toexamine the ‘dip’ angle of portable fired objectssuch as bricks or pots, as long it is possible toassume that they were fired in a horizontal posi-tion.

One fur ther d imens ion ofarchaeomagnetism is magnetic intensity, ameasure of the strength rather than the direc-tion of the magnetic field. Like magnetic align-ment, it may be retained after heating oracquired through deposition, and it also var-ies from area to area. A dated reference seriesof measurements is therefore needed for anyregion before it may be used as an independ-ent dating method. Insufficient variation hasbeen discovered for it to become a usefulchronological tool (Aitken 1990, 252–3).

The direction and intensity of the Earth’smagnetic field are of an interest that extendsbeyond dating, for they probably have an influ-ence upon global climate and cosmic radiation.The general pattern of major variations (such asNorth-South reversals) has been established bygeologists, and it has considerable implicationsfor archaeologists involved in the study of earlyhuman remains found in geological deposits inEast Africa and elsewhere. Furthermore, abso-lute dates for the major reversals have been de-termined by the potassium-argon method.Thanks to the occurrence of iron oxide particlesin sea-bed deposits, magnetic reversals have alsobeen correlated with the climatic changes indi-cated by oxygen-isotope variations recorded incores taken from ocean-floor sediments (above,p. 113). Thus, fossil bones or tools found in astratum with magnetic characteristics that canbe linked to a datable magnetic reversal may be

dated and placed into their correct environmen-tal context.

9.5 Cation-ratio dating (CR)

Prehistoric rock carvings (‘petroglyphs’) are notuncommon in arid areas where suitable surfaceshave escaped erosion by the action of rain andfrost. However, their age is notoriously difficult todetermine unless they are found in contact with adatable stratified deposit. Petroglyphs are com-monly covered by a so-called ‘rock varnish’, achemically changed layer that builds up afteraround 100 years through weathering, enhancedby the action of micro-organisms. Using a methodfirst put forward by Ronald Dorn in 1983, sam-ples are taken by scraping the ‘varnish’ frompetroglyph surfaces back to original rock surface.A separate cation (positively charged ion) leach-ing curve must then be established for differentgeographical areas, because local soil and mois-ture conditions affect the speed of its formation(Dorn 1988, 683). The date of the surface layerthat has formed over the carvings may sometimesbe checked by AMS radiocarbon dating, if suffi-cient carbon from micro-organisms was includedin the initial ‘glaze’.

Forty-six petroglyph samples from Piñon Can-yon, an arid site in south-eastern Colorado, weredated to between 300 and 2000+ years before thepresent (Loendorf 1991). The results were con-sistent with a relative sequence established on ty-pological grounds according to the style of thedesigns, and they also matched radiocarbon datesfrom associated sites. The first use of cation-ra-tio dating on carvings outside the USA was in thearid zone of south Australia, where it suggestedexploitation of the area for over 30,000 years(Dorn 1988). However, Lanteigne has sounded anote of caution, because many underlying as-sumptions have not yet been fully tested (1991).As with radiocarbon and the derivative tech-niques (notably obsidian hydration), many modi-fications will no doubt be required before thepotential and limitations are properly under-stood. It may also be possible to use this techniqueto date stone artefacts found on the surface ofdeserts; it would at least provide a minimum agefor the time that has elapsed since they were ex-posed by erosion.

Dating the past

129

10 The authenticity ofartefacts(Jones 1990)

It is inevitable that major museums that buy itemsfor their collections become involved in expen-sive commercial dealings in the fine art market.The profits to be made not only stimulate illicitplundering of ancient sites, but encourage skilfulforgeries. Scientific dating techniques bring ob-vious benefits, for precise dates are rarely re-quired, simply an assurance that an artefact is nota modern fake. Thermoluminescence andarchaeomagnetism provide adequate checks onpottery and a variety of highly priced elaborateceramic ‘terracotta’ sculptures from Africa andSouth America. Where they survive, the remainsof clay cores left inside bronze statues or objectsafter they were cast in moulds also provide suit-able samples. It is very unlikely that a forger couldcreate artificially the precise levels of radioactiveenergy or magnetic conditions that should befound in genuine items. Radiocarbon dating bythe AMS technique now allows very small sam-ples to be taken from small wooden, bone or otherorganic artefacts without affecting their appear-ance. Dendrochronology is helpful in the studyof wooden panels used in furniture and earlypaintings, while paints and pigments may be ex-amined by means of various forms of radioactiveisotope dating.

11 Conclusions

Thus scientific dating is not just a boring ne-cessity that tidies things up by providing num-bers, it is vital for valid interpretation.(Aitken 1990, 1)

Traditional forms of archaeological dating havebeen strengthened immeasurably by the growthof an extraordinarily diverse range of scientifictechniques that helps to demonstrate the trulymulti-disciplinary nature of modern archaeology.Traditional methods have not been replaced,however. The definition of sequences by meansof stratigraphic excavation remains the basis forobservations about sites and for typological stud-

ies of artefacts. Scientific dating techniques addprecision and allow specific hypotheses about therelationships of sites, regional cultures or formsof artefacts to be tested. The transition from hunt-ing and gathering to agriculture and the emer-gence of early civilizations may be interpreted inmeaningful human terms now that we know—thanks to radiocarbon dating—when they oc-curred and how long the processes oftransformation took. Similarly, potassium-argondating (in association with several other methods)has provided a framework for the study of hu-man evolution at the important point when thereare the first clear signs that stone tools began tobe used.

Scientific dating techniques play more of asupporting role in historical periods, and theyare particularly valuable where there is doubtover historical dates, or where gaps exist in thehistorical framework. It must not be forgottenthat even absolute methods such as radiocar-bon had to be validated first by testing samplesof known historical date. Libby used finds fromEgyptian pyramids up to 5000 years old, datedby historical records of the reigns of pharaohs,to test the consistency of carbon-14 measure-ments beyond the range of tree-rings (Aitken1990, 58, fig. 3.2). The refinement of radiocar-bon dating, combined with dendrochronology,now feeds information back into this process;recent detailed scientific dating of the lateBronze Age around the Aegean confirms thesequences built up from artefact typologies andhistorical records over the last century (Man-ning & Weninger 1992). As with other scien-tific approaches to archaeology, the wholeprocedure is founded on cooperation, and theincreasing complexity of methods used to refinethe accuracy of scientific dating techniquesdemands ever closer collaboration between sci-entists, historians, prehistorians and excavatorsto produce results that benefit all in differentways. The best of the old should accompany thebest of the new.


130

1 Is archaeology a science?

So much of the evidence left behind by past cul-tures has been destroyed that archaeologists havea duty to employ whatever methods are availableto extract the maximum possible information fromwhat still survives. Since archaeology has alwaysborrowed concepts and techniques from other dis-ciplines, it is not surprising to find that science hashad just as dramatic an effect on archaeology as ithas had on all other aspects of modern life. To dem-onstrate the fundamental role of science in archae-ology it would almost be sufficient to refer readersto the many forms of scientific dating described inchapter 4. The ‘radiocarbon revolution’ exempli-fies the nature of most relationships between sci-ence and archaeology. Its development was entirelydependent upon advances in nuclear physics, andthe archaeological applications are a very minordiversion from the central issues of physics. Like-wise, the location devices described in chapter 2were not originally devised for archaeological re-search, but they were applied and adapted to ar-chaeological situations either by scientists whohappened to be interested in archaeology, or byarchaeologists who understood the scientific prin-ciples sufficiently well to be able to recognize theirpotential.

The use of scientific methods does not makearchaeology into a science. The many scientifictechniques that may be applied more or less di-rectly to the investigation of sites or objects fall intoa sub-discipline generally known as archaeologi-cal science. On a deeper level, a ‘scientific’ attitudeof mind rejects individual observations or subjec-tive conclusions, and demands that questionsabout the past should be posed in the form ofhypotheses that can be tested. In this manner, ar-chaeological research should ideally proceed as aseries of laboratory experiments designed to verify

or refute these hypotheses. However, this analogycontains a flaw: archaeological experiments arerarely (if ever) repeatable under laboratory condi-tions, because no two sites or artefacts are everexactly the same. In practice, there are few ques-tions about the past that would not benefit frominvestigation with the help of the natural or bio-logical sciences; indeed, many questions may onlybe answered with the assistance of scientific meth-ods. One of the principal educational virtues ofarchaeology is that it is truly multi-disciplinary,and that it defies all attempts to pigeon-hole it ei-ther as a science or as one of the humanities. Thischapter will attempt to illustrate parts of the widerange of information that science contributes toarchaeology.

2 The examination of objectsand raw materials(Henderson 1989; Bowman 1991)

Any archaeological object, whether found casuallyor during a controlled excavation, poses questionsabout its date, origin, function and method of manu-facture. A museum curator who deals with visitors’enquiries will know that some of these questionsmay be answered superficially by a combination ofcommon sense and experience. Scientific analysisoffers many insights into ancient objects, but, aswith dating methods, there must be full coopera-tion and communication between archaeologistsand laboratory scientists to ensure that the mostappropriate methods are applied to suitable sam-ples. There is little point in conducting analyseswithout clear questions in mind, and those questionsshould be the result of archaeological research.Occasionally an archaeologist will only require astraightforward ‘yes’ or ‘no’; the course of furtherscientific investigation might depend entirely on theanswer. In the case of the ‘Ice Man’ found in the Alps

5 Science and Archaeology

Science and Archaeology

131

(Spindler 1994; Barfield 1994), it was important toknow the composition of a metal axe found withthe body. Whereas the typological form of the axesuggested a date in the early Bronze Age, radiocar-bon dates indicated that the man died in the lateNeolithic period. Fortunately, metallurgical analy-sis was able to reveal very quickly that the axe wasmade of pure copper, which was acceptable at thisearly date, rather than bronze, an alloy of copperand other metals that was not introduced until theearly Bronze Age.

Thus, the initial question about the Ice Man’s axewas very simple; once it had been answered, fur-ther information from the analysis might help in theinvestigation of subsidiary questions, such as thelikely source of the copper, or the precise methodof its manufacture. If the axe had turned out to havebeen made from bronze, a very different range ofquestions would have been posed. The most impor-tant problem would not have been related to the axeitself, but to the possibility of errors in the radio-carbon dating of the body. If a Neolithic date for abronze axe had been confirmed, a further pro-gramme of analyses would have to be conducted tocheck whether other early ‘copper’ objects were alsomade from bronze. This example demonstrates howinteractions between archaeologists, analytical sci-entists and radiocarbon dating laboratories are es-sential if the right questions are to be posed, andthe implications of the answers are to be under-stood.

It is not only finished artefacts that provide in-teresting sources of information about raw materi-als. Mines and quarries illustrate extractionmethods, and industrial sites offer insights intomanufacturing processes through the excavation offurnaces, kilns or workshops. Waste products, suchas the slags that flowed out of metal smelting fur-naces and solidified, can be subjected to the micro-scopic and analytical procedures similar to thoseexplained below in relation to objects found onoccupation sites (Craddock 1991). A wide range oftechniques may be applied to sites as different asthe scatter of stone fragments discarded by a pre-historic hunter-gatherer making tools in a tempo-rary camp in East Africa, or the ruins of anineteenth-century lead mining complex in the Pen-nines in northern England. Many details of the veryrecent industrial past either went unrecorded, or therelevant documents have not survived. Scientific

excavation and the recovery of carefully chosensamples for analysis will help to clarify both typesof site.

2.1 Microscopic examination(Olsen 1988)

Not all questions of scientific analysis require com-plex analytical methods; traditional study by micro-scope allows many aspects of stone or metalartefacts to be examined. Geologists have usedmicroscopes to enhance visual observations for sev-eral centuries, and metallurgists may still learn a lotabout a metal object by examining a magnifiedcross-section. Archaeologists now use microscopesin use-wear analysis of artefacts, in particular tools,for patterns of wear or damage on working surfacesmay suggest how a tool was used. Evidence formanufacturing techniques may also be revealed bymicroscopic examination; decorated metal objectswere frequently ornamented by means of a rangeof engraving tools whose shapes may be identifiedwhen magnified. Visual examination in use-wearstudies is enhanced dramatically by a scanning elec-tron microscope (SEM), which projects a magnifiedimage onto a screen (Olsen 1988). An SEM sweepsa band of electrons over a surface to provide im-ages that possess a depth of focus unobtainable fromconventional microscopes (see fig. 5.13). This, com-bined with a dramatic increase in the power ofmagnification, reveals not only traces of use-wearon tools, but also traces of tissues from animal orstarches from plants that were cut by them. SEMphotomicrographs of cross-sections of pottery arealso very informative, for they reveal the texture andstructure of clays and glazes with remarkable clar-ity, revealing techniques of manufacture and deco-ration (Tite 1992).

Petrology (Kempe & Harvey 1983)Besides the more sophisticated analytical tech-niques described below, traditional geologicalmethods have much to offer the archaeologist,whether in the context of early prehistoric culturesthat relied heavily on the use of stone for tools, orin more sophisticated societies where fine build-ing stone was transported over long distances.Petrology involves the examination under a micro-scope of thin sections cut from samples of stone.Many minerals may be identified by eye, and dis-


132

tinctive rocks are recognizable without the help ofspectrographic analysis to measure their elements.Axes made in stone from volcanic outcrops inwestern Britain were distributed all over Englandin the Neolithic period, and more than 7500 ex-amples were studied by petrological microscopefrom the 1940s onwards (Clough 1988). However,the majority of neolithic (and earlier) stone axes,along with other stone tools, were made from flint,which has such a uniform appearance that itssource can only be traced with difficulty byspectrometry; inspection under a microscope is ofno help.

Geologists occasionally pour cold water on at-tractive theories held by archaeologists. It wasthought for several decades that the volcanicstones from outcrops in south-western Wales usedin the construction of Stonehenge had been trans-ported by water and overland in an impressivedisplay of prehistoric organization and technol-ogy. Although a recent petrological survey hasconfirmed the Welsh origin of these ‘bluestones’,some geologists claim that they were carried toWessex by glaciers during the last Ice Age. Thebuilders of Stonehenge would have been able tofind them locally where they had been dropped(along with many other ‘erratic’ boulders) whenthe glaciers melted (Williams-Thorpe & Thorpe1992). The identification of the sources of build-ing stone is also very important in historical ar-chaeology, right up to recent times: Roman villas,medieval cathedrals and nineteenth-century townhalls provide insights into the technical skills,communications, and prosperity of the societiesand individuals who created them. The sourcesof stone used in the Roman palace at Fishbourne(Sussex) demonstrate that a wide range of appro-priate building stones had been identified all oversouthern England within a few years of the Ro-man conquest, and that Roman engineers werecapable of selecting, quarrying and transportingspecific types of stone for different parts of thebuilding (Greene 1986, 154–6).

Petrological techniques are applicable tobricks and pottery if their clay includes distinc-tive minerals related to the geology of the areaswhere they were manufactured. The forms andfabrics of the millions of Roman amphorae thatwere traded all over the Empire have respondedparticularly well to petrology, with the result

that we now know where most types were manu-factured. As a result it is possible to study thesources and distributions of important agricul-tural products such as Italian wine, Spanish fishsauce or North African olive oil; Roman docu-mentary sources tell us very little about trade,but amphorae are very common finds on exca-vations and in shipwrecks. Petrological study ishelped by the fact that amphorae are occasion-ally found still bearing hand-written inscrip-tions, written in black ink before their shipment,giving details of their contents and origin (Pea-cock & Williams 1986).

Metallography (Tylecote & Gilmour 1986)The Ice Man’s axe required full analysis to deter-mine whether it was made of pure copper or an alloywith other metals, but many other aspects of met-alworking are suitable for study under a micro-scope. Indeed, according to Northover, ‘…it shouldbecome a rule that analysis and metallography (andhardness testing) should always be combined, sincea full interpretation of one is impossible without theother’ (1989, 214).

It was realized long ago that, before bronzewas made, unalloyed copper was used, because(like gold) it occurs naturally, and can be workedto a certain extent without smelting. However, allbut the simplest artefacts required the metal tobe poured molten into a mould. Their form, andsurface traces left by flaws in the casting, usuallymakes it clear if this happened, but a metallur-gist is able to determine the kind of mould used(metal, stone or clay) and to distinguish cold-worked from cast objects by examining a cross-section under a microscope (fig. 5.1). Thecrystalline structure of cold-worked objects isseverely distorted and flattened by hammering.Because these traces are more difficult to detectthan minerals in a petrological sample, the sec-tion has to be polished and etched to enhance theedges of crystals, and it may be necessary to useSEM magnification to reveal subtle distinctions(Bowman 1991, 86, fig. 5.11).

Iron objects also reveal their production tech-niques when studied in section. Since cast iron didnot appear until the medieval period in Europe, allthe traces visible in earlier iron objects are the re-sult of laborious hammering by smiths. Treatmentof the surface to harden it by quenching in water or


133

roasting in charcoal was important in the produc-tion of iron weapons, and it also leaves visible tracesthat can be confirmed by analysis of their carboncontent. Large or complex objects were constructedfrom several pieces, and their joins are more easilystudied in X-ray photographs than by the destruc-tive process of cutting sections for microscopic in-spection; this is also necessary for badly corrodedobjects (see fig. 5.6).

2.2 Analysis and characterization(Jones & Catling 1986; Hughes 1991)

Many of the dating methods discussed in chapter 4are based upon scientific procedures, but few in-volve analysis of exactly what artefacts are made

from, or how they were made. For example, al-though thermoluminescence dating of pottery re-quires analysis to measure the radioactivity of itsminerals, it is not necessary to know their geologi-cal origin. The basic aim of the analysis of an arte-fact is to identify the materials from which it wasmade, and to measure accurately the relative quan-tities of its constituent minerals or chemicals. Whenthis information is interpreted it may be possible todefine the sources of the raw materials, to suggest aplace of manufacture, and to deduce techniquesinvolved in the manufacture of an object. Analysisis not restricted to objects, however; structures suchas buildings offer many possibilities for the analy-sis of stone, bricks and mortar.

Many forms of analysis are conducted bymeans of spectrometry. A sample is subjected toradiation, and each element then emits radiationwith a distinctive wavelength that may be identi-fied and measured. The presence or absence ofsignificant elements may thus be detected, and theactual quantity of each is measured from theamount of radiation recorded. Besides their prin-cipal elements, most types of stone, metal or ce-ramic material also contain small quantities of

5.1 The structure of metals, seen under a micro-scope, can reveal the processes involved in theirmanufacture. The cross-section on the left (c. 0.5 mmacross) shows the crystalline structure of an Egyptianfigurine cast from an alloy of silver and copper; incomparison, the section on the right (c. 0.5 mmacross) shows considerable flattening and distortionby extensive hammering and reheating during themaking of a Roman silver bowl. Vera Bird/Janet Long,British Museum


134

impurities; it is possible to detect and measurethese ‘trace elements’ down to a few parts permillion by spectrographic means. Trace elementsare extremely helpful in tracing the origins of rawmaterials, and this approach has been very use-ful in characterization studies. Characterizationaims to provide individual ‘fingerprints’ forsources of raw materials (whether stone, metalores or clay deposits) by detecting significant traceelements. The results normally require compli-cated statistical processing to determine whethera distinctive combination of elements found atone source genuinely differs from that found atall other sources; if so, the figures can be plottedonto reference graphs (see fig. 5.3). Obviously, alarge number of specimens from known sourcesmust be analysed before any artefacts of unknownorigin are tested.

An important consideration in the choice ofanalytical methods is the size and nature of the sam-ple that is required. Traditional geological andmetallurgical examination under a microscope in-volves the removal of a portion of an object suffi-ciently large to be ground flat and mounted on amicroscope slide; many stone axes on display inmuseums show visible traces of this kind of sam-pling. Most spectrographic techniques based onradioactivity (such as neutron activation analysis)are conducted on very small samples drilled froman unobtrusive part of an artefact. X-ray fluores-cence is completely non-destructive, but it onlydetects the composition of the surface. Thus, thechoice of technique must depend on full consulta-tion between a museum curator or archaeologistand the laboratory where an analysis will be per-formed. Many major museums have their own labo-ratories; the British Museum in London is a leadingcentre for programmes of research involving activecooperation between the museum’s staff and scien-tists that benefit both sides (Bowman 1991).

Obsidian (Taylor 1976)This volcanic glass occurs widely in both the Newand Old Worlds, and it has attracted considerableattention from archaeological scientists. Like flint,it has excellent working properties for chip-ping,flaking and grinding into tools with sharp cuttingedges (fig. 5.2). In some parts of the world, such asNew Zealand, straightforward visual inspection ormicroscopic examination has proved sufficient to

isolate different sources. Around the Mediterraneanand the American Cordillera, however, there arenumerous varieties of obsidian that require moresubtle differentia-tion. Most analyses have at-tempted to study patterns of prehistoric trade byidentifying sources that supplied sites; this has beenparticularly successful in the Near East and aroundthe Mediterranean (fig. 5.3). Their distributionpatterns provide insights into extensive connectionsbetween early neolithic sites in the Near East as earlyas the seventh and sixth millennia BC.

This valuable information about undocumentedcultures could not have been gained without theuse of scientific analysis. However, the interpreta-tion of the results in human terms remains an ar-

5.2 Obsidian, a natural volcanic glass, formed animportant raw material for tool production in manyparts of the world and was often traded over longdistances. Since sources are rarely identifiable byeye, petrological analysis is required to study thedistribution of obsidian from different sources. Thisflake of semi-translucent obsidian is from Greece,and the flake, core and arrowheads are fromPatagonia. Hancock Museum, University of Newcastleupon Tyne


135

chaeological problem. Geology may reveal thesources of obsidian, but archaeologists must at-tempt to explain why any particular site shouldhave received its raw material from one sourcerather than another. Analysis will not indicatewhether artefacts arrived on a site as finished ob-jects, or if blocks of raw obsidian were broken upand fashioned into tools on each site. Experiencedarchaeological observers may answer this questionby looking for waste flakes chipped off cores dur-ing the manufacturing process. The nature of‘trade’ is also a matter for archaeological interpre-tation, with the help of economic anthropologists:was the raw material bartered for other goods ina commercial manner, or was there an elaboratesystem of gift-exchange conducted on a ceremo-nial basis?

Bronze Age metallurgy (Oddy 1991)Bronze usually consists of copper alloyed with tin,and varying percentages of other metals. The com-position of an artefact made of copper or bronze isusually examined by a technique known as atomicabsorption spectrophotometry. It requires a smallsample drilled from the artefact, which is tested byrepeatedly burning parts of the sample, and shin-ing a beam characteristic of each element’s wave-length through the flame. The quantity of anelement is indicated by the amount of light absorbedby atoms of that element in the sample, to an accu-racy as precise as five parts per million if necessary;this allows trace elements to be measured along withthe principal metals. It is theoretically possible touse trace elements to identify areas from which orescame, by analysing ores and products in a mannersimilar to the study of obsidian. Unfortunately it wasnormal practice for scrap objects to be used as asource of metal, in addition to freshly quarried ormined ores. The resulting mixtures obviously con-fuse any attempt to pinpoint the sources of metalalloys.

Programmes of analysis of finished objects havebeen carried out since the 1930s in Europe and else-where, and a clear general pattern has emerged,although the changes took place at different datesin different areas, according to the availability ofmetal ores. Pure copper (i.e. with only naturallyoccurring impurities) and copper alloyed with ar-senic were soon superseded by ‘true’ bronze madeby adding tin to the copper.

This change was normally accompanied by the useof more sophisticated moulds that required lessfurther work to be carried out to finish the arte-fact after casting. In some areas (at various dates)lead was also included as a major constituent alongwith tin. This required a balance to be achievedbetween two conflicting factors: lead made themetal easier to cast into long elaborate swords or

5.3 Sources of obsidian in the Near East andMediterranean: samples from known sources havebeen analysed, and the results processed to f indtrace elements that differ signif icantly. In thisdiagram, barium and zirconium are plotted againsteach other on a logarithmic scale, and formclusters of results. Artefacts from sites can betested against these clusters to determine theorigin of the raw material from which they weremanufactured. Some clusters are more convincingthan others; relationships between severaldifferent trace elements may be needed toimprove the characterization of each source. Thesame technique is also useful in the study ofpottery, and other kinds of stone. Audio VisualCentre, University of Newcastle, after Cann et al.,1969


136

axes with hollow sockets, but it could make itweaker in actual use. In Egypt, the alloys of cop-per used for making axes were carefully matchedto their function (fig. 5.4). Those used as weap-ons were primarily tin bronze, while tools mightbe copper or arsenical copper; lead bronze wasonly utilized for axes that were decorative ratherthan functional.

Roman coins (Oddy 1980)The use of information derived from the analysisof metals is not restricted to the prehistoric pe-riod. Roman metallurgy has been extensivelystudied through examinations of ores, ingots,coins and manufactured objects. Evidently alloyscould be prepared very accurately to match theproperties of the metal to the function of the endproduct (Greene 1986, 143–9). Of particular in-terest is how the principal Roman silver coin, thedenarius, gradually lost its content of silver as

successive emperors debased the metal in a des-perate attempt to outstrip inflation. UnderAugustus (27 BC–AD 14) denarii were as pure aspossible, and contained 98% silver, but by themiddle of the third century AD, when inflationand political chaos reigned, the silver content hadfallen to 2–3%. Coins left the mint with a thinsilver coating that soon wore off in use, leaving acoin that could only be differentiated from con-temporary bronze issues by its size. The differencebetween early and late examples is obvious to theeye, but analysis has charted the decline in detail;it provides an interesting commentary on the con-temporary economy and the political problems ofrulers who authorized debasements (Greene1986, 60–2).

Isotopic analysis (Sealy 1986)Individual elements can be examined in more de-tail to establish which isotopes (elements with anabnormal number of electrons) are present, andin what proportions. The same procedure is usedin AMS radiocarbon dating, where the propor-tion of carbon-14 in a sample relative to carbon-12 is measured. Studies of stable isotopes nowassist in the study of Greek and Roman architec-ture and sculpture. Although some forms of mar-ble may be distinguished visually or under a

5.4 Analysis of the composition of metal alloys revealsinteresting changes over time. This graph illustratesthe changing proportions of Egyptian axes made frompure copper and three different alloys of copper.Axes used as weapons were primarily tin-bronze, andthose used as tools might be copper or arsenicalcopper; decorative axes were cast in weaker lead-bronze. Sandra Hooper, after Cowell & La Niece 1991,f ig. 5.4


137

microscope, the fine white marbles of Italy,Greece and Asia Minor have always presenteddifficulties. Analysis of the oxygen and carbonisotopes contained in their chemical structurenow provides a method of separating them, andit allows styles of carving to be related to the areaswhere stone was quarried and prepared for use(Walker 1984; Hertz 1987). As in the case ofobsidian, isotopic analysis had to begin with sam-ples taken from quarries known to have been inuse in the past, to characterize each of the mainsources, before their products could be identified.In cases where quarries are not distinguishable bymeans of stable isotopes, trace elements providea successful alternative approach.

Isotopic studies of metals show some promisefor the examination of ancient trade (Gale 1991;Gale & Stos-Gale 1992). Four different isotopesoccur naturally in lead, and their ratios have beenused to characterize the sources of lead ores, andthose of copper that contained natural lead im-purities. The technique works well on ingots ofmetal that were lost before use, or objects thatwere made from fresh lead or bronze. It may evenbe possible to extend the technique to other arte-facts that contained lead, but its usefulness willalways be limited by the same problem of mix-ing scrap metal from several sources that alsocomplicates the characterization of trace elementsin bronze.

2.3 Provenance by date

Scientific dating and analysis occasionally con-verge. For example, geologists are able to deter-mine the age of deposits of obsidian formed byvolcanic activity by means of fission track dat-ing (above, p. 125). If fragments of obsidianfound on an archaeological site are examined bythe same method, the dates obtained do not ap-ply to the artefacts themselves, but to the forma-tion of the raw material at its volcanic source.This date can then be matched to a deposit ofobsidian formed at a corresponding date. As inobsidian characterization studies, accuracy de-pends on how many potential sources have beensampled and tested. Potassium-argon dating hasbeen used in a similar way to identify the origins(rather than the date) of hones and whetstonesfound on Viking period sites in Britain; again, the

raw material has been matched with volcanicrocks of the same type and geological age inNorway (Mitchell 1984).

3 Conservation(Cronyn 1989; Oddy 1992)

Although conservation is one of the most importantaspects of archaeological science experienced byvisitors to sites and museums, it is easily overlooked.

3.1 Ancient objects(Black 1987)

Whenever an ancient object is removed from theground during an excavation it is immediatelyplaced at risk, for the stable environment that hasprotected it from total decay since its burial has beenlost. Objects in museum collections also requireconstant attention, whether during storage or inpublic displays. It is essential that an exact identifi-cation of the composition and structure of an ob-ject is made before conservation begins. Thestructure may sometimes be revealed by visual in-spection, using a microscope if necessary, but aparticularly complex artefact (or one that has be-come encased in a thick layer of corrosion) mayrequire X-ray radiography to understand it (figs5.5–7). Further analysis may be necessary to findout exactly which metals or other substances areinvolved, for these will dictate the form of treatmentto be employed.

The most important task of conservation is toneutralize decay, whether caused by the corrosionof metals or the rotting of organic matter, and thisrequires a detailed knowledge of chemistry. Thenext stage is to stabilize the object so that decaywill not start up again; even when treated success-fully, objects that are intended for display in amuseum will have to be monitored carefully toensure that changes in temperature and humiditydo not trigger further deterioration. Ethical issuesare involved in conservation; a responsible archae-ologist must plan the finance and facilities neces-sary for the preservation of finds, and noexcavation is complete without at least ‘first aid’facilities to minimize the onset of decay until fulltreatment is carried out. This is particularly impor-


138

tant in the case of waterlogged or desiccated siteswhere artefacts made of wood, leather or textilesare likely to be found, for these organic materialsdecompose extremely quickly once they have beenremoved from a stable environment. The restora-tion of artefacts is also sensitive from an ethicalpoint of view. A severely corroded or damagedobject has little commercial value, but cleaning,stabilization and repair not only improve its displayquality for a museum, but also increase its monetaryvalue in the antiquities trade. Since the borderlinebetween a heavily restored genuine artefact and afake is sometimes difficult to draw, conservatorsmust keep detailed records and photographs of allwork that they have carried out. This is also impor-tant in cases where further treatment may be re-quired later, or where researchers need to know theoriginal form of the object. Video recordings nowprovide a convenient additional medium for record-ing conservation work.

5.5–7 Conservation of archaeological finds and theirpresentation for archaeological display involves manyscientific and practical skills. This corroded metalobject is an Anglo-Saxon buckle (seventh century AD)from the St Peter’s Tip cemetery, Broadstairs, Kent.Iron corrosion had encrusted the entire object, andits decorative inlay of silver and brass was only visiblein an X-ray (fig. 5.6). Conservation stabilized thedecay of the artefact, and removed corrosion fromthe decorated surface (fig. 5.7) while preserving it onthe back, where traces of the leather belt and textilescould be seen. English Heritage/Cathy Haith, BritishMuseum


139

3.2 Historic buildings andarchaeological sites(Hodges 1987)

Newly excavated structures soon suffer from expo-sure, and require permanent supervision if they areto be left on display. Wind, rain, frosts, plant growthand human erosion (by visitors or vandals) soondestroy apparently sound masonry structures.Buildings that have been visible for hundreds ofyears are increasingly vulnerable, for ancient stone-work is easily damaged by air pollution in modernurban environments. Famous monuments, such asthe Parthenon in Athens or Trajan’s Column inRome, have been disfigured by deposits of dirt,while fine details of their carvings have disappearedsince accurate drawings and photographs weremade in the nineteenth century. These problems willonly be solved by a combination of science and goodenvironmental management.

4 The Environment(Shackley 1982; Bell & Walker 1992)

It is late November, outdoor work has come toa standstill. The winter rye has been sown andthe young plants cover the fields already with athin green carpet. Most of the cattle are stillkept outside as long as the weather permits.The longer they can graze on the stubble of thesummer crops, the better. It saves time in clean-ing out the stalls and having to transport thefarmyard manure to the fields. Moreover, onehas to be economical with winter fodder nowa-days. The pregnant cows have already beenbrought in to protect them from the early nightfrosts…The sheep, by day herded on the heath,are penned in during the nights in the outfield.The pigs fattened on acorns and beech-nuts inthe forest are now again grubbing around thehouse, rooting up the soil of the garden plotand feeding on the remains of last summer’syield. (Groenman-van Waateringe &Wijngaarden-Bakker 1987, 4)

These ‘reflections of a farmer in the early 10th cen-tury’ are included in the report on an early medi-eval site excavated at Kootwijk, in the Netherlands.

They provide an illuminating example of the valueof environmental studies, for every detail is basedupon evidence found on the site; they also empha-size the very human purpose of archaeological sci-ence.

The tendency to treat environmental archae-ology as a separate discipline obscures the diver-sity of specialist skills that it draws upon. Forexample, an archaeologist engaged in studies ofthe early Stone Age requires a detailed knowl-edge of the plant and animal resources availableto hunter-gatherers, an understanding of the pre-vailing climatic conditions, and informationabout human diet, diseases and life expectancy(Smith 1992). For this reason, from the earlynineteenth century onwards excavators of pre-historic cave sites paid close attention to animaland plant remains, as well as to human bonesand artefacts. A drought that lowered lake lev-els in Switzerland in 1853 led to the examina-t ion of many neolithic and Bronze Agesettlements where not only wooden and bone ar-tefacts were preserved, but also many remainsof plants. The plentiful occurrence of water-logged conditions in Scandinavia stimulated re-search into the environmental setting of humansin the past, notably through the definition of aseries of ‘climatic zones’ characterized bychanges in plant species since the last Ice Age(above p. 112). Awareness of this work ledGrahame Clark to excavate a waterloggedmesolithic hunter-gatherer site at Star Carr(Yorkshire, England) in 1949, with the expresspurpose of recovering a wide spectrum of botani-cal and zoological evidence to understand theeconomy and society of its inhabitants (Trigger1989, 264–8).

Environmental factors have not always beentaken sufficiently seriously in Roman and medievalstudies, whether in the interpretation of sites or inthe consideration of climatic influences upon his-tory. The archaeology of Roman villas was for manyyears dominated by comparisons of architecturalplans, wall paintings and mosaic floors, with littlereference to the agriculture that formed the basis ofthe site’s prosperity. Only rarely did excavationextend beyond the principal dwelling house to itsbarns and animal sheds, let alone to a full exami-nation of excavated animal bones and plant remains(Greene 1986, 89–94).


140

4.1 The concept of ‘sites’

The dubious significance and serious limitationsof what archaeologists call ‘sites’ has been dis-cussed in relation to landscape archaeology (above,p. 53). Few techniques used in environmental ar-chaeology are restricted to individual ‘sites’ orexcavations; humans are just as much part of awider ecological system as other animals or plants.Broader aspects, such as climate and vegetation,provide a background setting for human activity,and the primary sources of evidence for these as-pects do not come from excavations of habitationsites. More specific aspects, such as food gather-ing and production or the exploitation of naturalresources, can be studied by means of evidencerecovered from occupation sites, in the form ofbones, shells and any organic materials that hap-pen to survive.

4.2 The survival of environmentalevidence

Archaeology in the New World has always takena close interest in environmental aspects of sitesbecause of the many different ways that nativeAmerican cultures developed in diverse settings,from tropical rain-forests to temperate woodlands,and from plains to deserts. The superb preserva-tion of organic remains on settlement sites in ariddesert conditions in the south-west of the UnitedStates made the observation and examination ofenvironmental evidence an obvious part of archae-ology, in contrast to most European sites, whereorganic remains had disappeared through decay.The American sites also provided samples of an-cient timber that could be used in early studies ofclimatic changes reflected by variations in tree-rings.

Although bones are normally found in tre-mendous numbers on most archaeological sites,with the addition of shells near coasts, dampacidic soils are likely to destroy everything ex-cept burnt bone. The most favourable condi-tions are alkaline subsoils and well-drainedsands or gravels. Arid, waterlogged or perma-nently frozen conditions also assist the preser-vation of other organic materials besides bones,and whole bodies of humans and animals havebeen preserved in many parts of the world.

Egyptian mummies are widely known, butequally striking are some Iron Age bodies foundin bogs in Scandinavia and Britain, including‘Tollund Man’ and ‘Lindow Man’, two indi-viduals who both met a particularly grisly endas sacrificial victims. Frozen bodies, such as theGreenland mummies (medieval Inuits) or thelate neolithic ‘Ice Man’ from the Alps, havebeen found complete with clothing and per-sonal possessions that would have decayedcompletely under normal conditions. Amongstthe Ice Man’s property were a bag of driedmushrooms and some sloes, which indicate thegood potential for the survival of very delicatesubstances (Spindler 1994).

Plant remains are also preserved in arid or wetconditions: wood survives well, as do pips, seeds,and the fibrous matter from leaves, stalks, etc. Asurprising amount of information may still begained from sites with ordinary soils, too. It hasbecome common to be employ ‘flotation’ techniqueson excavations to improve the recovery of very smallbones from rodents, birds, reptiles and fish, as wellas small shells and the remains of insects and plants.Burning may convert plants into charcoal under theright conditions, and many species of wood, grain,and other plant material may be identified. All ofthese have considerable significance for the finalinterpretation of the economy and environment ofa site or other discovery.

Leather and textiles (some made from animalhair, others from plant fibres) were important rawmaterials in the past, but finds are absent from mostsites. The study of those that have survived is there-fore particularly significant if a fuller understand-ing of the exploitation of natural resources is to beachieved. The Basket-maker stages of the Anasazitradition in the American south-west take theirname from an organic raw material that normallyperishes under European climatic conditions. Everyarchaeologist should make regular visits to a mu-seum of ethnography or folk-life, and take note ofthe large number of significant items that would notsurvive on a normal archaeological site. Hairstyles,body paint, head-dresses and costume all play animportant role in the identity of cultures and per-sonal status (even in contemporary urban civiliza-tion), and it is impossible to have too manyreminders about this major loss of knowledge aboutthe past.


141

5 Climate(Roberts 1989; Wigley 1981)

Long-term climatic change has been a fundamen-tal factor in human development, seen at its mostdramatic during the periods of extensive glaciationthat have been known to geologists for more than200 years. Recent research has not only confirmedthe dates of major Ice Age episodes, but has alsogiven accurate indications of global temperatures.Evidence over a geological time scale comes fromvariations in oxygen isotopes in sea-bed deposits(above, p. 113), while more recent indications arederived from annual layers in ice-sheet cores (above,p. 113). Ice layers overlap with records from tree-rings, and the results can be correlated with precisedocumentary evidence in recent centuries.

While long-term change is obviously impor-tant from an archaeological point of view, short-term fluctuations may have had an importantimpact on human life in the past—especially infarming communities. Ice-sheet cores are veryinteresting in this respect, for they contain clearrecords of volcanic eruptions, represented bylayers containing high levels of dust and acid-ity. Volcanic ash in the upper atmosphere maycause severe disturbances to the weather byblocking solar radiation, and if these circum-stances were prolonged for many years theycould lead to changes in settlement patterns. Theabandonment of upland settlements in northernBritain around the twelfth century BC has beenattributed to this phenomenon, and supportingevidence comes from tree-rings that indicate aperiod of around twenty years’ appalling coldand wet weather from 1159 BC.

Vegetation is an important measure of regionalclimatic change that has direct archaeological im-plications. Plants are very sensitive to temperatureand moisture, and most species produce pollen.Fortunately, pollen grains resist decay well, so thatcores taken from bogs or lake beds contain excel-lent records of wind-blown pollen (below, p. 143).The general pattern of change since the last Ice Agehas been well known since the 1920s (see fig. 5.10),but samples are now dated by radiocarbon to pro-duce a detailed history of vegetation on a regionallevel. The interpretation of pollen analysis as anindicator of human, as opposed to climatic, influ-ences will also be examined below.

6 Rocks and soils(Davidson & Shackley 1976;Holliday 1992)

The earlier the period of archaeology that is be-ing studied, the more important geology islikely to be—particularly in phases related toIce Ages. Geology and geomorphology are es-sential for understanding the present landscapeand its past configurations, along with changesin sea level, erosion and the deposition of newland by sedimentation or volcanic activity. Thisinformation not only influences our conceptsabout the environmental context of human ac-tivities in the past, but also provides vitalinsights into the likelihood of finding sites andartefacts. The significance of early finds ofbones and flint tools in deep gravel beds wasonly fully appreciated when geologists hadstudied their formation and understood theprinciples of stratification. In cases such as thehand-axes discovered at Hoxne or in theSomme valley (above, p. 11), the artefacts andthe bones ‘associated’ with them had beeneroded from their original resting places andredeposited in river gravels. Geology is also akey component in the study of early hominidfossils in East Africa, where many finds havebeen made in layers of sediment separated byvolcanic material. It is very important to under-stand the stratification of these deposits, andto date them by the potassium-argon methodand magnetic reversals (above, p. 128); expe-ditions around Lake Turkana in Kenya are ac-tually planned according to the location ofdeposits already dated in this way (Leakey1992).

6.1 Soil science (Courty 1989)

A knowledge of geomorphology is also vital forreconstructing the wider environment of ancientsites in terms of natural resources. Surface depos-its and outcrops of rock, combined with evidencefor rivers and lakes, dictate the forms of vegeta-tion and animal life available to hunter-gather-ers or early farmers. An understanding of soilsadds further detail, for soils with differing col-ours, textures and other characteristics areformed and changed both by natural and human


142

activities (fig. 5.8). Maps of modern soils andtheir present agricultural potential are publishedin many countries, but they are not a reliable in-dicator of their state in the past. Modern field-work projects, such as the Neothermal Dalmatiasurvey outlined in chapter 2 (p. 54), take samplesof soils to determine the history of land-use. Soilscientists need to take deep samples and are par-ticularly keen to examine deposits that have beencut through by erosion or modern construction.When forests on hills are cleared for cultivationor grazing, an increase in erosion normally leadsto the deposition of sediments in valleys, cover-ing up earlier phases of valley-floor cultivationand settlement. Archaeological earthworks suchas ramparts or burial mounds usually preserve anearlier ground surface (see fig. 3.3) that may alsoprovide samples of pollen and/or molluscs (below,p. 151). Buried soils in these situations give im-portant information about the vegetation or formof cultivation that took place immediately beforethey were built. Dated structures (such as theAntonine Wall in Scotland, constructed in the

130s and 140s AD) act as a terminus ante quemfor a buried soil.

Other characteristics of soils give clear indi-cations of concentrations of settlement and ag-riculture, notably a high phosphate content(above, p. 50). Samples taken systematically overa wide area may help to define the limits of asettlement without extensive excavation; meas-urement of the levels of trace metals offers anaddition or alternative to phosphate testing(Bintliff 1990). The acidity of a soil is a usefulguide to the prospects for the survival of pollenand molluscs; if it is unfavourable, time need notbe spent on fruitless collection and processingof inadequate samples.

Soils also provide evidence of past climatethrough soil micromorphology (Courty 1989),which is based on microscopic analyses of soilstructures. The sizes and shapes of soil particlesdeposited by water during damp periods may bedistinguished from wind-blown material that ac-cumulated during periods of low rainfall. Soilsare classified into types that provide interestinginsights into human disturbance of the environ-ment. A good example is provided by ‘podsols’characteristic of heath and moorland. They onlysupport a thin surface layer of vegetation, andoverlie a layer of leached soil from which rainwater has washed iron and humus down to thesurface of the subsoil. However, the soils foundunder prehistoric earthworks erected on what isnow moorland are frequently ‘brownearths’typical of woodland, not podsols. Brownearthsare stable when covered by trees, but whenwoodland is cleared, rainfall causes a deterio-ration to poorer podsols. Thus the harsh inhos-pitable soil conditions characteristic of openmoorlands in much of upland Britain result fromhuman interference. Clearance for occupationand agriculture eventually exhausted them andrendered them uninhabitable for later farmers,and the process was possibly accelerated by thespell of rapid climatic deterioration caused byvolcanic activity (indicated by tree-rings: see p.111). The destruction of the world’s rain forestsin the twentieth century is leading to a similarresult, as the stable recycling of nutrients by treesis brought to an end, and alternations of extremewet and dry conditions break down the struc-ture of the soil.

5.8 The profile of a rendsina soil, commonly foundon chalk, limestone, or gravel, has a clear structureif it has been under pasture for a long time (left).Earthworm activity will have made stones sink to adistinct layer, along with any large artefacts, such ascoins or pottery, that have been deposited in thesoil. Since it takes a considerable time for thisstructure to re-form after it has been disturbed byploughing (right), a soil buried by an ancientearthwork can be examined to ascertain which kindof agriculture was responsible for its formation.Audio Visual Centre, University of Newcastle, after Evans1978, f ig. 29


143

7 Plant remains(Jones 1988; Pearsall 1989)

The conditions that favour the survival of plantremains have been indicated above (p. 86). Thelarger the sample, the more reliable the results oftheir study are likely to be. Botanical identificationsare time-consuming and expensive, but they areextremely important to the interpretation of an in-dividual site or vegetation in general. Nineteenth-century botanists concentrated on large fragmentsof plants, but the focus in the twentieth centurymoved to microscopic pollen grains. Large itemssuch as seeds and pips remain important, however,for they not only reveal the existence of plant spe-cies but also give insights into the collection andprocessing of wild fruits or crops from domesticatedplants. Finds of particular species of cereal grainshave implications for farming and harvesting meth-ods, and further enlightenment about soil conditionsis gained from studying seeds of weeds that grewamongst cereal crops (Veen 1992). One spectacu-lar example of the study of plant remains is the in-vestigation of the gardens of Pompeii, where thevolcanic eruption of AD 79 sealed vineyards, or-chards, vegetable plots and ornamental gardensunder a thick layer of ash (Jashemski 1979). Grape-pips, nuts and fruit stones were recovered, but abonus discovery was the existence of cavities in theearth where the roots of trees and other large plantshad decayed. It was possible to pour plaster intothese holes and then excavate the root system, whichcould be identified from its size and pattern. Oneopen area once thought to have been a cattle mar-ket was found to have been filled with vines andolive trees, and to have had open-air dining couchesamongst the foliage (Greene 1986, 94–7).

In contrast, indirect evidence of plants is alsorecovered in surprising ways. Impressions of grainare occasionally preserved on pottery; damp clayvessels were dried before firing, and their bases fre-quently picked up fragments of straw or grain fromdry material that was probably spread out to pre-vent them from sticking. The organic matter burntout completely during firing, leaving hollow voidsfrom which casts can be taken with latex or plas-ter. These are examined under a microscope to iden-tify the species present. Pottery can also be examinedto detect food residues absorbed into the clay dur-

ing cooking. Evidence for cabbage (or turnip) wasidentified from epicuticular leaf wax componentsat the Anglo-Saxon settlement at Raunds(Northants); no other evidence for soft plant tissueshad survived (Evershed 1991; 1992). Work on a siteon the Solomon Islands has recently demonstratedthat starch residues from plants are occasionallyvisible on stone tools, if they are examined underhigh magnification by a Scanning Electron Micro-scope (Loy 1992). This kind of evidence extends therange of information about plants to food prepa-ration and cooking.

The domestication of wild plants in the earlystages of settled farming is a process of profoundsignificance that may be investigated through theremains of plants. Research in Central America andsouth-east Asia indicates that a much wider rangeof species was involved than in Europe and the NearEast. Maize was a key food source in Mesoamerica,and finds of cobs in stratified deposits show thattheir size gradually increased. American food plantsare of particular interest because they spread to therest of the world after European contacts began in1492; the European diet would be very much lessvaried today without potatoes, tomatoes, peanuts,peppers and pineapples (Gowlett 1984, 170–1). Theprospects for future research are expanding nowthat methods for extracting ancient DNA fromancient seeds or cereal grains are beginning to suc-ceed (Allaby 1994). In a reflection on recentprogress, Jones concluded that:

…advances in molecular, chemical and radiomet-ric analyses have set a new agenda for bio-ar-chaeology. The direct examination of survivingfragments of past human food webs is graduallybeing liberated from those questions they wereleast suited to answer, and instead, throughanalyses with these newer techniques, can pro-vide the qualitative framework for a much morestructured examination of the particular place ofhumans in past food webs. (1992, 216)

7.1 Pollen analysis(Dimbleby 1985; Moore 1991)

The most productive technique that has been ap-plied to archaeological plant remains is undoubt-edly palynology—the study of pollen.


144

All hay-fever sufferers know that the air is fullof wind-borne pollen during the summer months.Fortunately for archaeologists each minute grain ofpollen has a tough outer shell of a different shapefor each species (fig. 5.9). These shells survive wellin soils whose acidity is sufficiently high to reducethe bacterial activity that would normally causethem to decay. The loss of pollen from alkaline soils,such as those of the densely occupied and farmedchalklands of England and northern France, is un-fortunate, but these soils are favourable to the sur-vival of molluscs that also provide environmentalinformation (below, p. 151). The toughness of pol-len grains allows them to be separated from sam-ples of soil collected on sites by straightforwardlaboratory methods, but they must then be identi-fied and counted under a microscope by an experi-enced palynologist—a very time-consuming task.Most grains are less than 100th of a millimetre indiameter (fig. 5.9); their abundance makes count-ing a tedious procedure, but it has the advantage

that statistically significant quantities are easilyobtained from small samples of soil.

Since palynology is able to monitor generalchanges in climate and vegetation over long peri-ods, it is of considerable interest to climatologists,ecologists, botanists and geographers as well as toarchaeologists. Samples of pollen taken from coresbored from deep peat bogs or lake sediments arestratified, with the earliest part lying deepest. Adeposit that has formed over thousands of yearsshould reflect overall changes from tundra to for-est or from forest to farmland, and indicate fluc-tuations in the prominence of individual plantspecies (fig. 5.10). Sufficient analyses have beenmade to give a fairly clear picture of the majorchanges of vegetation since the last Ice Age, and todefine a series of climatic ‘zones’ that formed avaluable form of dating before the arrival of theradiocarbon technique (above, p. 112). These zonesof climate and vegetation provide a general contextfor human activities, such as early Stone Age hunt-ing on the open tundra, or mesolithic hunting andgathering in forests. When a picture of backgroundvegetation is added to other plant remains, artefactsand animal bones from an excavated settlement,there is an increased possibility of accurate inter-pretation of past economies and the functions oftools and weapons. The application of palynologyis world-wide, and its value is not restricted to pre-historic times. It can be used to examine the envi-ronment of individual sites or regions in periodsbefore documents provided such information insufficient detail.

One key issue that may be studied throughpollen analysis is the appearance of the first set-tled neolithic farming communities. It representsa momentous stage in human development, bothin terms of exploitation of the environment andsocial organization. The neolithic economy re-quired permanent buildings to be erected and forland to be cleared of trees for pasture and arableland. The destruction of woodland is marked by

5.9 Different plant species have distinctive pollengrains whose tough outer shells can be identified byspecialists in the laboratory. This drawing showsimportant trees found in postglacial deposits: alder,birch, hazel, hornbeam, oak, elm, lime, beech andpine. Zeuner 1946, f ig. 21 (after Godwin)


145

a change in the ratio of tree-pollen (TP) to non-tree-pollen (NTP) in stratified deposits; thus,even if no neolithic sites have been discoveredin a particular region, the pollen record mayindicate their existence. TP declines, NTP rises,and tell-tale species of grasses and cereals appear,together with ‘weeds of cultivation’ that thrivedin the new conditions. Fine particles of charcoalmay also be detected, showing that forest clear-ance involved burning. The appearance of signsof a Neolithic economy in a pollen sample canbe dated by radiocarbon, and their presence mayprovide a spur to fieldwork to locate the settle-ment sites that belonged to the first farmingcommunities. In comparison with their use inprehistoric archaeology, environmental ap-proaches to historical periods are still in theirinfancy, but have great potential (Greene 1986,72–6; 126–8).

Because most pollen is deposited within a fewmiles of its source, it can provide a picture of theplant population in the immediate surroundings ofan individual site. A site that occupied a small clear-ing in a forest would have a high proportion of tree-pollen (TP) to non-tree-pollen (NTP), whereas asettlement in open country would show the reverse.TP and NTP may also be examined in terms of in-

5.10 Pollen diagrams are not easy to interpret, butthe method of presentation is similar to that ofseriation in that the thickness of the line for eachindividual species reflects its relative importance. Thisdiagram of vegetation in Jutland incorporates Dr JIversen’s results from numerous samples that revealedthe recolonization of Denmark by significant plantssince the last Ice Age. The earliest period is at thebottom of the diagram, where in stages I-III, grassesand small hardy trees (birch, willow, juniper) domi-nated the cold landscape. From V onwards, temperatetree species appeared as warmer conditions devel-oped, replacing open grassland. Human actions beganto have marked effects in VIII, when forest wascleared to supply fuel, timber and open land forcultivation and grazing. Grasses became commonagain, and signs of farming are evident from cerealsand the ‘weeds of cultivation’, such as plantain,associated with them. Audio Visual Centre, University ofNewcastle

dividual species or groups of related plants. NTPmay highlight different proportions of grasses andcereals that indicate the relative importance of graz-ing and grain production in the economy of a site,while pollen from plants such as legumes, flax andhemp may indicate other forms of food productionand raw materials for textiles. Samples of pollen


146

taken from soil buried beneath mounds or rampartsmay tell the archaeologist whether the land wasforested or covered with scrub before its occupa-tion. The soil that formed after the abandonmentof a site may show whether the land returned toscrub and then forest, or remained open, perhapsas part of the farmland of another settlement nearby.Further questions of direct relevance to an excava-tor may be answered by palynology. Mounds andramparts can be examined to see if their materialwas dug from the subsoil, in which case it will con-tain a mixture of contemporary and older fossilpollen, or whether they were formed by scrapingup turf or topsoil from the surface. This kind ofinformation may help in the interpretation ofditches, pits, etc., on a complex site, and clarify theirrelationships to the construction of earthwork fea-tures (Shackley 1982, 31).

7.2 Tree-rings(Schweingruber 1988)

Besides their value for dating (above, p. 109), tree-rings provide a continuous annual record of climate.The correlation between modern meteorologicalrecords of temperature and precipitation and thewidth of individual rings seems sufficiently close toallow them to be used to make estimates of condi-tions in the past before such records began(Schweingruber 1988, 170–5), but caution is stilladvised (Baillie 1992). At the opposite end of thescale, the pattern of tree-rings in an individual trunkis influenced by the location of the tree. Minor fluc-tuations in the immediate locality, such as fire dam-age, insect attack, clearance of surrounding trees,drought or flooding, may all leave tell-tale indica-tions in the rings (ibid. 176–83). A useful result ofthis degree of sensitivity is that timbers used in abuilding or a ship reflect the nature of the wood-land where they grew. Trees from dense forests dis-play different ring patterns from those that grew inopen spaces or hedgerows, for example. Ring pat-terns characteristic of a particular area allow theorigins of wood to be determined, revealing, forexample, that a Viking ship excavated in Denmarkwas constructed in Ireland.

An unexpected by-product of tree-ring datingis the detection of phases of exploitation of thelandscape, reflected by the age of tree trunks pre-served in river silts. During the Roman occupa-

tion of southern Germany, very large numbers oftrunks from mature trees up to 400 years oldended up in the Danube. It is likely that agricul-ture was intensified in response to the presenceof Roman forts and towns, and that there was anincreased demand for timber, both for buildingpurposes and fuel. These factors led to woodlandclearance and soil erosion, resulting in an increasein the amount of sediment that was washed intorivers. This caused flooding that swept awaymature trees growing some distance from thenormal course of the river (Schweingruber 1988,186).

Many other precisely dated climatic episodes aresuggested by tree-rings. That they are so closelydated provides an invitation for archaeologists toscrutinize all sorts of evidence to seek wider evidencefor changes caused by climatic phenomena. Evenmajor changes, such as the transition from huntingand gathering to farming in the British Isles, mayprove to have been associated with climatic eventsrevealed by tree-rings. Baillie concluded a paper onthese events with an optimistic judgement: ‘Thereappears to be unlimited potential for the reconstruc-tion of various aspects of past environmental changefrom tree-ring records’ (1992, 20).

8 Animal remains

8.1 Animal bones(Davis 1987; Rackham 1994)

The principal task of a zoologist confronted witha collection of ancient bones is to identify the spe-cies that are represented. The zoologist must haveexperience of archaeological samples, and mayneed to consult reference collections of bonesfrom other excavations. Domesticated animals inparticular differ considerably from their moderncounterparts, and closely related animals likesheep and goats are difficult to separate from eachother. Another important task is to estimate thenumber of animals involved. It is not sufficientsimply to count the bones, for while some animalsmay be represented by just one bone, there maybe many from others. It is customary to countspecific bones to estimate the minimum and maxi-mum number of individual animals required toproduce the sample; the larger the collection


147

available, the more accurate these estimates willbe. The approximate ages of individual animalsmay be ascertained by examining the state ofossification of particular bone structures, theeruption of teeth in jaw bones, and the amountof wear on teeth. Sex is more difficult to estab-lish, but statistical studies of large samples ofbones may help to divide them into groups ofdifferent sizes, of which the smaller is likely torepresent females (fig. 5.11).

Interpretation (Grayson 1984; Clutton-Brock 1988)It is important to understand the nature of anycontext from which bones have been recovered.Other finds, such as datable pot-sherds, may in-dicate whether it formed over a long or shortperiod, and the condition of the bones themselvesmay also tell something about the circumstancesof formation. Weathered, broken bones with signsof damage from rodents or scavenging animalsare readily distinguishable from those that wereburied immediately, and this information will ofcourse be valuable to the excavator as well as thebone specialist. Obviously, it would be a wasteof a specialist’s time to interpret bones from adisturbed deposit. It may be necessary to conductscientific tests to check the consistency of earlyprehistoric deposits that lack closely datable finds(above, p. 125).

Once reliable deposits have been recog-nized, and the species and numbers of the ani-mals identified, many further observations arepossible (fig. 5.12). A factor that needs to betaken into account is the effect of human se-lection on the sample of bones that survives.Hunters may well have butchered the carcassesof large animals where they were killed, leav-ing the majority of bones far from their livingsites (Smith 1992, 28–34; 105–7). Further-more, bones discarded by humans are fre-quently exploited by scavenging animals, suchas wild hyenas or domestic dogs; experimen-tal work has shown that different kinds ofbones have varying chances of surviving scav-enging (Marian 1992). Bones found on ahunter-gatherer site that was only occupied forpart of the year may give a very limited viewof the inhabitants’ exploitation of animals;other camps may have been associated with thesame groups of people at different times ac-

cording to the seasonal availability of animals,fish, molluscs and plants.

Where a site was a permanent human habita-tion belonging to a period after the introductionof farming, different questions can be asked aboutfood supply and diet. To what extent did the oc-cupants still exploit wild animals along with do-mesticated species, and how much meat could beobtained from the animals found? The age struc-ture of animal populations is particularly impor-tant for the analysis of farming practices. Theinhabitants may have enjoyed the luxury of eat-ing succulent young animals, or it may have beennecessary to maximize the use of cows and sheepfor milk and wool until they were several yearsold. Bones from towns present additional prob-lems: rubbish associated with houses may offer

5.11 Detailed measurements of neolithic animal bonesfrom a site at Troldebjerg, Denmark, revealed threedistinct groups that probably reflect differencesbetween wild and domesticated cattle, and separatesex groups within the latter. As with the characteriza-tion of obsidian (see fig. 5.3) careful statistical analysisof significant measurements, or ratios betweenmeasurements, may be required to reveal suchdifferences. Audio Visual Centre, University of Newcastle,after Evans 1978, f ig. 16


148

insights into diet, but care must be taken to dis-tinguish it from waste from butchers’ shops orindustrial workshops where bone was used as araw material for making artefacts. Sites occupiedover a long period with good stratified collectionsof bones offer additional possibilities for com-parisons between food supply and animal hus-

bandry practices at different times. However, caremust be taken to test the validity of any conclu-sions that are drawn, especially where collectionsof different scales are concerned; some straight-forward statistical probability tests measure thesignificance of any interesting observations (be-low, p. 155).

5.12 Studies of bones and other animal remainsfrom archaeological sites are of limited interestwithout a clear understanding of animal husbandrymethods. This flow-diagram presents the raisingand exploitation of cattle in early medieval Ireland;it is based on general principles of stock-manage-ment, combined with information about earlymedieval farming practices derived from contem-porary documents. An understanding of theprocesses involved may help to distinguish areas ofsites where specific activities, such as feasting,milking, butchering or craft production, tookplace. Mytum 1991, f ig. 5.8


149

The most important aspect of the study ofbones is that the nature of an excavated contextmust always be looked at very closely. Bonesfound on sites reflect living populations ofhunted or domesticated animals in differentways, and a sample recovered from an excava-tion may not be representative of the whole site.In one enlightening case in North America, theexcavated bones could be compared with thedocumented history of an eighteenth-centuryBritish fort. The number of animals representedby bones would only have been sufficient to feedthe fort’s occupants for a single day, but a gar-rison of varying strength was in occupation foreight years. The soldiers’ diet consisted largelyof boneless salt pork brought to the site fromelsewhere (Guilday 1970).

Besides identifying species and calculating thenumbers of animals represented by a collection ofbones, and their sex and age, specialists may beable to glean further information from bones.Hunting techniques may be deduced from injuries,and butchery practices are sometimes revealed byways that the bones were cut or broken. The ex-ploitation of animals for transport and traction hasimportant social implications. Evidence that horse-riding began around 4000 BC in the Ukraine wasfound through a study of surface wear on horseteeth, using a scanning electron microscope(Anthony Brown 1991) (fig. 5.13) Signs of rope-

marks on the horns of neolithic cattle found atBronocice in Poland (c. 3000 BC) probably indi-cate that the animals had been used to pull ploughsor vehicles. This interpretation was reinforced bythe discovery that many of the cattle bones fromthe site were from oxen, and that some were agedup to 5 or 10 years—much too old for the produc-tion of beef or dairy products (Milisauskas & Kruk1991).

DNA recovered from animal bones offers greatpotential for confirming difficult identificationsof species or sex, and for studying the processesof domestication by examining the genetic linksbetween wild and domesticated animals (Brown

5.13 Photomicrographs taken with a scanning electronmicroscope (SEM) provide images with a depth offocus that allows artefact surfaces, plant remains, etc.,to be studied in almost three-dimensional detail. Thetechnique has been used extensively in use-wearanalysis, and is proving valuable in the study of ancienttextiles, both for the identification of plant and animalfibres and for detecting signs of wear resulting frommanufacture or use. These images show heavily wornwoolen leg-bindings from Vindolanda, a Roman fortnear Hadrian’s Wall, enlarged 50 and 500 times. It ishoped that DNA studies may eventually add furtherprecision to the identification of the plants andanimals from which fibres were converted intotextiles. Dr J-P Wild/W D Cooke, Manchester AncientTextile Unit


150

1992, 19). Another fascinating line of research isoffered by the fact that animal blood and hairsmay sometimes be found on prehistoric tools. Aswith indirect evidence for plants (above, p. 143),DNA study of these traces extends the range ofinformation about animals to the stages of foodprocessing and consumption (Loy & Hardy1992).

In some cases studies of animal bones mergeimperceptibly into dating methods. Bones of pigsand elephants are much more common in East Af-rican geological deposits than those of humans,and their evolutionary development is well estab-lished. Thus, fleldworkers searching for humanfossils make a rough estimation of the age of adeposit without resorting to scientific datingmethods, and this will also highlight discrepan-cies between scientific dates and thebiostratigraphic date derived from the animals’evolutionary stages (Johanson & Shreeve 1991,92–101).

The following sections outline the study of fishbones and molluscs; space does not permit the dis-cussion of many other interesting kinds of ani-mal remains, such as insects and parasites. Theseremains reveal fascinating insights into the envi-ronment, living conditions, diet and health of hu-mans in the past (Buckland & Coope 1991; Jones1988).

8.2 Fish bones(Brinkhuizen & Clason 1986;Wheeler & Jones 1989)

Sieving and flotation techniques have improved therecovery of bones from small mammals, birds, rep-tiles, amphibians and fish. Unlike all the others, fishbones appear on archaeological sites on dry landas a direct result of human activity. Unfortunatelyfish bones have a much lower chance of survivalthan animal bones because of their small size andcartilaginous consistency. Interesting (if rather dis-tasteful) experiments have been conducted on thesurvival of modern fish bones that have passedthrough the digestive systems of pigs, dogs andhumans. Less than 10% of the bones of medium-sized fish survived, with the implication that theimportance of fish in the diet will be underestimatedon many sites, even when small fragments have beenrecovered by sieving (Jones 1986). If rats were com-

mon on a site, bones that they gnawed and digestedcould disappear altogether.

Nevertheless, the few bones that do surviveallow species to be identified; otoliths (‘earstones’) survive rather better, and they may beused to estimate the size and age of fish. Thisinformation gives insights into food gatheringstrategies, and the range of habitats reveals theextent of fishing, whether in local ponds andstreams or far out to sea. Otoliths also revealseasonal exploitation of fishing, because theyincorporate growth rings similar to those vis-ib le in mol lusc she l l s . Otol i ths f rom amesolithic midden at Cnoc Coig (Inner Hebri-des) demonstrated that most saithe (a cod-likefish) were caught in the autumn (Smith 1992,155–7). An indirect indication of fish and othermarine resources may be obtained by examin-ing carbon-12 and carbon-13 isotopes in thebones of humans or animals, for their ratio isinfluenced by a diet that contains seafood (be-low, p. 153).

8.3 Shells

Shells found in archaeological deposits fall into twodistinct categories. Some were brought to settle-ments from the sea-shore and discarded after theircontents had been eaten, and are informative aboutdiet and the exploitation of marine resources. Oth-ers belonged to land molluscs that lived on the site;many of these are extremely small and can only beseparated from samples of soil under laboratoryconditions, but they provide valuable insights intothe local environment.

Marine shells (Stein 1992)Large mounds of discarded shells (middens) arefound along many coastlines, providing evidenceof extensive marine exploitation in the past. Thefood potential of shells is fairly simple to calcu-late, but deeper insights may be gained by moredetailed observations. The size and shape of com-mon species, such as the limpet, show whetherthey were collected at random or whether particu-larly large examples were chosen at low tide; alimpet shell’s shape varies according to how farbelow the high water mark it lived. Non-randomcollection would obviously imply planned exploi-tation, perhaps indicating a greater dependence


151

on shellfish than on other food sources. Measure-ments of oxygen isotopes present in the edges ofshells reveals whether collecting took place allyear round or only seasonally (fig. 5.14). Sincethe proportions of these isotopes vary accordingto the prevailing temperature, middens associatedwith summer camps should show uniform pro-portions, while permanent sites should containthe range found in a whole year. As an alterna-tive to isotopic study, seasonal growth patternscan be seen in cross-sections of the shells of somespecies (Smith 1992, pl. vi b). Native Americansexploited soft-shell clams on the Atlantic coastsof North America, and the growth patterns foundin middens reveal a collection strategy based onintensive autumn harvesting (Lightfoot 1993).Oxygen isotope analysis of samples from deeplystratified sites may also document longer term cli-matic fluctuations that can be checked against thepresence or absence of particular species that aresensitive to temperature conditions. First-orderradiocarbon dating provides a useful means ofdating shell deposits found during fieldwork, ata fraction of the cost of conventional methods(above, p. 122).

Sea shells are not always evidence of diet,particularly when found inland. Large exam-ples may act as containers, spoons and eventools, while the Mediterranean murex providedpurple dye. Other exotic uses, such as charms,jewellery and even ceremonial trumpets, havebeen recorded by archaeologists and ethnogra-phers. Strontium isotope dating has demon-strated that spondylus shells found widelythroughout neolithic Europe really were mod-ern specimens gathered around the Mediterra-nean coasts , rather than foss i l examplescollected from geological deposits (Shackleton& Elderfield 1990). These examples demon-strate again that the scientific skills of marinebiologists are needed to identify shell speciesand subject them to various forms of analysis,but, as with other scientific and technical in-formation, the results require careful interpre-tation by archaeologists.

Land molluscs (Evans 1973)Land molluscs (mainly snails: Evans 1973) rangefrom large edible species to forms only visible andidentifiable with the help of a microscope. Species

recovered from ancient soils or geological depos-its reflect variations in the climate during succes-sive Ice Ages and warmer periods. They mirrorchanges in temperature in the same way as veg-etation (p. 144), and their distributions in the pastmay be compared with their modern habitats inexactly the same manner. Of more direct relevanceto archaeology is that hundreds of small shells maybe recovered from layers of soil. Samples are sortedinto groups of species that prefer grassland orwoodland, open or shaded localities, etc. Thus, thesnail species found in a ground surface buriedbeneath a structure such as a rampart or burialmound will indicate whether it was erected onopen heath (if light-loving grassland species aredominant) or in freshly cleared forest (if speciesthat live in dark and damp woodland conditionsare more numerous). A particularly gruesome il-lustration of this technique comes from the Romancity of Cirencester in Gloucestershire, where sev-

5.14 Analysis of mollusc shells can provide informa-tion about sea tempertures. The balance betweendifferent oxygen isotopes incorporated into thislimpet shell during two years’ growth showsfluctuations that reflect the temperature of the sea.Since the peaks of the graph indicate summers, theshell evidently stopped growing in winter. Thus,studies of large numbers of shells from humanoccupation sites can show whether harvesting tookplace all year round, or in one particular season.Audio Visual Centre, University of Newcastle, after Evans1978, f ig. 26


152

eral human skeletons were found in a roadsideditch within the city. The snail species found in-side a skull favoured a damp dark habitat, but notan underground one, implying that the body hadnot been buried, but had rotted where it lay in theover-grown ditch. This find clearly implies a de-cline in civic life at the end of the Roman period,and it has been suggested that the bodies were vic-tims of a plague that led to the desertion of the cityin the fifth century AD. This example typifies howscientific evidence may be used as the basis for ar-chaeological hypotheses that would be very diffi-cult to develop in any other way (Wacher 1974,313).

A further advantage of land snails is that theysurvive well on calcareous soils that do not favourthe preservation of pollen. In practice, they reflecta much more local environment, for pollen is scat-tered over many miles by wind. Ideally, bothsources of evidence should be examined togetherto establish the general and immediate environ-ment of a site.

9 Human remains(Ortner & Puschan 1985;Zivanovic 1982; Boddingtonet al., 1987; Price 1989)

The questions asked about human remains tendto be rather different from those asked aboutanimal bones or shells, which were normallydisposed of along with domestic rubbish. Evi-dence for early prehistoric people is very frag-mentary, especially the fossil bones from thegeological deposits in East Africa that are soimportant for tracing the emergence of modernspecies (Leakey 1992). Human remains wereregularly treated with more respect 35,000years ago, and even at this date complete bod-ies were buried with ‘grave goods’. Objectsplaced in graves help to date burials and mayindicate ritual activities or hint at the socialstatus of the deceased. Where soil conditionsallow, burials allow complete skeletons to berecovered for study, and this offers the possi-bility of establishing the cause of death, whichdemands the expertise of a pathologist. Thestudy of well-preserved bodies is like an exca-

vation itself, involving X-ray examination, dis-section and the study of all the materials en-countered, whether fibres of clothing, skintissues, or food remains (fig. 5.15). The mostfamous examples are of course Egyptian mum-mies (David 1987), but other notable findsrange from the bog bodies that are fairly com-mon in northern Europe, for example the IronAge Lindow Man from England (Stead 1986),to frozen bodies of medieval Inuit from Green-land (Hansen 1991) or the remarkable lateneolithic ‘Ice Man’ found in the Alps (Spindler1994). These bodies result from intentionalburial, ritual murder and accidental death; eachcategory offers different insights into ancientsocieties.

Multiple burials are common, where bod-ies have been jumbled together in collectivetombs over long periods, or where crematedbones were emptied into burial chambers in ir-retrievable confusion. To complicate mattersfurther, incomplete bodies were sometimesburied after the corpse had been exposed to theelements and scavenging birds and mammals.This practice is well known from native Ameri-can peoples and modern inhabitants of NewGuinea, as well as on many excavated prehis-toric sites. Expert work on burials at theAnglo-Saxon royal cemetery at Sutton Hoo(Suffolk) has demonstrated that careful exca-vation of soils of subtly differing colour andtexture may reveal outlines of decayed fleshand bones that have been destroyed by theacidic sandy soil (Carver 1992). Burials wherethe body was cremated, and the surviving frag-ments of bone were crushed and placed intoan urn or other container, are less favourablefor scientific study.

Given reasonably well-preserved remains, thetechniques employed in the study of human bonesare very similar to those applied to animal re-mains. Age may be estimated from a number ofosteological developments, such as the fusion ofthe skull bones and the growth of teeth. Sex maybe determined (with some difficulty) from thesizes of various parts of the skeleton, while stat-ure may be estimated from comparisons withmodern people—although discrepancies of sev-eral centimetres exist between different systemsof measurement. Pathologists examine deformi-


153

ties and evidence of disease ranging from malnu-trition, arthritis and dental decay to the erosionof bone through leprosy, as well as injuries, whetherhealed or fatal. Diet may be investigated throughthe analysis of carbon isotopes, or trace elementssuch as strontium, contained in bones; their ratiosor levels may indicate a preponderance of seafood,maize or rice. Samples derived from bone collagenlimit analysis to the last 10,000 years, but studies

of tooth enamel may allow these techniques to beextended to early hominids (Van der Merwe 1992).

Archaeologists need to scrutinize evidence pro-vided by the examination of human bones par-ticularly closely. It is very difficult to estimate theage structure and physical well-being (or other-wise) of a population because it is impossible totell whether burials recovered from a particularculture (if soil conditions allow their preserva-tion) represent the dead of all levels of society, orsimply a social élite. Were primitive people tall,healthy, ‘noble savages’, or diseased, short-lived,stunted individuals for whom life was ‘nasty,brutish, and short’? Roman gravestones fre-quently commemorate persons of advanced age,but memorials were presumably only erected forindividuals of high social standing whose lifestylefavoured longevity. Nevertheless, excavated re-mains and tombstones combine to confirm thegeneral impression that few individuals lived toa great age, while infant mortality was high andmany young women died in childbirth or as aresult of it (Hedges 1983). Only very rarely doarchaeologists uncover a large number of bodiesthat might represent a true cross-section of soci-ety—for example, a community of 486 nativeAmericans massacred around AD 1325 at CrowCreek, South Dakota (Willey 1992), or citizensof Herculaneum and Pompeii, who perished dur-ing the eruption of Vesuvius in AD 79. Even then

5.15 The ghostly image on the left, a computergenerated stacked-image axial scan of an Egyptianmummy, was produced using medical equipment. Thescanner records a series of sections through the bodyin digital form and stores them in a computer. Thedata may then be manipulated on a screen to presenta three-dimensional image (right) that can be rotatedfor viewing from different angles, or modified toshow tissues and bone of different densities. Thisinformation allowed an expert in facial reconstruction(Brian Hill, Newcastle Dental Hospital) to model thefeatures of the dead woman, last seen by embalmers3500 years ago. The finished head of Bakt-Hor-Nekht(with appropriate hair, jewellery and clothing) nowforms part of an attractive museum display that notonly generates public interest in Egyptology, but alsodemonstrates one way in which modern technologyenhances our perception of the past. Liz Watson,Hancock Museum; Michael Myres, Royal VictoriaInf irmary, Newcastle upon Tyne


154

there is always a high chance that young able-bodied individuals had already made their escape.

Ideally, evidence derived from the pathologicalstudy of human remains should be integrated withother information. A study of this kind has beenconducted in southern California, where sitesaround the Santa Barbara Channel dating from theeighteenth century AD back to 7200 BC were ex-amined (Lambert & Walker 1991). A decline intooth disease accompanied by a rise in arthritisseems to reflect a shift from the consumption ofvegetables and tubers to fish and sea-mammal meat,while an increase in deaths from infectious diseasesand poorer nutrition coincided with the growth ofsedentary villages. Evidence for violent conflictexisted at most times, represented by head injuriesand embedded projectile points, but it showed anotable increase when the bow and arrow wereintroduced around AD 500. Evidence for social andcultural change deduced from burials and settlementsites was also correlated with periods of tempera-ture change and drought known from tree-ring data.

9.1 Genetics(Jones 1993)

Work is progressing rapidly on the recovery ofDNA and blood proteins from bones or (whenpreserved) other body tissues (Brown 1992;Cattaneo 1991). At a basic level, DNA indicatesthe sex of a deceased individual—not alwayspossible from skeletal remains. It also offers thepossibility of studying whether bodies found in acemetery come from related family groups. Likeexisting studies of blood groups, this informationmight prove useful on a broader scale in chart-ing ethnic continuity or change-over periods whenartefacts seem to indicate the arrival of externalinfluences. Change may result from peaceful con-tact and trade, or migrations by invaders andsettlers (Sokal 1987). Great difficulties are in-volved, for it will take many years before suffi-cient samples have been studied to definerecognizable groups in historical periods; the situ-ation may be slightly easier in prehistoricpopulations that have not yet undergone toomuch confusing interaction. It is now well estab-lished that the genetic diversity of all nativeAmericans is so small that the entire populationof both continents probably descends from a rela-

tively small number of pioneers who crossed theBering Straits from Russia to Alaska in around20,000 BC (Jones 1993, 122–3). Argumentsabout the emergence and spread of early hominidsshould eventually be settled by means of mito-chondrial DNA inherited through the maternalline. At present it suggests that all modern hu-mans originated in a population living in EastAfrica little more than 100,000 years ago, andthat the rest of the world was gradually peopledfrom this source, replacing earlier species (Leakey1992, 218–28). Studies of DNA demand extremecare in the selection, preservation and handlingof specimens to avoid modern contamination(Hedges and Sykes 1992). Furthermore, ‘the in-terpretation must of course take into account thewhole range of archaeological evidence pertain-ing to the question being asked, and should alsoconsider ancient biomolecules other than DNA’(Brown 1992, 21).

9.2 The study of coprolites

Human coprolites (solid excreta) preserved on aridsites in the south-western United States, Mexico andSouth America have made a notable contributionto research into the natural resources available toearly native American cultures. The arid conditionsensure the preservation of fibrous matter that haspassed through the human digestive system, includ-ing fragments of bone, skin, scales, hair, feathers andmeat, as well as pieces of insects, parasites and theireggs. Plant fibres and seeds are also found, togetherwith microscopic pollen and ‘plant opals’ (distinc-tively shaped silica crystals formed by some plants).Even soft tissues from plants and animals can beextracted and identified by careful processing andsieving of rehydrated coprolites. Large collectionsrecovered from latrine deposits allow detailed sur-veys of the diet of the occupants of a site to be made.If deposits of different dates are recorded from aparticular site or area, long-term changes of diet maybe charted and related to variations in the availabil-ity of foodstuffs.

Coprolites may also contain eggs of parasiticworms that once infested the digestive tract of aliving human (Home 1985). Conversely, whereenvironmental conditions do not favour the survivalof coprolites, soil samples can be analysed to de-tect parasite eggs, whose presence may help to ex-


155

plain the function of latrines or pits. Parasites thatinfest a specific animal species are valuable indica-tors; eggs from a nematode associated with horseswere found in early military deposits on a Romansite at Carlisle, Cumbria, where some buildings hadalready been suggested to be stables (Jones 1988).

10 Statistics (Shennan 1988; Fletcher & Lock 1991)

Archaeology is full of intuitive statements based onexperience rather than calculation. Simple statisticsare useful for checking almost any statements thatinvolve comparisons, such as claims that the dimen-sions of a type of artefact change over time, or thatsettlement sites of a particular period tend to occupyone particular kind of soil. Probability testing isappropriate in these circumstances. In the latter case,the number of sites located on each soil type shouldbe counted, and a simple statistical analysis willcompare the totals with the numbers that wouldhave occurred if the distribution had been entirelyrandom. The results are expressed as a significancelevel; most sciences demand a level of at least 0.05,at which the figures observed have only a 1 in 20probability of occurring by chance. Statistical testsalso take account of the size of a sample, and datamay have to be rejected if the sample is too small.Similar tests may also be conducted on the distri-bution patterns of sites or artefacts, and again thebasis is a measurement of the difference from a ran-dom scatter. An understanding of probability is ofgrowing importance in radiocarbon dating, for thecalculation of dates from laboratory samples alwaysinvolves estimates of error. This has to be taken intoaccount when a calendar date is calculated from acalibration curve, for it has a built-in margin oferror.

Correlation is another common statisticalmeasure used by archaeologists. The relationshipbetween any two sets of numerical variables maybe tested by plotting them on a scatter diagram,and observing the pattern that results. A simpleexample might be to plot the length and breadthof Anglo-Saxon timber buildings (see fig. 3.30)to test the uniformity of their ratio. A straight lineon the graph would indicate that the builders

shared a uniform concept of proportions, irre-spective of size, whereas a wide scatter of vari-ables would indicate that buildings wereconstructed without any such guiding principles.The degree of correlation may be calculated, rang-ing from a maximum of 1.0 (if all buildings werealways exactly twice as long as they were wide)to 0.0 (if there was no relation at all). When thesample size is taken into account, the degree ofcorrelation may also be expressed as a level of sig-nificance. Julian Richards used extensive tests ofcorrelations and significance levels to investigateAnglo-Saxon burials, and discovered subtle rela-tionships between grave goods and the sex, ageor social standing of the deceased. Various fac-tors, such as the height of pots or their decora-tive motifs, were found to be significant in waysthat would not have been noticed without a care-ful numerical analysis (Richards 1987).

An awareness of probability and correlationalso leads to a better understanding of sampling.If a statistical study carried out on the dimensionsof a class of 400 axes, for example, found that theycould be divided into 3 separate classes accordingto their length in relation to their breadth, the re-sults would be expressed in terms of levels of prob-ability. This would estimate the likelihood that the3 divisions had occurred by chance in a sample of400 items. It would then be possible to work outthe minimum number of axes that would have tohave been measured to give a significant result.This is an important concept in planning research,whether it is related to objects or fieldwork; mostmodern field survey projects, such as the Dalma-tian project described in chapter 2, are designedwith a clear sampling strategy in mind (above, p.54). Sampling is the mathematical expression ofwhat is known colloquially as the law of dimin-ishing returns.

Scientific analyses carried out for purposessuch as the characterization of obsidian or clay(above, p. 135) produce bewildering columns offigures that can only be clarified by means of sta-tistical methods. Multivariate procedures havebeen designed to look for significant relation-ships or contrasts between elements or mineralsto define groups that may bring some order intothe data (see fig. 5.3). Multivariate statistics, no-tably cluster analysis, also lie behind some com-puterized exercises in the typological


156

classification of artefacts or the seriation of as-semblages found in stratified contexts or graves.The ready availability of powerful microcomput-ers facilitates and speeds up statistical analysis,so that a mass of confusing detail can be weighedup absolutely consistently. However, compli-cated computerized statistical procedures may bedangerous i f their results are accepteduncritically, and modern computer software al-lows sophisticated techniques to be carried outall too easily by archaeologists who lack any un-derstanding of statistics. Results are only as goodas the evidence they are based on, and the suit-ability of the tests that have been employed.Nevertheless, even a very basic awareness of sta-tistical concepts should encourage archaeolo-gists to design research questions in a way thatwill allow the results to be tested, and enable theoutcome to be stated as levels of statistical prob-ability, rather than subjective intuition.

10.1 Computers(Reilly & Rahtz 1992; Ross 1991)

Computers are so well integrated into archaeol-ogy, from prospecting and discovery to storageand publication, that it is no longer necessary todiscuss computing as a separate topic. Comput-ers have become so cheap, widely available and‘user friendly’ that their use requires no knowl-edge of electronics or programming. Many proc-esses that demanded the use of ‘mainframe’computers from the 1950s to the 1970s are nowperformed by an individual user’s desktop PC. Aremarkable combination of flexible software,memory, disk storage, processing speed, graph-ics and printing capabilities had become availableby the 1990s.

Most scientific dating methods, along with tech-niques used in the analysis of artefacts, employapparatus linked directly to computers that moni-tor the operation of the equipment and record andprocess the results. Computers not only save thetime of skilled laboratory staff, but are faster andmore reliable. Their importance is obvious in theestimation of the margin of error involved in meas-urements of radioactivity in samples of ancient car-bon, and the subsequent calibration of dates. The1993 issue of Radiocarbon included not only a re-vised calibration curve, but also a floppy disk with

a computer program to help with these complexcalculations! Computers are also an integral partof the technical equipment involved in processingaerial photographs and readings recorded duringgeophysical surveys (chapter 2; fig. 2.12). In bothcases the principal benefit comes from their abil-ity to ‘filter’ data by eliminating natural back-ground variations so that archaeological featuresare enhanced (above, p. 46). The adjustment ofoblique aerial photographs to fit an image on to auniform horizontal scale involves extraordinarilycomplicated mathematical procedures (Scollar1990).

Besides their involvement in statistics, labo-ratory science and cartography, the principalfunction of computers in archaeology is torecord, store and retrieve large quantities of in-formation, such as excavation records or mu-seum archives; this is a question of managementrather than science (Jones 1991). However, geo-graphical information systems (GIS) are a rathermore scientific application of computing thatcombines maps, environmental and archaeologi-cal data with statistical calculations to producegraphic visualizations of relationships betweenthese categories of information. GIS promises toprovide major advances in the analysis and in-terpretation of ancient landscapes (above, p. 56;see fig. 2.15).

11 Experimental archaeology(Coles 1979; Robinson 1990)

One welcome by-product of a scientific approachto archaeology has been the increasingly frequentuse of practical experiments to test hypotheses.Most have been one-off tests of specific ideas, but afew, for example the Butser Ancient Farm Project(fig. 5.16; Reynolds 1979; 1987), have developedinto long-term programmes observing a wholerange of variables over several decades. The strictdefinition of an experiment employed in the scien-tific world is rarely fulfilled in an archaeologicalcontext, for many factors are difficult to control ormeasure, let alone replicate on another occasion.Even when they are demonstrations or simulationsrather than true experiments, they may still producevaluable information.


157

11.1 Artefacts(Amick & Mauldin 1989)

Experimental archaeology is a useful compan-ion to scientific analysis in the study of arte-facts, for their composition and structure maysuggest methods of manufacture. Ancient tech-nology has been explored by reconstructions ofmetal casting procedures, the making and fir-ing of pottery, and various forms of stone work-ing. If one particular manufacturing techniquesuggested by an archaeologist is found to besuccessful in practice, the experiment only con-firms that it could have been used in the past,not that it actually was. For this reason it isimportant to adopt a more scientific approach;a single demonstration of one method of firingancient pottery is of limited value without aseries of comparative firings carried out usingdifferent fuels, kiln structures, methods of ar-ranging the pots in the kiln, etc. Again, it willnever be proved that the technique found to bemost effective today was the one employed inthe past, but the possibility of gross misinter-pretation will be reduced if some unsatisfactorytechniques are ruled out. It is essential that ap-propriate techniques, materials and equipmentknown to have been in use in the relevant soci-ety are employed.

Besides experiments concerning manufacture,the functions and efficiency of tools have also beenexamined. Again, scientific analyses may providerelevant information, and experiments are fre-quently carried out in association with microscopicuse-wear analysis. Patterns of wear or damage toworking surfaces may suggest how a tool was used;if a replica is used for the same purpose it can beexamined afterwards to see if the same effects areobservable. One danger inherent in experimentsabout the functions of artefact is that ‘efficiency’,measured in terms of time and energy, is a verymodern concept. Anthropological studies of pre-industrial societies reveal that ritual and symbolismmay override purely practical considerations. Inaddition, a twentieth-century archaeologist will notpossess either the physical fitness or the experienceof the original user of an artefact. Anthropologymight also help to limit the potential for other blun-ders contained in modern assumptions about thepurpose of artefacts. If a neolithic axe was assessed

purely according to its ability to cut through wood,the results would be meaningless if the axe was usedas a weapon, rather than a tool; perhaps forests werecleared by burning rather than being cut down.

11.2 Sites and structures(Morrison & Coates 1986)

Large-scale structures such as ships, buildings orearthworks may also be subjected to experimentsand simulation studies, normally by means of re-constructions based on evidence derived from ex-cavations. If possible this evidence should besupplemented with anthropological data, or, inthe case of more recent historical reconstructions,contemporary information and illustrationsfound in archives (see fig. 3.34). At the most ba-sic level, an exercise of this kind tests whether astructure postulated by an archaeologist couldhave stood up. Once this requirement has beensatisfied, other worthwhile questions include es-timates of human effort involved in construction,and the quantity and sources of the necessary rawmaterials.

The chief advantage of simulation studies is thatthey demand a much closer analysis of excavatedtraces than might otherwise be carried out.

5.16 The Butser Ancient Farm, Hampshire, is a large-scale, long-term experiment in which prehistoric andRoman agricultural practices and equipment areobserved under conditions of careful scientificobservation. Peter Reynolds


158

If the reconstruction of an excavated structureis found to be difficult or even impossible, the ex-cavator will be forced to review the evidence, andto look again for features that may have escapednotice or whose significance was not recognized. Incontrast, other excavated traces that have proveddifficult to explain may be clarified; a scatter of post-holes around a building may have supported scaf-folding or temporary props during construction, forexample.

Since archaeologists excavate the decayed re-mains of structures, the study of decay processes isfundamental to their interpretations, and manyexperiments have been designed to explore this area.In Denmark, reconstructed timber buildings havebeen burnt down and then re-excavated to learnhow the remains reflect the superstructure, to fa-cilitate the interpretation of ancient burnt remainson other sites. In England, long-term studies of theerosion of earthworks begun in the early 1960s willbe monitored well into the next century; at OvertonDown, Wilts, a bank and ditch were constructed ona chalk subsoil, and near Wareham, Dorset, a simi-lar structure was created on sand. The earthworkswere constructed to exact measurements, and ob-jects and other organic and inorganic samples wereburied at precisely recorded locations. Erosion, set-tlement, decay and the movement of objects byearthworms and other disturbances of the soil canall be charted by periodic excavations of small sec-tions of the earthworks. But because access to thesites is strictly limited, the effects of human erosionthat would have affected a real structure in every-day use have been excluded. Furthermore, moststructures in the past were probably carefully main-tained until obsolete, and not allowed to deterio-rate as soon as they were completed. These problemsillustrate the greatest weakness of experimentalarchaeology; experiments require controls, but pasthuman activity was not necessarily rational or con-sistent.

Many full-size reconstructions of ships have beenbuilt and tested since the nineteenth century. Greekand Roman warships have been particularly popu-lar because many accounts of naval battles survive

in classical literature. The most recent example isthe Olympias, an Athenian trireme constructed totest configurations of oarsmen and claims for thespeed and manoeuvrability made in Greek histori-cal texts (Morrison & Coates 1986; Welsh 1988).This ship raises a typical problem involved in re-constructions; many warships are illustrated inGreek paintings and carvings, and texts providedetails of the size of crews, but no illustration ordescription gives an unambiguous account of thearrangement of the oars and rowers within the hull.This problem is so serious that the validity of theentire experiment has been questioned by some crit-ics (e.g. Tilley 1992).

12 Conclusion

The scientific methods employed in archaeologicalresearch now impinge upon most areas of the sub-ject. The relationship between archaeology andscience remains clear, however; science suppliesincreasingly detailed and precise information uponwhich improved archaeological interpretations canbe based. Furthermore, the use of scientific evidenceand an awareness of scientific methods enhance thedesign and conduct of archaeological research.Archaeologists and historians ignore scientific evi-dence at their peril if they wish to understand thechronological framework of the past, or the mate-rial resources available to ancient societies, and thenatural environments where they lived. The devel-opment of new techniques, and the inevitable errorsthat they will contain at the outset, provides anopportunity for interaction between science andarchaeology when disagreements arise, for bothmust re-examine their own particular forms of dataand analysis. ‘For is it not, these days, a definingcharacteristic of real science that it istestable?…That archaeological science shouldsometimes give wrong answers, and that these canlater be shown to be indeed erroneous, must becounted one of the subject’s great strengths’ (Ren-frew 1992, 292).

Note: a guide to further reading that includes opicscovered in this chapter begins on p. 185.

159

Chapters 1–5 have clear subjects because they fo-cus upon specific principles and methods used inarchaeology, placed where appropriate into a his-torical perspective. Some references have been madeto changes in interpretation that have modified at-titudes to fieldwork, excavation or the collection ofscientific data. This final chapter will begin by look-ing more closely at the philosophy of the subject andexamine the importance of archaeological theory. Iwill then comment on a series of aspects of archae-ology that seem important in the closing years ofthe twentieth century. Some of them reflect my owninterests, and many may in the future be judged tohave been unproductive distractions.

Placing ‘theory’ into a separate chapter at the endof this book is an attempt to draw attention to itsspecial role. It does not mean that I consider theoryto be something separate from practice, any morethan I would draw a dividing line between scienceand dating, or excavation and fieldwork. If the ef-fect is still ‘to identify the discipline with its techni-cal instrumentation’—an accusation levelled at thefirst edition of this book by Shanks & Tilley (1987a,22)—this should be seen as a failure to devise abetter way of organizing the book, rather than anexpression of my own point of view.

1 Where is archaeology atthe end of the twentiethcentury?

Chapter 1 explored how ‘archaeology’ began inEurope after the Renaissance as a harmless distrac-tion for the educated rich. It commonly took theform of collecting, either of curiosities or works ofart; alternatively, archaeological sites might be re-corded, and even investigated by excavation, as partof an attempt to supplement the exploration of thepast through documents. Some aspects of archae-

ology, notably the study of human origins and earlyprehistory, developed into a respectable scientificpursuit in the nineteenth century. Archaeology un-derwent accelerated growth after 1900, so that to-day virtually every country in the world operatessome form of State-financed protection of ancientmonuments, as well as supporting research into thesubject in public universities and museums.

Archaeology has become a popular part of edu-cation at both school and university levels, perhapsbecause it involves a variety of practical and theo-retical work, and a mixture of scientific and aes-thetic approaches. It is also concerned with everydayobjects and structures, as well as the social élitesupon whom history has tended to concentrate.Museum displays are now aimed at general visitorsrather than specialists, and welldocumented ‘de-signer’ displays are displacing dull rows of pots withterse type-written labels. Mobility and leisure haveincreased to the extent that mass-tourism now regu-larly includes ancient sites and museums. Archae-ology receives extensive publicity through popularwriting and journalism; it even contributes to homeentertainment through television programmesshown at peak viewing times without causing anysurprise. This degree of familiarity makes it particu-larly important to keep on asking questions aboutexactly what archaeologists are trying to do, andfor whom; the answers are frequently uncomfort-able. Are we simply making what we want of thepast because we can do nothing about the present,let alone change the future?

1.1 Too much knowledge?

When human antiquity was established in themiddle of the nineteenth century, the academicworld was small and international, and not con-fined to universities and museums. John Evans,who visited Boucher de Perthes and witnessed thestratigraphic position of palaeolithic artefacts at

6 Making Sense of the Past

Making Sense of the Past

160

Amiens in 1859 (above, p. 13), was a busy pa-per-mill manager, and published articles andbooks on geology, pre-Roman coins, bronze im-plements and flints in his spare time (Evans 1943).As President of the Royal Society in London, henot only rubbed shoulders with most of theprominent geologists, physicists, biologists andother scientists of his day, but was acceptedamongst them on equal terms. The rapid ex-change of information and ideas between figuressuch as Evans, Lyell and Darwin that lay behindthis exciting phase in the development of archaeo-logical thinking would be more difficult today:cooperation between different disciplines tends tobe directed towards specific research projects withless ambitious objectives.

As the amount of information produced by ex-cavation, fieldwork and other forms of researchhas increased, so too have expectations about itsquality and detail. The many supporting tech-niques and sources of specialist information makeit more difficult, rather than easier, to complete areport for publication. The size and complexity ofthe end-product may be unwieldy, expensive, andimpossible for non-specialists to digest. Archaeolo-gists need to think about their duties to taxpay-ers, and find convincing justifications for yet moreanalyses of the material remains of the past. Au-thors of major archaeological reports have anobligation to communicate clearly without over-simplification, and to present the results in a formaccessible to the public as well as to specialists.Otherwise, what is the purpose of archaeologicalresearch?

Almost every subject in the humanities and sci-ences has undergone an information explosion inrecent decades. Librarians are particularly aware ofthe increase in the number and thickness of booksand periodicals, and of the search for alternativesto the printed page such as CD-ROMs, on-line com-puter databases and multimedia information stor-age and retrieval systems. It is difficult for specialiststo keep up to date in anything more than their ownnarrow field of research, and almost impossible forreliable ‘overviews’ or syntheses to be presented tothe public without gross simplification. These fac-tors have led to a fragmentation of academic archae-ology into restricted periods of the past, limitedgeographical areas or arcane philosophies of inter-pretation.

The difficulties that exist in communicating theresults of archaeology have undoubtedly contrib-uted to the flourishing of writers, such as Erich vonDäniken, who take a particular delight in deridingthe inability of ‘experts’ to find explanations thatseize the imagination of the public. AlthoughStukeley’s obsession with Druids demonstrates thatfinding supernatural or mystical connotations withantiquities is not new, views of this kind have neverbeen so widely disseminated (Williams 1991). Fewarchaeologists have sold as many paperbacks as vonDäniken; more recently, a meteorologist who linkedcrop circles to prehistoric ring-ditches or roundbarrows generated a reaction that no orthodox stu-dent of these monuments has ever achieved (Meaden1991).

2 Archaeological theory(Trigger 1989; Malina &

The abundance of information available about thepast, and the growth of unusual interpretations bypeople outside the academic archaeological world,require careful thought by archaeologists. The his-tory of the subject illustrates many attempts to es-tablish satisfactory conceptual frameworks intowhich to slot surviving remains, frequently in waysthat seem ludicrous today. However, the buildingof frameworks, whether deliberately or uncon-sciously, remains a prerequisite of serious research.Some archaeologists still consider that they exam-ine data objectively, and ‘let the facts speak forthemselves’. The majority now develop a conscioustheoretical approach and gather data or exploreexisting information with an explicit theoreticalframework in mind and a clear problem orienta-tion.

Theoretical frameworks now come and go withfrightening rapidity, according to the progress ofresearch or sometimes mere taste. Whereas in the1960s and 1970s the hottest area of debate wasbetween traditional and ‘new’ archaeology, discus-sion now centres upon the applicability of rivaltheoretical approaches. Regrettably, many archae-ologists, notably those who operate in familiar his-torical periods, still regard archaeological theoryas a sub-discipline that may be ignored. Their mis-


161

take is to overlook the fact that all investigationsof the past involve a theoretical perspective. Weare products of a social environment that has con-ditioned our outlook on the world, and our viewof the past is inevitably influenced by our percep-tion of the present. It is impossible to disentanglesubjectivity from supposedly ‘objective’ research,for our very choice of research topic probably re-flects our personal opinions. Awareness of ar-chaeological theory allows us to acknowledge thisproblem, even if there is very little that we can doabout it. Since most branches of archaeologicaltheory have their roots in philosophy or other dis-ciplines such as sociology or anthropology, theyoffer the possibility of gaining new and unexpectedinsights into familiar aspects of the past. Theorydoes not provide answers, but it suggests a widerrange of interesting questions that revitalize exist-ing data and stimulate a search for new and betterinformation.

This chapter will not attempt to explain the cur-rent state of archaeological theory, for this couldonly be an interim report. Indeed, it is a reflectionof the maturity of the subject that entire books arebeginning to be devoted to a retrospective analysisof archaeological theory as it has developed sincethe 1950s (Gibbon 1989). I will nevertheless con-sider the origins of some of the principal ideas in-volved, and provide some examples of theirapplication.

3 Social evolution(Friedman & Rowlands 1982;Renfrew 1984)

3.1 Early anthropology(Kuper 1988)

Before looking at theoretical archaeology, it is nec-essary to look back at early examples of the use ofanthropology or even abstract political philosophyto provide frameworks for the interpretation of thepast. At the most basic level, the way that earlyantiquaries drew analogies between artefacts fromthe New and Old Worlds is an obvious exampleof an informative interaction between the obser-vation of existing ‘primitive’ cultures and deduc-tions about the past (see fig. 1.10). It was notnecessary for antiquarians to make systematic in-

vestigations into anthropology, for classical Greekphilosophy contains extensive speculation aboutthe development of human societies and their tech-nologies (Blundell 1986). These writers influencedthe sociological approach of some French andScottish thinkers of the eighteenth century(Rousseau is the most widely known) who pro-posed schemes for stages of development in humansociety through hunting and fishing, pasturage,agriculture and commerce (Piggott 1976, 153).Any well-read antiquarian of the late eighteenthor nineteenth century who encountered ancientobjects and sites would probably approach theirinterpretation with a mixture of historical andbiblical knowledge, and some theoretical conceptof human development and progress. It is interest-ing to view John Frere and Boucher de Perthes inthis light (above, p. 11)

3.2 The impact of Darwin(Bowler 1989)

The Darwinian concept of evolution became uni-versally known in learned circles in the 1860s,even amongst those who rejected both its basisand implications. The great antiquity of humanancestors was established by geologists and ar-chaeologists at around the same time (above, p.15). As might be expected, Darwin’s biologicalconcepts were particularly welcomed by writerswho combined the evolutionary approach to pastsocieties with new anthropological and archaeo-logical observations. An American anthropolo-gist, Lewis Morgan, used his knowledge of nativeAmericans to define an elaborate series of stagesof development in a very influential book, AncientSociety (1877). This book took on wider signifi-cance when it was used as a source by Karl Marx;Friedrich Engels subsequently summarizedMarx’s interpretation of Morgan’s developmen-tal stages:

Savagery—the period in which man’s appropria-tion of products in their natural states predomi-nates; the products of human art are chieflyinstruments that assist this appropriation. Bar-barism—the period during which man learns tobreed domestic animals and to practise agricul-ture, and acquires methods of increasing thesupply of natural products by human agency.


162

Civilization—the period in which man learns amore advanced application of work to the prod-ucts of nature, the period of industry proper andart. (1884, quoted in Daniel 1967, 139–40)

3.3 ‘Culture history’(Trigger 1980)

It has been stressed in chapter 1 that the ‘evolution-ary’ approach to the typology of artefacts that ledto the acceptance of the Three-Age System in earlynineteenth-century Scandinavia was quite separatefrom Darwin’s, although both incorporated as-sumptions of linear progress and improvement(above, p. 28). Nevertheless, the combination ofevolutionary concepts related to animals (includinghumans), artefacts, economic systems and societiesmade a substantial contribution to twentieth-cen-tury archaeology. For example, in a popular work

entitled What Happened in History published in1942, the Australian Marxist prehistorianV.Gordon Childe employed chapter headings suchas ‘Palaeolithic Savagery’ and ‘The Higher Barba-rism of the Copper Age’. Childe was also influencedby historical accounts of the past; he classified re-current groups of related prehistoric artefacts andsettlements into ‘cultures’, and proposed that theyrepresented distinct ethnic or social groupings ofpeople. Support for this view could be obtainedfrom the work of archaeologists such as the Swede,Nils Åberg, who had studied some of the peoples(Goths, Franks, Saxons, etc.) involved in complexfolkmovements in Europe in the late Roman andearly medieval periods. Distinctive artefacts such asdecorated brooches, buckles or pottery found incemeteries could be matched with historical recordsof migrants or settlers in the areas where they werefound. This culture-historical approach will be ex-amined further below because of its contributionto nationalist and racist archaeology (p. 167).

4 Dif fusionism(Renfrew 1973)

The idea of diffusion supplied a critical connectingthread in earlier prehistoric periods where docu-ments were not available. Because of the geographi-cal focus of the Bible and Greek and Romanliterature, European scholarship had been hauntedby the civilizations of the Mediterranean and NearEast ever since the medieval period. Childe, likeMontelius and others before him, made an assump-tion that all innovations or improvements observedin European prehistory must have originated inthose areas where civilizations flourished at theearliest date (fig. 6.1). It must be emphasized thatthere was no logical inconsistency in asserting thatall impulses towards progress should have emanatedfrom those areas, particularly when Mesopotamiaand Egypt became better known in the nineteenthcentury. Diffusionism gained support from the con-cept of evolution, because it provided a powerfulmetaphorical explanation of ‘progress’ and the in-evitable spread of ‘improvements’ from advancedto less developed areas. This assumption becamefirmly rooted in twentieth-century archaeology,whether in a rational manner amongst scholars likeMontelius or Childe, or in the extreme views of EliotSmith, who claimed that the influence of Egyptian

6.1 Diffusionism: archaeologists like Oscar Monteliusor Gordon Childe envisaged a spread of culturalinfluences and innovations from the civilizations of theNear East into prehistoric Europe. This view wasbased on apparent connections between thetypologies of artefacts found in these regions, but, bythe 1970s, radiocarbon dates had broken thechronological sequence of links between stages 3 and4. A complete reconsideration of phenomena such asthe use of metals and the building of megalithic tombswas required when their origins and spread could nolonger be attributed simply to diffusion. Audio VisualCentre, University of Newcastle, after Renfrew


163

pyramids and embalming techniques extended tosimilar phenomena found in South America at amuch later date (Daniel 1981, 115).

The idea of diffusion illustrates the virtues ofexamining all aspects of archaeological explanationfrom a theoretical standpoint. The problem lay notin the idea of diffusion, for the history of technol-ogy provides many well-documented examples ofinvention followed by diffusion, but in the way thatit was accepted as a fundamental fact, rather thanas a theoretical proposition that might be supportedor refuted by further evidence. In consequence, itwas eventually rejected altogether. By the 1960sBritish archaeology was placed into a curious state,for many prehistorians considered that the majortechnological and social changes had resulted fromindependent development and internal evolution,rather than having been introduced by immigrantsor invaders. Ironically, the invasion-free centuriesof later prehistory were followed by periods markedby well-documented raids, invasions and coloniza-tion by foreign settlers: the Roman conquest in AD43; Anglo-Saxons in the fourth to fifth centuries;Vikings in the eighth to ninth centuries; and ofcourse the Norman Conquest in 1066. However,major changes in settlement sites, architecture andartefacts also took place in areas such as Ireland thatwere not occupied by Roman invasion or Anglo-Saxon settlement. Many of these changes happenedas a result of the arrival of Christianity, a phenom-enon of Near-Eastern origin that spread through-

out western Britain as a result of direct links withwestern Europe and the Mediterranean. This is aclear example of diffusion.

4.1 Diffusionism in disrepute:megalithic tombs(Renfrew 1983)

These large stone constructions are found in manyparts of western Europe, Scandinavia and Germany(see fig. 6.3), and they usually contain stone-linedchambers, some with long entrance passageways.Some are architecturally sophisticated, with enor-mous stones or elaborate vaulting, and they maybe decorated with geometric or curvilinear carving.The original covering mound of earth or stones hasfrequently disappeared, leaving distinctive uprightsand capstones that feature in many illustrations byearly antiquarians (fig. 6.2). Where the contents ofthe chambers survive, excavation usually revealscollective burials of large numbers of individuals.

Monumental stone architecture appeared inEgypt early in the second millennium BC and

6.2 This characteristic view of a megalithic tomb (KitsCoty House, Kent) makes it clear why they wereamong the first antiquities to attract attention andillustration, and why their size gave rise to tales ofconstruction by mythological heroes or giants. Theburial chamber has been exposed by erosion of anearth mound that originally covered the stonework.Stukeley 1776, II, pl. 32


164

reached remarkable sophistication in monu-ments such as Zoser ’ s s t ep pyramid a tSaqqara, c. 2650 BC. Tombs with stone pas-sageways and vaulted chambers exist in theAegean area around 1600 BC, and it was natu-ral to assume that these must have inspired themegalithic tombs that they resemble. StuartPiggott summarized the diffusionist interpre-tation in 1965:

Their distribution poses a problem: what likelyset of circumstances in antiquity can cause agroup of specialized ritual structures to be builtover such a sweep of territory? It is not, I think,inapposite to consider these monuments in thesame terms as one would the churches of Chris-tendom or the mosques of Islam… Creeds andbeliefs can be transmitted in many ways; con-quest or evangelism, fanaticism or fashion, bysaints or merchants…An eastern Mediterra-nean origin seems inherently likely, but is dif-ficult to document with any precision. (Piggott1965, 60)

However, the newly developed radiocarbondating technique was already beginning tocause problems for this kind of interpretation,because samples from megalithic tombs pro-duced dates around 3000 BC. Not only was thisembarrassingly early in relation to supposedprototypes in the Aegean, but it was even ear-lier than the first stone architecture in Egypt!It was suggested initially that there must beerrors in the dates from France. The diffusionistview survived into the second edit ion ofGrahame Clark’s authoritative World Prehis-tory (1969), but a new categorical statementappeared in its third edition (1977):

Despite…variations which were commonlyfavoured in particular regions, the occur-rence of collective tombs in most cases ofmegalithic construction over so great anextent of the Atlantic seaboard has oftenbeen attributed exclusively to seaborne dif-fusion. On the other hand the notion thatburial in collective tombs of monumentalconstruction originated in the east Mediter-ranean has recently been disproved by radio-carbon dating, (op. cit. 135)

Clark suggested that these tombs spread around thecoastal fringes by means of fishermen followinghake and mackerel, rather than the priests or trader/prospectors favoured by earlier writers. More fun-damental reinterpretation was made necessary bycalibration of radiocarbon dates according to thetree-ring curve (above, p. 117; see fig. 4.14). Manynew dates for megaliths were now available, includ-ing several from Britanny earlier than 4000 BC.Colin Renfrew stated the new position forcefully inBefore Civilization, a particularly stimulating bookon the impact of radiocarbon dating upon Europeanprehistory:

The implications are altogether clear: passagegraves, some with corbelled vaults, were alreadybeing built in these regions before 2500 b.c. inradiocarbon years, and thus by 3300 BC in cal-endar years. Quite obviously, if they have theirultimate origin in the Aegean, this must havebeen a long time before 3000 BC. Yet there areno collective built tombs in the Aegean until af-ter this date. The Breton dates, even without cali-bration, make nonsense of the diffusionist case.(1973, 89–90)

Scientific dating now shows that megalithic tombswere established in north-western France by 4700BC (see fig. 6.4; Scarre 1993). Calibrated radiocar-bon dates have also severed many other diffusionistlinks between the Aegean and the Balkans, westernand northern Europe.

4.2 New interpretations ofmegalithic tombs(Renfrew 1979)

The diffusionist account had been both intellectu-ally satisfying and methodologically sound. Ren-frew now proposed an alternative:

…we are no longer obliged to see the tombsas the result of a single movement, whether itoriginated in Iberia or in Brittany (as one couldnow argue, on the basis of the new dates).Instead our task is to create some social model,some simple picture of how it all came about.(1973, 124)


165

He divided the area where megaliths are foundinto four or possibly five regions where an independ-ent local origin could be claimed, and stressed thatthe tombs might be the only thing that they had incommon (fig. 6.3). The hypothesis that such tombscould be the products of small farming groups wasdeveloped and then tested against the geography oftwo Scottish islands where suitable tombs occurred;information about social structures and communityburial practices was supplied by studies of a mod-ern ‘megalithic’ culture in Borneo. Although Ren-frew’s fieldwork and anthropological analogiesdemonstrated that his explanation was possible, itwas always accepted that further analogies, new

fieldwork or excavation might generate alternativehypotheses.

Renfrew also tackled the problem of generalexplanation—why should similar monuments de-velop in as many as five different parts of the

6.3 Distribution map of megalithic tombs. Tholoi aretomb chambers with corbelled roofs originallythought to have been derived from examples aroundthe Aegean, but they are now known to be muchearlier. Radiocarbon dates suggest that the earliestmegalithic tombs are in Britanny, but rather thansimply proposing that their diffusion originated there,Renfrew has proposed five possible regions wherethey might have developed independently. Audio VisualCentre, University of Newcastle, after Fox 1959, f ig. 7


167

fringes of Europe (but nowhere else) at around thesame time? He proposed (very tentatively) thatpopulation pressures developed as farming com-munities reached the coastal limits of Europe,where they were influenced by indi genous hunter-fisher peoples; in other words, similar circum-stances produced similar results in separate, butcomparable, areas:

To make the coincidence understandable, all wehave to do is show that the development ofmonumental building in this way was, in thespecial circumstances of each individual area, thenatural result of intelligible processes operatingmore generally…. we can now begin to talkabout these monuments in human terms, as aproduct of living communities, and to give fullcredit to their builders, the world’s first archi-tects in stone, without any longer appealing byway of explanation to the convenient arrival ofwise men from the east. (op. cit. 142; 146)

I have discussed megaliths in some detail, becausethey demonstrate many of the problems involvedin making sense of the past. First, the impact ofradiocarbon dates and their calibration into ear-lier calendar years show just how quickly acceptedinterpretations may come crashing down in ruinswhen one prop is removed (fig. 6.4). Second, thishappened at a time when new approaches and pre-occupations were emerging amongst anthropolo-gists and archaeologists in the United States.Renfrew’s reinterpretation of megalithic tombsshows how a theory is tested in a real situation,and illustrates the continuing importance of an-thropological analogies. His ideas are not a re-placement for the diffusionist explanation, for anyhypothesis is liable to be reexamined, and, if nec-essary, rejected. This constant uncertainty, al-though very good for archaeology itself, increases

the difficulty of communicating comprehensibleresults to the public:

Diffusionist explanations may have been in er-ror, but they may also have been necessary, aninescapable stage in the progress of the disciplinethat enabled it to advance from the 1920s to the1950s and 1960s when independent chronolo-gies at last liberated European prehistory fromdependence on an axiom that had been untestedbecause it had been untestable. (Chippindale1989, 24)

5 Nationalism and racism(Gathercole & Lowenthal 1990;MacDougall 1982)

Concepts of social evolution and diffusion did notalways bring positive changes to archaeologicalinterpretation. Although Darwinism reinforcedrather than initiated racist interpretations of thepast, evolution and diffusion influenced the politi-cal outlook of the leading European governments.The situation in England in Pitt Rivers’ time hasbeen summarized forcefully by Donald Johanson:

For a society bursting with the self-congratula-tory fervor of a colonial power and the potencyof an industrial revolution, it took very little ef-fort to prop this biological premise under a so-cial construct and extend the ladder just a bit tojustify the white man’s inherent superiority overthe ‘lesser’ peoples of the earth. (Johanson &Shreeve 1991, 45–6)

Pitt Rivers did not devote his extensive intellectualenergy and financial resources to excavation andthe typological study of artefacts for purely aca-demic reasons. He considered that evidence for thegradual evolution of artefacts should be impressedupon museum visitors to counteract revolutionarytendencies in nineteenth-century politics (Bowden1991, 141–2). John Lubbock, a close associate ofDarwin, combined archaeology, anthropology andrigid evolutionary theory in his influential bookPrehistoric Times (1865), and concluded that non-European cultures were biologically inferior, andthat the most primitive peoples were doomed to

6.4 This diagram illustrates the stages of typologicaldevelopment of megalithic tombs in Brittany, alongwith radiocarbon dates expressed in calibrated yearsBC. It is superimposed upon a map of Quiberon Bay,where some of the best examples are found, andaccompanied by a typology of local pottery formsthat would help to relate undated tombs or settle-ment sites to the dated sequence. Boujot & Cassen1993, f ig. 3


168

inevitable extinction through natural selection.Thus, no moral responsibility need be felt for theirdecline or loss of identity through colonization,since they resulted from innate biological differ-ences (Trigger 1989, 110–18). Interestingly, AlfredWallace, the codiscoverer with Darwin of naturalselection, had first-hand knowledge of tribal peo-ples in South America and Asia, and consideredthat their mental abilities matched those of Euro-peans (ibid. 113). Pitt Rivers was more typical inhis view that savages were not capable of benefit-ing from the civilizing influence of superior racesby any means other than slavery (Bradley 1983,5–6). More recent uses of images of the past thatreflect or justify the political preoccupations andactions of the present will be explored in more de-tail below (p. 175).

The failing grip of the Ottoman Empire in thenineteenth century stimulated the exploration ofGreek civilization. Greece gained its independ-ence in 1831 and foreign excavators rapidlycleared the Acropolis of Athens, releasing theremains of such buildings as the Erechtheum andParthenon from the encumbrances of a harem andmosque respectively (see fig. 1.13). This actionillustrates the fact that archaeological research ishighly selective when combined with nationalism.Societies select the past that they wish to empha-size; the removal of physical reminders of Turk-ish rule and its religion, Islam, allowed the newGreek nation to underline its connections with theEuropean roots of classical culture (McNeal1991). Ironically, 150 years later, Bernal has takenan opposite view in stressing the non-Europeancontribution to ancient Greek culture, in a ma-jor work of scholarship with the arresting titleBlack Athena: The Afroasiatic roots of Classicalcivilization (1987; 1991).

North European archaeologists, notablyGustav Kossina (1858–1931), used recurrentgroups of distinctive prehistoric artefacts in aninnovative manner to define cultures of the kindobserved by ethnographers amongst living soci-eties. It was a small step from classifying culturesin material terms to equating them with social orpolitical units and calling them tribes or peoples(Trigger 1989, 148–206). Since most of the his-torically attested nations and linguistic groups ofEurope clearly originated in an undocumentedprehistoric period, nineteenth-century concepts of

nationalism were soon projected back into ear-lier times. Thus, a novel variant of diffusionismfrom the East was propounded by Kossina, whotraced all major European developments back toorigins in prehistoric Germany, credited to a pureNordic (‘Aryan’) race (Trigger 1989, 163–7). Hisviews were developed with enthusiasm after hisdeath because they gave welcome support to Naziclaims for the superiority of the Aryan race. Theoutcome of this enthusiasm was of course theinvasion of neighbouring countries to restoretheir rightful Teutonic ownership, and the attemptto enslave or exterminate inferior peoples, nota-bly Slavs, Jews and gypsies.

The extreme nature of this example should notblind us to the lesser prejudices contained in mostresearch; how many archaeologists considered thefeelings of native Americans or Australians whenexcavating the bones of their ancestors and plac-ing them on public exhibition in museums? Whydo pictorial reconstructions of prehistoric life al-ways seem to show men hunting or using toolswhile women cook and weave? The protests ofnative peoples and feminists have only very re-cently begun to force archaeologists to review theirperspectives on these aspects of the past, eventhough most had already done so in relation tonationalism and racism. The past is indeed pliableand potent.

6 Towards processualistarchaeology(Willey & Sabloff 1980;Lamberg-Karlovsky 1991)

The development of functional and thenprocessual approaches to archaeological datarepresented a replacement of the increasinglysterile preoccupation of culture-historical ar-chaeology with ethnicity by a vital new inter-est in how prehistoric cultures operated andchanged. (Trigger 1989, 288)

This quotation demonstrates simultaneously twoimportant facets of the study of theoretical archae-ology in the late twentieth century. First, it under-lines the continued importance of taking ahistorical view of the development of the subject


169

to place it into context, both in terms of its owncontent and its relationship to other subjects. Sec-ond, it underlines that it is necessary to acquire anextensive new vocabulary to participate in thesubject, for much of the discussion is conductedat a philosophical level, and is full of scientific orsociological terminology that makes it inaccessi-ble to outsiders.

The point that Trigger makes in this quotationis that much of the archaeology of the nineteenthand early twentieth centuries was dedicated to de-scription rather than exploration or explanation.This was of course inevitable in the early stages,because little or no meaningful explanation couldtake place until questions about dating had beensorted out. However, once the main sequences ofartefacts and cultures had been established, andafter they had been placed into a chronologicalframework through documentary evidence or ra-diocarbon dating, they could be studied in muchmore imaginative ways. In the ethnic approach, therelationships between cul-tures were discussed interms derived from historical archaeology: inva-sion, colonization, trade, etc. Alternatively, the de-velopment of cultures through time could be seenin evolutionary terms, from a biological, ecologi-cal, social, political or economic point of view.Thus, strict Marxist archaeologists saw in sites andartefacts the physical evidence for a progressionthat they already believed in on philosophicalgrounds.

6.1 Social evolution and Marxism(Bintliff 1984; McGuire 1992)

Abstract concepts such as evolution (whether bio-logical or social) and Marxism could be very com-forting for prehistoric archaeologists faced withcomplicated typologies or sequences of cultures.Evolution explained change in terms of competi-tion between humans and other animals, accom-panied by a need to adapt to the naturalenvironment. Marxism placed more emphasis onthe economy, and its influence upon ritual andsocial systems; it was the stresses between con-servative and progressive social forces that pro-voked change, and the experiences of the RussianRevolution in 1917 demonstrated that these couldbe rapid and dramatic. It is no coincidence that VG Childe talked of a Neolithic revolution (the

adoption of food production) and an urban revo-lution (settlement in towns rather than villages) inprehistory, both involving technological and socialchange.

The adoption of abstract evolutionary and po-litical concepts did have practical implications forarchaeology, however, for both emphasized theimportance of material evidence. After all, exca-vated sites could provide evidence for economicactivities such as farming, crafts and trade. Socialsystems should also be reflected in the dispositionand form of houses, and in the diversity of objectsfound in settlements or graves that indicated differ-ences in status or wealth. Evolutionary interestsstimulated the study of aspects of sites and theirsurroundings that could reveal the relationship ofthe inhabitants to their natural environment; pol-len analysis, soil science and the identification ofseeds from food plants all gained enhanced interestin the context of ecology and adaptation. Broaderclimatic and environmental factors could also helpto explain changes in human societies and econo-mies.

Thus, the middle decades of the twentieth cen-tury witnessed many overlapping schools of thoughtabout anthropology, archaeology and the goals ofthese subjects. However, most of these approachesdemanded closer attention to the excavation andrecording of sites and artefacts to fit them intobroader ‘functionalist’ views of the past (Trigger1989, 264–88). American archaeologists were at anadvantage, for they were concerned with indigenouspeoples unconnected with the Old World, some stillfollowing non-European lifestyles well into thetwentieth century. New World agriculture and civi-lization had developed independently (except in theminds of a few ‘hyperdiffusionists’ who claimedcontacts with ancient Egypt or other exotic cul-tures), but their study had been ignored by mostEuropean archaeologists. It is not surprising to findthat North America in particular became a vigor-ous debating ground for new approaches to ar-chaeological study that influenced the thinking ofthe rest of the world.

6.2 The ‘new’ archaeology(Clarke 1978; Binford 1972; 1989)

The ‘new’ archaeology has to be examined (withthe benefit of hindsight) in the light of


170

postprocessualism, which had become well estab-lished by the 1990s. Just as Wagner’s ‘new’ musicresembles that of other late Romantic composerswhen compared with what Schoenberg orStravinsky composed soon afterwards, processualarchaeology blends into the ecological and func-tional approaches that preceded it. It also bears simi-larities to some of the objectives of Russianarchaeology that developed quite independently inthe 1930s (Trigger 1989, 326). Perhaps the mostimportant key-words of the ‘new’ archaeology are‘system’, ‘hypothesis-testing’, ‘laws’, ‘process’ and‘explanation’.

Systems theory (originally developed in the1940s) was used to study interactions between asociety and its environment, and they were fre-quently presented in the graphic form used for elec-trical circuit diagrams. The details of the diagramwould be constructed from precise observations ofarchaeological and environmental evidence, and theresult might use the concept of ‘feedback’ to assesshow an interaction might cause changes in the sys-tem. Interpretations of change based upon systemstheory favoured modifications within societies, andhelped to discourage ‘historical’ explanations thatinvoked diffusion, invasion or migration. The for-mulation and testing of hypotheses reflect the grow-ing interest in scientific methodology, combinedwith the use of statistics, that characterized most ofthe social sciences in the 1960s. This approach con-trasted with the traditional method of reachingconclusions intuitively after looking at the data; inthe words of Lewis Binford, a leading light of the‘new’ archaeology:

What is argued…is that the generation of infer-ences regarding the fact should not be the endproduct of the archaeologist’s work… once aproposition has been advanced no matter bywhat means it was reached the next task is todeduce a series of testable hypotheses that, ifverified against independent empirical data,would tend to verify the proposition. (1972, 90).

Furthermore, just as in experimental science, thepurpose of hypothesis-testing was the establishmentof generally applicable laws:

In our search for explanations of differences andsimilarities in the archaeological record, our ul-

timate goal is the formulation of laws of culturaldynamics. (ibid. 100)

Doubts about the two-fold assumption that laws ofthis kind not only existed, but could be detectedarchaeologically, prevented many ‘old’ archaeolo-gists from embracing ‘new’ archaeology. Subse-quently, doubt has also been cast on the empiricalnature of the archaeological record and our abilityto interpret it.

Binford’s writings reflect an almost evangelicaldesire to propagate better modes of thought, be-cause archaeology could provide a depth of infor-mation about the past that anthropologists couldnever achieve, for the simple reason that they didnot live long enough to witness the unfolding of amajor process. He expressed dissatisfaction withthe chosen restrictions of traditional archaeologyand anthropology, and emphasized the role of‘theory’ as a conscious working process thatshould be combined with precise scientific think-ing. The sequence of the essential components ofhis approach are condensed into the followingparagraph, which is, incidentally, typical of hisdifficult written style:

Process, as I understand it, refers to the dynamicrelationships (causes and effects) operativeamong the components of a system or betweensystematic components and the environment. Inorder to deal with process we must seek expla-nations for observed phenomena, and it is onlythrough explanations of our observations thatwe gain any knowledge of the past. Explanationbegins for the archaeologist when observationsmade of the archaeological record are linkedthrough laws of cultural or behavioural function-ing to past conditions or events [my italics].(1972, 117)

Sceptics disagreed; Glyn Daniel predicted that the‘new’ archaeologists of the 1960s would becomethe disillusioned men of the 1980s and 1990s, prin-cipally because the pursuit of laws of cultural dy-namics was doomed to failure. However, mostarchaeologists would give the ‘new’ archaeologycredit for improving the recording and descriptionof archaeological information, because any at-tempt to reconstruct social and ecological systemsdemanded high-quality data. One result was rig-


171

orous analysis of archaeological sites, partlythrough better techniques of excavation but prin-cipally through an improved understanding of howsites were formed. The study of formation proc-esses (taphonomy) applies not only to the under-standing of artefacts and structures, but also thecontexts where economic and environmental data,such as bones or plant remains, are found. MichaelB Schiffer was a prominent early writer ontaphonomy (1976; 1987), and has subsequentlyedited a long series of volumes devoted to bring-ing together critical studies of archaeologicalmethod and theory. Studies of formation processesare taken further by archaeological experimentsinvolving reconstructions of ancient buildings, orreplications of crafts and food preparation meth-ods. Scientific observation of the processes in-volved and the residues that they leave behind mayhelp to illuminate remains encountered on anarchaelogical site (above, p. 157).

Processualism did not only look towards an-thropology and prehistory. Stanley South pio-neered the search for pattern recognition inartefacts from historical sites in North America,using rigorous data collection and statistical analy-sis (1977). He studied the changing nature of his-torical archaeology itself by drawing a seriationdiagram of the gradual emergence of publicationson theory and method, and the consequent reduc-tion of purely narrative site reports, in historicalarchaeology publications from 1960 to 1975 (ibid.20, fig. 3). His diagram is enhanced by the addi-tion of a snail, whose some-what bemused expres-sion parodies the reaction of many historians tothe move from a ‘traditional particularistic para-digm’ to an ‘emerging nomothetic paradigm’.Binford recognized the great interest of South’sbook to anthropologists concerned with other timeperiods, and concluded his foreword with thewords ‘Welcome historic sites archaeology to thescience of archaeology’ (xii).

7 Ethnoarchaeology(Hodder1982a; 1982b;Gould 1990)

One of the disadvantages of experimental ar-chaeology is that it is conducted by people withmodern ideas about work and efficiency. An al-ternative approach is to study living peoples,

and to incorporate additional informationmade available by the possibilities of discuss-ing artefacts, structures and processes with thepeople involved. Anthropological observationaimed at the understanding of the nature of ar-chaeological ev idence i s of ten cal ledethnoarchaeology. The demand for better un-derstanding of archaeological sites and socie-

6.5 Ethnoarchaeology: postgraduate students fromCambridge University observing a smith of the Tugentribe (Barengo district, Kenya) making an iron spear.Studies of the physical traces left by such activitiesmay help to elucidate Iron Age sites in other parts ofthe world, while an understanding of social and ritualaspects of craft production may broaden our conceptsof economic activities in pre-industrial, non-westernsocieties. Ian Hodder


172

ties caused Lewis Binford and Ian Hodder toconduct ethnoarchaeological investigationsamongst the Inuit (Eskimos) in Canada (Binford1978) and in several parts of Africa (fig. 6.5;Hodder 1982b). Such studies are a matter ofurgency, of course, for the number and rangeof ‘primitive’ lifestyles still available for obser-vation are diminishing rapidly. The implica-tions of ethnoarchaeology for taphonomy arenot always encouraging, for they underline thatmany important social and economic activitiestake place without leaving any helpful physi-cal traces; such difficulties must be recognizedand faced if archaeology is to progress.

Middle Range Theory involves the study of siteformation processes together with the identifi-cation, through ethnoarchaeology, of physicalevidence that reflects human behaviour (Trig-ger 1989, 361–7). This approach has had a con-siderable impact on the study of early hominidsin East Africa. Donald Johanson witnessed atfirst hand Binford’s intrusion into the smallgroup of specialists who had been working inEast Africa for many decades ‘…with thesweeping bravado of a Broadway veteran cometo show the yokels how to act. “It’s clear fromthe bone distributions and the way the limbshad been busted up,” Lew told me on the site.“The Olduvai toolmaker was no mighty hunterof beasts. He was the most marginal of scaven-gers.” So much for romance’ (Johanson &Shreeve 1991, 233; 235).

Thus, one side effect of the ‘new’ archaeologywas a revitalization of the relationship that hadexisted between archaeology and anthropologysince the nineteenth century. Anthropology is anideal source of information for generating hy-potheses and for offering explanations, for it mayprovide a comparative approach to activities suchas trade, agriculture, or burial, and the ways thatthey are carried out in societies with different hi-erarchies, religions and resources. Thus, Ren-frew’s propositions about the construction ofmegalithic tombs in neolithic Europe incorpo-rated observations about modern communities inBorneo (above, p. 165). However, most ‘primi-tive’ peoples who survived into the twentieth cen-tury had undoubtedly been affected by contactwith Europeans, even if indirectly. For this rea-

son explanations based upon modern anthropo-logical research that are applied to prehistoryshould only be regarded as hypotheses to be re-futed or falsified.

8 Post-processualism(Hodder 1991; Hodder &Shanks 1993)

Archaeology has come under many influencessince the 1970s, including structuralism, post-structuralism and forms of post-modernist criti-cism that, if taken to extreme lengths, reject notonly those values and judgements that havebeen made in the past, but even the possibilityof making any meaningful judgements at alltoday. It is hardly surprising that processualism,based on systems deduced from detailed obser-vations of sites and artefacts, came under force-ful attack from many direct ions. As hashappened in so many political and intellectualmovements in the past, yesterday’s revolution-aries have become today’s conservatives. How-ever, the ‘new’ archaeologists of the 1960s andlater did not maintain an unaltered set oftheory; the pursuit of general laws soon ap-peared unattainable, but many interesting ideaswere generated along the way. However, someof those who benefited from the advances intheory and methods now consider that theirachievements, and indeed those of their pred-ecessors, are under threat.

The lack of consensus is reflected by the pro-liferation of publications about archaeologicaltheory itself, rather than ‘real’ archaeology setinto a theoretical context, and also by the pas-sions generated by disputes about the validity ofdifferent theoretical approaches. New theoreticalapproaches continue to rely upon ‘mining-and-bridging’ exercises that borrow (mine) usefultheories from other disciplines, and modify themto provide explanations (bridges) in archaeologi-cal situations; ‘new’ archaeology had itself ben-efited from precisely this kind of activity in the1960s (Yoffee & Sherratt 1993, 3–4). One prob-lem is that post-processual archaeology presentsa rapidly moving target, far harder to criticizethan its processual predecessor. It thrives on con-tradictions, oppositions and confrontations withtraditional approaches:


173

Post-processual archaeology, then, involves thebreaking down of established, taken-for-granteddichotomies, and opens up study of the relation-ships between norm and individual, process andstructure, material and ideal, object and subject.Unlike processual archaeology it does not es-pouse one approach or argue that archaeologyshould develop an agreed methodology. That iswhy post-processual archaeology is simply ‘post-’. It develops from a criticism of that which wentbefore, building on yet diverging from that path.It involves diversity and lack of consensus. It ischaracterized by debate and uncertainty aboutfundamental issues that may have been rarelyquestioned before in archaeology. It is more anasking of questions than a provision of answers.(Hodder 1991, 181)

8.1 Context, structure, mind…(Renfrew 1982; Shanks 1991)

One major drawback of a strictly processual viewof the past is its failure to address some of the mostprofound aspects of human life related to thoughtand belief. Renfrew had confronted this issue in1982 by discussing ‘an archaeology of mind’, anddeploring the traditional division between mind andmatter:

The gap cannot be bridged, nor can ‘mind’ use-fully be considered when taken in isolation fromits ‘thoughts’. These ‘thoughts’, I assert, or someof them do find effective expression in the mate-rial record. (1982, 27)

He identified some of these expressions, such assystems of weights that indicate concepts of massand measurement, religious objects and struc-tures that imply ritual or ideology, deliberatedesigns of artefacts, and most forms of artisticdepiction.

This kind of thinking accords well with the con-textual archaeology developed by Ian Hodder as aresult of his ethnographic research in Kenya andelsewhere (Hodder 1982b; 1991). He generated amuch more sophisticated view of the role of mate-rial culture, in which artefacts and their ornamen-tation possessed a powerful role in societies, andwere used in an active manner to display, in sym-bolic form, feelings of allegiance, defiance, etc. At-

tention to symbols is also an important element instructuralism, an approach more commonly foundin anthropology and linguistics, where words, ideasor artefacts are linked into subtle patterns of mean-ing. A positive feature of philosophies of archaeol-ogy that seek meaning in psychological terms is thatthey are unlikely to neglect the active role of theindividual in society, whereas an excessivelyprocessual point of view may lead to a mechanicalnotion of human societies shaped entirely by envi-ronmental forces.

9 Reconstructingarchaeology…(Shanks & Tilley 1987; 1992)

Glyn Daniel stated that ‘man’s past is something tobe recorded, described, appreciated and understood’(1981, 192), as if there were no important problemsin gaining understanding, once ‘something’ hadbeen recorded and described properly. Readers ofthe editorials he wrote for Antiquity from the 1950sto the 1980s (published in book form in 1992) wereleft in no doubt that the past was to be appreciatedby connoisseurs in the same manner as a goodFrench wine. It should certainly be protected fromthe ‘lager-louts’ of theoretical archaeology, and keptout of the hands of those who placed themselvesbeyond the pale by showing an interest in treasurehunting or UFOs.

The title of this section is taken from a fasci-nating book, Re-constructing Archaeology:Theory and practice by Michael Shanks andChristopher Tilley, first published in 1987. Anupdated edition appeared in 1992 with an appen-dix that restated the authors’ central views withthe benefit of five years’ hindsight, and in the lightof criticisms that they had received. Re-construct-ing Archaeology was accompanied by a separatebook, Social Theory and Archaeology (1987).Despite their assertion that a simplified textbookversion of complex ideas ‘is merely a valorizationof anti-intellectualism’ (1992, 262), I will take thisrisk if only to encourage readers to look up thesebooks for themselves. Few authors of books onarchaeology rival Shanks & Tilley’s ability tostimulate and depress, fascinate and irritate, ex-cite and repel their readers.


174

Shanks & Tilley’s principal objection to the ‘new’archaeology was the separation between the devel-opment of theories and the case-studies carried outto test them:

We attempted to emphasise theory as a practice,as a process indelibly linking the archaeologist andthat which he or she constructs and re-constructs.Such a position puts an end to theory viewed assomething essentially divorced from and stand-ing beyond the practice of actually doing archae-ology. What we were working towards was atheory of and in practice, the notion that all ar-chaeology is theoretical practice. (248)

The rejection of ‘scientism’ may have struck at theheart of many archaeologists, but more shocks werein store. In answering their critics over accusationsof ‘…relativist pluralism, for apparently arguingthat validity depends simply on the present, forbeing nihilistic, negative and political’, Shanks &Tilley observed that the strongest objections camefrom the most imperialist capitalist nation-states(the USA and Canada) and Britain, ‘riddled withdeep-rooted class division’ (257). This underlinedthe political dimensions of their work:

Archaeology is a making of a past in a present….Taking the past seriously involves recognizing itsotherness not as a matter of exoticism but as ameans of undercutting and relativizing the legiti-macy of the present (260).

…We are seeking…to disrupt and to render di-shevelled prevailing contemporary archaeologi-cal discourses in order to foster fresh discoursesand new pasts, socially and politically relevantpasts. (262–3)

So where does all this lead?

We intend to develop further a critical sociologyof archaeology…Central questions to be askedhere are: who produces the past and why? Forwhom exactly is this production taking place? Inwhat circumstances? Who has the right to speakand expect to have their statements considered asworthy of attention and comment? (263)

Shanks & Tilley end on an upbeat note, expressing

a sentiment with which I agree, for it summarizesmy perception of the value of archaeology as anacademic discipline worthy of a place in education:‘We embrace a contradictory and fluid past which,even if not simple, will be intelligible’ (265). Oth-ers disagree, finding little comfort in Shanks &Tilley’s denial of relativism, and disliking a politi-cal agenda that threatens their own establishedposition:

Indeed, the post-processual school is no schoolat all…in that it does not attempt to formulate aconstructive archaeological agenda, launches nocoherent body of theory and method for inter-preting the past, and sets out deliberately toobfuscate the genuine gains made in over a cen-tury of systematic research. The ideological dan-ger posed by the grimmest processual scientismpales in comparison to the threat of those whoseek to undermine the framework of traditionalarchaeological practice and who, at their mostsystematically critical, are indeed nihilists.(Yoffee & Sherratt 1993, 8)

10 Current issues in archaeology

I hope that this extended discussion of the role oftheoretical archaeology has shown that its impor-tance is not restricted to understanding the past. Italso helps us to perceive our own points of viewmore clearly, and to guard against an unthinkingimposition of our own values and preoccupationson ancient societies. Although there is nothing newin recognizing that our opinions about the past tendto reflect our views on the present, it is not alwayseasy to bear it in mind.

One distinctive aspect of western society in thelate twentieth century is a reaction against modern-ism. It ranges from concern about environmentalproblems, such as global warming, that are seen asa result of industrialization, to the romantic appealof primitive lifestyles, whether those of modernpeoples such as the Yanamamo Indians or suppos-edly egalitarian and peaceful prehistoric societies.The confidence that allowed Victorian England torule and exploit colonies abroad while youngwomen and children worked in factories at home


175

has almost entirely disappeared. Prosperity andeducation have subsequently raised awareness of therights of individuals to a very high degree, and thesocial background of people who participate inarchaeology is now very diverse. As a result, sitesare studied and preserved by professional academ-ics and civil servants rather than the leisured church-men and indulgent landowners of the eighteenthcentury. Archaeology has also become involved inbroader issues, such as the market-led deregulatedeconomics of Thatcher and Reagan,multiculturalism, ‘green polities’, feminism, therights of indigenous peoples, and ethical questionsover the ownership of ‘cultural property’ rangingfrom classical Greek sculptures to relatively mod-ern ‘tribal’ artefacts from the Third World. Again,it is important to stress that a subject that involvesso many current issues has great educational poten-tial, and its exponents are increasingly unlikely tooperate in ‘ivory towers’ isolated from modern pres-sures.

11 Managing our heritage(Cleere 1989)

One context where it is particularly difficult to evadebroad ethical and political issues is sometimesknown as the heritage movement (or even the her-itage industry). This term embraces many differentapproaches to presenting the past to the public, andprovokes strong reactions and hot debate. A lessemotive description of the conservation and pres-entation of sites, landscapes and artefacts is culturalresource management (CRM). Heritage manage-ment has become an important source of employ-ment for archaeologists in the 1980s and 1990s, andit has absorbed a large number of specialists who,in previous decades, would have made their careersin universities or academic museums.

This expansion of employment opportunities (ina time of general economic recession) is intimatelybound up with the move towards a post-industrialsociety, with its emphasis on leisure and the servicesector. Many private industries have opened theirown museums, covering the industrial histories ofanything from pencils (Keswick, Cumbria) towhisky (Edinburgh) and even fishing-tackle(Alnwick, Northumberland). Furthermore, while

some ‘heritage’ projects have been financed by prop-erty developers whose activities tend to destroyarchaeological sites, a significant proportion hasbenefited from the injection of government fundsinto areas that have suffered most from the declineof heavy industries such as mining or ship-building.At the same time, governments committed (intheory) to efficiency, deregulation and lower pub-lic expenditure have had to establish an extensivebureaucracy at national and local levels to cope withnew legislation about ancient sites and with tighterguidelines to the authorities who grant planningpermission for developments affecting historicbuildings or archaeological sites.

11.1 Archaeology and the State

It is interesting to observe changes in the names ofstate bodies that have looked after ancient monu-ments over the years in Britain. When Pitt Riversinitiated the process of guardianship of scheduledsites in 1882, he acted (largely at his own expense)on behalf of His Majesty’s Office of Works. By the1950s this had been replaced by the rather moredemocratic-sounding Ministry of Public Buildingand Works, and the trend towards image-projectioncontinued in the 1970s with the invention of theDepartment of the Environment. Even more dra-matic changes took place during the Thatcher years,first with the creation of an impressively namedcommittee, the Historic Buildings and MonumentsCommission, and the removal of responsibility forancient monuments from the Department of theEnvironment in 1984 into the hands of EnglishHeritage. This organization was (theoretically) in-dependent of the government, but overall controlof the purse-strings remained in the hands of theDepartment of the Environment.

Management of sites by English Heritagebrought visible changes: visitors to ruined cas-tles were now greeted by enthusiastic custodi-ans, some of them young graduates, wearingsweat-shirts adorned with the logo of EH’s newcorporate image, in place of the dark uniformsand peaked caps of their dour predecessors.Most sites open to the public received bright newticket offices, shops containing an attractiverange of souvenirs and information, and, insome cases, cafés as well. The new atmosphereof customer orientation was even enhanced by


176

a period of sponsorship by a supermarket chain.Similar developments took place beyond Eng-land with the creation of Historic Scotland, andits Welsh counterpart, CADW. English Heritagenow offers a range of educational, cultural andhistorical events at major sites, provides sub-scribing members with a glossy magazine, andpublishes popular books about its sites and ar-chaeology in general in a joint venture with acommercial publisher. The result is that it is dif-ficult for a visitor to distinguish between a sitein the care of the State and one owned by anentirely private organization such as the Na-tional Trust.

English Heritage still retains responsibility forenforcing the legal protection of sites and historicbuildings, and it funds a large number of rescueexcavations and the publication of a backlog ofsites excavated over recent decades. However, intune with the vogue in the 1980s and 1990s forprivatization and competitive tendering, much ofthis work is carried out indirectly by teams of con-tract archaeologists and self-employed consult-ants. In the same spirit, English Heritage alsoattempts to place most of the financial burden forexcavation and publication on the shoulders ofdevelopers, who, as part of the process of apply-ing for planning permission, must commissionreports on the impact of their proposals upon theenvironment (including archaeology). The pro-duction of impact reports offers yet another op-portunity for independent consultants, who maythen bid for the contract to carry out whateverwork is required. Obviously, ethics intrude here,for competition drives down prices, and a devel-oper has a vested interest in obtaining an impactreport that minimizes the harmful effects of what-ever project is proposed. Fortunately, archaeol-ogy attracts publicity in the press, and fewrespectable developers wish to incur the cost ofdamaging their public image when a few thou-sand pounds devoted to an excavation mightenhance it.

Thus, in little over a century, State involvementin British archaeology has progressed from infor-mal supervision of a small number of sites,through cooperation with sympathetic landown-ers, to the provision of a major focus of leisureand tourism. Furthermore, visitors to sites arenow encouraged to feel that they are participat-

ing in their heritage, rather than being allowedthe privilege of access to Crown property. How-ever, we must not forget that Pitt Rivers sawmuseums as a means of discouraging populardesires for radical political change (above, p.167); what message lies behind the current styleof English Heritage’s presentation? I feel that theanswer lies in conservative Britain’s fears thatEuropean integration through the EEC threatensnational sovereignty; thus, the united kingdomsof England and Scotland, along with the Princi-pality of Wales, all have a heritage administra-tion bearing their own name. Is it cynical to seein the title English Heritage an exclusive label thatnot only counters pro-European sentiments, butdiscourages multiculturalism? Did the Secretaryof State for the Environment who chose this titlein 1984 wish to promote the heritage of peoplewho consider themselves English rather than Brit-ish, or all the inhabitants of a country calledEngland?

11.2 Museums: from art gallery to‘experience’(Hudson 1987; Lumley 1988)

The fears that I have voiced over the ‘ownership’of the past find much stronger expression in thecontext of museums. Unlike an ancient site or his-toric building, a museum is a modern creationwhose very form embodies the ideas of a designer.These ideas are then enhanced by the selection ofparticular items for display, and by the kind ofinformation offered in the form of labels andother documentation. Older museums are insepa-rable from art galleries in that the artefacts theydisplay are grouped by period, and presented asbeautiful objects to be contemplated with mini-mal documentation. A connoisseur of art is ex-pected to be well-informed about the significanceand history of a painting, and needs only to begiven that the name of the artist, and the placeand year it was painted. Likewise, archaeologi-cal museums assumed for many centuries that alla visitor needed to know were the origin, date andfunction of an artefact on display. In the twenti-eth century museums gradually began to makefewer assumptions about their visitors, and at-tempted to provide more information about ob-jects on view, normally by providing


177

commentaries in the form of pictures of contem-porary sites and extended text and labels. Under-standing was enhanced by placing objects intocontext with the help of models and reconstruc-tions, and might extend to establishing a displayincluding human figures as well as objects.

YorkThis trend has culminated in a few remarkable‘museums’ (this term is no longer the obvious de-scription) such as the Jorvik Viking Centre; in thewords of an advertisement:

Visit the Jorvik Viking Centre in York and a timecar will whisk you back 1000 years to a recon-struction of an actual street in Viking Age Jorvik.Complete with sights, sounds and smells, it’s anunforgettable experience.

The success of this approach may be measuredby the length of the queue of people who are pre-pared to wait several hours for this short, expen-sive experience. On the surface, there is littledifference between Jorvik and other tourist attrac-tions in the city. The Friargate Museum uses waxfigures to mix history (kings and queens, Thatcherand Major) with sensation (Dracula and unex-pected jets of water in the dark), while the YorkDungeon presents lurid scenes of life-size figuresundergoing horrific medieval tortures. TheFriargate Museum leads into a shop full of ex-ploding pens, plastic skeletons and other famil-iar joke-shop toys. In contrast, Jorvik ends witha conventional museum gallery full of artefacts,followed by a shop full of rather expensive buttasteful gifts, games and books, mostly with aclear historical relevance or educational purpose.Jorvik was set up in 1984 by the York Archaeo-logical Trust, a team of archaeologists that dealswith rescue archaeology in the city. The Trustformed a company to invest in this project in theexpectation of making a profit that could beploughed back into supporting its principal ar-chaeological activities, whose funding was alwaysunpredictable and liable to be cut back duringtimes of financial difficulties for English Herit-age or the local council (Addyman 1990).

Jorvik maintains a clear separation between re-constructed scenes and authentic artefacts displayedin glass cases, and it presents archaeological meth-

ods by means of ‘frozen’ scenes of the archaeologi-cal excavation upon which the reconstruction isbased, and the laboratory where the finds were clas-sified and conserved in 1980. It is interesting thatthese displays have already become historical inthemselves, for there have been technical advancesin both surveying methods and computerized cata-loguing of finds since they were prepared to pro-vide a ‘state of the art’ picture of archaeologicalmethods. A solution to this problem is to be foundin the ARC (Archaeological Resource Centre: fig.6.6), a separate ‘museum’ (it consciously avoids thatname), also created by the York ArchaeologicalTrust.

The ARC breaks two major conventions inmuseum display. First, there are no glass cases, andno permanent collections of specimens of any kind.Second, it is an integral part of the process of in-terpreting the past, for visitors handle and sortactual finds from excavations, and they see ‘real’archaeologists at work through the glass doors andwalls of offices in the same building. ARC invitesvisitors to learn by taking part, once they havewatched an introductory video about the processof excavation. They are encouraged to handle ar-tefacts or bones and to discuss them with ARCstaff; they may also try sewing leather shoes, orweaving on a replica of an early medieval loom.An assortment of computers allows visitors tocatalogue objects, to analyse the distributions offinds in relation to the plan of an excavated build-ing, or to use an interactive videodisk to study thedetails of photographs of a site. There is no narra-tive, no heavy-handed educational message, sim-ply an open invitation to participate; it appeals tochildren, adults, and even professional archaeolo-gists who manage to join in without standing ontheir dignity. Two clear messages are communi-cated: that archaeologists do not dig to discoverspectacular finds, and that the ordinary scraps ofpottery and bone they find really are interestingand informative.

The purpose of the ARC is not purely altruistic,of course. The citizens of York, and tourists fromelsewhere, pay taxes to support rescue archaeologyand finds processing, but they would not value theseactivities highly in a time of government spendingcuts. A pleasant outing to York is enhanced by avisit to the Jorvik Viking Centre, and the ARC dem-onstrates the kinds of archaeological methods that


178

were employed to interpret the remains that havebeen reconstructed in Jorvik’s Viking street scene.An understanding of the kind of work that archae-ologists actually do, gained by participation, is likelyto give visitors a more positive attitude towardsfunding it. Ironically, the ARC is located in a redun-dant church, whose owners are delighted that it hasfound a use. Does this suggest that the materialistsociety of the late twentieth century gains greatercomfort from handling the physical remains of thepast than from contemplating the spiritual prom-ise of life after death?

11.3 Controlling the present bymeans of the past?(Hewison 1987; McBryde 1985)

Post-processual archaeology has heightened ourawareness of the importance of symbolism in thepast, and the political implications of the interpre-tations we make today. The ‘heritage movement’has not escaped scrutiny, and museums have comein for particularly trenchant criticism. Some of themost severe opprobrium is directed towards mu-seums that present the recent industrial past, suchas the collection of buildings re-erected at theNorth of England Open Air Museum at Beamish,in County Durham. A publicity leaflet describedit as ‘…a working example of what life in theNorth of England was really like in the early1900s… Beamish leaves little to the imagination.The experience is authentic.’ This celebration ofcoal-mining and other heavy industries emphasizescheerful communities based around terracedhouses and busy local shops. It makes a poignantcontrast with the realities of life in contemporarynorth-eastern England, and evokes nostalgia anda sense of the loss. Yet it was the horrors of nine-teenth-century industrial towns that stimulatedmost of the reforms of government, health andsocial administration that came to be valued in thetwentieth century. Does a museum like Beamishsimply reinforce the politics of the New Right, byidealizing ‘Victorian values’, and glorifying aworld where trade unions and socialism had notyet challenged the comfortable class-values of land-owners and industrialists? It will be very interest-ing to observe how the Berlin Wall is exploited inthe twenty-first century to tell visitors about thedivision and reunification of Germany (Baker1993).

11.4 Stonehenge

Political issues of this kind are less obvious in rela-tion to sites or museums that belong to earlier peri-ods, but they always lurk beneath the surface.Beamish relates to the memories of local families(reinforced by the novels of Catherine Cookson),and Jorvik may be conceptualized as an analogy forthe shopping streets that surround it. What shouldbe made of older monuments such as Stonehenge?Why should we feel any interest in, or responsibil-

6.6 The Archaeological Resource Centre, York,introduces the work of archaeologists throughpractical activities, rather than static displays and glasscases. Visitors may handle and classify genuineartefacts and bones, try out ancient weaving andleather-working techniques, and make use ofcomputers that allow them to explore aspects ofexcavation and finds-processing in their own way. Theupper floor of this modified church building accom-modates archaeologists who may be observed atwork through glass walls by visitors. York Archaeologi-cal Trust


179

ity towards, this accidental survival from a distant,forgotten past?

A deserving winner of an archaeological ‘bookof the year’ award was Stonehenge Complete, byChristopher Chippindale (1983). It required an in-troduction of only ten pages of text and photo-graphs to describe Stonehenge, supplemented bya further eight at the end (‘Stonehenge: what isknown’, 264–72). Between these two sections are16 chapters that follow the site from its first sur-viving written record (AD 1130) up to the Druidicand ‘hippie’ festivals of the 1970s and 1980s. Ar-chaeological study began in the sixteenth century,accelerated in the seventeenth and eighteenth withAubrey and Stukeley (above, p. 20–1), and culmi-nated in serious excavation only after 1900.Stukeley dated Stonehenge to the pre-Roman pe-riod by means of various ingenious field observa-tions (Chippindale 1983, 77–82). This was refinedto the Bronze Age in the nineteenth century, butan age in years was only achieved in the twentiethcentury, first by cross-dating and finally by radio-carbon. Excavations between 1900 and the 1950srevealed that the site had a long sequence of phasesthat began, according to the current calibration ofradiocarbon years, around 3100 BC. Constructionof what we recognize as Stonehenge, with its mas-sive sarsen trilithons, began a thousand years later,and was modified several times over the next thou-sand years.

Thus, unlike Beamish, Jorvik or a medieval ca-thedral, everything that is currently known or be-lieved about Stonehenge is the cumulative resultof almost nine centuries of speculation, observa-tion and excavation. Since nothing about its build-ers or its original purpose links it directly to thepresent, its popular attraction stems more fromignorance than knowledge, and provides an inter-esting example of the problems of cultural resourcemanagement. Christopher Chippindale edited afurther book about the site in 1990, this time en-titled Who Owns Stonehenge? It contains chaptersby an extraordinarily diverse range of people: thereare three archaeologists, two English (an editor/museum curator and a university professor) andone Welsh (working in Australia); the director ofthe Centre for Earth Mysteries; and the SecularArch-Druid. Chippindale summarized the issues inan elegant sentence that contains reminders ofpost-processual concerns, and a hint of the rela-

tivism so despised by opponents of post-modern-ism in general:

Beneath its weathered old surface, a superficiallystraightforward site is just one item in a com-pound of powerful ingredients: archaeology, yes,and landscape history, but, over-powering thedelicacies of scholarship, a stronger and bubblingbrew of issues concerning intellectual freedom,rational and intuitive knowledge, preservation,presentation and access, the place and role ofreligious beliefs, the State and its dissidents, therights of dispossessed ethnic minorities, and eventhe concept of ownership. (1990, 9)

The preparation of Chippindale’s book coincidedwith a violent confrontation between police andpeople (sometimes described as ‘New Age travel-lers’) who had held a free festival near Stonehengefrom 1974 to 1984. The police sealed off all ap-proaches to the site over the mid-summer periodbecause English Heritage and the National Trusthad banned not only the festival, but the annualrituals performed by modern Druids within thestone circles (fig. 6.7). The 1980s closed withStonehenge in the summer ‘…festooned inbarbed-wire, surrounded by police, and patrolledby privately employed security guards. It haslooked like a concentration camp, the unaccept-able face of militarism in a democracy’ (ibid. 33).In 1993, English Heritage scrapped all existingplans for improving visitor access, parking andinformation, and called for a wide discussion ofthe possibilities. Stonehenge became public prop-erty in 1918, long after the number of visitors hadbegun to pose problems of conservation andmanagement. Archaeology may have achievedmany things in the twentieth century, but it willnot be surprising if the centenary of public man-agement at Stonehenge has to be celebrated invisitor facilities that fall a long way short of thedignity of the site.

Who Owns Stonehenge? does an important jobby bringing together many opinions in one book,and makes proposals for the future. Whether youbelieve them or not, the chapters by exponents ofearth mysteries and Druidism are just as serious asthose by the archaeologists, and make it clear thatsuch views will have to be taken into account if plansfor the management of the site are to succeed.


180

Meanwhile, we must reflect upon the irony thatStonehenge, of all sites, should have come to sym-bolize authoritarian repression and exclusive own-ership (in the names of English Heritage and theNational Trust) to those few individuals who valueit as a symbol of ‘otherness’ in cosmic or ritual termstoday.

12 Ethical issues

12.1 The antiquities trade (Greenfield 1989; Wilson 1989)

It does not take much imagination to realize thatthe raw materials of archaeology are finite. Destruc-tion of sites escalated rapidly from the eighteenthcentury onwards, as agriculture expanded to match

population growth, and by the end of the nineteenthcentury a serious conservation ethic had begun todevelop in several European countries. Neverthe-less, many archaeologists and museums regardedsites and artefacts from around the Mediterranean,the Near East and other parts of the world as yetanother resource to be exploited for the benefit oftheir own countries. This is no longer official policy,but it has left a powerful legacy, in the form of avast international market for works of art, antiqui-ties and ‘tribal’ material from the Third World.Public museums have a reasonably good record inrefusing to purchase items that lack proper docu-mentation about their origins and ownership, butprivate collectors are not always so scrupulous.

The result is that ancient sites and cemeteries allover the world are systematically plundered in thesearch for pots, jewellery, carvings or anything else that

6.7 In the 1980s and 1990s, Druids and New Agetravellers wanting to celebrate the summer solstice atStonehenge came into increasingly violent conflictwith police who were attempting to enforce measuresintroduced by English Heritage and the National Trustfor the protection of the ancient monument. Thisprint, which accompanied a newspaper feature aboutthe difficult issues involved, encapsulates the transfor-mation of an ancient site into a modern symbol ofstate power by echoing the form of the stones in

‘trilithons’ made from traffic cones and barriers thatexclude Druids from the focus of their rituals. Post-processual archaeology has done much to create awider awareness of the power of symbols, and tounderline ways in which the management of the pastis influenced by modern judgements of a political orphilosophical nature. David Bromley; Guardian, 15/6/1992


181

may be sold. This problem is most serious in the ThirdWorld, where antiquities form a valuable supplementto low incomes, along with other sought-after mate-rials such as drugs or ivory. The treasure-hunters ofAfrica and South America receive pitifully small re-wards in comparison with the high prices that antiq-uities command in London or New York. The problemis not restricted to the Third World: architectural sculp-tures, mosaics and wall paintings removed from Eu-ropean churches have turned up in salerooms andmuseums in the United States, and protected sites inBritain are regularly raided by illegal commercialmetal-detector operators.

It is difficult for archaeologists based in the formercolonial powers to denounce the antiquities tradewhen museums in their own countries are full of itemsfrom overseas possessions. The British Museum isperhaps the greatest example of a global collection,and many of its most famous items (not just the Elginmarbles from the Parthenon in Athens) were removedwith the permission of their owners in the past. Butwhy should modern Greece or Egypt recognize legalagreements made before they gained independencefrom Turkish rulers? And were ‘gifts’ and purchasesfrom Africa or India really made between equals,without political or military pressure? The relativismfor which post-processual archaeology is frequentlycriticized certainly helps to force archaeologists andmuseums to confront not only the politics of theiractivities, but the ethics of their approach.

12.2 Indigenous peoples(Layton 1989a; 1989b)

Prehistory is a Western concept according towhich those societies which have not devel-oped writing—or an equivalent system ofgraphic representation—have no history. Thisfits perfectly into the framework of evolution-ist thought typical of Western cultures…. Ar-chaeology has been up until now a means ofdomination and the colonial dispossession ofour identity. If it were to be taken back by theIndians themselves it could provide us withnew tools to understand our historical devel-opment, and so strengthen our present de-mands and our projects for the future.(Mamami Condori 1989, 51, 58)

This observation by a Bolivian Indian archaeolo-

gist emphasizes how worries over the ethics of col-lecting antiquities or ‘tribal’ art from the ThirdWorld are closely linked to growing sensitivitiesabout the rights of indigenous peoples. The generalissues were made more apparent during celebrationsof the bicentenary of the colonization of Australiain 1988, and in the 500th anniversary in 1992 ofthe ‘discovery’ of America by Columbus: in thewords of one poster, ‘Native Americans are notcelebrating’. Arguments about the ownership of‘cultural property’ in the form of antiquities andworks of art take on an interesting additional di-mension when they are focused on the question ofthe treatment of human remains. In both NorthAmerica and Australia, there have been demandsfor the reburial of remains excavated by archaeolo-gists—whatever their age. Furthermore, there havebeen requests to return material taken as specimensby nineteenth-century anthropologists and anato-mists.

Archaeologists sometimes justify the disturbanceof indigenous sites and burials on the grounds thatthey benefit native populations today by illustrat-ing their origins and early history. It is true that thiskind of evidence may support claims for the own-ership of land, but it assumes a very ‘western’ viewof the past, involving linear (rather than cyclical)time, and a notion of death as final: ‘…to mostAborigines this would be meaningless sophistry.Human bones are the remains of their ancestors, thelandscape itself the remains of ancestral beings andcreators’ (Haglund, quoted in Hubert 1989, 156).By the 1990s, agreements about the treatment ofburials had been reached between indigenous peo-ples and archaeologists in many parts of the world(e.g. Kucera 1991). The process has been stressfulfor many archaeologists, forced to confront theprejudices that lurked behind their ‘liberal’ self-image, but the recognition of the validity of otherviews of the world is likely to be beneficial in thelong term (Zimmerman 1989).

12.3 Gender(Gero & Conkey 1991; Claasen 1992)

A parallel issue to that of indigenous rights is gen-der, for it also confronts prejudices that may lie sodeep that they are unconscious (figs 6.8–9). Thefirst blunder that male archaeologists make is toassume that an interest in gender is equivalent to


182

feminism. After all, one approach to gender (orrace) is to promote equality, and to ignore genu-ine differences; an emphasis on ‘engendering’ ar-chaeology demands that it should be a seriousaspect of any inquiry. The feminist dimensionarises from the problem that while females do tendto see gender as a perspective, males frequentlyignore it:

Women archaeologists who study gender re-lations, either women in prehistory or womenin archaeology, expressly state their perspec-tive, which is in fact often considered to beonly a perspective by mainstream male archae-ologists, and they acknowledge that their viewis partial…. Feminist theorists admit theirposition, the context and perspective of theirresearch/writing…. In addition, post-processualists often show a total lack of an un-derstanding of themselves as genderedindividuals, as well as of gender’s part in thestructuring of individuals, culture and society.(Engelstad 1991, 512)

The absurdity of some male attitudes to archaeo-logical studies is illustrated by two interpretationsof early Stone Age art. Carved objects bearing tworounded protrusions were classified as ‘breast pen-dants’ for many years, but when rotated through90°, they could equally represent penises and tes-ticles, which were very frequently featured on pen-dants in the Roman period (Kehoe 1991). Thesecond example concerns ‘Venus’ figurines thatwere supposedly carved to exaggerate the femalebody in a religious symbolism of fertility around25,000 BC. While this ritual interpretation cannotbe tested, a study of the physical proportions ofmodern women shows that the figurines fit intothe normal range of obesity (Duhard 1991). Thebasis of their interpretation as ‘cult-figurines’ restsupon an idealized (predominantly male) image ofthe shape of women, rather than an observationof reality.

One reaction to the male dominance of themodern world is to propose that a different kindof society existed in the past. The concept of aGreat Goddess has been given progressivelygreater emphasis over recent decades by MarijeGimbutas, a major expert on European prehis-tory. She has welcomed the shift away from the

interpretation of megalithic stone monumentsthat dominated the 1960s and 1970s, when theywere seen as expressions of societies who pos-sessed sophisticated mathematical and astronomi-cal skills (very much associated with malepractitioners today). Instead, emphasis has beenplaced upon the purpose of their alignments withthe solar and lunar calendar: ‘An ideology basedon belief in an unending and returning cycle isdisclosed—birth, life, death, rebirth. The tombsand sanctuaries are permeated with the idea ofregeneration of life powers that depend on theCosmic Mother’ (Gimbutas, in Meaden 1991,10). The feminine attributes of the Goddess arecontrasted with the less attractive associations ofmachismo:

…She was Provider throughout the immenseperiod of time which was the Neolithic andBronze Ages, an era which for Britain, Ire-land and Brittany was largely a time of tran-quillity. She ruled over a classless, balancedsociety, until the convulsions of the Iron Agebrought widespread fortifications to hilltopsfollowing invasions by male-ruling, God-dominated warrior groups—the so-called‘heroic’ societies. Thus ended the serenity ofthe Age of the Goddess. So began the Age ofWars which has lasted to this day. (Meaden1991, 214)

Echoes of the ‘New Age’ philosophies of Califor-nia (where Gimbutas was based) are evident in thiskind of writing; utopian evocations of lost GoldenAges have a much longer history, stretching backto ancient Greece.

General problems of sexual inequality affectarchaeology in much the same way as other pro-fessions, for the same underlying reasons. Com-petition for careers in museums, universities orfield archaeology does not favour people who havestepped off the promotions ladder to raise fami-lies; this is where feminism must stimulate changesin attitudes. The transition from feminism and thearchaeology of women to an archaeology of gen-der remains a goal that ‘…will be partly realizedwhen gender is considered not as an optional is-sue, but as another structuring principle funda-mental to interpreting past societies’ (Gilchrist1991, 499).


183

12.4 The Green movement(Maclnnes & Wickham-Jones 1992)

Samples of tree-rings taken from oak trunks pre-served in the peat bogs of Ireland include a series ofexceptionally narrow rings that indicate an episodeof cold, wet weather beginning in 1159 BC andlasting around twenty years. It was almost certainlythe result of a volcanic eruption detected by a layerof acidity found in ice-sheet cores at 1100±50 BC.Minute particles of volcanic glass have been discov-ered in peat bogs from northern Britain sampled fortheir pollen content, and the chemical compositionof the volcanic minerals matches material from a

massive eruption of Hekla in Iceland. We knowfrom recent experience that volcanic eruptions sendenormous quantities of volcanic dust and ash intothe upper atmosphere. Mount St Helens in theUnited States and Mount Pinutabo in the Philippinesprovided vivid examples of this phenomenon inrecent years; they were modest in comparison withHekla, however.

The quantities of dust that Hekla poured intothe atmosphere were sufficient to shield the Earth’ssurface from solar radiation to such an extent thatthe climate was disrupted for the twenty-year pe-riod revealed by tree-rings. The effect upon hu-mans was even more dramatic, for it seems that

6.8–9 The 1951 Festival of Britain celebrated anoptimistic beginning to a new era, after a half-centurythat had included two world wars and the GreatDepression. It is interesting to see that while theSouth Bank Exhibition site in London looked forwardthrough ‘contemporary’ architecture, it also involved‘the people giving themselves a pat on the back’. It isinteresting to reassess a series of models showing theBritish family through the ages (now at the JewryWall Museum, Leicester, and still very popular withvisitors) from the perspective of the 1990s. Werethey intended to reassure people who feared that therecently reelected Labour government threatenedsocial upheavals? They imply that the nuclear familyunit had remained a constant feature of British societythroughout millennia of progress and change; thesame kinds of family images also featured heavily inadvertisements at the time. The Anglo-Saxon family isparticularly intriguing, for despite two recent wars, itseems to underline the Germanic origins of military

and cultural achievements in British history. Themother and daughter on the left also bear uncomfort-able similarities to propaganda images of blond Aryanfamilies produced in Hitler’s Germany. Both familieswear the full, billowing clothes introduced by Dior’s‘New Look’ in 1947, despite its condemnation forwastefulness by government ministers; the Iron Agewoman even echoes the facial expression andhairstyle of Dior’s models.

Most of the artefacts and costume accessoriesincluded in these groups are derived from a socialélite, not ordinary families. The figures also provokemixed reactions today because of the manner inwhich they present gender roles. Jewry Wall Museum,Leicester


184

settlement patterns and farming practices weredisrupted for long enough to cause the abandon-ment of most upland areas of northern Britain. Anarchaeologist, Colin Burgess, had already observeda rupture in the archaeological evidence for set-tlements and artefact types from this late BronzeAge period, and even went as far as suggesting thatit might have been the result of a natural catastro-phe such as a volcano. Baillie’s work on tree-ringsprovided independent confirmation of the propo-sition that an environmental catastrophe had beenresponsible for these profound changes in humanbehaviour.

There are clear lessons to be drawn from theeruption of 1159 BC. It is a timely reminder thatwe live in a fragile ecosystem, under constant threatfrom natural disasters that may strike from anyquarter. It also underlines the seriousness of currentconcern about global warming, for if a temporarydisturbance of the atmosphere by a single volcanomay put large tracts of land out of use for a genera-tion, what will the continuing build-up of green-house gases do over the next century?

Archaeology also demonstrates the effects ofhuman changes to the landscape. Irrigation inMesopotamia raised the salinity of soils too far forcrops to grow successfully; forest clearance in theApennines led to erosion that filled Mediterraneanvalleys with metres of silt, and turned the low-ly-ing areas of Rome into an unhealthy swamp.Mayan civilization may have collapsed as a resultof an excessive population’s demands on the lim-ited fertility of its land; Easter Island’s forest wasdestroyed by rival populations involved in com-petition to quarry, transport and erect huge stonestatues (Bahn & Flenley 1992). In other words,human societies in the past did not live in harmonywith their environments; they frequently exploitedand destroyed them. This knowledge does not give

us an excuse for doing the same thing; rather, itadds to our responsibility to explore non-destruc-tive agriculture, to utilize materials less wastefully,and to avoid non-productive competitive activi-ties—notably warfare.

13 Conclusion

The interpretation of the past from archaeologicalremains has come a long way since antiquarians firstmanaged to disprove Samuel Johnson’s depressingcontention that ‘We can know no more than whatold writers have told us’. The terminology and pre-occupations change, but the objectives and attitudesto the evidence (not to mention the disputes withopponents) have a familiar ring. One of the greatbenefits of a retrospective study of archaeology isto appreciate this point, and to take a more detachedview of any school of thought that claims to be‘new’, or to have exclusive possession of the truth.

An open-minded attitude is required not only tochanging fashions in archaeological interpretation,but to the subject as a whole. The lesson to be learntfrom the rapid advance of scientific techniques inrecent decades is that new and revolutionary evi-dence may appear at any moment from a completelyunsuspected source. My personal commitment tothe subject is greatly enhanced by this aspect: a dis-cipline that incorporates so much uncertainty andso many different academic approaches, while itignores conventional boundaries between the sci-ences and the humanities, is well worth studying atschool, university or as a leisure pursuit. If my bookhas conveyed any of this feeling to its readers, I willconsider it to have been a success.


185

Note: Many fundamental works are cited in headings withineach chapter; these should be consulted first. References topublications that only appear in this section are given in full;those that are described more briefly are included in thedetailed bibliography below.

General Works

Reference worksConcise accounts of many methods, sites and cultures maybe found in the Collins Dictionary of Archaeology (BahnP (ed), Glasgow; HarperCollins, 1992) or The PenguinDictionary of Archaeology (Bray W and Trump D (ed),1970); S Champion’s Dictionary of Terms and Techniquesin Archaeology (Oxford, Phaidon, 1980) is still worth pur-suing. Lavishly illustrated works that cover prehistoric andhistorical cultures and civilizations, as well as explainingarchaeological methods, are The Cambridge Encyclope-dia of Archaeology (Sherratt 1980) and Past Worlds: TheTimes Atlas of Archaeology (London, Times Books Ltd,1989).

Period outlinesBrian Fagan has written several readable overviews of pre-history, including World Prehistory: A brief introduction (2nded, HarperCollins, 1993) and The Journey from Eden (1990).The ‘feel’ of the search for the earliest humans is conveyedby Leakey & Lewin’s Origins Reconsidered (1992), whileMaisels covers The Emergence of Civilisation in the NearEast (1993). The role of archaeology in historical periods isclearly illustrated in M Grant’s The Visible Past: Greek andRoman history from archaeology (London, Weidenfeld andNicolson, 1990). Archaeology in Britain since 1945(Longworth H and Cherry J (ed), London, British Museum,1986) shows how much archaeology has enlarged the per-ception of the British past, while Fagan’s New Treasures ofthe Past (London, Quarto, 1988) takes a world-wide ap-proach.

Other introductionsJohn Gowlett’s Ascent to Civilization (2nd ed, New York,McGraw-Hill, 1993) manages to explain techniques andresults in an exemplary manner, supported by superb illus-

trations. Renfrew & Bahn’s Archaeology: Theories, meth-ods, and practice (1991) and Fagan’s In the Beginning (1993)are ‘heavy-weight’ textbooks that are more effective as ref-erence works than introductory reading. A lighter note isstruck by P Rahtz in Invitation to Archaeology (Oxford,Blackwell, 1991) and Paul Bahn’s Bluff your Way in Archae-ology (London, Ravette, 1989), while a book written by oneamateur for others is Amateur Archaeologist by S Wass (Lon-don, Batsford, 1992).

Periodicals and magazinesNews of new discoveries and techniques, discussions of theo-ries, book reviews and comments on political issues featurein the British quarterly periodical Antiquity. A more popu-lar approach is taken by Archaeology in America, Dossiersde l’archèologie et d’histoire in France, Archäologie der Weltin Germany, and Current Archaeology in Britain. The Na-tional Geographical Magazine includes archaeological arti-cles; a selection has been edited by C Lutyck (The Adventureof Archaeology, Washington, 1992)

Chapter 1:The idea of the pastGlyn Daniel’s A Short History of Archaeology (1981) is awell-illustrated introduction, although less detailed than 150Years of Archaeology (1975). The Idea of Prehistory wasreissued in 1988 (with additional chapters by Renfrew), andfurther insights may be gained from Antiquity and Man:Essays in honour of Glyn Daniel (Evans J D Cunliffe B andRenfrew C (ed), London, Thames & Hudson, 1982). A widerrange of issues is raised in Tracing Archaeology’s Past, ed-ited by A L Christensen (Southern Illinois Univ Press, 1989).

Avenues of investigationGrayson’s The Establishment of Human Antiquity (1983)remains the authoritative work on the conceptual break-through that took place in the nineteenth century, whileGould’s Time’s Arrow, Time’s Cycle (1987) explores theintellectual context of the discovery of geological time; seealso A Hallam’s Great Geological Controversies (2nd ed,Oxford Univ Press, 1989). Joan Evans’ Time and Chance(1943) explains how John Evans (her grandfather) becameinvolved in the momentous events of 1859.

Guide to Further Reading


186

Trigger’s A History of Archaeological Thought(1989) is a comprehensive guide, supplemented byMalina & Vasicek’s Archaeology Yesterday and To-day (1990). Stuart Piggott places British archaeologyinto the context of European thought in Ancient Brit-ons and the Antiquarian Imagination (1989), and thisbackground is covered by Sklen?ar in Archaeology inCentral Europe (1983).

Antiquarian f ieldwork in BritainStuart Piggott has written a full biography of WilliamStukeley (1985). Various editions of Richard Gough’s1789 translation of Camden’s Britannia may befound in reference libraries, while facsimiles of JohnAubrey’s Monumenta Britannica or A Miscellany ofBritish Antiquities (Parts 1 & 2) have been publishedwith editorial comments from Fowles J & Legg R(Boston, Little, Brown & Co, 1980–82). Avebury Re-considered: From the 1660s to the 1990s by Ucko P,Hunter M and Clark A (London, Unwin Hyman,1990) is a case-study of the continuing importanceof antiquarian records.

Touring and collectingThe northern European passion for Greek and Ro-man antiquities is illuminated in Haskell F and PennyN, Taste and the Antique: The lure of classical sculp-ture, 1500–1900 (Yale Univ Press, 1981), and Pen-ny’s study of an important collector, ThomasHoward, Earl of Arundel (Oxford, Ashmolean Mu-seum, 1985). The transformation of private collec-tions into public museums is explained in Museumsand the Shaping of Knowledge by Hooper-Greenhill(1992) and Impey & MacGregor’s The Origins ofMuseums (1985). One case is presented in detail inM Caygill’s The Story of the British Museum (Lon-don, 1992), while Archaeologists and Aesthetes: Thesculpture galleries of the British Museum in the 19thcentury by I Jenkins (London, British Museum, 1992)discusses a major disagreement about how thingsshould be displayed. Collecting was particularly im-portant to the development of the Three-Age System;the background is presented in O Klindt-Jensen’s AHistory of Scandinavian Archaeology (London,Thames and Hudson, 1975).

The discovery of civilizationsContemporary reports from the Illustrated Lon-don News recapture the excitement of nineteenth-century discover ies (Bacon E, The GreatArchaeologists, London, Seeker and Warburg,1976); Great Adventures in Archaeology, editedby R Silverberg (Penguin, 1985) is an anthology

of writings by discoverers. H Winstone’s generalaccount, Uncovering the Ancient World (London,Constable, 1986) is supplemented by his biogra-phies of Woolley of Ur (London, Secker and War-burg, 1990) and Howard Carter and the discoveryof the Tomb of Tutankhamun (London, Consta-ble, 1991). Other interesting biographies includeFlinders Petrie (Drower 1985) and The Find of aLifetime: Sir Arthur Evans and the discovery ofKnossos (S Horwitz, London, Weidenfeld &Nicolson, 1981). The interaction between discov-erers and collecting institutions is illuminated byJ L Fitton’s Heinrich Schliemann and the BritishMuseum (London, British Museum Occ Pap 83,1991).

Analytical approaches to discoveries includeMcDonald & Thomas’ Progress into the Past: Therediscovery of Mycenean civilization (1990) and R TRidley’s The Eagle and the Spade: Archaeology inRome during the Napoleonic era (Cambridge UnivPress, 1992). Discoveries outside Europe and the NearEast are featured in Fagan’s Elusive Treasure (1977)and Willey & Sabloff’s A History of American Ar-chaeology (1980). D K Chakrabarti has written AHistory of Indian Archaeology: From the beginningto 1947 (New Delhi, Munshiram Manoharlal, 1988),while A History of African Archaeology has beenedited by P Robertshaw (London, James Currey,1990).

Chapter 2:Discovery, f ieldwork and recording

The discovery and recording of sitesTraditional methods are described in Brown’s Field-work for Archaeologists and Local Historians (1987)and C Taylor’s Fieldwork in Medieval Archaeology(London, Batsford, 1974). For interpretation seeInterpreting Artefact Scatters: Contributions toploughzone Archaeology, edited by J Schofield (Ox-ford, Oxbow, 1990), or Shennan’s Experiments in theCollection and Analysis of Archaeological SurveyData (1985). Basic recording is covered in Surveyingfor Archaeologists by F Bettess (Univ of Durham,1992). The problem of maintaining systematic cata-logues of sites is the subject of County Archaeologi-cal Records: Progress and Potential, edited by IanBurrow (Assoc County Archaeol Officers, 1985), anda Danish volume, Sites and Monuments: Nationalarchaeological records, edited by Larsen (1992). Abrief case-study of an interesting fieldwork exercisethat rediscovered a port in northern France is David


187

Hill’s ‘Quentovic defined’ and ‘The definition of theearly medieval site of Quentovic’, Antiquity 64 (1990)51–8 and 66 (1992) 965–9.

Aerial photography and geophysical surveyingThe best introductions are Riley’s Air Photographyand Archaeology (1987) and Wilson’s Air Photo In-terpretation for Archaeologists (1982). Informativecase studies are included in The Impact of Aerial Re-connaissance on Archaeology (Maxwell G S (ed),London, Coun Brit Archaeol Res Rep 49, 1983), andInto the Sun, edited by D Kennedy (Sheffield, J Collis,1989). Important aspects of interpretation and re-cording are discussed in The Emerging Past: Air pho-tography and the buried landscape by Whimster(1989). Specific projects are described in Rome’sDesert Frontier from the Air by D Kennedy & D Riley(London, Batsford, 1990), The Archaeology of Dart-moor from the Air (Greeves T (ed), Exeter, DevonBooks, 1985) and R Palmer’s Danebury: An aerialphotographic interpretation of its environs (London,HMSO, 1984).

Geophysical surveying is explained clearly inClark’s Seeing Beneath the Soil (1990), and the detailedelectronic and mathematical background is presentedin Scollar’s Archaeological Prospecting and RemoteSensing (1990). Many case studies can be found inGeoprospection in the Archaeological Landscape,edited by Spoerry (1992). An excellent case-study ofsoil analysis in a fieldwork project has been publishedby ProschDanielson L and Simonsen A: ‘PCA of pol-len, charcoal & soil phosphate data as a tool in pre-historic land-use investigation at Forsandmoen, SWNorway’, Norwegian Archaeological Review 21.2(1988) 85–102.

Landscape archaeologyAston’s Interpreting the Landscape (1985) concen-trates on British fieldwork, and the results can be seenin interpretations such as C Taylor’s Village and Farm-stead: A history of rural settlement in England (Lon-don, Chapman and Hall, 1983). Oliver Rackham’s TheHistory of the Countryside (London, Dent, 1993) in-cludes fascinating studies of woodlands. The role ofenvironmental sciences is underlined by Alluvial Ar-chaeology in Britain, edited by Needham S andMacklin M (Oxford, Oxbow Monog 27, 1992), andby a ‘state of the art’ paper (‘An environmental his-tory of the upper Kennet valley, Wiltshire, for the last10,000 years’) by John Evans, in Proceedings of thePrehistoric Society 59 (1993) 139–95. Theoretical is-sues are debated in Space, Time, and ArchaeologicalLandscapes, edited by Rossignol J and Wandsnider L(New York, Plenum, 1992).

Regional f ield survey projectsSome British projects deserve study for their meth-ods as well as their results. Andrew Fleming’s TheDartmoor Reaves: Investigating prehistoric land di-visions (London, Batsford, 1988) examines uplands,while The Maddle Farm Project: An integrated sur-vey of prehistoric and Roman landscapes on the Berk-shire Downs by Gaffney V and Tingle M (Oxford,Brit Archaol Rep 200, 1989) includes intensively cul-tivated land. The context of an individual site isemphasized in The Stonehenge Environs Project(Richards 1990). The extent of work around theMediterranean and elsewhere is apparent in Archaeo-logical Field Survey: Britain and abroad edited byMacready S and Thompson F H (London, Soc ofAntiq, 1985), and Roman Landscapes (Barker &Lloyd 1991). Renfrew & Wagstaff’s An Island Pol-ity: The archaeology of exploitation in Melos (1982)contains valuable discussions of methodology focusedon an Aegean island.

Examples of a significant new development arecontained in Interpreting Space: GIS and archaeol-ogy (Allen 1990), while Gaffney V and Stancic Zhave made a study of one Croatian island, Hvar, inGIS Approaches to Regional Analysis (Ljubljana,Faculty of Arts, 1991). Many geographical andmathematical approaches are contained in The In-terpretation of Archaeological Spatial Patterning(Kroll E M and Price T D (ed), New York, Plenum,1991).

Chapters 3:Excavation

The development of excavation techniquesSeveral books by Glyn Daniel cited above (Chapter 1)chart the beginnings of methodical excavation. Theearly chapters of Harris’s Principles of Archaeologi-cal Stratigraphy (1989) explain how geological con-cepts were adopted by early excavators. Thedisappearance of evidence through careless clearanceand ‘excavation’ was documented by R Lanciani asearly as 1899 in The Destruction of Ancient Rome(London, Macmillan).

M W Thompson’s General Pitt-Rivers (1977) hasbeen followed by a further biography by Bowden(1991); unfortunately, the four massive volumes thatPitt Rivers distributed privately (Excavations inCranborne Chase, London, 1887–1898) are not easyto find in libraries. Mortimer Wheeler summarizedhis own techniques in Archaeology from the Earth,first published in 1954. His racy autobiography StillDigging (London, Michael Joseph, 1955) traces


188

much of his professional career, but JacquettaHawkes’ biography (1982) leaves many questionsunanswered.

Excavation procedureThe ‘classic’ work is Philip Barker’s Techniques of Ex-cavation (1993), but his earlier Understanding Ar-chaeological Excavation (1986) may be easier forbeginners. Harris’s Principles of ArchaeologicalStratigraphy (1989) remains the best account of thisdifficult aspect. Taphonomy is explored in Schiffer’sFormation Processes of the Archaeological Record(1987), and further relevant papers are contained inNatural Formation Processes and the ArchaeologicalRecord, edited by Nash D T & Petraglia M D (Ox-ford; Brit Archaeol Rep, 1987), and ArchaeologicalFormation Processes, edited by K Kristiansen (Copen-hagen, National Mus, 1988).

Popular syntheses of work on individual sites helpto explain the selection of a site, excavation strategy,and interpretation. The English Heritage/Batsford se-ries contains a growing number of these, includingWharram Percy: Deserted medieval village (BeresfordM and Hurst J 1990); two Iron Age hillforts, Danebury:The anatomy of an Iron Age hillfort (B W Cunliffe 1993)and Maiden Castle (Sharpies 1991); and Flag Fen: Pre-historic fenland centre (Pryor 1993). Two medieval sitesfeaturing an interesting combination of excavation,documentary history and architectural analysis areNorton Priory by J P Greene (1989) and A Temple forByzantium: The discovery and excavation of a palacechurch in Istanbul (R M Harrison, London, HarveyMiller, 1989). Dealing with developers and the publicin a busy urban setting are aspects of Richard Hall’s TheViking Dig: Excavations at York (London, Bodley Head,1984). Ian Hodder’s article ‘Writing archaeology: sitereports in context’ (Antiquity 63 (1989) 268–74) pro-vides an interesting perspective on the excavator’s crea-tive role, while technicalities of publishing a site areoutlined in the Council for British Archaeology’s Sign-posts for Archaeological Publication (London, 1991).For some long-term problems that might easily be over-looked see Preserving Field Records: Archival tech-niques for archaeologists and anthropologists, edited byKenworthy M A (Philadelphia, Univ PennsylvaniaMuseum, 1985). The continuing interaction of excava-tion and destruction by modern development is featuredin Rescue Archaeology: What’s next?, edited by MytumH and Waugh K (York Univ, Archaeol Monog 6, 1987);it follows on from Barri Jones’ Past Imperfect: The storyof rescue (London, Heinemann, 1984).

Excavation: special casesFor town sites see M Carver’s Underneath EnglishTowns: Interpreting urban archaeology (London,Batsford, 1987) and Arguments in Stone: Archaeo-logical research and the European town in the firstmillennium (Oxford, Oxbow Monog 29, 1993), alsoP Ottaway’s Archaeology in British Towns from theEmperor Claudius to the Black Death (London,Routledge, 1992). Waterlogged sites are the subjectof B & J Coles’s Peoples of the Wetlands: Bogs, bod-ies and lake dwellings: a world survey (London,Thames and Hudson, 1989), while The sea Remem-bers: Shipwrecks and archaeology by PThrockmorton (London, Weidenfeld and Nicolson,1987) illustrates many marine sites. All aspects ofmarine ‘fieldwork’, excavation and recording arecontained in the Nautical Archaeology Society’s Ar-chaeology Underwater, edited by Dean (1992). At theopposite extreme, A M Rosen’s Cities of Clay: Thegeoarchaeology of tells (Chicago Univ Press, 1986)concerns mud-brick sites. Rodwell’s Church Archae-ology (1989) discusses cemeteries as well as churchbuildings, and an interesting burial site is exploredin A Boddington’s Raunds: Furnells church and cem-etery (London, English Heritage, 1989). A volume ofstudies on Burial archaeology has been edited byRoberts (1989). Retrieval of Objects from Archaeo-logical Sites (Payton R (ed), Denbigh, Archetype,1992) is a collection of case-studies about the con-servation of structures and artefacts revealed by ex-cavation. Many interesting discussions of stone andtimber structures are included in Structural Recon-struction, edited by Drury (1982).

Chapter 4:Dating the past

Dating methods are also included in the historicalsurveys and general books cited at the beginning of thissection and in relation to chapter 1 (above, p. 185).

Historical datingSouth’s Method and Theory in Historical Archaeol-ogy (1977) is a complex but rewarding work, andtheoretical dimensions can also be appreciated in Ar-chaeology as long-term History, edited by Ian Hodder(Cambridge Univ Press, 1987). M Baillie’s A SliceThrough Time: Dendrochronology and precision dat-ing (London, Batsford 1995) shows how scientificmethods are reducing the difference between prehis-tory and history. A controversial discussion of histori-cal dating in Mediterranean, Egyptian andNear-eastern archaeology is Centuries of Darkness,edited by P James (London, Jonathan Cape, 1991); an


189

orthodox view is set out in Warren P M and HankeyV, The Absolute Chronology of the Aegean Bronze Age(Bristol, Classical Press, 1989). Greek and Roman ar-chaeology features in W R Biers’ Art, Artefacts andChronology in Classical Archaeology (London,Routledge, 1992). Pèrin’s La datation des tombesmèrovingiennes (Geneva, Librairie Droz, 1980) usesburials, typology, and historical texts to chart earlymedieval settlement in France. My own book Interpret-ing the Past: Roman pottery (London, British Museum,1992) includes a chapter on dating methods, while ABurnett’s Interpreting the Past: Coins (London, Brit-ish Museum, 1991) outlines the potential of numis-matics.

TypologyGraslund’s The Birth of Prehistoric Chronology(1987) is the fundamental work on nineteenth-cen-tury developments, and Montelius’s Dating in theBronze Age with Special Reference to Scandinavia(Stockholm, Royal Academy of Letters, History andAntiquities, 1885) is available in a translation by HClarke. Pitt Rivers’ contribution is discussed in bi-ographies by Bowden (1991) and Thompson (1977).Modern theoretical discussions include Archaeologi-cal Typology by L S Klejn (Oxford, Brit ArchaeolRep S153, 1982), Archaeological Typology andPractical Reality by Adams W and Adams E (Cam-bridge Univ Press, 1991), and Essays on Archaeo-logical Typology edited by Whallon R and Brown JA (Evanston, Kampsville Seminars in Archaeology1, 1982).

Scientif ic dating techniquesMartin Aitken’s Science-based Dating in Archaeol-ogy (1990) is essential reading. Its structure preventscomplexities of the background science from gettingin the way of general principles, and it contains com-prehensive scientific and archaeological references.Clear accounts of scientific dating may also be foundin Gowlett’s Ascent to Civilization (1993) and in Sci-ence and the Past, edited by Bowman (1991). Usefulinformation can also be gained from Current Scien-tific Techniques in Archaeology, edited by Parkes(1986), New Developments in Archaeological Sci-ence, edited by Pollard (1992), and Dating and AgeDetermination of Biological Materials, edited byZimmerman M R and Angel J L (Beckenham, CroomHelm, 1986).

Dating and interpretation are inseparable in thestudy of early human evolution. Johanson’s Lucy’sChild (1991) or Leakey & Lewin’s Origins Recon-sidered (1992) introduce issues that are explored

further in The Origin of Modern Humans and theImpact of Chronometric Dating, edited by Aitken M,Stringer M and Mellars P (Princeton Univ Press,1993). For the conflict between traditional and sci-entific dating over Akrotiri, a Bronze Age town de-stroyed by a volcano, see Thera and the Aegean WorldIII, edited by Hardy and Renfrew (1990). Spindler’sThe Man in the Ice (1994) is another example of therole of scientific dating, while its use in testing au-thenticity is included in Spencer’s Piltdown: A scien-tific forgery (1990).

Tree-ring datingM G Baillie’s Tree-ring Dating and Archaeology (Lon-don, Routledge, 1981) remains an excellent account;see also D Eckstein’s Dendrochronological Dating(Strasbourg, European Science Foudation, 1984), Treerings by Schweingruber (1987) and Applications ofTree-ring Studies: Current research indendrochronology and related subjects, edited by RWard (Oxford, Brit Archaeol Rep S33, 1987). All stud-ies of radiocarbon dating (see below) also discussdendrochronology in relation to calibration. Varves,pollen analysis, sea-bed deposits and ice-sheet coresfeature in books on environmental archaeology andclimatology—and of course in Aitken’s Science-basedDating.

Radiocarbon datingRadiocarbon dating by Bowman (1990) is an ideal in-troduction. Advances in high-precision measurementare discussed in Archaeological Aspects of Accelera-tor Dating, edited by Gowlett J and Hedges R (Ox-ford, OUCA Monog 11, 1986); see also RadiocarbonDating: Recent applications and future potential, ed-ited by J J Lowe (Quaternary Research Assoc, 1991).The technique’s dramatic impact on European prehis-tory is explained in Renfrew’s Before Civilization(1973). Taylor’s Radiocarbon Dating: An archaeologi-cal perspective (1987) contains a comprehensive as-sessment, including its history, and Taylor has alsoedited Radiocarbon after Four Decades (with Long &Kra, 1992).

Archaeology, Dendrochronology and the Radiocar-bon Calibration Curve, edited by B Ottaway (Edin-burgh Univ Press, 1984), contains papers of enduringinterest, as does the special radiocarbon section inAntiquity 61 (1987), which includes Pearson’s impor-tant paper ‘How to cope with calibration’. Finally, fulldetails of calibration methods, including a computerprogram, can be found in a Radiocarbon 35.1, editedby Stuiver, Long & Kra (1993).


190

Other scientif ic dating methodsThere are monographs on Thermoluminescence Dat-ing by G A Wagner (Strasbourg, European ScienceFoundation, 1983) and Aitken (1985), while Tarling’sPalaeomagnetism (1983) remains the fundamentalwork on magnetic dating. Reports on refinements indating techniques appear in the periodicalsArchaeometry or Journal of Archaeological Science,and less technical accounts are likely to be included inAntiquity, or magazines such as Nature or ScientificAmerican.

Chapter 5:Science and archaeology

Note: scientific aspects of archaeological fieldwork arecited in relation to chapter 2 above.

General worksImportant articles appear in the journals Archaeometryand Journal of Archaeological Science, and many de-velopments are reported in Antiquity. A historicalperspective can be gained from M S Tite’s article

‘Archaeological science—past achievements and fu-ture prospects’ in Archaeometry 32.3 (1991) 139–51.Archaeological science has become very specialized,and it is scarcely possible for any individual author tocover it. The most readable general survey is Scienceand the Past, edited by Bowman (1991); Parkes’ Cur-rent Scientific Techniques in Archaeology (1986) ismuch more detailed. An outstanding exhibition cata-logue is Les mystères de l’archéologie: les sciences à larecherche du passé (Presses Universitaires, Lyon, 1990).Many books contain collections of papers (frequentlygenerated by conferences) that tend to be overtakenby newer work, but of lasting interest are New Devel-opments in Archaeological Science, edited by Pollard(1992) and Scientific Analysis in Archaeology, editedby Henderson (1989). Volumes with a more specificarea of interest are Jones & Catling’s Science in archae-ology (1986), mainly relevant to Greece, and Sciencein Egyptology by David (1987). For the overall con-text of scientific work see Research Priorities in Ar-chaeological Science, edited by P Mellars (London,Council Brit Archaeol, 1987).

The examination of objects and raw materialsScanning Electron Microscopy in Archaeology, editedby Olsen (1988), demonstrates the wide applicationof microscopic examination. The study of artefacts hasgenerated many publications, including L Hurcombe’s

Use Wear Analysis: Theory, experiments and results(Sheffield Archaeol Monog 4, 1992), R Grace’s Inter-preting the Function of Stone Tools (Oxford, BritArchaol Rep S474, 1989) and The Interpretative Pos-sibilities of Microwear Studies, edited by B Graslund(Uppsala, Societas Archaeologica, 1990).

Studies of stone include The Scientific Study of Flint,edited by Sieveking G and Hart M (Cambridge UnivPress, 1986), Marble: Art historical and scientific per-spectives on ancient sculpture (Malibu, Getty Museum,1990) and Clough’s Stone Axe Studies (1988). Thebroader context of analytical results is examined in RTorrence’s Production and Exchange of Stone Tools:Prehistoric obsidian in the Aegean (Cambridge UnivPress, 1986) and Bradley & Edmonds’ Interpreting theAxe Trade (1993). For pottery see Recent Develop-ments in Ceramic Petrology, edited by Middleton Aand Freestone I (London, British Museum Occ Pap 81,1991) and Greek and Cypriot Pottery: A review ofscientific studies by R E Jones (British School at Ath-ens, 1985).

A range of spectrographic techniques is presentedin Neutron Activation and Plasma EmissionSpectrometric Analysis in Archaeology, edited byHughes (1991). Analysis and characterization of met-als feature in Aspects of Early Metallurgy, edited byOddy (1991) and Bronze Age Trade in the Mediterra-nean, edited by Gale (1991). These studies are alsoimportant for ancient coinage—see Scientific Studiesin Numismatics, edited by Oddy (1980) and Metal-lurgy in Numismatics, edited by Archibald & Cowell(London, Spink/Royal Numismatic Soc, 1993).

Conservation is an important area of archaeologi-cal science, and recent books include Cronyn’s Ele-ments of Archaeological Conservation (1989) andSease’s A Conservation Manual for the Field Archae-ologist (Los Angeles, UCLA Inst Archaeol, 1992).Several case-studies are described in The Art of theConservator, edited by Oddy (1992). The role of analy-sis in detecting forgeries should not be overlooked; seeFake? The art of deception, edited by Jones (1990).

The environmentThis subject has become so large and diverse that in-troductory books are scarce. Shackley’s Environmen-tal Archaeology (1982) and Using EnvironmentalArchaeology (London, Batsford, 1985) remain useful,as does John Evans’ An Introduction to Environmen-tal Archaeology (1978). Environmental archaeologyis integrated with a wider geographical picture inChanging the Face of the Earth by Simmons (1989)


191

and Late Quaternary Environmental change by Bell& Walker (1992). The many general collections ofpapers include Issues in Environmental Archaeology,edited by N Balaam and J Rackham (London, Insti-tute of Archaeolgy, 1992); Conceptual Issues in Envi-ronmental Archaeology, edited by J Bintliff (EdinburghUniv Press, 1988); Palaeoenvironmental Investiga-tions, edited by N Fieller (Oxford, Brit Archaeol RepS258 and S266, 1985); and Recent Developments inEnvironmental analysis, edited by E Webb (Oxford,Brit Archaeol Rep S416, 1988). Theoretical perspec-tives are included in Modelling Ecological Change, ed-ited by D Harris and K Thomas (London, Institute ofArchaeology, 1991).

Informative studies about Britain include Environ-mental Archaeology: A regional review, edited by HKeeley (London, English Heritage, 1984 and 1987)and Martin Jones’ England before Domesday (Lon-don, Batsford, 1986). Work that complements field-work projects abroad (chapter 2) can be found inMan’s Role in the Shaping of the Eastern Mediterra-nean Landscape, edited by S Bottema (Rotterdam,Balkema, 1990) or P Horden and N Purcell, The Medi-terranean World: Man and environment in antiquityand the middle ages (Oxford, Basil Blackwell, 1991).The role of climatology may be examined in ClimaticChange in Later Prehistory, edited by A F Harding(Edinburgh Univ Press, 1982) or Climate and History,edited by Wigley (1981).

Soil scienceSoils add an important dimension to environmentalarchaeology; three books of collected papers on thistopic are Man-made Soils, edited by W Groeneman-van Waateringe & M Robinson (Oxford, Brit ArchaeolRep S410, 1988); Soils and Micromorphology in Ar-chaeology, edited by Courty (1989); and Soils in Ar-chaeology, edited by Holliday (1992).

Plant remainsG W Dimbleby’s Plants and Archaeology (London, JBaker, 1978) remains a useful introduction. Detailedbooks include Pearsall’s Paleoethnobotany (1989), JGreig’s Archaeobotany (Strasbourg, European SciFoundation, 1989) or R Brookes’ Phytoarchaeology(Leicester Univ Press, 1991). Useful collections of stud-ies include Current Paleoethnobotany, edited by CHastorf & V Popper (Chicago, Prehist Archaeol &Ecology Series, 1988), Phytolith Systematics: Emerg-ing issues, edited by G Rapp & S Mulholland (NewYork, Plenum, 1992) and Progress in Old WorldPalaeoethnobotany, edited by W van Zeist (Rotterdam,Balkema, 1991).

Detailed British studies are contained in Archaeologyand the Flora of the British Isles, edited by M Jones(1988), while M Van der Veen’s Crop Husbandry Re-gimes (1992) includes very interesting methods and in-terpretations. Wider questions of human exploitation ofplants are addressed in Foraging and Farming, edited byD Harris and G Hillman (London, Unwin Hyman, OneWorld Archaeology 13, 1989) and New Light on EarlyFarming: Recent developments in palaeoethnobotany,edited by Jane Renfrew (Edinburgh Univ Press, 1991);see also D Zohary & M Hopf’s Domestication of Plantsin the Old World: The origin and spread of cultivatedplants in west Asia, Europe and the Nile valley (Oxford,Clarendon, 1993). Two interesting case-studies are Cornand Culture in the Prehistoric New World, edited by SJohannessen & C Hastorf (Oxford, Westview Press,1993), and Pharaoh’s Flowers, a study of plants fromTutankhamun’s tomb by F Hepper (London, Royal Bo-tanic Gardens Kew/HMSO, 1990).

Pollen analysis is explained thoroughly inDimbleby’s The Palynology of Archaeological Sites(1985) and in a more general text, Pollen Analysis byMoore, Webb & Collinson (1991).

Animal remainsThe Archaeology of Animals by Davis (1987) is very read-able, and is complemented by Rackham’s Interpretingthe Past: Animal bones (1994). Technical aspects arecovered in Ageing and Sexing Animal Bones from Ar-chaeological Sites, edited by B Wilson (Oxford, BritArchaeol Rep 109, 1982), R Lyman’s VertebrateTaphonomy (Cambridge Univ Press, 1994) and SHillson’s Teeth (Cambridge Univ Press, 1986). Clutton-Brock has written A natural History of DomesticatedAnimals (1988), and edited The Walking Larder: Pat-terns of domestication, pastoralism and predation (1988).

L Binford challenged traditional interpretations ofanimal remains in Bones: Ancient man and modernmyths (New York, Academic Press, 1981). Other ex-amples of interpretation can be found in excavationreports; an early hunter-gatherer site is Starr CarrRevisited: A re-analysis of the large mammals by ALegge & P Rowley-Conwy (Univ of London, Dept ofExtra-Mural Studies, 1988), while a historical exam-ple from York is Bones from Anglo-Scandinavian Lev-els at 16–22 Copper gate by T O’Connor (London,CBA, Archaeol of York 15.3, 1989). The contrastbetween rural and urban assemblages may be pursuedfurther in Groeneman-van Waateringe &Wijngaarden-Bakker’s Farm Life in a Carolingian Vil-lage (1987) and Diet and Craft in Towns:


192

The evidence of animal remains from the Roman tothe post-medieval periods, edited by D Searjeantson& T Waldron (Oxford, Brit Archaol Rep 199, 1989).

Birds, f ish, shells and insectsCohen A and Searjentson D have produced A Manualfor the Identification of Bird Bones from Archaeo-logical sites (Univ of London, Dept of Extra-MuralStudies, 1987). Wheeler & Jones have written anoverview of Fishes (1989), and many further studiesare included Fish and Archaeology, edited byBrinkhuizen & Clason (1986). Studies of marinemolluscs feature in Deciphering a Shell Midden, ed-ited by Stein (1992) and W F Buchanan’s Shellfish inPrehistoric Diet: Elands Bay, S W Cape coast, S Af-rica (Oxford, Brit Archaol Rep S455, 1988). LandSnails in Archaeology by J G Evans (1973) remainsan important source. In addition to H Kenward’spioneering The Analysis of Archaeological InsectAssemblages (London, CBA, The Archaeology ofYork 19.1, 1978), see Buckland & Coope’s A Bibli-ography and Literature Review of Quaternary En-tomology (1991). Insect remains are integrated intoa full review of site finds at York in R Hall & HKenward’s Environmental Evidence from theColonia: Tanner Row and Rougier Street (London,CBA, Archaeology of York 14.6, 1990).

Human remainsAn excellent introduction is A Chamberlain’s Inter-preting the Past: Human remains (London, BritishMuseum, 1994), followed up with studies of indi-vidual finds such as Spindler’s The Man in the Ice(1994) or Brothwell’s The Bog Man and the Archae-ology of People (London, British Museum, 1986)and Lindow Man, edited by Stead (1986). An Egyp-tian example is The Mummy’s Tale: The scientificand medical investigation of Natsef-Amun, editedby A R David & E Tapp (London, O’Mara Books,1992), and an unusual group of medieval Inuit (Es-kimos), accompanied by clothing, is beautifully il-lustrated in The Greenland Mummies, edited byHansen (1991). Forensic science and the pathologyof diseases and injuries are important; seeBoddington’s Death, Decay and reconstruction(1987), K Manchester’s The Archaeology of Disease(Univ of Bradford, 1983), Ortner & Puschan’s Iden-tification of Pathological Conditions in HumanSkeletal Remains (1985) or Health in Past Societies,edited by H Bush & M Zvelebil (Oxford, BritArchaol Rep S567, 1991). Particularly interestingdeductions from site finds are to be found in TWhite’s Prehistoric Cannibalism at Mancos5MTUMR-234–6 (Princeton Univ Press, 1992) and

J Hedges’ Tomb of the Eagles: A window on StoneAge tribal Britain (1984), which analyses a commu-nity found in a Scottish megalithic burial.

Bones in particular are the subject of D Ubelaker’sHuman Skeletal Remains (Washington, Taraxacum,1984), T White’s Human Osteology (London, Aca-demic Press, 1990) and Histology of Ancient HumanBone, edited by G Grupe & A Garland (Berlin,Springer Verlag, 1992). Analyses are presented in TheChemistry of Prehistoric Human Bone, edited byPrice (1989) and Prehistoric Human Bone: Archae-ology at the molecular level, edited by J Lambert &G Grupe (Berlin, Springer Verlag, 1993). The inves-tigation of evolution through genetics is examined inThe Human Revolution: Behavioural and biologicalperspectives on the origins of modern humans, ed-ited by P Mellars & C Stringer (Edinburgh Univ Press,1989) and Fagan’s The Journey from Eden (1990).Technical aspects can be found in Ancient DNA:Recovery and analysis of genetic material, edited byB Herrmann & S Hummel (Berlin, Springer Verlag,1993).

Statistics and computingGood introductions are Orton’s Mathematics in Ar-chaeology (Cambridge Univ Press, 1982) and Fletcher& Lock’s Digging Numbers (1991). Shennan’s Quan-tifying Archaeology (1988) and Quantitative Researchin Archaeology: Progress and prospects, edited by MAldenderfer (London, Sage, 1987), are more detailed,while full complexities are displayed in To Pattern thePast: Proceedings of the symposium on mathematicalmethods in archaeology, Amsterdam, 1984, edited byB Voorrips & S Loving (PACT 11, Strasbourg, Coun-cil of Europe, 1985). Mathematical approaches to ar-tefacts and site distributions were very popular duringthe 1970s: see Hodder & Orton’s Spatial Analysis inArchaeology (1976) and parts of the influential Ana-lytical Archaeology by David Clarke (revised byChapman, 1978). A particularly interesting combina-tion of statistics and typology is Richards’ The Signifi-cance of Form and Decoration of Anglo-SaxonCremation Urns (1987), which shows what subtletiesmay be revealed by these methods.

Books on computing become obsolete very quickly;samples of current work are found in conference pro-ceedings such as Computing the Past: CAA 92, editedby Andresen J, Madsen T and Scollar I (Aarhus UnivPress, 1993) or Computer Applications and Quanti-tative Methods in Archaeology, 1991, edited by G Lock& J Moffett (Oxford, Brit Archaol Rep S577, 1992).Rather more general are Computing for Archaeolo-


193

gists, edited by Ross (1991) and Archaeology and theInformation Age, edited by Reilly & Rahtz (1992).

Experimental archaeologyJohn Coles’s book Experimental Archaeology (1979)still provides a thorough introduction to the potentialand pitfalls of the experimental approach, which con-tinues to flourish, and may involve anything fromShennan’s Experiments in the Collection and Analy-sis of Archaeological Survey Data (1985) to the recon-struction of a Greek warship (Welsh, Building theTrireme, 1988). The latter project has been compre-hensively documented: see An Athenian Trireme Re-constructed: The British sea trials of Olympias, 1987,edited by J Morrison & J Coates (Oxford, Brit ArchaolRep S486, 1989). Another important area is illustratedby C Sussman’s A Microscopic Analysis of Use-wearand Polish Formation on Experimental Quartz Tools(Oxford, Brit Archaol Rep S395, 1988). Nothing avail-able in English approaches the superbly illustratedExperimentelle Archäologie in Deutschland, edited byM Fansa (1990).

Chapter 6:Making sense of the past

Archaeological theoryIt will only become apparent in future decades whichcurrent themes are of enduring significance. The in-fluence of evolution and anthropology began in thenineteenth century, and their impact can be studied insome of the books cited in Chapter 1, but the histori-cal overviews by Piggott and Daniel do not do justiceto the rapid changes of the later twentieth century.Fortunately there is now a wide choice of books: Trig-ger’s A history of Archaeological Thought (1989) cov-ers the entire history of the subject, while Malina &Vas?íc?ek’s Archaeology Yesterday and Today (1990)is almost as comprehensive. Theory also receives ex-tensive treatment in two textbooks: Renfrew & Bahn’sArchaeology: Theories, methods, and practice (1991)and Fagan’s In the Beginning (1993). Willey &Sabloff’s A History of American Archaeology (1980)is also useful in outlining the background of modernAmerican theory.

Modern developments are followed in Ian Hodder’sReading the Past (1991) and a collection of regionalstudies that he edited: Archaeological Theory in Eu-rope: The last three decades (London, Routledge,1991); Gibbon’s Explanation in Archaeology (1989)is also useful. The development of Binford’s thinkingcan be followed in two collections of his papers, InPursuit of the Past: Decoding the archaeological record

(London, Thames and Hudson, 1983) and DebatingArchaeology (1989).

Post-processualismThe route from processualism throughethnoarchaeology to post-processualism is outlinedin Hodder’s Reading the Past (1991), and typicalpapers are contained in The Meaning of Things:Material culture and symbolic expression, edited byI Hodder (London, Unwin Hyman, One World Ar-chaeology 6, 1988). Other elements are representedin Critical Traditions in Contemporary Archaeology,edited by V Pinsky & A Wylie (Cambridge Univ Press,1989), Archaeology After Structuralism, edited by IBapty & I Yates (London, Routledge, 1991) and TheAncient Mind: Elements of cognitive archaeology,edited by C Renfrew & A Zubrow (Cambridge UnivPress, 1994). Further insights may be gained fromInterpretive Archaeologies, edited by Hodder &Shanks (1993) or C Tilley’s Reading Material Cul-ture: Structuralism, hermeneutics and post-structur-alism (Oxford, Basil Blackwell, 1990). A collectionof backlashes is Archaeological Theory: Who sets theagenda?, edited by Yoffee & Sherratt (1993), whileHodder’s Theory and Practice in Archaeology (1992)is an attempt to reconcile abstract thoughts and prac-tical action.

One work by a historian that has provoked in-terest amongst archaeologists is The Sources of So-cial Power 1: A history of power from the beginningto AD 1760, by M Mann (Cambridge Univ Press,1986). A European school of historical thought hasinfluenced many writers on later prehistory and his-torical archaeology: see The Annales School andArchaeology, edited by J Bintliff (Leicester UnivPress, 1991) or Archaeology, Annales, andEthnohistory, edited by A Knapp (Cambridge UnivPress, 1992).

Heritage, archaeology, the public and the StateD Lowenthal’s The Past is a Foreign Country (Cam-bridge Univ Press, 1985) and Hewison’s The HeritageIndustry (1987) set a pessimistic critical tone for pub-lic presentations of the past. However, this has notdeterred practical approaches such as G Binks’ Visi-tors Welcome (London, HMSO/English Heritage1988) or the Institute of Field Archaeology’s Archaeo-logical Resource Management in the UK, edited by JHunter & I Ralston (Stroud, Alan Sutton, 1993).Among dozens of books that examine museums areHudson’s Museums of Influence (1987) and Hooper-Greenhill’s Museums and the Shaping of Knowledge(1992), while their future is discussed in Museums


194

2000: Politics, people, professionals and profit, editedby P Boylan (1993). Ethical aspects are the concern ofGreenfield’s The Return of Cultural Treasures (1989)and The Elgin Marbles: Should they be returned toGreece? by Hitchens C, Browning R and Binns G (Lon-don, Chatto and Windus, 1987).

Political issuesShanks & Tilley demand that archaeologists shoulddevelop a politically aware attitude in Re-construct-ing Archaeology (1992), and many individuals in-volved in ethnoarchaeology have become supportersof the rights of indigenous peoples. Two books editedby Layton contain papers on these themes: Who needsthe past? and Conflict in the Archaeology of Living

Traditions (both 1989). They resulted from sessionsat the World Archaeological Congress of 1986, whichwas riven by political disputes about the participationof South African archaeologists; see Ucko’s Academicfreedom and Apartheid: The story of the World Ar-chaeological Congress (Gloucester, Duckworth, 19).Another contemporary concern, gender, is explored inH Moore’s Feminism and Anthropology (Cambridge,Polity Press, 1987), Engendering Archaeology, editedby Gero & Conkey (1991) and Exploring Genderthrough Archaeology, edited by Claasen (1992). Abiography of one pioneering woman who made a ca-reer in archaeology is Born to Rebel: The life of HarrietBoyd Hawes by M Allsebrook (Oxford, Oxbow,1992).

195

Adam J-P 1994, Roman Building:Materials and Techniques, Lon-don, Batsford

Addyman P V 1990, ‘Reconstruc-tion as interpretation: the exam-ple of the Jorvik Viking Centre,York’, in Gathercole P &Lowenthal D, The Politics of thePast. London, Unwin HymanOne World Archaeology 12257–64

Addyman P V 1992, York Archaeo-logical Trust: 21 years of archae-ology in York, reprinted fromYorks Philos Soc Ann Rep

Aitchison T Ottaway B & Al-Ruzaiza AS 1991, ‘Summariz-ing a group of 14C dates on thehistorical time scale: with aworked example from the LateNeolithic of Bavaria’ , Antiq-uity 65 108–16

Aitken M J 1985,Thermoluminescence Dating,London, Academic Press

Aitken M J 1990, Science-basedDating in Archaeology, London,Longman

Alcock L 1972, ‘By South Cadburyis that Camelot’: the excavationof Cadbury Castle 1966–1970,London, Thames and Hudson

Allaby R Jones M & Brown T1994, ‘DNA in charred wheatgrains from the Iron Age hillfortat Danebury, England’, Antiq-uity 68 126–32

Allason-Jones L & Bishop M C1988, Excavations at RomanCorbridge: The hoard, HBMCLondon, English HeritageArchaeol Rep 7

Allen K et al (ed) 1990, Interpret-ing Space: GIS and archaeology,

London/New York Taylor andFrancis

Amick D S & Mauldin R P (eds)1989, Experiments in LithicTechnology, Oxford, BritArchaeol Rep S528

Anderson A 1991, ‘The chronologyof colonization in New Zea-land’, Antiquity 65 767–95

Andrew C 1992, The RosettaStone, London, British Museum

Anthony D W & Brown D R 1991,‘The origins of horseback rid-ing’, Antiquity 65 22–38

Arensberg C 1968, The Irish Coun-tryman: An anthropologicalstudy, New York, Peter Smith

Aries P 1985, Images of Man andDeath, Harvard Univ Press

Aston M 1985, Interpreting theLandscape, London, Batsford

Aston M & Rowley T 1974, Land-scape Archaeology: An intro-duction to fieldwork techniqueson post-Roman landscapes,Newton Abbot, David andCharles

Bacon E 1976, The Great Archae-ologists, London, Seeker andWarburg

Bahn P & Flenley J 1992, EasterIsland, Earth Island, London,Thames and Hudson

Bailey R N Cambridge E & BriggsH D 1988, Dowsing andChurch Archaeology,Wimborne Intercept

Baillie M G 1989, ‘Do Irish bogoaks date the Shang dynasty?’,Current Archaeol 117 310–13

Baillie M G 1992,‘Dendrochronology and envi-ronmental change’, in Pollard AM, New Developments in Ar-chaeological Science, Oxford,Clarendon 5–23

Baker F 1993, ‘The Berlin Wall:production, conservation andconsumption of a 20th-centurymonument’, Antiquity 67709–33

Barfield L 1994, ‘The iceman re-viewed’, Antiquity 68 10–26

Barker G & Lloyd J (ed) 1991,Roman Landscapes: Archaeo-logical survey in the Mediterra-nean region, British School atRome Monog 2

Barker P A 1986, UnderstandingArchaeological Excavation,London, Batsford

Barker P A 1993, Techniques ofArchaeological Excavation,London, Batsford, 3rd edition

Basalla G 1988, The Evolution ofTechnology, Cambridge UnivPress

Becker B 1993, ‘An 11,000-yearGerman oak and pine dendrochronology for radiocarboncalibration’, Radiocarbon 35.1201–13

Bell M & Walker M 1992, LateQuaternary EnvironmentalChange: Physical and humanperspectives, London, Longman

Berghaus P (ed) 1983, DerArchäologe: GraphischeBildnisse aus dem PorträtarchivDiepenbroick, Münster,Westfälischen Mus für Kunstand Kulturgesch

Bernal M 1987, Black Athena: TheAfroasiatic roots of Classicalcivilization 1: The fabrication ofancient Greece 1785–1985,London, Free Association Press

Bernal M 1991, Black Athena: TheAfroasiatic roots of classicalcivilisation 2: The archaeologi-cal and documentary evidence ,London, Free Association Press

BibliographyNote: items cited in the Further Reading section above are not repeated here.

Bibliography

196

Bersu G 1940, ‘Excavations at Lit-tle Woodbury, Wiltshire’, ProcPrehist Soc 6 30–111

Bethell P & Maté I 1989, ‘The useof soil-phosphate analysis inarchaeology: a critique’, inHenderson J, Scientific Analysisin Archaeology, Oxford, Ox-ford Univ Committee ArchaeolMonog 19, 1–29

Bidwell P et al (ed) 1988, PortaeCum Turribus, Oxford, BritArchaeol Rep S206

Binford L R 1972, An Archaeologi-cal Perspective, New York Semi-nar Press

Binford L R 1989, Debating Ar-chaeology, London, AcademicPress

Binford L R 1992, ‘Seeing thepresent and interpreting thepast—and keeping thingsstraight’, in Rossignol J &Wandsnider L, Space, Time, andArchaeological Landscapes,New York, Plenum 43–59

Bintliff J L (ed) 1984, EuropeanSocial Evolution: Archaeologi-cal Perspectives, University ofBradford Press

Bintliff J L et al 1990, ‘Trace metalaccumulation in soils on andaround ancient settlements inGreece’, in Bottema S et al,Man’s Role in the Shaping of theEastern Mediterranean Land-scape, Rotterdam, Balkema159–72

Birley E 1959, ‘Excavations atCorstopitum, 1906–58’,Archaeologia Aeliana 37 1–32

Birley E & Richmond I 1938, ‘Ex-cavations at Corbridge, 1936–1938’, Archaeologia Aeliana 15243–94

Bishop M C & Dore J N 1989,Corbridge: Excavations of theRoman fort and town, 1947–80, HBMC London, EnglishHeritage Archaeol Rep 8

Black J (ed) 1987, Recent Advancesin the Conservation and Analy-

sis of Artefacts, London, Inst ofArchaeology

Blatherwick S & Gurr A 1992,‘Shakespeare’s factory: archaeo-logical evaluations on the site ofthe Globe Theatre, Southwark’,Antiquity 66 315–88

Blundell S 1986, The Origins ofCivilization in Greek and Ro-man Thought, London,Routledge

Boddington A et al 1987, Death,Decay and Reconstruction: Ap-proaches to archaeology and fo-rensic science, Manchester UnivPress

Bogucki P 1993, ‘Animal tractionand household economies inNeolithic Europe’, Antiquity 67492–503

Bonde N & Christiansen A E1993, ‘Dendrochronologicaldating of the Viking Age shipburials at Oseberg, Gokstadand Tune, Norway’, Antiquity67 575–83

Booth W Ipson S & Haigh B 1992,‘An inexpensive PC-basedimaging system for applicationsin archaeology’, in Lock G andMoffett J, Computer Applica-tions and Quantitative Methodsin Archaeology 1991, Oxford,Brit Archaeol Rep S577 197–204

Boucher de Perthes J de C 1847–57,Antiquités celtiques etantediluviennes, 2 vols, Paris

Boujot C & Cassen S 1993, ‘A pat-tern of evolution for theNeolithic funerary structures ofthe west of France’, Antiquity67 477–92

Bowden M 1991, Pitt Rivers: Thelife and archaeological work ofLt-Gen. Augustus Henry LaneFox Pitt Rivers, CambridgeUniv Press

Bowler P J 1989, The Invention ofProgress: The Victorians and thepast, Oxford, Basil Blackwell

Bowman S (ed) 1991, Scienceand the Past, London, BritishMuseum

Bowman S 1990, RadiocarbonDating, London, British Mu-seum

Bradley B & Small C 1985, ‘Look-ing for circular structures in posthole distributions: quantitativeanalysis of two settlements frombronze age England’, J ArchaeolSci 12 285–98

Bradley R 1993, Altering the Earth:The origins of monuments inBritain and continental Europe,Edinburgh, Soc of Antiqs ofScotland Monog 8

Bradley R & Edmonds M 1993,Interpreting the Axe Trade: Pro-duction and exchange inneolithic Britain, CambridgeUniversity Press

Brinkhuizen D C & Clason A T (ed)1986, Fish and Archaeology:Studies in osteometry,taphonomy, seasonality andfishing methods, Oxford, BritArchaeol Rep S294

Brown A E 1987, Fieldwork forArchaeologists and Local Histo-rians, London, Batsford

Brown T A & K A 1992, ‘AncientDNA and the archaeologist’,Antiquity 66 10–23

Brunskill R W 1985, Timber Build-ing in Britain, London, Gollancz

Buckland P & Coope G R 1991, ABibliography and LiteratureReview of Quaternary Ento-mology, Sheffield, J R Collis

Cann J R et al 1969, ‘Obsidiananalysis and the obsidian trade’,in Brothwell D and Higgs E, Sci-ence in Archaeology, London,Thames and Hudson, 78–91

Carver M (ed) 1992, The Age ofSutton Hoo: The seventh cen-tury in north-western Europe,Woodbridge, Boydell andBrewer

Casson S 1939, The Discovery ofMan, London, Hamish Hamilton

Cattaneo C et al 1991, ‘Identifica-tion of ancient blood and tis-sue—ELISA and DNA analysis’,Antiquity 65 878–81

Caygill M 1992, The Story of

Bibliography

197

theBritish Museum, London,British Museum

Chapman J C & Shiel R 1991, ‘Set-tlement, soils and societies inDalmatia’, in Barker G & LloydJ, Roman Landscapes: Archaeo-logical survey in the Mediterra-nean region, Brit School RomeMonog 2 62–75

Chapman J C & Shiel R 1993, ‘So-cial change and land use in pre-historic Dalmatia’, Proc PrehistSoc 59 61–104

Chapman J C et al 1987, ‘Settle-ment patterns and land use inneothermal Dalmatia, Yugosla-via: 1983–1984 seasons’, J FieldArchaeol 14 123–146

Childe V G 1942, What Happenedin History, Harmondsworth,Penguin

Chippindale C 1983, StonehengeComplete, London, Thames &Hudson

Chippindale C 1988, ‘The inven-tion of words for the idea ofprehistory’, Proc Prehist Soc 54303–14

Chippindale C 1989, ‘“Social ar-chaeology” in the nineteenthcentury: is it right to look formodern ideas in old places?’, inChristensen A L, Tracing Ar-chaeology’s Past: Thehistoriography of archaeology.Southern Illinois Univ Press 21–33

Chippindale C et al, 1990, WhoOwns Stonehenge?, London,Batsford

Claasen C (ed) 1992, ExploringGender Through Archaeology:Selected papers from the 1991Boone Conference, Madison,Wisconsin, Monogs in WorldArchaeology 11, PrehistoryPress

Clark A J 1990, Seeing Beneath theSoil, London, Batsford

Clark J G D 1977, World Prehis-tory in New Perspective, Cam-bridge Univ Press

Clarke D L (revised Chapman R)

1978, Analytical Archaeology,London, 2nd ed. Methuen

Cleere H (ed) 1989, Archaeologi-cal Heritage Management in theModern World, London, UnwinHyman, One World Archaeol-ogy 9

Clough T H McK 1988, Stone AxeStudies 2, London, Council BritArchaeol Res Rep 67

Clutton-Brock J (ed) 1988, TheWalking Larder: Patterns of do-mestication, pastoralism andpredation, London, UnwinHyman, One World Archaeol-ogy 2

Clutton-Brock J 1988, A NaturalHistory of Domesticated Ani-mals, Cambridge Univ Press

Coates J F et al 1990, The TriremeTrials 1988, Oxford, OxbowBooks

Coe M 1992, Breaking the MayaCode, London, Thames andHudson

Coles B (ed) 1992, The WetlandRevolution in Prehistory,WARP, Prehistoric Society

Coles B & Coles J M 1986, SweetTrack to Glastonbury: TheSomerset Levels in prehistory,London, Thames and Hudson

Coles J M 1979, Experimental Ar-chaeology , London, AcademicPress

Coles J M 1984, The Archaeologyof Wetlands, Edinburgh UnivPress

Coles J M & Lawson A J (ed) 1987,European Wetlands in Prehis-tory, Oxford, Clarendon

Collingwood R G & Wright R P1965, The Roman Inscriptionsof Britain, Oxford, Clarendon

Courty M-A et al (ed) 1989, Soils andMicromorphology in Archaeol-ogy, Cambridge Univ Press

Cowell M & La Niece S 1991,‘Metalwork: artifice and art-istry’, in Bowman S, Science andthe Past, London, British Mu-seum 74–98

Craddock P 1991, ‘Mining andsmelting in antiquity’, in Bow-

man S, Science and the Past,London, British Museum 57–73

Craster H 1914, A History ofNorthumberland, 10: The par-ish of Corbridge, Newcastleupon Tyne, A Reid

Crawford O G S 1929, Air-photog-raphy for Archaeologists, Lon-don Ordnance Survey

Cronyn J M 1989, Elements ofArchaeological Conservation,London, Longman

Cunliffe B W (ed) 1982, The Pub-lication of Archaeological Exca-vations, London, Council BritArchaeol

Cunliffe B W 1971, Excavations atFishbourne 1961–1969, 2 vols,Rep Res Comm Soc Antiq, Lon-don 26

Daniel G 1967, The Origins andGrowth of Archaeology,Harmondsworth, Penguin

Daniel G 1975, 150 Years of Ar-chaeology, London, Duckworth

Daniel G 1981, A Short History ofArchaeology, London, Thamesand Hudson

Daniel G 1992, Writing for Antiq-uity, London, Thames and Hud-son

Daniel G & Renfrew C 1988, TheIdea of Prehistory, EdinburghUniv Press

David A R 1987, Science inEgyptology, Manchester UnivPress

Davidson D A & Shackley M L(eds) 1976, Geoarchaeology,London, Duckworth

Davis S J M 1987, The Archaeol-ogy of Animals, London,Batsford

Dean M et al (ed) 1992, Archaeol-ogy Underwater: The NASguide to principles and practice,London, Nautical Archaeol Soc/Archetype

Dearden B & Clark A 1990, ‘Pont-de-l’Arche or Pitres? A locationand archaeomagnetic dating forCharles the Bald’s fortificationson the Seine’, Antiquity 64567–619

Bibliography

198

Dimbleby G W 1985, ThePalynology of ArchaeologicalSites, New York/London, AcadPress

Dixon P 1988, Recording TimberBuildings, London, Council BritArchaeol, Practical Handbooksin Archaeology 5

Dorn R I et al 1988, ‘Cation-ratiodating of rock engravings fromthe Olary Province of arid SouthAustralia’ , Antiquity 62 681–689

Drower M S 1985, Flinders Petrie:A life in Archaeology, London,Gollancz

Drury P J (ed) 1982, StructuralReconstruction: Approaches tothe interpretation of the exca-vated remains of buildings, Ox-ford, Brit Archaeol Rep 110

Duhard J-P 1991, ‘The shape ofPleistocene women’, Antiquity65 552–61

Dunnell R C 1992, ‘The notionsite’, in Rossignol J andWandsnider L, Space, Time, andArchaeological Landscapes,New York, Plenum 21–41

Edis J MacLeod D & Bewley R1989, ‘An archaeologist’s guideto classification of cropmarksand soilmarks’, Antiquity 63112–146

Edwards N 1990, The Archaeologyof Early Medieval Ireland, Lon-don, Batsford

Engelstad E 1991, ‘Images ofpower and contradiction: femi-nist theory and post-processualarchaeology’ , Antiquity 65502–14

English Heritage 1989, The Management of ArchaeologicalProjects, London

Evans A 1921, The Palace of Minosat Knossos, I: The neolithic andearly middle Minoan ages, Lon-don, Macmillan

Evans A C 1986, The Sutton HooShip Burial, London, BritishMuseum

Evans A J 1899–1900, ‘Knossos I:the palace’, Ann Brit SchoolAthens 63–70

Evans C 1989, ‘Archaeology andmodern times: Bersu’sWoodbury 1938 and 1939’,Antiquity 63 436–50

Evans J 1860, ‘On the occurrenceof flint implements in undis-turbed beds of gravel, sand andclay’, Archaeologia 38 280–307

Evans J 1943, Time and Chance:The story of Arthur Evans andhis fore-bears, London,Longmans Green

Evans J 1956, A History of theSociety of Antiquaries , Oxford

Evans J D 1982, ‘Introduction: onthe prehistory of archaeology’,in Evans J D, Cunliffe B & Ren-frew C, Antiquity and Man:Essays in honour of Glyn Dan-iel, London, Thames and Hud-son 12–18

Evans J G 1973, Land Snails inArchaeology, London, SeminarPress

Evans J G 1978, An Introductionto Environmental Archaeology,London, Elek

Evershed P et al 1992, ‘The survivalof food residues: new methodsof analysis, interpretation andapplication’, in Pollard A M,New Developments in Archaeo-logical Science, Oxford,Clarendon 187–208

Exploring…1991, Exploring OurPast: Strategies for the archae-ology of England, London, Eng-lish Heritage

Fagan B 1977, Elusive Treasure:The story of early archaeologistsin the Americas, New York,Scribner

Fagan B 1990, The Journey fromEden, London, Thames andHudson

Fagan B 1993, In the Beginning: Anintroduction to archaeology, 7thed, HarperCollins

Fairclough G 1992, ‘Meaningfulconstructions: spatial and func-

tional analysis of medievalbuildings’. Antiquity 66 348–66

Fairweather A & Ralston I 1993,‘The Neolithic timber hall atBalbridie, Grampian Region,Scotland: the building, the date,the plant macrofossils’, Antiq-uity 67 313–23

Feyler G 1987, ‘Contribution àl’histoire des origines de la pho-tographic archéologique: 1839–1880’, Mel Ecole FrançaiseRome 99.2 1019–47

Fieller N R J et al (ed) 1985,Palaeo-environmental Investi-gations: Research design, meth-ods and data analysis, 2 vols:Oxford, Brit Archaeol RepS258, S266

Flannery K 1976, The Early Meso-american Village, New York,Academic Press

Fletcher M & Lock G 1991, Dig-ging Numbers: Elementary sta-tistics for archaeologists,Oxford, Oxford Univ Commit-tee Archaeol Monog 33

Fox C 1959, The Personality ofBritain, 4th ed, Cardiff, NatMus of Wales

Frere J 1800, ‘Account of flintweapons discovered at Hoxnein Suffolk’, Archaeologia 13204–5

Friedman J & Rowland M J (ed)1982, The Evolution of So-c ia l Systems, London,Duckworth

Gale N H (ed) 1991, Bronze AgeTrade in the Mediterranean,Studies in Medit Archaeol 90,Paul Astroms Forlag

Gale N H & Stos-Gale Z A 1992,‘Lead isotope studies in theAegean (the British Academyproject)’, in Pollard A M, NewDevelopments in Archaeologi-cal Science, Oxford, Clarendon63–108

Gamble C 1986, The PalaeolithicSettlement of Europe, Cam-bridge University Press

Gathercole P & Lowenthal D (eds)1990, The Politics of the Past,

Bibliography

199

London, Unwin HymanOneWorld Archaeology 12

Gero J M & Conkey M W (eds)1991, Engendering Archaeol-ogy: Women and prehistory,Oxford, Basil Blackwell

Gibbon G 1989, Explanation inArchaeology, Oxford, Blackwell

Gilchrist R 1991, ‘Women’s archae-ology? Political feminism, gen-der theory and historicalrevision’, Antiquity 65 495–501

Gillam J P 1957, ‘Types of Romancoarse pottery vessels in north-ern Britain’, Archaeol Aeliana35 180–251

Gingell C 1982, ‘Excavation of anIron Age enclosure atGroundwell Farm, Blunsdon StAndrew, 1976–7’, WiltsArchaeol Mag 76 33–75

Glover E Glover I & Vita-Finzi C1990, ‘First-order 14C dating ofmarine molluscs in archaeol-ogy’, Antiquity 64 562–6

Gough R (trans and ed) 1789,Camden W, Britannia, London

Gould R A 1990, Recovering thePast, Albuquerque, Univ ofNew Mexico Press

Gould S J 1987, Time’s Arrow,Time’s Cycle; Myth and meta-phor in the discovery of geologi-cal time, Harmondsworth,Penguin

Gowlett J 1984, Ascent to Civiliza-tion: The archaeology of earlyman, New York, McGraw-Hill

Graslund B 1987, The Birth of Pre-historic Chronology: Datingmethods and dating systems inn i n e t e e n t h - c e n t u r yScandinavian archaeology,Cambridge Univ Press

Grayson D K 1983, The Establish-ment of Human Antiquity, Lon-don Academic Press

Grayson D K 1984, QuantitativeZooarchaeology: Topics in theanalysis of archaeological fau-nas, London, Acad Press

Green J 1990, Maritime Archaeol-ogy: A technical handbook,London, Academic Press

Greene J P 1989, Norton Priory:The archaeology of a medievalreligious house, CambridgeUniv Press

Greene J P & Greene K 1970, ‘Atrial excavationon a Romano-British site at Clanacombe,Thurlestone’, Proc DevonArchaeol Soc 28 130–6

Greene K 1978, ‘Apperley DeneRoman fortlet: a re-examination1974–5’, Archaeol Aeliana 629–59

Greene K 1986, The Archaeologyof the Roman Economy, Lon-don, Batsford

Greene K 1987, ‘Gothic materialculture’, in Hodder I, Archaeol-ogy as Long Term History,Cambridge Univ Press, 117–31

Greenfield J 1989, The Return ofCultural Treasures, CambridgeUniv Press

Greeves T 1989, ‘Archaeology andthe Green Movement: A case forperestroika’, Antiquity 63 659–65

Griffith F M 1990, ‘Aerial recon-naissance in mainland Britain inthe summer of 1989’, Antiquity64 14–33

Grimm J L 1970, ArchaeologicalInvestigation of Fort Ligonier1960–1965, Pittsburgh, Annalsof Carnegie Museum 42

Groeneman-van Waateringe W& Wijngaarden-Bakker L1987, Farm Life in aCarolingian Village: A modelbased on botanical and zoo-logical data from an exca-vated site , Assen, VanGorcum/Univ Amsterdam

Guilday J E 1970, ‘Animal remainsfrom archaeological excavationsat Fort Ligonier’, Ann CarnegieMus 42 177–86

Hallam A 1989, Great GeologicalControversies, 2nd ed, OxfordUniv Press

Hamerow H 1993, Excavations atMucking 2: The Anglo-Saxonsettlement, London, EnglishHeritage Archaeol Rep 21

Hansen J P H et al (ed) 1991, TheGreenland Mummies, London,British Museum

Hanson W S & Rahtz P A 1988,‘Video recording on excava-tions’, Antiquity 62 106–11

Hanson W S et al 1979, ‘TheAgricolan supply base at RedHouse, Corbridge’, ArchaeolAeliana 71–97

Harris D & Thomas K (eds) 1991,Modelling Ecological Change,London, Institute of Archaeol-ogy

Harris E C 1977, ‘Units of archaeo-logical stratification’, NorwegArchaeol Rev 10 84–94

Harris E C 1989, Principles ofArchaeological Stratigraphy,2nd ed, London, AcademicPress

Harris E C Brown M R & BrownG J (eds) 1993, Practices of Ar-chaeological Stratigraphy, Lon-don, Academic Press

Haselgrove C et al 1990, ‘Stanwick,North Yorkshire, part 3: exca-vations on earthwork sites1981–86’, Archaeol J 147 37–90

Hatt G 1957, ‘Nørre Fjand, anearly Iron Age village in westJutland’, Arkaeol Kunsthist SkrDan Vid Selsk 2.2

Hawkes J 1982, MortimerWheeler: Adventurer in archae-ology, London, Weidenfeld &Nicolson

Hedeager L 1992, Iron-age Socie-ties: From tribe to state in north-ern Europe, 500 BC to AD 700,Oxford, Blackwell

Hedges J W 1983, Isbister: A cham-bered tomb in Orkney, Oxford,Brit Archaeol Rep 115

Hedges R E M & Sykes B C 1992,‘Biomolecular archaeology:past, present and future’, in Pol-lard A M, New Developments inArchaeological Science, Oxford,Clarendon 267–83

Henderson J (ed) 1989, ScientificAnalysis in Archaeology, Ox-

Bibliography

200

ford, Oxford Univ CommitteeArchaeol

Hewison R 1987, The HeritageIndustry: Britain in a climate ofdecline, London, Methuen

Hillam J et al 1990, ‘Dendro chro-nology of the English Neolithic’,Antiquity 64 210–20

Hind J G F 1980, ‘The Romano-British name for Corbridge’,Britannia 11 165–71

Hodder I 1982, The Present Past:An introduction to anthropol-ogy for archaeologists, London,Batsford

Hodder I 1982, Symbols in Action:Ethnoarchaeological studies ofmaterial culture , CambridgeUniv Press

Hodder I 1991, Reading the Past:Current approaches to interpre-tation in archaeology, 2nd edi-tion, Cambridge Univ Press

Hodder I 1992, Theory and Prac-tice in Archaeology, London,Routledge

Hodder I & Orton C 1976, SpatialAnalysis in Archaeology, Cam-bridge Univ Press

Hodder I & Shanks M (eds) 1993,Interpretive Archaeologies:Finding meaning in the past,London, Routledge

Hodges H (ed) 1987, In Situ Ar-chaeological Conservation,Santa Monica, J Paul Getty Mus

Hodges R 1991, Wall-to-wall His-tory: The story of RoystonGrange, London, Duckworth

Hole F et al 1969, Prehistory andHuman Ecology of the DehLuran Plain, Ann Arbor, Univof Michigan Mus of Anthro-pology

Holliday V T (ed) 1992, Soils inArchaeology: Landscape evolu-tion and human occupation,Washington, Smithsonian Inst

Hooper-Greenhill E 1992, Muse-ums and the Shaping of Knowl-edge, London, Routledge

Home P D 1985, ‘A review of theevidence of human endoparasit-ism in the pre-Columbian new

world through the study ofcoprolites’, J Archaeol Sci 12299–310

Horsley J 1732, Britannia Romana,London, Osborne & Longman

Hoskins W G 1988, The Making ofthe English Landscape, London,Hodder and Stoughton

Hubert J 1989, ‘A proper place forthe dead: a critical review of thereburial issue’, in Layton R,Conflict in the Archaeology ofLiving Traditions, London,Unwin Hyman, One World Ar-chaeology 8 131–66

Hudson K 1987, Museums of In-fluence, Cambridge Univ Press

Hughes M 1991, ‘Tracing tosource’, in Bowman S, Scienceand the Past, London, BritishMuseum 99–116

Hughes M J et al (ed) 1991, Neu-tron Activation and PlasmaEmission Spectrometric Analy-sis in Archaeology: Techniquesand applications, London, Brit-ish Museum Occ Pap 82

Humphreys S C & King H (ed)1981, Mortality and Immortal-ity: The anthropology and ar-chaeology of death, London,Academic Press

Impey O & MacGregor A (ed)1985, The Origins of Museums:The cabinet of curiosities in six-teenth- and seventeenth-centuryEurope, Oxford, Clarendon

Jashemski W F 1979, The Gardensof Pompeii, New York, CaratrasBros

Jenkins 1980, The Boat Beneaththe Pyramid: King Cheops’royal ship, London, Thames &Hudson

Johanson D & Shreeve J 1991,Lucy’s Child: The discovery ofa human ancestor, London, Pen-guin

Jones A 1986, ‘Fish bone survivalin the digestive systems of thepig, dog and man: some experi-ments’, in Brinkhuizen D C andClason A T, Fish and Archaeol-ogy: Studies in osteometry,

taphonomy, seasonality andfishing methods. Oxford, BritArchaeol Rep S294 53–61

Jones A et al 1988, ‘The worms ofRoman horses and other findsof intestinal parasite eggs fromunpromising deposits’, Antiq-uity 62 275–6

Jones M (ed) 1990, Fake? The Artof Deception, London, BritishMus

Jones M K (ed) 1988, Archaeologyand the Flora of the British Isles:Human influence on the evolu-tion of plant communities, Ox-ford, Oxford Univ CommitteeArchaeol Monog 14

Jones M K 1992, ‘Food remains,food webs and ecosystems’, inPollard A M, New Develop-ments in archaeological science,Oxford, Clarendon 209–219

Jones R E & Catling H W (eds)1986, Science in Archaeology,British School of Athens

Jones S 1993, The Language of theGenes, London, HarperCollins

Judd N M 1964, The Architec-ture of Pueblo Bonito, Wash-ington, Smithsonian MiscColl 147.1

Keeley H C M (ed) 1987, Environ-mental Archaeology: A regionalview 2, London, English Herit-age

Kehoe A B 1989, ‘Contextualizingarchaeology’, in Christensen AL, Tracing Archaeology’s Past:The historiography of archaeol-ogy, Southern Illinois Univ Press97–106

Kehoe A B 1991, ‘No possible,probable shadow of doubt’,Antiquity 65 129–31

Kempe D R C & Harvey A P (eds)1983, The Petrology of Ar-chaeological Artefacts, London,Clarendon

King A C & Potter T W 1990, ‘Anew domestic building-façadefrom Roman Britain’, J RomArchaeol 3 195–204

Klein R G & Cruz-Uribe K 1984,The Analysis of Animal Bones

Bibliography

201

from Archaeological Sites, Chi-cago Univ Press

Knowles W & Forster R 1910, TheRomano-British Site atCorstopitum: An account of theexcavations during 1909, Lon-don and Newcastle upon Tyne

Kucera V et al 1991, ‘“Shared Prin-ciples”: a cooperation agree-ment between a NativeAmerican group and archaeolo-gists’. Antiquity 65 917

Kuper A 1988, The Invention ofPrimitive Society: Transforma-tions of an illusion, London,Routledge

Lamberg-Karlovsky C C (ed) 1991,Archaeological Thought inAmerica, Cambridge Univ Press

Lambert P M & Walker P L 1991,‘Physical anthropological evi-dence for the evolution of socialcomplexity in coastal southernCalifornia’, Antiquity 65 870–8

Lane-Fox A 1875, ‘On the evolu-tion of culture’, Notices ProcRoy Inst Grt Brit 7 496–520

Lanteigne M P 1991, ‘Cation-ratiodating of rock-engravings: acritical appraisal’, Antiquity 65292–4

Larsen C U (ed) 1992, Sites andMonuments: National archaeo-logical records, Copenhagen,National Museum of Denmark

Layton R (ed) 1989, Who needs thePast? Indigenous values and ar-chaeology, London, UnwinHyman, One World Archaeol-ogy 5

Layton R (ed) 1989, Conflict in theArchaeology of Living Tradi-tions, London, Unwin Hyman,One World Archaeology 8

Lazreg N Ben & Mattingly D 1992,Leptiminus (Lamta): A Romanport in Tunisia, Michigan, JRom Archaeol Suppl Ser 4

Leakey R & Lewin R 1992, OriginsReconsidered, Little, Brown

Lightfoot K G et al 1993, ‘Prehis-toric shelflsh-harvesting strate-gies: implications from thegrowth patterns of soft-shell

clams (Mya arenaria)’, Antiq-uity 67 358–69

Lloyd Seton 1980, Foundations inthe Dust: The Story of Mesopo-tamian Exploration, London,Thames and Hudson

Loendorf L L 1991, ‘Cation-ratiovarnish dating and petroglyphchronology in southeasternColorado’, Antiquity 65 246–55

Longworth H & Cherry J (eds)1986, Archaeology in Britainsince 1945, London, BritishMuseum

Loy T H & Hardy B L 1992, ‘Bloodresidue analysis of 90,000-year-old stone tools from TabunCave, Israel’, Antiquity 66 24–36

Loy T H et al 1992, ‘Direct evi-dence for human use of plants28,000 years ago: starchresidues on stone artefacts fromthe northern Solomon Islands’,Antiquity 66 898–912

Lynn C J 1989, ‘Deer Park farms’,Current Archaeol 113 193–8

MacDougall H A 1982, RacialMyth in English History, Mon-treal, Harvest House

Maclnnes L & Wickham-JonesC (ed) 1992, All NaturalThings: Archaeology and thegreen debate, Oxford, Ox-bow Monog 21

Maisels C K 1993, The Near East:Archaeology in the ‘Cradle ofcivilization’, London, Routledge

Maisels C K 1993, The Emer-gence of Civilisation fromHunting and Gathering toAgriculture, Cities and theState of the Near East, Lon-don, Routledge

Malina J & Vas?íc?ek Z 1990, Ar-chaeology Yesterday and To-day: The development ofarchaeology in the sciences andhumanities , Cambridge UnivPress

Mamami Condori C 1989, ‘His-tory and prehistory in Bolivia:what about the Indians?’, in

Layton R, Conflict in the Ar-chaeology of Living Traditions,London, Unwin Hyman, OneWorld Archaeology 8 46–59

Manning S W 1991, ‘Approximatecalendar date for the first humansettlement of Cyprus?’, Antiq-uity 65 870–8

Manning S W & Weninger B 1992,‘A light in the dark: archaeologi-cal wiggle matching and theabsolute chronology of the closeof the Aegean Late Bronze Age’,Antiquity 66 636–63

Manning W H 1981, Report on theExcavations at Usk 1965–1976:The fortress excavations 1968–71, Cardiff, Univ Wales Press

MAP 1991, Management of Ar-chaeological Projects, London,English Heritage

Marian C W et al 1992, ‘Hyaenasand bones…’, J Archaeol Sci 19101–21

Mattingly D J 1992, ‘The field sur-vey: strategy, methodology, andpreliminary results’, in Lazreg NBen & Mattingly D, Leptiminus(Lamta): A Roman port in Tu-nisia , Michigan, J RomArchaeol Suppl Ser 4 89–112

McBryde I (ed) 1985, Who Ownsthe Past?, Oxford Univ PressAustralia

McDonald W A & Thomas C G1990, Progress into the Past:The rediscovery of Myceneancivilization, 2nd ed,Bloomington, Indiana UnivPress

McGuire R H 1992, A MarxistArchaeology, London, Aca-demic Press

McNeal R A 1991, ‘Archaeologyand the destruction of the laterAthenian acropolis’, Antiquity65 49–63

Meaden T 1991, The Goddess ofthe Stones: The language of themegaliths, London, SouvenirPress

Milisauskas S & Kruk J 1991, ‘Uti-lization of cattle for tractionduring the Neolithic in south

Bibliography

202

eastern Poland’, Antiquity 65562–6

Millett M & James S 1983, ‘Exca-vations at Cowdery’s Down,Basingstoke, Hampshire 1979–81’, Archaeol J 140 151–279

Millon R 1967, ‘Teotihuacan’, Sci-entific American June 38–48

Mitchell J G et al 1984, ‘Potas-sium-argon ages of schisthonestones from the Viking agesites at Kaupang, Aggersborg,and Hedeby’, J Archaeol Sci 11171–6

Modderman P 1975, ‘Elsloo, aneolithic farming community inthe Netherlands’, in Bruce-

Mitford R L S, Recent Archaeologi-cal Excavations in Europe, Lon-don, 260–86

Montelius O 1903, Dietypologische Methode: Die al-tere Kulturperioden im Orientund in Europe, Stockholm,Selbstverlag

Moore P D Webb J A & CollinsonM E 1991, Pollen Analysis,Oxford, Blackwell Scientific

Morrison J S & Coates J F (eds)1989, An Athenian TriremeReconstructed: The British seatrials of Olympias, 1987, Ox-ford, Brit Archaeol Rep S486

Mytum H 1991, The Origins ofEarly Christian Ireland, Lon-don, Routledge

Needham S P 1986, ‘Radiocarbon:a means to understanding therole of Bronze Age metal work’,in Gowlett J & Hedges R, Ar-chaeological Results from Accel-erator Dating, Oxford, OxfordUniv Committee ArchaeolMonog 11 143–50

Northover J P 1989, ‘Non-ferrousmetallurgy in British archaeol-ogy’, in Henderson J, ScientificAnalysis in Archaeology, Ox-ford, Oxford Univ CommitteeArchaeol Monog 19, 213–36

Oates D J 1976, The Rise of Civili-sation, Oxford, Elsevier/Phaidon

Oddy W A (ed) 1980, Scientific

Studies in Numismatics, Lon-don, British Museum

Oddy W A (ed) 1991, Aspects ofEarly Metallurgy, 2nd ed, Lon-don, British Museum Occ Pap17

Oddy W A (ed) 1992, The Art ofthe Conservatory London, Brit-ish Museum

Olsen S L (ed) 1988, Scanning Elec-tron Microscopy in Archaeol-ogy, Oxford, Brit Archaeol RepS452

Ortner D J & Puschan W G J 1985,Identification of PathologicalConditions in Human SkeletalRemains, Smithsonian Inst Press

Ottaway B S (ed) 1984, Archaeol-ogy, Dendrochronology and theRadiocarbon CalibrationCurve, Edinburgh Univ Press

Pader E J 1981, An Analysis ofSymbolic and SocialRelationsfrom Mortuary Remains, Shef-field, Dept Archaeol and Prehist

Parker A J 1992, Ancient Ship-wrecks of the Mediterraneanand the Roman Provinces, Ox-ford, Brit Archaeol Rep S580

Parkes P A 1986, Current ScientificTechniques in Archaeology,London, Routledge

Pearsall D M 1989,Paleoethnobotany: A handbookof procedures, London, Aca-demic Press

Pearson G W 1987, ‘How to copewith calibration’, Antiquity 6198–103

Petrie W F 1899, ‘Sequences in pre-historic remains’, J Royal AnthrInst 29 295–301

Petrie W F 1920, Prehistoric Egypt,London, British School of Ar-chaeology in Egypt

Piggott S 1965, Ancient Europefrom the Beginnings of Agricul-ture to Classical Antiquity, Ed-inburgh Univ Press

Piggott S 1976, Ruins in a Land-scape: Essays in antiquarianism,Edinburgh, Univ Press

Piggott S 1979, Antiquity Depicted:Aspects of archaeological illus-

tration, London, Thames andHudson

Piggott S 1982, ‘“Vast perennialmemorials”: the first antiquar-ies look at megaliths’, in EvansJ D, Cunliffe B & Renfrew C,Antiquity and Man: Essays inhonour of Glyn Daniel, Lon-don, Thames and Hudson 19–25

Piggott S 1985, William Stukeley,an 18th century Antiquary, Lon-don, Thames and Hudson

Piggott S 1989, Ancient Britons andthe Antiquarian Imagination:Ideas from the Renaissance tothe Regency, London, Thamesand Hudson

Pitt Rivers A 1887–98, Excavationsin Cranborne Chase, 1887,1888, 1892, 1898, London, pri-vately published

Pollard A M (ed) 1992, New De-velopments in ArchaeologicalScience: A joint symposium ofthe Royal Society and the Brit-ish Academy, February 1991,Oxford Univ Press (= Proc BritAcad 77)

Popham M R 1970, The Destruc-tion of the Palace at Knossos:Pottery of the Late Minoan IIIAPeriod, Gothenburg, StudMedit Archaeol 12

Prestwich J 1860, ‘On theoccurence of flint implements,associated with the remains ofextinct mammalia, in beds of alate geological period’, ProcRoyal Soc London 10 50–9

Price T Douglas (ed) 1989, TheChemistry of Prehistoric Hu-man Bone, Cambridge UnivPress

Pryor F 1993, Flag Fen: Prehistoricfenland centre, London, EnglishHeritage/Batsford

Rackham J 1994, Interpreting thePast: Animal bones, London,British Museum

Rackham O 1987, The History ofthe Countryside, London, Dent

Reilly P & Rahtz S (ed) 1992, Ar-chaeology and the Information

Bibliography

203

Age: A global perspective, Lon-don, Routledge, One WorldArchaeology 21

Renfrew C (ed) 1983, The Mega-lithic Monuments of WesternEurope: The latest evidence,London, Thames and Hudson

Renfrew C 1973, Before Civiliza-tion: The radio-carbon revolu-tion and prehistoric Europe,London, Cape

Renfrew C 1979, Investigations inOrkney, London, Soc of AntiqsRes Rep 38

Renfrew C 1982, Towards an Ar-chaeology of Mind, CambridgeUniv Press

Renfrew C 1984, Approaches toSocial Archaeology, Edinburgh,Univ Press

Renfrew C 1992, ‘The identity andfuture of archaeological sci-ence’, in Pollard A M, New De-velopments in ArchaeologicalScience, Oxford, Clarendon285–93

Renfrew C & Bahn P 1991, Ar-chaeology: Theories, methods,and practice, London, Thamesand Hudson

Renfrew C & Wagstaff J M (ed)1982, An Island Polity: Thearchaeology of exploitationin Melos, Cambridge, UnivPress

Reynolds P J 1979, Iron Age Farm:The Butser experiment, Lon-don, British Museum

Reynolds P J 1987, Ancient Farm-ing, Princes Risborough, ShireArchaeology 50

Richards J 1990, The StonehengeEnvirons Project, London,HBMC Archaeol Rep 16

Richards J D 1987, The Signifi-cance of Form and Decorationof Anglo-Saxon CremationUrns, Oxford, Brit ArchaeolRep 166

Riley D N 1987, Air Photographyand Archaeology, London,Duckworth

Roberts C A et al (ed) 1989, BurialArchaeology: Current research,

methods and developments,Oxford, Brit Arch Rep 211

Roberts N 1989, The Holocene: Anenvironmental history, Oxford,Blackwell

Robinson D E (ed) 1990, Experi-mentation and Reconstructionin Environmental Archaeology,Oxford, Oxbow Monograph

Rodwell K & W 1981, ‘Barton onHumber’, Current Archaeol 78208–15

Rodwell W 1989, Church Archae-ology, London, Batsford/Eng-lish Heritage

Rodwell W & K 1985, Rivenhall:Investigation of a villa, church,and village, 1950–1977, Lon-don, Council Brit Archaeol ResRep 55

Ross S et al (ed) 1991, Computingfor Archaeologists, Oxford,Oxford Univ Comm ArchMonog 18

Rossi P 1984, The Dark Abyss ofTime: The history of the earthand the history of nations fromHooke to Vico, Chicago UnivPress

Rule M 1982, The Mary Rose: Theexcavation and raising of HenryVIII’s flagship, London,Conway Press

Scarre C et al 1993, ‘New radiocarbon dates from Bougon and thechronology of French passage-graves’ , Antiquity 67 856–9

Schiffer M B 1976, Behavioral Ar-chaeology, New York, Aca-demic Press

Schiffer M B 1987, FormationProcesses of the ArchaeologicalRecord, Albuquerque, Univ ofNew Mexico Press

Schliemann H 1880, Ilios: The cityand country of the Trojans, Lon-don, John Murray

Schweingruber F H 1987, TreeRings: Basics and applica-tions of dendrochronology,Dordrecht, Reidel

Scollar I et al (eds) 1990, Archaeo-

logical Prospecting and RemoteSensing, Cambridge Univ Press

Sealy J 1986, Stable Carbon Iso-topes and Prehistoric Diets inthe South-western Cape Prov-ince, South Africa, Oxford, BritArchaeol Rep S293

Shackleton J & Elderfield H 1990,‘Strontium isotope dating of thesource of Neolithic EuropeanSpondylus shell artefacts’, An-tiquity 64 312–15

Shackley M L 1982, Environmen-tal Archaeology, London, AllenUnwin

Shanks M 1991, Experiencing thePast: On the character of ar-chaeology, London, Routledge

Shanks M & Tilley C 1987, SocialTheory and Archaeology, Cam-bridge, Polity Press

Shanks M & Tilley C 1992, Recon-structing Archaeology: Theoryand practice, 2nd ed, London,Routledge

Sharples N M 1991, Maiden Cas-tle, London, Batsford

Shennan I & Donoghue D 1992,‘Remote sensing in archaeologi-cal research’, in Pollard A M,New Developments in Archaeo-logical Science, Oxford,Clarendon 223–32

Shennan S 1985, Experiments inthe Collection and Analysis ofArchaeological Survey Data:The east Hampshire survey,Sheffield, Univ of Archaeol

Shennan S 1988, Quantifying Ar-chaeology, Edinburgh UnivPress

Sherratt A 1980, The CambridgeEncyclopedia of Archaeology ,Cambridge Univ Press

Simmons I G 1989, Changing theFace of the Earth: Culture, en-vironment and history, Oxford,Basil Blackwell

Sklenar K 1983, Archaeology inCentral Europe: The first 500years, Leicester Univ Press

Smith C 1992, Late Stone AgeHunters of the British Isles,London, Routledge

Bibliography

204

Smith R A 1920, A Guide to theAntiquities of the Bronze Age,2nd ed, London, British Mu-seum

South S 1977, Method and Theoryin Historical Archaeology, NewYork, Academic Press

Spencer F 1990, Piltdown: A sci-ent i f ic forgery, London,Natural Hist Mus Spindler K1994, The Man in the Ice,London, Weidenfe ld andNicolson

Spoerry P (ed) 1992,Geoprospection in the Archaeo-logical Landscape, Oxford,Oxbow Monog 18

Stead I M 1991, ‘The SnettishamTreasure: excavations in 1990’,Antiquity 65 447–65

Stead I M et al (ed) 1986, LindowMan: The body in the bog, Lon-don, British Museum

Stein J K (ed) 1992, Deciphering aShell Midden, London, Aca-demic Press

Stove G C & Addyman P V 1989,‘Ground probing impulse radar:an experiment in archaeologicalremote sensing at York’, Antiq-uity 63 337–42

Stringer C B 1986, ‘Direct dates forthe fossil hominid record’, inGowlett J & Hedges R, Ar-chaeological Results from Accel-erator Dating, Oxford, OxfordUniv Com mittee ArchaeolMonog 11 45–50

Stuart J & Revett N 1762–1830,The Antiquities of Athens, 5vols 1762, 1787, 1794, 1816,1830

Stuiver M Long A & Kra R (eds)1993, Calibration 1993, = Ra-diocarbon 35.1, Tucson, Univ ofArizona

Taylor R E (ed) 1976, Advances inObsidian Glass Studies: Archaeological and geochemicalperpectives, Park Ridge NJ,Noyes Press

Taylor R E 1987, RadiocarbonDating: An archaeological per-spective, London, Acad Press

Taylor R E Long A & Kra R S (eds)1992, Radiocarbon after FourDecades: An interdisciplinaryper spective, California Conf,1990, Berlin, Springer

Thompson M W 1977, GeneralPitt-Rivers: Evolution and ar-chaeology in the nineteenth cen-tury, Bradford-on-Avon,Moonraker

Tilley A 1992, ‘Three men to aroom—a completely differenttrireme’, Antiquity 66 599–610

Tite M S 1991, ‘Archaeological sci-ence—past achievements andfuture prospects’, Archaeometry32.3 139–51

Tite M S 1992, ‘The impact of elec-tron microscopy on ceramicstudies’, in Pollard A M, NewDevelopments in Archaeologi-cal Science, Oxford, Clarendon111–31

Trigger B G 1980, Gordon Childe:Revolutions in archaeology,London, Thames and Hudson

Trigger B G 1989, A History ofArchaeological Thought, Cam-bridge Univ Press

Tylecote R F & Gilmour B J J 1986,The Metallography of Early Fer-rous Edge Tools and EdgedWeapons, Oxford, BritArchaeol Rep 155

Van der Merwe N J 1992, ‘Lightstable isotopes and the recon-struction of prehistoric diets’, inPollard A M, New Develop-ments in Archaeological Science,Oxford, Clarendon 247–64

Veen M van der 1992, Crop Hus-bandry Regimes: Anarchaeobotanical study of farm-ing in northern England 1000BC–AD 500, Collis, SheffieldArchaeol Monog 3

Wacher J S 1974, The Towns ofRoman Britain, London,Batsford

Wagstaff J M (ed) 1987, Landscapeand Culture: Geographical andarchaeological perspectives, Ox-ford, Blackwell

Walker S 1984, ‘Marble origins by

isotopic analysis’, World Ar-chaeology 16.2 204–21

Weiss R 1969, The RenaissanceDiscovery of Classical Antiq-uity, Oxford, Blackwell (new ed1988)

Welsh F 1988, Building theTrireme, London, Constable

Wheeler A & Jones A 1989, Fishes,Cambridge Univ Press

Wheeler R E M 1923, Segontiumand the Roman Occupation ofWales, London, Hon SocCymmrodorion

Wheeler R E M 1954, TheStanwick Fortifications, Ox-ford, Rep Res Comm Soc AntiqLondon 17

Wheeler R E M 1954, Archaeologyfrom the Earth,Harmondsworth, Penguin

Wheeler R E M 1955, Still Digging,London, Michael Joseph

Wheeler R E M & Wheeler T V1936, Verulamium: A Belgicand two Roman cities, Oxford,Res Rep Soc Antiq London 11

Whimster R 1989, The EmergingPast: Air photography and theburied landscape , London,RCHME

Wigley T M L et al (ed) 1981, Cli-mate and History, CambridgeUniv Press

Willey G R & Sabloff J A 1980, AHistory of American Archaeol-ogy, 2nd ed, San Francisco, Free-man

Willey P 1992, Prehistoric Warfare onthe Great Plains: Skeletal analysisof the Crow Creek massacre vic-tims, Hamden CT, Garland

Williams S 1991, Fantastic Archae-ology: The wild side of NorthAmerican prehistory, Washing-ton, Smithsonian Institute Press

Williams-Thorpe O & Thorpe R S1992, ‘Geochemistry, sourcesand transport of the Stonehengebluestones’, in Pollard A M,New Developments in Archaeo-logical Science, Oxford,Clarendon 133–61

Wilson D M 1989, The British

Index

205

Museum: Purpose and politics,London, British Museum

Wilson D R 1982, Air Photo Inter-pretation for Archaeologists,London, Batsford

Wood J T 1877, Discoveries atEphesus, London, LongmanGreen

Woolley L 1907, ‘Corstopitum:provisional report on the exca-vations in 1906’, ArchaeologiaAeliana 3 161–86

Woolley Sir L 1953, Spadework,London, Lutterworth Press

Worsaae J J A (trans W J Thomas)1849, Primeval Antiquities ofDenmark, London, Parker

Yoffee N & Sherratt A (eds) 1993,Archaeological Theory: Whosets the agenda?, CambridgeUniv Press

Zeuner F E 1946, Dating the Past:An introduction togeochronology, London,

MethuenZimmerman L J 1989, ‘Made radi-

cal by my own: an archaeologistlearns to accept reburial’, in Layton R, Conflict in the Archaeol-ogy of Living Traditions, Lon-don, Unwin Hyman, One WorldArchaeology 8 60–7

Zivanovic S 1982, Ancient Dis-eases: The elements ofpalaeopathology, London,Routledge

|Aborigines, Australian, 29, 86, 181Accelerator Mass Spectrometry

(AMS), see radiocarbon datingaerial photography, 20, 23, 37–8, 41–

52, 57, 79–81, 156American civilizations, 18, 23, 35,

169, 184Americans, native, 21, 23, 26, 86,

100, 109, 111, 140, 151–4, 161,168–9, 174, 181

amino acid racemization dating, 125–6

analysis, scientific, see characterisa-tion, isotopic analysis, neutronactivation analysis, Spectrometry,trace elements, use-wear analysis,X-ray fluorescence

animals, 89, 115–8, 139–40, 143,146–55; extinct, 11–14, 51;husbandry, 49–50, 148; see alsobones, fish, insects, shells, zoology

anthropology, 10, 15–6, 21–5, 135,157, 161–73, 181

antiquarians, antiquarianism, 8–10,16–38, 61, 69, 89–90, 102, 161–3; see also Aubrey, Boucher dePerthes, Camden, Cyriac, Evans,Frere, Leland, Lhwuyd,Montelius, Peyrère, Schliemann,Stukeley, Thomsen, Winckelmann,Woolley, Worsaae

architecture, 21–3, 59, 90–100, 111;Greek and Roman, 16–17, 23, 28,31–2, 90–100, 132, 136;medieval, 21, 62, 92–6, 110, 132,

Index163; see also buildings, buildingmaterials

artefacts: antiquarian studies, 13,162, 167; early human, 11–15,18, 24, 32, 51, 58, 85, 101, 113,122, 129, 141, 159; experimentswith, 157; recording, 38–9, 49,56, 58, 61–3, 80–5, 169; use fordating, 24, 28–9, 36, 59, 63, 66,72, 84, 92, 101–7, 112, 129, 162;see also conservation, typology

Ashmolean Museum, Oxford, 20, 24,26, 34, 70

astronomy, 16, 20, 182Athens, 30–1, 90, 139, 168, 181Aubrey, John, 19–24, 30, 36, 51, 93,

179Avebury, Wiltshire, 21–2, 53, 79 barrows, see burial moundsBible, 8–18, 25, 31–3, 35, 101, 122,

162Binford, Lewis, 170–2biology, 16, 18–19, 81, 130, 143,

151, 160–1, 167–9biostratigraphic dating, 150birds, 152; see also bonesblood, 150, 154bones: animal, 11–14, 28, 51, 58,

84–5, 89, 115, 125, 139, 140–1,144, 146–50, 154, 171–2; boneartefacts, 9, 85, 106, 112, 129,139, 148; bird, 140, 150; datingmethods, 115, 120–2, 125–6,128; fish, 57, 140, 150; human,11, 15, 88–9, 113, 122, 125–6,128, 139–40, 150, 152–4, 168,181; see also otoliths, teeth

botany, 20–4, 139, 143–4; see alsoplants

Boucher de Perthes, Jacques, 11–15,18, 25, 32, 34, 61, 85, 159, 161

British Museum, London, 32–3, 50,60, 134, 181

bronze, 9, 24–7, 60, 101, 105–6,129, 131–2, 135–7, 160

Bronze Age, 13, 25, 26, 28–9, 53, 55,101, 105–6, 120, 122, 131, 135,139, 179, 182, 184; Greek, 33,103, 114, 129

building materials: brick, 90, 98–9,133; clay, 86, 89–93, 98–9, 128;concrete/mortar, 90–4, 99–100,133; mud-brick, 32, 59, 64, 89–90, 98–9; stone, 32, 59, 64, 69–74, 89–96, 99–100, 111, 131– 3,163–4, 167, 184; timber, 44, 47,49, 64–7, 71–5, 86–100, 109– 11,145–6, 155, 158; wattle anddaub, 98

buildings, structures: foundations, 44,49, 62, 75, 91–8; post-holes, 42,47, 65–7, 83, 91–2, 96–8, 100,158; reconstruction, 69, 97, 99–100, 157, 171, 177; roofing, 92,97–100, 111, 165; see alsoarchitecture

burial mounds, 21–3, 32, 53, 58–62,89–90, 109, 142, 151, 160

burials, 14, 18, 21, 26–9, 38, 57, 60–1, 85–9, 94–5, 107, 152–4, 163–4, 172, 181; cemeteries, 60, 88–9,138, 152–4, 162, 180; crema-

Index

206

tions, 88–9, 152; grave-goods, 61,88–9, 104, 107, 152, 155–6, 169;inhumations, 89; see alsoexcavation techniques, bones,human remains, megalithic tombs,pyramids, reburial, ship-burials,Sutton Hoo

Butser Ancient Farm, 156–7 Camden, William, 19–21, 26, 30, 37,

69camp sites, 85, 124, 131, 147, 151castles, 94, 175catchment analysis, 52–3cation-ratio dating, 128caves, 27, 58, 85, 123, 124, 125, 139cemeteries, see burialscharacterization, 133–7, 155Childe, V.Gordon, 162–3, 169churches, cathedrals, 17, 21, 50, 52,

69, 79, 89, 93–5, 100, 110, 164,178, 181

civilizations: discovery, 10, 30–6, 59;emergence, 115, 129, 162; seealso American, Egypt, Greek,Mesopotamia, Minoan,Mycenean, Roman

clay, see building materials, potteryclimate, 109–14, 139–44, 169, 183–4coins, 13, 17, 19–21, 28, 49, 50, 61,

76, 96, 136, 142, 160; use fordating, 27–8, 60, 64, 67–8, 70,73–4, 84, 93, 102, 106

collectors, collections, 9–10, 13–17,23–32, 36, 61, 129, 159, 180–1;see also museums

computers, computing, 39, 44–6, 56–7, 81–3, 110, 156, 177–8; see alsoGIS

concrete, see building materialsconservation: of artefacts, 60, 80, 87,

137, 138; of sites and landscapes,88, 175, 179–80

copper, 27, 104–5, 131–7Copper Age, 27, 162coprolites, 154Corbridge, Northumberland, 59–60,

64, 69–77, 96crannogs, 87crop-marks, see aerial photographycross-dating, 102–7, 112, 162, 179cultural resource management, see

heritagecultures, culture history, 86, 162,

168–9Cyriac of Ancona, 17, 19, 37 Darwin, Charles, 15, 28–9, 31, 101,

108–9, 160–2, 167–8; see alsoevolution

dating: absolute, 14, 101, 107–11,114, 123, 125, 128–9; byassociation, 11, 13–14, 27–28,

58, 64, 72–6, 85–6, 102, 104–7,111, 113, 115, 122, 125, 141;Irrelative, 14, 27, 36, 59, 101,107, 110, 128; terminus antequem, 67–8, 101, 103, 142;terminus post quem, 67– 8, 74,93, 101–2, 107, 110–11; see alsoarchitecture, coins, cross-dating,historical, sequence, seriation,stratigraphy

dating, scientific: 107–129; see alsoamino acid racemization,biostratigraphy, cation-ratio,electron spin resonance, fission-track, fluorine/uranium/nitrogen,geological, ice-sheet cores,magnetic, obsidian hydration,optical, potassium-argon,radiocarbon, seriation,thermoluminescence (TL), tree-ring, uranium series, varves,volcanic ash

decipherment, 32, 35, 101–2dendrochronology, see tree-ring

datingdiet, 139, 143, 147–54diffusionism, 105, 162–70disease, 139, 150–4; see also

pathologyDNA, 143, 149–150, 154domestication, 143, 146–9dowsing, 50Druids, 10, 21–3, 160, 179–80 ecology, 51, 86, 140, 144, 169–70economics, ancient, 16, 38, 51, 53,

94, 115, 135–6, 144–5, 162, 169,171–2; see also trade

EDM (electronic distance meter), 39,48

Egypt, 14, 16, 29–32, 36–7, 59, 63,101, 107, 140, 152–3, 162, 169,181; artefacts, 14, 35, 104, 106,129, 133, 136; historical records,17, 32–3, 101–4, 129; sites andmonuments, 16, 32–3, 58–9, 89,94, 98, 163–4

electron spin resonance (ESR) dating,123–5

English Heritage, 76, 79, 83, 138,175–80

environmental archaeology, 51–7, 80,83–4, 109–13, 128, 139–56, 169–71, 184; see also climate, plants,pollen, soil science, tree-rings

ethnicity, 154, 162, 168–9, 179; seealso race

ethnoarchaeology, 53, 85, 88, 171–2Evans, John, 11–13, 35, 159–60Evans, Arthur, 33–5, 59, 70, 106evolution, 8–9, 15, 19, 28–9, 101,

108–9, 129, 150, 161–2, 167–9;see also social evolution

excavation techniques: 19th century,23–7, 30, 33–6, 58–63, 69, 85,90; burials, 32, 49, 50, 58–62,88– 9, 95, 152; open area, 64–8,74–6, 81, 91–8; sections, 14, 35,62–9, 72–3, 76, 81–4, 99; stonestructures, 71, 90–4; timberstructures, 67, 75, 91, 95, 97, 98,100, 110; trenches, 64–7, 72–3,76, 80–1, 87, 90, 93, 96;underwater, 50, 87–8

experimental archaeology, 130, 147,150, 156–8, 171; see also buildingreconstruction, Butser, use-wear

fakes, 124–9, 138feminism, 168, 175, 182; see also

genderfieldwork, 22–4, 32–58, 62, 74, 77–

80, 102, 112, 142, 145, 151, 155,159–60, 165; antiquarian, 17–24,35; field survey, 38, 41, 53, 56,77, 155; fieldwalking, 38, 42, 54

fish, fishing, 57, 147, 150, 154, 161,164, 167; see also bones

fission-track dating, 122, 125, 137flint, 11–14, 24–6, 34, 39, 54, 85,

96, 107, 123–4, 132, 141, 160flotation, 80, 140, 150fluorine/uranium/nitrogen dating,

125fluxgate gradiometer, 49formation processes, 85, 89, 147,

170–2fossils, 15, 21, 61, 64, 101, 109Frere, John, 11–15, 25, 32, 61, 161 gender, 181–3genetics, 149, 154; see also DNAgeography, 51, 56–7, 144, 156, 165;

see also GISgeological dating, 14, 107–9, 113–

15, 122–3geology, 8–19, 25, 31, 35, 39, 44–5,

54–57, 61, 64, 80, 85, 101, 107–9, 112–14, 122–8, 131–7, 141,150–2, 160–1; see also fossils,petrology, stone, stratification

geophysical prospecting andsurveying, 37, 42, 46–50, 65, 77,80–1, 130, 156; see also dowsing,metal detectors, radar, sonar

GIS (Geographical InformationSystems), 56–7, 83, 156

glaciers, 112, 114, 132graves, grave-goods, see burialsGreek archaeology, 28, 30–1, 34, 59,

90, 101, 158; see also architectureGreek civilization, 9, 17, 30–1, 34–5,

168

Index

207

Hadrian’s Wall, 19, 22, 69–72, 92,103, 149

Hadrian, Emperor, 16–17, 89, 103Harris Matrix, 68, 82Herculaneum, Italy, 103, 153heritage, 58, 88, 138, 175–80historical archaeology, 10, 17, 32–3,

59, 87–8, 101–3, 106, 114, 132,169, 171, 175; see also Greekarchaeology, medieval archaeol-ogy, Roman archaeology

historical dating, 32, 72–6, 101–4,122, 127–9

Hodder, Ian, 171–3hominids, 86, 101, 113, 122, 125,

141, 150, 153–4, 172Hoxne, Suffolk, 11, 141human antiquity, 9–16, 19, 30, 34– 6,

61, 109, 113, 159, 161human remains, 88, 121, 131, 140,

152–4, 174, 182; see also bones,hominids, mummies, pathology

hunter-gatherers, 28, 53, 55, 85, 98,115, 121, 129, 131, 139, 141,144–7

Ice Age, 51, 54–5, 106, 109, 112–

13, 132, 139, 141, 144–5, 151Ice Man, 130–2, 140, 152ice-sheet cores, 103, 111–13, 141,

183Indians, see Americans, nativeindigenous peoples, rights of, 86,

169, 175, 181industrial archaeology, 102, 131,

175, 178inscriptions, 17, 23, 28, 32–5, 38, 51,

70, 73, 76, 102, 132insects, 140, 146, 150, 154Inuits, 140, 152, 171iron, 9, 25–7, 60–1, 101, 105, 120,

127–8, 132–3Iron Age, 20–1, 25, 27–9, 46, 54–5,

63–6, 96, 101, 120, 140, 152,171, 182–3

isotopic analysis, 151–3 Jorvik, see York Knossos, Crete, 33–6, 59–60, 106 landscape archaeology, 43–5, 51–7,

80, 140, 156, 184lead, 109, 114, 123, 135–7leather, 80, 138, 140, 177Leland, John, 19–21Lhwuyd, Edward, 21, 26Libby, Willard F, 115–16, 119, 129Lubbock, John, 28, 167Lyell, Sir Charles, 15, 109, 160

magnetic dating, 112, 127–9magnetic reversals, 113, 122, 128,

141magnetic susceptibility, 47–9magnetometers, 46–9, 80marble, 17, 92, 99, 137Marx, Karl, 161–2, 169Mary Rose, 50, 88, 94mathematics, 51, 56, 155–6, 182; see

also statisticsmedieval archaeology, 21, 39, 46, 50,

52, 64, 79, 86–7, 93, 106, 139,148; see also architecture

megalithic tombs, 89, 162–7, 172,182

Mesolithic, see Stone AgeMesopotamia, 35, 59, 98–9, 102,

104, 162, 184metal detectors, 49–50, 61, 181; see

also treasure huntingmetals, see bronze, copper, iron, lead,

silver, tinmetalworking, 129, 131–3, 135–7,

170microscopes, 110, 125, 131–4, 137,

142–4, 151, 157; SEM, 131–2,143, 149

Middle Range Theory, 172Minoan civilization, 34–6, 59–60,

103, 106molluscs, see shellsMontelius, Oscar, 28–9, 104–5, 162–

3Morgan, Lewis, 161mortar, see building materialsmud brick, see building materialsmummies, 140, 152–3museums, 20, 24, 26–7, 30, 32–6, 49,

70, 74–7, 80, 84, 129–30, 134,137–8, 153, 156, 159, 167– 8,175–83

Mycenaean civilization, 33–4, 104 nationalism, 18, 31, 77, 162, 167–8Neanderthals, 115Near East, 18, 24, 29–37, 58–9, 63,

64, 89, 92, 98–9, 101, 105, 134–5, 162, 163

Neolithic, see Stone Ageneutron activation, 134new archaeology, 51, 160, 169–73 obsidian, 125–8, 134–7, 155obsidian hydration dating, 125–8Olduvai Gorge, Tanzania, 86, 102,

172optical dating, 124otoliths, 150 palynology, see pollen analysispathology, 89, 149–54; see also

disease

Petrie, Flinders, 59, 63, 104, 107petrology, 131–5phosphate analysis, 49–50, 142; see

also soil sciencePeyrère, Isaac de la, 18, 25photography, 33, 37 69, 80–4, 90,

133, 138–9; photomicrographs,126, 131, 149; see also aerialphotography, video

Piltdown Man, 125Pitt Rivers, General, 13–14, 27–30,

36, 59–63, 70, 83, 94, 96, 104–5,167–8, 175–6

place-names, 19, 51, 80plants, 51, 86, 112, 116–18, 131,

139–45, 154, 169, 171; plantopals, 154; see also botany,cereals, pollen analysis, seeds,trees

politics, 16, 30, 38, 77, 136, 161,167–9, 172–81; see also Marx,nationalism

pollen analysis, 50–1, 101, 112–4,120, 141–6, 152, 154, 169, 183

Pompeii, Italy, 33, 59, 74, 103, 143,153

population, 38, 45, 55, 139, 153–4,165, 180, 184

post-holes, see buildingspost-processualism, 169, 172–4, 178–

82potassium-argon dating, 107–9, 122–

5, 128–9, 137, 141pottery, 28, 36, 39, 54–5, 59–60, 64,

66–74, 76, 80–4, 87, 92–6, 103–7, 112, 123–9, 131–5, 143, 147,155, 157, 162, 166, 177

prehistoric archaeology, 9–10, 18,27–30, 36, 38, 51, 59, 86–7, 101–2, 112–15, 121–4, 162–9, 181–2

Prestwich, Joseph, 11–15processualism, 51, 53, 168, 170–4public, and archaeology, 24, 26–7,

31, 49, 76, 79, 100, 153, 159–60,167–8, 175, 179

pyramids, 89, 94, 99, 129, 163–4 race, racism, 23, 162, 167–8, 182;

see also ethnicityradar, 50radioactivity, 109, 111–19, 122–5,

128–9, 133–7, 141, 150–1, 153,156, 183

radiocarbon dating, 9, 29, 36, 97,105–31, 136, 141, 144–5, 162,164–7, 169, 179; AMS, 106, 115–6, 118, 120–1, 128–9, 136;calibration, 9, 113–23, 126, 155–6, 164–7, 179; first-order, 151

reburial of human remains, 181relativism, 174, 179, 181remote sensing, see aerial photogra-phy

Index

208

Renaissance, 16–19, 21, 23–4, 28,30–1, 36, 90, 93, 159

rescue archaeology, 23, 41, 46, 64,69, 74–9, 83, 87, 176–7

resistivity, 46–8, 80Roman archaeology, 23, 31, 38, 90,

101; forts, 42–3, 46, 64–6, 71–6,100, 102–3, 110; roads, 23, 69–70, 72–4; villas, 44, 93, 139; seealso architecture

Roman civilization, 9, 16–17, 30–1,35, 162, 168

Roman Empire, 16–17, 21, 31, 38,54–5, 69–70, 100

Rome, Italy, 16–17, 38, 89, 99, 139,184

Rosetta Stone, 32 Schliemann, Heinrich, 30–4, 59, 63sea-bed cores, 113, 122, 128, 141sections, see excavationseeds, 80, 140, 143, 154, 169SEM, see microscopessequence dating, 107seriation, 107, 145, 156, 171settlement patterns, 38, 43, 45, 51–

5, 111–12, 114, 141, 184shadow sites, see aerial photographyshells, 113, 120–6, 140, 142, 144,

147, 150–2; see also snailsship-burials, 50, 60–1, 112ships, shipwrecks, 38, 49–50, 60–1,

88–9, 94–5, 100, 112, 132, 146,157–8

Silbury Hill, Wiltshire, 21–2, 53, 79silver, 133, 136, 138simulation studies, 156–7Sites and Monuments Records, 39–

41snails, 151–2; see also shellssocial evolution, 16, 18, 101, 161–3,

167–9sociology, 161, 174soil science, 49–50, 52–6, 80–1, 84,

89, 140–2, 146, 150, 154, 169,184

soil-marks, see aerial photographysonar scanners, 38, 50spatial archaeology, 51, 56, 94spectrometry, 116, 123, 132–5stalagmite, 123–5statistics, 51–2, 54, 57, 107, 117–19,

122, 134, 146–8, 155–6, 170–71;see also mathematics

stone, see flint, obsidian, marbleStone Age: Mesolithic, 28, 55, 86,

112, 139, 144, 150; Neolithic, 13,28, 36, 44, 53–7, 62, 64, 66, 96–8, 101, 105, 107, 110, 119– 21,131–2, 134, 139–40, 144–5, 147,149, 151–2, 157, 169, 172, 182;

Palaeolithic, 25, 28, 51, 90, 122,124, 139, 144, 159, 162, 182

Stonehenge, 19–23, 53, 79, 90, 132,178–80

stratification, archaeological, 28, 32–6, 50, 58–69, 72–6, 81–5, 90– 4,98, 101–2, 107, 120, 125, 128–9,143, 148, 151, 156; see alsoexcavation, Harris Matrix

stratification, geological, 11, 14–15,58, 61, 64, 85, 101, 122, 141,144–5, 159

structuralism, 172–3Stukeley, William, 19, 21–4, 36–7,

39, 51, 69, 160, 163, 179surveying, 22, 33, 37–9, 43–50, 52,

62, 65, 72, 80–3, 96; see alsoEDM

Sutton Hoo, Suffolk, 50, 60–1, 89,95, 152

symbolism, 157, 173, 178–82systems theory, 52–3, 170–2 Tacitus, 16, 18, 21, 102–3taphonomy, see formation processesteeth, 123–6, 147–9, 152–4; see also

bonestells, 36–7, 59, 63–4, 98–9temples, 32–3, 76, 90terminus ante quem, terminus post

quem, see datingtheory, archaeological, 51–3, 159–

61, 168–74, 178–82; see alsoethnoarchaeology, MiddleRange Theory, ‘new’ archaeol-ogy, post-processualism,processualism, structuralism,systems theory

Thera (Santorini), 103, 114thermoluminescence (TL) dating, 94,

122–4, 129, 133Thomsen, C J, 26–8, 30, 38, 61, 101Three-Age System, 8–9, 25–30, 61,

101, 162; see also Bronze Age,Copper Age, Iron Age, Stone Age

timber, 94–5, 109–10, 114, 140; seealso building materials, wood,trees

tin, 135–6tourism, 16–17, 59, 77, 100, 159,

176–7towns, see urban archaeologytrace elements, 134–7, 153trade, 10, 16, 52–3, 88, 106, 132–7,

154, 161, 164, 169, 172travellers, explorers, 9, 16–19, 24–5,

30, 33, 37, 69, 90treasure hunting, 30, 49, 61, 69, 70,

88, 173, 181; see also metaldetectors

Treasure Trove, 50tree-ring dating, 87, 94, 101–4, 108–

21, 129, 140–1, 146, 164;environmental implications, 141–2, 146, 154, 183–4

trees, 50, 57, 109–16, 121, 139, 142–6, 183–4

Troy, 30–4, 59typology, 27–9, 36, 38, 52, 59, 76,

103–7, 128–9, 131, 156, 162,166–7, 169; see also artefacts

underwater archaeology, 38, 50, 58,

87–8uranium series dating, 109, 114, 116,

121–5urban archaeology, 63–4, 68, 78, 81,

93–4, 96, 147, 151, 177use-wear analysis, 131, 149, 157 varves, 101, 112–14, 121Verulamium (St Albans), Herts, 63–

4, 93video, 45, 83–4, 110, 138, 177; see

also photographyvolcanic ash, use for dating, 103,

114, 122, 141volcanoes, 35, 60, 103, 111, 113–4,

122, 137, 141–3, 153, 183–4 weather, see climatewetland sites, 86–87; see also

crannogsWheeler, Mortimer, 63–5, 68, 72, 93,

103Wheeler, Tessa, 63–4Winckelmann, Johann, 23, 28wood, 80, 86, 98, 108, 110–12, 116,

120–1, 140, 146, 157; artefacts,9, 27, 76, 87, 106, 121, 129,138– 9; see also timber, trees

woodland, 45, 54, 139–40, 142,144–6, 151, 157, 184; see alsotrees

Woolley, Leonard, 59, 70Worsaae, J, 27–8, 30, 32, 61 X-ray examination, 133, 137–8, 152X-ray fluorescence, 134 York, 50, 69, 94, 177; ARC

(Archaeological Resource Centre),177–8; Jorvik Viking Centre,177–9; York ArchaeologicalTrust, 177

zoology, 139, 146