. 1
( 7)


This page intentionally left blank
Archaeological Theory and Scienti¬c Practice

Is archaeology an art or a science? This question has been hotly debated
over the last few decades with the rise of archaeological science. At the
same time, archaeologists have seen a change in the intellectual character
of their discipline, as many writers have adopted approaches in¬‚uenced
by social theory. The discipline now encompasses both archaeologi-
cal scientists and archaeological theorists, and discussion regarding the
status of archaeology remains polarised. Andrew Jones argues that we
need to analyse the practice of archaeology. Through an analysis of
archaeological practice, in¬‚uenced by recent developments in the ¬eld of
science studies, and with the aid of extensive case studies, he develops
a new framework, which allows the interpretative and methodological
components of the discipline to work in tandem. His reassessment of
the status and character of archaeology will be of interest to students,
scholars and professionals.

is a Lecturer in the Department of Archaeology,
Southampton University. He has worked extensively on British pre-
history (especially the Neolithic and Bronze Age). Among his many
research interests are the history of representation in archaeology, the
role of art and memory in archaeological research, and the archaeology
of animals and food. He has contributed to a number of journals and
edited volumes. This is his ¬rst book.
Topics in Contemporary Archaeology

Series Editor
Richard Bradley, University of Reading

This series is addressed to students, professional archaeologists and
academics in related disciplines in the social sciences. Concerned with
questions of interpretation rather than the exhaustive documentation of
archaeological data, the studies in the series take several different forms:
a review of the literature in an important ¬eld, an outline of a new area
of research or an extended case study. The series is not aligned with
any particular school of archaeology. While there is no set format for
the books, all books in the series are broadly based, well written and
up to date.
°µ¬©¤   ° ®¤© ¦  µ®©© ¦ ©¤§
The Pitt Building, Trumpington Street, Cambridge, United Kingdom

©¤§ µ®©© °
The Edinburgh Building, Cambridge CB2 2RU, UK
40 West 20th Street, New York, NY 10011-4211, USA
477 Williamstown Road, Port Melbourne, VIC 3207, Australia
Ruiz de Alarcón 13, 28014 Madrid, Spain
Dock House, The Waterfront, Cape Town 8001, South Africa


© Cambridge University Press 2004

First published in printed format 2001

ISBN 0-511-03136-X eBook (Adobe Reader)
ISBN 0-521-79060-3 hardback
ISBN 0-521-79393-9 paperback
Archaeological Theory and Scienti¬c Practice

Andrew Jones
Department of Archaeology, Southampton

List of illustrations page viii
List of tables x
Preface xi
Acknowledgements xiv

1 The archaeology of ˜two cultures™ 1
2 Science as culture: creating interpretative networks 23
3 Archaeology observed 39
4 Materials science and material culture: practice, scale
and narrative 63
5 Material culture and materials science: a biography
of things 83
6 A biography of ceramics in Neolithic Orkney 103
7 Making people and things in the Neolithic: pots, food
and history 145
8 Before and after science 168

References 183
Index 203


2.1 GC retention graph of sample SF 2,000 from a GC
integrator page 31
2.2 Histogram of sample SF 2,000 from GC plot 32
3.1 Exploding excavations 43
3.2 The transformation of artefacts as data from excavation
to laboratory 48
4.1 The changes of perception allied to changes in analytical
scale 67
4.2 Artefacts and their contexts as boundary objects 75
4.3 A schematic view of the Haya furnace 79
6.1 A selection of Neolithic material culture exhibiting similar
curvilinear motifs 108
6.2 Map showing Orkney archipelago 109
6.3 The spatial layout of the Later Neolithic house in Orkney 110
6.4 The spatial homology between passage grave, house
and henge 112
6.5 The distinction between incised Grooved ware and applied
Grooved ware 113
6.6 Map of the central area of Mainland Orkney indicating
position of principal monuments 114
6.7 Plan of the Later Neolithic settlement at Barnhouse 116
6.8 Graph of fabric plotted against wall thickness 121
6.9 Large Grooved ware vessel from Barnhouse 123
6.10 Two medium-size vessels from Barnhouse with
characteristic decorative schemes 124
6.11 Medium-size vessel from Barnhouse with serpentine
applied cordons 125
6.12 Sherds from small vessels from Barnhouse with passage
grave art motifs 126
6.13 Plan of the central area at Barnhouse 127

List of illustrations ix

6.14 Schematic diagram indicating the distinction in the use
of temper between inner houses and peripheral houses
at Barnhouse 128
6.15 The location of dyke rock sources in the Barnhouse
environs 129
6.16 The spatial location of sherds in a typical house at
Barnhouse 134
6.17 House 2, Barnhouse 136
6.18 The Grooved ware from house 2, Barnhouse 137
6.19 Plan of structure 8, Barnhouse 139
6.20 The Grooved ware from structure 8, Barnhouse 140
6.21 Plan of the early phase at Barnhouse indicating the
depositional relationships between individual houses
and middens 141
6.22 Schematic plan of Barnhouse indicating the relationships
between shell-tempered pottery and the central area, and
rock-tempered pottery and the periphery of the settlement 143
7.1 The Grooved ware from the ditch at the Stones of Stenness
henge 150
7.2 Plan of the Stones of Stenness henge indicating the
position of Grooved ware sherds 151
7.3 Plan of Quanterness passage grave 152
7.4 Relationship between Quanterness passage grave and
the Barnhouse settlement 153
7.5 The Quanterness Grooved ware 155
8.1 The relationship between the resistance and
accommodation of material and human agency 172
8.2 The relationship between the observational techniques
of science and the representations of the concrete nature
of the world by scientists 173
8.3 The intimate relationship between material agency
and its context of in¬‚uence depends upon the nature
and extent of the scale of analysis 180

6.1 Presence/absence of tempering agents in different
houses at Barnhouse page 128
6.2 Simpli¬ed results of GC examination of sherds
from large Grooved ware vessels at Barnhouse 132
6.3 Simpli¬ed results of GC examination of sherds from
medium Grooved ware vessels at Barnhouse 133
6.4 Simpli¬ed results of GC examination of sherds from
small Grooved ware vessels at Barnhouse 133
7.1 Depositional contexts for animal species in Later
Neolithic Orkney 159


Since the contents of this book are concerned so much with issues of
biography, it makes sense to begin by saying something about the biogra-
phy of both text and author. The subject matter “ the relationship between
archaeological theory and archaeological science “ arose from my doc-
toral research between 1993 and 1997 at Glasgow University, which was
supervised by Colin Richards and Richard Jones. The examination of
the pottery assemblage from the Late Neolithic settlement at Barnhouse,
Orkney comprised the central focus of the original thesis (see Richards
forthcoming, and chapters 6 and 7 this volume). However I felt that wider
and more fundamental questions lay behind my use of the techniques of
materials science within a framework informed by interpretative archae-
ology and anthropology. It was for this reason that I began to write the ¬rst
two chapters of the book in Glasgow, after the completion of the thesis.
At this time the subject matter was written from a personal perspective
derived from attempts to balance an interest in archaeological theory with
the practical application of scienti¬c techniques. This perspective altered
when I took up a teaching appointment at University College Dublin,
where amongst other things I was able to observe the pragmatic applica-
tion of scienti¬c analysis alongside archaeological theory under the aegis
of the Irish Stone Axe Project, directed by Gabriel Cooney and Stephen
Mandal. I began to see that the issues examined in the volume were more
fundamental to archaeological practice, and in Dublin I completed the
third chapter.
I was persuaded more ¬rmly of the subject matter of the book when I
took up a post-doctoral position at the McDonald Institute for Archaeo-
logical Research, Cambridge. In Cambridge I came into contact with a
growing number of people who were attempting to utilise both archae-
ological theory and archaeological science. My perspective on the topic
had shifted over the course of the book™s inception in Glasgow to its com-
pletion in Cambridge some two years later. No longer did it appear to
derive solely from personal experience; instead, it had become a topic that
was of wider concern to a growing number of archaeological scientists

xii Preface

and archaeological theorists. This was encouraging although, of course,
this state of affairs had prevailed throughout, since in reality we are never
writing in isolation, but are always situated in a wider discourse.
My immersion in this discourse is not solely con¬ned to my engage-
ment with issues of science and society, and science and the arts in the
academic world; these issues have an increasing impact upon the world
which we all inhabit. I write at a time in which faith in science as a
force of emancipation has diminished and public con¬dence in the sci-
ences has waned. Genetically Modi¬ed Organisms, the Human Genome
project and the issues surrounding the cloning of human tissue from stem
cells are at present regular topics of discussion in the media. The terms
in which these critical issues are discussed remain polarised in the frame-
work that I describe in chapter 1, with scientists in the media occupying
a position of certainty and knowledge sealed off from the wider concerns
of the public. Meanwhile, while concern grows for the ethical issues asso-
ciated with the newfound capabilities of the biological sciences, there is a
lingering assumption of the inexorable and progressive nature of science.
The discussion of these issues is then caught in a problematic trap: while
it is realised that at the ethical level society ought to have an impact upon
science, there remains the feeling that science proceeds outside the in¬‚u-
ence of the social. On a lighter note, the signi¬cance of the relationship
between science and society, and in particular science and the arts, is also
being increasingly stressed in the form of a number of major visual arts
exhibitions at venues ranging from the Hayward Gallery and the Natural
History Museum, London to the Victoria and Albert Museum.
While the wider issues concerning the sciences in relation to society and
the arts have affected me both negatively and positively, on an academic
and personal level my perspective on the philosophical implications of
these issues has fundamentally altered during the course of writing this
book. I have become convinced of the necessity of taking account not
only of the philosophical implications of our practices, an area tradition-
ally studied by philosophers of science (Embree 1992; Kelly and Hannen
1988; Wylie 1992), but also to understand the historical precedents and
trajectories of these philosophical distinctions. In this respect I have been
especially in¬‚uenced by Barkan and Bush (1995), Fabian (1983, 1991)
and Stocking (1996), amongst others. Although the history of science is
a relatively unexplored ¬eld in archaeology and remains fairly implicit in
my text, I nevertheless feel it is critically important to be aware of the
historical depth of the philosophical distinctions that we employ on a
routine basis in our contemporary practices. Moreover I believe that it
is important to re¬‚ect upon this awareness in the reformulation of our
philosophical frameworks. That is really what this book is about, since
Preface xiii

the aim is to examine the philosophical distinctions that divide the arts
and humanities from the natural sciences. In this regard it would have
been relatively simple to write an account that ˜took sides™. Radically
different accounts could have been written had I taken up the view of
the natural sciences in de¬ning positivism or empiricism as de¬nitional
knowledge (for the most famous recent examples of this approach see
Gross and Levitt 1994; Sokal and Bricmont 1998). Similarly, in taking
up a perspective ¬‚avoured by post-modernism it would have been pos-
sible to write an account which considered knowledge to exist in solely
representational form. Both of these approaches would have fallen foul
of the epistemological traps that ensnare our discussion of topics such
as rationalism and relativism, objectivity and subjectivity. With Fabian
(1991, 193) I believe that ˜it is a bad sort of critique that ¬rst needs to
pledge allegiance to one or another school™; instead, I have attempted to
develop a position that examines the nature of the connections between
each order of knowledge, and my account is meant to alienate neither
archaeological scientists nor archaeological theorists.
Due to the exigencies of space, this book focuses upon materials sci-
ence. However I am aware that excellent work of a similar vein is also being
undertaken in many other ¬elds of archaeological science, such as envi-
ronmental archaeology (Albarella forthcoming), soil micromorphology
(Boivin 2000), stable isotope analysis (Richards and Hedges 1999) and
Geographical Information Systems (Lock and Stancic 1995), to name
but a few. Furthermore some of the themes addressed in this book are of
wider concern to ¬eld archaeology, and these have been recently exam-
ined by Bender, Hamilton and Tilley (1997) and Hodder (1996, 1999).
In terms of my theoretical emphasis, I have focused upon issues such as
biography, consumption, technology and identity that are of pertinence
to interpretative archaeologists and anthropologists alike. It goes with-
out saying that the application of techniques derived from archaeological
science to the examination of theoretical issues need not focus on these
areas of interest alone. Rather it is the imaginative application of both
existent and novel techniques to a plethora of theoretical issues that will
promote the creation of fresh interpretative networks between researchers
in different ¬elds.


Over the course of writing this book I have enjoyed chatting to numer-
ous individuals about these topics. These people include Dean Arnold,
John Barrett, Kishor Basa, Robin Boast, Dusan Boric, Richard Bradley,
Emmet Byrne, Gabriel Cooney, Jo Sofaer-Derevenski, Chris Doherty,
Bryan Hanks, Yannis Hamilakis, Andy Hoaen, Cornelius Holtorf,
Andrew ˜Bones™ Jones, Stephanie Koerner, Mark Knight, Helen Loney,
Kirsi Lorenz, Stephen Mandal, Lesley McFadyen, Stephanie Meece,
Karen Milek, Preston Miracle, Lise Nordenborg-Myhre, Mike Parker-
Pearson, Jenny Rose, Hannah Sackett, Katerina Skourtopolou and Mike
I must also thank the staff of the Glasgow University Archaeological
Research Division (GUARD) for teaching me a number of valuable
lessons regarding the nature of contract archaeology. My time in Dublin
was made all the more pleasant by discussions with Gabriel Cooney,
Seamas Caul¬eld and Muiris O™Sullivan, and the hospitality (intellectual
and otherwise) of Margaret Coughlan and Tadhg O™Keeffe.
I am especially grateful to Dean Arnold, Jo Sofaer-Derevenski, Yannis
Hamilakis and Mark Knight, all of whom commented upon parts of
the text, and to David Williams for the loan of thin-sections. I am also
grateful both to Chatto and Windus publishers and the executors of the
estate of F. R. Leavis for their permission to quote from F. R. Leavis™
The two cultures? The signi¬cance of CP Snow, and to Cambridge Univer-
sity Press and the executors of the estate of C. P. Snow for their permis-
sion to quote from Two cultures. Further thanks must go to Colin Renfrew
for permission to use the illustration of the Quanterness Grooved ware
Special thanks must go to two people: Richard Bradley, who has been
a constant source of encouragement throughout the writing of the book
and whose editorial comments have been both incisive and an inspiration
to further work, and Hannah Sackett, who has kept me on the straight
and narrow during the writing of this book.

Acknowledgements xv

Finally, the book is dedicated to two remarkable individuals. My
father, Dr Edward Jones, a scientist with strong opinions on the cer-
tainty and reliability of scienti¬c knowledge and a ¬‚air for art, and my
mother, Felicity Jones, a historian with a sense of the value of history who
also keeps an eye on the empirical matters of life.
1 The archaeology of ˜two cultures™

I have had, of course, intimate friends among both scientists and writ-
ers. It was through living among these groups and much more, I think
through moving regularly from one to the other and back again that I got
occupied with the problem of what, long before I put it on paper, I chris-
tened to myself as the ˜two cultures™. For constantly I felt I was moving
among two groups “ comparable in intelligence, identical in race, not
grossly different in social origin, earning about the same incomes, who
had almost ceased to communicate at all, who in intellectual, moral and
psychological climate had so little in common. (C. P. Snow 1959, 2)

The only presence science has is as a matter of external reference, en-
tailed in a show of knowledgeableness. Of qualities that one might set
to the credit of scienti¬c training there are none. As far as the internal
evidence goes, the lecture was conceived and written by someone who
had not had the advantage of an intellectual discipline of any kind. I was
on the point of illustrating this truth from Snow™s way with the term
˜culture™ “ a term so important for his purposes. By way of enforcing his
testimony that the scientists ˜have their own culture™, he tells us: ˜This
culture contains a great deal of argument, usually much more rigorous,
and almost always at a higher conceptual level, than literary persons™ ar-
guments™. But the argument of Snow™s Rede Lecture is at an immensely
lower conceptual level, and incomparably more loose and inconsequent
than any I myself, a literary person, should permit in a group discussion
I was conducting, let alone a pupil™s essay. (F. R. Leavis 1962, 14“15)

The extracts above are taken from two Cambridge lectures. The ¬rst, de-
livered by the late Sir Charles Snow, a scientist and author, sketches
the problem which he considers to be inherent to twentieth-century
academia, that of the ˜two cultures™, divided conceptually between those
who study science, and those who study the arts. The outline of the lec-
ture, as indicated from this extract, suggested that the two disciplines were
simply not talking to each other. This extract illustrates quite clearly the
point that I wish to make in this opening chapter with regard to contem-
porary archaeology; that is, that archaeological scientists and theoretical
archaeologists are quite simply speaking in different languages and have

2 Archaeological theory and scienti¬c practice

quite different visions of what the study of archaeology entails. This para-
doxical disciplinary position has served to force both a vigorous critique
of positivism on the side of those practising interpretative or theoretical
approaches (see Thomas 1990) and a whole-hearted rejection of post-
structuralist theory on the part of those practising scienti¬c archaeology.
Here the position can be summed up by Dunnell™s assertion that ˜the ef-
fort, rigour and cost of physical analyses are lost in a humanistic approach
where they serve only to aspire story telling™ (1993, 164).
Of course, as Snow™s extract indicates, the division of intellectual labour
between the arts and sciences remains a long-standing problem. How-
ever, very few disciplines attempt to bridge the intellectual gap between
these bodies of knowledge. The question I wish to ask in this ¬rst chapter
is do we bridge the gap or do we in fact practise two different kinds of
archaeology, each of which produces different orders of knowledge about
the past? The aim of this book will be an attempt to examine the problems
facing contemporary archaeology as a discipline that is essentially split
in its theoretical and methodological aims. The second question I wish
to consider is whether this split is theoretically and methodologically
surmountable, or whether the two orders of knowledge are ultimately
The second quotation is from a lecture delivered some years later by
the late F. R. Leavis, a professor of English Literature and a prominent
literary critic. This second lecture inveighed against the coarse-grained
nature of Snow™s argument, against Snow himself and, to some extent,
against science itself as the talisman with which to heal all ills. This ex-
tract illustrates the intensity that the debate between scientists and artists
often reaches. Such intensity of debate certainly has its parallels in the
archaeological literature since the 1960s.
While Snow was both writer and scientist, his sympathy lay with sci-
ence. His interpretation of the problematic relationship between science
and the arts was simplistic; he saw science as the way forward, believ-
ing it to be more rigorous than the arts, and more capable of providing
both truth and answers for society™s problems. Science would emerge as a
latter-day holy grail, enabling the gap between rich and poor to be ¬nally
bridged. Leavis™ main point concerned the quality and rigour of Snow™s
argument, and he rightly noted that science by itself held little promise
if it was not linked to a clearer understanding of society. As we shall see,
the debates between the arts and the sciences over rigour, truth and the
application of science have considerable resonance with the problems we
need to face in examining the position of science and interpretation in
the wider archaeological programme.
The archaeology of ˜two cultures™ 3

The intellectual division outlined above is not peculiar to the subject
of archaeology; rather, the epistemological division between arts and sci-
ences is a major concern in the construction and understanding of all
forms of knowledge. When discussing the different intellectual positions
taken up in constructing different orders of knowledge, we ¬nd that there
are a plethora of terms used to de¬ne these interpretative positions. The
de¬nition of terms is a traditional issue of contention for those criticising
opposing knowledge claims (for example see Reyna 1995). Therefore,
in the proceeding section I wish to clearly outline the major problems in
our discussion of differing domains of knowledge, to de¬ne the terms in
which they are discussed, and to examine the ways in which they relate to
each other. This clari¬cation exercise is necessary before we proceed on
to consider how these varying theoretical positions have been discussed
within archaeology. In the account below it will not be possible to de¬ne
the precise details of each theoretical position; rather I intend to pro-
vide a broad overview of the epistemological problems which face both
the natural and social sciences. Overall, I want to critically evaluate the
practice of science and examine ways in which theoretical or interpreta-
tive archaeologists may engage with science. Meanwhile, I also wish to
demonstrate the necessity of social theory within archaeology, and sug-
gest ways in which scienti¬c archaeologists may critically engage in social

Objectivity and subjectivity
Conventionally, within Western philosophical traditions “ at least since
Descartes and the early work of Kant (see Toulmin 1990 for discussion
of the historical origins of Cartesian dualisms) “ the world has been per-
ceived to be composed of two things with differing properties, generally
described as objects and subjects. Nature “ the world of objects “ is seen as
an inanimate and immutable essence that existed prior to its description
by subjects. Subjects, on the other hand, are perceived as animate and
are therefore invested with the ability to act and describe the inanimate
world of objects. This section will consider the processes and methods by
which scientists, philosophers and sociologists investigate this apparent
According to an objectivist position the world consists of objects which
exist ˜out there™, beyond the internal world of human subjects. The re-
lations pertaining between these objects can be adequately described,
discussed and studied by perceiving them and then representing them
through language. The relationship between our language terms and the
4 Archaeological theory and scienti¬c practice

existence of objects in the world is seen as unproblematic and one-to-one.
The core concept on which much of the empirical position of objectivism
rests is that of phenomenalism. According to this position, the world can
only be perceived through its direct apprehension by the senses. Through
the description of externally perceived objects, language allows a direct
representation of what actually exists in the external world (see Rorty
1991, 1“20). The position of objectivism allows for the possibility of an
outsider™s view that is able to accurately describe the nature of the world
(Putnam 1975). This view can be taken up simply because, as think-
ing and acting subjects, we have a privileged and external view of nature.
When we view objects in this objective manner our sense data correspond
exactly with what is found in nature, and the language we use to describe
these sense data accurately depicts these data using words. The use of
these words in language allows us then to de¬ne the boundaries around
objects and establish the relations of sameness and difference between
described objects. What is more, the relationships between objects per-
ceived in this way are generally seen as causal; in other words, they can be
described by simple cause and effect systems. This generalised position
broadly encompasses a number of epistemological positions, and each is
characterised by the a priori assumption that this general division of the
world exists. For instance, logical positivism holds that through building
observation-based theories about the world, and testing those theories
against the observed world, we are able to adequately describe the true
nature of the world (Hempel 1965).
These positions are viewed as essential theoretical tools for the natu-
ral sciences. The objective existence of a prior natural world is essential
for carrying out science. This is because it is only by assuming the real
existence of the natural world that scientists can feel secure that their
knowledge provides a description of the world that is valid and consis-
tent. Since the goal of science is the steady and cumulative accretion of
knowledge, in order for science to be carried forward and reproduced it
must accept the notion of nature as a constant. This constant, the nat-
ural world, can always be drawn on to back up arguments concerning
the real nature of observations (Latour 1987, 94“100). There are two
important points we must draw from this: ¬rst, in order to carry out
science we must believe in the constancy of the natural world; and sec-
ond, we must take up a detached position to accurately describe that
But there are problems with this view. What if we cannot extricate our-
selves from the world in order to describe it? If we consider this possibility,
we then have to consider that maybe our senses are conditioned by the
position that we take up within the world. If this is allowed as a possibility,
The archaeology of ˜two cultures™ 5

then it is also probable that we are not accurately describing our world,
but categorising it in speci¬c ways. If this is the case then our language
cannot accurately represent the objects in the world. If we take all these
possibilities into consideration, we can no longer consider the world as a
constant. This is especially important if we wish to extend our analysis to
animate subjects in order to examine their role in constructing society.
I will consider each of these points in order to explain the nature of
subjectivism. At this point I wish to focus on the ways in which various
processes of acculturation affect the way in which we describe and inter-
pret the world. The main point here is that we can never step back from
the world in order to describe and know it since the very apparatus we
use to do so, our senses and our language, is determined by the cultural
world in which we live.
I will commence my discussion with the problem of perception. Here
the most important issue is the cognitive categorisation of our sense data
and the subsequent categories we use to describe these data. Recent work
by cognitive psychologists has questioned the notion that the categories
we use to describe the world are essentialist in form. They concede that
the mind has a particular and given structural organisation. However,
the way in which this structure is ordered is dependent upon the cultural
uses of devices such as metaphors in constructing relationships between
perceived objects (Lakoff and Johnson 1980; Lakoff 1987).
If we consider the way in which we categorise the world to be deter-
mined not by a priori categories in the world, but by the metaphors we
employ to describe those categories, then we reach a point at which the
description of the world is contingent or emergent. Rather than viewing
the world of objects and subjects as static, we have to see them as ¬‚uid
and dependent for their apparently solid nature on our descriptions of
them. Rather than accurately using sense data to describe objects, we
are using culturally contingent values or metaphors. If we take this as a
valid observation, then the language we use to describe those objects is
also contingent. This point was made apparent through the early work
of linguists such as Saussure (1916 [1966]) and philosophers such as
Wittgenstein (1953). Importantly, Saussure noted that there was an arbi-
trary relationship between objects and the precise words used to describe
them in language. There is nothing in the properties of objects that is re-
¬‚ected in the words used to refer to them in social language. For Saussure,
language was an abstract code distinct from the world of objects.
This appraisal of language has given rise to two further important no-
tions: most notably, structuralism (L´ vi-Strauss 1966), the study of how
such abstracts are ordered culturally, which is essentially a study of the
codes employed in constructing culture; and semiotics, which has given
6 Archaeological theory and scienti¬c practice

rise to a deeper understanding of how symbols are used. Rather than
considering symbols as entirely abstract, the focus is on how meaning
is created through the codi¬ed use of such symbols (Eco 1979; Ricoeur
1976). This presents us with a double problematic: we are not only con-
ditioned by the cultural world around us, and are therefore not perceiv-
ing the world directly, but we are also investigating the manipulation of
objects as cultural symbols, as cultural meanings. Thus we arrive at a
position where neither the senses used to report the natural world, nor
the cultural devices used to describe it (language), relate to the objective
world in a simple way. Rather they are determined by our cultural ex-
pectations. This leads us on to a further important point concerning our
interpretation of the world.
I have outlined the problems surrounding our cultural understanding
of the world of inanimate objects, and have observed that our subjective
examination of objects is bound up with the manipulation of cultural
meanings; however, further problems arise when we turn to consider
the world of animate subjects. First, our positions as interpreters are not
divorced from the subject that we are interpreting “ human society “ since
the very apparatus we use to describe society are the cultural meanings
from which society is composed. We are then in a situated relationship in
relation to our subject of investigation, and we must be extremely careful
about our interpretations with regard to this relationship. The study of
this situated relationship and the nature of the interpretations we make
while a part of this relationship are essential components of the process
of hermeneutics (Ricoeur 1981). What is more, while I have observed
that for natural scientists the world of objects is composed of static en-
tities with ¬xed relations between them, for social scientists society can
be considered to be composed of social relations; however, these social
relations are never static or constant. We cannot objectify them; rather
they are created through a continuous dynamic, described as social prac-
tice (Bourdieu 1977). If we are to study society, we cannot appeal to
an objecti¬ed and constant nature. We are not considering something
which has a priori existence; rather we are considering something which
is continually being made and remade.

Rationality and relativism
To reiterate, then, we are confronted by a world-view that divides off
objects from subjects. While the relationship between the two is seen
as problematic, there are two broad methods for achieving knowledge of
the world. The ¬rst, natural science, studies nature and uses its privileged
position as an active subject in taking up a detached view of the external
The archaeology of ˜two cultures™ 7

world. The second, social science, studies society and therefore cannot
take up any such privileged position. Rather, it realises the conditional
nature of the knowledge it produces while attempting to describe society.
I now wish to explore in more detail two further theoretical positions taken
up by natural science and social science, that is relativism and rationalism.
Both of these positions focus on the nature of belief, certainty and the
concept of truth. However, both positions rely on the assumption that
the world is divided up into inanimate, essential nature and animate,
contingent society, or objects and subjects.
Rationalism covers a wide series of debates (see Wilson 1971); however
it is broadly concerned with the nature of belief, and how we arrive at
that belief. Here rationalism overlaps considerably with the theoretical
position of objectivism. According to a rationalist view, if we consider a
priori that there is a nature that can be described, then the description
of nature must follow a rational path. This in itself requires a speci¬c
form of reasoning that involves building up a series of law-like statements
about the world. These statements follow an identical form in whatever
context we care to consider them. For instance, if we believe p as a reason
for believing q, then we will believe that p will equal q, wherever and
whenever we observe either p or q. The connection between these two
articles of belief is immutable and incontrovertible. A correlate of this is
that if our knowledge is rationally constructed, then our beliefs can be
considered as either true or false.
However, the relativist views things otherwise. Hollis and Lukes (1982,
5“10) de¬ne a series of relativist positions, including moral relativism,
conceptual relativism, perceptual relativism, relativism of reason and rel-
ativism of truth. In the interests of space, the discussion here will focus on
conceptual relativism, since this has most bearing on the issues discussed
above. Put simply, relativist positions encompass the belief that ˜people of
different cultures live in different worlds™ (Berger and Luckmann 1966;
Sperber 1982). As Sperber (1982, 154) indicates, this does not mean
that people literally live in parallel worlds, rather that they inhabit dif-
fering cognisable worlds. This basic position encompasses the idea that
knowledge may be culturally constructed and that the very act of rea-
soning itself is culturally speci¬c. Beliefs on a given topic can vary and
the relations between knowledge are not, then, absolute. This position is
particularly acute if we consider the way in which the world is categorised.
This view raises a series of problems. If the process by which beliefs are
constructed cannot be viewed as following the same rational process in all
parts of the world, how are we to assess competing knowledge claims? In
other words, we can have no absolute rational knowledge and therefore no
absolute incontrovertible truth. If we consider the possibility that belief
8 Archaeological theory and scienti¬c practice

is culturally contingent “ a conceptual position “ then this opens up the
possibility that truth itself is contingent, a moral relativist position. It
is due to the fact that conceptual relativism blurs with moral relativism
in this way that the entire concept of relativism has received such bad
press. If we throw out the possibility of an absolute transcendent system
of knowledge, i.e. rationalism, then we also dispose of an absolute truth.
Therefore, one of the major issues in the debate between relativism and
rationalism centres on our ability to assess knowledge between systems.
Again we are faced with differences in the order and goals of differing
forms of knowledge, characterised by the natural and social sciences. It
is essential for science to retain the idea of nature being ˜out there™, prior
to human experience, for if nature was constantly changing we would
be unable to observe it accurately and objectively. However, it is also
essential to retain the notion of a science that is ordered according to
unassailable universal laws, since if we consider the possibility that these
laws change according to cultural context, then we lose certainty in the
application of these laws in the generation of further scienti¬c theories. If
scientists had to continually check and recheck the reliability and validity
of these laws, science would be unable to get on with the task of scienti¬c
and technological advancement. The belief in the generation of valid laws
characterises a rationalist or positivist science.
On the other hand, the concept of some form of relativism allows his-
torians of science, and anthropologists and archaeologists studying other
cultures to consider the possibility of other knowledge systems as dis-
crete and coherent forms of knowledge, which each generate their own
forms of logic. If the social sciences were to take up a rationalist posi-
tion, it would be necessary to consider the beliefs of other periods of
history, or other cultures, as irrational or misguided. This would amount
to a form of rational imperialism which would debilitate the enterprise
of understanding other cultures. Furthermore, due to the hermeneutic
involved in the interpretation of cultural knowledge, the critical stance of
anthropology is seen as an important viewpoint, since by studying other
cultural systems we are able to critically re¬‚ect on our own. As Strathern
(1995) has recently noted in relation to the issues of global and local cul-
ture, knowledge is generated through our ability to shift between different
contexts. In this regard Tambiah (1990, 111) describes the interpretative
position of anthropologists as a ˜double subjectivity™. The anthropologist
must subjectively enter the minds of the people they are studying in order
to understand them according to their own categories, while simultan-
eously translating those categories as if distanced from them. The critical
distance that an anthropological viewpoint provides enables us to con-
trast a variety of different orders of knowledge, but this position brings
The archaeology of ˜two cultures™ 9

with it a whole series of problems. How are we to judge our knowledge
systems against others? Can we utilise a single benchmark against which
to judge other cultures? Is there any point at which knowledge may be
considered as commensurable? Is there a core set of real or essential facts
about the world from which the beliefs of other cultures are constructed?
The problem of how we go about judging knowledge claims has been
tackled on a number of levels. I wish to examine this issue from a variety of
angles by examining the problem of external perspective, as well as the
dif¬culties surrounding the internal constructs used within rational state-
ments. The major issue in the debate between rationalists and relativists
concerns the nature of the paradigms, or the worlds, in which knowledge
is constructed. Can we view these differing worlds as being composed of a
core set of beliefs around which alternate paradigms are constructed, or
do we simply classify alternate beliefs as equally true, equally false or
equally true-or-false(Hacking 1982, 49).Each view leads us to an impasse.
First, we will consider the possibility that there is a core set of beliefs
about the world that are incontrovertible. Such a view would propose that
each alternate viewpoint was viewing the same set of data from differing
perspectives, but that each of these perspectives could be bridged through
an act of translation (Hollis 1982). For instance, both Kuhn (1970) and
Feyerabend (1975) claim that differing paradigms can be observed within
the history of science, and that such paradigms are incommensurable. In
other words, the science practised by one set of practitioners, at a given
period in time, could not be comprehended by another set of practitioners
at another time. Each set of practitioners occupied differing worlds and
the knowledge generated in each world was relative to that world. Here
Kuhn (1970) indicated that each group was practising their own rational
methods, but from our viewpoint the knowledge of each group stands in
a relative relationship to the other.
For the rationalist, the view that these paradigmatic understandings can
be translated and understood by us supposes that the two systems cannot
be incommensurable. If we can translate between these two domains of
knowledge, there must be some common ground by which the two belief
systems can be compared. The assumption is that there is an external
viewpoint from which we can measure the validity of either belief. But
how do we externally measure the validity of either system? As Rorty
(1991a, 49“50) indicates, there can be no position by which we can judge
alternate paradigmatic positions, since such a position would involve tak-
ing up what he calls an ˜ethnocentric™ viewpoint in assuming that what
we described as true or rational was actually true. On the other hand, if
we take the relativist view that each paradigm or world has equally valid
belief systems, we still run into a problem if we also believe that in each
10 Archaeological theory and scienti¬c practice

world we are viewing the same external reality. This is partly because such
a view presupposes that cultural beliefs are simply an adjunct to external
reality (see Berger and Luckmann 1966; Ingold 1990; Richards 1990). A
number of writers have observed that there is a tendency amongst both
relativists and rationalists to employ both forms of belief system simul-
taneously (see Elkana 1981, 3). Elkana (1981, 3“4) describes this intel-
lectual position as ˜two-tier thinking™, and I will consider this in more
detail below.
At this stage I simply wish to note that both the rationalist and relativist
viewpoints leave us with a series of problems. The position of rationalism
ultimately relies on the notion that there is a set of rational core beliefs
which must relate to external reality in a precise and determined way.
Belief systems that do not accept the existence of these rational core be-
liefs are either classi¬ed as irrational or are considered to be translatable
to an immutable system of understanding. Meanwhile, the relativist belief
leaves us with the possibility that each paradigm or world is incommen-
surable, and therefore each system of beliefs has its own coherence and
rationale. Each discrete belief system is seen to relate to and to construct
external reality in its own manner. The former position is most applica-
ble to the study of the natural world, since it relies on the concept of a
constant and immutable nature. The latter position is most applicable to
the study of a constantly changing set of social relations, since it relies
on the notion of cultural or social difference. The relationship between
the two points is problematic since, any attempt to ¬nd a ˜bridgehead™
(Hollis 1982) must rely on the notion of an overall external and neutral
viewpoint by which to judge them.

Piecing together the past
The previous section outlined the problems involved in the broad ap-
proaches of both the natural and social sciences. In this section I want to
examine the way in which the issues of objectivity, subjectivity, rational-
ism and relativism have been considered within archaeology as a means
of understanding the underlying roots of the divided state of scienti¬c
and theoretical approaches to the past. I wish to consider the ways in
which archaeologists relate to, and interpret, the material residues of the
past: the archaeological record. Linda Patrik (1985) has undertaken the
most detailed account of the contrasting approaches to the archaeological
record. Here I will draw out some of Patrik™s observations regarding the
differing approaches to the archaeological record and set them against
some of the generalised observations I have already made regarding the
natural and social sciences.
The archaeology of ˜two cultures™ 11

The notion of an archaeological record is used as a heuristic device
by archaeologists, as a means of conceptualising the way in which they
perceive past human action to operate in relation to the remains of the
past. However, as Patrik observes, archaeologists have no uni¬ed notion
of what they understand the archaeological record to be (1985, 29“31).
She de¬nes two contrasting conceptions of the archaeological record:
the physical model and the textual model. The physical model is char-
acterised as considering the archaeological record to be composed of
physical objects and features that are static effects of past causes; the
record is perceived as a direct record of physical objects and processes.
Given this, both the features and spatial order of the record are seen to
be due to physical and behavioural processes that exhibit causal regu-
larities “ in other words, they can be seen to operate according to uni-
versal or probabilistic laws. By way of contrast, the textual model views
the record to be composed of physical objects and features that are ma-
terial signs or symbols of past concepts. The record is seen to record
human actions, ideas and events of human importance. Following on
from this, the structure of the record is viewed as being composed of
rule-guided behaviour which is expressed in culturally speci¬c ways; or
in other words, the record is viewed as contextually speci¬c (for a summa-
tion of these views see Patrik 1985, 36). Both viewpoints have their roots
in the contrasting epistemological and methodological positions outlined

The physical model of the archaeological record
Those archaeologists taking up the physical model treat the archaeolog-
ical record as the natural substrate on which the objective knowledge
of the world is founded. The physical remains of the past are objects
that have been separated from us by the passing of time. This approach
to the archaeological record is typi¬ed by two main schools of thought:
culture-historical archaeology and new or processual archaeology.
The ¬rst school of thought is exempli¬ed by the work of Vere Gordon
Childe. Childe was expressly concerned with distinguishing objective re-
ality from subjective experience. He was speci¬cally interested in distin-
guishing between the nature of knowledge itself and the observation of
˜reality™ within the archaeological record. Childe was aware of the situ-
ated position of the archaeological observer, since he noted that while
archaeologists are concerned with observing cultures, their principle in-
strument of observation is itself culture (1949, 5). He realised that the
categories we employ to understand other cultures are necessarily derived
from our own. McNairn (1980, 135) has observed that this intellectual
12 Archaeological theory and scienti¬c practice

realisation was critical to Childe™s approach to archaeology. Since, due to
the relative nature of the conceptual frameworks of culture, Childe was
uninterested in reconstructing past cultures using our own conceptual
frameworks; rather he was concerned with examining what he described
as ˜true knowledge™. This consisted primarily of practical or technical
behaviour, since he proposed, quite reasonably, that technical behaviour
was a distillation of cultural knowledge (see chapter 5 for a fuller discus-
sion of this notion). Despite this theoretical position, Childe™s writings
embody some of the most imaginative reconstructions of the past ever to
have been written. However apart from his use of Marxist approaches,
Childe™s theoretical approach to the study of past societies remains rela-
tively implicit.
Much of Childe™s writing on the nature of knowledge took special
care to distinguish different orders of knowledge, speci¬cally the magi-
cal and the scienti¬c (Childe 1956). He was careful to distinguish the
uniformitarian principles, which are observable in the present, from the
cultural principles of action that he supposed to be both lost and more-
over inconsequential to our understanding of past action. A similar posi-
tion concerning the de¬nition and veri¬cation of orders of archaeologi-
cal knowledge was proposed by Hawkes (1954). Hawkes suggested that
the degree of certainty or veri¬ability concerning statements about the
past depended on the degree to which those statements were grounded
in archaeological evidence. As statements moved further from empiri-
cal statements about archaeological remains, the inferences drawn from
them could be made with less and less certainty. Interpretation progressed
from the bottom up by a process of inductive reasoning based on the self-
evidential nature of the evidence. It is notable in this regard that those
domains of the archaeological record which remain closer to material
˜reality™ are most amenable to archaeological scientists. Since techno-
logical data are thought to be grounded in the material aspects of the
record, the technological dimensions of archaeological materials provide
the subject matter of much archaeological science.
A similar theoretical position persisted amongst new or processual ar-
chaeologists. This school of thought is exempli¬ed by the work of Lewis
Binford. Although, for processualists, knowledge about the past was also
based on the foundations of the archaeological record, the possibility of
reconstructing past social systems from this record was viewed with less
pessimism. For instance, Binford noted that due to the static nature of
the material remains of the past, the remains could not be treated as
if they spoke for themselves. He correctly noted that the mere obser-
vation of remains was inadequate as a means of understanding the past
(Binford 1983a and b). Rather, as with the subjectivist viewpoint outlined
The archaeology of ˜two cultures™ 13

above, he suggested that the observations by which the remains of the past
are apprehended are necessarily theory laden. Like Childe before him,
Binford™s main concern was with distinguishing between objective reality
and subjective experience. However, unlike Childe, he sought to work
from the objective reality to describe the subjective experience of past
cultures (Binford 1962; Binford and Binford 1968). Rather than aban-
doning the interpretation of past cultures, Binford instead sought to make
theory an explicit component of the process of interpretation.
In order to provide a ¬rmer basis on which to discuss past social sys-
tems, much new archaeology was concerned with producing an objective
account of material evidence prior to its involvement in interpretation.
Given a view of the archaeological record as the static remains of past
physical processes, the obvious step to take in the attempt to move from
the physical remains of the archaeological record towards more gener-
alised statements about past activity involved the creation of universalis-
ing laws. While new archaeologists realised their situated position with
regard to the material remains of the past, it should be noted that all
attempts to formulate universalising laws concerning the nature of the
archaeological record were distinguished by a single objective. That ob-
jective was the attempt to be divested of this situated relationship with
the aim of providing a more empirical account of the evidence.
The mode by which such laws were generated took a number of di-
vergent courses. The ¬rst involved ethnoarchaeology, the observation of
living populations creating material remains in the present. Here the sup-
position was that by observing the actions of various traditional popula-
tions in the ˜ethnographic present™, the archaeologist could provide an
objective model of the kind of probable formation processes which make
up the archaeological record (see Binford 1978; Gould 1980). I wish to
say little about this ¬eld of archaeological enquiry here; however, I will
note that such an activity, while providing increased knowledge concern-
ing site formation, says little about the social structure which brought
the site into being, to say nothing of the responsibilities and moralities
involved in the exercise (Gosden 1999, 58“61).
Given the problems of ethnoarchaeology, a further ¬eld of enquiry
emerged, involving the generation of laws that would more closely inform
the archaeologist about site formation processes. This second avenue of
study included two main strands of enquiry; the ¬rst of these, proposed
by Binford (1983), became known as Middle Range theory. By employ-
ing the principle of the uniformitarian behaviour of certain mechanical
and physical laws in the past and present, Binford sought to generate a
set of laws relating to the behaviour of objects which could be observed
in the present and related to the past. The central requirement of Middle
14 Archaeological theory and scienti¬c practice

Range research involved archaeologists effectively stepping out of the pre-
suppositions that bound them to the interpretation of the archaeological
record. In other words, it supposed a position of absolute objectivity. As
Barrett (1990, 33“4) notes, Middle Range research also relies on the
objective observation of these laws in the present, a position of extreme
A further attempt to formulate laws concerning site formation was
made under the aegis of behavioural archaeology (Schiffer 1976). Here
it was conceded that the record need not be a direct re¬‚ection of what
happened in the past, and may be subject to numerous distortions, both
natural and anthropogenic. Therefore, one of the primary and essential
tasks for archaeological enquiry was the formulation of laws that would
distinguish between the ˜real™ remains of the past and those that are the
result of disturbance. The generation of laws of this nature is especially
important with regard to attempts to understand the natural formation
processes by which the archaeological record is itself created.
We can observe from this that those archaeologists who view the record
as the trace of physical processes place a signi¬cant emphasis on distin-
guishing between the ˜real™ as opposed to the distorted archaeological
record. As I have observed earlier, there are dif¬culties with this approach.
However, we can observe its legacy in the work of many archaeological
scientists, most of whom take post-depositional changes, or taphonomy,
into account in presenting their data. Indeed, taphonomic problems are
used as one of the major means of refutation when criticising the validity
of a body of scienti¬c data. Furthermore, taphonomy is often used as
an explanatory device in interpreting the structure of certain aspects of
the archaeological record (e.g. Todd and Rapson 1988). The absolute
necessity of making an adequate assessment of taphonomic factors con-
cerns much of the practice of faunal and botanical analysis (for example
see Chaplin 1971; Dimbleby 1985; Evans and O™Conner 1999, 78“92).
In these accounts the physical appearance of the archaeological record
has very little to do with anthropogenic or cultural processes, and greater
weight is placed on its formation by physical processes.
While much emphasis was placed on investigating the material nature
of the archaeological record, in the attempt to move beyond the physical
nature of the archaeological record much new or processual archaeology
was also concerned with the interpretation of past social systems. A num-
ber of approaches were utilised in order to understand the nature of past
social systems, but systems theory, ecological theory and cultural evo-
lution were the most often employed. Systems theory was derived from
biology, economics and computing (Watson et al. 1971). It conceived of
any operating system, whether biological or man-made, as being divided
The archaeology of ˜two cultures™ 15

into inter-related parts, or subsystems, which each acted to create a holis-
tic working system. When utilised in archaeology, isolated elements of the
system could be studied as interacting processes. This enabled archae-
ologists to consider the relationship between differing elements of the
archaeological record. These elements were not conceived as possessing
simple one-to-one causative properties; rather, systems were perceived
as operating through a series of interacting elements. The analysis of
such complex systems required the mathematical or statistical treatment
of data in order to make them amenable to study. The interactions be-
tween elements in complex systems were often viewed according to a
model based on feedback mechanisms, a notion derived from cybernetics
(Watson et al. 1971; Renfrew 1984). By modelling these feedback mecha-
nisms, it was possible to explain or predict change in the system. Since
the effect of variables could be modelled in this way, the state of systems
could be described graphically.
Cultural evolution and ecological theory were both linked to a broad
systems theory approach. A cultural evolutionary approach was tradition-
ally used as a means of classifying social systems as they move through
time, into bands, tribes, chiefdoms and states (Sahlins and Service 1960).
This approach allowed given states within the system to be characterised
and was a powerful tool in explaining how speci¬c social formations came
into being (Renfrew 1973, 1979). While systems theory explained the
way in which various elements of the system interacted, cultural evolu-
tion was a means of ordering the changes within a system and provided a
useful conceptual anchor around which to arrange past social formations.
Ecological theory was more closely applicable to the explanation of stasis
and change. If the relationship between social systems and the environ-
ment was conceived as open, then the environment could be viewed as
having a feedback effect on the system. Such a view was founded on the
notion of culture as an ˜extrasomatic™, or extra-bodily, adaptive mecha-
nism for coping with the environment (Binford 1965, 205). According
to ecological theory societies were bounded systems and, like the more
general aspects of systems theory, ecological approaches enabled the re-
lationships between variable components of a system to be modelled (see
Evans and O™Conner 1999, 17“60). As such, ecological theory was par-
ticularly useful as a means of examining the relations between compo-
nents of settlement systems, between members of a trade or exchange
system, or more generally between social groups and the environment
(see Watson et al. 1971, 91“107).
Each of the general theories that are applied to either culture or society
by new or processual archaeologists relies on a number of core assump-
tions. The problem with the physical model of the archaeological record is
16 Archaeological theory and scienti¬c practice

that, when we move away from statements concerning the regularities ob-
served between static material remains and natural processes to consider
the relationship between material remains and human action, it becomes
necessary to also consider social relations as static. If we consider the
archaeological record to be composed of static objects, and the relations
between these objects as best studied through a theoretical framework
which relies on the use of universal or probabilistic laws, then we run into
a number of problems when we attempt to study society.
As noted above, an objective or rationalist viewpoint relies on the notion
of nature as static and constant, yet society is neither static nor constant.
However, the application of systems theory and its cognate theories, cul-
tural evolution and ecological theory, all depend upon the notion of stasis.
While systems theory attempts to describe the dynamics of systems, re-
lationships between elements of the system are examined as if they were
in distinct stable states. One of the core concepts of such a theory is the
notion of homeostasis, the process of remaining stable. Ecological the-
ory similarly relies on systematising and creating mathematical models as
a means of understanding settlement patterns or trade systems. Finally,
cultural evolution itself relies on the notion of stable bounded social for-
mations such as tribes and chiefdoms.
Each of these theories carries with it a rationalist notion of being able
to accurately map and model the world in a systematic way. In each
case human societies are ordered according to a series of given abso-
lutes, such as water, food source, etc. We can see then that those ar-
chaeologists who consider the archaeological record to be composed of
relations between static objects must adopt two strategies for understand-
ing the past. First, they systematise the physical objects from which the
record is composed, by creating general laws applicable to the forma-
tion of the record. Second, because of adopting this viewpoint, they then
¬nd it essential to build on this approach by systematising social systems
and creating laws that model the patterning of human behaviour in the
past (see Toulmin 1990 for a discussion of the history of these ideas of
It is important that we should note the legacy of systems theory, and
especially ecological theory, on the practice of archaeological science. An
understanding of taphonomic factors provides a methodological back-
drop for much archaeological science, especially faunal and botanical
analysis. Ecological theory (Jones 1992) coupled to classical economics
provides the theoretical framework within which the relations between
humans and the environment are then studied (for example see Higgs
1975; Jarman, Bailey and Jarman 1982). Meanwhile, the systematic and
law-like nature of systems theory and ecological theory provides much of
The archaeology of ˜two cultures™ 17

the impetus for the study of exchange systems. Archaeological science
operates within a framework which allows the possibility of systemati-
cally classifying the composition of the material nature of artefacts, and
due to this there is a concurrent emphasis on a theoretical framework
within which exchange can be systematically modelled. Thus archaeolog-
ical science provides a number of physical and chemical characterisation
techniques, such as petrology (Middleton and Freestone 1991), Neutron
Activation Analysis (Hughes et al. 1991; Neff 1992) and isotopic analysis
(Gale and Stos-Gale 1992), which allow a more or less accurate means of
distinguishing materials of different types. These methodologies are also
considered as a further means of classifying material. When these tech-
niques are employed to make more interpretative statements concerning
the past, such statements are often made within a theoretical framework
which also allows for a systematic description of the past (for examples
see Renfrew, Dixon and Cann 1966).
The point here is not necessarily that all archaeological scientists are
theoretically reliant on systems theory and ecological theory, but rather
that the legacy of the concerns which were bound up with these theoreti-
cal frameworks have had an important bearing on the areas of study which
archaeological scientists ¬nd of interest at present. Thus, much archae-
ological science is dedicated to the investigation of exchange networks
or the interpretation of the relations pertaining between humans and the
environment. While much archaeological science is eminently suitable
for these tasks, one of the important questions I wish to consider in this
volume is this: can archaeological science be employed to consider theo-
retical questions framed under a different philosophical background? In
order to consider this possibility, I will turn to consider the textual model
of the archaeological record.

The textual model of the archaeological record
As noted above, those archaeologists taking up a textual view of the past
believe that the archaeological record is composed of the material remains
of signs or symbols. The symbols are elements of a codi¬ed symbolic
structure, and such signs or symbols are viewed as having operated in past
communication systems. Therefore, the existence of material remains
noti¬es us of events of signi¬cance. Given this, the record is viewed as
structured, since each sign or symbol is seen to be a single element in
a wider structure. The task of archaeologists with this viewpoint is to
translate and read the past (Hodder 1986; Tilley 1991).
An important point to note here is that, following on from the approach
of structural linguistics, the relationship between the material world and
18 Archaeological theory and scienti¬c practice

the world of signs or symbols is not conceived of as one-to-one. Rather,
the archaeological record is viewed as a representation. Like the earlier
approach of Schiffer (1976), reading the archaeological record depends
on translation (Patrik 1985, 50). However, unlike Schiffer, this transla-
tion does not involve distinguishing between the ˜real™ and the extran-
eous; instead this act of translation embraces the problems of discourse
(see Ricoeur 1981).
Barthes (1977) notes the problem, common to the study and interpre-
tation of cultural signs, that there is an interpretative distance between
author and reader. Due to this interpretative distance, the validity of an
interpretation is bound up as much with the expectations of the reader as
with what the author wished to express (Ricoeur 1981, 131“45). There-
fore, rather than searching for the underlying ˜real™ archaeological record,
those archaeologists adopting a textual view of archaeology are interpret-
ing the record at a distance; the interpretation of the archaeologist in
the present is as valid as the meaning attached to the object in the past.
The reading of the archaeological record therefore cannot be objective;
at no point can the archaeologist step out of his or her interpretative re-
lationship with the archaeological record in order to create a series of
generalising laws. Rather, the process of reading the material remains of
the past is conditional: it is conditioned by the prejudices, presupposi-
tions and cultural values of the interpreter in the present (Shanks and
Tilley 1987, 105“6). The interpretation of the past is then a political ex-
ercise (Shanks and Tilley 1990; Tilley 1989a). Archaeologists are bound
up in an interpretative relationship, or hermeneutic, with the object of
their study. Objects themselves are only given meaningful status through
There are a number of important correlates of this view. First, the
meanings attached to objects are not ¬xed. Instead, the assignment of
meaning to objects involves a process of dialectics, a movement between
the presuppositions of the interpreter and the material constraints of the
object (Shanks and Tilley 1987, 110“12). Second, if we consider the view
that meaning may be considered to change according to context (Hodder
1986; Barrett 1987a) then, rather than considering the assignation of
meaning to an object as an entirely open and relativist exercise, mean-
ing will be tied down and ¬xed by context. The process of interpreting
the archaeological record is not a process of assigning ¬xed meanings to
objects but rather of reading the patterns of structured differences be-
tween objects. If we consider the archaeological record as a text, then we
can also consider this text to have a grammatical structure. Rather than
considering the archaeological record to be composed of static objects
with a ¬xed relationship to each other, the record is viewed as composed
The archaeology of ˜two cultures™ 19

of objects whose relationships are constantly changing according to their
contextual relationships.
Interestingly this emphasis on the archaeological record as a text privi-
leges the symbolic nature of the record over the physical nature of the
record. The textual approach allows sophisticated examinations of the
symbolic nature of objects; however, since the relationship between
the object and its cultural use as a symbol is considered to be arbitrary,
the physical nature of the object has little effect on the way in which
it is employed symbolically. This problem has been addressed in recent
years (see especially Gosden 1994; Tilley 1996) and will be developed
further in chapters 4 and 8. Nevertheless, it is easy to see why this view of
the representational nature of the archaeological record has had little
effect on archaeological scientists. This is because the contingent nature
of meaning makes objective assertions concerning the physical nature of
material remains dif¬cult to verify (see O™Conner 1991 and Rowley-
Conwy 2000 for clear delineation of these views). An initial proposal that
archaeological assemblages be treated as culturally structured (Moore
1982; Richards and Thomas 1984) has had a relatively low impact on
the analysis of archaeological materials. I feel that (contra Rowley-Conwy
2000) the notable exception to this is the analysis of certain kinds of
deposit in animal and human osteological studies (Hesse 1995; Marciniak
1999). Here the concept of structured deposition is viewed as having im-
portant implications for our understanding of the signi¬cance of differing
modes of deposition (see papers in Anderson and Boyle 1996; Hill 1995;
Kovacik 2000; Renouf 2000; Serjeantson 2000). In contrast, there has
been little discussion of the problems and possibilities of structured
deposits in relation to the study of plant remains or palynology (with
the notable exception of Butler 1995; Hastorf 1991).
This view of the changing nature of the meaning attached to objects
engenders a quite different approach to the interpretation of society. If
we view the relationships between objects to be in a constant state of
¬‚ux, this then allows us to see how the meanings attached to objects may
change over time. By conceptualising the meanings of and relationships
between objects as constantly open to change and re-contextualisation,
it is possible to see how this view allowed contextual archaeologists to
reconsider the relationship between material culture and society. Rather
than studying societies as bounded and static social systems, archaeolo-
gists adopting a textual metaphor viewed societies as constantly chang-
ing. Here the dominant views of society were derived from either Giddens
(1984) or Bourdieu (1977, 1990).
Both authors broadly view societies to be composed of a set of social
structures that are informed by a set of structuring principles. These
20 Archaeological theory and scienti¬c practice

structuring principles are not unlike the set of codi¬ed rules of struc-
turalist thinking. However, structures are not ¬xed, as in structuralism,
but are constantly being reproduced as individuals draw on them and re-
work them. This process is described by Giddens (1984) as structuration.
Therefore, the reproduction of society is not due to an innate tendency
towards homeostasis, as with earlier approaches, rather society is per-
ceived to be in a constant state of ¬‚ux. The constant process of making
and remaking the social structures of societies is what carries forward
social change. However, it is also this process which allows societies to
remain stable. According to this view of society, the structures upon which
individuals immersed within society draw, in order to act within the social
world, tend towards dispositions of stability, and it is this process which
is characterised by Bourdieu as habitus. However, while the process of
drawing on these structuring principles may provide stability, this process
may also provide the instrument of social change. Here the most impor-
tant concept embodied in such a notion of society is that of agency; that
is, that societies are created and recreated through the active involvement
of knowledgeable and active human subjects (see Johnson 1989; Barrett
1994; Dobres and Robb 2000). In conclusion, then, society is not con-
sidered as static; it is considered to be in a constant process of ¬‚ux. Both
the theoretical positions of Bourdieu and Giddens provide interpretative
archaeologists with a more re¬ned view of both social stability and so-
cial change. Finally, rather than utilising external factors to explain either
social change or stability, the notion of agency allows us to understand
more clearly how societies are shaped internally.

At this point I want to step back from the details of these issues and pro-
vide an overview of some of the problems concerned with the attempt to
harmonise each view of the archaeological record. As we have observed,
each model of the archaeological record involves taking up an epistemo-
logical position which resonates strongly with the wider issues that were
considered at the beginning of this chapter. Thus those archaeologists
taking up a position which views the record as physical can be broadly
characterised as objectivist, empiricist and rationalist, while those archae-
ologists who consider the archaeological record as textual can be broadly
characterised as subjectivist and relativist. The labels are fairly broad,
but at this point they serve as a means of characterisation. Many of the
dif¬culties of integrating scienti¬c archaeology with an interpretative ar-
chaeology based on a number of post-structuralist positions arise from
the con¬‚icting nature of these positions.
The archaeology of ˜two cultures™ 21

I do not wish to say too much here about how this integration of knowl-
edge may be achieved; rather I will focus on the problems and bene¬ts
of both archaeological science and interpretative approaches to the study
of past societies. A number of authors have noted that while many ar-
chaeologists have abandoned the central tenets of objectivism and ra-
tionalism as valid means of judging the past, archaeological science has
retained these theoretical standpoints (Thomas 1990; Edmonds 1990).
Moreover, it has also been suggested that archaeological science is reliant
on methodological critique as a substitute for interpretation (Thomas
1990). While this may be so, neither of these points should be entirely
unexpected since, as I have already indicated, the natural sciences are
epistemologically bound to an objective and rational theoretical position.
Methodology, on the other hand, is simply a procedure for distinguishing
and de¬ning the objective nature of the data. The precise de¬nition of
methodology within the literature of archaeological science has its place
within the theoretical framework of objectivism and empiricism. While
I am not proposing a return to empiricism or positivism, I feel that it
is the rigorous application of precisely de¬ned methodologies that lends
scienti¬c discourse its strength. It is this aspect of science in archaeology
that is essential to retain.
I am not interested in considering methodology as a validatory mech-
anism; rather it is a device, which allows certain aspects of the archae-
ological evidence to be reproduced with reasonable accuracy. Science
operates most comfortably within the wider ¬eld of archaeology when
we are able to employ well-de¬ned and rigorous scienti¬c techniques to
the archaeological record. I am thinking here of some of the techniques
routinely used in archaeology which have been imported from the phys-
ical sciences (Tite 1972), or from chemistry (Pollard and Heron 1996).
In other words, these are instrumental scienti¬c techniques that can be
usefully employed in order to provide a more detailed characterisation of
the archaeological object. The problems arise when we utilise a broader
empiricist philosophy as a means of understanding society. As noted ear-
lier, as soon as we begin to frame society as a possible object of scienti¬c
study we must begin to place arti¬cial constraints on our understanding.
We reify it, or make it into a static object. It is this method of studying
society that Bourdieu describes as social physics (1990, 26“7, 135). The
process of objectifying society and the concurrent methodologies and in-
terpretations derived from this process are a reasonable description of
what may be considered as scientism (Edmonds 1990; Barrett 1990).
It was precisely because of the problems inherent in viewing societies
as static objects that there was such a major shift in the interpretative
framework employed in archaeology (see Hodder 1982a). Rather than
22 Archaeological theory and scienti¬c practice

attempting to systematise society by viewing it through the lens of the
natural sciences, archaeology moved to the interpretative position which
had prevailed for a considerable period within a number of the social
sciences (see Leach 1973; Gellner 1982; Miller 1982). One of the major
strengths of an interpretative archaeology that embraces a variety of post-
structuralist approaches is the rigorous nature of its theoretical frame-
work. We are now able to begin to reconsider the complexities involved
both in understanding how societies operate and in understanding the
way in which the social operates to structure the archaeological record.
What is more, interpretative approaches also allow a more critical under-
standing of our social position as interpreters.
In conclusion, I would like to propose that in order to begin to con-
sider the possibilities of relating scienti¬c archaeology with interpretative
archaeology we must retain an aspect of each. The strengths of scien-
ti¬c approaches are re¬‚ected in their methodological rigour and repro-
ducibility, while the strengths of interpretative approaches are re¬‚ected
in their theoretical rigour and their ability to provide a coherent and sat-
isfying account of society. The problems of embracing these two aspects
of contemporary archaeology are manifold, and we must move through
a dif¬cult epistemological mine¬eld in order to provide a more satisfying
account of the past which encompasses both approaches. That will be
the subject of the next chapter.
2 Science as culture: creating
interpretative networks

Reviewing the record
The previous chapter broadly reviewed both the physical and the textual
approaches to the archaeological record, and was intended as an assess-
ment of many of the debates prevalent in the archaeological literature.
This re-examination had an important aim: by broadly characterising the
two main approaches to the archaeological record and providing a brief
account of the problems with each position, I drew out the differences
between the two approaches. These distinctions are crucial since I feel
that the source of the rift between archaeological scientists and theoretical
archaeologists lies, at a fundamental level, with the starkly different philo-
sophical approaches each group employs as a means of understanding the
past. On one side, we have a viewpoint which regards the archaeological
record as the product of physical processes which can be examined empir-
ically and objectively using the sense data derived from the description
of objects. These descriptions and measurements can then be built up
into generalising laws that can be applied in all archaeological contexts.
On the other side, we have a viewpoint which considers the archaeolog-
ical record to be the product of meaningful social action. As such it can
be considered to be composed of a structured set of differences, like a
text. In this case, each sentence of the text, or part of the archaeological
record, is contextually distinct.
There are two important points here. First, if we focus our intellectual
efforts not just on the objective description of objects, but on the inter-
pretation of past societies through the medium of objects, then we are
placing ourselves in a situated, hermeneutic position in relation to these
past societies. Second, if we regard societies as something more than static
objects which can be classi¬ed into cultural evolutionary stages, and we
dispose of the view of society as something we can describe mathemati-
cally, then we must begin to consider each society as culturally different
and historically contextual. We cannot then apply a self-evident rationalist

24 Archaeological theory and scienti¬c practice

procedure to describe these societies, since this would be denying each
society its cultural speci¬city.
If, on the other hand, we consider the material remains of the archae-
ological record to be the result of meaningful actions, then we must also
consider the cultural logic which brought these material remains into be-
ing. In order to do this we must consider the possibility that our interpre-
tations re¬‚ect the cultural speci¬city of the remains. By interpreting the
remains in so many differing ways, we abandon the idea of a universal and
rational mode of cultural understanding. This then opens up the problem
of cultural relativism (Kohl 1993), which is especially problematic in an
archaeological context. If we consider the possibility that differing orders
of knowledge may be responsible for the patterns of material we observe
in the archaeological record, how do we validate our interpretations of
this material evidence, and which paradigm provides the best interpreta-
tion of the evidence? If we abandon the notion of a rationalist yardstick for
interpretation, does this mean that all interpretations are equally valid?
Here it is essential to return to the issue of paradigms and the com-
mensuration of knowledge. If we consider Kuhn™s view of paradigms, then
each body of knowledge generated within a paradigm is structured by the
concerns of that paradigm and is therefore distinct. This approach serves
to perpetuate the dichotomy that is evident between the two models of
the archaeological record, since according to Kuhn neither domain of
knowledge is commensurable with the other. Interestingly, Patrik (1985,
546) considers the possibility of utilising both models of the archaeolog-
ical record simultaneously. The physical model would provide informa-
tion regarding the physical behaviour of artefacts, taphonomy and site
formation, while the textual model would be used as a means of under-
standing the structured nature of the archaeological record, which would
thereby allow a clearer interpretation of the social action represented by
the record. Although this seems a pro¬table way forward, there are prob-
lems with employing both models of the archaeological record in this way.
This is because, according to such a view, we would retain the notion of
a core set of beliefs regarding the ˜reality™ of the archaeological record
and onto these our textual or interpretative approach would construct its
culturally speci¬c worlds of meaning.
As an alternative I wish to reconsider Tambiah™s notion of a ˜double
subjectivity™ (1990, 1“11). This concept allows us to consider interpre-
tation as an act that results from the process of translation. Such a trans-
lation occurs, for example, when anthropologists place understandings
generated by culturally speci¬c categories within a framework generated
under the preconceptions of Western thought processes. What might this
activity look like in the context of archaeology? One possibility may be to
Science and culture: interpretative networks 25

adopt the interpretative methods considered by Barrett (1990). Barrett™s
ideas regarding the interpretative process arise out of a critique of the em-
piricism inherent in Binford™s project of Middle Range research. Rather
than simply applying empirical observations to data, as Binford™s con-
ception of Middle Range research requires, Barrett suggests a movement
back and forth between the theoretically informed notions of how society
is reproduced (notions which embrace an understanding of how human
action relates to material conditions) and the material evidence. The di-
alectic involved in this process allows for the possibility of making more
theoretically informed statements concerning the nature of past social

. 1
( 7)